Theoretical and Experimental Aerodynamics by Dr. Mrinal Kaushik

Mrinal Kaushik
Theoretical and
Experimental
Aerodynamics
Theoretical and Experimental Aerodynamics
Mrinal Kaushik
Theoretical and Experimental

Aerodynamics
123
Mrinal Kaushik
Department of Aerospace Engineering
Indian Institute of Technology Kharagpur
Kharagpur, West Bengal, India
ISBN 978-981-13-1677-7 ISBN 978-981-13-1678-4 (eBook)

https://doi.org/10.1007/978-981-13-1678-4
Library of Congress Control Number: 2018951404
© Springer Nature Singapore Pte Ltd. 2019

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction
on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic
adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and
regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed
to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty,
express or implied, with respect to the material contained herein or for any errors or omissions that may have been
made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional
affiliations.
This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore
Preface
This book is elegantly composed for an average student who is interested in learning the
fundamentals of aerodynamics. Even if they do not have any background in aerodynamics,
this book is for them. The concepts are well explained and are almost complete in themselves
so that the reader need not refer any other source of information. Moreover, in each chapter, I
have tried to maintain an adequate balance while providing an elaborate discussion on the flow
physics with their corresponding mathematical formulations. Indeed, I was inspired to write
this text while delivering the lectures in aerodynamics courses to undergraduate and post-
graduate students at IIT Kharagpur, where a strong need was felt to have a book which can
explain the subject from the student’s perspective.
This manuscript has been developed to provide a complete exposure to aerodynamics
beginning with definitions and concepts and to cover applied areas of practical importance.
The text is broadly divided into two parts: Theoretical Aerodynamics and Applied Aerody-
namics. The first part is spread over 12 chapters, while the second part consists of 2 chapters.
The sequential organization of the book is discussed below in brief.
In aerodynamics, one has to deal with the effects of the earth’s atmosphere on flying
objects; therefore, the properties of the atmosphere are discussed in Chap. 1. The basic
concepts in aerodynamics, equations of motion, potential flows, and incompressible flows over
two- and three-dimensional wings are given in Chaps. 2–7. Thermodynamic aspects of fluid
flows, one-dimensional isentropic flows, flow across shock waves, and high Mach number
flows are elaborated in Chaps. 8–10. The fundamental aspects of boundary layers, which are
essential to understanding the boundary layer separation, sudden rise of drag, and ways of its
reduction, are discussed in Chap. 11. A thorough discussion on the design and operation of
various general- and special-purpose wind tunnels is provided in Chap. 12. Besides these,
some advanced concepts of high Mach number flows particularly useful for researchers and
practicing engineers are discussed in Chaps. 13 and 14. These topics may be quite useful for
an advanced course, and hopefully, their inclusion will be appreciated to the existing literature
on the subject.
Some additional topics of interest are also discussed in Appendix A. To help estimate
uncertainty in an experimental investigation, the procedure for uncertainty analysis is laid
down in Appendix B. The properties of the International Standard Atmosphere and isentropic
flows are tabulated in Appendices C and D, respectively. In addition, to help evaluate the
understanding of the subject, a complete set of 600 multiple choice questions with their
answers are provided in Appendix E. I am confident that the readers will find them extremely
useful while studying the subject.
I thank the almighty for his mercy and blessings. I sincerely thank Ms. Swati Meherishi and
Ms. Avni of Springer for providing their generous support and advice that helped immensely
in improving the quality of the book.
Kharagpur, India Mrinal Kaushik
v
Contents
Part I Theoretical Aerodynamics

1 The Standard Atmosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.1.1 Role of Atmosphere in the Aerodynamics . . . . . . . . . . . . . . . . . 4
1.2 Composition and Structure of the Atmosphere . . . . . . . . . . . . . . . . . . . . . 4
1.2.1 Primary Layers in the Atmosphere . . . . . . . . . . . . . . . . . . . . . . . 4
1.2.2 Secondary Layers in the Atmosphere . . . . . . . . . . . . . . . . . . . . . 6
1.3 Interpretation of the Altitude . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.4 Variation of Pressure in the Standard Atmosphere . . . . . . . . . . . . . . . . . . 7
1.5 Relation Between Geopotential and Geometric Altitude . . . . . . . . . . . . . . 8
1.6 Distribution of Properties in Troposphere and Stratosphere . . . . . . . . . . . . 8
1.7 Physical and Optical Properties of the Atmosphere . . . . . . . . . . . . . . . . . . 10
1.7.1 Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.7.2 Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.7.3 Mass and Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.7.4 Scattering Phenomena . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.7.5 Absorption and Emission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.7.6 Refractive Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.8 Static Stability Analysis of Troposphere Layer . . . . . . . . . . . . . . . . . . . . . 12
1.9 Definition of Wind . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.9.1 Cyclones and Anticyclones in Temperate Zones . . . . . . . . . . . . 16
1.9.2 Hurricanes or Typhoons in Tropics . . . . . . . . . . . . . . . . . . . . . . 16
1.10 Geostrophic and Ageostrophic Winds . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.11 Ekman Spiral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
1.12 Global and Local Winds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
1.12.1 Land and Sea Breeze . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
1.12.2 Zonal Wind Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
1.12.3 Some Specific Names of the Wind . . . . . . . . . . . . . . . . . . . . . . 20
1.12.4 Measuring the Winds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2 Basic Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.1 Aerodynamics: An Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.2 The Aircraft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.2.1 Aerodynamic Forces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.2.2 Aerodynamic Moments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.3 Parametric Studies in Aerodynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.3.1 Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.3.2 Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.3.3 Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.3.4 Flow Velocity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.3.5 Stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
vii
viii Contents
2.3.6 Coefficient of Viscosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

2.4 The Airfoil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.4.1 Nomenclature of an Airfoil and the Wing . . . . . . . . . . . . . . . . . 34
2.4.2 Pressure Distribution Around an Airfoil . . . . . . . . . . . . . . . . . . . 35
2.4.3 Generation of Forces and Moments . . . . . . . . . . . . . . . . . . . . . . 37
2.4.4 Center of Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.4.5 The Aerodynamic Center . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.5 Dimensional Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.5.1 Concept of Flow Similarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.6 Types of Flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
2.6.1 Continuum and Non-continuum Flows. . . . . . . . . . . . . . . . . . . . 44
2.6.2 Steady and Unsteady Flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2.6.3 Uniform and Nonuniform Flows . . . . . . . . . . . . . . . . . . . . . . . . 46
2.6.4 Incompressible and Compressible Flows . . . . . . . . . . . . . . . . . . 47
2.6.5 Inviscid and Viscous Flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.6.6 Mach Number Flow Regimes . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.7 Hodograph Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3 Governing Equations of Fluid Flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.2 Review of Vector Relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.2.1 Scalar (or Dot) Product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.2.2 Vector (or Cross) Product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.2.3 Orthogonal Coordinate Axes . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.2.4 Scalar and Vector Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.2.5 Scalar Product of Two Vectors . . . . . . . . . . . . . . . . . . . . . . . . . 60
3.2.6 Vector Product of Two Vectors . . . . . . . . . . . . . . . . . . . . . . . . . 60
3.2.7 Gradient of a Scalar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
3.2.8 Divergence of a Vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3.2.9 Curl of a Vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3.2.10 Line Integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.2.11 Surface Integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.2.12 Volume Integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.2.13 Stokes Curl Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.2.14 Gauss Divergence Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.2.15 Gradient Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.3 Eulerian and Lagrangian Viewpoints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.3.1 Local and Material Derivatives . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.4 Primary and Auxiliary Laws for Continuous Media . . . . . . . . . . . . . . . . . 64
3.5 Flow Analysis Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.5.1 Finite Control Mass Approach . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.5.2 Finite Control Volume Approach . . . . . . . . . . . . . . . . . . . . . . . . 65
3.5.3 Infinitesimal Fluid Element Approach . . . . . . . . . . . . . . . . . . . . 65
3.5.4 Microscopic Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.6 Integral and Differential Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.7 One-, Two-, and Three-Dimensional Flows. . . . . . . . . . . . . . . . . . . . . . . . 66
3.8 The Continuity Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.8.1 Some Important Observations . . . . . . . . . . . . . . . . . . . . . . . . . . 67
3.9 Graphical Representation of Fluid Flows . . . . . . . . . . . . . . . . . . . . . . . . . 68
3.9.1 Timelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
3.9.2 Streamlines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
3.9.3 Pathlines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.9.4 Streaklines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Contents ix
3.10 Angular Velocity, Vorticity, and the Shear Strain Rate . . . . . . . . . . . . . . . 70

3.11 The Navier–Stokes Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.11.1 The Euler Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.11.2 Velocity–Vorticity Form of the Navier–Stokes Equation . . . . . . 77
3.11.3 The Crocco’s Theorem. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
3.12 Rotational Flows and the Circulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
3.13 Irrotational Flows and the Potential Function . . . . . . . . . . . . . . . . . . . . . . 79
3.14 Stream Function and the Concept of Vector Potential . . . . . . . . . . . . . . . . 80
3.14.1 Concept of Streamline in Three Dimensions . . . . . . . . . . . . . . . 82
3.14.2 Axisymmetric Flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
3.14.3 Physical Interpretation of Lagrange Stream Function . . . . . . . . . 83
3.15 The Cauchy–Riemann Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
3.16 The Bernoulli’s Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.16.1 Steady Bernoulli’s Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
3.16.2 Unsteady Bernoulli’s Equation . . . . . . . . . . . . . . . . . . . . . . . . . . 91
3.17 Using the Bernoulli’s Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.17.1 Airspeed Measurement Using Pitot-Static Probe . . . . . . . . . . . . 92
3.17.2 Pressure Coefficient and the Compressibility Correction
Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.17.3 The Airspeed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
3.18 Reynolds Transport Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
3.18.1 Physical Significance of RTT . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
3.19 The Energy Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
4 Potential Flow Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
4.2 Potential Flows and the Laplace’s Equation . . . . . . . . . . . . . . . . . . . . . . . 107
4.3 Standard Solutions of the Potential Flow . . . . . . . . . . . . . . . . . . . . . . . . . 108
4.3.1 Uniform Potential Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
4.3.2 Line Source (or Line Sink) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
4.3.3 Source–Sink Combination and the Doublet Potential . . . . . . . . . 112
4.3.4 Line Vortex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
4.4 Superposition of Standard Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
4.4.1 A Source in the Uniform Flow (Axisymmetric Flow
over a Semi-infinite Body) . . . . . . . . . . . . . . . . . . . . . . . . . .... 115
4.4.2 A Pair of Source and Sink in the Uniform Flow
(Axisymmetric Flow over a Closed Body) . . . . . . . . . . . . . .... 117
4.4.3 A Doublet in the Uniform Flow (Flow over a Circular
Cylinder Without Circulation) . . . . . . . . . . . . . . . . . . . . . . .... 117
4.4.4 A Point Vortex in the Uniform Flow . . . . . . . . . . . . . . . . . .... 119
4.4.5 A Doublet and a Vortex Flow in the Uniform Flow
(Flow Past a Circular Cylinder with Circulation) . . . . . . . . . . . . 119
4.5 Implications of Kutta–Joukowski Theorem in the Lift Generation . . . . . . 123
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
5 Thin Airfoil Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
5.2 Circulation and Vorticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
5.3 The Kutta Condition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
5.4 Kelvin’s Circulation Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
5.4.1 The Starting Vortex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
x Contents
5.5 Summary of the Lift Generation Mechanism by the Airfoil . . . . . . . . . . . 132

5.6 Classical Thin Airfoil Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
5.6.1 Aerodynamic Characteristics of a Thin Symmetric Airfoil . . . . . 135
5.6.2 Aerodynamic Characteristics of a Thin Cambered Airfoil . . . . . 138
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
6 Finite Wing Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 145
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 145
6.1.1 Relation Between Trailing Edge Vortices and Spanwise
Load Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
6.2 Concept of Vortex. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
6.2.1 Forced Vortex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
6.2.2 Free Vortex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
6.2.3 Vortex Line, Vortex Tube, and the Vortex Filament . . . . . . . . . 149
6.3 Helmholtz’s Theorems of Vortex Motion . . . . . . . . . . . . . . . . . . . . . . . . . 150
6.3.1 Helmholtz First Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
6.3.2 Helmholtz Second Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
6.3.3 Helmholtz Third Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
6.3.4 Helmholtz Fourth Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
6.4 Biot and Savart Law of Vortex Motion. . . . . . . . . . . . . . . . . . . . . . . . . . . 151
6.4.1 Application of Biot and Savart Law: Velocity Induced
by a Straight Vortex Filament . . . . . . . . . . . . . . . . . . . . . . . . . . 154
6.5 Vortex System and the Evolution of Prandtl’s Lifting Line Theory . . . . . 155
6.5.1 Lanchester–Prandtl Wing Theory . . . . . . . . . . . . . . . . . . . . . . . . 158
6.5.2 Symmetric Elliptical Aerodynamic Load Distribution . . . . . . . . 161
6.5.3 Symmetric General Aerodynamic Load Distribution . . . . . . . . . 163
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
7 Panel Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
7.2 Hess and Smith Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
7.2.1 Line Source Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
7.2.2 Perturbation Velocity Components Due to Source
Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
7.3 Vortex Panel Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
7.3.1 Panel of Constant Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
7.3.2 Panel with Linearly Varied Vortex Strength . . . . . . . . . . . . . . . 176
7.3.3 Transformation of Panel Coordinates . . . . . . . . . . . . . . . . . . . . . 177
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
8 Thermodynamics of Fluids in Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
8.1.1 Concept of System, Surroundings, and the Universe . . . . . . . . . 181
8.2 Internal Energy and the First Law of Thermodynamics. . . . . . . . . . . . . . . 181
8.2.1 Concept of Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
8.2.2 Concept of Enthalpy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
8.2.3 Different Forms of the First Law for an Adiabatic Flow . . . . . . 184
8.2.4 Concept of Specific Heats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
8.2.5 Coefficient of Thermal Conductivity . . . . . . . . . . . . . . . . . . . . . 185
8.3 Energy Equation for an Open System . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
Contents xi
8.4 Entropy and the Second Law of Thermodynamics . . . . . . . . . . . . . . . . . . 186

8.4.1 Thermodynamic Efficiency and Clausius Inequality . . . . . . . . . . 186
8.5 Combined Forms of the First Law and the Second Law . . . . . . . . . . . . . . 188
8.6 Maxwell’s Thermodynamic Relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
8.6.1 Internal Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
8.6.2 Enthalpy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
8.6.3 Helmholtz Free Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
8.6.4 Gibbs Free Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
8.7 Effects of Fluid Compressibility on Specific Heats . . . . . . . . . . . . . . . . . . 190
8.8 Thermal and Calorical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
8.8.1 Thermally Perfect Gas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
8.8.2 Mayer’s Relation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
8.9 The Perfect Gas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
8.9.1 Ratio of Specific Heats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
8.9.2 Limitation of Air as a Perfect Gas . . . . . . . . . . . . . . . . . . . . . . . 193
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
9 Compressible Flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
9.1 Introduction to Elastic and Inviscid Compressible Flows . . . . . . . . . . . . . 199
9.1.1 One-Dimensional Flow Approximation . . . . . . . . . . . . . . . . . . . 199
9.2 Governing Equations of Compressible Flows . . . . . . . . . . . . . . . . . . . . . . 199
9.3 Effects of Acoustic Speed on the Fluid Compressibility . . . . . . . . . . . . . . 201
9.3.1 Speed of Sound . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
9.3.2 Speed of Sound in a Perfect Gas . . . . . . . . . . . . . . . . . . . . . . . . 202
9.3.3 Effect of Molecular Weight on the Acoustic Speed . . . . . . . . . . 202
9.3.4 Concept of Mach Number . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
9.3.5 The Mach Angle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
9.3.6 Classification of Flow Regimes Based on the Mach
Number . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
9.4 One-Dimensional, Steady, and Isentropic Flow of a Perfect Gas . . . . . . . 204
9.4.1 Stagnation (or Total) Properties in an Isentropic Flow . . . . . . . . 205
9.4.2 The Dimensionless Velocity ðM Þ . . . . . . . . . . . . . . . . . . . . . . . 206
9.4.3 Effect of Area Variation on Compressible Flow
(Area–Velocity Relation) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
9.4.4 Mass Flow Rate (m ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
9.4.5 Mass Flow Rate in a Choked Streamtube . . . . . . . . . . . . . . . . . 209
9.4.6 Physical Interpretation of Flow at the Throat . . . . . . . . . . . . . . . 210
9.4.7 Area Ratio for the Convergent–Divergent Streamtube . . . . . . . . 210
9.4.8 Types of Characteristic Speeds along a Streamtube . . . . . . . . . . 211
9.4.9 Mass Flow Rate Variation with Pressure . . . . . . . . . . . . . . . . . . 212
9.5 The Adiabatic Flow Ellipse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
9.6 Processes Causing a Change of State in Compressible Flows . . . . . . . . . . 214
9.6.1 Fanno Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
9.6.2 Rayleigh Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
9.6.3 Normal Shock Wave . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
9.7 One-Dimensional Flow Across a Normal Shock . . . . . . . . . . . . . . . . . . . . 215
9.7.1 Governing Equations of the Normal Shock . . . . . . . . . . . . . . . . 216
9.7.2 Normal Shock Equations for a Perfect Gas . . . . . . . . . . . . . . . . 216
9.7.3 Working Relations for Normal Shock Wave . . . . . . . . . . . . . . . 217
9.7.4 The Prandtl–Meyer Relation . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
9.7.5 The Rankine–Hugoniot Relation . . . . . . . . . . . . . . . . . . . . . . . . 221
9.8 Supersonic Pitot Probe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
9.8.1 Rayleigh Supersonic Pitot Probe Formula . . . . . . . . . . . . . . . . . 222
9.9 Convergent–Divergent Nozzle (de Laval Nozzle) . . . . . . . . . . . . . . . . . . . 222
xii Contents
9.10 Two-Dimensional Flow Across an Oblique Shock Wave . . . . . . . . . . . . . 226

9.10.1 Governing Equations of the Oblique Shock . . . . . . . . . . . . . . . . 226
9.10.2 Minimum and Maximum Wave Angles . . . . . . . . . . . . . . . . . . . 228
9.10.3 h b M Relation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
9.10.4 Weak Oblique Shocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
9.10.5 Isentropic Compression in Supersonic Flow by Turning . . . . . . 230
9.11 The Prandtl–Meyer Expansion Fan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
9.11.1 Isentropic Expansion in Supersonic Flow by Turning . . . . . . . . 232
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
10 Hypersonic Flows. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
10.1.1 Shock Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
10.1.2 Entropy Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
10.1.3 Viscous-Inviscid Flow Interaction . . . . . . . . . . . . . . . . . . . . . . . 238
10.2 Oblique Shock Relations in Hypersonic Flow . . . . . . . . . . . . . . . . . . . . . . 240
10.3 Mach Number Independence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
10.4 Expansion Wave Relations in Hypersonic Flow . . . . . . . . . . . . . . . . . . . . 242
10.5 Hypersonic Similarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
10.6 Newtonian Theory. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
10.6.1 Lift and Drag Coefficients for a Flat Plate at an Angle
of Attack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
10.6.2 Modified Newtonian Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
11 Boundary Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
11.2 Boundary Layer Thickness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
11.2.1 Displacement Thickness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
11.2.2 Momentum Thickness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
11.2.3 Kinetic Energy Thickness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
11.3 Similarity Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
11.4 Boundary Layer Separation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
11.4.1 Physics of Separation Bubbles in Boundary Layers . . . . . . . . . . 256
11.5 Boundary Layer Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
11.5.1 Continuity Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
11.5.2 x-Momentum Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
11.5.3 y-Momentum Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
11.6 von Karman Momentum Integral Equation . . . . . . . . . . . . . . . . . . . . . . . . 260
11.7 K. Wieghardt Energy Integral Equation . . . . . . . . . . . . . . . . . . . . . . . . . . 262
11.8 Laminar Boundary Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
11.8.1 Incompressible Laminar Flow over a Flat Plate
(Blasius Solution). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
11.8.2 Compressible Laminar Flow over a Flat Plate . . . . . . . . . . . . . . 266
11.8.3 Stagnation Point Flow (or Hiemenz Flow) . . . . . . . . . . . . . . . . . 267
11.8.4 Mixing of Two Uniform Laminar Flows . . . . . . . . . . . . . . . . . . 269
11.9 Application of von Karman Momentum Integral Equation . . . . . . . . . . . . 270
11.9.1 Karman–Pohlhausen Approximate Solution Method . . . . . . . . . 270
11.10 Laminar–Turbulent Transition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
Contents xiii
11.11 Turbulent Boundary Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272

11.11.1 Mean Motion and Perturbations . . . . . . . . . . . . . . . . . . . . . . . . . 272
11.11.2 Governing Equations for Turbulent Flows . . . . . . . . . . . . . . . . . 274
11.11.3 Prandtl’s Mixing Length Hypothesis . . . . . . . . . . . . . . . . . . . . . 275
11.11.4 Regimes in Turbulent Boundary Layer . . . . . . . . . . . . . . . . . . . 277
11.11.5 Skin Friction Coefficient for Turbulent Boundary Layer
over a Flat Plate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 278
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 280
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 282
12 Wind Tunnels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
12.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
12.2 Types of Wind Tunnels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
12.2.1 Mach Number Regimes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
12.2.2 Tunnel Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
12.2.3 Operational Fluid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
12.2.4 Special Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
12.3 Experimental Models and Similitude . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
12.3.1 Geometric Similarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
12.3.2 Kinematic Similarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
12.3.3 Dynamic Similarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
12.4 Subsonic Wind Tunnels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
12.4.1 Effuser or Contraction Cone. . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
12.4.2 Test Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
12.4.3 Subsonic Wind Tunnel Diffuser . . . . . . . . . . . . . . . . . . . . . . . . . 290
12.4.4 Driving Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
12.4.5 Losses in Subsonic Wind Tunnels . . . . . . . . . . . . . . . . . . . . . . . 293
12.4.6 Energy Ratio of a Subsonic Wind Tunnel . . . . . . . . . . . . . . . . . 295
12.5 High-Speed Wind Tunnels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296
12.5.1 Intermittent-Blowdown Wind Tunnels . . . . . . . . . . . . . . . . . . . . 296
12.5.2 Intermittent-Indraft Wind Tunnels . . . . . . . . . . . . . . . . . . . . . . . 297
12.5.3 Continuous Supersonic Wind Tunnels . . . . . . . . . . . . . . . . . . . . 297
12.5.4 Losses in Supersonic Wind Tunnels . . . . . . . . . . . . . . . . . . . . . 299
12.5.5 Supersonic Wind Tunnel Components . . . . . . . . . . . . . . . . . . . . 299
12.6 Hypersonic Wind Tunnels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
12.7 Special Purpose Tunnels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
12.7.1 Shock Tube . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
12.7.2 Gun Tunnel(or Shock Tunnel) . . . . . . . . . . . . . . . . . . . . . . . . . . 311
12.7.3 Ludwieg Tube . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
12.7.4 Plasma Wind Tunnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
Part II Applied Aerodynamics

13 Supersonic Jets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
13.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
13.2 Free Turbulence Theories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
13.2.1 Semi-Empirical Theories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
13.2.2 Turbulence Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
13.3 Computational Techniques for Studying the Jets . . . . . . . . . . . . . . . . . . . . 328
13.3.1 Preliminary Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
13.3.2 Reynolds-Averaged Navier–Stokes (RANS). . . . . . . . . . . . . . . . 329
13.3.3 Large Eddy Simulation (LES) . . . . . . . . . . . . . . . . . . . . . . . . . . 330
xiv Contents
13.3.4 Direct Numerical Simulation (DNS) . . . . . . . . . . . . . . . . . . . . . 332

13.3.5 Some Specific Computational Studies on the Jets . . . . . . . . . . . 336
13.4 Experimental Techniques for Studying the Jets . . . . . . . . . . . . . . . . . . . . . 340
13.4.1 Pressure Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
13.4.2 Optical Flow Visualization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344
13.5 Aerodynamic Mixing Enhancement and Jet Controls . . . . . . . . . . . . . . . . 344
13.5.1 Classification of Jet Controls . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
13.6 Acoustics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
13.6.1 Pressure Mode Acoustics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
13.6.2 Vorticity Mode Acoustics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
13.6.3 Entropy Mode Acoustics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
13.7 Jet Noise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
13.7.1 Subsonic Jet Noise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
13.7.2 Supersonic Jet Noise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
14 Shock Wave and Boundary Layer Interactions . . . . . . . . . . . . . . . . . . . . . . . . . 361
14.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
14.1.1 Transonic Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
14.2 Supersonic Flow Field Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362
14.2.1 Simple Wave and Non-simple Region . . . . . . . . . . . . . . . . . . . . 362
14.2.2 Reflection of an Oblique Shock Wave from a Solid Wall . . . . . 362
14.2.3 Reflection of an Oblique Shock Wave from a Free
Pressure Boundary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362
14.2.4 Oblique Shock Wave Cancelation . . . . . . . . . . . . . . . . . . . . . . . 362
14.2.5 Interference of Shock and Expansion Waves . . . . . . . . . . . . . . . 364
14.2.6 Shock–Shock Interference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364
14.2.7 Types of Shock–Shock Interference (Edney Classification) . . . . 365
14.3 The SBLIs Phenomena . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366
14.4 The Oswatitsch Relation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
14.5 Some Important Studies on SBLIs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370
14.6 SBLIs Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373
14.6.1 Classification of Control Techniques . . . . . . . . . . . . . . . . . . . . . 373
14.7 Experimental Techniques for Studying the SBLIs . . . . . . . . . . . . . . . . . . . 376
14.8 Evaluation of Effectiveness of Some Specific Passive Controls . . . . . . . . . 376
14.8.1 SBLI Control Using Cavity Covered with Porous Surface . . . . . 377
14.8.2 SBLI Control Using Ramped-Vane Micro-Vortex
Generators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390
Appendix A: Supplemental Readings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393
Appendix B: The Uncertainty Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
Appendix C: The Standard Atmosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411
Appendix D: Isentropic Table ð° ¼ 1:4Þ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
Appendix E: Multiple Choice Questions in Aerospace Engineering . . . . . . . . . . . . 437
Appendix F: Letter of Acknowledgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507
About the Author
Mrinal Kaushik is Assistant Professor of aerospace engineering

at IIT Kharagpur. He earned his Ph.D., M.Tech, and B.Tech in
aerospace engineering in 2012, 2003, and 1999, respectively,
from IIT Kanpur. Prior to joining IIT Kharagpur in 2013, he also
worked at Defence Institute of Advanced Technology (Pune),
General Motors India Technical Center (Bangalore), Indian Space
Research Organization (Trivandrum), and Tata Consultancy
Services (Mumbai) for several years.
He has rich experience in teaching courses, namely industrial
aerodynamics, missile aerodynamics, high-speed aerodynamics,
and low-speed aerodynamics. At IIT Kharagpur, he has intro-
duced and developed two new courses, Introduction to Aircraft
Armaments and Physics of Fluid Flow Experiments and revised
the syllabus of Industrial Aerodynamics. His research interests are
shock–boundary layer interactions, jet controls, aeroacoustic and
high-enthalpy flows. In these areas, he has published
peer-reviewed research articles in the international journals and
refereed conference proceedings.
He has authored more than five books and numerous research
papers; in 2016, he was acknowledged for his book writing by the
Defense Minister of India. In the same year, his biography was
included in the prestigious Marquis Who’s Who in the World. For
his academic excellence, he received the Young Researcher
Award from the Venus International Foundation, India (2015).
He was the first Deputy Director of the Post-Induction Training
School of Defence Research and Development Organization
(2012).
xv
Nomenclature
aT Coefficient of thermal expansion

a Static pressure coefficient
b Compressibility
d** Kinetic energy thickness
d* Displacement thickness
d Boundary layer thickness
ƞd Isentropic efficiency of the inlet
ƞF Fan efficiency
C Circulation
Specific heat ratio
j von Karman constant
lt Eddy viscosity
l Mach angle
l Viscosity
”t Kinematic turbulent viscosity
” Kinematic viscosity
! Circular frequency
/ Velocity potential
…D Ratio of stagnation pressure
… Acoustic power
w Stream function
r Normal stress
¿D Ratio of stagnation temperature
¿ Shear stress
h Momentum thickness
a Speed of sound
CA Axial force coefficient
CD Drag coefficient
CL Lift coefficient
CM Moment coefficient
CN Normal force coefficient
CF Correction factor
c Chord
Dt Degree of turbulence
D Drag
Ek Ekman number
Eu Euler number
Fr Froude number
g Gravitational acceleration
ha Absolute height
hG Geometric height
xvii
xviii Nomenclature
H Shape factor
It Intensity of turbulence
IL Acoustic intensity level
I Acoustic intensity
Kmesh Mesh factor
Kn Knudsen number
lm Mixing length
L Lift
Mc Convective Mach number
M Mach number
NPR Nozzle pressure ratio
Prms Root mean square pressure
p Pressure
Rc Contraction ratio
rd Stagnation pressure ratio
Re Reynolds number
Ro Rossby number
SPL Sound pressure level
Uc Convective velocity
urms Root mean square velocity
U Internal energy
Wb Weber number
xCP Center of pressure
! Aerodynamic force
FA
c8 Specific heat capacity at constant volume
Cf Skin friction coefficient
cp Specific heat capacity at constant pressure
h Specific enthalpy
R Radius of Earth
Re Reynolds number
u, v, w Velocity components along x-, y-, and z-directions
Part I
Theoretical Aerodynamics
The Standard Atmosphere
1
The aerodynamic forces and moments that act upon an air- may be. That is, the material response of a fluid to the linear
borne vehicle are functions of the pressure, temperature, and stress is zero.
density of the surrounding air at that altitude. To provide an In addition, the difference in physical properties among
accurate prediction of aircraft performance, it is vital to mea- all the three states can also be witnessed at microscopic level.
sure these thermodynamic properties precisely. Thus, before In solids, the molecules are tightly packed together so that
discussing the aerodynamic aspects associated with an air- they find very small space to move freely and thus experi-
borne vehicle, it is essential to have a basic understanding of ence strong binding force. Consequently, the solid molecules
the atmosphere in which the vehicle is flying. In this chapter, possess low kinetic energy and in addition, they also have
we will study the various characteristics and phenomena asso- small vibrational energy, therefore, molecules remain fixed
ciated to the earth’s atmosphere. to their position causing a definite shape of the solids. In liq-
uids, the intermolecular spacing is large and despite of acting
strong intermolecular forces, the liquid molecules are allowed
1.1 Introduction to move freely to give the liquid its fluidity. In gases, due
to large intermolecular spacing and because of weak inter-
We know that the matter exists in three different states namely, molecular forces, the gas molecules have complete freedom
solid, liquid, and gas. The state of a substance depends on to move throughout the gas. Therefore, the gas molecules have
the intermolecular spacing among its molecules and at the enough kinetic energy to overcome the intermolecular forces
way in which these elementary particles move. A solid is the that holds the particles in solids and liquids, and if unrestricted
substance that has a fixed volume and has a fixed shape. In they will spread out as far as possible. As a result, a gas has
solids, the molecules are arranged in a regular pattern and due neither a definite shape nor a definite volume and once con-
to strong intermolecular cohesive forces they remain fixed at fined, it conforms to the shape and volume of the container.
their location, but vibrate. Since they cannot move from their Note that due to similar physical characteristics, the liquids
position and thus, solids are rigid. A liquid has fixed volume and gases are grouped together as fluid and the branch of sci-
but does not have fixed shape. The liquids conform the shape ence that studies the dynamics of both liquids and gases is
of the container which occupies them. In contrast, a gas has referred to as fluid dynamics.
no fixed volume and shape. They conform the shape as well The flow of a gas is, however, different from the flow of
as the volume of the container occupying them. Both liquids a liquid and moreover the properties of various gases such as
and gases are collectively termed as fluids. nitrogen, helium, etc., are different from each other. Thus, the
We can understand the technical difference between a solid fluid dynamics can be further classified into different streams.
and the fluid from solid mechanics point of view. The defor- The branch of science which studies the flow of liquids is
mation in a solid is produced when a force is applied in the called hydrodynamics and the branch of science which studies
tangential direction, called shear force, over the surface of the flow of gases is known as gas dynamics. Moreover, the
solid. From Hooke’s law, this deformation is directly propor- scientific stream which is dedicated to study the flow of air is
tional to the applied shear stress (i.e., shear force per unit known as aerodynamics.
area) within the elastic limit of the solid. In contrast, when Aerodynamics is that branch of science which finds several
the shear force is applied over the surface of a fluid, it deforms real-time applications in engineering. Using its theoretical as
continuously till the force is applied, no matter how small it well as experimental results, one can predict the aerodynamic
© Springer Nature Singapore Pte Ltd. 2019 3

M. Kaushik, Theoretical and Experimental Aerodynamics,
https://doi.org/10.1007/978-981-13-1678-4_1
4 1 The Standard Atmosphere
forces and moments generated on the object moving through atmospheric properties but considering these diversified
the air. For example, we may be interested in calculating set of data is not practical. Therefore, an International
forces and moments either over whole aircraft or any part of Standard Atmosphere is defined by Fédération Aéronautique
it. The branch of science that studies the flow field over an air- Internationale (FAI), responsible for maintaining the records
craft is known as aircraft aerodynamics. We may be interested for aeronautics and astronautics. The standard atmosphere
in studying the complete building aerodynamics or calculat- enumerates mean values of pressure, temperature, density,
ing forces on its various structures. We are also concerned in and other properties as functions of altitude. These values
knowing the aerodynamic drag on racing cars and bikes and are obtained from experimental balloon and sounding rocket
the effect of inverted wings in racing cars. These investiga- measurements combined with mathematical modeling of
tions come under the purview of industrial aerodynamics. The the atmosphere. It provides tables of common reference
aerodynamics finds important applications in spacecraft too; conditions utilized by aerospace engineers worldwide.
apart from studying the aerodynamics aspects, we are also
interested in studying the heat transfer characteristics of the
1.2 Composition and Structure of the
vehicle flying at hypersonic speeds. For example, one may be
Atmosphere
interested in designing the suitable “heat shield” for a space-
craft entering into the atmosphere at Mach 30. The branch
The atmosphere is very important to us as it protects life on
that studies the thermal aspects of trans-atmospheric vehicles
the earth. The International Standard Atmosphere as adopted
flying at high Mach numbers is referred to as aerothermody-
by FAI is broadly classified into two layers: primary and sec-
namics. All these fields of aerodynamics are grouped under
ondary. Based on the temperature distribution, the primary
the heading of external aerodynamics. Furthermore, we are
layer is divided into multiple layers, as shown in Fig. 1.2.
also interested in evaluating the flow properties inside the
The outermost primary layer shields the life on Earth by
rockets and air-breathing engines. One must know the flow
absorbing harmful ultraviolet solar radiations, warming the
conditions inside the test section of a wind tunnel, or inside a
earth’s surface by retention of heat. This phenomenon is pop-
pipe. These application areas, which are usually not exposed
ularly known as the greenhouse effect. The dry air is the main
to the atmosphere are studied under the heading of internal
constituent of the atmosphere which is composed of nitro-
aerodynamics.
gen (78.09% by volume), oxygen (20.5% by volume), argon
Irrespective of whether a problem belongs to the inter-
(0.93% by volume), and small traces of other gases.
nal or external aerodynamics, the investigation is carried out
considering the effect of atmosphere on the flow domain. For
example, to evaluate the aerodynamics forces and moments 1.2.1 Primary Layers in the Atmosphere
on an aircraft, one must know the thermodynamic and flow
properties of surrounding air. Spacecraft too experiences the The five main primary layers of the earth’s atmosphere are tro-
effect of atmosphere during their lift-off from the launch pad posphere, stratosphere, mesosphere, thermosphere, and exo-
till they move out to the outer space. Therefore, it is impera- sphere.
tive that to study the aerodynamics, we must have a thorough
understanding about the atmosphere first. 1.2.1.1 Troposphere
It is the layer closest to the Earth extending up to about 10–
1.1.1 Role of Atmosphere in the Aerodynamics 15 km above the earth’s surface. The troposphere is wider at
the equator but narrower at the poles. In this region, the move-
We have seen that the aircraft aerodynamics studies all those ment of free air caused on a large scale by thermal currents
vehicles which fly in the sensible atmosphere, whereas the in first 10 miles above the Earth surface. The air is heated by
spacecraft aerodynamics is dedicated towards the vehicles fly- the Sun’s radiation indirectly. The 40% part of the incident
ing outside the sensible atmosphere (Fig. 1.1). The sensible radiation lies in infrared range, absorbed by earth’s surface
atmosphere can be viewed as the envelope of gases surround- and then re-radiated to the troposphere. The radiations are
ing the Earth or any other planet. Technically, it can be defined predominantly absorbed by water vapor and carbon dioxide
as the thin boundary layer wrapping the earth’s surface. present in small traces in dry air. In this layer, the tempera-
The earth’s atmosphere is a dynamically changing system ture decreases at a rate of 6.5 ◦ C per kilometer, referred to as
in which the distribution of thermodynamic properties such as environmental lapse rate or simply lapse rate.
pressure and temperature depends on many factors: longitude
and latitude of the location on the earth, altitude, season in the 1.2.1.2 Tropopause
year, time in a day, and sunspot activities. It is obvious that It is the thin buffer layer between troposphere and stratosphere
the atmospheric properties are not constant due to these vari- and responsible for most weather phenomena occurring in the
able parameters. The design of aerospace vehicles requires atmosphere. Because of this layer, the water vapor cannot go
1.2 Composition and Structure of the Atmosphere 5
Aerospace Engineering
Aeronautical Science Space Technology

(Aircraft, Aerostat, Air−Balloons, etc.) (Spacecraft, Rockets, Lunar Probes, etc.)
Fig. 1.1 Classification of aerospace engineering
Exosphere
Ionosphere
Thermosphere
Mesopause
Mesosphere
Altitude (km)
Stratopause
Stratosphere
Tropopause
Troposphere
Temperature (K)
Fig. 1.2 Temperature variation with altitude in different atmospheric layers
higher and gets entrapped otherwise Earth would have lost all occur only in the tropopause. Thus, making flying possible
the water! for passenger aircraft in stratosphere without any appreciable
The solar radiations fall on the uneven surface of the Earth turbulence!
thereby heating the air in the vicinity of the Earth surface.
The warmer air rises and cooler air rushes in to fill the gap, 1.2.1.3 Stratosphere
producing the convection current. However, this warm air can- The layer just above the tropopause is known as stratosphere,
not go further up because the air in stratosphere is compara- roughly extending from 15 to 50 km. The temperature in the
tively more warmer and lighter preventing convection beyond bottom part is almost constant with altitude while it increases
the troposphere. It is interesting to note here that because of in the upper part of the stratosphere due to absorption of sun-
this convection prevention mechanism, the cloud phenomena light by the ozone gas in this region. One should note that the
presence of thick ozone layer is essential for life on the Earth 35 km. This layer has the ozone grouping of around 2–8 ppm
as it absorbs harmful ultraviolet solar radiations, responsi- (parts per million), substantially higher than the lower area
ble for skin cancer in humans. However, this layer is getting of the environment. The thickness of the ozone layer is
depleted especially by the use of chlorofluorocarbon (CFC) emphatically subject to the season in a year and geological
refrigerants such as CCl2 F2 (commonly known as R-12 or area.
Freon-12).
1.2.2.2 Ionosphere
1.2.1.4 Stratopause This layer begins at 50 km and extends up to 600 km, com-
This is the sandwiched layer between stratosphere and meso- prises of the mesosphere, the thermosphere, and a small por-
sphere which extends from 50 to 55 km. The temperature tion of the exosphere during the daytime. Ionosphere contains
increases with altitude in the upper stratosphere and reaches to charged particles called ions, responsible for radio communi-
its maximum value in this region. In stratopause, the temper- cation which are produced due to solar radiation. This layer
ature is almost constant with altitude (i.e., isothermal region). is responsible for a natural phenomena called aurora (also
referred to as polar or natural light) that occurs at higher alti-
1.2.1.5 Mesosphere tude. Aurora is a natural display of light, triggered when the
It is the middle layer of the atmosphere and hence termed solar wind muddles the magnetosphere adequately. Conse-
as mesosphere. This thin rarefied layer is located just above quently, the charged particles such as electrons and protons
the stratopause, however thick enough in slowing down and levitate into upper part of the atmosphere and lose their energy.
burning the meteors entering into the atmosphere. It extends The resulting ionization and stimulation of atmospheric ingre-
up to about 80 km above the earth’s surface. dients by these charged particles radiate the multicolored
light.
1.2.1.6 Mesopause
The mesosphere is bounded by an isothermal region, known 1.2.2.3 Homosphere and Heterosphere
as mesopause. It extends up to 90 km above the earth’s surface. The homosphere and heterosphere refer to the state of mixing
of the gases present in the atmosphere. They are classified
1.2.1.7 Thermosphere whether the atmospheric gases are thoroughly mixed. The
This layer is located just above the mesopause. The atmo- surface-based homosphere comprises the troposphere, strato-
sphere becomes highly rarefied at this altitude. Despite receiv- sphere, mesosphere, and the lowest part of the thermosphere,
ing extraordinary large amount of energy from the Sun, it where the chemical composition of the atmosphere does not
feels cold to us in reality. It is because of very low air density depend on molecular weight because the gases are mixed by
at this altitude. Thus, the probability of hitting by the fewer turbulence. This relatively homogeneous layer ends at the tur-
molecules and transferring any appreciable heat to our skin is bopause found at about 100 km.
extremely less. The thermosphere is further divided into two Above this altitude lies the heterosphere, which includes
regions. the most of the thermosphere and the exosphere. Here, the
chemical composition varies with altitude. This is because
1.2.1.8 Exosphere the distance that particles can move without colliding with
It is the outermost layer of the atmosphere, beginning from one another is large compared with the size of motions that
600 km and goes up to 10,000 km above the earth. Due cause mixing. This allows the gases to stratify by molecular
to highly rarefied region and lesser effects of gravity, the weight, with the heavier ones, such as oxygen and nitrogen,
molecules escape in outer space. present only near the bottom of the heterosphere. The upper
part of the heterosphere is composed almost completely of
hydrogen, the lightest element.
1.2.2 Secondary Layers in the Atmosphere
1.2.2.4 Atmospheric Boundary Layer
Notwithstanding the above vital layers of the atmosphere, The atmospheric boundary layer, also called planetary bound-
which are principally reliant upon the temperature, several ary layer, is the part of the troposphere that is adjacent to the
secondary layers can likewise be recognized in the light of surface of earth. It is about 1 km thick, which is in direct con-
different properties. tact with the ground, land or sea. The properties of this layer
change primarily through the turbulent diffusion. The gases
1.2.2.1 Ozone Layer in atmospheric boundary layer are dense and finely mixed
The ozone layer is predominantly situated in the lower during the daytime, whereas at night, it becomes thoroughly
fragment of the stratosphere, stretching out from 15 to stratified with sporadic mixing.
1.3 Interpretation of the Altitude 7
hG
ha
Fig. 1.3 Absolute and geometric altitudes
1.3 Interpretation of the Altitude The above assumptions make it convenient to use the
hydrostatic law along with other thermodynamic tools in
In literal sense, the altitude is termed as the height above relating pressure, temperature and other thermodynamic
the ground. However, in aerospace sciences, a total of six properties with altitude. Before we proceed further, it is
altitudes namely absolute, geometric, geopotential, pressure, essential to note up to what extent above the earth’s surface,
temperature, and density are used. It should be noted here the atmosphere exists to create aerodynamic forces. In
that the atmospheric data are tabulated with respect to sea reality, there is no fixed altitude at which atmosphere ends
level conditions as the reference state. The actual height of precipitously, rather it gets progressively thinner with the
the flying object above the sea level is termed as geometric increase in altitude. The Theodore von Karman was the
altitude. However, when the height is measured from the cen- first who calculated that at an altitude of about 100 km
ter of Earth by adding radius of Earth to the geometric altitude, above the sea level, the atmosphere becomes too thin to
then it is called absolute altitude. Thus, the relation between generate the lift. At this height, an aircraft will maintain the
the two is lift only if it flies at orbital velocity at which a spacecraft
revolves around the Earth without losing altitude. However,
ha = hG + R (1.1)
below this line, the lift on an aircraft can be maintained at
suborbital velocities. This level is commonly referred to as
where R is the radius of earth (Fig. 1.3).
Karman Line which demarcates between the Earth’s sensible
We know that the local acceleration of gravity varies
atmosphere and outer space. Eventually, it constitutes the
with the absolute altitude ha . Recall, Newton’s law of gravi-
realm of aeronautics and any other aerial activities above this
tation,
line come under the domain of astronautics.
2 Even though an aircraft can fly up to Karman Line, the
R
g = g0 (1.2) human physiological considerations further restrict its oper-
ha
ation. This limit is known as Armstrong Line or Armstrong
or Limit, named after Harry G. Armstrong who first recognized
2 this phenomenon. At this altitude, the ambient pressure is so
R low (0.0618 atm or 6.3 kPa) that the water starts boiling even
g = g0 (1.3)
hG + R at normal temperature of human body (37 ◦ C or 98.6 ◦ F). In
literature, this altitude has been variously reported between
where g0 is the gravitational acceleration at sea level and g is
the range of 18.8–19.4 km. This physiological limit necessi-
the local gravitational acceleration at a given absolute altitude
tates the wearing of pressurized suits in order to survive.
ha . The expressions which relate pressure, temperature, and
other thermodynamic properties with the altitude are obtained
based on the following assumptions. 1.4 Variation of Pressure in the Standard
Atmosphere
• The atmosphere is homogeneous.
• The changes in thermodynamic properties are
We know that pressure varies with altitude in the atmosphere.
reversible.
In this section, we will investigate the variation of pressure in
• The long stretch of fluid is substantially at rest.
the stationary atmosphere. Let us consider a small cylindrical
air column of radius r and height dhG as shown in Fig. 1.4. It

is assumed that the air column is stationary in the atmosphere, 2
h hG R
i.e., it is not moving. The air element is in equilibrium under
dh = dhG
the following forces. 0 0 R + hG

1. Pressure force (p + dp) r 2 acting in downward direc- or
tion at the
top surface.
hG 1
2. Weight ρgr 2 dh h = R2 dhG
G 2acting
in downward direction.
0 (R + hG )2
3. Pressure force pr acting in upward direction at the
bottom surface. or
hG
From the force balance, we have −1
h = R2
R + hG 0
(p + dp) r 2 + ρgr 2 dhG = pr 2
Hence,

dp = −ρgdhG (1.4) R
h= hG (1.8)
R + hG
Equation (1.4) is indeed a hydrostatic law of pressure. Here,
g is gravitational acceleration due to gravity at a given geo- This is the relation between geopotential altitude (h) and geo-
metric altitude hG . However, if g is assumed to be constant metric altitude (hG ).
throughout and is replaced with g0 (i.e., sea level value) in
Eq. (1.4), then we have
1.6 Distribution of Properties in
Troposphere and Stratosphere
dp = −ρg0 dh (1.5)
Based on experimental evidence, the temperature distribution
in the standard atmosphere is shown in Fig. 1.2. In this figure,
1.5 Relation Between Geopotential and the isothermal as well as gradient regions are clearly seen.
Geometric Altitude In this section, we will formulate the variations of pressure,
temperature, and density with altitude for both the aforesaid
A closer look at Eq. (1.5) suggests an interesting observation. regions.
Unknowingly, we have defined a new altitude where acceler- First, we will consider the isothermal layer (i.e., tropopause
ation due to gravity is constant equal to its value at sea level. in our case) with known values of pressure, temperature, and
This fictitious altitude h is known as geopotential altitude. density at the base of isothermal layer. Recall, Eq. (1.5)
Reader should note that because of small difference between
g and g0 , the hG is also slightly from h. dp = −ρgo dh
Dividing Eq. (1.4) with (1.5),
Dividing by equation of state,
g dhG
1= p = ρRT (1.9)
g0 dh
or We obtain
g dp −ρg0 dh ρgdh
dh = dhG (1.6) = =−
g0 p ρRT RT
or
2
R dp −g0 dh
dh = dhG (1.7) = (1.10)
R + hG p RT
Integrating Eq. (1.7) between sea level and a given point, Integrating between p1 and p; and h1 and h limits,
1.6 Distribution of Properties in Troposphere and Stratosphere 9
p + dp
dh
p
Fig. 1.4 Small cylindrical air column
where1 a is called the environmental lapse rate for gradient

layers. Equation (1.14) can be rewritten as,
p h
dp g0
=− dh dT
p RT dh = (1.15)
p1 h1 a
From Eq. (1.10), we have

p g0
ln =− (h − h1 )
p1 RT dp −g0 dh
=
or p RT
p −g0 or
= exp (h − h1 ) (1.11)
p1 RT dp −g0 dT
=
p RT a
Again from Eq. (1.9),
Integrating the above equation between the known values of
p ρRT ρ pressure, temperature, and density (p1 , T1 , ρ1 ) at the base of
= =
p1 ρ1 RT ρ1 gradient layer and some point at altitude h, also located in the
gradient layer.
ρ −g0
= exp (h − h1 ) (1.12) T
ρ1 RT p dp −g0 dT
=
p1 p aR T
Equations (1.11) and (1.12) are the pressure and density dis- T1
tributions in isothermal regions of the standard atmosphere
such that or
p ρ p −g0 T
= ln = ln
p1 ρ1 p1 aR T1
or or
p = (const) ρ (1.13) −g0

p T aR
= (1.16)
p1 T1
Now let us consider the gradient layers of the atmosphere
(troposphere and stratosphere in our case) with known val- From equation of state (Eq. (1.9))
ues of pressure, temperature, and density (p1 , T1 , ρ1 ) at the
base of the layer. Assuming the linear variation in temperature
distribution, the gradient of temperature will be 1 The lapse rate is defined as the rate of change of temperature with
altitude. It is of two types. First one is the environmental lapse rate,
(T − T1 ) dT which is measured experimentally. The other one is adiabatic lapse rate,
= =a (1.14) defined as the temperature change experienced by an air parcel when
(h − h1 ) dh displaced adiabatically in vertical direction.
p ρ RT atmospheric pressure. It is dependent upon the Earth’s gravi-

=
p1 ρ1 RT1 tational force acting on the column of gases over a particular
location. A barometer is the instrument, which is used to mea-
or
sure the atmospheric pressure that corresponds to 760 mm of
p ρ T Hg (mercury). In SI units, it is equal to 1 atm or 101325
=
p1 ρ1 T1 Pa. From Eq. (1.11), it is evident that the pressure decreases
exponentially with the increase of altitude.
Thus,
−g0
ρ T T aR 1.7.2 Temperature
=
ρ1 T1 T1
The variation of temperature is shown in Fig. 1.2. It is evident
Finally, that up to an altitude of 11 km above the sea level, the temper-
− g0 +1 ature decreases with increase of altitude. However, beyond
ρ T aR this altitude, the temperature stabilizes throughout the rest of
= (1.17)
ρ1 T1 troposphere. In stratosphere, the temperature further begins
to rise at an altitude of about 20 km and beyond. Essentially,
Using Eqs. (1.16) and (1.17), the pressure and density ratios it is the region where the ozone layer exists. The ultraviolet
can be related as, radiations from the Sun are absorbed in this layer and thus the

g0 temperature increases consequently. Once again, the temper-
p ρ g0 −a R
ature increases in thermosphere apparently at an altitude of
=
p1 ρ1 90 km (Fig. 1.2).
We know that for an ideal gas, the speed of sound is a
Therefore, the pressure and density are related as function of temperature only. Interestingly, the variation in

g0 speed of sound with altitude shows the similar trend as that
p = (constant) ρ g0 −aR
(1.18) of temperature.
In troposphere, the environmental lapse rate a = 6.5 ◦ Ckm−1

and R = 287 Kg−1 JK−1 . Hence, 1.7.3 Mass and Density
p = kρ1.235 (1.19) At sea level conditions, the density of air is found to be

1.2 kgm−3 which decreases with the increase in altitude.
where k is an arbitrary constant. Thus, the movement in static The total mean mass of the atmosphere is estimated as
air from one altitude to another is a polytropic process with 5.148 × 1018 kg, which varies with the changes in water
process exponent n = 1.235. From Eq. (1.14), the tempera- vapor loading. The mean mass of the water vapor is roughly
ture variation in gradient layers of the atmosphere can be calculated as 1.27 × 1016 kg.
written as
T = T1 + a (h − h1 ) (1.20) 1.7.4 Scattering Phenomena
The phenomena of light scattering is also known as Rayleigh

scattering. We know that the light is an electromagnetic radia-
1.7 Physical and Optical Properties of the tion which can be viewed as either a wave or as a particle. For
Atmosphere scattering of light by the particles, wave theory is most advan-
tageous. Rayleigh scattering is an elastic scattering of waves in
In this section, we discuss some important physical and optical
which the kinetic energy is conserved after collision. It occurs
properties associated with the atmosphere surrounding the
when the electromagnetic radiations interact with the gaseous
earth.
particles which are much smaller in size than the wavelengths
of these waves. It should also be noted that for scattering the
1.7.1 Pressure light rays, the refractive index of these atmospheric parti-
cles must be closer to one. The Rayleigh scattering occurs
Pressure is defined as the force per unit area which acts because of the polarizability of an individual molecule. The
perpendicular to a surface. Thus, the pressure exerted by intensity of wave scattering through a single molecule can be
the gases in the atmosphere on a surface is referred to as defined as
1.7 Physical and Optical Properties of the Atmosphere 11
8π 4 α2
face temperature of about 6000 K emits the radiation peak near
I = I0 1 + cos 2
θ (1.21)
λ4 R 2 to 500 nm. On the other hand, the Earth with surface temper-
ature of approximately 290 K emits radiation peaks near to
where I = resulting intensity, I0 = original intensity, α = 10,000 nm which is beyond the visible range of human eye.
polarizability, R = distance to the particle, and θ = scattering
angle.
Equation (1.21) shows that the scattering intensity is 1.7.6 Refractive Index
inversely proportional to the wavelength and thus, shorter the
wavelength the more is the scattering and vice versa. Since The refractive index of an electromagnetic radiation is defined
the blue and green have shorter wavelengths and hence they by
scatter easily therefore, the sky looks blue. However, at very √
high altitudes, the Sun rays are incident at 90◦ with respect to n= r μr (1.24)
the scattering molecules and thus, the light is being scattered
less and less. Consequently, we see the longer wavelengths where r = relative permittivity and μr = relative permeabil-
such as red and yellow. Because of this, the Sun appears ity of the material. Further, we know that the refractive index
yellow in color. (n) of the air is very close to, but slightly greater than 1. Thus,
the refractivity of the atmosphere can be defined as
1.7.5 Absorption and Emission N = 106 (n − 1) (1.25)
In order to understand the phenomena of absorption and emis- The refractivity depends upon many parameters including
sion of radiations, the light is viewed as the propagation pressure (p), temperature (T), and humidity (e). A commonly
of discrete parcels of energy, called photons. The absorp- used expression is
tion of electromagnetic radiation is defined as the process
77.6 4810 e
in which incident radiation is retained by the surface. Differ- N= p+ (1.26)
T T
ent gases present in the atmosphere absorb different wave-
lengths of radiation. The oxygen and ozone gases absorb the where p is in mbar, T is in K and e is the partial pressure
radiation of wavelength shorter than 300 nm, whereas water of water vapor in mbar. From Eq. (1.26), it can be seen that
vapor absorbs all the wavelength greater than 700 nm. When refractivity is directly proportional to pressure and humid-
a gaseous molecule absorbs a photon, it excites the transla- ity but it is inversely proportional to the temperature. If we
tional, rotational, and vibrational modes of energies of the assume the atmosphere as stratified medium composed of thin
molecule. Absorption of photons increases the temperature planar layers, each having different refractive index, which
of the atmosphere. changes with altitude as shown in Fig. 1.5.
The emission of an electromagnetic radiation is the oppo- From Snell’s law,
site to that of absorption. The surface temperature decreases
by emitting the radiations. Further, the Stefan–Boltzmann law n0 sin α0 = n1 sin α1 = n2 sin α2 = · · · = nn sin αn = · · ·
states that the maximum emissive power at a given tempera- (1.27)
ture is the black body emissive power (Eb ).
From Eq. (1.26), it is evident that as we go to higher altitudes
Eb = σT4 (1.22) then the refractivity decreases due to decrease in pressure.
Further, from Eq. (1.27) if
where σ = Stefan–Boltzmann constant. Furthermore, accord-
ing to Wein’s displacement law the maximum wavelength n0 > n1 > n2 > · · · > nn · · · .
(λmax ) associated to the emitted radiations decreases with the
increase in object’s temperature. That is, Then,
λmax T = 2.898 × 10−3 (1.23) α0 < α1 < α2 < · · · < αn · · · .
From Eqs. (1.22) and (1.23), it can be seen that the hot objects i.e., because of the refraction phenomena the light, initially
will emit more amount of radiations with lower wavelengths, launched vertically, bend towards the Earth’s surface. In other
whereas the relatively colder objects emit less radiations with words, the light ray follows the long optical path. This is why,
longer wavelengths. This is why the Sun, which has the sur- in some occasions, the observers on-board ships can see other
αn
nn
α 50
α n−1
n
50
α4
α
49
n
α3 3
α3
n
2
α2
α2
n1
α1
Fig. 1.5 A schematic diagram of stratified layers in the atmosphere
vessels just over the horizon because the light is refracted and
in the same direction following the curvature of the earth’s
surface. p = cρ
The above two equations can be written in generalized form

1.8 Static Stability Analysis of Troposphere as
Layer
p = c1 ρn (1.28)
The stability of atmosphere in hydrostatic equilibrium with
respect to vertical displacements is known as static stability. Further, if the upward movement of the air parcel is very
It measures the gravitational resistance on the air mass when fast, then the process can be considered as adiabatic without
displaced in vertical direction. The concept of static stability any appreciable amount of heat transfer across the parcel’s
in the atmosphere can be understood with the help of a lump boundaries. The thermodynamic condition of the parcel can
of fluid better known as the air parcel. Let us assume that be related as
the initial perturbation is given to an air parcel, in turn, the
following possibilities may arise (Fig. 1.6). p∗ = c2 ρ∗γ (1.29)
• If the displaced air parcel returns to its undisturbed loca- From the equation of state, for ambient air, the temperature
tion, the original equilibrium atmosphere is referred to as ratio is related to the static pressure at the altitudes, z1 and z2
statically stable. as
• If the displaced air parcel continues to move upward fur-
ther, the atmosphere is referred to as statically unstable. n−1
Tb pb n
• If the displaced air parcel takes up the new position and = (1.30)
Ta pa
remains there, the atmosphere is termed as neutrally stable.
Similarly, for the air parcel, the temperature ratio can be
Consider an air parcel that undergoes a forced displacement
related to the static pressure at different altitudes by using
in the atmosphere from an altitude z1 to z2 , as depicted in
equation of state as
Fig. 1.7. The state of the air parcel is denoted by the variables
with asterisks (∗) and the ambient conditions are denoted by γ−1
Tb∗ p∗b γ
the quantities without it. We know that for the troposphere and = (1.31)
the stratosphere layers, the ambient conditions are related by Ta∗ p∗a
Eqs. (1.13) and (1.18), respectively. That is,
Here, it should be noted that before the forced displace-
ment of air parcel, it was in thermodynamic (i.e., thermal,
p = cρ 1.235
1.8 Static Stability Analysis of Troposphere Layer 13
Air parcel Statically unstable
Initial perturbation Statically stable Neutrally stable
Fig. 1.6 Schematic diagram of the different types of stability
z − direction
p b ρ b Tb
p b* ρ b* T b* z2
Increasing Movement
of Air parcel
Altitude
Air parcel
p a ρ a Ta
p a* ρ a* T a* z1
Fig. 1.7 Schematic diagram of upward movement of air parcel
mechanical and chemical) equilibrium with its surroundings or

at altitude z1 . So, we have, p∗a = pa , Ta∗ = Ta and ρ∗a = ρ∗a . γ−n
Tb∗ pb γn
However, once the air parcel is reached to the altitude z2 , it = (1.33)
is still under mechanical equilibrium with its surroundings, Tb pa
required to ascertain the expansion work to be zero. That is,
From the equation of state and by using Eq. (1.32), we get
p∗b = pb (1.32) Tb∗ ρb
= ∗
Tb∗ Tb ρb
From algebraic rearrangement, we can write Tb as
or
Tb∗ T∗ T∗ γ−n
= b∗ × a ×
Ta ρ∗b pa γn
Tb Ta Ta Tb = (1.34)
ρb pb
But we have seen that, Ta∗ = Ta at the altitude z1 , thus Based on the relative magnitudes of n and γ in Eq. (1.34), we
can consider the following cases:
γ−1 n−1
Tb∗ T∗ Ta p∗b γ pb n
= b∗ × = × 1. If n < γ, then γ−n ∗
Tb Ta Tb p∗a pa γn > 0 which implies ρb > ρb , thereby
showing that the air parcel becomes heavier than the sur-
or rounding air at the new altitude. Consequently, the parcel
is prevented from further uprising in the atmosphere and
γ−1 n−1
Tb∗ p∗b γ pa n will have a tendency to come back to its original position
= ×
Tb p∗a pb reflecting the static stability. Hence, a statically stable air
resists the vertical motion of an air parcel.
2. Further, if n > γ, then γ−n γn < 0 which implies ρb < ρb

∗ In Fig. 1.10, −
→
r is in the fixed frame of reference and −
→r 1
thereby, showing that the air parcel is lighter than the sur- is in rotating frame of reference. The velocities in the fixed
rounding air at the new altitude and because of this reason, and rotating frame of reference, are related by the following
the parcel is free to keep rising further in the atmosphere. relation.
This behavior of the air parcel indicates static instability.
d−
→r1 − d−
→
ω ×−
→ r
Therefore, an unstable air promotes the vertical movement +→ r = (1.36)
of the air parcel. dt dt
3. However, if n = γ then γ−n γn = 0 which, in turn, implies The acceleration will be related as
ρ∗b = ρb . That is, the air parcel will have to stay at the new

altitude and this state of the atmosphere is referred to as d2 −
→r d d
the neutral stability. Note that, the exponent n is defined as 2
= +−
→
ω ω −
+−
→ →
r
dt dt 1 dt 1
g0
n= (1.35)
g0 − aR → −
d2 −
→r d2 −
→
r1 ω ×→
d − r 1 d−
→
r1 − → −
2
= 2
+ +−
→
ω × +→ ω ×→
ω × − r
dt dt
and for n = γ, the above equation gives a lapse rate (aNS ) dt dt
as

g0 γ − 1 d2 −
→r d2 −
→
r1 d−
→r 1 → −
aNS = =
9.8 × 0.4
2
= 2
+ 2 −→
ω × +−
→ ω ×→
ω × − r
R γ 287 × 1.4 dt dt dt
or or
d−
→v d−
→
v1 → − → −
(a)NS = 9.75 ◦ Ckm−1 = ω ×→
+2 − v 1 +− ω ×−
ω × → →
r
dt dt
where −
→
Here, aNS is also known as adiabatic lapse rate.
v = velocity relative to the fixed frame of reference,
−
→
v 1 = velocity relative to the rotating frame of reference, and
From the above discussion, it is clear that the disturbed air −
→
parcel will take either of the above three possible states of ω = velocity relative to the rotating frame of reference.
equilibrium (schematically shown in Fig. 1.6). −
d−
→v d→v → − → −
= ω ×→
+2 − v 1 +− ω ×−
ω × → →
r
dt dt 0
1.9 Definition of Wind (1.37)
→ −
Wind is the movement of free air at a large scale caused by the In Eq. (1.37), 2 − ω ×→ v 1 is termed as Coriolis force which
thermal currents in the first 15 km above the earth’s surface. moves the air particle to the right of its direction of motion in
The large-scale movement of air takes place because of dif- northern hemisphere and to the left in southern hemisphere.
ferential heating between polar and tropical regions and also If θ is the latitude then in the scalar form, the Coriolis force
due to rotation of Earth. will be given by Eq. (1.38).
Over the globe, the hot areas such as tropics will have
→ −
a warm column of air above the surface, whereas the cold Fc = −ω →
v sin θ (1.38)
areas like poles have a dense cold column above them. Con-
sequently, the pressure in tropics is found to be more than that The appearance of this force may be explained in the follow-
of the polar regions leading to the air circulation (Fig. 1.8). ing way. If a particle moves from the equator to the north, it
The idealized square circuit is circular in reality which crosses the latitude with decreasing circular radius. In order
is the basic weather pattern in both horizontal and vertical to preserve its angular momentum, the particle must rotate
planes. It is also seen that the general circulation of air takes faster than the Earth and thus will be deflected to its right.
place from poles to tropics. Moreover, we know that the shape A particle which travels from the north to the equator will
of the Earth is an oblate spheroid. If we assume that the axis of be retarded relative to the Earth and, hence, it will also be
rotation of the earth passes through the poles, then the rotation deflected to the right. However, in southern hemisphere, the
of Earth would be from west to east, as shown in Fig. 1.9. deviation will be towards left. From these deviations of the
From Fig. 1.9 it is seen that the easterly velocity of Earth particle, one may conclude that a force exists, a Coriolis force!
is ωR sin θ and acceleration towards the axis of the Earth is The effect of the earth’s rotation becomes apparent in the mid-
ω 2 R sin θ. latitudes. By virtue of earth’s rotation, each air particle in the
1.9 Definition of Wind 15
P1 P2
Warm air
Cold air
Fig. 1.8 Large-scale air movement
North
Pole star ω Zenith
East
ω
R
O β
Fixed longitude
Fig. 1.9 Non-rotating frame at earth center
ω
P
r1
r
θ
O
Fig. 1.10 Relation between fixed and rotating frames of reference
atmosphere has an angular momentum directed from west to 1.9.2 Hurricanes or Typhoons in Tropics
east. When the air particles move towards poles maintaining
the same height above the Earth’s surface, its radius of gyra- It has been observed that under specific conditions, the
tion decreases. Hence, to conserve the angular momentum, atmospheric disturbances can originate near the equator and
the velocity (in the west to east direction) also increases. This develop within a few days into vortices of 500–1000 km in
effect is small at equator but large in temperate zones. diameter. These tropical storms can generate very high wind
A fluid that would flow radially into a sink without the velocities near the center of rotation. If the tangential velocity
influence of the earth’s rotation takes on a counterclockwise exceeds 120 kmh−1 , the storm is called hurricanes. They
rotation in the northern hemisphere due to the effect of Cori- are also called typhoons in western Pacific and cyclones in
olis force. This rotary fluid flow is called cyclone. On the Australia and Indian Ocean. Some local names are Cardonazo
other hand, a source flow deflected in clockwise direction in the west coast of Mexico, Baguio in Philippines, and
in the northern hemisphere is termed as anticyclone. Low- Willy-Willy in Australia.
pressure regions are cyclonic and high-pressure regions are The hurricanes are tropical depressions smaller in area but
anticyclonic. more severe than the temperate zone depressions. They orig-
The rotation of Earth is capable of generating vortices in inate over warm ocean surface with a minimum temperature
the atmosphere and the ocean through a redistribution of vor- of 27 ◦ C. The energy is transmitted to the air by the latent heat
ticity. These vortices are different from those generated in of evaporation of water sucked into the low-pressure center.
non-rotating system, where the pressure always has a mini- Near the center, the humid or moist air is deflected upward and
mum at the center, however, the pressure in the vortex of a cools through adiabatic expansion which leads to the conden-
rotating system can have a minimum or maximum, depending sation of humidity. The condensed water returns to the sea in
on the direction of rotation. In cyclonic rotation, the pressure the form of torrential rainfall. Consequently, the latent heat is
has a minimum but, it has a maximum in an anticyclonic rota- thereby freed and warms the air which flows upward through
tion. buoyancy. During this process the heat energy is converted
into kinetic energy. The pressure difference, between the cen-
1.9.1 Cyclones and Anticyclones in Temperate ter of storm and its ambient, is found to be 90 mb. The lower
Zones the pressure, more intense is the storm. The rotation of the
tropical storm is due to Coriolis force, which causes the air
The cyclones are also termed as depressions. They are majorly being sucked into the central low-pressure region to circulate.
found in temperate zones. The depressions and anticyclones The rotating air rises upward which is saturated with water,
are caused by meeting of warm and cold streams of air, as cools and condenses, forming clouds. Hurricanes do not occur
illustrated in Fig. 1.11. in a belt of 5◦ –8◦ on either side of the equator because the
The cold and hot air fronts mix in circulatory manner and Coriolis force vanishes at the equator. The paths followed
this process continues for several days until the cold front by hurricanes are entirely random and the main affected areas
catches the warm air. Then, the depression dissolves into a from them are North America and Chinese coast. The destruc-
uniform mass of cold air. An idealized depression has a belt tive feature is not only their speed but also the tidal surge
of rain or snow surrounding the warm air. The depression is which follows.
imposed upon general flow, which is in the direction of warm
air stream. Depressions are also called extra–tropical cyclones
due to their anticlockwise rotation. They form when the warm
1.10 Geostrophic and Ageostrophic Winds
air adjacent to the Earth’s surface moves updward and in turn
Let us consider an atmospheric motion, in which only the pres-
the local empty space created is filled by cold air, rushing
sure and Coriolis forces are acting (Fig. 1.13). At steady state
towards it. Thus, upward movement of air must be compen-
and in the absence of friction, the Coriolis force is balanced
sated by a downward movement of air somewhere else caus-
by the pressure gradient force, leaving a resultant motion
ing a condition complementary to the original depression. The
approximately along the lines of equal pressures, referred to
slow downward movement of air compressed it and warms it,
as isobars.2 The wind in the direction of isobars is called
so that Earth’s warm air beneath cannot escape by convection.
geostrophic wind. It is given by Eq. (1.39).
These are of high pressure with widely spaced isobars and low
winds, which may consist of fine weather, uniform cloud, fog,
or smog. Further, if this region of high pressure has circulat-
2 The relation between streamlines and isobars is important in the inter-
ing winds, it is called anticyclone (Fig. 1.12). Furthermore, the
pretation of weather maps. If only pressure and Coriolis forces are acting
slow thermal currents rise and fall after the processes called in atmospheric motions, the isobars coincide with the streamlines for a
convergence and divergence, which result in succession of steady flow. This state which represents equilibrium between pressure
high- and low-pressure areas common in midlatitude (tem- and Coriolis forces is called geostrophic.
perate zones) regions of northern and southern hemispheres.
1.10 Geostrophic and Ageostrophic Winds 17
Cold air Direction of motion of center
Rain
Warm air
Fig. 1.11 Schematic of an idealized depression
Stormy cloudy weather Calm clear weather

Diverging surface Converging surface
Hot rising air Cold subsiding air
Converging surface Diverging surface

(Low pressure zone)
(High pressure zone)
Cyclonic flow Anticyclonic flow
Fig. 1.12 Cyclonic and anticyclonic flows
1 ∂p dz
vg ∝ (1.39) +g = −2vg sin (1.41)
ρ ∂x dn
We can see that the geostrophic wind is inversely proportional or

to the spacing of isobars with low pressure at its left. The total g dz
acceleration experienced by air/water parcel will be due to vg = − (1.42)
2 sin dn
both pressure gradient and Coriolis force. Thus,
The above velocity is known as geostrophic wind. In all the

dv dz dz cases in which the total acceleration is small compared to
= −g − g − 2v sin (1.40)
dt ds dn Coriolis acceleration for a given pressure gradient, the wind
will be nearly geostrophic. These conditions are generally met
where = angular velocity of the earth, and = earth’s at higher altitudes, where the surface features affect less
latitude. Further, in the absence
of friction, the fluid parcel (Fig. 1.14).
does not accelerate dv dt = 0 . When the force due to pressure The assumption of a geostrophic state does not hold under
gradient normal to flow alone balances the Coriolis force, then the following flow situations.
L
Pressure gradient force
1000 mbar
Geostrophic
1004 mbar wind
Coriolis force
1008 mbar
H
Fig. 1.13 Geostrophic wind balance
Geostrophic wind
Low pressure
Pressure gradient
High pressure
Isobars
Coriolis force
Fig. 1.14 Balance of forces over air/water parcel
Pressure gradient Real wind

force
L
1000 mbar
Geostrophic wind
Friction Coriolis force

1004 mbar
1008 mbar
H
Fig. 1.15 Imbalance of forces due to friction over air/water parcel
• We know that due to the frictional forces exerted by the sur- Consequently, the pressure gradient force becomes more
face of Earth, an atmospheric boundary layer forms in the dominant and the wind is no more geostrophic. As a result,
troposphere layer,3 which slows down the wind. The reduc- the total wind deflects slightly towards lower pressure.
tion in wind speed subsequently reduces the Coriolis force. The amount of deflection which the surface wind has with
respect to the geostrophic wind is known as ageostrophic
3 The height (thickness) of atmospheric boundary layer can vary depend- wind (Fig. 1.15).
ing on the type of terrain, wind, and vertical temperature profile. The • The geostrophic state assumption is also not valid when the
time of day as well as the season of year also affect its height. However, streamlines are strongly curved. However, a state similar to
usually it exists from the Earth’s surface to about 300 m to 2 km above
the geostrophic one can be reached if the centrifugal force
it.
1.10 Geostrophic and Ageostrophic Winds 19
is taken into account. In a flow where the pressure, Coriolis, ing objects to a force to the right of the direction of motion in
and centrifugal forces are in equilibrium, the streamlines the northern hemisphere and to its left in southern hemisphere.
coincide with the isobars. Such a flow is called gradient When the surface water molecules move by the force of
flow (Fig. 1.16). the wind, they, in turn, drag deeper layers of water molecules
below them. Each layer of water molecules is moved by
Near a solid or liquid surface, the assumption of a geostrophic the friction from the shallower layer and each deeper layer
or a gradient flow is not valid because the adherence of the moves more slowly than the layer above it, until the move-
fluid at the surface causes a boundary layer. We know that in ment ceases at a depth of about 100 m. Like the surface water,
the absence of rotation, the streamlines do not change their the deeper water also deflects in different directions in both
direction in a parallel flow. The direction of fluid motion in hemispheres due to Coriolis effect. Consequently, this rela-
boundary layer is same as the outer flow. However, under the tive velocity between the layers creates a spiral effect, i.e., a
influence of Coriolis force, the boundary layer profile will twist around the flow and opposite in direction to the surface
be twisted. If the axis of rotation is perpendicular to wall, the current (Fig. 1.17), which is known as Ekman spiral.
boundary layer will be twisted to its right in the northern hemi-
sphere. This twisted boundary layer is known as Ekman Layer
named after its inventor V.N. Ekman (1874–1954). If the axis
1.12 Global and Local Winds
of rotation is parallel to the wall, a Stewartson Layer develops
Due to uneven heating over the earth’s surface, the winds are
named after fluid dynamicist Kaith Stewartson (1925–1983).
caused by setting the moving convection currents. The large-
In both atmospheric and oceanic motions, the centrifugal
scale convection currents are responsible for the global winds,
and the Coriolis forces caused due to the rotation of Earth are
whereas the local winds are caused by small-scale currents.
always present. From Fig. 1.16, it is evident that the pressure
We know that the shape of Earth is an oblate spheroid.
gradient and the centrifugal force act in same direction and
Because of this, some parts of the Earth receive the solar radi-
therefore, the present discussion is restricted to Coriolis force
ations directly as compared to the other parts. For instance,
only. Although, these forces are weak in nature but, whether
the surface at the equator receives the daylight more directly
or not they may be neglected that depend on the magnitude
than the surface at the poles. As the hot air over the equator
of other coexistent forces. In the presence of inertial and fric-
rises, the cold air from the poles surges towards the equator
tional forces, two non-dimensional numbers can be defined.
to occupy the void created by the hot air. This trade of hot
Rossby Number and cold air that happens between the equator and the poles
create the global wind belts.
inertial force v v The local winds are generated because of the small-scale
Ro ≡ ≡ = (1.43)
coriolis force Lf L (2 sin φ) convection currents caused by heating; along a coast and in
the mountains. These winds blow over a much smaller area
where v = characteristic velocity, L = characteristic length,
and change their direction as well as speed within a short
and f = 2 sin φ = Coriolis parameter, where φ is latitude.
period of time than their global counterpart.
Ekman Number
frictional force ν ν 1.12.1 Land and Sea Breeze

Ek ≡ ≡ 2 = 2 (1.44)
coriolis force L f L (2 sin φ)
We know that during the daytime, land heats up faster than the
where ν = kinematic viscosity. water. The warmer air over land rises, while the cooler air over
Furthermore, if the Coriolis force is very small (f = 0), the ocean rushes in to take its place. This wind is called a sea
the effect of Earth’s rotation on fluid motion can be neglected. breeze. However, after sunset, the land cools down faster than
Consequently, both Rossby number and Ekman number will the water. The warmer air over the ocean rises while cooler
be infinitely large. In reality, aforesaid condition is obtained air over land rushes in to take its place. This wind is called a
only at equator. At other locations, the effect of Earth’s rota- land breeze.
tional motion can never be ignored. The specific heat of water is an important factor in moder-
ating global climate, global, and local weather patterns. The
high specific heats of water allows the oceans to function as
1.11 Ekman Spiral huge energy sinks that can transfer large amounts of energy
from one area to another. Conversely, the low specific heats
The Swedish scientist V.N. Ekman (1874–1954) was the first of dry ground cause its temperature to increase more rapidly
who put forward the theory of Ekman Boundary layer in 1902. in response to heating from the Sun, and to decrease more
It is a consequence of the Coriolis effect, which subjects mov- rapidly when the Sun goes down. Further, in the regions hav-
Centrifugal force Centrifugal force
Coriolis force Pressure gradient
Wind Wind
Pressure Coriolis
gradient force
Low High
Fig. 1.16 Typical gradient flows in the northern hemisphere
o
45
Direction of wind
Net water
transport
Surface water currents

Fig. 1.17 Formation of Ekman spiral in the ocean
ing land areas closer to large bodies of water, this difference i.e., lower part of troposphere. The thin belts of high-speed
in heating and cooling rates of land and water results in air winds in upper troposphere and lower stratosphere are known
movement (i.e., termed as wind) (Fig. 1.18). as jet streams. They blow from west to east all the way around
the earth.
1.12.2 Zonal Wind Directions
1.12.3 Some Specific Names of the Wind
The zonal wind directions at different latitudes on the Earth
is depicted in Fig. 1.19. In the region of westerlies, the wind All over the world, the local winds are known by different
has opposite directions to the general flow in both hemisphere. names. Some of them are discussed below.
Doldrums are the regions of little or no winds. The subtropical
high-pressure belts are also areas of light and variable winds. 1. Bricfielder: It is the hot and dry air that carries copious
The trade winds found in tropical zones are steady and of red dust from the desert of the southern Australia.
medium to low velocity. It should be noted that westerlies and 2. Cape Doctor: It is the cold and dry wind which blows
subtropical claws are the regions of very disturbed weather from the southeast direction in South Africa. This wind
with many cyclones and anticyclones. decontaminates the city of Cape Town.
The polar easterlies, temperate westerlies, and tropical 3. Chinook: It is the hot air that rushes eastward down the
trade winds are the winds which blow near the earth’s surface, rocky mountains of Canada and USA.
1.12 Global and Local Winds 21
Cold air Warm air Warm air Cold air
Ocean Land Ocean Land
(a) Sea breeze (b) Land breeze
Fig. 1.18 Schematic diagram of a the sea breeze and, b the land breeze
Polar high
Sub−polar low
Westerlies
Sub−tropical high
Tropical zone
(Trade winds)
Doldrums low
Tropical zone
(Trade winds)
Sub−tropical high
Westerlies
Sub−polar low
Polar high
Fig. 1.19 The zonal wind directions on the earth
4. Coromuel: It is warm and strong wind which blows from 10. Harmattan: It is a dry and dusty trade wind, which
afternoon to early morning in Mexico and California blows in West Africa. This northeasterly wind blows
(USA). from the great Sahara desert into Gulf of Guinea through
5. Hawk: It is strong cool breeze blowing westward through Nigeria.
Chicago from Lake Michigan (USA).
6. Loo: It is hot summer wind that blows across northern
India. 1.12.4 Measuring the Winds
7. Sirocco: It is the wind that reaches to hurricane speeds
while crossing Mediterranean sea and enters the Southern Wind is often measured in terms of wind shear, which is a dif-
Europe. They carry enormous amount of dust and sand ference in wind speed (v) and the direction over a set altitude
throughout the Northern Africa and contributes to wet (H) in the atmosphere. It is measured both in horizontal as well
weather once they reach to Europe. as in vertical directions. The wind shear (WS ) is defined as
8. Khamsin: It is a hot, dry, and dusty wind which blows
from south or southeast direction in Northern Africa WS = vH (1.45)
(Egypt). In fact it is a counterpart of Sirocco.
9. Haboob: The strong sandy dust storm that occurs along where v is in ms−1 and H is in km. Under normal conditions,
the southern edges of the great Sahara desert in Sudan. the wind blows faster and higher in the atmosphere, creating
high wind shear at high altitudes. Further, the amount of force interested in studying the heat transfer characteristics of the
that the wind generates is measured according to Beaufort vehicle flying at hypersonic speeds.
scale. This scale has 17 levels of wind force. Some of the The aircraft aerodynamics studies all those vehicles which
important levels are described as flies in the sensible atmosphere, whereas the spacecraft aero-
dynamics is dedicated towards the vehicles flying outside the
• “0” describes a calm condition in which smoke rises ver- sensible atmosphere. The sensible atmosphere can be viewed
tically. as the envelope of gases surrounding the Earth or any other
• “12” describes a hurricane. planet. Technically, it can be defined as the thin boundary
• “13” to “17” are reserved for tropical typhoons. layer wrapping the earth’s surface.
The Anemometer is a device used for measuring the wind The atmosphere is very important to us as it protects life on
speed. The strength of tornadoes is measured according to the earth. The International Standard Atmosphere as adopted
Fujita scale, which has 28 categories establishes by United by FAI is broadly classified into two layers: primary and sec-
States in 2007. The hurricane’s strength is measured using ondary. Based on the temperature distribution, the primary
Saffir–Simpson scale. It has five categories, with most layer is divided into multiple layers namely, troposphere,
powerful wind speed of 252 kmh−1 under category “5”. In stratosphere, mesosphere, thermosphere, and exosphere. The
our solar system, the Jupiter’s famous Great Red Spot is outermost primary layer shields life by absorbing harmful
actually a centuries-old hurricanes like storm swirling at ultraviolet solar radiations, warming the surface by reten-
about 644 kmh−1 . Also, the strongest wind belongs to the tion of heat (popularly known as green house effect). The
outermost planet Neptune, where the wind swirls at about dry air is the main constituent of the atmosphere which is
2100 kmh−1 . composed of nitrogen (78.09% by volume), oxygen (20.5%
by volume), argon (0.93% by volume), and small traces of
other gases. Notwithstanding the aforesaid primary layers of
the atmosphere, which are principally reliant upon the tem-
Summary
perature, several secondary layers can likewise be recognized
We know that the matter exists in three different states namely, in the light of different properties. They are ozone layer, iono-
solid, liquid, and gas. The state of a substance depends on sphere, homogeneous and heterogeneous layer, and atmo-
the intermolecular spacing among its molecules and at the spheric boundary layer.
way in which these elementary particles move. A solid is the In literal sense, the altitude is termed as the height above
substance that has a fixed volume and has a fixed shape. In the ground. However, in aerospace sciences, a total of six
solids, the molecules are arranged in a regular pattern and due altitudes namely absolute, geometric, geopotential, pressure,
to strong intermolecular cohesive forces they remain fixed at temperature, and density are used. It should be noted here
their location, but vibrate. Since, they cannot move from their that the atmospheric data are tabulated with respect to sea
position and thus, solids are rigid. A liquid has fixed volume level conditions as the reference state. The actual height of
but does not have fixed shape. The liquids conform the shape the flying object above the sea level is termed as geomet-
of the container which occupies them. In contrast, a gas has ric altitude. However, when the height is measured from the
no fixed volume and shape. They conform the shape as well center of Earth by adding radius of Earth to the geometric alti-
as the volume of the container occupying them. Both liquids tude, it is called absolute altitude. The relation between the
and gases are collectively termed as fluids. two is
Aerodynamics is that branch of science which finds several ha = hG + R
real-time applications in engineering. Using its theoretical as
well as experimental results, one can predict the aerodynamic where R is the radius of earth.
forces and moments generated on the object moving through An altitude where acceleration due to gravity is constant
the air. For example, we may be interested in calculating equal to its value at sea level is known as geopotential altitude.
forces and moments either over whole aircraft or any part of The relation between geometric (hG ) and geopotential (h)
it. The branch of science that studies the flow field over an air- altitudes is given by
craft is known as aircraft aerodynamics. We may be interested
in studying the complete building aerodynamics or calculat- R
h= hG
ing forces on its various structures. We are also concerned in R + hG
knowing the aerodynamic drag on racing cars and bikes and
the effect of inverted wings in racing cars. These investiga- The stability of atmosphere in hydrostatic equilibrium with
tions come under the purview of industrial aerodynamics. The respect to vertical displacements is known as static stability.
aerodynamics finds important applications in spacecraft too; It measures the gravitational resistance on the air mass when
apart from studying the aerodynamics aspects, we are also displaced in vertical direction. The concept of static stability
Summary 23
in the atmosphere can be understood with the help of a lump It has been observed that under specific conditions, the
of fluid better known as the air parcel. Let us assume that an atmospheric disturbances can originate near the equator and
initial perturbation is given to an air parcel, thus following develop within a few days into vortices of 500–1000 km in
possibilities may be considered. diameter. These tropical storms can generate very high wind
velocities near the center of rotation. If the tangential veloc-
• If the displaced air parcel returns to its undisturbed loca- ity exceeds 120 kmh−1 , the storm is called hurricanes. They
tion, the original equilibrium atmosphere is referred to as are also called typhoons in western Pacific and cyclones in
statically stable. Australia and Indian Ocean. Some local names are Cardon-
• If the displaced air parcel continues to move upward fur- azo in the west coast of Mexico, Baguio in Philippines, and
ther, the atmosphere is referred to as statically unstable. Willy-Willy in Australia.
• If the displaced air parcel takes up the new position At steady state and in the absence of friction, the Cori-
and remains there, the atmosphere is termed as neutrally olis force is balanced by the pressure gradient force, leav-
stable. ing a resultant motion approximately along the lines of equal
pressures, referred to as isobars. The wind in the direction
Wind is the movement of free air at a large scale caused by the of isobars is called geostrophic wind. When the force due to
thermal currents in the first 15 km above the earth’s surface. pressure gradient normal to flow, alone balances the Coriolis
The large-scale movement of air takes place because of dif-
force, then the geostrophic wind speed vg is given by
ferential heating between polar and tropical regions and also
due to rotation of Earth.
and −→ g dz
If −
→
ω is the earth’s angular velocity
v is the parti- vg = −
2 sin dn
cle velocity, the term 2 −ω ×−
→ →
v 1 is called the Coriolis force
which moves the air particle to the right of its direction of
where = angular velocity of the earth, and = earth’s
motion in northern hemisphere and to the left in southern
latitude.
hemisphere. If θ is the latitude then in the scalar form the
The Swedish scientist V.N. Ekman (1874–1954) was the
Coriolis force will be given by
first who put forward the theory of Ekman Boundary layer in
→ − 1902. It is a consequence of the Coriolis effect, which sub-
ω →
Fc = − v sin θ jects moving objects to a force to the right of the direction
of motion in the northern hemisphere and to its left in south-
If a particle moves from the equator to the north, it crosses ern hemisphere. When the surface water molecules move by
latitude with decreasing circular radius. In order to preserve the force of the wind, they, in turn, drag deeper layers of
its angular momentum, the particle must rotate faster than water molecules below them. Each layer of water molecules
the Earth and thus will be deflected to its right. A particle is moved by the friction from the shallower layer and each
which travels from the north to the equator will be retarded deeper layer moves more slowly than the layer above it, until
relative to the Earth and, hence, it will also be deflected to the the movement ceases at a depth of about 100 m. Consequently,
right. However, in southern hemisphere, the deviation will be this relative velocity between the layers creates a spiral effect,
towards left. From these deviations of the particle, one may i.e., a twist around the flow and opposite in direction to the
conclude that a force exists, a Coriolis force! The effect of surface current, which is known as Ekman spiral.
the earth’s rotation becomes apparent in the midlatitudes. By We know that during the daytime, land heats up faster than
virtue of earth’s rotation each air particle in the atmosphere has the water. The warmer air over land rises, while the cooler air
an angular momentum directed from west to east. When the over the ocean rushes in to take its place. This wind is called
air particles move towards poles maintaining the same height a sea breeze. However, after sunset the land cools down faster
above the Earth’s surface, its radius of gyration decreases. than the water. The warmer air over the ocean rises while
Hence, to conserve the angular momentum, the velocity (in cooler air over land rushes in to take its place. This wind is
the west to east direction) also increases. This effect is small called a land breeze.
at equator but large in temperate zones. Wind is often measured in terms of wind shear, which is a
A fluid that would flow radially into a sink without the difference in wind speed and the direction over a set altitude
influence of the earth’s rotation takes on a counterclockwise in the atmosphere. It is measured both in horizontal as well
rotation in the northern hemisphere due to the effect of Cori- as in vertical directions.
olis force. This rotary fluid flow is called cyclone. On the The anemometer is a device used for measuring the wind
other hand, a source flow deflected in clockwise direction speed. The strength of tornadoes is measured according to
in the northern hemisphere is termed as anticyclone. Low- Fujita scale, which has 28 categories establishes by United
pressure regions are cyclonic and high pressure regions are States in 2007. The hurricane’s strength is measured using
anticyclonic. Saffir–Simpson scale. It has five categories, with most pow-
erful wind speed of 252 kmh−1 under category ‘5’. In our (a) 9.75 ◦ C per kilometer
solar system, the Jupiter’s famous Great Red Spot is actu- (b) 6.5 ◦ C per kilometer
ally a centuries-old hurricanes like storm swirling at about (c) 11.83 ◦ C per kilometer
644 kmh−1 . Also, the strongest wind belongs to the outermost (d) 0.65 ◦ C per kilometer
planet Neptune, where the wind swirls at about 2100 kmh−1 .
2. For the neutrally stable state of the atmosphere, the adi-
abatic lapse rate is
Exercises
(a) 9.75 ◦ C per kilometer
Descriptive Type Questions (b) 6.5 ◦ C per kilometer
(c) 0.975 ◦ C per kilometer
1. Determine the flight velocity of an aircraft flying at con- (d) 6.05 ◦ C per kilometer
stant speed under standard sea level conditions, if the
3. At an altitude of about 100 km above the sea level, the
pitot pressure reading is 1.5 × 103 Nm−2 .
atmosphere becomes too thin to generate the lift. This
2. An aircraft flying at Mach 1.5 at an altitude of 19 km.
level of the atmosphere is commonly referred to as
Calculate the freestream Reynolds Number, if the char-
acteristic length of the flight is 11.5 m.
(a) Chandrasekhar Limit
3. Assuming standard sea level conditions, calculate the
(b) Armstrong Limit
weight of the air contained within a glass box of dimen-
(c) Karman Line
sion 5 m × 4 m × 6 m.
(d) Prandtl Line
4. The temperature and pressure at a point on the fuselage
is 283 K and 9.31 × 104 Nm−2 , respectively. Determine 4. The total mean mass of the atmosphere which varies
the density at this point. with the changes in water vapor loading is roughly cal-
5. Predict the density and specific volume at the nose of culated as
an aircraft, if the air pressure and temperature are 1.22
atm and 302 K, respectively. (a) 7.35 × 1012 kg
6. Estimate the Mach number of the missile moving with (b) 5.148× 1012 kg
the velocity 350 ms−1 at an altitude of 6096 m. (c) 7.35 × 1027 kg
7. The 500 kg of air is fed into a vacuum tank of volume (d) 5.148× 1018 kg
300 m3 . Calculate the pressure in the tank if the temper-
ature of the air is maintained constant at 350 K. 5. In earth’s atmosphere, the estimated mean mass of the
8. In standard atmosphere, suppose the temperature, pres- water vapor is approximately equal to
sure, and density at an altitude of 9 km are 229.74 K,
3.08 × 104 Nm−2 and 4.67 × 10−1 kgm−3 , respectively. (a) 1.27 × 1016 kg
Calculate the temperature, pressure, and density at an (b) 5.148× 1018 kg
altitude of 15 km. (c) 1.27 × 1012 kg
9. Assume that by covering a distance 115 km along the (d) 5.148× 1012 kg
sea surface, the height of surface is increased by 1.4 m
at a latitude of 35 ◦ N. At the sea surface, calculate the 6. The wind that would flow radially into a sink without
geostrophic wind speed. the influence of the earth’s rotation takes on a counter-
10. Consider the wind direction from east to west. Deter- clockwise rotation in the northern hemisphere due to
mine (a) In northern hemisphere, what would be the the effect of Coriolis force. This rotational motion of
compass direction of the averaged Ekman transport? (b) the wind is called
What is the compass direction of the current at the sur-
(a) cyclone
face of the Ekman layer (z = 0)? (c) Whether the wind
(b) anticyclone
is easterly or westerly? (d) What is the typical depth of
(c) geostrophic wind
the Ekman layer?
(d) ageostrophic wind
Objective Type Questions 7. The wind that would flow radially outward from a source
without the influence of the earth’s rotation takes on a
1. The atmospheric layer adjacent to the Earth surface is clockwise rotation in the northern hemisphere due to
called Troposphere. The environmental lapse rate in this the effect of Coriolis force. This rotational motion of
layer is the wind is termed as
Exercises 25
(a) cyclone (I) Hurricanes do not occur in a belt of 5◦ –8◦ on either

(b) anticyclone side of the equator because the Coriolis force vanishes
(c) geostrophic wind at the equator.
(d) ageostrophic wind (II) On Beaufort scale, the level “16” refers to a hurri-
cane.
8. Under the influence of Coriolis force, the boundary layer (III) The tangential velocity lesser than 120 kmh−1 is
profile is twisted. If the axis of rotation is perpendicular called hurricane.
to wall, the boundary layer will be twisted to its right in Which of the above statements are TRUE?
the northern hemisphere. This twisted boundary layer is
known as (a) I only
(b) I and II
(a) Karman Layer (c) II and III
(b) Stewartson Layer (d) All
(c) Ekman Layer
(d) Prandtl Layer Keys
9. In the above question, if the axis of rotation is parallel

1. (b)
to the wall the twisting of the boundary layer profile is
2. (a)
known as 3. (c)
(a) Karman Layer 4. (d)
(b) Stewartson Layer 5. (a)
(c) Ekman Layer 6. (a)
(d) Prandtl Layer 7. (b)
8. (c)
10. Consider the following statements with respect to hur- 9. (b)
ricanes. 10. (a)
Basic Concepts
2
In the study of aerodynamics, one needs to know various fun- build lighter-than-air1 vehicles that lead to the development of
damental definitions, point-by-point scientific classification present day aerostats2 that includes both unpowered balloons
and thorough understanding of underlying physical princi- and powered airships.3 Indeed, the first man-made aerostat
ples, related mechanics, and suitable mathematical concepts. was the hot gas balloon, which was flown successfully by the
Certainly, these conceptions are common to other fields as French Montgolfier brothers in 1783. Thus, it is regarded as
well, and it is the aim of this chapter to identify and discuss the first manned flight off the ground! Subsequently, other
those concepts that are appropriate to aerodynamics and that forms of aerostats such as blimps, zeppelins, hybrid airships,
are to be utilized in the subsequent chapters. etc., were developed and in fact, some of them are still flying
in many countries. However, none of them could become a
reliable model for human transportation due to technical and
2.1 Aerodynamics: An Overview safety issues. Perhaps, because of this reason, original search
for the vehicles heavier-than-air received special attention and
Aerodynamics is the branch of science devoted to study the the first major breakthrough in this direction was achieved by
motion of air flow past the solid objects. The field that studies Sir George Cayley (1773–1857). He is regarded as the father
the motion of air, particularly its interactions with the air- of aeronautical science, who proposed to utilize an engine to
craft, is known as aircraft aerodynamics. However, the scope make forward movement and anticipated that such movement
of aerodynamics is not just limited to the aircraft, rather it would produce lift through the wings. For more details on the
comprises all those phenomena where the air flow past a struc- history of aviation and aerodynamics in particular, one should
ture whether stationary or in motion, in the earth’s sensible refer to Anderson [1997].
atmosphere. The word aerodynamics is made of two Greek In the present text, we will restrict ourselves in studying
words: aer (aero) means air + dynamikós (dynamics) refers to the aerodynamic forces and moments acting on an aircraft
power. Essentially, the study of aerodynamics revolves around only. Before beginning the actual discussion on the funda-
estimating the forces and moments acting on an airborne vehi- mental concepts of aerodynamics, let us acquaint ourselves
cle. From the very beginning, we were interested to emulate with different parts of an aircraft and their roles in the flight.
a bird and while thinking of human flight centered on the
imitation of birds, several attempts have been made to bring
such a device into reality. But even today it seems unachiev- 2.2 The Aircraft
able with existing technologies. In a bird’s flight, the flapping
wings perform a dual role; they serve the purpose of both In this section, we will discuss various parts of an aircraft and
propulsive and aerodynamic devices. Any man-made device their functions. The aircraft is a flying object that operates in
which imitates the flight of a bird is known as ornithopter. In the sensible atmosphere. Depending upon the mission require-
recent historical research, it has been revealed that Leonardo ment, aircraft varies in shapes and sizes. A typical schematic
da Vinci (1452–1519), an Italian polymath, has sketched sev- diagram of a turbojet aircraft is shown in Fig. 2.1. The whole
eral ornithopter’s designs in his paintings. However, it is not
clear whether any of these designs were actually built and 1 The term lighter-than-air refers that the density of buoyant gas is lesser
tested by da Vinci. Clearly, it all started with the quest for the than the air.
typical shape that will make anything airborne in a sustained 2 An aerostat is a vehicle that gains its lift through the use of a buoyant
manner. In the quest for human flight, the initial focus was to gas.
3 An airship, also called the dirigible balloon, is a type of aerostat or
lighter-than-air aircraft that can navigate through the air under its own
© Springer Nature Singapore Pte Ltd. 2019 power. 27
28 2 Basic Concepts
aircraft has a median plane of symmetry, dividing the machine aerodynamic force on the control surface and thus stabilize
into two halves. Basically, these parts are mirror image of each the aircraft in the axis of rotation associated with the primary
other, if this plane of symmetry is considered as a mirror. control.
The aircraft requires lift to fly in the air, and thus a pair
of wings are attached to the long cylindrical section, known
as fuselage (or the body of the aircraft that holds all the parts 2.2.1 Aerodynamic Forces
together). Although each component of an aircraft contributes
in generating the lift, the major portion comes from wings According to the Archimedes principle, an object will be
only. That is, the wings are main lifting surfaces which con- buoyed up by a force equal in magnitude to the weight of
tribute maximum to the lift. There are two wings attached fluid displaced by the object. Similarly, an aircraft flying in
to each side of the fuselage, as shown in Fig. 2.1. The wing the air experiences a buoyant force equal in magnitude to the
fixed at the right side of fuselage is termed as starboard wing, weight of displaced air. The forces acting on the flying air-
−
→
and the wing located at the left is called the port-wing. To craft are known as aerodynamic forces. Let T P be the tractive
produce lift, the aircraft must be pushed through the air. That force produced by the propulsive system which enables the
is, it requires thrust to propel the aircraft in forward direc- aircraft to fly at a constant speed −
→v , as shown in Fig. 2.2. If
tion. Also, when the aircraft flies, its motion is retarded by −
→
the weight of the aircraft is W, acting in downward direc-
an opposing force called drag. In turbine-powered aircraft, to −
→
tion and F B is the forcedue to buoyancy acting vertically
generate thrust and overcoming the drag, engines are mounted −
→ − →
upward, then W − F B will be the resultant force acting
beneath the wings. Furthermore, to control and maneuver the
on the aircraft. Also, this net force will always be acting on
aircraft, wings of smaller sizes are attached at the rear end,
the aircraft irrespective of whether the aircraft is at rest or
commonly referred to as tail section. A pair of small-size fixed
in motion. According to Newton’s first law of motion, if the
wings mounted horizontally are called horizontal stabilizer,
aircraft is in an unaccelerated flight, the resultant force acting
and a vertically placed fixed part is known as vertical stabilizer
on it will be zero. Thus,
(also called the vertical tail). As the name suggests, both stabi-
lizers provide stability to the aircraft. In case of a sudden gust, −
→ −
→ −→ −→
nose of the aircraft use to swing sideways from the original TP+ W − FB + FA=0 (2.1)
direction of motion, called yaw. Likewise, the up-and-down
motion of the nose is known as pitch. The horizontal stabilizer −
→
In Eq. (2.1), the additional term F A is known as aerodynamic
prevents the pitching motion, whereas the vertical stabilizer force. It will be same whether the aircraft is flying in stagnant
is responsible for preventing the yawing motion of the nose. air, or it is at rest and air flows past it. That is, the aerodynamic
In addition to fixed parts, both wing and tail sections have forces and moments are generated solely due to the relative
small moving parts attached to the fixed sections by hinges. motion between air and the aircraft. Indeed, it is the concept
The hinged part of the vertical tail is known as rudder, whose of relative velocity that enables the possibility of using exper-
primary purpose is to control yaw. The hinged portion of imental data obtained in the wind tunnel (where the air flows
the horizontal stabilizer is called the elevator, used to con- past a stationary model), applicable to actual flight conditions.
trol pitch. The hinged part near the wing tip4 is called aileron, The relative motion of air flow past the aircraft is being
which controls the roll motion of the aircraft. Note that move- observed in the earth-fixed Cartesian coordinate space, xyz.
ment of ailerons is complementary to each other. That is, if The aerodynamic force is resolved into horizontal and vertical
one aileron deflects downward then the other one will deflect components. The horizontal component acting in thedirec-
upward in the same proportion. Consequently, one wing will −
→
tion of freestream (along x-axis) is termed drag D , and
be lifted but other will sink, causing the roll motion. The wings
the vertical component, − perpendicular to the freestream direc-
have additional hinges near to root section that are called flaps. →
They are deployed downward during landing and take-off to tion, is known as lift L (along z-axis). If α is the gliding
−
→
increase the amount of force generated by the wings. We can angle, then the magnitude of aerodynamic force F A can be
see an additional plate-like structure mounted on the top sur- resolved into its components as
face of the wings, known as spoilers. They are used to disrupt
− →
the flow over the wing and thus decreases the lift when spoil- → −
L = L = F A cos α (2.2)
ers are deployed. The secondary flight control surfaces that − →
→ −
are attached to the rear end of the primary control surfaces are D = D = F A sin α (2.3)
called trim tabs. They are deployed to “trim” or nullify the
4 The wing tips consist of those points of the wings which are at the
greatest distance from the plane of symmetry. Thus, the tip can be a
point, a line, or an area according to the design of the airfoil.
2.2 The Aircraft 29
x
O
Starboard wing
(Right wing)
z Vertical stabilizer
Spoiler
Right aileron
Fuselage Landing flap

Rudder
Trim tab
Elevator
Landing flap
Trim tab Tail plane

Engine
(Horizontal stabilizer)
Spoiler Left aileron
(Left wing)
Port wing
Fig. 2.1 Schematic diagram of an aircraft
where the scalar quantities L and D represent the magnitudes 2.2.2 Aerodynamic Moments
−
→ −
→
of the vectors L and D , respectively. For simplicity, to repre-
sent the lift and the drag, we will use only the scalar variables For a general motion of an aircraft, in addition to lift and
L and D, respectively, throughout the book. Note that even drag, another component of aerodynamic force (side force)
though the drag force is undesirable, it cannot be completely is experienced which acts along the y-axis with respect to a
eliminated, however, can be minimized by keeping the angle earth-fixed reference frame (Fig. 2.3). Clearly, the side force
of attack (α) as low as possible. In other words, a small gliding is perpendicular to both lift and drag. In Sect. 2.4.3, we will
angle will make the aircraft more streamlined. see that the pressure and shear stress distributions over a rigid
Furthermore, it is not necessary that the aircraft is always in body in motion generate aerodynamic forces and moments.
the steady and level flight; instead, it may be in a vertical loop In Fig. 2.3, the component of moment experienced by a rigid
where the lift may be directed vertically downward or even aircraft in xz plane is known as pitching moment. The two
act horizontally. Therefore, it is essential to provide a stan- other components of moment acting in xy and yz planes are
dard definition of aerodynamic forces which is valid for entire referred to as the yawing moment and rolling moment, respec-
flight path of an aircraft including maneuvers. Therefore, by tively.
definition, the component of aerodynamic force which acts
in the direction of freestream and opposite to the motion of
aircraft is termed as the drag. Similarly, the component which 2.3 Parametric Studies in Aerodynamics
acts perpendicular to the freestream and helps flying the air-
craft is known as the lift. In the study of aerodynamics, we get acquainted with more
and more newer variables throughout our life span. If one
30 2 Basic Concepts
L
FA
θ
v
TP
W−F
B
Fig. 2.2 Aerodynamic forces acting on an aircraft in level and steady flight
Lift
Yawing moment
Side force
Rolling moment
O
Drag
Pitching moment
Fig. 2.3 Aerodynamic moments acting on an aircraft in steady and level flight
starts preparing a list of these variables, then by no means with the wall of the container is perfectly elastic. When a
one can claim the list to be exhaustive. However, there are particle collides with the wall, it exerts a force on the wall.
some variables that are of fundamental in nature and thus Since the linear momentum of these particles, in the direc-
essential to take a note of them. In this section, we will tion perpendicular to the wall, is assumed to be conserved
review five most important variables in aerodynamics: pres- even after their interactions with the wall. Consequently, a
sure, density, temperature, flow velocity, and coefficient of normal force is exerted on a surface due to the time rate of
viscosity. change of linear momentum of the gas molecules impacting
that surface. By definition, the force per unit area is termed as
pressure. At this stage, readers must acknowledge the fact that
2.3.1 Pressure even though the pressure is defined as “force per unit area”
one need not have the surface area exactly equal to 1 m2 or
From kinetic theory of gases, we know that the gases are com- 1 ft 2 to define the pressure. Let us consider an elemental area
posed of a large number of particles that behave like hard, dA enclosing a point P at which pressure is required. If an
spherical objects, and in a state of constant random motion. elemental force dF acts on dA, then the pressure p is defined
Also, the collision of these particles among themselves or as
2.3 Parametric Studies in Aerodynamics 31
dF substance, and for a fluid in motion, one part of the fluid may
p = lim (2.4)
dA→0 dA be traveling at different velocities from another part. Further,
fluid velocity at a point is quite different and independent of
In Eq. (2.4), on the macroscopic basis, dA, however, is close
the instantaneous velocity of the molecule near to that point.
to zero but should be sufficiently larger than the average spac-
Rather, we consider the motion of the center of gravity of
ing between molecules from the microscopic point of view.
the volume ∂∀ , instantaneously surrounding that point (P),
Thus, pressure is the limiting form of the force per unit area
and define the fluid velocity at the point P as the instanta-
where the elemental area has shrunk to zero. Therefore, pres-
neous velocity of this center of gravity. Therefore, the fluid
sure is a point property and its value changes from point to
velocity at a point is the instantaneous velocity of the fluid
point.
particle which at that moment is passing through that point.
One must remember that here “fluid particle” means a small
mass of fluid of fixed identity and size of comparable ∂∀ .
2.3.2 Density
Unlike scalar quantities such as the pressure, temperature,
and density, the flow velocity is a vector quantity as it has
Like pressure, density is also a point property defined as the
both magnitude and direction.
mass per unit volume. From similar arguments made on the
pressure, we can say that the fluid volume exactly equal to
1 m3 or 1 ft3 is not required to define the density. Let us con-
2.3.5 Stress
sider an elemental volume d∀, enclosing a point P, has a small
mass dm. The density at P will be
Consider a circular area δA, in its plane surrounding a point
dm P. One side of surface exerts force on the other side and by
ρ = lim (2.5) Newton’s third law the latter side exerts force on the first side
dV→0 d∀
an equal but an opposite force. These forces are called surface
From Eq. (2.5), it is evident that the density is a limiting form forces.
of the mass per unit volume where d∀ will be close to zero Further, the surface forces acting on δA may be resolved
around P but should be sufficiently larger for the reasons as into a component normal to δA and along the plane of δA
discussed for pressure. (Fig. 2.4). The normal force per unit area, in limiting condi-
tion, is termed as pressure. That is,
2.3.3 Temperature dFnormal

p= lim
(2.7)
δA→δA dA
From kinetic theory of gases, the temperature T is directly pro-
portional to the average kinetic energy of the fluid molecules. In differential form,
If KE denotes mean kinetic energy of fluid molecules, then
dF
temperature is given by p= (2.8)
dA
3 Similarly, the shear stress τ is defined in terms of the limit of
KE = κT (2.6)
2 the tangential component of force per unit area acting on δA.
where κ = 1.38 × 10−23 m2 kgs−2 K is the Boltzmann con- dFtangential
stant. From the above equation, it is evident that at high tem- τ= lim
(2.9)
δA→δA dA
perature the random motion of the molecules will be high and
vice versa. In differential form,
dF
τ= (2.10)
2.3.4 Flow Velocity dA
In fluid dynamics, our prime focus is to know about the fluid 2.3.5.1 State of Stress in Inviscid Fluid (The
flow velocity. However, the concept of velocity in fluid motion Hydrostatic Pressure)
is slightly more subtle than that of a moving solid body. That It can be observed that the normal stress or pressure is
is, if a solid body is in translational motion at a particular independent of the orientation of the plane δA. Let us first
velocity, then all parts of the body are simultaneously trans- consider an inviscous fluid in which no viscous stresses (tan-
lating at that velocity. On the other hand, a fluid is a “flabby” gential or shear stresses) exist, even though there is relative
32 2 Basic Concepts
δ Fnormal
δA P
Free molecular Region of

region continuum
δ Fnormal
δA
δA
δ A’
Fig. 2.4 Definition of stress
motion within the fluid. Consider a two-dimensional acceler- py = pα (2.12)

ating fluid element as shown in Fig. 2.5.
Let px and py , respectively, are the components of pressure Thus, we finally have
force in x- and y-directions, and gx and gy are the components
of body forces due to gravity. From Newton’s second law of
motion in x-direction, we have px = py = pα (2.13)

Fx = max Since α-direction is arbitrary and can represent any direction.
Thus, we can say that in an inviscid fluid whether in motion or

where Fx is the resultant force, m is the mass of the body, not, the fluid pressure at a point is the same in all directions.
and ax is the acceleration in x-direction. Thus, This is called hydrostatic state of stress.
2.3.5.2 State of Stress in Viscous Fluid

dxdy dxdy
px dy − pα (dx)2 + (dy)2 cos α + ρgx = ρax Unlike inviscid fluid flow, in case of viscous fluid in motion,
2 2
all the stresses will be inserted. That is, px , py , and pα will
all be different and they can be expressed, same as solids,
But, from the geometry graphically
= by Mohr’s circle of stress. In this case, the stress
matrix σ will consist of six components, given in Eq. (2.14).
(dx)2 + (dy)2 cos α = dy
⎡ ⎤
σx x τx y τx z
Now since dxdy is negligible as dx and dy are very small. =
σ =⎣ τx y σ yy τ yz ⎦ (2.14)
Hence, τx z τ yz σzz
px = pα (2.11) where σxx , σyy , and σzz are the normal stresses, and τxy ,
τyz , and τxz denote shear stresses. At high Reynolds number,
Similarly, applying Newton’s second law in y-direction, we the shear stresses are found to be very small as compared
get to normal stresses. Consequently, the variation of normal
2.3 Parametric Studies in Aerodynamics 33
2 2
(dx) + (dy)
dy p
α
α α
px
dx
py
x
Fig. 2.5 Schematic representation of the hydrostatic pressure on a fluid element
stresses with their orientation is correspondingly small in fluid where the coefficients a, b and c are arbitrary constants; T0 is
flows. the reference temperature; and μ0 is the viscosity of liquid at
the temperature T0 .
In contrast, the viscosity of a gas increases with tempera-
2.3.6 Coefficient of Viscosity ture as shown by the following relations:
It is the property of the fluid due to which fluid layer has shear- ⎧ n
ing stresses between them. For fluid flow in two dimensions, ⎪
⎨ TT Power law
μ 0
the shear stress (τ ) is found to be proportional to rate of strain = 3 (2.18)
μ0 ⎪
⎩ T 2 T0 +S Sutherland law
known as Newton’s law of viscosity, given by Eq. (2.15). T0 T+S

∂u ∂v where μ0 is the known viscosity at a known absolute temper-
τ =μ + (2.15)
∂y ∂x ature T0 ; and S is called the Sutherland constant. For air, n =
0.7 and S = 110 K; a more useful form of Sutherland formula
where μ is known as dynamic viscosity, generally, expressed is given below:
in Nsm−2 or in m−1 kgs−1 . Sometimes, viscosity is also writ-
ten as 3

T2
μ = 1.46 × 10−6 ;
μ T + 111
ν= (2.16)
ρ
for 0.01 atm < pstatic < 100 atm; 0 K < T < 3000 K
where ν is kinematic viscosity. The SI unit of kinematic vis- (2.19)
cosity is m2 s−1 . In C.G.S. system, kinematic viscosity is also
given in terms of Stokes, named after George Gabriel Stokes. From Eqs. (2.17) and (2.19), it is evident that by increasing
Temperature has a profound effect on the viscosity, temperature the viscosity of a liquid decreases, whereas the
whereas the pressure has very less effect. The viscosity viscosity of a gas increases. Why? Let us try to understand
of gases and most liquids increases slowly with pressure. this phenomenon as follows. In the motion of both liquids
Nevertheless, the change in viscosity due to pressure is and gases, molecules are free to move from one layer to
negligibly small, and hence, the effects of pressure are another, traveling at different velocities. A molecule which
neglected throughout our discussion. moves from a layer of lower velocity to a layer at higher veloc-
Unlike gases, the temperature has opposite effect on liq- ity has to be accelerated. Similarly, a molecule which moves
uids. With increase of temperature, the viscosity of a liquid from a layer traveling at higher velocity to a layer at lower
decreases. Empirically, viscosity of a liquid is expressed in velocity must be slowed down. Therefore, the molecules dif-
the following form: fusing across the fluid layers transport net momentum, eventu-
ally leading to shear stresses between the layers. These shear
2
μ T0 T0 stresses will be zero if the layers are traveling at same veloc-
ln =a+b +c (2.17)
μ0 T T ity or if the fluid is at rest. Further, it is known that a bind-
34 2 Basic Concepts
ing cohesive force acts among the fluid molecules and for various locations on the wing usually depend on its distance
relative motion to occur these cohesive forces must be over- from the plane of symmetry.
come through shear stresses. Thus, the overall shear stress in
a fluid is the summation of the above two factors: shear stress
due to molecular diffusion, and shear stress due to overcom- 2.4.1 Nomenclature of an Airfoil and the Wing
ing the cohesive forces among the molecules. In liquids, the
shear stresses are due more to overcoming the cohesive forces The purpose of this section is to describe the parameters used
than due to diffusion because of lower molecular velocity. to define an airfoil. Consider the following variables with
On the other hand, the shear stresses in gases are generated respect to an airfoil, as shown schematically in Fig. 2.6.
mainly due to transport of momentum across the layers as
the velocity of gas molecules is much higher but the cohe- Suction Surface and Pressure Surface
sive forces are relatively weak. Therefore, whenever temper- The upper surface of an airfoil is also known as the suction
ature is increased intermolecular spacing increases, but the surface because it is the high velocity or the low static pressure
strength of cohesive force reduces. Hence, viscosity of liquid region. The bottom surface of an airfoil also called the pres-
decreases with increase of temperature. In the case of a gas, sure surface has comparatively higher static pressure. This
the viscosity is more due to momentum transfer. As the tem- difference of pressure between the upper and lower surfaces
perature increases, the additional kinetic energy is imparted causes the pressure gradient which eventually generates lift
to gas molecules and thus, the net exchange of momentum on the airfoil.
between the layers increases. Consequently, viscosity of the
gas increases. Leading Edge and Trailing Edge
The Sutherland’s relation closely represents the variation The leading edge is the point located at the front of the airfoil
of μ with temperature for air over wide temperature range. which has maximum curvature (or minimum radius). Simi-
However, the success of the equation representing the varia- larly, the trailing edge is defined as the point of minimum
tion of μ with T for several gases does not establish the validity curvature at the rear end of the airfoil.
of Sutherland’s molecular model for gases. In other words, the
Chord and Chord Line
equation is not sufficient enough to represent the core of the
molecule as a sphere or take molecular attraction into account The chord is essentially a line segment which joins two points
of first order only. The increase in exponential of μ along with on a curve and the chord line refers to an imaginary straight
T, when compared to that of non-attractive rigid sphere, has line joining the leading and trailing edge of an airfoil.
to be explained partly due to the softness of the repulsive field
Chord Length
at small distance and partly to the attractive forces for having
more than first-order effect. Hence, Sutherland’s relation can It is the distance between the trailing edge and the point on
be taken only as a simple interpolation formula for restricted the leading edge, where the chord intersects the latter.
temperature. Furthermore, the viscosity of air is independent
Camber
of pressure, for the temperature below 3000 K. For collision
cross section of atom–atom and atom–molecule, the trans- It is the maximum distance between the mean camber line
port properties of high air temperature should be taken into and the chord line, measured perpendicular to the chord line.
account.
Mean Camber Line
The mean camber line or mean line is the locus of points mid-
way between the upper and lower surfaces. Its shape depends
2.4 The Airfoil on the thickness distribution along the chord.
In a three-dimensional body, if a section is cut with a plane Airfoil Thickness

(parallel to the plane of symmetry), the intersection of the The thickness of an airfoil varies along the chord. It is mea-
body surface with that plane is termed as profile. For an air- sured by following either of the two conventions: The airfoil
craft wing, this profile is better known as airfoil. When the thickness can be measured either perpendicular to the camber
flow past an airfoil, then the latter produces the aerodynamic line (described as the American convention) or it can be mea-
forces, namely, lift and drag. Alternatively, for a given area the sured perpendicular to the chord line (described as the British
profile which generates a maximum amount of lift is called an convention).
airfoil. The shape and orientation of an airfoil section sliced at
2.4 The Airfoil 35
Upper camber
Thickness Mean camber
Leading edge Trailing edge
Lower camber Chord line

Chord
Fig. 2.6 Schematic layout of a typical airfoil with its nomenclature
Pitch Angle angle is the name given to the negative Dihedral angle, that is,
when there is a downward angle from horizontal of the wing
It is the angle between the chord of an airfoil section and
or tailplane of a fixed wing aircraft.
the plane of rotation. Pitch angle is also known as the setting
angle. Mean Aerodynamic Chord of the Wing
Angle of Inclination The width of the wing is greatest where it meets the fuselage
at the wing root and progressively decreases toward the tip.
It is the angle between the relative velocity vector and the
Consequently, the chord also changes along the span of the
plane of rotation.
wing. The average length of the chord is known as the mean
Angle of Incidence aerodynamic chord (MAC).
It is the angle between the relative velocity vector and the Swept Wing
chord line of the airfoil. It is also known as the angle of attack.
A swept wing describes an aircraft that has wings inclined
Wing Area either backward or forward relative to its root. The wing sweep
It is the projected area of the planform and is bounded by the delays the formation of shock waves and thus reduces the
leading and trailing edges and the wing tips. Note that the associated wave drag, thereby improving the performance of
wing area is not the total surface area of the wing. the aircraft. Thus, the swept wings are, generally, used in
the jet aircraft flying at high subsonic or supersonic Mach
Wing Span numbers.
The wing span (or just span) of an airplane is the distance
from one wingtip to the other wingtip. 2.4.2 Pressure Distribution Around an Airfoil
Aspect Ratio
We know that when the fluid flow past an airfoil then the
The aspect ratio of a wing is defined as the ratio of its span pressure distribution around the airfoil surface is not uniform
to its mean chord. It is equal to the square of the wingspan due to the geometry of the airfoil. Figure 2.7 shows the typi-
divided by the wing area. Thus, a long and narrow wing has cal pressure distributions for a given airfoil section at varied
a higher aspect ratio than a short and wide wing. angles of attack. Since it is often convenient to deal with the
Root Chord and Tip Chord nondimensional form of the static pressure over the airfoil
p−pa
The root chord is the chord length of the airfoil near the fuse- surface, and hence, the coefficient of pressure Cp = 1
2 ρUa
2
lage, whereas the tip chord is the chord length of the airfoil has been used in the figure, where p is the pressure over the
near the wing tip. airfoil surface, pa is the freestream pressure and Ua is the
freestream velocity. Note that the length of arrows is directly
Taper Ratio
proportional to the value of Cp and the direction of arrows
It is defined as the ratio of tip to root chords. The wing taper indicates the positive or negative values of Cp .
is always represented in terms of taper ratio. At conventional angles of incidence, in comparison to
Dihedral Angle and Anhedral Angle freestream static pressure, the suction (upper) surface experi-
ences a decrease in pressure over a large portion of it, whereas
The Dihedral angle is the upward angle from horizontal of a lesser decrease in pressure is felt by the lower (pressure)
the wing or tailplane of a fixed wing aircraft. The Anhedral
36 2 Basic Concepts
Negative pressure − suction
C
p
Forward stagnation point
− ve
0 x/ c
1
Positive pressure − pushing
+ ve (b) α = 0 deg
(a) Variation of pressure coefficient over the airfoil



(d) α = 14 deg
(c) α = 8 deg
Fig. 2.7 Typical pressure distributions around an airfoil at varied angles of attack
surface. This uneven pressure acting on upper and lower sur- We will now examine the pressure distribution over an airfoil
faces leads to the nonuniform pressure distribution around the section at varying angles of attack, as illustrated in Fig. 2.7.
airfoil. The maximum pressure occurs at stagnation points
Airfoil at α = 0o
where the flow is brought to rest, at the trailing edge, and
somewhere near to leading edge depending upon the angle The pressure distribution around the airfoil section at zero
of attack. At this point, the flow bifurcates into upper and degree angle of attack is shown in Fig. 2.7b. From the figure,
lower stream that pass over the airfoil. Note that, for finite it is evident that there are small regions at the nose and tail
flow speeds downstream of the leading edge, the flow must where Cp is positive, but over a large portion of the airfoil Cp
accelerate. Consequently, the pressure must decrease initially is negative. The reduction of static pressure on the pressure
below the stagnation value which in turn accelerates the flow (or suction) surface tends to draw the section upward, while
around the nose and the leading portion of the surface. The that on the lower (or pressure) surface has opposite effect.
rate of acceleration increases with increase in the angle of Since the suction effect is larger and thus there is a resultant
attack. force upward on the section, that is, lift.
On the upper surface, the pressure reduces continuously Airfoil at α = 8o
from the stagnation value through the freestream value to
a position where a peak negative value is reached. There- When the angle of attack is increased to 8o , as shown in
after, the flow is continuously retarded, increasing the pres- Fig. 2.7c, the pressure on the suction surface is decreased
sure through the freestream value to a small positive value further while the pressure on the lower surface is increased.
toward the trailing edge. On the other hand, the flow over the Consequently, the lift on the section is increased.
pressure surface accelerates much less rapidly than the suc- Airfoil at α = 14o
tion surface, reducing the pressure much more slowly from
The pressure distribution around the airfoil at α = 14o is
the freestream value to some small negative value. Thereafter,
shown in Fig. 2.7d. Since the increase in pressure on the lower
the subsequent decleration causes increase in pressure from
surface remains almost constant, and hence, the maximum
free stream value to a positive value hear the trailing edge.
contribution to the lift comes from the increased reduction of
pressure on the upper surface.
2.4 The Airfoil 37
2.4.3 Generation of Forces and Moments dNu = pl dsl cos β − τl dsl sin β (2.24)
dAu = pl dsl sin β + τl dsl cos β (2.25)
In the earlier discussion, we have qualitatively examined the
pressure distribution around the airfoil, neglecting the effect where sl is the linear distance measured along the lower sur-
of shear stress over the airfoil surface. In reality, when an face. Thus, the total normal and axial forces per unit span will
airstream flow past a body, irrespective of how complex the be
body shape may be, the aerodynamic forces and moments

on the body are due entirely to pressure and shear stress dis- TE TE
N =− pu cos β + τu sin β dsu + pl cos β − τl sin β dsl
tribution over the body surface. Both pressure (p) and shear LE LE
stress (τ ) have dimensions of force per unit area (e.g., New- (2.26)

ton per square meter). The pressure acts normal to the sur-
TE TE
A = −pu sin β + τu cos β dsu + pl sin β + τl cos β dsl
face, whereas shear stress acts tangential to the surface. Shear LE LE
(2.27)
stress is because of the “pulling action” on the surface, which
is caused by friction between the body and the air.
Let us now evaluate the aerodynamic moment about the lead-
The net impact of pressure and shear stress integrated over
ing edge of the body. By convention, the moment is considered
the complete body surface is a resultant aerodynamic force
positive if it tends to increase α, and negative if the moment
(R) and moment (M) on the body, as shown in Fig. 2.8a. Fur-
tends to decrease α. Consider again the elemental strips at
ther, the relative velocity of an undisturbed flow, far away
upper and lower surfaces at P and Q, respectively. Thus, the
from the body, is va , better known as freestream velocity. By
moment per unit span due to p and τ on the elemental strip
definition, the components of R in the directions, perpendic-
on the upper surface will be
ular to and along the Ua , are known as lift (L) and drag (D),
respectively. The linear distance between leading edge and
dMu = pu cos β + τu sin β xdsu + −pu sin β + τu cos β ydsu
trailing edge is termed as chord (c). (2.28)
The resultant force (R) can also be resolved into the com-
ponents perpendicular to and along the chord directions. The Likewise on the bottom surface,
component of R, in the direction perpendicular to c, is known
as the normal force, N, and the component along c is termed
dMl = −pl cos β + τl sin β xdsl + pl sin β + τl cos β ydsl
as the axial force, A. The angle formed in between the chord (2.29)
direction and the freestream velocity is called the angle of
attack (α), shown in Fig. 2.8a, b. From Fig. 2.8a, following
To calculate total moment per unit span about the leading
relations can be inferred.
edge, we have
L = N cos α − A sin α (2.20) TE

=
MLE pu cos β + τu sin β x + −pu sin β + τu cos β y dsu
D = N sin α + A cos α (2.21) TE
LE

+ −pl cos β + τl sin β x + pl sin β + τl cos β y dsl (2.30)
LE
Now let us consider a two-dimensional body (Fig. 2.8b) on
which we will select an elemental strip of arc length, ds, and
of span unity. Cartesian axes (xy) system is chosen with its Pitching Moment About a General Point
origin located at the leading edge (LE); x-axis and y-axis,
Suppose the pitching moment about a general point s at some
respectively, are parallel and perpendicular to the chord.
distance from the leading edge is known, and we are interested
Furthermore, the angles measured in clockwise direction are
in determining the pitching moment at a distance b, i.e., further
considered positive. Examining both Fig. 2.8a, b, we see that
away from point s. Thus, we write
the elemental
normal and axial forces per unit span, shown
with , acting at the point P on the elemental surface over = M − L cos α + D sin α s = M − L cos α + D sin α b
MLE s s s b b b
the upper body surface are
(2.31)
dNu = −pu dsu cos β − τu dsu sin β (2.22)
If ls = lb = l and ds = db = d, then rewriting equation (2.31)
dAu = −pu dsu sin β + τu dsu cos β (2.23) as
where su is the linear distance measured along the upper sur-

Ms = Mb − Lb cos α + Db sin α (b − s) (2.32)
face. Similarly, taking elemental strip at the lower body sur-
face (forces per unit span acting at point Q), we get
Dividing the above equation by 21 ρv2 Sc, we obtain
38 2 Basic Concepts
N
L
R
M
α
Ua α
D
A
(a) Normal and axial forces acting on the body.
dx
pu β
y
ds
β
β
P tu
x
α Q
β
Ua
τl
β
p
l
(b) Pressure and shear stress distribution over the body surface.
Fig. 2.8 Schematic diagram of forces over a two-dimensional body

TE TE
b s
CMs = CMb − (CL cos α + CD sin α) − (2.33) LS = cos α − pu cos β + τu sin β dsu + pl cos β − τl sin β dsl
LE LE
c c
TE TE

− sin α −pu sin β + τu cos β dsu + pl sin β + τl cos β dsl
Now rewriting the lift and drag from Eqs. (2.20) and (2.21), LE LE
per unit span (2.36)
TE
TE
LS = N cos α − A sin α (2.34) DS = sin α − pu cos β + τu sin β dsu +
LE
pl cos β − τl sin β dsl
LE
DS =
N sin α + A cos α
(2.35)
TE
TE

+ cos α −pu sin β + τu cos β dsu + pl sin β + τl cos β dsl
LE LE
Introducing Eqs. (2.26) and (2.27), in the above equations, the (2.37)
lift and drag per unit span at the point P are obtained as
Substituting Eqs. (2.36) and (2.37) in Eq. (2.31), we have

TE
MS = MLE
+ cos2 α − pu cos β + τu sin β dsu
LE
TE

+ pl cos β − τl sin β dsl
LE
2.4 The Airfoil 39
TE
TE
+ sin2 α − pu cos β + τu sin β dsu + pl cos β − τl sin β dsl 2.4.4 Center of Pressure
LE LE
TE
TE The aerodynamic forces on an airfoil section may be repre-
+ sin α cos α −pu sin β + τu cos β dsu + pl sin β + τl cos β dsl
LE LE sented by lift, drag, and pitching moment. At each value of the
(2.38) lift coefficient, there will be a particular point about which the
pitching moment coefficient is zero. The aerodynamic effects
We can obtain the working forms of the above equations by on the airfoil section may be represented by the lift and the
substituting dx = ds cos β and dy = −ds sin β. drag alone acting at that point, termed as the center of pres-
Force and Moment Coefficients sure.
In the above section, we have seen that the normal and axial
Moreover, it is sometimes convenient to express these equa- forces per unit span on the body are caused due to pressure
tions in their nondimensional forms as and shear stress distributions over the body surface (Eq. (2.26)
and Fig. 2.8a). These loads, also, generate a moment about
• Lift coefficient (CL ): CL = L
q a Sr , the leading edge, given by Eq. (2.30). Note that the resultant
• Drag coefficient (CD ): CD = qaDSr , single force (R) (or its components N and A ) should be
• Normal force coefficient (CN ): CN = N
q a Sr ,
placed on the body such that it produces the same effect as
• Axial force coefficient (CA ): CA = qaASr , and the distributed loads.
Consider a two-dimensional airfoil, shown in Fig. 2.9; the
• Moment coefficient (CM ): CM = qaM S r lr , distributed load produces a moment about the leading edge
given by Eq. (2.30). Therefore, N and A must be placed on
where qa = 21 ρa Ua2 is the freestream dynamic pressure, ρa is the airfoil at such a location (say xCP ) to generate the same
the freestream density, Sr is the reference area, and lr is the moment about the leading edge. In terms of per unit span, we
reference length. For different shapes, the reference areas and get
reference lengths are different. For example, for an aircraft

wing, Sr will be the planform area, and lr will be the mean MLE = −xCP N (2.44)
chord length. However, for a sphere, Sr will be the cross-
sectional area and lr will be the sphere diameter. It should be or
noted that the above force and moment coefficients are valid

for three-dimensional bodies, such as aircraft wings, sphere, MLE
xCP = − (2.45)
cylinder, etc. For two-dimensional bodies, these coefficients N
are calculated per unit span.
Consider an airfoil for which the reference area (Sr ) and where MLE is the moment about leading edge force per unit
reference length (lr ) are defined as span, and N is the normal force per unit span. From Fig.
2.9,
it can be seen that the axial force per unit span A does
Sr = b × c (2.39) not contribute to the moment about the leading edge as itacts
lr = c (2.40) along the chord, whereas the positive normal component N
produces a negative pitch-down moment about the leading

The lift L , drag D , and moment coefficients (CM ) per edge as it decreases angle of attack (α). This is consistent
unit span will be expressed as with the negative sign in Eq. (2.45). In Fig. 2.9 and Eq. (2.45),
xCP is defined as the center of pressure.
L For very small angle of attack, i.e., α ≈ 0, then sin α ≈ 0
CL = (2.41)
qa c and cos α ≈ 1. Thus, we have
D
MLE
CD = (2.42) xCP ≈ − (2.46)
qa c L
M where L is the lift per unit span. Let us examine Eqs. (2.45)
CM = (2.43)
qa (c) 2 and (2.46). We can see that as N and L decrease, xCP
increases and if the forces approach zero, the xCP tends to
Note that we have used with the quantities for per unit in infinity. This is the reason, the center of pressure is not always
order to distinguish them from the overall quantities. a suitable concept to use in aerodynamics. However, it can still
40 2 Basic Concepts
M LE
A
x cp
Fig. 2.9 The center of pressure for an airfoil
be used to define the forces and moment caused by the dis- angle of attack (α = 0), as schematically shown in Fig. 2.10d.
tributed load on a body; the resultant force can be located at Note that, in Eq. (2.46), the contribution in the moment due
any point on the body, as long as the value of the moment about to drag was not considered since the line of action of drag
that point is also known. Consider the forces and moments passes through the leading edge. However, for α = 0, the
on an airfoil, as shown in Fig. 2.10. In Fig. 2.10a, forces are contribution in moment due to drag must be considered, and
placed with a finite moment at the leading edge; in Fig. 2.10b, hence, Eq. (2.47) will be modified as
the resultant force with a finite moment is placed at quarter-
chord location; and in Fig. 2.10c, the resultant with a zero MLE = MAC − L cos α − D sin α xAC = − L cos α + D sin α xCP
moment about that point is placed at the center of pressure. (2.49)
In terms of per unit span, we can write the quantitative equiv-
1
alence of these three cases as follows: Dividing above by ρV2 Sc
, we get
xAC xCP
c CMLE = CMAC − (CL cos α − CD sin α) = − (CL cos α + CD sin α)
MLE = − L + Mc/4 = −xCP L (2.47) c c
4
or
xCP xAC CMAC
= − (2.50)
2.4.5 The Aerodynamic Center c c (CL cos α + CD sin α)
By definition, the aerodynamic center is the point about which But, for small values of angle of attack (α ≈ 0), we have
the pitching moment coefficient (CM ) for an airfoil does not cos α ≈ 1 and sin α ≈ 0. Thus, Eq. (2.50) becomes
vary with the angle of attack (α). That is,
xCP xAC CMAC
= − (2.51)
dCM c c CL
=0 (2.48)
dα
Note that xCP is always behind xAC as CMAC is invariably
i.e., cm is invariant with respect to α. For an inviscid and negative in Eq. (2.51). Further, recall Eq. (2.33)
incompressible flow past the flat plate (or curved surface), the
aerodynamic center is located at the distance 4c̄ behind the b s
CMs = CMb − (CL cos α + CD sin α) − (2.52)
leading edge, where c̄ is the chord length of the plate. The c c
viscosity of the fluid moves the aerodynamic center forward
(toward leading edge), while the compressibility shifts it back- Let us consider the point b as the aerodynamic center
ward (toward trailing edge). At supersonic Mach numbers, the (b = xAC ). Thus, rewriting the above equation fetches
aerodynamic center is located exactly at 2c̄ . It should also be x s
AC
noted that the aerodynamic center is a fixed point which gen- CMs = CMAC − (CL cos α + CD sin α) − (2.53)
c c
erally lies within the profile of the airfoil, unlike the center
of pressure that may lie outside the airfoil section. Moreover, For small angles of attack, without introducing severe error,
the location of center of pressure changes with variation in it can be assumed that sin α ≈ 0 and cos α ≈ 1. Thus, we
the lift coefficient. obtain x s
AC
The location of center of pressure for an airfoil at zero CMs = CMAC − CL − (2.54)
degree incidence angle (α = 0) is given by Eq. (2.46). We c c
will now determine the center of pressure for an airfoil at an Differentiating above equation with respect to CL , we get
2.4 The Airfoil 41
L
(a)
L M LE
M AC
AC
M LE D
D
x
AC
(b) L
c
M c/4
D (d)
M LE L
c/4
L
(c)
cp
D
x
D cp
x c
cp
Fig. 2.10 Forces and moments on an airfoil at zero and finite angles of attack
dCMs dCMAC xAC s These physical quantities are obtained from the capable mean
= − − (2.55)
dCL dCL c c of dimensional analysis.
Consider a body in the freestream at a given angle of attack
dCM
But, from the definition of aerodynamic center, dCL = 0. (α), as shown in Fig. 2.8a. We can suppose that the aerody-
Thus, namic force (R) depends on the body geometry and fluid/flow
x properties, more specifically,
dCMs AC s s xac
=0− − = − (2.56)
dCL c c c c̄ • Freestream velocity (Ua ),
• Freestream density (ρa ),
xAC s dCMs
= − (2.57) • Fluid viscosity (μa ) or (νa ),
c c dCL • Length scale of the body, such as chord (c), and
• Compressibility of the medium which can be related to the
If the lift coefficient is zero (CL = 0), then Eq. (2.54) leads to
speed of sound aa .
CMs = CMAC (2.58)
In functional form, we can write the above
Furthermore, if s = 0, then CMs = CMLE , i.e., point s is
R = f (ρa , Ua , c, μa , aa ) (2.60)
exactly located at the leading edge. Then, Eq. (2.57) becomes
xAC dCMLE Equation (2.60) is a general relation and cannot be directly

=− (2.59) used for the calculation of R. In order to get functional
c dCL
relationship, we can collect data by varying each parame-
ter one at a time. However, such an exercise would prove
2.5 Dimensional Analysis to be immensely expensive and tedious. Instead, if we could
reduce the number of parameters using the laws of dimen-
In Sect. 2.4.3, we have evaluated the forces and moments on a sional analysis, then it would be extremely beneficial. This
body, and their corresponding force and moment coefficients. technique defines a set of dimensionless parameters that gov-
However, we must know about the physical quantities which ern the aerodynamic forces and moments; consequently, it
cause variations in these aerodynamic forces and moments. will reduce the number of independent variables as shown in
Eq. (2.60). These thoughts are officially incorporated within
42 2 Basic Concepts
the Buckingham pi theorem,5 which, indeed, is a formaliza- In above equation, by replacing the physical quantities with
tion of Rayleigh’s method of dimensional analysis. their dimensions, we get
Now let us return back to our consideration of resultant
α1 β1
force on a given body at a given angle of attack, Eq. (2.60) 1 = ML−3 LT−1 (L)γ1 MLT−2
can also be expressed in the following form:
or
g1 (ρa , Ua , c, μa , aa , R) = 0 (2.63)
1 = (M)α1 +1 (L)−3α1 +β1 +γ1 +1 (T)−β1 −2
The total number of variables in Eq. (2.63) is equal to 6, i.e.,
n = 6. Further, from Buckingham pi theorem, there exists Since 1 is dimensionless, terms on the right side of the above
three fundamental dimensions in the present problem, i.e., equation must also be dimensionless. Thus, the exponents of
r = 3. They are the dimension of mass (M), the dimension of M, L, and T should be individually zero. That is,
length (L), and the dimension of time (T). Since n − r =3, and
thus, Eq. (2.63) can be re-expressed in terms of three dimen- α1 + 1= 0
sionless products, shown below:
g2 (1 , 2 , 3 ) = 0 (2.64) −3α1 + β1 + γ1 + 1 = 0 =⇒ β1 + γ1 = −4
where
−β1 − 2 = 0 =⇒ β1 = −2
1 = g3 (ρa , Ua , c, R) (2.65)
On solving the above equations, we have
2 = g4 (ρa , Ua , c, μa ) (2.66)
=⇒ α1 = −1; β1 = −2; γ1 = −2
3 = g5 (ρa , Ua , c, aa ) (2.67)
Therefore, dimensionless 1 product can be written as
Let us further assume the following relationship for 1
R
α1 β1 γ1 1 =
1 = (ρa ) (Ua ) (c) R (2.68) ρa va2 c2
The term ρ vR2 c2 is dimensionless in which c2 has the dimen-

a a
sion of area and can be suitably replaced with reference area,
5 Let r is equal to the number of fundamental dimensions required to Sr . Moreover, if 21 is multiplied in the denominator, the term
express the physical variables. In mechanics, the physical quantities are
usually described in terms of mass, length, and time; hence, r = 3. Let 1
R
will still be dimensionless. Thus, 1 can be rewritten
2 ρa v a S r
2
Q1 , Q2 , . . ., Qn represent n physical variables in the physical relation as
f1 (Q1 , Q2 , . . . , Qn ) = 0 (2.61) R R
1 = = = CR
2 ρa Ua Sr
1 2 qa Sr
The above physical relation may be re-expressed as a function of (n − r)
dimensionless products, commonly called products.
where CR is the force coefficient. Repeating the similar exer-
f2 (1 , 2 , . . . , n−r ) = 0 (2.62) cise for the other two dimensionless products, we get
where each product is a dimensionless product of a set of r physical

ρa Ua c 1
2 = =
variables plus one other physical variable. Let Q1 , Q2 , . . ., Qr be the μa Re
selected set of r physical variables. Then
and
1 = f3 (Q1 , Q2 , . . . , Qr , Qr+1 )
Ua
2 = f3 (Q1 , Q2 , . . . , Qr , Qr+2 ) 3 = =M
aa
........................................
n−r = fn−r+2 (Q1 , Q2 , . . . , Qr , Qn ) where Re is the Reynolds number and M is the Mach number.
Therefore, Eq. (2.64) can be expressed as
The repeating variables Q1 , Q2 , . . ., Qr are chosen such that they include
all the r dimensions used in the problem. Also, the dependent variable
R should appear only one of the products.
2.5 Dimensional Analysis 43

R ρa Ua c Ua ric similarity; if the quantities are associated to the motion,
g2 , , =0 (2.69)
1
ρ U 2S
2 a a r
μa aa then it is termed as kinematic similarity; and if the quantities
refer to forces, then the similarity is known as dynamic sim-
or ilarity. Note that the two flows will be called similar only if
all these similarities exist simultaneously.
g2 (CR , Re, M) = 0 (2.70) In investigations of physical similarity, the full size
or actual scale systems are called prototypes, while the
or laboratory-scale systems are known as models. It should be
noted that the use of same fluid with both prototype and
CR = g6 (Re, M) (2.71) model is not necessary, and also the model need not be
smaller than the prototype always.
Comparing Eqs. (2.63) and (2.71), it can be observed that
using Buckingham pi theorem the number of independent 2.5.1.1 Geometric Similarity
variables is reduced from 5 to 2. Thus, it has saved a huge Geometric similarity refers to the similarity of shape, in which
amount of resources and the wind tunnel testing time. Further- the ratio of linear dimension of prototype and the correspond-
more, the nondimensional parameters Re and M are known ing linear dimension of model is same everywhere. This ratio
as similarity parameters. of linear dimensions is called scale factor. Therefore, geomet-
Since the lift and drag are the components of resul- rically similar objects are similar in shape but their size differs.
tant force, and moment is generated due to these forces, If l1 and l2 , respectively, are the linear dimensions of any
therefore, object, then the condition for geometric similarity is

CL = g7 (Re, M) (2.72) lp lp
= = lr (2.78)
lm 1 lm 2
CD = g8 (Re, M) (2.73)
where lp and lm , respectively, refer to the prototype and model,
CM = g9 (Re, M) (2.74) and lr is termed as scale factor or model ratio. One should note
that the roughness of surface should also be geometrically
However, one should note that the above analysis was carried similar.
out for a given angle of attack (α). If α is allowed to vary then
2.5.1.2 Kinematic Similarity
CL = g10 (Re, M, α) (2.75) It refers to the similarity of motion, since motion of an object
CD = g11 (Re, M, α) (2.76) is described by two dimensions: length and time. Thus, kine-
matic similarity between the model and the prototype implies
CD = g11 (Re, M, α) (2.77) the similarity of length (i.e., geometric Similarity) and simi-
larity of time intervals. If the ratio of corresponding lengths
of model and prototype is denoted by lr , and if the ratio of
corresponding time intervals is given by tr , then the mag-
2.5.1 Concept of Flow Similarity nitudes of corresponding velocities will be given by ltrr , and
the magnitudes of corresponding accelerations will be given
The essential requirement to establish physical similarity
by tl2r .
between two flows is that the physics behind them must be r
For kinematic similar flows over the model and the proto-
same. That is, the flow in a horizontal pipe is quite differ-
types, the streamline patterns formed at corresponding time
ent from the flow in an open channel. This is because the
intervals are geometric similar. Moreover, for the flows past a
pipe flows are governed by the viscous and pressure forces,
nonporous solid surface, the streamlines represented by solid
whereas the open-channel flows are predominately influenced
boundaries will be kinematic similar only if the boundaries
by gravity force. Thus, the flows which are governed by the
are geometrically similar. In other words, the kinematic sim-
same physical principle, but operating under different condi-
ilarity ensures the geometric similarity; however, vice versa
tions, will be called similar if there are some specified phys-
is not always true. For kinematic similar flows, though, the
ical quantities whose ratios between these flows are found
streamlines in the near vicinity of solid surface imply geo-
constant everywhere. If the specified quantities are related to
metric similarity but, in the far field regions, away from the
geometrical dimensions, then the similarity is called geomet-
surface, same need not be true.
44 2 Basic Concepts
2.5.1.3 Dynamic Similarity wind tunnel testing. For a scale aircraft model, the measured
It implies the similarity of forces. For dynamic similarity, the values of CL , CD , and CM through wind tunnel testing will be
ratio of magnitudes of forces for model is same as the magni- same as for actual atmospheric flight as long as Re and M of
tude ratio of forces for prototype, at corresponding locations. the wind tunnel test section are the same as for actual flight.
In general, the governing forces acting
in fluid flows are vis-
Fμ , force due to pressure
cous force due to fluid viscosity
difference Fp , gravity force Fg , force due to surface ten- 2.6 Types of Flows
sion (Fst ), and elastic force due to fluid compressibility (Fe ).
According to the Newton’s second law of motion, the resul- In the study of aerodynamics, we are encountered with numer-
tant (Fr ) of these forces will cause the acceleration of a fluid ous and distinct types of flow. In this section, we will itemize
element. Thus, in vector notation, we write and compare these flows, and will describe their most impor-
tant physical phenomena in brief.
−
→ −
→ −
→ −
→ −
→ −
→
F r = F μ + F p + F g + F st + F e (2.79)
But, according to the Newton’s third law of motion, the inertia 2.6.1 Continuum and Non-continuum Flows
−
→ −
→
force F i must be equal and opposite to F r . Hence,
Any substance which looks continuous is actually composed
−
→ −
→ of a large number of molecules which are in constant motion
F i= − F r (2.80) and collision. In kinetic theory of gases (also referred to as sta-
tistical mechanics), the motion of the substances is studied by
where negative sign shows that the forces are acting in oppo- writing the governing equations for each individual molecule
site direction to each other. Moreover, to establish dynamic (or their statistical groups) with respect to some known initial
similarity, the magnitude ratios of these forces have to besame
→
conditions. This approach, although, has its own merits, but
−
for both model and prototype. Usually, the inertia force F i it is not suitable for practical purposes. Indeed, in the prob-
is taken as the common one to describe the ratio as lems of engineering interest, we are not concerned in study-
ing the motion of individual molecules, but rather in gross
−
→ −
→ −
→ −
→ −
→
| F μ | | F p | | F g | | F st | | Fe| behavior of the fluid thought of as a continuous substance.
→ , −
− → , − → , − → , and −→ Even though the postulation of a continuous fluid (contin-
| Fi| | Fi| | Fi| | Fi| | Fi|
uum) is only a convenient assumption, it interestingly turns
After performing the order of magnitude analysis, the above out to be a valid approach to many practical problems, where
ratios lead to some important dimensionless parameters such only macroscopic information is of interest. Here, it is also
as inherent that the given fluid element can be divided indefi-
−
→
nitely into smaller and smaller particles in such a way that the
| F μ| properties are constant in these sub-divided particles as well.
• −
→ ≡ Re, where Re stands for Reynolds number.
| F i| However, the validity of continuum assumption lies in the
−
→
| F p| fact that whenever the smallest fluid element has the sufficient
• −
→ ≡ Eu, where Eu stands for Euler number.
| F i| number of molecules making the statistical averages mean-
−
→ ingful. Furthermore, the changes in properties are assumed to
| F g|
• −
→ ≡ Fr, where Fr stands for Froude number.
| F i| be gradual over a period of time. Consequently, any property
−
→ can be defined as continuous function of geometrical positions
| F st |
• −→ ≡ Wb, where Wb stands for Weber number. in space and time (Fig. 2.11).
| F i|
−
→ The advantage of continuum assumption is that instead
| F e|
• −
→ ≡ M, where M stands for Mach number. of studying the instantaneous states of large number of
| F i|
molecules, we have to deal with only certain properties
Indeed, for dynamic similarity, one needs to match these describing the gross behavior of the matter. For compress-
dimensionless numbers between model and prototype. In ible fluids, the relevant properties are the pressure (p),
Sect. 2.5, we have emphasized two dimensionless numbers: temperature (T), density (ρ), velocity (v), shear stress (τ ),
Reynolds number (Re) and Mach number (M), which are coefficient of viscosity (μ), internal energy (u), entropy (s),
dominant similarity parameters in many aerodynamic appli- and coefficient of thermal conductivity (κ). The macroscopic
cations of practical importance. That is, the flows over geo- approach with continuum hypothesis is valid only when
metrically similar objects at same Re and M are dynamically the mean free path (λ) of the molecules is much less than
similar, and hence, CL , CD , and CM will also be same for the smallest significant dimension (l) of the problem under
both objects. This is an important result for the validity of investigation. The mean free path length of molecules is
2.6 Types of Flows 45
δm
δV
Free molecular Region of

region continuum
δm
δV
p
δV
δ V0
Fig. 2.11 Concept of density
defined as the average distance covered by the molecules characteristics length scale. This is the other extreme where
between two successive collisions. For λ l, the molecules the mean free path of the molecules is of the same order or
impact the body surface so frequently that the body cannot more than the characteristic dimension of the problem. Under
distinguish the individual molecular collisions, and thus the this situation, the molecular collisions with the body surface
surface feels the fluid as a continuous medium. This type of occur only intermittently, and the surface can feel distinctly
flow is called continuum flow. Further, the ratio of mean free each molecular impact. Clearly, the continuum hypothesis
path to the characteristics dimension of the object is known breaks down at these altitudes. This type of flow is termed
as Knudsen number (Kn), given by Eq. (2.81). as free molecular flow or rarefied flow. Under such condi-
tions, the fluid is analyzed at microscopic level which is the
λ
Kn = (2.81) subject matter of rarefied gas dynamics. The space shuttle
l encounters free molecular flow at the time of re-entry to the
We should note that the fluid element must have sufficient outermost edge of the earth’s atmosphere. The air density is so
number of molecules to define a constant value of density low that λ becomes the order of the shuttle size. The classifi-
averaged over the fluid element. At standard conditions of cation of flow regimes based on the Knudsen number is given
pressure and temperature, the elemental fluid volume (δV) below:
as small as 10−9 mm3 contains about 3 × 107 molecules,
sufficient enough to define the constant density (ρ). • Fluid continuum (Kn < 0.01),
• Slip flow (0.01 < Kn < 0.1),
δm • Transition flow (0.1 < Kn < 3), and
ρ= lim (2.82)
δVδV0 δV • Free molecular flow (Kn > 3).
However, when we go at higher altitudes, then these consider- The Knudsen number (Kn) can be related with the Mach num-
ations do not hold true due to very less number of molecules ber (M) and Reynolds number (Re) as
present in the given elemental volume. Also, the molecules
get farther apart and start having their own identities eventu- √ M
Kn = 1.26 γ (2.83)
ally leading to high mean free path length compared to the Re
46 2 Basic Concepts
where γ is ratio of specific heats for the gas. At hypersonic 2.6.3 Uniform and Nonuniform Flows
Mach numbers (M > 5) and for large values of Reynolds
number (Re), the Knudsen number becomes In a fluid flow, if the flow parameters do not change with
respect to space coordinates (say xyz) but can change with
M respect to time (t) is known as uniform flow. That is, if the
Kn = √ (2.84)
Re parameters have the same value at all points at a particular time
instant, the flow is referred to as uniform flow. However, if the
flow parameters vary and are different at different points in a
flow field, the flow is termed as nonuniform flow. For example,
2.6.2 Steady and Unsteady Flows consider three points A, B, and C in a velocity field −→v , which
are monitored by a stationary observer at two different times
In the dynamics of rigid bodies, the description of motion t1 and t2 . The flow will be called “uniform” if
of each particle can be given separately and precisely. For At t = t1 ,
example, if a large number of particles are in motion, and we − → −
are interested in specifying the velocity of nth particle, then →
v A 1 = −
v B 1 = →
v C 1
we can do so by writing the following scalar equations:
and at t = t2 ,
un = un (t) (2.85) − → −
→
v A 2 = −
v B 2 = →
v C 2
vn = vn (t) (2.86)
wn = wn (t) (2.87) The above relations can be expressed in differential form as
→
Thus, the particular particle is identified easily with the use of d −v
=0
subscript. In contrast, such a representation is not possible for ds
fluids as it has an infinite number of particles, which makes
That is, for an uniform flow, the spatial gradient of velocity
this approach unmanageable. Therefore, spatial coordinate
along the flow direction s is zero at a particular time instant t.
system is employed to help identify particles in a flow. The
Interestingly, this is in contrast to steady flows, where the flow
velocity of all particles in a flow can therefore be expressed d|−
→v|
in the following manner: parameters do not change with respect to time, i.e., dt = 0,
but the velocities at the points A, B and C need not be equal
u = u (x, y, z, t) (2.88) to each other at any point of time. Thus, it is evident that
steadiness of flow and uniformity of flow does not necessar-
v = v (x, y, z, t) (2.89) ily go together. Furthermore, by combining the above dis-
w = w (x, y, z, t) (2.90) cussed flows, uniform, nonuniform, steady and unsteady, we
can classify any flow into one of the following four types:
By specifying coordinates (x, y, z) and time t, one can deter-
1. Steady uniform flow: The flow conditions do not change
mine the velocity components of a fluid element at the partic-
with respect to position or with time. The fluid flow in
ular position and specified time. Therefore, it is evident that
a pipe of constant diameter with constant velocity is an
spatial coordinates have replaced the subscript “n” in rigid
example to this type of flow.
body velocity fields. This approach is called field approach.
2. Steady nonuniform flow: The flow conditions change spa-
Furthermore, if the properties and flow characteristics at each
tially but do not change with respect to time. The flow in a
location in space remain constant with time, the flow is called
convergent–divergent nozzle with constant velocity at the
steady flow. In contrast, a time-dependent flow is called an
inlet is steady but nonuniform due to varied geometry of
unsteady flow. The steady flow velocity field would then be
the nozzle.
given as
3. Unsteady uniform flow: At a particular instant of time,
if the flow conditions do not change spatially but change
u = u (x, y, z) (2.91)
with respect to time, they are called unsteady uniform flow.
v = v (x, y, z) (2.92) For example, in urban areas, water is supplied through the
pipes of constant diameter. The mass flow rate of the water
w = w (x, y, z) (2.93)
is maintained constant by pumping station. The sudden
stoppage of pump leads to unsteadiness in the flow.
4. Unsteady nonuniform flow: If the flow conditions change Since, ρ = ∀1 , where ∀ is the specific volume. Thus,
spatially as well as with respect to time, they are called Eq. (2.97) can be written as
unsteady nonuniform flow. The unsteady nonuniform flow
is caused due to generation of waves in an open-channel 1 d∀
βT = − (2.98)
flow. ∀ dp T
However, if the system is made insulated, i.e., no exchange of

2.6.4 Incompressible and Compressible Flows heat with the surrounding is possible, then the compression
takes place isentropically. Therefore, β is termed as isentropic
Whenever a substance is compressed by applying the pres- compressibility (βs ), defined as
sure, its density changes. The gases undergo a large change in
density whenever pressure is applied, whereas liquid shows 1 dρ
βs = (2.99)
relatively lower density change. In contrast, when solids are ρ dp s
compressed, virtually no change in density is noticeable. The
or
amount by which a substance can be compressed is measured

in terms of a specific property, known as compressibility. 1 d∀
Let us consider a small fluid element of specific volume βs = − (2.100)
∀ dp
∀ on which the pressure p is applied from all sides. If the
s
pressure is increased by an infinitesimal amount dp, and the where ∀ = ρ1 and the subscript s denotes that the partial
corresponding infinitesimal decrease in specific volume of the derivative is taken at constant entropy. In general, compress-
fluid element is d∀ , then the compressibility (β) of the fluid ibility of the gases is of several orders of magnitudes higher
element will be given by than that of liquids. From Eq. (2.96), it is evident that for a
given pressure change dp, the corresponding change in den-
1 d∀
β= (2.94) sity dρ will be negligibly small for liquids and thus, liquids
∀ dp are termed as incompressible. On the other hand, for a given
Thus, the compressibility is defined as the fractional change pressure change dp, the corresponding change in density dρ
in volume per unit change in pressure. Since the fluid density will be large for the gases (because of large β) and thus, gases
is the inverse of specific volume, i.e., ρ = ∀1 , thus Eq. (2.94) are usually compressible. However, the deviation from this
is also written as occurs in the low speed flow of a gas, where the change in
pressure dp from one point to another is very small as com-
1 dρ pared to the pressure itself despite of large β and the value of
β= (2.95)
ρ dp dρ will be dominated by small dp. Under this condition, the
density of gas (ρ) will be constant, and hence, we can treat
or the low speed flow of a gas as incompressible.
A more elegant definition of impressibility is given in terms
dρ = βρdp (2.96) of the Mach number. In Sect. 2.5.1.3, we have seen that the
Mach number is essentially a ratio of elastic force to inertia
That is, change in pressure of a fluid and the corresponding force. From order of magnitude analysis of the forces, this
change in its density is given by Eq. (2.96). Compressibility, ratio turns out to be the ratio of local flow speed to local
although, can be defined either by Eq. (2.94) or by Eq. (2.95), speed of sound. That is,
but it is yet not enough. When a gas is compressed by increas-
ing the pressure, temperature of the gas increases and thus, the v
M= (2.101)
heat transfer through the system (gas) boundary is inevitable. a
If the gas temperature is held constant by some suitable heat
transfer mechanism, then β is referred to as isothermal com- Moreover, in engineering applications, the flow with frac-
pressibility (βT ), given by tional density change of 5% is assumed a constant density
flow, i.e., an incompressible flow. At standard sea level con-
ditions (p = 101325 Pa; T = 288.15 K), this corresponds to
1 dρ
βT = (2.97) M = 0.3 for air (γ = 1.4). Thus, the flow with M < 0.3 is
ρ dp T
treated incompressible, whereas all the flows with M > 0.3
fall under compressible flow regime.
48 2 Basic Concepts
M a < 0.8
Fig. 2.12 Subsonic flow (Ma < 1)
2.6.5 Inviscid and Viscous Flows gives rise to a rule of thumb that Ma < 0.8 for subsonic flow
over slender bodies. For blunt bodies, Ma must be even lower
We have seen that the fluid molecules are free to move in to ensure fully subsonic flow. Moreover, a small disturbance
random fashion within the fluid boundaries. The movement travels with acoustic speed and since the flow speed in sub-
of these molecules causes the transport of mass, momentum, sonic regime is less than the speed of sound, thus, the presence
and energy from one location to another in the fluid. Essen- of disturbance is felt in domain throughout.
tially, this transport of matter at microscopic scale gives rise
to the phenomena of mass diffusion, viscosity, and thermal 2.6.6.2 Transonic Flow
conduction. All real fluid flows which manifest the effects of It is essentially a mixed region of subsonic and supersonic
these transport phenomena are called viscous flows. Contrar- flows. As stated earlier, if Ma is close to unity, flow over slen-
ily, a flow which does not have viscosity, thermal conduction, der bodies such as airfoil will be accelerated to supersonic
or diffusion is termed as inviscid flow. Mach numbers terminated by weak shock waves, downstream
Theoretically, a flow is assumed to be inviscid when the of which flow becomes subsonic again, shown schematically
Reynolds number (Re) tends to infinity. However, many flows in Fig. 2.13a. Furthermore, if Ma is increased slightly above
with high but finite Re are also assumed to be inviscid where unity, a bow-shock stands ahead of the body behind which
the effect of mass diffusion, viscosity, and thermal conduc- flow becomes subsonic (Fig. 2.13b). This subsonic flow fur-
tivity is limited to a very thin region adjacent to the solid ther expands over the body to low supersonic Mach numbers
surface, commonly known as boundary layer. The remainder ceased by weak shocks, resembling a “fishtail” pattern. Thus,
of the flow outside the boundary layer is essentially inviscid. 0.8 < Ma < 1.2 generally accepted for transonic flow over
The inviscid flow theory adequately predicts the pressure dis- slender bodies.
tribution over the body surface, streamline patterns away from
the body, and the flow field in the far regions away from body. 2.6.6.3 Supersonic Flow
Since the aerodynamic drag due to viscosity is vital, inviscid When the Mach number is more than unity at each and every
theory by itself does not envisage the total drag. point in a domain, the flow is termed as supersonic flow.
It is essentially dominated by shock and expansion waves
across which flow properties and streamlines change discon-
2.6.6 Mach Number Flow Regimes tinuously (Fig. 2.14), in contrast to a subsonic flow, where the
change in flow properties and streamlines are rather smooth.
Among all the criteria of categorizing and describing differ- In supersonic flows, the flow speed is greater than the speed
ent aerodynamic flows, the classification based on the Mach of sound, any disturbances traveling with acoustic speed can-
number is presumably the most ubiquitous. If Ma is the not travel upstream. Consequently, the upstream flow remains
freestream Mach number at an arbitrary point in a flow field, unaware of the disturbances downstream. It is the fundamen-
then using Ma as the criterion we can define following speed tal reason behind the occurrence of shock waves in supersonic
regimes. flows. Readers will be able to appreciate this difference more
clearly in Chaps. 9–14.
2.6.6.1 Subsonic Flow
When the flow speed is less than the speed of sound, we call
it as subsonic flow. Thus, in principle, we may say that the 2.6.6.4 Hypersonic Flow
flow with Mach number less than one (Ma < 1) is a subsonic When the flow Mach number increases above 1, for a fixed
flow. However, as the freestream Mach number approaches value of wedge angle (α), the shock wave comes closer to the
close to unity, the local Mach number somewhere in the flow wedge (Fig. 2.15). The shock becomes thin but its strength
domain becomes supersonic. Such is the case with flow over is considerably increased, leading to high temperature in the
slender bodies (airfoil) as shown in Fig. 2.12. Thus, Ma < 1 region between the shock and the wedge. At this high tem-
does not guarantee a totally subsonic flow over the body. This perature, nitrogen (N2 ) and oxygen (O2 ) molecules in the air
(a)
(b)
Fig. 2.13 Transonic flow (0.8 < Ma < 1.2)
Expansion fan
Shock wave
1.2 < M a < 5.0 α Wedge
Fig. 2.14 Supersonic flow (1.2 < Ma < 5)
begin to dissociate. Also, the shock starts interacting with the flow, the flow field is called hypersonic flow (here again, as a
viscous layer (boundary layer) on the wedge surface, lead- rule of thumb Ma > 5 for hypersonic flow).
ing to high viscous dissipation. Moreover, the impingement
of shock wave on the boundary layer imposes an adverse
pressure gradient on the boundary layer. Such a flow, where
2.7 Hodograph Transformation
shock–boundary layer interactions at high Mach number, and
A Hodograph is a vector diagram, also known as velocity
dissociation of air at high temperature begin to dominate the
diagram, which shows the changes in velocity with respect to
50 2 Basic Concepts
Oblique shock wave
5.0 < M a
α Wedge
Shock−viscous layer
interactions
Fig. 2.15 Hypersonic flow (Ma > 5)
Physical plane Hodograph plane
Fig. 2.16 Schematic of a typical Hodograph transformation
position or time. It was first used by James Bradley, but the From Eq. (2.107), we can infer that
practical development of Hodograph was later carried out by
p
Sir William Rowan Hamilton (1805–1865). The Hodograph =A (2.108)
transformation finds vast applications in aerodynamics, as it ργ
can transform the nonlinear equations to the linear ones. To
where A is an arbitrary constant. Further, differentiating
understand this transformation, let us consider the following
Eqs. (2.105) and (2.106) and substituting Eq. (2.108) in the
Euler’s differential equations (Fig. 2.16):
results, we get
∂ρ ∂
+ (ρu) = 0 (2.102) ρt + ρux + uρx = 0
∂t ∂x (2.109)
∂u ∂u 1 ∂p ut + uux + Aγργ−2 ρx = 0 (2.110)
+u =− (2.103)
∂t ∂x ρ ∂x
Note that both A and γ are constants. Let us introduce another
∂ p ∂ p
+u =0 (2.104) constant c, such that
∂t ργ ∂x ργ
Rewriting the above equations in simplified notations, we Aγ = c (2.111)

have
In physical plane, the position coordinates x and t are the inde-
ρt + (ρu)x = 0 (2.105) pendent variables, while in Hodograph plane, the components
of the velocity vector ρ and u are the independent variables.
px
ut + uux + =0 (2.106) The dependent variables (including position) are determined
ρ from the velocity components. Let us begin with Eq. (2.109),

p p rewriting as
+u =0 (2.107)
ργ t ργ x ∂ (ρ, x) ∂ (t, u) ∂ (t, ρ)
+ρ +u =0 (2.112)
∂ (t, x) ∂ (t, x) ∂ (t, x)
2.7 Hodograph Transformation 51
The application of Hodograph transformation results in ∂u ∂v

− =0 (2.125)
∂y ∂x
∂ (ρ, x) ∂ (t, u) ∂ (t, ρ)
+ρ +u = 0 (2.113)
∂ (ρ, u) ∂ (ρ, u) ∂ (ρ, u) In physical domain, the solutions of equation (2.124) and
(2.125) are
or
xu + ρtρ − utu = 0 (2.114)
u = u (x, y) (2.126)
or v = v (x, y) (2.127)
xu = −ρtρ + utu (2.115) where x and y are independent variables. Writing u and v in
partial derivatives forms
Similarly, rewriting Eq. (2.110) and applying the Hodograph
transformation, we get ∂u ∂u
du = dx + dy (2.128)
∂ (u, x) ∂ (t, u) ∂ (t, ρ) ∂x ∂y
+u + cργ−2 =0 (2.116) ∂v ∂v
∂ (ρ, u) ∂ (ρ, u) ∂ (ρ, u) dv = dx + dy (2.129)
∂x ∂y
In simplified notations
In simplified notations,
γ−2
−xρ + utρ − cρ tu = 0 (2.117)
du= ux dx + uy dy (2.130)
or dv= vx dx + vy dy (2.131)
xρ = utρ − cργ−2 tu (2.118)
Solving for dx and dy, we get
If u and ρ are independent variables, then
vy du − uy dv
dx = (2.132)
xuρ = xρu (2.119) ux vy − uy vx
By taking the partial derivatives of Eq. (2.115) with respect and

−vx du + ux dv
to ρ and Eq. (2.118) with respect to u, we obtain dy = (2.133)
ux vy − uy vx
xuρ = utuρ − tρ − ρtρρ (2.120)
Note that the denominators in the above equation can be rep-
γ−2
xρu = tρ + utρu − cρ tuu (2.121) resented in the determinant form as

u u
Thus, from Eq. (2.119) ≡ x y
vx v y
utuρ − tρ − ρtρρ = tρ + utρu − cργ−2 tuu (2.122)
Thus, Eqs. (2.132) and (2.133) become
On simplification v u
y y
dx = du− dv (2.134)

cργ−2 tuu − ρtρρ − 2tρ = 0 (2.123) v
x ux
dy = − du + dv (2.135)

In general, the Hodograph transformation can be applied to
any flow regimes. However, it is more accurate in solving the Let us now express the variables x and y as the functions of u
flow equations of transonic flows due to their mixed type of and v as
behavior. We can understand it further through an example.
Consider a two-dimensional transonic flow with perturbation x = x (u, v) (2.136)
velocity components u and v. The governing equation for this y = y (u, v) (2.137)
flow can be written as
∂u ∂u The respective differentials will be
(γ + 1) M2 ∂u
1 − M2 + = u (2.124)
∂x ∂y Ua ∂x dx = xu du + xv dv (2.138)
But, from irrotationality condition dy = yv du + yv dv (2.139)
52 2 Basic Concepts
Comparing Eqs. (2.134) and (2.138), we get The aircraft requires lift to fly in the air, and thus a pair
of wings is attached to the long cylindrical section, known as
vy
xu = (2.140) fuselage. Although each component of an aircraft contributes
in generating the lift, the major portion comes from wings
uy
xv = − (2.141) only. That is, the wings are main lifting surfaces which con-

tributes maximum to the lift. The wing fixed at the right side
Similarly, from Eqs. (2.135) and (2.139), we obtain of fuselage is termed as starboard wing, and the wing located
at the left is called the port wing. To produce lift, the air-
vx craft must be pushed through the air. When the aircraft flies,
yu = (2.142)
its motion is retarded by an opposing force called drag. In
vx
yv = − (2.143) turbine-powered aircraft, to generate thrust and overcoming
the drag, engines are mounted beneath the wings.
Introduction of xu , xv , yu , and yv in Eqs. (2.124) and (2.125) To control and maneuver the aircraft, wings of smaller
leads to sizes are attached at the rear end, commonly referred to as tail
section. A pair of small-size fixed wings mounted horizontally
∂y ∂x (γ + 1) M2 ∂y is called horizontal stabilizer, and a vertically placed fixed
1 − M2 + = u (2.144) part is known as vertical stabilizer. As the name suggests,
∂v ∂u Ua ∂v
∂x ∂y both stabilizers provide stability to the aircraft. In case of a
− =0 (2.145) sudden gust, nose of the aircraft uses to swing sideways from
∂v ∂u
the original direction of motion, called yaw. Likewise, the up-
Equations (2.144) and (2.145) are the transonic Hodograph and-down motion of the nose is known as pitch. The horizontal
equations. Note that these transformed governing equations stabilizer prevents the pitching motion, whereas the vertical
are linear in nature; x and y are dependent variables, while u stabilizer is responsible for preventing the yawing motion of
and v are independent variables. the nose.
The property of a fluid due to which the fluid layer has
shearing stresses between them is known as coefficient of
Summary viscosity (μ). With increase of temperature, viscosity of liquid
decreases. Empirically, viscosity of a liquid is expressed in the
Aerodynamics is the branch of science devoted to study the following form:
motion of air flow past the solid objects. The field that studies
2
the motion of air, particularly its interactions with the aircraft, μ T0 T0
ln =a+b +c
is known as aircraft aerodynamics. However, the scope of μ0 T T
aerodynamics is not just limited to the aircraft, rather it com-
prises all those phenomena where the air flow past a structure where the coefficients a, b, and c are arbitrary constants; T0
whether stationary or in motion, in the earth’s sensible atmo- is the reference temperature; and μ0 is the viscosity of liquid
sphere. The word aerodynamics is made of two Greek words: at the temperature T0 .
aer (aero) means air + dynamikós (dynamics) refers to power. In contrast, the viscosity of a gas increases with tempera-
Essentially, the study of aerodynamics revolves around esti- ture as shown by the following relations:
mating the forces and moments acting on an airborne vehicle.
From the very beginning, we were interested to emulate a bird ⎧ n
and while thinking of human flight centered on the imitation ⎪
⎨ TT Power law
μ 0
of birds, several attempts have been made to bring such a = 3
μ0 ⎪
⎩ T 2 T0 +S Sutherland law
device into reality. But even today it seems unachievable with T0 T+S
existing technologies. In a bird’s flight, the flapping wings
perform a dual role; they serve the purpose of both propul- where μ0 is the known viscosity at a known absolute tem-
sive and aerodynamic devices. Any man-made device which perature T0 ; and S is called the Sutherland constant. For air,
imitates the flight of a bird is known as ornithopter. n = 0.7 and S = 110 K; a more useful form of Sutherland
The aircraft is a flying object that operates in the sensi- formula is given below:
ble atmosphere. Depending upon the mission requirement,
3
aircraft varies in shapes and sizes. The whole aircraft has a −6 T2
median plane of symmetry, dividing the machine into two μ = 1.46 × 10 ;
T + 111
halves. Basically, these parts are mirror image of each other,
if this plane of symmetry is considered as a mirror. for, 0.01 atm < pstatic < 100 atm; 0 K < T < 3000 K
Summary 53
According to the Archimedes principle, an object will be ilarity. Note that the two flows will be called similar only if
buoyed up by a force equal in magnitude to the weight of all these similarities exist simultaneously.
fluid displaced by the object. Similarly, an aircraft flying in In investigations of physical similarity, the full size
the air experiences a buoyant force equal in magnitude to or actual scale systems are called prototypes, while the
the weight of displaced air. The forces acting on the flying laboratory-scale systems are known as models. It should be
aircraft are known as aerodynamic forces. The aerodynamic noted that the use of same fluid with both prototype and
force is resolved into horizontal and vertical components. The model is not necessary, and also the model need not be
horizontal component acting in the direction of freestream is smaller than the prototype always.
termed drag, and the vertical component, perpendicular to the Whenever a substance is compressed by applying the pres-
freestream direction, is known as lift. sure, its density changes. The gases undergo a large change in
In a three-dimensional body, if a section is cut with a plane density whenever pressure is applied, whereas liquid shows
(parallel to the plane of symmetry), the intersection of the relatively lower density change. In contrast, when solids are
body surface with that plane is termed as profile. For an air- compressed, virtually no change in density is noticeable. The
craft wing, this profile is better known as airfoil. When the amount by which a substance can be compressed is measured
flow past an airfoil, then the latter produces the aerodynamic in terms of a specific property, known as compressibility. If
forces, namely, lift and drag. Alternatively, for a given area, the pressure is increased by an infinitesimal amount dp, and
the profile which generates a maximum amount of lift is called the corresponding infinitesimal decrease in specific volume
an airfoil. The shape and orientation of an airfoil section sliced of the fluid element is d∀ , then the compressibility (β) of the
at various locations on the wing usually depend on its distance fluid element will be given by
from the plane of symmetry.
At conventional angles of incidence, in comparison with 1 d∀
β=
freestream static pressure, the suction (upper) surface experi- ∀ dp
ences a decrease in pressure over a large portion of it, whereas
a lesser decrease in pressure is felt by the lower (pressure) sur- When a gas is compressed by increasing the pressure, temper-
face. This uneven pressure acting on upper and lower surfaces ature of the gas increases and thus, the heat transfer through
leads to the nonuniform pressure distribution around the air- the system (gas) boundary is inevitable. If the gas temper-
foil, due to which lift and drag are produced on the airfoil. ature is held constant by some suitable heat transfer mecha-
The aerodynamic forces on an airfoil section may be rep- nism, then β is referred to as isothermal compressibility (βT ),
resented by lift, drag, and pitching moment. At each value given by
of the lift coefficient, there will be a particular point about
1 dρ
which the pitching moment coefficient is zero. The aerody- βT =
namic effects on the airfoil section may be represented by the ρ dp T
lift and the drag alone acting at that point, termed as the center
However, if the system is made insulated, i.e., no exchange
of pressure.
of heat with the surrounding is possible, then the compres-
The aerodynamic center is the reference point about which
sion takes place isentropically. Thus, β is termed as isentropic
the pitching moment coefficient does not change with changes
compressibility (βs ), defined as
in the angle of attack.
The essential requirement to establish physical similarity
1 dρ
between two flows is that the physics behind them must be βs =
ρ dp s
same. That is, the flow in a horizontal pipe is quite differ-
ent from the flow in an open channel. This is because the
The fluid molecules are free to move in random fashion within
pipe flows are governed by the viscous and pressure forces,
the fluid boundaries. The movement of these molecules causes
whereas the open-channel flows are predominately influenced
the transport of mass, momentum, and energy from one loca-
by gravity force. Thus, the flows which are governed by the
tion to another in the fluid. Essentially, this transport of matter
same physical principle, but operating under different condi-
at microscopic scale gives rise to the phenomena of mass dif-
tions, will be called similar if there are some specified phys-
fusion, viscosity, and thermal conduction. All real fluid flows
ical quantities whose ratios between these flows are found
which manifest the effects of these transport phenomena are
constant everywhere. If the specified quantities are related to
called viscous flows. Contrarily, a flow which does not have
geometrical dimensions, then the similarity is called geomet-
viscosity, thermal conduction, or diffusion is termed as invis-
ric similarity; if the quantities are associated to the motion,
cid flow.
then it is termed as kinematic similarity; and if the quantities
Among all the criteria of categorizing and describing dif-
refer to forces, then the similarity is known as dynamic sim-
ferent aerodynamic flows, the classification based on the
54 2 Basic Concepts
Mach number is presumably the most ubiquitous. If Ma is where the values of stresses are given in Nm−2 . Calculate
the freestream Mach number at an arbitrary point in a flow the thermodynamic pressure at this point.
field, then using Ma as the criterion we can define the follow- 6. Consider the following stress-tensor field in MPa.
ing speed regimes: ⎡ ⎤
16x + 10 −5x 2 0
• If Ma < 1, the flow is called subsonic. τij = ⎣ −5x 2 10y 2 + 6x y 0⎦
• If 0.8 < Ma < 1.2, the flow is called transonic. 0 0 0
• If Ma > 1, the flow is termed as supersonic.
• If Ma > 5, the flow is known as hypersonic. Show the above bulk stress distribution as a scalar field.
Find the value of bulk stress at the point P (0, 10 m, 2 m).
Hodograph is a vector diagram, also known as velocity dia- 7. Consider a blunt-nosed body flying at 4.5 kms−1 at an alti-
gram, which shows the changes in velocity with respect to tude of 50 km above the sea level. Calculate the coefficient
position or time. It was first used by James Bradley, but the of viscosity (μair ) of the air at the stagnation point down-
practical development of Hodograph was later carried out by stream of the normal portion of the bow-shock formed
Sir William Rowan Hamilton (1805–1865). The Hodograph upstream of the body.
transformation finds vast applications in aerodynamics, as it 8. The lift acting on a high speed aircraft depends upon
can transform the nonlinear equations to the linear ones. its length (l), speed of the aircraft (v), fuselage diam-
eter (d), angle of attack (α), fluid density (ρ), and the
speed of sound (a). Using Buckingham pi theorem, express
Exercises the nondimensional lift as a function of other nondimen-
sional group of parameters. For the wind tunnel test on a
Descriptive Type Questions 1
scale model of the full-scale vehicle, what airspeed
10th
will be required in the test section? Also, calculate the lift
1. An aircraft is flying at 200 ms−1 at an altitude of 10,000 m acting on the model as compared to the full-scale vehicle.
above the sea level. If the Reynolds number based on the Assume the pressure and temperature encountered dur-
mean aerodynamic chord length of the wing is found to ing the wind tunnel tests are same as experienced by the
be 1.225 × 107 , calculate the Knudsen number for the full-scale vehicle.
flow field around the wing in this case. 9. If the change in volume of mass of the fluid is 3001 th of the
2. Consider the following relationship between the pressure actual volume when pressure is increased to 6 MPa, then
and the density for a liquid: calculate the compressibility and the Bulk modulus.
n 10. Calculate the pressure that must be applied to a fluid to
p ρ
= (k + 1) −k reduce its volume by 2%. Assume the Bulk modulus of
pa ρa the fluid is 2.5 Gpa.
where pa and ρa , respectively, are the pressure and den-
sity at ambient conditions, k is a constant, and n is the Multiple Choice Questions
exponent. Calculate the isothermal compressibility (βT )
and isothermal bulk modulus of elasticity (ET ). 1. A man-made device, which imitates the flight of a bird, is
3. The velocity profile over a flat plate is given by v = 2y2/3 , known as
where v is the velocity in ms−1 and y is the perpendicu-
lar distance (m). Calculate the shear stress at y = 0 and (a) aerostat
y = 82 mm. Assume μ = 8.5 poise. (b) insectothopter
4. A fuel with kinematic viscosity of four centistokes and (c) quadcopter
specific weight of 7500 Nm−3 fills the space between a (d) ornithopter
large stationary parallel plate of 500 mm2 , which are kept
2 mm apart. If one plate is pulled with an uniform veloc- 2. In fluid flows, ratio of the surface tension force and the
ity of 4 ms−1 , determine the force and power required to inertia force gives which of the following nondimensional
maintained this speed. number?
5. Consider the following stress-tensor matrix at a point in
a fluid flow. (a) Reynolds number
⎡ ⎤ (b) Euler number
−5000 4000 2000 (c) Weber number
τij = ⎣ 4000 3000 −2000⎦ (d) Strouhal number
2000 −2000 −6000
Exercises 55
3. Which of the following Knudsen number (Kn) range rep- 8. Which of the following has the unit “Stokes”?
resents the “slip flow”?
(a) specific volume
(a) Kn < 0.01 (b) specific weight
(b) 0.01 < Kn < 0.1 (c) dynamic viscosity
(c) 0.1 < Kn < 3 (d) kinematic viscosity
(d) Kn > 3
9. Consider the flow of water between two infinitely large
4. With the increase of temperature, the viscosity of a perfect parallel plates kept 20 cm apart. If the upper plate is mov-
gas ing at a constant speed of 2 ms−1 in horizontal direction
parallel to water surface while the lower plate is at rest.
(a) decreases The
shear stress developed at the lower plate surface is
(b) increases μwater = 8 × 10−3 Nsm−2 .
(c) remains same
(d) increase or decrease depends on the pressure of the (a) 0.16 Nm−2
gas. (b) 8 × 10−4 Nm−2
(c) 4 × 10−4 Nm−2
5. By increasing the temperature of a liquid, its viscosity will (d) 0.08 Nm−2
(a) decrease 10. In which of the following flow regime, the Hodograph
(b) increase transformation accurately solves the flow equations?
(c) remain same
(d) increase or decrease depends on the liquid density. (a) subsonic flow
(b) transonic flow
6. The location of aerodynamic center of a thin airfoil in (c) supersonic flow
subsonic flow, measured from the leading edge (in terms (d) hypersonic flow
of chord c), is
(a) 0.15c Keys
(b) 0.25c
(c) 0.50c 1. (d)
(d) 0.75c 2. (c)
3. (b)
7. Consider an airfoil of chord length 1 m placed in a Mach 4. (b)
1.5 freestream. The location of aerodynamic center, mea- 5. (a)
sured from the leading edge along the chord line, is 6. (c)
7. (b)
(a) 0 8. (d)
(b) 0.25 9. (d)
(c) 0.50 10. (b)
(d) 1.5
Governing Equations of Fluid Flows
3
To calculate the aerodynamic forces and moments acting on 3.2 Review of Vector Relations
a flying vehicle, it is imperative to describe the flow pattern
around the vehicle. The flow pattern thus obtained depends on In aerodynamics, we come across many quantities which have
the geometry of the vehicle, its orientation with respect to the both magnitude and direction, such as force and velocity.
oncoming freestream, altitude, and flying speed. Often, these These are vector quantities, and as such, the mathematics of
patterns are very complex which necessitates to carry out both aerodynamics is most conveniently expressed in vector nota-
the experimental and theoretical studies to characterize the tion. In this section, we will set forth the basic relations impor-
resultant flow. Moreover, the theoretical studies may utilize tant for us from the vector algebra and the vector calculus.
simplifying approximations in order to obtain any solution at
all, and they must be matched with experimental data. There-
fore, we must understand the fundamental laws that govern 3.2.1 Scalar (or Dot) Product
the fluid flow phenomena so that we can relate the theoretical −
→ −
→
results with experimental observations. Consider two vectors A and B are at an angle α to each
other, as shown in Fig. 3.1. The scalar (dot) product of the
−
→ −
→
vectors A and B are given as
3.1 Introduction
−
→ − → − →
→ −
A · B = A B cos α (3.1)
Fluid mechanics is the branch of science which studies the
statics and dynamics of fluids, and thus, it is further classified − −
→ →
into fluid statics and fluid dynamics. As the name suggests, the where A and B , respectively, denotes the magnitudes of
fluid statics studies the fluid at rest, whereas the fluid dynamics −
→ −
→
vectors A and B , and α is smaller of the included angle.
investigates the effect of forces on the fluid in motion. To Note that the dot product of two vectors is a scalar. Further,
analyze a moving fluid, the fundamental laws such as the −
→ −
→
Eq. (3.1) shows that the projection of A on B is
conservation principles of mass, momentum, and energy are
−
used. Moreover, in arriving at the basic equations of motion →
= B cos α (3.2)
one can either define the motion of individual molecules or can
predict the gross behavior of molecules in a control volume.
−
→ − →
However, defining the motion of each and every molecule in Similarly, the projection of B on A is
the fluid is a cumbersome task, and usually not preferred as −
→
one is usually interested in knowing the average properties of = A cos α (3.3)
the flow in a selected control volume. The appropriate size of
the chosen control volume is very important, as the number −
→ − → − → −
→ −
→
of molecules in this elemental volume must be sufficiently Thus, A · B = A times projection of A on B (or) vice
large to have a meaningful statistical description of the flow. versa. Furthermore, dot product may be positive or negative
That is, the elemental control volume must be a continuous depending upon α. That is, for α < 90◦ , the dot product
media. Before we proceed further in actually deriving the is positive, whereas for α > 90◦ , the dot product is nega-
fundamental equations of fluid motion, let us review some tive. Moreover, dot product follows the commutative law, i.e.,
−
→ − → − → − →
elementary laws which are useful in our discussion. A · B = B · A

58 3 Governing Equations of Fluid Flows
α
O
Β
Fig. 3.1 Schematic diagram representing dot product
A
α
O
B
n
C
Fig. 3.2 Schematic diagram representing vector product
3.2.2 Vector (or Cross) Product cylindrical and spherical coordinate systems only. Thus, we
are interested in three most common orthogonal coordinate
−
→ −
→
The cross product of the vectors A and B is defined as system: Cartesian, cylindrical, and spherical, as schematically
(Fig. 3.2) shown in Fig. 3.3.
The Cartesian coordinate system is shown in Fig. 3.3a. The
−
→ − → − →
→ − −
→
A × B = A B sin α · n̂ = C (3.4) x, y, and z axes are mutually perpendicular, and î, ĵ, and k̂ are
unit vectors in the x-, y-, and z-directions, respectively. If
−
→
or A is a given vector in Cartesian space, it can be expressed
−
→
−
→ − → −→ as A = Ax î + Ay ĵ + Az k̂, where Ax , Ay and Az are scalar
A × B= C (3.5) −
→
components of A in x, y, and z-directions, respectively.
The cylindrical coordinate system is shown in Fig. 3.3b.
−
→ −
→ −
→
where C is perpendicular to the plane containing A and B , The location of point P in space is given by three coordinates
and n̂ is the unit vector a direction which obeys “right-hand (r, α, z), where r and α are measured in the xy plane. êr is the
−
→ −
→
rule”. That is, if we rotate A into B by curling our fingers, unit vector in the r direction, êα is the unit vector in the α direc-
−
→ −
→
then the right-hand thumb will point in the direction of C . tion, and ez is the unit vector in the z-direction. If A is a given
−
→
Also, vector in cylindrical space, then A = Ar êr + Aα êα + Az êz ,
−
→
− → − − where Ar , Aα and Az are scalar components of A in r-, α-,
→ − → → − →
C = A B sin α = A × B and z-directions, respectively.
The coordinates between Cartesian and cylindrical axes
where || represents magnitudes of the vectors. Moreover, vec- can be transformed by the relations,
tor product is noncommutative. That is,
x = r cos α
−
→ − → −
→ − →
A × B =− B × A y = r sin α
z=z
3.2.3 Orthogonal Coordinate Axes or inversely,

The mathematical description of a flow in three dimensions r= x 2 + y2
requires a three-dimensional coordinate system to describe y
the problem uniquely. Some aerodynamic problems best fit in α = arctan
x
a rectangular system, whereas others can be treated suitably in
z=z
3.2 Review of Vector Relations 59
y y
P (x, y, z)
r r
j α x
A
P (r,α, z) eα
x er
k O i
ez
z z
(a) Cartesian Coordinates (b) Cylindrical Coordinates
P (r, α, β)
z er A
eβ
eα
α r
x
(c) Spherical Coordinates
Fig. 3.3 Typical orthogonal coordinate system
The spherical coordinate system is shown in Fig. 3.3c. Here or inversely,

again, a general point P in space is given by three coordi-
nates (r, α, β), where r is the distance of P from the ori- r=x 2 + y 2 + z2
gin, α is the angle measured from the z-axis and is in the z
α = arccos
rz plane, and β is the angle measured from the x-axis and x + y 2 + z2
2
is in the xy plane. êr is the unit vector in the r direction, êα x
β = arccos
is the unit vector in the α-direction, and eβ is the unit vec- x 2 + y2
−
→
tor in the β-direction. If A is a given vector in spherical
−
→
space, then A = Ar êr + Aα êα + Aβ êβ , where Ar , Aα and
−
→
Aβ are scalar components of A in r-, α-, and z-directions, 3.2.4 Scalar and Vector Fields
respectively.
The coordinates between Cartesian and spherical axes can A scalar quantity given as a function of coordinate space and
be transformed by the relations, time is called scalar field. For example, pressure, density, and
temperature are scalar fields, given in Cartesian space as
follows:
x = r sin α cos β p = p (x, y, z, t) (3.6)
y = r sin α sin β ρ = ρ (x, y, z, t) (3.7)
z = r cos α T = T (x, y, z, t) (3.8)
Similarly, a vector quantity given as a function of coordinate 3.2.6 Vector Product of Two Vectors
space and time is called a vector field. For example, velocity
−
→ −
→
is a vector field Consider again the vectors A and B from Eqs. (3.13) and
(3.14),
−
→
v = vx î + vy ĵ + vz k̂ (3.9)
−
→
A = Ax î + Ay ĵ + Az k̂
where −
→
B = Bx î + By ĵ + Bz k̂
vx = vx (x, y, z, t) (3.10)
Their vector product is defined as
vy = vy (x, y, z, t) (3.11)
vz = vz (x, y, z, t) (3.12)
î ĵ k̂
−
→ − →
A × B = Ax Ay Az
Analogous expressions can be written for vector fields in Bx By Bz
cylindrical and spherical space.
Analogous expressions of vector products in cylindrical and
spherical coordinates can be defined by replacing the unit
3.2.5 Scalar Product of Two Vectors vectors and scalar components along the respective axes.
−
→ −
→
Consider two vectors A and B in Cartesian coordinate sys-
tem as 3.2.7 Gradient of a Scalar
−
→
A = Ax î + Ay ĵ + Az k̂ (3.13) Consider a scalar field such as pressure,
−
→
B = Bx î + By ĵ + Bz k̂ (3.14)
p = p1 (x, y, z) = p2 (r, α, z) = p3 (r, α, β) (3.16)
The scalar or dot product of these vectors is
The gradient of p at a given point (x, y) in space is defined as a
−
→ − → vector whenever its magnitude is the maximum rate of change
A · B = Ax Bx + Ay By + Az Bz (3.15) of p per unit length of the coordinate space at the given point
and its direction is that of the maximum rate of change of p at
Similar expressions for the scalar products of two vectors can the given point (Fig. 3.4). Consider ∇p at a given point and if
be obtained in cylindrical and spherical space. n̂ is the unit vector along any arbitrary direction, s. Then the
rate of change of p per unit length in the s direction will be
Direction of maximum
change in pressure at point (x, y)
y
p Gradient lines
(x, y)
Gradient lines
Isobars
Isobars
x
Fig. 3.4 Schematic representation of the gradient of a scalar
dp −
→
= ∇p.n̂ (3.17) It should be noted that in above equations, V can represents
ds any vector quantity. However, to get an insight on the physical
−
→
significance
− of the divergence, consider V to be flow velocity
where dp
ds is known as directional derivative in the s direction. →v . Let us further assume that a small fluid element (δV)
of fixed mass, is moving along a streamline with velocity −→
From Eq. (3.17), it is evident that gradient of a scalar field is
v.
simply the component of ∇p in that direction.
When the fluid mass moves through the space, its volume
For pressure field in Cartesian space, p = p (x, y, z), the −
→→
changes. It can be shown that ∇ .− v essentially denotes the
gradient of p (∇p) is given by the following expression.
time rate of change of the volume of a moving fluid element
∂p ∂p ∂p of fixed mass, per unit volume of that element. That is,
∇p = î + ĵ + k̂ (3.18)
∂x ∂y ∂z
−
→− 1 D (δ∀)
∇ .→
v = (3.24)
Similarly for pressure field in cylindrical coordinate, i.e., δ∀ Dt
p = p (r, α, z), we have
∂p 1 ∂p ∂p
∇p = êr + êα + êz (3.19)
∂r r ∂α ∂z 3.2.9 Curl of a Vector
Likewise for pressure field in spherical coordinate, i.e., Consider a vector field,
p = p (r, α, β), we get
−
→ − → −
→ −
→
∂p 1 ∂p 1 ∂p V = V (x, y, z) = V (r, α, z) = V (r, α, β) (3.25)
∇p = êr + êα + êβ (3.20)
∂r r ∂α r sin α ∂β −
→
Here again− for
convenience we will assume V to be the flow
→
velocity v . Likewise the divergence of a vector case, once
3.2.8 Divergence of a Vector again we visualize a fluid element moving along a streamline.
The fluid element may rotate with an angular velocity, −→ω , as
−
→ − → −
→
The divergence− (positive)
of a vector field V = V (x, y, z) is it translates along
the streamline.
It can be shown that ω is
→− → −
→ −
→ −→ − →
a scalar field ∇ . V , which measures how much V spreads half of curl of v ; ∇ × v . That is,
−
→
out at each point or for a negative divergence, how much V
1 −
→ −
−
→ ∇ ×→
converges to the point. That is, the divergence represents the
ω = v (3.26)
volume density of the outward flux of a vector field from 2
a small elemental volume around a given point. Mathemati-
cally, the divergence is the scalar product of the ∇ operator In Cartesian space, if −
→
v = vx î + vy ĵ + vz k̂, then
and the vector field on which it acts.
−
→ −
→− →
In Cartesian space, if V = Vx î + Vy k̂ + Vz k̂, then ∇ . V î ĵ k̂
−
→ −
∂ ∂ ∂
is ∇ ×→
v = ∂x (3.27)
∂y ∂z
vx vy vz
−
→−→ ∂ ∂ ∂
∇.V = Vx + Vy + Vz (3.21)
∂x ∂y ∂z
In cylindrical coordinate system, if − →
v = vr êr + vα êα + vz êz
−
→ then
In cylindrical coordinates, if V = Vr êr + Vα êα + Vz êz , then
−
→− →
∇ . V is êr r eˆα êz
−
→ −
∇ ×→
1
v = ∂r ∂ ∂ ∂
∂α ∂z (3.28)
−
→−→ 1 ∂ 1 ∂ ∂ r
∇.V = (rVr ) + Vα + Vz (3.22) vr rvα vz
r ∂r r ∂α ∂z
−
→ If −
→
v = vr êr + vα êα + vβ êβ is the velocity field in spherical
Similarly for spherical space, if V = Vr êr + Vα êα + Vβ êβ ,
−
→− → space then
then ∇ . V is

êr r eˆα (r sin α) êβ
−
→−→ 1 ∂ 2 1 ∂ 1 ∂ −
→ −
∇ ×→
∇.V = 2 rVr + (Vα sin α) + 1 ∂ ∂ ∂
r ∂r r sin α ∂α
V
r sin α ∂β β v = 2 ∂r ∂α ∂β (3.29)
r sin α
(3.23) vr r vα (r sin α) vβ
ds v
(a) n (b)
b
ds
v
C C
a
Fig. 3.5 Schematic diagram for line integrals
C Surface (S)
n dS
p
dS
Volume ( V )
Surface (S)
Fig. 3.6 Schematic diagram of an open and the closed surfaces
3.2.10 Line Integral clockwise direction, the area enclosed by the curve is always
on your left. Moreover, if −→v denotes velocity field then the
Consider the vector field line integral represents the flow circulation (). That is,
˛
−
→ − → −
→ −
→ −
→
V = V (x, y, z) = V (r, α, z) = V (r, α, β) (3.30) = − →
v · ds (3.33)
c
Also, let us consider an open curve C in space connecting
two points a and b as shown in Fig. 3.5a. If ds is an elemental
length of the curve, and n̂ be a unit vector tangent to the curve 3.2.11 Surface Integral
−
→ −
→
such that ds = n̂ds. Thus, the line integral of vector V along
the curve C from a to b is Consider an open surface s bounded by the closed curve C as
shown in Fig. 3.6a. At point P over an elemental area dS and n̂
ˆb be a unit vector normal to the surface pointing outward. Let us
−
→ − → −
→
V · ds (3.31) define a vector elemental area as dS = n̂dS. The surface inte-
a gral over the surface s can be defined in following three ways.
If p is scalar field, i.e., p = p (x, y, z) over S, we get
However, if the curve C is closed as shown in Fig. 3.5b, then ¨
the line integral will be −
→
pdS (3.34)
˛ b
−
→ − → S
V · ds (3.32)
a The outcome of the above integral will be a vector. However,
−
→ − →
The sign convention is represented as if we traverse along if V = V (x, y, z), i.e., a vector field over S, we have
¨
C in anticlockwise direction then it is considered as positive, −
→− →
otherwise negative. For a closed curve, if you move in counter- V .dS (3.35)
S
−→
Note that the above integration is essentially a scalar or dot vector field V are related through Gauss divergence theorem,
product of two vectors and hence, the result is a scalar. In ¨ ˚
addition, the vector or cross product may also be defined as −
→− → −
→− →
V .dS = ∇ . V d∀ (3.43)
¨
−
→ − → S ∀
V × dS (3.36)
S
3.2.15 Gradient Theorem
For the closed surface S which encloses the volume ∀, the
surface integrals over the closed surface are If p is the scalar field then by the gradient theorem,
"
−
→ ¨ ˚
= pdS (3.37) −
→
pdS = (∇p) d∀ (3.44)
S
S ∀
"
−
→− →
= V .dS (3.38)
S 3.3 Eulerian and Lagrangian Viewpoints
"
−
→ − → The continuum hypothesis allows us to surmise that at any
= A × dS (3.39) instant t, there exists a fluid particle corresponding to every
S point in space occupied by the fluid. Suppose a property such
as temperature T of the fluid is to be specified, it could be
done in two different ways. In the first approach, the prop-
3.2.12 Volume Integral erty is specified as a function of the position in space and
time, i.e., T = T (x, y, z, t). This is termed as Eulerian or
Consider a volume ∀ in space. If p is the scalar field in this field description, which essentially specifies the temperature
space, the volume integral over ∀ is of that fluid particle which happens to be at the location
˚ (x, y, z) at the given time t = t1 . At time t = t2 , the tem-
= pd∀ (3.40) perature T (x, y, z, t2 ) is the temperature, not of the same par-
∀ ticle, but of a different particle—the one that happens to be at
location (x, y, z) at t = t2 . Clearly, the Eulerian method does
−
→ not elucidate the temperature–time history of any specific
For the vector field V , in space, the volume integral over the
−
→ particle.
volume ∀ of vector V is
The other approach of specifying a property in a mov-
˚
−
→ ing fluid consists of identifying the fluid particles with some
= V d∀ (3.41)
labels, following them around, and specifying their proper-
∀ ties as a function of time. Usually, the particles are labeled by
the space point they occupied at some initial time t0 . Thus,
T (x0 , y0 , z0 , t) refers to the temperature at time t of a particle
3.2.13 Stokes Curl Theorem which was at location (x0 , y0 , z0 ) at t0 . This approach of iden-
tifying material points and following them along is termed as
Consider again the open surface S bounded by the closed
−
→ the Lagrangian or the particle or the material description.
curve C, shown in Fig. 3.6a. Let V be a vector field. The
Stokes theorem states that outflow of vorticity from the sur-
face is equal to the line integral of velocity. That is, 3.3.1 Local and Material Derivatives
¨ ˛
−→ − → −→ −
→ − →
∇ × V · dS = V · ds (3.42) Consider a small fluid element moving through a flow field,
s as shown in Fig. 3.7. Suppose the velocity field in Cartesian
c
space is given by
3.2.14 Gauss Divergence Theorem −

→
v = uî + vĵ + wk̂ (3.45)
Consider again the volume ∀ bounded by the closed surface where u = u (x, y, z, t), v = v (x, y, z, t), and w = w (x, y ,
S, shown in Fig. 3.6b. The surface and volume integrals of the z, t) are scalar components of velocity along x-, y- and z-
Point ’1’ Point ’2’
Fluid element at
Fluid element at
t = t2
t = t1
Fig. 3.7 A moving fluid element in the three-dimensional space
directions, respectively. Moreover, the density field is given Thus, we write

by ρ = ρ (x, y, z, t). Let the fluid element is positioned at
point “1” at time t = t1 , where the fluid density is Dρ ∂ρ ∂ρ ∂ρ ∂ρ
=u +v +w + (3.55)
Dt ∂x ∂y ∂z ∂t
ρ1 = ρ (x1 , y1 , z1 , t1 ) (3.46)
where DρDt is the material or Lagrangian derivative of density,
If the element is moved to point “2” at time t = t2 , where the ∂ρ
and ∂t is the local derivative of density. In vector notations,
fluid density becomes Eq. (3.55) becomes
ρ2 = ρ (x2 , y2 , z2 , t2 ) Dρ ∂ρ − →
−
+ →
(3.47)
= v .∇ ρ (3.56)
Dt ∂t
Expanding ρ = ρ (x, y, z, t) using Taylor series about point
−
→
“1” as follows: where −→v . ∇ denotes the convective derivative. In general,
Eq. (3.56) is written as
∂ρ ∂ρ ∂ρ
ρ2 = ρ1 + 1 x + 1 y + 1 z
∂x ∂y ∂z D () ∂ () − →
−
= + →
v . ∇ () (3.57)
∂ρ Dt ∂t
+ 1 t + higher order terms (3.48)
∂t
Dividing above series by t throughout and neglecting higher

order terms, we have
3.4 Primary and Auxiliary Laws for
ρ2 − ρ1 ∂ρ x ∂ρ y ∂ρ z ∂ρ Continuous Media
= 1 + 1 + 1 + 1
t ∂x t ∂y t ∂z t ∂t
(3.49) From the engineering applications point of view, four basic or
primary laws must be satisfied for a continuous media. They
Taking limit t → 0 both sides, we get are as follows:
ρ2 − ρ1
lim 1. Conservation of mass (continuity equation),
t→0 t
2. Newton’s second law (momentum equation),
∂ρ x ∂ρ y ∂ρ z ∂ρ
= lim 1 + 1 + 1 + 1 3. Conservation of energy (first law of thermodynamics),
t→0 ∂x t ∂y t ∂z t ∂t
and
(3.50) 4. Increase of entropy principle (second law of thermody-
namics).
But
In addition to these primary laws, there are numerous
Dρ ρ2 − ρ1
= lim (3.51) secondary (or auxiliary) laws, frequently called constitutive
Dt t→0 t
relations, that apply to particular type of medium (or flow
x processes). The equation of state for perfect gas, Newton’s
u = lim (3.52)
t→0 t law of viscosity, etc., are some common examples of
y subsidiary laws. Moreover, the Hooke’s law for elastic
v = lim (3.53)
t→0 t solids is also an auxiliary law, which is widely used in solid
z mechanics.
w = lim (3.54)
t→0 t
3.5 Flow Analysis Techniques 65
3.5 Flow Analysis Techniques surface. The control volume may be fixed in space with fluid
crossing it across its control surface. Alternatively, the control
The fundamental laws applicable to fluids are essentially the volume may be moving with the fluid but always composed of
same as those in solid mechanics. Newton’s second law of same set of fluid particles. In either case, the control volume is
motion, in fact, applicable to all the substances irrespective sufficiently large and have a finite space of the fluid flow. The
of their physical state, representing as it does, an equality fundamental principle is applied to the fluid mass inside the
between the rate of change of momentum and the forces acting control volume and the fluid mass crossing the control surface.
on a body. Since the rate of change of momentum envisaged Thus, one should see the ease provided by the control volume
in Newton’s second law refers to a specific body of matter, approach; instead of analyzing the complete flow domain, we
therefore, for fluids, the rate of change of momentum alludes can focus on the fluid in this finite region only.
to the substantial rate of change and not the local one. Simi-
larly, the principle of conservation of mass also applies to both
solids and fluids, though in solid mechanics it is rarely used 3.5.3 Infinitesimal Fluid Element Approach
unless the body is disintegrated or different parts are coalesc-
ing to form a composite body during motion or deformation, Another way of modeling the fluid flow is to choose an
whereas the conservation of mass is an essential analytical infinitesimal fluid element of differential volume, d∀. Even
tool in fluid flow analysis. But in contrast to a solid, the fun- though the size of fluid element is very small but it is large
damental laws are not directly applicable to a fluid due to its enough to contain a sufficient number of fluid molecules so
“squishy” behavior, that is, a body of matter in the fluid is that it can be viewed as a continuous medium. The fluid ele-
not readily identifiable. Therefore, we must have an appropri- ment may be fixed in space with fluid crossing its boundary.
ate technique which aids in treating the fluid as a free body. Alternatively, it may be moving with the fluid with velocity
Generally, there are four different fluid modeling techniques −
→v equal to the flow velocity at each point. Again, instead
which are efficiently used throughout the modern evolution of of looking at the whole flow field at once, the fundamental
aerodynamics. They are reviewed in the following sections. principles are applied to just the fluid element itself.
3.5.1 Finite Control Mass Approach 3.5.4 Microscopic Approach

From the above discussion, it is evident, at least in principle, In reality, the motion of a fluid is a ramification of the mean
that even with fluids, we may select a body of matter, con- motion of its atoms and molecules. Therefore, the fourth
sidering it as a free body identifying all the external forces approach could be a microscopic model wherein the funda-
acting on it; writing the Newton’s second law along with all mental laws are applied directly to the atoms and molecules,
the constitutive relationships and solving the problem, as is using suitable statistical averaging to define the resulting fluid
done in the solid mechanics. It is termed as the control mass properties. This approach comes under the purview of kinetic
or system approach. A control mass or system may change theory, which has its own merits but requires huge computa-
in size, shape, or location in space, but it must always com- tional resources and sometimes unnecessary for flow analysis.
posed of the same material. For example, in the treatment of Thus, it is beyond the scope of our discussion.
cardiovascular disease, counterpulsation of a gas-filled bal-
loon is used; a balloon inflates in the aorta during diastole
to improve diastolic pressure and deflates during systole to 3.6 Integral and Differential Analysis
reduce left ventricular afterload. One may consider the entire
mass of the gas in the balloon as a control mass. The loca- The fundamental laws of fluid motion are as pertinent to indi-
tion of the bounding surface, shape as well as volume of this vidual particles as to a group of particles. That is, the laws
control mass1 is changed with time. obtained for finite control volumes are equally applicable
when the control volume dwindles to a point in the flow field.
The method of analysis, where large control volumes are used
3.5.2 Finite Control Volume Approach
to obtain aggregate forces or transfer rates, is termed as inte-
gral analysis. On the other hand, when the analysis is applied
n this approach, a closed volume is drawn within a finite region
to individual points in the flow field, the resulting equations
of the flow. This volume is termed as control volume (C∀)
are differential equations and the method is termed as differ-
which is bounded by a closed surface (CS) known as control
ential analysis. Both these approaches are used in aerodynam-
ics, depending upon the requirement of the study. If we are
1 Note the similarity between the concept of control mass in fluid mechan-
interested in knowing the detailed description of flow field as
ics with the concept of closed system in thermodynamics.
we would if we are designing the wing section of a super- 1. Variation in duct cross section is not too excessive.
sonic aircraft, we would resort to differential analysis. If only 2. Curvature of streamlines is not too excessive.
overall forces are required, as, for example, while calculating 3. No appreciable change in velocity profile along the duct.
the power requirements of a compressor, the simpler integral
analysis suffices.
The foundations of aerodynamics are built on three conser- 3.8 The Continuity Equation
vation laws of matter: mass, momentum, and energy. These
laws can be stated in differential or integral forms, applica- The principle of conservation of mass states that in a closed
ble to an extended region. In integral form, the expression of system the mass of the system (m) remains invariant over the
these laws depend on whether they are related to a control period of time. That is,
volume fixed in space, or to a material volume, composed of
dm
same fluid particles and whose control surface moves with the =0 (3.58)
fluid. dt
Consider a stationary control volume (C∀) of dimensions dx,

dy and dz such that d∀ = dxdydz with the origin located at
3.7 One-, Two-, and Three-Dimensional
the center of the volume, as shown in Fig. 3.8. Thus, if we
Flows
move a distance dx 2 in the positive x-direction then at this
location the yz plane is designated as x+ face, while covering
In every analysis, a hypothetical substance is assumed which
the same distance in the negative x-direction we find the yz
lends itself to mathematical treatment while yielding results
plane termed as x− face. Similarly, the face located at a dis-
of practical value. In this regard, continuum hypothesis is
tance dy2 from the origin in the positive y-direction, is called
already discussed. At this stage, we will introduce simpli-
fied flows which when used with caution, will allow the use y+ face, and the face at distance dy 2 from the origin in negative
of highly developed theory on the problems of engineering y-direction is known as y face. Likewise, z+ and z− faces
−
interest. are located at a distance dz 2 from the origin in the positive-

In reality, all properties and flow characteristics are and negative z-directions, respectively. Let us assume that the
usually expressed as a function of three space coordinates fluid crosses this control volume with a velocity − →
v (x, y, z, t)
and time. That is, in general, fluid flows are of three- having the velocity components u (x, y, z, t), v (x, y, z, t) and
dimensional in nature. For example, the velocity field is w (x, y, z, t), respectively, in x-, y- and z-directions in Carte-
expressed in as a function of location (x, y, z) and time t, sian coordinate system. That is,
i.e., −
→
v =− →v (x, y, z, t). −
→
A two-dimensional flow is distinguished by the condi- v (x, y, z, t) = uî + vĵ + wk̂
tion that all properties and flow characteristics are func-
tions of two space coordinates and time, and hence do not From the conservation of mass principle, we can write
change along the third space coordinate direction. For exam-
ple, when the velocity field is expressed, in Cartesian coor- Rate of mass accumulation within the control volume
dinates, as −→v =− →v (x, y, t), independent of z-direction at a + Rate of mass leaving the control volume
given instant. All the planes normal to z-axis, will at a given − Rate of mass entering the control volume = 0
instant, have the same streamline pattern. The flow past an
airfoil of infinite aspect ratio is a theoretical example of the The first term in the above relation can be expressed as
flow in two dimensions.
A one-dimensional flow is a further simplification, where ∂ (ρdxdydz)
= (3.59)
all properties and flow characteristics are assumed to be ∂t
expressible as function of one space coordinate and time.
The position is usually the location along the flow direc- Moreover, using the Taylor’s series expansion the second and
tion. The flow across a normal shock wave and the flow in third terms gives the net rate of mass flow in x-direction is
a convergent–divergent duct are the mathematical examples
∂ (ρu) dx ∂ (ρu) dx
of one-dimensional flows. However, in reality, the flow in = ρu + dydz − ρu − dydz
a convergent–divergent duct can never be one-dimensional ∂x 2 ∂x 2
as the flow velocity varies over the duct cross section. Nev- ∂ (ρu)
= dxdydz (3.60)
ertheless, if the deviation is not too large, one-dimensional ∂x
flow may be assumed to exist. For instance, in ducts, one-
Similarly, the net rate of mass flow in y-direction is
dimensional assumption is often valid where
An airfoil theoretically said to have an infinite aspect ratio
Aspect ratio = wing span ^2 /Planform area

3.8 The Continuity Equation 67
y+ _
y z
+
_
x
x
x
dy
z+
_
z y dz
dx
Fig. 3.8 Infinitesimal control volume in a fluid flow
∂ (ρv) dy ∂ (ρv) dy where ∇ is called del or nabla operator such that

= ρv + dxdz − ρv − dxdz −
→
∂x 2 ∂x 2 ∇ . () = ∂() ∂() ∂()
∂x + ∂y + ∂z . From Eq. (3.57), we can write
∂ (ρv) Eq. (3.65) as
= dxdydz (3.61)
∂x
Dρ −
→ →
Likewise, the net rate of mass flow in z-direction is + ρ ∇ .−
v =0 (3.66)
Dt
∂ (ρw) dz ∂ (ρw) dz The generalized differential equations given by Eqs. (3.65)
= ρw + dxdy − ρw − dxdy
∂x 2 ∂x 2 and (3.66) are called the conservation and nonconservation
∂ (ρw) forms of the continuity equation. In Eq. (3.65), the terms are
= dxdydz (3.62)
∂x briefly defined as follows:
Thus, the overall rate of mass flow passing through the control ∂ρ
= rate of change of mass inside C∀
volume will be ∂t
→ −
−
∂ (ρu) ∂ (ρv) ∂ (ρw) ∇ . ρ→
v = difference in mass leaving and entering C∀
= + + dxdydz (3.63)
∂x ∂y ∂z
Putting all the terms together we obtain (Table 3.1)

3.8.1 Some Important Observations
∂ρ ∂ (ρu) ∂ (ρv) ∂ (ρw)
(dxdydz) + + + dxdydz = 0
∂t ∂x ∂y ∂z The conservation of mass principle or the continuity equation
is one of the most important laws of fluid mechanics. Some
Dividing by dxdydz throughout important facts about this equation, worth to remember, are
discussed below.
∂ρ ∂ (ρu) ∂ (ρv) ∂ (ρw)
+ + + =0 (3.64)
∂t ∂x ∂y ∂z 1. For a compressible barotropic fluid, the density is not a
constant and is the function of pressure, ρ = f (p). How-
In vector notation ever, at steady flow conditions, we have ∂ρ
∂t = 0. In turn,
∂ρ − → → Eq. (3.65) leads to
+ ∇ . ρ−
v =0 (3.65)
∂t
Table 3.1 The mass flow rate leaving and entering into a control volume
Directions Rate of mass leaving C∀ Rate of mass entering C∀

x-direction ρu + ∂(ρu) dx
∂x 2 dydz ρu − ∂(ρu) dx
∂x 2 dydz

y-direction ρv + ∂(ρv) dy
∂x 2 dxdz ρv − ∂(ρv) dy
∂x 2 dxdz

z-direction ρw + ∂(ρw) dz
∂x 2 dxdy ρw − ∂(ρw) dz
∂x 2 dxdy
→ −
−
∇ . ρ→
v =0 (3.67) to each other, but differ conceptually for unsteady flows. A
schematic diagram of streamlines, streaklines, and pathlines
In Cartesian coordinates, we can write above equation as in an unsteady flow are shown in Fig. 3.9.
∂ (ρu) ∂ (ρv) ∂ (ρw)

+ + =0 (3.68) 3.9.1 Timelines
∂x ∂y ∂z Real lines
2. For an incompressible fluid, since the density remains con- The line joining a set of fluid particles at a given instant is
stant (ρ = constant) then, ∂ρ∂t = 0. Substituting this into known as timeline, shown schematically in Fig. 3.9a. Obser-
Eq. (3.65), we get vations of a timeline at different time instants provide the
information of the flow field. They are particularly useful in
−
→−
∇ .→
v =0 (3.69) checking the uniformity of the flow.
In Cartesian space, we have

3.9.2 Streamlines Imaginary lines
∂u ∂v ∂w
+ + =0 (3.70)
OF COURSE ∂x ∂y ∂z A streamline is an imaginary line in a fluid flow drawn
in such a manner that it is everywhere tangent to the
Interestingly, Eq. (3.70) is valid for both steady and velocity vector at the particular instant in time at which
unsteady flows. the observation is made. Let the local velocity vector be
3. Note that Eq. (3.65) can also be expressed in cylindrical −
→v (x, y, z, t) = uî+ vĵ + wk̂. Consider
(r, θ, z) and spherical (r, θ, φ) coordinate systems. We can −
→ an infinitesimal arc-
length vector, ds = dxî + dyĵ + dzk̂ , along a streamline.
write the continuity equation in cylindrical (r, θ, z) coor-
−
→
dinate space as (The derivations of these equations are left Since ds is parallel to −→v , we have
as an exercise to the reader),
−
→ −
ds × →
v =0
∂ρ 1 ∂ (ρrvr ) 1 ∂ (ρvθ ) ∂ (ρvz )
+ + + =0 (3.71)
∂t r ∂r r ∂θ ∂z In determinant form
Also, in spherical (r, θ, φ) coordinates
î ĵ k̂

dx dy dz = 0
∂ρ 1 ∂ ρr 2 vr 1 ∂ (ρvθ sin θ) 1 ∂ ρvφ
+ 2 + + =0 u v w
∂t r ∂r r sin θ ∂θ r sin θ ∂φ
(3.72)
or
(wdy − vdz) î + (udz − wdx) ĵ + (vdx − udy) k̂ = 0
3.9 Graphical Representation of Fluid Above relation will be valid only when each term is separately
Flows equal to zero. Thus,
The quantitative and qualitative information of fluid flow can (wdy − vdz) = 0
be obtained through flow visualization, graphical represen- (udz − wdx) = 0
tation, and mathematical analysis. However, the visual rep-
resentation of flow fields is an important tool in modeling (vdx − udy) = 0
the flow phenomena. In general, there are four basic types
or
of line patterns used to visualize the flow, namely, timelines,
pathlines, streaklines, and streamlines. dx dy dz
A timeline is a set of fluid particles that form a line at = = (3.73)
u v w
a given instant. Timelines are instantaneous lines and can
easily be generated experimentally. Streamlines, streaklines, On solving Eq. (3.73), one can obtain the equation of stream-
and pathlines are the loci, defined by the vector fields in a line. To evaluate the constants of integration it is sufficient
fluid flow. These are the important concepts for visualizing or to specify some known point (x0 , y0 , z0 ) through which the
describing the flow fields. In steady flows, they are equivalent streamline passes. Furthermore, for the streamlines in two
3.9 Graphical Representation of Fluid Flows 69
At time
t=0
At time
v t = t 1 t = t 2 t = t3
(a) Timelines
Instantaneous streamlines
Streaklines at successive
time intervals
P
Pathline of fluid element at Q
v Instantaneous velocity vector
(b) Streamlines, streaklines and pathlines
Fig. 3.9 Graphical representation for fluid flow
dimensions, dz = 0 and w = 0, thus Eq. (3.73) reduces to at all points along its length, and there can be no flow through
the lateral surface of the streamtube. Consequently, the mass
dx dy flow rate remains constant at each cross section of the stream-
= (3.74)
u v tube. In short, the streamtube acts like a nonporous surface of
zero wall thickness and infinitesimal cross section. Assume
In polar coordinate system, if dr and rdθ, respectively, are the
that these cross sections are small enough so that the fluid
infinitesimal lengths along r- and θ-directions, then Eq. (3.74)
velocities across each cross section are constant. For steady
can be represented as
flow, if the fluid velocities at stations “1” and “2” are v1 and
dr rdθ v2 , respectively, then
= (3.75)
u v
ρ1 A1 v1 = ρ2 A2 v2 (3.76)
where u and v are scalar velocity components along r- and θ-
directions. Here, one should note that no two streamlines can where ρ1 and ρ2 are fluid densities at stations “1” and “2”,
intersect each other, because at the point of intersection two respectively; and A1 and A2 are the cross section areas at
different tangents can be drawn each on these streamlines, stations “1” and “2”, respectively. In addition, if the fluid is
showing two different velocities at the same point, which is incompressible (ρ1 = ρ2 ), then
impossible.
A1 v1 = A2 v2 (3.77)
3.9.2.1 Streamtube
A streamtube is a tubular region of fluid surrounded by stream-
That is, as the flow velocity increases, the cross-sectional area
lines, as shown in Fig. 3.10. Since the streamlines do not inter-
of the streamtube decreases. Also, when a large number of
sect each other, same streamlines pass through a streamtube
Can a streamline intersect itself?

v2
Set of streamlines
A2
A1
v1
Fig. 3.10 Streamlines forming a streamtube
streamtubes are arranged adjacent to each other such that the at some earlier time. Consider a continuous injection of the
arrangement forms a finite cross section, the arrangement of dye at a fixed point P in space, shown in Fig. 3.9. The con-
streamtubes is known as bundle of streamtubes. nection of all the elements passing through P over a period of
time is called the streakline. Thus, a streakline is depicted by
the movement of all the fluid elements along its length. This
3.9.3 Pathlines Real lines is in contrast to a pathline, where the movement of only one
fluid element in time is focused. Consequently, the trajectory
A pathline may be defined as a line in the flow field describing equations for a pathline are applied to all the fluid particles
the trajectory of a given fluid particle. It is essentially the elucidating the streakline.
path traversed by the fluid particle through the coordinate
space over a period of time. Figure 3.9 shows the pathline of a
fluid element R at Q (x, y, z, t), which earlier passed through
3.10 Angular Velocity, Vorticity, and the
the point P (x0 , y0 , z0 , t0 ). This path is fully described by the
Shear Strain Rate
three position functions xR (t), yR (t), zR (t), which can be
calculated by integrating the three components of velocity In this section, we will examine the change of shape and ori-
fields u (x, y, z, t), v (x, y, z, t), w (x, y, z, t) along the path. entation of a fluid element as it moves along a streamline.
The integration is carried out between the time t0 to time t, Consider an element which is rectangular in shape when the
where t0 shows the initial time at which the fluid particle was motion has just started. If the flow velocity significantly varies
at point P. across the extent of the fluid element the corners may not
ˆ t move in unison, and thus the element may rotate and in addi-
xR (t) = x0 + u (xR (t) , yR (t) , zR (t) , t) dt (3.78) tion, its shape may become distorted. Indeed, the edges of the
t0 element undergo some combination of tilting and stretching.
ˆ t
However, in our discussion, we will consider only the tilting
yR (t) = y0 + v (xR (t) , yR (t) , zR (t) , t) dt (3.79)
t0 motion because of its greater implications in aerodynamics.
ˆ t If the adjacent sides of the fluid element rotate equally
zR (t) = z0 + w (xR (t) , yR (t) , zR (t) , t) dt (3.80) and in the same direction, we have pure rotation; however,
t0
if the adjacent sides rotate equally but in opposite directions,
it leads to pure shearing motion. The absence of rotational
3.9.4 Streaklines Imaginary lines motion eventually simplifies the governing equations of fluid
motion to a large extent. Additionally, the shearing action
A streakline may be defined as the instantaneous line of all the between the adjacent elements together with the fluid viscos-
fluid elements that have passed through the point of injection ity produces shearing stresses, which are responsible for the
3.10 Angular Velocity, Vorticity, and the Shear Strain Rate 71
Fig. 3.11 Schematic diagram of a fluid element under different flow conditions
drag and separation phenomena. The quantum of distortion nent over the top surface of the fluid element can be written
and rotation are subject to the velocity field, our task is to as u (y0 + dy) = u (y0 ) + ∂u∂y dy. Similarly, the y-component
examine this dependency. If there is a velocity gradient in the of velocity on the left and right sides of the fluid element are
direction of the fluid other than the direction of motion, shear v (x0 ) and v (x0 + dx), respectively. Again, for small dx the
strain, and rotation occurs. velocity on the right face of the fluid element using Taylor’s
Consider the fluid element in a velocity field with the veloc- series will be v (x0 ) + ∂v
∂x dx. After the time dt, the velocity
ity gradients in x- and y-directions, as shown in Fig. 3.11a. The gradients will cause the fluid element to deform and rotate.
x-component of velocity at the bottom and top surfaces of Note that the amount of translation, deformation, and rotation
the fluid element are u (y0 ) and u (y0 + dy), respectively. For that occurs during the time dt has been enlarged in Fig. 3.11b
small dy, using Taylor’s series expansion the velocity compo- for clarity. However, in reality, the fluid element will appear
to change almost infinitesimal during the small time dt. The ∂u

ωOB = (3.88)
deformation and rotation of the fluid element during the time ∂y
dt will travel along with the fluid element and superimposes
the changes that occur, shown in Fig. 3.11c. The angular velocity about z-axis may be considered as the
The bottom side of the fluid element OA rotates to the average of ωOA and ωOB . That is,
small angle dα due to the gradient of the y-component of the
1
velocity in x-direction, and the left side of the fluid element ωz = (ωOA + ωOB )
2
OB rotates to the small angle dβ due to the gradient of x-
component of the velocity in the y-direction. The distance or
moved by the point A relative to the point O is the difference 1 ∂v ∂u
ωz = − (3.89)
in the velocity between the right and the left sides. That is, 2 ∂x ∂y
∂v Following the similar procedure as above, we can obtain the

dx dt (3.81) average angular velocity about x-axis as
∂x
Again, the distance moved by the point B relative to the point 1 ∂w ∂v

ωx = − (3.90)
O is the difference in velocity between the top and the bottom 2 ∂y ∂z
sides, which is given as
Likewise, the average angular velocity about the y-axis is
∂u
dy dt (3.82) 1 ∂u ∂w
∂y ωy = − (3.91)
2 ∂z ∂x
Let −
→
ω be the angular velocity and considering the counter-
clockwise direction as positive, the angular velocity of side Thus, the angular velocity vector −
→
ω = ωx î + ωy ĵ + ωz k̂
OA is becomes
dα
ωOA = (3.83)
dt −
→ 1 ∂w ∂v ∂u ∂w ∂v ∂u
ω = − î + − ĵ + − k̂
2 ∂y ∂z ∂z ∂x ∂x ∂y
Since time dt is very small, angle dα will also be very small. (3.92)
Therefore,
In vector notations, the angular velocity is
∂v
∂x dx dt ∂v
dα ≈ tan (dα) = = dt (3.84) 1 −
→ −
dx ∂x −
→
ω = ∇ ×→
v (3.93)
2
Introducing Eq. (3.84) in Eq. (3.83), we get
In aerodynamic applications,
− another useful parameter to
→
∂v work with is the vorticity ζ , defined as the twice of the
ωOA = (3.85)
∂x angular velocity. That is,
The angular velocity of side OB is equal to the negative time −
→ ∂w ∂v ∂u ∂w ∂v ∂u
rate of change of angle dβ, ζ = 2−
→
ω = − î + − ĵ + − k̂
∂y ∂z ∂z ∂x ∂x ∂y
dβ (3.94)
ωOB = − (3.86)
dt
In vector notations, the vorticity becomes
The negative sign in Eq. (3.86) is due to the fact that the side
−
→ − → →
OB rotates in clockwise direction, which is the negative direc- ζ = ∇ ×−
v (3.95)
tion for the angular velocity. Furthermore, for small dβ
i.e., in a velocity field the curl of velocity is equal to the
∂u vorticity. Equation (3.95), leads to two important conclusions.
∂y dy dt ∂u −
→ →
dβ ≈ tan (dβ) = = dt (3.87) If ∇ × − v = 0 at every point in the flow field, it is called
dy ∂y
the rotational flow. This is the case of finite vorticity and
−
→ →
From Eqs. (3.87) and (3.86), we have angular velocity. However, if ∇ × − v = 0 the flow field is
3.10 Angular Velocity, Vorticity, and the Shear Strain Rate 73
termed as an irrotational flow. In this case, the flow field has From the above derivations, one can see that the angular veloc-
zero vorticity everywhere and thus the fluid elements have no ity, the vorticity, and the shear strain rate depend exclusively
−
→
angular velocity, i.e., ζ = 2−→
ω = 0. Consequently, the fluid on the velocity gradients of the flow field. These velocity gra-
elements in an irrotational flow have only the translational dients can be represented in matrix form as follows:
motion. If we equate Eq. (3.92) to zero, we get ⎡ ⎤
∂u ∂u ∂u
∂x ∂y ∂z
∂w ∂v ⎢ ∂v ∂v ∂v ⎥
= (3.96) =⎣ ∂x ∂y ∂z ⎦ (3.106)
∂y ∂z ∂w ∂w ∂w
∂x ∂y ∂z
∂u ∂w
= (3.97)
∂z ∂x
Note that the sum of diagonal terms ∂u + ∂v
+ ∂w
in
∂v ∂u ∂x ∂y ∂z
= (3.98) −
→− →
∂x ∂y Eq. (3.106) is equal to ∇ . v , which from Eq. (3.24) essen-
tially denotes the stretching of a fluid element. In addition, all
i.e., all the three equations above must be satisfied simulta- other off-diagonal terms which appear as the cross derivatives
neously to have an irrotational flow field. Moreover, it can in angular velocity, vorticity, and shear strain rate relations are
be seen that the velocity gradients are still present in an irro- associated with tilting motion of a fluid element.
tational flow field which may cause deformation in the fluid
elements. For an irrotational and inviscid flow, it is possible
to predict the velocity field using potential flow theory. 3.11 The Navier–Stokes Equation
Furthermore, the velocity gradient in a flow can also be
related to the rate of shear strain of the fluid element. For this, In this section, we derive the conservation of momentum equa-
the two-dimensional shear strain can be assumed as the aver- tions commonly referred to as the Navier–Stokes equations.
age change in the angle between two lines which are initially They are obtained in Eulerian frame of reference and used
perpendicular (such as the lines OA and OB in Fig. 3.11c). to solve fluid mechanics and aerodynamics problems. In the
Therefore, the shear strain during the time dt is beginning, let us consider the Newton’s second law of motion
in Lagrangian frame of reference.
1
= (dα + dβ) (3.99) −
→
2 F = m−
→
a (3.107)
Hence, the time rate of change of shear strain in xy plane −
→
becomes It is a vector equation, where F is the sum of forces acting on
1 dα dβ the system and − →a is the acceleration of the system. Newton’s
εxy = + (3.100)
2 dt dt second law can also be expressed as
But from Eqs. (3.84) and (3.87) −

→ d −
F = m→
v (3.108)
dt fixed mass system
dα ∂v
= (3.101)
dt ∂x The equation above is applied to the systems which have their
dβ ∂u mass fixed. Our approach is to relate the equation above in
= (3.102)
dt ∂y Lagrangian frame of reference to fluid motion occurring in
Eulerian reference frame. In component form Eq. (3.108) can
Introducing the equations above in Eq. (3.100), we obtain
be rewritten as
1 dv du d
εxy = + (3.103) Fx = (mu) (3.109)
2 dx dy dt fixed mass system
Similarly, the shear strain rates εyz and εxz , respectively, in d

Fy = (mv) (3.110)
yz and xz planes are dt fixed mass system
d
1 ∂w ∂v Fz = (mw) (3.111)
εyz = + (3.104) dt fixed mass system
2 ∂y ∂z
1 ∂u ∂w where Fx , Fy and Fz are the forces in x-, y- and z-directions,
εxz = + (3.105)
2 ∂z ∂x respectively. Thus,
−
→
F = Fx î + Fy ĵ + Fz k̂ (3.112) Using Taylor series expansion, we can evaluate the equations
above, however, assuming small C∀, we retain the terms up
Consider again the stationary control volume (C∀) of dimen- to first order only neglecting higher order terms. For x+ face,
sions dx, dy and dz such that d∀ = dxdydz with the origin the rate of momentum
passing through the face out of C∀ is
located at the center of the volume, as shown in Fig. 3.8. Thus, ρuu + ∂(ρuu)
∂x 2
dx
dydz, while the rate of momentum passing
if we move a distance dx
2 in the positive x-direction then at this
location the yz plane is designated as x+ face, while covering through the face into C∀ is ρuu − ∂(ρuu) dx
∂x 2 dydz. Similarly,
the same distance in the negative x-direction we find the yz the rate of momentum coming out of C∀, and entering into
plane termed as x− face. Similarly, the face located at a dis- C∀ are ρuv + ∂(ρuv) ∂y 2
dy
dxdz and ρuv − ∂(ρuv) dy
∂y 2 dxdz,
tance dy2 from the origin in the positive y-direction, is called respectively. Likewise, the rate of momentum leaving C∀
y+ face, and the face at distance dy
2 from the origin in negative
through z+ face and entering into C∀ through z− face,
y-direction is known as y− face. Likewise, z+ and z− faces are ρuw + ∂(ρuw) ∂z
dz
2 dxdy and ρuw − ∂(ρuw)∂z
dz
2 dxdy,
are located at a distance dz 2 from the origin in the positive- respectively.
and negative z-directions, respectively. Let us assume that Now, putting all the terms together in the relation,
the fluid crosses this control volume with the velocity vector
−
→v (x, y, z, t) in Cartesian coordinate system. That is, Rate of increase of momentum (mu) in C∀ + Rate of mu leaving
C∀ − Rate of mu entering C∀ = Sum of forces applied to C∀ in
x-direction
−
→
v (x, y, z, t) = uî + vĵ + wk̂ (3.113)
We obtain,

where u is the component of velocity in x-direction with î ∂ (ρu) ∂ (ρuu) dx ∂ (ρuu) dx
dxdydz + ρuu + dydz − ρuu − dydz
unit vector, v is the component of velocity in y-direction with ∂t ∂x 2 ∂x 2

∂ (ρuv) dy ∂ (ρuv) dy
ĵ unit vector, and w is the velocity component in z-direction + ρuv + dxdz − ρuv − dxdz
∂y 2 ∂y 2
with k̂ unit vector.
∂ (ρuw) dz ∂ (ρuw) dz
+ ρuw + dxdy − ρuw − dxdy = Fx
In Eulerian frame of reference, for stationary control vol- ∂z 2 ∂z 2
ume, the balance of momentum can be expressed as
On simplifying,
Rate of increase of momentum m− →v in C∀ + Rate of m−
→v leaving
C∀ − Rate of m− →v entering C∀ = Sum of forces applied to C∀
∂ (ρu) ∂ (ρuu) ∂ (ρuv) ∂ (ρuw)
dxdydz + + + dxdydz = Fx
Furthermore, in component form the momentum balance in ∂t ∂x ∂y ∂z
(3.120)
x-, y- and z-directions can be written as
Rate of increase of momentum (mu) in C∀ + Rate of mu leaving Dividing by dxdydz = d∀ throughout
C∀ − Rate of mu entering C∀ = Sum of forces applied to C∀ in

x-direction ∂ (ρu) ∂ (ρuu) ∂ (ρuv) ∂ (ρuw) Fx
Rate of increase of momentum (mv) in C∀ + Rate of mv leaving + + + =
∂t ∂x ∂y ∂z d∀
C∀ − Rate of mv entering C∀ = Sum of forces applied to C∀ in
y-direction
(3.121)
Rate of increase of momentum (mw) in C∀ + Rate of mw leaving
C∀ − Rate of mw entering C∀ = Sum of forces applied to C∀ in But from chain rule of the calculus,
z-direction
∂ (ρuu) ∂ (ρu) ∂u
=u + ρu (3.122)
No we focus on x-momentum carried through all the six faces ∂x ∂x ∂x
of the control volume. Let us first examine a two-dimensional
face of C∀ across which x-momentum transfer is occurring, and
as shown in Fig. 3.12. From the figure, we can write the com- ∂ (ρu) ∂u ∂ρ
ponents of momentum crossing C∀ as follows: =ρ +u (3.123)
∂t ∂t ∂t
m̊x+ ux+ = (ρux+ dydz) ux+ = (ρux+ ) (ux+ dydz) (3.114) Thus, rearranging the terms in Eq. (3.121)
m̊x− ux− = (ρux− dydz) ux− = (ρux− ) (ux− dydz) (3.115)
∂ρ ∂ (ρu) ∂ (ρv) ∂ (ρw)
m̊y+ uy+ = ρvy+ dydz uy+ = ρuy+ vy+ dydz (3.116) u + + +
∂t ∂x ∂y ∂z
m̊y− uy− = ρvy− dydz uy− = ρuy− vy− dydz (3.117)
∂u ∂u ∂u ∂u Fx
m̊z+ uz+ = (ρwz+ dydz) uz+ = (ρuz+ ) (wz+ dydz) (3.118) +ρ +u +v +w = (3.124)
∂t ∂x ∂y ∂z d∀
m̊z− uz− = (ρwz− dydz) uz− = (ρuz− ) (wz− dydz) (3.119)
3.11 The Navier–Stokes Equation 75
m y+ u y+
mx− u x− x m x+ u x+
y− u y− m
Fig. 3.12 x-momentum transfer across the two-dimensional face
Note that the first term in the equation above denotes the weight of C∀, i.e., ρgdxdydz will be the body force. Thus, the
conservation of mass, thus component of body force in x-direction is,

∂ρ ∂ (ρu) ∂ (ρv) ∂ (ρw) Fx
+ + + =0 b
= ρgx (3.128)
∂t ∂x ∂y ∂z d∀
Hence, Eq. (3.124) reduces to The other forces are surface forces. Pressure is an important
surface force acting on C∀. We will examine the pressure
∂u ∂u ∂u ∂u Fx force acting in x-direction. Consider the pressure forces acting
ρ +u +v +w = (3.125)
∂t ∂x ∂y ∂z d∀ on x+ and x− faces, shown in Fig. 3.13. Note that the direc-
tion of pressure is toward the face as they are compressive
From Eqs. (3.57), (3.125) can also be expressed as in nature. Thus, the net pressure force acting in
x-direction is,
Du Fx
ρ = (3.126)
Dt d∀ ∂p
(px− Ax− ) − (px+ Ax+ ) = − dxdydz
∂x
Dt = ax , i.e., component of acceleration in x-direction.
But Du
Similarly, ay and az are acceleration components in y- and or
z-directions, respectively. Thus, the total acceleration can be (px− Ax− ) − (px+ Ax+ ) ∂p
expressed as =− (3.129)
d∀ ∂x
−
→
a = ax î + a y ĵ + az k̂ (3.127) Thus, net pressure force per unit volume in x-direction is equal
to the gradient along x-direction. Basically, the stresses that
where act on the surface when multiplied by the surface area given
surface forces. Stresses can be normal or shear stresses. In a
Du ∂u ∂u ∂u ∂u
ax = = +u +v +w given coordinate direction, there will be a net normal force
Dt ∂t ∂x ∂y ∂z on the two opposite faces and a net shear force on each of the
Dv ∂v ∂v ∂v ∂v other two sets of opposing forces. Stresses are usually denoted
ay = = +u +v +w
Dt ∂t ∂x ∂y ∂z using two subscripts, the first subscript represents the normal
Dw ∂w ∂w ∂w ∂w direction of the face upon which they act, and the second one
az = = +u +v +w
Dt ∂t ∂x ∂y ∂z denotes the direction of the stress. In addition, the sign con-
vention adopted for the stresses are as follows. The value of
Now let us evaluate the forces per unit volume acting on the stress is positive in the positive coordinate direction for the
control volume. In general, there are two kinds of forces act face with outward facing normal in the positive coordinate
on C∀; body forces and surface forces. By virtue of mass, the direction. They are again positive if the stress acts in the neg-
[ p − (dp/dx) (dx/2)] dydz [ p + (dp/dx) (dx/2)] dydz

x
Fig. 3.13 Pressure force acting in x-direction on the control volume
ative coordinate direction on the face with outward normal in Rearranging the terms,
negative coordinate directions.
From Eq. (3.129), it can be inferred that the net normal and ∂u ∂u ∂u ∂u ∂p
ρ +u +v +w = ρgx −
shear stresses in a particular coordinate direction are essen- ∂t ∂x ∂y ∂z ∂x
tially the gradient of those stresses in that direction. Thus, ∂ ∂u ∂v ∂w
+ μ + +
Eq. (3.125) becomes ∂x ∂x ∂y ∂z
∂ ∂u ∂ ∂u ∂ ∂u
∂u ∂u ∂u ∂u ∂σxx ∂τxy ∂τxz + μ + μ + μ
ρ +u +v +w = ρgx + + + ∂x ∂x ∂y ∂y ∂z ∂z
∂t ∂x ∂y ∂z ∂x ∂y ∂z

(3.130)
But ∂u ∂v ∂w
∂x + ∂y + ∂z = 0 is the continuity equation for an
incompressible fluid. Thus, we have
Similarly, following the same steps as carried out for the
x-momentum equation, we can obtain y- and z-momentum
∂u ∂u ∂u ∂u ∂p
equations as ρ +u +v +w = ρgx −
∂t ∂x ∂y ∂z ∂x
∂ ∂u ∂ ∂u ∂ ∂u
∂v ∂v ∂v ∂v ∂τyx ∂σyy ∂τyz + μ + μ + μ
ρ +u +v +w = ρgy + + + ∂x ∂x ∂y ∂y ∂z ∂z
∂t ∂x ∂y ∂z ∂x ∂y ∂z
(3.131)
With constant viscosity, x-component of momentum equation
∂w ∂w ∂w ∂w ∂τzx ∂τzy ∂σzz becomes
ρ +u +v +w = ρgz + + +
∂t ∂x ∂y ∂z ∂x ∂y ∂z
(3.132) ∂u ∂u ∂u ∂u
ρ +u +v +w
∂t ∂x ∂y ∂z
All we need to do is define the stresses/strain relationships for
∂p ∂2u ∂2u ∂2u
the fluids. Recall, the stress–strain relationship for a Newto- = ρgx − +μ + 2+ 2 (3.136)
nian fluid given as ∂x ∂x2 ∂y ∂z
Similarly, y-component of momentum equation is

∂u ∂u ∂v
σxx = −p + 2μ ; τxy = τyx = μ + (3.133)
∂x ∂y ∂x ∂v ∂v ∂v ∂v
∂v ∂v ∂w ρ +u +v +w
σyy = −p + 2μ ; τyz = τzy = μ + (3.134) ∂t ∂x ∂y ∂z
∂y ∂z ∂y ∂p ∂2v ∂2v ∂2v
∂w ∂w ∂u = ρgy − +μ + 2+ 2 (3.137)
σzz = −p + 2μ ; τzx = τxz = μ + ∂y ∂x2 ∂y ∂z
∂z ∂x ∂z
(3.135) Likewise, z-component of momentum equation is
Introducing Eq. (3.133) into Eq. (3.130) and simplifying, ∂w ∂w ∂w ∂w

ρ +u +v +w
∂t ∂x ∂y ∂z
∂u ∂u ∂u ∂u ∂p
ρ +u +v +w = ρgx − ∂p ∂2w ∂2w ∂2w
∂t ∂x ∂y ∂z ∂x = ρgz − +μ + + (3.138)
∂ ∂u ∂ ∂u ∂v ∂ ∂w ∂u ∂z ∂x2 ∂y2 ∂z2
+ 2μ + μ + + μ +
∂x ∂x ∂y ∂y ∂x ∂z ∂x ∂z
Equations (3.136)–(3.138) are the conservation forms of well ∂w ∂w ∂w ∂w ∂p

ρ +u +v +w = ρgz − (3.144)
known Navier–Stokes equation in Cartesian space. Note that ∂t ∂x ∂y ∂z ∂z
the three momentum equations and the continuity equa-
tion have total four unknowns (u, v, w, p) that need to be The above equations are the conventional Euler equations for
determined. The acceleration term or the rate of change of inviscid fluid flows in Cartesian coordinate system. In vector
momentum in the volume makes the Navier–Stokes equa- notation, we can rewrite these equations in their nonconser-
tions nonlinear. The viscous term essentially makes them a vation form as
D−
→
second-order partial differential equation. Furthermore, these
v −
→
equations can be represented in vector notations as ρ = ρ−
→
g − ∇p (3.145)
Dt
∂−
→ − → → −
→
+ −
→
v .∇ −v = ρ−
→
v
ρ g − ∇ p + μ∇ 2 −
→
v (3.139) Mathematically, Eqs. (3.145) and (3.65) can be solved to
∂t obtain the velocity field and the pressure in the fluid flow.
However, in reality, Eq. (3.145) can be regarded as the equa-
or
tion for velocity only as the pressure does not appear explicitly
−
→
∂−
→v −
−
→ → −
→ −
→ ∇p in the continuity equation (Eq. (3.65)). Therefore, for majority
+ v .∇ v = g − + ν∇ 2 −
→
v (3.140) of cases, the solution of Euler equations can only be achieved
∂t ρ
using computational techniques.
where ν = μρ is the kinematic viscosity of the fluid. Fur-
thermore, in terms of material or substantial derivative,
Eq. (3.140) can be rewritten as
3.11.2 Velocity–Vorticity Form of the
Navier–Stokes Equation
−
→
D−
→v −
→ ∇p
= g − + ν∇ 2 −
→
v (3.141) The Navier–Stokes equation, given by Eq. (3.140), can also
Dt ρ be represented in terms of vorticity. Neglecting the body force
term, rewriting Eq. (3.140), we get
Equations (3.140) and (3.141) are called the conservation and
nonconservation forms of the Navier–Stokes equation. We −
→
∂−
→v − → → ∇p
will now briefly describe the different terms in the Navier– + −
→
v .∇ −v =− + ν∇ 2 −
→
v (3.146)
Stokes equation. ∂t ρ
− → →
• − →v .∇ −
Taking curl of the above equation, we have
v refers the inertial or nonlinear term which
characterizes the Navier–Stokes equation. This term is
→ ∂−
− →v −
→ − → → −
→ − → −
→
+ ∇ ×ρ −
→
v .∇ −v = −ρ ∇ × ∇ p + μ ∇ × ∇ 2 −
→
responsible for transferring the turbulent kinetic energy ρ∇ × v
in −
the cascade. ∂t
→ (3.147)
• − ∇ρ p is the pressure gradient term which ascertains the
incompressibility of the fluid. or
• ν∇ 2 −→v is the viscous dissipation term emerged due to −
→
∂ζ − → −
→ −
→− →→ −
→ − → −
→
Reynolds stresses of the Newtonian fluid. It is proportional ρ +ρ −
→ v = −ρ ∇ × ∇ p + μ ∇ × ∇ 2 −
v .∇ ζ + ρ ζ .∇ − →
v
to the viscosity which have a major role in laminar flow ∂t
(3.148)
regime.
→ −
− → →
where ζ = ∇ × − v , is the vorticity. We know that the
3.11.1 The Euler Equation curl of a gradient is identically zero and thus substituting
−
→ − →
∇ × ∇ p = 0 into the above equation, we get
In some application areas of aerodynamics, it is reasonable to
−
→ −
neglect the viscous stresses. Thus, ignoring the viscous terms ∂ζ → −
→ −
→− →→ −
→
in Eqs. (3.136), (3.137) and (3.138), we get ρ +ρ −
→ v = μ ∇ × ∇2−
v .∇ ζ + ρ ζ .∇ − →
v
∂t
(3.149)
∂u ∂u ∂u ∂u ∂p
ρ +u +v +w = ρgx − (3.142) Since the curl and the Laplacian operator
∂t ∂x ∂y ∂z ∂x −
→ −
→ − commute with each
−
→ →
other, i.e., ∇ × ∇ v = ∇ ∇ × v . Thus, rewriting the
2 2
∂v ∂v ∂v ∂v ∂p
ρ +u +v +w = ρgy − (3.143) above equation
∂t ∂x ∂y ∂z ∂y
Physically, this term accounts for the way fluid elements carry their momentum as they move,
impacting neighboring elements and transferring momentum across the flow.
The energy cascade refers to the process through which turbulent kinetic energy is transferred
from larger scales to smaller scales.
The cascade is a fundamental feature of turbulence, and this term drives this transfer process by
continuously breaking up larger structures into smaller, more chaotic ones.
Streamlines
ζ
v
Fig. 3.14 A two-dimensional flow field in xy plane
−
→ −
∂ζ → −
→ −
→− →→ −
→ of these, which is nothing but the special form of the equa-
ρ +ρ −
→
v .∇ ζ + ρ ζ .∇ −
v = μ∇ 2 ζ (3.150) tion of motion. Consider a two-dimensional incompressible
∂t
and inviscid flow under conservative body forces in the xy
This is the velocity–vorticity form of the Navier–Stokes equa- plane, as shown in Fig. 3.14. For the steady flow conditions,
tion. The significance of various terms in this equation can be the Crocco’s theorem states that
described as follows:
−
→
• ρ ∂∂tζ is the unsteady term. 1
− → −→
Velocity
×
Vorticity
=
Gradient of
• ρ − → vector vector Density stagnation pressure
v . ∇ ζ is the convection term (i.e., the velocity con-
(3.151)
vects the gradient of vorticity).
−
→− →→
• ρ ζ .∇ − v is the vortex stretching term. What is a vortex? In mathematical notations,
−
→
• μ∇ 2 ζ is the diffusion term, which shows that the viscosity −
→ 1−→
−
→v × ζ = ∇ p0 (3.152)
diffuses away the Laplacian of the vorticity. ρ
From Eq. (3.150), we have two key conclusions: If pb is the pressure component due to body forces, then the
−→ stagnation pressure is defined as
1. This equation shows that if there is no vorticity ζ = 0
in the beginning (at time t = 0), then no vorticity will 1
p0 = p + ρv2 + pb (3.153)
be created later. Note that this result is in contrast to the 2
Navier–Stokes given by Eq. (3.140), where the initial zero
For a two-dimensional
− flow as illustrated in Fig. 3.14, the vor-
velocity does not necessary imply a zero velocity later. In →
fact, the flow field can have the velocity later because of ticity vector ζ is normal to the xy plane and velocity vec-
→
the pressure. tor − v is tangential to the streamline.
− From Eq. (3.152), the
→
2. If the flow field is two-dimensional, then no vorticity will gradient of stagnation pressure ∇ p0 is normal to both − →v
be created in the flow. That is, a two-dimensional flow may −
→ −
→
be rotating about some point or it may be diffusing but still and ζ , and thus ∇ p0 lies in the xy plane. According to
no vorticity will be created. How? the Crocco’s theorem, stagnation pressure is constant along
each streamline, whereas between the streamline p0 is con-
stant only if the vorticity content is zero. We can understand
3.11.3 The Crocco’s Theorem it further from an example. Consider the boundary layer flow
past a solid surface. In the boundary layer, although the static
In addition to the velocity–vorticity form of the Navier–Stokes pressure is constant everywhere but the velocity is variable
equation, there exist several other theorems that relate vortic- and so the stagnation pressure. Also, the presence of strong
ity to the dynamics of fluid flows. Crocco’s theorem is one velocity gradient near to the wall causes the strong stagnation
Elemental Surface (dS)
P Curve (C)
Fig. 3.15 An elemental surface ds bounded by a curve C
pressure gradient which reveals the presence of vorticity in integrated over any surface bounded −by that curve.
Also, it can
→ →
the wall zone. be seen that if the curl of velocity ∇ × − v = 0 is vanished
everywhere over any surface bounded by C, then circulation
will be zero ( = 0). If the curve C becomes very small such
3.12 Rotational Flows and the Circulation that the circulation around this curve is given as d, then
Eq. (3.155) yields
The circulation
→ () is defined as the line integral of a velocity
−
→ → − → − → → −
field − v around a closed curve. The term circulation was d = ∇ × − v . ds = ∇ × −
→
v .n̂ ds
first used independently by Frederick Lanchester, Wilhelm
Kutta, and Nikolai Joukowski. To define the circulation, let us or −
→ −
∇ ×→
take a closed curve C in a given flow field, as illustrated in d
−
→ v .n̂ = (3.156)
Fig. 3.15. Let ds be the directed line segment around a point ds
P on C. Thus, ˛ where ds is the infinitesimal area enclosed by the infinitesimal
−
→
= − →v . ds (3.154) curve C.
c Figure 3.16 shows that at a point P in the flow the compo-
nent of vorticity normal to ds is equal to the circulation per
i.e., circulation is the line integral of velocity around a closed unit area, where the circulation is taken around the boundary
curve in the flow. It is a kinematic property depending only of ds. Thus, the vorticity can alternatively be defined as the
on the velocity field and the choice of the curve C. Mathe- circulation per unit area.
matically, if the line integral is carried out in anticlockwise
direction along the curve C, then is considered positive.
ζ= (3.157)
The circulation is also related to vorticity. Consider the s
figure as illustrated in Fig. 3.15 which shows an open surface
bounded by the closed curve C is in a flow field. Let the
velocity at any point P on the surface (including any point on
the curve C) is − →
v . Using the Stokes curl theorem (Eq. (3.42)),
3.13 Irrotational Flows and the Potential
we write
˛ ¨ − Function
−
→ −
→ − →
= − →v . ds = ∇ ×→ v . ds (3.155)
c s
We have seen earlier that a flow is called irrotational when the
vorticity is zero at every point in the flow field. Thus, from
where ds represents the elemental surface. Equation (3.155) Eq. (3.94)
−
→ − → →
shows that the circulation about a curve is equal to the vorticity ζ = ∇ ×− v =0 (3.158)
v
n
− dΓ
P ds
I don't what is a stream
C function and a potential
function to define them!!
Fig. 3.16 An elemental area (ds) enclosed by an infinitesimal curve C
−
→
where − →
v is the velocity field and ζ is the vorticity. If φ is a vθ =
1 ∂φ
(3.169)
scalar function, then we consider the following vector identity r ∂θ
1 ∂φ
→ −
− → v = (3.170)
∇ × ∇φ =0 (3.159) r sin θ ∂
The velocity potential is analogous to the stream function in

i.e., the curl of the gradient of a scalar function is identically the sense that the derivatives of φ also yield the velocity com-
zero. Comparing Eqs. (3.158) and (3.159), we get ponents. However, there exist some well-defined differences
between φ and ψ. They are as follows:
−
→ −
→
v = ∇φ (3.160)
1. The velocity components are obtained by differentiating φ
Equation (3.160) states that for an irrotational flow, there in the same direction as that of the velocities, whereas ψ
exists a scalar function φ such that the velocity is given by (or ) is differentiated normal to the velocity direction.
the gradient of φ. We denote φ as the velocity potential, 2. φ is defined only for an irrotational flow, but ψ is valid for
which is a function of spatial coordinates; φ = φ (x, y, z) or both rotational and irrotational flows.
φ = φ (r, θ, z) or φ = φ (r, θ, ). 3. φ can be used in both two- and three-dimensional flows,
In Cartesian (x, y, z) system, from Eq. (3.160) we can write whereas ψ is applicable for two-dimensional flows only.
∂φ ∂φ ∂φ An irrotational flow field which allows to define a velocity

uî + vĵ + wk̂ = î + ĵ + k̂ (3.161) potential (φ) is called potential flow. Due to existence of φ the
∂x ∂y ∂z
analysis of an irrotational flow is simplified to a great extent.
Thus, the velocity components are For example, to obtain three unknown velocity components
(say u, v and w in Cartesian space) in a flow field, we must
∂φ have three different equations. However, if φ is known for the
u= (3.162)
∂x given problem, then u, v and w can be obtained by simply
∂φ taking the gradients in x-, y- and z-directions, respectively.
v= (3.163)
∂y
∂φ
w= (3.164) 3.14 Stream Function and the Concept of
∂z
Vector Potential
Similarly, in cylindrical (r, θ, z) coordinates the velocity com-
ponents are In a two-dimensional incompressible flow field, the equation
−
→→
of continuity reduces to ∇ .− v = 0, where − →v is the veloc-
∂φ ity vector field. Whenever the flow field is termed as two-
vr = (3.165)
∂r dimensional, it reflects that the velocity component in the third
1 ∂φ direction is identically zero. That is, the velocity components
vθ = (3.166)
r ∂θ in other two directions do not depend on the coordinate in that
∂φ third direction. If the velocity filed is expressed in Cartesian
vz = (3.167)
∂z coordinates as − →v =− →v (x, y, t), then the third direction is the
−
→
Likewise, in spherical (r, θ, ) space z-direction, so that vz = 0, and ∂∂zv = 0. It is sometimes bene-
ficial to define a scalar field termed as stream function denoted
∂φ by the symbol ψ, also known as Lagrange stream function. It
vr = (3.168)
∂r
3.14 Stream Function and the Concept of Vector Potential 81
dl
C = C [x(s), y (s)] dy
dx
ψ = c1
n
C ψ= c2
ψ
v
Fig. 3.17 Schematic diagram of the fluid flow between two streamlines
provides the flexibility to solve the governing equation with define this vector field in terms of stream function as
one variable ψ along with the associated boundary conditions
−
→
for velocity and pressure fields. This is clearly an advantage V (x, y, t) = ψ (x, y, t) k̂ (3.174)
over solving the full nonlinear Navier–Stokes and continuity
equations. Thus, once the stream function is obtained then it is But the velocity vector field is given by the relation
possible to calculate velocity components and pressure distri-
−
→ −
→ − →
bution over the body in a fluid flow. The stream function is also v = ∇ × V (3.175)
used to plot the streamlines, which represent the trajectories
of particles in a steady flow. However, along with advantages, Thus,
a penalty is also associated with this simplified approach. The
−
→ −
→
solution procedure becomes tedious as the governing equa- v = ∇ × ψ (x, y, t) k̂ (3.176)
tion is a fourth-order partial differential equation satisfied by
the stream function and thus it requires four boundary condi- Writing in terms of Cartesian coordinate system, we have
tions to solve. On the other hand, the Navier–Stokes equation
is a partial differential equation of second order, and hence, it î ĵ
k̂
requires two boundary conditions to solve. −
→ ∂ ∂ ∂
v = uî + vĵ = ∂x ∂ y ∂z
Consider the streamlines in two dimensions, as shown in
0 0 ψ (x, y, t)
Fig. 3.17. Let the equation of these streamlines is y = f (x, t)
dy
and thus, the slope at a point on the streamline is dx . If Simplifying above,
u (x, y, t) and v (x, y, t), respectively, are the velocity com-
ponents of the velocity field − →v in x- and y-directions. Thus, ∂ψ ∂ψ
uî + vĵ = î − ĵ
∂y ∂x
dy v
= (3.171)
dx u Thus,
Further, if the velocity components u and v are the known ∂ψ
functions of x, y, and t then integrating the equation above, u= (3.177)
∂y
we get
∂ψ
f (x, y, t) = c (3.172) v=− (3.178)
∂x
where c is an arbitrary constant of integration. Usually, the Introducing above velocity components into the continuity
function f (x, y, t) is written as ψ (x, y, t), thus equation for incompressible flows,
ψ (x, y, t) = c (3.173) −
→− ∂u ∂v ∂2ψ ∂2ψ
∇ .→
v = + = − =0 (3.179)
∂x ∂y ∂x∂y ∂x∂y
We will now deduce the relationship between the velocity
components and the stream function. Consider the vector field i.e., Lagrange stream function identically satisfies the incom-
−
→ − →
in a two-dimensional flow as, V = V (x, y, t). We can also pressible flow condition. Note that from mathematical point of
NOTE
view, the incompressibility condition (Eq. (3.69)) is the nec- or

essary and sufficient condition for the existence of a stream
r̂ θ̂ k̂
function. Furthermore, the dot product of velocity vector and
gradient of stream function leads to vr r̂ + vθ θ̂ = ∂r
∂ 1 ∂
r ∂θ
∂
∂z

(3.184)
0 0 ψ (r, θ, t)
−
→ −
→ ∂ψ ∂ψ
v .∇ ψ = u +v
∂x ∂y Thus,
or 1 ∂ψ
vr = (3.185)
r ∂θ
−
→ −
→ ∂ψ ∂ψ ∂ψ ∂ψ ∂ψ
v .∇ ψ = − =0 (3.180) vθ = −
∂y ∂x ∂x ∂y ∂r
(3.186)
i.e., the velocity field and the gradient of stream function are Note that the stream function (ψ) is a point function and thus
orthogonal to each other
(Fig. 3.17).
Consider again a para- it can be represented as exact differential. That is,
metric curve t (s) = x (s) , y (s) for the constant ψ, as illus-
trated in Fig. 3.17, so that ∂ψ ∂ψ
dψ = dx + dy (3.187)
∂x ∂y
∂ψ ∂ψ dx ∂ψ dy
=0⇒ +
∂s ∂x ds ∂y ds Substituting Eqs. (3.177) and (3.178) into Eq. (3.187), we get
∂ψ ∂ψ
But ∂x = −v and ∂y = u. Thus, dψ = udy − vdx (3.188)
dx dy But, from Eq. (3.74), we have

−v +u =0
ds ds
dx dy
= ⇒ udy − vdx = 0
In vector notations, u v
−
→ Therefore,
−
→
v×
d t
=0
ds
What do we want to convey? dψ = 0 (3.189)
−
→
where is the vector tangent to the parametric curve t (s),
d t
ds This is an important conclusion, which shows that along a
where ψ remains constant. Clearly, this imaginary line along
streamline, ψ remains invariant.
which the stream function has a constant value is essentially
the streamline. What is meant by invariant?
Similar to Cartesian coordinate system, we now define
the velocity components in cylindrical coordinates. Con-
3.14.1 Concept of Streamline in Three
sider again the two-dimensional flow field, as illustrated in
Dimensions
Fig. 3.17. Suppose the velocity field is given as −
→
v (r, θ) with
The two-dimensional concept of streamline can be extended
vz = 0. That is,
to
− three-dimensional flows as well. Consider a vector potential
→
−
→
v (r, θ) = vr r̂ + vθ θ̂ (3.181) ψ in Cartesian space, which satisfies the incompressibility
condition given by Eq. (3.69), as
If the vector field is given as
−
→
ψ = ψx î + ψy ĵ + ψz k̂ (3.190)
−
→
V = ψ (r, θ, t) k̂ (3.182)
where ψx , ψy and ψz are the components of vector potential in
Then the velocity field will be, x-, y- and
z-directions, respectively.
We know that the velocity
vector − →v = uî + vĵ + wk̂ is given by
−
→ −
→ − → − →
v = ∇ × V = ∇ × ψ (r, θ, t) k̂ (3.183)
−
→ → −
− →
v = ∇ × ψ
3.14 Stream Function and the Concept of Vector Potential 83
or In components form,

î k̂ 1 ∂
ĵ
vr = −
−
→ ∂ ∂ ∂ (3.194)
v = ∂x ∂y ∂z (3.191) r ∂z
1 ∂
ψx ψy ψz vz = (3.195)
r ∂r
Substituting above velocity components into the continuity

3.14.2 Axisymmetric Flows equation for incompressible flows
So far we have considered the flows occurring
−
→− 1 ∂ ∂vz 1 ∂2 1 ∂2
in the two-dimensional plane. The velocity field ∇ .→
v = (rvr ) + =− + =0
−
→ −
→ r ∂r ∂z r ∂r∂z r ∂z∂r
v = ∇ × ψ (x, y, t) k̂ was assumed to be invari- (3.196)
ant along z-axis with the z-component of velocity, i.e.,
vz = w = 0. However, the flows in which flow properties Clearly, the incompressible flow condition is again automat-
remains invariant in azimuth or θ-direction (in cylindrical ically satisfied. Moreover, Stokes stream function () has
polar coordinates), are termed as axisymmetric flows. A flow the properties similar to Lagrange stream function (ψ) and
in a circular pipe or a flow past a sphere are the examples of it is also constant along the direction of the flow. Also, the
flows with axial symmetry. Thus, for an axisymmetric flow volume flux or volume flow rate between two streamtubes
the velocity vector is given by with = inner and = outer will be, 2π (outer − inner ).
That is, the volume flux between two axisymmetric surfaces
−
→
v (r, z) = vr (r, z) r̂ + vz (r, z) k̂ (3.192) is again directly proportional to the difference of their stream
functions.
where vr (r, z) and vz (r, z) are the velocity components in
radial and cylindrical or longitudinal axis directions, respec- 3.14.3 Physical Interpretation of Lagrange
tively, and the velocity component in azimuth direction is Stream Function
zero, i.e., vθ = 0. Thus, the axisymmetric flows in cylindrical
coordinates have only two nonzero velocity components with From Eq. (3.173), a streamline is described by setting the
two effective coordinate directions. stream function equal to a constant, i.e., c1 , c2 , c3 , etc. Two
For axisymmetric incompressible flows, the vector field is different streamlines given by ψ = c1 and ψ = c2 , respec-
defined as tively, are illustrated in Fig. 3.17. We know that the vol-
−
→ 1 ume flow rateor the volume flux through an arbitrary curve
V = (r, z, t) θ̂ (3.193) C x (s) , y (s) with s ∈ [0, 1] which connects both stream-
r
lines, is
Note that we have used a different Greek letter for Stokes ˆ1
stream function to distinguish it from Lagrange stream func- q= −
→
v .n̂ds (3.197)
tion (ψ). This stream function is named in the honor of George
0
Gabriel Stokes. In fluid dynamics, Stokes stream function is
used to describe the streamlines and flow velocity in a three-
Consider a small element of length dl =
dxî + dyĵ on C and
dimensional incompressible flow with axisymmetry. A sur- −
→
face with constant encloses a streamtube everywhere tan- since the velocity vector v = uî + vĵ is parallel to dl, thus
gential to the flow velocity vector. The volume flow within
this streamtube is constant and all the streamlines are located dy dx
n̂ds = dyî − dxĵ = î − ĵ ds (3.198)
on its surface. Also, the velocity field associated with is ds ds
solenoidal, i.e., it has zero divergence.
Furthermore, the velocity field associated with axisymmet- The volume flux becomes
ric flow is given by
ˆ1 dy dx
−
→ q= uî + vĵ . î − ĵ ds (3.199)
−
→
v (r, z) = ∇ ×
1
(r, z, t) θ̂ ds ds
r 0
Substituting the velocity components u and v in terms of df ∂u ∂v

= +i (3.206)
stream function dx ∂x ∂x
ˆ1 and
∂ψ ∂ψ dy dx
q= î − ĵ . î − ĵ ds (3.200)
∂y ∂x ds ds
0 df ∂u ∂v
= +i (3.207)
dy ∂y ∂y
or
df df
ˆ1 ˆ1 Substituting dx and dy into Eq. (3.205), we get
∂ψ dy ∂ψ dx ∂ψ
q= + ds = ds (3.201)
∂y ds ∂x ds ∂s
0 0 df 1 ∂u ∂v ∂u ∂v
= +i −i +i (3.208)
or dz 2 ∂x ∂x ∂y ∂y
1 ∂u ∂v ∂v ∂u
= + +i − (3.209)
ˆψ2 2 ∂x ∂y ∂x ∂y
q= dψ = ψ2 − ψ1 (3.202)
df
ψ1 However, in complex or Argand plane dy will be zero along
df
the x-axis (real axis) and will be zero along the y-axis
dx
Thus, the volume flow rate between the two streamlines is (imaginary axis). Therefore,
equal to the difference of their stream functions. It can be
inferred that when the streamlines are close to each other, the df 1 ∂u ∂v
= +i (3.210)
velocity of the flow between them will be higher. dz 2 ∂x ∂x
and
3.15 The Cauchy–Riemann Equations
df 1 ∂v ∂u
= −i (3.211)
In the analysis of complex numbers, the Cauchy–Riemann dz 2 ∂y ∂y
equations form a necessary and sufficient condition for a com-
plex function to be complex differentiable. They are named If f (x, y) is the complex differentiable function, then the value
after the famous mathematicians Augustin Cauchy and Bern- of the derivative must be same for a given dz, irrespective of
df
hard Riemann. The purpose of this section is to review the its inclination. Hence, for the same dz we have
mathematical foundation behind these equations and their
application in the aerodynamics. Consider a complex func- 1 ∂u ∂v 1 ∂v ∂u
+i = −i (3.212)
tion f (x, y) given as 2 ∂x ∂x 2 ∂y ∂y
∂u ∂v ∂v ∂u
⇒ − +i + =0 (3.213)
f (x, y) = u (x, y) + iv (x, y) (3.203) ∂x ∂y ∂x ∂y
where z = x + iy; x, y ∈ R. Thus, We know that a complex number will be zero if and only if
both its real and imaginary parts are equal to zero. Thus,
dz = dx + idy (3.204) ∂u ∂v
= (3.214)
∂x ∂y
Using chain rule, f (x, y) can be written as ∂u ∂v
=− (3.215)
∂y ∂x
df ∂f ∂x ∂f ∂y
= +
dz ∂x ∂z ∂y ∂z In aerodynamics, many large scale steady and incompressible
1 ∂f ∂f flows such as the flow past an aircraft can be efficiently ana-
= −i (3.205) lyzed by using the functions of complex variables. If these
2 ∂x ∂y
flows are inviscid as well then they behave similar to the ideal
The partial differentiation of f (x, y) with respect to x and y fluids. Moreover, from the continuity equation of incompress-
will be, ible flows (Eq. (3.70)), we have
3.15 The Cauchy–Riemann Equations 85
−
→− ∂u ∂v
∇ .→
v = + =0 precisely, let us consider a two-dimensional, incompressible,
∂x ∂y and irrotational flow in Cartesian coordinates.
We know that along a streamline, ψ (x, y) = constant, and
Since the divergence of a curl is always zero, we can have a hence dψ = 0. Further, ψ is an exact differential, and thus,
function ψ (x, y), called stream function that obeys we write
∂ψ ∂ψ ∂ψ
u= (3.216) dψ = dx + dy = 0 (3.223)
∂y ∂x ∂y
∂ψ
v=− (3.217) or
∂x
Note that u and v satisfies the incompressibility condition, dψ = −vdx + udy = 0 (3.224)
given in Eq. (3.70). Since the function ψ (x, y) is orthogonal
to the gradient of ψ (x, y), they are parallel to the velocity Thus, we obtain
vector. Consequently, ψ (x, y) coincides with the streamlines
of the flow. dy v
=− (3.225)
In addition, if there is no vorticity in the flow field, i.e., dx ψ = constant u
−
→ −
∇ ×→ v = 0, the flow is said to be potential flow. From vec-
tor calculus, it is known that curl of a gradient is identically Similarly, along a potential line, φ = constant, which gives
zero, and thus, we can have another function φ (x, y), called dφ = 0. Again φ is an exact differential, so
potential function that obeys
∂φ ∂φ
dφ = dx + dy = 0 (3.226)
∂φ ∂x ∂y
u= (3.218)
∂x or
∂φ
v= (3.219)
∂y dφ = udx + vdy = 0 (3.227)
Interestingly, Eqs. (3.217), (3.218) and (3.219) suggest that
Hence, we get
∂φ ∂ψ
= (3.220) dy u
∂x ∂y = (3.228)
dx φ = constant v
∂φ ∂ψ
=− (3.221)
∂y ∂x
From Eqs. (3.225) and (3.228), it can be seen that
These equations are exactly similar to the Cauchy–Riemann
dy dy
equations given by Eq. (3.215), which suggest to have a com- × = −1 (3.229)
plex function, called the complex potential, of the form dx ψ = constant dx φ = constant
i.e., the streamlines and equipotential lines are orthogonal to

(z) = φ (x, y) + iψ (x, y) (3.222)
each other.
For an incompressible and irrotational flow, (z) will be ana-
lytic2 everywhere in the domain of analysis. Furthermore,
3.16 The Bernoulli’s Equation
Eqs. (3.218) and (3.219) show that the function φ (x, y) is per-
What is Fluid potential energy and pressure relation?
pendicular to the gradient of φ (x, y), and therefore, the curves
According to Bernoulli’s principle, an increase in the flow
traced by the constant potential functions (iso−φ) are orthog-
speed simultaneously results with a decrease in pressure or
onal to the curves depicted by the constant stream functions
a decrease in fluid’s potential energy. This theorem is named
(iso−ψ), depicted in Fig. 3.18. To understand this result more
after Daniel Bernoulli (1700–1782), a Swiss mathematician
and physicist, who published it in his book Hydrodynamica
in 1738. We will discuss both steady and unsteady forms of
2A complex function f (z) is said to be analytic in a region R of the
the Bernoulli’s equation in the following subsections.
complex plane, if
• f (z) has a derivative at each point of R.

• f (z) is single valued.
Potential line
Streamline ψ
3 φ
ψ 3
2
φ
ψ 2
1
φ
1
Fig. 3.18 The streamlines and potential lines in a flow field
Streamline
A fluid element moving in a streamline (p + δ p ) δ s δ t

δs n
δn δ t = particle thickness
(p −δ ps )δ n δt
n
δ Wn
δW
s δn θ δz
δs
δ Ws θ
θ δz θ
Normal to streamline
Along streamline
(p −δ p )δ s δt
n (p + δ ps ) δ n δ t
τδ s δ t = 0
Gravity (g)
Fig. 3.19 A fluid element moving along a streamline
3.16.1 Steady Bernoulli’s Equation where δm = differential mass of the fluid element, and as =
acceleration of the fluid element along the streamline direc-
To derive the steady Bernoulli’s equation, we will use the tion. Also,
streamline (s, n) coordinate system, where s is along the
streamline direction and n is the direction perpendicular to dv
as =
the streamline (normal direction). Let us assume a steady, dt
inviscid, and incompressible flow occurring along the stream-
Using chain rule,
line with negligible surface tension effects, as sketched in
Fig. 3.19. Consider a small fluid element of size δs × δn in dv ds dv
the sn plane, which is moving along the streamline. The vol- as = =v
ds dt ds
ume of the fluid element is δ∀ = δs × δn × δt, where δt is
the thickness of fluid element in the direction perpendicular Thus, Eq. (3.230) becomes
to the sn plane.
From Newton’s second law of motion, the sum of all the dv
δFs = δm × v
forces acting on the fluid element along the streamline direc- ds
tion can be written as
But δm = ρδ∀

δFs = δm × as (3.230) dv
δFs = ρδ∀ × v
ds
3.16 The Bernoulli’s Equation 87
or Canceling δ∀ both sides,

dv ∂p dv
δFs = ρv δ∀ (3.231) + ρv + ρg sin θ = 0 (3.236)
ds ∂s ds
Note that Eq. (3.231) is valid for both incompressible and But, from Fig. 3.19 sin θ = dz
ds , thus
compressible flows. Now let us evaluate term at the left-
hand side in Eq. (3.231) by collecting all the forces acting ∂p dv dz
+ ρv + ρg = 0 (3.237)
on the fluid element in streamline direction. We will first con- ∂s ds ds
sider the gravitational force due to which fluid element has its
own weight, δW, acting in the vertical direction (z-direction) Since p = p (s, n), the differential pressure dp can be written
as shown in Fig. 3.19. Taking component of δW along the as
streamline direction, we have
∂p ∂p
dp = ds + dn
δWs = −ρgδ∀ sin θ (3.232) ∂s ∂n
Why there is a negative sign here? However, for the fluid flow along a streamline dn = 0, thus
The second important force acting the fluid element is the
pressure. We know that the pressure is not constant throughout dp ∂p
the fluid due to weight of the fluid column. Thus, the pressure =
ds ∂s
(p) can be written as a function of streamline coordinates,
s and n, i.e., p = p (s, n). Consider p be the pressure acting Also, the second term in Eq. (3.237) can be rewritten as
at the origin O. Thus, the pressure acting over the face at the
right side of the element can be written as (p + δps ), while the dv 1 d v2
ρv = ρ
pressure acting at the left face of the element will be (p − δps ). ds 2 ds
For the small fluid element, using Taylor series expansion and
considering the terms up to first order only, we can write δps as Thus, Eq. (3.237) becomes

∂p δs dp 1 d v2 dz
δps = (3.233) + ρ + ρg = 0 (3.238)
∂s 2 ds 2 ds ds
Thus, the net pressure force acting on the fluid element along or What happened to sin(theta)?
the streamline direction will be dp 1 2
+ d v + gdz = 0 (3.239)
ρ 2
δFps = (p − δps ) δnδt − (p + δps ) δnδt = −2δps δnδt
Upon integration, we have
or
ˆ
∂p δs dp 1 2
δFps = −2 δnδt + v + gz = C (3.240)
∂s 2 ρ 2
or Since the flow is incompressible (ρ = constant), Eq. (3.240)

becomes
∂p
δFps = − δ∀ (3.234)
∂s 1
p + ρv2 + ρgz = C (3.241)
2
From Eqs. (3.232) and (3.234), total force (δFs ) along the
streamline direction becomes Equation (3.241) is the well known Bernoulli’s equation.
Here, C is the constant of integration, also referred to as
∂p the Bernoulli constant. Furthermore, note that while deriving
δFs = δWs + δFps = −ρg sin θ − δ∀ (3.235)
∂s Eq. (3.241), we have not stipulated whether the flow is rota-
tional or irrotational, and thus, this equation remains valid in
From Eqs. (3.231) and (3.235), we have either case. Since the constant C has different values along
different streamlines in a rotational flow, the pressure and
dv ∂p velocity cannot be directly coupled for these streamlines. Con-
ρv δ∀ = −ρg sin θ − δ∀
ds ∂s sequently, in a rotational flow, the Bernoulli’s equation can be
applied only along a streamline. However, if the flow is irro- incompressible to hypersonic. Thus, the incompressible
tational, the constant C has a single value for entire flow field, Bernoulli’s equation (Eq. (3.244)) can be written as
and hence, the Bernoulli’s equation can be used between any
two points irrespective of the fact whether the points lie on qinc = p0 − pa (3.245)
the same or on different streamlines. This result is indeed an
outcome of Eq. (3.152), which states that stagnation pressure Reader should bear in mind that Eq. (3.245) has been obtained
will be constant across the streamlines only if the vorticity from the incompressible form of the Bernoulli’s equation, and
content in between them is zero, i.e., when the flow is irrota- hence, it is valid only for incompressible flows. For compress-
−
→ − →
tional ∇ × v = 0 . ible flows, the Bernoulli’s equation is not valid,
and thus, the
Note that for the flow situations where the effect of gravity generalized form of the dynamic pressure qc = 21 ρv2 will
can be neglected, Eq. (3.241) reduced to also be not equal to p0 − pa .
Consider again a baseball which is moving through the
1 freestream and simultaneously spinning about its center, as
p + ρv2 = C1 (3.242)
2 shown in Fig. 3.21. Let us assume the freestream velocity is
Ua and the absolute velocity of the baseball is v. The figure
Equation (3.242) is the most commonly used form of the reveals that the relative velocity at the top surface (point A)
incompressible Bernoulli’s equation. It reveals that whenever will be (v − Ua ) as both the velocities are in opposite direc-
pressure decreases at point the flow velocity increases at that tion, and the relative velocity at the bottom surface (point
point. We can understand more about its application by con- B) will be (v + Ua ) as they are in same direction. Thus,
sidering some examples. Assume an infinitely long cylindrical Eq. (3.242) shows that the pressure at A is higher than the
object is placed in a steady and uniform flow of freestream pressure at B; consequently, there will be a net aerodynamic
velocity Ua , as shown in Fig. 3.20. For both upstream and force (Fb ) acting in downward direction. Note that Fb always
downstream locations sufficiently far away from the object, act even in the absence of gravity.
the flow velocity (u, v) approaches to (Ua , 0) and the pressure We now examine a situation where an airfoil is moving
approaches to the freestream pressure (pa ). The Bernoulli’s through the freestream, as depicted in Fig. 3.22. Also, it is
equation given by Eq. (3.242) is valid for each streamline. assumed that the effect of the airfoil is felt in the near field
However, the value of terms at the left side will be same only. Figure 3.22 shows that due to airfoil shape, to attain the
and equal to pa + 21 ρa Ua2 , for each of these streamlines at far freestream conditions (at PP ) a fluid particle at the top surface
upstream locations (say at PP ). In other words, the Bernoulli has to cover comparatively a longer distance than if it moves
constant (C1 ) on the right side is same for all the streamlines. along the bottom surface. In order words, the fluid velocity at
Thus, we can write the following equation which holds for all the top surface is higher than the fluid velocity at the bottom
the streamlines: surface. From Eq. (3.242), the pressure at the bottom surface
1 1 is higher than the pressure at the top surface, and hence, a net
p + ρa v2 = pa + ρa Ua2 (3.243) force acts in upward direction. This net force which pushes
2 2
the aircraft upward is known as lift.
where ρa denotes the freestream density.
From Fig. 3.20, it can be seen that a streamline (S0 ) always
exists which divides the ongoing flow over the upper part 3.16.1.1 Physical Significance of the Bernoulli’s
of the cylindrical object (denoted by the streamline S+ ) and Equation
ongoing flow around its lower portion (streamline S− ). The The Bernoulli’s equation given by Eq. (3.241) is used exten-
dividing streamline (S0 ) terminates at a point “A” on the body sively in fluid mechanics and aerodynamics. This equation
surface, where the flow velocity is reduced to zero. This point can be represented in several forms, and the purpose of this
is called the stagnation point, where the pressure, p0 , also section is to discuss the physical meaning of the terms present
known as stagnation pressure, is given by in each of these equations. Recall, Eq. (3.241)
1 1
p0 = pa + ρa Ua2 (3.244) p + ρv2 + ρgz = C
2 2
Clearly, the stagnation pressure (p0 ) is higher than the where p denotes the static pressure, 21 ρv2 is the dynamic pres-
freestream static pressure (pa ) by a factor 21 ρa Ua2 , called sure, and ρgz is the hydrostatic pressure. Thus, it is evident that
the dynamic pressure and usually denoted by the symbol for a fluid flow the sum of the static pressure, dynamic pres-
q. Also, note that the term q ≡ 21 ρa Ua2 is the freestream sure, and the hydrostatic pressure along a streamline remains
dynamic pressure by definition and hence valid for all flows; invariant.
GENERALLY WHEN WE TALK ABOUT THE PRESSURE

WE TALK ABOUT THE STAGNATION PRESSUER
P Q
p
a
+
S
S0 A
Ua Ua
_
S
Divides the flow into upper and lower flow!!

P’ Q’
Fig. 3.20 A two-dimensional cylindrical object in the steady uniform flow
A
(v − Ua )
r
ω
(v + U a )
B
Fig. 3.21 Schematic diagram showing a baseball in translational and rotational motion
P
Streamlines
Fluid Element Airfoil

P’
Fig. 3.22 Schematic diagram of an airfoil moving through the air
See here , every energy seems to be ok but

the energy due to the pressure seems some
what unusual.
the energy due to pressure is because we are

seeing that the flow is moving up pushing the
flow now we who is pushing here is the
pressure's role here.
Fig. 3.23 Schematic diagram of the fluid flow in a streamtube
We will now examine the application of Bernoulli’s equa- is not just equal to the sum of kinetic and potential energies
tion for the flowing fluid in a varied area streamtube, as but includes an additional term, p. To understand this term
sketched in Fig. 3.23a. Let us apply Eq. (3.241) between the further, let us consider the flow in a streamtube at two different
points “1” and “2”, we have times, t and t + t, as schematically shown in Fig. 3.23b.
Suppose a fluid element denoted by the region A is entering
1 1 into the streamtube at the location “1” at time t. Since the fluid
p1 + ρv12 + ρgz1 = p2 + ρv22 + ρgz2 (3.246)
2 2 elements move in succession, the element before pushes the
or element next to them. Therefore, after the time interval t, the
fluid element exits the streamtube at location “2” (depicted
1
(p1 − p2 ) + ρ v12 − v22 + ρg (z1 − z2 ) = 0 (3.247) by the region B). The region of the streamtube (shown by
2 C) is common to both the cases. Thus, for the motion of a
where fluid element between “1” and “2” the contribution to the
total energy coming from the region C, cancels each other,
(p1 − p2 ) ≡ net work done by power force per unit volume whereas there are no regions available to both A and B to
1 get canceled. In other words, A shows the amount of energy
ρ v12 − v22 ≡ change in kinetic energy per unit volume
2 added to the fluid by the external power force (pressure in this
ρg (z1 − z2 ) ≡ change in potential energy per unit volume case) in pushing the fluid into the streamtube, and B shows the
amount of energy lost as the fluid comes out of the streamtube.
For a moving fluid element, it is now evident that the sum of Since the magnitudes of the regions A and B are not the same,
work done by power force per unit volume, the kinetic energy there will always be a net work done by external power force.
per unit volume, and the potential energy per unit volume is This work is, in fact, depicted by the pressure term, p.
a constant along a streamline. Thus, the Bernoulli’s equation Furthermore, if we divide Eq. (3.241) by ρg throughout,
essentially represents the law of conservation of energy, which we have the following form of the Bernoulli’s equation:
Bernoullis also represents the law of conservation of linear momentum

→ −
− → −→ → − → →
−
p
+
v2
+ z = C2 (3.248) ∇ →v .−
→
v = −v .∇ −v + →
v .∇ − v
ρg 2g −
→ → −
→ →
+−
→
v × ∇ ×− v +− →v × ∇ ×−
v
where C2 is again a Bernoulli’s constant, different from C and (3.250)
C1 . In Eq. (3.248), the terms can be explained as follows:
p We know that for an irrotational flow the curl of velocity

≡ pressure head vector is zero. Thus, from Eq. (3.158), we have
ρg
v2 −
→ −
≡ velocity head ∇ ×→
v =0 (3.251)
2g
z ≡ elevation head Thus, Eq. (3.250) reduces to
− → →
3.16.1.2 Application of Bernoulli’s Equation in −
→
v .∇ −v =∇
1−→
v .−
→
v (3.252)
Ideal and Real Fluids 2
One should note that in deriving Eq. (3.241), we have ignored
all the dissipative losses as the flow is assumed to be inviscid. Introducing Eq. (3.252) into Eq. (3.249), we get
For an inviscid flow in a pipe or an inviscid flow over a flat
D−
→v ∂−
→v −
→ 1−→
surface, the incompressible Bernoulli’s equation is still appli- = +∇ v .−
→
v (3.253)
cable even though the wall exerts a force on the flowing fluid. Dt ∂t 2
This is because the force being perpendicular in direction to
the fluid actually does no work along the flow direction. On Substituting the above equation into Eq. (3.145), we have
the other hand, in real fluids, the wall exerts a force parallel to
∂−
→v −
→ 1−→ −
→
flow direction due to viscosity and thus opposes the motion ρ +∇ v .−
→
v = ρ−
→
g − ∇p (3.254)
of fluid. Thus, in a real fluid, there will be a drag acting on ∂t 2
it which causes severe dissipative losses especially near to
For an irrotational flow the vector potential is given by
the wall boundary where the viscous effects are predominant. −
→ −
→
Clearly, the Bernoulli’s equation is not valid inside a boundary v = ∇ φ and thus Eq. (3.254) becomes
layer flow. ⎡ −→ ⎤
∂ ∇φ −
→ 1−→ − → −
→
ρ⎣ +∇ ∇ φ. ∇ φ ⎦ = ρ−
→
g − ∇ p (3.255)
∂t 2
3.16.2 Unsteady Bernoulli’s Equation
The purpose of this section is to derive the unsteady form Since the gravity acts in the negative z-direction
of Bernoulli’s equation. Consider again a fluid element in an −
→
unsteady, incompressible, and inviscid flow moving along a g = −gk̂ (3.256)
streamline. Let us assume the velocity vector field is shown
−
→
by −→ Also,
v . So that the total derivative of velocity DDtv , using
−
→
ρ−
→
Eq. (3.57) can be expressed as
g = − ∇ (ρgz) (3.257)
D−
→v ∂−
→v − → →
= + −
→
v .∇ −v (3.249) Introducing above into Eq. (3.255), we get
Dt ∂t
⎡ −→ ⎤
−
→ → −
→ →
Consider the vector identity3 and set A = −
v and B = −
v, ∂ ∇φ −
→ 1 −
→ −
→ −
→ −
→
ρ⎣ +∇ ∇ φ. ∇ φ ⎦ = − ∇ (ρgz) − ∇ p
we get ∂t 2
(3.258)
or
3 If −
→ −
→ ⎡ −
→ ⎤
A and B are two vectors, then ∂ ∇φ
⎣ρ −
→ 1−→ − → −
→ −
→ ⎦
→ −
− →−→ −
→−→ −
→ −
→− → −
→ −→ −→ − → −→ −→ − →
+ ρ∇ ∇ φ. ∇ φ + ∇ (ρgz) + ∇ p = 0
∇ A. B = A.∇ B + B .∇ A + A × ∇ × B + B × ∇ × A ∂t 2
−
→
(3.259)
where ∇ is the del or nabla operator.
or where p0 is measured pa is measured through the static port

(B). Also, note that the freestream velocity Ua is the true
→ ∂φ
− 1−→ − → airspeed (TAS) relative to the surrounding air, vtrue , yields
∇ ρ +ρ ∇ φ. ∇ φ + ρgz + p = 0 (3.260)
∂t 2
2 (p0 − pa )
Upon integration, we get the following unsteady Bernoulli’s vtrue = (3.262)
ρa
equation for an incompressible, inviscid, and irrotational flow.
∂φ 1−→ − →
ρ +ρ ∇ φ. ∇ φ + ρgz + p = f (t) (3.261)
∂t 2 3.17.2 Pressure Coefficient and the
Compressibility Correction Factor
where f (t) is a time-dependent arbitrary function (so-called
constant of integration). In Chap. 2, we have seen that it is often convenient to express
variables in nondimensional coefficient form, i.e., the coeffi-
3.17 Using the Bernoulli’s Equation cient of pressure, defined as
(p − pa )
The Bernoulli’s equation is widely used in solving many fluid Cp = (3.263)
2 ρa Ua
1 2
flow problems. The operating principle for a large number of
instruments in the aircraft is governed by this equation. Thus,
where Cp = pressure coefficient; p = static pressure at some
it is imperative to examine whether such an equation, devel-
point in the flow where the velocity is v; pa = freestream
oped after considering so many assumptions, is still pertinent
static pressure; ρ = freestream density; and Ua = freestream
to the practical engineering equipment. In this section, we
velocity. Now, for an incompressible flow
will discuss the applicability of Bernoulli’s equation for the
measurement of aircraft speed. 1 1
p + ρa v2 = pa + ρa Ua2
2 2
3.17.1 Airspeed Measurement Using or
Pitot-Static Probe
1 2
p − pa = ρa Ua − v2
The measurement of airspeed is generally carried out using 2
a pressure probe called the pitot-static probe. The schematic
view of this probe is given in Fig. 3.24. As illustrated in the Thus, the pressure coefficient becomes
figure, the probe has a port (A) at its nose and others on its side
surface (B) which are located some distance downstream of v2
Cp,inc = 1 − (3.264)
the one at the nose. The port A facing the freestream essen- Ua2
tially measures the stagnation pressure as the flow entering
into the port comes to a halt isentropically (i.e., through a Examining Eq. (3.264), we can consider the following cases:
reversible adiabatic process). Consequently, all its dynamic
• If p > pa and v < Ua , then Cp,inc is positive.
pressure gets converted into the static pressure, and thus, an
• If p = pa and v = Ua , then Cp,inc is zero.
enhanced static pressure is experienced by the port, which is
• If p < pa and v < Ua , then Cp,inc is negative.
basically the stagnation pressure. However, the side ports B
cannot sense the dynamic pressure because the flow actually Furthermore, considering Fig. 3.20 again and defining the
slides over them. Hence, the side ports experience only the pressure coefficient at the stagnation point A as
static pressure. Moreover, the side ports are constructed at a
sufficient distance away from the front port so as to minimize p0 − pa
the flow acceleration effects on the pressure measurement. Cp0 ,inc = (3.265)
2 ρa Ua
1 2
Thus, the static pressure felt at the port B will be equal to the
freestream static pressure (pa ). It follows immediately from Eq. (3.244) that for an incom-
Let us apply the Bernoulli’s equation (Eq. (3.245)) between pressible flow
two points on a streamline such that one is in the freestream
and other is at the front port. Thus, Cp0 ,inc = 1 (always) (3.266)
1
pa + ρa Ua2 = p0 + 0 In Sect. 2.6.4, we have seen that whenever the Mach number
2
is greater than 0.3, the flow becomes compressible. Hence,
3.17 Using the Bernoulli’s Equation 93
Static pressure port
Stagnation pressure port
Static pressure
Stagnation pressure
Pressure transducer
(to measure differential pressure
Fig. 3.24 Schematic layout of a typical pitot-static probe
Eq. (3.266) which is obtained by assuming the incompress- Thus, Eq. (3.270) becomes
ibility condition, does not hold for compressible flows. To ⎛ ⎞
7
calculate the stagnation pressure coefficient in compressible 1 ⎝ M2 2
flow, our first step is to rewrite the dynamic pressure as fol- Cp0 ,c = 1 + − 1⎠ (3.271)
0.7M2 5
lows:
1 2 1 ρ From binomial series

q= ρv = γp v2 (3.267)
2 2 γp
(1 + x)n = 1 + nx + n(n−1) 2
2! x +
n(n−1)(n−2) 3
3! x + · · · ; −1 < x < 1
ρ
But γp = 1
a2
, where a is the speed of sound. Thus,
Solving Eq. (3.271), we get
v 2 !"
1 2 1
q= ρv = γp (3.268) 2
2 2 −1
7 7
2 2 a 1 7 M2 M2
Cp0 ,c = 1 + +
0.7M2 2 5 2 5
or 7 # $
3
2 2 −1 2 −2
7 7
M2
1 + + ··· − 1
q= γpM2 (3.269) 3 5
2
where M is the Mach number. Considering the fluid as air or

(γ=1.4), the stagnation pressure coefficient can be written as
M2 M4 M6
Cp0 ,c = 1 + + + + ··· (3.272)
(p0 − pa ) 1 p0 4 40 1600
Cp0 ,c = = −1 (3.270)
2 γpM
1 2 0.7M2 p One should keep in mind that the above stagnation pressure
coefficient for compressible flows is valid only in the Mach
From isentropic relation for air, we have number range, 0 ≤ M ≤ 1. For supersonic flows, Eq. (3.272)
is not applicable as the flow is completely modified due to
p0 7
= 1 + 0.2M2
2 shock wave formation. The variation of stagnation pressure
p coefficient with Mach number is given in Table 3.2.
Supersonic Motion equation

Table 3.2 Variation of Cp0 with M

M 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Cp0 1.0 1.0125 1.01 1.0227 1.0406 1.0641 1.0973 1.1286 1.1704 1.2192 1.2756
Fig. 3.25 Variation of Cp0 with M less than unity
The variation of stagnation pressure coefficient with Mach 3.17.3 The Airspeed
number, plotted using the MATLAB program, is shown in
Fig. 3.25. From the figure, it is evident that at M = 0.3 the The speed of an object (aircraft) relative to the surrounding
associated error in assuming the flow to be incompressible is air mass is called the airspeed. The pitot-static probe
just 2%, but rises rapidly with increase of Mach number. For is the commonly used instrument to measure airspeed in the
example, the associated error is around 6% at M = 0.5, but it laboratory and on the aircraft. However, there exist subtle
is about 28% close to sonic speed. differences in the requirements for the two applications.
Listing 3.1 The MATLAB program to generate. Figure 3.25

1 clc ;
2 c l o s e all ;
3 M = 0:0.1:1.0;
4 Cp = 1.0 + ( M . ^ 2 ) / 4 + ( M . ^ 4 ) / 4 0 + ( M . ^ 6 ) / 1 6 0 0 ;
5 plot (M , Cp , ’ bo - ’ , ’ M a r k e r S i z e ’ ,10 , ’ LineWidth ’ ,0.6);
6 grid on ;
7 x l a b e l ( ’M ’ , ’ FontSize ’ ,15 , ’ FontName ’ , ’ Times ’ , ’ FontAngle ’ , ’ italic ’);
8 y l a b e l ( ’ C_ { p_0 } ’ , ’ FontSize ’ ,15 , ’ FontName ’ , ’ Times ’ , ’ FontAngle ’ , ’ italic ’);
9 a x i s ([0 1.0 1 1 . 3 ] ) ;
10 set ( gca , ’ XTick ’ ,0:0.1:1.0 , ’ FontSize ’ ,12 , ’ FontName ’ , ’ Times ’);
11 set ( gca , ’ YTick ’ , 1 : 0 . 0 5 : 1 . 3 ) ;
12 set ( gca , ’ X M i n o r T i c k ’ , ’ on ’)
13 set ( gca , ’ Y M i n o r T i c k ’ , ’ on ’)
14 set ( gcf , ’ P a p e r P o s i t i o n M o d e ’ , ’ manual ’);
15 set ( gcf , ’ P a p e r U n i t s ’ , ’ inches ’);
16 set ( gcf , ’ PaperSize ’ , [7 3 . 5 ] ) ;
17 x S i z e = 7;
18 ySize = 3.5;
19 xLeft = (7 - xSize )/2;
20 yTop = (3.5 - ySize )/2;
21 set ( gcf , ’ P a p e r P o s i t i o n ’ ,[ xLeft yTop xSize ySize ]);
22 f i g u r e ();
3.17 Using the Bernoulli’s Equation 95
To measure airspeed the correct value of density should be Here, we should be able to distinguish between the above two
used in Eq. (3.262). This requirement is, although, feasible airspeeds; indicated airspeed (IAS) and equivalent airspeed
in the controlled laboratory environment where the density (TAS). The term indicated airspeed is used for the uncorrected
is either almost invariant or its variation can be easily reading of airspeed obtained using an actual airspeed indica-
accounted. However, to calculate the airspeed in actual flight tor, whereas equivalent airspeed is the uncorrected reading of
applications, one should consider the variation of density airspeed shown by an error-free airspeed indicator.
with ambient atmospheric pressure. There are five common
conventions of airspeeds which are used in aerospace 3.17.3.3 Calibrated Airspeed (CAS)
applications. It is the indicated airspeed (IAS) which is obtained after cor-
recting the instrument errors, static port position errors (PEC),
• True airspeed (TAS),
and the installation errors. During the turbulence-free flight
• Indicated airspeed (IAS),
operation, the position and the instruments errors are usually
• Calibrated airspeed (CAS),
small, and thus, the difference between indicated airspeed
• Equivalent airspeed (EAS), and
(IAS) and calibrated airspeed (CAS) is also small.
• Ground speed (GS).
The calibrated airspeed (CAS) becomes important when
the flaps are moved which, in turn, changes the pitot-static
3.17.3.1 Indicated Airspeed (IAS) probe pitch affecting the pressure readings as the probe is no
It is the speed of an aircraft as reflected directly on its flight more aligned with the incoming airflow.
deck instrument (pitot-static probe indicator) in the cockpit.
Since the indicator cannot essence the local air density at the
3.17.3.4 Ground Speed (GS)
flying altitude, it is by design calibrated to use the standard
It is defined as the speed of an aircraft relative to the ground.
sea level air density to calculate the airspeed. We have seen
Ground speed is essentially a vector addition of true airspeed
(TAS) and the speed of wind (− →
that the airspeed is estimated from the difference between
v wind ) at the flying altitude of
the stagnation pressure and the static pressure measured by a
aircraft. That is,
pitot-static probe, which is essentially a differential pressure
−
→
GS = TAS + −
→
gauge
that
shows the pressure readings in the units of speed v wind (3.275)
kmh−1 , instead of pressure. However, these readings are
subject to errors in the measurement of static pressure due to
This speed is, in fact, measured by the radar system of the
the inability to mount the static ports at appropriate positions,
aircraft traffic controller (ATC).
where the pressure is true static pressure at all airspeeds and
attitudes. The correction in mounting these ports is known
as position error correction (PEC), which changes with flight 3.18 Reynolds Transport Theorem
speed and the aircraft geometry.
Reynolds transport theorem (RTT), also known as
3.17.3.2 Equivalent Airspeed (EAS) Leibniz–Reynolds transport theorem, is essentially a
It is the speed at standard sea level (SL) conditions that pro- three-dimensional generalization of Leibniz integral rule.
duces same incompressible dynamic pressure which is gen- RTT is used to evaluate the derivatives of integrated quanti-
erated at true airspeed (TAS) and the altitude at which the ties. Moreover, the conservation laws of mass, momentum,
vehicle is flying. Since the density of air (ρa ) varies with alti- and energy are adopted from classical mechanics and ther-
tude and thus causes errors in a true airspeed indicator. One modynamics, where system approach is mainly followed,
solution to this problem could be to remove the dependency whereas in aerodynamics, it is often more convenient to work
of airspeed on local air density. By replacing ρa with ρa,SL in with control volumes as it is difficult to identify and follow a
Eq. (3.262), we obtain the expression for equivalent airspeed system of fluid particles. Thus, it is essential to relate control
(EAS). Thus, system approach and the control volume approach for certain
%
fluid and flow properties, which indeed is achieved through
2 (p0 − pa ) 2 (p0 − pa ) ρa RTT.
vequivalent = = ×
ρa,SL ρa ρa,SL In thermodynamics, we generally make a distinction
(3.273) between those properties of a substance whose measure
depends on the amount of the substance present and those
or properties whose measure is independent of the amount of
the substance present; former is called extensive property
√ ρa
vequivalent = vtrue × σ where, σ = (3.274) and latter is known as intensive property. The common
ρa,SL examples of extensive properties are as follows: volume,

weight, momentum, and energy. It is evident that changing Bsystem t+t = (BII )t+t + (BI )t+t + (BIII )t+t − (BI )t+t
the amount of mass directly changes the measure of these
properties. Thus, for each extensive property, we can But
introduce by distributive measurements the corresponding
intensive properties.
For example, one can define volume (BII )t+t + (BI )t+t = (Bcontrol volume )t+t
per unit mass ∀ = m∀ and energy per unit mass (e)
such that Hence,
˚
∀= ∀ ρd∀ (3.276) Bsystem t+t = (Bcontrol volume )t+t + (BIII )t+t − (BI )t+t
˚
E= eρd∀ (3.277) Again subtracting Bsystem t both sides in the above equation,
we have
Clearly, ∀ and e do not depend on the amount of matter

Bsystem t+t − Bsystem t = (Bcontrol volume )t+t − Bsystem t
present and are hence the intensive quantities related to exten-
+ (BIII )t+t − (BI )t+t
sive properties ∀ and E by distributive measure. They are also
referred to as specific quantities, i.e., specific volume and or
specific energy. Furthermore, properties such as pressure and

temperature are by their mass-independent nature already in Bsystem t+t − Bsystem t = (Bcontrol volume )t+t − (Bcontrol volume )t
the category of intensive property. + (BIII )t+t − (BI )t+t
−
→ − →
Consider an arbitrary flow field V = V (x, y, z, t) as seen
from some reference xyz wherein we observe a system of fluid Dividing above by the differential time t, we have
of finite mass at times t and t + t, as shown in Fig. 3.26 by the
curves ABCM and ANCD, respectively. Also, the streamlines Bsystem t+t − Bsystem t (Bcontrol volume )t+t − (Bcontrol volume )t
=
corresponds to those at time t. Let us consider an arbitrary t t
(BIII )t+t − (BI )t+t
extensive property B of the fluid for the purpose of relating the +
t
rate of change of this property for the control system with the
variations of this property associated with the control volume. In the limiting case, when t → 0
That is, & '
Bsystem t+t − Bsystem t
˚ lim =
t→0 t
B= βρd∀ (3.278)
(Bcontrol volume )t+t − (Bcontrol volume )t (BIII )t+t − (BI )t+t
lim +
t→0 t t

where β is the distribution of B per unit mass β = m
B
. One
should note that the volume in space occupied by the con- But
trol system (ABCM) at time t is also the control volume at & '
Bsystem t+t − Bsystem t dB DB
time t. lim = =
t→0 t dt system Dt

Bsystem t = (Bcontrol volume )t (3.279)
Thus,
That is, both control system and control volume are the same
DB (Bcontrol volume )t+t − (Bcontrol volume )t
at time t. At time t + t, the control surface moves out of Dt
= lim
t→0 t
system
the control volume and occupies the new position designated
(BIII )t+t − (BI )t+t
as ANCD. We have divided up our system at time t and the +
t
system at time t + t into three regions: I, II, and III, as
shown in Fig. 3.26. Thus, the fluid property B associated with Using sum4 law of limits, we can rearrange the above equation
the control system at time t + t consists of region II and as
region III. We have

Bsystem t+t = (BII )t+t + (BIII )t+t
Adding and subtracting

the property associated with the 4 TheSum Law states that the limit of the sum of two functions is the
region I (BI )t+t , we have sum of the limits. That is, if lim x→a f (x) = L, and lim x→a g (x) = M.
Further, if both L and M exist then lim x→a [ f (x) + g (x)] = L + M.
3.18 Reynolds Transport Theorem 97
System at time t + Δ t
System at time ’t’

A
Streamlines at time ’t’
III
II
N
I B
D
y
M
x C
Control volume at time ’t’
Fig. 3.26 A schematic view of moving system at times t and t + t
˝
DB (Bcontrol volume )t+t − (Bcontrol volume )t ( βρd∀)t+t
= lim the term limt→0 III
approximates the aver-
Dt system t→0 t t

+ lim
(BIII )t+t
− lim
(BI )t+t age rate of efflux of B across ABC during the˝interval t.
t→0 t t→0 t ( I βρd∀)t+t
Similarly, considering the last term limt→0 t
From Eq. (3.278), in Eq. (3.280) which approximates the amount of B that has
passed into the control volume during t through the control
DB
Dt
surface AMC. Hence, the last two integrals of Eq. (3.280) give
system
& ˝ ˝ ' the net rate of efflux of B from the control volume at time t.
βρd∀ control volume t+t
− βρd∀ control volume t
= lim Thus, net rate of efflux of B is
t→0 t
& ˝ ' & ˝ ' & ˝ ' & ˝ '
III βρd∀ t+t I βρd∀ t+t
III βρd∀ t+t I βρd∀ t+t
+ lim − lim (3.280) = lim − lim
t→0 t t→0 t
t→0 t t→0 t
Let us consider each term of the limiting process (3.282)
above separately.
˝ The first term
in Eq. (3.280), i.e.,
˝ We will now compute a more compact and workable form
( βρd∀)t+t −( βρd∀)t
limt→0 t denotes the partial of Eq. (3.282). To do so, consider Fig. 3.27, where a part of
derivative with respect to time. We have the
− control surface, dS, is in the steady-state velocity field
→v . The elemental area dS is also an interface of fluid that
& ˝ ˝ ' is just touching the control surface at times, t and t + t,
βρd∀ control volume t+t − βρd∀ control volume t
lim shown in Fig. 3.27a and b, respectively. Note that in differen-
t→0 t tial time interval (dt) the elemental area (interface) has trav-
˚
∂ eled a distance, vdt, in the direction tangent to the streamline
= βρd∀CV (3.281)
∂t at that point (Fig. 3.27a). Consequently, the volume of fluid
˝ that occupies the region swept out by dS in time dt, thus form-
( III βρd∀)t+t
In Eq. (3.280), the second term limt→0 t
ing a streamtube, is
essentially represents the amount of property moving out of d∀ = vdt (ds cos α)
the control volume through the control surface, ABC. Thus,
−
→
Fig. 3.27 Interface d S at the control surface at different times
¨
−
→
or Befflux rate = βρ−
→
v .d S
d∀ ANC
= v (ds cos α)
dt
Similarly, the influx rate of property B through the control
Multiplying above by ρ then gives instantaneous mass flow surface AMC at time t is
rate of fluid moving out of the control volume through the ¨
−
→
indicated area ds. We have Binflux rate = − βρ−→
v .d S
dm d∀ ANC
=ρ = ρv (ds cos α)
dt CS dt
Negative sign in the above expression indicates
that at the
control surface AMC, velocity vector −
→v and unit normal
In vector notation
n̂ to dS, are in opposite direction, as schematically shown
dm −
→ in Fig. 3.27b. Thus, the net efflux rate of B through the entire
= ρ−
→
v .d S control surface is
dt CS
−
→ (Bnet efflux rate )CS =Befflux rate − Binflux rate
where d S = n̂d S. The efflux rate of property B through the
control surface ABC at time t + dt can be estimated as
or
3.18 Reynolds Transport Theorem 99
⎛ ⎞ ! −→
¨ ¨ β ρ−→ . dS shows the net flux of property (B)
−
→ −
→ 3. v
βρ−
→
v .d S − ⎝− βρ−
→
v .d S ⎠
CS
(Bnet efflux rate )CS =
through the control surface (CS).
ANC ANC
or
3.19 The Energy Equation
¨ ¨
−
→ −
→
(Bnet efflux rate )CS = βρ−
→
v .d S + βρ−
→
v .d S From the first law of thermodynamics, we know that the rate
ANC ANC of change of energy of a fluid particle is equal to the sum of
the rate of heat addition and the rate of work done. The rate
In the limit as t → 0, the above relation becomes exact;
of change of energy is ρ DE Dt where E is the total energy of
thus, the terms on the right side can be expressed as
the particle. The change in total energy per unit mass (e) is
" →
− written as
1 2
(Bnet efflux rate )CS = β ρ−
→
v .d S (3.283) e = ui + u + v 2 + w2 (3.285)
2
CS

where the integral is a closed surface integral over the entire where ui and 21 u2 + v2 + w2 , respectively, are the internal
control surface. Also, one should note that even though we energy and the kinetic energy per unit mass of the system
have assumed a steady flow while deriving Eq. (3.283) but it is (fluid particle). Note that the potential energy (gravitation) is
equally applicable to unsteady flows as well. Now introducing not included in Eq. (3.285) and generally treated separately
Eqs. (3.281) and (3.283) into Eq. (3.280), we have as a source term. We will now derive the energy equation by
setting the total derivative equal to the change in energy as a
˚ "
DB ∂ →
− result of work done by viscous stresses and the net heat con-
= (βρ) d∀ + β ρ−
→
v .d S duction. Subsequently, by subtracting kinetic energy equa-
Dt system ∂t
C∀ CS tion one arrives at the conservation equation for the internal
(3.284) energy.
To calculate the work done by surface stresses, consider
Equation (3.284) is known as Reynolds transport theorem, the control volume with the origin located at the center of
which essentially relates the system approach to the con- the volume as shown in Fig. 3.28a. Since the work done by a
trol volume approach. One should note that while deriving force is the dot product of force vector and velocity vector,
−
→→
Eq. (3.284), we have assumed a fixed control volume in an w = F .− v . Thus, the work done by x-component of stresses
arbitrary reference frame xyz and the velocity field was mea- can be expressed as
sured relative to this reference frame. Clearly, the velocity
∂ ∂ ∂ ∂
vector −
→v is indeed measured relative to the control volume. = − (px u) + (uσxx ) + uτxy + (uτxz ) dxdydz
∂x ∂x ∂y ∂z
(3.286)
Similarly, the work done by y- and z-components of stresses
3.18.1 Physical Significance of RTT are
We have introduced the Eulerian and Lagrangian viewpoints ∂ ∂ ∂ ∂

= − py v + vτyx + vσyy + vτyz dxdydz
in Sect. 3.3. These approaches permit us to monitor fluid ∂y ∂x ∂y ∂z
particles passing through a fixed position in coordinate (3.287)
space (Eulerian approach) or to follow an individual particle ∂ ∂ ∂ ∂
= − (pz w) + (wτzx ) + wτzy + (wσzz ) dxdydz
(Lagrangian approach). The Reynolds transport theorem ∂z ∂x ∂y ∂z
enables a way to transfer the conservation laws for mass, (3.288)
momentum, and energy obtained from Lagrangian point of
Add Eqs. (3.286), (3.287) and (3.288) and divide by dxdydz
view (control system) to the Eulerian point of view (control
to get the work done per unit volume by the surface stresses.
volume). The significance of each term in Eq. (3.284) is
mentioned below:
∂ ∂ ∂
=− (px u) + py v + (pz w)
1. DB ∂x ∂y ∂z
Dt system represents the time rate of change of property
(B)˝ within a system. ∂ ∂ ∂
∂
+ (uσxx ) + uτxy + (uτxz )
2. ∂t C∀ (βρd∀) denotes the time rate of change of property ∂x ∂y ∂z
(B) within the control volume (C∀). ∂ ∂ ∂
+ vτyx + vσyy + vτyz
∂x ∂y ∂z
Fig. 3.28 Schematic diagram of control volume with energy exchange
∂ ∂ ∂
+ (wτzx ) + wτzy + (wσzz ) into the control volume at x− face is qx − dx d
qx . 21 dx dydz.
∂x ∂y ∂z Thus, the net rate of heat flux entering into the control volume
in x-direction is
or
∂ ∂ ∂
= −∇ (pv) + (uσxx ) + uτxy + (uτxz ) d 1 d 1
∂x ∂y ∂z = qx − qx . dx dydz − qx + qx . dx dydz
dx 2 dx 2
∂ ∂ ∂
∂x ∂y ∂z
dqx
∂ ∂ ∂ =− dxdydz
+ (wτzx ) + wτzy + (wσzz ) (3.289) dx
∂x ∂y ∂z
Similarly, the rate of heat fluxes in y- and z-directions is
Next, we will calculate energy flux due to heat conduction in dq
− dyy dxdydz and − dq z
dz dxdydz. Summing the terms in all the
all three x-, y- and z-directions. Consider again the control directions and dividing by dxdydz gives the net rate of heat
volume with the origin located at the center of the volume, as transfer to the fluid particle per unit volume.
shown in Fig. 3.28b.
The heat flux (q) has three components; qx , qy and qz in x-, dqx dqy dqz
y- and z-directions, respectively, at the origin. Using Taylor =− − −
dx dy dz
series, the
heat dflux 1crossing
out of the control volume at x+
= −∇q (3.290)
face is qx + dx qx . 2 dx dydz, while the heat flux entering
3.19 The Energy Equation 101
But, from Fourier’s law of heat conduction, Du −

→
ρ = −p ∇ v
Dt
dT ∂ ∂ ∂
qx = −k + (σxx ) + uτxy + (τxz )
dx ∂x ∂y ∂z
∂ ∂ ∂
+ τyx + σyy + τyz
dT ∂x ∂y ∂z
qy = −k
dy ∂ ∂ ∂ → −
− →
+ (τzx ) + τzy + (σzz ) + ∇ k ∇ T + Su
∂x ∂y ∂z
(3.295)
dT
qz = −k
dz Introducing the stresses in the equation above, we obtain
→ − →
where k is coefficient of thermal conductivity. Hence, in vec- Du − → → −
→→ −
tor notation ρ + ∇ ρu− v = −p ∇ − v + ∇ k ∇ T + + Su
Dt
q = −k∇T (3.291) (3.296)
where is the viscous dissipation term which is defined as
Thus, the heat flux due to conduction becomes " #
2 ∂u ∂v 2 ∂u ∂w 2 ∂w ∂v 2
= μ − + − + −
−∇q = ∇ (k∇T) (3.292) 3 ∂x ∂y ∂x ∂z ∂z ∂y
" #
∂u ∂v 2 ∂w ∂u 2 ∂v ∂w 2
Equating the total derivative for the energy in a fluid parti- +μ + + + + +
∂y ∂x ∂x ∂z ∂z ∂y
cle equal to the previously derived work and heat flux terms
(3.297)
results in the following energy equation:
DE → ∂ ∂ ∂ Equation (3.296) is essentially the nonconservation form

ρ = −∇ p− v + (uσxx ) + uτxy + (uτxz ) of the energy equation for the viscous fluid flow. Also,
Dt ∂x ∂y ∂z
Eq. (3.297) reveals that is always positive.
∂ ∂ ∂
∂x ∂y ∂z
∂ ∂ ∂
+ (wτzx ) + wτzy + (wσzz ) Summary
∂x ∂y ∂z
+ ∇ (k∇T) + SE (3.293) Fluid mechanics is the branch of science which studies the
statics and dynamics of fluids, and thus, it is further classified
where SE is the source term that includes energy sources such into fluid statics and fluid dynamics. As the name suggests, the
as potential energy, sources due to heat production from chem- fluid statics studies the fluid at rest, whereas the fluid dynamics
ical reactions, etc. Furthermore, we now derive kinetic energy investigates the effect of forces on the fluid in motion. To
equation. Consider x-, y- and z-momentum equations. Mul- analyze a moving fluid, the fundamental laws such as the
tiply x-momentum equation by u, y-momentum equation by conservation principles of mass, momentum, and energy are
v, and z-momentum equation by w and adding used. Moreover, in arriving at the basic equations of motion
one can either define the motion of individual molecules or can
D 1 2
ρ u + v 2 + w2 = −−→v .∇p predict the gross behavior of molecules in a control volume.
Dt 2 However, defining the motion of each and every molecule in
∂ ∂ ∂ the fluid is a cumbersome task, and usually not preferred as
+u (σxx ) + uτxy + (τxz )
∂x ∂y ∂z one is usually interested in knowing the average properties of
∂ ∂ ∂ the flow in a selected control volume. The appropriate size of
+v τyx + σyy + τyz the chosen control volume is very important, as the number
∂x ∂y ∂z
of molecules in this elemental volume must be sufficiently
∂ ∂ ∂
+w (τzx ) + τzy + (σzz ) (3.294) large to have a meaningful statistical description of the flow.
∂x ∂y ∂z
That is, the elemental control volume must be a continuous
Subtracting Eq. (3.294) in Eq. (3.293) media.
−
→ −
→
The scalar (dot) product of the vectors A and B are given a given point. Mathematically, the divergence is the scalar
as product of the ∇ operator and the vector field on which it
−
→
acts. In Cartesian space, if V = Vx î + Vy k̂ + Vz k̂, then
−
→− → − →
→ − −
→
A . B = A B cos α ∇. V is
− − −
→− → ∂ ∂ ∂
→ → ∇.V = Vx + Vy + Vz
where A and B , respectively, denote the magnitudes of ∂x ∂y ∂z
−
→ −
→
vectors A and B , and α is smaller of the included angle. In Cartesian space, if the flow velocity is given as
−
→ −
→
The cross product of the vectors A and B is defined as −
→ = v î + v k̂ + v k̂; the curl of the velocity vector
v−
→ −
x y z
−
→ − → −
→ − → −
→ ∇ ×→ v will be given by
A × B = A B sin α · n̂ = C

î ĵ k̂
−
→ −
→ −
→ −
→ −
∂ ∂ ∂
where C is perpendicular to the plane containing A and B , ∇ ×→
v = ∂x
and n̂ is the unit vector a direction which obeys “right-hand ∂y ∂z
−
→ −
→ v x v y vz
rule”. That is, if we rotate A into B by curling our fingers,
−
→
then the right-hand thumb will point in the direction of C . The continuum hypothesis allows us to surmise that at any
A scalar quantity given as a function of coordinate space instant t, there exists a fluid particle corresponding to every
and time is called scalar field. For example, pressure, density, point in space occupied by the fluid. Suppose a property such
and temperature are scalar fields, given in Cartesian space as as temperature T of the fluid is to be specified, it could be
follows: done in two different ways. In the first approach, the prop-
p = p (x, y, z, t) erty is specified as a function of the position in space and
time, i.e., T = T (x, y, z, t). This is termed as Eulerian or
ρ = ρ (x, y, z, t)
field description, which essentially specifies the temperature
T = T (x, y, z, t) of that fluid particle which happens to be at the location
(x, y, z) at the given time t = t1 . At time t = t2 , the tem-
Similarly, a vector quantity given as a function of coordinate perature T (x, y, z, t2 ) is the temperature, not of the same par-
space and time is called a vector field. For example, velocity ticle, but of a different particle—the one that happens to be
is a vector field, at location (x, y, z) at t = t2 . The other approach of specify-
ing a property in a moving fluid consists of identifying the
−
→
v = vx î + vy ĵ + vz k̂ fluid particles with some labels, following them around, and
specifying their properties as a function of time. Usually, the
where particles are labeled by the space point they occupied at some
vx = vx (x, y, z, t) initial time t0 . Thus, T (x0 , y0 , z0 , t) refers to the temperature
vy = vy (x, y, z, t) at time t of a particle which was at location (x0 , y0 , z0 ) at t0 .
vz = vz (x, y, z, t) This approach of identifying material points and following
them along is termed as the Lagrangian or the particle or the
For a pressure field in Cartesian space, p = p (x, y, z), the material description. →
gradient of p at a given point (x, y) in space is defined as For a given velocity field − v , the relation between the
Dρ
∂p ∂p ∂p total or material or Lagrangian derivative of density Dt
∇p = î + ĵ + k̂
∂x ∂y ∂z and the local derivative of density ∂ρ
∂t is given by
∂ρ − →
It is a vector whenever its magnitude is the maximum rate
Dρ −
of change of p per unit length of the coordinate space at the = + →
v .∇ ρ
Dt ∂t
given point and its direction is that of the maximum rate of
change of p at the given point. From the engineering applications point of view, four basic or
The divergence (positive) of a vector field primary laws must be satisfied for a continuous media. They
−
→ − → −
→− →
V = V (x, y, z) is a scalar field ∇ . V , which measures are as follows:
−
→
how much V spreads out at each point or for a negative 1. Conservation of mass (continuity equation),
−
→
divergence, how much V converges to the point. That is, 2. Newton’s second law (momentum equation),
the divergence represents the volume density of the outward 3. Conservation of energy (first law of thermodynamics), and
flux of a vector field from a small elemental volume around
Summary 103
4. Increase of entropy principle (second law of thermody- For an element which is rectangular in shape when the
namics). motion has just started. If the flow velocity significantly varies
across the extent of the fluid element the corners may not move
In addition to these primary laws, there are numerous in unison, and thus the element may rotate and in addition, its
secondary (or auxiliary) laws, frequently called constitutive shape may become distorted. If the adjacent sides of the fluid
relations, that apply to particular type of medium (or flow element rotate equally and in the same direction, we have pure
processes). The equation of state for perfect gas, Newton’s rotation; however, if the adjacent sides rotate equally but in
law of viscosity, etc., are some common examples of opposite directions, it leads to pure shearing motion.
subsidiary laws. Moreover, the Hooke’s law for elastic The fundamental governing equations for an incompress-
solids is also an auxiliary law, which is widely used in solid ible flow are the continuity and momentum equations. For
mechanics. steady incompressible flow, the continuity equation in differ-
The fundamental laws of fluid motion are as pertinent to ential form is written as
individual particles as to a group of particles. That is, the
∂ρ − → →
laws obtained for finite control volumes are equally applica- + ∇ . ρ−
v =0
ble when the control volume dwindles to a point in the flow ∂t
field. The method of analysis, where large control volumes
For steady and incompressible flows, the momentum equation
are used to obtain aggregate forces or transfer rates, is termed
(also known as Navier–Stokes equation) can be written as
as integral analysis. On the other hand, when the analysis is
−
→
∂−
→
applied to individual points in the flow field, the resulting → −
v ∇p
equations are differential equations and the method is termed + −
v .∇ →
v =−
→
g − + ν∇ 2 −
→
v
∂t ρ
as differential analysis.
In reality, all properties and flow characteristics are usually Circulation () is defined as the line integral of a velocity
expressed as a function of three space coordinates and time. field around a closed curve.
That is, in general, fluid flows are of three-dimensional in ˛
−
→
nature. A two-dimensional flow is distinguished by the con- = − →v . ds
dition that all properties and flow characteristics are functions
c
of two space coordinates and time, and hence do not change
along the third space coordinate direction. A one-dimensional and circulation per unit area is known as vorticity (ζ).
flow is a further simplification, where all properties and flow
characteristics are assumed to be expressible as function of
ζ=
one space coordinate and time. s
The quantitative and qualitative information of fluid flow
In vector notations, the vorticity is defined as
can be obtained through flow visualization, graphical rep-
−
→ − → →
ζ = ∇ ×−
resentation, and mathematical analysis. However, the visual
v
representation of flow fields is an important tool in modeling
the flow phenomena. In general, there are four basic types
If ζ = 0, the flow is called an irrotational flow.
of line patterns used to visualize the flow, namely, timelines,
According to Bernoulli’s principle, an increase in the flow
pathlines, streaklines, and streamlines.
speed simultaneously results with a decrease in pressure or
The line joining a set of fluid particles at a given instant is
a decrease in fluid’s potential energy. This theorem is named
known as timeline.
after Daniel Bernoulli (1700–1782), a Swiss mathematician
A streamline is an imaginary line in a fluid flow drawn
and physicist, who published it in his book Hydrodynam-
in such a manner that it is everywhere tangent to the velocity
ica in 1738. The Bernoulli’s theorem states that in a fluid
vector at the particular instant in time at which the observation
flow, the sum of the static pressure, dynamic pressure, and
is made.
the hydrostatic pressure along a streamline remains invariant.
A pathline is defined as a line in the flow field describing
For steady flow, the Bernoulli’s equation is written as
the trajectory of a given fluid particle. It is essentially the path
traversed by the fluid particle through the coordinate space 1
over a period of time. p + ρv2 + ρgz = constant
2
A streakline is defined as the instantaneous line of all the
fluid elements that have passed through the point of injection For an incompressible, inviscid, and irrotational flow, the
at some earlier time. unsteady form of Bernoulli’s equation is
∂φ 1−→ − → Exercises
ρ +ρ ∇ φ. ∇ φ + ρgz + p = f (t)
∂t 2
Descriptive Type Questions
The speed of an object (aircraft) relative to the surround-
ing air mass is called the airspeed. The pitot-static probe is 1. Express the Kinetic energy of the fluid for the two-
the commonly used instrument to measure airspeed in the dimensional motion in terms of stream function.
laboratory and on the aircraft. However, there exist subtle 2. Express the governing equations for an ideal fluid in terms
differences in the requirements for the two applications. To of streamline coordinates.
measure airspeed the correct value of density should be used 3. Consider an Eulerian flow field. In Cartesian coordinate
in calculations. This requirement is, although, feasible in the space, let the velocity and density fields are given by
controlled laboratory environment where the density is either −
→v = 3î + 2xĵ and ρ = 4y2 , respectively. Calculate the
almost invariant or its variation can be easily accounted. How- rate of change of density in Lagrangian reference frame,
ever, to calculate the airspeed in actual flight applications, one i.e., following a moving fluid element.
should consider the variation of density with ambient atmo- 4. The Lagrangian flow field description is given by
spheric pressure. There are five common conventions of air-
speeds which are used in aerospace applications: x = x0 exp k (t − t0 )
y = y0 exp −k (t − t0 )
• True airspeed (TAS),
• Indicated airspeed (IAS), z = z0
• Calibrated airspeed (CAS),
• Equivalent airspeed (EAS), and where k = 0.2 s−1 is a constant. At time t = t0 = 0, if
• Ground speed (GS). a particle is located at x0 = 1 m, y0 = 1 m, and z0 = 0,
then calculate the position of the particle at t = 2 s.
Reynolds transport theorem (RTT), also known as 5. For a given flow field, if the velocity components are given
Leibniz–Reynolds transport theorem, is essentially a by
three-dimensional generalization of Leibniz integral rule.
The conservation laws of mass, momentum, and energy u = −A exp {−ky} cos kx
adopted from classical mechanics and thermodynamics, v = −A exp {−ky} sin kx
where the system approach is mainly followed, whereas
in aerodynamics, it is often more convenient to work with where A and k are the constants. Find the stream func-
control volumes as it is difficult to identify and follow a tion for this flow. Check whether this flow is rotational or
system of fluid particles. Thus, it is essential to relate control irrotational?
system approach and the control volume approach for certain 6. Derive the equation of static equilibrium for a large mass
fluid and flow properties, which indeed is achieved through of fluid, where different parts are held together by means
RTT. In mathematical form, it is expressed as of gravitational attraction.
˚ "
→
7. Draw the streamline for the fluid motion, given by
DB ∂ −
= (βρ) d∀ + β ρ− →v .dS ψ = xy, where x ≥ 0 and y ≥ 0.
Dt system ∂t
C∀ CS 8. Find out the equations of streamlines for the velocity field
u = x2Ax and v = x2Ay , where A is a constant.
( +y2 ) ( +y2 )
The nonconservation form of the energy equation for the vis- 9. Find the stream function for the potential function

cous fluid flow is (x+a)2 +y 2
φ = 2 ln
1
.
(x−a)2 +y2
Du − → → −
→→ − → −
→ 10. A rocket is moving vertically upward against the gravita-
ρ + ∇ ρu−
v = −p ∇ −
v + ∇ k ∇ T + + Su tional force of the earth. If the rocket motor is operating
Dt
under steady flow conditions, so that the pressure in the
where is the viscous dissipation term which is defined combustion chamber and the exhaust gas velocity at the
as nozzle exit do not change with respect to time. Calculate
" #
2 ∂u ∂v 2 ∂u ∂w 2 ∂w ∂v 2 the forward velocity of the rocket at any instant t in a fixed
= μ − + − + − reference frame. Neglect the frictional effects caused by
3 ∂x ∂y ∂x ∂z ∂z ∂y
" # the earth’s atmosphere.
∂u ∂v 2 ∂w ∂u 2 ∂v ∂w 2
+μ + + + + +
∂y ∂x ∂x ∂z ∂z ∂y
Exercises 105
Multiple Choice Questions 5. Consider the motion of two circular cylinders of different
diameters in still air. If the velocities of these cylinders are
1. Consider the following statements: inversely proportional to their radii, then the drag experi-
(I) The line joining a set of fluid particles at a given instant enced by them will be
is known as timeline.
(II) A streamline is an imaginary line in flow field drawn in (a) same
such a manner that it is everywhere tangent to the velocity (b) more for the cylinder with larger diameter
vector. (c) less for the cylinder with larger diameter
(III) A streakline is defined as the instantaneous line of (d) more for the cylinder with smaller diameter
all the fluid elements that have passed through the point
of injection at some earlier time. 6. Consider the following statements with respect to the con-
(IV) A pathline is defined as the trajectory of a given fluid tinuity equation.
particle. (I) It is based on the postulation of the conservation of
Which of the above statements is/are FALSE? mass.
(II) It is not applicable to compressible fluids.
(a) I only (III) It can only be used for the analysis of conserved
(b) I and III quantity.
(c) II and IV Which of the above is/are correct?
(d) None
(a) I only
2. Let ρa is density of air at an altitude and ρa,SL is the (b) I and II
ρa
air density at standard sea level and σ = ρa,SL . Which of (c) I and III
the following express the relationship between equivalent (d) II and III
airspeed and true airspeed?
7. The smoke rings exiting the chimney are an example of
(a) vequivalent = vtrue × σ
√ (a) Streamline
(b) vtrue = vequivalent × σ
√ (b) Streakline
(c) vequivalent = vtrue × σ
(d) vtrue = vequivalent × σ 2 (c) Pathline
(d) Streamtube
3. For high-speed aircraft, the indicated airspeed has to be
corrected for compressibility. The instrument, used to 8. In a fluid flow, if the streamline pattern is invariant even
achieve it, is after sometime then the flow is
(a) Altimeter (a) steady

(b) Mach meter (b) unsteady
(c) Gyroscope (c) uniform
(d) Pitot probe (d) nonuniform
4. If the velocity components in x- and y-directions are 9. Which of the following statement is true?
u = 2x + 3y and v = 6x + 2y, respectively. The magni-
(a) Streamlines and pathlines are same in an incompress-
tude of vorticity will be
ible flow.
(a) 0 (b) Streamlines and streaklines are same in any fully
(b) 3 developed flow.
(c) 4 (c) Streaklines and pathlines are same in a uniform flow.
(d) 9 (d) Streamlines and streaklines are same in a steady flow.
10. In a steady flow field, the material derivative of velocity 7. (b)

of any fluid particle 8. (a)
9. (d)
(a) may or may not be zero 10. (a)
(b) is always zero
(c) is never zero
(d) is independent of convective derivative of flow veloc-
ity
Keys
1. (d)
2. (c)
3. (b)
4. (b)
5. (a)
6. (c)
If we can find a solution to a function, without putting any constraints and not taking any
arbitrary conditions.That type of solution is called a closed-form solution
Potential Flow Theory

Motivation to study this chapter:
4
We can get the velocity and pressure from the four governing equations,
but we should also remember that the Navior-Stokes equation is non -linear, we
cannot get the solutions that easily. Unless we put many constraints and conditions in
place. But the potential flow theory allows us to
find complicated flows because it satisfies the Laplace equation and Laplace is a linear
function.
In an irrotational flow field, one can use either velocity can still be complicated and thus we want even more simpler
potential or stream function to characterize the flow and equations to solve. It is achieved by applying an additional
both must fulfill Laplace equation. Besides, the investiga- kinematic constraint on the flow (i.e., assuming an irrotational
tion of this equation is substantially less demanding than the flow field), which leads to a much simpler equation that can
coordinate approach of completely viscous Navier–Stokes be solved directly to obtain the velocity. Such type of flow is
equations. The linearity of Laplace equation enables to add known as the potential flow and resulting equation is called
various basic solutions to obtain more complicated solutions. the Laplace’s equation.
The investigation of these solutions of Laplace equation is summery is that why we don't want to use much complicated
Navier stokes equation
referred to as potential flow theory, which has a great deal of
practical applications in characterizing complex flow fields. In 4.2 Potential Flows and the Laplace’s
this chapter, we will discuss the velocity potential and stream Equation
function for a couple of rudimentary flow fields, for exam-
ple, uniform flow, source/sink, doublet flow, and vortex; they Most of the problems of engineering importance, we are inter-
can be superimposed to get complex flow fields of practical ested in analyzing a low-speed flow over the aircraft wings
importance. and other objects of finite dimensions. Whenever, the flow
past a solid surface, the viscous effects are confined within
a thin layer adjacent to the surface, which is known as the
4.1 Introduction boundary layer. However, a large portion of the flow out-
side the boundary layer, is essentially inviscid and irrotational
In a flow field, it is a matter of interest to know the veloc- (i.e., fluid elements are not rotating while in the translational
ity and the pressure at each location in the flow domain. →In motion). An incompressible flow which is inviscid and irrota-
Cartesian coordinate system, the velocity vector field − v tional is called the potential flow. Note that, the irrotationality
has three velocity components u, v, and w along the three is a valid approximation for several flow field applications.
mutually perpendicular directions x, y, and z, respectively. To Recall, the irrotationality condition of a potential flow, given
calculate four different variables (three velocity components by Eq. (3.158), is
and the pressure), we actually need four different equations.
−
→ −
The classical approach in doing so is to solve the Navier– ∇ ×→
v=0 (4.1)
Stokes equations, which is a vector equation and comprises
of three different scalar equations, and the conservation of i.e., the curl of velocity vector field is zero. From the vector
mass equation. This approach, though, looks simple is a cum- calculus, it is also understood that the curl of the gradient of
bersome task in reality primarily because the Navier–Stokes a scalar is always vanished and thus the velocity field can be
equations are nonlinear. The approach, however, can be made represented as the gradient of a scalar function φ, known as
easier if we place a large number of constraints on the system the velocity potential. Recall, Eq. (3.160)
(such as assuming a low Reynolds number flow) and trans-
form the nonlinear Navier–Stokes equations into relatively −
→ −
→
v= ∇φ
simple and linear Stokes equation, which is comparatively
easy to solve for the velocity field and the pressure. More- But, Eq. (3.69) shows the incompressibility condition as
over, in some cases, the solutions to these Stokes equation
−
→−
∇ .→
v=0
108 4 Potential Flow Theory
Therefore, for the flows which are both incompressible and ∂v ∂u

− =0 (4.10)
irrotational, Eqs. (3.160) and (3.69) can be combined to yield ∂x ∂y
→ −
− → Introducing Eqs. (3.177) and (3.178) into Eq. (4.10), we get
∇. ∇φ =0

∂ ∂ψ ∂ ∂ψ
or − − =0 (4.11)
∂x ∂x ∂y ∂y
∇ φ=0
2
(4.2)
or
where ∇ 2 is the Laplacian operator. Equation (4.2) is called ∂2ψ ∂2ψ
+ =0 (4.12)
the Laplace’s equation in terms of velocity potential. This ∂x2 ∂y2
equation is usually mentioned in the following three coordi-
nate systems. or
Cartesian Coordinates: ∇ 2 ψ= 0 (4.13)
φ = φ (x, y, z) (4.3) Equation (4.13) is again the Laplace’s equation in terms of

∂2φ ∂2φ ∂2φ stream function. Hence, the stream function and the potential
∇2φ = + + =0 (4.4) function both satisfies the Laplace’s equation. From Eqs. (4.2)
∂x2 ∂y2 ∂z2
and (4.13), we have the following important conclusions:
Cylindrical Coordinates: 1. For an irrotational, inviscid, and two-dimensional incom-
pressible flow both stream function and velocity potential
φ = φ (r, θ, z) (4.5)
exist that satisfies the Laplace’s equation. Conversely, any
1 ∂ ∂φ 1 ∂2φ ∂2φ solution of the Laplace’s equation represents the velocity
∇2φ = r + 2 2 + 2 =0 (4.6)
r ∂r ∂r r ∂θ ∂z potential and the stream function for an irrotational, invis-
cid, and two-dimensional incompressible flow.
Spherical Coordinates: 2. It is a linear elliptic equation which allows the law of super-
position of its solutions.
φ = φ (r, θ, ) (4.7) 3. The solutions that obey the Laplace’s equation are called

1 ∂ ∂φ ∂ ∂φ ∂ 1 ∂φ harmonic functions. Thus, both the stream function (ψ)
∇ φ= 2
2
r sin θ
2
+ sin θ + =0
r sin θ ∂r ∂r ∂θ ∂θ ∂ sin θ ∂ and the velocity potential (φ) are the harmonic functions.
(4.8) 4. The Laplace’s equation can be viewed as the special case
of Helmholtz differential equation
We have seen earlier that in a two-dimensional incompressible
flow the velocity components in terms of the stream function ∇ 2 φ + k2 φ = 0 (4.14)
is given by Eqs. (3.177) and (3.178) as
with k = 0. Furthermore, it is also a special case of Pois-
∂ψ son’s equation
u=
∂y
∂ψ ∇ 2 ψ = −4πρ (4.15)
v=−
∂x
with ρ = 0.
Introducing the above velocity components into Eq. (3.69),
we get

∂ ∂ψ ∂ ∂ψ ∂2ψ ∂2ψ 4.3 Standard Solutions of the Potential Flow
+ − = − =0 (4.9)
∂x ∂y ∂y ∂x ∂x∂y ∂y∂x
In general, there exists four fundamental solutions (two-
i.e., the stream function ψ identically satisfies the continuity dimensional elementary flows) of the Laplace’s equation,
equation. In addition, the existence of ψ inherently reflects combining which all other steady flow conditions can be mod-
that the flow obeys the conservation of mass principle, and eled. They are uniform potential flow, source (or sink) flow,
hence Eqs. (3.177) and (3.178) can be used instead of using the doublet potential, and vortex flow. We will now examine these
continuity equation. Furthermore, if the flow is both incom- basic flows in the following sections.
pressible and irrotational, then from Eq. (3.158) we write
4.3 Standard Solutions of the Potential Flow 109
4.3.1 Uniform Potential Flow This is the velocity potential for an uniform potential flow.
Reader should note also that while deriving Eq. (4.23) we
We have seen earlier that in an uniform flow, the flow param- have not stipulated anything about the fluid density. That is,
eters do not change with respect to space coordinates, but can whether the flow is incompressible or compressible and hence
change with respect to time (t). In other words, if the parame- Eq. (4.23) is valid in either case. Further, from Eqs. (3.162)
ters have the same value at all points at a particular time instant and (3.163), we obtain the velocity components as
the flow is referred to as uniform flow. Moreover, each fluid
particle moves with constant speed so as the cross-section of ∂φ
u= = Ua (4.24)
each streamtube remains unchanged. ∂x
Consider an uniform flow of velocity Ua along x-direction ∂φ
v= =0 (4.25)
as illustrated in Fig. 4.1. It can be shown that an uniformflow ∂y
−
→→
is feasible only if the flow is incompressible ∇ .− v = 0 and
−→ − Let us further assume the flow to be incompressible and two-
irrotational ∇ × → v = 0 . Clearly, an uniform flow quali- dimensional so that the stream function (ψ) can be defined
fies to be called the potential flow which, in turn, allows to for this flow. Thus, from Cauchy–Riemann Eqs. (3.214) and
have the potential function (φuniform ). Recall, Eqs. (3.162) and (3.215), we write
(3.163)
∂ψ
∂φ = u = Ua (4.26)
u= ∂y
∂x ∂ψ
∂φ − = v=0 (4.27)
v= ∂x
∂y
Integrating Eq. (4.26) with respect to y and Eq. (4.27) with
From Fig. 4.1, we write respect to x, we get
u = Ua (4.16)
ψ = Ua y + f (x) (4.28)
v=0 (4.17)
ψ = C + g (y) (4.29)
Integrating Eq. (4.16) with respect to x, and Eq. (4.17) with
From similar reasons as given for the potential function, the
respect to y, we obtain
stream function can be written as
φuniform = Ua x + f (y) (4.18)
ψ= Ua y (4.30)
φuniform = C + g (x) (4.19)
Equation (4.30) is the stream function for an incompressible
where f (y) is a function of y only, g (x) is the function of x uniform flow. By setting this stream function equal to a con-
only, and C is a constant. Furthermore, Eqs. (4.18) and (4.19) stant, i.e., Ua y = constant, we get the equation of the stream-
show the same potential function (φuniform ) and thus by com- lines, sketched as the straight lines along the x-direction, are
parison it can be concluded that shown in Fig. 4.1. Similarly, Ua x = constant is the equation
of equipotential lines parallel to y-axis, depicted as verti-
g (x) = Ua x (4.20) cal straight lines in Fig. 4.1. Clearly, these streamlines and
f (y) = constant (4.21) equipotential lines are orthogonal to each other.
Velocity Field in the Polar Coordinates
Thus, φuniform becomes
The potential function (φ) and stream function (ψ) can also
φuniform = Ua x + C (4.22) be expressed in terms of polar (r, θ) coordinates. Substituting
x = r cos θ and y = r sin θ into Eqs. (4.23) and (4.30), we get
In aerodynamic applications, one is more interested in obtain-
ing the velocity field rather than knowing the actual value of φ = Ua r cos θ (4.31)
the potential function. Moreover, the derivative of a constant ψ = Ua r sin θ (4.32)
is always zero and thus dropping C in Eq. (4.22), we obtain
In addition, the radial and tangential components of velocities
φuniform = Ua x (4.23) can be calculated using Eqs. (3.185) and (3.186) as
Curve ’C’
h y
l
U a r
θ
x
ψ= constant
φ= constant
Fig. 4.1 Schematic diagram of the uniform flow
−
→ −
1 ∂ψ =− ∇ ×→
v .dS
vr = = Ua cos θ (4.33)
r ∂θ s
∂ψ
vθ = − = −Ua sin θ (4.34) −
→ →
∂r Clearly, for an irrotational flow if we introduce ∇ × − v =0
into the above equation, we get = 0 everywhere. Moreover,
Circulation it is also observed that both Eqs. (4.23) and (4.30) satisfy the
We will now calculate the circulation in an uniform potential Laplace’s equation. This is an important conclusion which
flow. Recall, Eq. (3.154) reveals the uniform flow as a fundamental solution of the
Laplace’s equation.
ffi
−
→ −
→
=− v . ds
c 4.3.2 Line Source (or Line Sink)
Consider the closed curve C shown by dashed lines in Fig. 4.1,
The two-dimensional radially symmetrical flow of an ideal
where h and l, respectively, are the length and width of the
fluid, emanating outward from a common point, is called a
curve. From Eq. (3.154), we write
simple source, as depicted in Fig. 4.2. In cylindrical coor-
ffi dinates, a source at the origin can be considered as a line
−
→ −
→
v . ds = −Ua l − 0 (h) + Ua l + 0 (h) = 0 (4.35) perpendicular to the z-plane along which the fluid is being
c
emitted. For fluid flows, a sink is a negative source and is a
or point of inward radial flow at which the fluid is considered
=0 (4.36) to be absorbed or annihilated (Fig. 4.2). Moreover, the source
(or sink) which produces a planar flow is called a line source
i.e. for the uniform flow the circulation around a rectangular (or line sink) since it must extend indefinitely in a line per-
closed curve is always zero. Although, this result is obtained pendicular to the plane of the flow. However, when a source
for a rectangular curve, it is true for any arbitrary closed curve. (or sink) produces the three-dimensional flow, it is called a
Note that, Ua is constant in both magnitude and direction, and point source (or point sink).
hence Consider the two-dimensional, incompressible fluid flow
ffi ffi from a source, where all the streamlines are straight lines ema-
−
→
=− − →v . ds = −Ua ds = Ua × 0 = 0 nating from the origin O as shown in Fig. 4.2a. Let the velocity
c c along each streamline (i.e., radial component of velocity) vary
inversely with the distance from O, such that, we write
This result is indeed consistent with our irrotational flow
assumption. From Eq. (3.155), the circulation can also be c
vr = (4.37)
defined as r
ψ = constant
φ = constant
ψ = constant − Vr
r
Vr
r φ = constant
O θ O θ
Discrete
point source Discrete
point sink
Source Flow Sink Flow

Fig. 4.2 Schematic diagram of a line source and line sink
where c is a constant. Further, Fig. 4.2 shows that the velocity Integrating Eq. (4.41) with respect to r, and Eq. (4.42) with
components in radial and tangential directions are vr and vθ , respect to θ, we get
respectively, where vθ = 0.
q
Reader should bear in mind that the source O, depicted in φ= ln r + f (θ) (4.43)
Fig. 4.2, is just a part of an entire line of sources along the 2π
z-axis, and the two-dimensional flow shown in the figure is φ = constant + f (r) (4.44)
the same in any plane perpendicular to the z-axis, i.e., for any q
plane given by z = constant. To derive the exact expression Comparing the above equations, we obtain f (r) = 2π ln r, and
for Eq. (4.37), let us assume that q is volume flow rate per unit f (θ) = constant. Thus, the potential function (φ) can be given
span across a circle of radius r as shown in Fig. 4.3. Thus, as
q
ˆ2π ˆ2π ˆ2π φ= ln r + constant (4.45)
−
→ c 2π
q= v .n̂dA = vr rdθ = rdθ = 2πc (4.38)
r As stated earlier, we are more interested in obtaining the
0 0 0
q velocity field rather than knowing the actual value of φ and
⇒c= (4.39) hence the constant in Eq. (4.45) can be dropped without the
2π
loss of generality. Therefore, for a two-dimensional incom-
Thus, Eq. (4.37) becomes pressible flow field from a line source the potential function
will become
q q
vr = (4.40) φ= ln r (4.46)
2πr 2π
In Eq. (4.40), if q is negative then the flow will direct inward
Stream Function
(toward the origin O). This type of flow is called a sink flow.
If the flow is also irrotational, then the stream function (ψ) is
Velocity Potential
obtained from Eqs. (3.214) and (3.215) as follows:
From Eqs. (3.218) and (3.219), the radial and tangential veloc-
ities in terms of velocity potential for this source is written 1 ∂ψ q
= vr = (4.47)
as r ∂θ 2πr
∂φ q and
= vr = (4.41)
∂r 2πr
∂ψ
1 ∂φ − = vθ = 0 (4.48)
= vθ = 0 (4.42) ∂r
r ∂θ
Vθ Vr
r
θ
x
O
Fig. 4.3 Volume flow rate from a line source
P P
Δθ
r1 r1 r2
r2 dθ
q −q q −q
θ1 θ2 d
θ
l Source l Sink
Source Sink
(a) Source−Sink Pair (b) Doublet − Limiting Case
Fig. 4.4 Schematic diagram of a the source–sink combination, and b the doublet
Integrating Eq. (4.47) with respect to θ, and Eq. (4.48) with at origin O. Clearly, these streamlines and equipotential lines
respect to r, we obtain will be orthogonal to each other. Interestingly, these results
q are consistent with the diagram of source flow as shown in
ψ= θ + f (r) (4.49) Fig. 4.2.
2π
ψ = constant + f (θ) (4.50)
4.3.3 Source–Sink Combination and the
Comparing the above equations and dropping the constant,
Doublet Potential
we finally have
q
ψ= θ (4.51)
2π Consider the case where a source of strength q is placed near to
a sink of equal and opposite strength −q, as shown in Fig. 4.4.
Equation (4.51) is the stream function for a two-dimensional Suppose l is the distance by which the source (So ) is separated
source flow. The equation of the streamlines can be obtained from the sink (Si ). Let us further assume that the line So P is
by setting Eq. (4.51) equal to a constant. That is, subtending an angle θ1 and the line Si P is subtending an angle
q θ2 with the line joining the source and the sink as shown in
ψ= θ = constant (4.52) Fig. 4.4a.
2π
Since in an incompressible and irrotational flow field, the
i.e., for the fixed source strength we have θ = constant, stream functions at an arbitrary point P due to the source and
q q
which represents the equation of a streamline in polar coor- the sink are 2π θ1 and 2π θ2 , respectively. Thus, from the law
dinates. In addition, Eq. (4.46) gives an equipotential line as of superposition the overall stream function (ψo ) at P for this
r = constant, which is the equation of a circle with its center combination is written as
Equipotential lines
Streamlines
Source Sink
Fig. 4.5 Streamlines and equipotential lines for a source and sink combination
q q d
ψo = (θ1 − θ2 ) = − θ (4.53) dθ = (4.57)
2π 2π r2
where θ = θ2 − θ1 , as shown in Fig. 4.4a. Similarly, the but
overall potential function (φo ) for this source–sink combina-
tion at P will be d = l sin θ (4.58)
r2 = r1 − l cos θ (4.59)
q q q r1
φo = ln r1 − ln r2 = ln (4.54)
2π 2π 2π r2 Substituting above equations into Eq. (4.57), we get
where r1 and r2 , respectively, are the distances of source d l sin θ
and sink from the point P. In addition, the streamlines and dθ = = (4.60)
r2 r1 − l cos θ
equipotential lines are found by choosing ψo = constant and
φo = constant as schematically drawn in Fig. 4.5. Thus, Eq. (4.56) becomes
We now examine the limiting case, let l → 0 while
the absolute magnitudes of source and sink strengths are q l sin θ
ψ = lim − (4.61)
increased in such a manner that the product ql remains l→0 2π r1 − l cos θ
invariant. Under limiting condition, this combination of
singularities is called the doublet. The strength of doublet is or
κ sin θ
denoted by κ and is defined as ψ = lim − (4.62)
l→0 2π r1 − l cos θ
κ= ql (4.55) or
κ sin θ
ψ=− (4.63)
2π r1
Stream Function
The stream function for a doublet of constant strength is But, for l → 0 we have r2 ≈ r1 . Dropping the subscript in the
obtained by calculating the value of Eq. (4.53) in the limit above relation, we write
l → 0 as κ sin θ
−q ψ= − (4.64)
ψd = lim dθ (4.56) 2π r
l→0 2π
This is the stream function for a two-dimensional doublet
Note that for l → 0, θ becomes dθ. Thus, from Fig. 4.4b we potential flow. Also, the equation of streamlines for a doublet
write flow are obtained by setting Eq. (4.64) equal to a constant.
y Streamlines
k
2π C 2
k
2π C 1
k
x
Equipotential Eccentric non−intersecting

lines circles
Fig. 4.6 Streamlines and equipotential lines for a doublet flow
That is, Equation (4.68) is again an equation of a circle of diame-

κ
κ 2πC
κ sin θ ter 2πC , whose center is located at a distance of 2 2
ψ=− = constant = C1 (say) (4.65) 2
2π r from the origin. That is, the equipotential lines due to a two-
dimensional doublet flow are the sets of eccentric and non-
or
κ intersecting circles with their centers located on the x-axis
r=− sin θ (4.66)
2πC1 (Fig. 4.6), where each circle corresponds to a different value
of the constant C2 . Interestingly, both the equipotential lines
From trigonometry recall that, in the polar coordinate system and the streamlines are orthogonal to each other as observed
the above equation essentially represents a circle of diame- in Fig. 4.6.
κ
− 2πC
κ
ter − 2πC 1
, whose center is located at 0, 2
1
. Hence,
the streamlines for a two-dimensional doublet are a family of 4.3.4 Line Vortex
eccentric and nonintersecting circles with their centers located
on the y-axis as shown in Fig. 4.6, where each circle corre- A two-dimensional, steady flow which circulates about a point
sponds to a different value of the constant C1 . is called the line vortex (Fig. 4.7). Since there is no radial com-
ponent of velocity in a vortex and thus in polar coordinates,
Potential Function we write
Following the similar steps, the velocity potential at point
P (x, y) for a doublet of constant strength is obtained as vr = 0 (4.69)
vθ = f (r) (4.70)
κ cos θ
φ= (4.67)
2π r It can be shown that only one function vθ (r),
− which satisfies
→ →
Further, the equation of equipotential line is obtained by the irrotationality condition ∇ × − v = 0 is
equating Eq. (4.67) to a constant. We write
K
vθ = (4.71)
κ cos θ r
φ= = C2 (constant)
2π r
where K is a constant. Equation (4.71) shows that the tangen-
or tial velocity (vθ ) is inversely proportional to the distance from
κ
r= cos θ (4.68) the center O. This is sometimes called a free vortex.
2πC2
vθ
Streamlines
O
Equipotential Lines
Fig. 4.7 Schematic diagram of the vortex flow
Stream Function and Velocity Potential Thus, the stream function of a vortex can be obtained from the
equation for the velocity potential for a source by replacing
Since the flow field is irrotational, thus the stream function
q by . Similarly, the velocity potential of a vortex from the
(ψ) may be obtained from Eqs. (3.214) and (3.215) as follows:
equation for the stream function of a source by replacing q
1 ∂ψ by . Therefore, we can say that a vortex is essentially the
vr = 0 = (4.72) reversed image of a source. The streamlines and equipotential
r ∂θ
K ∂ψ lines are obtained by setting Eqs. (4.79) and (4.80) equal to
vθ = = − (4.73) the constants as illustrated in Fig. 4.7. From the figure, it can
r ∂r
be seen that the streamlines are the circles of constant radius
Integrating above, we get (r) and the potential lines are radial spokes of constant θ.
ψ = −K ln r (4.74)
4.4 Superposition of Standard Solutions
Again from Eqs. (3.214) and (3.215), we obtain
From Eqs. (4.2) and (4.13), one should note that the Laplace’s
φ = Kθ (4.75) equation is a second-order linear partial differential equation.
The linearity of this equation allows solutions to be con-
Suppose the flow field has a constant circulation () in anti- structed from the superposition of simpler and elementary
clockwise direction and hence, from Eq. (3.154) solutions. For a two-dimensional incompressible and irrota-
tional flow, if ψ1 and ψ2 are
the solutions
(stream functions)
˛
of the Laplace’s equation ∇ 2 ψ = 0 then their linear com-
= v .d−
−
→ →
s = vθ (2πr) (4.76)
bination ψ1 + ψ2 will also be a solution. This is an unique
c feature that makes this equation a powerful tool for analyz-
or ing the fluid flow problems. It allows the complicated flow
pattern in an irrotational and incompressible flow to be mod-
K= (4.77) eled by adding together a number of elementary flows, which
2π
are also irrotational and incompressible. We will now look
Thus, the constant K is called the strength of the vortex. Sub- at some examples of superposition of flows in the following
stituting Eq. (4.77) into (4.73), the tangential component of sections.
velocity becomes

vθ = (4.78) 4.4.1 A Source in the Uniform Flow
2πr
(Axisymmetric Flow over a Semi-infinite
In addition, the stream function and the potential function for Body)
a line vortex flow can be obtained by introducing Eq. (4.77)
into Eqs. (4.74) and (4.75). Thus, A number of physically interesting flow fields can be obtained
by combining the flow fields of suitable distributions of

ψ= − ln r (4.79) sources and sinks with that of a uniform flow. A simple exam-
2π ple of such a combination is that of a single source with a uni-

φ= θ (4.80) form flow as shown in Fig. 4.8. Let the source strength be q
2π
Uniform Stream
R A To infinity
θ
B
(D) Source
C To infinity
Source
Fig. 4.8 Schematic diagram of a source in the uniform flow
and the velocity of the uniform freestream be Ua . The stream Stagnation Points
function for this combination can be written as
The stagnation points in the flow can be obtained by setting
Eqs. (4.84) and (4.85) equal to zero. That is,
ψ = ψuniform + ψsource (4.81)
q
Substituting for ψuniform and ψsource , respectively, from Ua cos θ + =0 (4.86)
2πr
Eqs. (4.32) and (4.51), we have Ua sin θ =0 (4.87)
q
ψ = Ua r sin θ + θ (4.82) Solving the above equations, we observe that only one stag-
2π
nation point
exists, which is located upstream of the source at
q
Note that both Eqs. (4.32) and (4.51) are the solutions of (r, θ) ≡ 2πU a
, π and it is denoted by point B in Fig. 4.8.
Laplace’s equation. Thus, from the law of superposition the Also, it can be seen that the distance DB decreases with
stream function given by Eq. (4.82) is also a solution of the increase of the freestream velocity (Ua ). Furthermore, sub-
Laplace’s equation. In other words, Eq. (4.82) describes a pos- stituting the coordinates of B into Eq. (4.83), we get
sible case of irrotational and incompressible flow. The stream-
lines for the combined flow are obtained by setting Eq. (4.82) q q
ψ = Ua sin π + π = constant (4.88)
equal to a constant. That is, 2πUa 2π
q or
ψ = Ua r sin θ + θ=C (4.83) q
2π ψ= = constant (4.89)
2
where C is an arbitrary constant. Using Eq. (4.83), the stream- i.e., the streamline that passes through the stagnation point is
line pattern is sketched in Fig. 4.8. Note that the source is described by ψ = q2 . This streamline is depicted by the curve
located at point D. ABC in Fig. 4.8. Since the flow field is inviscid, the velocity
Velocity Field at the surface of a solid body will always be tangent to the
body (i.e., the normal component of velocity is zero). Thus,
The radial and tangential components of velocities, obtained any streamline of the combined flow which passes through B
from Eqs. (3.185) and (3.186), are could be replaced by a solid surface of the same shape. Hence,
the streamline ABC, which contains the stagnation point at B
1 ∂ψ q
vr = = Ua cos θ + (4.84) is a dividing streamline; it separates the fluid coming from the
r ∂θ 2πr
freestream and the fluid emanating from the source at D. In
∂ψ
vθ = − = −Ua sin θ (4.85) other words, the uniform freestream is not able to penetrate
∂r
ABC, consequently, the flow inside ABC is unaffected by the
We have seen that the radial component of velocity due uniform flow. Therefore, the streamline ABC can be replaced
q by a solid wall of the same shape which extends downstream
to a line source is 2πr (Eq. (4.40)), and the component of
freestream velocity in radial direction is Ua cos θ (Eq. (4.33)). to the infinity, forming a shape of semi-infinite body which is
Interestingly, the direct sum of these two velocities is given by also known as Rankine’s half-body.
Eq. (4.84). This is an important conclusion worth to remem- The general equation for the streamline through the point
ber, that is, the law of superposition not only allows to add B is
elementary φ and ψ but also their derivatives (i.e., velocities) q q
ψ = Ua r sin θ + θ= (4.90)
as well, to construct more complex flows. 2π 2
4.4 Superposition of Standard Solutions 117
q
Note that, the radial distance R to any point on this streamline ψsource–sink = (θ1 − θ2 )
will be 2π
q (π − θ) where θ1 is the angle subtended by the line joining the source
R= (4.91)
2πUa sin θ and point P with the x-axis, and θ2 is the angle subtended
by the line joining the sink and point P with the x-axis
Moreover, we can see that as x → ∞, the streamline ABC (Fig. 4.9). Since the stream function for an uniform flow,
becomes parallel to the x-axis. Also, at sufficiently far away given by Eq. (4.32) is ψuniform = Ua r sin θ. Thus, Eq. (4.94)
location (i.e., at theoretical ∞) the perpendicular distance becomes
of ABC from the x-axis, is known as the maximum half-
q
width of the body. Note that, the perpendicular distance of ψo = Ua r sin θ + (θ1 − θ2 ) (4.95)
2π
this streamline is given by
From Eq. (4.95), the equation of streamlines will be
q (π − θ)
y = R sin θ = (4.92)
2πUa q
ψ = Ua r sin θ + (θ1 − θ2 ) = constant (4.96)
2π
But, as x → ∞, the angle θ → 0. Therefore,
Figure 4.9 shows the streamlines of the combined flow. Fur-
q thermore, the equation of streamline passing through the stag-
ymaximum = (4.93)
2Ua nation points is given by ψ = 0. Thus, from Eq. (4.96) we
write
From the above discussion, it can be concluded that the flow q
Ua r sin θ + (θ1 − θ2 ) = 0 (4.97)
past a semi-infinite body can be mathematically described as 2π
the combination of a source in an uniform freestream. Like-
Equation (4.97) is the equation of an oval, which is satisfied
wise, more complex flow fields can also be analyzed consid-
by the point A located at θ = θ1 = θ2 = π and the point B
ering the suitable combination of elementary solutions of the
located at θ = θ1 = θ2 = 0. Thus, A and B are the two stag-
Laplace’s equation. Some of these cases are discussed in the
nation points on the stagnation streamline (Fig. 4.9). Note
subsequent sections.
that the stagnation streamline (ψ = 0), shown as dark line
in Fig. 4.9, is a dividing streamline. All the streamlines inside
the oval originate at the source on the left, and flow into the
4.4.2 A Pair of Source and Sink in the Uniform
sink on the right. The net volume outflow from the oval is
Flow (Axisymmetric Flow over a Closed
zero. Again, this dividing streamline could be replaced by a
Body)
solid oval body of the same shape. The flow outside the oval,
In the previous section, we have seen that a combination of interpreted as the inviscid, incompressible, and irrotational,
source and uniform freestream actually represents the flow then corresponds to the flow past this body. Since this prob-
past a semi-infinite body. Thus, it is a half-body that stretches lem was first solved in the nineteenth century by the famous
to infinity in the downstream direction. Clearly, this body is Scottish engineer W.J.M. Rankine; hence, the shape given by
not closed at the downstream end. However, if a sink of equal Eq. (4.97) and shown in Fig. 4.9 is called a Rankine oval.
strength as the source is placed at some distance downstream
of the source, then the resulting body shape will be closed. 4.4.3 A Doublet in the Uniform Flow (Flow over
The purpose of this section is to examine this fact through the a Circular Cylinder Without Circulation)
combination of a pair of source and sink and an uniform flow.
We now superimpose an uniform freestream of velocity Ua Let us now superimpose a uniform flow with a doublet, as
in the x-direction, with a line source located at (−b, 0) and depicted in Fig. 4.10. Thus, adding the stream functions of
a line sink of equal and opposite strength located at (+b, 0), the uniform flow and the doublet flow given by Eqs. (4.32)
as shown in Fig. 4.9. Suppose the strengths of these source and (4.64), respectively, we have
and the sink are +q and −q, respectively. The overall stream
function (ψo ) for this combination of flows is written as k sin θ
ψ = Ua r sin θ − (4.98)
2π r
ψo = ψuniform + ψsource–sink (4.94) k
= Ua r sin θ 1 − (4.99)
2πUa r 2
The stream function for a pair of source and sink is
or
Stagnation streamline
U a
r r
1 r 2
A θ1 θ θ2 B
Source O Sink
b b
Rankine Oval
Fig. 4.9 Schematic diagram of a pair of source and sink in the uniform flow
R r
k θ
Uniform flow Doublet Non−lifting flow over a cylinder
Fig. 4.10 Schematic diagram depicting the non-lifting flow over a circular cylinder

R2 But, over the surface of the cylinder, where r = R, we have
ψ = Ua r sin θ 1 − 2 (4.100)
r
vr = 0 (4.103)
where R2 = k
2πUaand k is strength of the doublet. This com- vθ = −2Ua sin θ (4.104)
bination of flows basically corresponds to an uniform flow
past a stationary circular cylinder of radius R. Clearly, the maximum flow speed of 2Ua occurs over the
Surface Flow Velocities and the Pressure cylindrical surface at the location, θ = ±90◦ . Furthermore,
from the incompressible Bernoulli’s equation the static pres-
The radial and tangential components of velocities can be sure distribution over the cylinder is written as
obtained from Eqs. (3.185) and (3.186) as
1
1 ∂ψ R2 p (θ) = p0 − ρ vr2 + vθ2 (4.105)
vr = = Ua cos θ 1 − 2 (4.101) 2
r ∂θ r
Substituting vr and vθ into the above equation and using
∂ψ R2 the
vθ = − = −Ua sin θ 1 + 2 (4.102) freestream value for the total pressure p0 = pa + 21 ρUa2 , we
∂r r obtain
1
p (θ) = pa + ρUa2 1 − 4 sin2 θ (4.106)
2
Cp
+1
0 θ
π π 3π
−1
2 2
−2
−3
Fig. 4.11 A typical plot between Cp and θ for a doublet in the uniform flow
In addition, the corresponding pressure coefficient will be 4.4.5 A Doublet and a Vortex Flow in the
given by Uniform Flow (Flow Past a Circular
p (θ) − pa Cylinder with Circulation)
Cp (θ) = 1 (4.107)
2 ρUa
2
The flow past a rotating circular cylinder can be mathemati-
or cally represented as the combination of a doublet potential, a
Cp (θ) = 1 − 4 sin2 θ (4.108) vortex flow, and an uniform flow. That is, the stream function
of this combination of potential flows can be given as
This is the pressure coefficient for the flow past a nonrotating
circular cylinder. A typical variation of Cp with θ is shown in ψ= ψuniform + ψdoublet + ψvortex (4.114)
Fig. 4.11.
Note that the superposition of an uniform flow and a doublet
potential is essentially shows the flow past a nonrotating cir-
4.4.4 A Point Vortex in the Uniform Flow cular cylinder, which is given by Eq. (4.100). Thus, the overall
stream function for the flow past a rotating circular cylinder
Let us superimpose a vortex flow with the uniform flow as can be written as
schematically shown in Fig. 4.12. Mathematically, this com- ψ= ψNon-lifting cylinder + ψvortex (4.115)
bination can be expressed as
From Eqs. (4.100) and (4.74), we write

ψ = Ua r sin θ + ln r (4.109)
2π R2
ψ = Ua r sin θ 1 − 2 + ln r (4.116)
r 2π
Again, the radial and tangential components of velocities can
be obtained using Eqs. (3.185) and (3.186) as
where R2 = 2πU k
a
and k is strength of the doublet. From
1 ∂ψ Figs. 4.10 and 4.13, it is also evident that a vortex essentially
vr = = Ua cos θ (4.110) introduces asymmetry in the streamline pattern.
r ∂θ
∂ψ Surface Velocities
vθ = − = −Ua sin θ − (4.111)
∂r 2πr
The radial and tangential components of velocities is obtained
The location of the stagnation points can be obtained by equat- from Eqs. (3.185) and (3.186) as follows:
ing Eq. (4.110) and Eq. (4.111) to zero. That is,
1 ∂ψ R2
vr = = Ua cos θ 1 − 2 (4.117)
vr = Ua cos θ = 0 (4.112) r ∂θ r

∂ψ R2
vθ = −Ua sin θ − =0 (4.113) vθ = − = −Ua sin θ 1 + 2 − (4.118)
2πR ∂r r 2πr
Clearly, we get the stagnation point in the flow field at the We can see that the tangential velocity component depends
location, where r = R and θ = ±90◦ along with the circula- on the circulation (). Moreover, on surface of the cylinder
tion = 2πRUa . where r = R, we have
Uniform flow Vortex flow Resulting flow
Fig. 4.12 The superimposition of a vortex with the uniform flow
Non−lifting flow over a cylinder Vortex flow Lifting flow over a cylinder
R r θ R r θ
Fig. 4.13 Schematic diagram depicting the lifting flow over a circular cylinder
vr = 0 (4.119) • If < 4πUa R, we obtain two stagnation points S1 and S2

on the surface of the cylinder (Fig.
4.15), at β and π − β,
vθ = −2Ua sin θ − (4.120)
2πR respectively, where β = arcsin 4πU aR
.
• If = 4πUa R, both S1 and S2 coincide at π2 or − π2 and
Since the radial component of velocity is zero and thus the tan-
we get only one stagnation point S on the surface as shown
gential component of velocity at the cylinder surface will be
in Fig. 4.16.
given by vθ,s alone. Thus, the streamline pattern for this flow
• If > 4πUa R, the stagnation point will no longer be
depend upon the location of the stagnation points vθ,s = 0
located on the surface, but will appear in the flow (point
given by
S ) as shown in Fig. 4.17.

sin β = (4.121)
4πUa R Coefficient of Pressure
From the incompressible Bernoulli’s equation, the static pres-
Thus, the velocity at the surface vθ,s can be written as
sure distribution over the cylinder is written as
vθ,s = −2Ua sin θ + 2Ua sin β (4.122) 1
p (θ) = p0 − ρ vr2 + vθ2 (4.123)
2
As evident from Eq. (4.120), the location of stagnation points
is a strong function of the circulation (). Thus, consider the Introducing vr and vθ into above and from the freestream
following cases: value for the total pressure p0 = pa + 21 ρUa2 , we get
• If = 0, the stagnation points lie at θ = 0 (point A) and 1

θ = 2π (point B) as illustrated in Fig. 4.14. p (θ) = pa + ρUa2 1 − 4 sin2 θ (4.124)
2
A B
Fig. 4.14 Two stagnation points A and B for = 0
C
S 1 S 2
Fig. 4.15 Two stagnation points S1 and S2 for < 4πUa R
C
S
Fig. 4.16 One stagnation point S for = 4πUa R
C
S’
Fig. 4.17 Stagnation point located at S for > 4πUa R

Furthermore, the corresponding pressure coefficient calcu- dA= Rdθ (4.133)

lated from Eq. (4.107), will be
Thus, drag coefficient becomes
p (θ) − pa
Cp (θ) = ˆ2π
2 ρUa
1 2
1
CD = − Cp cos θdθ (4.134)
2
or 0
ˆ2π 2
2 1 2
2 = −1 + 4 sin θ +2
+ sin θ cos θdθ
Cp (θ) = 1 − 4 sin2 θ − − sin θ 2 2πUa R πUa R
2πUa R πUa R 0
(4.125) (4.135)
Forces on the Cylinder Upon integration, we get

The resultant force (per unit span) is obtained by integrating
the pressure forces over the surface of the cylinder of radius R. CD = 0 (4.136)
−
→
Let F be the net force per unit span and D and L, respectively,
Similarly, the expression of lift coefficient will be
are the drag (per unit span) and lift (per unit span) acting on
the cylinder. In vector notation, we write ˆ2π
1
CL = − Cp sin θdθ (4.137)
−
→ 2
F = Dî+Lĵ (4.126) 0
ˆ2π 2
1 2
where î and ĵ are the unit vectors along x- and y-directions, = −1 + 4 sin2 θ + + sin θ sin θdθ
2 2πUa R πUa R
respectively. The resultant force is calculated from the relation 0
(4.138)
given as
˛ ˛ On solving,
−
→
F =− pn̂dA = − (p − pa )n̂dA (4.127)

CL = (4.139)
One should note that, the freestream pressure (pa ) is a con- Ua R
stant and thus its subtraction from the surface static pressure From Eqs. (4.136) and (4.139), we obtain the lift per unit span
(p) does not change the result of the integration. This is in (L ) and the drag per unit span (D ) as
accordance with the following relation, which is valid for any
closed surface. ˛ L = ρUa (4.140)
(constant)n̂dA = 0 (4.128)
how? D =0 (4.141)
Bifurcating Eq. (4.127) into its components

1 2 along x- and y- i.e., the drag on a rotating circular cylinder is zero! But, we
directions, and dividing them by 2 ρUa 2R, we get the fol- know that all the real objects experience a finite amount of
lowing expressions for the drag and lift coefficients. drag in an uniform flow. So where the discrepancy lies? To
˛ answer this, one should understand that the total drag on a
1 body is an outcome of two aspects; drag due to pressure and
CD = − Cp nx dA (4.129)
2R drag due to viscosity. Equation (4.141) has considered only
˛ the drag due to pressure and ignored the viscous drag com-
1
CL = − Cp ny dA (4.130) ponent. Moreover, the axisymmetry of the circular cylinder
2R eventually leads to a zero pressure drag on a rotating circu-
From the Fig. 4.18, one can write lar cylinder (d’Alembert’s paradox). Furthermore, the lift on
the circular cylinder expressed by Eq. (4.140), is known as
nx = cos θ (4.131) the Kutta–Joukowski Theorem, which is essentially valid for
a two-dimensional body of any shape, not just for a circular
ny = sin θ (4.132) cylinder.
Also, the infinitesimal area dA on the circular cylinder can be

written as
4.5 Implications of Kutta–Joukowski Theorem in the Lift Generation 123
y ny
n
dθ nx
θ
x
R
Fig. 4.18 Schematic diagram of a circular cylinder in polar (r, θ) coordinates
4.5 Implications of Kutta–Joukowski the circulation in Kutta–Joukowski theorem is obtained from

Theorem in the Lift Generation the same pressure distribution.
In Sect. 4.4.5, the result given by Eq. (4.140) is derived for

a circular cylinder, however, in reality it is applicable for Summary
any cylindrical bodies of a general cross-section. Indeed, the
→
In Cartesian coordinate system, the velocity vector field −
Kutta–Joukowski theorem states that the lift per unit span on
v
a two-dimensional body is directly proportional to the circu-
has three velocity components u, v, and w along the three
lation around the body.
mutually perpendicular directions: x, y, and z, respectively. To
Moreover, we have also seen that the lifting flow over a
calculate four different variables (three velocity components
circular cylinder can be mathematically depicted as the super-
and the pressure), we actually need four different equations.
position of a uniform flow, a doublet, and a vortex. Among
The classical approach in doing so is to solve the Navier–
them, the uniform flow and the doublet are irrotational at all
Stokes equations, which is a vector equation and comprises
the points in the flow field, while, the vortex contains infi-
of three different scalar equations, and the conservation of
nite vorticity at the origin. If we choose an arbitrary curve
mass equation. This approach, though, looks simple is a cum-
which does not enclose the origin, then Eq. (4.140) will lead
bersome task in reality primarily because the Navier–Stokes
to = 0. However, if the curve encloses the origin as well
equations are nonlinear. The approach, however, can be made
then will be finite, equal to the strength of the vortex. The
easier if we place a large number of constraints on the system
same conclusion can be drawn for the flow around an airfoil.
(such as assuming a low Reynolds number flow) and trans-
To understand this concept further, consider an incompress-
form the nonlinear Navier–Stokes equations into relatively
ible flow around an airfoil section, as shown in Fig. 4.19. If
simple and linear Stokes equation, which is comparatively
a curve C1 is chosen such that it encloses the airfoil then the
easy to solve for the velocity field and the pressure. More-
circulation per unit span will be finite, given by Eq. (4.140).
over, in some cases the solutions to these Stokes equation
However, if any other curve, for example, C2 is drawn not
can still be complicated and thus we want even more simpler
enclosing the airfoil, then will again be zero. This may
equations to solve. It is achieved by applying an additional
be because the flow field external to an airfoil is irrotational
kinematic constraint on the flow (i.e., assuming an irrotational
and hence the circulation around the curve not enclosing the
flow field), which leads to a much simpler equation that can
airfoil is consequently zero. Likewise, it will be shown later
be solved directly to obtain the velocity. Such type of flow is
that the flow past an airfoil section may also be mathemati-
known as the potential flow and resulting equation is called
cally described by distributing the vortices either on the air-
the Laplace’s equation.
foil surface or interior to it. Again these vortices have nonzero
vorticity content and any selected curve that encloses the air- ∇ 2φ = 0
foil will demonstrate a finite , equal to the total strengths
of the vortices allotted on the surface or interior to the air- In general, there exists four fundamental solutions (two-
foil. dimensional elementary flows) of the Laplace’s equation,
Note that, to apply the Kutta–Joukowski theorem the curve combining which all other steady flow conditions can be mod-
around the body can be of any shape and size, but it must be eled. They are, uniform potential flow, source (or sink) flow,
enclosing the body completely. Also note that, the lift on the doublet potential, and vortex flow.
airfoil is of course produced due to pressure and shear stress The source is a potential flow field in which flow emanating
distributions over the airfoil surface only and the circulation from a point spreads radially outwards. In contrast, a sink is
theory is not at all different from these distributions. Rather, the potential flow field in which the flow is directed toward
Finite circulation
Ua
Ua
Vorticity distribution along the

surface of an arbitrary airfoil
C1
Ua
C2
Zero circulation
Vorticity distribution along mean camber line
Fig. 4.19 Schematic representation of the circulation around an airfoil
a point from all the directions. The velocity potential for a The Laplace’s equation is a second-order linear partial differ-
two-dimensional source of strength q is given as ential equation; its linearity allows solutions to be constructed
from the superposition of simpler and elementary solutions.
q
φ= ln r For a two-dimensional incompressible and irrotational flow, if
2π ψ1 and ψ2are the solutions (stream functions) of the Laplace’s

Similarly, the stream function for a source flow is calculated equation ∇ 2 ψ = 0 then their linear combination ψ1 + ψ2
as will also be a solution.
The stream function for the flow due to the combination
q of a source of strength q at the origin, in an uniform flow of
ψ= θ
2π velocity Ua in x-direction will be
A doublet flow is the potential flow field formed when a source q
ψ = Ua r sin θ + θ
and a sink of equal strengths are placed close to each other in 2π
such a way that the product of their strength and the distance The streamline passing through the stagnation point S is
between them remain constant. The velocity potential for a termed as stagnation streamline forming a shape of semi-
doublet flow is given by infinite body which is also known as Rankine’s half-body.
The Kutta–Joukowski theorem states that the lift per unit
κ cos θ
φ= span on a two-dimensional body is directly proportional to the
2π r circulation around the body. To apply this theorem, the curve
Likewise, the stream function for a doublet flow is obtained around the body can be of any shape and size, but it must be
as enclosing the body completely. Also note that, the lift on the
airfoil is of course produced due to pressure and shear stress
κ sin θ
ψ= − distributions over the airfoil surface only and the circulation
2π r theory is not at all different from these distributions. Rather,
A line vortex is a two-dimensional steady flow, which circu- the circulation in Kutta–Joukowski theorem is obtained from
lates about a point. The velocity potential for a line vortex is the same pressure distribution.
given as
Exercises
φ= θ
2π
In a similar manner as above, the stream function for a line
vortex is calculated as 1. Express the equations for an ideal fluid in terms of stream-
line coordinates.

ψ= − ln r 2. Show that the three-dimensional source flow is irrota-
2π
tional.
Exercises 125
3. Show that, the uniform flow is irrotational in nature and (c) a family of parabolas
it is physically possible for an incompressible flow. (d) a family of hyperbolas
4. A source of strength 25 m3 s−1 and a sink of equal strength
are placed 2 m apart on the x-axis. If an uniform flow of 4. Since the curl of the gradient of a scalar is always van-
velocity 50 ms−1 in negative x-direction is superimposed ished and thus for a fluid flow the velocity field can be
on this source–sink combination, then calculate the size represented as the gradient of a scalar function φ known as
−
→
of the Rankine oval body thus formed. the velocity potential −
→v = ∇ φ. Which of the following
5. What will be the shape of the streamlines, if the velocity is TRUE?
of the non-lifting flow over a circular cylinder is doubled? −
→−
6. Acquire the equation of the streamlines for the velocity (a) ∇ .→ v =0
−
→ −
field with the tangential and radial components of veloc- (b) ∇ ×→ v=0
ity are vθ = Ar and vr = 0, respectively, where A is a (c) ∇2φ = 0
constant. (d) all the above
velocity potentials φ1 = x − y and
7. Suppose two 2 2
5. The length and radius of the line source are 50 and 5 mm,
φ2 = r 2 cos 2θ are the solutions of the Laplace
1
respectively. For the line source, if the radial velocity is

equation. Prove that their linear combination, i.e.,
5 ms−1 the volume flow rate will be
φ3 = (φ1 + φ2 ) satisfies the Laplace equation.
8. What are the minimum and maximum values of the coef- (a) 3.25 × 10−3 m3 s−1
ficient of pressure on a nonrotating circular cylinder in an (b) 5.35 × 10−3 m3 s−1
uniform freestream when k = 4πUa . (c) 7.85 × 10−3 m3 s−1
9. Estimate the circulation of a circular cylinder with (d) 9.75 × 10−3 m3 s−1
radius 0.1 m is placed in an uniform flow of velocity
Ua = 3 ms−1 , which has the lift coefficient CL = 0.4. 6. For a two-dimensional irrotational flow, which of the fol-
10. Determine the radius of vortex flow having tangential lowing relation is correct?
velocity vθ = 16 ms−1 and circulation 8 m2 s−1 .
∂v ∂u
(a) + =0
∂x ∂y
Multiple Choice Questions ∂v ∂u
(b) − =0
∂x ∂y
1. Superposition of the stream functions of a uniform flow ∂u ∂v
(c) + =0
and a line source in a potential flow generates a dividing ∂x ∂y
streamline called as ∂u ∂v
(d) − =0
∂x ∂y
(a) rotating circular cylinder
(b) circular cylinder 7. The inherent feature of the free vortex flow streamlines is
(c) Rankine’s half-body
(d) Rankine’s oval (a) concentric
(b) sinusoidal
2. For a flow, if the potential function is given by φ = Cθ, (c) non-concentric
where C1 is an arbitrary constant. The stream function of (d) linear
the flow will be given as
8. The characteristic of ideal fluid flow will not be satisfied
C1 by
(a) r
C1
(b) ln r (a) sink
(c) C1 ln r (b) uniform flow
(d) C1 ln θ (c) superimposition of elementary flows
d) forced vortex
3. For a two-dimensional flow, if the velocity components
in x- and y-directions are given by u = ay and v = bx. In 9. The stream function (ψ) for a doublet flow is given by
this flow, the streamlines represent k sin θ
(a)
2π r
(a) a family of straight lines k sin θ
(b) a family of circles (b) −
2π r
2π sin θ 3. (d)
(c)
k r 4. (d)
k sin θ
(d) 5. (c)
4π r 6. (b)
10. The solutions of the Laplace’s equation are termed as 7. (a)
8. (d)
(a) harmonic functions 9. (b)
(b) Laplacian functions 10. (a)
(c) higher order functions
(d) exponential functions
Keys
1. (c)
2. (c)
Thin Airfoil Theory
5
Thin airfoil theory is a straightforward hypothesis of airfoils 5.2 Circulation and Vorticity
that relates angle of attack to lift for an incompressible and
inviscid flow past an airfoil. This theory idealizes the flow The circulation and vorticity are the two essential measures
past an airfoil as two-dimensional stream around a thin air- of rotation in a fluid. Circulation is a scalar quantity, obtained
foil which can be envisioned as tending to an airfoil of zero through the integration, which is a macroscopic measure of
thickness and infinite wingspan. rotation over a finite region in the fluid flow, whereas vorticity
is a vector field which provides a microscopic measure of the
rotation at any point in the fluid.
5.1 Introduction Circulation () is defined as the line integral of the tangen-
tial velocity around a closed curve C fixed in the flow field.
An aircraft moves in the air by overcoming the gravity with That is, ‰
−
→
a lifting force, provided by the aircraft’s wing. The cross- =− − →v . ds (5.1)
sectional geometry of the wing influences the flow of air and
c
the combined geometry of the wing and the reaction of the air
causes any general solution of the wing-sectional properties By convention, will be positive if the line integration is
to become too complicated, making it impossible to utilize carried out in counterclockwise direction; otherwise negative.
or almost difficult to ascertain. To deal with finding the flight Physically the circulation can be understood as the amount of
properties of wing sections, a more ameliorated way is to the force that pushes along a closed boundary or path. That is,
consider an inviscid and incompressible flow past the wing it is the total push what we get when traversing along a path
surface. A vortex superimposed on the airstream simulates such as a circle.
−
→
the process of lift generation by the wing section. The vortex The vorticity ζ is the tendency for fluid elements to
distribution along the wing will simulate the actual properties “spin”. It can be related to the amount of “circulation” or
of the wing and allow to have a simple approach of calculating “rotation” (more precisely local angular rate of circulation)
the properties of the wing. This hypothesis, better known as as
−
→
thin airfoil theory, was first conceived by Max Munk which ζ = 2−
→ω =∇ ×− →v (5.2)
was later refined by the team lead by Hermann Glauert in
1920s. This theory actually calculates the distribution of vor- where −→
ω is the angular velocity of rotation of the fluid ele-
tices which are compatible with the thin representation of an ment. Consider a rotational fluid element as shown in Fig. 5.1.
airfoil. This distribution can be used to find the lift, moment The circulation around the curve can be written as
and the pressure over the airfoil section. The fundamental

proposition of this theory is that an airfoil in the uniform flow ∂v ∂u
d = ux + v + x y − u + y x − vy
can be replaced by a vortex sheet along the chord line, and ∂x ∂y
the strength of this vortex sheet is determined by the condi-
∂v ∂u
tion that the camber line ought to likewise be a streamline. To = − xy (5.3)
∂x ∂y
compel the mean camber line to be a streamline, the aggregate
of all velocity components normal to the mean camber line Introducing Eq. (5.2) into Eq. (5.3), we get
must be equivalent to zero.
d = 2ωz xy (5.4)

128 5 Thin Airfoil Theory
y
u Δy
u+
y
c
Δy
v v
xΔx
v+
x
0 u Δx
Fig. 5.1 Schematic diagram of a rotational fluid element
Equation (5.4) is the product of angular velocity and area Note that the integration in the above equation is carried out
bounded by the curve. In vector notations, Eq. (5.4) becomes in counterclockwise direction to have positive . Since dn is
‰ ˆ ˆ arbitrarily small, Eq. (5.7) becomes
−
→ −
→ −→
= v . ds = 2ωz dA = ∇ ×−
→
v z .d A (5.5)
lim = lim (v1t − v2t ) dl (5.8)
c A A dn→0 dn→0
This is the well-known Stokes theorem, which relates the line From Eqs. (5.6) and (5.8) and dropping the subscript “t”, we
integral of the vorticity to its area integral. Therefore, the get
circulation around a closed contour can be viewed as the sum γ = v1 − v2 (5.9)
of vorticity enclosed by the contour.
This is the statement of Kutta condition in terms of the vortex
sheet distributed on a straight line.
5.3 The Kutta Condition Let us apply Eq. (5.9) for the flow over an airfoil, as shown
in Fig. 5.3. In this figure two types of airfoils are shown, a thick
In prior discourse, we have seen that for the flow past an airfoil with finite angle at the trailing edge (Fig. 5.3a), and a
airfoil at an angle of attack an endless number of potential thin airfoil with cusped trailing edge (Fig. 5.3b). Furthermore,
flow solutions are conceivable for different values of . Be considering again the philosophy of simulating the airfoil with
that as it may, it is likewise realized that the unique value of vortex sheet placed either on the surface or on the camber line,
lift on the airfoil is chosen by setting a specific value of . as sketched in Sect. 4.5. At the trailing edge of the airfoil,
Therefore, out of several values of , a specific value is picked Eq. (5.9) can be written as
up through an additional condition known as Kutta condition,
named after Wilhelm Kutta (1902) who first observed this. As γ (TE) = v1 − v2 (5.10)
per this condition, uniqueness of is guaranteed for the flow
over any airfoil section at an angle of attack if the flow leaves Figure 5.3a shows the airfoil that have a finite angle at the
the trailing edge smoothly. trailing edge. If v1 and v2 are the flow velocities at the top
In order to apply the Kutta condition in theoretical analysis, and bottom surfaces of the airfoil, they have two different
we should have a precise mathematical description about the directions at the trailing edge, i.e., at the same point. However,
nature of flow at the trailing edge of an airfoil section. We will physically this is not possible and the only solution is for both
first consider the distribution of vortices over a straight line v1 and v2 to be zero at the trailing edge. Consequently, the
forming the vortex sheet,1 as schematically shown in Fig. 5.2. trailing edge becomes the stagnation point and Eq. (5.10) will
The strength of such a vortex sheet is variable along the sheet lead to γ (TE) = 0. For the airfoil with cusped trailing edge,
and is denoted by γ, the circulation around the sheet will be as shown in Fig. 5.3b, both v1 and v2 are in the same direction
‰ and may be finite. But, from the Bernoulli’s equation, it is
−
→ −
→ revealed that for a single and unique value of pressure right
= v . ds = γdl (5.6)
c
at the trailing edge, both v1 and v2 are finite and equal in
magnitude and direction. From Eq. (5.10), we again obtain
or the result that γ (TE) = 0. In this manner, for the flow past an
= (v1t dl + v1n dn − v2t dl − v2n dn) (5.7) airfoil section, the Kutta condition can be expressed in terms
of the strength of vortex sheet as
1A vortex sheet is a surface across which there is a discontinuity in the
fluid velocity. For instance, a vortex sheet is formed by shearing activity γ (TE) = 0 (5.11)
between the layers of a fluid in motion.
5.3 The Kutta Condition 129
Vortex sheet
v 1t C
γ
v 1n v 2n dn
v 2t
dl
Fig. 5.2 Schematic diagram of a vortex sheet enclosed by a curve
(a) v1
v1
v2
Ua v1 = 0
v2 = 0
v1
v2 v2
(b)
v1
v1 v1
Ua v1 = v2
v2 v2
v2
Fig. 5.3 Flow past the trailing edges of a thick and the thin airfoils
Before we close this discussion, one more interesting obser- theorem,2 which states that a time rate of change of circulation
vation should be highlighted. In Sect. 4.5, we have quoted for a fluid line of fixed identity is governed by the torques
several times that the lift is produced due to both pressure and due to the forces acting on the line. Mathematically, we can
shear stress distribution on the body surface immersed in a express this theorem as
flow. However, we always felt that the lift is only an outcome j j j
D dp −
→− → μ 2− −
→
of distribution of pressure acting normal to the surface and =− + F . dr + ∇ →
v . dr (5.12)
the shear stress has virtually no role in lift generation. So, Dt ρ ρ
it is natural to think whether these statements are contradic- ı dp
tory in nature? In reality, they are not rather both pressure where the first term denotes the torque due to the pressure
ρ
as well as shear stress are responsible for lift on the airfoil. ı−→− →
forces, the second term F . dr represents the torque due to
ı −
→
the body forces, and the third term μρ ∇ 2 −→
Shear stresses are required to keep boundary layer attached to
v . dr is essentially
the surface throughout from the leading to the trailing edge,
the torque due to the viscous forces. The effects of these forces
thereby, it is ensured that the flow leaves the trailing edge
on the circulation are discussed as follows:
smoothly.
• The viscous forces, which are responsible for generating
the vorticity, changes the circulation as well. However, at
5.4 Kelvin’s Circulation Theorem the point when the fluids start rotating as if it were a solid
body, paradoxically the viscous forces are vanished alto-
We have observed that a right amount of circulation around
gether and there will be no force due to viscosity to make
an airfoil is required so that the flow leaves the trailing edge
further change in the circulation.
smoothly. Specifically, Kutta condition states that among the
infinite number of solutions possible for the potential flow
around an airfoil corresponding to each , only a single and 2 Kelvin’s circulation theorem does not hold under following conditions:
right value of ensures that the flow leaves the trailing edge
smoothly. Indeed, the optimum value of is selected based • For a fluid in motion under nonconservative body forces.
• For a non-barotropic fluid.
on a theorem commonly referred to as Kelvin’s circulation
• If the viscous stresses are present in the fluid flow.
−
→
• The body force F is also a circulation changing fac- where −→v is the tangential velocity vector and ds is an elemen-
tor. If the forces are irrotational, i.e., the body forces are tal length measured in counterclockwise direction around the
ı−→− → curve C. The governing equation (Euler equation) can be writ-
conservative, then F . dr = 0, whereas for the rotational
ten as
D−→
body forces (i.e., nonconservative) the pressure term is,
ı− →− → v 1−→ −
→
however, nonzero F . dr = 0 . =− ∇p+ ∇ (5.15)
Dt ρ
Furthermore, if the line of action of a body force passes
through the center of mass (i.e., center of gravity) of a fluid D
where Dt is the material or substantial derivative, ρ is the fluid
particle, it produces no torque on it, thus, no vorticity. The density, p is the pressure, and is the potential function under
centrally directed forces such as gravity are of this type, conservative body force.
that is, they are irrotational in nature. The other example of Differentiating Eq. (5.14) with respect to time, we write
an irrotational force is the electrostatic force. But, for the
j j −
→
two forces, namely, Coriolis force in the rotating frame of D D−
→v −→ −
→ D ds
reference and the electromagnetic forces, the line of action = . ds + v. (5.16)
Dt C Dt C Dt
need not pass through the center of gravity and thus they
are indeed rotational. In addition, a fluid which is in the Note that the differential operator acting upon the integrand
motion under the action of a nonconservative force, will be is the substantial (material) derivative D
Dt , since the contour
full of vorticity. For this reason, the atmospheric boundary of integration is a material contour moving with the fluid par-
layer around the Earth is completely filled with vorticity ticles.
due to the Coriolis force. We will solve the above equation by considering the terms
• The third circulation changing factor is the torque due separately. To solve the first term, Eq. (5.15) is introduced in
to the pressure forces. Incidentally,ıif the fluid is incom- Eq. (5.16) and subsequently applying the Stokes theorem, we
pressible (ρ = constant), the term dp ρ will not be able get
to change the circulation. Moreover, for a barotropic fluid
j ˆ
(ρ = ρ (p)), the lines of constant density and the lines of D−
→v −→ −
→ 1−→ −
→
constant pressure coincide with each other, in turn, the net . ds = ∇ × − ∇ p + ∇ .n̂ds
C Dt A ρ
pressure force acts through the center of gravity of the fluid ˆ
1 −→ −
→
particle, thereby, produces no torque on it. = ∇ ρ × ∇ p .n̂ds (5.17)
A ρ
2
From the above discussion, we may conclude that, for a fluid But, for a barotropic fluid it is known that
flow, if the fluid is barotropic, body forces are conservative and
viscous forces are negligible, then the circulation will forever −
→ −
→
∇ρ× ∇p=0 (5.18)
remains zero, and therefore the vorticity also will remain zero.
Thus, the resulting flow will be irrotational forever. In differ-
In addition, since the curl of any gradient is necessarily zero,
ent words, for a barotropic3 fluid in motion under the action of −
→ − →
i.e., ∇ × ∇ F = 0 for any function F. Therefore, we have
conservative body forces, the circulation () around an arbi-
trary closed curve, moving along with the fluid, is independent
−
→ 1−→ −
→
of time. Thus, from Eq. (5.12), we obtain ∇ × − ∇p+ ∇ =0 (5.19)
ρ
D
=0 (5.13) Hence, the first term in Eq. (5.16) becomes
Dt
j
An alternative derivation of the above equation can also be D−
→v −→
. ds = 0 (5.20)
given as follows. Consider an incompressible and inviscid C Dt
flow, as shown in Fig. 5.4a. A hypothetical arbitrary curve C −
−
→ →− → →
is selected specifying the region that comprises an identified By using the relation DDtds = ds . ∇ −v , the second term is
set of fluid particles. By definition, the circulation C around written as
the closed material contour C is given by j j j
− →− → → −
→
j −
→v.
Dds
= − →v . ds . ∇ − v =
1
∇ |− →v |2 . ds
−
→ −
→ Dt 2 C
C = v . ds (5.14) C c
C (5.21)
and applying the Stokes theorem (Eq. (3.42)), we have
3 The condition of barotropicity infers that the fluid density is a function
of pressure alone, i.e., ρ = ρ ( p).
5.4 Kelvin’s Circulation Theorem 131
C
(a) Γ
ΓC C2
C1 ΓC 1 2
(b)
At time t = t 1 At time t = t 2
Fig. 5.4 Schematic representation of the Kelvin’s circulation theorem
j
−
→ Dds Consider an airfoil section initially
v. =0 (5.22) →at rest,
as depicted in
C Dt Fig. 5.5a. Since the velocity is zero − v = 0 everywhere
in
the flow field, consequently, the circulation C1 about the
Introducing Eqs. (5.20) and (5.22) into Eq. (5.16), we obtain curve C1 specified around the airfoil will also be zero. When
the airfoil is set into motion, the flow curls around the trailing
D
=0 (5.23) edge, eventually leading to very high velocity at the trailing
Dt
edge. That is, in the beginning when the flow is unsteady,
which is same as given in Eq. (5.13), it shows that the time rate a small region of large velocity gradients are present at the
of change of circulation around a closed curve consisting of trailing edge. Because of this, a high vorticity region fixed to
same fluid particles is zero. For example, consider again two an identified set of fluid particle is formed in the neighbor-
arbitrary curves C1 and C2 at time instants t = t1 and t = t2 , hood of the trailing edge. As the fluid elements begin to move
respectively, as sketched in Fig. 5.4b, in a fluid flow without downstream, these vortices are also flushed downstream from
body forces. By definition, the circulation around the curve C1 the trailing edge. Initially, this thin sheet of vorticity content
ı →− →
is given by C1 = C1 − v . ds , and the circulation around the is highly unstable and it tends to curl, as depicted in Fig. 5.5b.
ı →− → This vortex is referred to as starting vortex. After some time,
curve C2 will be C2 = C2 − v . ds . Since C1 and C2 contain
when the flow becomes steady and it leaves the trailing edge
the same set of fluid particles, therefore, the circulation around
smoothly, the region of high-velocity gradients no longer exist
these curves will remain constant as the fluid particles proceed
at the trailing edge. That is, once the Kutta condition is satis-
downstream. Under these conditions, it is clearly visible that
fied, the vorticity is no longer produced at the trailing edge.
C1 = C2 .
However, the starting vortex initially shed is still being moved
steadily downstream continuously. Thus, at steady state, if the
original set of fluid particles enclosed within the curve C1 are
5.4.1 The Starting Vortex moved downstream and they have now formed the curve C2,
then using Kelvin’s theorem (Eq. (5.13)) the circulation C2
In this section, we will investigate the physics behind the
around the curve C2 must be the same as that around the curve
generation of circulation around an airfoil using the Kelvin’s
C1 when the flow was just started, namely, zero. That is,
circulation theorem. We will also examine the phenomena of
shedding of the starting vortex from the trailing edge.
C1 = C2 = 0 (5.24) This is an important result which states that the circulation
around an airfoil in a steady, incompressible, and inviscid
Let us now divide the curve C2 into the curve C3 (encloses flow field is equal but opposite to the circulation around the
the airfoil) and the curve C4 (encloses the starting vortex), as starting vortex shed from the trailing edge.
shown in Fig. 5.5b. Again, by definition, C3 is the circulation
around the curve C3 and C4 is the associated circulation
around the curve C4 . From the integral calculus, it can be 5.5 Summary of the Lift Generation
shown4 that the sum of these circulations is equal to the total Mechanism by the Airfoil
circulation C2 around the curve C2 . That is,
Let us now summarize the essence of all the vital concepts
C2 = C3 + C4 (5.25) studied so far in generating the lift by an airfoil. We have seen
that, soon after the flow is started, an infinitely large veloc-
From Eqs. (5.24) and (5.25), we get ity is developed at the trailing edge of an airfoil which leads
to large velocity gradients in the trailing edge neighborhood.
C3 = −C4 (5.26) In turn, a region of strong vorticity develops that tends to
curl around in counterclockwise direction downstream of the
trailing edge. This is called the starting vortex. As a response
4 Consider the flow field around an airfoil section, as shown in Fig. 5.5.
to this counterclockwise circulation around the starting vor-
Note that the curve C2 or abcd is divided into curves, abda and bcdb by tex, a circulation in clockwise direction around the airfoil is
drawing a curvilinear segment bd. The circulation C2 around the curve generated. Moreover, as the flow progresses, the vortices con-
abcd in counterclockwise direction can be written as tinuously shed from the trailing edge merge with the starting
ˆ ˆ ˆ ˆ vortex, making it more stronger. This results in a bigger coun-
−
→ −
→ b
−
→ −
→ c
−
→ −
→ d
−
→ −
→ a
−
→ −
→
C2 = v . ds = v . ds + v . ds + v . ds + v . ds terclockwise circulation around the starting vortex. However,
abcda a b c d
this has to be adjusted by the nature. Hence, the clockwise
Similarly, the circulation C3 around the curve abda can be written as circulation around the airfoil turns out to be further strong,
influencing the flow at the trailing edge all the more intently
ˆ ˆ ˆ approach the Kutta condition, decreasing the vortex shedding
−
→ −
→ b
−
→ −
→ d
−
→ −
→ a
−
→ −
→
C3 = v . ds = v . ds + v . ds + v . ds
a b d
from the trailing edge. Once the starting vortex gains just
abda
the right strength such that, the counterclockwise circulation
Likewise, the circulation C4 around the curve bcdb is around the starting vortex and clockwise circulation around
the airfoil exactly balance each other, no more vorticity will
ˆ ˆ ˆ
−
→ −
→ c
−
→ −
→ d
−
→ −
→ b
−
→ −
→ be fetched into the starting vortex, ensuring a smooth flow
C4 = v . ds = v . ds + v . ds + v . ds
bcdb b c d leaving the trailing edge. Under this condition, the Kutta con-
dition is precisely fulfilled and the starting vortex does not
Adding the circulations C3 and C4 , we get gain further strength or grow in size. Consequently, a steady
ˆ ˆ ˆ ˆ ˆ ˆ circulation prevails around the airfoil.
b
−
→ −
→ d
−
→ −
→ a
−
→ −
→ c
−
→ −
→ d
−
→ −
→ b
−
→ −
→
C3 + C4 = v . ds + v . ds + v . ds + v . ds + v . ds + v . ds
a b d b c d
But, 5.6 Classical Thin Airfoil Theory

ˆ ˆ
d
−
→ −
→ b
−
→ −
→ In this section, we will develop the fundamental equations
v . ds = − v . ds
b d necessary for the lift and moments calculation for a symmet-
ric airfoil. For a thin airfoil case, the airfoil is mathematically
Thus,
described by a vortex sheet placed along the camber line.
ˆ ˆ ˆ ˆ Here, we wish to calculate the variation of the strength of vor-
b
−
→ −
→ a
−
→ −
→ c
−
→ −
→ d
−
→ −
→
C3 + C4 = v . ds + v . ds + v . ds + v . ds tex sheet (γ) such that the camber line becomes a streamline
a d b c
ˆ ˆ ˆ ˆ of the flow and the Kutta condition (Eq. (5.11)) is satisfied at
b
−
→ −
→ c
−
→ −
→ d
−
→ −
→ a
−
→ −
→
= v . ds + v . ds + v . ds + v . ds the trailing edge, that is, γ (TE) = 0. Once we calculate the
a b c d
particular (γ) that obeys the aforesaid conditions, the total
Finally, we obtain circulation () around the airfoil is obtained by integrating
(γ) from the leading to trailing edge. Subsequently, the lift is
C2 = C3 + C4 calculated using the Kutta–Joukowski theorem (Eq. (4.140)).
5.6 Classical Thin Airfoil Theory 133
C1
(a)
ΓC = 0
1
The velocity field is

zero everywhere
Airfoil
C2
(b)
ΓC = 0
ΓC d 2
a 4 ΓC
3
Ua
C3
c
C4 Starting vortex
b
Airfoil
Fig. 5.5 The generation of starting vortex and the circulation around an airfoil
We know that an airfoil is defined either in terms of its the chord length is denoted by c. In addition, w (s) is the
chord line or its camber line. However, in general, the dis- component of velocity normal to the camber line induced by
tances of the upper and the lower surfaces of the airfoil are the vortex sheet.
measured from the chord line, as illustrated in Fig. 5.6. From Moreover, if the airfoil is thin, the chord line and camber
the figure, the shape of the camber line zc (x) is written as line are close to each other, and thus if seen from a distance,
the vortex sheet seems to fall roughly on the chord line. Thus,
1 it is usual to place the vortex sheet along the chord line, as
zc (x) = [zx (x) + zl (x)] (5.27)
2 depicted in Fig. 5.8a. Note that the vortex strength per unit
length will now be given as γ = γ (x) and the Kutta condi-
and the airfoil thickness t (x) will be
tion as γ (c) = 0. However, these conditions are still satisfied
such that the camber line only represents a streamline rather
t (x) = zu (x) − zl (x) (5.28)
than the chord line. In other words, we wish to calculate the
strength of vortex sheet γ (x) on the chord line such that the
Note that, for a thin airfoil, the thickness (t) at every location
camber line becomes a streamline in the flow field.
(x) measured from the leading edge is much smaller than the
Since there can be no flow across a streamline, the velocity
chord (c). That is,
normal to it will be zero. Likewise, for the camber line to
become a streamline of the flow, the component of velocity
tc
normal to it has to be zero at all the points along its length.
From the vector algebra, the resultant velocity at any point
Also, for a thin airfoil, the distribution of vortices over the
in the flow field will be the vector addition of the freestream
airfoil’s surface when viewed from a distance looks almost
velocity (Ua ) and the velocity induced by the vortex sheet,
similar to the case as if they were placed along the camber line.
w (s). Thus, if the component of freestream velocity normal
Consider an airfoil that is simulated by placing a vortex sheet
to the camber line is denoted by Ua,n , then for camber line to
along the camber line, as shown in Fig. 5.7. The airfoil is at an
be a streamline, we can write
angle of attack α in the freestream of velocity Ua . The distance
along the camber is measured in curvilinear coordinate s and
Ua,n + w (s) = 0 (5.29)
Thickness
z Camber line
z u(x)
Chord line
LE x
TE
z c (x)
z l (x)
Chord (c)
Fig. 5.6 Nomenclature used for an airfoil
Leading edge
s γ (s) Trailing edge
Ua
(a) Vorticity distribution on the airfoil surface
z
Camber line, z = z(x)
w’

s
x
α 0 c
Ua
(b) Vorticity distribution on the camber line of the airfoil
Fig. 5.7 Simulation of an airfoil by placing vortices on the airfoil surface and the camber line
z
Camber line, z = z (x)
w’(s)

s
0α x
Ua c
α
U a,n
tan−1 dz
dx
Fig. 5.8 Simulation of an airfoil by distribution of vortices on the chord line
The above equation should hold all along the length of the
vθ = (5.35)
camber line. Also, note that, Eq. (5.29) is a simple scalar addi- 2πr
tion due to the fact that both Ua,n and w (s) are in the same
Note that, in the above equation, the vortex strength (total
direction, i.e., normal to the camber line.
circulation) is calculated around a closed curve in counter-
Calculation of Ua,n clockwise direction. Thus, for the elemental vortex sheet of
the length dξ located at ξ and having the
vortex
strength per
We now derive the expression for Ua,n . Figure 5.7 shows
that
unit length γ (ξ), the velocity induced dw at a distance x
the slope of the camber line is defined as θ = tan−1 dz c
dx and from the leading edge will be
since the camber line is at an incidence α, the effective angle
of attack (αeff ) will be d
dvθ = dw = (5.36)
2πr
αeff = α − θ (5.30)
But, in our case d = γ (ξ) dξ and r = (x − ξ). In turn, the
dz
For a thin airfoil, θ = tan−1
dx ≈
dz
dx , and thus Eq. (5.30) above equation is rewritten as
becomes
dzc γ (ξ) dξ
αeff = α − (5.31) dw = − (5.37)
dx 2π (x − ξ)
In addition, the component of freestream velocity in the nor- The negative sign in the above expression shows that, in our
mal direction will be given by case, the circulation along the camber line is in clockwise
direction. Subsequently, to obtain the velocity w (x) induced
Ua,n = Ua sin (α − θ) (5.32) at x by all the elemental vortex sheets along the chord line, one
can integrate the above equation from the leading edge (LE)
Again for small angle of attack and for a thin airfoil case, the to the trailing edge (TE), that is, between the limits, x = 0 to
above equation gives x = c. Hence,

dzc ˆc ˆc
Ua,n = Ua α − γ (ξ) dξ
dx w (x) = dw = − (5.38)
2π (x − ξ)
0 0
For the convenience, dropping the subscript of z in the above
equation, we finally get Again from the approximation stated by Eq. (5.34), the
expression for w (s) is
dz
Ua,n = Ua α − (5.33)
dx ˆc
γ (ξ) dξ
w (s) = − (5.39)
2π (x − ξ)
Calculation of w (s) 0
Now, we wish to calculate an expression for the component Introducing Eqs. (5.33) and (5.39) into Eq. (5.29), we finally
of velocity induced by the vortex sheet, w (s), normal to the have
camber line in terms of the vortex sheet strength per unit ˆc
dz 1 γ (ξ) dξ
length, γ (s). Further, as discussed earlier, it is consistent with Ua α − = (5.40)
dx 2π (x − ξ)
the thin airfoil theory to assume the placement of vortex sheet 0
on the chord line if the camber line and chord line are close
to each other for a thin airfoil (Fig. 5.8). Thus, it is reasonable This is the fundamental governing equation of the thin airfoil
to have the approximation that theory, which indeed is a mathematical representation of the
fact that the camber line is a streamline of the flow.
w (s) ≈ w (x) (5.34)
Since vortices are continuously distributed on the chord line 5.6.1 Aerodynamic Characteristics of a Thin
and thus to obtain w (x), we can consider an element of vortex Symmetric Airfoil
sheet of length dξ at a distance ξ from the leading edge along
the chord line, as shown in Fig. 5.8b. The strength of this The purpose of this section is to calculate the aerodynamic
elemental vortex sheet will be γ (ξ) dξ. Recall, Eq. (4.78), we properties for a thin symmetrical airfoil using the fundamental
have governing equation (Eq. (5.40)) as obtained in the previous
section. We know that, a symmetric airfoil is the one which rule, i.e., differentiating the numerator and denominator of
does not possess the camber and its chord line coincides with this equation and applying the limits separately. In turn, we
the camber line, in turn, the
slope of the camber line will have
dz
essentially be zero dx = 0 . Thus, Eq. (5.40) becomes − sin π
γ (π) = 2αUa =0 (5.47)
cos π
ˆc
1 γ (ξ) dξ which shows that the Kutta condition is indeed satisfied at the
= Ua α (5.41)
2π (x − ξ) trailing edge of a thin symmetrical airfoil.
0
From Fig. 5.9a, the total strength of the vortex sheet placed
This is the basic equation for an inviscid, barotropic, and along the camber line of a thin symmetric airfoil, that is, the
incompressible fluid flow past a symmetric airfoil at small α. overall circulation () is given by
Since the above expression is in terms of x and ξ in Cartesian
ˆc
space and it will be more convenient to transform the above
= γdξ (5.48)
equation in polar coordinate (r, θ) system. Thus, we require
the following trigonometric substitution for ξ to transform it 0
into the polar coordinates.
Again, since the above equation is in terms of Cartesian space,
c i.e., (x, ξ), it should be transformed into the polar coordinates
ξ= (1 − cos θ) (5.42) (r, θ) before obtaining the solution. Introducing Eq. (5.44)
2
into the above integral, gives
Note that ξ being a general point is associated with the general
angle θ. However, x is a fixed location on the chord line, at ˆπ
c
which the total induced velocity has to be obtained, and hence = γ (θ) sin θdθ (5.49)
2
it must be associated with a fixed value of angle, say, θ0 . Then, 0
c Substitute γ (θ) from Eq. (5.46),

x= (1 − cos θ0 ) (5.43)
2
ˆπ
Differentiating Eq. (5.42) with respect to θ, we get c
= 2αUa (1 + cos θ) dθ
2
c 0
dξ = sin θdθ (5.44)
2 ˆπ
c
= 2αUa (1 + cos θ) dθ
Replacing the variables in Eq. (5.41), using the above trans- 2
0
formations, and integrating from the leading edge (θ = 0) to
the trailing edge (θ = π), yields Finally, the circulation () becomes
ˆπ
1 γ (θ) sin θdθ = παUa c (5.50)
= Ua α (5.45)
2π cos θ − cos θ0
0 Recall, the Kutta–Joukowski theorem (Eq. (4.140)), the lift
per unit span acting on the airfoil becomes
The solution for γ (θ) in the above integral requires tedious
procedure to follow, which is beyond the scope of this book.
L = ρUa = ρUa2 παc (5.51)
Rather, the solution is written directly as
Since the lift coefficient (CL ) is defined as
1 + cos θ
γ (θ) = 2αUa (5.46)
sin θ L
CL = (5.52)
In order to satisfy the Kutta condition at the trailing edge, qa c
where θ = π, we must have
where qa = 21 ρUa2 is the freestream dynamic pressure. Intro-
ducing Eq. (5.51) into the above relation gives the lift coeffi-
γ (TE) = 0
cient as
CL = 2πα (5.53)
But, Eq. (5.46) assumes an indeterminate form of 00 at θ = π,
and thus the solution can be obtained by using L’ Hospital’s
dL
(a) Chord line
M LE

dξ
ξ
Chord (c)
dL
Chord line
(b) Mx
dξ
x
ξ
Chord (c)
Fig. 5.9 The lift and the moments on a thin symmetric airfoil
This is indeed a standard result for the airfoil, which states This is the total moment generated due to the lift at a general
that the lift coefficient is linearly proportional
to the angle of point on the chord line. The negative sign in this expression
attack. Subsequently, the lift slope a0 = dC L
dα is given as
in fact shows that the moment is acting in the direction that is
opposite to the circulation caused by the vorticity.
dCL Likewise, the differential moment about the leading edge,
a0 = = 2π (5.54) is
dα
dMLE = −ξρUa γ (ξ) dξ (5.58)
Since, for a symmetrical airfoil, the chord line and camber
line coincide with each other, where the vortex sheet is placed. and thus the total moment about the leading edge will become
That means, the vorticity is distributed along the chord line.
Moreover, if each element of this vorticity causes some dif- ˆc

ferential lift dL , given as MLE = −ρUa ξγ (ξ) dξ (5.59)
0
dL = ρUa γ (ξ) dξ (5.55)
Again, since the above equation is represented in terms of ξ,
Moreover, the above differential lift (dL) causes the differen- transforming Eq. (5.59) with the aid of Eqs. (5.42) and (5.44),
tial moment dMx about the point x from the leading edge, as we get
shown in Fig. 5.9b, is given by
ˆπ
c c
dMx = − (ξ − x) dL = −ρUa γ (ξ) (ξ − x) dξ (5.56) MLE = −ρUa (1 − cos θ) γ sin θdθ (5.60)
2 2
0
Thus, the total moment per unit span (Mx ) about the point
x caused by the total lift is obtained by integrating equation Introducing Eq. (5.46) into Eq. (5.60), the moment about the
(5.56) over the entire chord line, i.e., between the limits, x = 0 leading edge for a thin symmetrical airfoil becomes
(leading edge) and x = c (trailing edge). That is, ˆπ
c2
ˆc ˆc MLE = 2αρUa2 (1 − cos θ) (1 + cos θ0 ) dθ
4
Mx = dMx =− ρUa γ (ξ) (ξ − x) dξ (5.57) 0
ˆπ
0 0 ρUa 2
2
= −α c 1 − cos2 θ dθ
2
0
Solving above integral, we finally have dzUnlike

the symmetrical airfoil, slope of the camber line
dx , evaluated at a fixed point x, is finite for a cambered
π
MLE = −αqa c2 (5.61) airfoil. Following the same steps as carried out to obtain
2 Eq. (5.45), the fundamental equation for a cambered airfoil
In addition, the moment coefficient (CM ) is defined by can be written as
ˆπ
M 1 γ (θ) sin θdθ dz
CM = = Ua α − (5.66)
qa Sc 2π cos θ − cos θ0 dx
0
where qa = 2 ρUa
1 2 is the freestream dynamic pressure. Thus,
In the above equation dx dz
= 0 for a cambered airfoil, thus it
from Eq. (5.61) the moment coefficient about the leading edge
is retained and since the point x corresponds to a fixed angle
will become
M πα θ0 , thereby, dx
dz
also becomes a function of θ0 . Again due to
CMLE = LE2 = − (5.62) the complexity involved in solving for γ (θ) in Eq. (5.66), we
qa c 2
can directly state the solution as
Further, using the lift coefficient (Eq. (5.53)), CMLE can be ∞

rewritten as 1 + cos θ
CL γ (θ) = 2Ua A0 + An sin nθ (5.67)
CMLE = − (5.63) sin θ
4 n=1
We know that the moment coefficient about any point on a Note that Eq. (5.67) satisfies the Kutta condition as γ (π) = 0.
thin symmetric airfoil can be written as Also, the first term is the same as Eq. (5.46) for a symmetric
airfoil and the second term is indeed a Fourier sine series
CL x with coefficients An (n = 1, 2, 3 . . .). Here, A0 and An should
CMx = − + CL (5.64)
4 c have the particular values such that the camber line becomes
a streamline. To calculate these specific values of A0 and An ,
At
quarter-chord
point x = 4c , the moment coefficient we will substitute Eq. (5.67) into Eq. (5.66). Thus, we get
CMc/4 is zero. That is,
ˆπ ∞ ˆπ
1 (1 + cos θ) dθ 1 sin nθ sin θdθ dz
CL 1 A0 + An =α−
CMc/4 =− + CL π cos θ − cos θ0 π cos θ − cos θ0 dx
4 4 0 n=1 0
(5.68)
or Consider the first term on left side of 5.68, the value of the
CMc/4 = 0 (5.65) integral term is
ˆπ
That is, for a thin symmetric airfoil, the moment coefficient (1 + cos θ) dθ
about the quarter-chord point is zero. Recall that, by defini- =π (5.69)
cos θ − cos θ0
tion, the center of pressure (xCP ) is defined as the point about 0
which the moments are zero. Thus, from Eq. (5.65), it can be
and the value of the integral in second term is given as
stated that the center of pressure for a thin symmetric airfoil
lies at the quarter-chord point. ˆπ
sin nθ sin θdθ
= −π cos nθ0 (5.70)
cos θ − cos θ0
5.6.2 Aerodynamic Characteristics of a Thin 0
Cambered Airfoil
From Eqs. (5.69) and (5.70), we can rewrite Eq. (5.68) as
A cambered airfoil is the one in which the camber line and the ∞
chord line do not merge with each other. This is in contrast dz
= (α − A0 ) + An cos nθ0 (5.71)
to the symmetric airfoil case where these lines get merged. dx
n=1
A camber essentially represents the asymmetry introduced
between the suction and pressure surfaces of an airfoil in which is in the form of a Fourier cosine series expansion for
dz
which the suction surface is more convex (bulges outward) the ´function dx . In order to solve, we will multiply Eq. (5.71)
π
than the pressure surface. In other words, the suction surface by 0 cos mθ0 dθ0 on either side where m = n in general.
has more positive camber as compared to the pressure surface. Thus,
ˆπ ˆπ
dz
∞
From Eq. (4.140), the lift per unit span L will become
cos mθ0 dθ0 = α − A0 + An cos nθ0 cos mθ0 dθ0
dx
n=1 ⎡ ⎤
0 0
ˆπ ∞ ˆπ
(5.72)
L = ρUa2 c ⎣A0 (1 + cos θ) dθ + An sin θ sin nθdθ⎦
If m = 0, then the above equation reduces to
0 n=1 0
ˆπ (5.79)
dz
dθ0 = π (α − A0 )
dx From any standard text on calculus, we can refer the following
0
results:
or ˆπ
ˆπ
1 dz (1 + cos θ) dθ = π (5.80)
A0 = α − dθ0 (5.73)
π dx
0 0
ˆπ
π
if n = 1
However, if m = 0, then Eq. (5.72) becomes sin θ sin nθdθ = 2 (5.81)
0 i f n = 1
0
ˆπ ∞ ˆπ
dz
cos mθ0 dθ0 = An cos nθ0 cos mθ0 dθ0 (5.74) Thus, if we set n = 1, the lift per unit span (Eq. (5.79)) reduces
dx
0 n=1 0 to π
L = ρUa2 c A0 π + A1 (5.82)
From any standard text on integrals, we can write 2
ˆπ In addition, the expression for lift coefficient will become

π
if n = m
= 2
π
cos nθ0 cos mθ0 dθ0 (5.75)
0 i f n = m L
0 CL = = 2 A0 π + A1 (5.83)
2 ρUa c
1 2 2
By setting n = m in Eq. (5.74), we get
Introducing the value of A0 from Eq. (5.73) and the value of
ˆπ A1 using Eq. (5.76) when n = 1, into the above expression,
dz π gives
cos nθ0 dθ0 = An
dx 2
⎡ ⎤
0
ˆπ
1 dz
or CL = 2π ⎣α + (cos θ0 − 1) dθ0 ⎦
ˆπ π dx
2 dz 0
An = cos nθ0 dθ0 (5.76)
π dx
0 The lift curve slope or lift slope (a0 ) will be
From Eqs. (5.73) and (5.76), it can be observed that A0 is a dCL

a0 = = 2π (5.84)
function of shape of the camber line as well as angle of attack, dα
whereas An depends only on the shape of the camber line.
We will now develop the expressions for aerodynamic lift It is interesting to note that the result given by Eq. (5.84) is the
and moments coefficients for a cambered airfoil. In doing so, same as obtained for a symmetric airfoil (Eq. (5.54)). Thus, it
the total circulation caused by the entire vortex sheet from can be said that the lift slope (a0 ) will always be equal to 2π
leading edge to trailing edge, placed on the camber line, is for any shape of the airfoil.
Furthermore, to evaluate the total moment MLE about
ˆc ˆπ the leading edge of a cambered airfoil, the similar exercise as
c
= γ (ξ) dξ = γ (θ) sin θdθ (5.77) carried out for the symmetric airfoil, will lead to the following
2
0 0 expression:
ˆc

Introducing Eq. (5.67) into the above equation, we get MLE = −ρUa ξγ (ξ) dξ (5.85)
⎡ ⎤
ˆπ ∞ ˆπ 0
= cUa ⎣A0 (1 + cos θ) dθ + An sin θ sin nθdθ⎦

Transformation of the above integral in polar coordinates
0 n=1 0 using Eqs. (5.42) and (5.44), gives
(5.78)
dL
Camber line
(a) M LE
Leading edge dξ Trailing edge
ξ
Chord (c)
dL
Camber line
(b) Mx
Leading edge dξ Trailing edge
x
ξ
Chord (c)
Fig. 5.10 The lift and the moments on a thin cambered airfoil

ˆπ MLE π A2
c2 CMLE = 1 =− A0 + A1 − (5.89)
MLE = −ρUa (1 − cos θ) γ (θ) sin θdθ (5.86)
2 ρUa c
2 2 2 2
4
0
This is the moment coefficient about the leading edge (x = 0)
Substitution of γ (θ) from Eq. (5.67), in the above expression of a cambered airfoil. If we wish to calculate the total moment
leads to about any arbitrary point (say x) on the camber line, as illus-
trated in Fig. 5.10b, then it can be obtained as follows. The
ˆπ
ρUa c2 vorticity, in a symmetric airfoil, is continuously distributed on
MLE = − 2Ua A0 (1 − cos θ) (1 + cos θ) dθ
2 2 the camber line. Consider an elemental vorticity segment of
0 length dξ at a point ξ on the camber line, shown in Fig. 5.10.
∞ ˆπ The differential lift dL caused by this segment is
+ An (1 − cos θ) sin nθ sin θdθ (5.87)
n=1 0 dL = ρUa γ (ξ) dξ (5.90)
Consider the first term on the right side of this equation, the
In turn, the differential moment per unit span dMx acting
value of the integral term will be about the point x will be
ˆπ
π dMx = −(ξ − x) dL (5.91)
(1 − cos θ) (1 + cos θ) dθ =
2
0 From Eq. (5.90), the above equation becomes
´π
In the second ´term, the value of 0 sin nθ sin θdθ is π2 when dMx = −ρUa (ξ − x) γ (ξ) dξ (5.92)
π
n = 1; and − 0 sin nθ cos θ sin θdθ = − π4 for n = 2. From
these values, Eq. (5.87) reduces to Moreover, the total moment per unit span (Mx ) about the point
1 π π π x caused by the total lift is obtained by integrating equation

MLE = − ρUa2 c2 A0 + A1 − A2 (5.88) (5.92) over the entire chord line, i.e., between the limits; x = 0
2 2 2 4
(leading edge) and x = c (trailing edge). Thus,

In turn, the moment coefficient CMLE about the leading edge
becomes
ˆc only on the shape of the camber line and are independent of

Mx = −ρUa (ξ − x) γ (ξ) dξ the angle of attack (α). Consequently, from Eq. (5.100), the

0 moment coefficient CMc/4 about the quarter-chord point will
ˆc ˆc also be independent of α, i.e.,
= −ρUa ξγ (ξ) dξ + ρUa x ξγ (ξ) dξ (5.93)
dCMc/4
0 0 =0 (5.101)
dα
In the above equation, the first term shows the moment per
However, by definition, the aerodynamic center is that point at
unit span (MLE ) about the leading edge, while the second
which the pitching moment coefficient for an airfoil does not
term is the contribution
to the moment caused by the lift per
vary with the lift coefficient (in turn, angle of attack). There-
unit span L about the leading edge. That is,
fore,
from Eq. (5.101), it is clear that the quarter-chord point
xc/4 is indeed the theoretical location of the aerodynamic
Mx = MLE

+ xL (5.94)
center (xAC ) for a thin cambered airfoil (see Sect. 2.4.4).
Furthermore, the location of the center of pressure (xCP )
So, the moment coefficient about the point x will be
can be obtained using Eq. (2.46) as
x
CMx = CMLE + CL (5.95)
c MLE CMLE
xCP =− =− c (5.102)
L CL
Dividing Eq. (5.83) by 4, we get
From Eq. (5.98), the above equation gives
CL π A1
= A0 + (5.96)
4 2 2 c π
xCP = 1+ (A1 − A2 ) (5.103)
4 CL
Solving above for A0 , gives
which shows that the center of pressure for a thin cambered
CL A1 airfoil varies with the lift coefficient (i.e., angle of attack).
A0 = − (5.97)
2π 2 Moreover, Eq. (5.102) reflects that if the lift on a cambered
airfoil decreases, the location of the center of pressure shifts
Introducing this relation into Eq. (5.89), we get
toward the trailing edge. In the limit, it may even leave the
airfoil and moves toward infinity if the lift approaches zero.
π CL A1 A2
CMLE = − − + A1 − Because of this reason, the center of pressure is not always
2 2π 2 2
an appropriate location to consider the aerodynamic forces
or and moments system and a more preferred location is the
aerodynamic center (as it does not depend upon the changes
CL π in the angle of attack).
CMLE = − + (A1 − A2 ) (5.98)
4 4
Substituting Eq.(5.98) into Eq. (5.95), we get the moment Summary

coefficient CMx about an arbitrary point on the camber line
as An aircraft moves in the air by overcoming the gravity with
1 x π
CMx = −CL − − (A1 − A2 ) (5.99) a lifting force, which is essentially provided by the aircraft’s
4 c 4
wing. The cross-sectional geometry of the wing influences the
flow of air and the combined geometry of the wing and the
For the quarter-chord point x = 4c , the above equation gives
reaction of the air causes any general solution of the wing-
π sectional properties to become too complicated, making it
CMc/4 = − (A1 − A2 ) (5.100) impossible to utilize or almost difficult to ascertain. To deal
4
with finding the flight properties of wing sections, a more
Interestingly, in contrast to a thin
symmetrical airfoil, where ameliorated way is to consider an inviscid and incompress-
the moment coefficient CMc/4 about the quarter-chord point ible flow past the wing surface. A vortex superimposed on the
was zero (Eq. (5.65)), for a thin cambered airfoil CMc/4 is airstream simulates the process of lift generation by the wing
finite. Therefore, the quarter-chord point xc/4 and the center section. The vortex distribution along the wing will simulate
of pressure location
(xCP ) are not the same for a thin cam- the actual properties of the wing and allow to have a sim-
bered airfoil xc/4 = xCP . However, since A1 and A2 depend ple approach of calculating the properties of the wing. This
hypothesis, better known as thin airfoil theory, was first con- either along the surface or along the camber line resulting a
ceived by Max Munk which was later refined by the team led vortex sheet, then in terms of the strength of vortex sheet, the
by Hermann Glauert in 1920s. Kutta condition can be expressed as
The circulation and vorticity are the two primary measures
of rotation in a fluid. Circulation is a scalar quantity, obtained γ (TE) = 0
through the integration, which is a macroscopic measure of
rotation over a finite region in the fluid flow, whereas vorticity Since an airfoil is described numerically either by setting the
is a vector field which provides a microscopic measure of the vortices along the airfoil surface or along the camber line. In
rotation at any point in the fluid. thin airfoil theory, these vortices are really disseminated by
Circulation is defined as the line integral of the tangential setting them along the mean camber line of an airfoil section.
velocity component around a closed curve fixed in the flow This arrangement essentially forms a vortex sheet which is
field. That is, placed along the chord line. In addition, the strength of vortex
‰ sheet (γ) is balanced with the end goal that when the uniform
−
→
=− − →v . ds stream is superimposed on this vortex sheet, then the camber
c
line turns into a streamline. The Kutta condition is naturally
fulfilled by this aforesaid flow configuration. The strength of a
−
→ vortex sheet is computed from the accompanying fundamental
and the vorticity ζ is defined as
equation of thin airfoil hypothesis.
−
→
ω =∇ ×−
ζ = 2−
→ →
v ˆc
dz 1 γ (ξ) dξ
Ua α − =
The concept of vortex sheet was introduced to provide the dx 2π (x − ξ)
0
mathematical description of an inviscid and incompressible
flow past an airfoil section. Suppose the strength of a vortex This is the fundamental governing equation of the thin airfoil
sheet per unit length is γ (s), where s is the curvilinear length theory, which is a mathematical representation of the fact that
measured from the leading edge on the airfoil surface. The the camber line is a streamline of the flow.
velocity potential induced by this sheet at an arbitrary point In order to obtain, the aerodynamic loads and moments
P (x, y) will be given as coefficients, the governing equation of thin airfoil theory has
ˆ been applied for a symmetric and the cambered airfoils. The
1
φ (x, y) = − θγ (s) ds major outcomes from these analysis are summarized as fol-
2π
lows:
and the circulation associated to this vortex sheet is For a symmetric airfoil section, the lift coefficient (CL ) is
ˆ given by
= γ (s) ds
CL = 2πα
It was also noticed that tangential velocity across this vortex which states that the lift coefficient is linearly proportional to
sheet is variable, if v1 and v2 , respectively, are the tangential the angle of attack. Subsequently, the lift slope is given as
velocities on the upper and lower surfaces, then the strength
of sheet per unit length can be described as dCL
a0 = = 2π
dα
γ (s) = v1 − v2
Moreover, for a symmetric airfoil section, the locations of the
The Kutta condition is the criteria by which a specific value center of pressure (xCP ) and aerodynamic center (xAC )
the
of is chosen for the flow around an airfoil at an angle of are the quarter-chord point x 4c .
attack so that the flow leaves the trailing edge smoothly. For
quarter-chord point x = 4 , the moment
c
In addition,
at the
a thick airfoil with finite angle at the trailing edge, the flow coefficient CMc/4 is zero, i.e., CMc/4 = 0.
velocities on the upper and lower surfaces leaving the trailing Furthermore, for a cambered airfoil section, the lift coef-
edge will be zero. Consequently, the trailing edge will become ficient (CL ) is defined as
a stagnation point. For the thin airfoil with cusped trailing ⎡ ⎤
edge, the flow velocities leaving the upper and lower surfaces ˆπ
1 dz
are nonzero and equal in magnitude and direction. If these CL = 2π ⎣α + (cos θ0 − 1) dθ0 ⎦
π dx
airfoils are mathematically described by distributing vortices 0
Summary 143
and the lift slope (a0 ) is (a) leading edge

(b) trailing edge
dCL
a0 = = 2π (c) center of pressure
dα (d) aerodynamic center
Interestingly, it has been observed that the lift slope (a0 ) will
2. For the flow past an airfoil, consider the following state-
always be equal to 2π for any shape of the airfoil.
ments.
For a cambered airfoil, it has been found that the aero-
dynamic center is located at the quarter-chord point, and the (i) There is a slight upflow before reaching the airfoil.
center of pressure varies with the lift coefficient (i.e., angle of (ii) There is a downflow after passing the airfoil.
attack). That is, (iii) The streamlines are farther to each other on the suc-

c π tion surface.
xCP = 1+ (A1 − A2 )
4 CL
Which of the above is/are true?
(a) (i) and (ii)

(b) (ii) and (iii)
Exercises (c) (i) and (iii)
(d) All
3. An airfoil creates a lift (per unit span) of 100 Nm−1 entails
1. Obtain the camber line that gives a uniform lift distribu- the circulation, = 1.2 m2 s−1 . The ambient pressure
tion across the chord of an airfoil. and temperature are 102 kPa and 299 K, respectively. If
2. Calculate CL and CM for a thin symmetrical flat plate at the specific gas constant is 287 J kg−1 K−1 , the freestream
the angle of attack, α = 2.2◦ . velocity is
3. Estimate the pressure distribution for an arbitrary airfoil
at an angle of attack. (a) 70.1 ms−1
4. Obtain the aerodynamic characteristics of a symmetric bi- (b) 81.3 ms−1
convex airfoil at zero angle of attack. Plot the distribution (c) 85.6 ms−1
of velocity component and the pressure over the airfoil (d) 90.4 ms−1
surface.
5. Plot the distribution of the velocity components and the 4. Assume that a positive cambered airfoil attains zero lift,
pressure over the parabolic arc airfoil placed at an angle the angle of attack experienced by the airfoil is
of attack. Also, obtain its aerodynamic characteristics.
6. Consider a flapped flat plate airfoil, i.e, an airfoil com- (a) zero
posed of two straight segments at an angle. The point of (b) greater than zero
contact of these two segments is referred to as the hinge. (c) less than zero
Calculate the hinge moment coefficient. (d) same for all the value of lift
7. Using complex variables, formulate the thin airfoil theory
directly in terms of the complex velocity that itself is an 5. The pressure coefficient at a point P on the surface of the
analytic function. airfoil in Mach 0.4 freestream is 0.6. If the freestream
8. Show that the lift is very high when the freestream flow Mach number is increased to 0.8, the new pressure coef-
is along the axis of the metacentric parabola. ficient at the same point will be
9. Prove that the line of action of the lift passes through the
(a) 0.7
centroid of the circulation for any given incidence.
(b) 0.8
10. Derive an expression for the lift coefficient per unit wing
(c) 0.9
chord in terms of the circulation and the freestream veloc-
(d) 1.0
ity.
6. For an airfoil, if the CL versus CD curve is approximately
symmetric about the line given by CL = 0.3. It is a
Multiple Choice Questions
(a) symmetric airfoil
1. On an airfoil, the decrease in pressure on suction surface (b) positively cambered airfoil
is greater than the decrease in pressure on the pressure
surface. In both cases, the effect is more near
(c) negatively cambered airfoil 10. For a positively cambered airfoil, if the angle of attack
(d) flat plate is increasing then the distance between the aerodynamic
center and the center of pressure will be
7. Consider a thin airfoil at an angle of attack α in the Mach
0.8 freestream. The expression for the lift coefficient using (a) increasing
Prandtl–Glauert rule is (b) decreasing
(c) increasing up to wing stall
(a) 6.24 α (d) remain same
(b) 8.36 α
(c) 9.55 α Keys
(d) 10.46 α
1. (a)
8. Consider a NACA 2412 airfoil of chord length 2.4 m. The 2. (a)
location of maximum camber, measured from the trailing 3. (a)
edge along the chord line, is 4. (c)
5. (c)
(a) 1.44 m 6. (b)
(b) 1.55 m 7. (d)
(c) 1.67 m 8. (a)
(d) 1.73 m 9. (c)
10. (d)
9. A subsonic aircraft flying at Mach 0.6 uses NACA 24012
airfoil of chord length 2.2 m. The maximum thickness of
the airfoil is
(a) 0.190 m
(b) 0.225 m
(c) 0.264 m
(d) 0.315 m
Finite Wing Theory
6
One of the most vital uses of potential flow theory was the The curling of the flow around wing tips has another con-
analysis of lifting surfaces such as the wings of an aircraft, sequence on the aerodynamic characteristics of the wing. This
since the boundary conditions on a complicated geometry “flow slip” establishes a circulatory motion that trails down-
can substantially muddle the attempt to solve the problem stream of the wing, i.e., a trailing vortex is produced at each
by analytical means which, in turn, necessitates some sim- wing tip, shedding downstream of the wing. These trailing
plifying assumptions to obtain the solution. In this chapter, vortices, shed from the wing tip, induce a small downward
these assumptions will be linked to the definition of the three- component of air velocity (opposite to the direction of lift) in
dimensional thin wing problems. the vicinity of the wing itself. This downward component of
air velocity is known as downwash, which is usually denoted
by w. The downwash can also be viewed as the result of
6.1 Introduction the lifting wing “pushing down” on the air, which results in
the air having added downward momentum in the neighbor-
In Chap. 5, we have discussed the properties associated with hood of the wing. The downwash combines with freestream
an airfoil section which are indeed the same as the properties velocity (Ua ) to produce a local relative wind which is canted
of a wing of infinite span. But all the real aircraft have the downward in the vicinity of each airfoil section, as shown in
wings of finite span and thus, it is natural to think whether Fig. 6.3.
the knowledge of airfoil properties can help in finding the We know that the angle between the chord line and the
properties of the finite wings. An airfoil, in fact, is simply the freestream direction is known as the geometric angle of attack
cross section of a wing and hence, it is obvious to assume that (α). However, by examining Fig. 6.3 it is evident that the local
both airfoil and the wing behave exactly in similar fashion. wind speed is inclined below the freestream velocity (Ua ) by
However, we should note that unlike an airfoil which is a two- the angle αi , referred to as the induced angle of attack. If the
dimensional object, a wing is essentially a three-dimensional downwash induced is wi and the freestream velocity is Ua ,
body, that is, there will be a component of flow in the span- then the induced angle of attack (αi ) is written as
wise direction. We know that an aircraft is airborne because

of the lift produced by its wings due to the higher pressure −1 −wi
αi = tan (6.1)
on the lower (bottom) surface and a lower pressure on the Ua
suction (top) surface. This pressure difference between the
lower and upper surfaces produces the lift. Also, it causes Note that the negative sign in the above expression reflects that
the flow to curl around near the wing tips being forced from the downwash is in downward direction, i.e., in the direction
the pressure surface toward the suction surface. Consequently, opposite to that of the lift force. Using small-angle approxi-
there will be a spanwise component of flow in the direction mation, induced angle can also be written as
from the wing tip to the wing root which, in turn, causes the
streamlines on the top surface to bend toward the wing root, as −wi
αi ≈ (6.2)
shown in Fig. 6.1. Likewise, on the bottom surface the span- Ua
wise component of flow will be in the direction from wing
root to the wing tip. Therefore, it is now evident that the flow The downwash thus produced by shedding trailing edge vor-
over the wings is indeed three-dimensional and hence their tices from the wing tips and its effect on the inclination of the
aerodynamic properties are quite different from those of its local relative wind has two major consequences on the local
airfoil sections (Fig. 6.2). airfoil section. The actual angle of attack as seen by the air-

146 6 Finite Wing Theory
(a) Reduced upper pressure
Wing−tip − p
vortex
Increased lower pressure
Freestream (U a )
Upper surface
(b) streamlines
Flow−slip at Wing Flow−slip at

the Wing−tip the Wing−tip
Lower surface
streamlines
Fig. 6.1 Front and top views of the flow pattern on a three-dimensional wing
Freestream (Ua )
Lift (L)
Wing−tip trailing
edge vortex shedding
Wing
Ua
Wing−tip trailing Downwash (w)

edge vortex shedding
Effective velocity (Ueff )
Fig. 6.2 Schematic diagram of the wing tip vortex shedding and the downwash
foil locally is, in fact, lower than the geometric angle of attack Also, the effective freestream velocity (Ueff ) will now become
(α). In Fig. 6.3, this angle of attack is referred to as effective
1/2
angle of attack (αeff ) for a three-dimensional wing. That is, Ueff = Ua2 + w2 (6.4)
αeff = α − αi (6.3)
However, for small downwash (w ≈ 0)
6.1 Introduction 147
L
(a)
Chord line
α Trailing edge
Ua
Di
(b) L α F
Chord line
α eff
α Ua Trailing edge
αi −w
Ueff
Fig. 6.3 Schematic diagram of a a two-dimensional airfoil section, and b a three-dimensional wing
Ueff ≈ Ua (6.5) Again from small-angle approximations
Since the effective angle of attack is lower than the geomet- sin αi ≈ αi (6.9)
ric angle of attack, the lift generated by a three-dimensional
wing is less than the two-dimensional wing (airfoil section) Thus, the drag per unit span becomes
if the downwash is absent. Consequently, a finite wing has to
wi
operate at a greater geometric angle of attack to achieve the Di = −ρUa (y) = L αi (6.10)
Ua
same lift per unit span as compared to its airfoil section.
In addition to lowering the angle of attack, downwash is If the wing experiences variable distribution of circulation
also associated with an additional drag penalty. We know that over the entire span, as sketched in Fig. 6.4, the total lift on
the local lift force acts normal to the local relative wind speed, the wing by semi-span (b) can be written as
and the tilt caused by the downwash also tilts the lift force by
the same angle αi , as depicted in Fig. 6.3. Consequently, the ˆb
local lift has a component along the freestream direction, i.e., L= ρUa (y) dy (6.11)
there will be a drag created due to downwash. This additional −b
drag is known as induced drag (Di ). In addition, the lift per
unit span produced by the wing is and the induced drag will be
L = ρUa (y) cos αi (6.6) ˆb

wi
Di = ρUa (y) dy (6.12)
Ua
Introducing small-angle approximations −b
cos αi ≈ 1 But αi (y) = Uwai is the induced angle, and thus the above
expression can be rewritten as
The lift per unit span becomes
ˆb
L = ρUa (y) (6.7) Di = ρUa (y) αi (y) dy (6.13)
−b
Similarly, the drag per unit span on the wing will be
which is the required expression for the induced drag (Di ) in
Di = ρUa (y) sin αi (6.8) terms of the induced downwash (wi ) and the circulation ().
Γ− dΓ
Γ0
Γ
y
f(y) dΓ
x
s
Fig. 6.4 Spanwise load distribution over a straight wing
6.1.1 Relation Between Trailing Edge Vortices creation and ill effects. Although today this misconception
and Spanwise Load Distribution is clarified, both danger and mystery have remained charac-
teristics of many vortex flows. For examples, typhoons and
The change in circulation from section to section is equal hurricanes are considered as the biggest natural menace due
to the strength of the vortices shed between these sections. to their devastation effects over the large portion of Asia and
Figure 6.4 shows that the circulation around the section y is America. Besides, the great astronomical riddles, such as the
and it is reduced to − d at the section y + dy. Thus, origin of the solar system and structure of our galaxy (Milky
between the sections y and y + dy, the trailing edge vortices Way), are still unsolved vortex problems.
of strength d will be shed. If γ (y) is the strength of the So what is a vortex? The answer to this question is neither
vortices per unit length, then straightforward nor unique. However, a vortex is generally
described in two ways. The first definition states that a vortex
d = −γ (y) dy (6.14) is the rotational motion of a multitude of fluid particles around
a common center; however, the paths of the rotation of these
Next, consider the influence of the trailing vortex filaments of particles need not be circular but may be asymmetrical as
strength d shed from the section at y. At some other point well.
y1 along the span the downwash created will be given by For a plane in which the pathlines of fluid elements are
γ (y) dy perpendicular to the axis of rotation, by first definition the
dw (y1 ) = − (6.15) vortices thus formed are referred to as cylindrical vortices.
4π(y − y1 )
However, most of the vortices in nature are having the path-
Note that, the above equation is derived by using Eq. (6.53). lines that are not perpendicular, in fact, skewed to the axis of
Integrating above equation between the limits −b and b, we rotation, and thus form a spiral shape. This is because the fluid
get particles have a component of velocity parallel to the axis of
ˆb rotation; consequently, these vortical flows are called spiral
1 γ (y) dy
w (y1 ) = − (6.16) vortices.
4π (y − y1 )
−b The second definition of the vortex was given by Cauchy
and Stokes. They referred to the angular velocity of any matter
This is the spanwise load distribution over the entire wing. (fluid particles in our case) at a point in the space (flow field)
as the vorticity. Though each vortex has the vorticity, vice
versa is not always true. For instance, the parallel shear flow,
6.2 Concept of Vortex as sketched in Fig. 6.5, has vorticity but it is not a vortex.
From the beginning of mankind till date the vortices have

fascinating humans. In day-to-day activities, everyone has 6.2.1 Forced Vortex
observed the whirling of leaves when strong wind blows or
the eddies at the surface of a water body. Until recently, they We know that, in rigid body rotation, all the elements are in
were through that the supernatural powers are behind their combined translational and rotational motions and there is no
6.2 Concept of Vortex 149
θ θ
Fig. 6.5 Parallel shear flow with vorticity, without forming a vortex
relative motion between the individual elements. When a fluid

vθ = (6.18)
is rotated about a common axis at constant angular velocity, 2πr
the entire fluid mass moves as if it were a solid body. The
where is the circulation around a closed contour obtained in
individual particle, though, may be in motion but they are not
counter-clockwise direction. Note that, at the center (r = 0),
being deformed as no shear stresses are acting between the
the tangential velocity tends to infinity. Thus, a singularity
fluid layers. This type of fluid rotation is known as the forced
exists at the center of a free vortex.
vortex.
In addition, the difference in velocity between the inward
The forced vortex behavior becomes dominant when the
and external edges of a fluid particle being conveyed in the
viscous forces acting between the fluid layers are greater
flow field makes it hold its rotational orientation irrespective
than the inertial forces. Since the ratio of inertia force to
of its location in the vortex. Consequently, a free vortex is
viscous
force is expressed in terms of the Reynolds num- additionally called an irrotational vortex. To maintain a free
ber Re = ρUμa L , a low Reynolds number flow indicates vortex flow, no external energy supply is needed; the flow
relatively larger viscous stresses, signifying the tendency to particles rotate under certain energy added earlier to them.
behave more like a solid body, that is, similar to a forced vor- Since there is no exchange of energy happening between the
tex. Consequently, in a forced vortex flow, the angular velocity outer source and the free vortex, and without dissipative loss
(ω) of rotation remains constant and the tangential velocity of mechanical energy, the total energy of the flow remains
(vθ ) increases linearly with radius of rotation (r). Thus, constant.
From the above discussion on the forced and free vortex
vθ = ωr (6.17) flows, an interesting observation can be made. In a forced
vortex flow, the angular velocity of rotation is assumed to be
In this vortical flow, the fluid particles in addition to transla- constant while the circulation is not considered; however, in
tional motion about a common axis spin about their own axis an irrotational flow, the circulation around each contour in the
as well. Hence, a forced vortex is a case of combined transla- vortex is assumed to be same but the angular velocity is not
tional and rotational motions. For this reason, a forced vortex taken into consideration.
is also called the rotational vortex. To maintain the solid body
like rotation of a forced vortex, a continuous supply of energy 6.2.3 Vortex Line, Vortex Tube, and the Vortex
or torque is essential. Filament
A vortex line, also called vorticity line, is defined as a line

6.2.2 Free Vortex whose tangent is everywhere parallel to the local vorticity
→
−
vector ζ . Thus, it is essentially a line that points in the
For a flow with high Reynolds number, the effect of viscosity direction of a vortex vector. A vortex line in vorticity field is
becomes small. In the limit, for excessively high Reynolds analogous to a streamline in the velocity field. Moreover, a
number, the flow can be conveniently assumed as inviscid. vortex line associated with a fluid is orthogonal to a streamline
This type of fluid rotation is called a free vortex. associated with a fluid particle. The vortex lines are defined
According to Newton’s second law of motion, in the by the following relation:
absence of all influential forces, the momentum of a body
will be conserved. Under steady conditions, if no influential dx dy dz
= = (6.19)
forces are present the angular momentum associated with a ζx ζy ζz
free vortex flow will remain invariant. It can be easily shown
that, for a free vortex flow, the tangential velocity (vθ ) is where ζx , ζy , and ζz are the components of the vorticity vector
inversely proportional to the radius (r). That is, along x-, y-, and z-directions in Cartesian space. Furthermore,
similar to a bundle of streamlines which is called a stream- where ∀ represents the volume of a subsection of the vortex
tube, a bundle of vortex lines is referred to as a vortex tube. It tube and S is its bounding surface. Introducing Eq. (6.20) into
is indeed a cylindrical tube in space whose surface elements (6.21), we get
are composed of vortex lines passing through a closed curve, ¨ ˚
defined by the circumference of an arbitrarily selected tube’s −
→− → →−
− →
ζ .dS = ∇ . ζ d∀ = 0 (6.22)
cross section. The circulation around this curve is propor-
S ∀
tional to the number of vortex lines that pass through the area
enclosed within the curve. By definition, the vorticity vector is oriented normal to its
The vortex tube strength, also called the vortex flux, is the bounding surface, i.e., S and along the sides of the tube.
integral of the vorticity across a cross section of the tube. Interestingly, in Eq. (6.22) the only nonzero components come
Since the vorticity is divergence free, the vortex flux is the from top and bottom surfaces of the vortex tube. Thus,
same at every cross section along the tube. In an inviscid
¨ ˆ ˆ
flow, a vortex tube moves with the fluid, that is, a vortex tube −
→− → −
→− → −
→− →
associated with some fluid elements will remain associated ζ .dS = ζ .dS + ζ .dS = 0 (6.23)
with those elements. Finally, a vortex filament is the limiting S S1 S2
case of a vortex tube as the cross-sectional area of the tube
shrinks to zero. That is, a vortex filament is the vortex tube of where S1 and S2 are areas of the surface bounding the vol-
ume ∀ at the bottom and the top, respectively.
− Also, note that
infinitesimal cross section. →
the area vector n̂ and the vorticity vector ζ are in oppo-
site direction at S1 , while they are in same direction at S2 .
6.3 Helmholtz’s Theorems of Vortex Motion Therefore, Eq. (6.23) can be written as
¨ ¨ ¨
In aerodynamics, the three-dimensional motion of fluid in −
→− → −
→ −
→
ζ .dS = − ζ .n̂dS + ζ .n̂dS = 0 (6.24)
the proximity of vortex filaments is described by Helmholtz’s
s s1 s2
theorems, named after physicist Hermann von Helmholtz
(1821–1894). They are valid for inviscid and incompressible or
flows and the flows in the absence of nonconservative forces. ¨ ¨
These theorems, indeed, opened the door for the renowned −
→ −
→
ζ .n̂dS = ζ .n̂dS (6.25)
Prandtl (1921) to propose the classical lifting line theory, one
s1 s2
of the most valuable contributions in aerospace science.
It has been observed that, due to viscous effects, the Further, by using Stokes curl theorem (Eq. (3.42)), we have
strength of a vortex reduces gradually. Since all the real fluids
¨ ˛
possess viscosity, consequently, the Helmholtz’s theorems are −
→ −
→ −
→
not applicable to real fluids. The four Helmholtz’s theorems ζ .n̂dS = v .dS (6.26)
are described as follows. S C
¸ →− →
But, by definition C − v .dS = , where is the circulation
6.3.1 Helmholtz First Theorem around the closed curve C bounding the area S. Thus, if C1
and C2 are the closed contours bounding the surfaces S1 and
The strength of a vortex tube, defined by the circulation around S2 , respectively, then Eq. (6.25) shows the desired result.
the circumference of any tube cross section, remains constant
along its length. This theorem is applicable to any fluid in C1 = C2 (6.27)
which the vorticity field is divergence free, i.e.,
That is, the circulation (strength of the vortex tube) is constant
→−
− → − → −
→ →
∇. ζ = ∇. ∇ ×−
v =0 (6.20) along its length.
Proof This theorem can be proved by using Gauss divergence

−
→ 6.3.2 Helmholtz Second Theorem
theorem. Recall Eq. (3.43) and substitute the vector field V
−
→
as the vorticity ζ . This theorem state that a vortex line can never end in a fluid
¨ ˚ abruptly. The vortex lines and vortex tubes should form a
−
→− → →−
− →
ζ .dS = ∇ . ζ d∀ (6.21) closed path. They must extend to infinity and start (or end) at
S ∀ solid boundaries.
6.3 Helmholtz’s Theorems of Vortex Motion 151
−
→ − → →
6.3.3 Helmholtz Third Theorem ζ = ∇ ×−
v
According to this theorem, an inviscid, incompressible, and The step-by-step procedure adopted will be described next.
barotropic ρ = ρ (p) fluid that is initially irrotational will We know that for an incompressible fluid, the continuity equa-
remain so forever. In turn, such a fluid cannot produce vor- tion (Eq. (3.69)) is given by
ticity. An exception to this is observed for the inviscid and
−
→−
incompressible flow across a bow-shock, where vortices are ∇ .→
v =0 (6.32)
produced.
Since the divergence of velocity for an incompressible fluid is
Proof This theorem can be proved readily by using Kelvin’s zero, using vector calculus, the velocity −
→
v may be expressed
circulation theorem (Sect. 5.4). It states that, for a barotropic as the curl of some other vector field, say of −
→
g (r, t). Thus,
fluid in motion under conservative body forces, circulation we write
−
→ −
→ →
() around an arbitrary closed curve, moving along with the v = curlg = ∇ × − g (6.33)
Dt = 0. From this the-
fluid, is independent of time, that is, D
orem, we can write It is known that the curl of any gradient of a vector is always
¨ −
→ zero, the vector −→
g is undefined to the extent of the gradient
D ∂ζ → −
− → → −→
= + ∇ × ζ ×−
v .dS = 0 (6.28) of the scalar function of position and time. From Eq. (6.33),
Dt ∂t it follows that
S
−
→ − → −
− → →
Since the above result is valid for any arbitrary area S, thus to ∇ ×→ v = ∇ × ∇ ×− g
→ −
− →→
= ∇ ∇ .− g − ∇2− →
hold the equality, the integrand itself must be zero. Thus, we
g (6.34)
have
−
→
∂ζ → −
− → →
=− ∇ × ζ ×−
Let us further assume that
v (6.29)
∂t
−
→−
∇ .→
g =0 (6.35)
or
−
→
∂ζ → → −
− →
= ∇ × −
Note
→ that the above assumption is permitted as the function
v × ζ −
→
g −
(6.30)
∂t r , t is undefined to the extent of a gradient vector. From
Eqs. (5.2), (6.34), and (6.35), we get
−
→
Therefore, if vorticity is zero initially, i.e., ζ = 0, then it will
−
→ → −
→
be zero for all the time. ∇2−
→
g = −∇ × −
v =−ζ (6.36)
This is the Poisson’s equation for vector −→g , which, in turn,

6.3.4 Helmholtz Fourth Theorem
can be considered as a vector potential. By using Eq. (6.33),
the velocity field may be obtained, once −→g is determined as
This theorem states that, for a vortex line of unique identity,
a solution of equation (6.36). →
the ratio of vorticity to the product of fluid density and the
In Cartesian space, if the vector −
→
g −r , t and the vorticity
length of the line remains constant as time proceeds. That is, → −
− →
vector ζ s , t are expressed in terms of their respective
ζ components as
= constant (6.31)
ρl −
→
g = gx , gy, gz (6.37)
Thus, if a vortex line is stretched, the vorticity increases. −
→
ζ = ζx , ζy, ζz (6.38)
In turn, Eq. (6.36) can be written as

6.4 Biot and Savart Law of Vortex Motion
∇ 2 gx = −ζx
For many aerodynamic applications, it is desired to calculate
the velocity induced by a concentrated distribution of vorticity ∇ 2 gy = −ζy (6.39)
(a vortex) with an arbitrary orientation. To obtain the velocity ∇ gz = −ζz
2
v (r, t) in terms of the vorticity ζ (r, t), we must invert the
equation given by Eq. (5.2).
(r − s)
ζ
r
Fig. 6.6 Coordinates used in the derivation of velocity induced from a vortex distribution
The solution† of Eq. (6.36) can be expressed as ∇ 2 φ = −ζx (r)

→ − ∇φ = ∇gx
˚ −
→ ζ →s ,t r − r1
g −
−
→ 1
r ,t = − d∀ (6.40) ∇ψ = − 3
4π →r −−→
s r −−
−
→ →
R r 1
→ →
− ∇ 2 ψ = 0 everywhere except at r = r1
where ζ − s , t d∀ is an element of the vortex distribution
where it becomes infinite
at the point s and R in the region in which the vorticity is
distributed, as shown in Fig. 6.6.
Since the point r = r1 is a singular point, in the sense that
† Proof: ψ, ∇ψ, and ∇ 2 ψ become infinite at that point, we now set a
small sphere of radius σ around that point with center at r1
We wish to obtain the solution for the following equation.
and apply Green’s theorem in the region contained between
→ →
− the sphere and an arbitrarily drawn large surface χ. To cover
g =−ζ −
∇2−
→ r all the space, we remove χ to infinity and shrink the sphere
to the point r1 . By doing so, we obtain
or in Cartesian space, the following set of equations:
→ ˚ ‹
∇ 2 gx = −ζx −
ζx (r) r − r1 ∇gx
r − → d∀ = −χ→∞
lim gx + .n̂dS
→ −
R r − r 1 −→
r −−→r 1
3 −→
r −−→r 1
−
→ χ
∇ 2 gy = −ζy r ‹
→ r − r1 ∇gx
∇ 2 gz = −ζz −r − lim
σ→0
gx
−→
r −−→ 3 + −
r 1
→
r −−→
r 1
.n̂dS
σ
→ (6.43)
Consider all space and assume that the field − g −
→ r , t van-
ishes sufficiently strongly at infinity (Fig. 6.7). Moreover, it Since we have earlier assumed that the field dies out suffi-
is sufficient to consider the solution to one of the equations ciently strongly at infinity, we require that both gx and ∇gx
of the system, given in Eq. (6.39). We will use the nomencla- vanish, as we approach toward infinity, in such a way that
ture as shown in Fig. 6.6 and apply the Green’s theorem to the limit of the surface integral over χ vanishes as χ → ∞.
construct the solution. Thus, Consequently, the limit of integral over χ is set equal to
˚ ‹ zero.
−
→ →
−
φ∇ ψ − ψ∇ φ d∀ =
2 2
φ ∇ ψ − ψ ∇ φ .n̂dS In spherical coordinates, consider the integral over the
∀ s sphere with origin at the point r1 . That is,
(6.41) ¨
where ∀ denotes the volume and S is the area of the bounding ∂gx
− gx + σ sin θdθdφ
surface. In Eq. (6.41), we can identify φ for the above equation ∂σ
using gx (r), as given in Eq. (6.39). In addition, let us consider
the function ψ as In the limiting case, for σ → 0 the above integral gives
→ − −1
ψ (r) = −
r −→
r 1 (6.42) −4πgx (r1 )
Therefore, we write Thus, we get

6.4 Biot and Savart Law of Vortex Motion 153
(r − r 1)
r
r1 σ
Fig. 6.7 Coordinates used in obtaining the solution of Poisson’s equation for vector −
→
g
˚ → →
ζx −r → → −
− ζ −s ,t
d∀ = 4πgx − → 1
− r1 dv r , t =
curlr − d∀ (6.47)
→
r −− → r 1 4π → r −−→s
R
Note that the subscript r on the curl emphasizes that the curl
Note that, in the above integral, the integration variable is r
is to be taken with respect to the coordinates of the point r.
and if we interchange the roles between r and r1 , the solution
Consider a vortex filament of strength (circulation), as
of equation (6.36) may be written as
sketched in Fig. 6.8. Select a volume element d∀ of this fila-
˚ →
1 ζ − r 1 ment as the cylinder formed by a cross-sectional surface n̂dS,
gx (r) = −→ d∀ (6.44)
4π r −− →r 1 and the length of an element is dl along the filament (Fig. 6.8).
R The contribution to the vector potential −→g at a field point −
→
r,
−
→
from the vortex element at s , is given by
Now the variable
→ of integration
→ is r1 . Likewise, the other com-
ponents, gy − r and gz − −
→
r , are obtained and their solutions → −
− → 1 ζ s →
−
are similar to that of gx −→ dg r = − n̂dS. dl (6.48)
4π →
r −− →
r . From these results, we can con-
s
clude that the solution for −→g is
˚ −
→
−
→ 1 ζ (r1 ) However, the elemental length vector dl can be written
g (r) = − d∀
4π →
r −− → r 1 as
−→
R −
→ ζ
dl = − dl (6.49)
→
Replacing r1 = s, we obtain the same solution as given by ζ
Eq. (6.40).
and
−
→
Now, by substituting Eq. (6.40) into (6.33), the velocity field ζ .n̂dS =
is obtained as
⎡ − ⎤ Equation (6.48) can be rewritten as
˚ →
−
→ → ζ ,
⎣− →
−
→ s t
v = ∇ ×−
1
g = ∇ × − d∀⎦ (6.45) −
→
4π →r −− → s → −
− → dl
dg r = − (6.50)
4π r − −
→ →
R
s
−
→
Suppose elemental vector dg is the contribution to −
→
g at −
→
r
−
→ −
→ The contribution to the velocity at the point r from the element
due to vortex element ζ d∀ located at s . of the filament is given by
−
→
→ −
− → 1 ζ s ,t −
→
dg r , t = → − d∀ → −
− −
→
dv →
(6.46) dl
4π −
r −→s r = ∇r× − (6.51)
4π r − −
→ →
s
→ →
−
Similarly, let the elemental velocity dv −
r , t is the contri- −
→
While taking the curl, dl and −→
s are considered to be fixed.
bution to −
→v at −→
r .
In turn, the above equation becomes
dl
n Γ
dS
s r
Fig. 6.8 Coordinates used in the derivation of Biot and Savart law
→ −
− → →
→ −
− → → −
dl × r − s where q = −q . Using the geometry, as shown in Fig. 6.9, the
dv r = → − 3 (6.52)
4π − r −→
s
above equation can be rewritten in scalar form as
sin θ
This is the well-known Biot and Savart law of vortex motion. dvi = dl (6.55)
The velocity at location −
→r due to whole vortex filament is 4π q2
obtained by integrating Eq. (6.52) over the entire length of the −
→
filament. Thus, Here again, dl = dl . From the figure, we can deduce the
following relations:
ˆ −
→ −
−
→ dl × →r −−→s
v (r) = − 3 (6.53) d = q sin θ
4π →r −−
→s
(6.56)
and
6.4.1 Application of Biot and Savart Law: d
= tan (π − θ) = − tan θ
Velocity Induced by a Straight Vortex l
Filament or
d
In this section, we will calculate the velocity induced by a l= − (6.57)
tan θ
straight line segment by using Biot and Savart law (Eq. (6.53)).
From Helmholtz’s first theorem, we know that a vortex line Differentiating the above expression with respect to θ gives
can neither start nor end in a fluid except at the solid bound-

aries. Here, we wish to calculate the induced velocity by a d
dl = dθ (6.58)
segment that is a section of a continuous vortex line. This sin2 θ
vortex segment has constant circulation () along the length
of the line and it is placed at an arbitrary orientation in Carte- Introducing above relations into Eq. (6.55), we obtain
sian space, as depicted in Fig. 6.9. Note that the direction of
induced velocity is obtained by applying sin2 θ d
the right-hand curl dvi = × sin θ × dθ
rule.1 Also, let the distance − → r −− →s , between the vortex 4π d2 sin2 θ
segment and the point
− Q, given in Eq. (6.53) is −
→q . Thus, the
→ −
→ or
velocity induced dv by a segment dl on this line at the

point Q can be written as dvi = sin θdθ (6.59)
4πd
−
→ −
−
→ dl × →q Assume the line segment ends, denoted by A and B, are sub-
dv = (6.54)
4π q3 tending the angles θ1 and θ2 , respectively, at the point Q. The
above equation can be integrated over the segment (A to B)
of the straight vortex line as
ˆθ2
1 Inthe expression of Biot and Savart law for calculating the induced
−
→ vi = sin θdθ
velocity, if we curl our fingers from the line segment dl toward r̂, the 4πd
θ1
thumb shows the direction of induced velocity.
6.4 Biot and Savart Law of Vortex Motion 155
y Γ
P A B
O
θ1 θ θ2
x
q
d
dvi
Q
l
Fig. 6.9 Velocity induced at point P by a straight vortex filament
Solving the above integral, we finally get 6.5 Vortex System and the Evolution
of Prandtl’s Lifting Line Theory

vi = (cos θ1 − cos θ2 ) (6.60)
4πd The concept of vorticity is essential to explain a variety of
fluid flow phenomena. Remember that a vorticity is essen-
This result is of utmost importance in aerodynamics, which
tially a measure of the moment of momentum of the fluid
states that the velocity induced by a straight vortex line seg-
particle about its own center of mass. Suppose if a small
ment is a function of its strength , the distance d, and the
sphere of fluid rotating about an axis passing through its center
view angles θ1 and θ2 . Using this result, we will now con-
freezes and turns into a solid sphere. During the freezing, the
sider two particular cases to derive the induced velocity in the
moment of momentum would be conserved and so it turns
proximity of the vortex line.
out that the angular velocity of the solid sphere, just at the
time of its formation, is exactly half of the vorticity of the
6.4.1.1 Velocity Induced by an Infinite Vortex fluid before freezing. Further, the dynamical theorems of the
Segment Kelvin and Helmholtz relate the changes in the vorticity of the
A vortex is called infinite vortex when its both ends stretching fluid particle to the moments of the forces acting on the fluid
to infinity. It is indeed a two-dimensional case in which θ1 = 0 particle.
and θ2 = π. Thus, from Eq. (6.60), the velocity induced at a Consider an airfoil which begins at rest and moves through
point in the vicinity of an infinite vortex becomes the air, as shown in Fig. 6.10. The circulation around any
curve, say C, in the flow field is to begin with and remains

vi = (6.61) zero (C = 0), as shown in Fig. 6.10a. Consequently, the flow
2πd
remains irrotational everywhere, except in the thin boundary
6.4.1.2 Velocity Induced by a Semi-infinite Vortex layer adjacent to the airfoil where the vorticity is produced by
Segment viscosity.
As name suggests, a semi-infinite vortex is that vortex when It is a curious fact that the viscosity helps generate lift on
one of its end stretches to infinity. For example, suppose the the airfoil. To understand this, let us consider an airfoil at rest
vortex line segment end B, shown in Fig. 6.9, stretches to with no circulation anywhere in the flow field, as sketched
infinity (θ2 = π). Thus, from Eq. (6.60), the velocity induced in Fig. 6.10a. When the airfoil begins to move the viscous
at a point located directly opposite to the segment end A, such effect near the trailing edge results in shedding of a vortex,
that θ1 = π2 , will be called the starting vortex (Fig. 6.10b), the circulation around a
closed
C1 enclosing the starting vortex will be nonzero
curve
C1 = 0 as illustrated in Fig. 6.10c. If we consider a larger
vi = (6.62) curve (C2 ) in the inviscid region that encloses both the vortex
4πd
and the airfoil, then Kelvin’s circulation theorem (Eq. (5.23)
which is exactly the half of the velocity induced by an infinite tells us that the circulation around the curve C2 will also be
vortex segment. zero, i.e., C2 = 0. Suppose a curve C1 encloses the airfoil,
(a) C
(b)
Bound vortex
Ua
Starting vortex
C’ 1
ΓC 2 C2
(c)
Bound vortex
Ua C1
Starting vortex
Γ C’
1
ΓC 1
C3 Bound vortex
(d) Γ C 3= 0
Ua
C4
A
Γ C 4= 0
Ua = 0 Starting vortex
Γ C5 = 0 Induced velocity ( w )
C5 Β
Bound vortex shedded as

starting vortex
B
Induced velocity ( wΑ)
Fig. 6.10 Schematic diagram of the bound vortex on an airfoil and the starting vortex downstream
then the circulations around the vortex and the airfoil neces- an equal and opposite circulation around the airfoil. This cir-
sitate that culation around the airfoil may be ascribed to a fictitious vor-
tex, which is necessary for the production of the lift. Since
C1 + C1 = C2 (6.63)
this fictitious vortex remains attached to the airfoil in motion,
it is known as bound vortex. Note that a bound vortex is dif-
Since C2 = 0, thus
ferent from a free vortex, which is free to move throughout
C1 = −C1 (6.64) the flow domain. Further, if the airfoil stops then the bound
vortex, denoted by B as shown in Fig. 6.10d, also sheds down-
This is an important result of aerodynamics. It states that the stream again because of the viscous action at the sharp trailing
circulation around the starting vortex, denoted by A, requires edge. Since the viscosity has acted on either of these vortices,
6.5 Vortex System and the Evolution of Prandtl’s Lifting Line Theory 157
the circulations around the curves C4 and C5 , respectively, surface. Moreover, since the pressure is a continuous func-
enclosing the starting vortex and the shed bound vortex in tion, it must be equal at the wing tips. In turn, the tendency
anticlockwise direction are found to be nonzero. But the curve of the fluid elements is to move from the lower surface of the
(C3 ) encloses both vortices, and passing through on the fluid wing around the wing tips to the upper surface, so that pressure
where friction has never acted, the circulation remains zero. becomes equal both sides. This motion of fluid elements from
Thus, the equal and opposite circulations of vortices produce the region of higher pressure to the region of lower pressure
zero net vorticity through the entire bounded region C3 . The around the wing tips is known as flow slip. Because of this,
vortices A and B of equal and opposite strength are convected there is a spanwise component of flow on the upper surface of
away downstream by the fluid. The vortex A will induce the the wing in the direction, from tip to root. Similarly, there is
downwash wA on the vortex B, which induces downwash wB a spanwise component of the flow on the lower surface in the
on the vortex A, as illustrated in Fig. 6.10d. direction, from root to tip. Also, there exists the approaching
Moreover, when all the torque producing factors (pressure freestream of air moving toward the wing. When these two
forces, body forces, and viscous forces) in Kelvin’s theorem motions are combined, the resultant motion is such that the
are absent, the geometrical interpretation of the dynamics of flow is inclined inward, i.e., toward the wing root on the top
vortex motion can be given in terms of Helmholtz’s theorem. surface, and it is inclined outward, i.e., toward the wing tip
For more details on these theorems, the readers are advised to on the bottom surface. Consequently, for the flow leaving the
refer Sect. 6.3. However, for convenience, an essence of these trailing edge of the wings, the pathlines from the upper are
four Helmholtz’s theorems is given as follows. indeed inclined to the pathlines from the lower surface, in turn,
the helical paths or vortices result. A complete line of vortices
• The first theorem states that the strength (i.e., circulation) is shed from the wings; the vortex strength being strongest at
of a closed-loop filament remains constant along its length. the wing tip and reduces swiftly to zero at mid-span. After
• The second theorem states that a vortex line can never end traveling a short distance downstream, these vortices merge
in a fluid suddenly; rather, they form closed loops. and form two distinct helical shape vortices. These vortices,
• The third theorem states that a fluid line which at any shed from the trailing edge of the wings, are known as the trail-
instant is a vortex line remains a vortex line. ing edge vortices or trailing vortices, as shown in Fig. 6.11.
• The fourth theorem states that, on a vortex line of fixed Figure shows that the rotations of vortices shed from either
identity, the ratio of vorticity to the product of the fluid wings are opposite in nature, one being clockwise and other
density and the length of the vortex line remains constant being anticlockwise. However, irrespective of their orienta-
as time proceeds. tions, the direction of vortex rotation is such that the flow is
always turned toward the wake region (at the rear of the air-
The above laws can be verified by considering the exam- craft). Furthermore, these trailing vortices induce additional
ple of smoke rings, which are indeed the vortex filaments. downward component of velocity behind the wing (within
Since these filaments can never end, they form closed loops. the wingspan). If we are fixed and view these vortices from
Besides, on a vortex line, each vortex induces a component the rear of an aircraft in motion, we observe that all the fluid
of velocity on the other vortices (downwash); consequently, within the system of vortices is moving downward (referred
these smoke rings are able to convect away on their own. to as downwash), whereas all the fluid external to this system
Therefore, these smoke rings which are indeed the fluid mass will move upward (known as upwash). The velocity gradient
are carried away with the vorticity. Moreover, these vortic- established between the upwash and the downwash is some-
ities (smoke rings) never die but they fade away due to the times very large and produce extreme motions, which might
viscosity. be disastrous for an aircraft flying through it as it may lead to
So far in our discussion on the vorticity, we have con- structural failure.
sidered an airfoil which is assumed to be a wing of infinite Although all these vortices discussed above can be treated
span (i.e., two-dimensional wing). But the wings of actual air- separately, in real flows, they all are present together and form
craft are of finite length consisting of a series of airfoils with a complex system of vortices. The horseshoe vortex model is
varied cross section. That is, an actual wing is essentially a simplified delineation of this complex vortex system for a
three-dimensional in shape. wing of finite span. It is indeed a closed loop that consists of
When an aircraft is in motion, its wings experience a higher a line of bound vortices of constant circulation traveling with
pressure region on its lower surface (or pressure surface) and the wing and essential for producing the lift, and two lines of
a lower pressure on its upper surface (or suction surface). This trailing wing tips vortices, in turn, having a shape resembling
pressure difference indeed creates a lift on the wing. Note that vaguely to a horseshoe, as shown in Fig. 6.12. Finally, the
the wings experience a pressure higher than the freestream vortex loop is closed by the line of starting vortices shed,
static pressure at the bottom surface, while they experience a when the wing begins to move through the fluid. This starting
pressure lower than the freestream static pressure at the top vortex gets dissipated under the action of viscosity.
Rear view of an airplane

Wing−tip vortex
shedding
Upwash Upwash
Downwash
Fig. 6.11 The wing tip vortex shedding, the upwash, and the downwash behind a flying aircraft
Ua
Bound Trailing vortex

vortex
Downwash
Starting vortex
Upwash
Fig. 6.12 A simplified horseshoe vortex model
Furthermore, despite the fact that simplified vortex model becomes a lifting line. Also, it was assumed that the lifting
is unreasonable as it expect a uniform circulation at all the sec- line which consists of the distributed bound vortices has the
tions of a finite wing, nevertheless, with a reduced effective circulation variable in order to account for the change in lift
wingspan, yet same mid-plane circulation provides a satisfac- (or circulation) over the span. Consequently, the free vortices
tory model for the flows induced sufficiently far away from (trailing vortices) are formed and shed downstream toward
the aircraft. the infinity. These trailing vortices are indeed responsible for
inducing the downwash at the lifting line which, in turn, mod-
ifies the local angle of attack at each airfoil section.
6.5.1 Lanchester–Prandtl Wing Theory In addition, the aerodynamic forces are essentially
described at the aerodynamic center than any other location
The lifting line theory is a mathematical model which predicts on the chord line, because it is the point about which the
the aerodynamic properties for a wing of finite span. This pitching moment coefficients do not vary with the angle of
theory was independently developed by Lanchester (1907) attack, i.e., dC
dα = 0. Moreover, an airfoil can be numerically
M
and by Prandtl (1918) and therefore, it is also known as simulated by placing the continuous distribution of vortices
the Lanchester–Prandtl lifting line theory or the Lanchester– along the surface of the airfoil section, which forms a vortex
Prandtl wing theory. In this model, it was assumed that the sheet. This sheet, which can be discretized into a series of
strength of vortices reduce over the whole wing because they straight segments of varied strengths, is called the vortex
continuously shed downstream a trail of vortices forming a panel. This panel, which is a straight line segment in two
vortex sheet from the entire trailing edge, rather than, just dimensions, indeed forms a quadrilateral panel for a three-
from the wing tips. dimensional wing where each element of the vortex sheet
This theory simulates a finite wing by distributing an infi- (in streamwise direction) can be viewed as a straight vortex
nite number of horseshoe vortices over the complete span. segment running along the y-axis (i.e., spanwise direction).
The bound vortices placed on a line passing through the aero- To obtain the downwash induced by the trailing vortices
dynamic centers of each airfoil section of the wing, in turn, at some point downstream of the wing, let us consider a
z
wb
+β
Ua
LE
TE
Γ (y) x
−β
c (y)
2
z Γ max
c (y) y
+β θ2
Γ = Γ (y)
(b) Quadrilateral panel dy − d Γ (y) dy
Ua dy
θ1 O
(c/2, y)
−β
L
x
Panel span = 2 β
y
Downwash distribution − d Γ (β) d β
Ua over the panel dy
+b
Di
−b
(a) Finite wing of span ’2b’
(c) Horseshoe vortex system
Fig. 6.13 Replacement of a finite wing with the system of horseshoe vortices
quadrilateral panel of dimensions; c (y) along the x-axis and and

2β along the y-axis, as shown in Fig. 6.13. In the quadrilateral
(y − β)
panel, consider a spanwise section of a typical horseshoe
cos θ2 = − 1/2
vortex system with strength (β) = − d(β) c 2
dy dβ is + (y − β)2
2
chosen at the quarter-chord location c(y) 4 and its effect
From above substitutions, wb becomes
is examined at the control point 3c(y) 4 . In addition, a ⎡ ⎤
continuous distribution of bound vortices on this vortex (β) ⎢ (y + β) (y − β) ⎥
segment give rise to a lifting line that induces a downwash in dwb = − ⎣
c 2 1/2
+ 1/2 ⎦
c(y) 2 c 2 2
4π 2 + (y + β) + (y − β)
the vicinity of it. By using Eq. (6.60), the downwash (wb ) 2 2
(6.66)
at the control point c(y)
2 , y due to this segment is given by
2
However, for a wing of large aspect ratio, the term 2c can
(β)
wb = − (cos θ1 − cos θ2 ) (6.65) be neglected, and hence, the above equation reduces to
4π c(y)
2 ⎡ ⎤
(β) d (β) ⎦
From the geometry of the horseshoe vortex system, as shown dwb = − (2) = − ⎣ (6.67)
c(y)
in Fig. 6.13, we have 4π 2 2π c(y)
2
(y + β) Note that this is the downwash induced by a single horse-

cos θ1 = 1/2 shoe vortex system. However, the lifting line consists of an
c 2
2 + (y + β)2 infinite number of horseshoe vortices, each with an infinitely
⎡ ⎤
small strength d. Thus, considering the contributions from ˆb − d(y) dy
(y) ⎦ 1 dy
all these horseshoe vortices, the induced downwash for the −⎣ − + Ua α = 0
2π 2 c(y) 4π (y − β)
complete lifting line, evaluated at y, will become −b
⎡ ⎤ (6.72)
(y) Dividing the above equation by Ua gives
wb = − ⎣ ⎦ (6.68)
2π c(y) ⎡ ⎤
2 ˆb − d(y) dy
(y) 1 dy
−⎣ ⎦− +α=0
Next, we wish to evaluate the downwash induced by the trail- 2π c(y) U a
4πUa (y − β)
2 −b
ing vortices shed by the wing. Since the strength of bound (6.73)
vortices may vary along the spanwise direction and since This equation is called Prandtl’s lifting line integro-
Helmholtz’s theorem states that no vortex filament can begin differential equation for the spanwise load distribution
or end in a flow abruptly, except at the solid boundaries, there- (y), which indeed is a combination of different angles, as
fore, any local change in this circulation must be shed down- illustrated in Fig. 6.3. Comparing the terms of Eqs. (6.3) and
stream. Indeed, a change in circulation along the lifting line of (6.73) gives the induced angle of attack (αi ) and effective
a horseshoe vortex system will shed downstream in the form angle of attack (αeff ) as
of trailing vortex and the strength of each trailing vortex is
equal to the change in circulation along the lifting line. More- ˆb − d(y) dy
1 dy
over, for a wing of finite span, the trailing vortex shedding αi = (6.74)
4πUa (y − β)
is assumed to be shed from a semi-infinite vortex line of the −b
strength (y) = d(y) dy dy. Thus, the strength of a vortex (y)
αeff = (6.75)
line starting at y = β will be (β) = d(β) dy dy. 2π c(y)
Ua
2
π
By using Eq. (6.62) and for θ1 ≈ 2 and θ2 ≈ π, the down-
wash at the control point can be written as Furthermore, the variations in the lift distribution over the
wing span change the downwash induced at each airfoil sec-
(y) 1 tion, which in turn alters the induced angle of attack (αi ). Con-
w (y) = (6.69)
4π (y − β) sequently, the effective angle of attack (αeff ) also varies over
the wingspan, i.e., αeff = αeff (y). Thus, an airfoil located at
which is the downwash induced by a single trailing vortex
y = β (Note that this is also the location of a trailing vor-
filament. To calculate the downwash induced by the com-
tex filament for the quadrilateral panel considered, as shown
plete wing, we must integrate equation (6.69) over the entire
in Fig. 6.13b) will experience an effective angle of attack
wingspan, i.e., in the limits −b ≤ y ≤ b, yields
αeff (β). Besides, for a cambered airfoil section, αeff is mea-
sured from the zero-lift angle (αL=0 ) which is in contrast
ˆb − d(y) dy
1 dy to
a symmetric airfoil, where it is measured from the origin
wi (y) = (6.70)
4π (y − β) CL (α=0) . From Eq. (5.84), for a cambered airfoil section,
−b the lift curve slope (a0 ) is given by
Furthermore, from the boundary condition, we know that no dCL
flow can penetrate the solid surface of a wing. That is, the a0 = = 2π
dα
flow velocity normal to the wing surface will essentially be
zero. Clearly, in our case, the sum of normal velocity com- For an airfoil section located at y = β, from the above relation,
ponents induced by the wing (wb ), trailing vortices (wi ), and the local lift coefficient (CL ) can be expressed as
the normal velocity component of the freestream flow (Ua α)

will have to be zero on the solid surface of the wing. CL αeff(β) − CL (αL=0 ) = a0 αeff(β) − αL=0 = 2π αeff(β) − αL=0
(6.76)
wb + wi + Ua α = 0 (6.71)
However, since CL (αL=0 ) = 0, the above equation
where the subscripts ‘b’ and ‘i’ stand for the bound and the reduces to
induced influences, respectively. Furthermore, by substituting
CL αeff(β) = 2π αeff(β) − αL=0 (6.77)
Eqs. (6.68) and (6.70) into (6.71), we get
y 2 21
Moreover, the local lift coefficient (CL ) for the airfoil section
located at β is defined by (y) = max 1 − (6.84)
b
L where max is the value of circulation at the origin O. Rewrit-
CL αeff(β) = (6.78)
qa c (β) ing the above expression yields
2
where L is the lift per unit span, qa = 21 ρUa2 is the freestream (y) y 2
dynamic pressure, and c (β) is the chord length of the panel. + =1 (6.85)
max b
From the Kutta–Joukowski theorem (Eq. (4.140)), the above
expression can be rewritten as which is essentially the equation of an ellipse. Thus, Eq. (6.84)
is more precisely referred to as the elliptical distribution of
ρUa (β) 2 (β) circulation. However, since the circulation curve forms only
CL αeff(β) = 1 = (6.79)
2 ρUa c (β)
2 Ua c (β) a half ellipse over the wingspan, consequently, the area under
this curve is πb2max , i.e., half of the total ellipse area.2
Suppose that a two-dimensional airfoil section has a local lift In order to obtain the total downwash at a specific location
curve slope a0 and the local effective angle of attack as seen y = β, by the complete trailing vortex sheet, let us differen-
by it is αeff . Introducing Eq. (6.79) into (6.77), we obtain tiate equation (6.84) with respect to y, we get
(β) y 2 −1/2
αeff(β) = + αL=0 (6.80) d (y) max y
πUa c (β) = 1− −2 2 (6.86)
dy 2 b b
Recalling the generalized form of the effective angle of attack Introducing Eq. (6.86) into (6.70) yields
from Eq. (6.3) and considering the airfoil section to be located
at y = β, the local effective angle of attack turns into ˆb
max y 2 −1/2 y
wi (y) = 1− dy (6.87)
αeff (β) = α (β) − αi (β) (6.81) 4πb2 b y−β
−b
By using Eqs. (6.80) and (6.74), the above equation yields the Note that, for y = β, the above integral has a singularity
geometric angle of attack at y = β as which can be removed from Cauchy principle value, provided
´∞ 2 y
ˆb d(y) that the limb→∞ −∞ 1 − yb y−β dy exists; it can be
(β) 1 dy dy
α (β) = + αL=0 (β) + solved by using the following substitution:
πUa c (β) 4πUa (β − y)
−b
(6.82) y = b cos θ (6.88)
which is the fundamental equation of the Prandtl’s classi-
cal lifting line theory. It reflects that, for a given cambered On differentiating the above, we get
airfoil section, the local geometric angle of attack is equal
to the sum of the effective angle of attack and the induced dy = −b sin θdθ (6.89)
angle experienced by the section locally. For a given finite
wing with known geometric angle of attack in a freestream of From Eq. (6.88), the elliptic distribution of circulation, given
given velocity, (β) can be readily calculated from Eq. (6.82). by Eq. (6.84), in transformed coordinates becomes
Also, a general boundary condition that satisfies the solution 1/2
of equation (6.82) is given by (θ) = max 1 − cos2 θ
(y = ±b) = 0 (6.83) or
(θ) = max sin θ (6.90)
6.5.2 Symmetric Elliptical Aerodynamic Load Introducing Eq. (6.88) into (6.87) and noting that at wing tips,
Distribution θ = π (y = −b) , and θ = 0 (y = b), yields
Consider the distribution of circulation, as shown in 2 In Cartesian space, suppose a general ellipse is defined by the equation
Fig. 6.13c. By inspecting the figure, the circulation (y) x2 2
+ by2 = 1, where a and b are the lengths of semi-major and semi-
a2
over the finite wing can be expressed as minor axes, respectively. The area enclosed by this ellipse will be πab.
ˆ0 To obtain the total lift, denoted by L, for the entire wingspan,

max −1/2 b cos θ (−b sin θ) dθ
wi (θ) = 1 − cos2 θ the above equation should be integrated between the limits;
4πb2 b (cos θ − cos β)
π −b ≤ y ≤ b. That is,
or ˆb y 2 21
ˆπ L = ρUa max 1− dy (6.97)
max cos θ b
wi (θ) = − dθ (6.91) −b
4πb (cos θ − cos β)
0
Solving above,
Note that the form of above integral is similar to a Glauert
πb
identity, i.e., L= ρUa max (6.98)
4
ˆπ
cos nθ π sin nθ This is the total lift acting on the complete wingspan for
dθ = (6.92)
cos θ − cos θ0 sin θ a symmetric elliptic lift distribution. Next, we wish to cal-
0 culate the overall induced drag (Di ) on the wing. Recall
Eq. (6.13)
where n = integers (including “0”). Comparing Eq. (6.91)
with (6.92), we find n = 1. Thus, ˆb
ˆπ Di = ρUa (y) αi (y) dy
cos θ π sin θ
= =π (6.93) −b
(cos θ − cos β) sin θ
0
Introducing Eqs. (6.84) and (6.95) into above gives
In turn, Eq. (6.91) yields the total induced downwash for the
ˆb y 2 21
entire wingspan as ρmax
2
Di = 1− dy (6.99)
4b b
max −b
wi (θ) = − (6.94)
4b
Substituting yb = cos θ in the above integral and by changing
Finally, by substituting Eq. (6.94) into (6.2), the induced angle the limits of integration yields
of attack (αi ) becomes
ˆ0 ˆ0
max ρmax
2 ρ 2
αi = (6.95) Di = −b sin θdθ = − max
2
sin2 θdθ
4bUa 4b 4
π π
Note that Eqs. (6.94) and (6.95) are very important results of
Interchanging the limits of above integral, we have
finite wing theory. They state that, for a symmetric elliptical
lift distribution over the wingspan, both induced downwash ˆπ
(wi (θ)) and induced angle (αi ) are not a function of “y”; ρmax
2
Di = sin2 θdθ
rather, they are constant along the span. Also, both wi (θ) and 4
0
αi approach zero if 2b → ∞, that is, when the wingspan is
infinitely large. On solving, the overall induced drag for a symmetric elliptical
loading is given by
6.5.2.1 Aerodynamic Loads and Coefficients for an
Elliptic Lift Distribution π 2
Di = ρ (6.100)
From Kutta–Joukowski
theorem (Eq. (4.140)), the lift per unit 8 max
length L is expressed as
Further, by definition, the lift coefficient is expressed as

L (y) = ρUa (y)
L
CL = (6.101)
2 ρUa S
1 2
Substituting the expression for (y) from Eq. (6.84), we get
y 2 21 where S is the planform area of the wing. Substituting the

L (y) = ρUa max 1 − (6.96) total lift L from Eq. (6.98), the expression for CL is
b
π b max lift downstream in the freestream direction, and thereby intro-

CL = (6.102)
2 S Ua duces a component of the lift along the freestream direction.
This additional force component is known as induced drag or
Solving the above equation for max yields
more often called drag due to lift. Since the same pressure
2Ua SCL imbalance present between the top and bottom surfaces of a
max = (6.103) wing is the mechanism behind generating the lift as well as
πb
the wing tip vortices (i.e., induced drag); thus, the induced
Furthermore, the coefficient of induced drag can be defined drag and the lift on a finite wing are undoubtedly interlinked
as to each other.
In addition, since the coefficient of induced drag is
Di inversely proportional to the aspect ratio of the wing,
CDi = (6.104)
1
2 ρU 2
aS therefore, in order to reduce the induced drag we must have a
wing with the highest possible aspect ratio. However, use of
Inserting the expression for an overall induced drag (Di ) given very large aspect ratio wings is difficult as they have severe
by Eq. (6.100), in the above equation, gives CDi as structural limitations.
π max
2
CDi = (6.105)
4 SUa2 6.5.3 Symmetric General Aerodynamic Load
Distribution
Substituting Eq. (6.103) into (6.105), we obtain
A more general solution for the unknown spanwise distribu-
1 S 2 tion of the circulation (y), which satisfies Eq. (6.82), can be
CDi = C (6.106)
π b2 L described by a trigonometric expansion (Fourier expansion)
A very important and useful geometric parameter for a finite series in θ coordinates as
wing is the aspect ratio (AR), which is defined as the ratio ∞

of square of the semi-wingspan and the planform area of the (θ) = 2bUa An sin nθ (6.110)
wing. Thus, n=1
b2 which is obtained through the substitution, as shown in

AR = (6.107) Eq. (6.88). Moreover, we see that for n = 1, 3, 5,…, i.e., odd
S
numbers, the terms in the above Fourier expansion are indeed
In turn, Eq. (6.106) takes the form symmetric in nature, as depicted in Fig. 6.14. For this reason,
the general distribution of circulation on a finite wing (shown
C2L in Eq. (6.110)) is known as the symmetric general load distri-
CDi = (6.108)
πAR bution.
Also, note that all the terms in this Fourier series satisfy
This is an important result of finite wing theory, which reveals the Kutta condition (Eq. (5.11)) at the wing tips, i.e.,
that the coefficient of induced drag is directly proportional to
the square of the coefficient of lift and inversely proportion (0) = (π) = 0 (6.111)
to the aspect ratio of the wing. In other words, the coefficient
of induced drag is a function of both the lift coefficient and The expression for ddy to substitute into Eq. (6.82) can be
aspect ratio. That is, obtained by differentiating equation (6.110) with respect to
y. That is,
CDi = CDi (CL , AR) (6.109)
∞

d d dθ 1
However, note that this dependence of the induced drag on = =− 2bUa nAn cos nθ
dy dθ dy b sin θ
the lift is not unusual because the induced drag is essentially n=1
an outcome of the lift produced on a finite wing. We have dis-
or
cussed earlier that the production of wing tip vortices due to
pressure difference between the suction and the pressure sur- ∞

d cos nθ
faces causes the variations in the geometric angles of attack = −2Ua nAn (6.112)
dy sin θ
(α) at each airfoil section. A decrease in α basically tilts the n=1
A 1 sin θ
A n sin n θ
Circulation ( Γ )
A 3 sin 3 θ
0 π π
2
A 5 sin 5θ
Polar coordinate ( θ )
Fig. 6.14 Schematic representation of the first three terms of the general lift distribution (θ)
Introducing Eqs. (6.110) and (6.112) into (6.82) yields By comparing Eq. (6.82) with (6.116), we can write the
∞ induced angle (αi ) as
2b
α (β) = An sin nβ + αL=0 (β)
πc (β) ∞
1 sin nβ
n=1
αi (β) = nAn (6.117)

∞ 2 sin β
ˆ0 −2Ua nAn cos nθ
sin θ
n=1
1 n=1
+ (−b sin θ) dθ
4πUa b (cos β − cos θ) This is the induced angle of attack for a symmetric general lift
π distribution on a finite wing, which is obtained at a particular
(6.113)
location (y = β). Since 0 ≤ β ≤ π, and since it is a dummy
variable, a more generalized expression for induced angle (αi )
Solving the above equation and interchanging the limits of
is written by replacing β with θ. Thus,
integration, we have
∞
∞
1 sin nθ
∞
ˆπ nAn cos nθ αi (θ) = nAn (6.118)
α (β) =
2b
An sin nβ + αL=0 (β) +
1 n=1 2 sin θ
n=1
πc (β) 2π (cos θ − cos β)
n=1 0
(6.114) Furthermore, from Eqs. 4.140 and (6.110), the sectional or
local lift coefficient (CL ) is obtained as
By changing the order of integration and the summation,
∞
Eq. (6.114) becomes L ρUa (θ) 4b
CL = = = An sin nθ
2 ρUa S
1 2
2 ρUa c (θ)
1 2 c (θ)
∞ ∞ ˆ π n=1
2b 1 cos nθ (6.119)
α (β) = An sin nβ + αL=0 (β) + nAn
πc (β) 2π (cos θ − cos β)
n=1 n=1 0 Similarly, from
Eq. (6.10), the sectional- or local-induced drag
(6.115) coefficient CDi can be defined as
Interestingly, the integral term on the right side of Eq. (6.115)
CDi = αi CL (6.120)
is a Glauert Identity, which can be replaced with its value
given by Eq. (6.92), therefore
Introducing Eqs. (6.118) and (6.119) into the above expres-
∞
∞

sion, we get
2b 1 sin nβ ∞ ∞
α (β) = An sin nβ + αL=0 (β) + nAn
πc (β)
n=1
2
n=1
sin β 2b sin nθ
(6.116) CDi = An sin nθ nAn (6.121)
c (θ) sin θ
n=1 n=1
Furthermore, the lift coefficient for the complete wing is or

obtained by integrating equation (6.119) in the spanwise
∞

direction, i.e., in the limits −b ≤ y ≤ b. Thus,
CDi = πAR nAn2 (6.129)
n=1
ˆb ˆπ
∞
CL (y) c (y) dy 2b2
CL = = An sin nθ sin θdθ Rewriting Eq. (6.129) by expanding the Fourier series up to
S S
−b 0 n=1 the first term, we obtain
(6.122)
b2 ∞

But = AR is the aspect ratio of a wing. Again, interchang-
S
ing the order of integration and the summation in Eq. (6.122) CDi = πAR A12 + nAn2
n=2
yields
∞

An2
∞ ˆπ = πAR A12 1+ n (6.130)
n=2
A12
CL = 2AR An sin nθ sin θdθ (6.123)
n=1 0 By using Eq. (6.125) and from small algebraic manipulations,
the above expression becomes
From any standard text on the integral calculus, we can recall
that ∞

C2L An2
CDi = 1+ n (6.131)
ˆπ π πAR A12
for n = 1 n=2
sin nθ sin θ= 2 (6.124)
0 for n = 1 or
0
From Eq. (6.124), it is clear that the Fourier series expansion, C2L
CDi = (1 + δ) (6.132)
shown in Eq. (6.123), eventually leads to the first term only πAR
as all other terms are vanished. In turn, Eq. (6.123) fetches
∞ An2
where δ = n=2 n A12
. Since δ ≥ 0, (1 + δ) ≥ 1. Further,
CL = πA1 AR (6.125)
if we define e = 1+δ 1
, e ≤ 1 is the span efficiency factor.
This is the wing lift coefficient for the general lift distribution In turn, Eq. (6.133) can also be expressed as
on a finite wing.
In similar fashion, the wing coefficient of
induced drag CDi is obtained by integrating equation (6.121) C2L
CDi = (6.133)
for the complete wingspan. That is, we write πeAR
ˆb ˆπ
∞
∞

which is the wing coefficient of induced drag for a sym-
Cdi (y) c (y) dy 2b2
CD i = = An sin nθ nAn sin nθ dθ metric general load distribution on a finite wing. Comparing
S S
−b n=1 n=1
0 Eqs. (6.108) and (6.133), we find that for an elliptic load dis-
(6.126)
tribution, the span efficiency factor becomes unity, i.e., e = 1
Since Eq. (6.126) consists of two summation series, thus from
and δ = 0, whereas it is less than or equal to one (e ≤ 1) for a
the following standard results of integral calculus
general load distribution. In turn, a finite wing under the ellip-
ˆπ tic load distribution will experience the minimum induced
π
2 for m = n drag.
sin mθ sin nθ= (6.127)
0 for m = n
0
we conclude that the terms involving the product of dissimilar Summary

coefficients such as A1 A2 , A3 A4 , A2 A3 ,...etc., in Eq. (6.126)
will be vanished and only terms with similar coefficients, i.e., The properties associated with an airfoil section which are
A12 , A22 , A32 , ėtc., will be finite. Thus, Eq. (6.126) gives indeed the same as the properties of a wing of infinite span.
These properties are different for the wings of finite span
∞
πb2 2 attached to a real aircraft. This is because, unlike an airfoil
CDi = nAn (6.128) which is a two-dimensional object, a wing is essentially a
S
n=1 three-dimensional body, that is, there will be a component of
flow in the spanwise direction. That is, the flow over the wings For the symmetric aerodynamic load distribution, defined as
is three-dimensional in nature and hence their aerodynamic 2 21
(y) = max 1 − yb , a summary of important relations
properties are quite different from those of its airfoil sections.
The downwash produced by shedding trailing edge vor- is described below.
tices from the wing tips and its effect on the inclination of the For a symmetric elliptical lift distribution over the
local relative wind has two major consequences on the local wingspan, both induced downwash and induced angle are
airfoil section. The actual angle of attack as seen by the air- constant along the span.
foil locally is, in fact, lower than the geometric angle of attack
max
(α). This angle of attack is referred to as effective angle of wi (θ) = −
4b
attack (αeff ) for a three-dimensional wing. That is,
max
αeff = α − αi αi =
4bUa
Also, the effective freestream velocity (Ueff ) will now The total lift acting on the complete wingspan for a symmetric
become elliptic lift distribution is
1/2
Ueff = Ua2 + w2 πb
L= ρUa max
4
However, for small downwash (w ≈ 0)
and the expression for CL is
Ueff ≈ Ua
π b max
CL =
Besides, the downwash induced by these trailing edge vor- 2 S Ua
tices from the wing tips leads to an additional component of
The overall induced drag for a symmetric elliptical loading is
drag known as induced drag.
given by
The concepts of vortex sheets and vortex filaments are
advantageous in evaluating the aerodynamic characteristics π 2
of wings of finite span. From a directed line segment (dl) of Di = ρ
8 max
a vortex filament, the induced velocity can be calculated by
using the below mentioned Biot–Savart law. In addition, the coefficient of induced drag is
ˆ −
→ −
dl × →r −−→ π max
2 1 S 2
−
→ s CDi = = C
v (r) = − 3 4 SUa2 π b2 L
4π →r −−
→s
This relation can also be written as
The velocity induced by a straight vortex filament of the finite
length is given as C2L
CDi =
πAR

vi = (cos θ1 − cos θ2 ) 2
4πd where AR = bS is the aspect ratio of a finite wing.
In the lifting line model, developed by Ludwig Prandtl, a wing For the symmetric general aerodynamic load distribution,

∞
is numerically described by an infinite number of horseshoe given by (θ) = 2bUa An sin nθ, a summary of impor-
vortices and these bound vortices pass through the aerody- n=1
tant relations is described as follows.
namic centers of the airfoils, which in turn creates the lifting
The generalized expression for the induced angle (αi ) is
line. Besides, the trailing edge vortices starting at the lifting
line and shed downstream toward the infinity are basically ∞
1 sin nθ
responsible for inducing the downwash at the lifting line, and αi (θ) = nAn
consequently, modify the local angles of attack. The circula- 2 sin θ
n=1
tion distribution (y) is calculated from the accompanying
relation For a wing of finite span, the lift coefficient is given by
ˆb d(y) dy
(β) 1 dy CL = πA1 AR
α (β) = + αL=0 (β) +
πUa c (β) 4πUa (β − y)
−b
Summary 167
and the coefficient of induced drag is (a) elliptical

∞
(b) straight rectangular
C2L C2L
CDi = πAR nAn2 = (1 + δ) = (c) straight tapered
πAR πeAR (d) tapered sweptback
n=1

∞ An2 2. For a wing of finite span, if the lift coefficient is doubled
where δ = n=2 n A12
and e = 1
1+δ is the span effi- then the induced drag coefficient
ciency factor.
(a) will remain same
(b) will be increased two times
(c) will be increased three times
Exercises (d) will be increased four times
Descriptive Type Questions 3. For an aircraft with tapered straight wing, the root chord
and tip chord are 0.9 and 0.5 m, respectively. If the total
1. Show that the integral on the right-hand side of Kelvin’s wingspan is 14 m, then the aspect ratio of wing will be
circulation theorem vanishes, if the fluid is barotropic.
2. Prove that the elliptical spanwise lift distribution leads to
a constant downwash along the span. (a) 10
3. Consider a vortex sheet, where the velocities above and (b) 20
below the sheet are 6 and 4 ms−1 , respectively. The ele- (c) 0.1
ment of this vortex sheet is 0.4 m wide, which rolled up (d) 30
into a line vortex after some time. Calculate the strength
of the vortex. 4. The wing sweep back in an aircraft provides which type
4. Determine the wing loading of an aircraft weighing 2000 of static stability?
kg, if the wing planform area is 18 m2 .
(a) longitudinal stability
5. Plot the curve between the induced drag coefficient and
(b) lateral stability
the lift coefficient for an elliptical load distribution over
(c) directional stability
the wing of aspect ratio 7.5.
(d) both (b) and (c)
6. If the aspect ratio of a glider, having elliptical plan-
form wing, is 6.5. Calculate the change in minimum 5. If the aspect ratio of a wing of finite span is increased, the
angle of glide, if the aspect ratio is doubled. Assume lift curve slope will
CD = 0.03 + 0.07C2L .
7. An airplane is flying at 150 ms−1 in a steady level flight. (a) decrease
If the aircraft weighs 80 kN and have elliptical wing of (b) increase
span 16 m, determine the induced drag. (c) remain same
8. Describe the motion of a vortex pair (a) when their circu- (d) cannot say
lations are equal and in the same direction, and (b) when
their circulations are equal but in opposite directions. 6. When the “flap” of a wing is deflected, then the lift curve
9. Determine the flow field due to (a) a vortex filament, slope will
which is in the form of a circular ring (vortex ring), (b) a
plane vortex sheet, which is formed by a distribution of (a) increase
horseshoe vortex filament, and (c) an infinite row of point (b) decrease
vortices of equal strength distributed along a straight line (c) remain same
at equal intervals. (d) increase or decrease depending upon the shape of the
10. Find the path of a vortex bounded by two walls perpen- flap.
dicular to each other.
7. For any given length, the swept back wings experience
lift than the straight wings.
Multiple Choice Questions (a) less
(b) more
1. According to Prandtl’s lifting line theory, which of the (c) same
following shape of the wing has minimum induced drag? (d) cannot say
8. Different types of aircraft have different stalling charac- (c) eight times the original
1
teristics. Which of the following is TRUE in a “benign (d) 8th of the original
stall”?
Keys
(a) One wing stalls slightly before the other.
(b) The nose drops gently and the wings remain level 1. (a)
throughout. 2. (c)
(c) The nose rises, pushing the wing deeper into the 3. (b)
stalled state. 4. (d)
(d) All the above are FALSE. 5. (b)
6. (c)
9. In the normal cruise motion (straight level flight) of an
7. (a)
aircraft, if the cruise speed is increased by three times
8. (b)
then the induced drag will be
9. (d)
(a) increased by three times 10. (c)
(b) increased by nine times
1
(c) 3rd of the original
1
(d) 9th of the original
References
10. For a finite wing, if the coefficient of lift is doubled while
the aspect ratio is reduced to half then the induced drag Lanchester FW (1907) Aerial flight, Vol. 1: aerodynamics. Archibald
Constable & Co, London
will be Prandtl L (1918) Tragftii geltheorie, 1. Mit-teilung. Nachr Ges Wiss
Gottingen 1918: 451–477. Also NACA TN9 (1920)
(a) doubled Prandtl L (1921) Uber die Eindringungsfestigkeit (Harte) plastischer
(b) four times the original Baustoffeund die Festigkeit von Schneiden. ZAMM 1: 15–21
von Helmholtz H (1868) Über discontinuierliche Flüssigkeits-
Bewegungen. Monatsberichte der Königlichen Preussische Akademie
der Wissenschaften zu Berlin, vol 23. pp 215–228
Panel Methods
7
The panel methods are the numerical techniques for solv- eled by placing the vortex panels in the form of straight line
ing incompressible potential flows over thick two- and three- segments possessing the strength, which are distributed over
dimensional geometries by replacing them with mathematical panels that model the airfoil contour itself.
models comprising of sources and vortex panels. The surface
to be analyzed will be depicted by the panels consisting of
sources and vortices, in turn, these methods are referred to as 7.2 Hess and Smith Method
source panel and vortex panel methods, respectively. In this
chapter, we will discuss both source and vortex panel methods In our earlier discussion, we have seen that there are a large
along with their applications. number of panel methods depending upon the types of singu-
larities, variation of singularity strength, and the distribution
of these singularities. The first practical panel method was
7.1 Introduction used by Hess and Smith, which is based on the distribution of
sources and vortices on the surface of the geometry.
In our discussion on potential flows, we have seen that the If a body is placed in an uniform flow, the total potential
two-dimensional potential flow around an oval body (Rankine function (φ) for the flow can be obtained by the superposition
oval) is generated by the superposition of a source, a sink of the potential functions of the uniform flow (φa ), the source
in a uniform flow along x-axis. In a similar fashion, three- (φS ) and the vortex distributions (φV ) as
dimensional flow around a Rankine body can be generated by
φ = φa + φ S + φ V (7.1)
placing a point source, a point sink in a uniform flow along x-
axis. Thus, it can be shown that the potential flow around some
The source distribution and vortex distribution possess the
configurations may be modeled by placing sources and sinks
varying strengths k (s) and ζ (s), respectively, where s is the
in the interior of the body. One can model the potential flow
arc length coordinate that spans the complete surface of the
around the slender bodies of any shape by placing distributed
airfoil section. The potential created by the distributions of
sources/sinks along x-axis inside the body surface. However,
sources and vortices is
the result obtained by these techniques will be approximate
unless the body is infinitely thin and the slope of the body k (s)
φS = ln rds (7.2)
is very small. This method was first practically conceived by 2π

Hess and Smith in developing Douglas aircraft in the early ζ (s)
φV = − θds (7.3)
1960s. 2π
A two-dimensional panel method refers to the computa-
tional method for calculating the flow around any geometry where the various parameters are defined in Fig. 7.1. Note
such as the wing section. These methods are based on the that, the above integrals is to be carried out along the com-
replacement of the geometry of wing sections by singularity plete surface of the airfoil section. Also, from superposition
panels, such as source panels, doublet panels, and vortex pan- principle any such distribution sources/sinks and vortices sat-
els. Essentially, the wing section is divided into piece-wise isfy Laplace’s equation, however, we are interested in finding
straight line segments or panels or boundary elements and those k (s) and ζ (s) such that the flow tangency boundary con-
vortex sheets of constant strengths are placed on each panel. dition and Kutta conditions are satisfied. Out of many choices
For non-lifting surfaces, a surface distributions of sources is possible, Hess and Smith made a simplified assumption. They
applied, whereas for lifting case the airfoil section is mod- have assumed the vortex strength to be constant over the whole
170 7 Panel Methods
P(x,y)
y
s θ
i th Panel
x
Ua
α
Fig. 7.1 Airfoil geometry parameters for panel methods
Nodes
n−2
n−1 n
1
4 3 2
8 6 5
Panel 7
Fig. 7.2 Schematic illustration of panels and nodes
airfoil section and from the Kutta condition the value of vor- t̂i = cos θi î + sin θi ĵ
tex strength is fixed. Although, the source strength which is
constant over each panel is allowed to vary from panel to The midpoint coordinates of the panel are given by
panel but, together with the constant vortex distribution the
flow tangency boundary condition is satisfied everywhere. xi + xi+1
xi =
Using the panel decomposition from Fig. 7.2, we can dis- 2
cretized Eq. (7.1) as
and
yi + yi+1
n
yi =
k (s) ζ(s) 2
φ = Ua (x cos α + y sin α) + ln r − θ ds
j 2π 2π
j=1 The velocity components at the above midpoints are
(7.4)
Consider ith panel located between the ith and (i + 1)th node.
ui = u xi , yi
The orientation of the panel to the x-axis is given by
and
yi + 1 − yi
sinθi = vi = v xi , yi
li
xi + 1 − xi
cosθi = The flow tangency boundary condition is written as −
→
u .−
→
n =
li 0, or for each panel
where li is the length of the panel. The normal vector to this
−ui sin θi + vi cos θi = 0
panel will be (Fig. 7.3)
where i = 1, 2, 3, . . . . . ., n. While the Kutta condition is sim-
n̂i = − sin θi î + cos θi ĵ
ply given as
and the tangential vector to this panel is given by
u1 cos θ1 + v1 sin θ1 = −un cos θn − vn sin θn (7.5)
7.2 Hess and Smith Method 171
y∗
y x∗
lj j+1
x θj
j
Fig. 7.3 Sketch of plane coordinate system
1 2 1
Note that the negative signs are due to the tangential vectors ∗ t=l
ln x∗ − t + y∗2
2
usij =− |t=0j
at the first and the last panels are in opposite directions. The 2π
source strength on the panel i is given by ∗ 1 y∗ t=l
vsij = tan−1 ∗ | j (7.9)
2π x − t t=0
k (s) = ki where i = 1, 2, 3, . . ., n
On interpreting the results of Eq. (7.9) with simple geometry
Thus, the velocities at the midpoints of each panel can be given in Fig. 7.4, it is revealed that
obtained by superimposing the contributions from all the
∗ 1 rij+1
sources and the vortices, i.e., usij =− ln
2π rij

n
n
∗ ε l − ε0 βij
ui = Ua cos α + kj usij + ζ uvij vsij = = (7.10)
2π 2π
j=1 j=1
n n Similarly, for the velocity field induced by the vortex on the
vi = Ua sin α + kj vsij + ζ vvij (7.6) panel j at the midpoint of panel i will be
j=1 j=1
lj
where usij and vsij are the components of velocity at the mid- ∗ 1 y∗ βij
uvij =− dt =
point of panel i induced by a source of unit strength at the 2π (x∗ − t) + y∗2
2 2π
0
midpoint of panel j. Similarly, we can interpret for uvij and
vvij . According to the following transformation, the integra- lj
∗ 1 x∗ − t 1 rij + 1
tion of Eq. (7.4) can be performed by noticing that the local vvij =− dt = ln (7.11)
2π (x∗ − t) 2
+ y∗2 2π rij
velocity components can be expanded into an absolute one as 0
follows:
Using Eq. (7.6) and undoing the local coordinate transforma-
∗ ∗
u = u cos θj − v sin θj tion of Eq. (7.7)
v = u∗ sin θj + v∗ cos θj (7.7)

n
Bij kj + Bin+1ζ = ci (7.12)
The local velocity components at the midpoint of ith panel j=1
due to a unit strength source distribution on the jth panel is
where
lj
∗ 1 x∗ − t
usij = dt Bij = −usij sin θi + vsij cos θi
2π (x∗ − t)2 + y∗2 ∗

0 = −usij cos θj sin θi − sin θj cos θi
∗

lj + vsij sin θj sin θi + cos θj cos θi
∗ 1 y∗
vsij = dt (7.8)
2π (x∗ − t)2 + y∗2 which yields
0
rij+1
where x∗ and y∗ are the coordinates of the midpoint of panel i 2πBij = sin θi − θj ln + cos θi − θj βij
in the local coordinate system of panel j. Integrating Eq. (7.8), rij
we get
Similarly, for the vortex strength coefficient
172 7 Panel Methods
(x *, y * ) or (x , y )
y*
β ij r ij+1
r ij
εl x*
ε0 j+1
lj
j
Fig. 7.4 Geometric interpretation of source and vortex-induced velocities

n
rij+1 On solving, it is easy to construct the tangential velocities at
2πBin+1 = cos θi − θj ln − sin θi − θj βij
rij the midpoint of each panel according to the following expres-
j=1
sion:
The right-hand side of Eq. (7.12) is given by

n
kj rij+1
Uti = Ua cos (θi − α) + sin θi − θj βij − cos θi − θj ln
ci = Ua sin (θi − α) 2π rij
j=1
n
ζ rij+1
Note that the flow tangency boundary condition gives us n + sin θi − θj ln + cos θi − θj βij
2π rij
equations. Thus, we require an additional relation that will be j=1
obtained by using the Kutta condition in order to solve the

system of equations. From Eq. (7.5), we write with the known tangential velocity components, we can cal-
culate the pressure coefficient at the midpoint of each panel

n according to the following formula:
Bn+1,j kj + Bn+1,n+1ζ = cn+1
j=1 Uti2
Cp (xi , yi ) = 1 − (7.13)
Ua2
After some manipulations, we get
rmj+1
Using the above relation, the force and moment coefficients
2πBn+1,j = sin θm − θj βmj − cos θm − θj ln can be computed assuming that this value of pressure coeffi-
rmj
m=1,n
cient is constant over each panel and by performing the dis-

n
rmj+1 crete summation.
2πBn+1,n+1 = sin θm − θj ln + cos θm − θj βmj
rmj
m=1,n j=1
cn+1 = −Ua cos (θ1 − α) − Ua cos (θn − α) 7.2.1 Line Source Distribution
where m=1,n are carried out only over the first and last Consider the source distribution for a two-dimensional body
panels, and not on the entire range [1, n]. These expressions distributed along the x-axis, as shown in Fig. 7.5. Let ki is the
set up a matrix of the form Bx = c, where the matrix B is of source strength at xi , then the stream function at the generic
size (n + 1) × (n + 1). point P (x, y) is given by

⎡ ⎤⎡ ⎤ ⎡ ⎤ ki y
B11 ··· B1i ··· B1n
B1,N +1 k1 c1 = Ua y + arctan (7.14)
⎢ .. .. ..
.. ⎥⎢ .. ⎥ ⎢ .. ⎥ 2π x − xi
⎢ . . .. ⎥⎢ . ⎥ ⎢ . ⎥ i
⎢ ⎥⎢ ⎥ ⎢ ⎥
⎢ Bi1··· ···
Bii Bi,n+1 ⎥
Bin ⎢ ⎥ ⎢ ⎥
⎢ ⎥⎢ ki ⎥ = ⎢ ci ⎥ where the first part of Eq. (7.14) is due to the uniform flow.
⎢ .. .. ..
. ⎥⎢ . ⎥ ⎢ . ⎥
⎢ . . ..
. ⎥⎢ .. ⎥ ⎢ .. ⎥
⎢ ⎥⎢ ⎥ ⎢ ⎥ If the source is distributed continuously, then we can replace
⎣ Bn1··· ···
Bni Bn,n+1 ⎦⎣ kn ⎦ ⎣ cn ⎦
Bnn summation by an integral in Eq. (7.14). The stream function
Bn+1,1 · · · Bn+1,n · · · Bn+1,n Bn+1,n+1 ζ cn+1 due to the elementary source distribution is given by
7.2 Hess and Smith Method 173

1 y Since we are dealing with potential flow with linear governing
ψS = k (t) arctan dt (7.15)
2π x−t equation, the body boundary condition can be linearized as
and the velocity potential due to the source distribution is dY
uS ∼
= Ua (7.22)
dx
1
φS = k (t) ln x − t 2 + y2 dt (7.16) y
2π In Eq. (7.21), the function, f (t) = acts like a distri-
(x−t)2 +y2
bution. This can be understood by assuming a slender body
The upstream stagnation point makes an angle θ = π and
such that y → ξ. Not surprisingly, when we use it for thin
the downstream stagnation point makes θ = 0o with the x-
airfoil, i.e., for Y → 0, then f (t) acts like a delta function.
axis. Both the stagnation points are located on the x-axis
Hence, we can write Eq. (7.21) as
(y = 0). Thus, for the upstream stagnation point and the
y
downstream stagnation point, we can write arctan x−t =π ∞
y
and arctan x−t = 0, respectively. From Eq. (7.14), if the body k (x) Y
lim vS (x, Y) ∼
= dt (7.23)
is represented by the streamline = 0, then the upstream Y→0 2π (x − t)2 + Y2
−∞
stagnation point is given as
Approximating Eq. (7.23), we have taken k (t) out of the inte-
c
k (t) dt gral and replaced it, by its value at t = x and extended the
=0=0− π (7.17)
2π limit from −∞ to +∞. Equation (7.23) can be simplified as
0
k (x) t−x ∞ k (x) π π
and the downstream stagnation point is given by vS arctan |t=−∞ = − −
2π Y 2π 2 2
(7.24)
c Hence,
k (t) dt
=0=0− 0 (7.18) k (x)
2π lim vs =
0 Y→0 2
Note that, Eq. (7.18) is insignificantly satisfied for a closed From Eq. (7.22), we get
body, the source distribution must be such that as given by
dY
Eq. (7.17). That is, k (x) = 2Ua
c dx
k (t) dt = 0 (7.19)
and Eq. (7.20) gives
0
c
For any closed body, this is a necessary condition for the 1 x−t
uS = k (t) dt (7.25)
source/sink distributions, i.e., the sum of source strength must 2π (x − t)2 + Y2
0
be equal to the sum of all the sinks needed to represent a closed
body. Now for Y → 0, Eq. (7.25) becomes
c
1 dt 2Ua Y
7.2.2 Perturbation Velocity Components Due to uS ∼
= k (t) = dt
Source Distribution 2π (x − t) 2π (x − t)
0
We can calculate the perturbation components of the induced This approximation gives satisfactory results, unless t is not
velocity due to source distribution as sketched in Fig. 7.5, by too close to x. On the surface of slender body considered
using Eqs. (7.15) and (7.16). Y → 0, and thus
c x−ξ x+ξ
1 x−t k (t) (x − t) k (x)
uS (x, y) = k (t) dt (7.20) 2πuS (x, Y (x)) dt + dt
2π (x − t)2 + y2 (x − t) (x − t)2 + Y2
0
0 x−ξ
c
1 y c
vS (x, y) = k (t) dt (7.21) k (t)
2π (x − t)2 + y2 + dt
0 (x − t)
x+ξ
174 7 Panel Methods
Ua
*
*
* P (x,y)
* ki
*
* θi
xi *
c *
*
*
*
*
x
Fig. 7.5 Source distribution over slender body along the x-axis
Performing a coordinate transformation Z = (x − t), the sec- 7.3 Vortex Panel Methods
ond integral in the above equation becomes
The vortex panel method is a technique for computing ideal
ξ flows, i.e., the flows in which the effects of compressibility
−ZdZ
k (x) and viscosity are negligibly small. The source panel method is
Z2 + Y2 (x) suitable for non-lifting flows only, whereas the vortices have
−ξ
circulation and hence the vortex panel method can be used
As the integrand here is odd and the limits are even, therefore, for lifting surfaces. This method essentially models the flow
the integral vanishes. past an airfoil section as the summation of a uniform flow and
a series of vortex panels or vortex sheets arranged to form
c a closed polygon with a shape that approximates, as nearly
Ua Y (t) dt
uS (x, Y (x)) (7.26) as possible, the actual curved shape of the airfoil section.
π (x − t) Here, it is assumed that each panel represents a planar vortex
0
sheet with linearly varying strength in such a way that the end
Recall, the pressure coefficient as strength of each panel is the same as the starting strength of
the next panel.
2
u
Cp = 1 − (7.27)
Va
7.3.1 Panel of Constant Strength
From the linear perturbation theory for streamwise velocity
component, the above equation can be reduced to Consider a panel of constant strength ζ, as shown in Fig. 7.6.
Let an arbitrary point is P(x, y) in the close proximity of the
2us small panel segment of strength d. The differential stream
Cp ≈ (7.28) function can be written as
Ua
1
In Eq. (7.25), if k (t) = constant (K) then for y → ξ, we have dψ = − ln (rλ) d
2π
r
c =− ln (rλ) dλ (7.30)
K dt K x 2π
uS = = ln (7.29)
2π (x − t) 2π x − c
0 Integrating Eq. (7.30) along the length of the panel, we get
From the above equation, it is evident that for constant l

strength source distribution, at x = 0 or c, the induced veloc- ζ
ψ=− ln (x − λ) + y dλ
2 2
ity displays a logarithmic singularity. 2π
0
7.3 Vortex Panel Methods 175
P (x,y)
r1 r2
dΓ θ2
θ1
λ= 0 dλ λ= l
Fig. 7.6 Schematic diagram of a single panel

ζ r2 which eventually leads to
= x ln − l ln (r2 ) + y (θ2 − θ1 ) (7.31)
2π r1
ζn = 0
From Fig. 7.6,
where n is the number of panels. The purpose of this section
y
θ1 = tan−1 is to find all the values of ζ, such that, the upwash at midpoint
x−l of each panel is going to be zero. That is, for all i
y
θ2 = tan−1
x wi = 0 (7.37)
r1 = x2 + y2
Here, we have four constraints for four panels, i.e., say n con-
r2 = (x − λ)2 + y2 straints from Eq. (7.37). In addition, we have another con-
straint from Eq. (7.36), so n + 1 equations are required to
The x-component of velocity in terms of the stream function solve the system but, we have only n variables available. Thus,
is it means that we have an over-constrained system. To solve
∂ψ
u= (7.32) this, let us begin with the influence coefficient.1 The first panel
∂y causes some upwash on the second panel, i.e., the upwash on
the second panel is due to the influence of first panel. If only
Differentiating Eq. (7.31) with respect to y
the first panel is considered, then the coefficient of upwash on
the second panel due to the first one is
∂ψ ζ ζ d r2 d
u= = (θ2 − θ1 ) + x ln − l ln (r2 )
∂y 2π 2π dy r1 dy w2
Cw21 = (7.38)
d
+y (θ2 − θ1 ) (7.33) ζ
dy
Arranging the influence coefficient matrix in such a way that
The second term in Eq. (7.33) has no contribution and hence
becomes zero. Thus, Eq. (7.33) becomes Cw11 Cw12
[Cw ] =
Cw21 Cw22
ζ
u= (θ2 − θ1 ) (7.34)
2π where Cwij = wζji . Once the influence coefficients are deter-
mined, then we need to plug in into the system of equations
Similarly, the y-component of velocity will be
as follows:
[Cw ] {ζ} + Ua {α − θ} = 0 (7.39)
∂ψ ζ r2
v=− =− ln (7.35)
∂x 2π r1
Equation (7.39) shows the set of n-equations for n unknowns
and Eq. (7.36) shows the additional constrain. Therefore, we
7.3.1.1 Application of the Constant Strength Panel
have to drop one of the equations from n equations and
Let us discretize the camber line into the number of panels
Eq. (7.36) have to be plugged in, so that, we can find the
with constant vortex strength ζ. Note that, in vortex panel
−
→ vortex strength ζ. Once ζ is found, then we can plug in back
method the boundary condition U a .n̂ = 0 must be satisfied
into our panel to find the flow over the entire panel.
at all the points. Since the chosen point is the midpoint of each
panel and hence, the upwash over the midpoint of the every
panel is given by w1 , w2 , w3 , and so on. Besides, the Kutta
condition is given as 1 Aninfluence coefficient is a measure of how one panel affects the
ζTE = 0 (7.36) upwash of another panel.
176 7 Panel Methods
ζ3
ζ2 ζ4
w2 w3
w
ζ1 w1
x2 x3 4
x2 x4
Ua x1 x5 ζ5= ζ TE= 0
α
Fig. 7.7 Discretized camber line into the number of panels
7.3.2 Panel with Linearly Varied Vortex Strength ζ2

uψ2 = −y log (r1 ) + y log (r2 ) + x (θ2 − θ1 )
2πl
Consider the linearly varying vortex strength from ζ1 at x1 ,
The net x-component of velocity, by using superposition prin-
and it goes down to zero at x2 . Neglecting any contribution
ciple, is
due to ζ2 , the stream function at any arbitrary point, shown in
Fig. 7.8a is given by (Fig. 7.7) ζ1
u= y log (r1 ) − y log (r2 ) + (l − x) (θ2 − θ1 )
2πl

ζ1 1 1 ζ2
ψ1 = − x (x − 2l) − y2 log (r1 ) + (l − x)2 − y2 log (r2 ) + −y log (r1 ) + y log (r2 ) + x (θ2 − θ1 ) (7.42)
2πl 2 2 2πl

(l − x) l
+ (l − x) y (θ2 − θ1 ) − (7.40) If we set ζ1 = ζ2 = ζ, then Eq. (7.42) becomes
2
ζ
Similarly, assuming the vortex strength from ζ2 at x2 , which u= (θ2 − θ1 ) (7.43)
2π
goes down to zero at x1 (Fig. 7.8b), the stream function at the
same arbitrary point due to ζ2 ignoring the contribution from Interestingly, if the vortex strength is assumed to be constant,
ζ1 will be then Eqs. (7.43) and (7.34) are the same. In a similar fashion,
the y-component of velocities due to ψ1 and ψ2 are obtained
as
ζ2 1 2 1 2
ψ2 = x − y2 log (r1 ) − x − l2 − y2 log (r2 )
2πl 2 2
∂ψ1 ζ1 l
xl vψ1 = − =− − (x − l) log r1 + (x − l) log (r2 ) − y (θ2 − θ1 ) +
+ xy (θ2 − θ1 ) − (7.41) ∂x 2πl 2
2
∂ψ2 ζ2 l
vψ2 =− =− x log (r1 ) − x log (r2 ) + y (θ2 − θ1 ) −
∂x 2πl 2
The superposition of Eqs. (7.40) and (7.41) gives the net
stream function, i.e., and the resultant y-component of velocity becomes

ψ= ψ1 + ψ2 ζ1 l
v=− − (x − l) log r1 + (x − l) log (r2 ) − y (θ2 − θ1 ) +
2πl 2

Recall ζ2 l
− x log (r1 ) − x log (r2 ) + y (θ2 − θ1 ) − (7.44)
2πl 2
∂ψ ∂ψ1 ∂ψ2
u= = + Again, by setting ζ1 = ζ2 = ζ, Eq. (7.44) leads to
∂y ∂y ∂y
∂ψ ∂ψ1 ∂ψ2 ζ r2
v=− =− − v=− ln (7.45)
∂x ∂x ∂x 2π r1
The x-component of velocity due to ψ1 will be
ζ1 7.3.2.1 Application of the Linearly Varied Vortex

uψ1 = y log (r1 ) − y log (r2 ) + (l − x) (θ2 − θ1 ) Strength Panel
2πl
We have seen that wi = 0 for all i. For n panels, we have nζ
and the x-component of velocity due to ψ2 is values and nw values. Recall, the system of equation
[Cw ] {ζ} + Ua {α − θ} = 0
7.3 Vortex Panel Methods 177
(a) ψ1(x,y)
r1
ζ1
(c) ψ(x,y)
x1 x2
ζ1 r1
ψ2(x,y) r2
ζ2
(b) + =
r2
x1 x2
ζ2
x1 x2
Fig. 7.8 Linear superimposition of panels

Consider a linear panel as sketched in Fig. 7.7. The coefficient Now, we can transform from prime axis x , y into (ξ, η)
of upwash on the midpoint of second panel due to ζ1 is as depicted in Fig. 7.9d. Now, we will be using the direction
cosine matrix in order to perform this transformation. That is,
−
→
Cw21 = U a (ζ1 , ζ2 , x1 , x2 , x2 ) .n̂2 (7.46)
ξ = x cos βi + y sin βi
where x2 is the midpoint of second panel. Since we are inter- η = −x sin βi + y cos βi
ested in the first panel alone, we can neglect ζ2 from Eq. (7.46).
Similarly, the coefficient of upwash on the midpoint of the To get the generic point (x, y) into the panel coordinates, we
second panel due to ζ2 is have to do this for each panel separately as each panel has
−
→ −→ its own coordinate system. Once we have ψ (ξ, η), ut (ξ, η)
Cw22 = U a ζ1 , ζ2 , x1 , x2 , x2 + U a ζ2 , ζ3 , x2 , x3 , x2 .n̂2
and vt (ξ, η), then we can transform back from (ξ, η) frame
(7.47) to (x, y) frame. This transformation can be done by using
In the above equation, we are not interested in finding either again the direction cosine matrix, however, this time oriented
ζ1 or ζ3 , and hence they can be ignored. By setting ζ1 = 1 through a negative βi as shown in Fig. 7.9e. That is,
in Eq. (7.46) and ζ2 = 1 in Eq. (7.47) and solving for all ζ,
we can calculate the freestream velocity (Ua ) and the stream u = ut cos βi − vt sin βi
function (ψ) everywhere in the flow field. Once these variables
v = ut sin βi + vt cos βi
are found then subsequently the pressure coefficient, and the
lift and drag coefficients can be easily calculated by using
After this, we have to simply realize that ξ and η refers to x
them.
and y, respectively.
7.3.3 Transformation of Panel Coordinates

Summary
Whenever the cambered airfoil is discretized, the panels are
no more aligned along the x-axis and, in fact, they have some The fluid flow characteristics around any complex three-
slope. Since, all our equations are applicable for the pan- dimensional vehicle configurations such as an aircraft, auto-
els oriented along the x-axis, therefore, it is essential to use mobile, etc., can be determined using the panel methods. The
some suitable transformations for panel coordinates in order aerospace industry has pioneered in developing and adopting
to solve the problem. Let us transform our panel from x-and these methods which are ideal for comparatively easy surface
y-axes as shown in Fig. 7.9a into ξ- and η-axes as sketched in modeling and analysis due to their time effective calculations.
Fig. 7.9b. Before the actual transformation, the intermediate They show superior performance in modeling fully attached,
step is to find the prime axis x and y shown in Fig. 7.9c. high Reynolds number subsonic flows. Also, they are profi-
cient in calculating the lift and form drag on the wing and/or
x = x − xi over an entire aircraft as long as viscous effects are negligibly
y = y − yi small. Although, the panel methods were originally conceived
for incompressible flows (M < 0.3) only, however, by incor-
178 7 Panel Methods
(a) (b)
η
x (x,y)
ξ (ξ,η)
y
ξ i= 0 ξ i+1= l
ηi= 0 ηi+1= 0
(c) (d) (e)
y’ (x,y) y’ u
η ξ
v βi
vt
βi ut
x’ x’
(xi , yi ) (x i+1 , yi+1)
Fig. 7.9 Transformation of panel coordinates
porating compressibility corrections they can be extended to Find the complex potential, if the source is located at a
high subsonic speeds (0.3 < M < 1). distance 0.3l from the bottom plate.
The panel methods are indeed powerful tools in analyz- 6. Consider a vortex panel with linearly varying vortex
ing the complex incompressible flow configurations, but they strength. Find the value of vortex strength at a point, where
experience serious limitations in some situations. They are the local velocity (uS ) is 10 ms−1 and θ2 − θ1 = 5o .
incapable of modeling the viscous effects, boundary layer 7. In a fluid flow, the freestream velocity and the local veloc-
characteristics and the phenomena where the flow separates. ity due to source distribution are 10 and 4 ms−1 , respec-
The lack of modeling the viscosity leads to further con- tively. Using the linear perturbation theory with suitable
straint; they cannot be used in modeling the forced vortices assumptions, find the pressure coefficient.
and/or rotational flows. The panel methods are incapable 8. Consider the vortex panel of constant strength ζ =
in modeling the flows at supersonic and hypersonic Mach 5 ms−1 . Find the y-component of velocity if rr21 = 0.75.
numbers. 9. For a panel with vortex distribution, the x-component of
velocity over the panel is 12 ms−1 . If θ1 = 120o and θ2 =
50o , then calculate the vortex strength of the panel.
Exercises 10. Consider a two-dimensional body of chord 15 mm with
source distribution in the direction along the x-axis.
Descriptive Type Questions Derive an expression for the x-component of induced
velocity at a generic point P (25 mm, 30 mm).
1. For the constant strength source distribution, calculate
the x-component of velocity by the direct integration of
Eq. (7.20). Multiple Choice Questions
2. Using Hess and Smith method, show that the velocities at
midpoints of each panel can be obtained by superimpos- 1. Panel methods are ideal for computing
ing the contributions from all the sources and the vortices.
(a) incompressible viscous flows
3. Consider a vortex panel of constant strength, ζ =
(b) compressible viscous flows
2 mm2 s−1 . Find the stream function for the given
(c) compressible inviscid flows
conditions; λ = 9 mm, r1 = 5.83 mm, r2 = 5 mm, θ1 =
(d) incompressible inviscid flows
30.96o and θ2 = 36.86o , and evaluate its value at an
arbitrary point P (5 mm, 3 mm). 2. The vortex panel method for an airfoil assumes that
4. For a flat plate at an angle of attack (α), calculate the
influence coefficient matrix aij by using the lumped vortex I. The strength of each panel varies linearly over the
element. To solve the problem divide the chord into five panel.
panels of equal lengths and assume the collocation point II. There exists a component of velocity perpendicular
to be located at 41th of chord from the trailing edge. to the airfoil.
5. Consider the fluid flow between two parallel plates sep-
arated by a distance l, due to a source of strength k (s).
Exercises 179
III. The strength of the panel is continuous across the 8. For a body in uniform flow, the velocity potential (φv )
panel joints. created by by the distribution of vortices is
IV. The strength of the panel is discontinuous at the trail-
ing edge. (a) − ζ(s) θds
ζ(s)2π
(b) θds
Choose the correct option from the following. 2πζ(s)
(c) − 2π ds

(a) I and III (d) ζ(s)2π ds
(b) II only
9. In vortex panel method, which of the following boundary
(c) I, II and III
condition must be satisfied at points?
(d) II and IV
−
→
(a) U a × n̂ = 0
3. Consider an airfoil system with N control points giving −
→
(b) U a .n̂ = 0
N equations for the N + 1 strengths. The final equation −
→
Ua
to solve this system is obtained by (c) n̂
=0
(d) −
→
n̂
=0
Ua
(a) Kutta condition
(b) Helmholtz method 10. For a panel with source distribution, the freestream veloc-
(c) The system can be solved without a final. ity and the pressure coefficient are 50 ms−1 and 0.9,
(d) Kelvin’s circulation theorem respectively. The perturbation velocity component due to
source distribution will be
4. The solution of the panel method becomes invalid when
(a) 12.5 ms−1
(a) the flow develops local supersonic flow (b) 18.5 ms−1
(b) when the flow is tangential to the panels (c) 22.5 ms−1
(c) when Neumann boundary conditions are applied (d) 35.5 ms−1
(d) the solution is always valid
Keys
5. It is more accurate to compute the wing lift on the wing
body interface using
1. (d)
(a) constant vortex panels 2. (b)
(b) line sources 3. (a)
(c) constant pressure panels 4. (a)
(d) line doublets 5. (c)
6. (a)
6. In order to simulate a closed body, the sum of the strengths 7. (b)
of all the sources should be the sum of all the sink’s 8. (a)
strength. 9. (b)
10. (c)
(a) equal to
(b) more than
(c) less than
(d) cannot say
7. For a constant strength source distribution, the induced

velocity at x = 0 or x = c has a
(a) linear singularity

(b) logarithmic singularity
(c) parabolic singularity
(d) hyperbolic singularity
Thermodynamics of Fluids in Motion
8
The purpose of the present chapter is to review those aspects 8.1.1 Concept of System, Surroundings,
of thermodynamics that are essential to compressible flows. and the Universe
This is not the slightest bit expected to be a thorough discus-
sion of thermodynamics; rather, it is an overview of only those A system is defined as a quantity of matter or a region in space,
key ideas and relations that will be of direct use in subsequent chosen for study; and the mass or the region outside the system
chapters. For the readers who are already familiar with the is known as its surroundings. A real or an imaginary surface
thermodynamics, this discussion will be a ready reminder of separating the system and surroundings is called the boundary,
some important relations. This discussion is similarly helpful may be fixed or movable. The system and its surrounding are
to those readers who are not comfortable with thermodynam- collectively called the universe.
ics, as it is fairly independent in giving essential concepts and The systems are classified into two: a closed system and
governing equations much of the time utilized as a part of an open system. A closed system, also called control mass,
subsequent chapters. has fixed amount of mass and does not allow the transport of
mass across the boundary. The energy, however, in the form of
heat or work can cross the boundary. Furthermore, if energy
8.1 Introduction is also not allowed to cross the boundary, the system is called
an isolated system. The boundary of a closed system may be
In a fluid flow, if the temperature change associated with the fixed or movable.
freestream is more than 5%, then it is imperative to study the An open system, also referred to as control volume, is
heat transfer aspects. For the low-speed flows (M < 0.5), the a properly selected region in space that encloses the device
thermodynamic considerations are not required due to large involving the mass transfer. The common examples are com-
heat capacity of the fluid compared to its kinetic energy. The pressor, turbines, nozzles, etc. The mass and energy are
temperature will remain constant even if the whole kinetic allowed to cross the control volume (C∀). The boundary of an
energy is converted into heat and thus, the static and the open system is known as control surface (CS) that can be real
stagnation temperatures of the fluid are equal. But, when or imaginary. For example, the inner surface of a compressor
the freestream Mach number is greater than 0.5, the change forms a real part, whereas entrance and exit sections form the
in energy of the flow will be substantially large. Hence, the imaginary part.
kinetic energy of the flow should also be taken into account
along with internal energy. In addition, for M > 0.5, the dif-
ference between static and stagnation1 temperatures is large 8.2 Internal Energy and the First Law
and therefore, the thermodynamic concepts should also be of Thermodynamics
considered in the analysis of compressible fluids. Entropy and
temperature are the two such fundamental thermodynamic The first law of thermodynamics is essentially the law of con-
concepts. servation of energy, which states that the total energy of an
isolated system remains constant. The energy can be neither
created nor destroyed; however, it can change its form. From
the macroscopic view, a system possesses potential energy
(PE) by virtue of its position and kinetic energy (KE) by virtue
1A stagnation state is achieved when a fluid is brought to rest traversing
of its motion, whereas on the microscopic scale the internal
through a reversible and adiabatic (i.e., isentropic) process.
energy (U) is defined as the sum of potential and kinetic ener-
182 8 Thermodynamics of Fluids in Motion
The Universe
Surroundings
System
System
Boundary
Fig. 8.1 Schematic of a system, surroundings, and boundary
gies of the atoms and molecules, which comprises the system. At this stage, we must acknowledge the sign convention fol-
It should be noted that, regardless of scale of observation, the lowed in Eq. (8.3). The sign convention for heat added to a
potential, kinetic, and internal energies are always measured system is positive and negative for heat extraction; work done
with respect to an arbitrary reference frame. Hence, the abso- by a system on its surroundings is positive, whereas it is nega-
lute description of these energies is not defined and only cor- tive when carried out on the system. Indeed, there is no restric-
responding changes should be accounted. Let us consider an tion if one is interested to follow their own sign convention
arbitrary system, which undergoes a change of state from “1” but it should be adopted uniformly in that particular thermo-
to “2”. If the corresponding changes in the internal energy, dynamic analysis in order to avoid confusion.
potential energy and kinetic energy, respectively, are desig- Since both Q and W are path functions, they depend on
nated as U, PE, and KE, then the change in total energy, how a thermodynamic system changes its state from initial to
E, of the system can be defined as final. Consequently, the change in heat transfer or change in
work performed is inexact differentials. On the other hand,
E = U + PE + KE (8.1) U is path independent and it is a state variable (also called
state function or point function). Therefore, dU is an exact
The first term, U, in the above expression is of fundamental differential. For a given dU, an infinite number of processes
importance in thermodynamics. Fortunately, to detect internal (paths) are possible to cause a change of state. However, for
energy differences, one need not require to use microscopic compressible fluids, we are interested in the following three
scale of observations, and change in internal energy (U) can processes only.
be evaluated by noting differences in macroscopic properties
of a system in different states. Let us consider an arbitrary • Reversible Process—It is the process that can be reversed
closed system with fixed mass of gas at rest. There is no trans- without leaving any trace on the surroundings. That is, both
fer of mass is taking place across the system boundaries. Let Q the system and the surroundings are returned to their initial
be the amount of heat added to the system across its boundary states at the end of the reverse process.
(by thermal conduction, convection, or direct radiation) and • Adiabatic Process—It is the process in which no heat trans-
W denotes the work done by the system on its surroundings. fer is occurring across the boundary of the system.
Mathematically, • Isentropic Process—It is the process that is both reversible
and adiabatic.
U = Q − W (8.2)
A system which allows the transport of the matter (mass and
Equation (8.2) is known as the first law of thermodynamics. energy) across its boundaries is known as open system. The
It states that “the heat added to the system minus work done mass that enters or leaves at different levels from an open
by the system is equal to the change in internal energy of the system will have kinetic energy by virtue of its motion and
system.” Expressing in differential form, the potential energy due to difference in levels. The open
systems are capable of delivering the work (called shaft work)
dU = δQ − δW (8.3) continuously as the matter (within the system) that transforms
energy is being replaced continuously.
8.2 Internal Energy and the First Law of Thermodynamics 183
Piston
ds
Moving Boundary P Mass ’ No’

Gas (m = constant)
Energy ’Yes’
Cylinder
Fig. 8.2 Schematic layout of a piston–cylinder assembly (closed system)
8.2.1 Concept of Work Upon integration,
Consider a piston–cylinder assembly as shown in Fig. 8.2. U2 V2

The cylinder contains a gas of constant mass (m). The initial dU = Q12 − p d∀ (8.6)
pressure of the gas is p, and volume is ∀; and A is the cross- U1 V1
sectional area of the piston. If the piston is allowed to move
a distance ds in a quasi-steady manner, the differential work or
done (W) during this process will be U2 − U1 = Q12 − p (∀2 − ∀1 ) (8.7)
δW = F.ds = pAds = pd∀ (8.4) Rearranging the terms,
It should be noted that p is the absolute pressure which is (U2 + p∀2 ) − (U1 + p∀1 ) = Q12 (8.8)
always positive. However, the change in volume d∀ is posi-
tive (increasing) during an expansion process and negative Since U + P∀ is an expression consisting of only state func-
(decreasing) during a compression process. Consequently, tions, a new state variable (or thermodynamic property),
the boundary work is positive during an expansion process, enthalpy (H) is given by
whereas it is negative during a compression process. The total
work done during the entire process is obtained by adding all H = U + p∀ (8.9)
the differential works from the initial state to the final state.
Thus, Differentiating both sides,

W = δW = pd∀ dH = dU + pd∀ + ∀dp
or
8.2.2 Concept of Enthalpy dH = δQ + ∀dp (8.10)
Substituting Eq. (8.4) into Eq. (8.3) and assuming constant Hence,
pressure,
H2 − H1 = H = Q12 (8.11)
dU = δQ − pd∀ (8.5)
dH = δQ (8.12)
In terms of specific quantities, where CV is the molar heat capacity at constant volume, usu-
ally expressed in Jmol−1 K−1 . In differential form,
h = u + p∀ (8.13)
dU = C∀ dT (8.20)
where h is the specific enthalpy, u is the internal energy
−1 −1
∀
in Jkg K ), and ∀ is the specific volume
(expressed In integral form,
∀ = m . Also,
T2
h2 − h1 = h = q12 (8.14) U = C∀ dT (8.21)
T1
or
In terms of specific quantities,
dh = δq (8.15)

∂u
c∀ = (8.22)
∂T ∀
8.2.3 Different Forms of the First Law
for an Adiabatic Flow where c∀ is specific heat capacity at constant volume,
expressed in Jkg−1 K−1 .
Recall Eqs. (8.5) and (8.10)
du = c∀ dT (8.23)
dU = δQ − pd∀
dH = δQ + ∀dp and
T2
For an adiabatic flow, δQ = 0. Hence,
u = c∀ dT (8.24)
dU = −pd∀ (8.16) T1
dH = ∀dp (8.17)
8.2.4.2 Specific Heat at Constant Pressure
Again from Eq. (8.12),
8.2.4 Concept of Specific Heats
dH = δQ
The specific heat is defined as the energy required to raise
Writing above equation in partial derivative form at constant
the temperature of a unit mass of a substance by one degree;
pressure,
however, this energy depends on how the process is executed.
In thermodynamics, we are interested in two kinds of specific
∂H δQ
heats: specific heat at constant volume (C∀ ) and specific heat = (8.25)
∂T ∂T
at constant pressure (Cp ). p p
But, by definition
8.2.4.1 Specific Heat at Constant Volume
From Eq. (8.5)
∂H
Cp = (8.26)
∂T p
dU = δQ − pd∀
where Cp is the molar heat capacity at constant pressure, gen-
Taking partial derivative of above equation with respect to erally expressed in Jmol−1 K−1 . Further, in differential form
temperature at constant volume,
dH = Cp dT (8.27)

∂U δQ
= (8.18)
∂T ∀ ∂T ∀
In integral form,
But, by definition T2

H = Cp dT (8.28)
∂U
C∀ = (8.19) T1
∂T ∀
8.2 Internal Energy and the First Law of Thermodynamics 185
In terms of specific quantities, Q12 − W12 = E2 − E1

∂h or
cp = (8.29)
∂T p
m m
Q12 − W12 = U2 + v22 + mgz2 − U1 + v12 + mgz1
2 2
where cp is the specific heat capacity at constant pressure, (8.36)
usually, expressed in Jkg−1 K−1 .
To evaluate the W12 , for the time being, let us replace the open
dh = cp dT (8.30)
system with a cylinder–piston assembly. Under quasi-steady
state, when a force F is applied on the piston at station “1”,
and
a small fluid mass moves a small distance d1 . Thus, the work
T2 done on the fluid mass W1 , at station “1”, is
h = cp dT (8.31)
T1 W1 = −F1 d1 = −p1 A1 d1 = −p1 ∀1
where F1 is the force applied at station “1”, p1 is the pressure,

8.2.5 Coefficient of Thermal Conductivity and A1 is the cross-sectional area of the piston. By similar
arguments, the work done by the fluid W2 at station “2” can
We know that heat transfer always takes place from high to low be written as
temperature. The Fourier’s law of heat conduction states that
W2 = p2 ∀2
“the heat flux per unit area in a given direction is proportional
to the temperature gradient in the same direction.” That is,
Further, due to continuous replacement of mass in the open
dT system, a useful work also called shaft work can be extracted.
qx = −κ (8.32) Therefore, the total work done by the system will be
dx
where qx is the heat flux per unit area in x-direction and κ W12 = WS + p2 ∀2 − p1 ∀1 (8.37)
is the proportionality constant, known as coefficient of ther-
mal conductivity. The negative sign in the above expression where WS is the shaft work and (p2 ∀2 − p1 ∀1 ) is called flow
denotes that the heat flows in the direction of decrease in work. Hence, the energy equation becomes
temperature. Fourier’s law can be expressed, in vector nota-
tions, as Q12 − W12 = Q12 − WS − p2 ∀2 + p1 ∀1
m m
q = −κ∇T (8.33) = U2 + v22 + mgz2 − U1 + v1‘2 + mgz1
2 2
where q is the vector heat flux per unit area and ∇T denotes Upon rearranging the terms,
the gradient of the temperature.
m 2
Q12 − WS = (U2 + p2 ∀2 ) + v + mgz2
2 2
8.3 Energy Equation for an Open System m
− (U1 + p1 ∀1 ) + v12 + mgz1
2
Let us consider an open system, schematically shown in
Fig. 8.3. The mass enters at station “1” with velocity v1 and where H1 = (U1 + p1 ∀1 ) and H2 = (U2 + p2 ∀2 ), respec-
leaves at station “2” with velocity v2 . Further, the elevations tively, are the total enthalpy at “1” and “2”. Thus,
of stations, “1” and “2”, respectively, are z1 and z2 , measured 1 1
from an arbitrary datum. The total energies of the flow at these H1 + mv12 + mgz1 = H2 + mv22 + mgz2 + WS − Q12
2 2
stations are (8.38)
1 2 In compressible flow applications, the change in potential

E1 = U1 + mv + mgz1 (8.34)
2 1 energies is very small as compared to the change in kinetic
1
E2 = U2 + mv22 + mgz2 (8.35) energy, and hence, can be neglected. Thus,
2
where U1 and U2 are internal energies of the fluid mass at 1 1

H1 + mv12 = H2 + mv22 + WS − Q12 (8.39)
stations “1” and “2”. For an open system the first law gives 2 2
Q
d2
Z2 V2
d1
2
V1 WS
Z1
Fig. 8.3 Schematic layout of an open system
Although the shaft work (WS ) is finite for the flow processes in the amount of energy lost by the hot object is equal to the
turbines and compressors, but for the flow around an airplane, amount gained by the cold object. Now let us consider the
WS = 0. Thus, if the process is assumed to be adiabatic as process in opposite direction—the hot object getting warmer
well, then Q12 = 0. Therefore, by gaining heat from the cold object. Interestingly, the first
law still allows this process as long as the amount of heat
1 1 taken by the hot object is equal to the heat given by the cold
H1 + mv12 = H2 + mv22 (8.40)
2 2 object. We know that it never happens. This limitation of the
or first law in stipulating the direction of the process is resolved
by introducing another condition, known as the second law
1 of thermodynamics. The two most important forms of second
H + mv2 = H0 = constant (8.41)
2 law are given below.
Kelvin–Planck Statement—It is impossible to construct a
where H0 is the stagnation enthalpy and H is the static cyclically operating device, which produces no other effect
enthalpy. Hence, the sum of static enthalpy and the flow than the extraction of heat from a single thermal reservoir
kinetic energy is constant in an adiabatic flow. Further, in and delivers an equivalent amount of work.
terms of specific quantities, Eqs. (8.40) and (8.41) are Clausius Statement—It is impossible to construct a device
that operates in a cycle and produces no effect other than
1 1
h1 + v12 = h2 + v22 (8.42) the transfer of heat from a low-temperature body to a high-
2 2
temperature body.
and The first statement regards to a heat engine and the sec-
v2 ond one regards to a heat pump. These statements, although,
h0 = h + (8.43) have no mathematical proofs, but so far their violation is not
2
recorded.
where h is the specific static enthalpy, and h0 is the specific
stagnation enthalpy. Equation (8.43) is the most useful form
of energy equation for compressible flows. 8.4.1 Thermodynamic Efficiency and Clausius
Inequality
8.4 Entropy and the Second Law of Consider a heat engine operating between a high-temperature
Thermodynamics thermal source and a low-temperature thermal sink, as shown
in Fig. 8.4. The heat source and heat sink are the thermal
Let us consider a case where a cold object is kept in contact reservoirs of very large masses that they can transmit or
with a hot object. From our experience, we know that the absorb, respectively, an unlimited amount of heat with-
cold object will get heated up and the hot object will cool out experiencing a change in temperature. Let QSource and
off. This process obeys the first law of thermodynamics since TSource , are the heat output from and absolute temperature of a
8.4 Entropy and the Second Law of Thermodynamics 187
Heat Source TSource
Q
Source
Heat Engine WE
Q Sink
Heat Sink TSink
Fig. 8.4 The heat engine
high-temperature heat source, respectively; and QSink and amount of molecular disorder will have a high value of
TSink are the heat input to and absolute temperature of entropy, and conversely a low entropy value suggests a low
a low-temperature heat sink, respectively. Assuming the degree of disorder in the system. By definition,
working substance as an ideal gas and all the processes to be
reversible, the thermodynamic efficiency can be written as δQrev
dS = (8.49)
T
W QSource − QSink
η= = In terms of specific quantities,
QSource QSource
or δqrev
ds = (8.50)
QSink T
η =1− (8.44)
QSource
Mathematically, T is an integrating factor, since the inexact
It can be shown that the maximum efficiency is achieved if differential δQrev is made exact by dividing the absolute tem-
the process works on Carnot cycle. perature. Clearly, to achieve theoretical Carnot efficiency, heat
transfer must take place in a reversible process, caused by an
TSink infinitely small temperature differential. In the real world,
η =1− (8.45) however, this method would require an infinite amount of
TSource
time to transport a finite amount of energy. Moreover, the
Rearranging Eqs. (8.44) and (8.45) and eliminating, heat transfer through a finite temperature difference is an
irreversible process. It can be observed that the work done
QSource QSink by a cyclic engine operational with irreversible heat transfer
= (8.46)
TSource TSink method is less than that in reversible case. Thus,
or
Wirrev < Wrev (8.51)
QSource QSink
− =0 (8.47) Recall Eq. (8.3)
TSource TSink
Generalizing the above for smaller Carnot cycles, we get dU = δQ − δW
δQrev Since the internal energy is a state function, independent

=0 (8.48)
T of whether the process path is reversible or irreversible.
That is,
At this stage, we must introduce a new thermodynamic prop-
erty, called entropy(s), which is the measure of amount of
dU = δQrev − δWrev = δQirrev − δWirrev (8.52)
molecular disorder within a system. A system with high
or 8.5 Combined Forms of the First Law and

the Second Law
δQrev = δQirrev + δWrev − δWirrev (8.53)
For a reversible flow process, let us introduce Eq. (8.49) into
or Eqs. (8.5) and (8.10). We have
δQrev δQirrev δWrev − δWirrev dU = TdS − pd∀ (8.60)

= + (8.54)
T T T
Introducing above equation into Eq. (8.49) dH = TdS + ∀dp (8.61)
δQrev δQirrev δWrev − δWirrev From the above equations, it is clear that increase in internal
dS = = + (8.55)
T T T energy of a system is equal to the heat added to it in an iso-
choric process, and increase in the enthalpy is equal to heat
For a work producing device, such as heat engine,
added in an isobaric process. At constant pressure, the entropy
δWirrev < δWrev and both are positive quantities as they are
change is
performed by the system. Hence,
dH
δQirrev dS = (8.62)
dS> (8.56) T
T
Integrating at constant temperature,
It is a widely used mathematical form of the second law of
thermodynamics for a closed system. Dropping the subscripts 2 2
1
from both Eqs. (8.49) and (8.58), a somewhat more general dS = dH
expression representing either reversible or irreversible pro- 1 T 1
cess can be written as
Substituting Eq. (8.27) into above
δQ
dS (8.57) 2 1 2
T dS = Cp dT (8.63)
1 T 1
and
or
δQ
0 (8.58) 2
T 1
S = Cp dT (8.64)
T 1
where equality sign is used for the heat transfer in a reversible
process. Equation (8.58) is known as the Clausius inequality. Similarly, at constant volume, the entropy change will be
If the process is adiabatic (δQ = 0), then
dU
dS 0 (8.59) dS = (8.65)
T
where dS > 0 represents an irreversible process, and dS = 0 Again, integrating at constant temperature
denotes a reversible process. Both Eqs. (8.58) and (8.59)
2 1 2
imply that a process will always proceed in a direction such
dS = dU
that the entropy of the universe (i.e., system plus surround- 1 T 1
ings) always increases or at least remains unchanged. That is,
in an adiabatic process, the entropy of the universe can never Substituting Eq. (8.20) into above
decrease. However, in a nonadiabatic process, we can extract
2 1 2
heat from the system and thus decrease the entropy of the S = dS = C∀ dT (8.66)
system. 1 T 1
8.5 Combined Forms of the First Law and the Second Law 189

or ∂T ∂p
=− (8.70)
∂∀ S ∂S ∀
1 2
S = C∀ dT (8.67)
T 1
8.6.2 Enthalpy
8.6 Maxwell’s Thermodynamic Relations Recall Eq. (8.61)
In this section, four important and useful thermodynamic dH = TdS + ∀dp

functions will be considered to deduce Maxwell’s relations.
They are the internal energy U, the enthalpy H, the Helmholtz
∂H
free energy (or simply the free energy) F, and the Gibbs free =∀ (8.71)
∂p S
energy (or the Gibbs function) G.
and

8.6.1 Internal Energy ∂H
=T (8.72)
∂S p
The internal energy (U) is a thermodynamic potential, which
means that it provides information through a differentiation. Again, if dH is an exact differential, hence
From Eq. (8.60)
∂T ∂∀
= (8.73)
dU = TdS − pd∀ ∂p S ∂S p

∂U 8.6.3 Helmholtz Free Energy
= −p (8.68)
∂∀ S
It is defined as
and
F = U − TS (8.74)
∂U
=T (8.69)
∂S ∀ Upon differentiating both sides,
Since dU is an exact2 differential, thus

∂2 z ∂2 z
2 Iff (x, y, z) = 0 ⇐⇒ z = z(x, y), y = y(z, x), and x = x(z, y). Let =
∂x∂ y ∂ y∂x
us consider z = z(x, y). We have
Thus, differentiating M partially with respect to y, and N with respect
∂z ∂z to x
z= dx + dy
∂x ∂y
y x
∂2 z ∂M
=
If ∂x∂ y ∂y x

∂z and
=M
∂x y
∂2 z ∂N
=
and ∂ y∂x ∂x y

∂z Therefore,
=N
∂y z

∂M ∂N
Then, =
∂y x ∂x y
z = Md x + N dy This is the condition of an exact (perfect) differential. The above result

shows that if x and y are thermodynamic properties then z will also be
If z is a continuous function, then a property.

dF = dU − SdT − TdS (8.75) ∂S
Cp = T (8.84)
∂T p
or
∂S
C∀ = T (8.85)
∂T ∀
dF = −pd∀ − SdT (8.76)
where Cp and C∀ are the principle specific heat capacities at
Since dF is an exact differential constant pressure and constant volume, respectively. Further,
we know that the entropy is a point function, which can be
∂p ∂S
= (8.77) represented in terms of temperature and volume. That is,
∂T ∀ ∂∀ T
S = S (T, ∀) (8.86)
8.6.4 Gibbs Free Energy Thus, we write
It is defined as G = F + p∀. Using Eq. (8.74), we get
∂S ∂S
dS = dT + d∀ (8.87)
∂T ∀ ∂∀ T
G = U − TS + p∀ (8.78)
Furthermore, the volume can also be defined in terms of pres-
and sure and temperature, i.e.,
dG = dU − TdS − SdT + pd∀ + ∀dp (8.79) ∀ = ∀ (T, p) (8.88)
or Since ∀ is a thermodynamic property, d∀ will be an exact

differential. Thus, we write
dG = ∀dp − SdT (8.80)

∂∀ ∂∀
d∀ = dT + dp (8.89)
Since dG is also an exact differential ∂T p ∂p T

∂∀ ∂S
=− (8.81) Introducing Eq. (8.89) into Eq. (8.87), we obtain
∂T p ∂p T

∂S ∂S ∂∀ ∂∀
Equations (8.70), (8.73), (8.77), and (8.81) are known as dS = dT + dT + dp
∂T ∀ ∂∀ T ∂T p ∂p T
Maxwell’s relations.
(8.90)
8.7 Effects of Fluid Compressibility on or

Specific Heats
∂S ∂S ∂∀ ∂S ∂∀
dS = + dT + dp
∂T ∀ ∂∀ T ∂T p ∂∀ T ∂p T
In this section, we will examine the effects of the compress-
(8.91)
ibility on the ratio of specific heats. Consider a reversible sys- Hence,
tem, as shown in Fig. 8.1, which exchanges an infinitesimal
amount of heat (δQ) with its surroundings. From the second
∂S ∂S ∂S ∂∀
law of thermodynamics, change in entropy (dS) of the system = + (8.92)
∂T p ∂T ∀ ∂∀ T ∂T p
is given as
δQ
dS = (8.82) Multiplying above by T throughout, we get
T rev

∂S ∂S ∂S ∂∀
or T =T +T (8.93)
∂T p ∂T ∀ ∂∀ T ∂T p
(δQ)rev = TdS (8.83)
From Eqs. (8.84) and (8.85), the above equation can be written
From the definition of specific heats, the above equation as
immediately results in
8.7 Effects of Fluid Compressibility on Specific Heats 191

∂S ∂∀ Hence, for a perfect gas, we get
Cp − C∀ = T (8.94)
∂∀ T ∂T p
1
αT = (8.104)
But, from the definition of coefficient of thermal expansion T
(αT ), we write
and

1 ∂∀ 1
= αT (8.95) βT = (8.105)
∀ ∂T p
p
Substituting Eq. (8.95) into Eq. (8.94), we get Substituting above equations into Eq. (8.102), we finally have
cp − c∀ = R (8.106)
∂S
Cp − C∀ = ∀TαT (8.96)
∂∀ T This is called the Mayer’s relation.
Let us now examine the effects of fluid compressibility on
From Eq. (8.77), the above equation can be written as the ratio of specific heat capacities. Dividing Eqs. (8.84) and
(8.85), we have
∂p ∂S
Cp − C∀ = ∀TαT (8.97) ∂T p
∂T ∀
Cp
= (8.107)
C∀ ∂S
By chain rule, the thermodynamic properties p, T, and ∀ can ∂T ∀
be written as
Again from chain rule, entropy (s), temperature (T), and pres-

∂p ∂∀ ∂T sure (p) can be written as
= −1 (8.98)
∂∀ T ∂T p ∂p ∀
∂S ∂T ∂p
= −1 (8.108)
or ∂T p ∂p s ∂s T

∂∀ or
∂p ∂T p ∂p
= − ∂S ∂T S
∂T ∂∀
(8.99) = − (8.109)
∀
∂p T
∂T p
∂p
∂S T
From Eqs. (8.95) and (2.98), the above equation can be written Similarly, entropy (s), temperature (T), and volume (∀) can
as be written as
∂∀
∂p αT ∂S ∂T S
= = − (8.110)
∂T ∀ βT
(8.100) ∂T ∀ ∂∀
∂S T
where βT is called the isothermal compressibility. Introducing Introducing Eqs. (8.109) and (8.110) into Eq. (8.107), we have
Eq. (8.100) into Eq. (8.97), we obtain

∂p ∂∀
Cp ∂T S ∂S
αT2 = T (8.111)
Cp − C∀ = ∀T (8.101) C∀ ∂p ∂∀
βT ∂S T ∂T S
In terms of specific quantities, the above equation can be writ- But

ten as ∂p
∂T S ∂p ∂T ∂p
= =
αT2 ∂∀ ∂T ∂∀ ∂∀
cp − c∀ = T∀ (8.102) ∂T S
S S S
βT
and
From thermodynamics, we know that the equation of state for
a perfect gas is written as
∂∀
∂S ∂∀ ∂S ∂∀
T = =
p∀ = RT (8.103) ∂p ∂S T ∂p T ∂p T
∂S T

∂u
Thus, we obtain But ∂T ∀ = c∀ , i.e., specific heat at constant volume. Thus,

Cp ∂p ∂∀ du = c∀ dT (8.119)
= (8.112)
C∀ ∂∀ S ∂p T
Considering the enthalpy as a function of temperature and
From Eqs. (2.98) and (2.100), the above equation can also be pressure, i.e.,
written as
Cp βT h = h(T, p) (8.120)
= (8.113)
C∀ βs
In terms of exact differentials,
i.e., the ratio of specific heats is same as the ratio of isother-
mal compressibility to isentropic compressibility of the fluid. ∂h ∂h
dh = dT + dp (8.121)
Furthermore, in terms of specific quantities, we write ∂T p ∂p T
cp βT For a constant pressure (isobaric) process,

= =γ (8.114)
c∀ βs
∂h
dh = dT
∂T p
8.8 Thermal and Calorical Properties

∂h
A relation between thermal properties such as pressure (p), But ∂T p = cp , i.e., specific heat at constant pressure.
temperature (T), and density (ρ) is known as the thermal equa- Hence,
tion of state. For a perfect gas, the thermal equation of state
is the ideal gas law, i.e., dh = cp dT (8.122)
p∀ = RT (8.115)
8.8.1 Thermally Perfect Gas

where ∀ = 1
ρ is the specific volume of the gas. R = R̄u

M
A gas that obeys thermal equation of state p∀ = RT is called
is the specific gas constant and R̄u =8314 Jkg−1 K−1
is the the thermally perfect gas, and for this gas both internal energy
universal gas constant. For any gas, the commonly referred and enthalpy are the functions of temperature alone. That is,
calorical properties are internal energy (u), enthalpy (h), and
entropy (s). Any relation between the calorical properties u, u = u (T) (8.123)
h, and s, and the thermal properties p, T, and ρ are called the
calorical equation of state. and
From the state principle, we know that by specifying
any two independent intensive thermodynamic properties, all h = h (T) (8.124)
other thermodynamic properties are fixed. Let us represent
the internal energy as a function of temperature and specific Thus, from Eqs. (8.119) and (8.122), we get
volume, i.e.,
c∀ = c∀ (T) (8.125)

u = u T, ∀ (8.116)
and
In terms of exact differentials, Eq. (8.116) can be expressed
cp = cp (T) (8.126)
as

∂u ∂u Thus, from Eqs. (8.117) and (8.121), we obtain
du = dT + d∀ (8.117)
∂T ∀ ∂∀ T
du = c∀ dT (8.127)
For a constant volume (isochoric) process, Eq. (8.117) reduces
to and

∂u dh = cp dT (8.128)
du = dT (8.118)
∂T ∀
8.8 Thermal and Calorical Properties 193
Therefore, the above equations are universally valid as long Such a gas is termed as calorically perfect gas. A perfect gas
as the gas is thermally perfect. is always thermally as well as calorically perfect. That is,
it must satisfy both the thermal equations of state: p= ρRT
∂u
as well as the calorical equations of state: C∀ = ∂T and
8.8.2 Mayer’s Relation ∀
∂h
Cp = ∂T . Further, a calorically perfect gas will always be
p
The state equation a thermally perfect gas is thermally perfect but vice versa is not true. Hence, thermal
perfectness is a precondition for caloric perfectness. Further-
P∀ = RT more, for a perfect gas, cp = cp (T) and c∀ = c∀ (T). That is,
both cp and c∀ are function of temperature. But it is interesting
In the differential form, to note that, despite cp and c∀ being a function of temperature,
their ratio γ is a constant and thus, independent of tempera-
Pd∀ + ∀ dp = RdT ture. For a calorically perfect gas, cp , c∀ as well as γ are
constant and independent of temperature.
Also,
h = u + p∀ 8.9.1 Ratio of Specific Heats
Upon differentiation, From kinetic theory of gases, the (γ) can be represented in
terms of degrees of freedom (n) of gas molecules as
dh = du + pd∀ + ∀ dp
n+2
γ= (8.130)
n
dh − du = pd∀ + ∀ dp = RdT
The monotonic gas has only three-translational degrees of
or freedom (i.e., n = 3). Thus,
cp dT − c∀ dT = RdT 3+2 5
γ= = = 1.67
3 3
or
The diatomic gases, such as nitrogen and oxygen, have n = 5,
cp − c∀ = R (8.129) i.e., they have three-translational degrees of freedom and two
rotational degree of freedom; thus
Note that Eq. (8.129) is the same relation as given by
7
Eq. (8.106), commonly referred to as Mayer’s relation. Thus, γ= = 1.4
5
for a thermally perfect gas even though cp and c∀ are the
functions of temperature, their difference remains a constant However, the polyatomic gases such as Freon and gaseous
which is equal to the specific gas constant (R). compounds of Uranium have large values of n; consequently,
γ is slightly greater than unity. Therefore, the specific heat
ratio (γ) varies from 1 to 1.67, depending on the molecular
8.9 The Perfect Gas behavior of the gas, i.e., 1 ≤ γ ≤ 1.67.
A perfect gas is the gas which has intermolecular spacing

so large that the intermolecular forces are neglected. For a 8.9.2 Limitation of Air as a Perfect Gas
perfect gas, both cp and c∀ are constant and independent of
temperature, i.e., The air is composed of nitrogen (78.09% by volume), oxy-
gen (20.5% by volume), argon (0.93% by volume), and small
cp = cp (T) cp = constant traces of other gases. The thermal and calorical properties of
air vary with the temperature.
and When the temperature of air is less than 500 K, the air is
both thermally and calorically perfect, i.e., it obeys equation
c∀ = c (T) c∀ = constant of state,
p = ρRT Summary
Also, cp and c∀ both are independent In a fluid flow, if the temperature change associated with the
of temperature and
c
remain constant and thus, their ratio γ = c∀p takes a constant freestream is more than 5%, then it is imperative to study the
heat transfer aspects. For the low-speed flows (M < 0.5), the
value of 1.4.
thermodynamic considerations are not required due to large
When the temperature is more than 500 K but less than
heat capacity of the fluid compared to its kinetic energy. The
2000 K, air retains its thermal perfectness, i.e., it obeys
temperature will remain constant even if the whole kinetic
equation of state. However, it becomes calorically imperfect
energy is converted into heat and thus, the static and the
because both cp and c∀ are now functions of temperature,
stagnation temperatures of the fluid are equal. But, when
i.e.,
the freestream Mach number is greater than 0.5, the change
cp = cp (T) in energy of the flow will be substantially large. Hence, the
kinetic energy of the flow should also be taken into account
and along with internal energy. In addition, for M > 0.5, the dif-
c∀ = c∀ (T) ference between static and stagnation temperatures is large
and therefore, the thermodynamic concepts should also be
But the ratio of specific heats still turns out to be constant, considered in the analysis of compressible fluids.
i.e., A system is defined as a quantity of matter or a region in
cp space, chosen for study; and the mass or the region outside the
γ= = constant system is known as its surroundings. A real or an imaginary
c∀
surface separating the system and surroundings is called the
When the temperature of air is more than 2000 K but remains boundary, which may be fixed or movable. The system and
less than 4000 K, the air is no more thermally perfect. The its surrounding are collectively called the universe.
oxygen starts dissociating into oxygen ions and free electrons. The systems are classified into two: a closed system and
That is, an open system. A closed system, also called control mass,
has fixed amount of mass and does not allow the transport of
O2 ⇐⇒ 2O+ + 2e− mass across the boundary. The energy, however, in the form of
heat or work can cross the boundary. Furthermore, if energy
Thus, for the temperature greater than 2000 K, air becomes is also not allowed to cross the boundary, the system is called
both thermally and calorically imperfect. As the air tempera- an isolated system. The boundary of a closed system may be
ture increases beyond 4000 K, the nitrogen also starts disso- fixed or movable.
ciating into nitrogen ions and free electrons and is complete The first law of thermodynamics is essentially the law of
around 9000 K. That is, conservation of energy, which states that the total energy of
an isolated system remains constant. The energy can be nei-
N2 ⇐⇒ 2N+ + 2e− ther created nor destroyed; however, it can change its form.
Mathematically, the first law of thermodynamics is expressed
During re-entry phase of space vehicles about at an altitude of as
70 km and approximately 20 min prior to touchdown, the free dU = δQ − δW
electrons and ions produced during the dissociation of oxy-
gen and nitrogen molecules form an electron cloud, respon- It states that “the heat added to the system minus work done
sible for communication blackout. This phase is commonly by the system is equal to the change in internal energy of
known as radio blackouts, ionization blackouts, or re-entry the system.” Moreover, for a given dU, an infinite number
blackouts, generally lasts for several minutes. In this phase, of processes (paths) are possible to cause a change of state.
the spacecrafts re-enter into the Earth’s atmosphere at high However, for compressible fluids, we are interested in the
Mach numbers. Because of high compression, the ionized air following three processes only.
is found in the state of plasma which interrupts the radio sig-
nals. Unfortunately, this is the most critical phase and if any • Reversible Process—It is the process that can be reversed
problems occur during this phase, the communication black- without leaving any trace on the surroundings. That is, both
out prevents the diagnostic telemetry from reaching to the the system and the surroundings are returned to their initial
Earth. Such was the case with the catastrophic failure of the states at the end of the reverse process.
Space Shuttle Colombia during re-entry on February 1, 2003.
Summary 195
• Adiabatic Process—It is the process in which no heat trans- Since dSirrev > 0, the above equation is also written as
fer is occurring across the boundary of the system.
δQ
• Isentropic Process—It is the process that is both reversible 0
and adiabatic. T
This is known as the Clausius inequality.

The Fourier’s law of heat conduction states that “the heat
A relation between thermal properties such as pressure,
flux per unit area in a given direction is proportional to the
temperature, and density is known as the thermal equation of
temperature gradient in the same direction.” That is,
state. For a perfect gas, the thermal equation of state is the
dT ideal gas law, given as p = ρRT
qx = −κ For any gas, the commonly referred calorical properties are
dx
internal energy, enthalpy, and entropy. Any relation between
The most useful form of energy equation for compressible the calorical properties and the thermal
properties iscalled

flows is ∂u ∂h
the calorical equation of state: c∀ = ∂T and cp = ∂T .
∀ p
v2 A perfect gas is the gas which has intermolecular spacing
h0 = h +
2 so large that the intermolecular forces are neglected. For a
perfect gas, both cp and c∀ are constant and independent
This limitation on the direction of the process is imposed
of temperature. Such a gas is termed as calorically perfect
by the second law of thermodynamics. The two most impor-
gas. A perfect gas is always thermally as well as calorically
tant forms of second law are given below. The first state-
perfect. That is, it must satisfy both the thermal equation
ment regards to a heat engine and the second one regards to a
of state and the calorical equations of state. Moreover, a
heat pump. These statements, although, have no mathematical
calorically perfect gas will always be thermally perfect
proofs, but, so far their violation is not recorded.
but vice versa is not true. Hence, thermal perfectness is a
Kelvin–Planck Statement—It is impossible to construct a
precondition for caloric perfectness.
cyclically operating device, which produces no other effect
From kinetic theory of gases, the specific heat ratio (γ)
than the extraction of heat from a single thermal reservoir
can be represented in terms of degrees of freedom (n) of gas
and delivers an equivalent amount of work.
molecules as
Clausius Statement—It is impossible to construct a device
that operates in a cycle and produces no effect other than n+2
γ=
the transfer of heat from a low-temperature body to a high- n
temperature body.
Suppose QSource and TSource are the heat output from
and absolute temperature of a high-temperature heat source,
respectively; and QSink and TSink are the heat input to and
Exercises
absolute temperature of a low-temperature heat sink, respec-
tively. Assuming the working substance as an ideal gas and all
the processes to be reversible, the thermodynamic efficiency
1. The gas at a pressure 150 kPa and the volume 7000 cm3 is
can be written as
quasi-statically compressed isothermally until the volume
QSink becomes 2500 cm3 . Calculate the work transfer and the
η =1− final pressure.
QSource
2. In a cyclic process, the heat transfers are +15.1, −23.5,
It can be shown that the maximum efficiency is achieved if −4.51, and +30.3 KJ. Calculate the net work for this pro-
the process works on Carnot cycle. cess.
3. The properties of a certain fluid are related by
TSink u = 180 + 0.617t and p∀ = 0.319(t + 273), where t
η =1−
TSource is the temperature in ◦ C, ∀ is the specific volume in
m3 kg−1 , u is the specific internal energy in kg−1 KJ, and
Entropy can be regarded as a measure of disorder or random- p is the pressure in kNm−2 . Calculate the specific heats
ness in a system. The change in entropy is given by at the constant pressure and the constant volume.
4. The heat engine delivers 7.8 kW and gains heat at the rate
δQ
dS = + dSirrev of 1300 KJmin−1 . Calculate the rate of heat rejection and
T the thermal efficiency.
5. Determine the location of absolute zero on the Celsius 4. In a Carnot cycle, heat is supplied to the water at 400 ◦ C
scale, if 30 KJ is added to the Carnot cycle at a temperature and rejected at 25 ◦ C. The water evaporates to steam,
of 90 ◦ C and 15.1 KJ are rejected at 1 ◦ C. while receiving the heat at 400 ◦ C. During this process,
6. 0.3 kg of air at 320 ◦ C is heated reversibly at the constant the change in entropy is 1.5 KJkg−1 K−1 . If the cycle oper-
pressure to 1990 K. Calculate the available energy of heat ates on a stationary mass of 1 kg of water, the work done
added. Assume T0 = 29 ◦ C and Cp = 1.005 KJkg−1 K−1 . per cycle is
7. 50 kg of water at 80 ◦ C is mixed with 30 kg of water at
50 ◦ C. Calculate the decrease in available energy due to (a) 321 KJ
mixing, if the surrounding temperature is 18 ◦ C. (b) 467 KJ
8. The velocities of air at the inlet and the exit of a blower (c) 562 KJ
are 170 and 120 ms−1 , respectively. The blower consumes (d) 633 KJ
17 kW for handling 1.2 kgs−1 of air at 25 ◦ C. Determine
the exit air temperature by assuming adiabatic conditions. 5. For an isolated system
Assume Cp =1.005 KJkg−1 K−1 .
(a) dQ = 0 and dW = 0
9. A metal block of 500 g mass having Cp = 170 JK−1 at
(b) dQ < 0 and dW < 0
90 ◦ C is placed in a water at 8 ◦ C. Determine the total
(c) dQ > 0 and dW > 0
entropy change of the metal and the water (universe).
(d) dQ > 0 and dW = 0
10. In an adiabatic expansion, if the air pressure halved then
calculate ρρfi where ρi = initial density of air and ρf = 6. In a piston–cylinder assembly, the piston is allowed to
final density of air. move a distance of 20 mm in quasi-steady manner under
the force of 10 N. If the cross-sectional area of the piston is
1963 mm2 , the differential work done during this process
is
1. A gas of volume 5000 cm3 is compressed quasi-statically (a) 30 Nm
until the volume becomes 1800 cm3 . Assuming the gas (b) 10 Nm
law as p∀2 = constant, if the initial pressure of the gas is (c) 2 Nm
110 kPa, then the final pressure after compression is (d) 0.2 Nm
(a) 421.6 kPa 7. If a substance of 6 kg requires 50 J of heat addition to
(b) 639.1 kPa change its temperature by 280 K, the specific heat of the
(c) 715 kPa substance is
(d) 848.7 kPa
(a) 0.013 Jkg−1 K−1
2. Consider the fluid flow over a flat plate of thickness (b) 0.029 Jkg−1 K−1
10 mm. The temperatures of the fluid above and below (c) 0.034 Jkg−1 K−1
the plate are 400 and 300 K, respectively. If the thermal (d) 0.046 Jkg−1 K−1
conductivity of the flat plate is 0.12 Wm−1 K−1 , the heat
flux per unit area will be 8. The heat addition to a gas of mass 0.5 kg is 2000 J. If the
specific heat is 10 Jkg−1 K−1 , the change in temperature
(a) 600 Wm−2 is
(b) 800 Wm−2
(c) 1200 Wm−2 (a) 50 K
(d) 1600 Wm−2 (b) 250 K
(c) 350 K
3. Let 3 kg of water at 70 ◦ C is mixed adiabatically with (d) 400 K
4 kg of water at 35 ◦ C. If the mixing process is isobaric at
1 atmosphere, then the increase in entropy of the whole 9. From kinetic theory of gases, the specific heat ratio (γ)
mass of water is (Assume Cp = 4.2 kJkg−1 K−1 ) can be represented in terms of the degrees of freedom (n)
of the gas molecules as
(a) 0.01 KJK−1
(b) 0.02 KJK−1 (a) γ = n+2
n
(c) 0.04 KJK−1 (b) γ = n
n+2
(d) 0.06 KJK−1
Exercises 197
(c) γ = n−2
n Keys
(d) γ = n
n−2
10. In a cyclic process, the heat input and heat output are 1200 1. (d)
and 3500 J, respectively. Assuming the working substance 2. (c)
to be an ideal gas and all the processes as reversible, the 3. (c)
thermodynamic efficiency will be 4. (c)
5. (a)
(a) 54% 6. (d)
(b) 66% 7. (b)
(c) 72% 8. (d)
(d) 78% 9. (a)
10. (b)
Compressible Flows
9

In our discussions so far, it is assumed that the flow is subsonic, the isentropic relation ρpγ = constant , and the second law
incompressible, and nonviscous. Indeed, it has established our of thermodynamics enable us to make the quantitative study
framework of aerodynamic theory for low-speed flight, like of compressible fluid flow.
the chronicled development of aerodynamics as connected to
aircraft design. The incompressible flow theory was sufficient
in designing the aircraft till World War I where the maximum 9.1.1 One-Dimensional Flow Approximation
speed was in the range of 200–230 kmh−1 . However, dur-
ing the World War II, the speeds of propeller-driven aircraft We have seen earlier that a flow may either be one-, two-, or
were well above 650 kmh−1 , while the speed of jet-propelled three-dimensional. When the flow is described in terms of one
aircraft was close to 980 kmh−1 . At these higher speeds, the spatial coordinate and time, it is called one-dimensional flow.
incompressible flow theory was inadequate in analyzing the Reader should not get confused with one-dimensional flow
flow fields accurately. Presently, in the domain of compress- and uniform parallel flow in a confined duct. This is because in
ible flow, the density of air could not be viewed as constant the latter, there is no limit on the number of spatial coordinates
which in fact complicates the classical aerodynamic scenario. by which the flow can be described. Also, one-dimensional
In this chapter, the reader will be exposed to those vital con- flow need not be in straight line, as shown in Fig. 9.1. The
cepts which muddle the aerodynamic picture. flow will be one-dimensional as long as the flow properties
vary with only one spatial coordinate and time. Further, it
should be noted that even if one-dimensional approximation
9.1 Introduction to Elastic and Inviscid is not fully satisfied, the flow at any cross section can be
Compressible Flows idealized as one-dimensional which yields averaging of flow
properties at a particular cross section along the flow direction.
The fluid becomes compressible when it is subjected to a One-dimensional approach is quite useful in dealing the flow
pressure field causing them to flow, i.e., the fluid will be com- situations such as the flow in diffuser, nozzle, duct, etc.
pressed or be expanded to some extent because of the pressure
acting on them. The time rate of change of velocity of the fluid
elements in a given pressure gradient is a function of the fluid 9.2 Governing Equations of Compressible
density (ρ), whereas the degree of compression is determined Flows
by the isentropic bulk modulus of compression (κ). The term
compressible flows are defined as the variable density flow. Consider a compressible flow passing through a rectangular
The variations in fluid density for compressible flow require control volume, as shown in Fig. 9.2. Suppose the flow is one-
attention to density and other fluid property relationships. The dimensional and the properties of the flow, from the region
fluid equation of state, often unimportant for incompressible “1” to “2”, change along x-direction. We will consider the
flows, is vital in the analysis of compressible fluids. More- following assumptions to derive the fundamental equations:
over, the temperature variations for compressible flows are
usually significant, necessitating to include the energy equa- • Steady and uniform flow from the region “1” to “2”.
tion in the analysis. The fundamental governing equations • The flow cross-sectional areas perpendicular to the flow
of fluid flows, continuity, momentum, and energy equations direction in both regions are assumed to be equal.
together with thermodynamic relationships for perfect gases, • The flow is nonviscous with no body forces.

200 9 Compressible Flows
A Streamlines
s − direction
A
Streamlines
Fig. 9.1 One-dimensional flow
Control surface
v1 v2
’A’ Area
Τ1 T2
p p
1 2
ρ ρ
1 2
e e2
1
(1) (2)
Fig. 9.2 Schematic representation of a typical one-dimensional flow
• The control volume is assumed to be isolated with no But

exchange of heat with the surroundings, i.e., control vol-
p
ume is adiabatic. h= e+ (9.5)
ρ
Continuity Equation From the conservation of mass princi-
Thus, we have
ple on the selected control volume, we write v12 v2
h1 + = h2 + 2 (9.6)
2 2
−ρ1 v1 A + ρ2 v2 A = 0
ρ1 v1 = ρ2 v2 (9.1) Equations (9.1), (9.2), and (9.6) are the primary governing
equations for the one-dimensional compressible fluid flows.
Momentum Equation From Newton’s second law of motion, Increase of Entropy Principle (Second Law of Thermody-
we get namics)
ρ1 (−v1 A) v1 + ρ2 (v2 A) v2 = − (−p1 A + p2 A) The direction of a particular flow process is established by

the second law of thermodynamics, also known as increase
p1 + ρ1 v12 = p2 + ρ2 v22 (9.2)
of entropy principle. Mathematically,

Note that, in Eq. (9.2) pA + ρAv2 is sometimes defined as
s2 − s 1 ≥ 0 (9.7)
impulse function, denoted by F. That is,
where s1 and s2 , respectively, are the entropy at state “1” and
F = pA + ρAv2 (9.3)
state “2” of the system.
Energy Equation For the steady flow, the energy equation Equation of State
can be written as
The equation of state of the fluid may be written implicitly in
p1 v12
p2 v22 the form
+ e1 + = + e2 + (9.4)
ρ1 2 ρ2 2
h = h (s, ρ) (9.8)
9.2 Governing Equations of Compressible Flows 201
s = s (p, ρ) (9.9) 9.3.1 Speed of Sound
Equations (9.7) and (9.9) are the secondary or auxiliary equa- The sound waves are the infinitesimal pressure disturbances,
tions used in compressible flow analysis. and the speed at which these waves propagate in a medium
is known as the speed of sound or acoustic speed. Further,
the term compressible flow reflects the variation in density
9.3 Effects of Acoustic Speed on the Fluid due to pressure change from one point to another in the flow
Compressibility field. The change in density with respect to pressure has strong
effects on the wave propagation.
In Sect. 2.6.4, we have seen that the compressibility
of a fluid Consider the propagation of an infinitesimal pressure
is defined as the fractional change in volume ∂∀ ∀ per unit pulse in a piston–cylinder assembly, as shown in Fig. 9.3.
change in pressure (∂p). In Fig. 9.3a, let the rightward movement of the piston with
an infinitesimal velocity dv, a pressure pulse begins to
1 ∂∀ propagate toward right with velocity a. The pressure and
β=− (9.10)
∀ ∂p density are (p + dp) and (ρ + dρ) in the region where the
wave is already traversed. However, the fluid on the right
where ∀ is the volume and p is the pressure. For any fluid, the side of the wave is still motionless and pressure and density
magnitude of compressibility strongly depends on whether are p and ρ, respectively. It is sometimes convenient to make
the process is isothermal or adiabatic. The isothermal com- the wave stationary and allow the fluid flow across the wave.
pressibility, denoted as βT , is given by This situation is depicted in Fig. 9.3b where the observer is
traveling with the wave velocity and the fluid flows steadily
1 ∂∀ from right to left. The flow velocity is reduced from a to
βT = − (9.11)
∀ ∂p T a − dv, pressure rises from p to p + dp, and density rises
from ρ to ρ + dρ.
where the subscript T indicates the partial differentiation car- Let us choose a small control volume around the pressure
ried out at the constant temperature. Similarly, for an isen- pulse, as illustrated in Fig. 9.3b. The control surfaces across
tropic process, the fluid compressibility, depicted as βs , is the waves are assumed to be stationary, and the shear forces are
given as neglected in comparison to the pressure forces acting on the
1 ∂∀
βs = − (9.12) control volume. From the conservation of momentum princi-
∀ ∂p s ple between upstream (1) and downstream (2) locations, we
write
where the subscript s shows the constant entropy process. In
the above equation, the minus sign is put to make the com-
A (p + dp) − p = m̊ (a − dv − a) (9.15)
pressibility positive as the volume of the fluid decreases (i.e.,
the fractional change in volume is negative) by increasing the
But m̊ = ρAa, which is the mass flow rate crossing the wave.
pressure. Furthermore, in classical mechanics the speed of
Thus, we have
sound in a medium is defined as

∂p A (p + dp) − p = ρAa (a − dv − a) (9.16)
a2 = (9.13)
∂ρ s
where ρ is the density of the medium. Thus, the isentropic dp = −ρadv (9.17)
compressibility can be expressed as
This is indeed the Euler’s equation for the steady fluid flows.
1 Moreover, the application of mass balance on both sides of
βs = 2 (9.14)
ρa the wave front noting that the cross-sectional area A remains
constant gives
where a is the speed of sound, which is discussed in the fol-
lowing section. (ρ + dρ) (a − dv) = ρa
(a) (b)
Pressure Pulse Control Volume Cross−Sectional Area "A"
dv a
1 2
a − dv a
Fig. 9.3 Propagation of pressure pulse in the cylinder–piston assembly
or On differentiation, we get
dρ dv dp dρ
= (9.18) −γ =0 (9.23)
ρ a p ρ
Since it is already assumed that the piston velocity is very or
small, the pressure and temperature changes are also small.
Consequently, the flow process may be assumed nearly dp γp
= (9.24)
reversible. Also, the rapidity of the process makes it difficult dρ s ρ
for control volume to have any appreciable exchange of
heat transfer with the surroundings. In these limits of Using perfect gas equation, p = ρRT, we obtain
reversible and adiabatic process, the flow can be considered
dp
as isentropic. So that, from Eqs. (9.17) and (9.18), we write = γRT (9.25)
dρ s

dp
a =
2
(9.19) Introducing above equation into Eq. (9.20), the acoustic speed
dρ s in a perfect gas becomes
or
a= γRT (9.26)

dp
a= (9.20) where R is the specific gas constant.
dρ s
Note that the similar results will be obtained for the velocity
of propagation of a small disturbance from a cylindrical wave 9.3.3 Effect of Molecular Weight on the Acoustic
spreading from a line source and for spherical wave spread- Speed
ing from a point source. In addition, we can see that for the
fluids dρ → 0 implies a → ∞. Since the liquids are nearly Since the specific gas constant R is obtained by the expression,
incompressible, the acoustic velocity in liquids is generally
Ru
very high. R= (9.27)
M
where R̄u = universal gas constant and M = molecular weight

9.3.2 Speed of Sound in a Perfect Gas
of the gas. Hence, Eq. (9.26) can also be written as
As a particular case, let us now examine the speed of sound
in a perfect gas. We know that for a perfect gas the pressure Ru
a = γ T (9.28)
and density in an isentropic process are related as M
p
= constant (9.21) From Eq. (9.28), it is seen that speed of sound depends upon
ργ
the ratio of specific heats and molecular weight of the gas.
C Since γ is bounded parameter which varies as 1 < γ < 1.67,
where γ = Cvp , which is the ratio of principle specific heats.
the gases with lower molecular weight will have the large
Taking log of the above expression, we have
acoustic speed and vice versa. Consider the acoustic speed in
the following gases.
ln p − γ ln ρ = ln (constant) = constant (9.22)
9.3 Effects of Acoustic Speed on the Fluid Compressibility 203
• Since air is a diatomic gas, γ = 1.4. In addition, the specific • It is proportional to the ratio of kinetic energy to internal
gas constant of air is energy.
γ
v2 v2 v2
Ru 8314 v2 γ (γ − 1) 2
Rair = = = 287 Jkg−1 K−1 2
= 2
= 2
RT
= 2
= M
M 29 e cv T a2 2
(γ−1) (γ−1)
where M is the molecular weight of the √ air. The speed of
sound in air will be given as aair ≈ 49 T. At sea level • It is measure of the directed motion of a fluid compared to
conditions, the speed of sound in air is of the order of 330 the random thermal motion of its molecules.
ms−1 .
• For mono-atomic gases, such as hydrogen
γ = 1.67 and MH2 = 2 , the speed of sound at standard
9.3.5 The Mach Angle
atmospheric conditions is 1200 ms−1 .
When an object moves through the air or the air flows past the
• Similarly, in refrigeration gases, e.g., Freon-22, the acous-
object, each element of the solid surface inclines to deflect the
tic speed is of the order of 110 ms−1 .
air from the direction it might otherwise take. These local dis-
From the above discussion, we can draw an interesting conclu- turbances act as a point source which create spherically prop-
sion. In turbomachines, the speed of the rotor should be in the agating pressure waves into the external air. At each instant,
range from 270–450 ms−1 , to avoid the excessive stresses gen- these waves can be represented as the superimposed images of
erated due to rotation. The studies also reveal that the loss in all the waves, emitted from this source some earlier instants of
efficiency mounts rapidly when the rotor speed approaches the time. Moreover, the resulted pattern of waves may be either
sonic velocity. Thus, for air compressors the limiting design symmetrical or asymmetrical depending upon whether the
factor on rotational speed may be either stress or compressibil- source is stationary or in motion and the speed of a mov-
ity considerations. In hydrogen compressors, the fluid coming source in comparison to the acoustic speed. Consider the
pressibility will never be a factor, whereas compressibility is following cases:
a major design factor for the compressor working with Freon-
• When the fluid medium is incompressible (ρ = constant)
22 as fluid.
or the speed of the source is negligibly small as compared
to the speed of sound (v a), the resulted wave pattern
will be same as shown in Fig. 9.4a.
9.3.4 Concept of Mach Number
• When the object (source) moves through a compressible
The speed of sound is a property that varies from point to point medium, however, the speed of the object is still less than
and if there exists a large difference in the speeds between the speed of sound (v < a), the wave pattern will look
the body and the compressible fluid surrounding it, the com- similar to Fig. 9.4b.
pressibility of the fluid medium influences the flow around • When the source moves with speed of sound, the resulted
the body. Thus, both the inertial forces and elastic forces due pressure pattern will be similar to as illustrated in Fig. 9.4c.
to fluid compressibility should be accounted in the analysis. • When an object moves with a speed greater than the speed
The ratio of inertial force to elastic force is a nondimensional of sound, the emitted disturbance waves lie within a cone,
parameter, called the Mach number (M). From the order of having its vertex at the body at that instant. The presence
magnitude analysis, it can be shown that the ratio of these of the disturbance is not felt upstream of the cone vertex;
forces becomes disturbances are only transmitted downstream within the
cone. This cone within which the disturbances are confined
M=
v
=√
v is known as the Mach cone and the half angle (μ) of the
a γRT cone is called the Mach angle. From the geometry of the
figure, shown in Fig. 9.4, the expression for half angle is
i.e., the ratio of the local flow speed to the local speed of sound obtained as
is known as the Mach number.
1
μ = sin−1 (9.29)
9.3.4.1 Physical Significance of the Mach Number M
The Mach number is dimensionless parameter which can be
where M is given by
explained in the following ways:
v object speed
• It reflects the compressibility of the fluid medium. M= =
a acoustic speed
Ua Ua
(a) v = 0 or M = 0 (b) v < a or M < 1
Planar wave front

Ua Ua Zone of silence
μ
Zone
of
action
Zone of action Zone of silence

Zone of silence Mach line
(c) v = a or M = 1 (d) v > a or M > 1
Fig. 9.4 Schematic diagram of pressure waves in different flow regimes
9.3.5.1 von Karman’s Rules for Supersonic Flows 9.3.6 Classification of Flow Regimes Based on
von Karman’s proposed three rules for supersonic flows which the Mach Number
are applicable for small disturbances. These rules, however,
can be used for large disturbances but for qualitative purposes The flow regimes can be classified based on the value of the
only. Mach number.
Rule of Forbidden Signals • For 0 < M < 1, the flow is termed as subsonic. In a
The effect of pressure changes produced by a body, moving subsonic field, the presence of small disturbance, travel-
at a speed faster than the sound, cannot felt upstream of the ing with acoustic speed, will be felt throughout the flow
body. domain. Thus, the subsonic flows are essentially “pre-
warned” to the disturbance.
Zone of Action and Zone of Silence • For 0.8 < M < 1.2, the flow is termed as transonic flow.
A stationary point source in a supersonic stream produces • For M = 1, the flow is called sonic flow.
effects only on the points that lie on or inside the Mach cone, • For M > 1, the flow is called supersonic flow. Since the
extending downstream from the point source. flow speed is above the speed of sound, they are no more
“pre-warned”.
Rule of Concentrated Action • For M > 5, the flow is called hypersonic flow.
The proximity of circles representing the various flow situa-
tions is a measure of the intensity of the pressure disturbance For more details on flow regimes, the reader is advised to refer
at each point in the flow field. Sect. 2.6.6.
Thus, for the stationary source the intensity of the distur-
bances is symmetrical about the source (Fig. 9.4a). For sub-
sonic source, the intensity is asymmetrical (Fig. 9.4b). In case 9.4 One-Dimensional, Steady, and
of supersonic source, we have the rule of concentrated action: Isentropic Flow of a Perfect Gas
the pressure disturbance is largely concentrated in the neigh-
borhood of the Mach cone that forms the outer limit of the If the rate of change of fluid properties normal to the stream-
zone of action (Fig. 9.4d). For this reason only, the sound pro- line direction is negligible as compared to the rate of change
duced by a fighter aircraft moving at supersonic speed cannot along the streamlines, the flow can be assumed to be one-
be heard until the wave attached to the aircraft’s nose passes dimensional. For flow in ducts this means that all the fluid
over the ear of the observer. properties can be assumed to be uniform over any cross section
9.4 One-Dimensional, Steady, and Isentropic Flow of a Perfect Gas 205
of the duct. These properties which define the state of a sys- and
tem are called static properties, and the properties at a state
cp
which is achieved by decelerating the flow to rest through an γ= (9.33)
cv
isentropic process (i.e., reversible and adiabatic process) are
known as stagnation properties. This stagnation state corre- Thus,
sponding to any fluid state is arrived by proceeding from that γR
cp = (9.34)
state along a line of constant entropy, usually denoted with γ−1
subscript “0”.
For the compressible flows, changes in enthalpy and the Introducing Eq. (9.34) into (9.32), we get
kinetic energy are much larger than that in elevation. Thus,
between any two points along a streamline, the specific static T0 γ − 1 v2
=1+ (9.35)
enthalpy (h) and fluid velocity (v) are related by Eq. (8.43). T 2 γRT
v12 v2 Also,
h1 + = h2 + 2
2 2 v v
M= =√
where “1” and “2” denotes two arbitrary points on a stream- a γRT
line.
Hence, Eq. (9.35) becomes
T0 γ−1 2
9.4.1 Stagnation (or Total) Properties in an =1+ M (9.36)
T 2
Isentropic Flow
Furthermore, from isentropic relations
In the above discussion, if at one of the locations (location
“1”), the flow is brought to rest through a reversible and adi- γ γ
p0 T0 γ−1 ρ0
abatic process (i.e., isentropic process), and the enthalpy at = =
p T ρ
this location is the stagnation enthalpy. Since there is no heat
transfer, the value of enthalpy remains constant on any given we obtain
streamline. Let h0 represent the specific stagnation enthalpy
γ
of the flow, and Eq. (8.43) becomes p0 γ − 1 2 γ−1
= 1+ M (9.37)
p 2
1 1
h0 = h + v2 (9.30) ρ0 γ − 1 2 γ−1
2 = 1+ M (9.38)
ρ 2
Further, for a perfect gas, we know that h = cp T and
h0 = cp T0 . Thus, the above equation can be written as These are the well-defined isentropic relations for one-
dimensional fluid flows, by using which the thermodynamic
1 properties are tabulated in the form of isentropic table
cp T0 = cp T + v2 (9.31)
2 (Appendix B). The critical or sonic properties are attained
if the local fluid is imagined to expand or compress isen-
where T is the static temperature and T0 is the stagnation
tropically till it reaches M = 1. It is denoted by superscript
temperature. Dividing Eq. (9.31) by cp T gives
asterisks. For M = 1, Eqs. (9.36), (9.37), and (9.38) become
T0 v2 T0 1 + γ
=1+ (9.32) = (9.39)
T 2cp T T∗ 2
γ
p0 1 + γ γ−1
Again recall Eq. (8.129) = (9.40)
p∗ 2
1
cp − cv = R ρ0 1 + γ γ−1
= (9.41)
ρ∗ 2
For air (γ = 1.4), Eqs. (9.39), (9.40), and (9.41) give The constant in the above equation can be calculated at three
reference conditions:
T∗
=0.8333 (9.42)
T0 1. at zero flow speed (v = 0),
p∗ 2. at zero temperature (T = 0), and
=0.5282 (9.43) 3. at acoustic speed (v∗ = a∗ ).
p0
ρ∗
=0.6339 (9.44) Consequently, three useful forms of the energy equation are
ρ obtained as
2 2 2 2
9.4.2 The Dimensionless Velocity (M∗ ) v2 + a = a (9.50)
γ−1 γ−1 0
2 2
We have seen that the asterisk (∗) generally denotes the value v2 + a = vmax
2
(9.51)
γ−1
of a property at M = 1; however, this convention is not fol-
2 2 γ + 1 ∗2
lowed in defining the dimensionless velocity. In gas dynamics, v +
2
a = a (9.52)
the Mach number (M) is the quite useful parameter but it has γ−1 γ−1
its own associated disadvantages. First, M is not dependent
Note that Eqs. (9.50), (9.51), and (9.52) are known as kine-
upon the velocity alone; rather, it depends upon the state of
matic forms of the energy equation for a steady adiabatic flow.
the fluid particularly the temperature. Second, at very high-
Rearranging the terms in Eq. (9.52), we get
speed (v → ∞), the Mach number tends to become very large
(M → ∞). Therefore, it is often convenient to use the ratio
v2 2 a 2 γ+1
of flow speed to acoustic speed at the state where M = 1, + = (9.53)
a∗2 γ − 1 a∗ γ−1
that is, the state at which the flow speed is equal to the sonic
speed (v∗ = a∗ ). This ratio is called the dimensionless veloc- From Eq. (9.47), the above equation is written as
ity (M∗ ). Thus,

2 M∗2 γ+1
v∗ v
M = ∗ = ∗ (9.45) M∗2 + =
a v γ−1 M2 γ−1
Squaring both sides, we get or

v 2 v 2 a 2 2 + (γ − 1) M2 γ+1
M∗2 =
M∗2 = ∗ = × ∗ (9.46) (γ − 1) M2 γ−1
a a a
or
or
a 2 (γ+1) 2
∗2
M =M × ∗ 2
(9.47) 2 M
a M∗2 = (9.54)
(γ−1) 2
1+ 2 M
2
We will now evaluate the term aa∗ using the energy equa-
tion for a steady and adiabatic flow of a perfect gas. From Equation (9.54) is the required expression for the dimension-
Eq. (9.30), we write less velocity. It can be seen that M∗ is not only a simple
function of M, but is related to M in such a way that
1
h0 = h + v2 = constant (9.48)
2 • If M < 1 then M∗
<1
For perfect gas, • If M = 1 then M∗
=1
• If M > 1 then M∗
>1
1
cp T0 = cp T + v2 = constant • If M = 0 then M∗
=0

2 γ+1
• If M → ∞ then M∗ → γ−1
γR
But cp = γ−1 and a2 = γRT. Thus, above equation can be
written as Thus, it is evident that both M and M∗ act in similar fashion,
except for very large Mach numbers (M → ∞). This is an
a2 v2 important result that makes M∗, a bounded parameter. Also,
+ = constant (9.49) M∗ depends upon the sonic condition in contrast to M which
γ−1 2

depends upon the fluid temperature. One should note that dρ
= a2
all the above results are valid for the steady, adiabatic flow dp s
whether reversible or irreversible.
Suppose a flow with the given pressure, temperature, and Thus, Eq. (9.61) becomes
velocity is brought to the rest through an irreversible adiabatic 2
process, then the final pressure will be less than the isentropic dA v dv
= 2 −1 (9.62)
pressure; however, the final temperature will be the same as A a v
the isentropic temperature. Thus, Eq. (9.36) is valid for both
reversible and irreversible processes, whereas Eq. (9.37) and But v
a = M, thus
Eq. (9.38) are valid only for a reversible and adiabatic process.
dA 2 dv
= M −1 (9.63)
A v
9.4.3 Effect of Area Variation on Compressible This is the well-known area–velocity relation for a steady
Flow (Area–Velocity Relation) isentropic flow. For different flow regimes based on the Mach
number, let us examine the following two cases.
Consider a steady, adiabatic flow in a streamtube of variable
area. The mass flow rate at any cross section is given by 1. If the flow is subsonic (M < 1) at the inlet of a streamtube,
then for dvv > 0 we have A < 0, i.e., for an accelerating
dA
ρAv = constant (9.55) flow the area must decrease in the flow direction and vice
versa. Thus, for the subsonic flow a convergent passage
Taking logarithmic both sides, we get acts as a nozzle, while the divergent passage acts as the
diffuser, as shown in Fig. 9.5.
ln ρ + ln A + ln v = ln (constant) (9.56) 2. When the inlet flow is supersonic (M > 1) then for dv v >0
we have A > 0, i.e., for accelerating flow the area must
dA
On differentiation increase in the flow direction and vice versa. Thus, for a
dρ dv dA supersonic stream, a divergent passage acts as a nozzle and
+ + =0 (9.57) the convergent passage acts as the diffuser (Fig. 9.6).
ρ v A
3. Consider the subsonic flow at the inlet of a convergent
or nozzle, as shown in Fig. 9.5. From Eq. (9.63), it is evident
that for M = 1, we get
dA dρ dv dp dρ dv
=− − =− × − (9.58)
A ρ v ρ dp v dA
=0 (9.64)
A
Recall Euler’s equation (Eq. (9.17))
i.e., the location of sonic condition is either a minimum
dp = −ρvdv or maximum area location. Since the second- and higher
order derivatives of Eq. (9.63) are zero, we have to perform
or the “first derivative test” to ascertain the nature of this sta-
tionary point at which M = 1. Consider a small positive
dp quantity to evaluate dA
= −vdv (9.59) A at “slightly less” and “slightly
ρ more” than M = 1. Note that, in the present case we are
interested in knowing the sign of gradients and not their
Introducing Eq. (9.59) into (9.58), we obtain
actual value. Equation (9.63) shows that, for M = 1 − we
get dAA < 0; and for M = 1 + we have A > 0. Clearly,
dA
dA dρ dv 2 dρ dv
= vdv × − = v −1 (9.60) the stationary point is a minimum turning point or minima.
A dp v dp v
Interestingly, it happens to be the nozzle exit in this case.
Now, if the flow in streamtube is frictionless then for an isen- Therefore, for the subsonic flow at the inlet, the maximum
tropic process the above equation can be written as Mach number achieved is M = 1 at the exit of the nozzle.
To obtain supersonic flow at the exit, we must add a diffuser
or divergent section right at the minimum area location.
dA 2 dρ dv
= v −1 (9.61) This type of nozzle is called the convergent–divergent noz-
A dp s v
zle or de Laval nozzle, as depicted in Fig. 9.7. Note that a
But from Eq. (9.20) convergent–divergent nozzle with straight boundaries pro-
dp < 0, dv > 0 dp < 0, dv > 0

Flow
Flow
Subsonic Nozzle Subsonic Diffuser
Fig. 9.5 Schematic diagram of a subsonic nozzle and the diffuser
dp < 0, dv > 0 Flow dp < 0, dv > 0
Flow
Supersonic Nozzle Supersonic Diffuser
Fig. 9.6 Schematic diagram of a supersonic nozzle and the diffuser
Convergent section
M<1
Divergent section
v1
M>1
M=1
p1
v2
Throat
p2
Fig. 9.7 Schematic layout of a typical convergent–divergent nozzle with contoured boundaries
p √
duce uniform flow at the exit; however, this flow may not But ρ = RT and v = M γRT. Hence, Eq. (9.65) becomes
be unidirectional. Thus, to obtain both uniform and uni-
directional flows at the nozzle exit, the nozzle boundaries p
m̊ = AM γRT
are suitably contoured. RT
or
γ
9.4.4 Mass Flow Rate m̊ m̊ = p AM (9.66)
RT
The mass flow rate through a streamtube of cross-sectional From Eqs. (9.36) and (9.37) and with some algebraic arrange-
area A is given by ment, Eq. (9.66) gives

γ
m̊ = ρAv (9.65) AMp0 RT 0
m̊ = (γ+1)
1+(γ−1)M2 2(γ−1)
2
or the flow becomes choked and the streamtube eventually dis-

γ+1 charges a maximum mass flow rate. A further decrease in pb
m̊ γ γ − 1 2 − 2(γ−1)
= p0 M 1+ M (9.67) will have no effect on the mass flow rate. This is because
A RT0 2 the acoustic signals cannot travel upstream of the stream-
tube exit (throat) to cause changes in the flow conditions.
This is the mass flow rate per unit area in a streamtube as a
Thus, the fluid streamlines make a sudden change of direc-
function of stagnation pressure, stagnation temperature, and
tion at the exit and expand to a much greater area. Hence,
the Mach number.
accelerating the fluid to higher velocities requires a diver-
gent section to be attached at the end of the convergent por-
9.4.4.1 Maximum Mass Flow Rate in a Convergent
tion.
Streamtube
Let us examine the condition for the maximum mass flow
rate inside a convergent streamtube. Suppose the air is drawn
9.4.5 Mass Flow Rate in a Choked Streamtube
inside this streamtube from a large reservoir with constant
stagnation conditions by decreasing the back pressure (pres- We will now calculate the mass flow rate which is being dis-
sure at the downstream of the streamtube exit) as shown charged from a streamtube when the throat becomes choked.
in (Fig. 9.8). For an isentropic flow, the stagnation pressure The discharged mass flow rate m̊ from any streamtube is
and temperature will be constant throughout the streamtube. given by
From maxima–minima principle, differentiating Eq. (9.67)
with respect to M and setting the result equal to zero, we m̊ = ρvA (9.70)
have
Similarly, the mass flow rate at the throat is given by
d m̊A
=0 (9.68) m̊ = ρ∗ v∗ A∗ (9.71)
dM
On solving, we get where A∗ , ρ∗ , and v∗ are the area, density, and the velocity
at the throat, respectively. Further, from isentropic relation
M= 1 under choked conditions, we have
1
Again considering the second derivative of m̊
A and introducing ρ0 (γ + 1) (γ−1)
= (9.72)
M = 1, we get ρ∗ 2

T∗ (γ + 1) −1
d2 m̊ = 1+ (9.73)
A T0 2
<0
dM2
Also, we know that
Clearly, at sonic conditions,
the mass flow rate per unit area
will be maximum, i.e., m̊ A max at M = 1. We have seen
v∗ = a ∗ = γRT∗ (9.74)
earlier that by definition the minimum area location of any
streamtube is known as the throat; in case of a convergent Using Eqs. (9.72) and (9.74) in the (9.71), we obtain
streamtube, the exit plane can be considered as the throat,
−1
i.e., Ae = A∗ . Thus, the maximum mass flow rate per unit m̊ (γ + 1) (γ−1)
= ρ0 γRT∗ (9.75)
area is given by A∗ 2
− γ+1 or
m̊ m̊ γ γ+1 2(γ−1) −1
= ∗ = p0 (9.69) m̊ (γ + 1) (γ−1) T∗
A A RT0 2 = ρ0 γR T0 (9.76)
max
A∗ 2 T0
i.e., for given stagnation conditions the maximum mass flow
p0
rate per unit area is directly proportional to √pT0 . But ρ0 = RT0 , then Eq. (9.76) becomes
0
Note that, if the reservoir pressure (p0 ) is held constant
−1
and the back pressure (pb ) is simultaneously reduced, the m̊ p0 (γ + 1) (γ−1) T∗
Mach number at the exit of the streamtube will increase until ∗
= γR T0 (9.77)
A RT0 2 T0
M = 1 is reached. For the sonic state prevailing at the exit,
Large Reservoir
Convergent Streamtube
h0
T0
p
0
Fig. 9.8 Isentropic flow from a large reservoir in a convergent streamtube
Introducing Eqs. (9.73) into (9.77), we get From aerodynamics point of view, this streamtube can be
considered as the convergent–divergent nozzle, illustrated in
−(γ+1) Fig. 9.7, for our convenience.
m̊ γ (γ + 1) 2(γ−1)
= p0 (9.78) In general, the mass flow rate at any cross section (A) of
A∗ RT0 2
the streamtube can be defined as
Comparing Eqs. (9.78) with (9.69), it is evident that the mass
flow rate per unit area of a streamtube remains unaltered even m̊ = ρvA (9.79)
after the flow gets choked. Equation (9.78) clearly shows that
the choked mass flow rate can be increased by increasing the Similarly, the mass flow rate at the throat is given by
reservoir pressure p0 or decreasing the reservoir temperature
T0 . m̊= ρ∗ v∗ A∗ (9.80)
where A∗ , ρ∗ , and v∗ are the area, density, and the velocity at

9.4.6 Physical Interpretation of Flow at the the throat, respectively. Thus, for steady flow, we write
Throat
ρvA = ρ∗ v∗ A∗ (9.81)
We have seen that for a given stagnation pressure ratio the
mass flow rate becomes maximum when M = 1 at the throat. or
At these conditions, the flow is said to be choked. Let us A ρ∗ v∗
= (9.82)
qualitatively analyze the choking phenomena in a convergent– A∗ ρv
divergent nozzle, depicted in Fig. 9.9. For the constant stag-
nation conditions at nozzle inlet, a slight reduction in back Multiplying and dividing by ρ0 in the above equation, we get
pressure (pb ) causes fresh streamlines entering into the noz-
zle. Further reduction in pb fetches more and more streamlines A ρ∗ ρ0 v∗
= (9.83)
into the nozzle. As we keep on decreasing pb , then a situa- A∗ ρ0 ρ v
tion comes when no more fresh streamlines enter the nozzle.
That is, the clogging of streamlines debars entering the new Suppose the velocity at the throat is given as
ones. Obviously, this is dictated by the minimum area loca-
tion, which is throat only. v∗ = a ∗ (9.84)
Thus, Eq. (9.83) becomes

9.4.7 Area Ratio for the Convergent–Divergent
Streamtube A ρ∗ ρ0 a∗
= (9.85)
A∗ ρ0 ρ v
The purpose of this section is to derive the area–Mach number
relation for a varying area streamtube, as shown in Fig. 3.23a. Squaring both sides, we obtain
Divergent Section
Convergent Section
Throat
Streamlines
Fig. 9.9 Streamlines in a convergent–divergent nozzle
2 2 2 2
A ρ∗ ρ0 a∗ 9.4.8 Types of Characteristic Speeds along a
= (9.86)
A∗ ρ0 ρ v Streamtube
But In this section, we will examine three reference speeds.

1
ρ0 (γ − 1) 2 γ−1
= 1+ M (9.87)
ρ 2 9.4.8.1 Maximum Flow Velocity (vmax )
Rearranging Eq. (9.31) gives
1
ρ0 (γ + 1) γ−1
= (9.88) v = 2cp (T0 − T) (9.91)
ρ∗ 2

γ+1 Substituting Eq. (9.34) into (9.91), we get
v 2 2 M2
∗2
=M = (9.89) 1/2
a∗ 1 + γ−1 M2 2γR
2 v= (T0 − T) (9.92)
γ−1
Introducing Eqs. (9.87), (9.88), and (9.89) into Eq. (9.86), we
get The maximum velocity (vmax ) can be obtained by setting
T = 0 in Eq. (9.92), and we obtain
(γ+1)
A 1 2 (γ − 1) 2 (γ−1) 1/2
= 2 1+ M (9.90) 2γRT0
A ∗ M (γ + 1) 2 vmax = (9.93)
γ−1
This is the critical area ratio relation, also known as area– 2γRT0
Mach number relation. Thus, the throat area of a duct can be vmax = (9.94)
γ−1
calculated if the Mach number (M) at a given cross-sectional
area (A) is known. A typical plot between the area ratio of a This is the maximum speed attained by a gas when expanded
convergent–divergent duct with the Mach number is schemat- isentropically to the vacuum (i.e., when the final pressure is
ically shown in Fig. 9.10. The following points can be con- zero). In other words, it is the speed achieved by the gas when
sidered. the kinetic energy of molecules in random motion is com-
pletely transformed into directed kinetic energy. The static
• For any given area ratio AA∗ , there are two values of temperature corresponding to that state will be zero. It is also
Mach numbers: one subsonic (M < 1) and other super- called the absolute zero temperature at which the substances
sonic (M > 1). possess no thermal energy. Note that Eq. (9.94) is valid for
• The area ratio AA∗ is always greater than unity, i.e., both reversible and irreversible flows as Eq. (9.31) is univer-
A∗ > 1.
A
sally valid.
Isentropic flows
A
A*
M<1 M>1
M=1
M

Fig. 9.10 The plot of area ratio A
A∗ with the Mach number (M)
Rewriting Eq. (9.94), we obtain 9.4.9 Mass Flow Rate Variation with Pressure
2
vmax 2 Let us consider an isentropic fluid flow accelerating through
= (9.95)
ao γ−1 a convergent–divergent nozzle. From the steady-state energy
equation (Eq. (8.43)), we have
or
vmax γ+1 v12 v2
= (9.96) h1 + = h2 + 2
a∗ γ−1 2 2
Note that the subscripts “1” and “2” refers to inlet and exit
This is the relation between the critical speed and the maxi-
sections of the nozzle, respectively. If the fluid velocity at the
mum speed of the fluid flow.
inlet is much smaller than the exit velocity, i.e., v1 v2 , then
above becomes
9.4.8.2 Acoustic Speed at Stagnation v2 ≈ 2 (h1 − h2 ) (9.100)
Temperature (a0 )
Recall Eq. (9.26), where (h1 − h2 ) is the enthalpy drop across the nozzle. Fur-
ther, using Eq. (8.61) in terms of intrinsic properties, we write
a= γRT
Tds = dh − ∀dp (9.101)
The speed of sound at the stagnation temperature is obtained
by substituting T = T0 into the above equation. Thus, But for isentropic flow ds = 0, then
a0 = γRT0 (9.97) dh = ∀dp (9.102)
Integrating Eq. (9.102) from “1” to “2”, we get

9.4.8.3 Characteristic Velocity (v∗ )
We know that the fluid velocity and acoustic speed are equal
2 2
at sonic condition. Thus, we write
dh = ∀dp
1 1
v∗ = a ∗ = γRT∗ (9.98)
or
or 2
1/2 (h1 − h2 )= − ∀dp (9.103)
2γ
v∗ = RT0 (9.99) 1
γ+1
Let us further assume that the pressure and volume of the fluid
This is known as the characteristic velocity.
during expansion obey the gas law
p∀n = constant = K (say) (9.104) Clearly, the exit pressure (p2 ) influences the mass flow rate
for a given inlet condition. Again from the maxima–minima
where n is the polytropic index. On solving Eqs. (9.103) and principle, the maximum mass flow rate per unit area Am̊2
(9.104), we get max
through the nozzle can be obtained by differentiating equa-
tion (9.115) with respect to pp21 . Indeed, for a given inlet con-
2 2
n 1/n −1/n ditions, the mass flow rate will be maximum when the term
(h1 − h2 )= − ∀dp = − p∀ p dp (9.105) within the square bracket on the right side of Eq. (9.115) is
1 1 maximum. For convenience, assume pp21 = y.
d 2
Integrating above, we have
n+1
yn − y n = 0 (9.116)
2 dy
1/n p1−1/n
(h1 − h2 ) =− p2 ∀2 (9.106)
1 − 1/n 1 Finally, we obtain
n−1
n 1/n
n−1

=− p2 ∀2 p2 n − p1 n (9.107) 2
n
n−1
n−1 y= (9.117)
n+1
n 1/2
n−1
=− p2 ∀2 − p1 ∀1 p1 n (9.108)
n−1 or n
p2 2 n−1
Thus, = (9.118)
p1 n+1
n
(h1 − h2 ) = (p1 ∀1 − p2 ∀2 ) (9.109) This is the required pressure ratio for obtaining maximum
n−1
mass flow rate through a convergent–divergent nozzle. Fur-
Introducing Eqs. (9.109) into (9.100), v2 becomes thermore, replacing n with the ratio of specific heats γ into
Eq. (9.118), we get

2n γ
v2 = (p1 ∀1 − p2 ∀2 ) (9.110) p∗ p2 2 γ−1
n−1 = =
p1 p1 γ+1
or
For diatomic gases (γ = 1.4), one gets
2n p2 ∀2
v2 = p1 ∀1 1 − (9.111)
n−1 p1 ∀1 p∗ = 0.528p1 (9.119)
or i.e., the pressure at the nozzle throat is almost 50% of the inlet
pressure.
n−1
2n p2 n
v2 = p1 ∀1 1 − (9.112)
n−1 p1
9.5 The Adiabatic Flow Ellipse
Furthermore, the mass flow rate at the nozzle exit is
When the speed of sound is plotted as a function of the speed of
the flow for an adiabatic flow of a gas, it results in an ellipse
m̊ = ρ2 A2 v2 (9.113)
known as adiabatic flow ellipse. Let us consider a stream-
tube which is not exchanging any heat with the neighboring
or
m̊ v2 streamtubes. Since the total energy of the steady flow in the
= ρ2 v2 = (9.114) streamtube remains constant, thus we write
A2 ∀2
Substituting Eq. (9.112) into above equation, we obtain the v2

h+ = h0 = constant (9.120)
mass flow rate per unit exit area as 2
But for perfect gas h = Cp T. Thus, the above energy equation

2/n n+1
m̊ 2n p1 p2 p2 can be written as
=
n
− (9.115)
A2 n − 1 ∀1 p1 p1
v2 γR
+ T= constant (9.121)
2 γ−1
a
M = 0.3
M=1
I II
a0 III M=5
o
45
IV
v
vmax
(I) Incompressible Subsonic Flow

(II) Compressible Subsonic Flow
(III) Supersonic Flow
(IV) Hypersonic Flow
Fig. 9.11 Schematic diagram of a typical adiabatic ellipse in the steady flow

Also, from T = a2 da γ−1
γR , we have = M (9.127)
dv 2

v2 a2 or
+ = constant (9.122)
2 γ−1 2 da
M= − (9.128)
γ−1 dv
At T = 0 we have v = vmax . Hence, the constant in
2 . Thus, Eq. (9.122) becomes
Eq. (9.122) will be equal to vmax Note that the change of slope from point to point indicates the
change of Mach number, i.e., the change in speed of sound
v2 a2
+ = vmax
2
(9.123) and velocity. The following points can be considered.
2 γ−1
• At high-speed flows, the change in speed of sound essen-
or
v2 2 a2 tially results in change in the Mach number.
+ =1 (9.124) • At low-speed flows, the change in velocity mainly results
2
vmax γ−1 2
vmax
in change in the Mach number.
Comparing Eqs. (9.95) and (9.124), we obtain • For M < 0.3, the flow is treated as incompressible and the
change in the speed of sound is negligibly small.
v2 a2
2
+ 2 =1 (9.125)
vmax a0
9.6 Processes Causing a Change of State in
This is the equation of ellipse having the major axis vmax
Compressible Flows
and the minor axis of a0 , as shown in Fig. 9.11. Rewriting
All the flow processes aim at change of state of a system from
Eq. (9.125), we get
one state to another. For compressible flows, this is attained
by the following approaches.
v2
a = 2
a02 − 2
a02
vmax • In processes, where the heat transfer effects are consid-
ered negligible and the effects due to friction are relatively
Differentiating above with respect to v and simplifying,
small, then the flow may be considered reversible and adi-

da γ − 1 v abatic and hence isentropic. Thus, in isentropic flows, the
=− (9.126) fluid properties change with a change in the cross-sectional
dv 2 a
area of the streamtube.
But M = va . Thus, • Consider the flow in a streamtube of constant cross-
sectional area. If the change of state of fluid is caused
9.6 Processes Causing a Change of State in Compressible Flows 215
2
Fanno Flow Rayleigh Flow
T
Isentropic Flow across

Flow Normal Shock
1
Fig. 9.12 T − s diagram for Fanno flow and Rayleigh flow of a perfect gas
purely by friction, neglecting the heat transfer, the flow 9.6.3 Normal Shock Wave
is known as Fanno flow; however, if the change of state
of fluid is caused solely by heat transfer, ignoring the The abrupt change of flow speed from the supersonic level to
frictional effects, the flow is known as Rayleigh flow. For subsonic level must occur through a normal shock and thus
a perfect gas, the Fanno and Rayleigh curves are shown the intersection points of Rayleigh and Fanno curves, i.e.,
in Fig. 9.12. In this section, we will discuss these flows locations “1” and “2”, respectively, denotes the upstream and
briefly. For more details, readers are advised to refer downstream conditions across the shock wave.
advanced texts such as Shapiro (1953) and Liepmann and
Roshko (1957).
9.7 One-Dimensional Flow Across a Normal
Shock
9.6.1 Fanno Curve
It has been observed that a compressible fluid under specified
The equation of this curve is obtained by solving the conti- conditions may experience a sudden change of state. The com-
nuity equation (Eq. (9.1)) and the energy equation (Eq. (9.6)) mon examples are the phenomena associated with detonation
together with the equation of state. The momentum equa- waves, explosions, and the wave system formed at the nose of
tion (Eq. (9.2)) is not considered into the analysis and thus body moving at supersonic Mach numbers. In all these cases,
Fanno curve depicts the states with the same mass flow rate the wavefront is very steep and the flow experiences a large
per unit area and same stagnation enthalpy but different values increase in pressure in traversing the wave, which is called a
of impulse function. Clearly, the friction causes the change of shock wave. Due to the large pressure gradient in the shock
states from “1” and “2”, respectively, upstream and down- wave, the gas experiences a large increase in its density with
stream locations along the Fanno curve passing through these a corresponding change in its refractive index.
locations. Since the shock acts as a compression front, it cannot be a
reversible process. The energy for compressing the gas flow-
ing through the shock wave is derived from the kinetic energy
9.6.2 Rayleigh Curve possessed by the flow upstream to the shock. Because of the
irreversibility of the shock process, the kinetic energy of the
The Rayleigh curve is the locus of states defined by the gas downstream of the shock is smaller than that for isentropic
equation obtained through solving the momentum equation flow compression between same pressure limits. This reduc-
(Eq. (9.2)), the continuity equation (Eq. (9.1)), and the equation in kinetic energy due to the shock appears as a heating of
tion of state. Since the energy equation (Eq. (9.6)) is not con- the gas to a static temperature above that corresponding to the
sidered here, Rayleigh curve represents the states with the isentropic compression value. Consequently, the gas passing
same mass flow rate per unit area and same impulse func- the shock wave experiences a decrease in its available energy
tion but different values of stagnation enthalpy. Therefore, and, accordingly, an increase in its entropy.
heat transfer effects are required to cause the change of states The shock wave is a very thin compression in front of
from “1” and “2” along this curve. the thickness comparable to the mean free path of the gas
molecules in the flow field. Since the flow crosses the shock done on the control volume, the energy equation becomes
at very high Mach numbers and thus, the combination of high
velocity of the flow and extremely small thickness of the shock h0 = h01 = h02 (9.131)
wave makes the fluid elements pass through the wave in an
infinitesimal time, ruling out any appreciable exchange of where h0 is the specific stagnation enthalpy. Across the shock,
energy between fluid elements and the surroundings, render- the above equation will be
ing the shock process to be adiabatic.
There are several different types of shock waves, each hav- v12 v2
h1 + =h2 + 2 (9.132)
ing particular characteristics. A shock which is normal to the 2 2
direction of the flow is known as normal shock, while the
shock which is at an angle to the flow is termed as oblique For the given conditions upstream of the shock wave,
shock. Eqs. (9.129), (9.130), and (9.132) provide three nonlinear
equations for the four unknowns p, v, h, and p downstream
of the wave. To obtain a closed-form solution, we require
9.7.1 Governing Equations of the Normal Shock
an additional equation which is provided by the equation of
state of the fluid. For a perfect gas, the equation of state can
The purpose of this section is to develop the explicit relation-
be written as
ships for the property change of the perfect gas across the nor-
p = ρRT (9.133)
mal shock wave in terms of the upstream Mach number (M1 ).
The flow through a normal shock may be analyzed by consid-
To calculate the change of entropy across the normal shock,
ering the one-dimensional flow through a streamtube. Con-
let us recall Eq. (8.61)
sider a control volume, as shown in Fig. 9.13, which includes
the normal shock and a small amount of fluid both upstream
Tds = dh − ∀dp
and downstream of the shock. Let the conditions upstream
and downstream of the shock are designated with subscripts
For a perfect gas, we write
“1” and “2”, respectively (Fig. 9.13). We will consider the
following assumptions in deriving the governing equations. dT dp
ds = cp −R
T p
• One-dimensional, steady, and adiabatic flow with no fric-
tion. Integrating above, we get
• The change in properties across the shock is irreversible.
• Shock wave is perpendicular to the flow. T2 p2
s2 − s1 = cp ln − R ln (9.134)
• Shock thickness is very small. T1 p1
• The area of cross section just upstream and downstream of
the shock is constant. Equations (9.129) to (9.134) are the governing equations for
• No external work is done on the control volume. the flow of a perfect gas through the normal shock wave.
• Body forces are negligibly small.
The continuity equation is 9.7.2 Normal Shock Equations for a Perfect Gas
m̊
= ρ1 v1 = ρ2 v2 (9.129) For a perfect gas, the energy equation given by Eq. (9.132)
A
can be simplified as
Neglecting body forces, the application of Newton’s second
law of motion to the control surface gives v12 v2
cp T1 + = cp T2 + 2 = cp T0 (9.135)
2 2
m̊
p1 − p2 = (v2 − v1 ) = ρ2 v22 − ρ1 v12 or
A
or T01 = T02 (9.136)
p1 + ρ1 v12 = p2 + ρ2 v22 (9.130)
That is, the stagnation temperature of a perfect gas remains
Note that, as defined earlier, pA + ρAv2 is the impulse func- constant across the shock. Further, using the isentropic rela-
tion. Since the flow is adiabatic and there is no external work tion (Eq. (9.36)), the above equation takes the form
9.7 One-Dimensional Flow Across a Normal Shock 217
Ahead of shock Behind the shock

p p <p
01 T 01 02 01 T02 = T 01
p T ρ v p > p T >T
1 1 11 2 1 2 1
ρ>ρ v <v
M1 > 1 2 1 1 2
M2< 1
Normal shock Control surface
Fig. 9.13 Flow across a normal shock wave
γ−1 2
T2 1+ 2 M1 Eliminating pp21 from Eqs. (9.142) and (9.144), we get the
= γ−1 2
(9.137)
T1 1+ following relationship for the Mach numbers upstream and
2 M2
downstream of a normal shock wave
From the continuity equation (Eq. (9.129)) and equation of 1 1
state (Eq. (9.133)), we get M1 γ−1 2 2 M2 γ−1 2 2
1+ M1 = 1+ M2
1 + γM12 2 1 + γM22 2
T2 p2 ρ1 p2 v2
= = (9.138) (9.145)
T1 p1 ρ2 p1 v1
Solving the above equation explicitly for M2 , we obtain two
But solutions.
v = M γRT (9.139) M2 = M1 (9.146)
Thus, This is a trivial solution, which signifies no shock wave in the

control volume and thus, it is discarded. The other solution is
p1 p2
M1 T1 = M2 T2 (9.140)
T1 T2 (γ − 1) M12 + 2
M22 = (9.147)
or 2γM12 − (γ − 1)
2 2
T2 p2 M2 This equation expresses the Mach number on the downstream
= (9.141) to the shock in terms of the Mach number on the upstream of
T1 p1 M1
the shock. Note that, for M1 = 1 Eq. (9.147) gives M2 = 1.
From Eqs. (9.137) and (9.141), we obtain the equation in Consequently, Eqs. (9.144) and (9.137) show that there would
terms of p and M be no rise in pressure or temperature across the shock wave.
In this condition, the shock wave behaves as the sound wave
⎛ ⎞
1+ γ−1 2 across which, by definition, the pressure and temperature
p2 M1 2 M1
= ⎝ ⎠ (9.142) changes are negligibly small. Therefore, it can be said that
p1 M2 1+ γ−1 2
the sound wave represents a degenerated normal shock wave.
2 M2
Introducing Eqs. (9.133) and (9.139) into the momentum

equation (Eq. (9.130)), we get 9.7.3 Working Relations for Normal Shock Wave

p1 1 + γM12 = p2 1 + γM22 (9.143) Often the upstream Mach number (M1 ) is known in a given
problem and it is desired to determine the downstream Mach
number (M2 ) of the shock wave. Once M2 is calculated by
or
p2 1 + γM12 using Eq. (9.147), the working relations for a given gas are
= (9.144) obtained as follows:
p1 1 + γM22
9.7.3.1 Static Pressure Ratio or

The static pressure ratio across the normal shock wave ⎧ ⎫
⎨ 1+ γ−1 2
is derived by using Eqs. (9.144) and (9.147). Introducing T2 2 M1
2γ 2
(γ−1) M1 −1 ⎬
Eqs. (9.147) into (9.144), we get = (9.152)
T1 ⎩ (γ+1)2 2 ⎭
2(γ−1) M1
p2 1 + γM12
=
p1 (γ−1)M12 +2 9.7.3.3 Stagnation Pressure Ratio
1+γ 2γM12 −(γ−1) The stagnation pressure ratio across the normal shock is cal-
culated by using the following relation:
1 + γM12
= 2 2γ
γM1 + (γ−1) p02 p02 p2 p1
1+ 2γ = × × (9.153)
(γ−1) M1 −1
2
p01 p2 p1 p01
2γ

1 + γM12 M 2−1
=
(γ−1) 1
(9.148) Note that pp022 and pp011 are obtained by using the isentropic
2γ 2γ relation (Eq. (9.37)) as
2
(γ−1) M1 − 1 + γM12 + (γ−1)
γ
Simplifying the above equation and rearranging the terms, we p02 γ − 1 2 (γ−1)
= 1+ M2 (9.154)
get p2 2
γ
2γ p1 γ − 1 2 − (γ−1)
1 + γM 2 M 2−1 = 1+ M1 (9.155)
p2 1 (γ−1) 1 p01 2
=
p1 (γ+1) (γ+1)
(γ−1) + (γ−1) γM 2
1 Substituting above equations and Eq. (9.151) into (9.153), we

2γ get
1 + γM12 (γ−1) M12 − 1
= (γ+1) γ

(γ−1) 1 + γM1 2γ 2
2 γ−1 2 (γ−1)
p02 1+ 2 M2
= 1+ M1 − 1
2γ γ−1 2 γ+1
(γ−1) M1
2−1 p01 1+ 2 M1
= (γ+1)
(9.149) (9.156)
(γ−1)
or Introducing the value of M22 from Eq. (9.147) into the above
equation, we obtain
p2 2γ (γ − 1)
= M12 − (9.150) ⎡ ⎤ γ
p1 (γ + 1) (γ + 1) 1 + γ−1
(γ−1)M12 +2 (γ−1)

p02 ⎢ 2 2γM12 −(γ−1) ⎥ 2γ 2
=⎢ ⎥ 1+ M1 − 1
Thus, we have p01 ⎣ 1 + γ−1 2 ⎦ γ+1
2 M1
p2 2γ 2 (9.157)
=1+ M1 − 1 (9.151) ⎤ γ
p1 γ+1 ⎡ 2
M12 + (γ−1) (γ−1)
γ−1
⎢1 + 2 ⎥
2γ 2
2γ
⎢ 2
(γ−1) M1 −1 ⎥
9.7.3.2 Static Temperature Ratio =⎢
⎢ γ−1 2
⎥
⎥ 1+ M1 − 1
⎣ 1 + 2 M1 ⎦ γ+1
The ratio of static temperatures across a normal shock is cal-
culated by substituting Eq. (9.147) into (9.137). That is, (9.158)
⎡ ⎤ γ
1 + γ−1
(γ−1)
T2 M12 (γ+1)2
2
2γ 2
⎢ 2(γ−1) M1 ⎥
= 2 = ⎣ ⎦ 1+ M1 − 1
T1 (γ−1)M2 +2 1 + γ−1 2 2γ 2−1 γ+1
1 + γ−1 1 2 M1 M
(γ−1) 1
2 2γM12 −(γ−1)
(9.159)
1 + γ−1 2 M1
2 2γ
(γ−1) M1 − 1
2
= Simplifying above, we get
2γ
M 2 − 1 + γ−1 M2 + 2
(γ−1) 1 2 1 (γ−1) ⎡ ⎤

γ
(γ+1) 2 (γ−1)
γ−1 2 2γ p02 ⎣ M1 2γ 2
1 + 2 M1 (γ−1) M1 − 1 2
p01
=
γ−1 2
2
2γ γ−1
⎦ 1+
γ+1
M1 − 1
= 1+ 2 M1
2
(γ+1) M1 − (γ+1)
2γ γ−1
(γ−1) + 2 M12 (9.160)
or T2 p2
s2 − s1 = cp ln − R ln
T1 p1
⎡ ⎤ γ
(γ−1)
p02 ⎣
(γ+1) 2
M 2γ 2 (γ−1)
1
= 2 1
⎦ 1+ M1 − 1 where s1 and s2 are the entropy upstream and downstream of
p01 1 + γ−1 M2 γ+1 the normal shock. Note that Eq. (9.134) can also be written
2 1
in terms of stagnation pressures upstream and downstream of
(9.161)
the shock as
This is the stagnation pressure ratio across a normal shock p02
s2 − s1 = −R ln (9.168)
wave. p01
9.7.3.4 Density Ratio or

From the equation of state for a perfect gas (Eq. (9.133)), we p01
write s2 − s1 = R ln (9.169)
p02
ρ2 p2 T
= /
p1 T
2 (9.162) Introducing the values of TT21 and p2
from Eqs. (9.152) and
ρ1 1 p1
(9.151) into (9.134), we obtain
Substituting pp21 and T2
T1 , respectively, from Eqs. (9.151) and
⎧ ⎫
(9.152), we obtain ⎪
⎨ 1 + γ−1 2γ ⎪
⎬
2γ 2
2 2
2 M1 (γ−1) M1 − 1
s2 − s1 = cp ln − R ln 1 + M1 − 1
⎪
⎩ (γ+1)2 2 ⎪
⎭ γ+1
2(γ−1) M1
⎧ ⎫
⎪
⎨ (γ+1)2 2 ⎪
⎬ (9.170)
ρ2 2γ 2 2(γ−1) M1
= 1+ M1 − 1
ρ1 γ+1 ⎪ γ−1
⎩ 1 + 2 M12 2γ 2 ⎪
⎭
(γ−1) M1 − 1
Simplifying above with algebraic manipulation of the terms,
(9.163)
⎧ ⎫ we finally have
⎨ (γ+1) 2 ⎬
2γ 2 2 M1
= 1+ M1 − 1
γ+1 ⎩ 1 + γ−1 M2 2γ 2 (γ−1) ⎭
2 1 (γ+1) M1 − (γ+1) s2 − s1 γ 2γ (γ − 1) 1 2γ 2 − (γ − 1)
= ln + + ln M1
(9.164) R γ−1 (γ + 1) M12 (γ + 1) 1−γ (γ + 1) (γ + 1)
⎧ ⎫ (9.171)
⎨ (γ+1) 2 ⎬
2γ 2 M1
= 1+ M1 − 1 2
γ+1 ⎩ 1 + γ−1 M2 1 + 2γ M2 − 1 ⎭
γ+1
2 1 1
For 1 ≤ γ ≤ 1.67, the entropy change is plotted against the
(9.165)
upstream Mach number, as shown in Fig. 9.14. From the
or figure, it is once again established that the shock waves are
possible only in case of a supersonic flow.
ρ2 (γ + 1) M12
= (9.166)
ρ1 (γ − 1) M12 + 2
9.7.4 The Prandtl–Meyer Relation
This is the density ratio across a normal shock wave. Note
that, for high Mach numbers (M1 → ∞), Eq. (9.166) gives We have seen earlier that across a normal shock wave the stag-
ρ2
ρ1 → 6. Thus, at very high Mach numbers in contrast to
nation enthalpy (or the stagnation temperature for a perfect
the pressure and the temperature which become unbounded gas) of a gas remains constant. Let us apply Eq. (9.52) to the
across the shock wave,density is a narrow range bounded flow upstream and downstream of the shock. We obtain
parameter 0 < ρρ21 < 6 .
v12 1 2 γ+1 a∗2
Using the continuity equation (Eq. (9.1)), the above equa- + a = (9.172)
tion can also be written as 2 γ−1 1 γ−1 2

v22 1 2 γ+1 a∗2
v1 (γ + 1) M12 + a = (9.173)
= (9.167) 2 γ−1 2 γ−1 2
v2 (γ − 1) M12 + 2
The above equations can be written as
9.7.3.5 Entropy Change
For a perfect gas, the change in entropy across the shock wave a12 γ+1 a∗2 γ−1
= − v1 (9.174)
can be calculated by using Eq. (9.134). So that v1 2 v1 2
s2 − s 1 Shock waves are possible

R
0 M1
1
Shock waves are not possible
Fig. 9.14 Change in entropy across the normal shock with upstream Mach number

a22 γ+1 a∗2 γ−1
= − v2 (9.175)
v2 2 v2 2 or

γ + 1 ∗2 1 1 γ−1
Further, the momentum equation (Eq. (9.2)) can be rewritten a − + (v2 − v1 )
2 v1 v2 2
as
= γ (v2 − v1 )
p1 − p2 = ρ2 v22 − ρ1 v12 (9.176)
or
Using continuity equation (Eq. (9.1)), the above equation can
γ + 1 ∗2 v2 − v1 γ−1
be written as a + (v2 − v1 )
2 v2 v1 2
= γ (v2 − v1 )
p1 − p2 = ρ2 v2 v1 − ρ1 v12
= ρ1 v1 (v2 − v1 ) (9.177) or
or (γ + 1) a∗2 + (γ − 1) v2 v1 = 2γv2 v1
p1 p2
− = v2 − v1 (9.178) Simplifying above, we obtain
ρ1 v1 ρ2 v2
But (γ + 1) a∗2 = (γ + 1) v2 v1
γp1
= a12 or
ρ1
and a∗2 = v2 v1 (9.180)
γp2 This is the well-known Prandtl–Meyer relation. Rearranging

= a22
ρ2 the terms in Eq. (9.180), we have
Thus, v1 v2
=1 (9.181)
a∗ a∗
a12 a22
− = γ (v2 − v1 ) (9.179) But from the definition of characteristic velocity (Eq. (9.45)),
v1 v2
we write
Introducing Eqs. (9.174) and (9.175) into (9.179), we get v1
= M1∗ (9.182)
a∗
∗2
γ+1 a γ−1 γ + 1 a∗2 γ−1
− v1 − + v2 and
2 v1 2 2 v2 2
= γ (v2 − v1 ) v2
= M2∗ (9.183)
a∗
p
Substituting above into Eq. (9.180), we get h= e+ (9.192)
ρ
M1∗ M2∗ = 1 (9.184)
where e is the internal energy of the gas. Using above relation,
Eq. (9.191) becomes
This is another useful form of the Prandtl–Meyer relation.

Equations (9.180) and (9.184) show that if the flow upstream p1 1 ρ2 p2 − p 1 p2 1 ρ1 p2 − p1
e1 + + = e2 + +
of normal shock is supersonic then the downstream flow will ρ1 2 ρ1 ρ2 − ρ1 ρ2 2 ρ2 ρ 2 − ρ1
be subsonic. (9.193)
or
9.7.5 The Rankine–Hugoniot Relation
p1 p2 1 ρ2 p2 − p1 1 ρ1 p2 − p1
e2 − e1 = − + −
ρ1 ρ2 2 ρ1 ρ2 − ρ1 2 ρ2 ρ 2 − ρ1
Since the flow process across a shock wave is adiabatic (9.194)
and irreversible, the shock wave can be viewed as an irre-
versible adiabatic compressor. We will now calculate the work
or
required in this compression process through the shock wave.

From the continuity equation (Eq. (9.1)), we write p1 p2 1 p2 − p1 ρ2 ρ1
e2 − e1 = − + − (9.195)
ρ1 ρ2 2 ρ2 − ρ1 ρ1 ρ2
ρ2
v1 = v2 (9.185)
ρ1 or

Substituting above into the momentum equation (Eq. (9.2)), p1 p2 1 (p2 − p1 ) (ρ2 + ρ1 ) (ρ2 − ρ1 )
e2 − e1 = − +
we get ρ1 ρ2 2 (ρ2 − ρ1 ) ρ1 ρ2
(9.196)
2
ρ2
p1 + ρ1 v22 = p2 + ρ2 v22 (9.186) or
ρ1

p1 p2 (p2 − p1 ) 1 1
or e2 − e1 = − + + (9.197)
ρ1 ρ2 2 ρ1 ρ2
ρ22 2
p1 + v = p2 + ρ2 v22 (9.187) On simplification and rearranging the terms, we get
ρ1 2

or (p2 + p1 ) 1 1
e2 − e1 = − (9.198)
2 ρ1 ρ2
ρ22 2
v −ρ2 v22 = p2 − p1 (9.188) This is the well-known Rankine–Hugoniot equation, named
ρ1 2
in recognition of the work carried out by Scottish engineer
Simplifying above, we obtain W. J. M. Rankine and French engineer P. H. Hugoniot. This
equation essentially relates the thermodynamic states across
p2 − p1 ρ1 the normal shock wave. It is equivalent to the first law of ther-
v22 = (9.189)
ρ2 − ρ1 ρ2 modynamics which states that the change in internal energy
of a gas in an adiabatic process is equal the work done (i.e.,
Similarly, we can show that average pressure times the change in specific volume of the
gas).
p2 − p1 ρ2
v12 = (9.190)
ρ2 − ρ1 ρ1
9.8 Supersonic Pitot Probe
Introducing Eqs. (9.189) and (9.190) into the steady flow
energy equation (Eq. (9.6)), we write A pitot-static probe is the most important instrument which
measures both total and static pressures (i.e., the dynamic
1 ρ2 p2 − p1 1 ρ1 p2 − p1
h1 + = h2 + pressure) in a subsonic flow. When the flow is incompressible
2 ρ1 ρ2 − ρ1 2 ρ2 ρ2 − ρ1 (M < 0.3), we can calculate the velocity of the flow by set-
(9.191) ting these pressure values in the incompressible Bernoulli’s
equation but, for the flow with M > 0.3, the velocity can
But the enthalpy is given by
Pitot pressure is the total

pressure behind the shock
M>1
Equally spaced holes

measure the static pressure
Bow shock
Static pressure
Stagnation pressure
Fig. 9.15 A typical supersonic pitot probe
be obtained by using the isentropic relation, as given in

Eq. (9.37). Simplifying above, we obtain
In supersonic flows, a bow-shock forms ahead of the stag-
γ (γ−1)
2γ 2
1
nation port, as shown in Fig. 9.15. Thus, the total or stagna- p02 (γ + 1) 2 (γ−1)
tion pressure measured by the probe is not the actual stagna- = M1 1+ M1 − 1
p1 2 γ+1
tion pressure of the freestream. Rather, the probe measures (9.201)
the total pressure at downstream of the shock wave. The
probe, which measures the stagnation pressure in a supersonic Note that the static pressure (p1 ) in the above equation is
stream, does not have static holes and hence they are called obtained by alternate methods. For example, in supersonic
simply the pitot probe. Since the shock wave gives rise to wind tunnels, p1 is obtained through the static ports mounted
an increase in entropy, the isentropic relation cannot be used on the test section walls.
to calculate the velocity (Mach number) of the freestream.
Therefore, a different relation becomes necessary and this is
called the Rayleigh Pitot formula, derived in the next section. 9.9 Convergent–Divergent Nozzle (de Laval
Nozzle)
9.8.1 Rayleigh Supersonic Pitot Probe Formula In a converging nozzle, the flow Mach number is increased
from M = 0 near the inlet to M = 1 at the exit. Thus, it is
The Rayleigh Pitot probe formula essentially relates the stag- obvious that to induce a supersonic stream at the exit we
nation pressure downstream (p02 ) of the shock wave to the must add a diverging portion following the converging part.
freestream static pressure (p1 ). For simplicity, the portion of Such a nozzle is named convergent–divergent nozzle or de
bow-shock standing just at the probe nose is assumed to be Laval nozzle or simply Laval nozzle, when Carl G.P. de Laval
the normal shock. Thus, pp021 is written as who first used such a configuration in his steam turbines in
late nineteenth century. The schematic diagram of a typical

p02 p02 p01 convergent–divergent nozzle is shown in Fig. 9.16. To gen-
= (9.199)
p1 p01 p1 erate supersonic flow in the test section, a C-D nozzle (or a
Laval nozzle) must be placed upstream of the test section. The
p02 p01
Introducing p01 (Eq. (9.161)) and p1 (Eq. (9.37)), we get nozzle manages the speed of air entering the test section of
the wind tunnel so that the uniform Mach number is set up.
⎡ ⎤ γ
(γ−1)
Mach number is interestingly dictated by the range propor-
(γ+1) 2 1 γ
p02 M1 2γ 2 (γ−1) γ − 1 2 γ−1 tion of the nozzle. An optimally designed nozzle makes the
= ⎣ 2 ⎦ 1+ M1 − 1 1+ M1
p1 1 + γ−1 M2 γ+1 2
2 1 stream parameters uniform over the cross section. The design
(9.200) of an appropriately shaped nozzle contour to get the desired
9.9 Convergent–Divergent Nozzle (de Laval Nozzle) 223
A entry A exit
At
M<1 Convergent section Throat Divergent section M>1
M=1
Fig. 9.16 Schematic layout of a typical convergent–divergent nozzle
uniform stream at the nozzle exit depends on the method of section of the nozzle become frozen. That is, the subsonic
characteristics (MOC). Recall Eq. (9.90) which dictates the flow in the convergent portion of the nozzle remains unaf-
flow field in a nozzle. The variation of area ratio with the fected and mass flow remains constant for pe < pe3 . At this
Mach number is shown in Fig. 9.10. condition, sonic flow prevails at the throat and the nozzle is
From Eq. (9.90), it is evident that local area (A) of the referred to as “choked”. Further reduction of exit pressure
passage at any location must be larger than or at least equal below pe3 does not increase the mass flow once the throat is
to the throat area (A∗ ) and the case, A < A∗ , is physically choked. At thisstage, we must realize that the choked mass
impossible
A in an isentropic flow. Also, for each value of area flow rate m̊max is maximum only for a given stagnation state
ratio A∗ > 1 , there correspond two values of exit Mach (p01 and T01 ) and remains constant till the reservoir condi-
numbers: one being subsonic and other supersonic. Which tions are the same. However, by altering stagnation pressure
of these two values of (Me ) occurs depends on the operating and stagnation temperature, a different value of m̊max will be
back pressure (pb ), which is the controlling pressure to estab- obtained corresponding to the new stagnation state.
lish the flow. Further, in order to establish the flow through The phenomena of throat choking can be physically
the duct, the pressure at the exit must be lower than the pres- explained as follows. Let us consider a convergent–divergent
sure at the nozzle entry, i.e., pp01e < 1. Let us assume that the nozzle which is kept in still environment where pe = p01 ,
exit pressure is reduced to a value, pe1 (< p0 ). Under this in the beginning. To start the nozzle, the exit pressure (pe )
situation, a small favorable pressure gradient is established is slightly reduced than the nozzle inlet pressure (p01 ) and
across the nozzle which causes the flow through the nozzle thus, fresh streamlines start entering into the nozzle. Further
at low subsonic speeds. The local Mach number will increase reduction in pe allows more and more streamlines entering
continuously through the convergent portion of the nozzle, the nozzle and mass flow rate m̊ increases. However,
reaching a maximum at the throat. In other words, the static continuous decrease in pe results in a state when no more
pressure will decrease continuously in the convergent portion fresh streamline is able to enter and the flow becomes
of the nozzle, reaching a minimum at the throat. If we fur- choked and obviously, it will be dictated by the minimum
ther reduce the exit pressure to pe2 , then the pressure gradient area location, i.e., throat. Because of choking, the flow
will be stronger, flow acceleration will be faster, and varia- condition downstream of the throat will not be able to pass
tion of Mach number and static pressure through the duct will the information upstream and thus, the flow is virtually
be larger than the previous case. Similarly, if the nozzle exit clogged (or frozen).
pressure is reduced continuously, at some value (pe3 ), the flow From the aforesaid discussion, it is now evident that for
will attain sonic state at the throat and for this case, At = A∗ . pe < pe3 the flow conditions downstream of the throat do not
Now the flow traveling at acoustic speed will expand further affect upstream. However, it rapidly accelerates in the noz-
in the divergent portion as supersonic flow if the nozzle exit zle divergence to supersonic Mach numbers and to obtain an
pressure is less that of the pressure at the throat, i.e., pe < pt isentropic flow, the exit pressure (pe ) must be reduced to the
, and will decelerate as subsonic flow if pe > pt , as shown specified value pe8 , i.e., for pe = pe8 (< pt ) shock-free super-
by the isentropic diffusion curve (b) in Fig. 9.17. Also, when sonic flow is obtained as depicted by the curve (a) in Fig. 9.17.
the sonic condition prevails at the throat, the Mach number Recall for pe = pe3 (> pt ) we would get another isentropic
at the throat becomes unity and hence, the flow properties flow curve (isentropic diffusion), which depicts the subsonic
at the throat and indeed, the flow throughout the convergent flow in the nozzle divergence. Further, for all other values of
A entry A
exit
At
Throat
Me
Convergent section Divergent section
pe1
p pe2 1
t1
p pe3
pe t2
(b)
p m < m* Isentropic diffusion pe4
01 2
p*
pe5
Normal pe6
m = m* (a) shock
3
pe7
4
pe8
Isentropic expansion to supersonic velocities pe9 5
1 Subsonic velocity at exit and Me < 1; isentropic flow in divergence
2 Sonic condition at throat, shock waves and non−isentropic flow in diverge
3 Supersonic flow in divergence with oblique shock and separated jet
4 Oblique shock at exit without jet separation
5 Expansion wave at exit and Me > 1
Fig. 9.17 Pressure distribution in a convergent–divergent nozzle under different operating conditions
exit pressure between pe3 and pe8 , a non-isentropic solution being discharged is known as back pressure (pb ). Basically,
results. That is, a normal shock appears in the nozzle diver- the change in exit pressure is caused due to back pressure
gence when pe < pe3 . Since the flow becomes subsonic in monitoring.
the downstream of a normal shock wave, the static pressure If we further reduce the back pressure such that
increases to pe4 at the exit. Further reduction in exit pressure pe8 < pb < pe6 , the flow inside the nozzle becomes fully
moves the shock toward the nozzle exit and hence, the static supersonic and isentropic; however, increase in flow pressure
pressure rises to pe5 . For pe = pe6 , the shock stands exactly to pe7 , resulting in equilibrium with pb , is taking place across
at the nozzle exit, where pe6 is the static pressure downstream an oblique shock wave attached to the exit, outside the
of the shock at the design Mach number MD of the nozzle. nozzle, as shown in Fig. 9.18a. In this condition, the nozzle
At this stage, we must realize the fact that the subsonic flow is said to be operating under overexpanded state. Conversely,
downstream of a normal shock is always correctly expanded if the nozzle operates with pb < pe8 , the nozzle is said to
and hence, the exit pressures will be equal to the back pres- be underexpanded. Here, the exit pressure is higher than the
sure, i.e., pe = pb . The ambient pressure at which the flow is back pressure and thus, the flow undergoes expansion after
9.9 Convergent–Divergent Nozzle (de Laval Nozzle) 225
A entry A exit
At
Divergent section
Gas stream
Convergent section Throat
Oblique shock
M=1
Flow bending towards
nozzle centerline
(a) Flow with oblique shock wave at the nozzle exit.
A entry A exit
At
Divergent section
Gas stream Expansion fans

Convergent section Throat
M=1
Flow turning away from

the expansion waves
(b) Flow with expansion fans at the nozzle exit.
Fig. 9.18 Overexpanded and underexpanded convergent–divergent nozzles
leaving the nozzle. These expansion waves relax the exit the reduction in thrust is observed. In divergent portion of
pressure to pe9 to establish the equilibrium with pb outside an overexpanded nozzle, the static pressure drops below
the nozzle (Fig. 9.18b). the back pressure and hence, the equilibrium of pressure
As discussed, when a nozzle discharges the maximum is established by compressing the gas through a series of
mass flow rate m̊max and the gas is expanded completely shock waves. Although this phenomenon has been studied
to the back pressure, a fully supersonic and isentropic flow by many investigators, however, it is yet not completely
results in nozzle divergence. The subsequent decrease of back understood. Let us consider
a convergent–divergent nozzle
pressure has no effect on the flow conditions inside the nozzle; having fixed area ratio AA∗ is being operated under correct
however, beyond the nozzle exit the supersonic flow expands expansion (pe8 = pb ) and the gas flow is fully supersonic
in the same manner as does a jet emanating from a converging and isentropic in nozzle divergence. If we raise pb above
nozzle working with a supercritical pressure ratio. Since the than pe8 , the increase of back pressure cannot propagate
gas is not completely expanded inside the nozzle, the kinetic itself upstream due to flow exiting the nozzle at supersonic
energy of the flow at the exit Aexit will be smaller than the Mach numbers. It can, however, propagate itself upstream
fully isentropic case. Consequently, the thrust produced by an through the boundary layer on the wall and surrounding the
underexpanded nozzle will be less than the nozzle operated flow. Since the fluid velocities within the boundary layer
under correctly expanded state. vary from supersonic levels at its interface with the outside
Similarly, when an optimally expanded convergent– main fluid to zero at the nozzle wall, there exist zones where
divergent nozzle is operated under overexpanded condition, the gas speed is subsonic. The acoustic signals in the form of
A entry A exit
At
Gas stream
Throat Oblique shock waves
M=1
Separated flow
Fig. 9.19 Schematic layout of an overexpanded convergent–divergent nozzle operating with flow separation
pressure waves can propagate upstream in the boundary layer Mach number but the values of pressure, temperature, and
through these subsonic regions. Further, if pb is marginally density decrease.
greater than pe8 , the oblique shock wave will be formed just When a supersonic flow is forced to change direction sud-
at
the
corner of the nozzle exit and the wall static pressure denly at a sharp concave corner, an attached oblique shock
pe will sharply increase from pe to pb . However, if pb is forms at the corner. The angle by which the flow turns at
increased to a level which is considerably larger than pe8 , the concave corner is called the turning or wedge angle (θ),
the flow will detach itself from the nozzle wall and the gas and the angle made by the oblique shock with the incoming
will be compressed through the oblique shocks increasing (upstream) streamlines is called the wave angle (β).
the static pressure to pb , as shown in Fig. 9.19. The static
pressure of the flow at the pointwhere separation occurs 9.10.1 Governing Equations of the Oblique Shock
is known as separation pressure psep . It is obligatory to
alleviate or minimize the flow separation especially in the
Consider an inviscid uniform flow with negligible body forces
rocket engines which flies at exceedingly high altitudes.
passing through an oblique shock, as shown in Fig. 9.20. Let
the conditions upstream and downstream of the shock waves
are designated by the subscripts “1” and “2”, respectively. In
9.10 Two-Dimensional Flow Across an
vector notations, the momentum equation, for the component
Oblique Shock Wave
of tangential velocity, can be written as
→
In the earlier sections on normal shocks, the thermodynamic
−
and kinematic changes that occurred when the flow traversed vt ρ−
→
v .d A = 0 (9.202)
a normal shock wave were studied. A normal shock is a spe- cs
cial form of the pressure discontinuity in a compressible fluid.
In general, the discontinuities observed in supersonic flows Applying above equation on a control surface consisting of
inclined to the freestream are termed as oblique shocks. It the shock wave and some amount of fluid on either side of the
occurs in a supersonic flow because of the continuous com- wave, we get
pression waves caused by a concave corner in the flow, tend-
ing to merge at a finite distance from the surface. This con- v2t (ρ2 v2n ) − v1t (ρ1 v1n ) = 0 (9.203)
cave corner where the flow is turned into itself leading to a
shock wave (compression front) is also known as compres- Substituting Eq. (9.205) into the above, we obtain
sion corner. Across an oblique shock wave, the Mach number
decreases but the pressure, temperature, and density increase. v1t = v2t (9.204)
Contrarily, the flow across a convex or expansion corner
turns away from itself causing an expansion wave (also called i.e., for a supersonic stream passing through an oblique
expansion fan). All the incoming streamlines while passing shock, the tangential component of velocity remains constant.
through the expansion fan are deflected to the same angle Because of this, the velocity of the flow across an oblique
θ, resulting in uniform parallel flow downstream of the fan. shock decreases solely due to decrease in its normal compo-
Across an expansion fan, the flow experiences an increase in nent (vn2 < vn1 ). Therefore, the inclination of the flow ahead
9.10 Two-Dimensional Flow across an Oblique Shock Wave 227
Normal
Oblique shock
p >p
2 1
v1t T2 > T1
v1n
β
v2 ρ>ρ
β 2 1
θ M 2> M 1
v1 v2t
v2n
p T ρ M p T ρ M
1 1 1 1 2 2 2 2
Ahead of shock Behind the shock
Fig. 9.20 Upstream and downstream conditions for the flow across an oblique shock
of the shock is always more than that behind the shock, i.e., shock will be
Mn2
the flow turns toward the shock wave after crossing the shock. M2 = (9.211)
Moreover, with respect to the normal component of velocity, sin (β − θ)
the flow across an oblique shock wave can be viewed as the
The introduction of M1 sinβ in place of M1 in normal shock
flow across a normal shock and hence, the governing equa-
relations given by Eqs. (9.151), (9.152), (9.161), (9.166), and
tions for an oblique shock wave may be obtained by replacing
(9.171) leads to the following equations for an oblique shock.
the flow velocities with their normal components in the nor-
mal shock relations. Thus, Static Pressure Ratio:
ρ1 v1n =ρ2 v2n p2 2γ 2 2

(9.205) =1 + M1 sin β − 1 (9.212)
p1 γ+1
p1 + ρ1 v1n
2
=p2 + ρ2 v2n
2
(9.206)
2
v1n v2 Static Temperature Ratio:
h1 + =h2 + 2n (9.207)
2 2 ⎧ ⎫
⎨ 1+ (γ−1) 2
T2 2 M1 sin β
2 2γ
(γ−1) M1 sin β
2 2 −1 ⎬
Since the superposition of the tangential component of veloc- =
T1 ⎩ (γ+1)2 2 ⎭
2(γ−1) M1 sin β
2
ity on the flow field across the normal shock does not affect
the flow properties, relations between the flow properties (9.213)
upstream and downstream of the flow field through the oblique
shock can be obtained from the normal shock relations given
Stagnation Pressure Ratio:
in Sect. 9.7.3.
The component of upstream Mach number (M1 ) normal
⎡ ⎤ γ
to the shock wave is p02 ⎣
(γ+1) 2
M1 sin2 β (γ−1)
2γ 2 2 (γ−1)
1
= 2 ⎦ 1+ M1 sin β − 1
p01 1 + (γ−1) M2 sin2 β
2 1
γ+1
Mn1 = M1 sinβ (9.208) (9.214)
Using Eq. (9.147), the normal component of Mach number

behind the shock is given by Density Ratio:

2 +2
(γ − 1) M1n ρ2 (γ + 1) M12 sin2 β
2
M2n = (9.209) = (9.215)
2 − (γ − 1)
2γM1n ρ1 (γ − 1) M12 sin2 β + 2
Substituting Eq. (9.208) into the above equation, we have Entropy Change:

(γ − 1) M12 sin2 β + 2 s2 − s1 γ 2γ
= ln
2
M2n = (9.210) R γ−1 (γ + 1) M12 sin2 β
2γM12 sin2 β − (γ − 1)
(γ − 1) 1 2γ (γ − 1)
+ + ln M12 sin2 β −
(γ + 1) 1−γ (γ + 1) (γ + 1)
From the geometry of the flow field across the oblique shock, (9.216)
depicted in Fig. 9.20, the Mach number downstream of the
9.10.2 Minimum and Maximum Wave Angles Introducing Eqs. (9.215) into (9.224), we obtain
From the normal shock analysis, it is evident that a shock can tan (β − θ) (γ + 1) M12 sin2 β
= (9.225)
only occur when the incoming freestream Mach number is tan β (γ − 1) M12 sin2 β + 2
greater than one. The same is valid for an oblique shock wave;
the normal component of freestream Mach number (M1n ) From some trigonometric manipulations, we get
should also be greater than one. That is,
M12 sin2 β − 1
M1n = M1 sinβ> 1 (9.217) tan θ =2 cot β (9.226)
M12 (γ + cos 2β) + 2
Thus, the minimum wave angle (βmin ) for a given M1 can be Equation (9.226) is a well-known relation between the flow
found from deflection angle (θ), the wave angle (β) and the upstream
Mach number (M). A schematic representation of θ − β − M
M1 sinβmin = 1 (9.218) relation is shown in Fig. 9.21 from which the following obser-
vations can be deduced.
or
• For a given Mach number, there exists a maximum deflec-
1
βmin = sin−1 (9.219) tion angle, θmax . Therefore, at a given M1 , if θ > θmax , then
M1 no solution in the form of an oblique shock wave attached
to the body is possible; rather, the oblique shock will be
Note that the minimum oblique shock wave angle βmin for a
curved and detached.
given M1 is indeed same as the Mach angle (μ) (Eq. (9.29))
• If θ < θmax then, for a given θ and M, there will be two
formed by an isentropic pressure wave moving at M1 > 1.
possible values of β. The larger value of β corresponds to
That is, the oblique shock wave at minimum wave angle to
the strong shock solution, while the smaller value refers to
the incoming flow degenerates to an isentropic shock wave.
the weak shock solution. For the strong shock solution, the
The maximum oblique shock wave angle (βmax ) for a given
flow downstream of the shock becomes subsonic, whereas
M1 is 90o , which is again a limiting case referring to a normal
for weak shock solution the flow behind the shock may
shock. Thus, the oblique shock wave angle (β) varies as
remain supersonic.
• If θ = 0, then β =
1 2 or β = μm . The former case refers
sin−1 < β< 90o (9.220) to a normal shock, while the latter shows the limiting value
M1
of β, i.e., the shock vanishes and only Mach waves prevail
in the flow field.
9.10.3 θ − β − M Relation Another useful form of Eq. (9.225) can be obtained by rear-
ranging the terms
From the geometry of oblique shock flow field, shown in
Fig. 9.20, we write 1 γ + 1 tan (β − θ) γ − 1
v1n = − (9.227)
tan β = (9.221) M12 sin2 β 2 tan β 2
v1t
and Using trigonometric identities, we can reduce above as

v2n
tan (β − θ)= (9.222) γ + 1 2 sin β sin θ
v2t M12 sin2 β − 1= M1 (9.228)
2 cos (β − θ)
Dividing Eqs. (9.222) by (9.221) and using Eq. (9.204), we However, for small values of flow turning angles (θ → θ)
obtain especially at high Mach numbers, sin θ ≈ θ and
tan (β − θ) v2n
= (9.223) cos (β − θ) ≈ cos β. Thus, the above equation becomes
tan β v1n

γ+1 2
From Eq. (9.205), the above equation can be written as M12 sin2 β − 1≈ M1 tan β θ (9.229)
2
tan (β − θ) ρ1
= (9.224) In the limiting case of very high Mach numbers, i.e., M → ∞,
tan β ρ2
Fig. 9.21 suggests very low values of β, i.e., β 1. Therefore,
9.10 Two-Dimensional Flow across an Oblique Shock Wave 229
90
Strong shocks
80
θ = θ max
M <1
70 2
M =1
M >1 2
60 2
1.2 1.4 1.6 2 3 10 M = infinity
1
50
Shock Wave
angle (β) 40
Weak shocks
30
20
10
0 5 10 15 20 25 30 35 40 45 50
Deflection angle (θ)
Fig. 9.21 Schematic representation of θ − β − M relation
the above equation can be further reduced to an interesting But rearranging the terms in Eq. (9.212), we write
form
p2 − p1 2γ 2 2
γ+1 = M1 sin β − 1 (9.233)
β= θ (9.230) p1 γ+1
2
Equation (9.230) seems to be an excellent approximation even Introducing above equation into Eq. (9.231), we get
for moderate flow turning angles.
p γM12
≈ θ (9.234)
9.10.4 Weak Oblique Shocks p1 M12 − 1
We know that the compression of supersonic flow across a Thus,

Mach wave is isentropic. The purpose of this section is to
demonstrate that the flow passing through an oblique shock, p
under the assumptions of small flow turning angle θ and super- Strength of shock = =∝ θ (9.235)
p1
sonic flow (M2 > 1) downstream of the shock, can be nearly
isentropic. This type of oblique shocks is termed as weak in
strength. For a weak oblique shock, M2 > 1 even though it is Equation (9.234) reveals that the shock strength p2p−p 1
1
is
less than M1 , i.e., the Mach number upstream of the shock. proportional to the flow turning angle (θ). From the similar
Therefore, a weak shock wave across which the decrease in exercise as above, it can be shown that the changes in density
flow Mach number (M1 − M2 ) and the flow deflection angle and temperature are also proportional to θ. That is,
θ are small can be approximated as the Mach wave. Thus,
by replacing the wave angle β with the Mach angle μ in ρ
∝ θ (9.236)
Eq. (9.229) and by using Eq. (9.29), we get ρ
⎡ ⎤

γ + 1 M 2
and
M12 sin2 β − 1≈ ⎣ 1 ⎦ θ (9.231)
2 M2 − 1
1 T
∝ θ (9.237)
T
Note that the strength of a shock is defined as
Let us now examine the effect of small flow deflection angles
p p2 − p1
= (9.232) (θ ≈ θ) on the entropy change across the shock wave. Recall
p1 p1 Eq. (9.216)

s2 − s1 γ 2γ (γ − 1) or
= ln +
R γ−1 (γ + 1) M12 sin2 β (γ + 1)
s ≈ θ (θ)2 (9.245)
1 2γ (γ − 1)
+ ln M12 sin2 β −
1−γ (γ + 1) (γ + 1) Therefore, if a large number of weak shocks cause the com-
pression, the entropy increase is reduced tremendously com-
But for weak oblique shocks
pared to a single compression front for the same total turning
angle (θ). By making θ smaller and smaller, a smooth turn
M12 sin2 β − 1≈ M12 − 1 (9.238) with θ → 0 is produced in the limit; the entropy increase
becomes zero and the compression becomes nearly isentropic
Using above approximation, Eq. (9.216) can be written as (Fig. 9.22).
+ Note that the above discussion considers the flow near the
s2 − s 1 2γ 2 (γ−1)
1
wall region in a supersonic flow field. At sufficiently far away
= ln 1+ M1 − 1
R (γ + 1) from the wall, the Mach lines are merged with each other
γ , forming an oblique shock, and hence the flow is no longer
2γ
− γ−1 γ−1 2 γ−1
isentropic farther away from the wall.
× M M1 − 1 + 1 (9.239)
γ+1
Expanding the terms

in the above equation using the log- 9.11 The Prandtl–Meyer Expansion Fan
2 3
arithmic series log (1 + ) = − + · · · · · · · · · and
2 3 In earlier sections, we have seen that an oblique shock wave
retaining the term up to first order, above equation simplifies is generated whenever a supersonic flow is turned into itself.
to Contrarily, if a supersonic flow is allowed to move away from
2 3 itself, another category of wave is produced which is known as
s 2γ M1 − 1 expansion wave or expansion fan, as schematically shown in
≈ (9.240)
R (γ + 1)2 3 Fig. 9.23. The purpose of this section is to develop a method-
ology to calculate the changes in the flow properties across
Introducing Eq. (9.234), into the above equation, we get an expansion wave.
3 Note that the expansion fan is a continuous expansion
s (γ + 1) p zone where a supersonic flow passes through an infinite num-
≈ (9.241)
R 12γ 2 p1 ber of Mach waves, each subtending the Mach angle μ with
the freestream in the direction of the flow. As depicted in
or Fig. 9.23, the series of expansion fan is bounded upstream by
a Mach wave which makes the Mach angle μm1 with respect
s≈ (θ)3 (9.242)
to the oncoming flow direction, where μm1 = sin−1 M11 .
At downstream end, the expansion fan is boundedby another

9.10.5 Isentropic Compression in Supersonic Mach wave which makes the angle μm2 = sin−1 M12 with
Flow by Turning respect to the incoming freestream in the direction of the flow.
The expansion fan produced at a sharp convex corner, as
In a supersonic flow, if the flow deflection angle (θ) is divided shown in Fig. 9.23, is known as centered expansion fan. Lud-
into n segments such that each segment causes the flow turning wig Prandtl and Theodor Meyer were the first who developed
by a small angle (θ), then the theory of centered expansion fan and thus these waves
are also called Prandtl–Meyer expansion fans. Some of the
θ = nθ (9.243) peculiar characteristics associated with these waves are given
below.
Thus, the total entropy change of the flow passing through
these segments will be • The streamlines passing through an expansion wave are
smooth curved lines.
s≈ n (θ)3 ≈ nθ (θ)2 (9.244) • Expansion fan is a continuous expansion region that con-
sists of an infinite number of Mach waves. Each of these
9.11 The Prandtl–Meyer Expansion Fan 231
Oblique shock
Slip line
Mach wave
M>1
Δθ θ
Δθ
Δθ
Δθ
Fig. 9.22 Supersonic flow around a smooth concave corner
Mach wave at
upstream end
Mach wave at
downstream end
Μ1 > 1
p1
T1
μ m1 M2 > M1
ρ1
μ m2
S p <p
2 1
θ T2 < T1
ρ <ρ
2 1
Fig. 9.23 Schematic diagram of a centered expansion fan
waves makes an angle with respect to the freestream in the dv

−dθ = M2 − 1 (9.247)
direction of flow. v
• There is no change in the entropy of the flow as expansion
Integrating both sides, we get
is isentropic for each Mach wave.
• The flow properties such as static pressure, static temper-
dv
ature, and static density decrease across an expansion fan −θ + constant = M2 − 1 (9.248)
v
while the Mach number increases.

-
It can be shown that the fractional change in flow speed dvv ,
The integral term M12 − 1 dv
v in the above equation is
across a wave of infinitesimal strength, is called the Prandtl–Meyer function (ν (M)), i.e.,
dv dθ
= −√ (9.246) dv
ν (M)= M2 − 1 = −θ + constant (9.249)
v M2 − 1 v
or
To solve the above integral, let us replace dv with dM. If a is If we assume θ = 0 at M = 1, then the constant in the above
the speed of sound, then flow velocity is written as equation will be zero. Thus, we get
v = aM (9.250) γ+1 γ−1 2

ν (M) = tan−1 M − 1 − tan−1 M2 − 1 (9.259)
γ−1 γ+1
Taking log both sides and differentiating
dv da dM Note that, for very high Mach numbers (M → ∞), which cor-
= + (9.251) responds to expanding the supersonic flow to zero pressure,
v a M
the Prandtl–Meyer function becomes maximum, i.e., νmax .
√ Thus, Eq. (9.259) gives
Since a = γRT, Eq. (9.36) can be written as

a02 γ−1 π γ+1
=1+ M2 (9.252) νmax = −1 (9.260)
a2 2 2 γ−1
or
For air (γ = 1.4)
−1
γ−1 2
a2 = a02 1 + M (9.253) νmax = 130.5◦ (9.261)
2
Differentiating Eq. (9.253) with respect to M, we get i.e., a flow which is initially at Mach 1 can turn 130.5◦ ,
whereas a stream which is initially traveling at Mach 2.5
−2 can turn only 90◦ . Thus, the higher the initial Mach num-
γ−1 γ−1
ada = − Ma02 1 + M2 dM ber, the lower its turning ability. Furthermore, Eq. (9.259) can
2 2
be applied between any two points within an expansion fan;
(9.254)
however, it is usual to relate the flow conditions upstream and
downstream of the wave. Thus, the relation between the flow
Dividing above equation by Eq. (9.253) and manipulating the
deflection angle (θ) and the upstream and downstream Mach
terms, we have
numbers will be

γ−1
M2
da dM
2 ν (M2 ) = ν (M1 ) + θ (9.262)
= − (9.255)
a γ−1
1+ 2 M M 2
Thus, the downstream Prandtl–Meyer function ν (M2 ) is
Introducing Eq. (9.255) into (9.251), we get obtained by adding the upstream Prandtl–Meyer function
ν (M1 ) and the flow turning angle θ. Since the expansion pro-
dv 1 dM cess is isentropic, the flow properties can be calculated using
= (9.256) isentropic relations.
v γ−1
1 + 2 M2 M
The reader should also note that the expansion process
through Mach waves is isentropic everywhere except at the
Thus, by substituting Eq. (9.256) into (9.249), the Prandtl– vertex S, shown in Fig. 9.23, where a large number of waves
Meyer function becomes unite, thereby, causing a large change in flow properties.
√
M2 − 1 dM
ν (M) = (9.257)
γ−1
1+ 2 M M 2 9.11.1 Isentropic Expansion in Supersonic Flow
by Turning
Solving above using integration by parts, we finally have
We have seen that the expansion of a supersonic flow through
a centered expansion fan, produced due to a sharp turning, is

M2 − 1 dM non-isentropic at the vertex. However, if the turning is made
ν (M) =
1 + γ−1
2 M 2 M gradual, the Mach waves will be segregated from each other,
as sketched in Fig. 9.24. In this condition, the expansion pro-
γ+1 γ−1 2
= tan−1 M − 1 − tan−1 M2 − 1 + constant cess is isentropic everywhere even at the wall surface due to
γ−1 γ+1
(9.258) gradual turning in the wall.
9.11 The Prandtl–Meyer Expansion Fan 233
Continuous Expansion Waves
Fig. 9.24 Supersonic flow around a smooth convex corner
However, as far as the changes in flow properties are con- The speed of sound is a property that varies from point
cerned, it does not matter whether the turn is sharp or smooth; to point and if there exists a large difference in the speeds
we still get same values unless we increase the maximum between the body and the compressible fluid surrounding it,
turning angle. This is because, the final properties are just the the compressibility of the fluid medium influences the flow
function of total deflection angle and a gradual turn simply around the body. Thus, both the inertial forces and elastic
shows that the expansion happens over an extended distance. forces due to fluid compressibility should be accounted in
the analysis. The ratio of inertial force to elastic force is a
nondimensional parameter, called the Mach number.
Summary von Karman proposed three rules of supersonic flows
which are applicable for small disturbances. These rules, how-
The fluid becomes compressible when it is subjected to a ever, can be to large disturbances but for qualitatively purposes
pressure field causing them to flow, i.e., the fluid will be com- only.
pressed or be expanded to some extent because of the pressure
acting on them. The time rate of change of velocity of the fluid • Rule of Forbidden Signals: The effect of pressure changes
elements in a given pressure gradient is a function of the fluid produced by a body, moving at a speed faster than the
density, whereas the degree of compression is determined by sound, cannot felt upstream of the body.
the isentropic bulk modulus of compression. The term com- • Zone of Action and Zone of Silence: A stationary point
pressible flows are defined as the variable density flow. The source in a supersonic stream produces effects only on
variations in fluid density for compressible flow require atten- the points that lie on or inside the Mach cone, extending
tion to density and other fluid property relationships. The fluid downstream from the point source.
equation of state, often unimportant for incompressible flows, • Rule of Concentrated Action: The proximity of circles rep-
is vital in the analysis of compressible fluids. resenting the various flow situations is a measure of the
The sound waves are the infinitesimal pressure distur- intensity of the pressure disturbance at each point in the
bances, and the speed at which these waves propagate in a flow field.
medium is known the speed of sound or acoustic speed. Fur-
The flow regimes can be classified based on the value of the
ther, the term compressible flow reflects the variation in den-
Mach number.
sity due to pressure change from one point to another in the
flow field. The change in density with respect to pressure has • For 0 < M < 1, the flow is termed as subsonic. In a
strong effects on the wave propagation. subsonic field, the presence of small disturbance, travel-
In turbomachines, the speed of the rotor should be in the ing with acoustic speed, will be felt throughout the flow
range of 270–450 ms−1 , to avoid the excessive stresses gen- domain. Thus, the subsonic flows are essentially “pre-
erated due to rotation. The studies also reveal that the loss in warned” to the disturbance.
efficiency mounts rapidly when the rotor speed approaches the • For 0.8 < M < 1.2, the flow is termed as transonic flow.
sonic velocity. Thus, for air compressors the limiting design • For M = 1, the flow is called sonic flow.
factor on rotational speed may be either stress or compressibil- • For M > 1, the flow is called supersonic flow. Since the
ity considerations. In hydrogen compressors, the fluid com- flow speed is above the speed of sound, they are no more
pressibility will never be a factor, whereas compressibility is “pre-warned”.
a major design factor for the compressor working with Freon- • For M > 5, the flow is called hypersonic flow.
22 as fluid.
If the rate of change of fluid properties normal to the stream- The variation of flow area A through the nozzle relative to
line direction is negligible as compared to the rate of change the throat area A∗ for the same mass flow rate and stagnation
along the streamlines, the flow can be assumed to be one- properties of a perfect gas is
dimensional. For flow in ducts, this means that all the fluid
(γ+1)
properties can be assumed to be uniform over any cross section A 1 2 (γ − 1) 2 (γ−1)
= 2 1+ M
of the duct. These properties which define the state of a sys- A∗ M (γ + 1) 2
tem are called static properties, and the properties at a state
which is achieved by decelerating the flow to rest through an This is known as area–Mach number relation.
isentropic means (i.e., reversible and adiabatic process) are The three reference speeds for studying the compressible
known as stagnation properties. flows are vmax corresponding to a given stagnation state, the
For the compressible flows, changes in enthalpy and the speed of sound at the stagnation temperature a0 , and the crit-
kinetic energy are much larger than that in elevation. Thus, ical speed v∗ . They are given as
between any two points, “1” and “2”, along a streamline the
specific static enthalpy (h) and fluid velocity (v) are related 2γRT0
vmax =
by γ−1
a0 = γRT0
v2 v2 1/2
h1 + 1 = h2 + 2 ∗ 2γ
2 2 v = RT0
γ+1
For isentropic flow of a perfect gas,
When the speed of sound is plotted as a function of the speed
γ γ of the flow for an adiabatic flow of a gas, it results an ellipse
p0 T0 γ−1 ρ0
= = known as adiabatic flow ellipse. It is given by the following
p T ρ
relation:
T0 γ−1 2
=1 + M
T 2 v2 a2
γ + =1
p0 γ − 1 2 γ−1
2
vmax a02
= 1+ M
p 2
1
ρ0 γ − 1 2 γ−1
= 1+ M
ρ 2 Exercises
The parameter M∗ is defined as the ratio of the local velocity Descriptive Type Questions
to the velocity of sound at the choked state (M = 1). It is
expressed as 1. An aircraft is flying at Mach 0.8 at an altitude of 15,000 m,
where the ambient pressure and temperature are 12.044
(γ+1) 2 kNm−2 and 216.65 K, respectively. Calculate the corre-
2 M
M∗2 = sponding pressure and temperature on the leading edge
(γ−1) 2
1+ 2 M of the wing where the freestream velocity relative to the
wing is negligible.
The mass flow rate through a streamtube of cross-sectional 2. Suppose a hot gas stream enters at Mach 0.35 to the tur-
area A is given by bine inlet of a jet engine where the temperature and pres-
sure are 1400 K and 150 kPa, respectively. Find the critical
γ+1
m̊ γ γ − 1 2 − 2(γ−1) temperature, critical pressure, and critical flow speed that
= p0 M 1+ M correspond to these conditions. Assume the gas properties
A RT0 2
are the same as those of air.
The maximum mass flow rate per unit area is given by 3. For an isentropic flow, establish the following relations.
− γ+1

2a2
m̊ m̊ γ γ+1 2(γ−1) (a) vmax = γ−1 + v2
= ∗ = p0
A A RT0 2
(b) a∗ = 2a +v
2 2 (γ−1)
max
γ+1
T = (1− γ)M v
dT 2 dv
i.e., for a given stagnation conditions, the maximum mass (c)
flow rate per unit area is directly proportional to √pT0 . (d) 2
vmax = γ−1
2
a02
0
Exercises 235
4. Show that for sonic flow the deviation between the com- 2. For the steady, one-dimensional isentropic flow of a per-
pressible and incompressible flow values of the pressure fect gas, the relation between the critical speed of sound
coefficients of a perfect gas (γ = 1.4) is about 27.5%. and the maximum speed (vmax ) is
5. In the test section of a supersonic wind tunnel, a pitot-
γ+1
static probe indicates a static pressure of 0.75 bar while (a) vamax
∗ = γ−1
the difference between the static and stagnation pressure
a∗
(b) vmax = γ+1
is 120 mm of mercury. Calculate the Mach number and γ−1
the velocity of airstream in the test section. (c) a∗ = γ−1
vmax
6. The air enters the diffuser at Mach 0.7, having the inlet ∗
γ+1
γ
area of 0.16 m2 . Assume the flow to be isentropic and the (d) vamax = γ−1
diffuser is operated at standard sea level conditions. The
flow velocity at the diffuser exit is 120 ms−1 . Determine 3. Consider the steady, one-dimensional isentropic flow of
(a) the mass flow rate, (b) the stagnation pressure and a calorically perfect gas (γ = 1.4) through a streamtube.
temperature at the exit, (c) the static pressure at the exit, If the static temperature at throat is 400 K, the stagnation
(d) the exit area. temperature of the flow is
7. An intermittent wind tunnel operated at Mach 2.5 by
(a) 380 K
expanding air at standard sea level conditions through
(b) 420 K
the test section into the vacuum. Assuming a pitot probe
(c) 480 K
is placed behind the normal shock in the test section. Cal-
(d) 800 K
culate the following conditions downstream of the shock.
4. The flow of a calorically perfect gas at Mach 3 encoun-
(a) Static pressure, density, and temperature.
ters an oblique shock wave. If the wave angle is 70o , the
(b) Stagnation pressure and stagnation temperature.
component of upstream Mach number normal to shock
(c) Mach number.
wave will be
8. In a supersonic intake, the air at Mach 2.3 is deflected by
(a) 1.18
an oblique shock with a wave angle of 18o . Calculate the
(b) 1.91
pressure ratio and the temperature ratio across the shock
(c) 2.67
wave. Also determine the flow deflection angle and the
(d) 2.92
downstream Mach number.
9. Consider a uniform flow of air at Mach 1.5 at the pressure 5. A convergent–divergent nozzle has the chamber tempera-
50 kPa and the temperature 345 K passes over a sharp ture 350 K and the chamber pressure 120 bar. The nozzle
concave corner. Downstream of an oblique shock of wave is operating at correct expansion and the exhaust is being
angle 60o at the corner, calculate (a) p2 , (b) T02 , (c) the discharged into the ambience at 1 bar. Assuming the noz-
flow turning angle (θ). zle flowto be isentropic, the static
10. Air at Mach 2.2 is being deflected isentropically by 6o in temperature at the exit
will be Cp = 1.2 KJkg−1 K−1 .
the clockwise direction. If the pressure and temperature
before the deflection are 100 kPa and 98o C, respectively, (a) 729 K
estimate the Mach number, pressure, temperature, and (b) 891 K
density of the deflected flow. (c) 1137 K
(d) 1532 K
Multiple Choice Questions 6. Consider an isentropic flow at Mach 0.5 in a streamtube.

Somewhere in the streamtube if the cross-sectional area
1. At very high Mach numbers (M → ∞) that corresponds is increased by 3% the change in flow density will be
to the flow of a perfect gas (γ = 1.4) expanding into the
vacuum, the maximum value of Prandtl–Meyer function (a) 1%
(νmax ) is (b) 2.5%
(c) 3.5%
(a) 90 (d) 5%
(b) 130.5
(c) 180 7. The inlet and exit areas of a turbojet engine operating
(d) 210.5 at Mach 0.7 are 0.8 m2 and 0.5 m2 , respectively. If the
freestream density is 0.5 kgm−3 , the mass flow rate of air respectively. The specific heat of the fluid at constant pres-
entering into the engine is sure will be
(a) 52 kgs−1 (a) 0.89 KJkg−1 K−1

(b) 64 kgs−1 (b) 1.02 KJkg−1 K−1
(c) 73 kgs−1 (c) 1.30 KJkg−1 K−1
(d) 84 kgs−1 (d) 1.52 KJkg−1 K−1
8. Consider a streamtube of constant cross-sectional area Keys

1.2 m2 , through which the air moves adiabatically at a
volume flow rate of 65 m3 s−1 . Assume one-dimensional 1. (b)
inviscid flow is fetched from a reservoir which has the 2. (a)
total temperature 600
K. The air temperature
inside the 3. (c)
streamtube will be R = 287 Jkg−1 K−1 . 4. (c)
5. (b)
(a) 321 K
6. (a)
(b) 363 K
7. (d)
(c) 411 K
8. (d)
(d) 452 K
9. (b)
9. In the flow of a calorically perfect gas, the static and stag- 10. (c)
nation temperatures are 260 K and 460 K, respectively.
The Mach number for this flow is
(a) 1.12 References

(b) 1.96
(c) 2.31 Houghton EL, Carpenter PW (1993) Aerodynamics for engineering stu-
dents, 4th edn. Edward Arnold (Publishers) Ltd., UK
(d) 3.19
Liepmann HW, Roshko A (1957) Elements of gas dynamics. Wiley, USA
Shapiro AH (1953) The dynamics and thermodynamics of compressible
10. In a fluid flow with velocity 510 ms−1 , the static and fluid flow, vol 1. Wiley, USA
stagnation temperatures are found to be 500 and 600 K,
Hypersonic Flows
10
It is generally well accepted that all those flows, having the thermal energy leads to high temperature rise than the flows
Mach numbers more than 5, are termed as hypersonic flow. at lower Mach numbers. This rise in temperature causes rel-
However, it is natural to ask, why this limit is set to Mach 5 atively large changes in other thermodynamic parameters.
only? why not at Mach 4 or even at Mach 6? One may also ask, For a hypersonic flow across the shock, the kinetic energy
how and in what respect the shock wave produced at hyper- is converted into the enthalpy. In turn, the temperature of
sonic Mach number is distinct than the shock produced in a the flow between the shock wave and the body becomes
supersonic flow? To answer these questions, we must have a substantial and the properties of air are profoundly altered.
thorough understanding of the characteristics associated with Indeed, the phenomena like vibrational excitation and dis-
a hypersonic flow regime. In this chapter, we have attempted sociation of the molecules of oxygen and nitrogen happen,
to describe these attributes which, in turn, represent a “formal i.e., the gas becomes imperfect. Since this dissociation pro-
definition” of the hypersonic flow. cess is endothermic in nature, the temperature of the flow gets
reduced as compared to the value that might be obtained if
the gas were perfect. The dissociation also shows a significant
10.1 Introduction effect on the pressure distribution over the body.
A flow is called hypersonic if the flow Mach number is greater

than 5, i.e., the flow speed is 5 times or more than the acoustic 10.1.1 Shock Layer
speed. Note that, it is just a commonly acknowledged refer-
ence point to categorize the lower speed limit of a hypersonic In earlier discussion on the oblique shock waves, we have
flow, and it is not the case that “flow actually changes its seen that for a fixed flow deflection angle (θ), the increase of
color” if accelerated from Mach 4.99 to Mach 5.01. In fact, a Mach number (M) eventually decreases the wave angle (β).
hypersonic flow is most appropriately delineated as that flow That is, when the flow Mach number is increased the shock
regime where certain physical characteristics become promi- comes closer to the body surface. The flow field between the
nent when the Mach number is increased beyond a level. In shock wave and the body surface is defined as the shock layer,
some cases, these phenomena might become important above as illustrated in Fig. 10.1. At very high Mach numbers, shock
Mach 4, albeit, in other occasions, they may be insignificant layer will be very thin and may be even close to the body
even at Mach 7 or higher. giving rise to a possibility of an interaction with the bound-
Alternatively, a hypersonic flow is described as the flow in ary layer on the surface. This phenomenon is referred to as
which fluid particles have a very high level of kinetic energy shock–boundary layer interaction, which is more pronounced
as compared to their thermal energy. For a fixed thermody- at low Reynolds number where boundary layer is compara-
namic state, the ratio of kinetic energy to the internal energy tively thicker. Notwithstanding, even at high Reynolds num-
in the freestream is directly proportional to the square of bers where the shock layer is almost inviscid, a thin shock
Mach number. For a perfect gas and with constant specific layer may prove to be advantageous in analyzing the flow
heats, this ratio is equal to 21 γ (γ − 1) M2 . We can see that field. In fact, the flow field with a thin shock layer can be
for airstream at M = 10, the kinetic energy is almost 30 times solved by using the Newtonian flow model, which is easy
greater than the thermal energy. Thus, when a hypersonic flow and often used in hypersonic aerodynamics for approximate
is decelerated, the conversion of particle’s kinetic energy into calculations.

238 10 Hypersonic Flows
M1 = 28
Shock layer
β
θ = 15 deg
β = 18 deg
Fig. 10.1 Thin shock layer at hypersonic Mach number
Shock layer
Entropy layer
Hypersonic Blunt nose

flow
Boundary layer
Shock wave
Fig. 10.2 Entropy layer formation over a blunt-nosed body in hypersonic flow
10.1.2 Entropy Layer 10.1.3 Viscous-Inviscid Flow Interaction
In order to avoid, the thin shock layer to be close to the body a Whenever the flow past a solid surface, a thin viscous layer
blunt-nosed aerodynamic body configuration is used at hyper- develops over the surface. At high Mach numbers, the hyper-
sonic Mach numbers. As obvious, a blunt-nosed configuration sonic stream possesses a large amount of kinetic energy, which
is typically a case where θ > θmax , consequently, there will gets retarded by viscous actions within the boundary layer. A
be a detached bow-shock standing at some distance from the portion of the lost kinetic energy is utilized in increasing the
nose of the body (Fig. 10.2). A bow-shock can be assumed internal energy of the gas which, in turn, increases the tem-
as the combination of a strong normal shock at the center perature of the boundary layer. This phenomenon is known as
and multiple oblique shocks at the periphery, which finally viscous dissipation. In turn, the rise in temperature has pro-
degenerates into Mach waves at infinity. Further, we know found effect on the boundary layer characteristics. Since the
that the entropy of the flow increases across the shock and coefficient of viscosity of a gas increases with temperature
higher the shock strength larger the entropy increase. Thus, and this, by itself, will make the boundary layer thicker. In
the entropy increase along the streamline passing through the addition, due to constant static pressure within the bound-
normal shock near the centerline of the flow is higher than the ary layer, an increase in temperature causes a decrease in
p
neighboring streamlines passing through the peripheral weak density, following the equation of state ρ = RT . Thus, to
oblique shocks. Hence, there exists a strong entropy gradient accommodate the same mass flow rate at reduced density, the
in the nose region of the body, which results in an entropy layer boundary layer thickness should be increased. Both of these
downstream that persists all along the body. From Crocco’s reasons are responsible for the rapid growth of the bound-
theorem, this entropy layer can be viewed as full of vorticity. ary layer at hypersonic speeds than at lower Mach numbers.
Since the boundary layer on the body surface grows inside Furthermore, in the boundary layer theory for incompressible
the entropy layer and hence the vorticity and boundary layer flows, it is assumed that the viscous effects are confined in
interactions are inevitable. a thin region near to the wall; and the outer inviscid flow is
10.1 Introduction 239
Inviscid flow is
weakly affected
Shock wave
Inviscid flow is
strongly affected
Outer edge of boundary layer
M a
y δ
Strong interaction Weak interaction
Fig. 10.3 The hypersonic boundary layer over a flat plate
ρw ue x
almost unaffected due to boundary layer. In contrast, because Rex = (10.2)
of the large boundary layer thickness at hypersonic Mach μw
numbers, the outer inviscid flow gets either partly or strongly
where ρw = density of the gas at wall temperature, μw =
affected by the thick boundary layer. This is because of the
coefficient of viscosity at wall temperature, and ue = flow
greater displacement effects caused by a hypersonic boundary
velocity within the boundary layer. Equation (10.2) shows that
layer that changes the characteristics of the outer inviscid flow
greatly; the changes in outer flow lead to further changes in the x
boundary layer growth. In this way, a feedback loop is formed δ∝ (10.3)
ρw u e x
between the boundary layer and the outer inviscid flow. This μw
interaction is commonly referred to as viscous-inviscid flow

or
interaction. This interaction can be of two types. In the first x μw
one, an exceptionally thick boundary layer grows over the δ∝√ (10.4)
ue x ρw
surface, it is called the pressure interaction. In other one, a
shock wave impinges on the boundary layer, due to which After some algebraic manipulation, we get
the boundary layer is subjected to an adverse pressure gra-

dient present across the shock. Consequently, the boundary x ρe μw
δ∝ (10.5)
layer flow is either retarded or even separated from the surface ρe u e x ρw μe
μe
depending upon the boundary layer thickness and the wave
strength.
where ρe = density of the gas at freestream temperature, and
The viscous-inviscid flow interaction modifies the surface
μe = coefficient of viscosity freestream temperature. But,
pressure distributions around the flying object which, in turn,
Rea = ρeμuee x ≡ Reynolds number based on freestream con-
affect the lift and drag on the object. The heat transfer and the
ditions. Thus, Eq. (10.5) is written as
skin friction are also increased by this interaction.

x ρe μw
δ∝√ (10.6)
10.1.3.1 Viscous-Inviscid Flow Interaction over a Rea ρw μe
Flat Plate
Consider a boundary layer on a flat plate in a hypersonic flow, ρe
The density ratio ρw can be evaluated using the perfect gas
as shown in Fig. 10.3. The thickness of the boundary layer (δ) equation as
at a distance x from the leading edge is defined by ρe pe Tw
= (10.7)
ρw pw Te
x
δ∝√ (10.1)
Rex Similar to an incompressible boundary layer, if the static pres-
sure within the hypersonic boundary layer is also assumed to
where Rex = Local Reynolds number, which is defined as be constant, so that pw = pe . Thus, Eq. (10.7) becomes
Rex = ρvxμ .
The local Reynolds number based on the wall temperature ρe Tw
(Tw ) can be written as = (10.8)
ρw Te
Let us further assume a linear dependence between the tem- ρ2 (γ + 1) M12 sin2 β
=
perature and the viscosity, we have ρ1 (γ − 1) M12 sin2 β + 2
μw Tw The above equation can be written as

= (10.9)
μe Te
ρ2 (γ + 1)
= (10.13)
Introducing Eqs. (10.8) and (10.9) into Eq. (10.6), we obtain ρ1 (γ − 1) + M2 sin
2
2β
1
x Tw
δ∝√ (10.10) In the limit of high Mach number, i.e., M 1, the above
Rea Te equation reduces to
Assuming an adiabatic wall and from the no-slip condi- ρ2 (γ + 1)
tion, the wall temperature (Tw ) can be approximated as the = (10.14)
ρ1 (γ − 1)
freestream stagnation temperature. Thus, Eq. (9.36) gives
In similar fashion, the exact relation for the pressure ratio
Tw γ−1 2 across the shock is given by Eq. (9.212)
=1+ Ma (10.11)
Te 2
p2 2γ 2 2
Using above equation, Eq. (10.10) shows that =1+ M1 sin β − 1
p1 γ+1
x
δ∝√ M2 (10.12) If M 1, then for small wave angle we have,
Rea a
M2 sin2 β − 1 ≈ M2 sin2 β 1. Thus,
That is, the boundary layer thickness (δ) varies as the square
p2 2γ
of Ma and thus, δ will be excessively large at hypersonic Mach = M12 sin2 β (10.15)
numbers. p1 γ+1

p
By using the perfect gas relation T = ρR , the temperature
ratio across the shock is written as
10.2 Oblique Shock Relations in Hypersonic
Flow p2
T2 p1
= ρ2 (10.16)
T1 ρ1
We know that whenever the supersonic stream turns into itself,
a shock wave is produced. Shock is an extremely thin region
Introducing Eqs. (10.15) and (10.14) into the above equation,
which has the thickness of the order of 10−5 cm, where the vis-
we get
cosity and thermal conductivity are the important mechanism
making the shock process irreversible. Because of this irre-
2γ
versibility associated with the shock wave, stagnation pres- T2 γ+1 M12 sin2 β
= (γ+1)
sure across the shock decreases with increase of the Mach T1
(γ−1)
number, while the static pressure, static density, and static
temperature rise. The hypersonic shock wave remains sta- or
tionary if the static pressure downstream of the shock is suf-
ficiently high. T2 2γ (γ − 1) 2 2
= M1 sin β (10.17)
The relations obtained for the straight oblique shock T1 (γ + 1)2
waves at supersonic speeds are still valid at moderately
high hypersonic Mach numbers. But, for very high Mach In addition, the relationship between the wave angle β and
numbers (M → ∞), these exact relations eventually take the flow deflection θ is simplified. Recall, θ − β − M relation
some interesting and approximate forms. We will examine from Eq. (9.226)
these results now.
Consider the flow through an oblique shock, as illustrated M12 sin2 β − 1
tan θ = 2 cot β
in Fig. 10.4. Let the upstream and downstream conditions be M12 (γ + cos 2β) + 2
denoted by subscripts “1” and “2”, respectively. Also, note
that we will discuss only the fluid dynamics effects of Mach This relationship is plotted in Fig. 10.5. Note that, for small
numbers ignoring viscous and real gas effects. Rewriting the flow turning angle, the change in wave angle becomes smaller
density ratio for a perfect gas from Eq. (9.215) and smaller with increase of Mach number. For small angles,
10.2 Oblique Shock Relations in Hypersonic Flow 241
ρ ρ 2
1
v 1t v 1n
p v2 p 2
1
M 1 = v1 /a 1 θ M 2 = v2 /a 2
v1 v 2t
T1 v 2n T2
β
Fig. 10.4 Oblique shock wave in the hypersonic flow
Μ = 1.5
Μ=2
Μ=3
M = 10
M=
Ο
θ = 10
Fig. 10.5 Shock waves at different Mach numbers
we write It is interesting to note that for small flow turning angles, in

hypersonic limit, the wave angle is only 20% larger than the
sin β ≈ β (10.18) flow deflection angle.
cos 2β ≈ 1 (10.19) In aerodynamics, it is usual to describe the pressure dis-
tribution in terms of pressure coefficient instead of pressure

tan θ ≈ θ (10.20)
itself. For incompressible flows, the pressure coefficient Cp
1 1
cot θ = ≈ (10.21) is given by
tan θ θ
p2 − p1
Using above approximations, Eq. (9.226) becomes Cp = (10.26)
q1
2 M12 β 2 where p1 and q1 , respectively, are the static and dynamic
θ= (10.22)
β M12 (γ + 1) + 2 pressures of the freestream. Note that the above equation is
a definition which is valid for all the flow regimes, incom-
Including high Mach number approximation, i.e., M1 1 pressible subsonic to hypersonic. Moreover, in compressible
flows, q1 = 21 ρv2 is given in terms of Mach number, i.e.
2 β2 q1 = 21 ρv2 = γ2 pM2 . Thus, Cp becomes
θ= (10.23)
β (γ + 1)

2 p2
Therefore, Cp = −1 (10.27)
γM12 p1
γ+1
β= θ (10.24) Substituting the exact relation for p2
from Eq. (9.212) into the
2 p1
above, we obtain
For air (γ = 1.4),
2 2γ 2 2
Cp = M1 sin β − 1 (10.28)
β = 1.2θ (10.25) γM12 γ+1
Again, for M 1 10.4 Expansion Wave Relations in

Hypersonic Flow
4
Cp = sin2 β (10.29)
γ+1 Consider the centered Prandtl–Meyer expansion fan created
at the convex corner, as depicted in Fig. 10.6. We know that an
expansion fan comprises of an infinite number of waves pro-
10.3 Mach Number Independence duced at the expansion corner and spread in the downstream
direction. Let the Mach number upstream and downstream of
In Sect. 10.1, we have discussed the physical characteristics the waves are M1 and M2 , respectively. Rewriting Eq. (9.262),
that make a hypersonic flow different than the supersonic flow. we have
In this section, we will examine this difference between these
flow regimes from the mathematical view point. θ = ν (M2 ) − ν (M1 ) (10.33)
Consider again an oblique shock in the hypersonic
freestream,
as sketched in Fig. 10.4. Note that, the pressure where ν is the Prandtl–Meyer function. Recall, Eq. (9.259)
ratio pp21 across the shock wave (Eq. (10.15)) increases

with Mach number, and in limiting case, i.e., if M → ∞ γ+1 γ−1 2
ν (M) = tan−1 M − 1 − tan−1 M2 − 1
then pp21 → ∞. Similar is the case with the temperature ratio γ−1 γ+1

T2
T1 , given by Eq. (10.17), where the temperature ratio √
In the limit of high Mach numbers, M2 − 1 ≈ M. Thus,
increases with increase of Mach number. Also, TT21 → ∞
the above equation reduces to
when M → ∞.
In contrast, the other relationships such as the density

ρ2 γ+1 γ−1
ratio ρ1 and the coefficient of pressure Cp , given by ν (M) = tan−1 M − tan−1 M (10.34)
γ−1 γ+1
Eqs. (10.14) and (10.29), respectively, are independent of the
Mach number at very high hypersonic speeds (M → ∞). If,
however, we make the shock downstream static pressure (p2 ), Using trigonometric relation,1 we write
nondimensional
by dividing it with the upstream dynamic
π 1
pressure q1 = 21 ρv12 = γ2 p1 M12 , i.e., tan−1 M = − tan−1 (10.35)
2 M

p2 p2 2 Also, from series expansion for inverse tan function,2 we have
p∗2 = = (10.30)
q1 p1 γM12
1 1 1 1 1 1
tan−1 = − + − + + ··· .
Introducing Eq. (10.15) into the above equation, M M 3M3 5M5 7M7 9M9
(10.36)

2 2γ
p∗2 = M12 sin2 β (10.31) Substituting Eq. (10.36) into Eq. (10.35), we get
γM12 γ+1

or γ+1 π γ+1 1 π 1
ν (M) = − + ··· − − + ···
γ−1 2 γ−1M 2 M
4
p∗2 = sin2 β (10.32) (10.37)
γ+1
Including high Mach number assumption, we can ignore
which becomes independent of Mach number. This for sure
the terms M13 , M15 and so on. Therefore, the above equation
demonstrates the Mach number independence principle;
becomes
at high Mach numbers some aerodynamic parameters, for

example, pressure coefficient (and henceforth the lift and drag
γ+1 π γ+1 1 π 1
coefficients), the density ratio, and the shock pattern end up ν (M) = − − −
γ−1 2 γ−1M 2 M
noticeably independent of Mach number.
(10.38)

1 tan−1 y + tan−1 π
1
= 2.
y
2 arctan 1
y = 1
y − 1
3y3
+ 1
5y5
− 1
7y7
+ 1
9y9
+ ··· .
10.4 Expansion Wave Relations in Hypersonic Flow 243
M1
μ
1
M2
μ2
Fig. 10.6 Sketch of supersonic flow at the convex corner
Simplifying above, we get At high Mach numbers (M 1), Eq. (10.43) becomes
2γ
p2 M1 γ−1
γ+1π γ+1 1 π 1 = (10.44)
ν (M) = − − + (10.39) p1 M2
γ−1 2 γ−1 M 2 M
Thus, Eq. (10.33) becomes Introducing Eq. (10.41) into (10.44), we get
γ
p2 γ−1 γ−1
γ+1π γ+1 1 π 1 γ+1π = 1− M1 θ (10.45)
θ = − − + − p1 2
γ−1 2 γ−1 M2 2 M2 γ−1 2

γ+1 1 π 1 Interestingly, Eq. (10.45) is equivalent to the hypersonic shock
+ + −
γ − 1 M1 2 M1 wave relation given by Eq. (10.15). Moreover, the pressure
coefficient is given by
or

2 p2
γ+1 1 1 1 1 Cp = −1 (10.46)
θ= − + − γM12 p1
γ−1 M1 M2 M2 M1
or Substituting Eq. (10.45) into the above equation, we obtain

γ
2 1 1 2 γ−1 γ−1
θ= − (10.40) Cp = 1− M1 θ −1 (10.47)
(γ − 1) M1 M2 γM12 2
This is the relation for Prandtl–Meyer expansion waves in
Again we can see that Eq. (10.47) is analogous to the pressure
hypersonic flow. It shows that the flow deflection angle θ is a
coefficient obtained for hypersonic shock wave (Eq. (10.28)).
positive quantity as M2 > M1 across an expansion fan.
Furthermore, Eq. (10.40) can be rearranged in another use-
ful form as 10.5 Hypersonic Similarity

M1 γ−1
=1− M1 θ (10.41) In aerodynamics, similarity refers to identifying the parame-
M2 2
ters or the set of parameters that result in identical flow fea-
In addition, the flow through an expansion fan is isentropic, tures at different conditions. This enables to use one set of
i.e., p02 = p01 , thus by using Eq. (9.37), we can write parameters obtained under specific experimental conditions
γ
in predicting the results at another set of conditions. In this
γ−1 2 γ−1 section, we will not derive the criteria for hypersonic similar-
p2/p02 p2 1+ 2 M1
= = γ (10.42) ity mathematically, instead we will focus our discussion on
p1/p01 p1 γ−1 2 γ−1
1+ justifying the use of similarity.
2 M2
Consider a slender body in the hypersonic freestream with
or γ velocity Ua at a low angle of attack, as shown in Fig. 10.7. If
γ−1 2 γ−1
p2 1+ 2 M1
u and v are the perturbation components of velocity in x- and
= γ (10.43) y-directions, then at some distance downstream of the shock,
p1 γ−1 2 γ−1
1+ 2 M2
we write
Shock Wave
Ua
θ
Slender Body
Fig. 10.7 A slender body in hypersonic freestream
u = Ua + u (10.48) 10.6 Newtonian Theory

v= v (10.49)
In 1687, Sir Issac Newton proposed a fluid flow modeling
For the flow over a slender body u v, and thus the flow method where the flow was assumed as the stream of non-
will be deflected by a small flow deflection angle (θ). Hence, interacting particles. Essentially, this concept hypothesized
from Fig. 10.7 fluid to be inviscid which gave inferior results when applied
to predict the ship-hull drag. However, this model gives sur-

v prisingly good results when applied to hypersonic problems.
sin = sin θ (10.50) The basic assumption in this theory was that each stream-
Ua + u
line of particles approaching a body will be deflected to par-
But, from small perturbation theory we know that, u Ua . allel to the surface. That is, there will be a complete loss of
Thus, the component of momentum normal to the body surface,
while the component of momentum parallel to the surface

v will remain constant. From Newton’s second law, we know
sin ≈ sin θ (10.51)
Ua that the time rate of change of momentum is equal to the force
exerted on the surface. The total loss of normal component
Further, assuming v Ua , and for small turning angle of momentum eventually results in the force normal to the
surface (Fig. 10.8).
v Consider a stream with velocity Ua impinging upon a sur-
≈θ (10.52)
Ua face of area A inclined at the angle θ to the incoming hyper-
sonic freestream. From the geometry as shown in Fig. 10.9,
or
we write
v The change in velocity normal to surface is
≈ Ma θ (10.53)
aa
= Ua sin θ − 0
where aa and Ma are the speed of sound and the freestream = Ua sin θ (10.55)
Mach number, respectively. Equation (10.53) is essentially
a measure of the perturbation with respect to the freestream The mass flow rate incident upon the surface of area A is
velocity. In hypersonic flows, this disturbance indicator
is defined as the hypersonic similarity parameter (K). = ρa AUa sin θ (10.56)
That is,
Thus, the time rate of change of momentum this fluid mass
K = Mθ (10.54) will be
Thus, if two different flow problems have same values of K = ρa AUa sin θ (Ua sin θ)
then they are similar flows and will have like solutions. = ρa AUa2 sin2 θ (10.57)
10.6 Newtonian Theory 245
The fluid flow model given by Sir Issac Newton (1687). The fluid particles lose their component of momentum
normal to the solid boundary and thereby, moves further along the boundary.
2θ
Uniform flow of discrete

set of fluid particles
Fig. 10.8 The streamlines in a hypersonic flow
Ua
Ma > 5
A sin θ
Fig. 10.9 Sketch for Newtonian impingement theory
But, from Newton’s second law of motion, the time rate of Note that, the
term
on the left side of equality is the pressure
change of momentum is equal to the force F exerted on a coefficient Cp . Therefore,
surface. Thus,
F = ρa AUa2 sin2 θ (10.58) Cp = 2 sin2 θ (10.61)
or This is the well known Newtonian sine-squared law for pres-

F
= ρa Ua2 sin2 θ (10.59) sure coefficient. Equation (10.61) shows that the pressure
A coefficient is proportional to the sine-squared of the flow
In reality, Newton has assumed a stream of noninteracting deflection angle. At large Mach numbers and moderately
particles in rectilinear motion impinging upon the surface. small turning angles, the shock wave angle is almost equal
This means that he has ignored the random motion of particles. to the deflection angle where the fluid particles eventually hit
But, from kinetic theory of gases we know that the static the surface without any prior warning or deflection. This con-
pressure, experienced by a surface, is solely due to random dition matches well with Eq. (10.61) and thus it is found to be
motion of particles. Thus, the force per unit area exerted on the useful in analyzing the flow at hypersonic Mach numbers.
flat surface, given by Eq. (10.59), can be taken as the pressure
difference above the freestream static pressure, i.e., 10.6.1 Lift and Drag Coefficients for a Flat Plate
at an Angle of Attack
p − pa = ρa Ua2 sin2 θ (10.60)
Let us examine the applicability of Eq. (10.61) on a flat plate
where pa is the freestream static pressure acting on the surface with chord length c at an angle of attack α with respect to the
and p is pressure experienced by the surface. Equation (10.60) freestream of velocity Ua , as shown in Fig. 10.10.
can be rewritten as In the absence of friction and since the surface pressure acts
normal to the plate, the resultant aerodynamic force R also acts
p − pa
= 2 sin2 θ perpendicular to the plate. Moreover, N can be resolved into
2 ρa Ua
1 2
its components L and D, which are called the lift and drag,
A Sin α U α
a A
Fig. 10.10 An inclined flat plate in the hypersonic flow
respectively. From Eq. (10.61), the pressure coefficient on the Substituting Eq. (10.66) into above,
lower surface of the plate is given by
CL = 2 sin2 α cos α (10.71)
Cp (lower) = 2 sin α 2
(10.62)
and
But, on the upper surface, there is no flow and hence the
pressure coefficient will be CD = 2 sin3 α (10.72)
Cp (upper) = 0 (10.63) Therefore, the lift-to-drag ratio becomes

L CL
We have seen that the normal force coefficient is given by = = cot α (10.73)
D CD
N
CN = (10.64)
qa Sr 10.6.1.1 Calculations of Maximum Lift Coefficient

To obtain the value of maximum lift coefficient CL,max , let
where Sr = c (1). Since the normal force is produced due to us differentiate Eq. (10.71) with respect to α.
the difference of pressures in lower and upper surfaces, thus
dCL
c = 2 sin2 α (− sin α) + 4 sin α cos2 α = 0
1 dα
CN = Cp (lower) − Cp (upper) dx (10.65)
c 0 or
where x is the distance measured from the leading edge. Intro- 2
sin2 α =
ducing Eqs. (10.62) and (10.63) into the above equation, we 3
get
Thus,
CN = 2 sin α2
(10.66) α ≈ 55o (10.74)
In addition, the lift and drag coefficients can be defined as

CL,max = 0.77 (10.75)
L
CL = (10.67)
qa Sr
10.6.1.2 Concluding Remarks on Newtonian
D
CD = (10.68) Sine-Squared Law
qa Sr The results predicted by using the Newtonian theory are appli-
cable to inviscid hypersonic flow over a flat plate. The vital
Furthermore, from the geometry, as shown in Fig. 10.10, we
observations can be summarized as follows:
deduce that
• In hypersonic aerodynamics, the pressure coefficient Cp is
CL = CN cos α (10.69)
predicted only by a local deflection angle θ and only by
CD = CN sin α (10.70) the surfaces directly in frontal flow conditions.
10.6 Newtonian Theory 247
L
• The lift-to-drag ratio D increases by decreasing the angle Therefore, for very high Mach numbers, the pressure coeffi-
of attack and becomes infinitely large when α is almost cient given by the modified Newtonian theory becomes
equal to zero. This result is logical because, for an inviscid ⎧ ⎫
γ
flow, there will not be any shear or viscous drag on the ⎨ (γ + 1)2 (γ−1)
4 ⎬ 2
surface. Cp = sin θ (10.82)
⎩ 4γ (γ + 1) ⎭
• The lift coefficient (CL ) increases with α and becomes
maximum at around α ≈ 55o , which is almost a practical
condition. For air (γ = 1.4), the coefficient of sin2 θ is equal to 1.839,
• For small angles of attack (α < 15o ), we observe the non- whereas for γ = 1, the value of the coefficient is 2. For high
linear variation of CL , which is in contrast to the subsonic Mach numbers and high geometric altitudes (e.g., M > 40 and
and supersonic flows. hG > 90 km), γ approaches a value of one that eventually
leads to the Newtonian sine-squared law. Furthermore, the
results obtained by modified Newtonian theory are suitable for
blunt body configuration, because this theory calculates the
10.6.2 Modified Newtonian Theory exact pressure at stagnation point, while the direct Newtonian
method is more suitable for the slender bodies such as wedged
For blunt bodies, Lees (1955) has put forward the following or cones.
modified form of Eq. (10.61).
Cp = Cp,max sin2 θ (10.76)

Summary
where Cp,max = Cp02 , i.e., the coefficient of pressure at the
stagnation point behind the shock wave. By definition, A flow is called hypersonic if the flow Mach number is greater
than 5, i.e., the flow speed is five times or more than the
p02 − pa 2 p02 acoustic speed.
Cp02 = = −1 (10.77) For a fixed flow turning angle, the increase of Mach num-
2 ρa Ua
1 2 γMa2 pa
ber eventually decreases the wave angle. That is, when the
Thus, flow Mach number is increased the shock comes closer to
the body surface. The flow field between the shock wave and
2 p02 the body surface is defined as the shock layer. At very high
Cp,max = −1 (10.78)
γMa2 pa Mach numbers, shock layer will be very thin and may be even
close to the body giving rise to a possibility of an interaction
But, from the Rayleigh pitot formula (Eq. (9.201)), with the boundary layer on the surface. This phenomenon
is referred to as shock–boundary layer interaction, which is
γ
2γ 2 1
p02 (γ + 1) 2 (γ−1) (γ−1) more pronounced at low Reynolds number where boundary
= Ma 1+ Ma − 1 layer is comparatively thicker.
pa 2 γ+1
(10.79) Whenever the flow past a solid surface, a thin viscous layer
develops over the surface. At high Mach numbers, the hyper-
Introducing the above equation into Eq. (10.78) and on rear- sonic stream possesses a large amount of kinetic energy, which
ranging the terms, Eq. (10.78) becomes gets retarded by viscous actions within the boundary layer. A
⎧ ⎫
portion of the lost kinetic energy is utilized in increasing the
γ
2 ⎨ (γ + 1)2 Ma2 (γ−1)
1 − γ + 2γMa2 ⎬ internal energy of the gas which, in turn, increases the tem-
Cp,max = Cp02 = −1
γMa2 ⎩ 4γMa2 − 2 (γ − 1) γ+1 ⎭ perature of the boundary layer. This phenomenon is known
as viscous dissipation.
(10.80)
The thickness of the boundary layer (δ) at a distance x
from the leading edge is defined by
Note that, Cp,max is a function of the specific heat ratio (γ) and
the Mach number (M). Including high Mach number (M 1) x
approximation, Cp,max becomes δ∝√
Rex
⎧ γ ⎫
⎨ (γ + 1)2 (γ−1)
4 ⎬ For a hypersonic boundary layer,
Cp,max = (10.81)
⎩ 4γ (γ + 1) ⎭ x
δ∝√ M2
Rea a

That is, the boundary layer thickness (δ) varies as the square 2 1 1
θ= −
of Ma and thus, δ will be excessively large at hypersonic Mach (γ − 1) M1 M2
numbers.
We know that whenever the supersonic stream turns into where θ is the flow turning angle and M1 and M2 are the
itself, a shock wave is produced. Shock is an extremely thin Mach numbers upstream and downstream of the expansion
region which has the thickness of the order of 10−5 cm, fan. Also, for the same assumption, the pressure ratio across
where the viscosity and thermal conductivity are the important the expansion fan will be
mechanism making the shock process irreversible. Because γ
of this irreversibility associated with the shock wave, stagna- p2 γ−1 γ−1
= 1− M1 θ
tion pressure across the shock decreases with increase in the p1 2
Mach number, while the static pressure, static density, and
static temperature rise. The hypersonic shock wave remains In hypersonic flows, the similarity parameter (K) is defined
stationary if the static pressure downstream of the shock is as
sufficiently high.
In the limit of high Mach number, i.e., M 1, the den- K = Mθ
sity ratio across the shock produced at hypersonic Mach
number is Thus, if two different flow problems have same values of K
then they are similar flows and will have like solutions.
ρ2 (γ + 1) In this chapter, we have discussed two hypersonic local sur-
=
ρ1 (γ − 1) face inclination methods: Newtonian and modified Newtonian
theories. They are used to predict the local surface pressure as
and the pressure ratio is a function of the surface inclination angle with respect to the
incoming freestream direction. For high Mach numbers, the
p2 2γ
= M12 sin2 β pressure coefficient predicted by direct Newtonian method is
p1 γ+1
Cp = 2 sin2 θ
In addition, the temperature ratio across the shock is
and by the modified Newtonian approach is
T2 2γ (γ − 1) 2 2
= M1 sin β
T1 (γ + 1)2 Cp = Cp,max sin2 θ
Together with high Mach number (M1 1) and small angle where
approximations, the relation between shock wave angle and ⎧ γ ⎫
flow turning angle is ⎨ (γ + 1)2 (γ−1)
4 ⎬
Cp,max =
⎩ 4γ (γ + 1) ⎭
γ+1
β= θ
2
For air (γ = 1.4),

Exercises
β = 1.2θ
Thus, at hypersonic Mach numbers for small flow turning
angles, the wave angle is just 20% larger than the deflection 1. Using Newtonian impact theory, demonstrate that, at
angle. hypersonic speeds, the stagnation pressure is approxi-
mately equal to two times the dynamic pressure.
For hypersonic flows, the pressure coefficient Cp is
2. Consider a flat plate in the airstream at Mach 20 at 35o to
4 the freestream direction, as sketched in Fig. 10.10. If the
Cp = sin2 β temperature of the freestream is 220 K, then calculate the
γ+1
Mach number and static temperature downstream of the
For expansion waves at high but finite Mach numbers, we shock.
have 3. Consider a flat plate in hypersonic airstream at 35o to the
freestream direction. Utilizing Newtonian impact theory,
Exercises 249
calculate (a) the pressure coefficient, (b) the lift coeffi- 2. Consider a flat plate placed in hypersonic airstream
cient, (c) the drag coefficient. (γ = 1.2) at an inclination of 20o . According to modified
4. Consider a sphere of diameter 0.5 m flying at a speed of Newtonian theory, the maximum pressure coefficient
6 kms−1 at an altitude of 65 km above the sea level. Cal- will be
culate (a) the freestream Mach number and the Reynolds
number, (b) Assuming air to be in the thermodynamic (a) 0.8
equilibrium, determine the static pressure (p2 ) and static (b) 1
temperature (T2 ) downstream of the shock positioned (c) 1.2
ahead of the sphere. (d) 1.9
5. Consider a flat plate in uniform hypersonic airstream at an
angle of attack α. Using Newtonian flow theory, find the 3. An airstream (γ = 1.4) turns around an expansion corner
value of α for which the lift coefficient (CL ) is maximum. such that the upstream and downstream Mach numbers
Also, find the expressions for drag coefficient (CD ) and are 6 and 9, respectively. The flow turning angle is

lift-to-drag ratio CCDL when CL = CL,max . (a) 0.22o
6. Consider a blunt-nosed body flying in the air sufficiently (b) 0.28o
above the sea level. Let the entropy increase along the (c) 0.32o
stagnation streamline is 0.7 KJkg−1 K−1 . Find (a) the (d) 0.38o
freestream Mach number, (b) the strength of shock in the
proximity of stagnation streamline. 4. In the limit of high Mach number (M 1), the density
7. Consider a hypersonic vehicle flying at Mach 22 at an ratio of a gas (γ = 1.2) flow across a shock wave is
altitude of 65 km above the sea level. What will the air
temperature at the stagnation point on the forward end of (a) 3
the vehicle? Explain the accuracy of your answer. Can it (b) 5
be changed? Under what conditions? (c) 7
8. Utilizing Newtonian impact theory, develop an expression (d) 11
for the drag on a sphere. If a sphere of diameter 40 cm is
moving at Mach 8 through the air at an ambient pressure 5. For high Mach number and small angle approximations,
of 0.05 kPa, calculate the drag on the sphere. the relation between shock angle and the flow turning
9. In a hypersonic wind tunnel, the test section is designed angle is
to operate at Mach 25. If the stagnation temperature in
the settling chamber is 3500 K, calculate the temperature (a) β = γ+1 θ
2
in the test section. Also find the required minimum set- (b) β = γ−1 θ
tling chamber temperature so that the condensation can 2
γ+1
be avoided in the test section. (Note: The liquefaction (c) θ = 2 β

temperature of air is about 77 K.) (d) θ = γ−1 β
If a sphere 2.0 m in diameter is flying at 6.5 kms−1 at
2
10.
an altitude of 65,000 m above the sea level. Calculate 6. Airstream at Mach 7, encounters a shock with the wave
(a) the freestream Mach number Ma , (b) the freestream angle of 15o . If the static pressure upstream of the shock
Reynolds number Rea , (c) Assuming the air to be in ther- wave is 50 kPa, the static pressure downstream of the
modynamic equilibrium, find the values of static pressure shock will be
(p2 ), static temperature (T2 ), stagnation pressure (p02 ),
and stagnation temperature (T02 ) downstream of the (a) 155 kPa
bow-shock formed in front of the flying sphere. (b) 167 kPa
(c) 186 kPa
(d) 198 kPa
7. For the hypersonic flow over a flat plate at 10o , the pres-
1. The hypersonic similarity parameter (K) is defined as sure coefficient predicted by direct Newtonian method is
(a) M + θ (a) 0.028

(b) M − θ (b) 0.048
(c) Mθ (c) 0.068
(d) M
θ (d) 0.088
8. A gas (γ = 1.3) flow at Mach 6 turns around the expan- Keys

sion corner of turning angle 0.3o . The pressure ratio across
the expansion fan is
1. (c)
(a) 0.10 2. (d)
(b) 0.15 3. (b)
(c) 0.25 4. (d)
(d) 0.30 5. (a)
6. (a)
9. The temperature upstream of a shock wave in Mach 6.5 7. (b)
airstream is 500 K. If the wave angle is 20o , the air tem- 8. (c)
perature downstream of the shock is 9. (d)
10. (b)
(a) 330 K
(b) 368 K
(c) 381 K
(d) 392 K
Reference
10. Consider a hypersonic flow over a flat plate of length
2 m at zero angle of attack. If the Reynolds number is Lees L (1955) Hypersonic flow. In: Fifth international aeronautical con-
increased, the boundary layer thickness at a distance 1 m ference. Los Angeles, USA
from the leading edge will
(a) increase
(b) decrease
(c) remain same
(d) become infinitely large
Boundary Layers
11
This chapter presents the boundary layer hypothesis and sur- the surface. Initially, this shearing action occurs only at the
veyed its results. We have described the mathematical simpli- body surface and retards the layer of fluid close to the surface,
fications which allows to obtain the reduced form of the full causing the fluid elements present in this layer to come to rest.
governing equations of motion for a viscous fluid. Several These fluid elements subsequently interact with the elements
exact and approximate solutions for steady laminar bound- in the layer above them and retard their motion. In this way, as
ary layer flows are talked about at a length. What’s more, the the fluid near to the surface passes downstream, the retarded
physics behind the boundary layer transition and separation action penetrates farther away from the surface and the layer
are likewise inspected. of retarded fluid thickens up (Fig. 11.1).
11.1 Introduction
11.2 Boundary Layer Thickness
In several flow situations, the effects of viscosity can be
neglected everywhere except in a thin region adjacent to the The extent of viscous effects on a surface is measured in terms
solid surfaces. This thin region where the viscous action pre- of boundary layer thickness (δ), defined as the distance normal
dominates is known as the boundary layer. The boundary layer to the surface such that the local flow velocity (u) at that
theory was first proposed by Ludwig Prandtl in 1904. This distance is 99% of the freestream velocity (Ua ). That is
concept gave rise to a model of the flow past solid bound-
aries in which the pressure distribution was initially calcu- at y = 0; u (y) = 0 (11.1)
lated without accounting viscosity and heat conduction, and at y = δ; u (y) = 0.99Ua (11.2)
subsequently, the viscous effects at the wall were computed
using the boundary layer theory. Except for the flow situations
in which either the boundary layer separates or subjected to
the large pressure gradients, the inviscid (or potential) flow
external to the boundary layer is substantially independent of 11.2.1 Displacement Thickness
the boundary layer. But due to boundary layer, the potential
flow streamlines are displaced outward from the wall giv- The presence of boundary layer over the surface causes a
ing rise to displacement thickness. Also, the flow in viscous deficit in the mass flow rate. The decrease in mass flow is
layer depends strongly on the longitudinal pressure distribu- estimated in terms of a more useful parameter known as the
tion established by the potential flow. displacement thickness (δ ∗ ) , defined as the distance by which
Moreover, the fluid sticks to a solid surface due to vis- a surface would have to be displaced outwards in a hypothet-
cosity so that the velocity of fluid layer adjacent to the sur- ical inviscid flow so as to maintain the same mass flux as in
face is equal to the velocity of the surface itself. In other the real flow. That is, the displacement thickness is defined
words, the relative velocity of the fluid with respect to a solid as the distance perpendicular to the boundary, by which the
surface is zero. This condition is commonly referred to as freestream is displaced due to the formation of boundary layer.
“no-slip” condition. Whenever the flow past a surface, due to It essentially modifies the shape of a body immersed in a fluid
large velocity gradients between the wall and the freestream, to allow to use an inviscid solution for practical engineering
large shearing gradients and the stress develop adjacent to applications.

252 11 Boundary Layers
Ua
Ua
y δ
u(y)
x
Fig. 11.1 Schematic diagram depicting fluid flow over a flat plate

Due to the presence of boundary layer, the flow past a given To obtain the total reduction in mass flow rate m̊t through
point on the surface is reduced by a volume equivalent to the BC, we must integrate the above equation from 0 to δ. Thus,
area ABC, as shown in Fig. 11.2. The reduction in volume
is given by the integral (Ua − u) dy. If the area under the δ
curve ABC is equated to the area under the curve ABED, m̊t = ρb (Ua − u) dy (11.7)
whose volume may be calculated as δ ∗ Ua . That is 0
δ or
δ ∗ Ua = (Ua − u) dy δ
0
m̊t = ρb (Ua − u) dy (11.8)
or 0
δ u

But m̊t will also be equal to the reduction in mass flow rate if
δ∗ = 1− dy (11.3)
0 Ua the plate is moved by δ ∗ in a freestream of velocity Ua . We
get,
Thus, the displacement thickness for the boundary layer may δ
be defined as the distance the surface would have to move in ρb (Ua − u) dy = ρbUa δ ∗ (11.9)
y-direction to reduce the flow passing by a volume equivalent
0
to the real effect of the boundary layer.
or
Aliter δ
Consider a section “1–1” in the boundary layer of thickness (Ua − u) dy = Ua δ ∗ (11.10)
(δ) over the flat plate, as shown in Fig. 11.2. In the velocity 0
profile, let us choose a small elemental streamtube of width
b (in z-direction) and thickness dy at a distance y from the or
δ
surface of the plate. Assume Ua is the freestream velocity ∗ u
δ = 1− dy (11.11)
and u is the local velocity of the fluid passing through the Ua
elemental strip. 0
The mass flow rate passing through the elemental stream-

tube of cross-sectional area dA (= bdy) will be given as
11.2.2 Momentum Thickness
m̊s = ρubdy (11.4)
Momentum thickness (θ) is defined as the distance through
However, in absence of plate, the mass flow rate (through which the boundary layer must be displaced to compensate
the elemental streamtube) would have been, m̊ s = ρUa bdy. the reduction in momentum of the flowing fluid due to bound-
Thus, the reduction in mass flow rate through the elemental ary layer formation. In other words, momentum thickness is
streamtube will be defined as the thickness of a layer of fluid of velocity Ua for
which the momentum flux is equal to the deficit of momentum
m̊s = m̊ s − m̊s = ρUa bdy − ρubdy (11.5) flux through the boundary layer.
Consider again the small streamtube of thickness dy within
or the boundary layer, as shown in Fig. 11.2. The mass flow rate
m̊s = ρb (Ua − u) dy (11.6) through the elemental streamtube is (recall, Eq. (11.4))
11.2 Boundary Layer Thickness 253
1
C
Boundary Layer
Ua
dy
y
y
D E
δ*
x
A B
z x
1’
Fig. 11.2 Displacement thickness
m̊s = ρubdy The adverse pressure gradient is found to be stronger at higher

values of H. Also, the high value of adverse pressure gradi-
Thus, the rate of momentum deficit with respect to freestream ent reduces the critical Reynolds number at which laminar
is to turbulent flow transition occurs. Furthermore, for Blasius
velocity profile H = 2.59, which is typical for laminar flows
Ps = ρu (Ua − u) dy (11.12) while, H = 1.3–1.4 is typical for turbulent flows.
Moreover, the rate of momentum deficit for the thickness θ is

given by ρa Ua2 θ. Thus, for compressible flow, we get 11.2.3 Kinetic Energy Thickness
δ Kinetic energy thickness (δ ∗∗ ) is the distance measured per-

ρa Ua2 θ = ρu (Ua − u) dy (11.13) pendicular to the surface of the solid body through which the
0
boundary should be displaced to compensate for the reduction
in kinetic energy of the flowing fluid due to boundary layer
or formation. Consider again Fig. 11.2. For the mass flow rate
through small elemental streamtube, recall, Eq. (11.4)
δ
ρu u
θ= 1− dy (11.14) m̊s = ρubdy
ρa Ua Ua
0
The kinetic energy for the fluid passing through the elemental
If the flow is incompressible, then we have streamtube is
δ 1
u u Ks = (ρubdy) u2 (11.17)
θ= 1− dy (11.15) 2
Ua Ua
0
But the kinetic energy of fluid when there is no plate,
The concept of momentum thickness is quite useful in skin
1
friction calculations. From Eq. (11.15), it is evident that θ Ks = (ρubdy) Ua2 (11.18)
represents the distance through which the surface would have 2
to be displaced in order that, with no boundary layer, the total Loss in kinetic energy due to presence of boundary layer is
momentum at the section considered would be the same as
that which would actually occur. 1
Ks = (ρubdy) Ua2 − u2 (11.19)
Furthermore, to determine the nature of boundary layer, 2
another useful parameter referred to as shape factor (H) is
also used. It is defined as the ratio of displacement thickness Let δ ∗∗ be the distance by which plate is displaced to com-
(δ ∗ ) to momentum thickness (θ). pensate for the reduction in kinetic energy. Thus, the loss of
kinetic energy through δ ∗∗ of the fluid flowing with velocity
δ∗ (Ua ) is
H= (11.16)
θ
1 ∗∗ ρvL
Ks = ρbδ Ua Ua2 (11.20) Re = (11.25)
2 μ
or where L is the characteristic length of the problem investi-
1 gated.
Ks = ρbδ ∗∗ Ua3 (11.21)
2 In addition to viscosity, the compressibility of the gas
also affects the aerodynamic forces acting on the object. If
Thus, we have the object moves through the gas at relatively low speeds
1 1 (M < 0.5), the density of the fluid remains constant. How-
ρbδ ∗∗ Ua3 = (ρubdy) Ua2 − u2 (11.22) ever, if the object travels at high speeds, some portion of the
2 2
kinetic energy of the object is utilized in compressing the sur-
or rounding fluid and thus changes the fluid density, which in
δ 2 turn affects the amount of resulting force acting on the object.
∗∗ u u The effect of compressibility is more pronounced at higher
δ = 1− (11.23)
Ua Ua flow speeds. For the flow speeds near to and beyond the speed
0
of sound, the shock waves are generated that affect the lift
and drag on the object. Thus, in addition to inertia force, the
elastic force should also be considered while analyzing the
11.3 Similarity Parameters →
−
compressible flows. The ratio of inertia force F i to elastic
−
→
When an object moves through the atmosphere, the gas force F e is known as the Mach number.
molecules near the object are disturbed and moves around
it which, in turn, generates aerodynamic forces over the body. −
→
The magnitude of these forces depends on the speed of object, F i
− ≡M (11.26)
shape of object, mass of exhaust gas, compressibility, and →
F e
viscosity of the gas. The precise modeling of these aspects
requires similarity parameters, which are essentially the ratio
−
→ −
→
of forces governing the phenomena. If the values of the simi- In the above relation, if we replace F i and F e with their
larity parameters are the same for two different flows, then the order of magnitudes, we have
relative importance of the forces have to be suitably modeled
v
(Table 11.1). M= (11.27)
Whenever airstream flows past an object, the air molecules a
stick to the surface due to the viscosity of air. Consequently, a That is, the Mach number can be defined as the ratio of local
thin viscous layer forms over the surface known as boundary flow speed to the speed of sound. Thus, to establish the sim-
layer, whose effects must be considered while calculating the ilarity of flows over a model and the prototype, the Mach
drag acting on the object. The drag and lift acting on the number must be matched if the compressibility effects are
body are generally estimated through using wind tunnels. The taken into account. Further, the Mach number is also a scal-
body mounted in a wind tunnel for testing is termed as the ing parameter in many of the governing equations for shock
model, which is a scaled replica of the actual object, called and expansion waves. Furthermore, the experimental results
the prototype. To establish flow similarity between model and obtained through wind tunnel testing can be applied to actual
the prototype, the similarity parameters must be matched. flight conditions if the similarity of Mach number is ensured.
The Reynolds number is an important similarity parameter Clearly, the value of drag coefficient measured at subsonic
for viscous flows, defined as the ratio of inertia force to vis- speeds cannot be used at supersonic Mach numbers as the
cous force. From conservation of momentum principle, we compressibility of gas (air) alter the flow physics between
can define the inertia force as the product dv of density (ρ) × these two flow regimes.
velocity (v) × gradient of the velocity dx , and the viscous
viscosity coefficient (μ) × second
force can be definedas the
2
gradient of velocity ddxv2 . Thus, the Reynolds number will be 11.4 Boundary Layer Separation
ρ×v× dv The boundary layer assumption help in solving an initial-

Re = dx
(11.24) boundary value problem (Navier–Stokes equations) through a
d2 v
μ× dx2 stipulated procedure that solves continuous differential equa-
tions via step-wise changes in the values of the indepen-
On further simplifying, we get dent variables. Due to this, information progressively moves
11.4 Boundary Layer Separation 255
Table 11.1 Similarity parameters

Contents Compressibility Viscosity
Parameter Mach number (M) Reynolds number (Re)
Characteristic Springiness Stickiness
V ρVL
Equation a μ
flow velocity density×velocity×length
Definition speed of sound coefficient of viscosity
downstream within the boundary layer. If one moves toward moving along the flow direction and may stop and reverse its
the wall, the value of inviscid flow pressure impressed upon direction causing the boundary layer to deflect away from the
the boundary layer will remain constant, but the convection boundary. This is an onset of separation shown by the point
velocity reduces and becomes zero at the wall (no-slip con- C in Fig. 11.3.
dition).
Thus,
in the presence of adverse pressure gradient Furthermore, the above separation phenomena can be
dp
dx > 0 , convection becomes weak as one approaches the
described mathematically using Prandtl’s boundary layer
wall, while the adverse pressure gradient remains invariant at equations given by Eqs. (11.42) and (11.47), provided that the
those heights. However, with increase of adverse pressure gra- curved boundaries have no large variations in its curvature.
dient, a stage may come when the pressure gradient becomes We consider x as a curvilinear coordinate having the shape
more adverse as compared to convective acceleration in the of boundary and y as the normal distance away from the
proximity of wall. Consequently, downstream propagation of boundary. Adopting this convention, consider Fig. 11.3,
information would not be possible under such condition. which depicts the onset of separation. We have pointed out
So far, we have focused our attention in dealing with the earlier that at point C flow tend to reverse its direction as a
flow over the flat plate with a zero pressure gradient in order result of an adverse pressure gradient impressed upon the
to describe some fundamental concepts in a simplified man- boundary layer by the outer inviscid freestream. At this point,
ner. We will now study an important characteristic associated the velocity as well as its variation with respect to y both are
with the flow over surfaces other than the flat plate oriented clearly zero. That is
parallel to the main flow, known as the separation phenomena,
in which the shear layer adjacent to the wall is deflected side- (u)y = 0 = 0 (11.28)

ways, in the direction normal to the wall. In this condition, it ∂u
=0 (11.29)
is usually said that the flow is separated. ∂y y = 0
Let us now understand that under what circumstances the
flow past a surface separates? Consider an incompressible Thus,
flow around an airfoil at high angle of attack, as shown in
Fig. 11.3. Our region of interest is the suction surface of the τw = 0 (11.30)
airfoil. It has been revealed by various computational and
experimental studies that between points A and B, the invis- That is, shear stress at the wall is zero. Further, the point of
cid flow outside the boundary layer accelerates. Note that the separation is defined as the limit between forward and reverse
airfoil assumes its maximum thickness at point B and beyond flow inside the viscous layer, in near vicinity of the wall.
this point, there is a continual decrease in freestream veloc- Since the flow field vanishes at the separation point thus it
ity and thus the maximum velocity occurs at point B itself. is considered to be a singular point (point C). Also, the wall
According to the Bernoulli’s equation, it is understood that streamline gets bifurcated and thus this location is also known
the continual increase in velocity between A and B even- as half-saddle point.
tually decreases the static pressure from close to stagnation Furthermore, beyond the half-saddle point, the adverse
pressure at point A, with minimum pressure at point B. On the pressure gradient which acts in opposite direction resulting in
other hand, the flow experiences an increase in pressure due to a backflow is shown by point D in Fig. 11.3. The separation
continual decrease in flow velocity beyond the minimum pres- phenomena can also be explained using the second derivative
sure point (point B). Clearly, the boundary layer experiences of horizontal component of velocity (u) at the wall. Consider
a pressure gradient in the direction of flow which resists its the x-momentum equation given by Eq. (11.47) at the wall,
motion. Such a pressure gradient is termed as adverse pressure where u = v = 0. Thus,
gradient. The fluid element moving in the boundary layer is
subjected to this increasing pressure and hence slows down.
∂2u 1 dp
Further deceleration may lead to a situation when the fluid = (11.31)
∂y2 y=0 μ dx
element does not possess enough kinetic energy to continue
dp/dx > 0
Ua
Separation
streamline
ψ=0
ψ=0
Edge of
boundary
B C D
layer A τ w= 0 τw< 0
τw> 0
Ua B Point of separation
A C
dp
Fig. 11.3 Boundary layer flow over a flat plate, when dx >0
First, consider a situation due to a favorable pressure

2 gradi- 11.4.1 Physics of Separation Bubbles in
ent, i.e., dx < 0. From Eq. (11.31), we have ∂∂yu2
dp
< 0. Boundary Layers
y=0
As we proceed towards the freestream, the local velocity
The performance of almost all the aircraft are affected due
u approaches asymptotically, so ∂u ∂y decreases at continu-
to the undergoes transition formation of separation bubbles,
ously lesser rate in the y-direction. This means that ∂∂yu2
2
which usually occurs at low Reynolds numbers. This phenom-
y=0
remains negative near the edge of the boundary layer. Thus, it ena takes place over many bodies with large upper surface
is clearly seen that for a continual decrease in pressure gradi- curvatures. In aircraft, high local curvature over the forward
ent, part of the chord may initiate a laminar separation when the
2 the
curvature of a velocity profile is always negative, i.e.,
∂ u
< 0, as shown in Fig. 11.4a. Next consider the situ- airfoil is at a moderate angle of incidence. From Bernoulli’s
∂y2 y=0 equation, it can be observed that the pressure rises over the
dp
ation due to an adverse pressure gradient, where dx > 0. From suction surface of the airfoil due to drop in flow velocity,
Eq. (11.31), one can observe that the curvature
2 of velocity pro- leading to an adverse pressure gradient which resists the flow
∂ u
file at the wall must be positive, i.e., ∂y2 > 0, shown that causes the separation of laminar boundary layer. The
y=0
boundary layer leaves the surface approximately in tangen-
in Fig. 11.4b. However, near the edge of boundary layer the
tial direction, resulting in a bubble-shaped separation zone.
arguments made regarding ∂u ∂2u
∂y and ∂y2 in the previous case
2 As compared to an attached boundary layer, this separated
∂ u but still laminar flow is more sensitive to disturbances that
is still valid and the curvature is negative ∂y2
<0 .
y=0 grow at the rapid pace and particularly at low Reynolds num-
Therefore, it is obvious that for an adverse
pressure gradient, bers. In turn, the separated laminar boundary layer undergoes
∂2u
there must exist a point for which ∂y2 = 0. This point is a transition to turbulence and experiences an increased bound-
y=0
known as the point of inflection in the velocity profile of the ary layer thickness. Indeed, the transition does not occur at
boundary layer, as shown in Fig. 11.4b. However, the point a point, instead, it spreads over the small but finite region,
of separation means zero velocity gradient at the wall, i.e., located away from the airfoil at the outer boundary of the
∂u ∂2u
∂y = 0. Also, Eq. (11.31) shows that ∂y2 > 0 at the wall can
separated zone. Furthermore, the increase in boundary layer
only occur due to an adverse pressure gradient. But we have thickness may be sufficient enough for lower edge to come
already observed that at the edge of boundary layer, ∂∂yu2 < 0.
2 back and reattach as turbulent boundary layer on the surface.
The zone in which a turbulent flow reattaches the surface is
It is, therefore, evident that if there exists a point of sepa-
termed as reattachment point. In this way, a bubble of fluid
ration, there must exist a point of inflection in the velocity
gets entrapped in the volume enclosed between the separation
profile.
and reattachment points. This bubble is called the laminar sep-
11.4 Boundary Layer Separation 257
y y y
2
u u
u τw y y2

dp
(a) Under favorable pressure gradient dx
<0 .
y y y
Point of
inflection
u uy uyy

dp
(b) Under adverse pressure gradient dx
>0 .
Fig. 11.4 Velocity distribution and its shear derivatives within a boundary layer
aration bubble. A circulatory flow may be established within xδ (11.32)

the bubble; the direction of flow in the close proximity of
the surface may even be the opposite of the direction of the where δ is boundary layer thickness. Clearly, the momentum
outer flow. However, there is no energy exchange takes place thickness and kinetic energy thickness will also be smaller
between the outer flow and the circulatory flow of the bubble. than the length scale in axial direction, i.e.,
Thus, the laminar separation bubble becomes quite stable.
Since the bubble increases the thickness of boundary layer θx (11.33)
and thus the drag of the airfoil also increases. Also, the lam- ∗∗
δ x (11.34)
inar separation bubble formation affect the lift and moment
associated to an airfoil and thus it can pose challenges to the The structure and properties of boundary layer depend on
stability and controllability of the aircraft as well. whether the boundary layer is laminar or turbulent. We have
Furthermore, if the separated flow transition occurs quite seen that the thin viscous layer in close proximity to a sur-
far away from the airfoil surface such that the flow can not face, in which flow velocity increases from zero at the wall
reattach itself to the surface again, then the separated zone to freestream value at a distance δ above the wall, is termed
thickens which causes a high drag but a very low lift. A quite as boundary layer. In this manner, we can define two flow
efficient way of reducing the drag penalties associated with the regions, even if, the demarcation is not very sharp. The thin
formation of laminar separation bubble is to use mechanical boundary layer adjacent to wall has the large value of velocity
turbulators. They are solid metallic strips placed just before gradient normal to wall as compared to the velocity gradient in
the region of separation bubble, shed vortices into the flow to the axial direction, i.e., ∂u ∂u
∂y ∂y . Also, even with small value
cause transition into the turbulent boundary layer, before the
of viscosity, the shear stress τ = μ ∂u
∂y may assume large val-
laminar separation can occur.
ues. The region outside the viscous layer, velocity gradients
∂u ∂u
∂x and ∂y are of the same order but both are insignificant.
Also, the influence of viscosity is small in this region and
11.5 Boundary Layer Equations
thus, the flow is of inviscid in nature and may be considered
Since the viscous layer adjacent to the surface is very thin so as potential.
that the characteristic length along the axial direction (x) is Moreover, the growth of boundary
layer thickness (δ)
much larger than the transverse length scale (δ), i.e., in x-direction is slow dx 1 . Usually, δ decreases with
dδ
Ua
y
Ua δ
x
Ua
Fig. 11.5 Boundary layer over a flat plate at zero incidence
y y’ = y/δ
Rescale
δ
x’ = x/L
x
L 1
Fig. 11.6 Transformation of boundary layer to render the flow variables of order unity
increase of Reynolds number (Re), i.e., δ decreases with the boundary layer varies from zero to Ua over a length δ
decrease in viscosity (υ). From many exact solutions of which is an order of magnitude smaller than the length along
Navier–Stokes equations, it has been observed that the wall, thus L cannot be used as characteristics length in
√ the y-direction across the boundary layer. The appropriate
δ∼ υ (11.35) characteristics length in the y-direction, therefore is δ but
in x-direction it is still L. Thus, the boundary layer analysis
Since the boundary layer is very thin, it allows some sim-
requires two different characteristics lengths, one in stream-
plifications that results in the equations commonly referred
wise and other in the transverse direction. The velocity com-
to as Prandtl boundary layer equations. Without referring to
ponents in these two directions may also be expected to
whether the flow is laminar or turbulent, we will develop the
have different characteristics values. Assume u and v, respec-
boundary layer equation for the steady, two-dimensional, and
tively, are the velocity components in streamwise and trans-
incompressible fluid of constant viscosity that flows past a flat
verse directions, then the nondimensional variables u = ⊂ua ,
plate is shown in Fig. 11.5. Essentially, an order of magnitude
v = Vv , x = Lx and y = yδ are expected to be of order unity.
analysis of each term in Navier– Stokes equations pertaining
It is now clear that the value of U is of the same order
to the flows both inside and outside the boundary layer will
as Ua but the values of δ and V are as yet unknown. To
be performed.
determine them, we essentially utilize suitable transforma-
For convenience, assume the flat plate to be coinciding with
tions to render all the flow variables such as x , y , u , v , p
the x-axis and subjected to a uniform freestream of velocity ∂v
and all the derivatives such as, ∂u
∂x , ∂y , etc., are of order unity
Ua in the x-direction (Fig. 11.5). Let us further assume that
(Fig. 11.6).
the boundary layer thickness is very small as compared to the
length of flat plate, i.e.,
11.5.1 Continuity Equation
δL (11.36)
For a two-dimensional fluid flows, consider the following con-
where δ is boundary layer thickness, measured in y-direction
tinuity equation
perpendicular to the plate. Furthermore, the velocity across
11.5 Boundary Layer Equations 259
∂u ∂v due to change of momentum of any fluid element as it moves

+ =0 (11.37)
∂x ∂y between regions of different velocities in the flow field. Using
dimensionless variables, one can rewrite Eq. (11.43) as,
If the nondimensional variables are used in the above
equation, we get 1 ∂u Uδ 1 ∂u
ρU Uu + ρ Uv =
U ∂u V ∂v L ∂x L δ ∂y
+ =0 (11.38)
L ∂x δ ∂y U ∂p
2 1 ∂ 2 u 1 ∂ 2 u
−ρ + μ U + U
L ∂x L2 ∂x2 δ 2 ∂y2
The two terms in the above equation must be of the same order, (11.44)
otherwise one of them will drop out making the equation to
degenerate. That is ρU2
Dividing the above equation by L throughout, we have
U V
∼ (11.39) L 2
δ
∂p
1 ∂ 2 u ∂ 2 u
L ∂u ∂u
u +v =− + 2
+ δ (11.45)
or ∂x ∂y ∂x Re ∂x Re ∂y2
LV
≡ O(1) (11.40) where Re = ρUL
Uδ μ is the Reynolds number. For the large values
of Re, which is an essential requirement for boundary layer
2
Thus, the component of velocity in y-direction can be chosen approximation, the coefficient of ∂∂xu2 , i.e., Re
1
will be very
as 1 ∂ u 2
small as compared to one. Therefore, the term Re ∂x2
can
be neglected. Since the viscous forces play a dominant role
δ
V=U (11.41) inside the boundary layer and hence must be retained. To
L 2
achieve this, the coefficient of ∂∂yu2 should be of order unity.
The velocities in y-direction, therefore, are much smaller than That is
those in the x-direction. With these approximations, the conti- L 2
nuity equation in terms of original variables can be expressed δ
∼1
as Re
∂u ∂v
+ =0 (11.42) or
∂x ∂y
L 2
The results of this order of magnitude analysis will aid one δ
≡ O (1)
to perform the same for momentum conservation equations Re
considering one at a time, buttressed by the observation that
Thus, the boundary layer thickness can be chosen as
the viscous effects are confined in a narrow region in a shear
layer adjacent to the wall. For simplicity, we will consider the δ 1
steady flow condition, perceiving that identical analysis can =√ (11.46)
L ReL
be performed for unsteady flows as well, provided the thin
shear layer assumption holds. Thus, the x-momentum equation in terms of original quanti-
ties can be written as
11.5.2 x-Momentum Equation ∂u ∂u ∂p ∂2u

u +v =− +μ 2 (11.47)
∂x ∂y ∂x ∂y
Consider the x-component of momentum equation for the
steady two-dimensional flow,
2 11.5.3 y-Momentum Equation
∂u ∂u ∂p ∂ u ∂2u
ρ u +v =− +μ + 2 (11.43)
∂x ∂y ∂x ∂x2 ∂y Similarly, considering the y-component of momentum equa-
tion for the steady two-dimensional flow, we have
In Eq. (11.43), the first and second terms on the right side
2
are the pressure force and viscous force, respectively. The ∂v ∂v ∂p ∂ v ∂2v
term on left side represents the inertia force that is caused ρ u +v =− +μ + (11.48)
∂x ∂y ∂y ∂x2 ∂y2

∂p
Rewriting the above in terms of nondimensional quantities in practical fluid flow applications, the last equation ∂y =0
we get, is omitted and in x-component of momentum equation the
pressure gradient term is written as the ordinary derivative,
1 Uδ ∂v Uδ 1 Uδ ∂v dp
ρU u +ρ v = dx . The boundary conditions that satisfies the above equations
L L ∂x L δ L ∂y are

U2 ∂p Uδ 1 ∂ 2 v Uδ 1 ∂ 2 v
−ρ + μ +
δ ∂y L L2 ∂x2 L δ 2 ∂ 2 y2 at y = 0 : u = 0, v = 0 (11.55)

Dividing the above equation by ρU2
throughout, we have at y = δ : u = Ua , v = 0 (11.56)
δ
2 Also, for a steady two-dimensional inviscid flow, Eq. (11.43)

δ ∂v ∂v ∂p 1 δ 2 ∂ 2 v ∂ 2 v
u + v = + + 2 simplifies to
L ∂x ∂y ∂y Re L ∂x2 ∂y
1 ∂p dUa
For large values of Re and from Eq. (11.46), the terms − = Ua (11.57)
2 2 2 ρ ∂x dx
δ ∂v ∂v δ ∂ v 2
L u ∂x + v ∂y
1
and Re L ∂x2
+ ∂∂yv2 will be
where u = Ua and v = 0. Equation (11.57) essentially relates
very small and hence can be neglected. With these approx-
the pressure gradient distribution at the edge of boundary layer
imations, the y-component of momentum equation can be
with the velocity gradient estimated from the ideal flow anal-
written as
∂p ysis. Furthermore, by integrating equation (11.57), one can
=0 (11.49) obtain Bernoulli equation for a steady and incompressible
∂y
flow as
From Eq. (11.49), it is clear that the pressure in a direction
1
normal to the boundary layer is practically invariant. Thus, p + ρU2 = constant (11.58)
2

p = p x (11.50) where subscript ‘a’ is dropped for simplicity.
At the edge of boundary layer, it is assumed to be equal to

that at the outer edge of the boundary layer (inviscid region). 11.6 von Karman Momentum Integral
That is Equation

p (x ) = pe x (11.51) In 1921, Theodore von Karman put forward an integral repre-
sentation of boundary layers, which simplifies the governing
Because of this reason, it is said that the pressure is impressed partial differential equation to an ordinary differential equa-
upon the boundary layer by the outer flow, and it is a function tion. It is an alternative approach to solving the boundary
of only x coordinate (flow direction) and time t. This result layer problem by considering the integral form of momentum
paved a way for decoupling the viscous and inviscid part of equation. We will now develop the integral form of boundary
an actual flow in the presence of a boundary layer. In addition, layer equations for the two-dimensional steady compressible
one conceptual model can be used to analyze the fluid flow flows beginning with
in inviscid region (outside the boundary layer) to obtain the
pressure distribution that can be used to solve the following ∂ (ρu) ∂ (ρv)
+ =0 (11.59)
boundary layer equations. ∂x ∂y
∂u ∂v ∂u ∂u dp ∂τ
+ =0 (11.52) ρu + ρv =− + (11.60)
∂x ∂y ∂x ∂y dx ∂y
∂u ∂u ∂p ∂2u
u +v =− +μ 2 (11.53) where τ = μ ∂u
∂x ∂y ∂x ∂y ∂y , is the shear stress; and u and v, respectively,
∂p are the velocity components in x- and y-directions. Recall at
=0 (11.54) the boundary layer edge velocity is related to the pressure
∂y
gradient through the Bernoulli’s equation,
The above three equations are valid approximation of the
dp dUa
Navier–Stokes equation inside the boundary layer. However, = −ρa Ua (11.61)
dx dx
11.6 von Karman Momentum Integral Equation 261
Introducing above into Eq. (11.60), h is so chosen that it is independent of x and thus order of
differentiation and integration can be interchanged with each
∂u ∂u dUa ∂τ other. Therefore, the first term on left side in Eq. (11.67) can
ρu + ρv = ρa Ua + (11.62)
∂x ∂y dx ∂y be written as ⎡ h ⎤

d
Multiplying Eq. (11.59) by u and adding it to the above equa- − ⎣ ρu(Ua − u)dy⎦
tion, we get dx
0
∂ 2 ∂ dUa ∂τ Changing the limits of integration and rearranging,

ρu + (ρuv) = ρa Ua + (11.63)
∂x ∂y dx ∂y ⎡ ⎤
∞ ∞
d ⎣ 2 ρu u ⎦ dUa ρu
Integrating the above equation with respect to y between 0 ρa Ua 1− dy + ρa Ua 1− dy = τw
dx ρa Ua Ua dx ρa Ua
and h where h is a constant and sufficiently greater than the 0 0
(11.68)
boundary layer thickness (δ), so that all the variables assume
One should note that for compressible flows, the displacement
their freestream values much before reaching y = h.
thickness and momentum thickness are given by
h h
∂ 2 dua ∞
ρu dy + ρh vh ua = ρa ua dy − (τh − τw ) ∗ ρu
∂x dx δ = 1− dy (11.69)
0 0 ρa Ua
(11.64) 0
where the quantities shown with subscript h assume their ∞

ρu u
value at h after integration; and τw denotes the shear stress at θ= 1− dy (11.70)
ρa Ua Ua
the wall. But for freestream τ → 0 and thus putting τh = 0 0
in Eq. (11.64). We get
Substituting Eqs. (11.69) and (11.70) into Eq. (11.68), we
h h have
∂ 2 dUa d dUa ∗
ρu dy + ρh vh Ua = ρa Ua dy − τw ρa Ua2 θ + ρa Ua δ = τw (11.71)
∂x dx dx dx
0 0
(11.65) where
Also, integrating equation (11.59) with respect to y between d

ρa Ua2 θ ≡ rate of loss of momentum
0 and h, we have dx
dUa ∗
ρa Ua δ ≡ effect of freestream pressure gradient
h dx
∂
ρh vh = − (ρu) dy (11.66) τw ≡ wall shear stress
∂x
0
Writing Eq. (11.71) in terms of pressure gradient,
Introducing Eq. (11.66) into Eq. (11.65), we get
d dp
ρa Ua2 θ = τw + δ ∗ (11.72)
h dx dx
∂ 2 ∂ dUa
ρu − Ua (ρu) − ρa Ua dy = −τw
∂x ∂x dx Using isentropic relation for compressible flows,
0
1 dρe M2 due
Rearrange the terms, =− e (11.73)
ρe dx ue dx
h
∂ dUa and
using the
definitions for skin friction coefficient
− [ρu (Ua − u)] − (ρa Ua − ρu) dy = −τw ∗

∂x dx Cf = 1 τw 2 and shape factor H = δθ , simplifying
0 2 ρa Ua
(11.67) Eq. (11.71)
Note that, for all the heights greater than or equal to h, velocity
difference Ua − u ≈ 0. Thus, we can extend the upper limit to dθ θ dUa C
f
+ H + 2 − Ma2 = (11.74)
any height above h (we will replace h by ∞). Also, the limit dx Ua dx 2
where Momentum Equation
τw ∂u ∂u ∂u ∂Ua ∂Ua ∂2u

Cf = 1 ≡ coeffcient of skin friction
2 +u +v = + Ua +ν 2 (11.78)
2 ρa Ua ∂t ∂x ∂y ∂t ∂x ∂y
δ∗
H= ≡ shape factor
θ Continuity Equation
dθ
≡ inertia term
dx ∂u ∂v
θ dUa + =0 (11.79)
H + 2 − Ma2 ≡ pressure gradient term ∂x ∂y
Ua dx
Cf Let us multiply the momentum equation by 2u and the conti-
≡ wall skin friction term
2 nuity equation by Ua2 − u2 . We get
For a constant density incompressible flow, we can set
∂ (Ua − u) ∂Ua ∂u ∂u ∂2u
Ma = 0. Thus, the corresponding momentum integral 2u + 2uUa − 2u2 − 2vu = 2uν 2
equation becomes ∂t ∂x ∂x ∂y ∂y
(11.80)
dθ θ dUa Cf ∂u ∂v
+ (H + 2) = (11.75) Ua2 − u2 + Ua2 − u2 =0
dx Ua dx 2 ∂x ∂y
Equation (11.75) is known as momentum integral first derived (11.81)
by von Karman. Note that, till this step no specific assumption
Adding the equations above, one obtains
was made for τw and the velocity gradient at the boundary,
and thus this equation is valid for both laminar and turbulent

boundary layers. Also, the shape factor (H) is only pertinent
∂ Ua u − u2 ∂ u ∂ 2 ∂ 2
∂t
+ Ua2
∂t
1−
Ua
+
∂x
Ua u − u3 +
∂y
vUa − vu2
when there is a freestream pressure gradient dU dx = 0 . A
a
2
∂u ∂ ∂u
large shape factor stipulates that a boundary layer is near- = 2ν − 2ν u
∂y ∂y ∂y
ing to separation. However, if the zero pressure gradient (11.82)
is impressed upon the boundary layer, then Eq. (11.74) or
Eq. (11.75) become Integrating the equation above with respect to y between the
limits; 0 and ∞. We have,
dθ Cf
= (11.76)
dx 2 ∞
∂ Ua u − u2 ∂ u ∂ 2 ∂ 2
+ Ua2 1− + Ua u − u3 + vUa − vu2 dy
∂t ∂t Ua ∂x ∂y
For this case, the total drag coefficient, CD (L) = 1D(L)2 0
2 ρUa L ∞
L
∂u
2
∂

∂u
where D (L) = 0 τw dx. Thus, CD (L) can also be written = 2ν − 2ν u dy (11.83)
∂y ∂y ∂y
as 0
θ (L) Let us replace y-component of velocity v by using the conti-

CD (L) = 2 (11.77)
L nuity equation and change the order of integral and differen-
Thus, the total drag is completely determined by the down- tial operators in the equation above. Consider the Eqs. (11.3),
stream momentum thickness. (11.15) and (11.23), we obtain
∂δ ∗ ∂ 2 ∂ 3 ∗∗ 2ε
Ua2 + Ua θ + Ua δ = (11.84)
∂t ∂t ∂x ρ
11.7 K. Wieghardt Energy Integral Equation
∞ 2
In 1948, K. Wieghardt derived energy integral equation using where ε = μ 0 ∂u ∂y dy; ε is the energy dissipation rate
a similar approach as used by von Karman in deriving momen- due to viscosity across the boundary layer. δ ∗ , θ and δ ∗∗ ,
tum integral equation for laminar boundary layer. Considering respectively, are the displacement thickness, the momentum
the general forms of the Prandtl’s boundary layer equations thickness and the energy thickness of the boundary layer. For
for an incompressible flow as given below. the steady laminar flows, Eq. (11.84) reduces to
11.7 K. Wieghardt Energy Integral Equation 263
∞ 2
d 3 ∗∗ 2μ ∂u 11.8.1 Incompressible Laminar Flow over a Flat
Ua δ = dy (11.85) Plate (Blasius Solution)
dx ρ ∂y
0
One of the classic and most often used solutions of boundary
Equation (11.85) is the energy integral equation for the steady, layer theory is the semi-infinite flat plate solution, which can
incompressible laminar boundary layer flows. represent either laminar or turbulent flow. For the laminar flow
past a flat plate, the Prandtl boundary layer equations can be
solved exactly for velocity components, u and v, assuming
11.8 Laminar Boundary Layers that the freestream velocity Ua is constant, i.e., dU
dx = 0. The
a
solution was first mentioned by Prandtl’s student Blasius in

When an infinite expanse of fluid flows past a flat plate, the his doctoral thesis in 1908.
effect of viscosity brings the fluid in immediate contact with Consider two-dimensional, steady, incompressible bound-
it, to rest. This fluid, in turn, decelerates the adjoining layers, ary layer flow of a Newtonian fluid with zero pressure gradi-
thereby giving rise to a boundary layer. The boundary layer ent and with negligible body forces over a flat plate, shown
thickness (δ) grows in the downstream direction because of in Fig. 11.7. The governing equations of motion are,
more and more fluid brought under the influence of viscos-
ity as the flow proceeds. If the plate is very thin, the inviscid ∂u ∂v
velocity also referred to as freestream velocity (Ua ) is con- + =0 (11.86)
∂x ∂y
stant everywhere. Since the velocity is constant, the inviscid
dp ∂u ∂u ∂2u
pressure field is also constant in whole flow domain, i.e., dx u +v =ν 2 (11.87)
∂x ∂y ∂y
will be set to zero in Eq. (11.43). Furthermore, depending
upon the flow characteristics inside the boundary layers, one The boundary conditions become
can classify the boundary layers into the following two types.

1. Laminar boundary layers. at y = 0 : u = 0; v = 0 (11.88)
2. Turbulent boundary layers. ∂u
at y = ∞ : u = U; =0
∂y
A laminar boundary layer is the one where the fluid layers are
assumed to slide over one another smoothly. That is, in a lami- From Eqs. (11.86), (11.87) and (11.88), it is evident that
nar boundary layer, the fluid layers, also called laminas, slide these equations form the set of coupled, nonlinear partial
over adjacent layers such that there exists only the micro- differential equations, where velocity components (u, v) are
scopic exchange of matter (mass, momentum, and energy) unknown. To solve them, Blasius showed that the velocity
among the layers. However, if we try to observe the laminar u
profile Ua must be similar for all the values of x, when
flows on macroscopic scale, we will not feel any exchange of
plotted against a nondimensional distance from the wall. The
matter taking place between the fluid layers. Moreover, any
boundary layer thickness (δ), was an obvious choice for mak-
small disturbances or instabilities are quickly damped out by
ing the distance from the wall nondimensional. He assumed
the viscous forces that resist the relative motion of adjoining
the solution in the following form:
layers. Thus, a laminar boundary layer flow is an orderly flow
in which fluid elements move in an orderly manner such that u
the transverse exchange of momentum is negligibly small, = f (η) (11.89)
Ua
however, the axial exchange of momentum may be consid-
erable. In contrast, the flow in a turbulent boundary layer is where η ∝ yδ . Based on above postulation, Blasius further

three-dimensional random phenomena, exhibiting multiplic- assumed that δ ∝ U νx
and considered
a
ity of scales, possessing vorticity, and showing very high dis-
sipation. Also, the laminar boundary layer causes less skin
friction drag1 as compared to the turbulent boundary layer Ua
η=y (11.90)
and thus the flow inside a laminar boundary layer is more νx
prone to separation (or less stable).
We know that the velocity components in x- and y-directions
can be represented in terms of stream function (ψ) as,
∂ψ
u= (11.91)
1 The friction between the surface of the body and the fluid causes viscous
∂y
shear stress and this force is known as skin friction drag.
Ua
Ua
δ (x)
δ
y
x τ w (x)
x=0 x=L
Fig. 11.7 Boundary layer on a horizontal flat plate
∂ψ also,
v=− (11.92)
∂x ∂2u Ua2 d3 f
= (11.98)
∂y2 νx dη 3
The above equations satisfy the continuity equation identi-
cally. Substituting for u and v in Eq. (11.87), we obtain an Introducing Eq. (11.98) into Eq. (11.87), we get
equation in which ψ is the single independent variable. Thus,
let us consider a non-dimensional stream function ( f (h)) as d3 f d2 f
2 + f =0 (11.99)
ψ dη 3 dη 2
f (η) = √ (11.93)
νxUa
The boundary conditions become
With ψ defined by Eq. (11.93) and η defined by Eq. (11.90),
one can calculate each of the term in Eq. (11.87). The velocity df
at η = 0 : f = =0 (11.100)
components are given by dη
df
at η → ∞ : =1 (11.101)
∂ψ ∂ψ ∂η df Ua df dη
u= = = νxUa = Ua (11.94)
∂y ∂η ∂y dη νx dη
It can be seen that the governing equations; Eqs. (11.86) and
(11.87) are the second-order partial differential equations that
∂ψ ∂f 1 νUa
v=− =− νxUa + f = have been transformed into a nonlinear, third-order ordinary
∂x ∂x 2 x differential equation, given by Eq. (11.99). Furthermore, the

df 1 1 1 νUa boundary conditions in transformed coordinate space is given
=− νxUa − η + f by Eq. (11.100). This is Blasius equation, for which a solution
dη 2 x 2 x
in closed form is not possible and hence the solutions are
obtained only through numerical integration. Some classified
or
1 νUa df values of the velocity profile shape Uua = f (η) are given in
v= η −f (11.95)
2 x dη Table 11.2.
On differentiating the velocity components, u and v with 11.8.1.1 Boundary Layer Thickness
respect to spatial coordinates x and y, we have Recall that the boundary layer thickness (δ) is defined as the
value of y for which Uua = 0.99. Further, from Table 11.2 for
∂u Ua d2 f
Ua = 0.99155, we get η = 5.0. Thus, Eq. (11.90) gives
u
=− η 2 (11.96)
∂x 2x dη

Ua
and δ = 5.0
∂u d2 f νx
= Ua Ua/νx 2 (11.97)
∂y dη
11.8 Laminar Boundary Layers 265
Table 11.2 The Blasius velocity profile

η=y U a
νx
u
Ua
0.0 0.0
1.0 0.32979
2.0 0.62977
3.0 0.84605
4.0 0.95552
5.0 0.99155
∞ 1.00000
or Solving above, we have

δ 5.0
=√ (11.102)
x Rex νx
θ = 0.664 (11.104)
Ua
11.8.1.2 Displacement Thickness
Recall Eq. (11.3) 11.8.1.4 Energy Thickness
Again using Eq. (11.23),
∞
∗ u
δ = 1− dy ∞ 2
Ua ∗∗ u u
0 δ = 1−
Ua Ua
0
From algebraic manipulation,
or
∞
∗ u dy ∞ 2
δ = 1− dη 2νx df df
Ua dη ∗∗
0
δ = 1− dη (11.105)
Ua dη dη
0
dy
But dη = 2νx
Ua , we have
Solving the above equation, we obtain
∞
2νx df ∗∗ νx
δ = ∗
1− dη δ = 1.0444 (11.106)
Ua dη Ua
0
11.8.1.5 Shear Stress and Coefficient of Skin
On solving, we get Friction
The local wall shear stress and hence the coefficient of skin
νx
δ ∗ = 1.7208 (11.103) friction can also be calculated readily from function f (η). The
Ua wall shear stress can be written as
11.8.1.3 Momentum Thickness ∂u

τw = μ y=0 (11.107)
From Eq. (11.15) ∂y
∞ or
u u ∂η ∂u
θ= 1− dy τw = μ × η=0 (11.108)
Ua Ua ∂y ∂η
0

∂η Ua ∂u d2 f
Rearranging the terms, we have where ∂y = 2νx and ∂η = Ua dη 2 η=0
. Thus,
∞
2νx df df Ua d2 f
θ= 1− dη τw = μUa
Ua dη dη 2νx dη 2
0 η=0
√
d2 f
But dη 2 η=0
= 0.332 2. Therefore, the wall shear stress layer flows of a compressible fluid over a semi-infinite flat
becomes plate are as follows.

Ua 0.332ρUa2 11.8.2.1 Continuity Equation
τw = 0.332μUa = √ (11.109)
νx Rex ∂ (ρu) ∂ (ρv)
+ =0 (11.114)
∂x ∂y
ρUa x
where Rex = μ . Hence, the coefficient of skin friction
(Cf ), is given by 11.8.2.2 x-Momentum Equation

τw ∂u ∂u dpe ∂ ∂u
Cf =
0.664
=√ ρu + ρv =− + μ (11.115)
(11.110) ∂x ∂y dx ∂y ∂y
2 ρUa
1 2 Rex
Now, the drag acting on one side of the plate (let the span-wise 11.8.2.3 y-Momentum Equation
breadth is B and length is L) is given by ∂p
=0 (11.116)
∂y
L
D=B τw (x) dx (11.111) 11.8.2.4 Energy Equation
2
0
∂h ∂h ∂ ∂T dpe ∂u
ρu + ρv = k +u +μ
Introducing Eq. (11.109) into Eq. (11.111), we have ∂x ∂y ∂y ∂y dx ∂y
(11.117)
L However, for the boundary layer flow over a flat plate
Ua dx
D = 0.332μBUa √
ν x dpe
0 =0
dx
or
Thus, the conservation equations become

LUa ∂ (ρu) ∂ (ρv)
D = 0.664μBUa = 0.664μBUa ReL (11.112) + =0 (11.118)
ν ∂x ∂y
Thus, the drag coefficient is
∂u ∂u ∂ ∂u
ρu + ρv = μ (11.119)
1.328
D ∂x ∂y ∂y ∂y
CD = 1 =√ (11.113)
2 ρUa BL
2 ReL
∂p
=0 (11.120)
∂y
11.8.2 Compressible Laminar Flow over a Flat
Plate 2
∂h ∂h ∂ ∂T ∂u
ρu + ρv = k +μ (11.121)
In previous section, we have discussed the properties asso- ∂x ∂y ∂y ∂y ∂y
ciated with the two-dimensional, steady, and incompressible
laminar boundary layer flow over a flat plate. But these results From the above equations, the following observations can be
are valid only for low Mach number flows where the den- made for a compressible boundary layer flow:
sity remains invariant throughout the boundary layer. How-
ever, we have seen earlier that fluid compressibility should • The energy equation must be included.
be considered when the flow Mach number is greater than • Density is treated as a variable.
0.3, i.e., (M > 0.3). To derive the governing equations for the • Since fluid viscosity (μ) and coefficient of thermal con-
compressible flow past a flat plate, we will first consider the ductivity (k), in general, are the functions of temperature
Navier–Stokes equations, which are subsequently reduced to and thus they should also be treated as variables.
simpler forms using suitable approximations. Let x and y be
the distances along and normal to the wall, respectively. The It is sometimes
convenient
to deal with stagnation or total
2
origin is fixed at the leading edge of the plate. The conserva- enthalpy h0 = h + v2 , instead of static enthalpy used in
tion equations for steady, two-dimensional laminar boundary Eq. (11.121). From the boundary layer approximations, it is
consistent to assume the velocity component along the y- Equation (11.128) is an alternative form of energy equation
direction to be much smaller than the velocity component for the compressible boundary layer over a flat plate.
along the x-direction, i.e., v u. Thus, In Eq. (11.128), Pr is the local Prandtl number, which
is a function of temperature (T) and varies through-
V2 (u + v)2 out the boundary layer. For the compressible laminar
h0 = h + =h+
2 2 flow over a flat plate, the system of governing equa-
tions comprises of Eqs. (11.114), (11.116) and (11.128),
But v u, we have
which are essentially the nonlinear partial differential
u2 equations.
h0 ≈ h + One can deduce the expressions for boundary layer char-
2
acteristics adopting a similar approach those used for incom-
Now, let us obtain the energy equation in terms of h0 . First, pressible flows. For compressible flow over the flat plate, the

multiply Eq. (11.119) with u, boundary layer thickness δcomp can be defined as

∂ u 2/2 ∂ u 2/2 ∂ ∂u 5.0x Tw
ρu + ρv =u μ (11.122) δcomp =√ G Me, Pr, (11.129)
∂x ∂y ∂y ∂y Rex Te
Adding Eq. (11.122) to Eq. (11.121), we get The boundary conditions are:

∂ h + u 2/2 ∂ h + u 2/2
ρu + ρv at y = 0; T = Tw (11.130)
∂x ∂y
2
∂ ∂T ∂u ∂ ∂u at y = d; T = Te (11.131)
= k +μ +u μ (11.123)
∂y ∂y ∂y ∂y ∂y
where Tw = wall temperature, Te = temperature at the
Recall that for a calorically perfect gas, boundary layer edge, and d = local distance from the wall.
In Eq. (11.129), function G can be obtained through the
dh = Cp dT numerical simulation of the problem. Furthermore, the skin
friction coefficient Cf,comp can be written as
Thus,

∂T 1 ∂h 1 ∂ u2 1.328 Tw
= = h0 − (11.124) Cf = √ F Me , Pr, (11.132)
∂y cp ∂y cp ∂y 2 Rec Te
Introducing Eq. (11.124) into Eq. (11.123) and then add to In Eq. (11.132), function F is obtained through the numerical
Eq. (11.122), we obtain simulation of the problem.
2
∂h0 ∂h0 ∂ k ∂ u2 ∂u ∂ ∂u
ρu + ρv = h0 − +μ +u μ
∂x ∂y ∂y cp ∂y 2 ∂y ∂y ∂y 11.8.3 Stagnation Point Flow (or Hiemenz
(11.125) Flow)
Rearranging the terms,
It is a fluid flow in the close proximity of a surface at

k ∂ u2 μk ∂ u2 μ ∂h0 ∂u which fluid impinging the surface bifurcates into two different
h0 − = h0 − = −u streams. Due to no-slip condition, the fluid is stagnant on the
cp ∂y 2 μCp ∂y 2 Pr ∂y ∂y
(11.126) wall, however, the stagnation point refers to stagnation points
of inviscid Euler’s solutions. In 1911, Hiemenz was the first
2
∂u ∂ ∂u ∂ ∂u who formulated the two-dimensional stagnation point flow
μ +u μ = μu (11.127)
∂y ∂y ∂y ∂y ∂y problem and thus, this problem is also known as Hiemenz
flow. We will now consider a simplest example of this type
Now introducing Eqs. (11.126) and (11.127) into of flow, as shown in Fig. 11.8. It can be seen that a uniform
Eq. (11.125), we get flow impinges perpendicularly upon a horizontal flat plate of
infinite length and the flow bifurcates.
∂h0 ∂h0 ∂ μ ∂h0 1 ∂u The flow in near vicinity of stagnation point on a fully
ρu + ρv = + 1− μu
∂x ∂y ∂y Pr ∂y Pr ∂y curved surface can be modeled by a flow towards an infinite
(11.128)
Ua Ua
y
δ υ
x
Fig. 11.8 Stagnation point flow field
flat plate, if the local curvature effects are neglected. Let plate where f (y) is an unknown function of y, need to be deter-
be in xz plane with stagnation point at (x, y) = (0, 0). Thus, mined. Introducing Eq. (11.141) into Eq. (11.136), we get
from inviscid flow theory, the stream function (ψ) can be
written as ∂u
= f (y)
∂x
ψ = kxy (11.133)
where denotes the differentiation with respect to y. Upon
integration, we have
where k is an arbitrary constant. Again from potential theory,
the velocity components are: u = xf (y) + constant
u = kx (11.134) Applying the no-slip boundary condition, i.e.,

v = −ky (11.135)
The two-dimensional continuity and the Navier–Stokes equa- at x = 0 u = v = 0

tions can be written as
Thus, the constant turns out to be zero. Hence,
∂u ∂v
+ =0 (11.136)
∂x ∂y df (y)
2 u=x (11.142)
∂u ∂u ∂u ∂p ∂ u ∂2u dy
ρ +u +v = −ρgx − +μ +
∂t ∂x ∂y ∂x ∂x2 ∂y2
(11.137) Since, the Bernoulli’s
equation for potential flow can be
2 written as p + 21 ρ u2 + v2 = p0 . On substituting u = kx
∂v ∂v ∂v ∂p ∂ v ∂2v
ρ +u +v = −ρgy − +μ + 2 and v = ky, we can write p + 21 ρk 2 x2 + y2 = p0 . Let us
∂t ∂x ∂y ∂y ∂x2 ∂y assume the following solution for a completely viscous flow.
(11.138)
1 2 2
With the boundary conditions, p0 − p = ρk x + G (y) (11.143)
2
at y = 0 u = v = 0 (11.139) where G (y) is the another function of y. Introducing
at y → ∞ u = kx (11.140) Eqs. (11.142) and (11.141) into Eq. (11.143), one obtains

Our approach is to find the solutions of Eqs. (11.137) and f 2 − ff = k 2 + νf (11.144)
(11.138) by reducing the partial differential equation to ordi- 1
ff = k 2 G − νf (11.145)
nary differential equation, achieved by taking an assumption 2
that when suitably scaled the velocity profile retains its shape
throughout the flow field. The flow field characteristics sug- From Eqs. (11.144) and (11.145), it can be seen that the
gest that the component of velocity normal to the flow direc- NS equations have been reduced to ordinary differential
tion is independent of x, i.e., equations. In addition, the boundary conditions given
as Eqs. (11.139) and (11.140) will be transformed as
v = −f (y) (11.141) follows:

at y = 0 f = f = 0 (11.146) ∂u ∂u 1 ∂ ∂u
u +v = u (11.153)
at y → ∞ f = k (11.147) ∂x ∂y ρ ∂y ∂y
The above equations need to be solved using the following

It can be seen from Eq. (11.144), that f (y) depends upon
boundary conditions:
both k and ν in addition to y. Therefore, we should reduce
Eq. (11.144) into a more useful form, so that one needs to
y → ∞ u → u1 (11.154)
solve it once. Let us assume the following transformations of
f (y) and y. y → −∞ u → u2 (11.155)
f (y) = βφ (η) where η = αy (11.148) Let us define the nondimensional stream function f (η) and
the similarity parameter (η) is given by
where α and β are constants to be evaluated by introducing
Eq. (11.148) into Eq. (11.144). Noting that, ψ (x, y) = u1 δ (x) f (η) (11.156)
y
η= (11.157)
∂f dη dφ δ (x)
f = = β
∂y dy dη
where δ (x) is the thickness of shear layers mixing with each
or other. It is considered with the thin-layer approximation same
dδ
as before, i.e., dx 1. Now, let us evaluate the velocity com-
f = αβφ
ponents, u and v as follows:
Thus, Eq. (11.144) becomes ∂ψ
u=
∂y
α2 β 2 φ2 − α2 β 2 φφ = k 2 + να3 βφ
Introducing Eq. (11.156) into above
Let us further assume that α2 β 2 = k 2 = να3 , then α = k
ν
√ ∂ [u1 δ (x) f (η)]
and β = kν. The universal form would be u=
∂y
φ + φφ − φ2 + 1 = 0 (11.149) or
with boundary conditions, ∂ [f (η)] ∂η

u = u1 δ (x)
∂η ∂y
φ (0) = φ (0) = 0 (11.150)
∂η

φ (∞) = 1 (11.151) From Eq. (11.157), ∂y = 1
δ(x) . Thus,
In summary, φ(η) is the solution of boundary layer problem ∂ [f (η)]

u = u1 = u1 f
given by Eq. (11.149) along with the boundary conditions; ∂η
Eqs. (11.150) and (11.151), which has no closed-form solu-
∂η
tion. Moreover, Eq. (11.149) is nonlinear and has to be solved Also, ∂x = − δy2 dx
dδ
= − ηδ dx
dδ
. Hence,
numerically together with the boundary conditions.
dδ
v = u1 f η−f
dx
11.8.4 Mixing of Two Uniform Laminar Flows
It should be noted that y = 0 is not always the line of sym-
Consider two uniform laminar flows which are being mixed metry for these mixing layers. Now, we will try to write the
with each other. Our objective is here to demonstrate the exis- shear layer equation in terms of similarity parameter. Differ-
tence of a similarity solution subject to some constraints. Let entiating the velocity components, we get
us begin with the conservation of mass and x-component of
∂u ∂η u1 dδ
momentum equation. With thin shear layer approximation, = u1 f =− η f
the governing equations become ∂x ∂x δ (x) dx
∂u ∂v and
+ =0 (11.152)
∂x ∂y
∂u ∂η u1 ∂2u u1 11.9 Application of von Karman

= u1 f = f and 2 = 2 f
∂y ∂y δ (x) ∂y δ Momentum Integral Equation
Introducing the above into Eq. (11.153), we obtain Recall Eq. (11.75), the momentum integral equation formu-
lated by von Karman is
u1 dδ dδ u1 vu1
−u1 f ηf + u1 f η−f f = 2 f
δ dx dx δ δ dθ θ dUa Cf
+ (H + 2) =
or dx Ua dx 2
u12 dδ u12 dδ u12 dδ vu1 To evaluate the equation above, we must have a suitable veloc-
− ηf f + ηf f − ff = 2 f ity profile. Further, as there was no specific assumption taken
δ dx δ dx δ dx δ
in Eq. (11.75) with respect to the velocity profile and thus this
Simplifying the above equation equation is valid for both laminar and turbulent boundary lay-
ers. With the help of an appropriate velocity profile, one can
u1 δ dδ calculate various integral properties.
f + ff = 0 (11.158)
v dx There exists a variety of momentum integral solution
methods. We have discussed a solution method adopted by
u1 δ dδ
Let us assume that v dx as the similarity solution and also, Pohlhausen in the following section.
u1 δ dδ
=λ
v dx 11.9.1 Karman–Pohlhausen Approximate
Solution Method
where λ is an arbitrary constant. Integrating the above equa-
tion we get, In an approximate solution method, to determine the laminar
δ2 vx
=λ boundary layer characteristics over a thin flat plate, an approx-
2 u1 imate velocity profile is assumed. Let us consider a 4th order
quartic nondimensional velocity profile as used by Karman
If we set the value of λ to be 21 . The above equation becomes (or Pohlhausen).
1/2
vx u (x, y)
δ= = a0 + a1 η + a2 η 2 + a3 η 3 + a4 η 4 (11.160)
u1 Ua
Substituting above in Eq. (11.158), we get where η = yδ . As there are five unknowns in Eq. (11.160) and
thus to evaluate them we require five boundary conditions.
1 The three boundary conditions given by Eqs. (11.100) and
f + ff = 0 (11.159)
2 (11.101) are as follows:
One should note that, Eq. (11.159) is similar to (11.99), which df
should be solved with the boundary conditions at the far field at η = 0 : f = =0 (11.161)
dη
regions. That is df
at η = 1 : =1 (11.162)
dη
at η = 0 f = 0
Additional boundary conditions are obtained at the edge of
or boundary layer where the velocity derivatives are zero, i.e.,
∂k u
1 ∂yk
= 0 for k ≥ 1. Thus,
at η = 0 f = (1 + k)
2
d2 f
at η = 1 : f = =0 (11.163)
where k = u2
= f is a variable of the problem. dη 2
u1
11.9 Application of von Karman Momentum Integral Equation 271
Further, we can now examine the x-momentum equation 1.37

cf,L = 1
(11.172)
(Eq. (11.47)). On the wall (y = 0), the equation reduces to (ReL ) 2
2
∂ u 1 ∂pa
= (11.164)
∂y2 y=0 μ ∂x 11.10 Laminar–Turbulent Transition
From Eq. (11.57), we can write the equation above as For a uniform flow past the flat plate, boundary layer starts
2 growing as laminar flow beginning at the leading edge of the
∂ u Ua ∂Ua flat plate. However, due to instabilities, this laminar flow turns
=− (11.165)
∂y2 y=0 ν ∂x into transition flow which subsequently turns into turbulent
flow. The instabilities introduced in a laminar flow are ampli-
∂2u ∂Ua
Assuming ∂y2 y=0
= − Uνa ∂x = λ, the additional bound- fied and result in flows which are orderly in nature in some
parts, but also shows temporarily (and/or spatially) irregular
ary condition can be written as
fluctuations of all the flow quantities in other parts, one may
d2 f say of a transitional flow state. In this state, intermittent lam-
at η = 0 : =λ (11.166) inar and turbulent behavior is observed; some phases occur
dη 2
in which the flow behaves as laminar, and in other phases,
Solving the set of equations gives us the coefficients, one can the flow exhibits turbulent characteristics. Since the transi-
find tion phenomena are exceedingly complex and thus there is no
1 1 1 1 accurate theory available to predict the process. It is the area
a0 = 0 a1 = λ+2 a2 = − λ a3 = λ−2 a4 = − λ + 1
6 2 2 6 of active research.
Transition process strongly depends upon both space and
Thus, Eq. (11.160) becomes
time, and thus it is difficult to simulate. The rapidity of the

2η − 2η 3 + η 4 + λ 16 η (η − 1)3 for η ∈ [0, 1]
process further adds the complexity in analyzing the phenom-
u(x,y)
= ena using the computational fluid dynamics techniques. How-
Ua
1 for η ≥ 1 ever, in spite of formidable difficulties, tremendous advance-
(11.167) ment has been made, and transition of the boundary layer
can be accurately predicted for the simple engineering appli-
Let us assume constant freestream velocity, i.e., λ = 0. From cations. The transition of a boundary layer takes place due
Eq. (11.75), we obtain to the growth of small perturbations within the layer. The
37 boundary layer essentially acts as a complex nonlinear oscil-
θ= δ (x) (11.168) lator which gives initially a linear response to some external
315
2μUa inputs. The sources of these disturbances may be freestream
τw = (11.169) turbulence, acoustic waves, surface roughness, vibration, and
δ (x)
like. In the beginning, these inputs produce low-amplitude
where waves which propagate in the downstream direction and either
grows or decays depending on the local flow conditions. It
5.84x eventually leads to turbulent flows. In the initial phase of
δ (x) = (11.170)
(Rex ) /2
1
transition, these linear waves retain their amplitudes but sub-
sequently becomes two-dimensional. These waves in their
The local skin friction coefficient cf,x is initial phase are called Tollmien–Schlichting waves, named
after their inventors, which are held responsible in initiating
0.685 the transition to turbulent flow via a subsequent characteristic
cf,x = 1
(11.171)
(Rex ) 2 three-dimensional structure formation.
For the boundary layer over the flat plate, transition is usu-
In addition, the total or average skin friction coefficient cf,L ally predicted in terms of critical Reynolds number.
will be given as
ρvxcr
Recr = (11.173)
L μ
cf,L = cf,x dx
where xcr is called the critical length. It is the axial length
0 measured from the leading edge of the plate at which transi-
tion begins. Alternatively, xcr may also be considered as the
Therefore,
distance till the flow remains laminar. For the flow over a a random variation with time and space. Even in a steady tur-
flat plate, the value of critical Reynolds number is 5 × 105 , bulent flow, the velocity at a given point will fluctuate with
whereas, for pipe flows the critical Reynolds number based time, whereas, even in uniform turbulent flows, instantaneous
on the inner diameter of the pipe is 2300. velocity at all points along a streamline would vary randomly.
We know that in laminar flows, the lateral movement of fluid
particles takes place due to negligibly small molecular dif-
11.11 Turbulent Boundary Layers
fusion. In turbulent flows, the lump of fluid particles move
The turbulence or turbulent flows can be described as a ran- laterally and longitudinally, similar to a vortex, often called
dom three-dimensional phenomenon in which the perturba- eddying motion. These eddies may change their shape, may
tion components are superimposed over mean flow, i.e., it stretch, and rotate or break into two or more eddies. In a turbu-
is the flow which has irregular fluctuations. By definition, a lent flow, the instantaneous flow parameters can be expressed
turbulent flow can be described as the flow which is three- as
dimensional, random, exhibiting multiplicity of scales, pos-
sessing vorticity and shows very high dissipation. Most of the u = u + u (11.174)

naturally occurring flows are turbulent in nature. Moreover, v=v+v (11.175)

the majority of boundary layers encountered in engineering w=w+w (11.176)
practice are turbulent over most of their length. Thus, the tur- p = p + p (11.177)
bulent boundary layer analysis is regarded as of greater fun-
damental importance than that of the laminar boundary layer. where u, v, w and p are instantaneous values; u, v, w and p
However, in contrast to laminar flows, there is no exact theory are time-averaged values; and u , v , w and p are fluctuating
available for turbulent flows over the flat plate. But there exist components.
many computational results of boundary layer equations using A turbulent flow field is called stationary when statistical
various empirical eddy viscosity models. Nevertheless, simi- flow parameters do not change with time. Alternatively, these
lar to laminar boundary layer velocity distribution assuming types of flow fields are referred to as quasi-steady (Fig. 11.9).
a suitable velocity profile for turbulent boundary layers, an In such a situation, even if the instantaneous velocity is chang-
integral analysis can be performed whose results are widely ing, an average value of velocity can be deduced. At a given
accepted. point, one can define
T
11.11.1 Mean Motion and Perturbations 1
u= udT (11.178)
2T −T
In contrast to laminar flows, a turbulent flow is always three-
dimensional in nature. This is because of the superimposi- The total time domain 2T should be chosen with utmost care.
tion It should be large enough as compared to time scale of small
of three-dimensional perturbation
velocity components
u , v , w on the mean flow U, V, W . The mean flow may eddies but it should be smaller than the time scale of larger
be one–, two– and three-dimensional but fluctuation compo- eddies. The time scale of an eddy can be defined as the time
nents are always three-dimensional. We can understand this taken by an eddy to pass past a given point in the flow field.
by an example. Let us consider a pitot-static probe, which It can also be observed that
is used in an experiment to calculate the flow velocity. We T
1
know that the frequency response of a pitot-static probe is u dT = 0 (11.179)
1 2T −T
low to about Hz, so there is a potential for significant loss
30
of fluctuating pressure signal and consequent attenuation of That is,the time averaging of fluctuation
velocity component

1 t
peak pressures. Thus, the manometer will measure only the
is zero u = t 0 u (t)dt = 0 . Hence, the quantification of
steady values of pressure and will neglect the perturbation turbulence is done in terms of root mean square (rms) values.
components. It also shows that the turbulent flows consist of That is
steady component of velocity independent of time over which

the perturbation components are superimposed.
urms = (u ) , vrms = (v ) , wrms = (w )
2 2 2
Furthermore, it has been revealed through various studies
that with increase of Reynolds number, certain frequencies (11.180)
and wavelengths are amplified in any flow that causes break-
down of disturbances into chaotic motion. Essentially, the ran- Also, the mean velocity (Vm ) of steady components is given
dom motion of these fluctuations are termed as turbulence. It by
is an irregular condition in which various flow quantities show
11.11 Turbulent Boundary Layers 273
0.5
Quasi−Steady State
u
0.4
u (m/s)
0.3
40 80 120 140 180
t (Seconds)
Fig. 11.9 Schematic diagram of a turbulent flow under stationary state

2 2 2 tion components of flow variables such as pressure, tempera-
Vm = U + V + W (11.181)
ture, velocity, etc., are superimposed on the mean flow. Using
urms vrms the method of time averaging, the instantaneous flow variable
Further, the intensity of turbulence is equal to or is defined as
Vm Vm
wrms
or . In quasi-unsteady or nonstationary flow situation as
Vm f = f + f (11.185)
shown in Fig. 11.10, such an average u taken over time will be
different. Therefore, for nonstationary flow situations, these
where time average component is denoted by f and the
time averages are uncommon.
perturbation component (i.e., deviation from mean or time-
Similarly, we can now consider a quasi-uniform flow. Here
averaged component) is denoted by f . In mathematical
again, even though the instantaneous velocity is changing with
notations, the time averaging can be depicted as
respect to space coordinates, the average value of velocity can
be defined as τ
1
f ≡ lim f x, y, z, t = ti + t dt
1 x τ →∞ T
u= udx (11.182) 0
2x −x
The above averaging can be extended to whole space as where ti is the initial or reference time and τ is the time suf-
ficiently larger but lesser than infinite. The above approach is
known as Reynolds averaging. Let us consider f1 and f2 are
1
u= udx (11.183) the two time and space coordinate-dependent functions such
∀ ∀
that
Using statistics, one can also define the ensemble average of
f1 = f1 + f1 (11.186)
velocity as
f2 = f2 + f2 (11.187)
N
1
< u(t) > = lim ui (t)dN (11.184) Then by applying Reynolds rules, the following relations can
N→∞ N i=1
be deduced.
In which, N is the corresponding points where u is measured
for an infinite number of macroscopic flows. It is important f1 + f2 = f1 + f2 (11.188)
to know the relationship between time and space averages.
According to ergodic theorem, the time and space averages cf1 = cf1 (11.189)
are same if the process is ergodic. Since, the turbulence is an
ergodic process, thus it is sufficient to deal with time averaging where c is a constant.
of flow parameters.
f1 .f2 = f1 .f2 (11.190)
11.11.1.1 Reynolds Rules of Averaging
We have seen that the turbulent flow is a three-dimensional, ∂f ∂f
= (11.191)
complex, and highly unsteady phenomena in which fluctua- ∂n ∂n
0.5
Quasi−Unsteady State
u (m/s) 0.4
0.3
40 80 120 140 180
t (Seconds)
Fig. 11.10 Schematic diagram of a turbulent flow under nonstationary state
Similarly, we can approximate that 11.11.2 Governing Equations for Turbulent

Flows
f1 .f2 = f1 .f2 + f1 .f2 (11.192)
We can use Reynolds averaging approach on the continuity
11.11.1.2 Isotropic and Homogeneous Turbulence equation (Eq. (11.37)) and x-component of momentum equa-
The turbulence is said to be isotropic if the statistical averaging tion (Eq. (11.43)) given below.
of flow parameters are direction independent. That is, they
∂u ∂v ∂w
remain invariant with respect to reflection or rotation of axes. + + =0
∂x ∂y ∂z
Mathematically, 2
∂u ∂u ∂u ∂u ∂p ∂ u ∂2u ∂2u
ρ +u +v +w =− +μ + +
∂t ∂x ∂y ∂z ∂x ∂x2 ∂y2 ∂z2
u2 = v2 = w2 (11.193)
Substituting u, v and w in place of f in Eq. (11.185), we obtain
For example, in wind tunnel studies, the turbulence behind
a honeycomb structure is nearly isotropic. However, in those u = u + u
situations where mean velocity shows gradient, the turbulence
v = v + v
will be non-isotropic or anisotropic. Flow in a pipe near the
boundary layer is anisotropic. In this case, w = w + w
u2 = v2 = w2 (11.194) Introducing above into Eq. (11.37), we get

Qualitatively, if the turbulence has same structure in all parts ∂u ∂v ∂w ∂u ∂v ∂w
+ + + + + =0
of the flow, it is referred to as homogeneous turbulence. If ∂x ∂y ∂z ∂x ∂y ∂z
there exists two points a and b in a flow field, then according (11.198)
to the homogeneity of turbulence,
Taking a time average of the whole equation above, one can

see that the value of time average fluctuation components will
u2 = u2 (11.195)
a b be zero. Thus, Eq. (11.198) reduces to

∂u ∂v ∂w
v2 = v2 (11.196) + + =0 (11.199)
a b ∂x ∂y ∂z
Further, subtracting Eq. (11.199) from Eq. (11.198), we get

w2 = w2 (11.197)
a b
∂u ∂v ∂w
+ + =0 (11.200)
∂x ∂y ∂z

Similarly, introducing instantaneous velocities (u, v, w) and ∂p ∂σ zx ∂ 2 τ zy ∂ 2 σ zz
=− +μ + + (11.205)
pressure (p) into x-component of momentum equation, we ∂z ∂x ∂y ∂z
get
Hence, the Reynolds stresses can be depicted in the form of
∂ u + u ∂ u + u ∂ u + u ∂ u + u
ρ + u + u + v + v + w + w a matrix as
∂t ∂x ∂y ∂z

∂ p+p

2
∂ u+u 2
∂ u+u
∂ u + u
2
⎛ ⎞
=−
∂x
+μ + + u 2 u v u w
∂x2 ∂y2 ∂z2 ⎜ ⎟
(11.201) −ρ⎝ u v v 2 v w ⎠ (11.206)
u w v w w 2
If we take the time averaging of each term in the equation
above, we get Note that, in the Reynolds stress matrix above, there are total
nine components of stresses with six distinct quantities.
∂ u + u ∂ u + u ∂ u + u ∂ u + u
ρ + u + u + v + v + w + w
∂t ∂x ∂y ∂z
11.11.2.1 Turbulent Boundary Layer Equations
∂ p+p
∂ 2 u + u ∂ 2 u + u ∂ 2 u + u For the two-dimensional turbulent boundary layer, only one
=− +μ + +
∂x ∂x2 ∂y2 ∂z2
of the Reynolds stresses need to be considered which is
(11.202)
−ρu v . Thus, for two-dimensional turbulent boundary lay-
While simplifying the equation above, one should note that ers, the time-averaged boundary layer equations are
even though the time average of fluctuation velocity com-
∂u ∂v
ponents are zero, i.e., u = v = w = 0. But the product of + =0 (11.207)
these perturbation components need not be zero, i.e., u v = 0. ∂x ∂y

Thus, on simplification we get ∂u ∂u ∂u ∂p ∂τ
ρ +u +v =− + (11.208)
∂t ∂x ∂y ∂x ∂y

∂u ∂u ∂u ∂u
ρ +u +v +w One of the main difficulty associated with turbulent flows
∂t ∂x ∂y ∂z
is that there is no way to derive the Reynolds stresses from
∂p ∂σ xx ∂ 2 τ xy ∂ 2 τ xz the fundamental principles other than solving the three-
=− +μ + + (11.203)
∂x ∂x ∂y ∂z dimensional unsteady forms of Navier–Stokes equations.
Thus, we must have semiempirical theories available for
where modeling the Reynolds stresses to solve Eqs. (11.207) and
(11.208). One such semiempirical theory was proposed by
∂u
− ρu2
σ xx = μ Prandtl, where the concept of eddy viscosity was introduced.
∂x
This theory is commonly known as Prandtl’s mixing length
∂u
τ xy = μ − ρu v hypothesis, which is discussed in the following section.
∂y
∂u
τ xz = μ − ρu w
∂z 11.11.3 Prandtl’s Mixing Length Hypothesis

The normal stress −ρu2 and shear stresses −ρu v , In 1925, Ludwig Prandtl proposed a turbulence model where

−ρu w are known as Reynolds stresses or turbulent he assumed that the eddies in turbulent flow move around
similar to the manner in which the molecules move in a gas.
stresses. Furthermore, one can obtain the time-averaged y-
In kinetic theory of gases, the dynamic viscosity is defined as
and z-components of momentum equations as
1
∂v ∂v ∂v ∂v μ= ρurms λ (11.209)
ρ +u +v +w 3
∂t ∂x ∂y ∂z

∂p ∂τ yx ∂ 2 σ yy ∂ 2 τ yz where urms is the root mean square velocity of the molecules,
=− +μ + + (11.204) and λ is the mean free path of the molecules. Similarly, Prandtl
∂y ∂x ∂y ∂z
postulated that the eddy viscosity (μt ) can be written as
∂w ∂w ∂w ∂w
ρ +u +v +w
∂t ∂x ∂y ∂z τ
μt = du
= ρlm u (11.210)
dy
1
where lm is the mixing length of the eddies and u is the
u = (u1 + u2 ) (11.214)
turbulence velocity along x axis in the xy plane. It should 2
be noted that the turbulent viscosity is not homogeneous (it or
varies in space). However, it is assumed to be isotropic (same
du
in all directions). This assumption is valid for a wide variety
u = lm (11.215)
of flows except flows involving large separation or swirl. dy
Similar to kinematic viscosity (υ), a kinematic turbulent
viscosity (υt ) can also be defined as According to mixing length hypothesis, the transverse turbu-

lent velocity fluctuation v is of the same order of magnitude
μt
υt =
ρ as u . Thus,

To understand Prandtl’s hypothesis further, let us refer to the
du
simplest case of a parallel flow (xy plane) in which the velocity v = lm (11.216)
dy
varies only from streamline to streamline. The direction of
mean flow (u) is assumed parallel to the x-axis and we have Physically, the mixing length (lm ) can be interpreted in the
following way. The mixing length is that distance in the
u = u(y) transverse direction which must be covered by the fluid lump
v=0 traveling with its original mean velocity in order to make the
difference between its velocity and the velocity into new lam-
Since, the fluid passes along the wall in turbulent motion, fluid ina equal to the mean transverse turbulent fluctuation
in tur-

particle coalesce into lumps which moves bodily and cling bulent flow. Further, the turbulent shear stress τ = −ρu v
together for a given traversed length, both in the longitudi- is given by
nal and in the transverse direction, retaining their momentum
parallel to x-axis. Let us assume that a fluid lump is displaced 2
du du du
from the layer A, possessing the velocity u(y1 − l) to a layer τ= ρlm
2 = ρl 2
(11.217)
dy dy m
dy
B of velocity u(y) (Fig. 11.11). The lump is displaced over
a distance lm , known as Prandtl’s mixing length. Since, the We should note that the negative sign in shear stress expres-
lump has retained its momentum and hence, the velocity of sion is omitted in Eq. (11.217). The reason behind it should
lump will be smaller than the velocity prevailing at B. There- be explored by the reader!
fore, the difference in velocities will be From Eq. (11.210) the eddy viscosity can be written as

u1 = u(y1 ) − u(y1 − l) (11.211)
2 du
μt = ρlm dy (11.218)
Using Taylor series to expand u(y1 − l) and neglecting higher
order terms, we have In Eq. (11.218), it is evident that unlike dynamic viscosity,
the eddy viscosity varies from point to point and depends on
du the nature of the flow. Thus, it is clear that eddy viscosity
u1 ≈ lm (11.212)
dy and consequently the mixing length are not the properties
of fluid and they are dependent upon turbulent fluctuations.
In the above transverse motion, we have v > 0. Similarly, Moreover, it has been shown through various correlations and
assume that a fluid lump arrives at B from a layer C possessing experimental studies that the value of mixing length in the
the velocity u(y1 + l), retaining its original momentum. In regime of isotropic turbulence can be given as
this motion, the velocity of lump will be more than the velocity
prevailing at layer B. Therefore, the change in velocity will lm = χy (11.219)
be
du where y is the distance from the wall, and χ is known as von

u2 ≈ lm (11.213) Karman constant. In general, χ = 0.4 and thus lm = 0.4y.
dy
We should that in this case v < 0. The time averaging of u1

and u2 can be regarded as the longitudinal turbulent velocity

fluctuation u . Thus,
y
u(y)
u(y+1)
1
C
l m
u(y )
1
B
lm
u(y −1)
1
A
Fig. 11.11 Schematic diagram of fluid lump movement
η
Ua
Turbulent shear stresses
Turbulent μdu/dy
shear Viscous shear
Shaded area in separation
ρu v viscous shear
Buffer Us
layer
Wall 0
Laminar sub−layer τxy u
(δ s )
Fig. 11.12 Different regions in a turbulent boundary layer over the wall
11.11.4 Regimes in Turbulent Boundary Layer thus this outer layer is known as the outer boundary layer or
the fully turbulent layer.
Turbulent boundary layer can be classified into three different Consider if τw represents the wall shear stress, and δ and
regimes (Fig. 11.12). The bottommost layer, adjacent to wall Ua , respectively, are the boundary layer thickness and veloc-
shows the flow characteristics similar to the laminar flow and ity at the edge of boundary layer (y = δ). In 1930, Prandtl
thus known as laminar sub-layer. Basically, in close proxim- proposed that u must be independent of δ. Thus, one can rep-
ity to the wall, the velocity gradients are large and the viscous resent u as
shear mechanism is strong enough to transmit the shear stress
to the boundary, consequently, a very thin layer adjacent to u = f (μ, τw , ρ, y) (11.220)
the boundary is in laminar state even when the flow in the rest
of the boundary layer is turbulent. Although, the sub-layer From dimensional analysis, we have
is very thin in comparison to overall boundary layer thick-
1 1
ness, but its presence is vital in deciding whether a surface is y τw 2 τw 2
hydraulically rough or smooth. In addition, viscous sub-layer u = 1 (11.221)
ν ρ ρ
has strong damping effects on turbulence so that in the near
vicinity of the wall, the viscous shear stresses are exceedingly 1
τw
= u∗ is
2
larger than the Reynolds shear stresses. The layer beyond it is where 1 is a nondimensional function. Also, ρ
the transition or buffer layer, where the magnitudes of viscous
called the friction or shear velocity as its dimension is ms−1 .
stresses and Reynolds stresses are almost equal. The outer-
However, it is not a flow velocity and shear velocity is used
most part of boundary layer blended with the buffer layer has
to make the velocity nondimensional. Thus, we can write
the characteristics similar to that of the free shear layer, and

u yu∗ logarithmic velocity law:
= 1
u∗ ν ∗
u 1 yu
or ∗
= ln +C
u χ ν

u yu∗
u+ = = 1 (11.222) or
u∗ ν
1 +
u+ = ln y + C
The equation above is found to be valid in the range,
∗
χ
0 ≤ yuν ≤ 5. Thus, the approximate thickness of the viscous
sub-layer is For smooth wall flows, the value of constants χ and C, respec-
tively, are estimated from the experiments as 0.4 and 5.0.
5ν Therefore,
δsub-layer = (11.223)
u∗
u+ = 2.5 ln y+ + 5.0 (11.226)
Equation (11.223) shows that by increasing mean velocity the
boundary layer thickness decreases, i.e., boundary layer gets Equation (11.226) is called the logarithmic buffer layer, which
thinner, consequently, the velocity profile becomes nearly flat. ∗
is applicable in the range 5 < yuν < 50.
Also, for high Reynolds number (or low viscosity), the veloc-
ity distribution becomes more uniform. From the experiments,
it has been established that 11.11.5 Skin Friction Coefficient for Turbulent
∗ Boundary Layer over a Flat Plate
yu yu∗
1 =
ν ν
The calculation of skin friction drag over lifting bodies is
Therefore, the equation just above Eq. (11.223) can be written essentially similar to skin friction drag estimation over flat
as plate. The zero pressure gradient over the flat plate makes it
easier to consider for analysis. Moreover, as long as there is
u yu∗ no separation over bodies, even though the pressure gradi-
u+ = ∗
= = y+ (11.224)
u ν ent differs from zero but the skin friction coefficient remains
invariant. Let us recall Eq. (11.76), which is valid for both
The parameter uν∗ has the dimension of length and thus it laminar and turbulent boundary layers.
can be used to obtain dimensionless distance y from the wall
surface. dθ Cf
In 1933, Prandtl further deduced that in outer layer u is =
dx 2
independent of μ but its deviation from centerline velocity,
i.e., (Umax − u) should depend on δ and other flow parame- But skin friction coefficient (Cfx ) is defined by
ters. That is
τw
Cfx =
0.5ρUa2
(Umax − u)outer-layer = g (τw , δ, ρ, y)
Equation (11.76) can be written as,
Again from dimensional analysis, the equation above can be
deduced in the following form: dθ
Cfx = 2 (11.227)
y dx
(Umax − u)outer-layer
= 2 (11.225)
u∗ δ Let us assume the velocity distribution2 within the boundary
layer as
The quantity (Umax − u)outer-layer is called the velocity defect
or flow retardation. Also, Eq. (11.225) is known as velocity 1/7
u y
defect law. Equation (11.225) is valid whenever y+ is more = (11.228)
∗ Ua δ (x)
than 50, i.e., yuν > 50.
In 1937, C. B. Millikan found that the wall law
(Eq. (11.222)) and velocity defect law (Eq. (11.225))
although have different forms but they overlap smoothly 2 The turbulent velocity profile is much steeper than the laminar boundary
in buffer layer if the velocity profile obeys the following layer flows. Thus, fluid particles in turbulent flows possess high kinetic
energy than the laminar flows.
which is valid for Re 105 . The velocity profile given in Let us further assume for convenience that, the turbulent
Eq. (11.228) provides satisfactory results for the most of the boundary layer starts developing from the leading edge of
boundary layer, however, it fails to satisfy the boundary con- the plate itself so that we can use the following boundary
ditions at the wall. That is condition to evaluate the integration constant in the equation
above.

∂u 1 6 1
at y = 0 : = (y)− 7 Ua δ − 7 → ∞
∂y y=0 7 at x = 0 : δ (x) = 0 ⇒ C1 = 0
However, in realty, there exists a viscous sub-layer where the Thus,

velocity variation is assumed to be linear and it is tangen-
1
tial to one-seventh velocity profile where viscous sub-layer ν 4
5/4
blends with turbulent part of the boundary layer. Further, the δ = 0.2892x
Ua
momentum thickness (θ) can be estimated by
or
δ 1/7 1/7
y y 1
θ= 1− dy δ ν 5
δ (x) δ (x) = 0.37
0 x xUa
or or
7 δ 1
θ= δ (x) (11.229) = 0.37 [Rex ]− 5
72 x
Thus, Eqs. (11.227) and (11.229) leads to, where Rex = Uνa x . From Eq. (11.231), the local skin friction
coefficient becomes
dθ 14 dδ (x)
Cfx = 2 = (11.230)
dx 72 dx 0.059
Cfx = 1
(11.232)
For the calculations of wall shear stress, we use the following [Rex ]− 5
experimental relation valid for the Reynolds number below
Further, the total drag on the flat plate is calculated by inte-
than 107 , i.e., Re < 107 .
grating the shear stress over the plate area (A).
1/4
τw ν
= 0.0225 0.0361ρUa2 (l.w)
ρUa2 δ (x) Ua D= τw dA = (11.233)
(ReL )0.2
A
Thus, the local skin friction coefficient becomes
where l and w are the length and width of the plate, respec-
1/4
τw ν tively. Using Eq. (11.233), the average friction coefficient will
Cfx = = 0.045 (11.231)
0.5ρUa2 δ (x) Ua be,
From Eq. (11.230) and the equation above, we get D

Cf =
0.5ρUa (l.w)
1/4
7 dδ (x) ν
= 0.0225 Thus,
72 dx δ (x) Ua
or 0.073
Cf = (11.234)
(ReL )0.2
1/4
dδ ν
δ /4 = 0.2314
1
dx Ua The results obtained by using Eq. (11.234) are found to be

accurate if Re < 2 × 106 . Based on the experimental results
Upon integration, of Nikuradse, Schlichting obtained a semiempirical relation
for the average friction coefficient given by
1/4
5/4 ν
δ = 0.2892x + C1
Ua
0.455 δ
Cf = (11.235) u u
(log Re)2.58 θ= 1− dy
Ua Ua
0
Note that, to obtain Eq. (11.235), a fully turbulent bound-
ary layer over the flat plate is assumed to be developed right Kinetic energy thickness (δ ∗∗ ) is the distance measured per-
from the leading edge of the flat plate. However, in realty, the pendicular to surface of the solid body through, which the
turbulent boundary layer starts developing after the transition boundary should be displaced to compensate for the reduc-
region, i.e., at some distance downstream of the leading edge. tion in kinetic energy of the flowing fluid due to boundary
Because of this reason, Prandtl suggested to consider the fol- layer formation.
lowing relation for the mixed boundary layer over a thin flat
plate. δ 2
u u
0.455 α δ ∗∗ = 1−
Cf = − Ua Ua
0
(log Re)2.58 Re
The Reynolds number is an important similarity parameter for
where α depends on the critical Reynolds number (Recr ) at
viscous flows, defined as the ratio of inertia force to viscous
which boundary layer transition occurs. It is found that for
force.
Recr = 5 × 105 , the value of α is 1700. Thus, ρvL
Re =
0.455 1700 μ
Cf = − (11.236)
(log Re) 2.58 Re
The ratio of inertia force to elastic force is known as the Mach
number. From order of magnitude of analysis, the Mach num-
This result is applicable for a smooth flat plate.
ber is defined as
v
M=
a
Summary That is, the Mach number can be defined as the ratio of local
flow speed to the speed of sound.
A boundary layer is a thin viscous layer adjacent to a solid sur-
For very thin boundary layers, the Prandtl boundary layer
face, where the fluid viscosity predominates. Due to viscous
equations are
effects, fluid sticks to the solid surface such that the velocity
of fluid layer adjacent to the surface is equal to the velocity of ∂u ∂v
the surface itself. In other words, the relative velocity of the + =0
∂x ∂y
fluid with respect to a solid surface is zero. This condition is
∂u ∂u ∂p ∂2u
commonly referred to as no-slip condition. u +v =− +μ 2
The extent of viscous effects on a surface is measured in ∂x ∂y ∂x ∂y
terms of boundary layer thickness (δ), defined as the distance ∂p
=0
normal to the surface such that the local flow velocity (u) at ∂y
that distance is 99% of the freestream velocity (Ua ). That is
The boundary conditions that satisfies the above equations are
at y = 0; u (y) = 0

at y = δ; u(y) = 0.99Ua
at y = 0 : u = 0, v = 0

The displacement thickness (δ ∗ ) is defined as the distance at y = δ : u = Ua , v = 0
perpendicular to the boundary, by which the freestream is
displaced due to the formation of boundary layer. The momentum integral equation suggested by von Karman
is
δ dθ θ dUa C
f
∗ u + H + 2 − Ma2 =
δ = 1− dy dx Ua dx 2
Ua
0
where
The momentum thickness (θ) is defined as the distance
through which the boundary layer must be displaced to com-
pensate the reduction in momentum of the flowing fluid due
to boundary layer formation.
Summary 281
τw τw 0.664
Cf = 1 ≡ coeffcient of skin friction Cf = =√
ρ U2
2 a a 1
ρU 2 Rex
2 a
δ∗
H= ≡ shape factor
θ For compressible flow over the flat plate, the boundary layer

dθ
≡ inertia term thickness δcomp can be defined as
dx
θ dUa
H + 2 − Ma2 ≡ pressure gradient term 5.0x Tw
Ua dx δcomp = √ G Me , Pr,
Rex Te
Cf
≡ wall skin friction term
2 The boundary conditions are:
The energy integral equation derived by K. Wieghardt (1948)
for laminar boundary layer is given by at y = 0; T = Tw (11.237)

∂δ ∗ ∂ 2 ∂ 3 ∗∗ 2ε at y = d; T = Te (11.238)
Ua2 + Ua θ + Ua δ =
∂t ∂t ∂x ρ
where Tw = wall temperature, Te = temperature at the
∞ ∂u 2 boundary layer edge, d = local distance from the wall, and
where ε = μ 0 ∂y dy; ε is the energy dissipation rate
function G can be obtained through the numerical simulation
due to viscosity across the boundary layer. δ ∗ , θ and δ ∗∗ , of
respectively, are the displacement thickness, the momentum the problem. Furthermore, the skin friction coefficient
Cf,comp can be written as
thickness and the energy thickness of the boundary layer.
A laminar boundary layer is the one where the fluid layers
1.328 Tw
are assumed to slide over one another smoothly. That is, in a Cf = √ F Me , Pr,
Rec Te
laminar boundary layer the fluid layers, also called laminas,
slide over adjacent layers such that there exists only the micro- For a uniform flow past the flat plate, boundary layer starts
scopic exchange of matter (mass, momentum, and energy) growing as laminar flow beginning at the leading edge of the
among the layers. However, if we try to observe the laminar flat plate. However, due to instabilities this laminar flow turns
flows on macroscopic scale, we will not feel any exchange of into transition flow which subsequently turns into turbulent
matter taking place between the fluid layers. Moreover, any flow. The instabilities introduced in a laminar flow are ampli-
small disturbances or instabilities are quickly damped out by fied and result in flows which are orderly in nature in some
the viscous forces that resist the relative motion of adjoining parts, but also shows temporarily (and/or spatially) irregular
layers. Thus, a laminar boundary layer flow is an orderly flow fluctuations of all the flow quantities in other parts, one may
in which fluid elements move in an orderly manner such that say of a transitional flow state. In this state, intermittent lam-
the transverse exchange of momentum is negligibly small, inar and turbulent behavior is observed; some phases occur
however, the axial exchange of momentum may be consid- in which the flow behaves as laminar, and in other phases
erable. In contrast, the flow in a turbulent boundary layer the flow exhibits turbulent characteristics. Since the transi-
is three-dimensional random phenomena, exhibiting multi- tion phenomena is exceedingly complex and thus there is no
plicity of scales, possessing vorticity, and showing very high accurate theory available to predict the process. It is the area
dissipation. of active research.
For the laminar flow past a flat plate, the Blasius solutions For the boundary layer over the flat plate, transition is usu-
are ally predicted in terms of critical Reynolds number.
δ 5.0
=√
x Rex ρvxcr
Recr =
μ

νx
δ ∗ = 1.7208 where xcr is called the critical length. It is the axial length
Ua
measured from the leading edge of the plate at which transition
begins. For the flow over a flat plate, the value of critical
νx
θ = 0.664 Reynolds number is 5×105 whereas, for pipe flows the critical
Ua Reynolds number based on the inner diameter of the pipe is
2300.
∗∗ νx
δ = 1.0444
Ua
The turbulence or turbulent flows can be described as a Exercises

random three-dimensional phenomenon in which the pertur-
bation components are superimposed over mean flow, i.e., it Descriptive Type Questions
is the flow which has irregular fluctuations. By definition, a
turbulent flow can be described as the flow which is three- 1. The flat plate is kept at zero incidence in the uniform
dimensional, random, exhibiting multiplicity of scales, pos- flow field of 50 ms−1 . If the chord of the plate is 0.7 m,
sessing vorticity and shows very high dissipation. Most of the calculate the displacement thickness at the trailing edge.
naturally occurring flows are turbulent in nature. 2. Consider an adiabatic flat plate of chord 0.15 m which is
In a turbulent flow, the instantaneous flow parameters can immersed in a Mach 4 laminar flow. Calculate the skin
be expressed as friction drag on the plate per unit span.
3. An incompressible flow has the velocity components
u = u + u given by u = 5y and x = 8x2 ms−1 . Calculate the shear
v = v + v stress at x = 3 m and y = 2 m on the top face of flat plate
normal to x-axis.
w = w + w
4. The freestream velocity of a fluid over a flat plate
p = p + p is 7 ms−1 and the kinematic viscosity of the flow is
ν = 1.4 × 10−5 m2 s−1 . At a certain x location, the local
where u, v, w and p are instantaneous values; u, v, w and p velocity at a distance y = 5×10−3 m is 6 ms−1 . Deter-
are time-averaged values; and u , v , w and p are fluctuating mine the the skin friction coefficient at this particular
components. location.
The turbulence is said to be isotropic if the statistical aver- 5. Consider the water flow past a flat plate, as sketched in
aging of flow parameters are direction independent. That is, Fig. 11.13.
they remain invariant with respect to reflection or rotation of Suppose the boundary layer thickness is 6 mm at a dis-
axes. Mathematically, tance x = 0.5 m, measured from the leading edge of the
plate. Determine whether the boundary layer is laminar
u2 = v2 = w2 or turbulent?
6. For an incompressible boundary layer over a flat plate, if
In those situations where mean velocity shows the gradient, 3
the velocity distribution is given by Uua = 25 yδ − 21 yδ ,
turbulence will be non-isotropic or anisotropic. In this case, where δ is the boundary layer thickness. The freestream
velocity is 15 ms−1 and the dynamic viscosity of the fluid
u2 = v2 = w2 is 1.8×10−5 kg m−1 s−1 . Calculate the wall shear stress at
a streamwise location where the boundary layer thickness
For two-dimensional turbulent boundary layers, the time- is 5 mm.
averaged boundary layer equations are 7. Suppose a two-dimensional boundary layer is developed
over a flat plate under zero pressure gradient. Let ρa is the
∂u ∂v
+ =0 fluid density and Ua is the freestream velocity. Thus for
∂x ∂y boundary layer thickness δ (x) at a distance x measured

∂u ∂u ∂u ∂p ∂τ from the leading edge in the streamwise direction, show
ρ +u +v =− +
∂t ∂x ∂y ∂x ∂y that the wall shear stress (τw ) is given by
Turbulent boundary layer can be classified into three differ- δ

ent regimes. The bottommost layer adjacent to wall shows d
τw = ρ u (Ua − u) dy
the flow characteristics similar to the laminar flow and thus dx
0
known as laminar sub-layer. The layer beyond it is the transi-
tion or buffer layer, where the magnitudes of viscous stresses where u is the x-component of velocity and y is the dis-
and Reynolds stresses are almost equal. The outermost part tance normal to plate.
of boundary layer blended with the buffer layer has the char- 8. Consider an incompressible boundary layer developed
acteristics similar to that of the free shear layer, and thus this over a flat plate. If the velocity profile is given by
outer layer is known as the outer boundary layer or the fully
3 y

1 y 2
Ua = 2 δ − 2 δ . Determine the expressions for (a)
u
turbulent layer. the displacement thickness, (b) the momentum thickness,
(c) the boundary layer shape factor, and (d) the energy
thickness.
Exercises 283
Ua
Ua
y δ = 6 mm
u(y)
x
x = 0.5 m
Fig. 11.13 Water flow past a flat plate
9. A thin flat plate with sharp leading edge of length 1.5 m (c) 1.25
and breadth 3 m, is immersed to a uniform airstream (d) 1.33
of velocity 2.5 ms−1 . Find the drag on the one side
of the plate. For air, ρ = 1.23 kg m−3 and ν = 1.46 × 4. Consider an airstream of velocity 100 ms−1 flows
10−5 m2 s−1 . past a solid body. If the kinematic viscosity of air
10. Consider two infinitely long parallel plates separated by is 1.48 × 10−5 m2 s−1 and the Reynolds number is
0.02 m. The bottom plate is moving constantly at a speed 6.756 ×106 , the characteristic length of the solid body is
of 25 ms−1 in the direction of the plate. If an incom-
pressible viscous air of temperature 315 K is allowed in (a) 0.5 m
between these plates, find the shear stress on the top and (b) 0.75 m
bottom plate. Assume that, there is no pressure gradient (c) 1.0 m
exists in the direction of flow. (d) 1.25 m
5. The Reynolds number for an incompressible laminar flow

over a flat plate is 2000. The distance from the leading
edge, where the boundary layer thickness δ = 10 mm, is
1. Consider a flat plate of 15 cm by 10 cm, placed at right
(a) 46 mm
angles to an incoming freestream of velocity 25 ms−1 .
(b) 89 mm
If the freestream density is 1.225 kg m−3 , the resistance
(c) 93 mm
offered by the plate is
(d) 114 mm
(a) 4.4 N
6. The displacement thickness of the boundary layer in an
(b) 6.8 N
incompressible laminar flow over a flat plate is 10 mm at
(c) 8.2 N
a distance of 100 mm downstream of the leading edge. If
(d) 10.3 N
the kinematic viscosity of the flow is 1.48 ×10−5 m2 s−1 ,
2. Due to fluid viscosity, the velocity of fluid layer adjacent the freestream velocity is
to a surface is equal to the velocity of surface itself. This
(a) 43.82 ms−1
is referred to as
(b) 54.67 ms−1
(a) slip condition (c) 67.88 ms−1
(b) no-slip condition (d) 80.13 ms−1
(c) saddle point
7. Consider an incompressible laminar flow over a flat plate.
(d) half-saddle point
If the freestream velocity and kinematic viscosity of the
3. Consider a turbulent flow, in which the displacement flow are 50 ms−1 and 1.4×10−5 m2 s−1 , respectively, then
thickness, momentum thickness and energy thickness are at a distance 120 mm from the leading edge, momentum
10, 7.5 and 6 mm, respectively. The shape factor is thickness is
(a) 0.75 (a) 0.88 mm

(b) 0.80 (b) 1.52 mm
(c) 2.67 mm
(d) 3.84 mm Keys
8. Consider an incompressible laminar flow of velocity 1. (b)

50 ms−1 over a flat plate. At a distance of 309.7 mm from 2. (b)
the leading edge, the energy thickness is found to be 3. (d)
10 mm. The kinematic viscosity of the flow is 4. (c)
5. (b)
(a) 1.25 ×10−5 m2 s−1 6. (a)
(b) 1.48 ×10−5 m2 s−1 7. (d)
(c) 2.24 ×10−5 m2 s−1 8. (b)
(d) 4.28 ×10−5 m2 s−1 9. (b)
10. (c)
9. For an incompressible laminar flow over a flat plate, if the
Reynolds number is 1000, the drag coefficient is
(a) 0.02
(b) 0.04
(c) 0.2
(d) 0.4
10. For the compressible flow over a flat plate at low Reynolds
number (ReL ≈ 2000), the average skin friction coeffi-
cient will be equal to
(a) 0.01
(b) 0.02
(c) 0.03
(d) 0.04
Wind Tunnels
12
The wind tunnels are the essential designing instruments in 12.2.1 Mach Number Regimes
aerodynamics research. They are utilized to study the effect
of airflow past a solid object. Despite of great advancements Based on the desired flow speed in the test section, the wind
in computational fluid dynamics techniques in recent years, tunnels are divided into:
the data obtained from wind tunnel tests stay basic to control
detail design decisions for several engineering applications. • Subsonic wind tunnel (MTS < 0.8).
This chapter reviews diverse kinds of broadly useful wind • Transonic wind tunnel (0.8 < MTS < 1.2).
tunnels, characterized in view of the test area flow speeds. • Supersonic wind tunnel (1.2 < MTS < 5).
Plus, some unique reason wind tunnels are additionally talked • Hypersonic wind tunnel (MTS > 5).
about.
The classification based on Mach number is caused due to
compressibility effects. We know that for M < 0.3, the com-
12.1 Introduction pressibility effects can be ignored. However, for Mach num-
bers higher than this obtained in transonic and supersonic
A wind tunnel is the device that provides airstreams flow- tunnels, the flow must be assumed to be compressible. The
ing under controlled conditions so that models of interest can fluid compressibility has serious implications on the tunnel
be tested using them. Basically, they are used to test scale design; the test section has the minimum area in a subsonic
models of aircraft and spacecraft, however, some wind tun- tunnel whereas, the throat is the minimum area location for a
nels are big enough to hold the full-size (prototype) vehicles. supersonic wind tunnel.
The airstream past an object makes it seem like the object is
really flying. From an operational point of view, wind tun-
nels are generally classified as low-speed, high-speed, and 12.2.2 Tunnel Design
special-purpose tunnels. In this chapter, we will restrict our
discussion to wind tunnel design and operation only. For more The wind tunnels are also designated by the geometry of the
information on measurements and instrumentation, one can tunnel. When a tunnel is open at both ends and draws fresh air
refer to advanced texts such as Barlow et al. [1999] and Pereira from the ambient, it is referred to as open-circuit type shown
[2011]. in Fig. 12.1.
On the other hand, when a tunnel forms a closed circuit
12.2 Types of Wind Tunnels such that same fluid recirculates in a predetermined path
through the test section, it is termed as closed-circuit wind
The wind tunnels are designed for a specific purpose and tunnel (Fig. 12.2). These configurations are commonly used
hence there are several ways to classify them. Following in subsonic tunnels. Both open-circuit or closed-circuit tun-
are the most common ways in which the wind tunnels are nels may operate with either open test section walls or closed
classified. test section walls.

286 12 Wind Tunnels
Honeycomb Wire screen
Contraction cone
Air flow
Effuser Fan unit

Test section
Diffuser
Fig. 12.1 Schematic diagram of an open-circuit low-speed wind tunnel
Driving unit
Guide vanes
Test section
Fig. 12.2 Schematic diagram of a closed-circuit low-speed wind tunnel
Further, all these tunnels may be subclassified as inter- can be used to test the models at high Reynolds numbers in
mittent type or as continuous operation tunnels with longer transonic flows.
testing time availability. A blowdown tunnel is of intermit-
tent type, suitable for establishing supersonic flows in the test
section. It draws air from a high-pressure reservoir located 12.2.4 Special Purpose
upstream of the test section and discharges to a low-pressure
tank. A variation of blowdown tunnel is used for hypersonic The wind tunnels are also designated as special purpose for
flows, referred to as shock tube. The testing time available in which they are designed and built. They have layouts alto-
both supersonic and hypersonic tunnels is very less as com- gether different than that of low- or high-speed wind tunnels.
pared to their subsonic counterpart. Following are some typical examples of this type:
1. Spinning tunnels—since aircraft have a tendency to go to

12.2.3 Operational Fluid spin mode once it stalls, thus these facilities are used to
investigate those phenomena.
The tunnels can also be labeled based on the type of fluid used 2. Stability tunnels—sometimes, it is imperative to study the
in tunnel operation. For low-speed aircraft testing, air is com- stability of an aircraft and hence, these tunnels are designed
monly used as working medium. For visualizing compres- in such a way to allow the model to move freely within the
sion waves in the high-speed aircraft, studying flow patterns test section.
around submarines and underwater vehicles, the water is used 3. Low-density tunnels—to accurately simulate the high-
as operational fluid. Due to high-temperature rise in hyper- temperature effects at hypersonic Mach numbers or rar-
sonic (or high enthalpy) facilities,more stable gases such as efied state of gases in outer layer of the atmosphere, certain
nitrogen or helium have been used. Similarly, liquid nitrogen low-density tunnels are designed.
12.3 Experimental Models and Similitude 287
12.3 Experimental Models and Similitude the elasticity of the medium as well as the gravity. In order to
have a same dynamic pressure ratio at corresponding points
In order to obtain an accurate quantitative data from a model in both model and prototype, the ratios of various forces must
study, the physics behind model and prototype flow system be the same at these points. These forces are as follows:
operating under different conditions should be the same. The
similarity between two flow systems is established by equat- • Inertia force (Fi ).
ing the ratio of specified physical parameters calculated in • Viscous force (Fv).
each system. If the specified physical quantity is geometric • Gravity force Fg .
dimension, the similarity is called geometric similarity; if the • Pressure force Fp .
quantity is related to motion, the similarity is called kinematic • Compressibility force (Fc ).
similarity; and if the quantity denotes the force, then the simi-
larity is termed as dynamic similarity. Two fluid flows are said The inertia force is generally taken as the common one to
to be similar only if these three similarities exist together. describe these ratios. The ratio of inertia force to viscous force
is known as the Reynolds number (Re). Similarly, the ratio
of inertia force to pressure force is called the Euler number
12.3.1 Geometric Similarity (Eu). In addition, the ratio of inertia force to compressibility
force and the ratio of inertia force to gravitational force are
This similarity refers to the similarity of geometrical shape. called the Mach number (M) and the Froude number (Fr),
In two systems, if the ratio of any length in one system to the respectively.
corresponding length in other system is constant everywhere,
they are called geometrically similar systems. The ratio of Fi
Re = (12.1)
lengths is generally termed as scale factor. Clearly, these sys- Fv
tems have similar shapes but differ in size. The full-size actual Fi
Eu = (12.2)
systems are called the prototypes, while the scale size systems Fp
are referred to as models. However, it is not necessary that the Fi
model should always be smaller than the prototype. In some M= (12.3)
Fc
cases such as in microfludics, the model may be even bigger Fi
in size than the prototype. Also, it is not mandatory to use the Fr = (12.4)
Fg
same fluid with both the model and the prototype.
Geometrical similarity necessitates the similarity of sur- Note that, not all nondimensional parameters are matched
face roughness, i.e., if the model is one-tenth the size of proto- simultaneously in every experiments. It depends upon the
type in every linear dimension, then the roughness projections specific experiment and the types of forces generated. For
will also be in the same ratio. example, in a low-speed flow regime where the effect of
gravitational force is negligibly small, the simulation of
Reynolds number is important to ensure dynamic similarity,
12.3.2 Kinematic Similarity
while the similarity of Mach number is not required. In
contrast, in high-speed flows, the Mach number similarity is
This similarity is generally referred to as kinematic similitude.
necessary.
It is essentially the similarity of motion and since the motion is
described by both distance and time, the kinematic similarity
refers to the similarity of lengths (geometrical similarity) and 12.4 Subsonic Wind Tunnels
similarity of time intervals. Clearly, the kinematic similarity
ensures the geometrical similarity, however, the vice versa In Sect. 12.2.2, we have seen that the low-speed wind tunnels
is not necessarily true. In addition, when the fluid flows are are either open-circuit type or closed-circuit type. A typi-
kinematic similar, the streamline patterns are geometrically cal layout of an open-circuit low-speed tunnel is shown in
similar at corresponding instants. Fig. 12.1. A general utility low-speed wind tunnel has four
major components: the effuser, the test section, the diffuser,
and the driving unit.
12.3.3 Dynamic Similarity
It is essentially the similarity of forces, i.e., the ratio of 12.4.1 Effuser or Contraction Cone
dynamic pressures at corresponding points must be a invari-
ant. A body moving through the fluid is subjected to several It is a converging duct located upstream of the test section
forces and moments caused due to fluid viscosity, its inertia, in which the fluid gets accelerated from rest (or from very
288 12 Wind Tunnels

low speed) drawn at the upstream end of it to the required Dt 1 2
1
It = = u + v2 + w2 (12.7)
conditions at the test section. Thus, the effuser is also referred U U 3
to as contraction cone (Fig. 12.3). The contraction ratio (ac )
of an effuser is defined by the Eq. (12.5). where u is RMS (Root Mean Square) value of the time aver-
aged perturbation component of velocity along the x-axis over
Cross-sectional area at entry to contraction cone the time interval t. That is
ac =
Cross-sectional area at exit to contraction cone

(12.5) t
1
u = u2 dt = u2 (12.8)
The ac usually varies between 4 and 20 for a conventional t
0
subsonic wind tunnel. In addition to increase in flow speed,
the effuser brings down the swirl and eddies and produce a Similarly, v and w are the RMS values of fluctuation compo-
uniform flow at the exit. For this purpose, the honeycomb nent of velocities along y- and z-axes. In wind tunnel studies,
structures and wire gauze (screens) are employed in the con- the turbulence is generally assumed to be isotropic in nature,
traction cone. In order to establish uniform and unidirectional i.e., the time averaged perturbation component of velocity is
flow in the test section, we must overcome flow irregularities, same in all coordinate directions.
which is classified into the following types.
u2 = v2 = w2
12.4.1.1 Spatial Velocity Variations
The mean velocity variations over a cross section are elim- In this case, the x-component of perturbation velocity alone
inated by transferring excess total head from the regions of can be used to define the intensity of turbulence. Thus,
high velocity to those of low velocity. Eq. (12.7) will take the following form:
12.4.1.2 Flow Swirl u2

It = (12.9)
The flow may rotate about an axis resulting in the variation U
of direction of flow. The honeycomb structures are used to
In a wind tunnel, the turbulence in central region outside the
reduce the swirl.
boundary layer is nearly isotropic. However, behind the hon-
eycomb structure in the contractor of a wind tunnel, the tur-
12.4.1.3 Low-Frequency Pressure Oscillations bulence damps out very slowly. Thus, they are assumed to be
A surge is a moving wave front which results in a rapidly homogeneous.1
varied unsteady flow condition. The surges are created when
the velocity of a fluid suddenly changes and becomes unsteady 12.4.1.5 Effects of Honeycomb on Turbulence
or transient. The honeycomb structures are either natural or artificial, hav-
ing a geometry that requires minimum material to construct
12.4.1.4 Turbulence and consequently, it has minimum weight. It is sometimes
The deviation from the mean velocity occurs to some extent in called as flow straightener, which is used to minimize lateral
the test sections of all the wind tunnels. These deviations are turbulence caused by swirling motion in the airstream dur-
due to perturbation components of velocities, u , v , and w in ing entry to the contraction cone. The honeycomb cells may
three coordinate directions exhibit large fluctuations. These have circular, square, or hexagonal shapes of cross section as
deviations are measured in terms of degree of turbulence (Dt ) shown in Fig. 12.4.
or Intensity of turbulence (It ) of the wind tunnel. From Table 12.1, it can be seen that the hexagonal cells
Degree or level of turbulence is defined as the square root of incur minimum pressure loss. Thus, it is most efficient in elim-
the arithmetic mean of the mean square values of perturbation inating large-scale eddies. Further, besides the cross-sectional
components of velocities along x-, y-, and z-directions. shape, a major factor in choosing the correct cell size is the
ratio of cell length (Lc ) to cell diameter (Dc ). That is
1 2
Dt = u + v2 + w2 (12.6)
3
1 Turbulence has the same structure quantitatively
Intensity of turbulence (It ) is defined as the ratio throughout
the flow
of degree of field. That is, the root mean square values of u = u 2 , v = v 2
turbulence (D) and the freestream velocity U . Mathemati-
cally and w = w 2 can all be different, but each value must be constant
over the entire turbulent field.
12.4 Subsonic Wind Tunnels 289
v
1 A1
Control Volume A2
v
2
Fig. 12.3 Schematic diagram of a typical contraction cone
Circular cells Square cells Hexagonal cells
Fig. 12.4 Honeycomb structures with different cell cross sections
Lc to augment the flow uniformity; by imposing static pressure

β= (12.10)
Dc drop as well as reducing turbulence levels. The smaller size of
screen meshes breaks down the large eddies into smaller ones
• If β < 1, the honeycomb cells of low-length ratio are used which subsequently decays in the wake of screen. Further,
as screens to obtain uniform profile and to reduce the tur- multiple screens are found to be effective in bringing down
bulence intensity. the yaw angles to 10◦ . Also, when a number of screens are
• If β > 1, the honeycomb cells of high-length ratio are used employed, a considerable reduction in turbulence intensity
in eliminating lateral turbulence and eddies of the flow. has been observed.
However, the major drawback associated with this method It is well known that the size of eddies depends upon the
is the introduction of high turbulence intensity in the test flow Reynolds number (Re) and they are practically absent
section. when Red = Uνa d < 60 (where d is the wire diameter, Ua is the
freestream velocity and ν is the kinematic viscosity of the air).
The commonly used honeycomb structures are made of alu- Because of this reason, the screen is placed at the beginning of
minum. Metallic structures can also be used but the burrs must the contraction cone where the velocity is low and hence the
be removed from them otherwise they can introduce addi- Reynolds number is least. The screen effectiveness is often
tional turbulence. Alternatively, the polycarbonate materials measured in terms of pressure drop coefficient (K), defined
provide cost-effective structures. by Eq. (12.11).
12.4.1.6 Effects of Screen on Turbulence p1 − p2

K= (12.11)
2 ρv1
Bell and Mehta (1989) found that, as long as the yaw angles 1 2
of the airstream are less than 10◦ , the honeycomb is quite
efficient in the removal of flow swirl and lateral velocity where p1 and p2 , are the static pressures upstream and down-
components, making the flow more uniform. Whereas, the stream of screen, respectively, ρ is the flow density and
larger yaw angles than this cause the honeycomb’s cell to v1 is the flow velocity upstream of screen. Further, from
stall, which increases the pressure drop across it and intro- Eq. (12.11), it can be seen that the pressure drop depends on
duces nonuniformity in the flow. As a remedy, for large swirl the local flow velocity squared. Consequently, the velocity
angles, the screens of very fine mesh size are deployed before decrease is more in regions where larger velocity exists. This
the honeycomb structures. The screen serves two purposes is in contrast to smaller velocity regions where the decrease in
290 12 Wind Tunnels
Table 12.1 Pressure loss coefficients of honeycomb cells

Cross-sectional shape Loss coefficient
Circular 0.30
Square 0.22
Hexagonal 0.20
velocity is found to be less. Essentially, K is a drag coefficient acteristics. It is important that the cross-sectional area of the
which measures the normal force exerted on the solid area of test section remains same along its length. However, due to
the screen. Collar (1939), proposed a relationship between the formation of boundary layer over the test section walls,
pressure loss coefficient and upstream and downstream tur- the net area decreases and consequently increasing the flow
bulence. If U + u1 is the longitudinal velocity far upstream of speed. Therefore, a horizontal force starts acting on the model
the screen and U + u2 is the longitudinal velocity far down- in the downstream direction, commonly known as Horizon-
stream of the screen, then tal Buoyancy. Thus, a suitable divergence is provided to the
test section walls so that the net cross-sectional area of the
u2 2−K uniform flow is constant along the length of the test section.
= (12.12)
u1 2+K Model to be tested is placed here in the airstream, leaving the
downstream end of the effuser, and the required measurements
From Eq. (12.12), it is clearly seen that the nonuniformity of
and observations are made. If the test section is bounded by
the flow will be eliminated by a screen whose pressure drop
rigid walls, the tunnel is called a closed-throat tunnel. If it
coefficient is equal to 2 whereas, it will be reversed for K > 2.
is bounded by air at different velocity (usually at rest), the
In general,
tunnel is called open jet tunnel.
K = f (Re, α, β) (12.13)
12.4.3 Subsonic Wind Tunnel Diffuser
where α is the flow incidence angle measured from the normal
to the screen and β is the open area ratio of the screen. The The diffuser follows the test section, used to reconvert the
porosity or open area ratio of a screen with square mesh is kinetic energy of the airstream leaving the test section into
given by the pressure energy as efficiently as possible. Essentially, it is
2 a passage where the subsonic flow decelerates by increasing
d the cross-sectional area downstream. The schematic diagram
β = 1− (12.14)
s of a typical diffuser is shown in Fig. 12.6. The diffusers are
described by both area ratio (ad ) or divergence angle (α), also
where d is the wire diameter, and s is the spacing between
known as equivalent cone angle. The area ratio of a diffuser
wires (Fig. 12.5).
is defined as
The screens with low porosity have higher values of pres-
sure drop coefficient. Also, we should note that the screens Cross-sectional area at exit A2
ad = = (12.15)
reduce axial turbulence more than the lateral turbulence. This Cross-sectional area at entry A1
is because it has a relatively larger pressure drop in the flow
direction, which reduces the higher velocity more than the The pressure rise phenomenon can be understood by applying
lower velocity, promoting a more uniform axial velocity. On the conservation of mass on the control volume (C∀) as shown
the other hand, the honeycombs have small pressure drops in Fig. 12.6.
in the flow direction and thus less effective on axial velocity.
However, owing to their length, they reduce the lateral veloc- A1 v1 = A2 v2 (12.16)
ities. It is believed that both honeycombs and screens reduce
lateral and axial turbulence due to energy exchange between Since,
the axes. Perhaps because of this, only the turbulence tends
towards isentropic in nature downstream. A2 > A1
12.4.2 Test Section Thus,
The test section is also referred to as working section. It is v2 < v1

the portion of a wind tunnel which has constant flow char-
Fig. 12.5 Schematic diagram of screen with square mesh
A2
Control Volume
v2
A1
α
v1
Fig. 12.6 Schematic diagram of a typical diffuser of a subsonic wind tunnel
i.e., with increase in cross-sectional area the flow veloc- fuser would result in a loss of efficiency. For a diffuser with
ity decreases. Equation (12.17) is the differential form of losses occurring, the modified form of Bernoulli’s equation is
Bernoulli’s equation written for an ideal diffuser. given by Eq. (12.18).

dp v2 v2 dp
+d (12.17) η d + (12.18)
ρ 2 2 ρ
It can be seen that the reduction in velocity leads to static pres- where η = diffuser efficiency. The efficiency of a diffuser may
sure rise. Therefore, the flow experiences an adverse pressure be represented in the following ways.
gradient while passing through the diffuser. Consequently, the
boundary layer decelerates and thickens rapidly on the pas- 1. Polytropic efficiency ηp
sage walls and hence, it becomes prone to separation. The 2. Isentropic efficiency or diffuser effectiveness (ηi )
separation of the flow from the wall of the diffuser is known
as diffuser stall, which restricts the static pressure recovery. 12.4.3.1 Polytropic Efficiency
The flow separation causes vibrations, oscillating for loading, Let us assume that the flow process in a diffuser is steady
oscillations in test section velocities (often called surging) and adiabatic. On the Mollier diagram, let the states at the
and increased losses in tunnel downstream. Since the energy entry and exit of the diffuser are denoted by “1” and “2”,
losses in a subsonic wind tunnel vary as the velocity cube. respectively (Fig. 12.7).
Therefore, the reduction in velocity should take place in the The diffuser efficiency for the polytropic process between
shortest possible distance in order to keep the losses at a mini- points “1” and “2,” can be defined as
mum. Generally, the semi-divergence angle is kept in between
7 and 8◦ to minimize the boundary layer separation on the ηp =
Isentropic enthalpy change
(12.19)
diffuser walls. However, even with very optimized design the Actual enthalpy change
boundary layer thickening and/or flow separation cannot be
fully avoided and hence, the conversion of kinetic energy into Since the process is steady and adiabatic and hence, the total
pressure energy can never be fully efficient. Also, any change enthalpy remains constant. That is
from the design mass flow and pressure ratio across the dif-
h01 = h02
292 12 Wind Tunnels
p
01
p
02
0a 02
2 p
(1/2) v 2
2
02s
2
Enthalpy (h) 2
(1/2) va
2s p
1
p
a
Entropy (s)
Fig. 12.7 Mollier diagram for the flow process through a subsonic diffuser
or 12.4.3.2 Isentropic Efficiency (Diffuser

Effectiveness)
1 2
h2 − h1 = v1 − v22 (12.20) The efficacy of a diffuser is also evaluated in terms of isen-
2 tropic efficiency, which is also known as diffuser effective-
For equivalent isentropic (i.e., reversible and adiabatic) pro- ness. It is defined to measure the actual diffuser performance
cess between “1” and “2s”, the isentropic enthalpy change against the theoretical ideal. That is
is
Actual static pressure rise
ηi = (12.25)
1 2 Ideal static pressure rise
h2s − h1 = v1 − v2s
2
(12.21)
2
We will now derive a mathematical expression of the
From Eqs. (12.20) and (12.21), the diffuser efficiency can be Eq. (12.25). Consider the control volume as shown in
written as Fig. 12.6. For a steady and incompressible flow, applying the
2 conservation of mass between points “1” and “2”,
v1 − v2s
2
ηp = 2 (12.22)
v1 − v22 U1 A1 = U2 A2 (12.26)
For an incompressible low-speed flow, Eq. (12.21) may also where U1 and U2 are the averages of axial velocities over the
be written as cross sections “1” and “2.”

p2 − p1 1
h2s − h1 = (12.23) Uavg = udA (12.27)
ρ A A
Thus, the diffuser efficiency is given by Eq. (12.24). It is visible that the flow velocity varies from zero at the dif-
fuser walls to a maximum at the centerline. This velocity
(p2 − p1 ) distribution is described by the kinetic energy flux factor (β)
ηp = 2 (12.24)
2ρ v1 − v22
1 which, varies from entry to exit of diffuser. For the same mass
flow passing through a particular diffuser cross section of area
(A), it is defined as
3
Actual transport of fluid kinetic energy 1 u section is necessary. It is achieved by deploying guide vanes at
β= = dA
Minimum possible transport of fluid kinetic energy A A Uavg
the corners to reduce the sharp angle variations. Therefore, in
(12.28) an open-circuit wind tunnel, the final output power produced
by fans is invested in two such as overcoming various losses
The minimum value of β is equal to 1, corresponding to in the circuit and imparting kinetic energy to the flow thrown
the uniform flow. However, β increases as the flow becomes into the ambient atmosphere. This total output power is the
peaked. Thus, if the peak of velocity at exit is minimized, power required to run the wind tunnel.
then α is minimized. Hence, the kinetic energy of the flow If pglobal is the total pressure drop across the tunnel (in
exiting the diffuser will be minimized. In other words, there Pascals) and Q is the volume flow rate measured in m3 s−1 ,
is a maximized static pressure recovery possible for a uniform then the power required to run the tunnel (in W) is given by
flow exiting the diffuser.
In terms of α, the static pressure coefficient for an ideal Prequired = pglobal × Q (12.32)
and nonuniform flow is given as
2 Thus, the fan efficiency (ηF ) can be defined as
p2 − p1 A1
Cp = = β1 − β2 (12.29)
1 2 A2 circuit losses
ρv1 ηF = (12.33)
2 Pinput
where β1 and β2 , are the kinetic energy profile parame- where Pinput is the power input to the fan (or driving unit).
ters at entry and exit of the diffuser, respectively. For an
ideal uniform flow profiles, β1 = β2 = 1 and u = Uavg . Thus,
Eq. (12.29) becomes
12.4.5 Losses in Subsonic Wind Tunnels
2
A1
Cp = 1 − (12.30) In a wind tunnel, the pressure losses occur as consecutive
A2
pressure
losses
in the different sections. Overall pressure loss
Thus, the diffuser effectiveness will be given by Eq. (12.31). pglobal is equal to the pressure gain due to the fan or driving
unit. The pressure loss in a subsonic tunnel circuit occurs due
Cp to the following reasons.
ηi = 2 (12.31)
A1
1− A2 • Pressure losses in wire gage (or screens).
• Pressure losses in honeycomb structure.
• Pressure losses due to friction in the contraction cone (or
12.4.4 Driving Unit nozzle).
• Pressure losses due to friction in the constant area section
In absence of losses, theoretically, the steady and uniform (working section or test section).
flow could be obtained forever once established without the • Pressure losses in the diffuser due to friction and expan-
need of further power supply. However, the losses occur due to sion.
dissipation of flow kinetic energy into heat because of vortic- • Pressure losses in the corners (or guide vanes) due to fric-
ity, eddying motion and turbulence. Moreover, in open-circuit tion and expansion.
wind tunnels some amount of kinetic energy is always rejected
at the diffuser exit which is converted into heat in mixing If the local pressure drop across a wind tunnel component is
with the surrounding air. In order to compensate these energy pi , then the pressure loss coefficient (Ki ) is defined as
losses, an external power source or driving unit is essential
for the wind tunnel operation. Usually, a motor and a fan or pi pi
Ki = = 1 2 (12.34)
2 ρi vi
propeller combination serve the purpose of driving unit in qi
low-speed tunnels.
The rejection of kinetic energy at the diffuser exit is somewhere qi is dynamic pressure, ρi is density, and vi is velocity
what eliminated in closed-circuit wind tunnels. However, the at entry to tunnel component.
external power is still necessary to overcome the losses occur-
ring due to vorticity, eddying motion, turbulence and the skin
12.4.5.1 Pressure Losses in Screens
friction at the walls and other surfaces. If the velocity distribu-
In wind tunnels, the screens are placed at the following two
tion has to be uniform in the test section and particularly with
locations.
low swirl and eddying motion, a construction ahead of the test
294 12 Wind Tunnels
1. Before the contraction cone to reduce turbulence. evaluate the pressure loss coefficient (Kn ).
2. Just before the fan inlet section where the velocities are rel-
atively higher and thus has significant impact on pressure Ln
Kn = 0.32favg (12.40)
loss. (Dh )SC
However, the pressure losses for both of the above screens can where favg is the average friction factor, Ln is the length of
be estimated by an empirical relation (Eq. (12.35)) proposed contraction cone, and (Dh )SC is the settling chamber hydraulic
by Eckert et al. [1976]. The screen pressure loss depends diameter. In Eq. (12.40), the average friction factor can be
upon porosity or its complement solidity, the Reynolds num- estimated using Prandtl universal law of friction for smooth
ber based on wire diameter (Rew ), and mesh factor (Kmesh ). pipes at high Reynolds number using Eq. (12.41).
The average value of mesh factor, Kmesh 1.3, is generally √ 2
used for most of the cases. f = 2 log10 Re f − 0.8 (12.41)
2
σs ρvDh
Ks = Kmesh KRn σs + (12.35) where Re =
βs μ .
where σs is screen solidity, βs refers to screen porosity and 12.4.5.4 Pressure Losses in Test Section
Consider a constant area test section of uniform hydraulic
0.785 1 − Rew
for 0 ≤ Rew < 400 diameter (Dh ). The pressure loss (p) due to friction along
KRn = 354
the length of duct (L) is given by Eq. (12.42).
1.0 for Rew ≥ 400
(12.36)
p L v2
=f (12.42)
In literature, screen solidity is defined as the complement of L Dh 2
screen porosity. That is
where f is the friction factor given by Eq. (12.41). Combin-
ing Eqs. (12.34) and (12.42), one can obtain the pressure loss
σs = 1 − βs (12.37)
coefficient for a constant area test section as

12.4.5.2 Pressure Losses in Honeycombs L
KTS = f (12.43)
The loss of pressure in honeycomb structure is affected by its Dh
porosity, flow Reynolds number based on cell hydraulic diam-
eter and ratio of streamwise length to cell hydraulic diameter. 12.4.5.5 Pressure Losses in Diffuser
To evaluate pressure losses, Eckert et al. [1976] proposed the The energy losses in the diffuser of a subsonic wind tunnel
following relation. owe to two factors such as skin friction and expansion. Math-
2 2 ematically
Lh 1 1
Kh = λh +3 −1 (12.38)
Dh βh βh Kd = Kf + Kexp (12.44)
In Eq. (12.38), Consider a subsonic diffuser of cylindrical cross section

⎧ 0.4 ⎫ shown in Fig. 12.8. Here, the divergence angles (α) is in
⎪
⎨ 0.375 ⎪ between 6◦ and 8◦ which proves to be optimum for diffusers

Dh Re−0.1
for Re ≤ 275 ⎬
λh = 0.4 (12.39) resulting in minimum pressure loss. The optimum cone angle
⎪
⎩ 0.214
for Re > 275 ⎪
⎭
Dh also depends upon diffuser area ratio, which dictates the pres-
sure recovery as well as the pressure gradient. If the diver-
where is material roughness, Re is Reynolds number gence angle is too large, prevailing adverse pressure gradient
based on material roughness, and Dh is hydraulic diameter. in the diffuser will lead to flow separation.
The loss coefficient due to skin friction can be obtained as
12.4.5.3 Pressure Losses in Contraction Cone
The pressure losses in the nozzle (or contraction cone) are favg D1 4
Kf = 1− (12.45)
about 3–4% of the total loss. Hence, errors in estimating 8 tan α2 D2
the loss coefficient are less significant as compared to
high-speed wind tunnels. Thus, the approximated expression where favg average friction factor, α wall divergence angle of
(Eq. (12.40)) given by Wattendorf [1969] can be used to the diffuser, and D1 and D2 , are diffuser inlet and exit diame-
D2
D1 α
Fig. 12.8 Schematic diagram of a subsonic diffuser
ters, respectively. Further, from Fleigner’s formula for pres- 12.4.6 Energy Ratio of a Subsonic Wind Tunnel
sure change in gradually expanded flows, the loss coefficient
may be obtained as2 The ratio of the kinetic energy of the jet flow in the test section
to the energy losses incurred in the circuit is a measure of the
4 energy efficiency of a wind tunnel, termed as the energy ratio
D1 (E.R.).
Kexp = sin α 1 − (12.46)
D2
Kinetic energy in test section
The combined losses for the diffuser is the sum of the frictional E.R. = (12.49)
Energy losses in tunnel circuit
and expansion loss coefficients (Fig. 12.9). Hence,
Based on the denominator, other definitions of the energy
4
favg D1 ratio are also used. For example, in place of circuit losses,
Kd = + sin α 1− (12.47)
8 tan α2 D2 the electrical power input to the fan or driving unit might be
chosen. However, Eq. (12.49) emphasized the aerodynamic
aspects of the energy budget and aids in distinguishing the
12.4.5.6 Pressure Losses in Corners
circuit flow properties from the efficiency of the fan or other
In an open-circuit wind tunnel, the critical location for major
driving unit.
corner losses is after the test section. This is because of greater
The energy ratio for both closed-circuit as well as open-
dynamic pressure and need for uniform flow at the fan inlet.
circuit wind tunnels (excluding open test section walls) is
To minimize the recirculation or flow separation losses at the
always greater than unity and is typically in the range of 3–7
corners, guide vanes are installed in the closed-circuit tunnels.
for closed-circuit tunnels. The greater the energy ratio, better
With an efficient blade cross section as well as an appropriate
the energy efficiency of the tunnel. Since, its value is always
chord-to-gap ratio. Guide vanes with cambered airfoils and
less than one for a free jet configuration and thus, no large
straight leading edges are less sensitive to approaching flow
size wind tunnel uses the open test section layout. Now, recall
angularities than sharp leading edge vanes.
Eq. (12.34),
Similar to that of a flat plate, assuming the skin friction as
a function of Reynolds number the corner loss coefficient can pi pi
be evaluated using Eq. (12.48). Ki = = 1 2
2 ρi vi
qi
4.55
Kc = 0.1 + 2.58 (12.48) Hence, the total circuit losses is written as
log10 Rec

where Rec is the local Reynolds number based on the chord pi = Ki qi (12.50)
of guide vanes. i i
These local losses are generally referred to the test section

2 Fleigner’s formula: conditions, defining the coefficient of loss of the local com-
ponent referred to the test section dynamic pressure as
ρ
p = (sin α) v12 − v22 .
2
296 12 Wind Tunnels
K0
Pressure loss co−efficient (K)

K 02
K 01
1 2 3 4 5 6 7 8 9 10
Divergence angle
Fig. 12.9 Variation of total losses with divergence angle
pi qi qi tent type tunnels are essentially open-circuit system which is

KTS = = Ki
qi qTS qTS further divided into blowdown and indraft (or induction) tun-
nels. Due to large power requirements, most of the high-speed
From Eq. (12.50), wind tunnels are of intermittent type.
The intermittent-blowdown and intermittent-indraft tun-
Ki qi = KTS qTS nels are normally used for Mach numbers from 0.5 to 5.0
i i whereas, the intermittent pressure-vacuum tunnels are used
for higher Mach numbers. The continuous operation tunnel is
Thus, energy ratio can be expressed as
used throughout the speed range. The choice of a particular
qTS 1 tunnel depends upon the specific measurement requirements
E.R. = = (12.51) as well as the power consumption. This is because despite
KTS qTS KTS
i i having advantages both intermittent and continuous tunnels
have their own disadvantages also. Therefore, a single tun-
nel is not suitable for all the purposes and for entire flow
12.5 High-Speed Wind Tunnels regime.
We know that, if the fractional change in temperature associ-

T 12.5.1 Intermittent-Blowdown Wind Tunnels
ated to a fluid flow is more than 5% ≥ 5% , then the
T
temperature effects can no longer be neglected in the analysis. There can be several different configurations of blowdown
At standard sea level conditions, TT ≥ 5% corresponds to the tunnels, used to generate high subsonic to moderately high
flow velocity of 650 kmh−1 (M ≥ 0.5). Since the compress- supersonic Mach numbers. The most common type of blow-
ibility effects become predominant at M ≥ 0.3, the flow with down tunnel uses the pressure difference between a pressur-
M ≥ 0.5 are called high-speed flows and so the wind tunnel is ized reservoir and the atmosphere to attain supersonic Mach
referred to as high-speed wind tunnels. Further, it is observed numbers in the test section. In closed-blowdown tunnel con-
that the power required for running a low-speed wind tunnel figuration, in addition to a high-pressure reservoir at one end, a
is proportional to cube of test section speed. Although, the relatively lower pressure tank is also attached at the other end.
same relationship does not hold for a high-speed wind tun- The valves of both chambers are kept open in the beginning
nel, nevertheless, the power requirement still increases with of a run and the pressure differential causes air to flow in the
increase of Mach number. direction ofs lower pressure until both the tanks attain equi-
The high-speed wind tunnels are classified on the basis of librium. The woking-section or test section is located imme-
either test section speed or operational point of view. Based on diately after the supersonic nozzle. In order to minimize the
the range of test section speed or Mach number, the classifi- pressure losses, many blowdown tunnels have two throats.
cation of high-speed tunnels are given in Table 12.2. From the The second throat is used to reduce the flow Mach numbers
operational point of view, they are grouped into intermittent from supersonic to subsonic level before it discharges to low-
and continuous operation tunnels (Fig. 12.10). The intermit- pressure chamber or ambiance. The typical characteristics of a
12.5 High-Speed Wind Tunnels 297
Table 12.2 Classification of wind tunnels based on test section Mach number
Tunnel type Mach number range
Transonic tunnel 0.8 < M < 1.2
Supersonic tunnel 1.2 < M < 5
Hypersonic tunnel 5<M
Wind tunnels
Intermittent tunnels Continous operation tunnels
Blowdown tunnels Indraft (or Induction) tunnels
Fig. 12.10 Classification of wind tunnels based on operational procedure
intermittent-blowdown wind tunnel are schematically shown 12.5.2.1 Advantages

in Fig. 12.11. The advantages of indraft tunnels are the following:
12.5.1.1 Advantages • Unlike blowdown tunnels, the stagnation temperature is

The major advantages of blowdown tunnels are given below: almost constant throughout the test run.
• Since the driving unit is at the downstream end, the test
• These tunnels start easily. section flow is free of contaminants.
• They have simple layout and thus most economical to build • In comparison to blowdown tunnels, these tunnels can
as compared to other types. operate even at higher Mach numbers before a heater is
• They have superior design for propulsion and smoke flow required to prevent liquefaction of the flow during expan-
visualization. sion.
• Using a vacuum tank is safer than using a high-pressure
12.5.1.2 Disadvantages reservoir, which may face the risk of explosion.
The main disadvantages of blowdown tunnels are the follow-
ing: 12.5.2.2 Disadvantages
The main disadvantages of indraft tunnels are given below.
• These tunnels have very small test time. Consequently,
they require expensive and high precision instrumentation • Due to large suction requirement, these tunnels are five
for measurements. times more expensive over blowdown tunnels.
• During the tunnel run the stagnation temperature drops, • The Reynolds number is very low as the stagnation pres-
which causes a change of Reynolds number in the test sure is atmospheric.
section. • Although, these tunnels are capable of running without air
• The starting load of the tunnel is very high. drier, however, they can do so only up to Mach 1.7 without
flow condensation.
12.5.2 Intermittent-Indraft Wind Tunnels

12.5.3 Continuous Supersonic Wind Tunnels
In this type of tunnel, a vacuum is created at the downstream
end of the tunnel while the upstream end is open to the atmo- These tunnels are essentially a closed-circuit system, used to
sphere. When a desired vacuum is established, the valve is establish a wide range of Mach numbers in the test section.
opened and ambient air rushes from the outside and thus, the Unlike intermittent tunnels, the flow which passes through
flow is established into the test section. A typical schematic the tunnel does not exhaust to the atmosphere, instead, it
diagram of a indraft tunnel is shown in Fig. 12.12. returns repeatedly to the test section through a devious path
(Fig. 12.13).
298 12 Wind Tunnels
High pressure tank
Test section Low pressure chamber
Pressure regulating valve
Settling chamber
Laval nozzle
Second throat
Diffuser
Fig. 12.11 Schematic layout of the intermittent-blowdown tunnel
Valve
Environment
Vacuum tank
Settling chamber Test section

Pressure
evacuation
pump
Fig. 12.12 Schematic layout of the intermittent-indraft tunnel
Cooler
Compressor
Drier
Guide vanes
Settling chamber
Test section
Diffuser
Fig. 12.13 Schematic layout of the continuous supersonic tunnel

12.5.3.1 Advantages and correction factor (CF) is defined as

• These tunnels are beneficial as more controlled conditions
can be established in the test section as the tunnel is cutoff Diffuser losses
CF =
from the environment while running. Total losses
• The guide vanes in the corners and flow straighteners
Depending upon the shock pattern through which the pressure
before the test section ensure a relatively uniform flow.
recovery is attained in the diffuser, the correction factor varies
• The operation of these tunnels are less noisy.
from 0.6 to 0.85.
• The test conditions can be held constant for an extended
period of time and the test time for each run is compara-
tively longer than other wind tunnel types.
12.5.5 Supersonic Wind Tunnel Components
12.5.3.2 Disadvantages There are five main parts of an intermittent supersonic wind
• The ratio of charging to operating time is large and thus, tunnels. They are
power requirement is very high.
• Temperature stabilization requires large size cooler. 1. Air supply system and storage tanks.
• The construction of these tunnels is complicated and 2. The settling chamber or wide-angle diffusers.
expensive. 3. The convergent–divergent or Laval nozzle.
4. The test section following the nozzle.
5. The diffuser, located just after the test section.
12.5.4 Losses in Supersonic Wind Tunnels
12.5.5.1 Air Supply System and Storage Tanks
The aggregate energy losses in a closed-circuit supersonic The compressed air stored in large storage tanks is fetched
tunnel might be split into the losses occurring in accompa- to run the wind tunnel. The size of air storage tanks depends
nying parts: dissipative losses in return circuit, expansion upon the mass flow rate and test time required. Generally,
losses in diffuser, frictional losses in contraction cone and multiple tanks, placed horizontally or vertically are used for
test section, losses in guide vanes, thermal losses in cooling higher storage requirements. In the author’s lab, a two-stage
devices, shock wave losses in supersonic diffuser, and losses reciprocating compressor, capable of delivering 360 ft 3 min−1
due to tunnel balance drag. Among these, the losses occurring of air at a pressure of 500 psi is being used to fill a vertically
due to shock and model mounting mechanism usually amount placed storage tank, having a total capacity of 1000 ft3 . The
to 90% of the total power loss, in which shock wave losses photographic view of the storage tank located in the author’s
alone constitute to about 80% and the rest is due to associated lab is shown in Fig. 12.14.
drag on wind tunnel balance. Hence, all other losses com- The compressor is driven by a 150 hp three-phase induc-
prise to only about 10% of the total loss. Thus, in estimating tion motor. A cooling water circuit, driven by an independent
the power requirement to determine the overall pressure ratio pump, cools the compressed air through an intercooler. The
(p0c )exit
required for the supersonic wind tunnel operation (p0c ) , compressed air is then passed through a prefilter consisting of
inlet
(p0d )exit porous stone candles to remove solid contaminants like rust
the pressure ratio across the diffuser (p0d ) alone is taken
inlet particles and oil droplets. An activated carbon filter is used
into account and a correction factor (CF) is applied to take
for finer filtering. The compressed air is dried in a dual-tower
care of the remainder of the losses. That is
semi-automatic silica gel dryer. While one tower is in use,

(p0c )exit 1 (p0d )inlet a portion of the dried air is heated and used to reactivate the
= other. A diaphragm type back pressure valve operated by pres-
(p0c )inlet CF (p0d )exit
sure relief pilot permits the dryer to operate at 500 psi, while
where the pressure in the storage tanks builds up from atmospheric
to the required storage pressure.
(p0c )exit = stagnation pressure at the compressor exit.
12.5.5.2 Settling Chamber
(p0c )exit = stagnation pressure at the compressor inlet. The gas fetched from storage tanks enter the settling cham-
ber through the tunnel section with a gate valve followed by a
(p0d )inlet = stagnation pressure at the diffuser inlet. pressure regulating valve and a mixing length tube of 80 mm in
diameter (Fig. 12.15). The settling chamber is connected to the
mixing tube by a wide angle diffuser followed by screens or
(p0d )exit = stagnation pressure at the diffuser exit.
closely meshed grids for minimizing turbulence at the super-
300 12 Wind Tunnels
Fig. 12.14 Photographic view of the vertically placed storage tank
sonic nozzle inlet. The settling chamber has a constant area stream parameters uniform over the cross section. The design
circular section of 300 mm inside diameter and 600 mm length of an appropriately shaped nozzle contour to get the desired
and have tappings for measuring the stagnation pressure and uniform stream at the nozzle exit depends on the method of
stagnation temperature. characteristics (MOC).
In Chap. 9, we have seen that the area ratio AAthroat
exit
and
12.5.5.3 Convergent–Divergent Nozzle (de Laval
Nozzle) Mach number (M) relationship (Eq. (12.52)) dictates the flow
A nozzle is the passage through which the flow accelerates. field in a nozzle. The variation of area ratio with the Mach
In a converging nozzle, the flow Mach number is increased number is shown in Fig. 12.17.
from M = 0 near the inlet to M = 1 at the exit. Thus, it’s 2 (γ+1)/(γ−1)
obvious that to induce a supersonic stream at the exit we must Aexit 1 2 γ+1 2
= 2 1+ M
add a diverging portion following the converging part. Such A∗ M γ+1 2
a nozzle is named convergent–divergent nozzle or de Laval (12.52)
Nozzle or simply Laval Nozzle, when Carl G.P. de Laval
who first used such a configuration in his steam turbines in where γ = ratio of specific heats for air.
the late nineteenth century. The schematic diagram of a typ- From Eq. (12.52), it is evident that local area (A) of the
ical convergent–divergent nozzle is shown in Fig. 12.16. To passage at any location must be larger than or at least equal
generate supersonic flow in the test section, a convergent– to the throat area (A∗ ) and the case, A < A∗ , is physically
divergent nozzle must be placed upstream of the test section. impossible
A in an isentropic flow. Also, each value of area
The nozzle manages the speed of air entering the test section ratio A∗ > 1 , there correspond two values of exit Mach
of the wind tunnel so that the uniform Mach number is set up. numbers; one being subsonic and other supersonic. Which of
Mach number is interestingly dictated by the range propor- the two values of (Me ) occurs depends on the operating back
tion of the nozzle. An optimally designed nozzle makes the pressure (pb ), which is the controlling pressure to establish
Fig. 12.15 Photographic view of the settling chamber
Divergent section
At
M<1
A entry
Throat
A exit
M=1
M>1
Convergent section
Fig. 12.16 A typical convergent–divergent nozzle
Isentropic flows
A
A*
M<1 M>1
M=1
Mach number
A
Fig. 12.17 Variation of A∗ with Mach number
the flow. Further, in order to establish the flow through the pressure will decrease continuously in the convergent portion
duct, the pressure at the exit must be lower than the pres- of the nozzle, reaching a minimum at the throat. If we fur-
sure at the nozzle entry, i.e., pe/p01 < 1. Let us assume that ther reduce the exit pressure to pe2 , then the pressure gradient
the exit pressure is reduced to a value, pe1 (< p0 ). Under this will be stronger, flow acceleration will be faster and varia-
situation, a small favorable pressure gradient is established tion of Mach number and static pressure through the duct will
across the nozzle which causes the flow through the nozzle be larger than the previous case. Similarly, if the nozzle exit
at low subsonic speeds. The local Mach number will increase pressure is reduced continuously, at some value (pe3 ), the flow
continuously through the convergent portion of the nozzle, will attain sonic state at the throat and for this case, At = A∗ .
reaching a maximum at the throat. In other words, the static Now the flow traveling at acoustic speed will expand further
302 12 Wind Tunnels
in the divergent portion as supersonic flow if the nozzle exit sure downstream of the shock at the design Mach number MD
pressure is less than the pressure at the throat, i.e., pe < pt , of the nozzle.
and will decelerate as subsonic flow if pe > pt , as shown by At this stage, we must realize the fact that the subsonic flow
the isentropic diffusion curve (b) in Fig. 12.18. Also, when downstream of a normal shock is always correctly expanded
the sonic condition prevails at the throat, the Mach number and hence, the exit pressures will be equal to the back pres-
at the throat becomes unity and hence, the flow properties at sure, i.e., pe = pb . The ambient pressure at which the flow is
the throat and indeed, throughout the convergent section of being discharged is known as back pressure (pb ). Basically,
the nozzle becomes frozen. That is, the subsonic flow in the the change in exit pressure is caused due to back pressure
convergent portion of the nozzle remains unaffected and mass monitoring.
flow remains constant for pe < pe3 . This condition for sonic If we further reduce the back pressure such that
flow at the throat is referred to as choked flow. Further reduc- pe8 < pb < pe6 , the flow inside the nozzle becomes fully
tion of exit pressure below pe3 does not increase the mass flow supersonic and isentropic, however, increase in flow pressure
once the throat is choked. A this
stage we must realize that the to pe7 , resulting in equilibrium with pb is taking place across
choked mass flow rate m̊max is maximum only for a given an oblique shock wave attached to the exit, outside the
stagnation state (p01 and T01 ) and remains constant till the nozzle, as shown in Fig. 12.19a. In this condition, the nozzle
stagnation conditions are same. However, by altering stagna- is said to be operating under overexpanded state. Conversely,
tion pressure and stagnation temperature a different value of if the nozzle operates with pb < pe8 , the nozzle is said to
m̊max will be obtained corresponding to the new stagnation be underexpanded. Here, the exit pressure is higher than the
state. back pressure and thus, the flow undergoes expansion after
The phenomena of throat choking can be physically leaving the nozzle. These expansion waves relax the exit
explained as follows. Let us consider a convergent–divergent pressure to pe9 to establish the equilibrium with pb outside
nozzle which is kept in still environment where pe = p01 , the nozzle (Fig. 12.19b).
in the beginning. To start the nozzle, the exit pressure (pe ) As discussed, when a nozzle discharges the maximum
is slightly reduced than the nozzle inlet pressure (p01 ) and mass flow rate m̊max and the gas is expanded completely
thus, fresh streamlines start entering into the nozzle. Further to the back pressure, a fully supersonic and isentropic flow
reduction in pe allows more and more streamlines entering results in nozzle divergence. The subsequent decrease of back
the nozzle and mass flow rate m̊ increases. However, pressure has no effect on the flow conditions inside the nozzle,
continuous decrease in pe results in a state when no more however, beyond the nozzle exit, the supersonic flow expands
fresh streamline is able to enter and the flow becomes in the same manner as does a jet emanating from a converging
choked and obviously, it will be dictated by the minimum nozzle working with a supercritical pressure ratio. Since the
area location, i.e., throat. Because of choking, the flow gas is not completely expanded inside the nozzle, the kinetic
condition downstream of the throat will not be able to pass energy of the flow at the exit Aexit , will be smaller than the
the information upstream and thus, the flow is virtually fully isentropic case. Consequently, the thrust produced by an
clogged (or frozen). underexpanded nozzle will be less than the nozzle operated
From the aforesaid discussion, it is now evident that for under correctly expanded state.
pe < pe3 the flow conditions downstream of the throat do not Similarly, when a optimally expanded convergent–
affect upstream. However, it rapidly accelerates in the noz- divergent nozzle is operated under overexpanded condition,
zle divergence to supersonic Mach numbers and to obtain the reduction in thrust is observed. In divergent portion of
an isentropic flow the exit pressure (pe ) must be reduced to an overexpanded nozzle, the static pressure drops below
the specified value pe8 , i.e., for pe = pe8 (< pt ) shock free the back pressure and hence, the equilibrium of pressure
supersonic flow is obtained as depicted by the curve (a) in is established by compressing the gas through a series of
Fig. 12.18. Recall, for pe = pe3 (> pt ) we would get another shock waves. Although, this phenomena has been studied
isentropic flow curve (isentropic diffusion), which depicts the by many investigators, however, it is yet not completely
subsonic flow in the nozzle divergence. Further, for all other understood. Let us consider
a convergent–divergent nozzle
values of exit pressure between pe3 and pe8 a non-isentropic having fixed area ratio AA∗ , is being operated under correct
solution results. That is, a normal shock appears in the nozzle expansion (pe8 = pb ) and the gas flow is fully supersonic
divergence when pe < pe3 . Since, the flow becomes subsonic and isentropic in nozzle divergence. If we raise pb more than
in the downstream of a normal shock wave and hence, the pe8 , the increase of back pressure cannot propagate itself
static pressure increases to pe4 at the exit. Further reduction upstream due to flow exiting the nozzle at supersonic Mach
in exit pressure moves the shock toward the nozzle exit and numbers. It can, however, propagate itself upstream through
hence, the static pressure rises to pe5 . For pe = pe6 , the shock the boundary layer on the wall and surrounding the flow.
stands exactly at the nozzle exit, where pe6 is the static pres- Since the fluid velocities within the boundary layer vary from
A entry A
exit
At
Throat
Me
pe1
p pe2 1
t1
p pe3
pe t2
(b)
p m < m* Isentropic diffusion pe4
01 2
p*
pe5
Normal pe6
m = m* (a) shock
3
pe7
4
pe8
Isentropic expansion to supersonic velocities pe9 5
1 Subsonic velocity at exit and Me < 1; isentropic flow in divergence
2 Sonic condition at throat, shock waves and non−isentropic flow in divergence
3 Supersonic flow in divergence with oblique shock and separated jet
4 Oblique shock at exit without jet separation
5 Expansion wave at exit and Me > 1
Fig. 12.18 Pressure distribution in a convergent–divergent nozzle under different operating conditions
supersonic levels at its interface with the outside main fluid minimize the flow separation especially in the rocket engines
to zero at the nozzle wall and thus, there exist zones where which flies at exceedingly high altitudes.
the gas speed is subsonic. The acoustic signals in the form of
pressure waves can propagate upstream in the boundary layer
through these subsonic regions. Further, if pb is marginally 12.5.5.4 Test Section
greater than pe8 , the oblique shock wave will be formed just The model to be tested is mounted on a sting balance in the
at test section, located at the downstream end of the convergent–
the
corner of the nozzle exit and the wall static pressure
pe will sharply increase from pe to pb . However, if pb is divergent nozzle. It is also called as working section of the
increased to a level which is considerably larger than pe8 , the wind tunnel. The photographic view of the test section of
flow will detach itself from the nozzle wall and the gas will supersonic wind tunnel located in the author’s lab is shown in
be compressed through the oblique shocks increasing the Fig. 12.21. The design of test section is affected by the desired
static pressure to pb , shown in Fig. 12.20. The static pressure flow speed in the test section. Also, the required minimum area
of the flow at the point where of the test section is same as the second throat area to have
separation occurs is known
as separation pressure psep . It is obligatory to alleviate or the shock-free test section.
304 12 Wind Tunnels
Divergent section
Gas stream
At
A entry
Throat
Oblique shock
A exit
M=1
Convergent section
Flow bending towards

nozzle centerline
Divergent section
Gas stream
At
A entry Expansion fans
Throat
A exit
M=1
Convergent section
Expansion fans
Fig. 12.19 Overexpanded and underexpanded nozzles
Divergent section
Gas stream
At
A entry
Throat
Oblique shock waves
M=1
A
exit
Convergent section
Separated flow
Fig. 12.20 Schematic diagram of an overexpanded nozzle operation with flow separation
To determine the size of the model, we must consider 12.5.5.5 Supersonic Wind Tunnel Diffuser
both incidents and reflected shock waves that form the shape A convergent–divergent diffuser is viewed as the operation
similar to a rhombus and hence, known as test rhombus of a convergent–divergent nozzle in reverse and hence, it
(Fig. 12.22). The model must be accommodated within the is also known as the reverse nozzle diffuser. In supersonic
rhombus formed by the incident and reflected shocks, for wind tunnel, supersonic diffuser immediately follows the
proper measurements. Also, note that the reflection of shock test section. We will investigate the typical characteristics
waves is not specular, which means that the angle of incidence of the flow through a fixed geometry convergent–divergent
of the shock at the test section wall is not the same as that for diffuser assuming that the flow is quasi-one-dimensional and
reflected shock wave. Taking this fact into account, the chord quasi-steady, which is isentropic everywhere except across the
length of the model is so chosen that the reflected waves do shock. Also, it is assumed that the flow is being discharged
not interfere with the model. to the ambient where the back pressure is pb . The stagnation
Fig. 12.21 Photographic view of the test section in the supersonic wind tunnel at IIT Kharagpur, India
Incident shock Reflected shock
Test rhombus
Wall
Fig. 12.22 Model size determination with the test rhombus
conditions (p01 , T01 ) in the settling chamber upstream of the When a diverging duct is attached at the end of test section,
diffuser, where p01 and T01 , are the stagnation pressure and the reduction in pb will move the shock wave downstream in
stagnation temperature, respectively, are constant. The flow is the diffuser, which eventually make the test section shock
being discharged to the atmosphere where stagnation pressure free (Fig. 12.23b). This phenomena is called normal shock
(p02 ) and stagnation temperature (T02 ) are also constant. The recovery or shock swallowing. In this case, the subsonic flow
flow is produced by reducing the static pressure downstream downstream of the shock is further decelerated isentropically
of the diffuser exit known as the back pressure (pb ). to p02 . Consequently, the wind tunnel operation at a lower
Let us consider the flow in a Laval nozzle with a normal pressure ratio (σ2 ) is achieved. That is
shock standing at the exit, shown in Fig. 12.23. With the reduc-
− γ
tion in back pressure (pb ), shock moves in the test section and p01 2γ 2 (γ−1)
1
(γ + 1) M12 (γ−1)
will stand precisely at the exit for the minimum possible pb . σ2 = = 1 + M1 − 1

p02 γ+1 (γ − 1) M12 + 2
The losses will be severe, if the flow is directly fetched to the (12.54)
atmosphere soon after exiting the test section. For an isen-
tropic flow, the required operating pressure ratio σ1 = pp02 01 Since,
to run the tunnel will be inversely
proportional to the stag-
p02 p02 > p02
nation pressure ratio p01 across the normal shock wave.
That is,
and hence,
− γ
2γ 2 1
(γ−1) (γ + 1) M12 (γ−1)
σ1 =
p01
= 1+ M1 − 1 σ2 < σ1
p02 γ+1 (γ − 1) M12 + 2
(12.53) i.e., by attaching a diverging diffuser at the nozzle exit, a lower
operating pressure is in principle, achievable.
306 12 Wind Tunnels
Fig. 12.23 Schematic layout of a portion of supersonic wind tunnel circuit
It is however, possible to reduce the operating pressure Further, a small reduction in pb produces the flow inside
even further than the previous case. If the supersonic flow exit- the tunnel which, at first, is subsonic throughout. However,
ing the test section could be first isentropically compressed to subsequent reduction in pb , causes increased mass flow rate
sonic conditions at a downstream throat, commonly referred until the nozzle throat A1∗ becomes
choked. In this situation,
to as second throat, and then it could be decelerated sub- a maximum mass flow rate m̊max is being passed through
sonically in the diffuser. To obtain these flow conditions, the nozzle, which is said to be operating at its first critical
instead of a diverging duct, a convergent–divergent diffuser condition. That is, the ratio, pp01b , is sufficient enough to cause
is attached to the end of test section as shown in Fig. 12.24. the nozzle throat (first throat) choked and consequently, sonic
Let us assume that the flow is taking place through a fixed condition (Mt = 1) prevails, where Mt is the Mach number
geometry convergent–divergent diffuser under constant stag- at the throat. When the back pressure is reduced further, a
nation conditions (p01 , T01 ) in the settling chamber where normal shock appears just downstream of the throat, and if pb
p01 and T01 , are the stagnation pressure and the stagnation is decreased continuously, the shock will move down in the
temperature, respectively. The flow is being discharged to the nozzle divergence and pass rapidly through the test section.
atmosphere where stagnation pressure (p02 ) and stagnation With a normal shock in the test section, which occurs at start-
temperature (T02 ) are being held constant. The flow in above up, the tunnel is said to be operating in its most unfavorable
wind tunnel circuit is produced by reducing the static pressure and off-design condition (Fig. 12.24a). If the ratio pp01b is low-
downstream of the diffuser exit, known as the back pressure ered further then the diffuser swallows the normal shock to
(pb ). the divergent section, consequently, the test section is shock
Fig. 12.24 Schematic layout of a portion of a supersonic wind tunnel
free causing an isentropic flow throughout. The power con- 0.6847p01 A1∗
m̊∗ = m̊max = √ (12.56)
sumption in operating the tunnel is however, still higher due to RT01
stagnation pressure loss across the shock in the diffuser diver-
or
gence. Once, the ratio pp01b is slightly increased the shock will
move upstream to the diffuser throat, the position at which the p01 A∗
shock strength is minimum. Figure 12.24b shows the general m̊∗ ∝ √ 1 (12.57)
T01
running condition of a wind tunnel, which is also called the
most favorable condition. where A1∗ = nozzle throat area and R = gas constant. The
Furthermore, we know that the mass flow rate m̊ through above mass flow rate will have to be passed through the dif-
a nozzle can be obtained by fuser throat as well, therefore we can write
√
m̊ p0 M γ/RT0 p02 A∗
= γ+1
(12.55) m̊∗ ∝ √ 2 (12.58)
A γ−1 2 2(γ−1) T02
1+ 2 M
where p02 and T02 are the stagnation conditions at
For air (γ = 1.4) when the nozzle throat is choked, Eq. (12.55) downstream of the diffuser. Comparing, Eqs. (12.57) and
reduces to (12.58)
308 12 Wind Tunnels
p01 A1∗ p02 A∗ contraction ratio (ψ). The maximum permissible contraction
√ = √ 2 (12.59)
T01 T02 ratio (ψmax ) for starting of a wind tunnel is
!
Further, the flow across a shock wave is adiabatic, i.e., ATS ATS A∗
T01 = T02 . Hence, ψmax = ∗ = ∗ · ∗ 1 (12.63)
A2 min A1 A2 min
p01 A1∗ = p02 A2∗ (12.60)
Introducing Eq. (12.62) into Eq. (12.63), we have
or
ATS ATS p02
ψmax = ∗ = ∗ · (12.64)
p01 A∗ A2 min A1 p01 TS
= 2∗ (12.61)
p02 A1
ATS p02
From Eqs. (12.52) and (12.53), and p01 TS , respec-
Equation (12.61) is a unique relation for a supersonic wind A1∗
tunnel operation, which implies that the flow can pass through tively, it is clear that the maximum permissible contraction
the only if A2∗ > A1∗ , that is, if the diffuser ratio (ψmax ) is a function of test section Mach number. That
∗diffuser
throat area
∗ . In the limiting
A2 is larger than the nozzle throat area A is
1
case, when A2∗ = A2∗ min , the shock will be positioned in
the test section causing the maximum loss of pressure. Under ψmax = f (MTS ) (12.65)
such condition, Eq. (12.61) becomes
Due to viscous interactions of shock and boundary layer on the
∗
p01 A2 min diffuser wall which, in turn, modifies the flow field structure,
= (12.62) the 100% pressure recovery is never achieved even with an
p02 TS A1∗
optimally designed diffuser. The fixed geometry convergent–
divergent diffuser is though, quite efficient at design oper-
where pp0201
is the ratio of stagnation pressure across the ating conditions but causes severe losses under off-design
TS
normal shock in the test section. conditions due to formation of a normal shock ahead of it.
We can understand the above discussion with an example. This shock may be swallowed by two methods: accelerat-
Let us assume that a normal shock is positioned in the test ing the flow beyond the design Mach number (MD ) at the
section, as shown in Fig. 12.24a, where the Mach number diffuser inlet or by using a variable throat area diffuser. How-
upstream of the shock is MD = M1 = 2. ever, except for very low Mach numbers the power required
For M1 = 2.5, using Eq. (12.53), we have to accelerate the flow to swallow the shock beyond its design
speed is quite large and hence, this technique is totally imprac-
A1∗
≈ 0.7209 tical. Also, the use of variable throat area into a diffuser poses
A2∗ many mechanical difficulties in the wind tunnel operation.
Thus, because of these reasons, the convergent–divergent dif-
where A1∗ = throat area upstream of the shock in the test fusers are not commonly used in supersonic wind tunnels.
section, and A2∗ = throat area downstream of the shock. Furthermore, the construction of a fully efficient diffuser
Thus, it is evident
that an increase in throat area greater is however, impossible but the pressure losses can be min-
(1−0.7209)
than 0.7209 × 100 = 38.7% is required to swallow the imized by employing some suitable techniques. One such
shock. Further, to start a supersonic tunnel having a diffuser method is to modify the geometry of convergent–divergent
of fixed geometry, it is essential to keep ∗
∗
A2 slightly larger than diffuser to compress the incoming supersonic flow through a
the theoretical minimum, i.e., A2 min to account for inaccu- series of weak oblique shocks followed by a normal shock of
racies in eliminating frictional losses due to viscous effects considerably lower strength instead of through a single and
and errors caused due to deviation from the assumption of strong normal shock wave as shown in Fig. 12.25, before it
one-dimensional flow, and so on. However, it should be real- actually expands as a subsonic flow, will result in a better
ized that, despite being a simplified analysis with isentropic pressure recovery. In this flow arrangement, the static pres-
flow assumption the aforesaid results are applicable to actual sure at the diffuser exit, in principle, can be made equal to
flow through the diffuser where the boundary layer is sucked the ambient pressure (pb ). However, in practice, the shock–
through the wall porosities. boundary-layer interactions limit the static pressure rise and
Furthermore, the ratio of test section area (ATS ) to the creates additional stagnation pressure losses, which partially
second throat area A2∗ is generally known as diffuser reduce the benefits of oblique shock diffuser.
12.6 Hypersonic Wind Tunnels 309
01
A*1
Convergent−Divergent diffuser with modified throat
Weak normal
A*2 shock
02
Convergent−Divergent nozzle
Test section
Weak oblique shocks

Shock diamonds
Fig. 12.25 Supersonic wind tunnel circuit with the modified diffuser geometry
12.6 Hypersonic Wind Tunnels cooled by the water. In addition, the alloy made of beryllium
and copper, or made of titanium, zirconium, and molybdenum
A hypersonic wind tunnel is designed to generate the hyper- are generally used as material liners which provide strength
sonic flow field in the test section. In principle, the design and to the throat with high heat conductivity.
operation of a hypersonic tunnel is similar to that of a super-
sonic wind tunnel (Fig. 12.26). However, the test section speed
is much higher for a hypersonic tunnel, which usually varies 12.7 Special Purpose Tunnels
between Mach 5 and Mach 15. During the tunnel operation,
air is initially compressed to high pressure by using a mul- The wind tunnels are often designated by the special purpose
tistage compressor depending upon the stagnation pressure for which they were designed and built. In the following sec-
requirement. This high-pressure air is then passed through tions, we will discuss four special purpose tunnels which are
the dryer to remove the moisture content before it is stored in used to obtain high Mach numbers in the test section.
the large tanks.
For a given stagnation temperature, when the dry air
expands through the nozzle it tends to liquefy3 due to decrease 12.7.1 Shock Tube
in the static temperature. Thus, the air must be heated to a
sufficiently high value to avoid condensation effects in the A shock tube is an instrument which is used to replicate and
test section. In the conventional hypersonic tunnels, different direct the blast wave over the model or a sensor for simu-
types of heaters such as combustor, electric resistance, and lation of the actual explosion and their effects on a smaller
arcjets are used for this purpose. Use of such heaters makes it scale. The simple shock tube, as shown in Fig. 12.28, con-
possible to increase the temperature of the high-pressure air sists of a straight tube of uniform cross section separated by
but lower the power requirement. In some wind tunnels, the a thin diaphragm, which divides the tube into two compart-
working fluid which has lower boiling point than the air such ments and establishes a pressure difference between them.
as Helium is also used to achieve condensation free hyper- One compartment is known as driver section and the other
sonic flow. one is referred to as driven section. The driver section is filled
The axisymmetric convergent–divergent nozzles are gen- with a highly pressurized gas, while the driven section is either
erally used in hypersonic wind tunnels which expand the evacuated or filled with a low-pressure test gas. Since both the
high-pressure and high-temperature air to the desired test compartments are separated by a thin diaphragm and thus the
section Mach number. The schematic diagram of a typical gases are in thermal equilibrium. Usually, an internal spike
convergent–divergent nozzle is shown in Fig. 12.27. These is used to rupture the diaphragm, which removes the barrier
nozzles are properly contoured to obtain uniform flow at the between the gas chambers. Once the diaphragm is bursted, a
nozzle exit. For high Mach numbers of order 10, the throat shock wave is propagated into the driven section by means
size becomes extremely narrow and forming the shape itself of expansion wave centered at the origin. Consequently, the
becomes very difficult. To withstand the high temperature or test gas is compressed and heated by the shock wave while
high enthalpy flow conditions, the nozzle throat is usually the high-pressure gas is expanded and cooled by the prop-
agation of expansion fan. Between the above compression
3 Note and expansion waves, there may be the regions in which gas
that air liquefies whenever static temperature falls below than
90 K.
310 12 Wind Tunnels
Hypersonic nozzle
Cooler
Driving unit
Settling
Compressor Test section Diffuser
chamber
To environment
Heat accumulator Compressor
Reservoir
Fig. 12.26 Schematic layout of a typical hypersonic wind tunnel
Divergent section
d1 At
M<1
A entry
Throat
d2
M=1 A exit
M>5
Convergent section
Fig. 12.27 Schematic layout of the hypersonic nozzle
Turbomolecular
vacuum
Pressure gauge Vacuum pump pump Liquid Constant Vacuum
injection temperature gauge
nozzle system Test section
Driver section Driven section
Diaphgram
Nozzle
Gas cylinder
Pressure vessels
Fig. 12.28 Schematic layout of the shock tube
moves with varied speeds ranging from subsonic to super- are different. These gases are separated by an interface or con-
sonic Mach numbers depending on the initial pressure ratio tact surface, which is indeed a thin zone rather than a plane
across the diaphragm. Moreover, there exist two bodies of surface. The contact surface is assumed to travel at a con-
gases which are brought to the same pressure but due to dif- stant speed known as particle velocity. Therefore, behind the
ferent mixing process their temperature, density, and entropy shock wave, there is a region of steady flow of the gas is at
12.7 Special Purpose Tunnels 311
high pressure and high temperature, whereas, the flow behind expansion waves propagate through the nozzle and the long
the contact front is again steady with same pressure but at tube. The unsteady propagation of expansion wave sets up a
lower temperature. The steady-state characteristics of these steady subsonic flow toward the convergent–divergent noz-
zones, allow to use them for aerodynamic testing. However, zle, that expands the flow to supersonic velocities. Between
the duration of steady flow in these regions is limited to the the two disturbances, a steady flow is established until the
order of few milliseconds in tubes of sufficient lengths by the expansion fan gets reflected from the far end of the tube and
reflection of shock and expansion waves from the opposite reach to the nozzle again. A test time of about 100 ms can be
ends of the tube. achieved, which is sufficient enough for practical purposes.
12.7.2 Gun Tunnel(or Shock Tunnel) 12.7.4 Plasma Wind Tunnel
The gun tunnels also known as shock tunnels, are the wind Plasma wind tunnel, also called plasma arc tunnel, uses a high
tunnels that operate at Mach 25 or higher for the time inter- current electric arc to heat the test gas to very high tempera-
vals up to a few milliseconds. The air is used as working fluid ture (∼12000 K). The available testing time is of the order of
which is heated and compressed in a long tube. The gun tun- several minutes using either direct or alternating current. The
nel, as sketched in Fig. 12.29, consists of three independent tunnel is composed of an arc chamber, a convergent–divergent
sections: the driver section, the driven section, and the test nozzle which expands the flow typically below Mach 3, a
section. The driver and driven section are separated by a high- test section, and a vacuum or low-pressure chamber, shown
pressure diaphragm, while a low-pressure diaphragm sepa- in Fig. 12.31. During the tunnel operation, a cold test gas is
rates the driven tube and the test section. A lightweight piston fetched through the arc chamber and the nozzle. The elec-
is also placed in the driven section, beside the diaphragm, so tric arc produced between an electrode and some surface of
that when the high-pressure diaphragm bursts, the piston is the arc chamber, raises the test gas temperature to ionization
pushed through the driven tube, compressing the gas ahead of level yielding plasma; that is essentially a mixture of positive
it. Since the piston weight is very small and hence it is accel- charged ions, free electrons, and neutral atoms. To achieve
erated to supersonic velocities in the driven tube. This causes higher degree of ionization, Argon is frequently used as the
a shock wave to precede the piston through the driven tube test gas in place of air with a given power input.
and heat the gas. The shock will be reflected from the end of Plasma arc tunnels are particularly useful in studying the
the driven tube to the piston, inflicting additional heating of phenomena associated with re-entry vehicles, as they are
the gas. The piston comes to rest with equal pressure on its capable of generating high heat rates. Surface ablation test
two sides, and the heated and compressed driven gas bursts can be performed using this tunnel, which is generally not
the low-pressure diaphragm and flows through the nozzle. It is possible with low-temperature tunnels or high-temperature
evident that gun tunnels are limited in the maximum tempera- short-duration tunnels.
ture that can be achieved by the piston design. The maximum
temperatures of about 2000 K is generally obtained.
Summary
12.7.3 Ludwieg Tube A wind tunnel is the device that provides airstreams flowing
under controlled conditions so that models of interest can
Hubert Ludwieg, a German scientist was the first who probe tested using them. Basically, they are used to test scale
posed the concept of Ludwieg tube in 1955. It is relatively models of aircraft and spacecraft, however, some wind tunnels
cheaper and an efficient way of producing supersonic flow. are big enough to hold the full-size (prototype) vehicles. The
Essentially, Ludwieg Tube is a wind tunnel in which the test airstream past an object makes it seem like the object is really
section Mach numbers up to 4 can be easily obtained with- flying. From an operational point of view, wind tunnels are
out additional heating arrangements. The schematic layout generally classified as low-speed, high-speed, and special-
of a typical Ludwieg tube is shown in Fig. 12.30. A large purpose tunnels.
evacuated dump tank is separated from the downstream end Based on the desired flow speed in the test section, the
of a convergent–divergent nozzle by a thin diaphragm. The wind tunnels are divided into:
upstream end of the nozzle is connected to a long cylindrical
tube, whose cross-sectional area is significantly larger than the • Subsonic wind tunnel (MTS < 0.8).
nozzle throat area. In the beginning, both tube and the nozzle is • Transonic wind tunnel (0.8 < MTS < 1.2).
filled with a highly pressurized gas (air). Once the diaphragm • Supersonic wind tunnel (1.2 < MTS < 5).
is bursted, a shock wave propagates into the dump tank and the • Hypersonic wind tunnel (MTS > 5).
312 12 Wind Tunnels
High pressure air
Piston
Model
Driver Driven section
section
Nozzle throat
Diaphragms
Fig. 12.29 Typical layout of a gun tunnel
Pressure regulating Test section

valve
Discharged column
Ludwieg charge tube length
Hypersonic nozzle
Vacuum dump tank
Fig. 12.30 Schematic layout of a typical Ludwieg tube
Cooling water out

Cooling water in
Cooling water out Cooling water in
Low density
test section
Gas inflow
Model
Arc chamber Settling chamber Nozzle

Insulated electrode
Fig. 12.31 Schematic layout of a typical plasma arc tunnel

Summary 313
The classification based on Mach number is caused due to formation of boundary layer over the test section walls, the
compressibility effects. We know that for M < 0.3, the com- net area decreases and consequently increasing the flow speed.
pressibility effects can be ignored. However, for Mach num- Therefore, a horizontal force starts acting on the model in the
bers higher than this obtained in transonic and supersonic downstream direction, commonly known as Horizontal Buoy-
tunnels, the flow must be assumed to be compressible. The ancy. Thus, a suitable divergence is provided to the test section
fluid compressibility has serious implications on the tunnel walls so that the net cross-sectional area of the uniform flow
design; the test section has the minimum area in a subsonic is constant along the length of the test section. Model to be
tunnel whereas, the throat is the minimum area location for a tested is placed here in the airstream, leaving the downstream
supersonic wind tunnel. end of the effuser, and the required measurements and obser-
The wind tunnels are also designated as special purpose vations are made. If the test section is bounded by rigid walls,
for which they are designed and built. They have layouts alto- the tunnel is called a closed-throat tunnel. If it is bounded by
gether different than that of low or high-speed wind tunnels. air at different velocity (usually at rest), the tunnel is called
Following are some typical examples of this type: open jet tunnel.
The diffuser follows the test section, used to reconvert the
1. Spinning tunnels—since aircraft have a tendency to go to kinetic energy of the airstream leaving the test section into the
spin mode once it stalls, thus these facilities are used to pressure energy as efficiently as possible. Essentially, it is a
investigate those phenomena. passage where the subsonic flow decelerates by increasing the
2. Stability tunnels—sometimes it is imperative to study the cross-sectional area downstream. The diffusers are described
stability of an aircraft and hence these tunnels are designed by both area ratio (ad ) or divergence angle (α), also known as
in such a way to allow the model to move freely within the equivalent cone angle. The area ratio of a diffuser is defined
test section. as
3. Low-density tunnels—to accurately simulate the high-
temperature effects at hypersonic Mach numbers or rar- Cross-sectional area at exit A2
ad = =
efied state of gases in outer layer of the atmosphere, certain Cross-sectional area at entry A1
low-density tunnels are designed.
Theoretically, in the absence of losses, the steady and uni-
In order to obtain an accurate quantitative data from a model form flow could be obtained forever once established without
study, the physics behind model and prototype flow system the need of further power supply. However, the losses occur
operating under different conditions should be the same. The due to dissipation of flow kinetic energy into heat because of
similarity between two flow systems is established by equat- vorticity, eddying motion and turbulence. Moreover, in open-
ing the ratio of specified physical parameters calculated in circuit wind tunnels, some amount of kinetic energy is always
each system. If the specified physical quantity is geometric rejected at the diffuser exit which is converted into heat in
dimension, the similarity is called geometric similarity; if the mixing with the surrounding air. In order to compensate these
quantity is related to motion, the similarity is called kinematic energy losses, an external power source or driving unit is
similarity; and if the quantity denotes the force, then the simi- essential for the wind tunnel operation. Usually, a motor and
larity is termed as dynamic similarity. Two fluid flows are said a fan or propeller combination serve the purpose of driving
to be similar only if these three similarities exist together. unit in low-speed tunnels.
A general utility low-speed wind tunnel has four major In a wind tunnel, the pressure losses occur as consecutive
components: the effuser, the test section, the diffuser, and the pressure losses in the different sections. Overall pressure loss
driving unit. Effuser is a converging duct located upstream of is equal to the pressure gain due to the fan or driving unit.
the test section in which the fluid gets accelerated from rest The pressure loss in a subsonic tunnel circuit occurs due to
(or from very low speed) drawn at the upstream end of it to the following reasons:
the required conditions at the test section. Thus, the effuser
is also referred to as contraction cone. The contraction ratio • Pressure losses in wire gage (or screens).
(ac ) of an effuser is defined as • Pressure losses in honeycomb structure.
• Pressure losses due to friction in the contraction cone (or
Cross-sectional area at entry to contraction cone nozzle).
ac = • Pressure losses due to friction in the constant area section
Cross-sectional area at exit to contraction cone
(working-section or test section).
The test section is also referred to as workingsection. It is • Pressure losses in the diffuser due to friction and expan-
the portion of a wind tunnel which has constant flow char- sion.
acteristics. It is important that the cross-sectional area of the • Pressure losses in the corners (or guide vanes) due to fric-
test section remains same along its length. However, due to tion and expansion.
314 12 Wind Tunnels
The ratio of the kinetic energy of the jet flow in the test-section (p0c )exit = stagnation pressure at the compressor exit.
to the energy losses incurred in the circuit is a measure of the
energy efficiency of a wind tunnel, termed as the energy ratio (p0c )exit = stagnation pressure at the compressor inlet.
(E.R.).
Kinetic energy in test section (p0d )inlet = stagnation pressure at the diffuser inlet.
E.R. =
Energy losses in tunnel circuit
(p0d )exit = stagnation pressure at the diffuser exit.
The energy ratio for both closed-circuit as well as open-circuit
wind tunnels (excluding open test section walls) is always and correction factor (CF) is defined as
greater than unity and is typically in the range of 3–7 for
closed-circuit tunnels. The greater the energy ratio, better the Losses in diffuser
CF =
energy efficiency of the tunnel. Since, its value is always less Total losses
than one for a free jet configuration and thus, no large size
wind tunnel uses the open test section layout. Depending upon the shock pattern through which the pressure
We know that, if the fractional change in temperature asso- recovery is attained in the diffuser, the correction factor varies
from 0.6 to 0.85.
T
ciated to a fluid flow is more than 5% ≥ 5% , then the There are five main parts of an intermittent supersonic wind
T
temperature effects can no longer be neglected in the analysis. tunnels. They are:
At standard sea level conditions, TT ≥ 5% corresponds to the
flow velocity of 650 kmh−1 (M ≥ 0.5). Since the compress- • Air supply system and storage tanks.
ibility effects become predominant at M ≥ 0.3, the flow with • The settling chamber or wide-angle diffusers.
M ≥ 0.5 are called high-speed flows and so the wind tunnel is • The convergent–divergent or Laval nozzle.
referred to as high-speed wind tunnels. Further, it is observed • The test section following the nozzle.
that the power required for running a low speed wind tunnel is • The diffuser, located just after the test section.
proportional to cube of test section speed. Although, the same
A hypersonic wind tunnel is designed to generate the hyper-
relationship does not hold for a high-speed wind tunnel, nev-
sonic flow field in the test section. In principle, the design
ertheless, the power requirement still increases with increase
and operation of a hypersonic tunnel are similar to that of
of Mach number. The high-speed wind tunnels are classified
a supersonic wind tunnel. However, the test section speed is
on the basis of either test section speed or operational point
much higher for a hypersonic tunnel, which usually varies
of view.
between Mach 5 and Mach 15. During the tunnel operation,
The aggregate energy losses in a closed-circuit supersonic
air is initially compressed to high pressure by using a mul-
tunnel might be split into the losses occurring in accompa-
tistage compressor depending upon the stagnation pressure
nying parts: dissipative losses in return circuit, expansion
requirement. This high-pressure air is then passed through
losses in diffuser, frictional losses in contraction cone and
the dryer to remove the moisture content before it is stored in
test section, losses in guide vanes, thermal losses in cooling
the large tanks.
devices, shock wave losses in supersonic diffuser, and losses
For a given stagnation temperature, when the dry air
due to tunnel balance drag. Among these, the losses occur-
expands through the nozzle, it tends to liquefy due to decrease
ring due to shock and model mounting mechanism usually
in the static temperature. Thus, the air must be heated to a
amount to 90% of the total power loss, in which shock wave
sufficiently high value to avoid condensation effects in the
losses alone constitute to about 80% and the rest is due to
test section. In the conventional hypersonic tunnels, different
associated drag on wind tunnel balance. Thus, in estimating
types of heaters such as combustor, electric resistance, and
the power requirement to determine the overall pressure ratio
(p0c )exit arcjets are used for this purpose. Use of such heaters makes it
required for the supersonic wind tunnel operation (p0c ) , possible to increase the temperature of the high-pressure air
inlet
(p0d )exit but lower the power requirement. In some wind tunnels, the
the pressure ratio across the diffuser (p0d )inlet alone is taken
into account and a correction factor (CF) is applied to take working fluid which has lower boiling point than the air such
care of the remainder of the losses. That is, as Helium is also used to achieve condensation free hyper-
sonic flow.

(p0c )exit 1 (p0d )inlet The wind tunnels are often designated by the special pur-
= pose for which they were designed and built. We have dis-
(p0c )inlet CF (p0d )exit
cussed four special purpose tunnels which are used to obtain
where high Mach numbers in the test section.
Summary 315
A shock tube is an instrument used to replicate and direct of several minutes using either direct or alternating current.
the blast wave over the model or a sensor for simulation of The tunnel is composed of an arc chamber, a convergent–
the actual explosion and their effects on a smaller scale. The divergent nozzle which expands the flow typically below
simple shock tube consists of a straight tube of uniform cross Mach 3, a test section, and a vacuum or low pressure chamber.
section separated by a thin diaphragm, which divides the tube During the tunnel operation, a cold test gas is fetched through
into two compartments and establishes a pressure difference the arc chamber and the nozzle. The electric arc produced
between them. One compartment is known as driver section between an electrode and some surface of the arc chamber,
and the other one is referred to as driven section. The driver raises the test gas temperature to ionization level yielding
section is filled with a highly pressurized gas, while the driven plasma; that is essentially a mixture of positive charged ions,
section is either evacuated or filled with a low-pressure test free electrons, and neutral atoms. To achieve higher degree of
gas. Since both the compartments are separated by a thin ionization, Argon is frequently used as the test gas in place of
diaphragm and thus the gases are in thermal equilibrium. air with a given power input.
Usually, an internal spike is used to rupture the diaphragm,
which removes the barrier between the gas chambers. Once
the diaphragm is bursted, a shock wave is propagated into the Exercises
driven section by means of expansion wave centered at the
origin. Consequently, the test gas is compressed and heated Descriptive Type Questions
by the shock wave while the high-pressure gas is expanded
and cooled by the propagation of expansion fan. 1. Consider a suction type subsonic wind tunnel which
The gun tunnels also known as shock tunnels, are the wind draws air at standard sea level conditions (15 ◦ C and
tunnels that operate at Mach 25 or higher for the time intervals 101.325 kPa) by using a fan at the wind tunnel exit. If
up to a few milliseconds. The air is used as working fluid the airspeed in the test section is 75 ms−1 , calculate the
which is heated and compressed in a long tube. The gun tunnel pressure in the test section.
consists of three independent sections: the driver section, the 2. In a wind tunnel study, the root mean square (RMS) value
driven section, and the test section. The driver and driven of the perturbation component of velocity along the x-axis
section are separated by a high-pressure diaphragm, while is 2.5 ms−1 . If the freestream velocity is 25 ms−1 , calcu-
a low-pressure diaphragm separates the driven tube and the late the degree and the intensity of turbulence. Assume
test section. A lightweight piston is also placed in the driven turbulence to be isotropic.
section, beside the diaphragm, so that when the high-pressure 3. Consider the operation of a subsonic wind tunnel under
diaphragm bursts, the piston is pushed through the driven tube, the following two different configurations.
compressing the gas ahead of it. Since the piston weight is
very small and hence, it is accelerated to supersonic velocities (a) Wind tunnel consists of a nozzle and the test section and
in the driven tube. This causes a shock wave to precede the the flow is simply discharged to the atmosphere, i.e., it
piston through the driven tube and heat the gas. The shock has no diffuser after the test section.
will be reflected from the end of the driven tube to the piston, (b) Wind tunnel has a nozzle, the test section followed by
inflicting additional heating of the gas. a diffuser such that the flow after the diffuser exit is
Ludwieg Tube is a wind tunnel in which the test section discharged to the environment.
Mach numbers up to 4 can be easily obtained without addi-
tional heating arrangements. A large evacuated dump tank is In both the cases, assume the contraction cone ratio and
separated from the downstream end of a convergent–divergent the test section area, respectively, are 5.0 and 0.4 m2 . Find
nozzle by a thin diaphragm. The upstream end of the nozzle the overall pressure difference established across the wind
is connected to a long cylindrical tube, whose cross-sectional tunnel, if the test section speed is 90 ms−1 in both cases.
area is significantly larger than the nozzle throat area. In the Assume the diffuser exit area is 1.8 m2 for the case (b).
beginning, both tube and the nozzle is filled with a highly 4. An aircraft is flying at a speed of 80 ms−1 against a wind
pressurized gas (air). Once the diaphragm is bursted, a shock speed of 6 ms−1 . What will be the pressure difference
wave propagates into the dump tank and the expansion waves recorded by the pitot-static probe mounted on the aircraft?
propagate through the nozzle and the long tube. The unsteady (For air, ρair = 1.22 kg m−3 ).
propagation of expansion wave sets up a steady subsonic flow 5. A supersonic blowdown type wind tunnel has the test
toward the convergent–divergent nozzle, that expands the flow section area 0.04 m2 and pressure 0.03 MPa. If the test
to supersonic velocities. section Mach number is 3 and tunnel exhausts into the
Plasma wind tunnel also called plasma arc tunnel, uses a atmosphere, then find (a) the minimum stagnation pres-
high current electric arc to heat the test gas to very high tem- sure required for tunnel operation, (b) the corresponding
perature (∼12000 K). The available testing time is of the order stagnation density, and (c) mass flow rate.
316 12 Wind Tunnels
6. A pitot probe in Mach 2.5 freestream flow shows the (a) 64.12 ms−1
stagnation pressure reading as 5.2 × 104 Nm−2 . If (b) 75.84 ms−1
the freestream static and stagnation temperatures are (c) 79.66 ms−1
−183 ◦ C and −3 ◦ C, respectively. Calculate the static (d) 83.52 ms−1
pressure, static density, and the stagnation pressure
upstream of the bow-shock formed in front of the probe. 3. An aircraft is flying at 230 ms−1 at an altitude where the
7. A pitot probe immersed in the high-speed airstream mea- surrounding temperature is 250 K. A scaled model of the
sures the pressure at the stagnation point as 0.1 MPa. The same aircraft is tested in the wind tunnel. What will be
static pressure and static temperature of the airstream are the temperature inside the wind tunnel test section, if the
0.07 MPa and 263 K, respectively. Calculate the airspeed. airspeed is found to be 260 ms−1 ?
8. An airship 6 m in diameter and 50 m in length is flying
at 20 ms−1 . If the wind tunnel test is performed on a 201th (a) 250 K
scale model of the airship prototype, the calculate the (b) 274 K
airspeed required in the test section. Assume similar con- (c) 301 K
ditions of pressure and temperature in both the cases. (d) 319 K
9. In a transonic wind tunnel, cross-sectional area of the
4. Consider a Mach 2.2 wind tunnel where the scaled model
test section is 0.6 m2 . At a location, where the cross-
of a wing is being tested. If the angle of attack experienced
sectional area is 3 m2 , the local pressure and tempera-
by the wing is reduced from 8◦ to 3◦ , then which of the
ture are 0.4 MPa and 278 K. Assuming the test section
following will be TRUE?
flow to be one-dimensional and isentropic, calculate the
static pressure, static temperature and static density at a (a) induced drag will be decreased
location in the test section, where the Mach number is (b) induced drag will be increased
0.75. (c) skin friction drag will be decreased
10. Consider the operation of a blowdown type supersonic (d) pressure drag will be decreased
wind tunnel with cylindrical cross section. The area of
the first throat is 0.03 m2 and the tunnel is designed to 5. The inlet and exit diameters of a diffuser are 1 and 2 m,
operate at Mach 2.5. Calculate the minimum area of the respectively. The wall divergence angle of the diffuser is
second throat required, so that, the test section flow is 15◦ . From the Fleigner’s formula of pressure change in
completely isentropic. After the normal shock swallow- gradually expanded flows, the loss coefficient is
ing, the second throat area is reduced to 0.05 m2 . If the
diffuser exit area is 0.2 m2 , what will be the ratio of the (a) 0.11
back pressure (pb ) and the settling chamber stagnation (b) 0.21
pressure (p01 )? (c) 0.31
(d) 0.41
6. The type of tunnel used to accurately simulate the high

Multiple Choice Questions temperature effects at very high Mach numbers (i.e., rar-
efied state of gases in the outer layer of the atmosphere)
1. In supersonic wind tunnels, the oblique shock diffusers is
are preferred over normal shock diffusers because of
(a) low-density tunnels
(a) lower total pressure losses (b) stability tunnels
(b) higher speed (c) spinning tunnels
(c) lesser number of shock waves (d) supersonic tunnels
(d) lower thermal effects
7. The honeycomb cells may have circular, square or hexag-
2. In a wind tunnel, the contraction ratio and the cross- onal shapes of their cross sections. The pressure loss coef-
sectional area of the test section are 14:1 and 0.3 m2 , ficient for the hexagonal shaped honeycomb cells is
respectively. If the static pressure difference between the
settling chamber and the test section is 35 cm of water (a) 0.10
column, then the airspeed in the test section will be (b) 0.20
(Assume ρair = 1.2 kg m−3 , ρwater = 1000 kg m−3 and (c) 0.22
g = 9.81 ms−2 ). (d) 0.32
Exercises 317
8. A subsonic wind tunnel with the test section area of (a) 250 kW
0.25 m2 , is operated at 40 ms−1 . The pressure and the (b) 7.5 kW
temperature of air in the test section are 100 kPa and (c) 0.625 MW
296 K, respectively. The flow density inside the test (d) 10 MW
section is
Keys
(a) 1.32 kg m−3
(b) 1.28 kg m−3 1. (a)
(c) 1.22 kg m−3 2. (b)
(d) 1.17 kg m−3 3. (d)
4. (a)
9. In subsonic wind tunnels, screens are placed near the 5. (b)
entrance of the contraction cone, where the velocity is 6. (a)
found to be minimum and the Reynolds number is lowest. 7. (b)
If ‘d’ is the wire diameter and ‘s’ is the spacing between 8. (d)
wires, the screen porosity for square mesh will be given 9. (c)
as 10. (c)
2
(a) ds − 1
2
(b) 1 − ds
2 References
(c) 1 − ds
s 2
(d) d − 1 Barlow JB, Rae WH, Pope A (1999) Low speed wind tunnel testing.
Wiley, New York
10. A wind tunnel of test section area 0.3 m2 is being operated Bell JH, Mehta RD (1989) Boundary layer prediction for small low speed
at 50 ms−1 . If the total pressure drop across the tunnel is contractions. AIAA Journal. 27(3):372–374
Collar AR (1939) Aeronautical Research Council (Great Britain) Tech-
50 kPa, then the power required to run the tunnel will be nical Report. Number 1867. p.1
Eckert WT, Mort KW, Piazza JE (1976) An experimental investigation
of end treatments for non-return wind tunnels. NASA TMX-3402
Pereira JD (2011) Wind tunnels. Nova Science Publishers Inc., New
York
Wattendorf FL (1969) The AGARD history 1952–1968.
AGARD/Technical Editing and Reproduction, Paris/London
Part II
Applied Aerodynamics
Supersonic Jets
13
The control of high speed jets has turned into an interesting local wind blows in the same direction as of the plume. In this
research area due to its extensive variety of uses starting from co-flow arrangement which is initially buoyancy dominated
household applications to space technology. Among these, but becomes momentum dominated in the far field region, the
the passive control technique has turned into a conspicuous transition is generally speculated in terms of a dimensionless
research field because of its undeniable preference. The differ- number called the Richardson number (Ri).
ent passive controls for increasing aerodynamic mixing and
aeroacoustic advantages have been discussed in this chapter. Buoyancy term g ∇ρ
Ri = =
Additionally, the computational studies on jets performed by Flow shear term ρ (∇u)2
a few researchers in the past is likewise inspected at a length.
where g is the gravity, ρ is the density, and u is the flow
speed. As the jet stream moves away from its source, it slows
13.1 Introduction down due to mixing of low speed or stagnant ambient fluid
with high momentum jet flow. This mixing phenomena can
The free or submerged jet is defined as the momentum-driven be understood as follows. Basically, when the jet comes out
unrestricted fluid flow issuing from a nozzle or orifice into of a confined space, the jet boundary develops a tangential
quiescent ambience. If it is surrounded by another fluid flow- separation surface having either a stagnant or lower velocity
ing with a different velocity at its periphery, it is referred to as fluid flow on the other side. The tangential shear at the separa-
the co-flowing jet. When the jet is incident upon a solid or fluid tion surface forms large-scale eddies (or large size vortices),
boundary, it is referred to as a impinging jet. A number of stud- which are assumed to be good suction creator. These eddies
ies carried out in the past and the consistent research which which are generated all along the tangential shear surface in
is still being done clearly demonstrate the significance asso- large numbers, engulfs the zero or low momentum fluid inside
ciated with the jet flows. This owes to their extensive nature the jet flow. In turn, the mass flow at any cross-section of the
of use, from household appliances to rocket science. The jets jet gradually increases and accordingly the jet spreads along
find application in numerous engineering fields such as air- the downstream direction. Consequently, to conserve the total
craft, rockets, missile, propulsive systems of aircraft, thrust momentum, the jet centerline velocity decreases in the down-
augmenting ejectors, powder metallurgy, metal cutting, food stream direction.
processing, and so on. The instabilities (large-scale eddies) which moves in
Sometimes the jets are also referred as plumes and vice chaotic manner both along and across the jet stream are
versa, however they are not the same in reality. Unlike jets assumed to be responsible in exchanging the matter between
which are momentum driven, the plumes move away from the neighboring fluid flows. That is, they transfer mass,
its source solely because of buoyancy effects. That is, the jet momentum, and energy in transverse direction. In this pro-
moves forward because of pressure imbalance while plume cess, at the boundary of two fluids a zone of finite thickness
travels due to density difference. Rising of smoke streaks in with the continuous distribution of velocity, temperature, and
still environment is an example of plume. The flows which are mass concentration is developed. This region is popularly
both momentum as well as buoyancy dominated re referred known as turbulent jet boundary layer. In a low-speed
as buoyant jet or forced plume. Similar to a jet, the plume subsonic flow, the fluid boundary cannot withstand a pressure
also widens due to mass entrainment of the surrounding fluid difference across it and eventually becomes a free shear layer
across its edges. The shape of a plume is influenced if the with a constant static pressure everywhere. Further, the large

322 13 Supersonic Jets
size eddies created at jet boundary get easily fragmented into either correctly expanded or underexpanded. The supersonic
relatively small-scale eddies while transporting the matter jets are subclassified as overexpanded, correctly expanded,
toward jet centerline. Smaller eddies being the efficient and underexpanded jets.
mixing promoter assist the low momentum fluid entrained
at jet boundary to gain momentum from the jet mass of
high momentum and moves toward the axis of jet. Thus, the 13.2 Free Turbulence Theories
viscous action penetrates toward the jet axis. In reality, the
viscous activity finally reaches at the jet centerline at some We have seen earlier that for a free turbulent flow1 the con-
axial distance downstream. Up to this point, the jet velocity tinuity equation along with the Reynolds Averaged Navier–
(or the nozzle exit velocity) remains unaffected along the jet Stokes (RANS) equations of motion contain more variables
axis. That is, in the vicinity of the nozzle exit, there exists a than the available number of equations; thus they cannot be
finite zone, which is yet not affected by the mixing started fully solved. Although, with the advancement of computing
at the jet boundary leaving the zone nearly potential with a techniques the RANS equations may be solved to obtain the
constant axial velocity throughout. The velocity profile is instantaneous velocities for turbulent flows. But, the comput-
similar to a top-hat shape at the nozzle exit. A schematic ing resources and computation time will be very high. Further,
diagram of a typical subsonic jet is shown in Fig. 13.1a. For in many situations, it is suffice to know only about the mean
subsonic jets, the aforesaid zone is called potential jet core. flow characteristics and, therefore, this approach will be too
Once the viscous activity starts reaching toward the jet axis, expensive and inessential from practical point of view.
the top-hat velocity profile begins to shrink in size in the We know that the turbulent flows have great capability for
direction of jet propagation. Due to mixing process carried diffusion and, therefore, the matter associated with it such as
out by the vortices, the jet centerline velocity decreases mass, momentum, energy (heat), and vorticity can be trans-
continuously after the core region. Higher the mixing, larger ported. These transport processes were modeled by many
the decay in jet velocity. This decay is found to be inversely investigators: Boussinesq (1988), Prandtl (1925), von Kar-
proportional to the axial distance from the nozzle exit. This man (1930), Taylor, and Reichardt. They brought out some
phenomenon is generally known as characteristics decay. important semi-empirical relationships between the mean
The potential core region extends up to about 6 times the flow characteristics. Although, the underlying mechanism of
nozzle exit diameter (D) and the characteristics decay region the aforementioned transport processes may not be correctly
dominates from about 6D to about 12D (Fig. 13.1b). After described by these phenomenological theories; however, they
this, the jet decay is gradual and approaches to almost are good enough in predicting the results in a wide vari-
zero velocity at a far downstream location of about 30D. ety of flow situations. In addition to the transport of matter,
This region is referred to as fully developed region. Here, the turbulent motion also transports the quantities like tur-
even though the jet velocity decreases continuously but the bulent kinetic energy, turbulent viscosity, and eddy size and
velocity profile retains its shape at every downstream axial hence, additional algebraic/differential equations are deduced
location. Therefore, this region is also termed as self-similar from the equations of motion, termed as turbulence models.
region. For the sake of brevity, these approaches will be briefly dis-
Unlike subsonic jet, there is no constant axial velocity zone cussed in this section. For more details, reader is advised to
exists in a supersonic jet due to the presence of shock and refer advanced texts such as Schlichting (1951); Pai (1954);
expansion waves (Fig. 13.2). This region is highly viscous Abramovich (1963); Rajarathnam (1976); and Pope (2000).
dominated referred to as supersonic jet core or simply the jet
core. Here, the core can be defined as the axial extent up to 1 We should note that the problems in a free turbulent flow (such as for a
which waves dominate. In other words, the axial extent from free jet) are of a boundary layer nature. That is, the region of flow analysis
the nozzle exit to the beginning of characteristics decay zone does not extend far in transverse direction as compared to axial direction
and thus, the gradients in transverse direction are large. Consequently,
is known as the jet core. the free jet analysis can be performed using boundary layer equations.
The jets can be classified based on either pressure or Mach In a two-dimensional incompressible flow, these are
number prevalent at the exit of a nozzle or orifice. The jets
can be broadly classified into incompressible and compress- ∂u ∂v
+ =0
ible jets (Fig. 13.3). The jets with Mach number less than 0.3 ∂x ∂y
till which the compressibility effects are ignored are called
incompressible jets. The jet with Mach number more than
∂u ∂u ∂u 1 ∂τ
0.3 are referred to as compressible jets. The compressible +u +v =
∂t ∂x ∂y ρ ∂y
jets are further classified into subsonic (M < 1) jets, sonic
(M = 1) jets, and supersonic (M > 1) jets. The subsonic jets where τ is the turbulent shear stress. Further, it can be seen that the
are always correctly expanded, whereas the sonic jets can be pressure term is not considered in the momentum equation because of
the assumption that the pressure remains constant.
13.2 Free Turbulence Theories 323
Fig. 13.1 Flow field characteristics of a typical subsonic jet

13.2.1 Semi-empirical Theories τij = −ρui uj as
The studies carried out on the transport processes of mass, 2

momentum, energy, and vorticity brought out various phe- τij = 2μt S∗ij − ρKδij (13.1)
3
nomenological theories. In this section, we will discuss only
those which assumed the velocity field in incompressible flow or
because the latter is then independent of the temperature field.
2
−ρui uj = 2μt S∗ij − ρKδij (13.2)
3
13.2.1.1 Boussinesq Hypothesis
Joseph Valentin Boussinesq was the first who tried to deduce where S∗ij is the traceless mean rate of stress tensor defined as
additional equation to attain the closure of the system of equa-

tions. In 1877, he introduced the concept of eddy viscos- 1 ∂ui ∂uj 2 ∂uk
ity (μt ) to express the time-averaged turbulent shear stress S∗ij = + − δij
2 ∂xj ∂xi 3 ∂xk
Shock−Cell
P/P0
Supersonic core Characteristics decay Fully developed zone
X/D
Fig. 13.2 Schematic diagram of different velocity zones in a supersonic jet
Jets
Incompressible Compressible
Subsonic
Correctly−Expanded
Sonic Supersonic
Correctly−Expanded Underexpanded
Overexpanded Correctly−Expanded Underexpanded
Fig. 13.3 Classification of jets
K is the turbulent kinetic energy, given by where u is the time-averaged velocity at distance y from the
surface and τ is the time-averaged turbulent shear stress along
1 x−direction in xy plane.
K= uu
2 i i
13.2.1.2 Prandtl Mixing Length Theory
and δij is Kronecker delta. For thin shear layer, Eq. (13.1) can
In Chap. 11, although, we have described the theory put for-
be rewritten as
ward by Ludwig Prandtl (1925). However, for convenience, it
du is outlined in this section again. Prandtl proposed a turbulence
τ = μt (13.3) model where he assumed that the eddies in turbulent flow
dy
move around similar to the manner in which the molecules
move in a gas. In kinetic theory of gases, the dynamic viscos- In the above transverse motion, we have v > 0. Similarly,
ity is defined as assume that a fluid lump arrives at B from a layer C possessing
the velocity u (y1 + l), retaining its original momentum. In
1 this motion, the velocity of lump will be more than the velocity
μ= ρu rms λ (13.4)
3 prevailing at layer B. Therefore, the change in velocity will
be
where urms is the root mean square velocity of the molecules,
and λ is the mean free path of the molecules. Similarly, du
Prandtl postulated that the eddy viscosity (μt ) can be written u2 ≈ lm (13.8)
dy
as
We should note that in this case, v < 0. The time-averaging
τ
μt = = ρlm ui (13.5) of u1 and u2 canberegarded as the longitudinal turbulent
du

dy velocity fluctuation u . Thus,
where lm is the mixing length of the eddies and u is the 1

turbulence velocity along x−axis. Here, one should note that u = (u1 + u2 ) (13.9)
2
the turbulent viscosity despite of it being inhomogeneous in
nature is assumed to be isotropic, i.e., same in all directions. or
This assumption is valid for a wide variety of flows except
du
flows involving large separation or swirl. In addition, similar u = lm (13.10)
to the kinematic viscosity (ν), a kinematic turbulent viscosity dy
(νt ) can also be defined as
According to mixing length hypothesis, the transverse turbu-
μt
νt = lent velocity fluctuation v is of the same order of magnitude
ρ

as u . Thus,
To understand Prandtl’s hypothesis further, let us refer to the

simplest case of a parallel flow (xy plane) in which the velocity du
varies only from streamline to streamline. The direction of v = lm (13.11)
dy
mean flow (u) is assumed parallel to the x−axis and we have
Physically, the mixing length (lm ) can be interpreted in the
u = u(y) following way. In turbulent flows, the mixing length is essen-
v=0 tially the distance traveled by the fluid lump with its original
mean velocity in transverse direction which must be covered
Since, the fluid passes along the wall in turbulent motion, fluid so as to make the difference between its original velocity
particle coalesces into lumps which move bodily and cling and the velocity into new lamina equal to the mean trans-
together for a given traversed length, both in the longitudinal verse
turbulent fluctuation. Further, the turbulent shear stress

τ = −ρu v is given by
and in the transverse direction, retaining their momentum par-
allel to x−axis. Let us assume that a fluid lump is displaced
from the layer A possessing the velocity u (y1 − l) to a layer 2
du du du
τ= ρlm
dy dy = ρlm dy
2 2
B of velocity u (y) (Fig. 13.4). The lump is displaced over a (13.12)
distance lm known as Prandtl’s mixing length. Since, the lump
has retained its momentum and hence, the velocity of lump Note that, the negative sign in shear stress expression is omit-
will be smaller than the velocity prevailing at B. Therefore, ted in Eq. (13.12). The reason behind it is left as an exercise
the difference in velocities will be to the reader! Furthermore, from Eq. (13.5) the eddy viscosity
becomes
u1 = u (y1 ) − u (y1 − l) (13.6)

2 du
μt = ρlm dy (13.13)
Using Taylor series to expand u (y1 − l) and neglecting higher
order terms, we have
In Eq. (13.13), it is evident that unlike dynamic viscosity, the
du eddy viscosity varies from point to point and depends on the
u1 ≈ lm (13.7)
dy nature of the flow. For pipe flows, Nikuradse conducted a
series of experiments and obtained the following correlation
y
u(y)
u(y+1)
1
C
l m
u(y )
1
B
lm
u(y −1)
1
A
Fig. 13.4 Fluid lump movement between fluid layers
for mixing length (lm ) at large Reynolds number. belief where the lump dimension was small compared to trans-
verse dimension of mixing region. However, it is valid only
lm y 2 y 4
= 0.14 − 0.08 1 − − 0.06 1 − (13.14) in the case of free turbulent flows such as mixing of a jet with
R R R surrounding stagnant air.
where R is the pipe radius and y is the distance from wall
μt = ρc1 b (umax − umin ) (13.16)
surface. For small values of y, Eq. (13.14) reduces to,
lm y where c1 is a constant to be determined experimentally, b

= 0.4 is the width of mixing zone, and (umax − umin ) is the maxi-
R R
mum difference in time-averaged mean flow velocity. From
or Eq. (13.12), the turbulent shear stress for a free jet can be
written as
lm = κy (13.15)
du
τ = ρc1 b (umax − umin ) (13.17)
where κ = is known as von Karman constant for pipe flows. dy
In free shear flows, the mixing length (lm ) is proportional to
half-width (b) of the mixing zone (or shear layer). Several 13.2.1.3 von Karman’s Similarity Hypothesis
different forms of the equation (13.14) were proposed for the In order to determine the dependence of mixing length (lm ) on
free shear flows given in the Table 13.1. space coordinates, von Karman proposed a hypothesis which
Two major shortcomings were noticed in the mixing length states that the turbulent fluctuations are similar at all points
theory of Prandtl. in the flow field and they differ only by time and length scale
factors. It can be shown that the mixing length satisfies the
dy = 0, then
1. In Eq. (13.13), at the center of the pipe where du following equation.
we get μt = 0. However, at this point, the turbulent mixing
is still present and thus, the eddy viscosity should have a du
dy
finite value. lm = κ 2 (13.18)
2. In this theory, the effect of pressure fluctuations on the ddyu2
momentum transport was not considered. Due to pressure
fluctuations, the momentum of fluid lump will not be con- where κ is the von Karman constant. Introducing Eq. (13.18)
stant during its motion over the path lm . Therefore, it will into Eq. (13.12), the turbulent shear stress (τ ) is written as
not be preserved. 4
du
dy
Reichardt carried out an extensive experimental investiga- τ = ρκ2 2 (13.19)
tion on turbulent flows which confirmed above views. Real- d2 u
dy2
izing these limitations, Prandtl further proposed a modified
relationship for eddy viscosity (μt ) given in Eq. (13.16).
13.2.1.4 Taylor’s Vorticity Transport Theory
In proposing the new hypothesis, Prandtl assumed that the
G. I. Taylor also obtained a result similar to that contained
dimensions of the fluid lump are of the same order of magni-
in Eq. (13.12) on the basis of his vorticity transport theory.
tude as the width of the mixing zone, contrary to his earlier
In Prandtl’s mixing length theory, the underlying assumption
Table 13.1 Typical mixing lengths of free shear flows

For plane jet lm = 0.09b
For axisymmetric jet lm = 0.075b
For mixing layer lm = 0.07b

made was that the mean flow velocity (u) remains constant ∂u2 ∂uv ∂p
ρ + =− (13.22)
during transverse motion of a fluid lump; the Taylor’s vorticity ∂x ∂y ∂x
transporttheory
replaced it with the hypothesis that it is the
rotation dudy , which remains invariant during the transverse Since, the pressure term vanishes in free turbulence, thus, the
motion of the lump. Taylor deduced the following equation: temporal averaging of equation (13.22) will give

1 2 du du ∂u2 ∂u v
τ= ρlw (13.20) + =0 (13.23)
2 dy dy ∂x ∂y
Comparing Eqs. (13.12) and (13.20), we have From experimental data, Reichardt established the following
√ empirical relation:
lw = 2lm (13.21)
∂u2
From the above equation, it is evident that the mixing u v = − (x) (13.24)
√ length ∂y
obtained by Taylor’s vorticity transport theory is 2 times
higher than that obtained by Prandtl’s momentum transfer where (x) is called the momentum transfer length and is
theory. Based on his postulations, Taylor concluded that the evaluated experimentally. If (x) were constant, this equa-
diffusion of temperature difference and vorticity in the mixing tion would have been identical to one-dimensional heat con-
region downstream of a circular cylinder follows similar laws. duction equation (Fourier’s law of heat conduction in one
Interestingly, this inference is found to be in close agreement dimension). Further, introducing Eq. (13.24) into Eq. (13.23)
with the experimental observations. It can be interpreted as we get
follows. The axis of rotation of vortices align themselves pri- ∂u2 ∂u2
marily at right angles to the mainstream and to the direction of = (x) . (13.25)
∂x ∂y
velocity gradient. In contrast, in the vicinity of a solid wall, the
vortices in a flow field have axis parallel to the flow direction. Equation (13.25) is known as Reichardt’s fundamental equa-
Because of this reason, the temperature and velocity fields are tion.
similar to each other.
13.2.1.5 Reichardt’s Inductive Theory of Turbulent 13.2.2 Turbulence Models

Flows
The phenomenological theories discussed above made some Although, the semi-empirical theories discussed above are
hypothesis and deduced the velocity law by solving the dif- good enough to provide solutions to simple problems such
ferential equations to which the RANS reduced to under the as boundary layer on flat plate, open channel flows, and tur-
hypothesis. On the other hand, Reichardt’s performed exten- bulent jets and wakes. However, the complex flow situations
sive experimental investigations and proposed a simple for- involving a sudden change in flow direction, recirculatory
mula for free turbulent flow for the case of a two-dimensional flow zones, complex boundary conditions cannot be ana-
incompressible flow. He concluded that velocity distribution, lyzed by these theories. Similarly, problems involving heat
in most of the cases be represented by Gaussian distribu- exchange, chemical reactions, and transport of species in tur-
tion. The laminar stresses are negligible in the free turbulence bulent flows cannot be solved by phenomenological theories
regions. Thus, the theory which has the empirical formula- and requires turbulence modeling. A turbulence model is a
tion based on the experimental data is referred to as induc- computational strategy which together with mean flow equa-
tive theory. The momentum equation for time-averaged two- tions form a system of closure equations. In numerous engi-
dimensional frictionless flow with the aid of continuity equa- neering applications, it is pointless to determine the subtle ele-
tion can be written as ments of turbulent fluctuations and one may be intrigued just
to know how the turbulence influences the mean flow. In those cous flow was considered. Since, the jets are known to be
cases, a turbulence model permits the estimation of the mean unstable, presumably to infinitesimal disturbances, they con-
flow without first figuring the full time-dependent flow field. sidered the typical small disturbance as a Fourier component
Eventually, it helps in deducing the expressions for Reynolds with sinusoidal dependence. With this base, they developed
stresses. In different words, a turbulence model means a set a time-dependent solution using the classical linearized the-
of equations which, when solved with mean flow equations ory to analyze the stability characteristics in the limit of large
allow calculation of relevant correlations and, hence, helps in Reynolds number. Beavers and Wilson (1970) represented
the solution of equations simulating the behavior of real fluids the inviscid flow domain by discrete arrays of point vortices.
in important aspects. However, first realistic attempt to numerically compute and
The turbulence models are classified according to the num- investigate the free jet characteristics in subsonic flow regime
ber of differential equations used. Models in which only can be considered as started in 1974. Although in subsonic
algebraic equation is used for determining turbulent viscos- flow, the absence of waves might make the computations eas-
ity/diffusivity and do not use any additional differential equa- ier, understanding the vortex dynamics precisely and accu-
tion are called zero-equation models. Those which use addi- rate capturing of turbulent structures poses real challenges.
tional equation or equations are called one-equation, two- Grant (1974) used the time-independent Navier–Stokes equa-
equation, or multi-equation turbulence models. The additional tion to study the instabilities associated with the low subsonic
equation may be for transport of turbulent kinetic energy or axisymmetric jets. From the isodines and the stream-function
other such quantities. profiles at various Reynolds numbers, it was found that the
The jets can be studied by using both computational and jet field is dominated by the large-scale vortical structures.
experimental techniques. These techniques will be reviewed Due to initial perturbations, the wavelength of vortex shed-
in the following sections. ding in the jet field was found to be slightly varying. However,
the results obtained from the numerical model were found to
be in reasonably good agreement with the experiments. Mat-
13.3 Computational Techniques for suda et al. (1987) analyzed the underexpanded supersonic
Studying the Jets jet evolution numerically. They adopted the time-averaged
second-order Osher-upwind scheme to solve the Euler equa-
The advancement of computational techniques has made it tions for a compressible ideal gas at varying boundary con-
possible to study the highly complex fluid flow phenomena. ditions. Numerical results were found to be sensitive to the
However, the numerical simulation of jets is still a cum- suitability of the conditions imposed. By the density contours
bersome task because of the requirement for an appropri- plots, it was shown that the initially stable wave that dom-
ate turbulence model to accurately capture small- and large- inated jet structure can be destabilized by the formation of
scale eddies, and the shock cell structures in an imperfectly Kelvin–Helmholtz vortex rings. The numerical results were
expanded jet. Kaushik et al. (2015) performed a broad liter- compared with the experimentally obtained Schlieren pho-
ature survey of the computational fluid dynamics studies on tographs. The researchers observed a turbulent flow with vor-
jets carried out by various researchers. The major outcomes tices and shocks in the former, and a complete shock cell
of this study are presented in the following sections.2 structure in the latter. From this study, it was strongly sug-
gested that a time-independent or a time-converged solution
cannot be expected without making a suitable time-averaging
13.3.1 Preliminary Studies of the time-dependent solution. Agui and Hesselink (1988)
analyzed the instabilities of an acoustically excited subsonic
The early theoretical and numerical studies are limited in co-flowing jet both experimentally as well as computationally
number both due to the time-dependent nonlinear character- using vortex-element method. With the assumption that the jet
istics of the Navier–Stokes equations and unavailability of divergence does not play a vital role in the evolution and devel-
high-performance computing facilities. In early 60s, Batche- opment of streamwise vortices, they modeled a periodically
lor and Gill (1962) mathematically analyzed the stability of excited jet. By using the vortex-element method, the vorticity
steady axisymmetric parallel flows of uniform fluid in the distribution and the mass entrainment on the basis of surface-
absence of rigid boundaries. For this case, a jet at sufficiently to-volume ratio were computed. The numerical results were
high Reynolds number with small spreading angle of vis- compared with results obtained through experiments and it
was seen that the morphology of the flow possesses vor-
2 The content of this section is licensed under a Creative Commons Attri- tices, mainly concentrated into a single large-scale vortex
bution 4.0 International License with permission to reuse; the full article ring. Instead of vortex-element method, Martin and Meiburg
is originally authored by Dr. Mrinal Kaushik in American Journal of
(1991) used vortex filament method to simulate the subsonic
Fluid Dynamics, Vol. 5 No. A, 2015, pp. 1–11. (p-ISSN): 2168–4707,
(e-ISSN): 2168–4715, DOI: 10.5923/s.ajfd.201501.01. jet vortex dynamics. They assumed the flow to be inviscid
13.3 Computational Techniques for Studying the Jets 329
and explained the mechanisms governing the jet evolution equations are primarily used in analyzing turbulent flows
by spatially periodic calculations. The study was intended to with approximations based on the knowledge of flow proper-
the understanding of processes leading to the concentration, ties in order to give approximate time-averaged solutions to
reorientation, and stretching of the vorticity in the uncon- the Navier–Stokes equations. The RANS method attempts to
trolled jet and for the jet induced due to sinusoidal perturba- model all the turbulent fluid dynamic effects using a turbu-
tions in the azimuthal direction. They found that the evolution lence model. Currently, the RANS approach is being used for a
of the vortex rings depends on the ratio of the jet radius to its majority of applications. Two-equation models such as k − ε,
shear layer momentum thickness Rθ . Furthermore, they also Menter’s SST formulations and k − ω models are widely used
observed that the rate at which streamwise vorticity devel- to solve for turbulent kinetic energy in the jets.
ops is strongly affected by the ratio of the streamwise and Thies and Tam (1996) used the k − ε model with
azimuthal perturbation amplitudes. However, the assumption a correction factor as developed by Pope (1978) and
of inviscid flow in the vortex filament technique is obviously Sarkar and Lakshmanan (1991). They attempted to predict
unable to capture the viscous effects in the actual flow field. the flow characteristics of the jets having high convective
Miller et al. (1995) studied the entrainment of the subsonic Mach number. The axisymmetric and non-axisymmetric
jet flows originating from elliptic, rectangular, and triangular jet mean flow results over the Mach number range of
nozzles with aspect ratios of 1:1 and 2:1 using the compress- 0.4−2.0 were compared with the available experimental
ible Navier–Stokes equations, and a passive scalar conserva- data. Interestingly, even though this model was observed to
tion equation with the Fourier heat conduction and the Fickian reproduce the core length and shock cell length precisely, the
diffusion assumptions. These equations were solved numeri- quantitative data such as turbulent intensities were found to
cally without resorting to any turbulence or imposed sub-grid have some error. It is known that the standard k − ε model
models. The results were compared with those of a circular always overpredicts the growth rate of turbulent kinetic
jet of the same equivalent diameter to determine the relative energy in compressible mixing layers such as high-speed
efficiency of non-circular nozzles in mixing enhancement. jets. Although, the dynamic characteristics of fine-scale
Though the flow structure was predicted, the effectiveness turbulence may be same for all turbulent flows, the large
of the compressible or the time-independent Navier–Stokes turbulence structures (having dimensions comparable to
equation in the simulation of subsonic jet or in predicting the local length scale of the flow) are significantly influenced
three-dimensional vortical structures remained unexplained. by the local boundary conditions and geometry. Indeed,
There remains a necessity to obtain the accurate aerodynamic the authors proposed that there is no universally applicable
and aeroacoustic prediction tools for analyzing exhaust noz- turbulence model and it should be chosen depending upon
zle systems and the turbulent jet flow fields. To this end, initial the flow configuration. It creates the necessity to introduce
successful attempts were made using the Reynolds-Averaged the correction terms. Durbin (1996) proposed to limit the
Navier–Stokes (RANS) method. A brief summary of some of eigenvalues of the Reynolds stresses. He mentioned that
the remarkable attempts using RANS is discussed in the next these limits must be between zero and twice of the turbulent
section. kinetic energy. The Durbin-realizable model is found to be
effective up to Mach 3.
Further examination of the compressibility effects on tur-
13.3.2 Reynolds-Averaged Navier–Stokes bulent flows reveals that these effects can be categorized
(RANS) into two basic groups namely, the dilatational compressibil-
ity effects and structural compressibility effects. The dilata-
The robustness of any numerical scheme is very important tional effects are related to the variations of mean mass density
to simulate the supersonic-free jet. This is due to the fact or variations in the instantaneous mass density, whereas the
that the strong discontinuities present in the flow field have structural compressibility has a significant influence on the
a tendency to destabilize the flow. Researchers have investi- structure of turbulent flows. Zeman (1990) and Sarkar et al.
gated the effectiveness of various computational methods in (1991) modeled the dilatation dissipation based on the results
simulating the shear-driven supersonic jet flows. The appli- of Direct Numerical Simulation (DNS) of isotropic com-
cation of a reliable technique in computing the supersonic jet pressible turbulence subjected to strong compression. Heinz
flow field characteristics originated in 1996. To this end, the (2003) proposed a model to account for the structural com-
RANS method is the most conventional approach to calcu- pressibility effects by remodeling the turbulence production
late the shear flows, consisting of time-averaged equations term. These compressibility corrections terms are employed
for fluid flow motion. Reynolds (1895) proposed the method in the standard k − ε turbulence model. Various modified lin-
of Reynolds decomposition, where an instantaneous quantity, ear two-equation turbulence models, developed specifically to
for example, u is decomposed into ū and u , i.e., into its time- improve the predictions of jet flow, were compared to the stan-
averaged and fluctuating quantities, respectively. The RANS dard RANS turbulence models. Tandra et al. (2006), devel-
oped a modified model of k − ε with three additional terms and 10. The numerical results were also compared with the
to predict the flow properties of high subsonic and supersonic experimental data. They concluded that the ideal configura-
jets up to Mach 2.8. Combining the correction terms proposed tion for an efficient mixing should have four or less number
by Durbin (1996), Sarkar et al. (1991), and Heinz (2003), of injectors with a low aspect ratio orifice in order to prevent
the modified k − ε model was formulated. The performance the premature vortex interaction and excessive diffusion. It
of the developed numerical model in predicting the free-jet was found that the modification of inside supersonic waves
flows, the jet propagating between smooth flat plates, and the do not contribute to the mixing enhancement. The compari-
jet impinging on a solid surface was tested. The jet center- son plots, depicting the numerical and experimental results,
line velocity, centerline pressure, radial velocity and pressure exhibited some discrepancies, which necessitated the adop-
profiles for the jets with Mach numbers of 0.67, 2.22, 2.3, tion of improved numerical schemes.
and 2.8 were compared with the available experimental data.
A good agreement with the experimental data was obtained
demonstrating the efficacy of the modified k − ε model. Berg 13.3.3 Large Eddy Simulation (LES)
et al. (2006), compared the two-equation turbulence mod-
els of k − ω and standard k − ε in calculating the turbulent From the above discussion, it is clear that the important char-
flow properties of a rectangular Mach 0.17 free jet of aspect acteristics of jets such as core length, jet spreading rate,
ratio 2.0. The results concluded that the simulation by k − ε mixing and turbulent characteristics could not be predicted
model with parameters adapted from experimental results best accurately by any of the RANS models. These models were
predicted the jet centerline velocity decay. It was found that found to suffer from limitations for the jets, which have
the conventional density- based algorithms become inefficient significant three-dimensionality, compressibility, and high
and often unstable at low Mach numbers. temperature. Further, the RANS method can only provide
Birky and Page (2001) suggested the adaption of modi- time-averaged turbulent information. However, in spite of
fied pressure based methodology with compressibility cor- these limitations, the RANS approach frequently captures the
rections in simulating the sonic jet at underexpanded condi- trend exhibited in experiments, altering the geometry as well
tions. They proposed that the pressure-based approach with as the flow conditions. This fact coupled with the advantage
a correctly formulated compressible treatment can accurately that the RANS solutions are relatively inexpensive. That is,
compute flows containing both high and low Mach number these methods still have a place for designing nozzle systems.
zones. This numerical scheme uses a basic first-order upwind Further improvements in RANS approach for predicting the
scheme discretized in an implicit manner. They adapted a jet characteristics will be incremental at best. Hence, the next
higher order discretization scheme using a Total Variation logical choice is to adopt the unsteady Large Eddy Simulation
Diminishing (TVD) method as an explicit correction to the (LES) and DNS techniques, for more accuracy and fidelity.
basic upwind scheme. The high-speed jet issuing into a stag- At present, owing to the crucial role of turbulence dynamics
nant environment was selected as flow domain. For NPR 3.5, and complex geometries, the LES has a growing importance
the results obtained from the standard k − ε with and without in the field of jet simulation. Hilgers (2000), carried out the
compressibility corrections are compared with their experi- LES of a jet and found that the Strouhal number grows with
mental counterpart. By varying the NPR from 3.5 to 30, the amplitude of the major actuation parameter, which enhances
location of the Mach-disc was found to be in close agree- the mixing and jet spread rate. The Strouhal numbers, Sth
ment with the experimental data. It was observed that, the and Sta corresponding to the preferred Strouhal number of
k − ε model with compressibility correction produces a faster the jet and twice its value, respectively, were estimated for
decay as compared to the experimental results even though all the Reynolds numbers investigated. It was shown that the
the shock cell lengths were found to be same. The compress- jet spread was larger at higher Reynolds number. However, to
ibility correction to the turbulence model was found to have validate the accuracy of the results obtained they did not com-
almost no effect on the shock cell structure. pare it with the experimental data or with any other numerical
Chauvet et al. (2007), used the Spalart–Allmaras turbu- schemes.
lence model in predicting the jet flow characteristics. From the Terrier and Lu (2001) used the LES scheme in simulat-
series of RANS computations and the experiments conducted, ing the jet, which is controlled by a step incorporated right
they showed that the underexpanded sonic jets experienced after the throat of the nozzle. They performed both experi-
augmented mixing under the influence of radially injected mental and computational investigations. The control device
secondary jets. The Spalart–Allmaras model was used to cor- was found to provide good thrust recovery in comparison to a
relate the strong distortion of the jet interface under the action nozzle without the step at overexpanded conditions. The sim-
of the longitudinal vortices. The researchers studied the noz- ulation was performed with FALCON for a full-scale nozzle
zle configurations having 2, 4, and 8 numbers of radial injec- design using finite volume approach using structured grid in
tions and secondary orifice jets of aspect ratios 0.5, 1.0, 2.5, order to achieve the greater conservation qualities than the
comparable finite difference codes. The overexpanded nozzle this study. The modified LES variants are formulated to elim-
produced a complex shock–boundary layer interactions at the inate the minor discrepancies in the flow field.
nozzle exit. However, it was reported that when a step was Wang and McGuirk (2013) carried out LES of a supersonic
added near the downstream of the nozzle throat, it reduced jet issuing through a convergent–divergent nozzle of rectan-
the overexpansion, thereby yielding a better pressure recov- gular cross-section of high aspect ratio. They have adopted
ery in the divergent portion of the nozzle. The computed value rescaling/recycling method (R2M) of LES in generation of
of Cfg (thrust coefficient) showed a 2.5% improvement in the the inlet conditions. Two different models namely, Piomelli-
performance as compared to the plain jet. These results were Guerts Sub-Grid Scale (PGSGS) and Smagorinsky Sub-Grid
verified by the subscale model testing. Scale (SSGS) models were developed and compared with the
For the excited and forced jets, the LES scheme was tested results obtained from RANS model. At the overexpanded noz-
for accuracy and stability. Maidi and Lesieur (2005) used zle pressure ratio of 2.5, the instantaneous turbulent eddies
LES filtered structure functional model to simulate the high and the instantaneous axial velocity of the jets were visualized
Reynolds number turbulent flows (Re ≈ 36000) for Mach 0.7 for all the cases. The eddy structures were found to be more
and 1.4 jets. They used a basic velocity profile close to a complex in the results obtained through PGSGS modelings
top-hat, combined with small inlet perturbations. At Mach than obtained through SSGS modelings. Also, the appear-
0.7, they observed the shedding of quasi-axisymmetric vor- ance of the Kelvin–Helmholtz-like structures in the SSGS
tex rings with longitudinal hairpin-like vortices from the jets. simulation was interpreted as a failure of the model to cap-
However, these were reported to be disappeared in Mach 1.4 ture the five correct behaviors of the boundary layer within
jet and instead primary vortices undergoing alternate pair- the nozzle. The velocity profiles at the major and minor axes
ing was observed leading to a 27% increase in jet core length. at various axial locations were compared for both the PGSGS
From the LES analysis of Reynolds stresses, it was shown that and SSGS models. They found that the PGSGS model with
the turbulence intensities were decreased with an increase of adopted rescaling/recycling method was in good agreement
Mach number. Mrinal et al. (2006) extended the work to study with the available experimental data for better prediction of
the effect of different types of upstream forcing on Mach 0.7 the shear layer growth. It leads to the improved capturing of
and Mach 1.4 circular jets. They conducted the tests at four shock cells and jet core length.
different excitation frequencies and predicted the optimum Suto et al. (2004) tested two SGS models that are the
frequency at which the jet spreading was maximum. The high- Smagorinsky Model (SM) and the Dynamic Smagorinsky
est spreading rate was found for Mach 0.7 jet having varicose– Model (DSM) to carry out the LES of turbulent round jets
flapping excitation with a core length reduction of 27% in at four different Reynolds numbers in the range 1200–106 .
comparison to unexcited jets. For Mach 1.4 jet, the flapping The parameters of the DSM model were calculated using dif-
excitation at the second sub-harmonic frequency was found to ferent levels of space filtering, which is unlike the SM model
be the most efficient jet control technique with a core length where the parameters were empirically deduced. They found
reduction of 28%. It was interpreted to cause a quick transition that the quantitative accuracy of the DSM model was far better
to turbulence, which leads to the more enhanced jet mixing. than the SM model, which overpredicted the mean velocity
The application of LES for the prediction of jet charac- of the jet. The model coefficient was found to be large in the
teristics of an actuator controlled device was carried out by fully developed zone of the jet, whereas it was found to be
Faivre and Poinsot (2004). They developed an actuator device small in the developing zone. Hence, they recommended the
made up of four small jets feeding the primary jet flow, ori- optimization of the model coefficient in the DSM model for
ented to add an azimuthal component to the velocity field. The successful prediction of streamwise variation of the jet.
experimental study of three different configurations of the jet To gain an insight of the flow characteristics, most of the
(α = 90◦ , 45o at 30 mm and α = 90o at 8 mm from the nozzle LES of jets have focused mainly on benchmark experiments
exit) was quantified by Schlieren photographs, PLIF, and hot of round nozzles. Thus, the modified nozzle geometries as
wire anemometry results. It was found the most efficient con- jet controls and for aeroacoustic advantages may not give the
figuration in terms of mixing and jet spreading enhancement expected results if the nozzle geometry itself is not included.
device, where actuation jets are located close to the nozzle Therefore, the LES method was found to be effective only
and oriented to provide maximum swirl injection (α = 90◦ with the simplest turbulence models in the simulation of jets.
at 8 mm from the nozzle exit). The LES of both forced and Additional efforts are required to adapt current LES meth-
unforced actuators located at α = 90◦ were also performed ods to handle such complex geometries. The computational
for better understanding of the production of vortices and time and cost of LES also increases exponentially as com-
mixing phenomenon at this configuration. The instantaneous pared to RANS turbulence schemes making the LES one of
vorticity visualization has revealed the presence of secondary the laborious schemes. A promising numerical scheme is the
vortices due to actuator. However, the physics behind the pro- hybrid RANS/LES simulation, where the upstream nozzle
duction of secondary vortices was far from understanding in
wall boundary layers are calculated using RANS and the jet ing the edges and corners and developed the compatibility
regions are modeled using LES. conditions for inlet/outlet and wall/outlet boundaries. Both,
Many researchers intended to combine the advantages DNS and LES studies were performed. It was seen that three-
of both LES and RANS in affordable computational costs. dimensional NSCBC brought a drastic reduction of flow dis-
This paved the way to hybrid RANS-LES schemes. Geor- tortion and numerical reflection, even in the regions of strong
giadis and Bonis (2007), proposed an alternative method for transverse convection. The treatment for edges and corner
RANS in numerically simulation of the turbulent jets issu- boundaries with the combination of multidirectional charac-
ing from the exhaust nozzles of an aircraft engine. Since teristics and compatibility conditions yielded the stable and
the RANS turbulence models were reported with limitations accurate solutions.
for jets having significant three-dimensionality, compressibil- Furthermore, it is now established in the literature that
ity, and high temperature. The researchers proposed a new in LES approach, only larger energy-containing eddies are
approach of hybrid RANS/LES methods, where RANS was resolved and the effects of the smaller eddies are modeled
used to model the small-scale turbulence in wall boundary using a Sub-Grid Scale (SGS) model. In all the previous LES
layers and LES was adapted in large-scale mixing regions. studies on jets have utilized the linear SGS model, which is
In this method, the LES is coupled with an explicit sub- isotropic in nature. However, for the free-jet modelings the
grid scale (SGS) turbulence model. However, a more effi- shear action and mass engulfment between the jet core and
cient hybrid RANS/ILES (LES with an implicit SGS turbu- the ambient fluid are unstable and anisotropic. Therefore, to
lence model) approach to effectively simulate the subsonic enable the adequate modelings of turbulence anisotropy and
and supersonic jets from realistic nozzles is suggested by energy back-scatter, the use of more complex nonlinear SGS
Lyubimov (2008). The flow inside the nozzle was calculated models seems to be sensible. Hence, to compare the efficacy
using unsteady RANS approach with relatively coarse grids. of LES of subsonic plane jet using both linear and nonlin-
The ILES approach is realized in the jet region, where a fine ear SGS models, the centerline velocity decay, jet spread-
grid was employed. The fifth-order monotonic scheme with ing rate, mean velocity distributions and turbulence inten-
additional reduced viscosity for approximating the convec- sity distributions were investigated by Liu et al. (2008). In
tion terms in Navier–Stokes equations ensures the accuracy of this study, the researchers compared the results of mixed-
this method. The calculations were performed for jets issuing nonlinear SGS models namely, Kosovic (1997), Leray (1934)
from different nozzle configurations such as conical, chevron, and Lagrangian Averaged Navier–Stokes (LANS−α, 2001)
and coaxial. The effects of Mach number and jet temperature with the Smagorinsky (1963) and Yoshizawa (1993) linear
at the nozzle exit on the fluctuations in longitudinal veloc- SGS models. The effectiveness of all these five numerical
ity component in mixing layer was investigated for conical methods in predicting the jet flow characteristics were also
nozzles. The level of turbulence intensities was predicted for compared with the results obtained through DNS and experi-
both jet axis and in shear layers. The turbulence intensities ments. They observed that except for LANS−α model where
calculated with the nodes ranging from 5 × 105 to 1.2 × 106 the small scale vortex stretching is suppressed, the difference
in computational grids were found in accordance with the in results obtained through other models were insignificant.
experimental results. However, the improvements are needed Therefore, the researchers established that the corrections in
to incorporate for developing the proper interfaces between basic LES models do not show a clear improvement.
the RANS and LES simulation zones.
The LES method requires higher order numerical schemes, 13.3.4 Direct Numerical Simulation (DNS)
which do not lend themselves well to the complicated grid
topologies. Furthermore, the treatment of boundary condi- The Direct Numerical Simulation (DNS) technique involves
tions is one of the most critical issues in CFD. This is because the direct solving of the Navier–Stokes equations numerically,
of compressible solver’s accuracy, which is strongly sensi- without using any turbulence model. This essentially means
tive to boundary condition, and may be spoiled by spurious that the whole range of spatial and temporal scales of the
numerical reflections generated at open boundaries. A gen- turbulence must be resolved. All the spatial scales of the tur-
eral methodology in organizing the incoming and outgoing bulence must be resolved in the computational mesh from the
waves based on the Navier–Stokes Characteristic Boundary smallest dissipative scales (Kolmogorov micro-scales), up to
Conditions (NSCBC) was first suggested by Poinsot and Lele the integral scale associated with the motions containing most
(1992). Lodato et al. (2008) extended the three-dimensional of the kinetic energy. The DNS is found to be the most accurate
NSCBC to account for convection and pressure gradients in among all the available numerical schemes. In the following
boundary planes. The system is augmented with an additional sections, the work carried out by the various researchers using
transport equation for a passive scalar Z (diffusion accord- DNS have been discussed.
ing to the Fick’s law), which carries information on turbulent In a recent study performed by Ball et al. (2012), many
scalar mixing. They proposed a systematic strategy in deal- issues related to simple turbulent round jets were studied
both computationally and experimentally. The role played by passive scalar transport equations. They visualized the vor-
complex mathematical approaches such as LES and DNS on ticity magnitude contours for the unforced and forced shear
the effects of Reynolds number and inlet conditions (exclud- layers. Observations indicated that under forced conditions
ing swirl) were extensively investigated. They found that the when the actuators are co-planar to jets than being normal
structure of turbulence in a jet depends increasingly on DNS, the jet spreads quickly. The plot of centerline velocity for
whose application is being pushed to even higher Reynolds the unforced and forced cases showed the decrease in core
numbers. They also stated that the inlet conditions of the length in the latter to almost half of the former. The forcing
jet in far field affect its development and structure. For bet- frequency was not varied in this study. The quantification of
ter understanding of the flow field, they proclaimed that the the heat radiated in both the cases enabled them to conclude
study of vortical structures, their energy, and modes are of that the mixing was pronounced more in forced jet case. They
vital importance. However, the authors did not focus upon also observed a small amount of reverse flow near the nozzle
some of the vital parameters including the thermo-physico- lip in the forced case, which was not clearly explained.
chemical effects (such as density variation, heat release, etc.), The DNS for the vector controlled jets for improved jet
the effects of passive or active controls, scalar contaminant mixing was studied by Tsujimoto et al. (2011). They oscil-
transports, and the jets operating in absence of any geometric lated the inflow jet periodically at an oscillating angle of 5◦
modifications (such as lobes, tabs, etc., at the nozzle exit). perpendicular to the jet axis and varied the oscillating fre-
The three-dimensional spatially evolving turbulent jets quency from St = 0.01 to St = 0.4 in steps. The visualiza-
were numerically computed using DNS of the compressible tion studies of the instantaneous vortical structures showed
Navier–Stokes equation and the outcomes were validated with the exhaust jet, exhibiting three modes: the wavy mode
the available experimental data by Stanley et al. (2002). In this (St = 0.01), the bifurcating mode (St = 0.04), and the flap-
work, the DNS with higher order space and time accuracy ping mode (St = 0.4). The centerline velocity and Turbulent
through appropriate schemes to handle inflow and outflow Kinetic Energy (TKE) for the three modes were processed and
boundaries were claimed to successfully represent a spatially the jet mixing characteristics were quantified using passive
evolving plane jet. The evolution of mixing process in turbu- scalar statistical entropy along with fluctuation components
lent planar jets are characterized by the probabilistic density of statistical entropy. The researchers suggested that the mix-
functions for the passive scalar. They proposed that in the core ing efficiency was improved for the flapping and bifurcating
of the self-similar region of a fully turbulent jet the mixing modes as compared to their uncontrolled counterpart.
process is dominated by small-scale mixing, whereas near the The active control of jet using fluid injection was simulated
jet boundary the engulfment of co-flowing fluid is predomi- by Lardeau et al. (2002). They used DNS numerical scheme
nately done by the large-scale structures. They showed that with realistic inflow conditions to analyze the performance
the development from interacting shear layers near the nozzle of low Reynolds number jet mixing in the presence of two
to the self-similar jet can be captured, albeit at a moderate secondary control jets with 10% mass flux of the main jet ori-
Reynolds number. ented at an angle of 45◦ inward to the main jet. They studied
Gohil et al. (2012) adopted the DNS approach for various four cases by inducing swirl, pulsed and non-pulsed with con-
low Reynolds number flows through circular jets, in order trol jets in-phase and out-of-phase with the freestream. They
to understand the instabilities associated with them. They carried out vorticity visualizations and plotted the fluctuating
obtained the critical Reynolds number from the perspective of kinetic energy and also recorded the effects of the controls on
unsteadiness of axisymmetric shear layers within the compu- the mixing properties of the main flow by solving a passive
tational domain. When the Reynolds number was increased scalar equation. A rapid decrease in the mean longitudinal
above the critical value, the mode changed from helical to velocity was observed in the case of pulsed control jet with
axisymmetric. They introduced a small amount of perturba- no swirl present in phase with the primary jet. They observed
tion in the form of noise to the flow, which had a profound that the use of swirl effects and phase changes were the least
effect on the entrainment rate, jet half-width, and the vorticity favorable. It was also found that the pulsed secondary jets
thickness. However, it was seen that beyond 3% perturbations enhanced the rapid mixing in the main flow as compared to
in the flow made it turbulent throughout. The DNS approach other cases.
was implemented to calculate the complex shear flows such Tsujimoto et al. (2006) simulated the mixing enhancement
as jet issuing from a modified nozzle exit for the prediction in the parallel jets and combined jets in subsonic regime using
of flow structures and noise characteristics within acceptable DNS by varying the angular inclination of the jets having a
limits. Freund and Moin (1998) used DNS to study the effect fixed distance between them. A hybrid scheme of sixth order
of low Reynolds number fluidic actuators at the nozzle lip in the axial direction and the Fourier series in the azimuthal
in forcing a Mach 0.8 jet. Instead of simulating them explic- direction was adopted in domain discretization. The vortices
itly, they modeled the actuators and the mainstream by adding produced by the different cases of unexcited and excited jets
source-terms to the compressible form of Navier–Stokes and were quantitatively and qualitatively visualized by the DNS
results. The acquired turbulent intensity and velocity decay computationally impractical in the foreseeable future for all
plots were compared with available experimental data. The but the simplest configurations. Rembold et al. (2002), per-
parallel jets with varying distances were simulated and it was formed the DNS of a Mach 0.5 jet exiting from rectangular
found that these were ineffective in efficient mixing. This is nozzle at Re = 2000. For spatial discretization, they used
because, the interaction of the two jets begins at a far down- the fifth-order compact upwind-biased scheme for the con-
stream location where the vortex break down and up to this vective terms and a sixth-order compact central scheme for
location and the parallel jets behave as individual jets. The the diffusive terms. They triggered the transition process by
flow visualization and streamwise velocity contours revealed the most unstable symmetric linear Eigenmode of a parallel
that the impingement causes the abrupt generation of smaller laminar jet profile at the nozzle exit. They observed the tran-
eddies irrespective of the jet inclination angle and the presence sition process, in which the initial two-dimensional distur-
of large scale eddies before impingement. Also, the enhanced bances in the inflow region of the laminar jet rapidly become
mixing was observed in case of combined jets with low angle three-dimensional turbulence. The shear layer roll-up for the
of inclination. laminar jet and a subsequent rapid transition to small-scale
The DNS for pulsed jets at low Reynolds numbers was turbulence was observed from the instantaneous and statisti-
developed by Mouldon and Acharya (2010). They performed cally averaged results. Because of the vortex shedding from
the simulation by varying the pulsing frequency of the jet the jet periphery, the symmetrical rectangular jet cross-section
characterized by Strouhal numbers (St) at 0.2, 0.4, and 0.6 got distorted to an elliptic one. Furthermore, it is understood
for sine wave and compared it with the unpulsed cross jet. from the vortex theory that size of a vortex is proportional to
The visualization of the wake vortices, velocity vectors, tem- the radius of curvature of the surface from which it is shed.
poral frequency spectra, and jet trajectories for all the pulsed Since the relatively larger eddies are shed along the major axis
and unpulsed cases showed that the pulsing greatly increased and smaller ones along the minor axis, the jet spread rate was
the jet spreading rate and enhanced mixing properties of the found to be rapid along the major axis direction due to higher
jet. At St = 0.6, in addition to efficient lateral penetration of mass engulfment. This differential rate in spreading leads to
cross-flow into the jet, they observed the elimination of wake rotation of axes known as “axis-switch” (Fig. 13.5).
vortices in the jet. A significant asymmetry was observed in The impingement of a high-velocity jet on an object was
the flow at St = 0.2. However, the unpulsed cases were also studied extensively by the researchers due to their wide appli-
not clearly explained by the authors. cations including the rockets and Vertical Take-off and Land-
The passive control of jets using DNS approach was found ing (VTOL) aircraft. A typical schematic of the impinging jet
to give results with exceptional accuracy for various con- is shown in Fig. 13.6. The fluid stream at a high velocity and
trol devices. Cai et al. (2010) performed numerical simula- thus with high kinetic energy issuing from a nozzle is made
tion using DNS method and conducted experiments to com- to impinge on a flat plate. The characteristics of these jets
pare the mixing characteristics of circular nozzle, V-shaped involve complex flow evolution and interactions due to the
notched nozzle, and an A-shaped notched nozzle of the same vorticity and turbulence dynamics. Because of these factors,
circular cross-sectional area. The dye visualization technique the numerical modeling of the impinging jet becomes a chal-
revealed that in addition to azimuthal vortex rings created lenging task. Further, it is also seen that in all the previous
because of the shear layer, the streamwise vortex pairs (both studies, DNS of cold jets was performed ignoring the heat
inside and outside the vortex rings that spread outward as the transfer aspects. Hattori and Nagano (2004) computed the
vortex rings move downstream), greatly enhanced the entrain- plane turbulent impinging jets with heat transfer using DNS.
ment and mixing of the ambient fluid with the jet core. It was The flow characteristics and turbulence statistics along with
seen that the production of streamwise vortex pairs for the the heat transfer data were obtained to understand the influ-
notched circular nozzles owned peak and trough locations ence of the nozzle exit to plate distance (H) on the impinging
unlike their circular counterpart. From the results of instanta- jet. In terms of the Nusselt number, the local heat transfer rate
neous vorticity visualizations, they concluded that the mixing was found to increase with a decrease of H. The researchers
characteristics of the three types of nozzles is such that the observed the presence of a second peak in the developing
jet spreads faster in the peak-to-peak and trough-to-trough region of the wall jet for lower values of H. It was also seen
cross-planes than in other directions. Thus, the notches in the that the turbulence intensity in the direction normal to the wall
nozzle exit significantly enhanced jet mixing as shown by was promoted in the near wall region.
the centerline velocity decay plot. In contrast to the RANS From the above discussion, it is evident that with
and LES approaches, DNS methods successfully provided advancement of computational resources the researchers
the entire turbulent energy spectrum by resolving all turbu- used various numerical schemes namely, RANS, LES, and
lent scales ranging from integral length scale down to the DNS for the simulation of jets. All these techniques have
Kolmogorov scale. Although, it avoids the limitations asso- their own merits and demerits. It was found that the RANS
ciated with turbulence modeling, DNS methods may remain approach is having serious limitations when the jets have
1 2 3
Rectangular nozzle exit

Rectangular nozzle exit
Jet cross−section at 1 Jet cross−section at 2
Fig. 13.5 Schematic diagram of a typical axis-switch phenomena
D Confinement plate
H
Potential core
Impingement plate
Fig. 13.6 Schematic diagram of a jet impinging on a flat plate
significant three-dimensionality, compressibility, and high technology, the DNS scheme with extreme requirements on
temperature. The implementation of correction factors to the grid resolution and resulting computational expense is not
RANS approach, in general, improves the accuracy of the viable for realistic geometries and flow conditions in the fore-
mean velocity field and quantities such as jet core length. seeable future. However, the LES method offers a good trade
However, they still suffer from inaccuracies in the prediction off between accuracy and computational efficiency. It is clear
of turbulent kinetic energy. Though the LES approach has that the LES-based methods surpass the accuracy of RANS
delivered results with good accuracy, it is also found to result for most of the problems. In addition, LES provides accurate
in errors when applied to complicated flow structures. The unsteady turbulent information, which RANS approach fails
researchers investigated the DNS approach to calculate the to provide. However, there exists a barrier to utilize LES,
entire turbulent energy spectrum by resolving all turbulent instead of RANS, as the method of choice for complex jet
scales down to the Kolmogorov scale. However, from the flow fields. The hybrid RANS/LES approach may be an
standpoint of the availability of computational resources, efficient tool in predicting the flow field by combining the
the DNS approach suffers from being computationally advantages of both RANS and LES schemes which, still has
intractable for high Reynolds number cases and realistic limitations due to complex grid topologies at the interface.
geometries. Even with the skyrocketing advances in today’s These methods must be extended to handlecomplicated grid
schemes in order to make LES viable for future work.
13.3.5 Some Specific Computational Studies on the accuracy of the predictions. For k − ε model, the jet core
the Jets is clearly seen which extends up to XR = 14.6 from the nozzle
exit, where R is the radius of nozzle exit. The jet core length
This section presents the analysis of supersonic jets over (Lc ) is usually defined as the distance from the nozzle exit
the range of Mach numbers from 1.4 to 2.0. Both correctly to the beginning of centerline velocity decay. The tip of the
expanded (on-design) and underexpanded or overexpanded core is defined as the point where the value of the centerline
(off-design) jets are discussed. We know that in correctly jet velocity has dropped to 0.99 times the nozzle exit velocity.
expanded supersonic jet, the working fluid (air) undergoes The velocity decay has been noticed just after the supersonic
isentropic expansion throughout a converging–diverging noz- core. The decay is continued till it reaches freestream condi-
zle and the exit pressure equals the ambient pressure. But in tions. The classical relation of velocity decay is given by
overexpanded or underexpanded nozzles, the exit pressure
following isentropic expansion throughout the converging– Uc (x) Lc
= (13.26)
diverging nozzle is either more (underexpanded) or less (over- Ue x
expanded) than the ambient pressure. The outflow bound-
ary conditions are specified accordingly. The pressure outlet Here, it should be noted that the nozzle exit turbulent char-
boundary condition applied in subsonic jets is used for on- acteristics in the experiments are not reported. In Fig. 13.7, it
design supersonic jets. A convective boundary condition is is seen that k − ε model predicts the jet core length as 14.6R
used for underexpanded and overexpanded jets. Both cold and but the Spalart–Allmaras one-equation model considerably
hot jets are analyzed for on-design conditions. However, only underpredicts the core length to about 4.4R. The subsequent
cold jets are simulated for the off-design cases. The ambi- velocity decay predicted by both the models differs from the
ent condition for pressure and temperature are taken to be experimental data in the mixing zone. However, the Spalart–
101325 Pa and 300 K. The density of air is taken to be 1.225 Allmaras one-equation model shows the larger discrepancy.
Kgm−3 . The dynamic viscosity is computed using Sutherland The uncertainty about the initial turbulence statistics is one of
formula. The results obtained using the k − ε and Spalart– the reasons of these discrepancies. However, it is apparent that
Allmaras turbulence models are compared with the available k − ε model needs improvement and the Spalart–Allmaras
data. The computational mesh with 30000 cells over the 40R one-equation model is also not adequate.
× 10R domain is used. A fixed time-step of 10−4 s with 20
iterations per step is used for the computations. 13.3.5.2 Mach 1.4 Cold Jet Analysis Using k − ε
Turbulence Model (Off-design Conditions)
13.3.5.1 Mach 1.4 Cold Jet Analysis Using k − ε and Off-design Mach 1.4 supersonic jet is analyzed for three
Spalart–Allmaras Turbulence Models exit-to-ambient pressure ratios PPexit
a
representing underex-
(On-design Conditions) panded and overexpanded jets. The exit pressure ratios are
The nozzle exit conditions are specified in the Table 13.2. The taken to be 1.445 and 3.15 for underexpanded jets and 0.8 for
jet inlet temperature shows that the jet is cold. The nozzle exit overexpanded jet. The ambient conditions for pressure and
pressure is taken the same as the ambient pressure, i.e., the temperature are taken as 101325 Pa and 300 K, respectively.
nozzle is operating at the design condition. As mentioned The density of the ambient air is taken as 1.225 kgm−3 . Only
above, on-design operation of a jet is defined as the condition the k − ε model with standard model constants is used for
for which the nozzle exit static pressure is identical to the simulations of the off-design jets. The computational domain
ambient static pressure. The on-design Nozzle Pressure Ratio is kept the same as that of on-design conditions but to resolve
(NPR) defined as the ratio of settling chamber total pressure the rapidly changing flow field the mesh with 40000 cells is
to ambient static pressure PP01a , is a fixed value for each given used.
Mach number. For jets operating at on-design NPR, pressure Since at off-design operating conditions the initial jet is
gradients are very small in the entire flow domain and waves either overexpanded or underexpanded. As a consequence,
in the flow domain are either absent or typically weak. The the shock waves and expansion waves appear in the jet flow.
principal driving mechanism for on-design jet development These waves may be quite strong if the exit pressure ratio dif-
is turbulent mixing in the jet shear layer. fers significantly from the design pressure ratio. For a given
The Spalart–Allmaras one-equation turbulence model is nozzle geometry, the exit jet Mach number is fixed regardless
also used for the simulation of correctly expanded Mach 1.4 of NPR, assuming the NPR is high enough to fully estab-
jet. The inlet conditions and other parameters are kept identi- lish supersonic flow at the nozzle exit. At NPR values suffi-
cal. Figure 13.7 compares the variation of axial velocity along ciently close to the design point, only weak shocks are present
the jet centerline for inlet Mach number of 1.4 computed using in the jet flow. The driving mechanism is mainly being the
k − ε and Spalart–Allmaras turbulence models. The experi- exit pressure ratio. Figure 13.8 shows the density contours
mental data due to Panda and Seasholtz is used to evaluate in an underexpanded Mach 1.4 jet operating at an exit pres-
Table 13.2 The nozzle exit conditions

M Texit (K) Uexit ms−1 Total pressure (Pa) Total temperature (K)
1.4 216 411 321700 300
Fig. 13.7 Centerline velocity decay of Mach 1.4 jet
sure ratio of 1.445. The initial expansion fan from the nozzle in Fig. 13.10 for center-to-center distance of 2.5R clearly
lip reflects from the sonic line in the shear layer near the jet show the development of individual potential core and jet
boundary as shock, which, in turn, reflects back as expansion interaction further downstream. Subsequently due to entrain-
fan and the process continues downstream till the pressure is ment of ambient fluids, the jet diameter increases and the indi-
adjusted. The consequent cellular structure and formation of vidual jets merge together to form a single jet. It is observed
Mach disks in the major plane of symmetry is clearly visible that after merging, the jet is no longer straight. The velocity
in the figure. variation along the three representative lines namely, center-
lines of the top and bottom jets and the line midway between
13.3.5.3 Mach 2.0 Twin Jet Analysis Using k − ε the jets, are shown in Fig. 13.10. The figure shows that the
Turbulence Model (On-design Conditions) extent of constant velocity in twin jets as compared to a single
The interaction of two supersonic jets issuing from the nozzles jet is decreased considerably. The fluctuations in the core is
of equal radius (R) operating at the correctly expanded state is relatively increased but is increased significantly in the down-
analyzed. The center-to-center distance (P) between the jets stream till the end of the computational domain. The velocity
have been kept as 2.5R and 3R. The computational domain fluctuations on the center- lines of the two jets are asymmetric,
of size XR = 40 and R = 20 is discretized with a structured
Y indicating a strong nonlinear interaction.
mesh of 66000 cells for this case. One part of the mesh close Figure 13.11a and b, show similar results for the twin jets
to the jets is shown in Fig. 13.9. with P = 3R. The jet behavior in this case closely resembles
In this study only k − ε model with standard constants is P = 2.5R case. However, with increasing gap between the jets
used to simulate the twin jets. The velocity contours presented the interaction effect is reduced slightly.
Fig. 13.8 Mach 1.4 jet (underexpanded)
Fig. 13.9 Computational mesh consisting of 66000 cells
Based on the above studies on supersonic jets using differ- short core lengths. It gives a nondimensional core length of
ent turbulence models following conclusions may be drawn. about 4.4 for M =1.4 jet, whereas the k − ε model predicts
the core length of about 14.6.
1. The correctly expanded (on-design) supersonic jets have 3. The flow parameters in the supersonic core show small
considerably long jet core which increases with jet Mach amplitude oscillations.
number. At an initial jet Mach number of 1.4, the nondi- 4. For both underexpanded and overexpanded jets, the
mensional core length is about 14.6. flow parameters show large amplitude fluctuations for
2. The k − ε model with standard constants predicts the core quite a long distance. The amplitude depends on the exit
length and subsequent velocity decay fairly. However, pressure ratio, which increases as the ratio moves away
the Spalart–Allmaras one-equation model predicts far too from 1.0.
Fig. 13.10 Mach 2.0 twin jet (correctly expanded) at P = 2.5R
Fig. 13.11 Mach 2.0 twin jet (correctly expanded) at P = 3R

5. The off-design jet flows are characterized by a series of the incoming turbulence, closely spaced screen and wire
shocks and expansion fans. The Mach-disc appears if the gauges are installed inside the settling chamber. The settling
exit pressure ratio is considerably different from 1.0. chamber, generally, has a constant area circular cross-
6. The jet cores are clearly visible in optimally expanded section, mounted with pressure and temperature ports for
identical twin jets. However, the core is found to be shorter measuring the stagnation pressure and total temperature,
in twin jets than for isolated jet. The flow parameters show respectively. The experimental models are mounted at the
large fluctuations in the supersonic core as well as in the end of the settling chamber using a slot-holder mechanism.
far field. The stagnation pressure in the settling chamber (P01 ), is
7. The jet interaction decreases with increasing center-to- the controlling parameter in any jet study. It is maintained
center distance between the jets. at constant level during the experiments by controlling
the PRV.
13.4 Experimental Techniques for Studying

the Jets 13.4.1 Pressure Measurements
In this section, an overview of the experimental methods Generally, a pitot probe is used for measuring the total or stag-
used for studying the jets is reviewed. In jet studies, the nation pressure in jets. It is a simple and commonly used pres-
major parameters considered are the pressure or the velocity sure measuring device in experimental aerodynamics. The
decay along the jet centerline, and the pressure or velocity accuracy of pressure measurements depends upon the shape
profiles normal to jet axis. These plots are generated by of the Pitot probe, flow Reynolds number and the Mach num-
performing pressure survey along and normal to the jet axis. ber, orientation of the pitot probe with respect to the mean flow
The ambient pressure is considered as the static pressure at direction, and the turbulence intensity (Chue 1975). Before
which jet discharges from its source. Since the subsonic jets performing any measurements, it is essential to ensure that
are optimally expanded, the measured total pressure can be the pressure readings are not affected by geometrical block-
converted into Mach number or velocity using ambient static age of the pitot probe defined as the ratio of probe exit area to
pressure. However, it is not always the case with sonic and the nozzle exit area. In jet literature, it is established that the
supersonic jets. This is due to the presence of complex wave effect of blockage is negligibly small if this ratio comes out to
structure inside the jet core. Also, the pressure measurement be more than 64. Further, whenever pitot probe faces the jet
is usually done with pitot probe which measures the pressure boundary layer forms on the inner wall of the probe leading
behind the bow-shock (or curved-shock). Thus, in order to error in pressure readings. To ignore the viscous effect on
to evaluate the total pressure upstream of the shock, either pitot probe pressure measurements, the Reynolds number of
the Mach number or the static pressure must be known jet must be greater than the worrisome value of 500.
separately. But the precise calculation of both Mach number Another point to be noted is that in supersonic regime the
and static pressure is difficult due to the wave strength pressure measured by the pitot probe is not the actual total
variation at any jet cross- section. Therefore, the conversion pressure of the mean flow rather it is the total pressure down-
of total or stagnation pressure into Mach number or velocity stream of the bow-shock positioned at the probe nose. To
is usually not attempted and they are used as it is in order to obtain the actual total pressure the pressure loss across the
study the sonic and supersonic jet characteristics. shock must be accounted. Moreover, the flow Mach number
The schematic diagram of a typical experimental jet-test in the supersonic jet core varies from point to point due to
facility is shown in Fig. 13.12. The compressed air stored the presence of a series of shock waves and expansion fans.
in large storage tanks is fetched to run the experimen- Also, the strength of shock waves varies from one shock cell
tal facility. To remove the solid impurities such as rust to another. Because of these reasons, the measured total pres-
particles and oil droplets, the high-pressure air stored in sure is usually not corrected for the losses due to shocks. In
large reservoirs is allowed to pass over a prefilter unit addition, there may be some errors get associated to the mea-
which consists of porous stone candles. For finer filtration, sured pressure data due to probe interference with the shock
sometimes an activated carbon is also used. Subsequently, cell structures and this is why, the pressure data in supersonic
the compressed air is dried in twin-tower semiautomatic flows should be considered good enough for qualitative anal-
silica gel dryer. The air thus dried enters the bell-shaped ysis only. In the pressure measurements, the pitot probe is
settling chamber through a mixing length pipe. Also, in mounted on a rugged traverse which has multiple degrees of
order to regulate the mass flow and the pressure, a Gate freedom (i.e., translational and rotational degrees of freedom).
Valve and a Pressure Regulating Valve (PRV), respectively, The rotational degrees of freedom enables the probe-pitching
are installed before the mixing length pipe. For reducing and probe-yawing mechanisms.
13.4 Experimental Techniques for Studying the Jets 341
Probe holder
6D Transverse
Temperature port
Pressure regulating
valve Gate valve
y
Model holder
Settling chamber Dry air from

x storage tank
d = 0.6 mm
Y
Temperature port
Fig. 13.12 Schematic layout of a typical jet-test facility
13.4.1.1 Pressure Decay along the Jet Centerline The centerline pressure decay for Mach 1.7 circular jet at
It is well known that the jet centerline pitot pressure decay overexpanded, correctly expanded, and underexpanded states
is a measure of jet mixing with the entrained fluid mass. The are compared in Fig. 13.13. These results cover a range of
centerline pressure decay can clearly show the extent of the expansion levels beginning from 23.8% overexpansion level
jet core, which is defined as the axial distance up to which to an underexpansion level of around 14.3%, for Mach 1.7 jet.
the nozzle exit velocity is unaffected for subsonic jets and the When the nozzle exit pressure is lower than the ambient
axial extent up to which supersonic flow prevails for super- pressure to which the jet is discharged, it is termed as overex-
sonic jets. Thus, the axial distance from the nozzle exit to panded state. The results of centerline decay at overexpansion
the location where the characteristic decay begins represents corresponding to NPR 4 given in Fig. 13.13 show that the jet is
the core length of the jet. Here, the pressure measurements wave dominated possessing three prominent shock cells with
have been carried out with pitot probe and thus the measured compression and expansion waves of considerable strength. It
total pressure distribution is also termed as pitot pressure dis- is clearly seen that the pitot pressure jumps from 0.6 to almost
tribution. The pitot pressure (P02 ) distribution along the jet up to 0.9 at XD = 0, revealing the presence of oblique shock
centerline direction is made nondimensional by dividing them waves right at the nozzle exit. After crossing the compression
with settling chamber pressure (P01 ), which runs the jet. The front the pitot pressure decreases up to some axial locations
nondimensional pitot pressure distribution is plotted against from the nozzle exit reaching to its first minimum implying
the nondimensional axial locations X D along the jet center- the acceleration of supersonic flow. This is because the pitot
line, where D is the nozzle exit diameter. The experiments are probe essentially measures the total pressure downstream
conducted at different Nozzle Pressure Ratios (NPRs) defined of the detached shock wave, which is bound to decrease
as the ratio of settling chamber pressure (P01 ) to ambient pres- with increase of supersonic Mach number. The minimum
sure (Pb ) at which the jet is being discharged. That is pressure point corresponds to the first cross-over point,
where opposite family of oblique shock wave meets. Even
P01 though these individual oblique shocks may be of weak in
NPR =
Pb nature, their combined strength is strong enough causing the
flow to become subsonic. Thus, it is indeed the point of local
In all the experiments, NPR is the governing parameter which maximum of Mach number. The subsonic flow downstream
dictates the expansion level prevailing at the nozzle exit. of the first shock cross-over point gains the momentum
Based on the relative magnitudes of ambient pressure (Pb ) from the higher momentum flow surrounding it. The pitot
to which the jet is discharged and nozzle exit pressure (Pe ), pressure increases up to some axial locations reaching to its
the following jet expansion levels may arise. first peak. This zone is referred to as subsonic acceleration
again because of the fact that the increase of pitot pressure
Pe < Pb ⇒ overexpansion at subsonic speeds indicates increase of flow velocity (i.e.,
Pe = Pb ⇒ correct−expansion Mach number). The first pressure peak in the plot implies
Pe > Pb ⇒ underexpansion the attainment of sonic state, which again accelerates to
Fig. 13.13 Centerline pressure decay for Mach 1.7 jet
supersonic speeds reaching up to second minima or second downstream of the location at which the jet had become fully
maximum Mach number point in the plot. Behind this point developed for NPR 4.
the flow becomes subsonic again and begins to accelerate If the nozzle exit pressure is higher than the pressure in
to sonic level and further accelerates to supersonic levels. the ambient medium (i.e., back pressure), the jet is said to be
This cycle repeats leading to a periodic shock cell structure underexpanded. At an underexpansion level of about 14.3%,
whose axial extent is known as jet core length. At NPR 4, the which corresponds to NPR 6, the centerline decay for Mach
supersonic core extends to about 8D and after this, the flow 1.7 circular jet is shown in Fig. 13.13. To bring down the
begins to show a monotonic decrease of pitot pressure. The higher nozzle exit pressure to the ambient pressure level,
characteristic decay zone prevails from 8D to about 15D. expansion waves are formed at the nozzle exit. These waves
Beyond 15D, the jet is found to be fully developed. cross one another and are reflected from the boundaries of the
At NPR 5.25, which is the correctly expanded state for jet flow field as shock waves. These compression waves again
Mach 1.7 jet, the centerline decay is represented in Fig. 13.13. cross each other and are reflected on the boundaries of the jet
Despite the fact that for optimum expansion level, the nozzle as expansion waves. The cycle repeats leading to a shock cell
exit pressure is equal to the ambient pressure (no pressure structure which extends up to about 15D. For this NPR, the
imbalance), yet the pressure oscillations can be easily seen flow becomes fully developed only beyond 25D.
in the core. This is because, even though the jet is issuing at
correct-expansion, soon after exiting the confined area (i.e.,
13.4.1.2 Pressure Profiles
nozzle), the flow encounters a large space to relax. In jet lit-
To get an insight into the flow development and symmetry of
erature, it is popularly known as relaxation effect. Because of
the jet, the pitot pressure distributions measured in the direc-
this effect, the flow turns away from the nozzle axis in a bid
tion normal to the jet centerline are studied by constructing
to occupy the space available downstream of the nozzle exit.
pressure profiles. The pitot pressure variations in radial direc-
Since the supersonic flow is wave dominated and any change
tion for the uncontrolled jet, at different NPRs could be used
of flow properties can take place through these waves only.
to authenticate the observations made from centerline decay
Thus, for correctly expanded jets the relaxation phenomena
results. The pressure profiles would also reveal the symmetry
leads to the formation of expansion waves at the nozzle exit.
of the jet as well as the tendency for axis-switching. It is well
These waves will travel some distance downstream and be
known that, in a free jet propagation, the earlier axis-switching
reflected from the jet boundary as shock waves. The compres-
implies the rapid mixing of jet with the ambient fluid. Thus,
sion waves travel further and get reflected as expansion waves
to understand the jet spread, the pitot pressure distribution
from the jet boundary. The repetition of this cycle leads to a
(P02 ), along the radial direction (i.e., perpendicular to jet cen-
periodic shock cell structure inside the jet core. It is seen that,
terline) at different axial locations are made nondimensional
with the increase in the nozzle pressure ratio, the core for the
by dividing them with settling chamber pressure (P01 ). The
uncontrolled jet extends to about X D = 12.5, which is longer radial distance (R) is made nondimensional by dividing them
than that for NPR 4. The jet tends to become fully developed
by the nozzle exit diameter (D).
only beyond X D = 22.5 for this NPR, which is considerably
13.4 Experimental Techniques for Studying the Jets 343
Fig. 13.14 Pressure profiles for Mach 1.7 jet
The pitot pressure profiles for uncontrolled jet at NPR 4 downstream location given by X D = 4, the maximum pressure
at axial distances of XD = 0.5, 1, 2, 4, and 6 are presented in reduces to 0.8 but the pressure decrease becomes gradual as
Fig. 13.14a. It is seen that the jet exhibits symmetrical flow compared to the locations upstream of this point. At the far-
field about the jet axis at X D = 0.5 and the pitot pressure is thest downstream location X D =6 , interestingly, no constant
found to be minimum or in other words, the jet velocity is pressure level is observed, rather, a single peak pressure pro-
maximum at this location. Slightly away from the jet cen- file is obtained.
terline, the pressure level is found to be constant till approxi- The pressure profiles for the correctly expanded jets from
mately D R
= 0.5, but decreases suddenly from 0.5 to 0.6. After a nozzle are presented in Fig. 13.14. At NPR 5.25, which is a
the location D R
= 0.6, the pressure becomes nearly constant correctly expanded state for Mach 1.7 jet, the pressure profile
with a magnitude of 0.3P01 . At further downstream location at XD = 0.5, is found to be of uniform magnitude over some
D = 1, the pressure level depicts nearly a constant pressure
X radial distance from the jet centerline. At this NPR, the peak
region which implies that there exists a uniform Mach number pitot pressure is found to be about 0.86, and from normal
region in the vicinity of the jet centerline. This zone extends shock relation for Mach 1.7 jet the stagnation ratio PP02
01
should
R
up to D = 0.3. For the radial location greater than 0.3, a sharp be equal to 0.8557. This clearly reveals that the jet is being
decrease in pressure till D R
= 0.6, can be seen. Beyond this issued from the nozzle exit at correct-expansion.
location, the pressure maintains a constant level. Further, the At the highest tested NPR of the present study correspond-
pressure level around the jet centerline at X D = 2 is found
ing to 6, there exists a favorable pressure gradient at the exit
to be higher than D = 1. However, the radial extent of this
X of Mach 1.7 nozzle. Consequently, the expansion waves ema-
peak pressure is now limited to D R
= 0.25, only. At further nating from the nozzle exit are strong and hence, the flow
Mach number in the near field zone could be more than 1.7. Mach-disc) as seen in Fig. 13.16. It is well established that a
Also, at the location XD = 0.5, the pressure level is found to Mach-disc forms when the nozzle exit pressure is more than
be 0.8 which is indeed lower than for NPR 5.25. But, at other or equal to four times the back pressure. That is
downstream locations X D = 1, 2, 4, and 6, the pressure profile
is almost similar to that for correct-expansion case. However, Pe ≥ 4Pb
at this NPR, the spread of uniform Mach number zone is more
than for NPR 5.25.
13.5 Aerodynamic Mixing Enhancement
and Jet Controls
13.4.2 Optical Flow Visualization
In various technological applications, the performance can be
In addition to pressure measurements, the optical flow visu- improved by augmenting the mixing of jet with the ambient
alization techniques such as Schlieren and Shadowgraph are fluid to which it is discharged. One such application is the
used to visualize the wave strength in the jet core. In this increase of combustion efficiency, which can be improved by
section, the flow visualization images using Shadowgraph increasing the rate of mixing of fuel with air. For supersonic
method are discussed. The schematic diagram of a typical combustion ramjet (SCRAM) engines, even this augmented
Shadowgraph setup is shown in Fig. 13.15. mixing should be completed within the short distance to
For Mach 1.7 jet at an overexpanded state corresponding to minimize the size of the combustion chamber. Further, for
NPR 4, the oblique shock waves are clearly seen (Fig. 13.16a). an efficient combustion, both large and small-scale mixing
These shock waves cross each other at the jet axis and meet the is desired as the large scale eddies determine the rapidity of
barrel shock near to jet boundary. Since the reflection from the mixing and small-scale eddies ensure the effective molecular
fluid boundary is unlike and thus the shock waves get reflected level mixing. Also, with increased mixing of a nozzle exhaust
as expansion fans on reaching the barrel shock. Further, we the infrared signature of hot jet can be reduced, thereby
know that the point at which a shock is reflected as expansion increasing the stealth capability. The other technological
wave is termed as kink produced at the shock reflection point applications, where the mixing enhancement is required
is observed in Fig. 13.16a. The expansion fans, after cross- include thrust vector controls, gas dynamic lasers, thrust
ing each other, extend to the jet boundary and get reflected augmenting ejectors, etc.
in the form of compression waves. These reflected waves The jet mixing enhancement is achieved by controlling
once again cross each other at the jet centerline and reflect the jet which are active or passive in nature. Here, the control
back as an expansion wave from the barrel shock boundary. of jet is defined as the technique to modify the mixing
This wave reflection phenomena continues for some distance characteristics of the jet in order to achieve engineering
downstream. The distance between any two successive shock efficiency, technological ease, economy, and so on. The
reflection points is taken as the shock cell. At NPR 4, three controlled jets find a variety of applications from domestic
shock cells can be clearly seen. The first two cells are strong purposes to rocket science.
but later two are weak in nature. It is well known that the shear layer forms as soon as the jet
At NPR 5.25, which is correctly expanded state for a comes out of the nozzle exit. After the potential core region,
Mach 1.7 jet, the waves prevailing in an uncontrolled jet depending upon the initial velocity profile, the jet experiences
field are shown in Fig. 13.16b. It is interesting to note that, flow oscillations at the boundary. These flow perturbations
even at correct-expansion the core is wave dominated. This roll up forming eddies which grow in size while propagat-
is because, as explained in Sect. 13.4.1.1, the flow encoun- ing downstream. The vortices, thus, generated create suction
ters expansion waves because of relaxation effect soon after at the jet periphery, in turn, the ambient fluid gets entrained
exiting the nozzle. These expansion fans get reflected as com- into the jet and mixes with the jet fluid. This vortex interac-
pression waves and the process continues. For this case also, tion eventually leads to the transition of the flow from lam-
two prominent shock cells are observed. inar to turbulent. Formation of these coherent structures in
At NPR 6 (underexpanded state for Mach 1.7 jet), the shear layer is initiated by Kelvin–Helmholtz instability. The
expansion waves at the nozzle exit are stronger. Because of coherent structure is essentially a connected, large-scale fluid
the stronger expansion fans, the jet centerline Mach number lump with a phase correlated (coherent) vorticity over the
attains a much higher value leading to the longer shock cells. whole structure (Hussain 1983) and their shedding in jets is a
At NPR 6, as many as five shock cells are clearly seen. This strong function of initial conditions and thus can be controlled.
is a moderately underexpanded operation as there exists a Since the frequency associated with these vortical structures
cross-over point at the end of first shock cell. However, in is obtained easily and thus they are generally controlled by
highly underexpanded jets, the point of intersection becomes using acoustic excitation. The initial vortex shedding fre-
an intersecting zone, resembling a disc (popularly known as quency (f) also known as the most amplified frequency of
13.5 Aerodynamic Mixing Enhancement and Jet Controls 345
Light source
Condenser lens
Screen Test section Concave mirror
Fig. 13.15 Schematic layout of the Shadowgraph setup
Fig. 13.16 Shadowgraph pictures of the axisymmetric submerged jets
these structures depend upon the various parameters such as large eddies, even though, are efficient in ambient fluid engulf-
coherent structure shape, initial shear layer momentum thick- ment but they are poor mixing promoter. Also, these eddies
ness (θ0 ), and the jet exit velocity (U0 ). These factors are put owing to their large size have shorter lifespan and subse-
together ina nondimensional form called the Strouhal number
quently break down into small size vortices due to their large
Sti = fθ 0
U0 . The range of Strouhal number for initially ampli-
inertia. These smaller eddies which have higher life span are
fied normalized frequencies, predicted by the linear stability found to be better mixing promoter. For an efficient mixing,
theory, is 0 < Sti < 0.04. both large size and small size eddies should be present in
The coherent structures initially produced in the shear proper proportion. However, the mechanism behind the mix-
layer grow in size and coalesce to each other while they ing of supersonic streams is quite different from their subsonic
are convected downstream. Because of the mass engulfment counterparts. The mixing phenomena in subsonic shear layers
and merging of the vorticity, shear layer spreads and the fre- is primarily depends on the ratios of velocity and density in
quency associated with the large-scale eddies reduces. The mixing layers, whereas in supersonic shear layers apart from
the velocity and density ratios across the shear layers, the com- metric counterparts, experience an early “axis-switch” phe-
pressibility also plays a dominant role. The compressibility nomenon, and thus they are well-suited in augmenting the
of shear layer is defined in terms of a nondimensional param- supersonic jet mixing.
eter called the convective Mach number (Mc ). It is defined From the above discussion, it is clear that the jet flow
as the ratio of the relative speed of large-scale eddies with (more precisely shear flow) control methods aim to augment
respect to the freestream (U − Uc ) and the speed of sound the three-dimensional flow characteristics, and hence mass
(a). Thus, for two shear layers designated as “1” and“2”, the engulfment and mixing by modifying the natural develop-
convective Mach numbers will be defined as Mc1 = (U1a−U 1
c)
ment of large-scale coherent structures and their breakdown
and Mc2 = (U2a−U 2
c)
; U1 and U2 are the freestream veloci- into mixing promoting small-scale structures.
ties, a1 and a2 are the speeds of sound in respective streams,
and Uc is the convective speed of the eddies. In case, if
13.5.1 Classification of Jet Controls
both streams have equal static pressures and specific heat
ratios then these convective Mach numbers will be equal, i.e.,
1 −U2 )
All the flow control methods are broadly classified into active
Mc1 = Mc2 = (U (a1 +a2 ) . However, it is also observed that at and passive controls. In an active control technique, the addi-
low convective Mach numbers, the flow characteristics of a
tional source of energy is required to activate the control. The
supersonic shear layer is almost similar to a two-dimensional
pulsed jets, piezoelectric actuators, micro-jets, and oscillat-
incompressible shear layer. But, with increase of convective
ing jets are among the most effective control technique for
Mach number, the mixing layer essentially becomes more and
mixing enhancement. In contrast, the passive controls do not
more three-dimensional in nature, especially when Mc > 0.6.
require any additional energy source for its action, rather they
Another important phenomena associated to the compressible
make use of geometrical modification of the nozzle exit. The
shear layers
is the generation
of vorticity due to the baroclinic
grooves and cut-outs at the nozzle exit, placement of tabs at
torque − ρ2 (∇ρ×∇ρ) , produced when the density gradients
1
the nozzle exit, use of multi-jets, lobbed nozzles, etc., are
present in compressible shear layer is coupled with large pres- commonly used passive controls.
sure gradients across the shock waves.
The mixing process in jet shear layers is initiated by large 13.5.1.1 Active Controls
size coherent structures, which are formed when the lower Active flow control in the form of periodic excitation con-
momentum fluid outside fluid is ingested into the relatively sists of additional system and requires power input, but it is
higher momentum jet fluid. The smaller eddies embedded used in various applications due to its large potential payoff.
in the large structures have no role in mass entrainment but Many active jet control methods use energized actuators to
are responsible to its mixing with the jet fluid (Brown and dynamically manipulate flow phenomena based on open or
Roshko 1974). In various engineering applications, the con- closed-loop algorithms. The design of an active flow control
trol of jet essentially lies on understanding the complex topol- system requires knowledge of flow phenomenon and selec-
ogy of these coherent structures only. That is, investigating tion of appropriate actuators, sensors, and a control algorithm.
the techniques that affect the formation, mutual interaction, The role of an actuator is to inject perturbations at a pre-
amalgamation, and breakdown of these structures. In axisym- scribed frequency into the flow at locations where the flow
metric jets, the development of round and azimuthal coher- is most receptive to these inputs. The actuator leverages or
ent structures and their subsequent amalgamation governs disrupts the flow to bring about a desired effect. For example,
the growth of the jet shear and fluid mass ingestion (Crow the conventional excitation methods have relied on exciting
and Champagne 1971). Soon after exiting the nozzle, in the instability modes with their most amplified frequency band
near field itself, the jet essentially becomes three-dimensional to bring about jet mixing enhancement. For jet excitation,
and streamwise vortices predominately engulfs the ambient the conventional philosophy has been to energize the large-
fluid (Liepmann and Gharib 1992). Also, the phenomena like scale coherent structures or bring about vortex interactions
stretching of vortices, self-induction, and merging becomes that result in the engulfment of surrounding fluid, resulting in
active as soon as jet comes out of the nozzle (Hussain 1986). mixing enhancement.
When the azimuthal nonuniformities are introduced at the
nozzle exit, they further add the complexity to the progression
13.5.1.2 Passive Controls
of jet shear layer as well as the three-dimensional develop-
The passive control technique uses the geometrical modifi-
ment of the jet. For example, the modified nozzle geometry
cations of the nozzle from which the jet flow is issuing to
and perturbations introduced at the nozzle exit influence the
change the shear layer stability characteristics. Some exam-
jet shear layer progression, leading to a three-dimensional
ples of these modifications are passive controls in nonreacting
jet structure. These asymmetric jets inherently dominated by
and reacting flows, elliptic jets, trip wires in plane shear lay-
high azimuthally unstable modes as compared to their axisym-
ers, convoluted splitter plates, and noncircular jets such as
13.5 Aerodynamic Mixing Enhancement and Jet Controls 347
square jets, elliptic jets, etc. Note that, though the passive where S denotes the surface area. The noise level is then deter-
control methods are efficient in promoting mixing but due to mined either by the acoustic intensity level (IL) or by the
large amount of entropy production the uniformity of the low sound pressure level (SPL), measured in decibel (dB).
at the nozzle exit gets spoiled and thus leads to the thrust loss.
I
IL = 10 log10 (13.30)
Iref
13.6 Acoustics
where Iref =10−12 Wm−2 at 1000 Hz and I is the scalar
Acoustics is that branch of science that studies the propaga- acoustic intensity normal to the surface as obtained from
tion of sound waves as in the combustor, jet acoustics, meteo- Eq. (13.28).
rological acoustics, and environmental acoustics. The sound
waves may be generated in stagnant air even with very small
Prms
pressure-induced excitations. In turn, a noise could be pro- SPL = 20 log10 (13.31)
Pref
duced which is in audible range to humans. In such cases,
except the pressure, all other flow parameters remain constant. where Pref = 2.14 × 10−5 Nm−2 , corresponds to Iref in a
Contrarily, in some cases, the noise level may be extremely plane wave and Prms is the root mean square pressure.
high as in thunderstorms or in sudden explosions. However,
still no appreciable changes in flow variables other than the
pressure are seen in these cases as well. This phenomenon
13.6.1 Pressure Mode Acoustics
is termed as the pressure mode acoustics. Moreover, due
to circulatory motion significantly large velocity gradients
The wave equation is the most commonly used relation in
are established within the fluid, vortical waves are produced
acoustics, which can be derived from continuity and momen-
which in turn cause pressure disturbances. The noise produced
tum equations. The pressure can be represented as the sum of
by these vortical structures are classified as the vorticity mode
freestream pressure and the fluctuation component.
acoustics. Furthermore, in several occasions such as in hyper-
sonic flows past a spacecraft, a rapid increase in temperature
is encountered creating the entropy layer between the shock p = p0 + p
layer and the momentum boundary layer. This entropy layer
leads to pressure fluctuations and since the entropy waves are Assuming no source and sink of mass and in absence of body
predominant in this case, the phenomenon is categorized as forces, for a zero convection velocity the convected wave
the entropy mode acoustics. Nevertheless, it is certainly pos- equation is given by Eq. (13.32).
sible that the noise level may actually be the combination of
all three modes in a given physical situation. 1 ∂ 2 p
− ∇ 2 p = 0 (13.32)
Some basic definitions used in acoustics are summarized aa2 ∂t 2
in this section. The time-averaged value of a fluid property
(f) is defined as where aa is the speed of sound at freestream conditions. For
a nonzero convection velocity, the convected wave equation
ˆ
t+t is represented by Eq. (13.33).
1
f̄ =< f >= fdt (13.27)
t 2
t 1 ∂
+ v.∇ p − ∇ 2 p = 0 (13.33)
aa2 ∂t
From this result, the acoustic intensity (I) measured in Wm−2
is defined as Now, Eq. (13.32) is multiplied with eiωt both sides and sub-
sequently integrated over an appropriate time interval. We
ˆ
t+t
1 obtain the famous Helmholtz equation.
I =< pv >= pvdt (13.28)
t
t 2
ω
∇2 + p eiωt = 0 (13.34)
where f = pv. Thus, the acoustic power (π) will be aa
ˆ
π= I.nd = ρ0 a0 < u2 > S (13.29) where ω is the circular frequency.
S
13.6.1.1 Equation of Pressure Wave: Stationary between the normal vector and the radial direction, and n̂ is
Surfaces the outward normal vector.
Ribner (1981) proposed that the Kirchhoff’s law can also be
used to solve the problems of wave propagation in acoustics. 13.6.1.2 Equation of Pressure Wave: Surfaces
The idea of Kirchhoff’s formula is to surround the region of a Moving at Subsonic Speeds
nonlinear flow field and acoustic sources by a closed surface. The noise generated by high-speed propellers and helicopter
To determine p (x, t), consider the homogeneous Helmholtz rotors was predicted by Hawkings (1977). To account for
equation given by Eq. (13.34), whose solution is the Green’s motion, he modified the Eq. (13.37) using Prandtl–Glauret
function G (x, y; ω). It can be shown that transformations (x1 = x, y1 = βy and z1 = βz). Here, the
˛ subscript ‘1’ denotes the transformed coordinate. We have
∂p
p(x, ω) = G (x, y;ω) (y,ω) ˆ
p ∂r1 1 ∂p 1 ∂p

∂r1 ∂x1

∂yj 4πp(x, t) = − + − Ma dS1
S
r 2 ∂n1 r1 ∂n1 aa r1 β 2 ∂τ ∂n1 ∂n1
∂G (x, y;ω) S1
− p(y,ω) n̂j dS(y) (13.35)
∂yj
where Ma is the freestream Mach number, x , y , z
2
where n̂ is the unit normal on S directed into the fluid. Now represents the position of the source, r1 = x − x
using convolution theorem,3 the time domain solution can be 1 1
2 2 2
obtained from Eq. (13.35) (frequency domain solution) as +β y − y + z − z , β = 1 − Ma2 2 and
[r1 −Ma (x−x )]
˛
∂p
τ= aa β 2
is the retarded time.
p (x, t) = −G (x, y; t − τ ) (y, τ )
∂yj
S
13.6.1.3 Equation of Pressure Wave: Surfaces
∂G (x, y; t − τ )
−p (y, τ ) n̂j dS(y)dτ (13.36)
∂y j Moving at Supersonic Speeds
Unlike the stationary surfaces and the surfaces moving at sub-
Now from momentum equation after linearization and without sonic speeds, the retarded time for the surfaces moving at
∂v ∂p
body forces, ρa ∂τj = − ∂y j
. Thus, supersonic speeds, is not uniquely defined. Rather, it is given
˛ by
∂vj
p (x, t) = −G (x, y; t − τ ) ρa (y, τ )
∂τ
S
±r1 − Ma x − x
−p (y, τ )
∂G (x, y; t − τ )
n̂ j dS(y)dτ (13.37) τ± =
∂yj aa β12
Using the free space Green function, G (x, y; t − τ ) 1

where β1 = Ma2 − 1 2 . In this case, the pressure wave is
= 4π|x−y|
1
δ t − τ − |x−y|
aa , we obtain defined as
ˆ
˛ vn y, t − |x−y| 4πp(x, t) =
p ∂r1
−
1 ∂p
+
1 ∂p
±
∂r1
− Ma
∂x1
ρa ∂ a a r 2 ∂n1 r1 ∂n1 aa r1 β12 ∂τ ∂n1 ∂n1
dS1
p (x, t) = dS(y) ±
4π ∂t |x−y|
S1
S
(13.39)

j p y, t − |x−y|
1 ∂ aa where ± notations indicate to evaluate the above integral at
− nj dS(y) both τ + and τ − , retarded times.
4π ∂xj |x−y|
S
Which can further be reduced to following form: 13.6.2 Vorticity Mode Acoustics
ˆ
p ∂r 1 ∂p 1 ∂r ∂p
4πp (x, t) = − + (13.38) The noise generated by large and small-scale eddies is referred
r 2 ∂n r ∂n aa r ∂n ∂τ to as aerodynamic sound. The theory of aerodynamic sound
S
was developed by Lighthill (1952). He derived the exact and
where |x − y| = r, the distance between the observer and the nonhomogeneous wave equation, whose source terms are per-
∂r
source, τ is the retarded time, ∂n = cos θ where θ is the angle tinent only in turbulent region.
3 Convolution Theorem: According to this theorem, the Fourier transform
of a convolution F { f g} is equal to the pointwise product of Fourier

transforms of individual functions, i.e., F { f } .F {g}.
13.6 Acoustics 349

1 ∂2
13.6.2.1 Lighthill’s Acoustic Analogy − ∇ 2 H ( f ) aa2 (ρ − ρa )
For an ideal and linear fluid at rest, the momentum equation aa ∂t
2 2
is given by ∂ 2 H ( f ) Ti j
=
∂xi ∂x j

∂ (ρvi ) ∂ aa2 (ρ − ρa ) ∂Tij ∂ ∂H
+ =− (13.40) − ρvi v j − v̄ j + ( p − pa ) δi j − τi j (f)
∂t ∂xi ∂xj ∂xi ∂x j

∂ ∂H
where ρa is density, aa is speed of sound and Tij is the exter- +
∂t
ρ (vi − v̄i ) + ρa v̄i
∂x j
(f) (13.45)
nally applied stress. To eliminate the term ρvi in Eq. (13.40),
the continuity equation is used which results the famous In integral form,
Lighthill acoustic analogy equation as
˚
∂2 d3 y
1 ∂2 ∂ 2 Tij H ( f ) aa2 (ρ − ρa ) = Ti j
∂xi ∂x j 4π | x − y |
− ∇ 2
aa2 (ρ − ρa ) = (13.41) ¨ d S (y)
aa ∂t
2 2 ∂xi ∂xj ∂
ρvi v j − v̄ j + pi j
j
−
∂xi 4π | x − y |
¨
Since the assumed fluid was ideal and linear in which the ∂ d S j (y)
+ ρ (vi − v̄i ) + ρa v̄i (13.46)
momentum transfer was solely produced by the pressure. Tij ∂t 4π |x − y|
also referred to as Lighthill stress tensor is given by
where pij = (p − pa ) δij − τij and retarded time
|x−y|
Tij = ρvi vj + (p − pa ) − aa2 (ρ − ρa ) δij − τij (13.42) τ =t− aa .
Equation (13.46) comprises of following
three source terms:
¨
|v | ∂ d S j (y)
For low fluctuating Mach numbers aa 1 and high
[ρ (vi − v̄i ) + ρa v̄i ] =⇒ monopolesource
∂t 4π |x − y|
Reynolds numbers |vaν|L 1 , using order of magnitude
analysis it can be shown that Tij ≈ ρ0 vi vj , when the viscous
stresses are neglected. If the perturbation pressure in far field ¨ d S (y)
∂
j
is given p (x, t) = aa2 (ρ − ρa ), the solution to Lighthill equa- − =⇒ dipolesource
∂xi 4π |x − y|
tion can be written as

ˆ ρa vi vj y, t − |x−y| ˚
∂2 d3 y
∂ 2 aa Ti j
p (x, t) ≈ d3 y (13.43) ∂xi ∂x j 4π | x − y |
∂xi ∂xj 4π |x − y| ¨ d S (y)
∂
j
− =⇒ quadripolesource
∂xi 4π |x − y|
13.6.2.2 Ffowcs Williams–Hawkings Equation
These equations are based on the extension to Lighthill’s
analogy. Let f (x, t) be an indicator function that vanishes
on the surface S and satisfies f (x, t) > 0 in the fluid, where
Lighthill’s equation is to be solved and f (x, t) < 0 elsewhere. 13.6.3 Entropy Mode Acoustics
Let us multiply equation (13.40) and continuity equation with
H (f) and rearrange into the forms In a flow domain, if the temperature gradients are high, it may
have very large fluctuation components, which lead to entropy
∂ ∂ waves. Chung (1996) combined the first and second law of
(ρvi H (f))+ H (f) aa2 (ρ − ρa )
∂t ∂xi
thermodynamics and using Maxwell’s relations derived the
∂
=− H ( f ) Ti j + ρvi v j − v̄ j acoustics analogy equation given by
∂x j
+ ( p − pa ) δi j − τi j
∂H
(f)
∂
(H ( f ) (ρ − ρa ))

∂x j ∂t ∂ 1 ∂p 1 ∂ αT DS
− ∇. ∇ p = vi v j + T
+
∂
(H ( f ) ρvi ) = (ρ (vi − v̄i ) + ρa v̄i )
∂H
(f) (13.44) ∂t ρaa2 ∂t ρ ij ∂t c p Dt
∂xi ∂x j (13.47)
Eliminating Hρvi between the above two equations, we get
∂ρ
the well-known Ffowcs Williams–Hawkings equation in dif- where S is specific entropy and αT = − ρ1 ∂T is the coeffi-
ferential form cient of thermal expansion.
13.7 Jet Noise sible for entrainment, mixing and are also largely responsi-
ble for the production of Reynolds stresses. The presence of
The study of flow generated acoustic waves probably began coherent structure has rekindled interest in jet noise phenom-
with Gutin’s theory of propeller noise, which was developed ena. This is because of the prospect for a more deterministic
in 1937. Yet, it was not until 1952, when Lighthill introduced representation for the Lighthill’s quadruple source term in the
his acoustic analogy to deal with the problem of jet noise, aerodynamic noise theory. A detailed comparison between the
that a general theory began to emerge. Lighthill’s ideas were experiments on subsonic jet and Lighthill’s theory of jet noise
subsequently extended by Curle (1955), Powell (1960), and has been made by Lush (1971) and Ahuja (1973). Sarohia and
Ffowcs Williams and Hall (1970) to include the effects of solid Massier (1978) carried out an experimental study to investi-
boundaries. These extensions include Gutin’s analysis and, in gate the effects of large-scale structures in subsonic jet flows
fact, provide a complete theory of aerodynamically generated on jet mixing and their relation to jet noise production. They
sound that can be used to predict jet noise as well as blading found that a large organized turbulent structure exists to a far
noise. Availability of vast quantum of knowledge on jet noise downstream of the nozzle exit as 7D. High- speed Schlieren
reveals the amount of importance associated with jet noise motion pictures synchronized with near field measurements
control. In the past five decades, there have been many reviews indicate that a significant part of the near field pressure sig-
on aeroacoustic. Lighthill (1963), Ffowcs Williams (1969), nal was contributed by the interaction or merging of the large
Ffowcs Williams (1977), and Crighton (1975) reviewed the organized structures in the jet flow. The proposition of Zaman
acoustic analogy theories on subsonic jet noise. Review paper and Hussain (1981) was that it is the breakdown process of the
on “ perspective on jet noise” by Ribner (1981) brought out initial toroidal structures into substructures near the end of the
the different theories to model jet noise production and a large potential core and their interactions that produce most noise.
number of schemes for suppressing jet noise. Michalke (1984) In that sense, the noise production mechanism involves coher-
carried out a comprehensive survey on jet instability theory ent substructures, rather than coherent structures. The state of
for axisymmetric jets. Goldstein (1984) reviewed the different the flow as it leaves the orifice or nozzle is also important in
mechanism for generation of sound by high Reynolds number determining the jet noise because of its influence on the tur-
turbulent shear flows. Nevertheless, all the above investiga- bulent mixing process. The power spectrum of a subsonic jet
tions considered only subsonic jets and yielded that the noise is characterized by a broad single maximum with no discrete
generation mechanism in supersonic jets are cumbersome and frequencies showing noise purely due to turbulent mixing.
are altogether different from those of subsonic jets. Notably, The major difference between the model jet and the engine
the noise generation mechanisms of supersonic jets seem, by jet is in their temperatures. A low-density jet (heated jet) has
all accounts, to be greatly improved and comprehended in a much higher spread and hence, expected to have difference
contrast to subsonic jets. Moreover, the prediction of jet noise in the noise produced compared to its cold counterpart.
directivity and spectral characteristics are relatively straight- The main issue in reducing the aerodynamic noise pro-
forward at supersonic Mach numbers. Most of the advances duced by turbulence is connected with the lesser under-
in supersonic jet noise research have been made in the past standing of turbulence itself, which is one of the principal
25 or so years. Part of these works have been reviewed by noise sources of engine and airframe noise. The problem
Seiner (1984), Seiner (1984) Seiner and Krejsa (1989), and becomes more complicated due to the fact that the sound
Tam (1995). “takes place’not on the main energy carrying part of vorti-
cal flow, but on a rather low energetic, though well correlated
part of turbulence, which in Fourier space of wave numbers
13.7.1 Subsonic Jet Noise lies on the so-called radiating sphere. To simulate the radiating
part of turbulence, it is required on the one hand, elaboration
Since the acoustic analogy model proposed by Lighthill in of new turbulence models and, on the other hand, application
1952, progress in the understanding of jet noise phenomena of the most advanced computational methods.
has been both scant and sporadic. Realization of the presence
of organized large-scale structures in turbulent shear flows has
generated a growing interest in advancing the understanding 13.7.2 Supersonic Jet Noise
of their role in the production of jet noise. Bradshaw et al.
(1964) observed organized structures in jet flows and this was It has been observed that even though the noise characteristics
further confirmed by Crow and Champagne (1971), Brown of supersonic jets are quite complex, their noise generation
and Roshko (1974), Winant and Browand (1974). Browand mechanism is better understood than the subsonic jets. Also,
and Weidman (1976) recognized that the randomly merging the noise directivity and spectral characteristics of supersonic
and amalgamation process accompanying the evolution of jets are easier to predict. Further, it is now well accepted that
these orderly structures in turbulent shear flows are respon- all the turbulent jets contain both large-scale and small-scale
13.7 Jet Noise 351
eddies, capable of generating noise. However, the relative The intensity of turbulent mixing noise and as well as its
intensity of the noise produced by these structures strongly spectral and directional characteristics strongly depends upon
depends upon the jet Mach number and the jet temperature. the jet Mach number and the ratio of jet to ambient temper-
For cold subsonic jets, the convective Mach number, with ature. When the Mach number is increased, the difference
respect to the ambient speed of sound, of large-scale eddies is between peak levels of dominant part and background noise
subsonic and thus, they are ineffective noise producers. There- increases. However, at constant Mach and Strouhal numbers
fore, the dominant part of jet noise is created by small-scale when the jet temperature is increased, the maximum level of
eddies only. On the other hand, in supersonic jets especially at noise radiation does not increase appreciably. But the increase
elevated temperatures, the large-scale vortices convect down- in jet temperature broadens the dominant noise sector. Fur-
stream at supersonic Mach numbers. Due to this, they are ther, it has been observed that at low Strouhal numbers more
capable of generating intense noise which predominates over noise is radiated, leading to a widening of peak in the noise
the sound produced by the small-scale structures. Hence, it spectrum. Therefore, due to these reasons, the overall noise
can be fairly said that the large-scale eddies are the domi- is increased with an increase of temperature.
nant source of supersonic jet noise. Moreover, due to imper- Both the large turbulence structures and the fine-scale tur-
fect levels of expansion (overexpansion or underexpansion) bulence of the jet flow contribute to the turbulent mixing noise
of supersonic jet, a quasi- shock cell structure forms within of a supersonic jet. The former generates the dominant part
the core. Because of this, an additional component of noise is of the turbulent mixing noise. The latter is responsible for
generated referred to as shock-associated noise, which is fur- the background noise. Further, to predict the noise genera-
ther subclassified into broadband shock-associated noise and tion from supersonic jets, the large-scale eddies should be
the screech tone. The former, as the name suggests, is a broad- represented through an appropriate turbulence model. Tam
band noise, whereas the latter occurs at discrete frequencies. and Chen (1979) were the first to propose a stochastic model
Thus, the imperfectly expanded supersonic jet comprises of to describe the large-scale structures mathematically. The
three parts namely turbulent mixing noise, broadband shock- stochastic instability wave model, for large-scale eddies, is
associated noise, and screech tone. based on the observation that the spreading rate of high-speed
jets is very small. It essentially means that the flow variables
13.7.2.1 Characteristics of Turbulent Mixing Noise and the turbulence statistics change very slowly in the down-
Seiner (1984) measured the narrow band noise spectrum of stream direction, i.e., the turbulence statistics are nearly con-
an imperfectly expanded supersonic jet. The highest peak was stants locally. If all the turbulence statistics are true constants
the screech tone, which is generally accompanied by its har- (i.e., stationary in time and in the flow direction), then the sys-
monics. Sometimes, even the fourth- or fifth-harmonic could tem is in a state of equilibrium. Indeed, it is not completely true
be detected. The low- frequency zone at the left of screech tone because of the divergence of the jet flow. However, the rate of
is referred to as turbulent mixing noise and the peaks located divergence is so small that the turbulence of the jet flow may
at its right are termed as broadband shock associated noise. be regarded as in a quasi-equilibrium state. For a system in
Here, it should be kept in mind that the relative magnitudes dynamical equilibrium, statistical mechanics theory suggests
of these components strongly depends upon the direction of that the large-scale fluctuations of the system can be mathe-
noise level measurement. matically represented by a linear superposition of its normal
The studies carried out by Seiner et al. (1992) revealed modes. For the jet flow, the most important normal modes
that the broadband shock-associated noise is dominant in the are the instability wave modes (decomposed as Fourier series
upstream, whereas the turbulent mixing noise is more intense with respect to the azimuthal angle). Thus, if p(r, φ, x, t) is
in the downstream of the jet. They measured the noise direc- the pressure fluctuation associated with the large turbulence
tivity of Mach 2 correctly expanded jet at varied total temper- structures of the jet flow in a cylindrical coordinate system
ature. The study carried out at the stagnation temperature of centered at the nozzle exit with the x−axis pointing along the
500 K at select Strouhal numbers revealed that the dominant jet centerline, then p (and similarly for the other flow vari-
part of turbulent mixing noise is radiated in the angular sec- ables) may be expressed as
tor of about 45◦ to 60◦ , measured from the jet flow direction.
p̂n (r, x, ω) π
Within this dominant noise sector, the spectrum is dominated p (r, φ, x, t) = an (ω) × exp i θn + nφ − ωt +
p̂n (r0.5 , 0, ω) 2
by a single broadband peak which occurs at a Strouhal number (13.48)
of around 0.1 to 0.25. In the upstream of jet, the noise inten-
sity is found to be uniform and nearly flat over a large range ´x
where θn = 0 kn (x, ω) dx, an (ω) is the amplitude of the
of Strouhal numbers. This is the background noise believed to
instability wave, n is the azimuthal mode number, ω is the
be generated by the small-scale eddies, whereas the dominant
angular frequency of the wave, r0.5 is the radial distance from
part of the noise is believed to be produced by the large-scale
the jet axis to the half-velocity point, p̂n (r, x, ω) be the insta-
structures.
bility wave eigenfunction, and kn (x, ω) is the wave number fied the broadband shock-associated noise. Since then a large
(eigenvalue) of the instability function. number of studies by various researchers (Tanna 1977; Seiner
We know that at high Reynolds number the viscosity is and Norum 1979, 1980, 1980; Seiner and Yu (1984); and
not an effective parameter and the turbulent jet flow in the jet Yamamoto et al. 1984) have been carried out.
mixing layer immediately downstream of the nozzle lip has no Norum and Seiner (1982) measured the noise spectra of
intrinsic geometrical length scale. Thus, the flow in jet mixing an underexpanded jet at a constant axial location with 15◦
layer which do not have length or time scales, are referred to interval. It was found that in almost all the spectra the screech
as self-similar. Because of this self-similarity of turbulence tone is emitted at a frequency of 2.6 kHz. The turbulent mixing
statistics, the eddies attain an asymptotic state independent noise component is observed at the left of screech noise and
of the initial conditions at the nozzle lip. In other words, the broadband shock-associated noise component is noticed to
state of structures downstream has no memory of its past. Due its right. Thus, it was established that the shock-associated
to this loss of memory, it allows one to regard the large-scale noise component predominates in upstream of the jet. They
eddies in the jet flow as being generated by random excitations observed the following vital characteristics.
at the nozzle lip such that the excited spectrum has no intrinsic
length or time scales. • The frequency at which peak noise intensity occurs,
changes with the direction of radiation.
13.7.2.2 Noise Generation Mechanism • Each noise spectrum is made of several peaks.
From the stochastic instability wave model, it is understood • The half-width of spectral peak increases with nozzle inlet
that the large turbulence structures are statistically equivalent exit angle.
to instability waves of the jet. Consider an instability wave
Using the theories proposed by Prandtl (1904) and Pack
of a specific azimuthal mode at a given frequency spreading
(1950), Tam et al. (1985) derived a correlation to obtain the
downstream a jet column. This instability wave can be approx-
pressure distribution in the shock cells. From this theory, an
imated as a wavy wall having same wavelength and wave
approximate relation for calculating the shock cell length is
speed as that of the wave. For the wavy wall, it is commonly
accepted that if the wave propagates at supersonic Mach num- 1 D
2 j
bers, enormous noise in the form of Mach waves is produced Ls ≈ π Mj2 − 1 (13.51)
(Fig. 13.17). The wavy wall analogy proposes that the direc- σ1
tions of propagation of the most intense noise radiation and the where Mj is the fully expanded jet Mach number, Dj is the
most amplified instability wave of the jet are same and can fully expanded jet diameter, and σ1 (≈ 2.404826) is the first
be calculated by using Eq. (9.29). In addition, the Strouhal root of the zero-order Bessel function. The fully expanded
number at the peak of the radiated noise spectrum should also values depend upon the level of expansion prevalent at the
be equal to that of the most amplified wave. nozzle exit. When the flow is correctly expanded the jet Mach
The direction of radiation (θ) is related to phase velocity number Mj is equal to the nozzle exit Mach number, also
(C) and the ambient speed of sound (aa ) by the following referred to as design Mach number (Md ) and the jet diameter
Mach wave relation, i.e., Dj is equal to nozzle exit diameter (D). For underexpanded
jets, Mj > Md and Dj > D, however, for overexpanded jets,
aa =C cos θ (13.49) Mj < Md and Dj < D.
Let us consider a free jet exiting from a circular nozzle
Thus, the Eq. (13.49) can also be rewritten as under fully expanded state. The jet Mach number Mj and
uc design Mach number (Md ) can be represented in terms of
f= (13.50) pressure ratios as
θ
Ls 1 + uc cos
aa
γ−1 1
2
2 p0 γ
13.7.2.3 Characteristics of Broadband Mj = −1 (13.52)
Shock-Associated Noise (γ − 1) pa
Through various studies in the past, it is observed that the γ−1 1
2
weak interaction of the downstream propagated large-scale 2 p0 γ
Md = −1 (13.53)
eddies with quasiperiodic shock cells generates the noise com- (γ − 1) pe
ponent referred to as the Broadband shock-associated noise.
This noise component is radiated when the jet is imperfectly where p0 is stagnation pressure in the settling chamber, pe is
expanded. Harper and Fisher (1974) were the first who identi- static pressure at the nozzle exit, and pa is the ambient static
13.7 Jet Noise 353
Mach Waves
a
C
Fig. 13.17 Schematic of Mach wave radiation generated by a wavy surface
pressure. Further, combining the Eqs. (13.52) and (13.53), the fb Dj 0.54
= 1 (13.57)
expression for jet Mach number in terms of static pressure uj 2
Mj2 − 1 1 + 0.7Mj cos θ
ratio at the nozzle exit is
γ−1 1
2
From Eq. (13.57), it is evident that the peak Strouhal num-
2 pe γ (γ − 1) 2 ber for broadband shock-associated noise depends on fully
Mj = 1+ Md − 1
(γ − 1) pa 2 expanded jet Mach number Mj and angle of radiation or
observation angle (θ). Basically, the broadband shock noise
(13.54)
component is generated by the constructive scattering of the
large turbulence structures of the jet flow by the stationary
As can be seen from the Eq. (13.54) that the jet Mach num-
quasiperiodic shock cells in the jet core.
ber depends on the static pressure ratio at nozzle exit and the
design Mach number. However, the latter is only dependent on
13.7.2.4 Characteristics of Screech Noise
the geometrical area ratio of the nozzle and is independent of
Powell (1953) was the first to observe screech tones in super-
pressure. Furthermore, using the conservation of mass prin-
sonic jets. In his pioneering work on choked nozzles above the
ciple, Tam and Tanna (1982) obtained a relation for the jet
design point, a dominant peak in the spectrum was observed
diameter Dj in terms of the nozzle exit diameter (D).
which he called to be a powerful whistle or screech. It was
(γ+1) found that this peak may dominate the spectrum over tur-
(γ−1) 2 4 (γ−1) 1
1+ 2 Mj Md 2 bulent mixing noise and broadband shock- associated noise.
Dj = D (γ−1) 2
(13.55) As an explanation to this phenomenon, they proposed that an
1+ 2 Md
Mj
acoustic feedback mechanism is causing these discrete tones.
Based on the stochastic model, a theory to predict the spectra Upstream propagating acoustics are impinging on the noz-
of broadband shock-associated noise has been proposed by zle lip and forcing instabilities in the thin mixing layer at
Tam (1987). They decomposed the quasiperiodic shock cell the nozzle exit. These Kelvin–Helmholtz instabilities con-
structure into several time-independent modes and superim- vect downstream and grow rapidly in the mixing layer of the
posed the interaction of the instability wave with the distinct jet by extracting energy from the mean flow. Subsequently,
modes to a spectrum of broadband shape. In one-dimensional they interact with the periodic shock cell structure and radi-
analysis, the peak Strouhal number of the broadband shock- ate noise. In case of broadband shock-associated noise, the
associated noise is given by Eq. (??). acoustics are emanated mainly in the upstream direction and
transported outside the jet toward the nozzle. Once they reach
fb Dj uc Dj the nozzle lip, new instabilities are produced and shed down-
= (13.56) stream and the feedback loop is closed.
uj Ls 1 + C∞ucos
c
θ uj Although, many researchers (Davies and Oldfield 1962;
Krothapalli et al. 1986; McLaughlin et al. 1975; Westley and
where Ls is the shock cell spacing and uc is the convective Woolley 1975; Seiner and Norum 1979; Tam and Tanna 1982;
velocity of large-scale eddies. Further, under the assumption Nagel et al. 1983; Norum 1983; Seiner 1984; Gutmark et al.
from the empirical results, uc = 0.7uj and from Eq. (13.51), 1990; Seiner et al. 1992; Raman 1997; and Alkislar et al.
we obtain 2003) have investigated the screech phenomena at a length.
But it is yet, perhaps, the least understood and least predictable the direction of maximum acoustic radiation generated by
component of supersonic jet noise. Similar to the case of the passage of the large-scale structures through the shock
broadband shock-associated noise, it was observed that the cells, be in the direction of the nozzle lip. That is, the fun-
fundamental screech tones also predominately radiates in the damental screech frequency can be estimated by setting
upstream direction, whereas the principal direction of its first θ = π in Eq. (13.50).
harmonic is at 90◦ to the jet flow direction (Norum 1983). • The convective Mach number is 0.7 times the jet Mach
They found that whenever the shock cell structure oscillates, number, i.e., Mc ≈ 0.7 Mj
strong screech tones are radiated by the jet. Recall, it is in • The shock cell spacing was taken to about 20% smaller
contrast to broadband shock noise component, which is radi- than that given by Eq. (13.51), which corresponds to a
ated when the shock cell structure is stationary. Further, when shock cell located at four–five shock cells downstream.
the screech tone is emanated two types of jet motion may
occur. First is the toroidal-mode oscillations and second, is From the above discussion, it is evident that the noise atten-
the flapping or helical mode oscillations. The superposition uation is the most challenging task from a practical point of
of two equal counter-rotating helical mode oscillations cre- view. Since the aerospace community worldwide is focused
ate a flapping mode oscillation. When the shock cell struc- on the development of high-speed civil transport aircraft, it
tures perform helical mode oscillations, both the jet flow and becomes an exigency to gain a thorough understanding about
acoustic field exhibit axisymmetry, whereas in flapping mode the jet turbulence and noise. Nonetheless, the possibility of
oscillations the jet oscillates up and down across a flapping sonic fatigue in aircraft structures due to high speed and cabin
plane. At low supersonic Mach numbers, the screech tone noise issues further aggravate the criticality.
was found to be associated with toroidal-mode oscillations The experience gained from the studies carried out in the
(Norum 1983). However, with increase of jet Mach number past revealed that the augmented jet mixing is proved to be
the mode of oscillations changes from toroidal to flapping. advantageous in reducing the noise substantially. It may be
Furthermore, it is now well established that the intensity of because of the fact that the improved mixing reduces the tur-
screech tone is affected by the jet Mach number, the jet tem- bulence and consequently noise. Sensibly, these investiga-
perature, the nozzle lip thickness and the sound reflecting sur- tions expedited the streamlining of most of the current noise
faces near the jet. Massey et al. (1994) found that at low jet reduction methods into the development of enhanced mix-
Mach numbers, the intensity of screech tone decreases with ing devices, also known as jet controls. Some of the popular
increase of jet temperature. Perhaps because of this reason, jet controls such as grooves and cut-outs, ejectors, tabs, and
the screech tone is not perceived as a potential cause of sonic asymmetric nozzle geometries are discussed in the following
fatigue in jet engines. In the studies carried out by Norum sections.
(1983), it was demonstrated that the thicker nozzle lip acts a
good reflecting surface, which can increase the screech noise
by 10 dB. Tam et al. (1986) derived the following relation to Summary
calculate the screech tone peak Strouhal number.
The free or submerged jet is defined as the momentum-driven
− 1 1 1 −1
fs Dj 2 (γ − 1) 2 − 2 T0 2 unrestricted fluid flow issuing from a nozzle or orifice into
= 0.67 Mj2 − 1 1 + 0.7Mj 1 + Mj
uj 2 Ta quiescent ambience. If it is surrounded by another fluid flow-
(13.58) ing with different velocity at its periphery, it is referred to as
the co-flowing jet. When the jet is incident upon a solid or
where Dj is fully expanded jet diameter, defined by Eq. (13.55) fluid boundary, it is referred to as an impinging jet.
The flows which are both momentum as well as buoyancy
(γ+1)
(γ−1) 2 1 dominated, they are referred as buoyant jet or forced plume.
1+ 2 Mj
4(γ−1)
Md 2
Dj = D (γ−1) 2
The instabilities (large-scale eddies), which moves in
1+ 2 Md
Mj chaotic manner both along and across the jet stream are
assumed to be responsible in exchanging the matter between
To derive the Eq. (13.58), Tam et al. (1986) have taken the the neighboring fluid flows. That is, they transfer mass,
following assumptions. momentum, and energy in transverse direction. In this
process, a region of finite thickness with a continuous dis-
• The nozzle lip is the most vulnerable location of the feed-
tribution of velocity, temperature and species concentration
back loop where the upstream acoustic waves energize
is formed on the boundary between the two flowing fluids.
new instabilities. Thus, it was postulated that the feed-
This region is popularly known as turbulent jet boundary
back acoustic waves must have large enough amplitude to
layer. Since the fluid boundary cannot sustain a pressure
maintain the feedback loop. Therefore, it is essential that
difference across it, the jet boundary becomes a free shear
Summary 355
layer in which the static pressure is constant throughout The studies carried out on the transport processes of mass,
as far as the jet is subsonic. Further, the large size eddies momentum, energy, and vorticity brought out various phe-
created at jet boundary get easily fragmented into relatively nomenological theories. Some of the popular theories are
small-scale eddies while transporting the matter towards
jet centerline. Smaller eddies being the efficient mixing • Boussinesq hypothesis,
promoter assist the low momentum fluid entrained at jet • Prandtl mixing length theory,
boundary to gain momentum from the jet mass of high • von Karman’s similarity hypothesis,
momentum and moves toward the axis of jet. Thus, the • Taylor’s vorticity transport theory,
viscous action penetrates toward the jet axis. In reality, the • Reichardt’s inductive theory.
viscous activity finally reaches at the jet centerline at some
axial distance downstream. Up to this point, the jet velocity Although, the semi-empirical theories discussed above are
remains unaffected along the jet axis. Here, the mixing good enough to provide solutions to simple problems such as
initiated at the jet boundaries has not yet permeated into boundary layer on flat plate, open channel flows, and turbulent
this flow field, thus leaving a region that is characterized jets and wakes. However, the complex flow situations involv-
by a constant axial velocity, just downstream of the nozzle ing, a sudden change in flow direction, recirculatory flow
exit. The profile is similar to a top-hat shape at the nozzle zones, complex boundary conditions cannot be analyzed by
exit. For subsonic jets, this region is called potential jet core. these theories. Similarly, problems involving heat exchange,
Once the viscous activity starts reaching towards the jet axis, chemical reactions, and transport of species in turbulent flows
the top-hat velocity profile begins to shrink in size in the cannot be solved by phenomenological theories and requires
direction of jet propagation. Due to mixing process carried turbulence modeling. A turbulence model is a computational
out by the vortices, the jet centerline velocity decreases procedure which is used to close the system of mean flow
continuous after the core region. Higher the mixing, larger equations.
the decay in jet velocity. This decay is found to be inversely For analyzing the jets, the major parameters considered are
proportional to the axial distance from the nozzle exit. This pressure or velocity decay along the jet centerline and pressure
phenomenon is generally known as characteristics decay. or velocity profiles normal to jet axis. These are evaluated by
The potential core region extends up to about 6 times the performing pressure survey along and normal to jet axis. The
nozzle exit diameter (D) and the characteristics decay region ambient pressure is considered as static pressure at which the
dominates from about 6D to about 12D. After this, the jet jet discharges from its source. Since the subsonic jets are opti-
decay is gradual and approaches to almost zero velocity mally expanded, the measured total pressure can be converted
at a far downstream location of about 30D. This region is into Mach number or velocity using ambient static pressure.
referred to as fully developed region. Here, even though the However, it is not always the case with sonic and supersonic
jet velocity decreases continuously but the velocity profile jets. This is due to the presence of complex wave structure
retains its shape at every downstream axial location. Thus, it inside the jet core. Also, the pressure measurement is usually
is also termed as self-similar region. done with pitot probe which measures the pressure behind
Unlike subsonic jet, there is no constant axial velocity zone the bow- or curved-shock. Thus, in order to evaluate the total
exists in a supersonic jet due to the presence of shock and pressure upstream of the shock, either the Mach number or
expansion waves. This region is highly viscous dominated the static pressure must be known separately. But the precise
referred to as supersonic jet core or simply the jet core. Here, calculation of both Mach number and static pressure is diffi-
the core can be defined as the axial extent up to which waves cult due to the wave strength variation at any jet cross-section.
dominate. Therefore, the conversion of total or stagnation pressure into
The jets can be classified based on either pressure or Mach Mach number or velocity is usually not attempted and they
number prevalent at the exit of a nozzle or orifice. The jets are used as it is in order to study the sonic and supersonic jet
can be broadly classified into incompressible and compress- characteristics.
ible jets. The jets with Mach number less than 0.3 till which the In various technological applications, the performance can
compressibility effects are ignored are called incompressible be improved by augmenting the mixing of jet with the ambi-
jets. The jet with Mach number more than 0.3, are referred to ent fluid to which it is discharged. One such application is the
as compressible jets. The compressible jets are further clas- increase of combustion efficiency, which can be improved by
sified into subsonic (M < 1) jets, sonic (M = 1) jets, and increasing the rate of mixing of fuel with air. For Supersonic
supersonic (M > 1) jets. The subsonic jets are always cor- Combustion Ramjet (SCRAM) engines, even this augmented
rectly expanded, whereas the sonic jets can be either correctly mixing should be completed within the short distance to min-
expanded or underexpanded. The supersonic jets are sub- imize the size of the combustion chamber. Further, for an
classified as overexpanded, correctly expanded, and under- efficient combustion, both large- and small-scale mixing is
expanded jets. desired as the large- scale eddies determine the rapidity of
mixing and small-scale eddies ensure the effective molec- Eq. (13.28). Moreover, the sound pressure level (SPL) is cal-
ular level mixing. Also, with increased mixing of a nozzle culated by using
exhaust, the infrared signature of hot jet can be reduced,
thereby increasing the stealth capability. The other technolog- Prms
SPL = 20 log10
ical applications, where the mixing enhancement is required, Pref
include thrust vector controls, gas dynamic lasers, thrust aug-
menting ejectors, etc. where Pref = 2.14 × 10−5 Nm−2 , corresponds to Iref in a
All the flow control methods are broadly classified into plane wave and Prms is the root mean square pressure.
active and passive controls. In an active control technique, All the turbulent jets contain both large-scale and small-
the additional source of energy is required to activate the con- scale eddies capable of generating noise. However, the relative
trol. The pulsed jets, piezoelectric actuators, micro-jets, and intensity of the noise produced by these structures strongly
oscillating jets are among the most effective control technique depends upon the jet Mach number and the jet temperature.
for mixing enhancement. In contrast, the passive controls do For cold subsonic jets, the convective Mach number, with
not require any additional energy source for its action, rather respect to the ambient speed of sound, of large scale eddies is
they make use of geometrical modification of the nozzle exit. subsonic and thus, they are ineffective noise producers. There-
The grooves and cut-outs at the nozzle exit, placement of tabs fore, the dominant part of jet noise is created by small-scale
at the nozzle exit, use of multi-jets, lobbed nozzles etc., are eddies only. On the other hand, in supersonic jets especially at
commonly used passive controls. elevated temperatures, the large-scale vortices convect down-
Acoustics is that branch of science that studies the prop- stream at supersonic Mach numbers. Thus, the large-scale
agation of sound waves as in the combustor, jet acoustics, eddies are the dominant source of supersonic jet noise. More-
meteorological acoustics, and environmental acoustics. The over, due to imperfect levels of expansion (overexpansion or
sound waves may be generated in stagnant air even with very underexpansion) of supersonic jet, a quasi- shock cell struc-
small pressure-induced excitations. In turn, a noise could ture forms within the core. Because of this, an additional com-
be produced which is in audible range to humans. In such ponent of noise is generated referred to as shock- associated
cases, except the pressure all other flow parameters remain noise, which is further subclassified into broadband shock-
constant. Contrarily, in some cases, the noise level may be associated noise and the screech tone. The former, as the name
extremely high as in thunderstorms or in sudden explosions. suggests, is a broadband noise, whereas the latter occurs at
However, still no appreciable changes in flow variables other discrete frequencies. Thus, the imperfectly expanded super-
than the pressure are seen in these cases as well. This phe- sonic jet comprises of three parts namely turbulent mixing
nomenon is termed as the pressure mode acoustics. More- noise, broadband shock-associated noise, and screech tone.
over, due to circulatory motion significantly large velocity
gradients are established within the fluid, vortical waves are
produced which, in turn, cause pressure disturbances. The Exercises
noise produced by these vortical structures are classified as
the vorticity mode acoustics. Furthermore, in several occa- Descriptive Type Questions
sions such as in hypersonic flows past a spacecraft, a rapid
increase in temperature is encountered creating the entropy 1. Define jets. How it is different from a plume? Discuss
layer between the shock layer and the momentum bound- different zones of a typical subsonic jet.
ary layer. This entropy layer leads to pressure fluctuations 2. Discuss the various levels of expansion at the nozzle exit.
and since the entropy waves are predominant in this case, 3. Discuss the role of shear layers in subsonic and supersonic
the phenomena is categorized as the entropy mode acoustics. jet mixing.
Nevertheless, it is certainly possible that the noise level may 4. What do you mean by “jet control”? Discuss various avail-
actually be the combination of all three modes in a given able techniques to achieve it. Also, discuss their relative
physical situation. strengths and weaknesses.
The noise level is then determined either by the acous- 5. Discuss the “axis-switch” phenomena in jets. Why an early
tic intensity level (IL) or by the sound pressure level (SPL), “axis-switch” signifies better mixing?
measured in decibel (dB). 6. What is optical flow visualization? Write a short note on
various optical flow visualization techniques. Why it is
I
IL = 10 log10 actively being used in jet research?
Iref 7. Explain the phenomena of noise generation in a subsonic
jet. Compare it with the noise generation mechanism in a
where Iref = 10−12 Wm−2 at 1000 Hz and I is the scalar supersonic jet.
acoustic intensity normal to the surface as obtained from
Exercises 357
Multiple Choice Questions 6. If fj is the vortex shedding frequency, D is the nozzle exit
diameter, and U0 is the jet velocity at the nozzle exit, then
1. The shape of a plume is influenced if the local wind blows the Strouhal number (StD ) is defined as
in the same direction as of the plume. In this co-flow
fj D2
arrangement which is initially buoyancy dominated but (a) U0
becomes momentum dominated in the far field region, the √
fj D
transition is generally speculated in terms of a dimension- (b) U0
less number called (c) U0
fj D
fj D
(a) Biot number (d) U0
(b) Richardson number
(c) Prandtl number 7. Since all the turbulent jets contain both large-scale and
(d) Karman number small-scale eddies capable of generating noise. However,
the relative intensity of the noise produced by these struc-
2. The constant axial velocity zone in jets is known as tures strongly depends upon the jet Mach number and the
jet temperature. Now consider the following statements.
(a) potential core region (I). In cold subsonic jets, the large-scale eddies are effective
(b) characteristic decay region noise producers.
(c) fully developed region (II). Small-scale eddies are the dominant source of super-
(d) self-similar region sonic jet noise.
(III). In subsonic jets, the turbulent mixing is the major
3. The mixing length obtained by Taylor’s vorticity trans-
source of noise.
port theory is higher than that obtained by Prandtl’s
(IV). The jet noise at supersonic Mach numbers comprises
momentum transfer theory.
of both turbulent mixing noise and the shock-associated
noise.
(a) 2 times
Select the correct statements from the following.
(b) 3√times
(c) √2 times (a) I and II
(d) 3 times
(b) II and III
(c) III and IV
4. From a convergent–divergent nozzle, the overexpanded jet
(d) all
is obtained when
8. Consider the following statements.
(a) the nozzle exit pressure is higher than the ambient
(I). For a free turbulent flow, the continuity equation along
pressure.
with the Reynolds-Averaged Navier–Stokes equations of
(b) the nozzle exit pressure is lower than the ambient pres-
motion form a closed system of equations and thus can be
sure.
fully solved.
(c) the nozzle exit pressure is equal to the ambient pres-
(II). The density gradients in compressible shear layers,
sure.
coupled with large pressure gradients across shocks pro-
(d) the jet is exiting to a vacuum.
duce vorticity due to the baroclinic torque.
5. Consider the following expansion levels at the exit of a Which of the above statements is/are TRUE?
nozzle.
(a) I only
(I) overexpansion (b) II only
(II) correct-Expansion (c) I and II both
(III) underexpansion (d) none
A convergent nozzle can produce 9. Mixing in supersonic shear layers is critically dependent
upon the compressibility effects which is best described by
(a) I only the convective Mach number. The convective Mach num-
(b) I and II bers in two different streams (Mc1 and Mc2 ) are equal to
(c) II and III each other when both streams have equal static pressures
(d) all the above and specific heat ratios. If U1 and U2 are the mean veloci-
ties and a1 and a2 are the speeds of sound in these mixing Birky P, Page GJ (2001) Numerical predictions of turbulent under
layers, then Mc1 and Mc2 are given by expanded sonic jets using a pressure-based methodology. Proc Inst
Mech Eng, Part G: J Aerosp Eng 215:165–173
(U1 +U2 ) Boussinesq JV (1897) Théorie de l’écoulement tourbillonnant et
(a) Mc1 = Mc2 = (a1 +a2 ) tumultueux des liquides dans les lits rectilignes a grande section.
(U1 +U2 ) Gauthier-Villars et fils. (in French)
(b) Mc1 = Mc2 = (a1 −a2 ) Bradshaw P, Ferris O, Johnson RF (1964) Turbulence in noise producing
(U1 −U2 )
(c) Mc1 = Mc2 = (a1 −a2 )
region of a circular jet. J Fluid Mech 19:591–624
Browand FK, Weidman PD (1976) Large scales in the developing mixing
(U1 −U2 )
(d) Mc1 = Mc2 = (a1 +a2 )
layers. J Fluid Mech 76:127–144
Brown GL, Roshko A (1974) On density effects and large structure in
turbulent mixing layers. J Fluid Mech 64(4):775–816
10. The direction of noise radiation (θ) is related to the phase
Cai J, Tsai HM, Liu F (2010) Numerical simulation of vortical ows in
velocity (C) and the ambient speed of sound (aa ) by which the near eld of jets from notched circular nozzles. Comput & Fluids
of the following relation? 39:539–552
Chauvet N, Deck S, Jacquin L (2007) Numerical study of mixing
(a) C = aa cos θ enhancement in supersonic round jets. AIAA J 45(7):1675–1687
Chue SH (1975) Pressure probes for uid measurement. Prog Aerosp Sci
(b) aa = C cos θ
16(2):147–223
(c) C = aa sin θ Chung TJ (1996) Applied Continuum Mechanics. Cambridge University
(d) aa = C tan θ Press, London, UK
Crighton D (1973) Instability of an elliptic jet. J Fluid Mech 59:665–672
Crighton DG (1975) Basic principles of aerodynamic noise generation.
Keys Prog Aerosp Sci 16:31–96
Crow SC, Champagne FH (1971) Orderly structure in jet turbulence. J
1. (b) Fluid Mech 48(3):547–591
2. (a) Curle N (1955) The influence of solid boundaries on aerodynamic sound.
Proc R Soc Lond Ser A: Math Phys Sci 231:505–514
3. (c) Davies MG, Oldfield DE (1962) Tones from a choked axisymmetric
4. (b) jet. I. Cell structure, eddy velocity and source locations, Acustica.
5. (c) 12:257–267
6. (d) Domaradzki JA, Holm DD (2001) Navier–Stokes-alpha model: LES
equations with nonlinear dispersion. Modern simulation strategies
7. (c) for turbulent flow. ERCOFTAC Bulletin 48:2. Edwards Publishing
8. (b) Durbin PA (1996) On the k-ε stagnation point anomaly. J Heat Fluid
9. (d) Flow 17:89–90
10. (b) Elliott GS, Samimy M, Arnette SA (1992) Study of compressible mixing
layers using filtered rayleigh scattering based visualizations. AIAA
J 30(10):2567–2569
Elliott GS, Samimy M, Arnette SA (1993) The evolution of large scale
structure in compressible mixing layers. In: 9th symposium on tur-
References bulent shear flows, Kyoto, Japan
Faivre V, Poinsot T (2004) Experimental and numerical investigations
Abramovich GN (1963) The Theory of Turbulent Jets. MIT Press, Cam- of jet active control for combustion applications. J Turbul 5(1):025
bridge Ffowcs Williams JE (1969) Hydrodynamic noise. Ann Rev Fluid Mech
Agui JC, Hesselink L (1988) Flow visualization and numerical analy- 1:197–222
sis of a co-flowing jet: a three-dimensional approach. J Fluid Mech Ffowcs Williams JE (1977) Aeroacoustics. Ann Rev Fluid Mech
191:19–45 9:447–468
Ahuja AA (1973) Correlation and prediction of jet noise. J Sound Vib Ffowcs Williams JE, Hall LH (1970) Aerodynamic sound generation by
29(2):155–168 turbulent flow in the vicinity of a scattering half plane. J Fluid Mech
Alkislar MB, Krothapalli A, Lourenco LM (2003) Structure of a screech- 40(4):657–670
ing rectangular jet: a stereoscopic particle image velocimetry study. Franken PA, Westervelt PJ, Dyer I (1958) Jet noise reduction by induced
J Fluid Mech 489:121–154 flow. J Acoust Soc Am 30(8):761–764
Ball CG, Fellouah H, Pollard A (2012) The flow field in turbulent round Freund JB, Moin P (1998) Mixing enhancement in jet exhaust using
free jets. Prog Aerosp Sci 50:1–26 fluidic actuators: direct numerical simulations. Proc ASME Fluids
Bandyopadhyay PR, Watson RD (1988) Structure of rough-wall turbu- Eng Div Summer Meet. Washington DC, USA. FEDSM98–5235
lent boundary layers. Phys Fluids 31(7):1877–1883 Georgiadis NJ, DeBonis JR (2007) Navier-Stokes analysis methods for
Batchelor GK, Gill AE (1962) Analysis of the stability of axisymmetric turbulent jet flows with application to aircraft exhaust nozzles. Prog
jets. J Fluid Mech 14:529–551 Aerosp Sci 42:377–418
Beavers GS, Wilson TA (1970) Vortex growth in jets. J Fluid Mech Gohil TB, Saha AK, Muralidhar K (2012) Numerical study of instability
44:97–112 mechanisms in a circular jet at low reynolds numbers. Comput &
Berg JR, Ormiston SJ, Soliman HM (2006) Prediction of the flow struc- Fluids 64: 1–18
ture in a turbulent rectangular free jet. Int Commun Heat Mass Transf Goldstein ME (1984) Aeroacoustics of turbulent shear flows. Ann Rev
33:552–563 Fluid Mech 16:263–285
Grant AJ (1974) A numerical model of instability in axisymmetric jets.
J Fluid Mech 66(4):707–724
References 359
Gutmark EJ, Schadow KC, Bicker CJ (1990) Near acoustic field Matsuda T, Umeda Y, Ishii R, Yasuda A (1987) Numerical and experi-
and shock structure of rectangular supersonic jets. AIAA J 28(7): mental studies on choked under expanded jets. AIAA-87–1378
1163–1170 McLaughlin DK, Morrison GL, Troutt TR (1975) Experiments on the
Harper BM, Fisher MJ (1974) The noise from shock waves in supersonic instability waves in a supersonic jet and their acoustic radiation. J
jets. AGARD-CP-131. 11, pp. 1–13 Fluid Mech 69:73–95
Hattori H, Nagano Y (2004) Direct numerical simulation of turbulent Michalke A (1971) Instability of compressible circular free jet with con-
heat transfer in plane impinging jet. Int J Heat Fluid Flow 25:749– sideration of the influence of the jet boundary layer thickness. NASA-
758 TM-75190
Hawkings DL (1977) A generalized analysis of the vibration of circular Michalke A (1984) Survey on jet instability theory. Prog Aerosp Sci
rings. J Sound Vib 54(l): 67–74 21:159–199
Heinz S (2003) A model for the reduction of the turbulent energy redis- Miller RS, Madnia CK, Givi P (1995) Numerical simulation of non-
tribution by compressibility. Phys Fluids 5(11):3580–3583 circular jets. Comput & Fluids 24(1): l–25
Hilgers A (2000) Control and optimization of turbulent jet mixing. Cen- Morris PJ, Giridharan MG, Lilley GM (1990) On the turbulent mixing
ter for Turbulence Research, Annual Research Briefs, pp 47–54 of compressible free shear layers. Proc R Soc Lond A 431:219–243
Ho CM, Gutmark EJ (1987) Vortex induction and mass entrainment in Morrison GL, McLaughlin DK (1979) The noise generation by instabili-
a small-aspect ratio elliptic jet. J Fluid Mech 179:383–405 ties in low reynolds number supersonic jets. J Sound Vib 65:177–191
Hussain AKMF (1986) Coherent structures and turbulence. J Fluid Mech Mouldon F, Acharya S (2010) Direct Numerical Simulation of pulsed
173:303–356 jets in cross- ow. Comput & Fluids 39: 1745–1773
Kaushik M, Thakur PS, Rathakrishnan E (2006) Studies on the effect of Nagel R, Papathansiou A (1983) An experimental study of supersonic jet
notches on circular sonic jet mixing. J Propuls Power 22:211–214 shock-associated noise. Proceedings of the AIAA 8th Aeroacoustics
Kaushik M, Rakesh K, Humrutha G (2015) Review of computational Conference. Atlanta, USA. AIAA-1983-708
fluid dynamics studies on jets. Am J Fluid Dyn 5(3A):1–11 Norum TD (1983) Screech suppression in supersonic jets. AIAA J
Kosovic B (1997) Subgrid-scale modelling for the large-eddy simulation 21(2):235–240
of high-Reynolds-number boundary layers. J Fluid Mech 336:151– Norum TD, Seiner JM (1982) Broadband shock noise from supersonic
182 jets. AIAA J 20(1):68–73
Krothapalli A, Bagano D, Karamcheti K (1981) On the mixing of a Pack DC (1950) A note on Prandtl’s formula for the wavelength of a
rectangular jet. J Fluid Mech 107:201–220 supersonic gas jet. Q J Mech Appl Math 3:173–181
Krothapalli A, Hsia Y, Bagano D, Karamcheti K (1986) The role of Papamoschou D (1990) Communication paths in the compressible shear
screech tones on mixing of an underexpanded jet. J Sound Vib layers. AIAA Paper 90–0155
106:119–143 Papamoschou D (1991) Structure of the compressible turbulent shear
Krothapalli A, McDaniel J, Bagano D (1990) Effect of slotting on the layers. AIAA J 29(5):680–681
noise of an axisymmetric supersonic jet. AIAA J 28(12):2136–2138 Papamoschou D, Debiasi M (1999) Noise measurements in super-
Lardeau S, Lamballais E, Bonnet JP (2002) Direct numerical simulation sonic jets treated with the mach wave elimination method. AIAA
of a jet controlled by uid injection. J Turbul 3(2) J 37(2):154–160
Leray J (1934) On the motion of a viscous liquid filling space. Acta Poinsot T, Lele S (1992) Boundary conditions for direct simulations of
Mathematica 63:193–248 compressible viscous flows. J Comput Phys 101:104–129
Liepmann D, Gharib M (1992) The role of streamwise vorticity in the Pope SB (1978) An explanation of the turbulent Round-Jet/Plane-Jet
near-field entrainment of round jets. J Fluid Mech 245:643–668 anomaly. AIAA J 16(3):279–281
Lighthill MM (1952) On sound generated aerodynamically:i. general Pope SB (2000) Turbulent Flows. Cambridge University Press, New
theory. Proc R Soc Lond Ser A: Math Phys Sci 211:564–581 York
Lighthill MM (1954) On sound generated aerodynamically: turbulence Powell A (1953) On the mechanism of choked jet noise. Proc Phys Soc
as a source of sound. Proc R Soc Lond Ser A: Math Phys Sci 222:1–32 Lond 66:1039–1056
Lighthill MM (1963) Jet noise. AIAA J 1(7):1507–1517 Powell A (1960) Aerodynamic noise and the plane boundary. J Acoust
Lin JC, Selby GV, Howard FG (1991) Exploratory study of vortex- Soc Am 32(8):982–990
generating devices for turbulent flow separation control. AIAA Paper Prandtl L (1904) Beiträge zur Theorie der Dampfströmung durch Düsen.
910042 Z. Ver. Dtsch. Ing. 48(10):348–350 (in German)
Liu Y, Tucker PG, Kerr RM (2008) Linear and nonlinear model large- Prandtl L (1925) Bericht über Untersuchungen zur ausgebildeten Tur-
eddy simulations of a plane jet. Comput & Fluids 37: 439–449 bulenz. ZAMM. 5:136–139
Lodato G, Domingo P, Vervisch L (2008) Three-dimensional bound- Rajarathnam N (1976) Turbulent Jets. Elsevier Scientific Publishing Co.,
ary conditions for direct and large-eddy simulation of compressible Netherlands
viscous flows. J Comput Phy 227:5105–5143 Raman G (1997) Screech tones from rectangular jets with spanwise
Lush PA (1971) Measurements of subsonic jet noise and comparison oblique shock cell structures. J Fluid Mech 330:141–168
with theory. J Fluid Mech 46(3):477–500 Rembold B, Adams NA, Kleiser L (2002) Direct numerical simulation
Lyubimov DA (2008) Development and applications of the efficient of a transitional rectangular jet. Int J Heat Fluid Flow 23:547–553
hybrid RANS/ILES approach for the calculation of complex turbu- Reynolds O (1895) On the dynamical theory of incompressible viscous
lent jets. High Temp 46(2):243–253 fluids and the determination of the criterion. Philos Trans R Soc Lond
Maidi M, Lesieur M (2005) Large eddy simulations of spatially growing 186(A): 123–164
subsonic and supersonic turbulent round jets. J Turbul 6(38):1–20 Ribner HS (1981) Perspective on jet noise. AIAA J 19(12):1513–1526
Maidi M, Lesieur M, Metais O (2006) Vortex control in large-eddy sim- Sandham ND, Reynolds WC (1990) Compressible mixing layer: linear
ulations of compressible round jets. J Turbul 7(49):1–22 theory and direct simulation. AIAA J 28:618–624
Martin JE, Meiburg E (1991) Numerical investigation of three- Sarkar S, Erlebacher G, Hussaini MY, Kreiss HO (1991) The analysis
dimensionally evolving jets subject to axisymmetric and azimuthal and modeling of dilatational terms in compressible turbulence. J Fluid
perturbations. J Fluid Mech 230:271–318 Mech 227:473–493
Massey K, Ahuja K, Jones R, Tam CKW (1994) Screech tones of super- Sarkar S, Lakshmanan B (1991) Application of a reynolds stress turbu-
sonic heated free jets. Proceedings of the 32nd Aerospace Sciences lence model to the compressible shear layer. AIAA J 29(5):743–749
Meeting and Exhibit. Aerosp Sci Meet AIAA-1994-141
Sarohia V, Massier PF (1978) Experimental results of large scale struc- Tandra DS, Kaliazine A, Cormack DE, Tran HN (2006) Numerical simu-
tures in jet flows and their relation to jet noise production. AIAA J lation of supersonic jet flow using a modified k-ε model. Int J Comput
16(8):831–835 Fluid Dyn 20(1):19–27
Schlichting H (1951) Boundary Layer Theory. McGraw-Hill Co., New Tanna HK (1977) An experimental study of jet noise part i: turbulent
York, USA mixing noise. J Sound Vib 50(3):405–428
Seiner JM (1984) Advances in high speed jet aeroacoustics. AIAA Pap Terrier DA, Lu FK (2001) Numerical Study of Passive Control of an
84–2275 Over Expanded Nozzle for Fighter Aircraft. Proc 23rd Int Symp
Seiner JM, Krejsa EA (1989) Supersonic jet noise and the high speed Shock Waves. Texas, USA. ISSW24-3090
civil transport. AIAA Pap 89–2358 Thies AT, Tam CKW (1996) Computation of turbulent axi-symmetric
Seiner JM, Norum TD (1979) Experiments on shock associated noise of and non-axisymmetric jet flows using the K-epsilon model. AIAA J
supersonic jets. AIAA Pap 79–1526 34(2):309–316
Seiner JM, Norum TD (1980) Aerodynamics aspects of shock containing Tsujimoto K, Koji Ao, Shakouchiand T, Ando T (2011) Numerical
jet plumes. AIAA Pap 80–0965 investigation on flow structures and mixing performances of vector-
Seiner JM, Ponton MK, Jansen BJ, Lagen NT (1992) The effect of tem- controlled free jet using DNS. J Fluid Sci Technol 6(4):401–411
perature on supersonic jet noise emission. AIAA Pap 92–02–046 Tsujimoto K, Shakouchi T, Sasazaki S, Ando T (2006) Direct numerical
Seiner JM, Yu JC (1984) Acoustic near field properties associated with simulation of jet mixing control using combined jets. JSME Int J:
broadband shock noise. AIAA J 22:1207–1215 Ser B. 49(4):966–973
Pai S-I (1954) Fluid Dynamics of Jets, 1st edn. D Van Nostrand Company, von-Kármán T (1930) Mechanische Ähnlichkeit und Turbulenz
Torronto [Mechanical similarity and turbulence]. Nachrichten von der
Smagorinsky J (1963) General circulation experiments with the primitive Gesellschaft der Wissenschaften zu Göttingen Mathematisch-
equation: I. The basic experiment. Mon Weather Rev 91:99–164 physikalische Klasse (in German)
Stanley SA, Sarkar S, Mellado JP (2002) A study of the flow-field evo- Wang PC, McGuirk JJ (2013) Large Eddy Simulation of supersonic jet
lution and mixing in a planar turbulent jet using direct numerical plumes from rectangular convergent divergent nozzles. Int J Heat
simulation. J Fluid Mech 450:377–407 Fluid Flow 43:62–73
Suto H, Matsubara K, Kobayashi M, Kaneko Y (2004) Large eddy sim- Westly R, Woolley JH (1975) The near eld sound pressures of a choked
ulation of flow and scalar transport in a round jet. Heat Transf-Asian jet when oscillating in the spinning mode. AIAA-75-479
Res 33(3):175–188 Winant CD, Browand FK (1974) Vortex pairing: The mechanism of
Tam CKW (1995) Supersonic jet noise. Ann Rev Fluid Mech 27:17–43 turbulent mixing-layer growth at moderate reynolds numbers. J Fluid
Tam CKW, Chen P (1979) A statistical model of turbulence in twodi- Mech 63(2):237–255
mensional mixing layers. J Fluid Mech 192:303–306 Yamamoto S, Ozeki O (1984) Conducted noise measurements of auto-
Tam CKW, Jackson JA, Seiner JM (1985) A multiple scale model of the motive electrical and electronic devices using arti cial network. Elec-
shock cell structure of imperfectly expanded sonic jets. J Fluid Mech tronics and Communications in Japan (Part I: Communications).
153:123–149 67(8): 81–89
Tam CKW, Seiner JM, Yu JC (1986) Proposed relationship between Yoshizawa A (1993) Bridging between eddy-viscosity-type and sec-
broadband shock associated noise and screech tones. J Sound Vib ondorder turbulence models through a two-scale turbulence theory.
110:309–321 Phys Rev 48(1):273–281
Tam CKW, Tanna HK (1982) Shock associated noise of supersonic jets Zaman KBMQ, Hussain AKMF (1981) Turbulence suppression in free
from convergent-divergent nozzles. J Sound Vib 81(3):337–358 shear flows by controlled excitation. J Fluid Mech 103:133–159
Zeman O (1990) Dilatation dissipation: the concept and application in
modeling compressible mixing layers. Phys Fluids 2(2):178–188
Shock Wave and Boundary Layer
Interactions 14
The interaction of a shock wave with boundary layer is called difference between the two is that the shock wave is con-
the Shock–Boundary Layer Interactions (SBLIs). This inter- siderably thinner as compared to an ordinary boundary layer.
action has a profound effect on the performance of a vehicle Moreover, across a thin shock wave a large pressure gradient
flying at high Mach numbers, especially in hypersonic flow exists.
regime. Because of their ubiquitous presence at supersonic The shock waves when impinges on the solid boundaries,
and hypersonic speeds and their influence on the vehicle per- they tend to impose large pressure gradients on the boundary
formance, the studies on SBLIs have been a challenging task layer. In turn, the boundary layer is necessarily distorted to a
among the researchers for past seven decades. This chapter considerable extent. These effects on the boundary layer are,
explores the fundamental aspects of SBLIs with an emphasis moreover, propagated both upstream and downstream through
on understanding the physics behind these interactions. The the subsonic portion of the layer. The consequent changes
major outcomes of a few experimental studies performed in in the boundary layer simultaneously act in such a manner
the author’s laboratory are also discussed. as to propagate compression and expansion waves into the
external flow. These, in turn, tend to change the original shock
pattern. In addition, due to convergence of a shock wave and
14.1 Introduction the boundary layer, their mutual interactions of considerable
extent are highly probable. These interactions are popularly
We know that if the flow past a body a thin viscous region known as Shock–Boundary Layer Interactions (SBLIs). Since
adjacent to the surface develops and grows downstream. This both may be found in almost every supersonic flow and hence
thin region where the viscous effects predominate is called their interactions are not unusual. The common occurrence is
the boundary layer. Beyond the boundary layer, farther away witnessed when an externally generated shock wave advances
from wall, the flow is assumed to be inviscid and irrotational, over an area on which there is a boundary layer.
whereas the flow within the boundary layer might be rota-
tional owing to the shear stress present on the surface. From
the Newton’s law of viscosity,
the shear stress is given by 14.1.1 Transonic Interactions
the relation τ = μ dy . Thus, the shear stress, in general,
du
flows depend upon the viscosity of the fluid and the velocity The transonic normal shock wave and turbulent boundary
gradient in the flow field. No matter how small the viscosity layer interactions, referred to as transonic SBLIs comprise
is the fluid sticks to the surface of a solid body, i.e., the fluid of supersonic flow upstream of the shock and a subsonic flow
layer near to the surface has zero relative velocity with respect downstream. This mixed regions of the flow make the tran-
to the surface. However, at a very short distance away from sonic flow interactions distinct from the supersonic and hyper-
the body the flow velocity is equal to the freestream veloc- sonic interactions. Since the steady subsonic flow is always
ity. Therefore, a large velocity gradient is present inside the optimally expanded without having waves and thus, under-
boundary layer and despite low viscosity, the shear stresses of goes gradual changes compared to supersonic flow. Never-
sizable magnitude is generated on the surface, which in turn, theless, the downstream subsonic flow conditions can feed
make the flow rotational. upstream and affect the strength of shock, shape, and loca-
Similar to boundary layers, the shock waves are also the tion of shock wave which causes the interactions, while the
flow discontinuities in which both the viscosity as well as upstream flow conditions are shielded from such events by
the heat conduction play a dominant role. The significant supersonic outer flow. The transonic interactions can be found

362 14 Shock Wave and Boundary Layer Interactions
in the intake of jet engines, where incoming supersonic flow is by the same angle θ1 , making the flow parallel to the wall
compressed and decelerated to subsonic speeds. Other exam- CB. Let us assume that M2 and M3 are the gas flow Mach
ples are transonic-aircraft wings, where the shock wave over numbers in the regions 2 and 3 , respectively. Then,
the wing interacts the boundary layer; and transonic turbine
and compressor blade cascades, where the local regions of M1 > M2 > M3 (14.1)
supersonic flow induce shock waves which, in turn, interacts
with the boundary layer on the blade. In this chapter, we will From Eq. 14.1, it is clear that for the same flow deflection
confine ourselves mainly discussing the important character- angles (θ),ss the shock angles (β) of incident and reflected
istics of transonic SBLIs in airbreathing engine intakes.1 shock waves are different. The reflected shock wave CG
impinges upon the surface FE, another reflection occurs and
the entire phenomena repeats. Theoretically, the reflection of
14.2 Supersonic Flow Field Characteristics oblique shock wave continues till the shock upstream Mach
number and the shock angle are sufficient enough to keep the
Before we begin our actual discussion on SBLIs, some typical wave attached to the wall surface. These types of reflections
characteristics of a supersonic flow field will be reviewed in are termed as regular or like reflections.
the following sections. In a situation, where an attached wave is not possible, the
wave becomes normal to the wall and curves-out (Fig. 14.3).
It is tangential to the incident oblique shock wave (DE) which
14.2.1 Simple Wave and Non-simple Region is called the Mach shock wave (CE). The flow becomes sub-
sonic across the wave CE. The incident shock wave (DE) gets
The simple waves are straight Mach lines with constant con- reflected at the point E into the wave (EF). A slip-stream,
ditions on each one and which follow the simple relation across which the flow properties except pressure are discon-
between flow deflection angle (θ) and Prandtl–Meyer func- tinuous, is generated at the point E.
tion (ν). Supersonic expansion and compression with Mach
line which are straight is termed as simple region. Two oppo-
site families of waves, left and right running waves, depend 14.2.3 Reflection of an Oblique Shock Wave
upon the walls producing these waves are either left or right of from a Free Pressure Boundary
the waves. In the regions, where two simple waves of opposite
families interact with each other, the flow is non-simple and Consider a jet issuing out of a convergent–divergent nozzle as
thus linear relationship between ν and θ is no more valid. The shown in Fig. 14.4. In the region 1 , the static pressure (p1 )
schematic diagram of a typical supersonic field with simple would be the same as that of the pressure of the jet boundary.
waves and non-simple regions are depicted in Fig. 14.1. When the flow crosses the incident wave AC, static pressure
increases to p2 (> pa ). Also, the wave turns away the flow
toward jet boundary. However, at point C, the physical bound-
14.2.2 Reflection of an Oblique Shock Wave ary condition requires the static pressure to remain equal to
from a Solid Wall pa .
As p2 > pa , a wave CB is reflected back from the point
Consider an uniform stream with incident Mach number C in the jet in order to reduce the pressure back to the value
(M1 ), flowing between two frictionless walls ACB and DFE pa . In supersonic flow field, the reduction in static pressure
as shown in Fig. 14.2. The upper wall (ACB) and portion of is only possible through an expansion wave. Thus, the wave
the lower wall (DF) are parallel to each other. However, at CB is an expansion fan, turning the flow away from the wave.
point F, there is a concave corner such that the wall FE sub- Therefore, the boundary of jet at C turns outward.
tends the angle θ1 with respect to the incident stream, parallel
to the walls ACB and DF.
Due to concave corner, an oblique shock wave (FC) 14.2.4 Oblique Shock Wave Cancelation
emanates from point F and impinges upon the wall at C,
which deflects the flow at an angle of θ1 , toward the wall Figure 14.5 shows an incident oblique shock wave (DC)
ACB. The physical boundary condition requires that the flow impinging on the wall AC at point C, which deflects the flow
downstream of the shock has to be parallel to the wall. It toward the wave through the turning angle θ and increases the
is accomplished through a reflected wave (CG) emanating static pressure of incoming stream. At point C, the physical
from the point C, which turns the stream away from the wall boundary condition requires that the downstream flow to be
parallel to the wall. If the wall CB would have been existent,
1 The readers are advised to refer Appendix “A” for basic details about a reflected wave (expansion wave) will be emanated making
airbreathing engine intakes.
14.2 Supersonic Flow Field Characteristics 363
Right−running wave
Uniform stream Non−simple region
Left−running wave
Fig. 14.1 Simple waves and non-simple regions in isentropic supersonic field
C
A B
3
1 M2
M1 M3
Reflected shock wave
Incident shock wave 2
E
D
F
Fig. 14.2 Regular or like reflection of an oblique shock wave from a solid wall
C
A B
Wall Mach shock
E Slip line
Incident shock
Reflected shock
D F
Fig. 14.3 Reflection of Mach shock wave
C M3
Freestream boundary p = p
3 a
M2
pa p =p
1 a B
p >p
2 a Reflected expansion waves
M1
Incident oblique shock
Fig. 14.4 Unlike reflection of an oblique shock wave from a free pressure boundary
B
Frictionless surface
A Frictionless surface C
B’
M1 M2
Incident shock wave
Fig. 14.5 Phenomena of the shock wave cancelation
Bend shock waves
Expansion fans
Incident shock wave Weak reflections
Solid wall
Solid wall
Fig. 14.6 Interference of oblique shock and expansion waves
the gas flow parallel to it, similar to the case as discussed in actions (at point C and D) cause more attenuation of shock
Fig. 14.2. However, turning the wall CB through an angle θ wave and thus, it bends further (shock, CD). In addition to
(making the wall CB) with respect to the wall AC, the same attenuation of shock, these interactions when weak in nature,
physical boundary condition is satisfied. In this case, the flow result in partial reflection of the expansion fans along the
downstream of the shock wave will be parallel to wall without Mach Lines of opposite family. These Mach Lines are of neg-
forming the reflected wave CE. This is known as cancelation ligible strength and can be ignored. When the interactions
or neutralization phenomena of an oblique shock wave. of shock and expansion waves are quite strong, the reflected
waves cannot be neglected and there will be multiple slip-lines
downstream. In turn, there will be a whole region of vorticity
14.2.5 Interference of Shock and Expansion or an entropy field exists downstream of the interactions.
Waves
The interaction of an oblique shock wave (originated at A) 14.2.6 Shock–Shock Interference

with an expansion fan (centered at O) is shown in the Fig. 14.6.
When the shock wave intersects the expansion fan (say at Consider Fig. 14.7, which represents the interference of two
B), it gets attenuated. That is, the strength of shock wave oblique shock waves of different strengths. The incident shock
decreases, which leads to change of shock angle and conse- waves AO and BO are intersecting at point O, and are transmit-
quently, the shock starts bending (shock, BC). Further inter- ted into OC and OD, respectively. Due to unequal strengths of
14.2 Supersonic Flow Field Characteristics 365
Incident shock wave

A
D
2
p θ
2 2
M 1 M
M 4
2
p = p
4 4 5
Slip line
p θ O
1 1
θ = θ
4 5
5
1
M
3 M
5
M
1
p θ
3 3 Transmitted shock wave
3
C
B
Fig. 14.7 Shock–shock interference
S2 2 S4 4 Slip line
M1
1
C 5
S1 3 S3
Fig. 14.8 Schematic diagram of shock–shock interference of Type I
incident waves AO and BO, the flow properties in the regions acknowledged as six types, can be interpreted using the dis-
2 and 3 are not the same. However, the mechanical equi- continuity theory of shock waves.
librium at point O requires the static pressures in the regions
4 and 5 must be equal. The strength of transmitted waves
14.2.7.1 Shock–Shock Interference of Type I
OC and OD are adjusted in such a way that the flow across When two oblique shock waves of unequal strength from
them are turned toward the wave with same deflection angle opposite families (S1 and S2 ) intersect each other the shock
(θ4 = θ5 ) and have equal pressure (p4 = p5 ). Also, a stream- interference of Type I is produced (Fig. 14.8). Across the
line commonly called slip-stream, is emanated from point O shock (S1 ) the static pressure increases from p1 to p3 and the
which divides the flow field into two regions having equal flow undergoes an upward deflection of θ1 , whereas the shock
pressure but different fluid properties on either side of it. (S2 ) causes the pressure to jump from p1 to p2 with a down-
ward deflection (θ = θ2 ). Generally, the flows downstream of
the shocks S1 and S2 are not compatible because their pres-
14.2.7 Types of Shock–Shock Interference sures and directions are not same. In order to have a common
(Edney Classification) direction (i.e., θ3 = θ4 ), the flows across these waves must
be deflected and achieved through the transmitted shocks,
The distinctive features of shock-induced separation are the S3 and S4 , emanating from the point of intersection C.
shock patterns that occur in the adjoining inviscid flow as a Across the shocks S3 and S4 , the pressures increase from p3
consequence of the behavior of the boundary layer during the to p5 and from p2 to p4 , respectively. The physical boundary
interaction process. The shock patterns are produced when condition at point C requires the pressure p4 should be equal
two shock waves intersect or interfere with each other. These to p5 . A streamline is emanated from point C across
patterns, classified by Edney into what are now commonly which p4 = p5 but the other fluid properties are
different.
S2 2 S4 4
M=1
M1 C2 Slip line
1 S5 M<1 M>1
C1 Slip line
S1 S3 5
3
Fig. 14.9 Schematic diagram of shock–shock interference of Type II
14.2.7.2 Shock–Shock Interference of Type II Fig. 14.13. The flow field structure is comparatively less com-
The Type II interactions of the obliques shock waves plex than the other shock interference types. The oblique
(S1 and S2 ) essentially reflect the Mach
reflection
phenom- shock waves S1 and S2 intersect at the triple point C from
ena that produces two triple points 2 and 3 separated by which a shock S3 emanates, causing a pressure jump across
a near to normal shock wave (S5 ) (Fig. 14.9). it. The compatibility conditions for change of states require
the formation of expansion waves and a slip-stream in the
14.2.7.3 Shock–Shock Interference of Type III flow field (Fig. 14.13).
Whenever an oblique shock wave (S1 ) intersects a strong bow-
shock (S2 ), a slip-stream is generated which separates the
subsonic flow zone 2 and supersonic flow zone 3 . The slip- 14.3 The SBLIs Phenomena
stream first detaches then reattaches the wall surface. It should
be noted that the Type III interaction is feasible only at smaller The Shock–Boundary Layer Interactions (SBLIs) phenomena
shock angles (β ≤ 20o ) as long as the continuum hypothesis is all about the pressure jump caused by the shock, which is
holds. In this case, two triple points C1 and C2 are produced imposed on the boundary layer and its response. The SBLIs
(Fig. 14.10). occur when a shock wave and a boundary layer converge, and
since both are commonly present in supersonic and hyper-
sonic flows, these interactions are inevitable. Because of its
14.2.7.4 Shock–Shock Interference of Type IV
omnipresent behavior and boundless occurrence at high Mach
In Type IV interaction, up to zone 4, the flow field structure
number flows, it is imperative to have a thorough understand-
is similar to that of Type III interference with the generation
ing of these interactions and their impact on the vehicle’s
of slip-stream. However, in this type of interaction, the shear
performance.
layer does not reattach to the surface again, instead, a super-
The schematic diagram of a typical shock–boundary layer
sonic jet surrounded by the subsonic flow at the boundary,
interaction phenomena occurring in a supersonic compres-
develops. The static pressures in the outer subsonic regions
sion intake is shown in the Fig. 14.14. The wedge produces
are almost constant. Similar to the previous case, the flow field
an oblique shock that impinges upon the boundary layer over
structure have two triple points (C1 and C2 ).
the surface. Due to large pressure jump across the shock, an
In supersonic jet, the oblique shock (S4 ) impinges the
intense adverse pressure gradient is imposed upon the bound-
boundary producing a kink (point O1 ). However, to offset
ary layer, which might lead to thickening, or in the worst
the pressure jump across the shock, a centered expansion
case separation of the boundary layer. The separated bound-
wave must be formed. This expansion ray extends further and
ary layer diverts the external supersonic flow into itself thus,
impinges the opposite boundary of the jet and get reflected
a second oblique shock wave is formed known as induced
as a compression front, which, in turn, get reflected as an
separation shock wave. At some distance downstream, the
expansion wave, and so on (Fig. 14.11).
separated boundary layer turns toward the surface and reat-
taches to it. Consequently, the supersonic flow gets deflected
14.2.7.5 Shock–Shock Interference of Type V into itself, generating a third shock wave referred to as the
This type of interaction occurs when two weak oblique shocks reattachment shock. Moreover, between the points of separa-
of the same family interact above the sonic line (M = 1), pro- tion and reattachment, where the boundary layer turns back
ducing two triple points, C1 and C2 . Instead of a slip-stream, toward the surface, the supersonic flow is turned away from
a supersonic jet originates from the point C1 (Fig. 14.12). itself generating expansion waves.
Due to the local effects caused by SBLIs, the changes in
14.2.7.6 Shock–Shock Interference of Type VI boundary layer take a long time to subside, thereby, making
Type VI interaction occurs when the two weak oblique shock the flow prone to separation occurring farther downstream.
waves of the same family intersect each other as shown in In those cases, where the considerable separations occur,
14.3 The SBLIs Phenomena 367
S2 M<1 2
M1
1 C1 Slip line
M>1
S1 4
3 S3
S4 5
C2
Slip line
S5
Fig. 14.10 Schematic diagram of shock–shock interference of Type III
M<1
S
2
3
M1
1 C1 Slip line O1
7 8
4 S4
S1
S3 6 M<1
2
C2 O
2 5
M>1
S
5
Fig. 14.11 Schematic diagram of shock–shock interference of Type IV
SBLI could lead to significant changes in the shock structure We should note that the interaction of a shock wave with
and overall flow. They can introduce large-scale flow field either a laminar or a turbulent boundary layer is typical to
unsteadiness, inlet buzz, or engines unstart. They have the distinguish. If we consider a Reynolds or Favre averaged tur-
potential to cause intense heating, severe enough to destroy bulent flow in which perturbation components are filtered out,
the complete vehicle. The SBLIs can be witnessed in many then there is no basic difference between the two types of
situations such as in supersonic and hypersonic aircraft and flow with regards to the flow physics and overall topology.
spacecraft including missiles, in propeller and engine turbine The major difference is observed in the velocity profiles in the
blades running at high RPM, Ramjet and Scramjet engine boundary layer. In an interaction, the flow behavior is dictated
intakes, intake isolators, etc. The reverberations of a SBLI by the resistance of the boundary layer to the pressure jump
are numerous and often can be a critical parameter in deter- imparted by the shock wave; therefore, it is obvious that the
mining the performance of a vehicle or a propulsion system. turbulent boundary layer which has a shallow velocity distri-
Sometimes, they may only restraint the operating range of the bution and hence carrying more momentum will react less as
system, even then it can be dreadful. compared to the laminar boundary layer of which the velocity
S3 Slip line
M>1
4 M =1
M1
1 6 Slip line
C1 M<1
M>1
7
S5 5
S1 S4
C2
2
3
S2
Fig. 14.12 Schematic diagram of shock–shock interference of Type V
S3
3
C
5
M1 1
Expansion fans
S1 4
S2
Fig. 14.13 Schematic diagram of shock–shock interference of Type VI
profile is rather steep. It is seen that when a boundary layer is dation caused by the shock wave. The effect is predominant in
submitted to an adverse pressure gradient, the incompressible laminar flow which is viscous dominated. However, the resis-
shape factor2 increases. tant offered by the laminar boundary layer decreases with
In addition to velocity distribution, the shear stress also increase of Reynolds number, in turn, leading to decrease in
plays an important role during the SBLI—it opposes the retar- viscous stresses. Therefore, the laminar flows are less resistant
to the influence of the shock at high Reynolds number than at
2 The boundary layer shape factor is defined as the ratio of displacement lower values. In turbulent flows, the effect of fluid viscosity
thickness (δ ∗ ) to momentum thickness (θ). It reflects about the fullness and thus, shear stresses is less obvious and the interaction of
of the boundary layer profile, where a low value signifies a more fuller the shock wave with the boundary layer is a weak function of
profile. For the boundary layer over a flat plate, H is almost equal to
the Reynolds number.
2.59 for a laminar flow, whereas it lies in the range of 1.3–1.4 for the
turbulent flows.
14.4 The Oswatitsch Relation 369
Induced separation shock
Incident shock Expansion waves
M>1
Separation zone Re−attachment shock

Boundary layer
M<1
Compression waves
M>1
M<1
Sonic line
M=1
Viscous sublayer
Subsonic layer
Fig. 14.14 Schematic diagrams of shock–boundary layer interactions for two different cases
14.4 The Oswatitsch Relation where Ta is the upstream flow temperature, ds is the change
in entropy, and dh0 is the change in stagnation enthalpy.
Oswatitsch (1944) deduced an equation which relates the drag The thrust on the vehicle is produced by the propulsion
on a vehicle with the entropy and the stagnation enthalpy unit, which ejects the total enthalpy into the flow. In Eq. 14.2,
introduced by the vehicle into the flow field. Let us consider the total (or stagnation) enthalpy term on the right-hand side
an elemental mass (dm = ρ− →v .n̂ds), surrounded by a control either compensates or exceeds the entropy term. Obviously, in
surface (CS), is in uniform freestream (va ) and let F be the latter case, the thrust is generated on the vehicle which propels
net force acting on the elemental mass (dm) in the direction it. Further, the drag (D) on the vehicle can be expressed as
of drag (Fig. 14.15). Then, shown in Eq. 14.3.

1 1
F= (Ta ds − dh0 ) dm (14.2) D= (Ta ds) dm (14.3)
va s va s
Control surface (CS)

v
dm
Fig. 14.15 Schematic of an elemental mass in a freestream
Thus, it can be fairly said that the drag and entropy produc- or
tion are closely coupled to each other. Contrarily, for an adi-
abatic flow the energy equation in tensorial notations can be p02 s2 − s1
= exp (14.6)
expressed as p0a R
ds dvi Equation 14.6 reveals that the efficiency or stagnation pressure
ρT = τij (14.4)
dt dxi losses are also related to the entropy production (or viscos-
ity). Thus, it is evident that the pressure losses is an internal
where τij is the stress-tensor. In external flows, such as in case
aerodynamic equivalent to the drag in external flows.
of a freestream past a flat plate the only source of entropy pro-
duction is the shear stress which is, in turn, related with the
fluid viscosity. The effect of viscosity is predominant in the 14.5 Some Important Studies on SBLIs
regions of rapid velocity variations in boundary layers, and
in thin region of shock waves. The drag produced by entropy The research on the intriguing field of shock–boundary layer
generation in the viscous boundary layer is known as frictional interactions has a grand history of seven decades beginning
drag (or skin friction) and the drag arising due to entropy pro- in the early 1940s. Regardless of exceptional development
duction in shock waves is called wave drag. Further, the SBLI in the experimental and the computational facilities, these
augments the entropy production and consequently, the drag interactions and their detrimental effects on the flow field
increases. Therefore, the prime objective of SBLI controls is such as immense localized heating, pressure losses, unsteady
to minimize both frictional as well as wave drag components dynamic loads, etc., are not well understood.
to reduce overall drag on the vehicle. The first case of SBLI was reported in a transonic flow over
In internal aerodynamics problems such in supersonic an airfoil in the studies of Donaldson (1944). They observed
intakes, it is desired to minimize the drop in stagnation pres- the interaction of a normal shock and the incoming boundary
sure occurring due to SBLIs in order to maximize the overall layer over an airfoil at transonic speeds, and highlighted the
pressure recovery in the intake isolator region. The efficiency state of the boundary layer, whether laminar or turbulent that
is defined in terms of ratio of stagnation pressures given by governs the nature of interactions. To explore more on the
Eq. 14.5. SBLI phenomenon, Liepmann (1946), Ackeret et al. (1947)
p02 and Fage and Sargent (1947) tested various basic configura-
η= (14.5)
p0a tions such as flat plate and airfoils in transonic regime. Their
results were found to be in close agreement with the findings
where p02 is the mean stagnation pressure at engine level and of Donaldson (1944). It is seen that the systematic study of
p0a is the incoming freestream stagnation pressure. We know SBLIs at transonic speeds was very challenging due to the
that for the flow of a perfect gas across the shock wave, change existence of only a portion of supersonic regime embedded
in entropy is defined as in the whole transonic flow field. To further understand their
occurrence and the effects of SBLIs in a purely supersonic
p0a
s = R ln flow field was realized to be essential and thus, the researchers
p02 started performing the experiments in supersonic flows.
14.5 Some Important Studies on SBLIs 371
Various basic geometries such as flat plate, flat plate with rise. It can be argued that initiating flow control studies
ramp and step configurations including those of external without fully understanding the underlying flow physics of
shock generators and axisymmetric bodies with collars were the SBLIs will be a hit-or-miss affair and may waste both
investigated. These studies yielded much useful data on the time and resources and lead to unpleasant surprises. Thus,
effects of Mach number, Reynolds number, as well as shock urge to fully understand the SBLIs and their consequences
strength and reinforced the findings of earlier observations. kept increasing with the invention of new computational and
Fage and Sargent (1947) and Barry et al. (1951) studied the experimental methods.
nature of SBLIs over the flat plate in a fully supersonic flow. Holden (1986) reviewed a number of aerothermodynamics
Similar to the interactions in transonic flow, they showed problems associated with high Mach number flows. He stated
that the interactions in supersonic flows are also dominated that the thermal loads on the leading edges of swept wings
by the state of the incoming boundary layer. Liepmann and hypersonic intakes when the boundary layer becomes
et al. (1952) studied the characteristics of the shock wave turbulent may be a key feature in controlling the aerother-
impinging on a flat surface kept in a supersonic flow field mal design. When transition occurs in or near the regions of
with varied Mach numbers between 1.3 and 1.5. They SBLI, these flows are not only difficult to describe theoreti-
compared the characteristics of the reflected shocks in the cally but also sensitive to Mach number, Reynolds number,
laminar and turbulent boundary layers and confirmed that surface blowing, and roughness. In hypersonic flows, regions
the reflected shock pattern depends strongly on the nature of of shock wave and turbulent boundary layer interaction are
the boundary layer. The measurements of relative upstream one of the most stressing aerothermal problems as well as
influence in terms of pressure variation due to the impinging most difficult to compute precisely. At the heart of the prob-
shock were carried out. It was observed that the laminar lems associated with theoretical description of these flows, is
boundary layers have the maximum upstream influence up the description of turbulence development at the base of the
to 50δ, whereas the turbulent boundary layer has only up boundary layer where separation occurs. Also, as the bound-
to 5δ. They also reported that the flow separation is almost ary layer undergoes a radical thinning in a recompression pro-
inevitable in laminar flow, opposite to that of turbulent flows cess, the pressure can rise up to two orders of magnitude over
where no separation was observed. In addition to the flat several boundary layer thicknesses. It is now clear that only
plate, the compression corners were also studied. The similar through carefully conducted experimental studies involving
results were observed in case of a shock originating from the detailed mean and fluctuation measurements in the flow field
compression corner with only exception that the influence of as well as on the surface, can give the insight and detailed
the corner extends far upstream in laminar flow. Donaldson information may be generated to resolve these issues. Accord-
and Lange (1952) studied the pressure rise in the flow field ing to Dolling (2001), the major problem causing effects of
due to SBLIs in an internal flow through an axisymmetric SBLIs are their flow unsteadiness and heat transfer. The loca-
duct at Mach 3.03, which they assumed to be analogous to a tion at which peak heating occurs due to SBLI and its mag-
supersonic diffuser. The diameter of the duct was kept at 12δ nitude are the most vital areas to be focused upon. It is well
and the test conditions were assumed to be essentially same known that SBLIs cause severe heating particularly in hyper-
as of the two dimensional flat plate present in a supersonic sonic flows, with peak rates up to 10–100 times the temper-
flow. The shock is generated by a collar projected into the ature of that of the incoming attached boundary layer flow
flow and thus the shock interactions with both the laminar and many times the equivalent stagnation point value. Knight
and turbulent boundary layers were studied. The test results and Degrez (1998) reviewed the computational capabilities
indicated that the critical pressure rise across a shock wave of the high Mach number flows. They poorly concluded the
caused separation of the boundary layer and was proportional prediction of heat transfer distributions due to SBLIs in a
to the skin friction. For laminar boundary layers, they found numerical supersonic and hypersonic flow field except for
that the critical pressure rise was proportional to the inverse the weak interactions. For strong interactions, an error up to
of the square root of Reynolds number, whereas for turbulent 100% was observed even for the best available turbulence
flows it was the inverse of fifth power of Reynolds number. model as compared to experimental results. For the case of a
They concluded that it is desirable to have turbulent boundary strong SBLI at Mach number greater than 8, the experimental
layer at low Reynolds numbers since they are less prone to results and numerical predictions were compared. Depend-
separation even at higher pressure rise. Thus, an induced ing upon the location, the maximum deviation of the numeri-
transition from laminar to turbulent boundary layer may be cal predictions from the experimental results of heat transfer
useful unless the Reynolds number is very low such that varies between 40 and 150%. They concluded their review
the laminar flow itself tolerates a larger pressure rise. They by stating that, it is mandatory to obtain accurate experimen-
suggested the use of vortex generator to induce transition tal data to develop a reliable turbulence modeling scheme for
by increasing the local skin friction which may be used to SBLIs, especially in predicting the heat flux, the wall pressure,
prevent the shock-induced separation at a given pressure and the heat transfer fluctuations. To understand the physics
behind how and why the peak heating takes place, an exper- bulent separation appears to be inherently unsteady. Fluctu-
imental database of real time with continuous measurements ating pressure loads generated by translating shock waves,
of surface heat transfer is necessary. This knowledge would pulsating separated flows, and expansions/contractions of the
pave way for studies, which focuses on developing methods to global flow field are suggested to be severe enough to cause
reduce peak heating and thermal loads. Lee (1998) identified structural damage and should not to be ignored by design-
that the fluctuations of thermal loads near the reattachment of ers of supersonic and hypersonic vehicles. The importance of
the separated flow under the shock structure contain consider- unsteadiness is compounded by the largest loads, which can
able high-frequency components. His results were supported often occur in regions of high heat transfer. Pozefsky (1989)
by the distribution of properties of the surface temperature identified the locations of maximum loads in a generic hyper-
fluctuation, such as their RMS (Root Mean Square) levels sonic air-breathing vehicle. At all these locations, the flow
and the power spectra obtained for each interaction. Davis field was found to be dominated by SBLIs. Also, in many
and Strurtevant (2000), developed a new scaling parameter to situations, the unsteadiness was the dominant phenomenon
approximately account for wall temperature effects on sepa- and the mean flow properties behavior often could not be
ration length for a laminar nonreacting boundary layer. They understood without the knowledge of unsteadiness. From the
proved a linear relationship between separation length and experiments made in a Mach 3 compression ramp interac-
reattachment pressure ratio and a Reynolds number effect for tion, Bibko et al. (1989) investigated the behavior of wall
transitional interactions. At high enthalpy experiments, they pressure fluctuations and found to be closely related to the
observed a significant increase in scaled separation and this different flow regimes. It was evident that the influence of
increase was found to be attributed to recombination of the Mach number on pressure fluctuations was virtually inde-
free shear layer downstream of separation perhaps altering its pendent of the flow regime and an increase in Mach number
velocity profile. caused a decrease in pressure fluctuations. Dolling (1993)
To study the physical effect of elevated surface tempera- measured the wall pressure fluctuations in a two-dimensional
tures on the flow field, Reinartz et al. (2007), analyzed the SBLIs induced by a compression ramp at a nominal Mach
hypersonic flow over a heated 15o ramp both experimen- 3 freestream. The author detected the presence of large-
tally and computationally. The numerical simulation of the amplitude pressure fluctuations existing throughout the inter-
unsteady compressible Navier Stokes equations were solved action, particularly near the separation and reattachment. In
by the FLOW solver using RANS turbulence model. Differ- the upstream region of flow field, the unsteadiness of sepa-
ent surface temperatures are specified to investigate the impact rated shock wave structure was found to be responsible in gen-
on the SBLI and on the size of the shock-induced separation. erating an intermittent wall pressure signal. In this region, the
They performed computations in laminar as well as in lami- mean wall pressures resulted by the superposition of the rel-
nar and turbulent (transitional) combination. The preliminary atively low-frequency, large-amplitude, shock-wave-induced
three-dimensional simulations emphasized the strong depen- fluctuations on the pressure signal of the undisturbed bound-
dence of separated hypersonic flows over the cell distribution. ary layer. Erengil and Dolling (1993), showed a correlation
In contrast to earlier findings for a double wedge configura- between the wall pressure fluctuations beneath the incoming
tion, they stated that there is no upstream influence on the boundary layer and the separation shock foot velocity from
size of the separation bubble for the combined laminar and which it was inferred that the small scale motion of the shock
turbulent flow computations. However, they observed that the is caused by its response to the passage of turbulent fluc-
introduction of turbulence moves the point of reattachment tuations through the interaction. They also showed that the
upstream and increases the computed levels of peak heating large-scale motion is a result of the shock’s displacement due
and peak pressure. The level of the plateau as well as the peak to the expansion and contraction of the separation bubble.
pressure was found to remain unaffected by an increase in Further, Andreopoulos and Muck (1987) suggested that the
the surface temperature. The authors stated the existence of a frequency of the separation shock motion scales are based on
linear dependence of the size of the separation bubble on the the bursting frequency of the incoming turbulent boundary
ratio of wall temperature to total temperature. layer. These findings were contradicted by the observations
Thus, from the quantum of literature, it can be seen that of Thomas et al. (1995). They proved that no discernible sta-
the thermal loads produced by the SBLIs have drastic detri- tistical relationship exists between burst events and span wise
mental effects in the structure of the vehicle and its perfor- coherent shock front motion. Beresh et al. (2002) conducted
mance. In addition, the necessity to fully understand the heat Particle Image Velocimetry (PIV) studies and high-frequency
transfer effects is of utmost importance in order to control response wall pressure studies to investigate the relationship
them without affecting the performance of the vehicle. In between upstream turbulent boundary layer properties and the
addition to heat transfer losses, the unsteadiness associated unsteady separation shock behavior in a Mach 5 unswept com-
with SBLIs is another undesired phenomenon which causes pression ramp interaction. They suggested that no correlation
severe performance penalties in the flow. Shock-induced tur- was found between variations in the incoming boundary layer
14.5 Some Important Studies on SBLIs 373
thickness and the separation shock foot position, as has been generally, shock wave and shear layer interactions will
suggested in earlier work. A thorough understanding of the have entirely negative consequences. The increase in the
unsteadiness and wall pressure fluctuations is mandatory to fluctuation level they cause can be used to enhance fuel–air
initiate the process of SBLI controls. Korkegi (1975), reported mixing in ramjet combustion chambers or to accelerate the
that the three-dimensional swept shock wave and turbulent disorganization of hazardous flows such as wing trailing
boundary layer interaction always leads to separation and pos- vortices. Also, because the interactions in which separation
sible flow breakdown in rectangular diffusers and inlets. Also, occurs can lead to smearing or splitting of the shock system,
the critical pressure ratios across the shock are found to be the phenomenon can be used to decrease the wave drag
well below the incipient values in two-dimensional cases. associated with the shock. Thus, an optimized shock and/or
The above discussion provided a basic insight to the reader boundary layer controlled configurations must be tested
to understand the key concepts of the SBLIs. Apart from the experimentally on whether the interactions have detrimental
two-dimensional interactions, the requirement to gain insight or positive influences.
into the complex three-dimensional interaction characteris-
tics is increasing more than ever due to the advancements in
aircraft and missiles, where the obvious presence of three- 14.6.1 Classification of Control Techniques
dimensional SBLIs are witnessed. It is understood by con-
sidering a number of fundamental geometries based on the The SBLI control techniques are classified into passive and
shape of the shock wave generator namely, sharp unswept active types. In the former, an attempt is made to produce a
and swept fins, semi-cones, swept compression ramps, blunt beneficial result without the expenditure of externally sup-
fins, and double sharp unswept fins. Here, the shock wave gen- plied energy and includes well-established methods as the
erators has an overall size sufficiently large compared to the ubiquitous vortex generator, riblets, as well as the venting
boundary layer thickness, so that any further increase in size methods. Active control on the other hand is categorized as
does not affect the flow. In addition, numerous experimen- being either predetermined or interactive. In the former case,
tal investigations yielded detailed descriptions of the three- steady or unsteady energy inputs are made through some form
dimensional flow field structure for several canonical config- of actuator irrespective of the state of the flow field. In con-
urations. Despite the knowledge gained from the canonical trast, in an interactive method the power supplied to the actu-
configurations, their usefulness in predicting the flow field ator is varied continually, depending on input from a sensor
structure particularly the location and extent of separation for or sensors.
realistic flight vehicle configurations, is limited by the sheer Supersonic intake is one area where the payoff could be
complexity of typical shock wave interactions in three dimen- very large due to SBLIs. Currently, bleed is used to improve
sions. pressure recovery and mitigate flow distortion by reduc-
ing separation and unsteadiness. However, the inlet must be
increased in size to account for the bleed mass flow losses.
14.6 SBLIs Control
14.6.1.1 Active Control Techniques
The detrimental effects of SBLIs discussed in Sect. 14.5 are Viswanath (1988) reviewed the importance of active control
very much undesired since they have drastic effects on the techniques such as suction and bleed and their role in control-
flow field starting from flow separation and unsteadiness to ling the SBLIs in supersonic and hypersonic flows. Schulte
severe localized heating, which might lead to the failure of et al. (2001) explored and demonstrated the possibilities of
the whole system. Therefore, the consequences of the SBLIs manipulating the SBLIs with regard to an application for
occurrence almost invariably are detrimental in some respect hypersonic intake. Experimental and numerical investigations
which affects the efficiency and performance of the vehicle. of the active control of SBLIs in laminar region was studied
Thus, the need arises for controlling the phenomenon by some in a Mach 6 inlet. In SBLI, we know that the implementa-
appropriate methods which modify the flow, either before or tion of bleed leads to a reduction of the separation bubble
during the interaction process. thickness by almost 50%. They proposed that the achievable
The target of control techniques are mainly to either reduction of the heat loads on the wall surface is dominated by
prevent shock-induced separation or stabilize the shock the amount and the position of the boundary layer bleed. They
when it occurred in naturally unsteady configurations. demonstrated that the application of a optimally designed and
The upstream influence of the shock and the resistance positioned bleed system can remarkably increase the attain-
of a boundary layer to separate depend mainly on the able total pressure recovery.
momentum prior to interaction with the shock, which can Mitani et al. (2005) extended the boundary layer bleed and
be modified by appropriate boundary layer manipulation a two-staged injection of fuel for the application of Mach 4
techniques. It is not inevitable that the SBLIs or more and Mach 6 scramjet engines. The bleed system consisting of
a porous plate, an air cooler and a metering orifice with an upon which type of physical mechanism is addressed, appro-
on-and-off valve, was developed. For a Mach 4 engine, while priate control method is used to manipulate the overall flow
limiting the fuel rate, a bleed of 3% of the captured air doubled structure. Here, we should note that any modification in the
the operating range and the thrust. The engine operating range structure of a shock wave has an effect on the boundary layer
was found to be extended from 0.3 to 0.95. Similarly, a and vice versa. However, both control methods are found to
bleed of 6% of the captured air extended the start limit of the be effective in reducing the shock wave unsteadiness.
Mach 6 engine from 0.48 to 1.0, in addition to 1.5 times Many studies have successfully demonstrated the shock
increase in the thrust. controlling bump concept for transonic wings. A well-
Souverein and Debiève (2010) experimentally investigated designed bump is capable of generating a sizeable λ−shock
the effect of upstream injection by means of continuous air region while incurring only modest viscous flow penalties
jet vortex generators in controlling the SBLIs. The base- and it is well established that the flow behind a control
line interaction is of the impinging type, with a flow deflec- generated λ−shock exhibits significant reduction in pressure
tion angle of 9.5 at Mach 2.3. From the Stereoscopic Parti- losses as well. In a successful control bump design, the
cle Image Velocimetry (SPIV) and Hot Wire Anemometry boundary layer health behind the bump is comparable to
(HWA) investigations to study the unsteady characteristics of that observed without it. The surface bumps are also found
the reflected shock, it was shown that the vortex generators in delaying shock-induced buffet on transonic wings. As an
cause a reduction in length and height of the separation bub- alternative technique which has been extensively studied as
ble. However, the use of boundary layer bleed in a hypersonic shock control is the deployment of surface cavity. In this
inlet in controlling the SBLIs strongly depends on the purpose method, a shallow cavity is placed underneath the shock
of inlet. This is because a gain in total pressure recovery is location which alters the shock strength. The cavity increases
often combined with a marginal loss in the inlet mass, conse- the size of smeared region and expands the flow domain and
quently, reducing the thrust. Therefore, a new active control consequently, the losses are reduced. Instead of attempting
method was studied by Webb (2009), where they investigated to generate a large smeared shock foot with continuous
the ability of Localized Arc Filament Plasma Actuators to compression waves, it is generally easier to aim for a large
eliminate or reduce the extent of boundary layer separation. λ−structure, replacing the shock foot with an oblique shock
By varying the frequency, location, and the mode of actua- followed by a near normal shock wave.
tion, they quantified the effects of the actuators. The effec- The control of SBLIs by controlling the boundary layer is
tiveness of actuators were evaluated using the Schlieren flow the principle on which vortex generators operate. The bound-
visualization technique, the Particle Image Velocimetry (PIV) ary layer control alters the characteristics of the near wall
and unsteady pressure measurements. It was observed that flow field ahead of an SBLI to prevent or reduce the shock-
the actuators were capable of energizing the boundary layers induced separations. This approach is adapted to minimize
effectively in reducing the separation caused by the shock and the viscous drag and to delay the emergence of unsteady
boundary layer interactions. flow such as shock-induced buffet. The Micro-Vortex Gen-
erators (MVGs) are used in re-energizing the boundary layer
14.6.1.2 Passive Control Techniques with minimal parasitic drag. They are found to be effi-
To extend the range of operation of supersonic and hyper- cient in reducing the shock-induced separation. Holden and
sonic vehicles without using the complex feedback systems, Babinsky (2007) experimentally investigated the effect of
it is essential to employ passive techniques in controlling MVGs placed upstream the location where a normal shock
the Shock–Boundary Layer Interactions (SBLIs). Kaufman interacts with the turbulent boundary layer. Two configu-
et al. (1962) reviewed the use of flaps and fins as controls in rations were analyzed: the wedge shaped and the counter-
hypersonic vehicles to control these interactions and its unde- rotating vane type. Although, both vortex generators were
sirable effects on the performance of the vehicle. In recent quite effective in reducing the shock-induced separation, the
years, many passive control techniques are developed and vane-type configuration was found to be the most efficient.
employed to serve the purpose as SBLI controls: fins, cavi- However, both vortex generators are found to have some unde-
ties, and ramps. The passive control of SBLIs can be achieved sirable effects causing a complex wave pattern consisting of a
either by controlling the shock (Fig. 14.16a) or by control- series of weak shocks and expansion waves which increases
ling the boundary layer (Fig. 14.16b). The commonly used the total pressure losses. Babinsky (2007) investigated the
shock controls are contoured surface bump, cavity covered micro-ramp MVGs as the control devices in a Mach 2.5 intake.
with porous surface, etc., whereas the micro-vortex genera- By varying the height of the ramp, they found that the largest
tor is the demonstrated boundary layer control. The shock ramp size showed the strongest effect, however, it also had the
controls aim to change the shock structure to reduce stagna- greatest momentum deficit. This study was further extended
tion pressure loss across it, while the boundary layer control is by Babinsky et al. (2009) to evaluate the efficacy of these
employed to reduce viscous and separation losses. Depending controls on the flow characteristics and to prevent/delay the
14.6 SBLIs Control 375
Control
(a) Uncontrolled SBLI (b) Controlled SBLI
α
W
Fig. 14.16 Schematic layouts of some SBLI controls
separation in the region of reflected shock and boundary layer Each micro-ramp was found to produce counter-rotating lon-
interactions. By varying the height of the ramp, they observed gitudinal streamwise vortex pairs which promoted the mixing
the general flow features are in scale with the device height. and consequently, re-energized the boundary layer. Further,
Downstream of the micro-ramps, a relatively complex struc- the double row configuration was found to be effective in
ture of multiple pairs of counter-rotating streamwise vortices decreasing the probability of reversed flow and the flow sep-
were observed. Behind the ramp, a significant low momentum aration as compared to the single row configuration. How-
zone was also observed which leads to the viscous drag. The ever, both micro-ramp configurations were efficient in stabi-
extent and magnitude of this zone were found to be increas- lizing the shock motion near the interaction zone. In addition
ing with the size of device. They suggested the placement to these experimental investigations, to evaluate the efficacy
of smaller ramps in the regions of expected adverse pressure of MVGs some Computational Fluid Dynamics (CFD) stud-
gradients, which might delay the flow separation and reduce ies were also performed. Lee (2009) performed Large Eddy
the viscous drag. The extension of this work was carried out Simulation (LES) to investigate the effects of various vor-
by Blinde et al. (2009). They experimentally investigated an tex generator configurations such as micro-ramp, split-ramp,
array of micro-ramps arranged in a single and the double rows. and ramped-vane on the interactions of normal shock with the
boundary layer. The ramped-vane configuration has increased a compressor, reservoir tank, cooling and drier units, and an
the vorticity content and it has significantly reduced the sep- air delivery system. The compressed and dried air from the
aration length downstream as compared to other two vortex storage tank is supplied into the settling chamber through the
generators. Zhang et al. (2015) computationally investigated gate valve. The settling chamber is provided with both pres-
the potential of the swept micro-ramps in controlling the inter- sure and temperature ports (or taps) for readings. The settling
actions occurring at Mach 3.5 in a finite duct. By the pre- chamber total pressure (P0 ), which is a controlling parame-
compression, dividing, obstructing and energizing effects, a ter in the experiment, is maintained constant during the test
highly swept micro-ramps of the height equal to 0.24δ, was run by controlling the Pressure Regulating Valve (PRV). The
found to be efficient in controlling the SBLIs, where δ is the temperature of settling chamber is the same as that of the
boundary layer thickness. The swept micro-ramp’s effective- ambience.
ness was also authenticated for different shock impingement From settling chamber, the flow is fetched into the test
positions. The results suggested that the separation can be section, where the model is mounted using a cantilever mech-
well controlled when the shock impinges in the rear of the anism, located at the end of the test section. From Eq. 14.7,
controls. the test section Mach number (MTS ) is calibrated using the
static pressure (P) measured at ports mounted on the wall
using mechanical or electrical transducers. These transduc-
14.7 Experimental Techniques for Studying ers (generally, Piezoresistive transducers) experience change
the SBLIs in electrical resistivity when there is a change in mechanical
strain due to pressure. A Wheatstone bridge (circuit) converts
In many instances, the interactions of shock wave with bound- the resistance into output voltage. In turn, the static pressure
ary layer may be computed effectively using modern CFD values are obtained from the output voltage by means of spe-
techniques. But, these methods are prone to errors when the cific conversion factor of each transducer.
boundary layer separates. Thus, the extensive experiments
γ−1
need to be performed to understand the underlying principle P0 γ−1 2 γ
of interactions and their control. The major parameters con- = 1+ MTS (14.7)
P 2
sidered for analyzing these flows are the variation of total
pressure at the centerline and the wall surface pressure distri- The qualitative investigation of wave strength and the flow
bution. The total pressure distribution is measured using pitot structure are carried out using Schlieren or Shadowgraph
probe along the centerline, whereas the wall static pressure optical flow visualization techniques. The schematic lay-
is obtained through the static pressure ports mounted on the out of a typical Z-type Schlieren visualization setup with
wall. monochromatic light source and receptive mirrors, is shown
In subsonic flows, the measured total pressure can be in Fig. 14.18. In order to minimize the positioning errors of
converted into Mach number using pressure–Mach number the mirrors and the lenses, Eqs. 14.8 and 14.9, respectively,
relation, because the subsonic flows are always correctly are used.
expanded. However, in supersonic flows the pitot probe mea-
sures the pressure behind the bow-shock positioned at the 1 1 1
= + (14.8)
nose. Also, because of the complex wave pattern prevailing fi ui vi
in the flow field, the static pressure varies from point to point. 1 1 1
= − (14.9)
Thus, an accurate measurement of static pressure at these fi vi ui
points is almost impossible. Therefore, the researchers have
used the measured pressure data as it is to predict the interac-
tion characteristics.
In addition to pressure measurement, the wave strength 14.8 Evaluation of Effectiveness of Some
and boundary layer characteristics can be qualitatively stud- Specific Passive Controls
ied using the optical flow visualization techniques such as
Schlieren or Shadowgraph. These visualization images can From the above discussions on SBLIs and the survey of some
confirm the strength of waves and overall flow structure pre- major findings of the effectiveness of shock and boundary
dicted by the pressure plot. layer controls on weakening the shocks and delaying the
The shock wave and boundary layer interaction studies boundary layer separation, it is evident that the right geome-
are generally performed in the test section of a high-speed try of passive controls plays the vital role in controlling these
wind tunnel. The typical layout of a supersonic blowdown interactions. The potential benefits of flow control depends
wind tunnel facility is shown in Fig. 14.17. The experimen- on the application with an objective to obtain increased effi-
tal facility contains an air supply system which consists of ciency or improved performance of a vehicle or a component
14.8 Evaluation of Effectiveness of Some Specific Passive Controls 377
Driver unit
Reciprocating
compressor
Storage tanks
Induction
motor Water cooling Test section
unit
Diffuser
Gate valve
Pressure regulating
valve
Settling chamber
Fig. 14.17 Schematic layout of an open circuit supersonic wind tunnel
Monochromatic light source
Test−section f
1
Mirror 2 f2 Mirror 1
Knife edge
Mirror 3
f3
High speed camera Biconvex
lens
Fig. 14.18 Schematic diagram of Z-type Schlieren optical flow visualization technique
at a reduced cost. It is well established that any change in the 14.8.1 SBLI Control Using Cavity Covered with
structure of a shock using shock control affects the bound- Porous Surface
ary layer characteristics and vice versa. Therefore, these two
aspects—modifying of the shock structure without control- The high-speed intakes are places where the shocks are inten-
ling the boundary layer, and shedding the mixing promoting tionally generated to compress the air before entering to
the vortices within the boundary layer without controlling the combustion chamber (Fig. 14.19). However, higher shock
the shock—were studied as two specific investigations with strength and uncontrolled interactions generate large drag and
cavity covered with porous surface, and with ramped-vane in turn, decreases the propulsive efficiency. Further, these
micro-vortex generators located in a Mach 2.2 intake. The interactions lead to significant changes in the shock struc-
major findings of these two experimental studies are discussed ture causing smearing or splitting of the shock system. Thus,
below.
Cowl
Forebody
Shock
SBLIs
Fig. 14.19 Schematic diagram of a typical high-speed intake
the control of shock–boundary layer interactions is of utmost of the pores, porosity of the surface is varied as 10, 15.7 and
necessity. The potential benefits of flow control depend on the 22.5% as shown in Table 14.1.
application with an objective either to obtain increased effi- It is also known that the placement of cavity at the shock
ciency or improved performance of a vehicle at reduced cost. impact point attenuates the wave strength considerably, creat-
Controlling the flow is the process of manipulation or mod- ing a recirculation zone which can be modified by varying the
ification of the flow, either before or during the interaction contraction ratio (Rc ) , which plays a vital role in determining
process. The target of control techniques is mainly to either the characteristics of any internal flow. The contraction ratio
prevent shock-induced separation or stabilize the shock when of the cavity is defined as
it occurred naturally.
In addition to various other control techniques, the cavity 2 γ+1
Ac 1 2 γ−1 2 γ−1
with porous wall beneath the shock is established as an effi- Rc = = 2 1+ Ma
At Ma γ+1 2
cient shock control in the literature. The application of cavity (14.11)
covered with porous surface was used to control the interac-
tions in a supersonic stream. Shallow cavities of aspect ratio Moreover, the limiting contraction ratio for the intake, in terms
more than 2.0, covered with porous surface are proved to be of freestream Mach number (Ma ) is given by
the efficient controls. Keeping this in mind, Humrutha et al.
(2017) evaluated the efficacy of cavity of aspect ratio 2.0 in 1 1
At γ−1 2 2 2γ γ−1 γ−1
controlling the Mach 2.2 supersonic intake. The mixed com- = + −
Ac γ + 1 (γ + 1) Ma2 γ + 1 (γ + 1) Ma2
pression intake, where the oblique shock waves were gener-
(14.12)
ated internally as well as externally to compress the flow, was
designed to decrease the total pressure losses. The maximum
The efficacy of cavities in controlling the interactions are
pressure recovery in a supersonic intake is calculated using
tested at varied contraction ratios by varying the throat height
the Oswatitsch (1944) condition given by Eq. 14.10.
between the core and cowl. Thus, in addition to design con-
traction ratio (Rc = 1.14), the experiments were also per-
Ma sin β1 = M1 sin β2 = M2 sin β3 (14.10)
formed at off-design ratios (Rc = 1.19 and Rc = 1.25) for
comparison.
In the uncontrolled intake, a shock generated by cowl lip
impinges at Lx = 0.6, where x is axial distance measured from
the leading edge of the core, and L is the intake length. The
14.8.1.1 Wall Static Pressure Variation
schematic diagram of a typical uncontrolled intake is shown that the variation of nondimensional
It is well established
static pressure PP0 with the nondimensional axial distance
in Fig. 14.20. Since the cavities must be placed in the region of x
adverse pressure gradient, a shallow cavity of aspect ratio 2.0 L is good enough in predicting the control effectiveness.
covered with a thin porous surface is deployed at the shock The measured pressure (ps ) was made nondimensional with
impact point (Fig. 14.21). By varying the diameter and pitch the settling chamber pressure (p0 ) and the axial distance
Cowl
At
y
Ac
Core
x
O
L
Fig. 14.20 Schematic diagram of an uncontrolled intake
Fig. 14.21 Photographic view of cavity covered with porous surface controlled intake
Table 14.1 Percentage variation of porosity with diameter and pitch

Cases Diameter (mm) Pitch (mm) Porosity (%)
1 0.5 1.5 10.0
2 1.0 2.5 15.7
3 0.75 1.5 22.5
0 0.13 0.27 0.48 0.7 0.83 1

x/L
Fig. 14.22 The wall-mounted pressure ports along the intake length
(x), measured from the leading edge was made nondimen- ied by measuring the pressures at Lx = 0.13 and Lx = 0.27,
sional with the intake length (L), for all the intake config- downstream of the waves and to investigate the influence of
urations. The investigations were carried out at the design the cavity covered with porous surface in near field loca-
contraction ratio (Rc = 1.14). The measurements were also tions, the third and fourth pressure ports were specifically
performed at two other off-design contraction ratios: 1.25 and constructed at Lx = 0.48 and Lx = 0.7, upstream and down-
1.19 for comparison, however, at these contraction ratios the stream of the cavity, respectively. In addition, the effect of
intake did not start because of shock-induced spillage. The interactions in far field locations was observed by mount-
static pressures were measured at five different axial loca- ing the pressure port at 0.83L. The geometrical blockage
tions; 0.13L, 0.27L, 0.48L, 0.7L, and 0.83L (Fig. 14.22). The is kept constant for both uncontrolled and cavity-controlled
effects of first two ramp-generated shock waves were stud- cases.
Listing 14.1 The MATLAB program to generate Fig. 14.23.

1 clc ;
2 c l e a r all ;
3 str1 = ’ H :\ P r e s s u r e \ ’;
4 str2 = ’ H :\ P r e s s u r e \ Plot \ ’;
5 f i l e _ p a t h 1 = ’ Plain ’;
6 f i l e _ p a t h 2 = ’ C100 ’;
7 f i l e _ p a t h 3 = ’ C157 ’;
8 f i l e _ p a t h 4 = ’ C225 ’;
9 str3 = s t r v c a t ( f i l e _ p a t h 1 , f i l e _ p a t h 2 , f i l e _ p a t h 3 , f i l e _ p a t h 4 );
10 str5 = ’ T =10 mm ’;
11 str6 = ’ Text ’;
12 str7 = ’ XvsP ’;
13 str8 = ’ 001 ’;
14 str9 = ’. txt ’;
15 for j =[1 2 3 4]
16 f i l e n a m e = s t r c a t ( str1 , str3 (j ,:) , ’\ ’ , str5 , ’\ ’ ,
17 str6 , ’\ ’ , str7 , str8 , str9 );
18 p l o t _ f i g n a m e = s t r c a t ( str2 , ’\ ’ , str5 , ’\ ’ , str7 , str8 );
19 M = load ( f i l e n a m e );
20 l e n g t h ( M );
21 X = M (: ,1);
22 P = M (: ,2);
23 if j ==1
24 plot (X ,P , ’ k * - ’ , ’ M a r k e r S i z e ’ ,3 , ’ LineWidth ’ ,0.8);
25 hold on
26 grid on
27 e l s e i f j ==2
28 plot (X ,P , ’ ks - - ’ , ’ M a r k e r S i z e ’ ,3 , ’ M a r k e r F a c e C o l o r ’ ,[0 0 0] ,
29 ’ LineWidth ’ ,0.8);
30 e l s e i f j ==3
31 plot (X ,P , ’ ko - ’ , ’ M a r k e r S i z e ’ ,3 , ’ LineWidth ’ ,0.8);
32 e l s e i f j ==4
33 plot (X ,P , ’ kd - - ’ , ’ M a r k e r S i z e ’ ,3 , ’ M a r k e r F a c e C o l o r ’ ,[0 0 0] ,
34 ’ LineWidth ’ ,0.8);
35 end
36 end
37 x l a b e l ( ’ x / L ’ , ’ FontSize ’ ,13 , ’ FontName ’ , ’ Times ’ ,
38 ’ FontAngle ’ , ’ italic ’);
39 y l a b e l ( ’ P / P_0 ’ , ’ FontSize ’ ,13 , ’ FontName ’ , ’ Times ’ ,
41 a x i s ([0 1 0 0 . 5 ] ) ; set ( gca , ’ XTick ’ ,0:0.2:1 , ’ FontSize ’ ,9 ,
42 ’ FontName ’ , ’ Times ’);
43 set ( gca , ’ YTick ’ , 0 : 0 . 1 : 0 . 5 ) ;
44 l e g e n d ( ’ U n c o n t r o l l e d ’ , ’10.0% Porosity ’ , ’15.7% Porosity ’ , ’22.5% Porosity ’ ,
45 ’ Location ’ , ’ northwest ’);
46 set ( gca , ’ X M i n o r T i c k ’ , ’ on ’);
47 set ( gca , ’ Y M i n o r T i c k ’ , ’ on ’);
48 grid on ;
52 x S i z e = 7;
53 ySize = 3.5;
54 xLeft = (7 - xSize )/2;
55 yTop = (3.5 - ySize )/2;
57 s a v e a s ( gcf , p l o t _ f i g n a m e , ’ png ’);
58 hold off ;
At design contraction ratio (Rc = 1.14), the static pres- is shown in Fig. 14.26a. It is clearly visible that, the oblique
sure variation for the Mach 2.2 intake is shown in Fig. 14.23. shock generated due to cowl lip impinges at the shoulder of the
In both uncontrolled and controlled intakes, the equal rise in second ramp and, in turn, the boundary layer thickens. Also,
pressures have been observed for the flow across first and sec- the faint black line appearing near to the wall of uncontrolled
ond oblique shock waves, whereas the pressure remains con- intake core is the edge of the boundary layer. It should be noted
stant between second and third ports for the intakes controlled that by providing porosity to the surface the boundary layer
with cavities having 10 and 22.5% porosities. A significant thickness is reduced and the shock structure is modified in all
pressure drop is observed at Lx = 0.48 in the case of cav- the controlled configurations. However, the cavity with 15.7%
ity covered with 15.7% porous surface showing the upstream porosity decreased the thickness substantially (Fig. 14.26c).
influence of the cavity. Further, a similar trend is observed The above study revealed that the intake controlled by cav-
between the third and fourth ports where the cavity covered ity with 15.7% porous surface is found to be efficient in con-
with 15.7% porous surface decreases the wall pressure signif- trolling the shock-boundary layer interactions. It’s optimal
icantly. This maybe because of higher suction effect created performance may be because of higher boundary layer suc-
by the cavity covered with 15.7% porous. As we go far down- tion effect especially in near upstream and downstream of the
stream location Lx = 0.83, the wall pressure is increased in all interaction region.
the controlled configurations except the 15.7% porosity case
where a marginal decrease in pressure is observed. Further, the
variation of wall pressure with 10% surface porosity is almost 14.8.2 SBLI Control Using Ramped-Vane
similar to the uncontrolled intake, showing the inferior perfor- Micro-Vortex Generators
mance due to relatively lesser suction effect. Furthermore, for
the highest surface porosity of 22.5% studied in the present In addition to shock control, the SBLIs can also be controlled
investigation, the advantage of the larger boundary layer suc- by suitable manipulation of boundary layer characteristics. In
tion effect was nullified by the increase in pressure which this investigation, an attempt was made to control the inter-
maybe because of large number of weaker waves produced actions by shedding the vortices at the shock impact point in
due to increased porosity. a Mach 2.2 mixed compression intake. Two types of Micro-
The pressure variations at off-design contraction ratios Vortex Generators (MVGs); a conventional and an innovative
equal to 1.19 and 1.25 are shown in Figs. 14.24 and 14.25, ramped-vane type were experimentally studied. The conven-
respectively. Unlike the Rc = 1.14 case, large pressure ampli- tional vortex generators of optimum dimensions were repro-
tudes are found at the second pressure port for all the config- duced using the results of Anderson et al. (2006) as shown in
urations. Interestingly, the cavity with 15.7% surface poros- Fig. 14.27a. Also, an innovative design referred to as ramped-
ity vane MVG, which combines the benefits of both ramp- and
x is found to increase the pressure in near field location
L = 0.48 and it has maintained its superiority between third vane-type configurations was also investigated (Fig. 14.27b).
and fourth pressure ports. The figures reveal that at Rc = 1.19, The shock generated by the cowl lip impinges upon the bound-
the λ−shocks are formed at the entry to isolator region of both ary layer on intake core at Lx = 0.6. Therefore, the controls
uncontrolled intake and the intake controlled by cavity with are provided before the shock impact point, upstream to the
10% porosity. Even though, the λ−shock formation did not region of adverse pressure gradient.
occur in 22.5% porosity case, but, a significant pressure loss The heights of both conventional and ramped-vane MVGs
is observed due to larger number of shocks produced at pores. are varied as 200, 400, and 600 µm. The dimensions of MVGs
Once again, the cavity controlled with 15.7% porous surface are given in Table 14.2 and the schematic diagram of a typical
has maintained its superiority by eliminating the formation of ramped-vane MVG is shown in Fig. 14.28. Unlike the previ-
λ−shock and minimizing the pressure losses. ous study on cavity with porous surface controlled SBLIs,
From the above discussion on wall pressure variation, it this investigation was carried out only at the design contrac-
is evident that the cavity covered with 15.7% porous surface tion ratio (Rc = 1.14) of Mach 2.2 intake.
produced larger boundary layer suction in near field locations, The wall static pressures (P) were measured by piezoresis-
consequently, improving its controlling efficiency. tive transducers. Further, the qualitative investigation of the
flow development and shock structure were carried out using
14.8.1.2 Flow Visualization the Schlieren technique.
The Schlieren optical flow visualization images of uncon-
trolled and controlled Mach 2.2 intakes at the design contrac-
14.8.2.1 Wall Static Pressure Variation
tion ratio (Rc = 1.14), are shown in Fig. 14.26. However, for
For both conventional and ramped-vane
MVGs, the variations
the sake of brevity the visualization pictures of the intakes
of nondimensional static pressure PP0 with nondimensional
operating under off-design conditions will not be discussed.
The flow visualization picture of uncontrolled Mach 2.2 intake axial distances Lx are shown in Fig. 14.30. The pressure
Fig. 14.23 Variation of wall static pressure with axial distance at Rc = 1.14
Fig. 14.26 Schlieren images of uncontrolled and controlled Mach 2.2 intakes at Rc = 1.14
Fig. 14.27 Photographic views of the controlled mixed compression Mach 2.2 intake
Table 14.2 Dimensions of the MVGs

Half angle α 24o
Side length C 7.2H
Span-wise pitch S 7.5H
α
W
α
α
Fig. 14.28 Schematic diagram of the ramped-vane MVGs

Listing 14.2 The MATLAB program to generate Fig. 14.30.

1 clc ;
2 c l e a r all ;
3 str1 = ’ H :\ P r e s s u r e \ C R 1 1 4 \ C o n t r o l E f f e c t \ ’;
4 str2 = ’ H :\ P r e s s u r e \ C R 1 1 4 \ C o n t r o l E f f e c t ’;
5 f i l e _ p a t h 1 = ’ Plain ’;
6 f i l e _ p a t h 2 = ’ C R 1 1 4 _ M V G b _ H 2 0 0 ’;
9 str3 = s t r v c a t ( f i l e _ p a t h 1 , f i l e _ p a t h 2 , f i l e _ p a t h 3 , f i l e _ p a t h 4 );
10 str7 = ’ XvsP ’;
11 str8 = ’ 001 ’;
12 str9 = ’. txt ’;
13 for j =[1 2 3 4]
14 f i l e n a m e = s t r c a t ( str1 , str3 (j ,:) , ’\ ’ , str7 , str8 , str9 );
15 p l o t _ f i g n a m e = s t r c a t ( str2 , ’\ ’ , str7 , str8 );
16 M = load ( f i l e n a m e );
17 l e n g t h ( M );
18 X = M (: ,1);
19 P = M (: ,2);
20 if j ==1
21 plot (X ,P , ’ k * - ’ , ’ M a r k e r S i z e ’ ,3 , ’ LineWidth ’ ,0.8);
22 hold on
23 grid on
24 e l s e i f j ==2
25 plot (X ,P , ’ ks - - ’ , ’ M a r k e r S i z e ’ ,3 , ’ M a r k e r F a c e C o l o r ’ ,
26 [0 0 0] , ’ LineWidth ’ ,0.8);
27 e l s e i f j ==3
28 plot (X ,P , ’ ko - ’ , ’ M a r k e r S i z e ’ ,3 , ’ LineWidth ’ ,0.8);
29 e l s e i f j ==4
30 plot (X ,P , ’ kd - - ’ , ’ M a r k e r S i z e ’ ,3 , ’ M a r k e r F a c e C o l o r ’ ,
31 [0 0 0] , ’ LineWidth ’ ,0.8);
32 end
33 end
34 x l a b e l ( ’ x / L ’ , ’ FontSize ’ ,13 , ’ FontName ’ , ’ Times ’ ,
36 y l a b e l ( ’ P / P_0 ’ , ’ FontSize ’ ,13 , ’ FontName ’ , ’ Times ’ ,
37 ’ FontAngle ’ , ’ italic ’); axis ([0 1 0 0 . 5 ] ) ;
38 set ( gca , ’ XTick ’ ,0:0.2:1 , ’ FontSize ’ ,9 , ’ FontName ’ , ’ Times ’);
39 set ( gca , ’ YTick ’ , 0 : 0 . 1 : 0 . 5 ) ;
40 l e g e n d ( ’ U n c o n t r o l l e d ’ , ’200\ mum i n n o v a t i v e MVGs ’ ,
41 ’400\ mum i n n o v a t i v e MVGs ’ , ’600\ mum i n n o v a t i v e MVGs ’);
42 set ( gca , ’ X M i n o r T i c k ’ , ’ on ’);
43 set ( gca , ’ Y M i n o r T i c k ’ , ’ on ’);
44 grid on ;
48 x S i z e = 7;
49 ySize = 3.5;
50 xLeft = (7 - xSize )/2;
51 yTop = (3.5 - ySize )/2;
53 s a v e a s ( gcf , p l o t _ f i g n a m e , ’ png ’);
54 hold off ;
variation for the uncontrolled intake is also plotted for controlled with MVGs of height 200 µm. Their distinguished
comparison. The experiments were performed only at the effect owe to much efficient flow mixing caused due to shed-
design contraction ratio (Rc = 1.14). The static pressures ding of comparatively smaller size counter-rotating vortices.
were measured at five ports mounted at the locations: 0.13L,
0.27L, 0.48L, 0.7L and 0.83L shown in the Fig. 14.29. The 14.8.2.2 Flow Visualization
effects of first two ramp-generated shock waves were studied The Schlieren flow visualization images for both uncontrolled
by measuring the pressures at Lx = 0.13 and Lx = 0.27, and controlled Mach 2.2 intakes at the design contraction
downstream of the waves and to investigate the influence ratio (Rc = 1.14) are shown in Fig. 14.31. The oblique shock
of the cavity covered with porous surface in near field wave generated by the cowl lip impinges at the shoulder of
locations, the third and fourth pressure ports were specif- the second ramp and thus, thickens the boundary layer. In
ically constructed at Lx = 0.48 and Lx = 0.7, upstream and Fig. 14.31a, near the wall of the intake core, a faint black
downstream of the cavity, respectively. The far field effects line shows the boundary layer edge. The shock cell struc-
were studied by measuring the pressure at the port mounted ture in the intake isolator region is found to be similar in
at 0.83L. For all the uncontrolled and MVG controlled both conventional and ramped-vane MVG controlled intakes
intake configurations, the blockage was essentially constant. with a λ−shock formation at the cowl lip, however, the shock
The static pressure variation in the cowl of the intake strength is drastically reduced for ramped-vane MVGs. It is
controlled with MVGs of height 600 µm is shown in also observed that the ramped-vane type, cause a significant
Fig. 14.30a. It can be seen that the pressure rises at Lx = reduction in boundary layer thickness at all the tested MVG
0.48, for all the controlled configurations in comparison heights. They efficiently alter the momentum of the bound-
to uncontrolled intake. This may be due to the formation ary layer by remixing the flow downstream and thus, reduc-
of pressure hill caused by vortex generators. Further, the ing the boundary layer thickness. Also, the smallest tested
pressure measured at 0.7L is constant for all the intakes, MVGs (H = 200 µm) cause a number of weak oblique shock
which may be because of shedding of large size vortex waves which leads to the minimum pressure losses, exhibiting
structures from this longest tested MVG, capable of entrain- their efficacy in controlling the SBLIs. Clearly, these obser-
ing the surrounding fluid of higher momentum, towards vations support the findings of the pressure plots (Fig. 14.30).
the lower momentum fluid on the wall. The exchange of
momentum decreases the static pressure downstream of
the MVGs. Further, the static pressure measured at the Summary
farthest port, corresponding to Lx = 0.83, is found to be
increasing for both conventional and ramped-vane MVGs. When the flow past a body, a thin viscous region adjacent
When the height of MVGs was reduced to 400 µm, to the surface develops and grows downstream. This thin
the pressure variation at the first two ports are found to be region where the viscous effects predominate is called
similar to that of the MVGs of height 600 µm (Fig. 14.30b). the boundary layer. Beyond the boundary layer, farther
However, the effect of MVGs are felt at Lx = 0.48, where away from wall, the flow is assumed to be inviscid and
the static pressure is substantially increased for the con- irrotational, whereas the flow within the boundary layer
trolled configurations, whereas the pressure variation for might be rotational owing to the shear stress present on
uncontrolled intake is still found to be similar to that of the surface, which, in turn, make the flow rotational.
the previous case (H = 600 µm). At the axial locations, Similar to boundary layers, the shock waves are also
L = 0.7 and L = 0.83, both conventional as well as
x x the flow discontinuities in which both the viscosity as well
ramped-vane MVGs showed the inferior performance com- as the heat conduction play a dominant role. The signif-
pared to their uncontrolled counterpart, owing to the larger icant difference between the two is that the shock wave
entropy production by these MVGs of relatively longer size. is considerably thinner as compared to an ordinary bound-
On further decreasing the height of MVGs to 200 µm, the ary layer. Moreover, across a thin shock wave a large pres-
pressure plot is shown in the Fig. 14.30c. Here again, the trend sure gradient exists. When the shock waves impinge on the
is almost similar to the previous cases between the first two solid boundaries, they tend to impose such large pressure
pressure ports. Both the controls were effective in decreas- gradients on the boundary layer. Consequently, the bound-
ing the pressure as compared to plain intake. However, the ary layer is necessarily distorted to a considerable extent.
ramped-vane MVGs caused the maximum reduction of pres- In addition, due to convergence of a shock wave and the
sure, revealing its efficacy in controlling the interactions. In boundary layer, their mutual interactions of considerable
addition to near field locations, the effect of MVGs (H = extent are highly probable. These interactions are popularly
200 µm) is observed in far fields also, in contrast to their known as Shock–Boundary Layer Interactions (SBLIs). Since
longer counterparts. Further, the effect of ramped-vane MVGs in almost every supersonic flow both may be found and
are felt stronger and farther downstream in the case of intake hence their interactions are not unusual. The common occur-
0 0.13 0.27 0.48 0.7 0.83 1

x/L
Fig. 14.29 The wall-mounted pressure ports along the intake length
rence is witnessed when an externally generated shock wave Due to the local effects caused by SBLIs, the changes
advances over an area on which there is a boundary layer. in boundary layer take a long time to subside, thereby,
The transonic normal shock wave and turbulent boundary making the flow prone to separation occurring farther
layer interactions, referred to as transonic SBLIs, comprise downstream. In those cases, where the considerable sep-
of supersonic flow upstream of the shock and a subsonic arations occur, SBLI could lead to significant changes
flow downstream. This mixed regions of the flow make in the shock structure and overall flow. They can intro-
the transonic flow interactions distinct from the supersonic duce large-scale flow field unsteadiness, inlet buzz or
and hypersonic interactions. Since the steady subsonic engines unstart. They have the potential to cause intense
flow is always optimally expanded without having waves heating, severe enough to destroy the complete vehicle.
and thus, undergoes gradual changes compared to super- Oswatitsch (1944) deduced an equation which relates
sonic flow. Nevertheless, the downstream subsonic flow the drag on a vehicle with the entropy and the stagnation
conditions can feed upstream and affect the strength of enthalpy introduced by the vehicle into the flow field.
shock, shape and location of shock wave which causes Let us consider an elemental mass (dm = ρ− →
v .n̂ds),
the interactions, while the upstream flow conditions are surrounded by a control surface (CS), is in uniform
shielded from such events by supersonic outer flow. freestream (va ) and let F be the net force acting on the
The simple waves are straight Mach lines with constant elemental mass (dm) in the direction of drag. Then,
conditions on each one and which follow the simple relation
between flow deflection angle (θ) and Prandtl–Meyer function 1
F= (Ta ds − dh0 ) dm
(ν). Supersonic expansion and compression with Mach line va s
which are straight is termed as simple region. Two opposite
families of waves, left and right running waves, depend upon where Ta is the upstream flow temperature, ds is the change
the walls producing these waves are either left or right of the in entropy, and dh0 is the change in stagnation enthalpy.
waves. In the regions, where two simple waves of opposite In internal aerodynamics problems such in supersonic
families interact with each other, the flow is non-simple and intakes, it is desired to minimize the drop in stagnation pres-
thus linear relationship between ν and θ is no more valid. sure occurring due to SBLIs in order to maximize the overall
The distinctive features of shock-induced separation pressure recovery in the intake isolator region. The efficiency
are the shock patterns that occur in the adjoining inviscid is defined in terms of ratio of stagnation pressures given by
flow as a consequence of the behavior of the boundary p02
layer during the interaction process. The shock patterns η=
p0a
are produced when two shock waves intersect or interfere
with each other. These patterns, classified by Edney into where p02 is the mean stagnation pressure at engine
what are now commonly acknowledged as six types, can be level and p0a is the incoming freestream stagnation
interpreted using the discontinuity theory of shock waves. pressure. We know that for the flow of a perfect gas
The Shock–Boundary Layer Interactions (SBLIs) phe- across the shock wave, change in entropy is defined as
nomena is all about the pressure jump caused by the shock

which is imposed on the boundary layer and its response. p0a
The SBLIs occur when a shock wave and a boundary layer s = R ln
p02
converge and since both are commonly present in super-
sonic and hypersonic flows, these interactions are inevitable. or
Summary 387
Fig. 14.30 Variation of the wall static pressure with axial distance at Rc = 1.14
Fig. 14.31 Schlieren images of uncontrolled and controlled Mach 2.2 intakes at Rc = 1.14

p02 s2 − s1 structure for several canonical configurations. Despite the
= exp
p0a R knowledge gained from the canonical configurations, their
usefulness in predicting the flow field structure particularly
The above equation reveals that the efficiency or stagnation the location and extent of separation for realistic flight
pressure losses are also related to the entropy production (or vehicle configurations, is limited by the sheer complexity
viscosity). Thus, it is evident that the pressure losses is an of typical shock wave interactions in three dimensions.
internal aerodynamic equivalent to the drag in external flows. The detrimental effects of SBLIs are very much unde-
The discussion on literature provided a basic insight to the sired since they have drastic effects on the flow field
reader to understand the key concepts of the SBLIs. Apart starting from flow separation and unsteadiness to severe
from the two-dimensional interactions, the requirement to localized heating, which might lead to the failure of the
gain insight of the complex three-dimensional interaction whole system. Therefore, the consequences of the SBLIs
characteristics is increasing more than ever due to the occurrence almost invariably are detrimental in some
advancements in aircraft and missiles, where the obvious respect which affects the efficiency and performance
presence of three-dimensional SBLIs are witnessed. It of the vehicle. Thus, the need arises for controlling the
is understood by considering a number of fundamental phenomenon by some appropriate methods which modify
geometries based on the shape of the shock wave generator the flow, either before or during the interaction process.
namely, sharp unswept and swept fins, semi-cones, swept The SBLI control techniques are classified into passive
compression ramps, blunt fins, and double sharp unswept and active types. In the former, an attempt is made to
fins. Here, the shock wave generators have an overall size produce a beneficial result without the expenditure of
sufficiently large compared to the boundary layer thickness, externally supplied energy and includes well-established
so that any further increase in size does not affect the flow. methods as the ubiquitous vortex generator, riblets, as
In addition, numerous experimental investigations yielded well as the venting methods. Active control, on the other
detailed descriptions of the three-dimensional flow field hand is categorized as being either predetermined or
Summary 389
interactive. In the former case, steady or unsteady energy 5. Discuss the co-rotating and the counter-rotating vortex
inputs are made through some form of actuator irrespective generation by using micro-vortex generators. How they
of the state of the flow field. In contrast, in an interac- are efficient in controlling the shock–boundary layer
tive method the power supplied to the actuator is varied interactions at supersonic Mach numbers?
continually depending on input from a sensor or sensors.
In many instances, the interactions of shock wave with Multiple Choice Questions
boundary layer may be computed effectively using modern
CFD techniques. But, these methods are prone to errors when 1. According to Edney classification, the shock–shock inter-
the boundary layer separates. Thus, the extensive experiments ference patterns are classified into how many types?
need to be performed to understand the underlying principle
of interactions and their control. The major parameters con- (a) five types
sidered for analyzing these flows are the variation of total (b) six types
pressure at the centerline and the wall surface pressure distri- (c) seven types
bution. The total pressure distribution is measured using pitot (d) eight types
probe along the centerline, whereas the wall static pressure is
obtained through the static pressure ports mounted on the wall. 2. The impingement of a bow-shock on the laminar bound-
In addition to pressure measurement, the wave strength ary layer causes
and boundary layer characteristics can be qualitatively
(a) boundary layer thickening but no separation
studied using the optical flow visualization techniques
(b) boundary layer separation
such as Schlieren or Shadowgraph. These visualiza-
(c) the reduction of boundary layer shape factor
tion images can confirm the strength of waves and
(d) none of the above
overall flow structure, predicted by the pressure plot.
The shock wave and boundary layer interaction studies are 3. The equation, which relates the drag on a vehicle with
generally performed in the test section of a high-speed wind the entropy and the stagnation enthalpy introduced by the
tunnel. The experimental facility contains an air supply sys- vehicle into the flow field, is given by
tem which consists of a compressor, reservoir tank, cooling
and drier units, and an air delivery system. The compressed (a) Fliegner
and dried air from the storage tank is supplied into the settling (b) Blasius
chamber through the gate valve. The settling chamber is pro- (c) Oswatitsch
vided with both pressure and temperature ports (or taps) for (d) Reichardt
readings. The settling chamber total pressure (P0 ), which is
a controlling parameter in the experiment is maintained con- 4. Which of the following is correct for the shock-induced
stant during the test run by controlling the pressure regulating separation?
valve (PRV). The temperature of settling chamber is same
as that of the ambience. From settling chamber, the flow is (a) It is more pronounced in laminar boundary layer.
fetched into the test section where the model is mounted using (b) It is more pronounced in turbulent boundary layer.
a cantilever mechanism, located at the end of the test section. (c) It is more pronounced during boundary layer transi-
tion.
(d) It has similar effects in both laminar and turbulent
boundary layers.
Exercises
5. The shock impinging on a flat plate with an already thick-
Descriptive Type Questions ened boundary layer produces
1. What is shock–shock interference? Write a short note on (a) strong-reflected shock waves.
the Edney classification of shock–shock interference. (b) weak-reflected shock waves.
2. Write a short note on induced-drag prediction method (c) expansion waves.
using Oswatitsch’s equation. (d) weak Mach waves.
3. What is Shock–Boundary Layer Interaction (SBLI)? Dis-
cuss its consequences. Whether SBLI always have detri- 6. Which of the following is not a principle involved in
mental effects? shock–boundary layer interaction control?
4. Discuss the mechanism of shock control by using cavity
with porous upper surface. (a) mass injection
(b) localized boundary layer suction
(c) momentum re-energization Andreopoulos J, Muck KD (1987) Some new aspects of the shock-wave
(d) pre-heated walls and boundary-layer interaction in compression ramp flows. J Fluid
Mech 180:405–428
Anderson BH, Tinapple J, Surber L (2006) Optimal control of shock
7. Which of the following is not an example of the transonic wave turbulent boundary layer interactions using micro-array actua-
flow interactions? tion. Proc 3Rd AIAA Flow Control Conf. San Fransico, USA. AIAA-
2006-3197
(a) external compression inlet operating at Mach 2.0 Babinsky H (2007) Understanding Micro-Ramp Control for Shock
(b) flow over an airfoil at Mach 8.5 Boundary Layer Interactions. AFRL Contract Report. FA9550-06-
1-0387
(c) mixed compression intake operating at Mach 5 Babinsky H, Li Y, Ford CWP (2009) Microramp Control of Super-
(d) cascade compressor blades with a relative Mach sonic Oblique Shock-Wave and Boundary-Layer Interactions. AIAA
number 0.9 J 47(3):668–675
Barry RW, Shapiro AH, Neumann EP (1951) The Interaction of shock
8. The shock control technique waves with boundary layers on a flat surface. J Aerosp sci 18(4):229–
238 (Institute of the Aeronautical Sciences)
Blinde PL, Humble RA, van Oudheusden BW, Scarano F (2009) Effects
(a) reduces the stagnation pressure losses of micro-ramps on a shock wave and turbulent boundary layer inter-
(b) reduces the skin friction losses action. Shock Waves. 19(6):507–520
(c) reduces both stagnation pressure and skin friction Beresh SJ, Clemens N, Dolling DS (2002) The relationship between
upstream turbulent boundary layer velocity fluctuations and separa-
losses
tion shock unsteadiness. AIAA J 40(12):2412–2422
(d) improves the shock strength Bibko VN, Efimtsov BM, Kuznetsovn VB (1989) Spectrums of surface
pressure pulsations ahead of inside corners. Uchenie Zapiski TSAGI
9. The boundary layer control techniques (Scientific notes of TSAGI) 20(4):112–117 (in Russian)
Davis JP, Sturtevant B (2000) Separation length in high-enthalpy shock
(a) reduces the stagnation pressure losses and boundary-layer interaction. Phys Fluids 12(10):2661–2687
Dolling DS (2001) Fifty years of shock-wave and boundary-layer inter-
(b) reduces the skin friction losses action research: What next? AIAA J 39(8):1517–1531
(c) reduces both stagnation temperature and skin friction Dolling DS (1993) Fluctuating loads in shock wave turbulent boundary
losses layer interaction: Tutorial and update. Proc 31st AIAA Aerosp Sci
(d) reduces the shock strength Meet Exhib. Reno, Nevada. AIAA-93-0284
Donaldson C (1944) Effects of interaction between normal shock and
boundary layer. NACA-CB-4A27
10. Which of the following is an efficient passive method to Donaldson CD, Lange RH (1952) Study of the Pressure Rise Across
control boundary layer during SBLI? Shock Waves Required to Separate Laminar and Turbulent Boundary
Layers. NACA-TN-2770
(a) plasma actuators Erengil ME, Dolling DS (1993) Physical causes of separation shock
(b) surface bump unsteadiness in shock-wave and turbulent boundary-layer interac-
tions. AIAA-93-3134
(c) cavity with porous walls Fage A, Sargent RF (1947) Shock wave and boundary layer phenom-
(d) micro-vortex generator ena near a flat surface. Proc R Soc Lond, Ser A: Math Phys Sci
190(1020):1–20
Fage A, Sargent RF (1949) Effect on aerofoil drag of boundary-layer
Keys suction behind a shock wave. Aeronaut Res Counc Tech Rep
Humrutha G, Kaushik M, Shock Sinhamahapatra KP (2017) Control
1. (b) boundary layer interaction, in supersonic intake using cavity with
2. (b) porous surface. In: 35th AIAA applied aerodynamics conference,
June 7, Denver, Colorado
3. (c)
Holden H, Babinsky H (2007) Effect of microvortex generators
4. (a) on separated normal shock/boundary-layer interactions. J Aircraft
5. (c) 44(1):170–174
6. (d) Holden MS (1986) A review of aerothermal problems associated with
hypersonic flight. Proc AIAA 24th Aerosp Sci Meet. Nevada, USA
7. (c)
Kaufman LG, Hartofilis SA, Evans WJ, Oman RA, Meckler LH, Weiss
8. (a) D (1962) A Review of Hypersonic Flow Separation and Control
9. (b) Characteristics. Gruman Aircraft Engineering Corporation Contract
10. (d) Report. ASD-TDR-62–168
Knight DD, Degrez G (1998) Shock wave and boundary layer interac-
tions in high mach number flows: A critical survey of current numer-
ical prediction capabilities. Advis Rep 319, AGARD 2: 1.1–1.35
References Korkegi RH (1975) Comparison of shock induced two and three dimen-
sional incipient turbulent separation. AIAA J 13(4):534–535
Liepmann HW (1946) The interaction between boundary layer and shock
Ackeret F, Rott M (1947) Investigations of compression shocks and waves in transonic flow. J Aerosp Sci 13(12):623–638
boundary layers in gases moving at high speed. NACA-TM-1113
References 391
Liepmann HW, Roshko A, Dhawan S (1952) On reflection of shock Reinartz B, Ballmann J, Brown L, Fischer Ch, Boyce R (2007) Shock-
waves from boundary layer. NACA-TM-1100 wave and boundary-layer interactions in hypersonic intake flows.
Lee Y (1998) Experimental study on fluctuating thermal loads in swept Proc 2Nd Eur Conf Aerosp Sci. Brussels, Belgium
shock wave and turbulent boundary-layer interactions. KSME Int J Schulte D, Henckels A, Neubacher R (2001) Manipulation of shock and
12(4):719–727 boundary-layer interactions in hypersonic inlets. J Propuls Power
Lee S (2009) Large eddy simulation of shock and boundary layer interac- 17(3):585–590
tion control using micro-vortex generators. PhD Thesis, Dep Aerosp Souverein LJ, Debiève JF (2010) Effect of air jet vortex generators on a
Eng, University of Illinois. Illinois, USA shock wave boundary layer interaction. Exp Fluids 49(5):1053–1064
Mitani T, Sakuranaka N, Tomioka S, Kobayashi K (2005) Boundary- Thomas FO, Putnam CM, Chu HC (1995) On the mechanism of unsteady
layer control in Mach 4 and Mach 6 scramjet engines. J Propuls shock oscillation in shock wave and turbulent boundary layer inter-
Power 21(4):636–641 actions. Exp Fluids 18(1–2):69–81
Oswatitsch K (1944) Der Druckwiedergewinn bei Geschossen mit Viswanath PR (1988) Shock-wave-turbulent-boundary-layer interaction
Riickstossantrieb bei hohen Uberschallgeschwindigkeiten. KWI- and its control: A survey of recent developments. Sadhana. 12(1–
Ber. NACATM- 1140 (in German) 2):45–104
Pozefsky P (1989) Identifying sonic fatigue prone structures on a hyper- Webb NJ (2009) Control of the Interaction between shock wave and
sonic Transatmospheric Vehicle (TAV). Proceedings of the 12th a supersonic turbulent boundary layer by Localized Arc Filament
Aeroacoustic Conference. San Antonio,USA Plasma. PhD Thesis, Department of Mechanical Engineering, The
Ohio State University. Ohio, USA
Zhang Y, Tan HJ, Du MC, Wang DP (2015) Control of shock and bound-
ary layer interaction for hypersonic inlets by highly swept micro-
ramps. J Propuls Power 31(1):133–143
Supplemental Readings
A
A.1 Introduction the flow from supersonic to subsonic speeds in a transonic

SBLI. Although it is preferable to design inlets with weaker
We will now discuss the basic details about air-breathing terminating shock waves, constraints on overall system size
engine intakes which are frequently used in Ramjets and generally limit the compression that can be achieved through
Scramjets. In addition, a small discussion about supersonic oblique shock waves.
combustion is also provided which help the reader in under- Such strong interactions pose considerable problems for
standing the combustion at supersonic Mach numbers. We inlet efficiency. The strong normal or near-normal shock
have already referred them many times earlier in the text for waves incur considerable entropy increase and stagnation
purposes of explaining the various phenomena. Moreover, we pressure loss, which is a direct performance loss for the sys-
do not discuss all possible topics in this appendix but will tem. Further, since the boundary layer already has experienced
choose those that are essential for explaining the theories as a number of adverse pressure gradient regions in the previ-
presented in the text. ous SBLIs, in turn, the boundary layer becomes more vul-
nerable to flow separation, when encountered the final shock
wave. Flow separation has an obvious detrimental impact on
A.2 Air-Breathing Engine Intakes inlet performance. In addition to the introduction of addi-
For jet aircraft operating at supersonic speeds, it is neces- tional stagnation pressure losses, it introduces considerable
sary to decelerate and compress the incoming air to subsonic nonuniformity in the flow entering the subsonic diffuser or
speeds before entering into the combustor. It is achieved via the combustor.
the intake. The simplest form of compression is via a normal Moreover, any flow separations are also likely to introduce
shock ahead of a pitot inlet, but this incurs significant stag- considerable unsteadiness into the flow, which can lead to
nation pressure losses, rendering this form of intake imprac- unacceptable dynamic loads on the engine. If the terminal
tical for M > 2. A better approach is to generate a series (near-normal) shock oscillation is so extreme that it reaches
of oblique shock waves that can increase the pressure and the converging part of the inlet geometry, it becomes unstable.
reduce the Mach number before eventually changing the flow At this point, it moves rapidly upstream, making more of the
state to subsonic through a terminating near-normal shock. flow inside the inlet subsonic, until it is eventually expelled
For a given incoming flow Mach number, a series of multiple from the intake causing unstart (or buzz, if this phenomenon is
shock waves incurs a smaller entropy production and, thus, periodic). This is comparable to shock stall or shock-induced
lower losses than a single normal shock wave. Depending on buffet on transonic wings; such a violent event is extremely
whether the oblique shock waves are generated outside the damaging to the engine. To avoid the problems associated with
intake or within the inlet duct, such designs are referred to as strong transonic SBLIs in inlets, the researchers make use of
external or internal compression inlets. flow control to enable the boundary layer to stay attached even
In either case, the shock wave interacts with the boundary when the shock waves have considerable strength. The most
layer growing along the inlet surface. Most of the interactions popular control method is boundary layer suction, or bleed.
feature oblique shock waves with supersonic flow on both In any air-breathing engines, an inlet, a combustion cham-
sides of the interaction. However, in each inlet design, there ber, and a nozzle are the three main components. Furthermore,
is a final terminating, near-normal shock wave that switches it is established that 1% loss in inlet stagnation pressure even-

394 Appendix A: Supplemental Readings
tually leads to about 1–1.5% loss of engine gross thrust (Intake where πD is the ratio of stagnation pressure, and p0,exit and
Aerodynamics by J. Seddon and E. Goldsmith 1999). There- p0,entry , respectively, are the stagnation pressure at the inlet
fore, an efficient performance of the engine components is of exit and the stagnation pressure at the inlet lip. Also, to a very
prime importance for good performance of the whole engine. high degree of approximation the flow in the inlet is assumed
Also, designing the engine components with high accuracy is to be adiabatic, that is, no exchange of heat transfer between
more critical in the regions of increasing static pressure than the inlet and its surroundings. We have,
the areas where static pressure decreases. This is because of
boundary layer separation in the presence of adverse pressure T0,exit
τD = =1 (A.2)
gradient. Clearly, the inlet design is more challenging than the T0,entry
nozzles which are subjected to favorable pressure gradients.
where τD is the ratio of stagnation temperature, and T0,exit
and T0,entry , respectively, are the stagnation temperature at the
A.3 Engine Inlets exit and the stagnation temperature at the lip. The design of
subsonic inlet is greatly influenced by the two major require-
An inlet is the device which recovers pressure energy by ments; to prevent the separation of boundary layer at high
reducing the kinetic energy of the flow. Depending on the angles of attack and need of high mass flow during landing
flight Mach numbers, they are categorized into; subsonic or and takeoff; and to suppress the formation of both internal
supersonic inlets. Inlets find tremendous application predom- and external shock waves at transonic flight Mach numbers.
inantly in jet engines. But these two requirements are contradictory to each other,
because a thick inlet lip is the best suited for high angle of
attack engine operation, whereas a thin lip is suitable to high
A.3.1 Subsonic Inlets Mach number requirements. With the advent of modern com-
puting, it is now feasible to obtain analytical estimation of the
It is known that the flow entering into the compressor of a tur- complex flow fields and the associated losses to develop the
bojet engine must have the Mach number (M) in the range of best compromised inlet designs.
0.4−0.7, where the upper limit is suitable for transonic com-
pressors or fans. Further, if the engine has to operate for the A.3.1.1 Flow through the Inlet (Internal Flow)
subsonic level flight at M = 0.85, then the inlet must cause the Based on varied flight Mach numbers and mass flow require-
flow deceleration from 0.85 to about 0.6. It should be noted ments of the engine, an inlet has to operate at different
that the flow undergoes both external and internal deceleration freestream conditions. To investigate the inlet performance,
in an intake. The properly designed intake should minimize or let us consider two typical subsonic freestream conditions
eliminate boundary layer separation even during the pitch and and their corresponding thermodynamic processes on T − S
yaw motions of the aircraft. Also, there should be minimum diagrams of an aggregate fluid lump as shown in Fig. A.1.
stagnation pressure loss in an inlet and it must deliver a uni- In this figure, freestream conditions are depicted by sub-
form flow to the compressor. A nonuniform flow at the entry script “a” upstream of the inlet and Aa is the streamtube cross-
to the compressor not only affects its efficiency drastically but sectional area. The concept of streamtube introduced here is
also, it may lead to flow-induced vibrations thereby causing very useful and resembles an aerodynamic duct. The airflow
the failure of blades. In addition, as the diffuser is required to entering into the inlet may either undergo acceleration or decel-
have a stable operation in both subsonic and supersonic flow eration in the aforesaid aerodynamic duct. During level cruise
regimes, its design becomes more challenging. motion, where an aircraft flies at high Mach number with rela-
Typically, a subsonic inlet suffers mainly from the follow- tivelylowermassflowattheinlet,airexperiencessomedeceler-
ing three types of losses: ation external to intake, shown in Fig. A.1a. Thus, an increase in
cross-sectionalareaofthestreamtubefromfreestreamtointake
1. Losses due to wall friction. entry can be observed. In other flight modes, such as during
2. Losses due to shock waves (at high subsonic or transonic takeoff and landing, the mass flow requirement is high but the
flight conditions). aircraft speed is low. In these low-speed high-thrust flight oper-
3. Losses due to separation the flow. ations, the streamtube resembles a converging duct as shown
in Fig. A.1b, which illustrates the external acceleration of air
As the flow passes through the inlet, all the above factors
before entering into the inlet. Essentially, in both of the afore-
cause loss of stagnation pressure. That is,
said cases, the air undergoes a change of state outside the intake
p0,exit following an isentropic process as there is no physical surface
πD = <1 (A.1) involved to introduce friction.
p0,entry
Appendix A: Supplemental Readings 395
p
0a
p
T 0a 0a 02 02 p
2
1 2
02s
2
a
Enthalpy (h) 2s
Aa
p
1 1
p
a
a
Ta Inlet
Entropy (s)
(a) During level cruise motion (high Mach number flight or low air mass flow rate).
p
0a
p p
T0a 0a 02 2
1 2
02s
2
a
Enthalpy (h)
2s
Aa
p
a
a
p
1
1
Ta Inlet
Entropy (s)
(b) During landing or take-off (low Mach number flight or high air mass flow rate).
Fig. A.1 Streamline patterns and the corresponding h − s diagrams for subsonic inlets
The external acceleration, shown in Fig. A.1b, lowers the be too large; otherwise, it will cause shock-induced boundary
static pressure at the entry to the inlet and thus increases the layer separation resulting in high nacelle drag.
internal pressure rise across the diffuser for the flow velocities From the physics point of view, the flow process in an
at stations “a” and “2”. If the rise in internal pressure is too inlet and a diffuser is similar. In both, pressure rises and
large, the diffuser may stall due to boundary layer separation fluid momentum decreases, with now work being done as the
leading to increased loss of stagnation pressure. Conversely, fluid passes through the device. However, it should be noted
the external deceleration causes less increase in static pres- that despite having quantum of experimental and computa-
sure inside the diffuser and consequently, the boundary layer tional results on diffusers, they are not directly applicable as
is subjected to lower adverse pressure gradient and thereby, subsonic aircraft inlets. This is because the maximum pres-
less prone to separation (Fig. A.1a). Therefore, the inlet area sure recovery across a diffuser is accompanied by a highly
is chosen so as to minimize the external acceleration during nonuniform velocity profile at the exit and even it has some
takeoff with the result that external deceleration occurs dur- unsteadiness in the flow. Whereas in subsonic aircraft inlets,
ing level cruise operation. In these conditions, the upstream it is necessary to have steady and uniform flow velocity enter-
streamtube area Aa is less than the diffuser area at station “1”, ing the compressor. Therefore, the design of inlet does not
i.e., Aa < A1 and some amount of flow spillage takes place depend on the results obtained through research on diffuser;
over the inlet surface, which accelerates as it passes over the rather, it relies on the potential flow coupled with boundary
surface. Further, at high subsonic Mach number operations, layer analysis followed by wind tunnel testing over a wide
this acceleration and subsequent deceleration of the flow must range of test conditions.
Fig. A.2 Probable locations of boundary layer separation
In actual inlets, the boundary layer may separate in any and the static pressure. Further, Ai and Amax , respectively,
of the flow zones: “1”, “2”, or “3”, as shown in Fig. A.2. are the minimum and maximum areas of the inlet, subscript
The acceleration and subsequent deceleration of the flow at “a” depicts the freestream condition. For simplicity, we will
outer surface results in the boundary layer separation in zone assume the flow to be one-dimensional and incompressible.
“1”, which eventually leads to high nacelle drag. On inter- The net momentum flux out of C∀ will be
nal surfaces, two probable locations of flow separation are as
follows: zone “2” and zone “3”, depending on the duct geom- m̊s va + ρvi2 Ai − ρva2 Amax (A.3)
etry and operating conditions. The flow separation in zone “3”
occurs due to large adverse pressure gradient caused by the The mass flow rate m̊s escaping from the sides of C∀ is
flow acceleration around the nose of the centerbody and sub-
sequent deceleration at the rear end as the curvature decreases. m̊s = ρva Amax − ρvi Ai (A.4)
A.3.1.2 Flow over the Inlet (External Flow) Thus, Eq. (A.3) can be written as
It is seen that both internal and external flow deceleration
poses great challenges in the design of inlets and hence ρAi vi2 − vi va (A.5)
requires an optimization between the two. To investigate
the effect of external deceleration of subsonic flow on the Further, the net force in axial direction, acting on the control
inlet design, let us consider a typical streamline pattern over volume, is
the inlet as shown in Fig. A.3. The flow acceleration on the
external surface will cause a low-pressure zone which can pa Amax −pi Ai − Fx (A.6)
adversely affect the boundary layer in two ways. In a flow
domain where the entire flow is subsonic, the decrease in where Fx is the component of F in x−direction. In absence of
pressure will be followed by a region of rising pressure lead- friction, Fx can be expressed as
ing to boundary layer separation. Under this condition, one
might expect to have a point of minimum pressure or maxi- ˆ ˆmax
A
mum flow velocity downstream of which boundary layer may Fx = ˆ
p î .ndA= pdAx (A.7)
be separated. If the flow domain is partly supersonic, then it
inlet Ai
will end up abruptly forming a shock wave which will interact
the boundary layer on the wall causing the latter to separate.
where î and n̂, respectively, are the unit vectors in the flow
Thus, in this case, also, the local Mach number must be below
direction and in outward pointing direction normal to inlet
a limiting value in order to prevent the boundary layer separa-
area. Introducing above equation into Eq. (A.6), we have
tion. Whatever may be the case, the detachment of boundary
layer is an undesired phenomenon which will adversely affect ˆmax
A
the overall pressure recovery as the flow passes downstream
pa Amax − pi Ai − pdAx (A.8)
in the aircraft engine. It also generates a net rearward force or
drag on the body. Ai
In Fig. A.3, m̊s and F, respectively, are the mass flow rate
Combining this result with Eq. (A.3), the momentum equation
crossing from the sides of control volume (C∀) and net force
will be written as
acting on the inlet, v and p, respectively, are the flow velocity
minimum (pmin ) and the flow velocity is maximum (umax ).

Hence, the pressure must rise from pmin to ambient pressure
ˆmax
A
(pa ) at some downstream location. Thus, on the outer surface
pa Amax − pi Ai − pdAx = ρAi vi2 − vi va of the inlet the average pressure can be defined as
Ai
´ Amax
Ai (pa − p) dAx
From algebraic rearrangement of the terms, we have pa − p = = f (pa − pmin )
(Amax − Ai )
where f is a factor between 0 and 1. Therefore, Th can be

ˆmax
A
written as
(pa − p) dAx = ρAi vi2 − vi va + (pi − pa ) Ai
Ai Th = f (pa − pmin ) (Amax − Ai )
Applying Bernoulli’s equation between entry and exit of the Equation (A.9) can be rewritten as
inlet
2
f (pa − pmin ) (Amax − Ai ) vi
va2 − vi2 = 1−
pi − pa = ρ 2 ρva Ai
1 2 va
2
or
Thus,
2
1 − vvai
ˆmax
A

Amax
=1+
v2 − vi2 2
(pa − p) dAx = ρAi vi2 − vi va + ρ a Ai
f vvmax Cp,max
2 a
Ai
(pa −pmin )
or where Cp,max = 1 is the pressure coefficient which
2 ρva
2
must not be too large; otherwise, boundary layer will separate.

ˆmax
A 2 Hence, the above equation can be written as
1 vi
(pa − p) dAx = ρva2 Ai 1 −
2 va 2
Ai vi
Amax 1− va 1 − Cp,max
´ Amax =1+ (A.10)
It should be noted that the term Ai (pa − p) dAx is the Ai fCp,max
increment in component of thrust (Th) which acts on the
It should be noted that the value of factor, (f), depends on
front external surface of the inlet due to reduction in static
the shape of the inlet. For example, taking f = 0.5, we can
pressure. Thus,
show the size of external surface required to prevent boundary
ˆmax
A layer separation for a given vvai , as shown in Fig. A.4. Also, it
is interesting to note that to prevent excessive drag caused due
Th = (pa − p) dAx
to flow separation, the size of nacelle should be increased with
Ai
increase of external
deceleration, i.e., by decreasing the veloc-
vi
or ity ratio va . Nevertheless, the condition of higher drag for
the larger nacelle still holds even without separation. Further,
2
1 2 vi the effect of relatively moderate velocity ratios vvai > 0.8
Th = ρva Ai 1 −
2 va on minimum nacelle size is rather small. Alternatively, the
or employment of partial internal deceleration is found to be
quite useful to reduce the maximum diameter of the inlet
2 because it allows a reduction in both Ai and AAmax
Th vi .
= 1− (A.9) i
Even though the above analysis is carried out assuming a
2 ρua Ai
1 2 va
simplified flow field around the inlet, yet, it showed that the
From Eq. (A.9), it is evident that the thrust generated by the performance of an inlet depends on the pressure gradient on
engine can be increased by lowering the flow speed at the
both internal and external surfaces as well as on area ratio
Amax
entry to the inlet. Furthermore, we have seen that on the outer Ai . The pressure rise on the outer surface is limited by
surface of the inlet, there exists a point where the pressure is Amax
external compression and Ai whereas the internal pressure
CV va
ms
va
F
pa
va
vi Ai A max
pa
p
F a
va
Fig. A.3 Calculation of thrust over inlet surface
3.0
0.4
0.5
0.6 Cp
max
2.0
A max
Ai
1.0
0 0.2 0.4 0.6 0.8 1.0

vi
va

Amax vi
Fig. A.4 Variation of area ratio Ai with velocity ratio vmax for f = 0.5
rise depends on the flow deceleration between entry to the inlet If the flow velocity leaving the diffuser exit is assumed to
and entry to the compressor in turbojet engine (or combustion be small, i.e., v2 ≈ 0, then we have
chamber in case of a ramjet). For a more practical analysis,
one must consider compressibility effects. h02s − ha
ηd = (A.11)
h0a − ha
A.3.1.3 Performance Criteria of Inlets
Assuming the gas to be perfect (say, air), we can
The performance of an inlet is evaluated either in terms of
write
isentropic efficiency or in terms of stagnation pressure ratio.
To get a clear idea, let us look into them in detail in the fol-
h = Cp T (A.12)
lowing sections.
Isentropic Efficiency We have
The performance of an inlet is estimated in terms of isen- T02s − Ta
tropic efficiency, also known as diffuser efficiency (ηd ). It is ηd = (A.13)
T0a − Ta
defined as the ratio of enthalpy change of the flow between the
entrance and exit of the diffuser to the kinetic energy of the where 02s depicts the isentropic state that would be reached
flow. A typical flow process in an inlet, depicted by Mollier by isentropic compression to the actual outlet stagnation pres-
diagram is shown in Fig. A.5. sure. Further, from isentropic relations
p
0a
p
T0a 0a 02 02 p
2
1 v2
02s 2 2
2 2
1
Enthalpy (h) 1 2
2s
v a
2 1
p
1
1
p
a
Ta a
Inlet
Entropy (s)
Fig. A.5 Mollier (h − s) diagram of flow states in the subsonic inlet

T02 γ−1 tion pressure ratio (rd ) across the shock wave. For a typical
=1+ Ma2 (A.14)
Ta 2 subsonic diffuser, the variations of rd and ηd with flight Mach
number (M) are given in Fig. A.6.
where Ma is the freestream Mach number. Also,
γ−1
T02s p02s γ
= (A.15) A.3.2 Supersonic Inlets
Ta pa
Similar to flying at subsonic speed, for supersonic flight also,
Therefore, the isentropic efficiency of the inlet (ηd ) is given it remains necessary; at least for present designs, the flow
by Eq. (A.16). coming out of an inlet must be subsonic only. The com-
γ−1 pressors capable of ingesting supersonic stream, however,
p02 γ may provide very high mass flow rate per unit area and
pa −1
high-pressure ratio per stage. However, the development of a
ηd = (A.16)
γ−1 supersonic compressor without excessive loss of stagnation
2 M2
pressure across the shock waves is still far from reality. Thus,
till date, the Mach number of the airstream (in axial direc-
Stagnation Pressure Ratio
tion) approaching to a subsonic compressor must not be more
The diffuser effectiveness is also evaluated in terms of stag- than 0.4; however, for transonic stage, it can go up to a max-
nation pressure ratio (rd ), defined as imum of about 0.6. Here, the transonic stage refers to the
relative speed of the axial flow with respect to the blade
p02 tip, and thus, the absolute speed will still be less than 0.4
rd = (A.17)
p0a only.
The challenges posed by an incoming supersonic stream
From algebraic rearrangements
in the designing of compressor are not present in a ramjet.
γ Also, it is possible to have combustion at supersonic Mach
p02 p02 p0a γ−1 2 γ−1
= × = (rd ) 1 + Ma numbers eliminating aerodynamic losses such as shock asso-
pa p0a pa 2 ciated losses. The ramjet which permits combustion at super-
sonic speeds is better known as supersonic combustion ram-
Introducing above in Eq. (A.16), we have
jet (scramjet), the concept which could not be applied so far

2 γ−1 2 γ−1 due to difficulty in having a stable combustion without exces-
ηd = 1+ Ma (rd ) γ − 1 (A.18) sive aerodynamic losses. Therefore, at present, the supersonic
(γ − 1) Ma2 2
inlets are designed to decelerate a supersonic flow to subsonic
From Eq. (A.18), it is clear that for a given freestream Mach speeds bearable by existing compressors of turbojet engines
number (Ma ) the diffuser efficiency depends only the stagna- or fans of ramjet combustors.
1.0
rd
0.95
η
d
0.90
0.85
0 0.2 0.4 0.6 0.8 1.0
M
Fig. A.6 Performance curves of a typical subsonic diffuser
A.3.3 Hypersonic Inlets ing that delivers the air from inlet to combustor is called the
diffuser, whereas it is termed as isolator for scramjet engines.
The emerging hypersonic air-breathing propulsion systems, The burnt mixture in the combustor is subsequently expanded
currently under development, will provide a means for sus- in a convergent–divergent nozzle, exiting once again at super-
tained and accelerating flight within the atmosphere at hypersonic speeds. At hypersonic Mach numbers, the losses asso-
sonic Mach numbers. These propulsion systems could be used ciated with decelerating the incident airstream to subsonic
in long range cruise missiles to intercept the time-sensitive tar- speeds become quite large and hence, the supersonic combus-
gets, responsive hypersonic aircraft for global payload deliv- tion ramjet (scramjet) is preferred. In scramjets, the incoming
ery and reusable launch vehicles to achieve cost-effective flow is still compressed by the inlet, but the combustion is
space access. allowed to occur at supersonic Mach numbers.
The hypersonic propulsion systems are broadly classified The inlets which are used to compress the requisite amount
into; air-breathing and non-air-breathing. The liquid and solid of airstream at hypersonic Mach numbers are called hyper-
propellant rocket motors fall under the category of non-air- sonic inlets. The following precautions should be exercised to
breathing propulsion systems because they do not require improve the overall pressure recovery achieved in hypersonic
atmospheric oxygen in the combustion process. Instead, they inlets:
carry both fuel and oxidizer either separately in liquid fuel
tanks or combined within a solid propellant grain which • The static pressure rise can be maximized by minimizing
are burned within a high-pressure chamber to produce hot the viscous losses on the walls and by reducing the stag-
gaseous products that are expanded through an exhaust nozzle nation pressure loss.
to produce thrust. The commonly used air-breathing engines • The inlet contribution in overall drag should be minimized.
are; turbojet, turbofan, turboprop, and ramjet and scramjet, as • The inlet performance should not be affected much by
schematically shown in Fig. A.7. Because of material limita- varying the angle of attack.
tions on allowable turbine blade temperature, the maximum • The inlet must be able to withstand the back pressure
flight speed attained by a turbojet engine is usually limited caused due to heat transfer.
to Mach 3.5. The primary air-breathing engine used to fly
at Mach numbers approaching to 5, is the ramjet. At super- In hypersonic inlets, the incoming airstream should be com-
sonic speeds, a ramjet-powered vehicle utilizes an inlet that is pressed to about three times, before it is being ducted to the
designed to ingest the atmospheric air and compress it for an combustion chamber. The incident airstream is decelerated
efficient combustion. Once the air is compressed, it is ducted in a highly convergent duct and for a given Mach number the
into a combustion chamber where it is mixed with fuel, and duct can achieve two different flow configurations. The inlet is
the mixture is burnt to raise the temperature and pressure said to be unstarted, when a strong bow shock is located ahead
inside the engine. In subsonic combustion ramjets, the duct- of the inlet lip, causing the hypersonic airstream to become
Compressor Turbine
Combustor
Airstream Exhaust Gases
Diffuser Nozzle
(a) Turbojet engine.

Engine cowl
Airstream
Turbine Exhaust gases

Fan Burner
Compressor Crankshaft Nozzle
Burner
Inlet
(b) Turbofan engine.

HPT
Low pressure High pressure
compressor compressor Combustor LPT Nozzle
Airstream
Exhaust
Inlet gases
HPT− High Pressure Turbine

w LPT− Low Pressure Turbine
(c) Turboprop engine.
Fuel injector Flame holder

Airstream Exhaust gases
Diffuser Combustion Nozzle

chamber
Inlet
(d) Ramjet engine.
Fig. A.7 Schematic layout of some air-breathing engines
subsonic and permitting the requisite amount of flow spillage. In a hypersonic inlet, the pressure recovery is achieved
But, when there are no bow shock and no flow spillage and by compressing the flow by a series of oblique shock waves,
the flow is supersonic throughout in the inlet, the condition is as shown in Fig. A.8. The interaction of these shock waves
referred to as start condition. with boundary layer on the wall is termed as shock-boundary
Oblique shock wave
Separation bubble
M>5 M>1 M>1
Cowl Separated boundary layer
Fig. A.8 Schematic layout of a hypersonic inlet
layer interactions (SBLIs). The impingement of an oblique can be viewed as the working of a convergent–divergent noz-
shock wave on the boundary layer imposes an adverse pres- zle in reverse. This is why, the convergent–divergent diffuser
sure gradient, which decelerates the flow and eventually leads is also referred to as reverse nozzle diffuser. The major dif-
to boundary layer thickening. Interaction of boundary layer ference between the reverse nozzle diffuser and the second
with shock reflections is assumed to be responsible for the throat is that in the former the aerodynamic streamtube effect
flow separation and causing inlet unstart. However, the sepa- is felt ahead of it, whereas in the latter there is no such effect is
rated boundary layer subsequently reattaches to the surface at felt when used in supersonic wind tunnel. The application of
some downstream location. The recirculatory zone between supersonic diffuser is associated with many practical difficul-
the points of separation and reattachment of the boundary ties. For example, it has to operate successfully over a wide
layer is known as “separation bubble”, which acts as blockage range of flight Mach numbers without excessive nacelle drag.
and in the worst case may even lead to inlet unstart. Further, The losses due to interaction of shock and boundary layer on
the “bubble” would also increase the heat transfer as well as internal and external surfaces further aggravate the situation.
the wall friction, thereby, deteriorating the flow quality enter- Also, under certain conditions, the flow field becomes highly
ing into the combustion chamber. Therefore, one must have oscillatory.
a deep understanding in order to explore the techniques used The rise of static pressure by decelerating the flow is often
to minimize or eliminate the detrimental effects of boundary critical for the jet engine operation as the nozzle inlet pres-
layer separation. sure affects the exhaust velocity. It is established that 1% loss
Furthermore, the established techniques to start the super- in inlet stagnation pressure eventually leads to about 1–1.5%
sonic inlets such as variable intake geometry, bleeding, cowl loss of engine gross thrust (Intake Aerodynamics by J. Sed-
deflection, micro-vortex generators, etc., are not directly don and E. Goldsmith 1999). Thus, an efficient design of a
applicable to hypersonic inlets. The large temperature diffuser is quite crucial for the whole engine operation. Since
gradients present at hypersonic Mach numbers cause severe the maximum rise in static pressure is the isentropic stagna-
structural problems in any complex mechanical control tion pressure and therefore, it is highly desirable to have a
system requiring an efficient cooling mechanism. shock-free diffuser operation.
Furthermore, in supersonic wind tunnels, the normal shock
wave located in the test section should be pushed through the
A.4 Supersonic Diffusers second throat to minimize the pressure losses caused due to
compression front. This is achieved either by increasing the
It is known that for supersonic flow decreasing area of a duct operating stagnation pressure or by temporarily increasing the
will result in deceleration of the flow. This concept is exploited second throat area. In supersonic inlets, the position of shock
in the form of second throat in supersonic wind tunnels, where at the throat is achieved either by momentarily overspeeding
a convergent–divergent duct immediately follows the test sec- the inlet or by using the variable area geometry diffuser. At
tion. As the flow passes through the duct, it decelerates due this stage, we must realize that the positioning of the shock is
to formation of a normal shock. In order to keep the pressure an independent phenomenon and is not affected by the bound-
losses to a minimum, it is usual to position the shock ahead of ary layer at the wall. Therefore, in the foregoing discussion,
the second throat where the Mach number is slightly greater we will neglect the effects of boundary layer and will investi-
than one. Theoretically, the operation of a supersonic diffuser gate the starting problem of a convergent–divergent diffuser
assuming the flow to be isentropic and in one dimension.
Ai A th
Aa
(a)
M<1 M<1 M<1 M<1 (e)
(b) M = M’ − dM M<1 M=1 M<1
M<1 M<1 M=1 M<1 (f)
Weak Shock M = M’ + dM M=1 M<1

(c)
M=1 M<1 M=1 M<1 (g)
M = MD M=1 M<1
(d) Strong Shock
Weak Shock
M>1 M<1 M=1 M<1
Flow Spillage
Fig. A.9 Start-up and acceleration of a one-dimensional fixed-geometry convergent–divergent inlet
A.4.1 The Starting Problem where M is the freestream Mach number. From Eq. (A.10), it
is evident that for sufficiently high subsonic speeds, we have
Let us examine the flow through a fixed-geometry inlet which
is being accelerated from subsonic to supersonic speeds in Aa Aa Ai
= <
steps, shown schematically in Fig. A.9. We will assume the A∗ Ath Ath
flow to be isentropic and in one dimension. Thus, the losses
occur only across the shock wave. Despite it being a simpli- Thus, at high subsonic speeds, the streamtube capture area
fied analysis, the results are applicable to actual flow through (Aa ) is less than the inlet lip area (Ai ) and therefore, flow
diffusers where the boundary layer is sucked through wall spillage will occur around the inlet. As the flow further accel-
porosities. erates the freestream Mach number becomes one (M = 1), a
When the low subsonic flow passes through the inlet, it weak shock appears ahead of the inlet, as shown in Fig. A.9c.
undergoes deceleration before entering the inlet, as shown in At sonic or supersonic flight Mach numbers, the spillage
Fig. A.9a. Since the presence of inlet is felt upstream due to mechanism is essentially non-isentropic. In other words, in
subsonic speed, the streamtube capture area (Aa ) is dictated order to “sense” the presence of inlet and the flow around it,
by the downstream conditions. As the flow accelerates, the the spilled air must be reduced to subsonic velocity ahead of
streamtube upstream of the inlet gets adjusted, as shown in the inlet. Consequently, a bow shock stands sufficiently far
Fig. A.9b, consequently, the flow is accelerated to sonic veloc- away from the inlet causing the required spillage (Fig. A.9d).
ity at the minimum area location, i.e., at inlet throat (Ath ). The bow shock formation in this case can be understood as
Under this condition, the mass flow rate through the intake is follows. We will consider the case when the freestream first
limited by choking at the throat and since the flow is assumed attains the supersonic speed without formation of the shock
isentropic, Ath = A∗ . The upstream capture area (Aa ) can be wave. Under this condition, the entire flow at the inlet lip has
expressed as to enter without deviation and thus, the streamtube capture
area (Aa ) will same as the inlet lip area (Ai ). But Fig. A.10
γ+1 shows that at low supersonic Mach numbers the allowable
Aa Aa 1 2 γ−1 2 2(γ−1)
= = 1+ M (A.19) capture area (Aa ), which is limited by choking at At is less
A ∗ Ath M γ+1 2 than Ai . Consequently, there will be an accumulation of mass
Area−ratio
for isentropic flow
10.0
(f)
6.0
Aa Aa Area−ratio
A th for detached shock
A*
2.0 Ai
(a) (g)
Area−ratio
A th
(b) (e)
(c) (d)
1.0
0.8
0.4
0.1 0.2 0.4 0.6 1.0 MD M’ 6.0 10.0
Mach Number
A
Fig. A.10 Variation of A∗ with M for a one-dimensional fixed-geometry supersonic inlet
and a rise in pressure in the inlet. The pressure will be build up In turn, the shock will move upstream and exactly at design
rapidly causing a shock of sufficient strength moved upstream speed the shock will be positioned right at the throat, where the
against the supersonic flow and optimally located to allow the Mach number is just unity. The shock becomes very weak and
required spillage. hence, the isentropic flow condition is attained throughout the
Once the shock is formed the flow becomes non-isentropic inlet.
and thus, when the freestream velocity reaches to design Mach However, in actual flight operations, the shock must be
number (MD ), as schematically shown in Fig. A.9d, the isen- positioned slightly downstream of the throat1 as the opera-
tropic mass flow rate as of previous cases will not pass through tions at design conditions are unstable and a slight decrease
At . This follows from the equation,
which indicates that the in Mach number will push the shock in convergent portion,
choked mass flow rate m̊max through a given area At is and hence, the overspeeding operation will have to be repeated
proportional to p0 and also from the fact that fluid experi- to swallow the shock. Also, because of transient waves, the
ences a stagnation pressure loss in traversing the shock. From shock may be pushed further upstream or in the worst case
Fig. A.10, it can be seen that the inlet lip area (Ai ) is still upstream of inlet tip. Consequently, the losses will be signif-
large and the flow spillage is continued even beyond MD . icantly higher and the entire inlet will experience only sub-
However, once the flow is accelerated to a sufficiently high sonic flow. Furthermore, one must realize that except for mod-
Mach number M , the inlet is able to ingest the entire inci- est design Mach numbers, to swallow the shock at higher
dent mass flow without spillage. At the Mach number slightly speeds substantial overspeeding will be required. Thus, at
less than M , i.e., at M − dM shock will be positioned just high freestream Mach numbers, the flow overspeeding will
the inlet lip as shown in Fig. A.9e, whereas a slight increase not be a feasible solution to the problems associated with the
in Mach number, i.e., for M + dM shock will be pushed to starting of an inlet.
enter the inlet convergence. The shock, however, cannot find a An alternative approach to swallow the shock is to use a
stable position within the inlet convergence and thus, it moves variable geometry inlet at constant flight Mach number. Let
quickly downstream to come to rest within the divergent sec- us analyze the flow in the inlet from one-dimensional point of
tion (Fig. A.9f), an equilibrium position decided by the down- view. Suppose that the inlet is accelerated to the design Mach
stream conditions. Here, an isentropic flow is established to number (MD ) with starting shock present as shown schemat-
∗
the throat andsince the throat becomes choked (A < At ), the ically
in Fig. A.11a. Under this condition,
if the area ratio
Aa Aa
∗ > A will be given by point (f) of Fig. A.10. Ai Ai
area ratio A t Ath is decreased to a value Ath at which entire mass flow
That is, an incident supersonic flow decelerates from Ai to At downstream of the shock can be ingested, the shock will be
and subsequently, accelerates in the inlet divergence and thus,
having attained an isentropic flow throughout till the loca- 1 Keeping the shock, slightly downstream to the throat, maintains the
tion of shock the Mach number is reduced from M to MD . throat Mach number slightly greater than the unity.
(a) A th
A i / A th (f)
M=1 M<1 M=1 M<1
A / A* A i / A’th (d)
(b) 1.0
M>1 M=1 M<1
1.0 MD
M
Fig. A.11 Shock swallowing using variable area supersonic inlet
Low Backpressure Low Backpressure

(a) (b)
Ai
M=M M>M M<1 M=M D

M = M’i M<1
D D
Best Backpressure Best Backpressure
M=M M<1 M=M M = M’i M<1

D D
Fig. A.12 Fixed-geometry diffuser with the normal shock waves inside (a) simple diverging passage and (b) Kantrowitz–Donaldson inlet
swallowed to downstream location of the throat (Fig. A.11b). number at the throat, greater than unity. These requirements
Essentially, the variation in area ratio causes a temporary are achieved through an improved design of inlet called
increase in the throat area from At to At and thereby, establish- Kantrowitz–Donaldson inlet, shown in the Fig. A.12. This
ing an isentropic flow within the inlet convergence. Here, the configuration has the maximum internal convergence that will
throat will experience the Mach number greater than one, i.e., just permit the shock swallowing at MD . The inlet operates
> 1 and relatively a stronger shock forms at the father
Mth under supercritical condition with adequate mass flow rate
downstream location to the throat. If area ratio is brought such that the shock is stabilized slightly downstream of the
back to its original value, the isentropic flow can be achieved throat in a position, where it is unaffected by the upstream
throughout the inlet, consequently, the operating point moves and downstream disturbances (Fig. A.12b).
from (d) to (f) in Fig. A.11. Despite having an advantage in obtaining a stabilized
From the above discussion, it might appear to design a fixed shock system in inlet divergence, the Kantrowitz–Donaldson
geometry inlet so that the shock wave, positioned upstream inlet suffers a major drawback as the local Mach number of
to the inlet lip, would be swallowed just as the inlet is accel- the flow approaching to the compression front will be greater
erated to the design Mach number (MD ). This would, how- than at the throat. Consequently, the shock will be of higher
ever, require a larger throat area and would lead to a Mach strength.
The operating condition in which a shock is positioned temperature may experience a significant decrease in tem-
ahead of the inlet, and where there is a flow spillage, is called perature, and thus, the dissociated products may recombine
subcritical operation. When the shock is just at the inlet lip, themselves. Therefore, the chemical kinetics of recombina-
the operation is called critical operation. However, once the tion process would have a strong influence on the thrust as
shock is swallowed to the inlet divergence, it is termed as well as on the propulsive efficiency of the ramjet.
supercritical operation. To avoid the dissociation and stagnation pressure losses
associated with deceleration of supersonic or hypersonic flow
to subsonic level in ramjets, the concept of supersonic com-
A.5 Supersonic Combustion and Scramjet bustion ramjet (scramjet) is proposed. With combustion at the
Isolator supersonic Mach numbers, the static temperature of fluid is
relatively lower, and hence, the dissociation losses decrease,
The pressure loss associated with the deceleration of a super- since the dissociation of a gas depends on the static tempera-
sonic or hypersonic flow passing through the shock waves ture. On the other hand, heat transfer through the wall depends
may be substantial. Thus, the losses incurred in the ramjet mainly on the stagnation temperature, and thus, the wall cool-
combustor even at subsonic Mach numbers will be high. Fur- ing problem is not eliminated by performing the combus-
ther, if the ramjet engines are used to fly at hypersonic Mach tion at supersonic Mach numbers. As a remedy, the cooling
numbers, in addition to pressure losses caused by the shocks, caused by liquid hydrogen on its way from the fuel tank to the
the static temperature of the flow would also be extremely engine has been predicted as a means of keeping the engine
high. The high temperature not only makes the cooling of the and vehicle cool enough to sustain the period of hypersonic
vehicle difficult, but it also leads to the dissociation of oxygen flight.
and nitrogen gases in the air, which in turn increases the com- Furthermore, the supersonic combustion requires the mix-
bustion losses. Furthermore, for hypersonic flights (M > 8), ing of injected fuel with supersonic airstream without exces-
the temperature of the air in the combustor will be quite large sive losses due to shock waves. In literature, it has been
and will become a strong function of pressure. However, the demonstrated that the hydrogen fuel in its gaseous state when
rise in pressure is quite advantageous as it suppresses the injected at an angle to the airflow direction, the combined per-
dissociation of mixture in the combustion chamber despite turbations of fuel injection, fuel–air mixing and combustion
an increase in temperature. Consequently, the temperature of results in a complex shock cell pattern that is associated with
the combustion products is likewise pressure dependent. It is rather gradual pressure rise. Subsequently, after the initial
found that for a Mach 10 flight at the combustion pressure pressure rise, the mixture is accelerated to supersonic Mach
of 10 atm, there is no temperature rise due to combustion, numbers in the convergent–divergent nozzle. Sometimes, the
because all the heat released in the combustion is absorbed pressure rise associated with shock cell train may be high
in dissociation (Mechanics and Thermodynamics of Propul- enough to cause separation of the incoming boundary layer.
sion, Hill and Peterson, Addison-Wesley, 1992). Indeed, at These flow-induced disturbances may travel upstream and on
sufficiently high Mach numbers the temperature of the com- reaching the intake they can seriously distort the flow to such
bustion products may be even lower than that of the incoming an extent as to even cause breakdown of the flow through the
airstream. The low Mach number at which fuel and air are inlet. Therefore, it is essential to prevent the upstream prop-
converted into dissociation products may show that there is agation of perturbation in the combustor. This is achieved
sufficient residence time available in the combustion cham- by a constant area duct, positioned between the intake exit
ber to attain an equilibrium composition. However, when the and combustor inlet, in which a series of oblique shocks are
burnt gases are subsequently expanded in the nozzle, there’s present. These shock waves are relatively weak and thus, they
no guarantee to ascertain an equilibrium composition in suc- do not cause any significant pressure loss. However, they
cession with each step of temperature and pressure reduction, cause a considerable increase in static pressure of incoming
owing to the rapidity of the expansion process. If the expan- flow, which is an advantage from the combustion performance
sion is extremely rapid the mixture may be effectively frozen point of view. Also, these oblique shocks interact with the
with the initial high-temperature composition. Thus, it is evi- boundary layer on the constant area duct, resulting in a com-
dent that for flight Mach numbers of the order of 10, only an plex wave pattern which prevents the upstream propagation
insignificant amount of combustion energy of the fuel would of the perturbation from the combustor. Thus, the constant
be available for the acceleration of the combustion products area duct virtually isolates the inlet from getting disturbed by
to generate thrust. Additionally, due to the rapid accelera- the instabilities from the combustor and hence, it is referred
tion encountered in the nozzle, combustion products at high to as isolator.
The Uncertainty Analysis
B
B.1 Introduction • Each uncertainty component is quantified by a standard

deviation.
Uncertainty analysis or error analysis is essentially the • All biases are assumed to be corrected and any uncertainty
procedure which involves estimating the uncertainty asso- is the uncertainty of the correction.
ciated with the measured data in any experiment. The ISO • Zero corrections are allowed if the bias cannot be corrected
definition of uncertainty is “Parameter, associated with the and an uncertainty is assessed.
result of a measurement, that characterizes the dispersion • All uncertainty intervals are symmetric.
of the values that could reasonably be attributed to the
measurand”. As per ISO, the error components are classified into two major
The quality of the performed experiments and the obtained groups, depending upon the source of data. Type A compo-
data will be considered acceptable, based on proper docu- nents, which are evaluated by statistical methods and Type B
menting of the experimental data along with the uncertainty. components evaluated by other means (or in other laborato-
In any measured data, we attempt to find the true value by ries), which can be applied to both random error and bias. The
making multiple measurements or by using different meth- sources of uncertainty evaluated by Type B components are
ods of measurement. We may obtain slightly different values as follows:
in each measurement. This range of measured values includes
a true value, and it is represented as • Reference standards calibrated by another laboratory,
• Physical constants used in the calculation of the reported
Measured value = Best estimate ± Uncertainty value,
• Environmental effects that cannot be sampled,
Any uncertainty analysis must include both the accuracy and • Possible configuration/geometry misalignment in the
precision in the measurement. The difference between pre- instrument, and
cision and accuracy lies in the fact that the former indicates • Lack of resolution of the instrument.
the quality of the measurement without any correctness in
the measured value, whereas the latter estimates the devia-
tion of the measured value from the expected value. Further, B.2 Uncertainty Estimation in the Static and
the relative uncertainty accounts for precision, while the rel- Stagnation Pressures
ative error accounts for the accuracy in measurement. Rela-
tive uncertainty is given by the absolute value of the ratio of The general procedure for estimating the uncertainty in a cal-
uncertainty to the measured quantity. Furthermore, the rela- culated quantity using the measured data is discussed in this
tive error is the ratio of difference between the measured value section. In the experimental studies, discussed in Sects. 14.8.1
and expected value to the expected value. According to ISO and 14.8.2, the wall static pressure and the settling chamber
standard, the basic tenets in uncertainty quantification are as stagnation pressure are calculated using the voltage output of
follows: pressure transducers.

B.2.1 Standard Deviation and the Uncertainty B.2.2 Propagation of the Uncertainty
We know that the repeated measurements of a variable (x), Suppose we are interested in estimating the quantity F, which
will give slightly different values during each measurement. is a function of a number of variables such as x, y, z, etc. Thus,
A step-by-step procedure using statistical analysis, to deter- the overall uncertainty in F can be calculated by the measured
mine the standard deviation and uncertainty in the measured uncertainties in x, y, z, ..., as σx , σy , σz , ..., respectively.
variable is given below. In the experimental studies, discussed in Sects. 14.8.1 and
Generally, the average of a measured dataset can be con- 14.8.2, it is seen that the static and stagnation pressures are
sidered as the best estimate of the ideal value. The average or the functions of a single variable, i.e., the measured transducer
mean value of the repeatedly measured variable is given as output voltage V, i.e.,
X1 + X2 + ... + XN V = mP + V0
X= (B.1) (B.5)
N
This mean value is not accurate since there will still be some If the transducer voltage constant is V0 and the slope is m,
systematic errors due to the measuring equipment. Since it is the pressure can be written as
quite tedious to calibrate the equipment perfectly, we express
the uncertainty in the average value by calculating the average
deviation. V − V0
P= = VC1 + C2 (B.6)
m
Differentiating above, we get

|X1 − X| + |X2 − X| + ... + |XN − X|
d̄ = (B.2)
N
The standard deviation is the most common way to charac- dP

= C1 (B.7)
terize the spread of a dataset. This can be calculated as dV
Thus, the propagation of uncertainty is measured as

(X1 − X)2 + (X2 − X)2 + ... + (XN − X)2
S= (B.3) dP
N−1
σP = σV (B.8)
dV
when we report the mean value of N measurements, the uncer-
tainty associated with this mean value is the standard devia- In addition, the measured static or stagnation pressure is given
tion of the mean value, often called the standard error or the by
uncertainty in measurement.
= P ± σP (B.9)
S
σX = √ (B.4)
N
Appendix B: The Uncertainty Analysis 409
Listing B.1 The MATLAB program for standard deviation and

wall pressure calculations.
1 clc ; clear all ;
2 N = 5; % n u m b e r of test runs
3 str1 = ’ D :\ Text Data \1. Plain \ T =8 mm \ Text \ ’;
4 str2 = ’ D :\ Text Data \1. Plain \ T =8 mm \ Plot \ ’;
5 str3 = ’ Ori_XvsV ’;
6 str4 = ’ 001 ’;
7 str5 = ’. txt ’;
8 str6 = ’ _WError ’;
9 str7 = ’ 002 ’;
10 str8 = ’ 003 ’;
11 str9 = ’ 004 ’;
12 str10 = ’ 005 ’;
13 str11 = ’ XvsP ’;
14 t e x t n a m e = s t r c a t ( str1 , str11 , str6 , str5 );
15 f i l e n a m e 1 = s t r c a t ( str1 , str3 , str4 , str5 );
20 M1 = load ( f i l e n a m e 1 );
21 l e n g t h ( M1 );
22 x = M1 (: ,1);
23 v1 = M1 (: ,2);
24 M2 = load ( f i l e n a m e 2 );
25 l e n g t h ( M2 );
26 v2 = M2 (: ,2);
27 M3 = load ( f i l e n a m e 3 );
28 l e n g t h ( M3 );
29 v3 = M3 (: ,2);
30 M4 = load ( f i l e n a m e 4 );
31 l e n g t h ( M4 );
32 v4 = M4 (: ,2);
33 M5 = load ( f i l e n a m e 5 );
34 l e n g t h ( M5 );
35 v5 = M5 (: ,2);
36 % average
37 v = ( v1 + v2 + v3 + v4 + v5 )/ N
38 % d e v i a t i o n from the a v e r a g e
39 d1 =( v1 - v )./ v
40 d2 =( v2 - v )./ v
41 d3 =( v3 - v )./ v
42 d4 =( v4 - v )./ v
43 d5 =( v5 - v )./ v
44 % average deviation
45 ad = ( abs ( d1 )+ abs ( d2 )+ abs ( d3 )+ abs ( d4 )+ abs ( d5 ))/ N
46 % standard deviation
47 sd = sqrt ((( d1 .* d1 )+( d2 .* d2 )+( d3 .* d3 )...
48 +( d4 .* d4 )+( d5 .* d5 ))/( N -1))
49 % s t a n d a r d error or u n c e r t a i n i t y in v o l t a g e
50 se_v = sd / sqrt ( N )
51 % C o n v e r s i o n of v o l t a g e to p r e s s u r e
52 p =0.6 v
53 % s t a n d a r d error or u n c e r t a i n i t y in p r e s s u r e
54 dp_dv =0.6
55 se_P =( abs ( dP_dV )).* se_v
56 C =[ x ; v ; sd ; se ; p ; sep ]
57 f i l e I D = f o p e n ( textname , ’w ’);
58 f p r i n t f ( fileID , ’%4 s %6 s %6 s %6 s %6 s %6 s \ r \n ’ , ’x ’ , ’v ’ ,...
59 ’ sd ’ , ’ se ’ , ’p ’ , ’ sep ’);
60 f p r i n t f ( fileID , ’%1.4 f %1.4 f %1.4 f %1.4 f %1.4 f %1.4 f \ r \ n ’ , C );
61 f c l o s e ( f i l e I D );
The Standard Atmosphere
C

The properties of International Standard Atmosphere (ISA) h (m) hG (m) T (K) a ms−1 p (Pa) ρ kgm−3
are tabulated in SI units. 725 725.08 283.44 337.5 92913.24 1.142
750 750.09 283.28 337.4 92633.61 1.139
1. The geopotential (h) and geometric (hG ) altitudes are mea- 775 775.09 283.11 337.31 92354.66 1.136
800 800.1 282.95 337.21 92076.39 1.134
sured in meters (m). 825 825.11 282.79 337.11 91798.8 1.131
2. The temperature (T) values are given in Kelvin (K). 850 850.11 282.63 337.02 91521.88 1.128
3. The speed of sound (a) is given in ms−1 . 875 875.12 282.46 336.92 91245.65 1.125
4. The pressure (p) data are in Pascals (Pa). 900 900.13 282.3 336.82 90970.09 1.123
5. The density (ρ) values are represented in kgm−3 . 925 925.13 282.14 336.73 90695.2 1.12
950 950.14 281.98 336.63 90420.98 1.117
975 975.15 281.81 336.53 90147.44 1.114
1000 1000.16 281.65 336.43 89874.57 1.112

h (m) hG (m) T (K) a ms−1 p (Pa) ρ kgm−3 1025 1025.16 281.49 336.34 89602.37 1.109
0 0 288.15 340.29 101325 1.225 1050 1050.17 281.33 336.24 89330.83 1.106
25 25 287.99 340.2 101025.03 1.222 1075 1075.18 281.16 336.14 89059.97 1.103
50 50 287.83 340.1 100725.78 1.219 1100 1100.19 281 336.05 88789.76 1.101
75 75 287.66 340.01 100427.25 1.216 1125 1125.2 280.84 335.95 88520.22 1.098
100 100 287.5 339.91 100129.44 1.213 1150 1150.21 280.68 335.85 88251.35 1.095
125 125 287.34 339.81 99832.34 1.21 1175 1175.22 280.51 335.75 87983.13 1.093
150 150 287.18 339.72 99535.96 1.207 1200 1200.23 280.35 335.66 87715.58 1.09
175 175 287.01 339.62 99240.29 1.205 1225 1225.24 280.19 335.56 87448.69 1.087
200 200.01 286.85 339.53 98945.33 1.202 1250 1250.25 280.03 335.46 87182.45 1.085
225 225.01 286.69 339.43 98651.08 1.199 1275 1275.26 279.86 335.36 86916.87 1.082
250 250.01 286.53 339.33 98357.54 1.196 1300 1300.27 279.7 335.27 86651.95 1.079
275 275.01 286.36 339.24 98064.7 1.193 1325 1325.28 279.54 335.17 86387.68 1.077
300 300.01 286.2 339.14 97772.58 1.19 1350 1350.29 279.38 335.07 86124.06 1.074
325 325.02 286.04 339.04 97481.16 1.187 1375 1375.3 279.21 334.98 85861.1 1.071
350 350.02 285.88 338.95 97190.44 1.184 1400 1400.31 279.05 334.88 85598.78 1.069
375 375.02 285.71 338.85 96900.42 1.182 1425 1425.32 278.89 334.78 85337.12 1.066
400 400.03 285.55 338.76 96611.11 1.179 1450 1450.33 278.73 334.68 85076.1 1.063
425 425.03 285.39 338.66 96322.5 1.176 1475 1475.34 278.56 334.58 84815.73 1.061
450 450.03 285.23 338.56 96034.58 1.173 1500 1500.35 278.4 334.49 84556 1.058
475 475.04 285.06 338.47 95747.36 1.17 1525 1525.37 278.24 334.39 84296.92 1.055
500 500.04 284.9 338.37 95460.84 1.167 1550 1550.38 278.08 334.29 84038.49 1.053
525 525.04 284.74 338.27 95175.01 1.164 1575 1575.39 277.91 334.19 83780.69 1.05
550 550.05 284.58 338.18 94889.88 1.162 1600 1600.4 277.75 334.1 83523.54 1.048
575 575.05 284.41 338.08 94605.44 1.159 1625 1625.41 277.59 334 83267.02 1.045
600 600.06 284.25 337.98 94321.68 1.156 1650 1650.43 277.43 333.9 83011.15 1.042
625 625.06 284.09 337.89 94038.62 1.153 1675 1675.44 277.26 333.8 82755.91 1.04
650 650.07 283.93 337.79 93756.25 1.15 1700 1700.45 277.1 333.71 82501.3 1.037
675 675.07 283.76 337.69 93474.56 1.148
700 700.08 283.6 337.6 93193.56 1.145

412 Appendix C: The Standard Atmosphere

h (m) hG (m) T (K) a ms−1 p (Pa) ρ kgm−3 h (m) hG (m) T (K) a ms−1 p (Pa) ρ kgm−3
1725 1725.47 276.94 333.61 82247.33 1.035 3075 3076.48 268.16 328.28 69442.46 0.902
1750 1750.48 276.78 333.51 81994 1.032 3100 3101.51 268 328.18 69221.58 0.9
1775 1775.49 276.61 333.41 81741.3 1.029 3125 3126.53 267.84 328.08 69001.26 0.897
1800 1800.51 276.45 333.31 81489.22 1.027 3150 3151.56 267.68 327.98 68781.52 0.895
1825 1825.52 276.29 333.22 81237.78 1.024 3175 3176.58 267.51 327.88 68562.34 0.893
1850 1850.54 276.13 333.12 80986.97 1.022 3200 3201.61 267.35 327.78 68343.72 0.891
1875 1875.55 275.96 333.02 80736.78 1.019 3225 3226.63 267.19 327.68 68125.67 0.888
1900 1900.57 275.8 332.92 80487.22 1.017 3250 3251.66 267.03 327.58 67908.19 0.886
1925 1925.58 275.64 332.82 80238.28 1.014 3275 3276.68 266.86 327.48 67691.26 0.884
1950 1950.6 275.48 332.73 79989.97 1.012 3300 3301.71 266.7 327.38 67474.9 0.881
1975 1975.61 275.31 332.63 79742.28 1.009 3325 3326.74 266.54 327.28 67259.1 0.879
2000 2000.63 275.15 332.53 79495.22 1.006 3350 3351.76 266.38 327.18 67043.86 0.877
2025 2025.64 274.99 332.43 79248.77 1.004 3375 3376.79 266.21 327.08 66829.17 0.875
2050 2050.66 274.83 332.33 79002.94 1.001 3400 3401.82 266.05 326.98 66615.05 0.872
1875 1875.55 275.96 333.02 80736.78 1.019 3425 3426.84 265.89 326.88 66401.47 0.87
1900 1900.57 275.8 332.92 80487.22 1.017 3450 3451.87 265.73 326.78 66188.46 0.868
1925 1925.58 275.64 332.82 80238.28 1.014 3475 3476.9 265.56 326.68 65975.99 0.865
1950 1950.6 275.48 332.73 79989.97 1.012 3500 3501.92 265.4 326.58 65764.08 0.863
1975 1975.61 275.31 332.63 79742.28 1.009 3525 3526.95 265.24 326.48 65552.73 0.861
2000 2000.63 275.15 332.53 79495.22 1.006 3550 3551.98 265.08 326.38 65341.92 0.859
2025 2025.64 274.99 332.43 79248.77 1.004 3575 3577.01 264.91 326.28 65131.66 0.857
2050 2050.66 274.83 332.33 79002.94 1.001 3600 3602.04 264.75 326.18 64921.95 0.854
2075 2075.68 274.66 332.23 78757.73 0.999 3625 3627.06 264.59 326.08 64712.78 0.852
2100 2100.69 274.5 332.14 78513.14 0.996 3650 3652.09 264.43 325.98 64504.16 0.85
2125 2125.71 274.34 332.04 78269.16 0.994 3675 3677.12 264.26 325.88 64296.09 0.848
2150 2150.73 274.18 331.94 78025.79 0.991 3700 3702.15 264.1 325.78 64088.56 0.845
2175 2175.74 274.01 331.84 77783.04 0.989 3725 3727.18 263.94 325.68 63881.58 0.843
2200 2200.76 273.85 331.74 77540.9 0.986 3750 3752.21 263.78 325.58 63675.13 0.841
2225 2225.78 273.69 331.64 77299.37 0.984 3775 3777.24 263.61 325.48 63469.23 0.839
2250 2250.79 273.53 331.55 77058.46 0.981 3800 3802.27 263.45 325.38 63263.86 0.837
2275 2275.81 273.36 331.45 76818.14 0.979 3825 3827.3 263.29 325.28 63059.04 0.834
2300 2300.83 273.2 331.35 76578.44 0.976 3850 3852.33 263.13 325.18 62854.75 0.832
2325 2325.85 273.04 331.25 76339.34 0.974 3875 3877.36 262.96 325.08 62650.99 0.83
2350 2350.87 272.88 331.15 76100.85 0.972 3900 3902.39 262.8 324.98 62447.78 0.828
2375 2375.89 272.71 331.05 75862.96 0.969 3925 3927.42 262.64 324.88 62245.09 0.826
2400 2400.9 272.55 330.95 75625.68 0.967 3950 3952.45 262.48 324.78 62042.94 0.823
2425 2425.92 272.39 330.86 75388.99 0.964 3975 3977.48 262.31 324.68 61841.32 0.821
2450 2450.94 272.23 330.76 75152.91 0.962 4000 4002.51 262.15 324.58 61640.24 0.819
2475 2475.96 272.06 330.66 74917.42 0.959 4025 4027.54 261.99 324.48 61439.68 0.817
2500 2500.98 271.9 330.56 74682.53 0.957 4050 4052.58 261.83 324.38 61239.65 0.815
2525 2526 271.74 330.46 74448.24 0.954 4075 4077.61 261.66 324.28 61040.15 0.813
2550 2551.02 271.58 330.36 74214.55 0.952 4100 4102.64 261.5 324.18 60841.17 0.811
2575 2576.04 271.41 330.26 73981.45 0.95 4125 4127.67 261.34 324.08 60642.72 0.808
2600 2601.06 271.25 330.16 73748.94 0.947 4150 4152.71 261.18 323.97 60444.8 0.806
2625 2626.08 271.09 330.07 73517.02 0.945 4175 4177.74 261.01 323.87 60247.39 0.804
2650 2651.1 270.93 329.97 73285.7 0.942 4200 4202.77 260.85 323.77 60050.52 0.802
2675 2676.12 270.76 329.87 73054.96 0.94 4225 4227.8 260.69 323.67 59854.16 0.8
2700 2701.14 270.6 329.77 72824.81 0.938 4250 4252.84 260.53 323.57 59658.32 0.798
2725 2726.17 270.44 329.67 72595.25 0.935 4275 4277.87 260.36 323.47 59463 0.796
2750 2751.19 270.28 329.57 72366.28 0.933 4300 4302.9 260.2 323.37 59268.2 0.794
2775 2776.21 270.11 329.47 72137.89 0.93 4325 4327.94 260.04 323.27 59073.92 0.791
2800 2801.23 269.95 329.37 71910.09 0.928 4350 4352.97 259.88 323.17 58880.15 0.789
2825 2826.25 269.79 329.27 71682.87 0.926 4375 4378.01 259.71 323.07 58686.9 0.787
2850 2851.28 269.63 329.17 71456.23 0.923 4400 4403.04 259.55 322.97 58494.16 0.785
2875 2876.3 269.46 329.07 71230.17 0.921 4425 4428.08 259.39 322.86 58301.93 0.783
2900 2901.32 269.3 328.98 71004.69 0.919 4450 4453.11 259.23 322.76 58110.22 0.781
2925 2926.34 269.14 328.88 70779.79 0.916 4475 4478.15 259.06 322.66 57919.02 0.779
2950 2951.37 268.98 328.78 70555.47 0.914 4500 4503.18 258.9 322.56 57728.32 0.777
2975 2976.39 268.81 328.68 70331.72 0.911 4525 4528.22 258.74 322.46 57538.14 0.775
3000 3001.41 268.65 328.58 70108.54 0.909 4550 4553.25 258.58 322.36 57348.46 0.773
3025 3026.44 268.49 328.48 69885.95 0.907 4575 4578.29 258.41 322.26 57159.29 0.771
3050 3051.46 268.33 328.38 69663.92 0.904 4600 4603.32 258.25 322.16 56970.63 0.769
Appendix C: The Standard Atmosphere 413

4625 4628.36 258.09 322.05 56782.47 0.766 6100 6105.85 248.5 316.02 46537.67 0.652
4650 4653.4 257.93 321.95 56594.81 0.764 6125 6130.89 248.34 315.91 46377.94 0.651
4675 4678.43 257.76 321.85 56407.66 0.762 6150 6155.94 248.18 315.81 46218.66 0.649
4700 4703.47 257.6 321.75 56221 0.76 6175 6180.99 248.01 315.71 46059.82 0.647
4725 4728.51 257.44 321.65 56034.85 0.758 6200 6206.04 247.85 315.6 45901.43 0.645
4750 4753.54 257.28 321.55 55849.2 0.756 6225 6231.09 247.69 315.5 45743.48 0.643
4775 4778.58 257.11 321.45 55664.04 0.754 6250 6256.14 247.53 315.39 45585.96 0.642
4800 4803.62 256.95 321.34 55479.39 0.752 6275 6281.19 247.36 315.29 45428.89 0.64
4825 4828.66 256.79 321.24 55295.23 0.75 6300 6306.24 247.2 315.19 45272.25 0.638
4850 4853.69 256.63 321.14 55111.56 0.748 6325 6331.29 247.04 315.08 45116.06 0.636
4875 4878.73 256.46 321.04 54928.39 0.746 6350 6356.34 246.88 314.98 44960.29 0.634
4900 4903.77 256.3 320.94 54745.71 0.744 6375 6381.39 246.71 314.88 44804.97 0.633
4925 4928.81 256.14 320.83 54563.53 0.742 6400 6406.44 246.55 314.77 44650.08 0.631
4950 4953.85 255.98 320.73 54381.83 0.74 6425 6431.49 246.39 314.67 44495.62 0.629
4975 4978.89 255.81 320.63 54200.63 0.738 6450 6456.54 246.23 314.57 44341.6 0.627
5000 5003.93 255.65 320.53 54019.91 0.736 6475 6481.59 246.06 314.46 44188.01 0.626
5025 5028.97 255.49 320.43 53839.69 0.734 6500 6506.64 245.9 314.36 44034.85 0.624
5050 5054.01 255.33 320.33 53659.95 0.732 6525 6531.69 245.74 314.25 43882.12 0.622
5075 5079.05 255.16 320.22 53480.69 0.73 6550 6556.74 245.58 314.15 43729.81 0.62
5100 5104.09 255 320.12 53301.92 0.728 6575 6581.79 245.41 314.05 43577.94 0.619
5125 5129.13 254.84 320.02 53123.64 0.726 6600 6606.84 245.25 313.94 43426.5 0.617
5150 5154.17 254.68 319.92 52945.84 0.724 6625 6631.9 245.09 313.84 43275.48 0.615
5175 5179.21 254.51 319.82 52768.52 0.722 6650 6656.95 244.93 313.73 43124.88 0.613
5200 5204.25 254.35 319.71 52591.68 0.72 6675 6682 244.76 313.63 42974.71 0.612
5225 5229.29 254.19 319.61 52415.33 0.718 6700 6707.05 244.6 313.53 42824.97 0.61
5250 5254.33 254.03 319.51 52239.45 0.716 6725 6732.11 244.44 313.42 42675.65 0.608
5275 5279.37 253.86 319.41 52064.05 0.714 6750 6757.16 244.28 313.32 42526.75 0.606
5300 5304.41 253.7 319.3 51889.13 0.713 6775 6782.21 244.11 313.21 42378.27 0.605
5325 5329.45 253.54 319.2 51714.68 0.711 6800 6807.27 243.95 313.11 42230.21 0.603
5350 5354.5 253.38 319.1 51540.71 0.709 6825 6832.32 243.79 313 42082.57 0.601
5375 5379.54 253.21 319 51367.21 0.707 6850 6857.37 243.63 312.9 41935.34 0.6
5400 5404.58 253.05 318.9 51194.19 0.705 6875 6882.43 243.46 312.8 41788.54 0.598
5425 5429.62 252.89 318.79 51021.63 0.703 6900 6907.48 243.3 312.69 41642.15 0.596
5450 5454.67 252.73 318.69 50849.55 0.701 6925 6932.54 243.14 312.59 41496.18 0.595
5475 5479.71 252.56 318.59 50677.94 0.699 6950 6957.59 242.98 312.48 41350.62 0.593
5500 5504.75 252.4 318.49 50506.8 0.697 6975 6982.64 242.81 312.38 41205.47 0.591
5525 5529.8 252.24 318.38 50336.13 0.695 7000 7007.7 242.65 312.27 41060.74 0.59
5550 5554.84 252.08 318.28 50165.92 0.693 7025 7032.75 242.49 312.17 40916.42 0.588
5575 5579.88 251.91 318.18 49996.19 0.691 7050 7057.81 242.33 312.06 40772.51 0.586
5600 5604.93 251.75 318.08 49826.91 0.689 7075 7082.87 242.16 311.96 40629.02 0.584
5625 5629.97 251.59 317.97 49658.1 0.688 7100 7107.92 242 311.86 40485.93 0.583
5650 5655.02 251.43 317.87 49489.76 0.686 7125 7132.98 241.84 311.75 40343.25 0.581
5675 5680.06 251.26 317.77 49321.87 0.684 7150 7158.03 241.68 311.65 40200.97 0.579
5700 5705.1 251.1 317.66 49154.45 0.682 7175 7183.09 241.51 311.54 40059.1 0.578
5725 5730.15 250.94 317.56 48987.49 0.68 7200 7208.15 241.35 311.44 39917.64 0.576
5750 5755.19 250.78 317.46 48820.99 0.678 7225 7233.2 241.19 311.33 39776.59 0.575
5775 5780.24 250.61 317.36 48654.94 0.676 7250 7258.26 241.03 311.23 39635.93 0.573
5800 5805.28 250.45 317.25 48489.36 0.674 7275 7283.32 240.86 311.12 39495.68 0.571
5825 5830.33 250.29 317.15 48324.23 0.673 7300 7308.37 240.7 311.02 39355.84 0.57
5850 5855.38 250.13 317.05 48159.56 0.671 7325 7333.43 240.54 310.91 39216.39 0.568
5875 5880.42 249.96 316.94 47995.34 0.669 7350 7358.49 240.38 310.81 39077.34 0.566
5900 5905.47 249.8 316.84 47831.57 0.667 7375 7383.55 240.21 310.7 38938.7 0.565
5925 5930.52 249.64 316.74 47668.26 0.665 7400 7408.61 240.05 310.6 38800.45 0.563
5950 5955.56 249.48 316.63 47505.4 0.663 7425 7433.66 239.89 310.49 38662.6 0.561
5975 5980.61 249.31 316.53 47342.99 0.662 7450 7458.72 239.73 310.39 38525.14 0.56
6000 6005.66 249.15 316.43 47181.03 0.66 7475 7483.78 239.56 310.28 38388.09 0.558
6025 6030.7 248.99 316.33 47019.52 0.658 7500 7508.84 239.4 310.18 38251.42 0.557
6050 6055.75 248.83 316.22 46858.45 0.656 7525 7533.9 239.24 310.07 38115.16 0.555
6075 6080.8 248.66 316.12 46697.84 0.654 7550 7558.96 239.08 309.96 37979.28 0.553

7575 7584.02 238.91 309.86 37843.8 0.552 8950 8962.59 229.98 304.01 30971.81 0.469
7600 7609.08 238.75 309.75 37708.71 0.55 8975 8987.66 229.81 303.9 30856.96 0.468
7625 7634.14 238.59 309.65 37574.01 0.549 9000 9012.73 229.65 303.79 30742.46 0.466
7650 7659.2 238.43 309.54 37439.7 0.547 9025 9037.8 229.49 303.69 30628.3 0.465
7675 7684.26 238.26 309.44 37305.78 0.545 9050 9062.87 229.33 303.58 30514.48 0.464
7700 7709.32 238.1 309.33 37172.24 0.544 9075 9087.95 229.16 303.47 30401.01 0.462
7725 7734.38 237.94 309.23 37039.1 0.542 9100 9113.02 229 303.36 30287.87 0.461
7750 7759.44 237.78 309.12 36906.34 0.541 9125 9138.09 228.84 303.26 30175.08 0.459
7775 7784.5 237.61 309.02 36773.96 0.539 9150 9163.16 228.68 303.15 30062.63 0.458
7800 7809.56 237.45 308.91 36641.98 0.538 9175 9188.23 228.51 303.04 29950.52 0.457
7825 7834.62 237.29 308.8 36510.37 0.536 9200 9213.3 228.35 302.93 29838.75 0.455
7850 7859.68 237.13 308.7 36379.15 0.534 9225 9238.38 228.19 302.82 29727.31 0.454
7875 7884.75 236.96 308.59 36248.31 0.533 9250 9263.45 228.03 302.72 29616.21 0.452
7900 7909.81 236.8 308.49 36117.85 0.531 9275 9288.52 227.86 302.61 29505.45 0.451
7925 7934.87 236.64 308.38 35987.77 0.53 9300 9313.6 227.7 302.5 29395.03 0.45
7950 7959.93 236.48 308.27 35858.07 0.528 9325 9338.67 227.54 302.39 29284.94 0.448
7975 7985 236.31 308.17 35728.75 0.527 9350 9363.74 227.38 302.28 29175.18 0.447
8000 8010.06 236.15 308.06 35599.81 0.525 9375 9388.82 227.21 302.18 29065.76 0.446
8025 8035.12 235.99 307.96 35471.25 0.524 9400 9413.89 227.05 302.07 28956.67 0.444
8050 8060.18 235.83 307.85 35343.06 0.522 9425 9438.96 226.89 301.96 28847.91 0.443
8075 8085.25 235.66 307.74 35215.24 0.521 9450 9464.04 226.73 301.85 28739.48 0.442
8100 8110.31 235.5 307.64 35087.81 0.519 9475 9489.11 226.56 301.74 28631.38 0.44
8125 8135.38 235.34 307.53 34960.74 0.518 9500 9514.19 226.4 301.64 28523.62 0.439
8150 8160.44 235.18 307.43 34834.05 0.516 9525 9539.26 226.24 301.53 28416.18 0.438
8175 8185.5 235.01 307.32 34707.73 0.514 9550 9564.34 226.08 301.42 28309.07 0.436
8200 8210.57 234.85 307.21 34581.78 0.513 9575 9589.41 225.91 301.31 28202.28 0.435
8225 8235.63 234.69 307.11 34456.2 0.511 9600 9614.49 225.75 301.2 28095.82 0.434
8250 8260.7 234.53 307 34330.99 0.51 9625 9639.56 225.59 301.09 27989.69 0.432
8275 8285.76 234.36 306.89 34206.15 0.508 9650 9664.64 225.43 300.99 27883.88 0.431
8300 8310.83 234.2 306.79 34081.68 0.507 9675 9689.71 225.26 300.88 27778.4 0.43
8325 8335.89 234.04 306.68 33957.57 0.505 9700 9714.79 225.1 300.77 27673.24 0.428
8350 8360.96 233.88 306.58 33833.83 0.504 9725 9739.87 224.94 300.66 27568.4 0.427
8375 8386.02 233.71 306.47 33710.46 0.502 9750 9764.94 224.78 300.55 27463.89 0.426
8400 8411.09 233.55 306.36 33587.45 0.501 9775 9790.02 224.61 300.44 27359.69 0.424
8425 8436.16 233.39 306.26 33464.8 0.5 9800 9815.1 224.45 300.33 27255.82 0.423
8450 8461.22 233.23 306.15 33342.52 0.498 9825 9840.17 224.29 300.23 27152.26 0.422
8475 8486.29 233.06 306.04 33220.6 0.497 9850 9865.25 224.13 300.12 27049.03 0.42
8500 8511.36 232.9 305.94 33099.04 0.495 9875 9890.33 223.96 300.01 26946.11 0.419
8525 8536.42 232.74 305.83 32977.84 0.494 9900 9915.41 223.8 299.9 26843.51 0.418
8550 8561.49 232.58 305.72 32857 0.492 9925 9940.49 223.64 299.79 26741.23 0.417
8575 8586.56 232.41 305.62 32736.52 0.491 9950 9965.56 223.48 299.68 26639.26 0.415
8600 8611.62 232.25 305.51 32616.4 0.489 9975 9990.64 223.31 299.57 26537.61 0.414
8625 8636.69 232.09 305.4 32496.63 0.488 10000 10015.72 223.15 299.46 26436.27 0.413
8650 8661.76 231.93 305.29 32377.22 0.486 10025 10040.8 222.99 299.35 26335.24 0.411
8675 8686.83 231.76 305.19 32258.17 0.485 10050 10065.88 222.83 299.25 26234.53 0.41
8700 8711.9 231.6 305.08 32139.47 0.483 10075 10090.96 222.66 299.14 26134.13 0.409
8725 8736.97 231.44 304.97 32021.13 0.482 10100 10116.04 222.5 299.03 26034.04 0.408
8750 8762.03 231.28 304.87 31903.13 0.481 10125 10141.12 222.34 298.92 25934.26 0.406
8775 8787.1 231.11 304.76 31785.49 0.479 10150 10166.2 222.18 298.81 25834.8 0.405
8800 8812.17 230.95 304.65 31668.21 0.478 10175 10191.28 222.01 298.7 25735.64 0.404
8825 8837.24 230.79 304.54 31551.27 0.476 10200 10216.36 221.85 298.59 25636.79 0.403
8850 8862.31 230.63 304.44 31434.68 0.475 10225 10241.44 221.69 298.48 25538.24 0.401
8875 8887.38 230.46 304.33 31318.44 0.473 10250 10266.52 221.53 298.37 25440.01 0.4
8900 8912.45 230.3 304.22 31202.55 0.472 10275 10291.6 221.36 298.26 25342.08 0.399
8925 8937.52 230.14 304.12 31087.01 0.471 10300 10316.68 221.2 298.15 25244.45 0.398

10325 10341.76 221.04 298.04 25147.13 0.396 11700 11721.53 216.65 295.07 20266.83 0.326
10350 10366.84 220.88 297.93 25050.12 0.395 11725 11746.62 216.65 295.07 20187.1 0.325
10375 10391.92 220.71 297.82 24953.4 0.394 11750 11771.71 216.65 295.07 20107.67 0.323
10400 10417 220.55 297.71 24857 0.393 11775 11796.8 216.65 295.07 20028.56 0.322
10425 10442.09 220.39 297.6 24760.89 0.391 11800 11821.9 216.65 295.07 19949.76 0.321
10450 10467.17 220.23 297.49 24665.08 0.39 11825 11846.99 216.65 295.07 19871.27 0.32
10475 10492.25 220.06 297.38 24569.57 0.389 11850 11872.08 216.65 295.07 19793.08 0.318
10500 10517.33 219.9 297.27 24474.37 0.388 11875 11897.18 216.65 295.07 19715.21 0.317
10525 10542.42 219.74 297.16 24379.46 0.387 11900 11922.27 216.65 295.07 19637.64 0.316
10550 10567.5 219.58 297.05 24284.85 0.385 11925 11947.36 216.65 295.07 19560.38 0.315
10575 10592.58 219.41 296.94 24190.54 0.384 11950 11972.46 216.65 295.07 19483.42 0.313
10600 10617.67 219.25 296.83 24096.52 0.383 11975 11997.55 216.65 295.07 19406.76 0.312
10625 10642.75 219.09 296.72 24002.8 0.382 12000 12022.65 216.65 295.07 19330.41 0.311
10650 10667.83 218.93 296.61 23909.38 0.38 12025 12047.74 216.65 295.07 19254.35 0.31
10675 10692.92 218.76 296.5 23816.25 0.379 12050 12072.83 216.65 295.07 19178.6 0.308
10700 10718 218.6 296.39 23723.42 0.378 12075 12097.93 216.65 295.07 19103.14 0.307
10725 10743.09 218.44 296.28 23630.87 0.377 12100 12123.02 216.65 295.07 19027.98 0.306
10750 10768.17 218.28 296.17 23538.62 0.376 12125 12148.12 216.65 295.07 18953.11 0.305
10775 10793.25 218.11 296.06 23446.67 0.374 12150 12173.22 216.65 295.07 18878.54 0.304
10800 10818.34 217.95 295.95 23355 0.373 12175 12198.31 216.65 295.07 18804.27 0.302
10825 10843.42 217.79 295.84 23263.62 0.372 12200 12223.41 216.65 295.07 18730.28 0.301
10850 10868.51 217.63 295.73 23172.54 0.371 12225 12248.5 216.65 295.07 18656.59 0.3
10875 10893.59 217.46 295.62 23081.74 0.37 12250 12273.6 216.65 295.07 18583.19 0.299
10900 10918.68 217.3 295.51 22991.23 0.369 12275 12298.7 216.65 295.07 18510.07 0.298
10925 10943.77 217.14 295.4 22901.01 0.367 12300 12323.79 216.65 295.07 18437.24 0.296
10950 10968.85 216.98 295.29 22811.08 0.366 12325 12348.89 216.65 295.07 18364.7 0.295
10975 10993.94 216.81 295.18 22721.43 0.365 12350 12373.99 216.65 295.07 18292.45 0.294
11000 11019.03 216.65 295.07 22632.06 0.364 12375 12399.08 216.65 295.07 18220.48 0.293
11025 11044.11 216.65 295.07 22543.02 0.362 12400 12424.18 216.65 295.07 18148.79 0.292
11050 11069.2 216.65 295.07 22454.32 0.361 12425 12449.28 216.65 295.07 18077.38 0.291
11075 11094.29 216.65 295.07 22365.98 0.36 12450 12474.38 216.65 295.07 18006.26 0.277
11100 11119.37 216.65 295.07 22277.98 0.358 12475 12499.48 216.65 295.07 17935.42 0.288
11125 11144.46 216.65 295.07 22190.33 0.357 12500 12524.57 216.65 295.07 17864.85 0.287
11150 11169.55 216.65 295.07 22103.02 0.355 12525 12549.67 216.65 295.07 17794.56 0.286
11175 11194.64 216.65 295.07 22016.06 0.354 12550 12574.77 216.65 295.07 17724.55 0.285
11200 11219.72 216.65 295.07 21929.44 0.353 12575 12599.87 216.65 295.07 17654.81 0.284
11225 11244.81 216.65 295.07 21843.16 0.351 12600 12624.97 216.65 295.07 17585.35 0.283
11250 11269.9 216.65 295.07 21757.22 0.35 12625 12650.07 216.65 295.07 17516.16 0.282
11275 11294.99 216.65 295.07 21671.62 0.348 12650 12675.17 216.65 295.07 17447.25 0.281
11300 11320.08 216.65 295.07 21586.35 0.347 12675 12700.27 216.65 295.07 17378.6 0.279
11325 11345.17 216.65 295.07 21501.42 0.346 12700 12725.37 216.65 295.07 17310.22 0.278
11350 11370.26 216.65 295.07 21416.82 0.344 12725 12750.47 216.65 295.07 17242.12 0.277
11375 11395.35 216.65 295.07 21332.56 0.343 12750 12775.57 216.65 295.07 17174.28 0.276
11400 11420.44 216.65 295.07 21248.63 0.342 12775 12800.67 216.65 295.07 17106.71 0.275
11425 11445.53 216.65 295.07 21165.03 0.34 12800 12825.77 216.65 295.07 17039.4 0.274
11450 11470.62 216.65 295.07 21081.75 0.339 12825 12850.87 216.65 295.07 16972.36 0.273
11475 11495.71 216.65 295.07 20998.81 0.338 12850 12875.97 216.65 295.07 16905.59 0.272
11500 11520.8 216.65 295.07 20916.19 0.336 12875 12901.07 216.65 295.07 16839.07 0.271
11525 11545.89 216.65 295.07 20833.9 0.335 12900 12926.17 216.65 295.07 16772.82 0.27
11550 11570.98 216.65 295.07 20751.93 0.334 12925 12951.27 216.65 295.07 16706.83 0.269
11575 11596.07 216.65 295.07 20670.28 0.332 12950 12976.38 216.65 295.07 16641.1 0.268
11600 11621.16 216.65 295.07 20588.95 0.331 12975 13001.48 216.65 295.07 16575.62 0.267
11625 11646.25 216.65 295.07 20507.95 0.33 13000 13026.58 216.65 295.07 16510.41 0.265
11650 11671.34 216.65 295.07 20427.26 0.328 13025 13051.68 216.65 295.07 16445.45 0.264
11675 11696.43 216.65 295.07 20346.89 0.327 13050 13076.79 216.65 295.07 16380.74 0.263

13075 13101.89 216.65 295.07 16316.29 0.261 14450 14482.85 216.65 295.07 13135.82 0.211
13100 13126.99 216.65 295.07 16252.1 0.261 14475 14507.96 216.65 295.07 13084.13 0.21
13125 13152.09 216.65 295.07 16188.15 0.26 14500 14533.08 216.65 295.07 13032.66 0.21
13150 13177.2 216.65 295.07 16124.46 0.259 14525 14558.19 216.65 295.07 12981.38 0.209
13175 13202.3 216.65 295.07 16061.02 0.258 14550 14583.31 216.65 295.07 12930.31 0.208
13200 13227.41 216.65 295.07 15997.83 0.257 14575 14608.42 216.65 295.07 12879.43 0.207
13225 13252.51 216.65 295.07 15934.89 0.256 14600 14633.53 216.65 295.07 12828.76 0.206
13250 13277.61 216.65 295.07 15872.19 0.255 14625 14658.65 216.65 295.07 12778.28 0.205
13275 13302.72 216.65 295.07 15809.74 0.254 14650 14683.77 216.65 295.07 12728.01 0.205
13300 13327.82 216.65 295.07 15747.54 0.253 14675 14708.88 216.65 295.07 12677.93 0.204
13325 13352.93 216.65 295.07 15685.58 0.252 14700 14734 216.65 295.07 12628.05 0.203
13350 13378.03 216.65 295.07 15623.87 0.251 14725 14759.11 216.65 295.07 12578.37 0.202
13375 13403.14 216.65 295.07 15562.4 0.25 14750 14784.23 216.65 295.07 12528.88 0.201
13400 13428.24 216.65 295.07 15501.17 0.249 14775 14809.34 216.65 295.07 12479.58 0.201
13425 13453.35 216.65 295.07 15440.18 0.248 14800 14834.46 216.65 295.07 12430.48 0.2
13450 13478.45 216.65 295.07 15379.43 0.247 14825 14859.58 216.65 295.07 12381.57 0.199
13475 13503.56 216.65 295.07 15318.92 0.246 14850 14884.69 216.65 295.07 12332.86 0.198
13500 13528.67 216.65 295.07 15258.65 0.245 14875 14909.81 216.65 295.07 12284.34 0.198
13525 13553.77 216.65 295.07 15198.62 0.244 14900 14934.93 216.65 295.07 12236.01 0.197
13550 13578.88 216.65 295.07 15138.82 0.243 14925 14960.05 216.65 295.07 12187.86 0.196
13575 13603.99 216.65 295.07 15079.26 0.242 14950 14985.16 216.65 295.07 12139.91 0.195
13600 13629.09 216.65 295.07 15019.93 0.242 14975 15010.28 216.65 295.07 12092.15 0.194
13625 13654.2 216.65 295.07 14960.83 0.241 15000 15035.4 216.65 295.07 12044.57 0.194
13650 13679.31 216.65 295.07 14901.97 0.24 15025 15060.52 216.65 295.07 11997.18 0.193
13675 13704.42 216.65 295.07 14843.34 0.239 15050 15085.64 216.65 295.07 11949.98 0.192
13700 13729.52 216.65 295.07 14784.94 0.238 15075 15110.75 216.65 295.07 11902.96 0.191
13725 13754.63 216.65 295.07 14726.77 0.237 15100 15135.87 216.65 295.07 11856.13 0.191
13750 13779.74 216.65 295.07 14668.83 0.236 15125 15160.99 216.65 295.07 11809.48 0.19
13775 13804.85 216.65 295.07 14611.11 0.235 15150 15186.11 216.65 295.07 11763.02 0.189
13800 13829.96 216.65 295.07 14553.63 0.234 15175 15211.23 216.65 295.07 11716.74 0.188
13825 13855.07 216.65 295.07 14496.36 0.233 15200 15236.35 216.65 295.07 11670.64 0.188
13850 13880.17 216.65 295.07 14439.33 0.232 15225 15261.47 216.65 295.07 11624.72 0.187
13875 13905.28 216.65 295.07 14382.52 0.231 15250 15286.59 216.65 295.07 11578.99 0.186
13900 13930.39 216.65 295.07 14325.93 0.23 15275 15311.71 216.65 295.07 11533.43 0.185
13925 13955.5 216.65 295.07 14269.57 0.229 15300 15336.83 216.65 295.07 11488.05 0.185
13950 13980.61 216.65 295.07 14213.42 0.229 15325 15361.95 216.65 295.07 11442.85 0.184
13975 14005.72 216.65 295.07 14157.5 0.228 15350 15387.07 216.65 295.07 11397.83 0.183
14000 14030.83 216.65 295.07 14101.8 0.227 15375 15412.19 216.65 295.07 11352.99 0.183
14025 14055.94 216.65 295.07 14046.32 0.226 15400 15437.32 216.65 295.07 11308.32 0.182
14050 14081.05 216.65 295.07 13991.05 0.225 15425 15462.44 216.65 295.07 11263.83 0.181
14075 14106.16 216.65 295.07 13936.01 0.224 15450 15487.56 216.65 295.07 11219.51 0.18
14100 14131.27 216.65 295.07 13881.17 0.223 15475 15512.68 216.65 295.07 11175.37 0.18
14125 14156.39 216.65 295.07 13826.56 0.222 15500 15537.8 216.65 295.07 11131.4 0.179
14150 14181.5 216.65 295.07 13772.16 0.221 15525 15562.92 216.65 295.07 11087.6 0.178
14175 14206.61 216.65 295.07 13717.97 0.221 15550 15588.05 216.65 295.07 11043.98 0.178
14200 14231.72 216.65 295.07 13664 0.22 15575 15613.17 216.65 295.07 11000.53 0.177
14225 14256.83 216.65 295.07 13610.24 0.219 15600 15638.29 216.65 295.07 10957.25 0.176
14250 14281.94 216.65 295.07 13556.69 0.218 15625 15663.41 216.65 295.07 10914.14 0.175
14275 14307.06 216.65 295.07 13503.35 0.217 15650 15688.54 216.65 295.07 10871.19 0.175
14300 14332.17 216.65 295.07 13450.23 0.216 15675 15713.66 216.65 295.07 10828.42 0.174
14325 14357.28 216.65 295.07 13397.31 0.215 15700 15738.78 216.65 295.07 10785.82 0.173
14350 14382.39 216.65 295.07 13344.6 0.215 15725 15763.91 216.65 295.07 10743.38 0.173
14375 14407.51 216.65 295.07 13292.09 0.214 15750 15789.03 216.65 295.07 10701.11 0.172
14400 14432.62 216.65 295.07 13239.79 0.213 15775 15814.16 216.65 295.07 10659.01 0.171
14425 14457.73 216.65 295.07 13187.7 0.212 15800 15839.28 216.65 295.07 10617.07 0.171

15825 15864.41 216.65 295.07 10575.3 0.17 17200 17246.56 216.65 295.07 8513.89 0.137
15850 15889.53 216.65 295.07 10533.69 0.169 17225 17271.7 216.65 295.07 8480.4 0.136
15875 15914.66 216.65 295.07 10492.25 0.169 17250 17296.83 216.65 295.07 8447.03 0.136
15900 15939.78 216.65 295.07 10450.97 0.168 17275 17321.97 216.65 295.07 8413.8 0.135
15925 15964.91 216.65 295.07 10409.85 0.167 17300 17347.1 216.65 295.07 8380.69 0.135
15950 15990.03 216.65 295.07 10368.89 0.167 17325 17372.24 216.65 295.07 8347.72 0.134
15975 16015.16 216.65 295.07 10328.09 0.166 17350 17397.38 216.65 295.07 8314.88 0.134
16000 16040.28 216.65 295.07 10287.46 0.165 17375 17422.51 216.65 295.07 8282.16 0.133
16025 16065.41 216.65 295.07 10246.98 0.165 17400 17447.65 216.65 295.07 8249.58 0.133
16050 16090.54 216.65 295.07 10206.67 0.164 17425 17472.79 216.65 295.07 8217.12 0.132
16075 16115.66 216.65 295.07 10166.51 0.163 17450 17497.93 216.65 295.07 8184.79 0.132
16100 16140.79 216.65 295.07 10126.51 0.163 17475 17523.06 216.65 295.07 8152.59 0.131
16125 16165.92 216.65 295.07 10086.67 0.162 17500 17548.2 216.65 295.07 8120.51 0.131
16150 16191.04 216.65 295.07 10046.98 0.162 17525 17573.34 216.65 295.07 8088.56 0.13
16175 16216.17 216.65 295.07 10007.45 0.161 17550 17598.48 216.65 295.07 8056.74 0.13
16200 16241.3 216.65 295.07 9968.08 0.16 17575 17623.62 216.65 295.07 8025.04 0.129
16225 16266.43 216.65 295.07 9928.86 0.16 17600 17648.75 216.65 295.07 7993.46 0.129
16250 16291.55 216.65 295.07 9889.8 0.159 17625 17673.89 216.65 295.07 7962.01 0.128
16275 16316.68 216.65 295.07 9850.88 0.158 17650 17699.03 216.65 295.07 7930.69 0.128
16300 16341.81 216.65 295.07 9812.13 0.158 17675 17724.17 216.65 295.07 7899.48 0.127
16325 16366.94 216.65 295.07 9773.52 0.157 17700 17749.31 216.65 295.07 7868.4 0.127
16350 16392.07 216.65 295.07 9735.07 0.157 17725 17774.45 216.65 295.07 7837.45 0.126
16375 16417.2 216.65 295.07 9696.77 0.156 17750 17799.59 216.65 295.07 7806.61 0.126
16400 16442.33 216.65 295.07 9658.61 0.155 17775 17824.73 216.65 295.07 7775.9 0.125
16425 16467.45 216.65 295.07 9620.61 0.155 17800 17849.87 216.65 295.07 7745.3 0.125
16450 16492.58 216.65 295.07 9582.76 0.154 17825 17875.01 216.65 295.07 7714.83 0.124
16475 16517.71 216.65 295.07 9545.06 0.153 17850 17900.15 216.65 295.07 7684.47 0.124
16500 16542.84 216.65 295.07 9507.5 0.153 17875 17925.29 216.65 295.07 7654.24 0.123
16525 16567.97 216.65 295.07 9470.1 0.152 17900 17950.43 216.65 295.07 7624.13 0.123
16550 16593.1 216.65 295.07 9432.84 0.152 17925 17975.57 216.65 295.07 7594.13 0.122
16575 16618.23 216.65 295.07 9395.72 0.151 17950 18000.72 216.65 295.07 7564.25 0.122
16600 16643.37 216.65 295.07 9358.76 0.15 17975 18025.86 216.65 295.07 7534.49 0.121
16625 16668.5 216.65 295.07 9321.94 0.15 18000 18051 216.65 295.07 7504.84 0.121
16650 16693.63 216.65 295.07 9285.26 0.149 18025 18076.14 216.65 295.07 7475.32 0.12
16675 16718.76 216.65 295.07 9248.73 0.149 18050 18101.28 216.65 295.07 7445.91 0.12
16700 16743.89 216.65 295.07 9212.34 0.148 18075 18126.43 216.65 295.07 7416.61 0.119
16725 16769.02 216.65 295.07 9176.09 0.148 18100 18151.57 216.65 295.07 7387.43 0.119
16750 16794.15 216.65 295.07 9139.99 0.147 18125 18176.71 216.65 295.07 7358.36 0.118
16775 16819.29 216.65 295.07 9104.03 0.146 18150 18201.85 216.65 295.07 7329.41 0.118
16800 16844.42 216.65 295.07 9068.21 0.146 18175 18227 216.65 295.07 7300.58 0.117
16825 16869.55 216.65 295.07 9032.53 0.145 18200 18252.14 216.65 295.07 7271.85 0.117
16850 16894.68 216.65 295.07 8996.99 0.145 18225 18277.28 216.65 295.07 7243.24 0.116
16875 16919.82 216.65 295.07 8961.6 0.144 18250 18302.43 216.65 295.07 7214.74 0.116
16900 16944.95 216.65 295.07 8926.34 0.144 18275 18327.57 216.65 295.07 7186.36 0.116
16925 16970.08 216.65 295.07 8891.22 0.143 18300 18352.72 216.65 295.07 7158.08 0.115
16950 16995.22 216.65 295.07 8856.23 0.142 18325 18377.86 216.65 295.07 7129.92 0.115
16975 17020.35 216.65 295.07 8821.39 0.142 18350 18403.01 216.65 295.07 7101.87 0.114
17000 17045.48 216.65 295.07 8786.68 0.141 18375 18428.15 216.65 295.07 7073.93 0.114
17025 17070.62 216.65 295.07 8752.11 0.141 18400 18453.29 216.65 295.07 7046.09 0.113
17050 17095.75 216.65 295.07 8717.68 0.14 18425 18478.44 216.65 295.07 7018.37 0.113
17075 17120.89 216.65 295.07 8683.38 0.14 18450 18503.59 216.65 295.07 6990.76 0.112
17100 17146.02 216.65 295.07 8649.21 0.139 18475 18528.73 216.65 295.07 6963.25 0.112
17125 17171.16 216.65 295.07 8615.18 0.139 18500 18553.88 216.65 295.07 6935.86 0.112
17150 17196.29 216.65 295.07 8581.29 0.138 18525 18579.02 216.65 295.07 6908.57 0.111
17175 17221.43 216.65 295.07 8547.52 0.137 18550 18604.17 216.65 295.07 6881.39 0.111

18575 18629.31 216.65 295.07 6854.31 0.11 19975 20037.82 216.65 295.07 5496.51 0.088
18600 18654.46 216.65 295.07 6827.34 0.11 20000 20062.98 216.65 295.07 5474.89 0.088
18625 18679.61 216.65 295.07 6800.48 0.109 20025 20088.14 216.68 295.09 5453.35 0.088
18650 18704.75 216.65 295.07 6773.73 0.109 20050 20113.3 216.7 295.1 5431.9 0.087
18675 18729.9 216.65 295.07 6747.08 0.108 20075 20138.46 216.73 295.12 5410.53 0.087
18700 18755.05 216.65 295.07 6720.53 0.108 20100 20163.61 216.75 295.14 5389.25 0.087
18725 18780.2 216.65 295.07 6694.09 0.108 20125 20188.77 216.78 295.15 5368.06 0.086
18750 18805.34 216.65 295.07 6667.75 0.107 20150 20213.93 216.8 295.17 5346.95 0.086
18775 18830.49 216.65 295.07 6641.52 0.107 20175 20239.09 216.83 295.19 5325.93 0.086
18800 18855.64 216.65 295.07 6615.39 0.106 20200 20264.25 216.85 295.21 5304.99 0.085
18825 18880.79 216.65 295.07 6589.36 0.106 20225 20289.41 216.88 295.22 5284.14 0.085
18850 18905.94 216.65 295.07 6563.43 0.106 20250 20314.57 216.9 295.24 5263.38 0.085
18875 18931.09 216.65 295.07 6537.61 0.105 20275 20339.73 216.93 295.26 5242.69 0.084
18900 18956.23 216.65 295.07 6511.89 0.105 20300 20364.89 216.95 295.27 5222.09 0.084
18925 18981.38 216.65 295.07 6486.27 0.104 20325 20390.05 216.98 295.29 5201.58 0.084
18950 19006.53 216.65 295.07 6460.75 0.104 20350 20415.21 217 295.31 5181.14 0.083
18975 19031.68 216.65 295.07 6435.33 0.103 20375 20440.37 217.03 295.32 5160.79 0.083
19000 19056.83 216.65 295.07 6410.01 0.103 20400 20465.53 217.05 295.34 5140.52 0.083
19025 19081.98 216.65 295.07 6384.79 0.103 20425 20490.69 217.08 295.36 5120.34 0.082
19050 19107.13 216.65 295.07 6359.67 0.102 20450 20515.85 217.1 295.38 5100.23 0.082
19075 19132.28 216.65 295.07 6334.64 0.102 20475 20541.01 217.13 295.39 5080.21 0.082
19100 19157.43 216.65 295.07 6309.72 0.101 20500 20566.18 217.15 295.41 5060.26 0.081
19125 19182.58 216.65 295.07 6284.9 0.101 20525 20591.34 217.18 295.43 5040.4 0.081
19150 19207.73 216.65 295.07 6260.17 0.101 20550 20616.5 217.2 295.44 5020.62 0.081
19175 19232.89 216.65 295.07 6235.54 0.1 20575 20641.66 217.23 295.46 5000.92 0.08
19200 19258.04 216.65 295.07 6211 0.1 20600 20666.82 217.25 295.48 4981.29 0.08
19225 19283.19 216.65 295.07 6186.57 0.099 20625 20691.99 217.28 295.49 4961.75 0.08
19250 19308.34 216.65 295.07 6162.23 0.099 20650 20717.15 217.3 295.51 4942.29 0.079
19275 19333.49 216.65 295.07 6137.98 0.099 20675 20742.31 217.33 295.53 4922.9 0.079
19300 19358.64 216.65 295.07 6113.83 0.098 20700 20767.48 217.35 295.55 4903.59 0.079
19325 19383.8 216.65 295.07 6089.78 0.098 20725 20792.64 217.38 295.56 4884.36 0.078
19350 19408.95 216.65 295.07 6065.82 0.098 20750 20817.8 217.4 295.58 4865.21 0.078
19375 19434.1 216.65 295.07 6041.95 0.097 20775 20842.97 217.43 295.6 4846.14 0.078
19400 19459.25 216.65 295.07 6018.18 0.097 20800 20868.13 217.45 295.61 4827.14 0.077
19425 19484.41 216.65 295.07 5994.5 0.096 20825 20893.29 217.48 295.63 4808.22 0.077
19450 19509.56 216.65 295.07 5970.92 0.096 20850 20918.46 217.5 295.65 4789.37 0.077
19475 19534.71 216.65 295.07 5947.43 0.096 20875 20943.62 217.53 295.66 4770.6 0.076
19500 19559.87 216.65 295.07 5924.03 0.095 20900 20968.79 217.55 295.68 4751.91 0.076
19525 19585.02 216.65 295.07 5900.72 0.095 20925 20993.95 217.58 295.7 4733.29 0.076
19550 19610.18 216.65 295.07 5877.5 0.095 20950 21019.12 217.6 295.72 4714.75 0.075
19575 19635.33 216.65 295.07 5854.38 0.094 20975 21044.28 217.63 295.73 4696.28 0.075
19600 19660.48 216.65 295.07 5831.34 0.094 21000 21069.45 217.65 295.75 4677.89 0.075
19625 19685.64 216.65 295.07 5808.4 0.093 21025 21094.61 217.68 295.77 4659.57 0.075
19650 19710.79 216.65 295.07 5785.55 0.093 21050 21119.78 217.7 295.78 4641.32 0.074
19675 19735.95 216.65 295.07 5762.78 0.093 21075 21144.95 217.73 295.8 4623.15 0.074
19700 19761.1 216.65 295.07 5740.11 0.092 21100 21170.11 217.75 295.82 4605.05 0.074
19725 19786.26 216.65 295.07 5717.53 0.092 21125 21195.28 217.78 295.83 4587.02 0.073
19750 19811.42 216.65 295.07 5695.03 0.092 21150 21220.45 217.8 295.85 4569.07 0.073
19775 19836.57 216.65 295.07 5672.62 0.091 21175 21245.61 217.83 295.87 4551.19 0.073
19800 19861.73 216.65 295.07 5650.31 0.091 21200 21270.78 217.85 295.89 4533.38 0.072
19825 19886.88 216.65 295.07 5628.07 0.09 21225 21295.95 217.88 295.9 4515.64 0.072
19850 19912.04 216.65 295.07 5605.93 0.09 21250 21321.12 217.9 295.92 4497.98 0.072
19875 19937.2 216.65 295.07 5583.88 0.09 21275 21346.28 217.93 295.94 4480.38 0.072
19900 19962.35 216.65 295.07 5561.91 0.089 21300 21371.45 217.95 295.95 4462.86 0.071
19925 19987.51 216.65 295.07 5540.02 0.089
19950 20012.67 216.65 295.07 5518.23 0.089

21325 21396.62 217.98 295.97 4445.41 0.071 22700 22781.17 219.35 296.9 3586.03 0.057
21350 21421.79 218 295.99 4428.02 0.071 22725 22806.35 219.38 296.92 3572.09 0.057
21375 21446.96 218.03 296 4410.71 0.07 22750 22831.53 219.4 296.94 3558.21 0.056
21400 21472.12 218.05 296.02 4393.47 0.07 22775 22856.71 219.43 296.95 3544.39 0.056
21425 21497.29 218.08 296.04 4376.29 0.07 22800 22881.89 219.45 296.97 3530.62 0.056
21450 21522.46 218.1 296.06 4359.19 0.07 22825 22907.07 219.48 296.99 3516.91 0.056
21475 21547.63 218.13 296.07 4342.15 0.069 22850 22932.25 219.5 297 3503.25 0.056
21500 21572.8 218.15 296.09 4325.18 0.069 22875 22957.43 219.53 297.02 3489.64 0.055
21525 21597.97 218.18 296.11 4308.28 0.069 22900 22982.61 219.55 297.04 3476.09 0.055
21550 21623.14 218.2 296.12 4291.45 0.069 22925 23007.79 219.58 297.05 3462.6 0.055
21575 21648.31 218.23 296.14 4274.69 0.068 22950 23032.97 219.6 297.07 3449.16 0.055
21600 21673.48 218.25 296.16 4257.99 0.068 22975 23058.15 219.63 297.09 3435.77 0.054
21625 21698.65 218.28 296.17 4241.36 0.068 23000 23083.33 219.65 297.11 3422.43 0.054
21650 21723.82 218.3 296.19 4224.8 0.067 23025 23108.51 219.68 297.12 3409.15 0.054
21675 21748.99 218.33 296.21 4208.3 0.067 23050 23133.7 219.7 297.14 3395.93 0.054
21700 21774.16 218.35 296.23 4191.87 0.067 23075 23158.88 219.73 297.16 3382.75 0.054
21725 21799.34 218.38 296.24 4175.51 0.067 23100 23184.06 219.75 297.17 3369.63 0.053
21750 21824.51 218.4 296.26 4159.21 0.066 23125 23209.24 219.78 297.19 3356.56 0.053
21775 21849.68 218.43 296.28 4142.98 0.066 23150 23234.43 219.8 297.21 3343.54 0.053
21800 21874.85 218.45 296.29 4126.81 0.066 23175 23259.61 219.83 297.22 3330.57 0.053
21825 21900.02 218.48 296.31 4110.71 0.066 23200 23284.79 219.85 297.24 3317.66 0.053
21850 21925.19 218.5 296.33 4094.67 0.065 23225 23309.97 219.88 297.26 3304.8 0.052
21875 21950.37 218.53 296.34 4078.7 0.065 23250 23335.16 219.9 297.27 3291.98 0.052
21900 21975.54 218.55 296.36 4062.79 0.065 23275 23360.34 219.93 297.29 3279.22 0.052
21925 22000.71 218.58 296.38 4046.95 0.065 23300 23385.53 219.95 297.31 3266.51 0.052
21950 22025.89 218.6 296.39 4031.16 0.064 23325 23410.71 219.98 297.33 3253.86 0.052
21975 22051.06 218.63 296.41 4015.45 0.064 23350 23435.89 220 297.34 3241.25 0.051
22000 22076.23 218.65 296.43 3999.79 0.064 23375 23461.08 220.03 297.36 3228.69 0.051
22025 22101.41 218.68 296.45 3984.2 0.063 23400 23486.26 220.05 297.38 3216.18 0.051
22050 22126.58 218.7 296.46 3968.67 0.063 23425 23511.45 220.08 297.39 3203.72 0.051
22075 22151.75 218.73 296.48 3953.2 0.063 23450 23536.63 220.1 297.41 3191.32 0.051
22100 22176.93 218.75 296.5 3937.79 0.063 23475 23561.82 220.13 297.43 3178.96 0.05
22125 22202.1 218.78 296.51 3922.45 0.062 23500 23587 220.15 297.44 3166.65 0.05
22150 22227.28 218.8 296.53 3907.17 0.062 23525 23612.19 220.18 297.46 3154.39 0.05
22175 22252.45 218.83 296.55 3891.95 0.062 23550 23637.37 220.2 297.48 3142.17 0.05
22200 22277.63 218.85 296.56 3876.79 0.062 23575 23662.56 220.23 297.49 3130.01 0.05
22225 22302.8 218.88 296.58 3861.69 0.061 23600 23687.75 220.25 297.51 3117.9 0.049
22250 22327.98 218.9 296.6 3846.65 0.061 23625 23712.93 220.28 297.53 3105.83 0.049
22275 22353.15 218.93 296.61 3831.67 0.061 23650 23738.12 220.3 297.54 3093.81 0.049
22300 22378.33 218.95 296.63 3816.75 0.061 23675 23763.31 220.33 297.56 3081.84 0.049
22325 22403.51 218.98 296.65 3801.9 0.06 23700 23788.49 220.35 297.58 3069.92 0.049
22350 22428.68 219 296.67 3787.1 0.06 23725 23813.68 220.38 297.6 3058.04 0.048
22375 22453.86 219.03 296.68 3772.36 0.06 23750 23838.87 220.4 297.61 3046.21 0.048
22400 22479.03 219.05 296.7 3757.68 0.06 23775 23864.05 220.43 297.63 3034.43 0.048
22425 22504.21 219.08 296.72 3743.05 0.06 23800 23889.24 220.45 297.65 3022.7 0.048
22450 22529.39 219.1 296.73 3728.49 0.059 23825 23914.43 220.48 297.66 3011.01 0.048
22475 22554.57 219.13 296.75 3713.99 0.059 23850 23939.62 220.5 297.68 2999.37 0.047
22500 22579.74 219.15 296.77 3699.54 0.059 23875 23964.81 220.53 297.7 2987.78 0.047
22525 22604.92 219.18 296.78 3685.15 0.059 23900 23990 220.55 297.71 2976.23 0.047
22550 22630.1 219.2 296.8 3670.82 0.058 23925 24015.18 220.58 297.73 2964.73 0.047
22575 22655.28 219.23 296.82 3656.54 0.058 23950 24040.37 220.6 297.75 2953.27 0.047
22600 22680.45 219.25 296.83 3642.33 0.058 23975 24065.56 220.63 297.76 2941.86 0.046
22625 22705.63 219.28 296.85 3628.17 0.058 24000 24090.75 220.65 297.78 2930.49 0.046
22650 22730.81 219.3 296.87 3614.06 0.057 24025 24115.94 220.68 297.8 2919.17 0.046
22675 22755.99 219.33 296.89 3600.02 0.057 24050 24141.13 220.7 297.81 2907.9 0.046

24075 24166.32 220.73 297.83 2896.67 0.046 25450 25552.07 222.1 298.76 2342.93 0.037
24100 24191.51 220.75 297.85 2885.48 0.046 25475 25577.27 222.13 298.77 2333.94 0.037
24125 24216.7 220.78 297.87 2874.34 0.045 25500 25602.47 222.15 298.79 2324.98 0.036
24150 24241.89 220.8 297.88 2863.24 0.045 25525 25627.68 222.18 298.81 2316.06 0.036
24175 24267.08 220.83 297.9 2852.19 0.045 25550 25652.88 222.2 298.83 2307.17 0.036
24200 24292.27 220.85 297.92 2841.18 0.045 25575 25678.08 222.23 298.84 2298.32 0.036
24225 24317.46 220.88 297.93 2830.21 0.045 25600 25703.28 222.25 298.86 2289.51 0.036
24250 24342.66 220.9 297.95 2819.29 0.044 25625 25728.48 222.28 298.88 2280.73 0.036
24275 24367.85 220.93 297.97 2808.41 0.044 25650 25753.69 222.3 298.89 2271.98 0.036
24300 24393.04 220.95 297.98 2797.58 0.044 25675 25778.89 222.33 298.91 2263.27 0.035
24325 24418.23 220.98 298 2786.78 0.044 25700 25804.09 222.35 298.93 2254.59 0.035
24350 24443.42 221 298.02 2776.03 0.044 25725 25829.29 222.38 298.94 2245.95 0.035
24375 24468.62 221.03 298.03 2765.33 0.044 25750 25854.5 222.4 298.96 2237.34 0.035
24400 24493.81 221.05 298.05 2754.66 0.043 25775 25879.7 222.43 298.98 2228.76 0.035
24425 24519 221.08 298.07 2744.04 0.043 25800 25904.9 222.45 298.99 2220.22 0.035
24450 24544.19 221.1 298.08 2733.46 0.043 25825 25930.11 222.48 299.01 2211.71 0.035
24475 24569.39 221.13 298.1 2722.92 0.043 25850 25955.31 222.5 299.03 2203.24 0.034
24500 24594.58 221.15 298.12 2712.42 0.043 25875 25980.52 222.53 299.04 2194.8 0.034
24525 24619.77 221.18 298.14 2701.97 0.043 25900 26005.72 222.55 299.06 2186.39 0.034
24550 24644.97 221.2 298.15 2691.56 0.042 25925 26030.93 222.58 299.08 2178.02 0.034
24575 24670.16 221.23 298.17 2681.19 0.042 25950 26056.13 222.6 299.09 2169.68 0.034
24600 24695.35 221.25 298.19 2670.85 0.042 25975 26081.34 222.63 299.11 2161.37 0.034
24625 24720.55 221.28 298.2 2660.57 0.042 26000 26106.54 222.65 299.13 2153.09 0.034
24650 24745.74 221.3 298.22 2650.32 0.042 26025 26131.75 222.68 299.14 2144.85 0.034
24675 24770.94 221.33 298.24 2640.11 0.042 26050 26156.95 222.7 299.16 2136.64 0.033
24700 24796.13 221.35 298.25 2629.94 0.041 26075 26182.16 222.73 299.18 2128.46 0.033
24725 24821.33 221.38 298.27 2619.81 0.041 26100 26207.36 222.75 299.19 2120.32 0.033
24750 24846.52 221.4 298.29 2609.73 0.041 26125 26232.57 222.78 299.21 2112.2 0.033
24775 24871.72 221.43 298.3 2599.68 0.041 26150 26257.78 222.8 299.23 2104.12 0.033
24800 24896.91 221.45 298.32 2589.67 0.041 26175 26282.98 222.83 299.25 2096.07 0.033
24825 24922.11 221.48 298.34 2579.7 0.041 26200 26308.19 222.85 299.26 2088.05 0.033
24850 24947.31 221.5 298.35 2569.77 0.04 26225 26333.4 222.88 299.28 2080.07 0.033
24875 24972.5 221.53 298.37 2559.88 0.04 26250 26358.6 222.9 299.3 2072.11 0.032
24900 24997.7 221.55 298.39 2550.03 0.04 26275 26383.81 222.93 299.31 2064.19 0.032
24925 25022.9 221.58 298.4 2540.22 0.04 26300 26409.02 222.95 299.33 2056.29 0.032
24950 25048.09 221.6 298.42 2530.45 0.04 26325 26434.23 222.98 299.35 2048.43 0.032
24975 25073.29 221.63 298.44 2520.72 0.04 26350 26459.43 223 299.36 2040.6 0.032
25000 25098.49 221.65 298.46 2511.02 0.039 26375 26484.64 223.03 299.38 2032.8 0.032
25025 25123.68 221.68 298.47 2501.37 0.039 26400 26509.85 223.05 299.4 2025.03 0.032
25050 25148.88 221.7 298.49 2491.75 0.039 26425 26535.06 223.08 299.41 2017.29 0.032
25075 25174.08 221.73 298.51 2482.17 0.039 26450 26560.27 223.1 299.43 2009.58 0.031
25100 25199.28 221.75 298.52 2472.63 0.039 26475 26585.48 223.13 299.45 2001.91 0.031
25125 25224.48 221.78 298.54 2463.12 0.039 26500 26610.69 223.15 299.46 1994.26 0.031
25150 25249.67 221.8 298.56 2453.65 0.039 26525 26635.9 223.18 299.48 1986.64 0.031
25175 25274.87 221.83 298.57 2444.22 0.038 26550 26661.11 223.2 299.5 1979.05 0.031
25200 25300.07 221.85 298.59 2434.83 0.038 26575 26686.32 223.23 299.51 1971.49 0.031
25225 25325.27 221.88 298.61 2425.48 0.038 26600 26711.53 223.25 299.53 1963.97 0.031
25250 25350.47 221.9 298.62 2416.16 0.038 26625 26736.74 223.28 299.55 1956.47 0.031
25275 25375.67 221.93 298.64 2406.88 0.038 26650 26761.95 223.3 299.56 1949 0.03
25300 25400.87 221.95 298.66 2397.63 0.038 26675 26787.16 223.33 299.58 1941.56 0.03
25325 25426.07 221.98 298.67 2388.43 0.037 26700 26812.37 223.35 299.6 1934.15 0.03
25350 25451.27 222 298.69 2379.25 0.037 26725 26837.58 223.38 299.61 1926.77 0.03
25375 25476.47 222.03 298.71 2370.12 0.037 26750 26862.79 223.4 299.63 1919.41 0.03
25400 25501.67 222.05 298.72 2361.02 0.037 26775 26888 223.43 299.65 1912.09 0.03
25425 25526.87 222.08 298.74 2351.96 0.037 26800 26913.21 223.45 299.66 1904.8 0.03

26825 26938.42 223.48 299.68 1897.53 0.03 28200 28325.38 224.85 300.6 1538.79 0.024
26850 26963.64 223.5 299.7 1890.29 0.029 28225 28350.6 224.88 300.62 1532.96 0.024
26875 26988.85 223.53 299.71 1883.08 0.029 28250 28375.82 224.9 300.64 1527.15 0.024
26900 27014.06 223.55 299.73 1875.9 0.029 28275 28401.05 224.93 300.65 1521.36 0.024
26925 27039.27 223.58 299.75 1868.75 0.029 28300 28426.27 224.95 300.67 1515.59 0.023
26950 27064.49 223.6 299.77 1861.62 0.029 28325 28451.49 224.98 300.69 1509.85 0.023
26975 27089.7 223.63 299.78 1854.53 0.029 28350 28476.72 225 300.7 1504.13 0.023
27000 27114.91 223.65 299.8 1847.46 0.029 28375 28501.94 225.03 300.72 1498.43 0.023
27025 27140.13 223.68 299.82 1840.42 0.029 28400 28527.17 225.05 300.74 1492.75 0.023
27050 27165.34 223.7 299.83 1833.4 0.029 28425 28552.39 225.08 300.75 1487.1 0.023
27075 27190.55 223.73 299.85 1826.42 0.028 28450 28577.61 225.1 300.77 1481.47 0.023
27100 27215.77 223.75 299.87 1819.46 0.028 28475 28602.84 225.13 300.79 1475.86 0.023
27125 27240.98 223.78 299.88 1812.53 0.028 28500 28628.06 225.15 300.8 1470.27 0.023
27150 27266.19 223.8 299.9 1805.62 0.028 28525 28653.29 225.18 300.82 1464.7 0.023
27175 27291.41 223.83 299.92 1798.74 0.028 28550 28678.52 225.2 300.84 1459.16 0.023
27200 27316.62 223.85 299.93 1791.89 0.028 28575 28703.74 225.23 300.85 1453.64 0.022
27225 27341.84 223.88 299.95 1785.07 0.028 28600 28728.97 225.25 300.87 1448.13 0.022
27250 27367.05 223.9 299.97 1778.27 0.028 28625 28754.19 225.28 300.89 1442.65 0.022
27275 27392.27 223.93 299.98 1771.5 0.028 28650 28779.42 225.3 300.9 1437.19 0.022
27300 27417.49 223.95 300 1764.76 0.027 28675 28804.65 225.33 300.92 1431.76 0.022
27325 27442.7 223.98 300.02 1758.04 0.027 28700 28829.87 225.35 300.94 1426.34 0.022
27350 27467.92 224 300.03 1751.35 0.027 28725 28855.1 225.38 300.95 1420.95 0.022
27375 27493.13 224.03 300.05 1744.69 0.027 28750 28880.33 225.4 300.97 1415.57 0.022
27400 27518.35 224.05 300.07 1738.05 0.027 28775 28905.55 225.43 300.99 1410.22 0.022
27425 27543.57 224.08 300.08 1731.44 0.027 28800 28930.78 225.45 301 1404.89 0.022
27450 27568.78 224.1 300.1 1724.85 0.027 28825 28956.01 225.48 301.02 1399.57 0.022
27475 27594 224.13 300.12 1718.29 0.027 28850 28981.24 225.5 301.04 1394.28 0.022
27500 27619.22 224.15 300.13 1711.75 0.027 28875 29006.46 225.53 301.05 1389.01 0.021
27525 27644.43 224.18 300.15 1705.25 0.026 28900 29031.69 225.55 301.07 1383.76 0.021
27550 27669.65 224.2 300.17 1698.76 0.026 28925 29056.92 225.58 301.09 1378.53 0.021
27575 27694.87 224.23 300.18 1692.3 0.026 28950 29082.15 225.6 301.1 1373.32 0.021
27600 27720.09 224.25 300.2 1685.87 0.026 28975 29107.38 225.63 301.12 1368.13 0.021
27625 27745.31 224.28 300.22 1679.46 0.026 29000 29132.61 225.65 301.14 1362.96 0.021
27650 27770.52 224.3 300.23 1673.08 0.026 29025 29157.84 225.68 301.15 1357.82 0.021
27675 27795.74 224.33 300.25 1666.72 0.026 29050 29183.07 225.7 301.17 1352.69 0.021
27700 27820.96 224.35 300.27 1660.39 0.026 29075 29208.3 225.73 301.19 1347.58 0.021
27725 27846.18 224.38 300.28 1654.08 0.026 29100 29233.53 225.75 301.2 1342.49 0.021
27750 27871.4 224.4 300.3 1647.79 0.026 29125 29258.76 225.78 301.22 1337.42 0.021
27775 27896.62 224.43 300.32 1641.53 0.025 29150 29283.99 225.8 301.24 1332.37 0.021
27800 27921.84 224.45 300.33 1635.3 0.025 29175 29309.22 225.83 301.25 1327.34 0.02
27825 27947.06 224.48 300.35 1629.09 0.025 29200 29334.45 225.85 301.27 1322.33 0.02
27850 27972.28 224.5 300.37 1622.9 0.025 29225 29359.68 225.88 301.29 1317.34 0.02
27875 27997.5 224.53 300.38 1616.74 0.025 29250 29384.91 225.9 301.3 1312.37 0.02
27900 28022.72 224.55 300.4 1610.6 0.025 29275 29410.14 225.93 301.32 1307.42 0.02
27925 28047.94 224.58 300.42 1604.49 0.025 29300 29435.37 225.95 301.34 1302.48 0.02
27950 28073.16 224.6 300.43 1598.4 0.025 29325 29460.6 225.98 301.35 1297.57 0.02
27975 28098.38 224.63 300.45 1592.33 0.025 29350 29485.84 226 301.37 1292.68 0.02
28000 28123.6 224.65 300.47 1586.29 0.025 29375 29511.07 226.03 301.39 1287.8 0.02
28025 28148.82 224.68 300.48 1580.27 0.025 29400 29536.3 226.05 301.4 1282.94 0.02
28050 28174.04 224.7 300.5 1574.28 0.024 29425 29561.53 226.08 301.42 1278.1 0.02
28075 28199.27 224.73 300.52 1568.3 0.024 29450 29586.77 226.1 301.44 1273.29 0.02
28100 28224.49 224.75 300.53 1562.35 0.024 29475 29612 226.13 301.45 1268.49 0.02
28125 28249.71 224.78 300.55 1556.43 0.024 29500 29637.23 226.15 301.47 1263.7 0.019
28150 28274.93 224.8 300.57 1550.53 0.024 29525 29662.46 226.18 301.49 1258.94 0.019
28175 28300.15 224.83 300.59 1544.65 0.024 29550 29687.7 226.2 301.5 1254.2 0.019

29575 29712.93 226.23 301.52 1249.47 0.019 30950 31101.09 227.6 302.43 1015.83 0.016
29600 29738.17 226.25 301.54 1244.76 0.019 30975 31126.33 227.63 302.45 1012.02 0.015
29625 29763.4 226.28 301.55 1240.07 0.019 31000 31151.58 227.65 302.47 1008.23 0.015
29650 29788.63 226.3 301.57 1235.4 0.019 31025 31176.82 227.68 302.48 1004.46 0.015
29675 29813.87 226.33 301.59 1230.75 0.019 31050 31202.07 227.7 302.5 1000.7 0.015
29700 29839.1 226.35 301.6 1226.11 0.019 31075 31227.31 227.73 302.52 996.95 0.015
29725 29864.34 226.38 301.62 1221.49 0.019 31100 31252.56 227.75 302.53 993.22 0.015
29750 29889.57 226.4 301.64 1216.89 0.019 31125 31277.81 227.78 302.55 989.5 0.015
29775 29914.81 226.43 301.65 1212.31 0.019 31150 31303.05 227.8 302.57 985.8 0.015
29800 29940.04 226.45 301.67 1207.75 0.019 31175 31328.3 227.83 302.58 982.11 0.015
29825 29965.28 226.48 301.69 1203.2 0.019 31200 31353.54 227.85 302.6 978.43 0.015
29850 29990.51 226.5 301.7 1198.67 0.018 31225 31378.79 227.88 302.62 974.77 0.015
29875 30015.75 226.53 301.72 1194.16 0.018 31250 31404.04 227.9 302.63 971.13 0.015
29900 30040.99 226.55 301.74 1189.67 0.018 31275 31429.29 227.93 302.65 967.49 0.015
29925 30066.22 226.58 301.75 1185.19 0.018 31300 31454.53 227.95 302.67 963.88 0.015
29950 30091.46 226.6 301.77 1180.73 0.018 31325 31479.78 227.98 302.68 960.27 0.015
29975 30116.7 226.63 301.79 1176.29 0.018 31350 31505.03 228 302.7 956.68 0.015
30000 30141.93 226.65 301.8 1171.87 0.018 31375 31530.28 228.03 302.72 953.1 0.015
30025 30167.17 226.68 301.82 1167.46 0.018 31400 31555.52 228.05 302.73 949.54 0.015
30050 30192.41 226.7 301.84 1163.07 0.018 31425 31580.77 228.08 302.75 945.99 0.014
30075 30217.65 226.73 301.85 1158.7 0.018 31450 31606.02 228.1 302.77 942.46 0.014
30100 30242.88 226.75 301.87 1154.34 0.018 31475 31631.27 228.13 302.78 938.93 0.014
30125 30268.12 226.78 301.89 1150 0.018 31500 31656.52 228.15 302.8 935.43 0.014
30150 30293.36 226.8 301.9 1145.68 0.018 31525 31681.77 228.18 302.82 931.93 0.014
30175 30318.6 226.83 301.92 1141.37 0.018 31550 31707.02 228.2 302.83 928.45 0.014
30200 30343.84 226.85 301.94 1137.08 0.017 31575 31732.27 228.23 302.85 924.98 0.014
30225 30369.08 226.88 301.95 1132.81 0.017 31600 31757.52 228.25 302.87 921.53 0.014
30250 30394.31 226.9 301.97 1128.55 0.017 31625 31782.77 228.28 302.88 918.08 0.014
30275 30419.55 226.93 301.99 1124.31 0.017 31650 31808.02 228.3 302.9 914.66 0.014
30300 30444.79 226.95 302 1120.09 0.017 31675 31833.27 228.33 302.92 911.24 0.014
30325 30470.03 226.98 302.02 1115.88 0.017 31700 31858.52 228.35 302.93 907.84 0.014
30350 30495.27 227 302.04 1111.69 0.017 31725 31883.77 228.38 302.95 904.45 0.014
30375 30520.51 227.03 302.05 1107.52 0.017 31750 31909.02 228.4 302.97 901.07 0.014
30400 30545.75 227.05 302.07 1103.36 0.017 31775 31934.27 228.43 302.98 897.71 0.014
30425 30570.99 227.08 302.09 1099.22 0.017 31800 31959.52 228.45 303 894.36 0.014
30450 30596.23 227.1 302.1 1095.09 0.017 31825 31984.77 228.48 303.02 891.02 0.014
30475 30621.47 227.13 302.12 1090.98 0.017 31850 32010.03 228.5 303.03 887.7 0.014
30500 30646.72 227.15 302.14 1086.88 0.017 31875 32035.28 228.53 303.05 884.39 0.013
30525 30671.96 227.18 302.15 1082.81 0.017 31900 32060.53 228.55 303.06 881.09 0.013
30550 30697.2 227.2 302.17 1078.74 0.017 31925 32085.78 228.58 303.08 877.8 0.013
30575 30722.44 227.23 302.19 1074.7 0.016 31950 32111.03 228.6 303.1 874.53 0.013
30600 30747.68 227.25 302.2 1070.66 0.016 31975 32136.29 228.63 303.11 871.27 0.013
30625 30772.92 227.28 302.22 1066.65 0.016 32000 32161.54 228.65 303.13 868.02 0.013
30650 30798.17 227.3 302.24 1062.65 0.016 32025 32186.79 228.72 303.18 864.78 0.013
30675 30823.41 227.33 302.25 1058.66 0.016 32050 32212.05 228.79 303.22 861.56 0.013
30700 30848.65 227.35 302.27 1054.69 0.016 32075 32237.3 228.86 303.27 858.35 0.013
30725 30873.89 227.38 302.28 1050.74 0.016 32100 32262.55 228.93 303.32 855.15 0.013
30750 30899.14 227.4 302.3 1046.8 0.016 32125 32287.81 229 303.36 851.97 0.013
30775 30924.38 227.43 302.32 1042.87 0.016 32150 32313.06 229.07 303.41 848.8 0.013
30800 30949.62 227.45 302.33 1038.97 0.016 32175 32338.32 229.14 303.46 845.64 0.013
30825 30974.87 227.48 302.35 1035.07 0.016 32200 32363.57 229.21 303.5 842.49 0.013
30850 31000.11 227.5 302.37 1031.19 0.016 32225 32388.83 229.28 303.55 839.36 0.013
30875 31025.35 227.53 302.38 1027.33 0.016 32250 32414.08 229.35 303.59 836.24 0.013
30900 31050.6 227.55 302.4 1023.48 0.016 32275 32439.34 229.42 303.64 833.13 0.013
30925 31075.84 227.58 302.42 1019.65 0.016 32300 32464.59 229.49 303.69 830.04 0.013

32325 32489.85 229.56 303.73 826.96 0.013 33700 33879.21 233.41 306.27 675.07 0.01
32350 32515.1 229.63 303.78 823.88 0.012 33725 33904.47 233.48 306.32 672.61 0.01
32375 32540.36 229.7 303.83 820.83 0.012 33750 33929.74 233.55 306.36 670.15 0.01
32400 32565.61 229.77 303.87 817.78 0.012 33775 33955.01 233.62 306.41 667.71 0.01
32425 32590.87 229.84 303.92 814.75 0.012 33800 33980.28 233.69 306.45 665.27 0.01
32450 32616.13 229.91 303.97 811.73 0.012 33825 34005.54 233.76 306.5 662.84 0.01
32475 32641.38 229.98 304.01 808.72 0.012 33850 34030.81 233.83 306.55 660.43 0.01
32500 32666.64 230.05 304.06 805.72 0.012 33875 34056.08 233.9 306.59 658.02 0.01
32525 32691.9 230.12 304.1 802.73 0.012 33900 34081.35 233.97 306.64 655.62 0.01
32550 32717.15 230.19 304.15 799.76 0.012 33925 34106.61 234.04 306.68 653.23 0.01
32575 32742.41 230.26 304.2 796.8 0.012 33950 34131.88 234.11 306.73 650.85 0.01
32600 32767.67 230.33 304.24 793.85 0.012 33975 34157.15 234.18 306.78 648.48 0.01
32625 32792.93 230.4 304.29 790.91 0.012 34000 34182.42 234.25 306.82 646.12 0.01
32650 32818.19 230.47 304.34 787.99 0.012 34025 34207.69 234.32 306.87 643.77 0.01
32675 32843.44 230.54 304.38 785.07 0.012 34050 34232.96 234.39 306.91 641.43 0.01
32700 32868.7 230.61 304.43 782.17 0.012 34075 34258.23 234.46 306.96 639.1 0.009
32725 32893.96 230.68 304.47 779.28 0.012 34100 34283.5 234.53 307 636.77 0.009
32750 32919.22 230.75 304.52 776.4 0.012 34125 34308.77 234.6 307.05 634.46 0.009
32775 32944.48 230.82 304.57 773.53 0.012 34150 34334.04 234.67 307.1 632.15 0.009
32800 32969.74 230.89 304.61 770.67 0.012 34175 34359.31 234.74 307.14 629.86 0.009
32825 32995 230.96 304.66 767.83 0.012 34200 34384.58 234.81 307.19 627.57 0.009
32850 33020.26 231.03 304.7 764.99 0.012 34225 34409.85 234.88 307.23 625.29 0.009
32875 33045.52 231.1 304.75 762.17 0.011 34250 34435.12 234.95 307.28 623.02 0.009
32900 33070.78 231.17 304.8 759.36 0.011 34275 34460.39 235.02 307.32 620.76 0.009
32925 33096.04 231.24 304.84 756.56 0.011 34300 34485.66 235.09 307.37 618.51 0.009
32950 33121.3 231.31 304.89 753.77 0.011 34325 34510.93 235.16 307.42 616.27 0.009
32975 33146.56 231.38 304.94 750.99 0.011 34350 34536.21 235.23 307.46 614.03 0.009
33000 33171.82 231.45 304.98 748.23 0.011 34375 34561.48 235.3 307.51 611.81 0.009
33025 33197.08 231.52 305.03 745.47 0.011 34400 34586.75 235.37 307.55 609.59 0.009
33050 33222.34 231.59 305.07 742.73 0.011 34425 34612.02 235.44 307.6 607.39 0.009
33075 33247.6 231.66 305.12 739.99 0.011 34450 34637.29 235.51 307.64 605.19 0.009
33100 33272.87 231.73 305.17 737.27 0.011 34475 34662.57 235.58 307.69 603 0.009
33125 33298.13 231.8 305.21 734.56 0.011 34500 34687.84 235.65 307.74 600.81 0.009
33150 33323.39 231.87 305.26 731.86 0.011 34525 34713.11 235.72 307.78 598.64 0.009
33175 33348.65 231.94 305.3 729.17 0.011 34550 34738.39 235.79 307.83 596.48 0.009
33200 33373.92 232.01 305.35 726.49 0.011 34575 34763.66 235.86 307.87 594.32 0.009
33225 33399.18 232.08 305.4 723.82 0.011 34600 34788.93 235.93 307.92 592.17 0.009
33250 33424.44 232.15 305.44 721.16 0.011 34625 34814.21 236 307.96 590.03 0.009
33275 33449.7 232.22 305.49 718.51 0.011 34650 34839.48 236.07 308.01 587.9 0.009
33300 33474.97 232.29 305.53 715.88 0.011 34675 34864.76 236.14 308.06 585.78 0.009
33325 33500.23 232.36 305.58 713.25 0.011 34700 34890.03 236.21 308.1 583.66 0.009
33350 33525.49 232.43 305.63 710.63 0.011 34725 34915.31 236.28 308.15 581.56 0.009
33375 33550.76 232.5 305.67 708.03 0.011 34750 34940.58 236.35 308.19 579.46 0.009
33400 33576.02 232.57 305.72 705.43 0.011 34775 34965.86 236.42 308.24 577.37 0.009
33425 33601.29 232.64 305.76 702.85 0.011 34800 34991.13 236.49 308.28 575.29 0.008
33450 33626.55 232.71 305.81 700.27 0.01 34825 35016.41 236.56 308.33 573.21 0.008
33475 33651.82 232.78 305.86 697.7 0.01 34850 35041.68 236.63 308.38 571.15 0.008
33500 33677.08 232.85 305.9 695.15 0.01 34875 35066.96 236.7 308.42 569.09 0.008
33525 33702.35 232.92 305.95 692.61 0.01 34900 35092.23 236.77 308.47 567.04 0.008
33550 33727.61 232.99 305.99 690.07 0.01 34925 35117.51 236.84 308.51 565 0.008
33575 33752.88 233.06 306.04 687.55 0.01 34950 35142.79 236.91 308.56 562.97 0.008
33600 33778.14 233.13 306.09 685.03 0.01 34975 35168.06 236.98 308.6 560.94 0.008
33625 33803.41 233.2 306.13 682.53 0.01 35000 35193.34 237.05 308.65 558.92 0.008
33650 33828.67 233.27 306.18 680.03 0.01 35025 35218.62 237.12 308.69 556.91 0.008
33675 33853.94 233.34 306.22 677.55 0.01 35050 35243.89 237.19 308.74 554.91 0.008

35075 35269.17 237.26 308.79 552.92 0.008 36450 36659.74 241.11 311.28 454.32 0.007
35100 35294.45 237.33 308.83 550.93 0.008 36475 36685.03 241.18 311.33 452.72 0.007
35125 35319.73 237.4 308.88 548.95 0.008 36500 36710.32 241.25 311.37 451.12 0.007
35150 35345.01 237.47 308.92 546.98 0.008 36525 36735.61 241.32 311.42 449.52 0.006
35175 35370.28 237.54 308.97 545.02 0.008 36550 36760.89 241.39 311.46 447.94 0.006
35200 35395.56 237.61 309.01 543.06 0.008 36575 36786.18 241.46 311.51 446.35 0.006
35225 35420.84 237.68 309.06 541.11 0.008 36600 36811.47 241.53 311.55 444.78 0.006
35250 35446.12 237.75 309.1 539.17 0.008 36625 36836.76 241.6 311.6 443.21 0.006
35275 35471.4 237.82 309.15 537.24 0.008 36650 36862.05 241.67 311.64 441.64 0.006
35300 35496.68 237.89 309.2 535.31 0.008 36675 36887.34 241.74 311.69 440.09 0.006
35325 35521.96 237.96 309.24 533.4 0.008 36700 36912.63 241.81 311.73 438.53 0.006
35350 35547.24 238.03 309.29 531.49 0.008 36725 36937.93 241.88 311.78 436.99 0.006
35375 35572.52 238.1 309.33 529.58 0.008 36750 36963.22 241.95 311.82 435.45 0.006
35400 35597.8 238.17 309.38 527.69 0.008 36775 36988.51 242.02 311.87 433.91 0.006
35425 35623.08 238.24 309.42 525.8 0.008 36800 37013.8 242.09 311.91 432.39 0.006
35450 35648.36 238.31 309.47 523.92 0.008 36825 37039.09 242.16 311.96 430.86 0.006
35475 35673.64 238.38 309.51 522.04 0.008 36850 37064.38 242.23 312 429.35 0.006
35500 35698.92 238.45 309.56 520.18 0.008 36875 37089.67 242.3 312.05 427.84 0.006
35525 35724.2 238.52 309.6 518.32 0.008 36900 37114.97 242.37 312.09 426.33 0.006
35550 35749.48 238.59 309.65 516.46 0.008 36925 37140.26 242.44 312.14 424.83 0.006
35575 35774.76 238.66 309.7 514.62 0.008 36950 37165.55 242.51 312.18 423.34 0.006
35600 35800.04 238.73 309.74 512.78 0.007 36975 37190.84 242.58 312.23 421.85 0.006
35625 35825.33 238.8 309.79 510.95 0.007 37000 37216.14 242.65 312.27 420.37 0.006
35650 35850.61 238.87 309.83 509.13 0.007 37025 37241.43 242.72 312.32 418.89 0.006
35675 35875.89 238.94 309.88 507.31 0.007 37050 37266.72 242.79 312.36 417.42 0.006
35700 35901.17 239.01 309.92 505.5 0.007 37075 37292.01 242.86 312.41 415.95 0.006
35725 35926.46 239.08 309.97 503.7 0.007 37100 37317.31 242.93 312.45 414.49 0.006
35750 35951.74 239.15 310.01 501.9 0.007 37125 37342.6 243 312.5 413.04 0.006
35775 35977.02 239.22 310.06 500.11 0.007 37150 37367.9 243.07 312.54 411.59 0.006
35800 36002.3 239.29 310.1 498.33 0.007 37175 37393.19 243.14 312.59 410.15 0.006
35825 36027.59 239.36 310.15 496.55 0.007 37200 37418.48 243.21 312.63 408.71 0.006
35850 36052.87 239.43 310.19 494.78 0.007 37225 37443.78 243.28 312.68 407.28 0.006
35875 36078.16 239.5 310.24 493.02 0.007 37250 37469.07 243.35 312.72 405.85 0.006
35900 36103.44 239.57 310.29 491.27 0.007 37275 37494.37 243.42 312.77 404.43 0.006
35925 36128.72 239.64 310.33 489.52 0.007 37300 37519.66 243.49 312.81 403.01 0.006
35950 36154.01 239.71 310.38 487.78 0.007 37325 37544.96 243.56 312.86 401.6 0.006
35975 36179.29 239.78 310.42 486.04 0.007 37350 37570.26 243.63 312.9 400.19 0.006
36000 36204.58 239.85 310.47 484.32 0.007 37375 37595.55 243.7 312.95 398.79 0.006
36025 36229.86 239.92 310.51 482.6 0.007 37400 37620.85 243.77 312.99 397.4 0.006
36050 36255.15 239.99 310.56 480.88 0.007 37425 37646.14 243.84 313.04 396.01 0.006
36075 36280.43 240.06 310.6 479.17 0.007 37450 37671.44 243.91 313.08 394.62 0.006
36100 36305.72 240.13 310.65 477.47 0.007 37475 37696.74 243.98 313.13 393.25 0.006
36125 36331 240.2 310.69 475.78 0.007 37500 37722.03 244.05 313.17 391.87 0.006
36150 36356.29 240.27 310.74 474.09 0.007 37525 37747.33 244.12 313.22 390.5 0.006
36175 36381.58 240.34 310.78 472.41 0.007 37550 37772.63 244.19 313.26 389.14 0.006
36200 36406.86 240.41 310.83 470.73 0.007 37575 37797.93 244.26 313.31 387.78 0.006
36225 36432.15 240.48 310.87 469.06 0.007 37600 37823.22 244.33 313.35 386.43 0.006
36250 36457.44 240.55 310.92 467.4 0.007 37625 37848.52 244.4 313.4 385.08 0.005
36275 36482.72 240.62 310.96 465.74 0.007 37650 37873.82 244.47 313.44 383.74 0.005
36300 36508.01 240.69 311.01 464.09 0.007 37675 37899.12 244.54 313.49 382.4 0.005
36325 36533.3 240.76 311.06 462.45 0.007 37700 37924.42 244.61 313.53 381.06 0.005
36350 36558.59 240.83 311.1 460.81 0.007 37725 37949.71 244.68 313.58 379.74 0.005
36375 36583.87 240.9 311.15 459.18 0.007 37750 37975.01 244.75 313.62 378.41 0.005
36400 36609.16 240.97 311.19 457.56 0.007 37775 38000.31 244.82 313.67 377.1 0.005
36425 36634.45 241.04 311.24 455.94 0.007 37800 38025.61 244.89 313.71 375.78 0.005

37825 38050.91 244.96 313.76 374.47 0.005 39200 39442.69 248.81 316.21 309.59 0.004
37850 38076.21 245.03 313.8 373.17 0.005 39225 39468 248.88 316.26 308.53 0.004
37875 38101.51 245.1 313.85 371.87 0.005 39250 39493.31 248.95 316.3 307.47 0.004
37900 38126.81 245.17 313.89 370.58 0.005 39275 39518.62 249.02 316.35 306.42 0.004
37925 38152.11 245.24 313.94 369.29 0.005 39300 39543.93 249.09 316.39 305.37 0.004
37950 38177.41 245.31 313.98 368.01 0.005 39325 39569.24 249.16 316.43 304.33 0.004
37975 38202.71 245.38 314.03 366.73 0.005 39350 39594.55 249.23 316.48 303.29 0.004
38000 38228.01 245.45 314.07 365.45 0.005 39375 39619.86 249.3 316.52 302.25 0.004
38025 38253.31 245.52 314.11 364.19 0.005 39400 39645.18 249.37 316.57 301.21 0.004
38050 38278.61 245.59 314.16 362.92 0.005 39425 39670.49 249.44 316.61 300.18 0.004
38075 38303.92 245.66 314.2 361.66 0.005 39450 39695.8 249.51 316.66 299.16 0.004
38100 38329.22 245.73 314.25 360.41 0.005 39475 39721.11 249.58 316.7 298.14 0.004
38125 38354.52 245.8 314.29 359.16 0.005 39500 39746.43 249.65 316.75 297.12 0.004
38150 38379.82 245.87 314.34 357.91 0.005 39525 39771.74 249.72 316.79 296.1 0.004
38175 38405.12 245.94 314.38 356.67 0.005 39550 39797.05 249.79 316.83 295.09 0.004
38200 38430.43 246.01 314.43 355.43 0.005 39575 39822.37 249.86 316.88 294.09 0.004
38225 38455.73 246.08 314.47 354.2 0.005 39600 39847.68 249.93 316.92 293.08 0.004
38250 38481.03 246.15 314.52 352.97 0.005 39625 39872.99 250 316.97 292.08 0.004
38275 38506.33 246.22 314.56 351.75 0.005 39650 39898.31 250.07 317.01 291.09 0.004
38300 38531.64 246.29 314.61 350.53 0.005 39675 39923.62 250.14 317.06 290.09 0.004
38325 38556.94 246.36 314.65 349.32 0.005 39700 39948.94 250.21 317.1 289.11 0.004
38350 38582.24 246.43 314.7 348.11 0.005 39725 39974.25 250.28 317.15 288.12 0.004
38375 38607.55 246.5 314.74 346.91 0.005 39750 39999.57 250.35 317.19 287.14 0.004
38400 38632.85 246.57 314.79 345.71 0.005 39775 40024.88 250.42 317.23 286.16 0.004
38425 38658.16 246.64 314.83 344.51 0.005 39800 40050.2 250.49 317.28 285.19 0.004
38450 38683.46 246.71 314.88 343.32 0.005 39825 40075.51 250.56 317.32 284.22 0.004
38475 38708.77 246.78 314.92 342.14 0.005 39850 40100.83 250.63 317.37 283.25 0.004
38500 38734.07 246.85 314.96 340.95 0.005 39875 40126.14 250.7 317.41 282.29 0.004
38525 38759.38 246.92 315.01 339.78 0.005 39900 40151.46 250.77 317.46 281.33 0.004
38550 38784.68 246.99 315.05 338.6 0.005 39925 40176.77 250.84 317.5 280.37 0.004
38575 38809.99 247.06 315.1 337.43 0.005 39950 40202.09 250.91 317.54 279.42 0.004
38600 38835.29 247.13 315.14 336.27 0.005 39975 40227.41 250.98 317.59 278.47 0.004
38625 38860.6 247.2 315.19 335.11 0.005 40000 40252.72 251.05 317.63 277.52 0.004
38650 38885.9 247.27 315.23 333.95 0.005 40025 40278.04 251.12 317.68 276.58 0.004
38675 38911.21 247.34 315.28 332.8 0.005 40050 40303.36 251.19 317.72 275.64 0.004
38700 38936.52 247.41 315.32 331.66 0.005 40075 40328.68 251.26 317.77 274.7 0.004
38725 38961.82 247.48 315.37 330.51 0.005 40100 40353.99 251.33 317.81 273.77 0.004
38750 38987.13 247.55 315.41 329.38 0.005 40125 40379.31 251.4 317.85 272.84 0.004
38775 39012.44 247.62 315.46 328.24 0.005 40150 40404.63 251.47 317.9 271.92 0.004
38800 39037.74 247.69 315.5 327.11 0.005 40175 40429.95 251.54 317.94 271 0.004
38825 39063.05 247.76 315.54 325.98 0.005 40200 40455.27 251.61 317.99 270.08 0.004
38850 39088.36 247.83 315.59 324.86 0.005 40225 40480.58 251.68 318.03 269.16 0.004
38875 39113.67 247.9 315.63 323.75 0.005 40250 40505.9 251.75 318.08 268.25 0.004
38900 39138.97 247.97 315.68 322.63 0.005 40275 40531.22 251.82 318.12 267.34 0.004
38925 39164.28 248.04 315.72 321.52 0.005 40300 40556.54 251.89 318.16 266.44 0.004
38950 39189.59 248.11 315.77 320.42 0.004 40325 40581.86 251.96 318.21 265.54 0.004
38975 39214.9 248.18 315.81 319.32 0.004 40350 40607.18 252.03 318.25 264.64 0.004
39000 39240.21 248.25 315.86 318.22 0.004 40375 40632.5 252.1 318.3 263.74 0.004
39025 39265.52 248.32 315.9 317.13 0.004 40400 40657.82 252.17 318.34 262.85 0.004
39050 39290.83 248.39 315.95 316.04 0.004 40425 40683.14 252.24 318.38 261.96 0.004
39075 39316.14 248.46 315.99 314.95 0.004 40450 40708.46 252.31 318.43 261.08 0.004
39100 39341.45 248.53 316.03 313.87 0.004 40475 40733.78 252.38 318.47 260.19 0.004
39125 39366.76 248.6 316.08 312.8 0.004 40500 40759.1 252.45 318.52 259.32 0.004
39150 39392.07 248.67 316.12 311.72 0.004 40525 40784.42 252.52 318.56 258.44 0.004
39175 39417.38 248.74 316.17 310.66 0.004 40550 40809.74 252.59 318.61 257.57 0.004

40575 40835.07 252.66 318.65 256.7 0.004 41950 42228.05 256.51 321.07 213.45 0.003
40600 40860.39 252.73 318.69 255.83 0.004 41975 42253.38 256.58 321.11 212.74 0.003
40625 40885.71 252.8 318.74 254.97 0.004 42000 42278.72 256.65 321.16 212.03 0.003
40650 40911.03 252.87 318.78 254.11 0.004 42025 42304.05 256.72 321.2 211.33 0.003
40675 40936.35 252.94 318.83 253.25 0.003 42050 42329.38 256.79 321.24 210.62 0.003
40700 40961.68 253.01 318.87 252.4 0.003 42075 42354.72 256.86 321.29 209.92 0.003
40725 40987 253.08 318.91 251.55 0.003 42100 42380.05 256.93 321.33 209.23 0.003
40750 41012.32 253.15 318.96 250.7 0.003 42125 42405.38 257 321.37 208.53 0.003
40775 41037.64 253.22 319 249.86 0.003 42150 42430.72 257.07 321.42 207.84 0.003
40800 41062.97 253.29 319.05 249.02 0.003 42175 42456.05 257.14 321.46 207.15 0.003
40825 41088.29 253.36 319.09 248.18 0.003 42200 42481.39 257.21 321.51 206.47 0.003
40850 41113.62 253.43 319.13 247.34 0.003 42225 42506.72 257.28 321.55 205.78 0.003
40875 41138.94 253.5 319.18 246.51 0.003 42250 42532.06 257.35 321.59 205.1 0.003
40900 41164.26 253.57 319.22 245.68 0.003 42275 42557.39 257.42 321.64 204.42 0.003
40925 41189.59 253.64 319.27 244.86 0.003 42300 42582.73 257.49 321.68 203.74 0.003
40950 41214.91 253.71 319.31 244.03 0.003 42325 42608.06 257.56 321.72 203.07 0.003
40975 41240.24 253.78 319.36 243.21 0.003 42350 42633.4 257.63 321.77 202.4 0.003
41000 41265.56 253.85 319.4 242.4 0.003 42375 42658.73 257.7 321.81 201.73 0.003
41025 41290.89 253.92 319.44 241.58 0.003 42400 42684.07 257.77 321.86 201.06 0.003
41050 41316.21 253.99 319.49 240.77 0.003 42425 42709.41 257.84 321.9 200.39 0.003
41075 41341.54 254.06 319.53 239.96 0.003 42450 42734.74 257.91 321.94 199.73 0.003
41100 41366.86 254.13 319.58 239.16 0.003 42475 42760.08 257.98 321.99 199.07 0.003
41125 41392.19 254.2 319.62 238.35 0.003 42500 42785.42 258.05 322.03 198.41 0.003
41150 41417.51 254.27 319.66 237.56 0.003 42525 42810.75 258.12 322.07 197.76 0.003
41175 41442.84 254.34 319.71 236.76 0.003 42550 42836.09 258.19 322.12 197.1 0.003
41200 41468.17 254.41 319.75 235.97 0.003 42575 42861.43 258.26 322.16 196.45 0.003
41225 41493.49 254.48 319.8 235.17 0.003 42600 42886.76 258.33 322.21 195.81 0.003
41250 41518.82 254.55 319.84 234.39 0.003 42625 42912.1 258.4 322.25 195.16 0.003
41275 41544.15 254.62 319.88 233.6 0.003 42650 42937.44 258.47 322.29 194.52 0.003
41300 41569.47 254.69 319.93 232.82 0.003 42675 42962.78 258.54 322.34 193.87 0.003
41325 41594.8 254.76 319.97 232.04 0.003 42700 42988.12 258.61 322.38 193.23 0.003
41350 41620.13 254.83 320.02 231.26 0.003 42725 43013.46 258.68 322.42 192.6 0.003
41375 41645.46 254.9 320.06 230.49 0.003 42750 43038.79 258.75 322.47 191.96 0.003
41400 41670.78 254.97 320.1 229.72 0.003 42775 43064.13 258.82 322.51 191.33 0.003
41425 41696.11 255.04 320.15 228.95 0.003 42800 43089.47 258.89 322.55 190.7 0.003
41450 41721.44 255.11 320.19 228.19 0.003 42825 43114.81 258.96 322.6 190.07 0.003
41475 41746.77 255.18 320.23 227.42 0.003 42850 43140.15 259.03 322.64 189.45 0.003
41500 41772.1 255.25 320.28 226.66 0.003 42875 43165.49 259.1 322.69 188.82 0.003
41525 41797.43 255.32 320.32 225.91 0.003 42900 43190.83 259.17 322.73 188.2 0.003
41550 41822.76 255.39 320.37 225.15 0.003 42925 43216.17 259.24 322.77 187.58 0.003
41575 41848.09 255.46 320.41 224.4 0.003 42950 43241.51 259.31 322.82 186.97 0.003
41600 41873.42 255.53 320.45 223.65 0.003 42975 43266.85 259.38 322.86 186.35 0.003
41625 41898.75 255.6 320.5 222.91 0.003 43000 43292.19 259.45 322.9 185.74 0.002
41650 41924.08 255.67 320.54 222.16 0.003 43025 43317.53 259.52 322.95 185.13 0.002
41675 41949.41 255.74 320.59 221.42 0.003 43050 43342.88 259.59 322.99 184.52 0.002
41700 41974.74 255.81 320.63 220.68 0.003 43075 43368.22 259.66 323.03 183.91 0.002
41725 42000.07 255.88 320.67 219.95 0.003 43100 43393.56 259.73 323.08 183.31 0.002
41750 42025.4 255.95 320.72 219.21 0.003 43125 43418.9 259.8 323.12 182.71 0.002
41775 42050.73 256.02 320.76 218.48 0.003 43150 43444.24 259.87 323.16 182.11 0.002
41800 42076.06 256.09 320.81 217.76 0.003 43175 43469.58 259.94 323.21 181.51 0.002
41825 42101.39 256.16 320.85 217.03 0.003 43200 43494.93 260.01 323.25 180.92 0.002
41850 42126.72 256.23 320.89 216.31 0.003 43225 43520.27 260.08 323.29 180.32 0.002
41875 42152.05 256.3 320.94 215.59 0.003 43250 43545.61 260.15 323.34 179.73 0.002
41900 42177.39 256.37 320.98 214.87 0.003 43275 43570.96 260.22 323.38 179.14 0.002
41925 42202.72 256.44 321.02 214.16 0.003 43300 43596.3 260.29 323.43 178.56 0.002

43325 43621.64 260.36 323.47 177.97 0.002 44700 45015.84 264.21 325.85 148.79 0.002
43350 43646.99 260.43 323.51 177.39 0.002 44725 45041.19 264.28 325.89 148.31 0.002
43375 43672.33 260.5 323.56 176.81 0.002 44750 45066.55 264.35 325.94 147.83 0.002
43400 43697.67 260.57 323.6 176.23 0.002 44775 45091.9 264.42 325.98 147.35 0.002
43425 43723.02 260.64 323.64 175.65 0.002 44800 45117.26 264.49 326.02 146.88 0.002
43450 43748.36 260.71 323.69 175.08 0.002 44825 45142.61 264.56 326.07 146.4 0.002
43475 43773.71 260.78 323.73 174.5 0.002 44850 45167.97 264.63 326.11 145.93 0.002
43500 43799.05 260.85 323.77 173.93 0.002 44875 45193.33 264.7 326.15 145.46 0.002
43525 43824.4 260.92 323.82 173.37 0.002 44900 45218.68 264.77 326.2 144.99 0.002
43550 43849.74 260.99 323.86 172.8 0.002 44925 45244.04 264.84 326.24 144.53 0.002
43575 43875.09 261.06 323.9 172.23 0.002 44950 45269.39 264.91 326.28 144.06 0.002
43600 43900.43 261.13 323.95 171.67 0.002 44975 45294.75 264.98 326.33 143.6 0.002
43625 43925.78 261.2 323.99 171.11 0.002 45000 45320.11 265.05 326.37 143.13 0.002
43650 43951.12 261.27 324.03 170.55 0.002 45025 45345.46 265.12 326.41 142.67 0.002
43675 43976.47 261.34 324.08 170 0.002 45050 45370.82 265.19 326.46 142.22 0.002
43700 44001.82 261.41 324.12 169.44 0.002 45075 45396.18 265.26 326.5 141.76 0.002
43725 44027.16 261.48 324.16 168.89 0.002 45100 45421.54 265.33 326.54 141.3 0.002
43750 44052.51 261.55 324.21 168.34 0.002 45125 45446.89 265.4 326.58 140.85 0.002
43775 44077.86 261.62 324.25 167.79 0.002 45150 45472.25 265.47 326.63 140.4 0.002
43800 44103.21 261.69 324.29 167.24 0.002 45175 45497.61 265.54 326.67 139.95 0.002
43825 44128.55 261.76 324.34 166.7 0.002 45200 45522.97 265.61 326.71 139.5 0.002
43850 44153.9 261.83 324.38 166.16 0.002 45225 45548.33 265.68 326.76 139.05 0.002
43875 44179.25 261.9 324.42 165.61 0.002 45250 45573.69 265.75 326.8 138.6 0.002
43900 44204.6 261.97 324.47 165.08 0.002 45275 45599.05 265.82 326.84 138.16 0.002
43925 44229.94 262.04 324.51 164.54 0.002 45300 45624.41 265.89 326.89 137.71 0.002
43950 44255.29 262.11 324.55 164 0.002 45325 45649.76 265.96 326.93 137.27 0.002
43975 44280.64 262.18 324.6 163.47 0.002 45350 45675.12 266.03 326.97 136.83 0.002
44000 44305.99 262.25 324.64 162.94 0.002 45375 45700.48 266.1 327.01 136.39 0.002
44025 44331.34 262.32 324.68 162.41 0.002 45400 45725.84 266.17 327.06 135.96 0.002
44050 44356.69 262.39 324.73 161.88 0.002 45425 45751.2 266.24 327.1 135.52 0.002
44075 44382.04 262.46 324.77 161.35 0.002 45450 45776.56 266.31 327.14 135.09 0.002
44100 44407.39 262.53 324.81 160.83 0.002 45475 45801.93 266.38 327.19 134.65 0.002
44125 44432.74 262.6 324.86 160.31 0.002 45500 45827.29 266.45 327.23 134.22 0.002
44150 44458.09 262.67 324.9 159.79 0.002 45525 45852.65 266.52 327.27 133.79 0.002
44175 44483.44 262.74 324.94 159.27 0.002 45550 45878.01 266.59 327.32 133.37 0.002
44200 44508.79 262.81 324.99 158.75 0.002 45575 45903.37 266.66 327.36 132.94 0.002
44225 44534.14 262.88 325.03 158.24 0.002 45600 45928.73 266.73 327.4 132.51 0.002
44250 44559.49 262.95 325.07 157.72 0.002 45625 45954.09 266.8 327.44 132.09 0.002
44275 44584.84 263.02 325.12 157.21 0.002 45650 45979.46 266.87 327.49 131.67 0.002
44300 44610.19 263.09 325.16 156.7 0.002 45675 46004.82 266.94 327.53 131.25 0.002
44325 44635.54 263.16 325.2 156.19 0.002 45700 46030.18 267.01 327.57 130.83 0.002
44350 44660.89 263.23 325.25 155.69 0.002 45725 46055.54 267.08 327.62 130.41 0.002
44375 44686.25 263.3 325.29 155.18 0.002 45750 46080.91 267.15 327.66 130 0.002
44400 44711.6 263.37 325.33 154.68 0.002 45775 46106.27 267.22 327.7 129.58 0.002
44425 44736.95 263.44 325.38 154.18 0.002 45800 46131.63 267.29 327.75 129.17 0.002
44450 44762.3 263.51 325.42 153.68 0.002 45825 46157 267.36 327.79 128.75 0.002
44475 44787.66 263.58 325.46 153.18 0.002 45850 46182.36 267.43 327.83 128.34 0.002
44500 44813.01 263.65 325.51 152.69 0.002 45875 46207.72 267.5 327.87 127.94 0.002
44525 44838.36 263.72 325.55 152.2 0.002 45900 46233.09 267.57 327.92 127.53 0.002
44550 44863.72 263.79 325.59 151.7 0.002 45925 46258.45 267.64 327.96 127.12 0.002
44575 44889.07 263.86 325.64 151.21 0.002 45950 46283.82 267.71 328 126.72 0.002
44600 44914.42 263.93 325.68 150.72 0.002 45975 46309.18 267.78 328.05 126.31 0.002
44625 44939.78 264 325.72 150.24 0.002 46000 46334.55 267.85 328.09 125.91 0.002
44650 44965.13 264.07 325.77 149.75 0.002 46025 46359.91 267.92 328.13 125.51 0.002
44675 44990.48 264.14 325.81 149.27 0.002 46050 46385.28 267.99 328.17 125.11 0.002

46075 46410.64 268.06 328.22 124.71 0.002 47450 47806.05 270.65 329.8 104.78 0.001
46100 46436.01 268.13 328.26 124.32 0.002 47475 47831.43 270.65 329.8 104.45 0.001
46125 46461.37 268.2 328.3 123.92 0.002 47500 47856.8 270.65 329.8 104.12 0.001
46150 46486.74 268.27 328.35 123.53 0.002 47525 47882.18 270.65 329.8 103.79 0.001
46175 46512.11 268.34 328.39 123.13 0.002 47550 47907.56 270.65 329.8 103.47 0.001
46200 46537.47 268.41 328.43 122.74 0.002 47575 47932.94 270.65 329.8 103.14 0.001
46225 46562.84 268.48 328.47 122.35 0.002 47600 47958.31 270.65 329.8 102.82 0.001
46250 46588.21 268.55 328.52 121.96 0.002 47625 47983.69 270.65 329.8 102.49 0.001
46275 46613.57 268.62 328.56 121.58 0.002 47650 48009.07 270.65 329.8 102.17 0.001
46300 46638.94 268.69 328.6 121.19 0.002 47675 48034.45 270.65 329.8 101.85 0.001
46325 46664.31 268.76 328.65 120.81 0.002 47700 48059.83 270.65 329.8 101.53 0.001
46350 46689.67 268.83 328.69 120.42 0.002 47725 48085.21 270.65 329.8 101.21 0.001
46375 46715.04 268.9 328.73 120.04 0.002 47750 48110.58 270.65 329.8 100.89 0.001
46400 46740.41 268.97 328.77 119.66 0.002 47775 48135.96 270.65 329.8 100.57 0.001
46425 46765.78 269.04 328.82 119.28 0.002 47800 48161.34 270.65 329.8 100.25 0.001
46450 46791.15 269.11 328.86 118.9 0.002 47825 48186.72 270.65 329.8 99.94 0.001
46475 46816.52 269.18 328.9 118.53 0.002 47850 48212.1 270.65 329.8 99.62 0.001
46500 46841.88 269.25 328.94 118.15 0.002 47875 48237.48 270.65 329.8 99.31 0.001
46525 46867.25 269.32 328.99 117.78 0.002 47900 48262.86 270.65 329.8 99 0.001
46550 46892.62 269.39 329.03 117.4 0.002 47925 48288.24 270.65 329.8 98.68 0.001
46575 46917.99 269.46 329.07 117.03 0.002 47950 48313.62 270.65 329.8 98.37 0.001
46600 46943.36 269.53 329.12 116.66 0.002 47975 48339 270.65 329.8 98.06 0.001
46625 46968.73 269.6 329.16 116.29 0.002 48000 48364.38 270.65 329.8 97.75 0.001
46650 46994.1 269.67 329.2 115.93 0.001 48025 48389.77 270.65 329.8 97.45 0.001
46675 47019.47 269.74 329.24 115.56 0.001 48050 48415.15 270.65 329.8 97.14 0.001
46700 47044.84 269.81 329.29 115.19 0.001 48075 48440.53 270.65 329.8 96.83 0.001
46725 47070.21 269.88 329.33 114.83 0.001 48100 48465.91 270.65 329.8 96.53 0.001
46750 47095.58 269.95 329.37 114.47 0.001 48125 48491.29 270.65 329.8 96.22 0.001
46775 47120.96 270.02 329.41 114.11 0.001 48150 48516.67 270.65 329.8 95.92 0.001
46800 47146.33 270.09 329.46 113.74 0.001 48175 48542.06 270.65 329.8 95.62 0.001
46825 47171.7 270.16 329.5 113.39 0.001 48200 48567.44 270.65 329.8 95.32 0.001
46850 47197.07 270.23 329.54 113.03 0.001 48225 48592.82 270.65 329.8 95.02 0.001
46875 47222.44 270.3 329.59 112.67 0.001 48250 48618.2 270.65 329.8 94.72 0.001
46900 47247.81 270.37 329.63 112.32 0.001 48275 48643.59 270.65 329.8 94.42 0.001
46925 47273.19 270.44 329.67 111.96 0.001 48300 48668.97 270.65 329.8 94.12 0.001
46950 47298.56 270.51 329.71 111.61 0.001 48325 48694.35 270.65 329.8 93.83 0.001
46975 47323.93 270.58 329.76 111.26 0.001 48350 48719.74 270.65 329.8 93.53 0.001
47000 47349.3 270.65 329.8 110.91 0.001 48375 48745.12 270.65 329.8 93.24 0.001
47025 47374.68 270.65 329.8 110.56 0.001 48400 48770.51 270.65 329.8 92.94 0.001
47050 47400.05 270.65 329.8 110.21 0.001 48425 48795.89 270.65 329.8 92.65 0.001
47075 47425.42 270.65 329.8 109.86 0.001 48450 48821.27 270.65 329.8 92.36 0.001
47100 47450.8 270.65 329.8 109.52 0.001 48475 48846.66 270.65 329.8 92.07 0.001
47125 47476.17 270.65 329.8 109.17 0.001 48500 48872.04 270.65 329.8 91.78 0.001
47150 47501.55 270.65 329.8 108.83 0.001 48525 48897.43 270.65 329.8 91.49 0.001
47175 47526.92 270.65 329.8 108.48 0.001 48550 48922.81 270.65 329.8 91.2 0.001
47200 47552.29 270.65 329.8 108.14 0.001 48575 48948.2 270.65 329.8 90.91 0.001
47225 47577.67 270.65 329.8 107.8 0.001 48600 48973.59 270.65 329.8 90.62 0.001
47250 47603.04 270.65 329.8 107.46 0.001 48625 48998.97 270.65 329.8 90.34 0.001
47275 47628.42 270.65 329.8 107.12 0.001 48650 49024.36 270.65 329.8 90.05 0.001
47300 47653.79 270.65 329.8 106.79 0.001 48675 49049.74 270.65 329.8 89.77 0.001
47325 47679.17 270.65 329.8 106.45 0.001 48700 49075.13 270.65 329.8 89.49 0.001
47350 47704.55 270.65 329.8 106.11 0.001 48725 49100.52 270.65 329.8 89.21 0.001
47375 47729.92 270.65 329.8 105.78 0.001 48750 49125.9 270.65 329.8 88.92 0.001
47400 47755.3 270.65 329.8 105.45 0.001 48775 49151.29 270.65 329.8 88.64 0.001
47425 47780.67 270.65 329.8 105.11 0.001 48800 49176.68 270.65 329.8 88.37 0.001

48825 49202.07 270.65 329.8 88.09 0.001 49400 49786.04 270.65 329.8 81.92 0.001
48850 49227.45 270.65 329.8 87.81 0.001 49425 49811.43 270.65 329.8 81.66 0.001
48875 49252.84 270.65 329.8 87.53 0.001 49450 49836.82 270.65 329.8 81.4 0.001
48900 49278.23 270.65 329.8 87.26 0.001 49475 49862.21 270.65 329.8 81.15 0.001
48925 49303.62 270.65 329.8 86.98 0.001 49500 49887.61 270.65 329.8 80.89 0.001
48950 49329.01 270.65 329.8 86.71 0.001 49525 49913 270.65 329.8 80.64 0.001
48975 49354.4 270.65 329.8 86.43 0.001 49550 49938.39 270.65 329.8 80.38 0.001
49000 49379.78 270.65 329.8 86.16 0.001 49575 49963.79 270.65 329.8 80.13 0.001
49025 49405.17 270.65 329.8 85.89 0.001 49600 49989.18 270.65 329.8 79.88 0.001
49050 49430.56 270.65 329.8 85.62 0.001 49625 50014.57 270.65 329.8 79.63 0.001
49075 49455.95 270.65 329.8 85.35 0.001 49650 50039.97 270.65 329.8 79.38 0.001
49100 49481.34 270.65 329.8 85.08 0.001 49675 50065.36 270.65 329.8 79.13 0.001
49125 49506.73 270.65 329.8 84.81 0.001 49700 50090.76 270.65 329.8 78.88 0.001
49150 49532.12 270.65 329.8 84.55 0.001 49725 50116.15 270.65 329.8 78.63 0.001
49175 49557.51 270.65 329.8 84.28 0.001 49750 50141.55 270.65 329.8 78.38 0.001
49200 49582.9 270.65 329.8 84.01 0.001 49775 50166.94 270.65 329.8 78.13 0.001
49225 49608.29 270.65 329.8 83.75 0.001 49800 50192.34 270.65 329.8 77.89 0.001
49250 49633.69 270.65 329.8 83.49 0.001 49825 50217.73 270.65 329.8 77.64 0.001
49275 49659.08 270.65 329.8 83.22 0.001 49850 50243.13 270.65 329.8 77.4 0.001
49300 49684.47 270.65 329.8 82.96 0.001 49875 50268.52 270.65 329.8 77.15 0.001
49325 49709.86 270.65 329.8 82.7 0.001 49900 50293.92 270.65 329.8 76.91 0.001
49350 49735.25 270.65 329.8 82.44 0.001 49925 50319.32 270.65 329.8 76.67 0.001
49375 49760.64 270.65 329.8 82.18 0.001 49950 50344.71 270.65 329.8 76.43 0.001
49975 50370.11 270.65 329.8 76.18 0.001
50000 50395.51 270.65 329.8 75.94 0.001
Isentropic Table (γ = 1.4)
D
p T ρ a A p T ρ a A
M p0 T0 ρ0 a0 A∗ M p0 T0 ρ0 a0 A∗
0.00 1.0000 1.0000 1.0000 1.0000 ∞ 0.40 0.8956 0.9690 0.9243 0.9844 1.5901
0.01 0.9999 1.0000 1.0000 1.0000 57.8738 0.41 0.8907 0.9675 0.9207 0.9836 1.5587
0.02 0.9997 0.9999 0.9998 1.0000 28.9421 0.42 0.8857 0.9659 0.9170 0.9828 1.5289
0.03 0.9994 0.9998 0.9996 0.9999 19.3005 0.43 0.8807 0.9643 0.9132 0.9820 1.5007
0.04 0.9989 0.9997 0.9992 0.9998 14.4815 0.44 0.8755 0.9627 0.9094 0.9812 1.4740
0.05 0.9983 0.9995 0.9988 0.9998 11.5914 0.45 0.8703 0.9611 0.9055 0.9803 1.4487
0.06 0.9975 0.9993 0.9982 0.9996 9.6659 0.46 0.8650 0.9594 0.9016 0.9795 1.4246
0.07 0.9966 0.9990 0.9976 0.9995 8.2915 0.47 0.8596 0.9577 0.8976 0.9786 1.4018
0.08 0.9955 0.9987 0.9968 0.9994 7.2616 0.48 0.8541 0.9559 0.8935 0.9777 1.3801
0.09 0.9944 0.9984 0.9960 0.9992 6.4613 0.49 0.8486 0.9542 0.8894 0.9768 1.3595
0.10 0.9930 0.9980 0.9950 0.9990 5.8218 0.50 0.8430 0.9524 0.8852 0.9759 1.3398
0.11 0.9916 0.9976 0.9940 0.9988 5.2992 0.51 0.8374 0.9506 0.8809 0.9750 1.3212
0.12 0.9900 0.9971 0.9928 0.9986 4.8643 0.52 0.8317 0.9487 0.8766 0.9740 1.3034
0.13 0.9883 0.9966 0.9916 0.9983 4.4969 0.53 0.8259 0.9468 0.8723 0.9730 1.2865
0.14 0.9864 0.9961 0.9903 0.9980 4.1824 0.54 0.8201 0.9449 0.8679 0.9721 1.2703
0.15 0.9844 0.9955 0.9888 0.9978 3.9103 0.55 0.8142 0.9430 0.8634 0.9711 1.2549
0.16 0.9823 0.9949 0.9873 0.9974 3.6727 0.56 0.8082 0.9410 0.8589 0.9700 1.2403
0.17 0.9800 0.9943 0.9857 0.9971 3.4635 0.57 0.8022 0.9390 0.8544 0.9690 1.2263
0.18 0.9776 0.9936 0.9840 0.9968 3.2779 0.58 0.7962 0.9370 0.8498 0.9680 1.2130
0.19 0.9751 0.9928 0.9822 0.9964 3.1123 0.59 0.7901 0.9349 0.8451 0.9669 1.2003
0.20 0.9725 0.9921 0.9803 0.9960 2.9635 0.60 0.7840 0.9328 0.8405 0.9658 1.1882
0.21 0.9697 0.9913 0.9783 0.9956 2.8293 0.61 0.7778 0.9307 0.8357 0.9647 1.1767
0.22 0.9668 0.9904 0.9762 0.9952 2.7076 0.62 0.7716 0.9286 0.8310 0.9636 1.1656
0.23 0.9638 0.9895 0.9740 0.9948 2.5968 0.63 0.7654 0.9265 0.8262 0.9625 1.1552
0.24 0.9607 0.9886 0.9718 0.9943 2.4956 0.64 0.7591 0.9243 0.8213 0.9614 1.1451
0.25 0.9575 0.9877 0.9694 0.9938 2.4027 0.65 0.7528 0.9221 0.8164 0.9603 1.1356
0.26 0.9541 0.9867 0.9670 0.9933 2.3173 0.66 0.7465 0.9199 0.8115 0.9591 1.1265
0.27 0.9506 0.9856 0.9645 0.9928 2.2385 0.67 0.7401 0.9176 0.8066 0.9579 1.1179
0.28 0.9470 0.9846 0.9619 0.9923 2.1656 0.68 0.7338 0.9153 0.8016 0.9567 1.1097
0.29 0.9433 0.9835 0.9592 0.9917 2.0979 0.69 0.7274 0.9131 0.7966 0.9555 1.1018
0.30 0.9395 0.9823 0.9564 0.9911 2.0351 0.70 0.7209 0.9107 0.7916 0.9543 1.0944
0.31 0.9355 0.9811 0.9535 0.9905 1.9765 0.71 0.7145 0.9084 0.7865 0.9531 1.0873
0.32 0.9315 0.9799 0.9506 0.9899 1.9219 0.72 0.7080 0.9061 0.7814 0.9519 1.0806
0.33 0.9274 0.9787 0.9476 0.9893 1.8707 0.73 0.7016 0.9037 0.7763 0.9506 1.0742
0.34 0.9231 0.9774 0.9445 0.9886 1.8229 0.74 0.6951 0.9013 0.7712 0.9494 1.0681
0.35 0.9188 0.9761 0.9413 0.9880 1.7780 0.75 0.6886 0.8989 0.7660 0.9481 1.0624
0.36 0.9143 0.9747 0.9380 0.9873 1.7358 0.76 0.6821 0.8964 0.7609 0.9468 1.0570
0.37 0.9098 0.9733 0.9347 0.9866 1.6961 0.77 0.6756 0.8940 0.7557 0.9455 1.0519
0.38 0.9052 0.9719 0.9313 0.9859 1.6587 0.78 0.6691 0.8915 0.7505 0.9442 1.0471
0.39 0.9004 0.9705 0.9278 0.9851 1.6234

432 Appendix D: Isentropic Table (γ = 1.4)
M p0 T0 ρ0 a0 A∗ M p0 T0 ρ0 a0 A∗
0.79 0.6625 0.8890 0.7452 0.9429 1.0425 1.42 0.3055 0.7126 0.4287 0.8442 1.1262
0.80 0.6560 0.8865 0.7400 0.9416 1.0382 1.43 0.3012 0.7097 0.4244 0.8425 1.1320
0.81 0.6495 0.8840 0.7347 0.9402 1.0342 1.44 0.2969 0.7069 0.4201 0.8407 1.1379
0.82 0.6430 0.8815 0.7295 0.9389 1.0305 1.45 0.2927 0.7040 0.4158 0.8390 1.1440
0.83 0.6365 0.8789 0.7242 0.9375 1.0270 1.46 0.2886 0.7011 0.4116 0.8373 1.1501
0.84 0.6300 0.8763 0.7189 0.9361 1.0237 1.47 0.2845 0.6982 0.4074 0.8356 1.1565
0.85 0.6235 0.8737 0.7136 0.9347 1.0207 1.48 0.2804 0.6954 0.4032 0.8339 1.1629
0.86 0.6170 0.8711 0.7083 0.9333 1.0179 1.49 0.2764 0.6925 0.3991 0.8322 1.1695
0.87 0.6106 0.8685 0.7030 0.9319 1.0153 1.50 0.2724 0.6897 0.3950 0.8305 1.1762
0.88 0.6041 0.8659 0.6977 0.9305 1.0129 1.51 0.2685 0.6868 0.3909 0.8287 1.1830
0.89 0.5977 0.8632 0.6924 0.9291 1.0108 1.52 0.2646 0.6840 0.3869 0.8270 1.1899
0.90 0.5913 0.8606 0.6870 0.9277 1.0089 1.53 0.2608 0.6811 0.3829 0.8253 1.1970
0.91 0.5849 0.8579 0.6817 0.9262 1.0071 1.54 0.2570 0.6783 0.3789 0.8236 1.2042
0.92 0.5785 0.8552 0.6764 0.9248 1.0056 1.55 0.2533 0.6754 0.3750 0.8219 1.2116
0.93 0.5721 0.8525 0.6711 0.9233 1.0043 1.56 0.2496 0.6726 0.3710 0.8201 1.2190
0.94 0.5658 0.8498 0.6658 0.9219 1.0031 1.57 0.2459 0.6698 0.3672 0.8184 1.2266
0.95 0.5595 0.8471 0.6604 0.9204 1.0021 1.58 0.2423 0.6670 0.3633 0.8167 1.2344
0.96 0.5532 0.8444 0.6551 0.9189 1.0014 1.59 0.2388 0.6642 0.3595 0.8150 1.2422
0.97 0.5469 0.8416 0.6498 0.9174 1.0008 1.60 0.2353 0.6614 0.3557 0.8133 1.2502
0.98 0.5407 0.8389 0.6445 0.9159 1.0003 1.61 0.2318 0.6586 0.3520 0.8115 1.2584
0.99 0.5345 0.8361 0.6392 0.9144 1.0001 1.62 0.2284 0.6558 0.3483 0.8098 1.2666
1.00 0.5283 0.8333 0.6339 0.9129 1.0000 1.63 0.2250 0.6530 0.3446 0.8081 1.2750
1.01 0.5221 0.8306 0.6287 0.9113 1.0001 1.64 0.2217 0.6502 0.3409 0.8064 1.2836
1.02 0.5160 0.8278 0.6234 0.9098 1.0003 1.65 0.2184 0.6475 0.3373 0.8046 1.2922
1.03 0.5099 0.8250 0.6181 0.9083 1.0007 1.66 0.2151 0.6447 0.3337 0.8029 1.3010
1.04 0.5039 0.8222 0.6129 0.9067 1.0013 1.67 0.2119 0.6419 0.3302 0.8012 1.3100
1.05 0.4979 0.8193 0.6077 0.9052 1.0020 1.68 0.2088 0.6392 0.3266 0.7995 1.3190
1.06 0.4919 0.8165 0.6024 0.9036 1.0029 1.69 0.2057 0.6364 0.3232 0.7978 1.3283
1.07 0.4860 0.8137 0.5972 0.9020 1.0039 1.70 0.2026 0.6337 0.3197 0.7961 1.3376
1.08 0.4800 0.8108 0.5920 0.9005 1.0051 1.71 0.1996 0.6310 0.3163 0.7943 1.3471
1.09 0.4742 0.8080 0.5869 0.8989 1.0064 1.72 0.1966 0.6283 0.3129 0.7926 1.3567
1.10 0.4684 0.8052 0.5817 0.8973 1.0079 1.73 0.1936 0.6256 0.3095 0.7909 1.3665
1.11 0.4626 0.8023 0.5766 0.8957 1.0095 1.74 0.1907 0.6229 0.3062 0.7892 1.3764
1.12 0.4568 0.7994 0.5714 0.8941 1.0113 1.75 0.1878 0.6202 0.3029 0.7875 1.3865
1.13 0.4511 0.7966 0.5663 0.8925 1.0132 1.76 0.1850 0.6175 0.2996 0.7858 1.3967
1.14 0.4455 0.7937 0.5612 0.8909 1.0153 1.77 0.1822 0.6148 0.2964 0.7841 1.4070
1.15 0.4398 0.7908 0.5562 0.8893 1.0175 1.78 0.1794 0.6121 0.2931 0.7824 1.4175
1.16 0.4343 0.7879 0.5511 0.8877 1.0198 1.79 0.1767 0.6095 0.2900 0.7807 1.4282
1.17 0.4287 0.7851 0.5461 0.8860 1.0222 1.80 0.1740 0.6068 0.2868 0.7790 1.4390
1.18 0.4232 0.7822 0.5411 0.8844 1.0248 1.81 0.1714 0.6041 0.2837 0.7773 1.4499
1.19 0.4178 0.7793 0.5361 0.8828 1.0276 1.82 0.1688 0.6015 0.2806 0.7756 1.4610
1.20 0.4124 0.7764 0.5311 0.8811 1.0304 1.83 0.1662 0.5989 0.2776 0.7739 1.4723
1.21 0.4070 0.7735 0.5262 0.8795 1.0334 1.84 0.1637 0.5963 0.2745 0.7722 1.4836
1.22 0.4017 0.7706 0.5213 0.8778 1.0366 1.85 0.1612 0.5936 0.2715 0.7705 1.4952
1.23 0.3964 0.7677 0.5164 0.8762 1.0398 1.86 0.1587 0.5910 0.2686 0.7688 1.5069
1.24 0.3912 0.7648 0.5115 0.8745 1.0432 1.87 0.1563 0.5884 0.2656 0.7671 1.5187
1.25 0.3861 0.7619 0.5067 0.8729 1.0468 1.88 0.1539 0.5859 0.2627 0.7654 1.5308
1.26 0.3809 0.7590 0.5019 0.8712 1.0504 1.89 0.1516 0.5833 0.2598 0.7637 1.5429
1.27 0.3759 0.7561 0.4971 0.8695 1.0542 1.90 0.1492 0.5807 0.2570 0.7620 1.5553
1.28 0.3708 0.7532 0.4923 0.8679 1.0581 1.91 0.1470 0.5782 0.2542 0.7604 1.5677
1.29 0.3658 0.7503 0.4876 0.8662 1.0621 1.92 0.1447 0.5756 0.2514 0.7587 1.5804
1.30 0.3609 0.7474 0.4829 0.8645 1.0663 1.93 0.1425 0.5731 0.2486 0.7570 1.5932
1.31 0.3560 0.7445 0.4782 0.8628 1.0706 1.94 0.1403 0.5705 0.2459 0.7553 1.6062
1.32 0.3512 0.7416 0.4736 0.8611 1.0750 1.95 0.1381 0.5680 0.2432 0.7537 1.6193
1.35 0.3370 0.7329 0.4598 0.8561 1.0890 1.96 0.1360 0.5655 0.2405 0.7520 1.6326
1.36 0.3323 0.7300 0.4553 0.8544 1.0940 1.97 0.1339 0.5630 0.2378 0.7503 1.6461
1.37 0.3277 0.7271 0.4508 0.8527 1.0990 1.98 0.1318 0.5605 0.2352 0.7487 1.6597
1.38 0.3232 0.7242 0.4463 0.8510 1.1042 1.99 0.1298 0.5580 0.2326 0.7470 1.6735
1.39 0.3187 0.7213 0.4418 0.8493 1.1095 2.00 0.1278 0.5556 0.2300 0.7454 1.6875
1.40 0.3142 0.7184 0.4374 0.8476 1.1149 2.01 0.1258 0.5531 0.2275 0.7437 1.7016
1.41 0.3098 0.7155 0.4330 0.8459 1.1205 2.02 0.1239 0.5506 0.2250 0.7420 1.7160
2.03 0.1220 0.5482 0.2225 0.7404 1.7305
2.04 0.1201 0.5458 0.2200 0.7388 1.7451
Appendix D: Isentropic Table (γ = 1.4) 433
M p0 T0 ρ0 a0 A∗ M p0 T0 ρ0 a0 A∗
2.05 0.1182 0.5433 0.2176 0.7371 1.7600 2.67 0.0450 0.4122 0.1091 0.6421 3.0938
2.06 0.1164 0.5409 0.2152 0.7355 1.7750 2.68 0.0443 0.4104 0.1079 0.6406 3.1233
2.07 0.1146 0.5385 0.2128 0.7338 1.7902 2.69 0.0436 0.4086 0.1067 0.6392 3.1530
2.08 0.1128 0.5361 0.2104 0.7322 1.8056 2.70 0.0430 0.4068 0.1056 0.6378 3.1830
2.09 0.1111 0.5337 0.2081 0.7306 1.8212 2.71 0.0423 0.4051 0.1044 0.6364 3.2133
2.10 0.1094 0.5313 0.2058 0.7289 1.8369 2.72 0.0417 0.4033 0.1033 0.6350 3.2440
2.11 0.1077 0.5290 0.2035 0.7273 1.8529 2.73 0.0410 0.4015 0.1022 0.6337 3.2749
2.12 0.1060 0.5266 0.2013 0.7257 1.8690 2.74 0.0404 0.3998 0.1010 0.6323 3.3061
2.13 0.1043 0.5243 0.1990 0.7241 1.8853 2.75 0.0398 0.3980 0.0999 0.6309 3.3377
2.14 0.1027 0.5219 0.1968 0.7225 1.9018 2.76 0.0392 0.3963 0.0989 0.6295 3.3695
2.15 0.1011 0.5196 0.1946 0.7208 1.9185 2.77 0.0386 0.3945 0.0978 0.6281 3.4017
2.16 0.0996 0.5173 0.1925 0.7192 1.9354 2.78 0.0380 0.3928 0.0967 0.6268 3.4342
2.17 0.0980 0.5150 0.1903 0.7176 1.9525 2.79 0.0374 0.3911 0.0957 0.6254 3.4670
2.18 0.0965 0.5127 0.1882 0.7160 1.9698 2.80 0.0368 0.3894 0.0946 0.6240 3.5001
2.19 0.0950 0.5104 0.1861 0.7144 1.9873 2.81 0.0363 0.3877 0.0936 0.6227 3.5336
2.20 0.0935 0.5081 0.1841 0.7128 2.0050 2.82 0.0357 0.3860 0.0926 0.6213 3.5674
2.21 0.0921 0.5059 0.1820 0.7112 2.0229 2.83 0.0352 0.3844 0.0916 0.6200 3.6015
2.22 0.0906 0.5036 0.1800 0.7097 2.0409 2.84 0.0347 0.3827 0.0906 0.6186 3.6359
2.23 0.0892 0.5014 0.1780 0.7081 2.0592 2.85 0.0341 0.3810 0.0896 0.6173 3.6707
2.24 0.0878 0.4991 0.1760 0.7065 2.0777 2.86 0.0336 0.3794 0.0886 0.6159 3.7058
2.25 0.0865 0.4969 0.1740 0.7049 2.0964 2.87 0.0331 0.3777 0.0877 0.6146 3.7413
2.26 0.0851 0.4947 0.1721 0.7033 2.1153 2.88 0.0326 0.3761 0.0867 0.6133 3.7771
2.27 0.0838 0.4925 0.1702 0.7018 2.1345 2.89 0.0321 0.3745 0.0858 0.6119 3.8133
2.28 0.0825 0.4903 0.1683 0.7002 2.1538 2.90 0.0317 0.3729 0.0849 0.6106 3.8498
2.29 0.0812 0.4881 0.1664 0.6986 2.1734 2.91 0.0312 0.3712 0.0840 0.6093 3.8866
2.30 0.0800 0.4859 0.1646 0.6971 2.1931 2.92 0.0307 0.3696 0.0831 0.6080 3.9238
2.31 0.0787 0.4837 0.1628 0.6955 2.2131 2.93 0.0302 0.3681 0.0822 0.6067 3.9614
2.32 0.0775 0.4816 0.1609 0.6940 2.2333 2.94 0.0298 0.3665 0.0813 0.6054 3.9993
2.33 0.0763 0.4794 0.1592 0.6924 2.2538 2.95 0.0293 0.3649 0.0804 0.6041 4.0376
2.34 0.0751 0.4773 0.1574 0.6909 2.2744 2.96 0.0289 0.3633 0.0796 0.6028 4.0763
2.35 0.0740 0.4752 0.1556 0.6893 2.2953 2.97 0.0285 0.3618 0.0787 0.6015 4.1153
2.36 0.0728 0.4731 0.1539 0.6878 2.3164 2.98 0.0281 0.3602 0.0779 0.6002 4.1547
2.37 0.0717 0.4709 0.1522 0.6863 2.3377 2.99 0.0276 0.3587 0.0770 0.5989 4.1944
2.38 0.0706 0.4688 0.1505 0.6847 2.3593 3.00 0.0272 0.3571 0.0762 0.5976 4.2346
2.39 0.0695 0.4668 0.1488 0.6832 2.3811 3.01 0.0268 0.3556 0.0754 0.5963 4.2751
2.40 0.0684 0.4647 0.1472 0.6817 2.4031 3.02 0.0264 0.3541 0.0746 0.5951 4.3160
2.41 0.0673 0.4626 0.1456 0.6802 2.4254 3.03 0.0260 0.3526 0.0738 0.5938 4.3573
2.42 0.0663 0.4606 0.1439 0.6786 2.4479 3.04 0.0256 0.3511 0.0730 0.5925 4.3989
2.43 0.0653 0.4585 0.1424 0.6771 2.4706 3.05 0.0253 0.3496 0.0723 0.5913 4.4410
2.44 0.0643 0.4565 0.1408 0.6756 2.4936 3.06 0.0249 0.3481 0.0715 0.5900 4.4835
2.45 0.0633 0.4544 0.1392 0.6741 2.5168 3.07 0.0245 0.3466 0.0707 0.5887 4.5263
2.46 0.0623 0.4524 0.1377 0.6726 2.5403 3.08 0.0242 0.3452 0.0700 0.5875 4.5696
2.47 0.0613 0.4504 0.1362 0.6711 2.5640 3.09 0.0238 0.3437 0.0692 0.5862 4.6132
2.48 0.0604 0.4484 0.1346 0.6696 2.5880 3.10 0.0234 0.3422 0.0685 0.5850 4.6573
2.49 0.0594 0.4464 0.1332 0.6682 2.6122 3.11 0.0231 0.3408 0.0678 0.5838 4.7018
2.50 0.0585 0.4444 0.1317 0.6667 2.6367 3.12 0.0228 0.3393 0.0671 0.5825 4.7467
2.51 0.0576 0.4425 0.1302 0.6652 2.6615 3.13 0.0224 0.3379 0.0664 0.5813 4.7920
2.52 0.0567 0.4405 0.1288 0.6637 2.6865 3.14 0.0221 0.3365 0.0657 0.5801 4.8377
2.53 0.0559 0.4386 0.1274 0.6622 2.7117 3.15 0.0218 0.3351 0.0650 0.5788 4.8838
2.54 0.0550 0.4366 0.1260 0.6608 2.7372 3.16 0.0215 0.3337 0.0643 0.5776 4.9304
2.55 0.0542 0.4347 0.1246 0.6593 2.7630 3.17 0.0211 0.3323 0.0636 0.5764 4.9774
2.56 0.0533 0.4328 0.1232 0.6578 2.7891 3.18 0.0208 0.3309 0.0630 0.5752 5.0248
2.57 0.0525 0.4309 0.1218 0.6564 2.8154 3.19 0.0205 0.3295 0.0623 0.5740 5.0727
2.58 0.0517 0.4289 0.1205 0.6549 2.8420 3.20 0.0202 0.3281 0.0617 0.5728 5.1210
2.59 0.0509 0.4271 0.1192 0.6535 2.8688 3.21 0.0199 0.3267 0.0610 0.5716 5.1697
2.60 0.0501 0.4252 0.1179 0.6521 2.8960 3.22 0.0196 0.3253 0.0604 0.5704 5.2189
2.61 0.0493 0.4233 0.1166 0.6506 2.9234 3.23 0.0194 0.3240 0.0597 0.5692 5.2685
2.62 0.0486 0.4214 0.1153 0.6492 2.9511 3.24 0.0191 0.3226 0.0591 0.5680 5.3186
2.63 0.0478 0.4196 0.1140 0.6477 2.9791 3.25 0.0188 0.3213 0.0585 0.5668 5.3691
2.64 0.0471 0.4177 0.1128 0.6463 3.0073 3.26 0.0185 0.3199 0.0579 0.5656 5.4201
2.65 0.0464 0.4159 0.1115 0.6449 3.0359 3.27 0.0183 0.3186 0.0573 0.5645 5.4715
2.66 0.0457 0.4141 0.1103 0.6435 3.0647 3.28 0.0180 0.3173 0.0567 0.5633 5.5234
M p0 T0 ρ0 a0 A∗ M p0 T0 ρ0 a0 A∗
3.29 0.0177 0.3160 0.0561 0.5621 5.5758 3.91 0.0074 0.2464 0.0302 0.4964 9.8877
3.30 0.0175 0.3147 0.0555 0.5609 5.6286 3.92 0.0073 0.2455 0.0299 0.4955 9.9771
3.31 0.0172 0.3134 0.0550 0.5598 5.6820 3.93 0.0072 0.2446 0.0296 0.4945 10.0672
3.32 0.0170 0.3121 0.0544 0.5586 5.7358 3.94 0.0071 0.2436 0.0293 0.4936 10.1581
3.33 0.0167 0.3108 0.0538 0.5575 5.7900 3.95 0.0070 0.2427 0.0290 0.4926 10.2496
3.34 0.0165 0.3095 0.0533 0.5563 5.8448 3.96 0.0069 0.2418 0.0287 0.4917 10.3420
3.35 0.0163 0.3082 0.0527 0.5552 5.9000 3.97 0.0069 0.2408 0.0285 0.4908 10.4350
3.36 0.0160 0.3069 0.0522 0.5540 5.9558 3.98 0.0068 0.2399 0.0282 0.4898 10.5289
3.37 0.0158 0.3057 0.0517 0.5529 6.0120 3.99 0.0067 0.2390 0.0279 0.4889 10.6234
3.38 0.0156 0.3044 0.0511 0.5517 6.0687 4.00 0.0066 0.2381 0.0277 0.4880 10.7188
3.39 0.0153 0.3032 0.0506 0.5506 6.1260 4.01 0.0065 0.2372 0.0274 0.4870 10.8148
3.40 0.0151 0.3019 0.0501 0.5495 6.1837 4.02 0.0064 0.2363 0.0271 0.4861 10.9117
3.41 0.0149 0.3007 0.0496 0.5484 6.2419 4.03 0.0063 0.2354 0.0269 0.4852 11.0093
3.42 0.0147 0.2995 0.0491 0.5472 6.3007 4.04 0.0062 0.2345 0.0266 0.4843 11.1077
3.43 0.0145 0.2982 0.0486 0.5461 6.3600 4.05 0.0062 0.2336 0.0264 0.4833 11.2069
3.44 0.0143 0.2970 0.0481 0.5450 6.4198 4.06 0.0061 0.2327 0.0261 0.4824 11.3068
3.45 0.0141 0.2958 0.0476 0.5439 6.4801 4.07 0.0060 0.2319 0.0259 0.4815 11.4076
3.46 0.0139 0.2946 0.0471 0.5428 6.5409 4.08 0.0059 0.2310 0.0256 0.4806 11.5091
3.47 0.0137 0.2934 0.0466 0.5417 6.6023 4.09 0.0058 0.2301 0.0254 0.4797 11.6115
3.48 0.0135 0.2922 0.0462 0.5406 6.6642 4.10 0.0058 0.2293 0.0252 0.4788 11.7147
3.49 0.0133 0.2910 0.0457 0.5395 6.7266 4.11 0.0057 0.2284 0.0249 0.4779 11.8186
3.50 0.0131 0.2899 0.0452 0.5384 6.7896 4.12 0.0056 0.2275 0.0247 0.4770 11.9234
3.51 0.0129 0.2887 0.0448 0.5373 6.8532 4.13 0.0055 0.2267 0.0245 0.4761 12.0290
3.52 0.0127 0.2875 0.0443 0.5362 6.9172 4.14 0.0055 0.2258 0.0242 0.4752 12.1354
3.53 0.0126 0.2864 0.0439 0.5351 6.9819 4.15 0.0054 0.2250 0.0240 0.4743 12.2427
3.54 0.0124 0.2852 0.0434 0.5340 7.0471 4.16 0.0053 0.2242 0.0238 0.4735 12.3508
3.55 0.0122 0.2841 0.0430 0.5330 7.1128 4.17 0.0053 0.2233 0.0236 0.4726 12.4597
3.56 0.0120 0.2829 0.0426 0.5319 7.1791 4.18 0.0052 0.2225 0.0234 0.4717 12.5695
3.57 0.0119 0.2818 0.0421 0.5308 7.2460 4.19 0.0051 0.2217 0.0231 0.4708 12.6801
3.58 0.0117 0.2806 0.0417 0.5298 7.3135 4.20 0.0051 0.2208 0.0229 0.4699 12.7916
3.59 0.0115 0.2795 0.0413 0.5287 7.3815 4.21 0.0050 0.2200 0.0227 0.4691 12.9040
3.60 0.0114 0.2784 0.0409 0.5276 7.4501 4.22 0.0049 0.2192 0.0225 0.4682 13.0172
3.61 0.0112 0.2773 0.0405 0.5266 7.5193 4.23 0.0049 0.2184 0.0223 0.4673 13.1313
3.62 0.0111 0.2762 0.0401 0.5255 7.5891 4.24 0.0048 0.2176 0.0221 0.4665 13.2463
3.63 0.0109 0.2751 0.0397 0.5245 7.6595 4.25 0.0047 0.2168 0.0219 0.4656 13.3622
3.64 0.0108 0.2740 0.0393 0.5234 7.7305 4.26 0.0047 0.2160 0.0217 0.4648 13.4789
3.65 0.0106 0.2729 0.0389 0.5224 7.8020 4.27 0.0046 0.2152 0.0215 0.4639 13.5965
3.66 0.0105 0.2718 0.0385 0.5213 7.8742 4.28 0.0046 0.2144 0.0213 0.4631 13.7151
3.67 0.0103 0.2707 0.0381 0.5203 7.9470 4.29 0.0045 0.2136 0.0211 0.4622 13.8345
3.68 0.0102 0.2697 0.0378 0.5193 8.0204 4.30 0.0044 0.2129 0.0209 0.4614 13.9549
3.69 0.0100 0.2686 0.0374 0.5183 8.0944 4.31 0.0044 0.2121 0.0207 0.4605 14.0762
3.70 0.0099 0.2675 0.0370 0.5172 8.1691 4.32 0.0043 0.2113 0.0205 0.4597 14.1984
3.71 0.0098 0.2665 0.0367 0.5162 8.2443 4.33 0.0043 0.2105 0.0203 0.4588 14.3215
3.72 0.0096 0.2654 0.0363 0.5152 8.3202 4.34 0.0042 0.2098 0.0202 0.4580 14.4456
3.73 0.0095 0.2644 0.0359 0.5142 8.3968 4.35 0.0042 0.2090 0.0200 0.4572 14.5706
3.74 0.0094 0.2633 0.0356 0.5132 8.4739 4.36 0.0041 0.2083 0.0198 0.4563 14.6965
3.75 0.0092 0.2623 0.0352 0.5121 8.5517 4.37 0.0041 0.2075 0.0196 0.4555 14.8234
3.76 0.0091 0.2613 0.0349 0.5111 8.6302 4.38 0.0040 0.2067 0.0194 0.4547 14.9513
3.77 0.0090 0.2602 0.0345 0.5101 8.7093 4.39 0.0040 0.2060 0.0193 0.4539 15.0801
3.78 0.0089 0.2592 0.0342 0.5091 8.7891 4.40 0.0039 0.2053 0.0191 0.4531 15.2099
3.79 0.0087 0.2582 0.0339 0.5081 8.8695 4.41 0.0039 0.2045 0.0189 0.4522 15.3406
3.80 0.0086 0.2572 0.0335 0.5072 8.9506 4.42 0.0038 0.2038 0.0187 0.4514 15.4724
3.81 0.0085 0.2562 0.0332 0.5062 9.0323 4.43 0.0038 0.2030 0.0186 0.4506 15.6051
3.82 0.0084 0.2552 0.0329 0.5052 9.1148 4.44 0.0037 0.2023 0.0184 0.4498 15.7388
3.83 0.0083 0.2542 0.0326 0.5042 9.1979 4.45 0.0037 0.2016 0.0182 0.4490 15.8735
3.84 0.0082 0.2532 0.0323 0.5032 9.2817 4.46 0.0036 0.2009 0.0181 0.4482 16.0092
3.85 0.0081 0.2522 0.0320 0.5022 9.3661 4.47 0.0036 0.2002 0.0179 0.4474 16.1459
3.86 0.0080 0.2513 0.0316 0.5013 9.4513 4.48 0.0035 0.1994 0.0178 0.4466 16.2837
3.87 0.0078 0.2503 0.0313 0.5003 9.5372 4.49 0.0035 0.1987 0.0176 0.4458 16.4224
3.88 0.0077 0.2493 0.0310 0.4993 9.6237 4.50 0.0035 0.1980 0.0174 0.4450 16.5622
3.89 0.0076 0.2484 0.0307 0.4984 9.7110 4.51 0.0034 0.1973 0.0173 0.4442 16.7030
3.90 0.0075 0.2474 0.0304 0.4974 9.7990
Appendix D: Isentropic Table (γ = 1.4) 435
M p0 T0 ρ0 a0 A∗ M p0 T0 ρ0 a0 A∗
4.52 0.0034 0.1966 0.0171 0.4434 16.8449 5.14 0.0016 0.1591 0.0101 0.3989 27.9373
4.53 0.0033 0.1959 0.0170 0.4426 16.9878 5.15 0.0016 0.1586 0.0100 0.3983 28.1579
4.54 0.0033 0.1952 0.0168 0.4418 17.1317 5.16 0.0016 0.1581 0.0099 0.3976 28.3800
4.55 0.0032 0.1945 0.0167 0.4411 17.2767 5.17 0.0016 0.1576 0.0099 0.3970 28.6036
4.56 0.0032 0.1938 0.0165 0.4403 17.4228 5.18 0.0015 0.1571 0.0098 0.3963 28.8287
4.57 0.0032 0.1932 0.0164 0.4395 17.5699 5.19 0.0015 0.1566 0.0097 0.3957 29.0552
4.58 0.0031 0.1925 0.0163 0.4387 17.7181 5.20 0.0015 0.1561 0.0096 0.3950 29.2833
4.59 0.0031 0.1918 0.0161 0.4380 17.8674 5.21 0.0015 0.1555 0.0095 0.3944 29.5129
4.60 0.0031 0.1911 0.0160 0.4372 18.0178 5.22 0.0015 0.1550 0.0095 0.3938 29.7441
4.61 0.0030 0.1905 0.0158 0.4364 18.1693 5.23 0.0015 0.1545 0.0094 0.3931 29.9767
4.62 0.0030 0.1898 0.0157 0.4357 18.3218 5.24 0.0014 0.1540 0.0093 0.3925 30.2109
4.63 0.0029 0.1891 0.0156 0.4349 18.4755 5.25 0.0014 0.1536 0.0092 0.3919 30.4467
4.64 0.0029 0.1885 0.0154 0.4341 18.6303 5.26 0.0014 0.1531 0.0092 0.3912 30.6840
4.65 0.0029 0.1878 0.0153 0.4334 18.7862 5.27 0.0014 0.1526 0.0091 0.3906 30.9229
4.66 0.0028 0.1872 0.0152 0.4326 18.9433 5.28 0.0014 0.1521 0.0090 0.3900 31.1634
4.67 0.0028 0.1865 0.0150 0.4319 19.1015 5.29 0.0014 0.1516 0.0089 0.3893 31.4054
4.68 0.0028 0.1859 0.0149 0.4311 19.2608 5.30 0.0013 0.1511 0.0089 0.3887 31.6491
4.69 0.0027 0.1852 0.0148 0.4304 19.4212 5.31 0.0013 0.1506 0.0088 0.3881 31.8943
4.70 0.0027 0.1846 0.0146 0.4296 19.5828 5.32 0.0013 0.1501 0.0087 0.3875 32.1411
4.71 0.0027 0.1839 0.0145 0.4289 19.7456 5.33 0.0013 0.1497 0.0087 0.3869 32.3896
4.72 0.0026 0.1833 0.0144 0.4281 19.9095 5.34 0.0013 0.1492 0.0086 0.3862 32.6397
4.73 0.0026 0.1827 0.0143 0.4274 20.0746 5.35 0.0013 0.1487 0.0085 0.3856 32.8914
4.74 0.0026 0.1820 0.0141 0.4267 20.2409 5.36 0.0013 0.1482 0.0085 0.3850 33.1448
4.75 0.0025 0.1814 0.0140 0.4259 20.4084 5.37 0.0012 0.1478 0.0084 0.3844 33.3998
4.76 0.0025 0.1808 0.0139 0.4252 20.5770 5.38 0.0012 0.1473 0.0083 0.3838 33.6565
4.77 0.0025 0.1802 0.0138 0.4245 20.7469 5.39 0.0012 0.1468 0.0083 0.3832 33.9148
4.78 0.0025 0.1795 0.0137 0.4237 20.9179 5.40 0.0012 0.1464 0.0082 0.3826 34.1748
4.79 0.0024 0.1789 0.0135 0.4230 21.0902 5.41 0.0012 0.1459 0.0081 0.3820 34.4365
4.80 0.0024 0.1783 0.0134 0.4223 21.2637 5.42 0.0012 0.1454 0.0081 0.3814 34.6999
4.81 0.0024 0.1777 0.0133 0.4216 21.4384 5.43 0.0012 0.1450 0.0080 0.3808 34.9650
4.82 0.0023 0.1771 0.0132 0.4208 21.6144 5.44 0.0011 0.1445 0.0079 0.3802 35.2318
4.83 0.0023 0.1765 0.0131 0.4201 21.7916 5.45 0.0011 0.1441 0.0079 0.3796 35.5003
4.84 0.0023 0.1759 0.0130 0.4194 21.9700 5.46 0.0011 0.1436 0.0078 0.3790 35.7705
4.85 0.0023 0.1753 0.0129 0.4187 22.1497 5.47 0.0011 0.1432 0.0078 0.3784 36.0425
4.86 0.0022 0.1747 0.0128 0.4180 22.3306 5.48 0.0011 0.1427 0.0077 0.3778 36.3162
4.87 0.0022 0.1741 0.0126 0.4173 22.5128 5.49 0.0011 0.1423 0.0076 0.3772 36.5917
4.88 0.0022 0.1735 0.0125 0.4166 22.6963 5.50 0.0011 0.1418 0.0076 0.3766 36.8690
4.89 0.0022 0.1729 0.0124 0.4159 22.8811 5.51 0.0011 0.1414 0.0075 0.3760 37.1480
4.90 0.0021 0.1724 0.0123 0.4152 23.0671 5.52 0.0011 0.1410 0.0075 0.3754 37.4288
4.91 0.0021 0.1718 0.0122 0.4145 23.2545 5.53 0.0010 0.1405 0.0074 0.3749 37.7113
4.92 0.0021 0.1712 0.0121 0.4138 23.4431 5.54 0.0010 0.1401 0.0073 0.3743 37.9957
4.93 0.0021 0.1706 0.0120 0.4131 23.6331 5.55 0.0010 0.1397 0.0073 0.3737 38.2819
4.94 0.0020 0.1700 0.0119 0.4124 23.8243 5.56 0.0010 0.1392 0.0072 0.3731 38.5699
4.95 0.0020 0.1695 0.0118 0.4117 24.0169 5.57 0.0010 0.1388 0.0072 0.3725 38.8597
4.96 0.0020 0.1689 0.0117 0.4110 24.2109 5.58 0.0010 0.1384 0.0071 0.3720 39.1513
4.97 0.0020 0.1683 0.0116 0.4103 24.4061 5.59 0.0010 0.1379 0.0071 0.3714 39.4448
4.98 0.0019 0.1678 0.0115 0.4096 24.6027 5.60 0.0010 0.1375 0.0070 0.3708 39.7402
4.99 0.0019 0.1672 0.0114 0.4089 24.8007 5.61 0.0010 0.1371 0.0070 0.3703 40.0374
5.00 0.0019 0.1667 0.0113 0.4082 25.0000 5.62 0.0009 0.1367 0.0069 0.3697 40.3365
5.01 0.0019 0.1661 0.0112 0.4076 25.2007 5.63 0.0009 0.1363 0.0069 0.3691 40.6374
5.02 0.0018 0.1656 0.0112 0.4069 25.4027 5.64 0.0009 0.1358 0.0068 0.3686 40.9402
5.03 0.0018 0.1650 0.0111 0.4062 25.6062 5.65 0.0009 0.1354 0.0067 0.3680 41.2450
5.04 0.0018 0.1645 0.0110 0.4055 25.8110 5.66 0.0009 0.1350 0.0067 0.3674 41.5516
5.05 0.0018 0.1639 0.0109 0.4049 26.0172 5.67 0.0009 0.1346 0.0066 0.3669 41.8602
5.06 0.0018 0.1634 0.0108 0.4042 26.2249 5.68 0.0009 0.1342 0.0066 0.3663 42.1707
5.07 0.0017 0.1628 0.0107 0.4035 26.4339 5.69 0.0009 0.1338 0.0065 0.3658 42.4831
5.08 0.0017 0.1623 0.0106 0.4029 26.6444 5.70 0.0009 0.1334 0.0065 0.3652 42.7974
5.09 0.0017 0.1618 0.0105 0.4022 26.8563 5.71 0.0009 0.1330 0.0064 0.3646 43.1137
5.10 0.0017 0.1612 0.0104 0.4015 27.0696 5.72 0.0008 0.1326 0.0064 0.3641 43.4320
5.11 0.0017 0.1607 0.0104 0.4009 27.2843 5.73 0.0008 0.1322 0.0063 0.3635 43.7523
5.12 0.0016 0.1602 0.0103 0.4002 27.5005 5.74 0.0008 0.1318 0.0063 0.3630 44.0745
5.13 0.0016 0.1597 0.0102 0.3996 27.7182
M p0 T0 ρ0 a0 A∗ M p0 T0 ρ0 a0 A∗
5.75 0.0008 0.1314 0.0063 0.3624 44.3987 5.87 0.0007 0.1267 0.0057 0.3560 48.4481
5.76 0.0008 0.1310 0.0062 0.3619 44.7249 5.88 0.0007 0.1263 0.0057 0.3554 48.7991
5.77 0.0008 0.1306 0.0062 0.3613 45.0532 5.89 0.0007 0.1260 0.0056 0.3549 49.1522
5.78 0.0008 0.1302 0.0061 0.3608 45.3834 5.90 0.0007 0.1256 0.0056 0.3544 49.5075
5.79 0.0008 0.1298 0.0061 0.3603 45.7157 5.91 0.0007 0.1252 0.0055 0.3539 49.8649
5.80 0.0008 0.1294 0.0060 0.3597 46.0500 5.92 0.0007 0.1249 0.0055 0.3533 50.2244
5.81 0.0008 0.1290 0.0060 0.3592 46.3864 5.93 0.0007 0.1245 0.0055 0.3528 50.5861
5.82 0.0008 0.1286 0.0059 0.3586 46.7248 5.94 0.0007 0.1241 0.0054 0.3523 50.9501
5.83 0.0008 0.1282 0.0059 0.3581 47.0653 5.95 0.0007 0.1238 0.0054 0.3518 51.3161
5.84 0.0007 0.1279 0.0058 0.3576 47.4079 5.96 0.0007 0.1234 0.0053 0.3513 51.6844
5.85 0.0007 0.1275 0.0058 0.3570 47.7525 5.97 0.0007 0.1230 0.0053 0.3508 52.0549
5.86 0.0007 0.1271 0.0058 0.3565 48.0993 5.98 0.0006 0.1227 0.0053 0.3502 52.4276
5.99 0.0006 0.1223 0.0052 0.3497 52.8026
Multiple Choice Questions in Aerospace Engineering
E
1. High wing position gives (a) remain constant.

(b) increase isentropically to the static pressure at the noz-
(a) the same lateral stability as a low wing. zle exit.
(b) more lateral stability than a low wing. (c) decrease isentropically to the static pressure at the noz-
(c) less lateral stability than a low wing. zle exit.
(d) no stability change. (d) increase or decrease, depending upon the static pres-
sure at the nozzle exit.
2. The Mach number range of hypersonic flow (until the con-
tinuum hypothesis holds for air), is 6. For a flow across Normal shock wave,
(a) 5 ≤ M ≤ 25
(a) stagnation pressure remains constant.
(b) 5 ≤ M ≤ 40
(b) Mach number increases.
(c) 5 ≤ M ≤ 55
(c) stagnation temperature remains constant.
(d) 5 ≤ M ≤ 75
(d) both static pressure and density decreases.
3. What is the purpose of Elevons in Light Combat Aircraft?
7. The absolute ceiling is determined by
(a) rolling
(b) pitching (a) aircraft weight, altitude, and speed.
(c) both a & b (b) aircraft weight, speed, and cg position.
(d) none of the above (c) aircraft speed, altitude, and cg position.
(d) aircraft weight and cg position.
4. Which of the following statement is TRUE with respect to
International Standard Atmosphere? 8. The Stagnation or total pressure at a point is defined as the
pressure when the flow is brought to rest
(a) Cruising phase of civil aviation flights takes place in
TROPOSPHERE. (a) isothermally
(b) Cruising phase of civil aviation flights takes place in (b) isobarically
TROPOPAUSE. (c) isentropically
(c) Cruising phase of civil aviation flights takes place in (d) adiabatically
STRATOSPHERE.
(d) Cruising phase of civil aviation flights takes place in 9. How does CL vary with altitude?
MESOSPHERE.
(a) It increases with increasing altitude.
5. Considering a steady and inviscid flow in a convergent– (b) It decreases with increasing altitude.
divergent nozzle, with a normal shock placed in the diver- (c) It decreases with decreasing altitude.
gent portion. The static pressure along the nozzle, down- (d) It remains constant with increasing altitude.
stream of the normal shock will

438 Appendix E: Multiple Choice Questions in Aerospace Engineering
10. For a flow across Prandtl–Meyer expansion wave, (b) 0.04

(c) 0.24
(a) Density remains constant. (d) 0.12
(b) Temperature remains constant.
(c) Entropy remains constant. 17. For any two-dimensional, irrotational, and incompress-
(d) Mach number remains constant. ible flow
11. Among the choices given below, the specific impulse is (a) Both velocity potential and stream function satisfies
maximum for a Laplace’s equation.
(b) Velocity potential satisfies but stream function does
(a) solid rocket not satisfy Laplace’s equation.
(b) liquid rocket (c) Stream function satisfies but velocity potential does
(c) ramjet not satisfy Laplace’s equation.
(d) cryogenic rocket (d) Both velocity potential and stream function do not
satisfy Laplace’s equation.
12. For a flow across an oblique shock wave, which of the
statement is TRUE? 18. The induced drag coefficient for a finite wing with gen-
eral lift distribution is
(a) Component of velocity normal to shock wave
decreases while tangential component increases. (a) directly proportional to aspect ratio
(b) Component of velocity normal to shock wave (b) inversely proportional to aspect ratio
increases while tangential component decreases. (c) directly proportional to (aspect ratio)1/2
(c) Component of velocity normal to shock wave (d) inversely proportional to (aspect ratio)1/2
decreases while tangential component remains
unchanged. 19. The pressure distribution on the surface of the sphere is
(d) Component of velocity normal to shock wave is given by
unchanged while tangential component decreases.
(a) Cp = 1 − 45 sin2 θ
13. A flow has a velocity field given by v = 2xî−2yĵ . The (b) Cp = 1 − 49 sin2 θ
potential function, φ(x, y) for the flow is (c) Cp = 1 − 4sin2 θ
(d) Cp = 2sin2 θ
(a) 2x–2y + constant
(b) 2xy + constant 20. Combustion efficiency of a rocket engine is generally in
(c) x2 + y2 + constant the range of
(d) x2 − y2 + constant
(a) 80–89%
14. Thin airfoil theory predicts that lift slope is dCl
= 2π for (b) 70–79%
dα
(c) 40–49%
(a) any airfoil shape. (d) 90–99%
(b) symmetric airfoils only
(c) cambered airfoils only 21. In hot-wire anemometry, the hot-wire sensor is generally
(d) Joukowski airfoils only made of
15. The conventional altimeter is the instrument used for (a) stainless steel
(b) nickel
(a) pressure measurement (c) tungsten
(b) temperature measurement (d) copper
(c) density measurement
(d) velocity measurement 22. The Shadowgraph is an optical diagnostic technique
which is sensitive to changes of the
16. The maximum thickness-to-chord ratio for NACA 2412
airfoil is (a) fluid density.
(b) first derivative of the fluid density.
(a) 0.02 (c) second derivative of the fluid density.
(d) third derivative of the fluid density.
Appendix E: Multiple Choice Questions in Aerospace Engineering 439
23. “Washout” is (a) increase with increasing altitude.

(b) are independent of outside air temperature.
(a) a decrease in chord from root to tip. (c) increase in proportion to the airspeed.
(b) an increase in incidence from root to tip. (d) decrease in proportion to the ambient pressure at
(c) a decrease in incidence from root to tip. constant temperature.
(d) a decrease and then increase in angle of chord from
root to tip. 31. The effect of a tailwind on the time to climb to a given
altitude is
24. If KE is the mean molecular kinetic energy, T is the
temperature in Kelvin, and κ is the Boltzmann constant (a) none
then (b) increases
(c) decreases
(a) KE = 52κT
(d) depends on the airplane type.
(b) KE = 32κT
(c) KE = 12κT 32. Apart from lift, the forces that determine the angle of
(d) KE = 72κT climb are
25. The Air deviates from the perfect gas behavior when (a) weight and drag
(b) thrust and drag
(a) pressure decreases and temperature increases.
(c) weight and thrust
(b) pressure increases and temperature decreases.
(d) weight, drag, and thrust
(c) when both pressure and temperature decreases.
(d) when both pressure and temperature increases. 33. Which of the following statements are correct?
26. With increasing the aircraft weight, it will (a) Induced drag is independent of speed.
(b) Induced drag decreases with increasing angle of
(a) increase the stalling speed.
attack.
(b) not affect the stalling speed.
(c) Induced drag decreases with increasing speed.
(c) decrease the stalling speed.
(d) Induced drag increases with increasing speed.
(d) stalling speed becomes zero.
34. The Bernoulli equation for compressible flow is
27. In level flight Maximum horizontal speed occurs when
(a) thrust = minimum drag. (a) p + 21 ρv2 = p0

γ p γ p0
ρ + 2 ρv =
1 2
(b) thrust does not increase with increasing speed. (b) γ+1 γ+1 ρ0
γ p γ p0
γ−1 ρ + 2 ρv =
1 2
(c) maximum thrust = total drag. (c) γ−1 ρ0
(d) thrust = maximum drag. (d) γ ρp + 21 v2 = γ ρp00
28. In a level flight, at a constant mass and Mach number, a 35. The speed to attain the minimum power required for a
higher altitude requires a turbojet aircraft is
(a) higher angle of attack (a) less than the speed for minimum drag.
(b) lower CL (b) higher than the speed for minimum drag.
(c) lower CD (c) slower in a climb and faster in a descent.
(d) lower angle of attack (d) the same as minimum drag speed.
29. The maximum operating altitude for an aircraft with a 36. Which of the following is CORRECT?
pressurized cabin
(a) With increase of temperature, the viscosity of air
(a) is dependent on the aerodynamic ceiling. decreases and viscosity of water increases.
(b) is dependent on the outside air temperature. (b) With increase of temperature, the viscosity of air
(c) is only certified for four engine aircraft. increases and viscosity of water decreases.
(d) is the highest pressure altitude certified for normal (c) With increase of temperature, the viscosity of BOTH
operation. air and water increases.
(d) With increase of temperature, the viscosity of BOTH
30. At a constant Mach number, the thrust and fuel flow of
air and water decreases.
a jet engine
37. The thrust of a jet engine at constant RPM (a) diesel

(b) petrol
(a) is inversely proportional to the airspeed. (c) liquid hydrogen
(b) increases in proportion to the airspeed. (d) kerosene
(c) does not change with changing altitude.
(d) is independent of airspeed. 45. In a straight and level flight of an aircraft
38. If the thrust available exceeds the thrust required in level (a) Lift equals aircraft weight.
flight, the aircraft (b) Lift is more than weight.
(c) Lift is less than weight.
(a) will accelerate. (d) Lift depends upon the size of aircraft and its
(b) will descend if the airspeed remains constant. loading.
(c) decelerates if it is in the region of reversed command.
(d) will decelerate. 46. In propulsion engines, generally the combustion takes
place at
39. Which force compensates the mass in unaccelerated
straight and level flight (a) constant volume
(b) constant pressure
(a) lift (c) constant entropy
(b) thrust (d) constant temperature
(c) drag
(d) resultant from lift and drag 47. For a flow across Normal shock wave, when M1 → ∞
then which of the following is CORRECT?
40. Increased mass will cause the climb performance to
ρ2
(a) limM1 →∞ ρ1 =2
(a) improve. ρ2
(b) limM1 →∞ ρ1 =4
(b) be unchanged. ρ2
(c) limM1 →∞ ρ1 =6
(c) be unchanged if the short-field technique is used. ρ2
(d) degrade. (d) limM1 →∞ ρ1 =∞
41. Modern transport aircraft use which of the following 48. The specific impulse of a solid rocket motor is of the
engine for propulsion? order of
(a) turbojet (a) 650 seconds

(b) turbofan (b) 250 seconds
(c) rocket (c) 550 seconds
(d) turboshaft (d) 750 seconds

42. In a glide the lift force is 49. A flow field has the stream
3 function,
ψ = x2 −y2 and
velocity potential, φ = x + y3 . Which of the follow-
(a) greater than the weight . ing statements are correct?
(b) to the weight.
(c) less than the weight. (a) Flow is incompressible and irrotational.
(d) equal to drag. (b) Flow is incompressible and rotational.
(c) Flow is compressible and irrotational.
43. When a Rocket climbs up, the thrust (d) Flow is compressible and rotational.
(a) increases. 50. In straight and level flight the following forces act on the
(b) decreases. a/c
(c) remains constant.
(d) decreases and then increases. (a) thrust, lift, and drag only.
(b) lift and drag only.
44. Which of the following is a potential fuel in a supersonic
(c) thrust, lift, and weight only.
combustion ramjet (SCRAMJET), which can also serve
(d) all thrust, lift, drag, and weight.
as a coolant
51. A Ramjet engine does not have the following 58. In an irrotational, incompressible and two-dimensional
flow, the velocity potential (φ), satisfies which of the
(a) nozzle following relation (u and v are velocity components in
(b) combustion chamber Cartesian coordinates)?
(c) compressor
∂φ
(d) flame holder (a) u = ∂x and v = − ∂φ
∂y
52. In a climb the lift force is (b) u = − ∂φ

∂x and v= ∂φ
∂y
∂φ ∂φ
(c) u = ∂x and v = − ∂y
(a) greater than the weight.
(b) equal to the weight. (d) u = − ∂φ
∂x and v = − ∂φ
∂y
(c) than the weight.
(d) greater than trust. 59. If an aircraft maintains a constant radius of turn but the
speed is increased then
53. To maintain steady level flight, as the angle of attack is
increased, the airspeed must (a) The bank angle must be increased.
(b) The bank angle must be decreased.
(a) be increased. (c) The bank angle will remain constant.
(b) be decreased. (d) The bank angle is equal to zero.
(c) remain constant.
(d) no affect by airspeed. 60. Maximum range for a jet occurs when the airplane is
flying at
54. For International Standard Atmosphere, the relation 1/2
between geopotential and geometric altitude is CL
(a)
CD
(a) h = r+h G
hG 3/2 max
r CL
(b) CD
(b) h = h+h r
hG max
G
(c) h = r−h G
hG (c) CCDL
r 5/2 max
CL
(d) h = r+hr
G
hG (d) CD
max
55. Power available at absolute ceiling is
61. For maximum aerodynamic efficiency the wing plan
(a) more than the power required. shape would be
(b) less than the power required.
(c) zero. (a) elliptical
(d) equal to the power required. (b) triangular
(c) rectangular
56. Which statement is true as far as a rocket vehicle is con- (d) circular
cerned
62. Stall speed is affected by
(a) The vehicle velocity can exceed the exhaust velocity.
(b) The vehicle velocity cannot exceed the exhaust (a) weight, load factor, and power.
velocity. (b) load factor, angle of attack, and power.
(c) The vehicle velocity is always equal to the exhaust (c) angle of attack, weight, and air density.
velocity. (d) only load factor and power.
(d) The vehicle velocity can exceed exhaust velocity in
vacuum only. 63. Which of the following is true due the expansion of the
gas through a nozzle
57. Lift on a wing
(a) Mach number increases.
(a) acts through center of gravity.
(b) Pressure decreases.
(b) acts vertically upward.
(c) Temperature decreases.
(c) is perpendiculars to chord line.
(d) All of the above.
(d) acts vertically upward through center of pressure.
64. To improve the frequency of a structure, you will (c) rudder

(d) vertical tail
(a) increase stiffness.
(b) decrease mass. 72. For the flow past a “Golf Ball” kept in a uniform flow,
(c) both (a) & (b). which of the following statement is correct?
(d) increase mass.
(a) Skin friction drag is greater when boundary layer is
65. Spiral divergence is a form of laminar.
(b) Skin friction drag is greater when boundary layer is
(a) longitudinal dynamic instability. turbulent.
(b) lateral dynamic instability. (c) Form drag is greater when boundary layer is turbu-
(c) lateral static stability. lent.
(d) all are correct. (d) Both Skin friction drag and form drag are higher
when boundary layer is laminar.
66. The function of the horizontal stabilizer is to assist
73. If the angle of attack of an airfoil is increased, then the
(a) lateral stability center of pressure will
(b) directional stability
(c) longitudinal stability (a) moved forward
(d) none (b) moved backward
(c) stay the same
67. Anhedral is defined as (d) none
(a) the upward inclination of the aircraft wings in the 74. Dutch roll is
tip.
(b) neutrally stable. (a) a type of slow roll.
(c) the downward inclination of the aircraft wings in the (b) primarily a pitching instability.
tip. (c) a combined rolling and yawing motion.
(d) none. (d) a combined pitching and yawing motion.
68. Sweep-back wing provides 75. The viscosity of air at very high temperature is
(a) directly proportional to T1/2

(a) increased lateral stability
(b) inversely proportional to T1/2
(b) reduced drag at all speed
(c) directly proportional to T3/2
(c) less wing weight
(d) inversely proportional to T3/2
(d) all are correct
76. When the horizontal control surface is deflected, then
69. In aircraft design, it is usual to position the center of lift
the slope of Cm versus α curve is
relative to the center of gravity in the following location
(a) proportional to deflection
(a) forward of CG (b) inversely proportional to deflection
(b) aft of CG (c) depending on position
(c) to the side of CG (d) the same throughout
(d) below CG
77. Aircraft load factor is found from the relationship
70. Control of yaw is mainly influenced by
(a) lift/drag
(a) the vertical fin (b) lift/weight
(b) the tail-plane (c) weight/drag
(c) the ailerons (d) drag/lift
(d) the rudder
78. Which of the following statements are FALSE?
71. Differential control of an aircraft is associated with
(a) Local jump in radial velocity across the vortex sheet
(a) aileron is equal to the local sheet strength.
(b) trim tab
(b) Circulation around an airfoil takes a value so that the 85. Natural frequency of a structure depends upon
flow leaves the trailing edge smoothly.
(c) Time rate of change of circulation in a flow domain (a) mass and stiffness
is zero. (b) mass and damping
(d) Quarter-chord point is both the center of pressure (c) stiffness and damping
and aerodynamic center of the symmetric airfoil. (d) none of the above
79. In the airfoil, NACA 23015, what do the second and third 86. The Young’s Modulus of most of the metallic materials
digits signify? used in aerospace applications varies with temperature.
With increase in temperature it
(a) Multiply by 23 to get design lift coefficient in hun-
dredths. (a) increases.
(b) Maximum thickness in hundredths of chord. (b) decreases.
(c) Divide by 2 to get location of maximum camber in (c) first increases up to critical temperature and then
tenths of chord. decreases.
(d) Location of minimum pressure in hundredths of (d) remains constant.
chord.
87. During supersonic isentropic expansion, total tempera-
80. Consider a non-lifting flow over a circular cylinder. At ture along the length of the nozzle
which of the following angular positions will have the (a) increases.
value of pressure coefficient, Cp to be zero? (b) decreases.
(a) 90o (c) first decreases, then increases.
(b) 30o (d) remains the same.
(c) 0o
88. For stability point recommended position of CG of a
(d) 45o
control surface
81. Which of the following statements are FALSE? (a) ahead or on the hinge line
(a) For an incompressible flow, ∇2φ = 0. (b) after the hinge line
(b) For an incompressible flow, ∇ 2 ψ = 0. (c) anywhere
(c) Principle of superposition does not hold good for (d) none
incompressible, irrotational flows.
89. Solid rocket motors nozzles have a thermal protection
(d) Laplace equation is a linear PDE.
system based on
82. Name the angle between the reference line of an aircraft (a) regenerative cooling
and the projection of the wind vector on the yaw plane. (b) film cooling
(a) angle of attack (c) ablative cooling
(b) angle of side-slip (d) radiation cooling
(c) angle of incidence
90. If a disturbing force is removed from a body and the body
(d) flight path angle
immediately tends to return toward the equilibrium, then
83. PHUGOID motion is a form of it is
(a)longitudinal dynamic stability (a) statically stable

(b)directional dynamic stability (b) dynamically stable
(c)lateral dynamic stability (c) dynamically unstable
(d)all are correct (d) none of the above

91. What are the four typical loads are on the aircraft?
84. For a given area ratio A e
At , thrust can be maximized by
(a) tension, torsion, creep, elongation.
(a) maximizing combustion chamber pressure.
(b) elasticity, shear, compression, torsion.
(b) ensuring optimum expansion at nozzle exit.
(c) tension, compression, torsion, shear.
(c) increasing altitude.
(d) compression, buckling, elasticity, shear.
(d) all of the above.
92. In vacuum, the maximum exit velocity (ve )max depends (a) increases.
on which of the following? (b) decreases.
(c) first increases then decreases.
(a) area ratio AAt
e
(d) none of the above.

(b) pressure ratio pp0e 99. Which of the following fuel/oxidizer combinations is

(c) mass flow rate m̊ hypergolic?
(d) combustion chamber temperature (Tc )
(a) hydrazine (N2 H4 ) and di–nitrogen tetraoxide
93. A convergent–divergent nozzle is said to be choked when (N2 O4 )
(b) RP–1 and LOX
(a) Pressure at the exit is equal to ambient pressure (c) hydrazine (N2 H4 ) and LOX
(pe = pa ). (d) none of the above
(b) Area Ratio is maximized for optimum expansion.
(c) Pressure at some point in the nozzle equals Critical 100. Consider the following statements:
Pressure (p = p∗ ). I. Frozen flow in nozzle is isentropic.
(d) Nozzle operates shock-free. II. Equilibrium flow in nozzle is isentropic.
94. The famous TACHOMA suspension bridge failed due to (a) Only I is TRUE.
(b) Only II is TRUE.
(a) divergence (c) Both I and II are TRUE.
(b) flutter (d) Both I and II are FALSE.
(c) vibration
(d) fatigue 101. For overexpanded nozzle, pressure equalization takes
place through
95. Total temperature across a Normal shock wave
(a) normal shock wave
(a) increases. (b) subsonic diffusion
(b) decreases. (c) both (a) and (b)
(c) remains the same. (d) when exit pressure is equal to the ambient pressure
(d) first decreases, then increases. (pe = pa )
96. Mass flow rate is dependent on which of the following? 102. Which of the following factor affects Creep?

(a) pressure ratio pp0e (a) the duration of the load applied
(b) combustion chamber temperature (Tc ) (b) buckling of the material
(c) molecular weight of combustion gases (c) shear
(d) all of the above (d) strength of the material
97. Consider the following statements: 103. The construction of an aircraft wing is centered on a
I. Aerodynamic heating can take place in space. main member. The member is
II. Design of airframes at room temperature is adequate
for development of missiles. (a) skin
III. Infrared heating using IR lamps are commonly used (b) rib
for thermo-structural testing. (c) frame
Which of the above statements is/are true? (d) spar
(a) only I 104. The torsion constant (J) of a thin-walled closed tube of
(b) I and II thickness (t) and mean radius (r) is given by
(c) only III
(a) J = 2πrt 3
(d) all of the above
(b) J = 2πr 3 t
98. As flight Mach number increases, TSFC of Ramjet (c) J = 2πr 2 t 2
Engine (d) J = 2πr 4
105. The number of independent elastic constants in the con- I. At this Mach number, the aerodynamics drag on an
stitutive equations of a generalized anisotropic solid is airfoil begins to increase rapidly.
(a) 2 II. At this Mach number, the aerodynamics drag on an

(b) 9 airfoil begins to decrease rapidly.
(c) 21
(d) 36 III. This is always greater than critical Mach number.
106. In an impulse turbine, the degree of reaction is IV. The large increase in drag is caused by the formation
of a shock wave on the upper surface of the airfoil.
(a) 1
(b) 0.75
Which of the above is/are CORRECT?
(c) 0
(d) 0.5 (a) I and IV
(b) II and IV
107. A column of length l and flexural rigidity EI has both
(c) II, III, and IV
ends fixed. The critical buckling load for the column is
(d) I, III, and IV
π 2 EI
(a) 111. Consider the following statements:
(0.7l)2
π 2 EI
(b)
(0.5l)2
π 2 EI I. The velocity of a satellite in a circular orbit depends
(c)
(l)2 on altitude of the satellite from Earth’s surface.
π 2 EI
(d)
(2l)2
II. In an elliptical orbit, the velocity of a satellite at
108. Consider the following statements: apogee is lesser than that at perigee.
I. Airy stress function can be used only for two- III. Eccentricity of an elliptical orbit determines the
dimensional problems. shape of the ellipse.
II. Euler’s equations can be applied for viscous flows IV. Geosynchronous satellite has an orbital time period
as well. of 24 hrs.
III. In Schlieren flow visualization, the optical refrac- Choose correct answer from the following:
tive index is sensitive to the second-order derivative of
the flow density. (a) I and II
(b) I, II, and III
Which of the above is/are CORRECT? (c) I, III, and IV
(d) all
(a) I only
(b) I and II only 112. Steady flow occurs when
(c) all
(d) none (a) Pressure does not change along the flow.
(b) Velocity does not change.
109. Which is the point on the airfoil about which pitching (c) Conditions change gradually with time.
moment coefficient is not a function of angle of attack? (d) Conditions do not change with time at any point.
(a) aerodynamic center 113. The fluid forces considered in the Navier Stokes equa-
(b) center of pressure tion are
(c) center of gravity
(d) leading edge of airfoil (a) gravity, pressure, and viscous.
(b) gravity, pressure, and turbulent.
110. Consider the following statements about Drag diver- (c) pressure, viscous, and turbulent.
gence Mach number: (d) gravity, viscous, and turbulent.
114. Which one of the following statements is FALSE for a (a) 5 × 105
supersonic flow? (b) 2300
(c) 1.66 × 105
(a) Over a gradual expansion, entropy remains constant. (d) 200
(b) Over a sharp expansion corner, entropy can increase.
(c) Over a gradual compression, entropy may remain 121. In a low-speed wind tunnel, the honeycomb and wire-
constant. gauze structures are used because
(d) Over a sharp compression corner, entropy increases.
(a) Honeycomb reduces more lateral turbulence, while
115. The existence of a boundary layer near a solid surface wire-gauze reduces more axial turbulence.
was first proposed by (b) Honeycomb reduces more axial turbulence, while
wire-gauze reduces more lateral turbulence.
(a) Ludwig Prandtl
(c) Both honeycomb and wire-gauze reduces mostly
(b) Theodore von Karman
axial turbulence.
(c) Albert Einstein
(d) Both honeycomb and wire-gauze reduces mostly
(d) Isaac Newton
lateral turbulence.
116. According to the boundary layer theory, the flow field
122. The continuity equation
outside a boundary layer is
(a) viscous (a) Expresses relationship between hydraulic parame-

(b) inviscid ters of flow.
(c) can be inviscid or viscous (b) Expresses the relationship, between work and
(d) none of the above energy.
(c) Is based on Bernoulli’s theorem.
117. For a flow around a stagnation point, the boundary layer (d) Relates the mass rate of flow along a streamline.
thickness perpendicular to the flow direction
123. Which of the following is true?
(a) grows.
(b) decreases. (a) Induced drag ∝ 1
(EAS)2
(c) remains constant. (b) Induced drag ∝ (EAS)2
(d) first increases and then decreases. (c) Induced drag ∝ √ 1
√EAS
(d) Induced drag ∝ EAS
118. If n is the degree of freedom of the gas molecules, then
the ratio of specific heats will be
124. Consider the following statements:
I. A perfect gas must be both thermally and calorically
(a) γ = n+4
n perfect.
(b) γ = n−2
n II. A calorically perfect gas must be thermally perfect.
(c) γ = n+2
n III. A thermal perfect gas must be calorically perfect.
(d) γ = n
n+2 Which of the following is true?
119. Consider the following:
(a) I only
I. concept of pathline (b) I and II
II. concept of streakline (c) I and III
III. concept of streamline
(d) all
Which of the above concepts are REAL in nature?
125. The region of speed instability is
(a) I only
(b) I and II (a) The same as the region of reverse command.
(c) II and III (b) The region in which manual control is not possible.
(d) I and III (c) At speeds below the low buffet speed.
(d) The region above the thrust available and drag curve
120. In a cross-flow situation for a pipe flow, based on
intersection.
cylinder diameter, what is the typical value of critical
Reynolds number?
p02
126. Choose the CORRECT relation between flow Mach (b) p01
number (M) and characteristics Mach number (M∗ ), (c) p2 −p1
p1
defined as ratio of local flow speed and critical speed of p2 −p1
(d) p2
sound:
133. Which of the following does not fall under solid rocket
(a) If M → ∞ then M∗ → ∞
propellant group?
(b) If M → ∞ then M∗ → 0
(c) If M → ∞ then M∗ −→ γ+1 (a) double base
γ−1
(b) cast modified double base
(d) If M → ∞ then M −→ γ−1
∗
γ+1 (c) bipropellant
127. At a constant mass and altitude a decreased airspeed (d) composite propellant
requires:
134. A constant headwind component:
(a) a higher CL .
(a) increases the angle of climb path
(b) less thrust and a lower CL .
(b) increases the maximum rate of climb
(c) more thrust and a lower CL .
(c) decreases the angle of climb path
(d) more thrust and a lower CD .
(d) increases the maximum endurance
128. The temperature of a gas is produced due to
135. If an aircraft is put into a 2g turn from level flight, main-
(a) kinetic energy of molecules taining a constant speed:
(b) the heating value of gas
(a) CL will be increased by four times and CD will be
(c) cohesive and adhesive forces among the gas
doubled.
molecules
(b) CL will be doubled CD will be increased by four
(d) surface tension of molecules
times.
129. The most efficient methods of compressing the gas (c) CL will be doubled and C√D will be doubled.
and maximum work done on the gas are __________, (d) CL will be increased by 2 times, and CD will be
respectively. increased by four times.
(a) adiabatic and isothermal 136. Which of the following will increase the takeoff dis-
(b) isentropic and polytropic tance?
(c) isothermal and adiabatic
(a) Downhill slope because of decreased angle of
(d) isothermal and polytropic
attack.
130. Which of the following compressors are used in air- (b) Slush.
craft? (c) Headwind due to increased drag.
(d) Decreased takeoff mass.
(a) radial flow compressors
(b) centrifugal 137. For a light engine airplane the carriage of an additional
(c) axial flow passenger will cause the climb performance to be:
(d) combination of above
(a) degraded
131. The buckling load for a given material depends on (b) improved
(c) unchanged
(a) Slenderness ratio and area of cross section. (d) unchanged if a short-field takeoff technique is used
(b) Poisson’s ratio and modulus of elasticity.
(c) Slenderness ratio and modulus of elasticity. 138. In a Hypersonic flow (M > 5) past a wedge (ver-
(d) Slenderness ratio, area of cross section, and modu- tex angle, 2θ), the hypersonic similarity parameter is
lus of elasticity. defined as:
132. The strength of a normal shock is defined as (a) M2 θ

2
(b) Mθ
p01 (c) Mθ
(a) p02
(d) M
θ
139. For a perfect gas, the supersonic flow becomes a sub- 144. If Mcr is critical Mach number and, MDD is drag diver-
sonic flow across a normal shock wave. This is, in accor- gence Mach number then which of the following is
dance with which of the following law of thermodynam- TRUE for an airfoil kept in a subsonic flow?
ics?
(a) MDD < Mcr < 1.0
(a) zeroth law (b) Mcr < MDD < 1.0
(b) first law (c) 1.0 < Mcr < MDD
(c) second law (d) MDD < 1.0 < Mcr
(d) third law
145. For a propeller-driven airplane, choose the CORRECT
140. The free molecular flow is the regime in which the from the following:
fluid molecules are so widely dispersed that the inter-
molecular forces can be neglected. This situation is best (a) At absolute ceiling the maximum rate of climb RC
described by which of the following range of Knudsen of airplane is maximum.
number (Kn)? (b) At service ceiling the maximum rate of climb RC
of airplane is maximum.
(a) Kn < 0.01 (c) At absolute ceiling the maximum rate of climb RC
(b) 0.01 < Kn < 0.1 of airplane is zero.
(c) 0.1 < Kn < 5 (d) At service ceiling the maximum rate of climb RC
(d) Kn > 5 of airplane is 0.05 ms−1 .
141. Consider the following statements with respect to vis- 146. A low pressure altitude at an airport causes
cous flow past a full length aircraft and its model:
I. The geometrically similar flows must be dynamically (a) improved takeoff and degraded climb characteris-
similar. tics.
II. The dynamically similar flows must be geometrically (b) degraded takeoff and climb characteristics.
similar. (c) degraded takeoff and improved climb characteris-
Which of the above is true? tics.
(d) improved takeoff and climb characteristics.
(a) I
(b) II 147. If T represents thrust and v represents the velocity, and
(c) I and II
(d) none I. Trequired − Tavailable = 0
∂(Trequired −Tavailable )
142. For the flow over a flat plate, the turbulent boundary II. ∂v =0
layer thickness can be expressed as
Which of the following is true?
δ
(a) x = 5
1/2
Rex
δ (a) I
(b) = 5
1/2
x Rex (b) II
δ
(c) x = 0.16
1/7
(c) I and II
Rex
(d) none
δ
(d) x = 0.16
1/2
Rex
148. If the drag polar of a propeller aircraft is
143. “The circulation of a vortex tube remains constant in CD = CD0 + KC2L , for a given climb angle 0 < γ < π2
time”. Which of the following Helmholtz’s theorem and with shallow climb approximation, the maximum
describes this? climb rate represents

(a) first vortex theorem (a) CL =
2CD0
(b) second vortex theorem K
3CD0
(c) third vortex theorem (b) CL =
K
(d) fourth vortex theorem (c) CL = 3CD0 K
(d) CL = 2CD0 K
2
149. With reference to gliding flight, considering the exact (a) Cp = −2 u
Ua
form of dynamical equations the flattest glide occurs
when (b) Cp = −2 Uua
2
(c) Cp = −2 Uua
E2
(a) u = 4 E 2 max+1
(d) Cp = −2 Uua
max
E 2 +1
(b) u = 4 Emax2
max 155. In the wind tunnel operation, which of the following is
E2 FALSE?
(c) u = 4 E 2 max−1
max
E 2 +1 (a) The normal shock in the test section owes to about
(d) u = 4 E max
2 −1 80% of the total power requirements in supersonic
max
wind tunnels.
150. If ρ is the measured density around the airplane, and (b) Pitot probe does not measure the actual freestream
ρSL is the standard sea level density then which of the stagnation pressure in the test section of supersonic
following is TRUE for airspeeds of an airplane? wind tunnels.
(c) The starting load is less than the Running load in
(a) Vtrue = Vequivalent ρρSL supersonic wind tunnels.

(d) In the subsonic wind tunnel the test-section walls
(b) Vequivalent = Vtrue ρρSL
√ √ should not be made parallel but divergent.
(c) Vequivalent = Vtrue ρ × ρSL
(d) Vtrue = Vequivalent ρρSL 156. A low wing and a high wing configuration aircraft are
identical to each other in all respects except the loca-
151. Generally for closed throat wind tunnels the energy ratio tion of the wing. The longitudinal static stability of the
(ER) will be airplane with high wing configuration will be:
(a) ER < 3 (a) More than the airplane with low wing configuration.
(b) 2 < ER < 5 (b) Less than the airplane with low wing configuration.
(c) 3< ER <7 (c) Same as the airplane with low wing configuration.
(d) always greater than 5 (d) More if the elevator is deflected.
152. In the wind tunnel operation, if Re is the effective 157. Let an airplane in a steady level flight be trimmed at a
Reynolds number and Rec is the measured Reynolds certain speed. A level and steady flight at higher speed
number then Turbulence Factor is defined as could be achieved by changing
Re (a) engine throttle only

(a) Rec
(b) Rec (b) elevator only
Re
(c) × Rec
Re2 (c) throttle and elevator together
(d) Re × Rec2 (d) rudder only
153. In both intermittent and blowdown supersonic wind 158. The effect of propeller on the longitudinal static stability
tunnel operations, the axial flow compressors is used of the aircraft is to
because of
(a) increase longitudinal static stability.
(a) High pressure ratio and large mass flow rate. (b) decrease longitudinal static stability.
(b) High pressure ratio and small mass flow rate. (c) longitudinal static stability remains same.
(c) Low pressure ratio and small mass flow rate. (d) longitudinal static stability is minimum.
(d) It has small power requirements.
159. If the center of gravity of an airplane is moved forward
154. If u is perturbation velocity and Ua is freestream veloc- toward the nose of the airplane, the CL,max (maximum
ity. Using small perturbation theory the pressure coef- value of the lift coefficient) value for which the aircraft
ficient in two-dimensional planar flows will be can be trimmed (Cm = 0) will be
(a) decreased. (b) Isentropic flows only.

(b) increased. (c) Adiabatic flows whether reversible or irreversible.
(c) remains the same. (d) Defined only for Mach numbers greater than one.
(d) depend upon the rudder deflection.
166. For agivengas, the maximum mass flow rate per unit
160. If the contribution of only the horizontal tail of an air- area m̊Amax
is directly proportional to
plane was considered for estimating ∂C m
∂α , and if the tail
moment arm lt was doubled, then how many times the (a) √T0
original value would the new ∂C m P0
∂α , become (b) √P0
√T0
(a) 1.414 times (c) T0
(b) 1.732 times P2
(d) √0
T0
(c) 2 times
(d) 3 times 167. The flow across a non-isentropic compression front can
be made isentropic flow by optimizing the entropy gen-
161. If the vertical tail of an airplane is inverted and put below eration (S) and flow deflection angle (θ) across
the horizontal tail, the contribution to roll derivative, the compression front. The approximate linear relation
∂CL
∂β will be holds for this is
(a) negative (a) S ∝ θ
(b) positive (b) S ∝ (θ)3
(c) zero (c) S ∝ 1 3
(θ)
(d) imaginary
(d) S ∝ 1
(θ)4
162. PHUGOID motion can be excited by giving an initial
disturbance in 168. The Joukowski airfoil is studied in aerodynamics
because
(a) Speed only.
(b) By any one or both of the Elevator deflection and (a) It has a simple geometry.
Speed. (b) It is used in many aircraft during World War II.
(c) By any one or both of the Elevator deflection and (c) It is easily transformed into circle, mathematically.
Angle of attack. (d) It has highest lift curve slope among all airfoils.
(d) By any one or all of the Speed, Elevator deflection,
169. Consider a pin supported rod of length L, with E as
and Angle of attack.
Young’s modulus and I is the area moment of inertia
163. Spiral divergence arises because of then the critical load at which the rod buckles is
π 3 EI
(a) high directional and low roll stability. (a) 2L2
(b) high roll and low directional stability. π 2 EI
(b) L3
(c) high roll and high yaw stability. π 2 EI
(c) L2
(d) low roll and low yaw stability. π 3 EI
(d) L2
164. If the aircraft wing has an Anhedral of then the local
angle of attack (αl ) on the starboard side and the port 170. A supercritical airfoil has
side are given by
(a) higher wave drag.
(a) αl = α ± β (b) higher critical Reynolds number.
(b) αl = α ∓ β (c) higher critical Mach number.
(c) αl = α ± β (d) higher drag divergence Mach number.
(d) αl = α ± β 171. A uniform beam of length L, is simply supported at
its ends and carries a uniformly distributed load of w
165. The ratio of stagnation and static temperature is defined
per unit length between mid-span and a point from the
as TT0 = 1 + (γ−1) 2
2 M this relation holds for right hand end. An expression for the deflection at the
(a) Adiabatic flows only. mid-span is
19wL4
(a) y = 4096EI
(d) It cannot be predicted.
29wL4
(b) y = 384EI 177. In a steady state, the Bernoulli’s equation is applicable
19wL4
(c) y = 384EI in which of the following situation?
wL4
(d) y = 4096EI (a) Between any two points in both inviscid and poten-
tial flows.
(b) Between any two points in inviscid flow and only
I. A geostationary orbit can be geosynchronous, but
along a streamline in potential flow.
not all geosynchronous orbits are geostationary.
(c) Only along a streamline in both inviscid and poten-
II. An orbit can be both in a Sun-synchronous orbit and
tial flows.
in a repeat orbit at the same time.
(d) Only along a streamline in inviscid flow and
between any two points in potential flow.
Choose the CORRECT from the following:
178. For the maximum propulsive efficiency, the ratio of
(a) I
flight speed to the nozzle exhaust velocity is
(b) II
(c) I and II (a) 0.5
(d) none (b) 1.0
(c) 2.0
173. Which of the following criteria leads to maximum turn
(d) 4.0
rate and minimum radius in a level turn flight?
179. The skin and spar-webs are used in semimonocoque
(a) Highest possible load factor and highest possible
wing configurations because they are the primary car-
velocity.
riers of:
(b) Lowest possible load factor and lowest possible
velocity. (a) Shear stresses due to an aerodynamics moment
(c) Highest possible load factor and lowest possible component alone.
velocity. (b) Normal stresses due aerodynamics forces alone.
(d) Lowest possible load factor and highest possible (c) Shear stresses due aerodynamics forces alone.
velocity. (d) Shear stresses due aerodynamics forces and a
moment component.
174. Which of the following input is required to produce
constant roll rate using ailerons in an aircraft? 180. In the closed-circuit supersonic wind tunnel, the
convergent–divergent nozzle and test section are
(a) a step input
attached with convergent–divergent diffuser to swal-
(b) an impulse input
low the stalling normal shock. The condition which
(c) a ramp input
must be met is
(d) a sinusoidal input
(a) The diffuser throat area must be larger than nozzle
175. The Hohmann ellipse used as Earth–Mars transfer orbit
throat area and shock must be standing downstream
has
of diffuser throat.
(a) apogee at Earth and perigee at Mars. (b) The diffuser throat area must be smaller than nozzle
(b) both apogee and perigee at Earth. throat area and shock must be standing downstream
(c) apogee at Mars and Perigee at Earth. of diffuser throat.
(d) both apogee and perigee at Mars. (c) The diffuser throat area must be same as the nozzle
throat area and shock must be standing upstream of
176. An aircraft in a steady climb suddenly experiences a diffuser throat.
10% drop in thrust. After a new equilibrium is reached (d) The diffuser throat area must be smaller than nozzle
at the same speed, the new rate of climb is? throat area and shock must be standing upstream of
diffuser throat.
(a) Lower by exactly by 10%.
(b) Lower by more than 10%. 181. The effect of increasing the internal pressure on the
(c) Lower by less than 10%. buckling of the fuselage skin is
(a) It is delayed. 187. In the combustion chamber an efficient fuel injection

(b) It is advanced. system is required because
(c) It remains the same.
(d) Insufficient data to predict. (a) It increases the fuel flow rate in the combustion
chamber.
182. The process in which there is only work interaction (b) It increases the airflow rate in the combustion cham-
between the system and its surroundings is ber.
(c) It sprays the fuel in the form of small droplets.
(a) iIsothermal process (d) It acts as an igniter in initiating the combustion.
(b) diabatic process
(c) quasi-static process 188. An aircraft is flying at Mach 2, where the ambient tem-
(d) adiabatic process perature is found to be 200K. What is the stagnation
temperature on the surface of the aircraft? (Assume the
183. An oblique shock wave with a wave angle β is generated specific heat ratio of air to be 1.4)
from a wedge angle of θ. The ratio of the Mach number
downstream of the shock to its normal component is (a) 300
(b) 360
(a) sin (β − θ) (c) 375
(b) cos (β − θ) (d) 400
(c) sin (θ − β)
(d) cos (θ − β) 189. In choosing materials for structural members of an air-
craft, there is most often a trade-off between which of
184. According to Euler–Bernoulli beam theory, the nature the following characteristics?
of the governing equation for the static transverse
deflection of a beam under a uniformly distributed load (a) strength and weight
is (b) plasticity and weight
(c) strength and elasticity
(a) Second-order linear homogeneous partial differen- (d) plasticity and elasticity
tial equation.
(b) Fourth-order linear nonhomogeneous ordinary dif- 190. In the cantilever shown in Fig. E.1, what type of force
ferential equation. is being applied to the wall at point A?
(c) Second-order linear nonhomogeneous ordinary dif-
ferential equation. (a) compression
(d) Fourth-order nonlinear homogeneous ordinary dif- (b) tension
ferential equation. (c) shear
(d) torsion
185. Consider a spring–mass system. If a mass of 1 kg is
attached the spring elongates by 16 mm. The mass 191. Sometimes wrinkles are formed on webs. The reason
is pulled by 10 mm and then released from rest. of wrinkle formation is
The response−2of
the mass in mm is given by.
g = 9.81ms (a) tensile principal stress.
(b) compressive principal stress.
(a) x = 10 cos (24.76t) (c) buckling of thin sheet of web.
(b) x = 10 cos (16t) (d) fatigue in the webs.
(c) x = 10 sin (24.76t)
(d) x = 10 sin (16t) 192. Consider Fig. E.2. Assuming the same wedge angles
(θ), if M1 > M2 then for shock angles β1 and β2 , which
186. If an axial flow turbine is operating at high temperature of the following is true?
then which of the following material is suitable for its
construction? (a) β1 > β2
(b) β1 < β2
(a) nickel alloy (c) β1 = β2
(b) steel alloy (d) cannot say
(c) titanium alloy
(d) aluminum alloy
Fig. E.1 Cantilever beam
M 1 >> 1 β M >> 1 β
1 2 2
Fig. E.2 Oblique shocks on diamond airfoil
r r
t t
Fig. E.3 Thin-walled tubes
193. The presence of carbon monoxide in the combustion The ratio of torsional rigidity of thin-walled closed tube
products means to thin-walled open tube is
r 2
(a) high excess air (a) 2
(b) poor combustion t 2
(b) 3 rt
(c) high thermal efficiency 2
(c) rt
(d) all of the above 3
(d) 4 rt
194. Consider the following thin-walled tubes shown in
Fig. E.3. 195. Consider a simply supported beam of length L, shown
in Fig. E.4.
M
L/2
Fig. E.4 Simply supported beam
A counterclockwise concentrated bending moment M, 199. In scramjet, the flow at entry to the combustion cham-
is applied at mid-span of the beam. The shear force ber is
diagram of the beam is
(a) stagnant (M = 0)
(a) (b)
1 2
1 2
(d)
(c)
1 2
1 2
196. Under uniformly distributed load, the shape of a can- (b) low subsonic (M < 0.3)
tilever beam will be: (c) supersonic
(d) hypersonic
(a) straight line
(b) hyperbolic 200. In a centrifugal compressor, the stagnation pressure rise
(c) parabolic takes place
(d) elliptical
(a) in the diffuser only.
197. The overall efficiency of a rocket will be maximum, (b) in the impeller only.
when aircraft velocity is the exhaust jet velocity. (c) in the inlet guide vanes only.
(d) in the diffuser and impeller.
(a) equal to
(b) one-half 201. In the wind tunnel experiments, a pitot-static probe
(c) double measures
(d) four times
(a) static pressure
198. Let the slenderness ratio (λ) of a column is defined (b) stagnation pressure
in terms of effective length and minimum radius of (c) both static and stagnation pressures
gyration. The range of λ, for Euler’s crippling load for- (d) absolute pressure
mula is
202. The benefits of thin wing design are
(a) λ ≤ 12
(b) λ ≤ 30 (a) Reduced shock stall effect.
(c) 30 < λ ≤ 80 (b) Reduced low-speed stall effects.
(d) λ > 80
(c) Improved landing characteristics. Pressure distribution is integrated to obtain the drag
(d) Reduced Mach angle. coefficient Cd1 . The circular cylinder is then mounted
on a device called external balance which holds it from
203. An aircraft moving at supersonic Mach numbers dis- outside the test section and can measure all the forces
turbs the airstream acting on it when the tunnel is running. From the read-
ing of the balance the drag coefficient is obtained as
(a) behind the Mach lines. Cd2 . The two drag coefficient values obtained from the
(b) on both sides of the Mach lines. two separate measurements can be compared as follows
(c) ahead of the Mach lines.
(d) beyond the ends of the Mach lines. (a) Cd1 = Cd2
(b) Cd1 > Cd2
204. The pressure gradient normal to the body surface of a (c) Cd1 2Cd2
boundary layer flow is (d) Cd2 2Cd1
(a) negative 208. Which of the following statement is TRUE about the
(b) positive steady flow of a fluid in a streamtube?
(c) zero
(d) constant (a) Mass flow is conserved.
(b) The speed increases if the cross-sectional area
205. Under adverse pressure gradient a viscous flow can be increases.
assisted to remain attached to the boundary by (c) The density must be constant.
(d) The pressure must be constant.
(a) Moving the boundary tangential to the flow and
along flow direction. 209. Spoilers can also be used to assist
(b) Vibrating the boundary normal to the flow direction.
(c) Moving the boundary tangential to the flow and (a) flaps
opposite to the flow direction. (b) ailerons
(d) Injecting high momentum fluid into the flow tan- (c) slats
gentially. (d) rudder
206. If is Young’s modulus of elasticity (E) is the shear 210. The difference between air-breathing engine and rocket
modulus (G) is the bulk modulus of elasticity (K) and engine are
is the Poisson’s ratio (ν) then for isotropic materials
consider the following: (a) There is altitude limit.
(b) There is temperature limit.
I. G = E
2(1+ν) (c) Atmospheric air is used for formation of the jet.
II. G = E
3(1−ν)
211. It is better to operate a centrifugal compressor with
III. K = E
3(1−2ν)
respect to mass flow rate
IV. K = E (a) Close to the surge line.

(1−2ν)
(b) Left side of the surge line.
Which of the above are CORRECT? (c) Right side of the surge line.
(d) Both at left and right side of the surge line.
(a) I and IV
(b) I and III 212. Which of the following propellants has highest specific
(c) II and III impulse?
(d) II and IV
(a) Powdered Al + Ammonium Perchlorate.
207. A circular cylinder is mounted in the test section of (b) Monomethyl Hydrazine (MMH) + Nitrogen
a low-speed wind tunnel. The pressure distribution Tetroxide (N2 O4 ).
around it is measured when the tunnel is running. (c) Liquid Hydrogen (LH2 ) + Liquid Oxygen (LOX).
(d) Kerosene + Liquid Oxygen (LOX).
213. Which of the following tend to increase the Dutch roll 219. The rocket travels upward from the launch pad, it gets
tendency? continually easier to accelerate mainly because
(a) wings placed well above the center of mass. (a) the jet stream helps to push it.
(b) sweep-back wings. (b) the rocket continually loses mass as fuel is burned.
(c) dihedral wings. (c) the engines are more efficient at higher altitude.
(d) all of the above. (d) there is little gravity at higher altitude.
214. When the particles of a body or system move approxi- 220. If a hollow, circular torsion bar has a wall thickness (t)
mately perpendicular to the axis of the body the vibra- and a mean radius of (r) then its polar moment of inertia
tion is said to be (J) will be

(a) longitudinal vibrations. (a) 2πr 3 t 1 + 2rt
2
(b) undamped vibrations. (b) 2πr 3 t 1 + 2rt
(c) torsional vibrations. 2
(d) lateral or transverse vibrations. (c) 2πr 3 t 1 − 2rt
4
(d) 2πr 3 t 1 + 2rt
215. If the Airy’s stress function φ = Ay2 + Bxy, the normal
stress σyy is
221. The conditions that determine aerodynamic ceiling are
(a) 0
(a) When thrust available is equal to the thrust required.
(b) 2A
(b) When power available is equal to the power
(c) −2A
required.
(d) B
(c) When the high-speed buffet and low-speed buffet
216. The ratio of maximum shear stress to average shear are equal.
stress of a circular beam is (d) All three of the above.
(a) 2 222. In the following list, identify the monopropellant

3
3
(b) 2 (a) Hydrogen peroxide
(c) 3 (b) Hydrazine
4
4 (c) Mono methyl Hydrazine
(d) 3 (d) Dimethyl Hydrazine
217. Due to downwash, the nature of the lift distribution over 223. The ratio of maximum adiabatic flame temperature in
wings of an aircraft is air to the maximum adiabatic temperature in pure oxy-
gen, is always
(a) equal throughout
(b) parabolic (a) equal to one
(c) elliptical (b) much less than one
(d) hyperbolic (c) much greater than one
(d) uncertain
I. A geostationary orbit can be geosynchronous, but not 224. In a shock tube, the driver gas should have
all geosynchronous orbits are geostationary.
II. An orbit can be both in a Sun-synchronous orbit and (a) Low molecular weight at low temperature.
in a repeat orbit at the same time. (b) High molecular weight at low temperature.
Choose the CORRECT from the following: (c) High molecular weight at high temperature.
(d) Low molecular weight at high temperature.
(a) I
225. The hypersonic reentry capsules have a blunt leading
(b) II
edge because of
(c) I and II
(d) none
(a) aerodynamic heating.
(b) aerobraking. (c) T1

√
(c) payload consideration. (d) T
(d) more structural strength as compared to aerody-
namic shape. 231. When compared to turboprop, turbojet engines are char-
acterized by handling
226. Flow across a curved shock wave is
(a) high mass of air at low velocity.
(a) irrotational (b) high mass of air at high velocity.
(b) incompressible (c) low mass of air at high velocity.
(c) one-dimensional (d) low mass of air at low velocity.
(d) none of the above
232. Modern fighter class engines belong to which of the
227. Time history is required for the analysis of following category?
(a) lateral stability (a) turbojet engines.

(b) longitudinal stability (b) low bypass turbofan engines.
(c) static stability (c) high-bypass turbofan engines.
(d) dynamic stability (d) turboshaft engines.
228. Consider the following statements for a chocked nozzle 233. A typical propeller aircraft is performing a climb in the
flow: international standard atmosphere at constant equiva-
I. The nozzle must be a supersonic one lent airspeed and with the maximum power setting. The
II. Critical condition exists at the nozzle throat power available of this aircraft can be assumed indepen-
III. Flow inside the nozzle must be isentropic dent of airspeed and decreases with altitude according
IV. There is no heat transfer from the hot flowing gases to the following relation
to the wall
0.72
Which of the above is/are TRUE? ρ
Pa,max = Pa,max(sealevel)
ρ0
(a) I
(b) II It can be said that the aircraft is performing a flight.
(c) I, II, and III
(d) I, II, and IV (a) straight (quasi-rectilinear) and steady
(b) straight (quasi-rectilinear) and unsteady
229. Consider the following two statements: (c) curved and steady
I. Separation system is a part of Liquid Propellant (d) curved and unsteady
Rocket Propulsion system.
II. In solid propellant “grain” contains all the chemical 234. To describe the motion of an airplane, four coordinate
elements for complete burning. systems are used. In order to describe the attitude of an
Which of the above is/are CORRECT? airplane (body axes), with respect to the moving Earth
axis system, three Euler angles are used. The sequence
(a) I in which these angles are used is very important. The
(b) II correct sequence of these angles to obtain the orienta-
(c) Both I and II tion of the body axes, starting from the moving Earth
(d) none axes is
230. With all other things remaining constant and with T (a) angle of pitch (θ), angle of roll (φ), angle of yaw
the outside static air temperature expressed in Kelvin, (ψ).
the specific fuel consumption of a turbojet powered air- (b) angle of pitch (θ), angle of yaw (ψ), angle of roll
plane in a constant Mach number cruise in still air is (φ).
proportional to (c) angle of yaw (ψ), angle of roll (φ), angle of pitch
(θ).
(a) T (d) angle of yaw (ψ), angle of pitch (θ), angle of roll
(b) T12 (φ).
235. A pilot wants to perform a steady coordinated turn. The 240. Which of the following statement related to the “snow-
pilot initiates the turn by banking the aircraft. After ball effect” with respect to an aircraft is correct?
banking the aircraft the pilot must
(a) By reducing the empty weight of the aircraft more
(a) increase the pitch attitude. payloads can be transported or the range can be
(b) increase the thrust. increased.
(c) decrease the thrust. (b) Reducing the structural weight of an aircraft,
(d) increase the pitch attitude and the thrust. additional weight reductions due to the “snowball
effect” are also structural.
236. The decision airspeed is an important speed during take- (c) Once an initial weight reduction is achieved, all
off. For a given multi-engine aircraft, this airspeed is other systems of the aircraft can be reduced in
weight as well.
(a) independent of the balanced field length. (d) Given a combination of fuel and payload, an initial
(b) a function of weight, altitude, and temperature. weight reduction induces extra weight reductions.
(c) a function of aircraft weight.
(d) generally much higher than the rotation speed. 241. To limit the wing bending, which of the following solu-
tion should be applied?
237. Lift dumpers are used to reduce the landing ground run
distance (a) A high strength material as the skin material of the
wing.
(a) By decreasing lift on the wing, this makes it easier (b) A high strength material as the web plate material
for the pilot to put the aircraft on the ground. for the spars.
(b) By decreasing lift on the wing, this creates a larger (c) A high stiffness material as the skin material of the
normal force on the wheels and thus a larger ground wing.
drag when applying brakes. (d) A high stiffness material as the web material for the
(c) By disturbing the airflow on the wing, this creates ribs.
a lot of turbulence and thus aerodynamic drag.
(d) By increasing lift on the wing, this makes it easier 242. The efficiency of a jet engine is higher at
for the pilot to put the aircraft on the ground.
(a) low altitudes
238. Which of the following statements about material and (b) high altitudes
structural properties are true? (c) high speeds
(d) low speeds
(a) The material properties and structural properties are
identical. 243. Only rocket engines can be propelled to space because
(b) The material properties and structural properties are
complementary. (a) They can generate very high thrust.
(c) Structural properties depend on material properties (b) They have high propulsion efficiency.
and geometrical features. (c) These engines can work on many fuels.
(d) Material properties are much more important than (d) They are not air-breathing engines.
structural properties.
244. The thrust of a jet propulsion power unit can be
239. Even though the first metal aircraft appeared in early increased by
1930s, the real problems with metal fatigue showed up
about 20 years later. What is the reason for this time (a) injecting ammonia into the combustion chamber.
difference? (b) burning fuel after gas turbine.
(c) injecting water into the compressor.
(a) It takes quite some time to initiate and grow a (d) all of the above.
fatigue crack to a detectable size.
(b) In the early 1950s aircraft flew at higher altitudes 245. A turboprop is preferred to turbojet because
and had pressure cabins.
(c) In the early 1950s the aircraft were much bigger (a) it can fly at supersonic Mach numbers.
and much faster. (b) it can fly at high altitudes.
(d) The metal alloys of the 1930s were not fatigue- (c) it has high power for takeoff.
sensitive like the later ones.
(d) it has high propulsive efficiency at high Mach num- (c) III. The highest possible velocity.
bers. (d) IV. The lowest possible velocity.
246. The degree of reaction is usually kept for all types Which of the following combination is TRUE?
of axial flow compressors.
(a) I and III
(a) 0.4 (b) II and III
(b) 0.3 (c) I and IV
(c) 0.5 (d) II and IV
(d) 0.2
252. For pull-up maneuvers, if n stands for load factor and Ua
247. Adding dihedral to a glider can improve its is the freestream velocity then, which of the following
relation holds for turn radius R?
(a) pitch stability
Ua2
(b) roll stability (a) R = g(n+1)
(c) yaw stability Ua2
(b) R = g(n−1)
(d) static Stability Ua
(c) R = g(n+1)2
248. At apogee, the centrifugal force is Ua3
(d) R = g(n−1)
(a) less than the gravitational force. 253. For a given airfoil, which of the following relationship
(b) more than the gravitational force. is BEST suited to represent the lift coefficient (CL )?
(c) equal to the gravitational force.
(d) not dependent on the gravitational force. (a) CL =(α)
(b) CL =(α, Ma , Re)
249. For the flow past a NACA 2315 airfoil, which of the
(c) CL =(Ma , Re)
following relation is TRUE?
(d) CL =(α, Re)

(a) Cd ∝ M∞ 2 −1
(where α is angle of attack, Ma is Mach number, Re is
(b) Cd ∝ √ 12 the Reynolds number, and stands for function)
(2M∞ −1) 254. For a given aircraft, consider the following statements:
(c) Cd ∝ M∞ −1
2 2 I. For a given lift coefficient (CL ), the drag coefficient
(d) Cd ∝ M∞ − 1
(CD ) is much larger at supersonic speeds than at sub-
250. For the maximum endurance of a jet-propelled airplane sonic speeds.
consider the following statements? II. For a given lift coefficient (CL ), the drag coefficient
I. Flying with maximum aerodynamic efficiency. (CD ) is much smaller at supersonic speeds than at sub-
II. Having the lowest possible thrust-specific fuel con- sonic speeds. L
sumption. III. The values of D max
at supersonic speeds are
III. Having the highest possible thrust-specific fuel con- smaller than at subsonic
L speeds.
sumption. IV. The values of D max
at supersonic speeds are larger
IV. Carry a lot of fuel. than at subsonic speeds,
Which one of the following is true? Which of the following is/are TRUE?
(a) I and II (a) I and III

(b) I, II, and IV (b) I and IV
(c) I, III, and IV (c) II and III
(d) II and IV (d) II and IV
251. For an aircraft, to obtain the largest possible turn rate 255. Under which of the following condition, the free vibra-
consider the following statements. tion of a linear spring–mass–dashpot system undergoes
a non-oscillatory motion?
(a) I. The highest possible load factor.
(b) II. The lowest possible load factor. (a) undamped
(b) underdamped
(c) overdamped (a) vr = 4ve

(d) critically damped (b) vr = ve
(c) vr = v4e
256. Which of the following airfoil will be appropriate in (d) vr = 2ve
constructing the vertical tail of an aircraft?
262. In the flow across a normal shock, if M1 and M2 repre-
(a) NACA 2-series positively cambered airfoil sent the Mach number upstream and downstream of the
(b) symmetric airfoil shock, what is the asymptotic value of the downstream
(c) negatively cambered airfoil Mach number for an ideal gas having γ = 1.4.
(d) supercritical airfoil
√
(a) √0.5
257. Compared to the monoplane, the biplane wing experi- (b) 0.14
ences a in the lift and a in the drag. (c) 17
√
(d) 7
(a) increase, decrease
(b) decrease, increase
263. Which of the following is not a characteristic of a NACA
(c) increase, increase
5-digit airfoil?
(d) decrease, decrease
(a) Roughness has less effect.
258. If an elliptical wing has an aspect ratio (AR → ∞),
(b) Higher maximum lift coefficient.
the value of lift curve slope will be approximately
(c) Low pitching moment.
equal to
(d) High pitching moment.
(a) ∞
264. For a solid object subjected to the arbitrary loading,
(b) 8
which of the following material properties will be con-
(c) 5
stant irrespective of the loading?
(d) 6
(a) specific volume only
259. Which of the following inequalities are true for an air-
(b) density and elasticity
craft engine cruising at subsonic velocity? (Intake, com-
(c) mass only
pressor, turbine, and nozzle efficiencies are denoted by
(d) volume, mass, and density
ηi , ηc , ηt , and ηn , respectively)
265. For a solid rod of circular cross section under torsional
(a) ηt < ηc < ηi < ηn
vibration, the boundary conditions are changed from
(b) ηc < ηt < ηi < ηn
fixed-free to fixed-fixed. The fundamental natural fre-
(c) ηi < ηc < ηt < ηn
quency is changed by a factor of
(d) ηt < ηc < ηn < ηi
(a) 2.0
260. At an altitude a 35,000 feet, a supersonic nozzle is
(b) π2
designed to produce maximum thrust. At which alti-
(c) π
tude, the nozzle will operate in underexpanded condi-
(d) 21
tion?
266. For an axial compressor stage having the degree of reac-
(a) 40,000 feet
tion 0.5. Which of the following is NOT true?
(b) near service ceiling
(c) 35,000 feet (a) Velocity triangles at the entry and exit of the rotor
(d) 34,000 feet are symmetrical.
(b) The pressure drop is equally shared by the rotor and
261. Through a single impulse out-of-plane maneuver, a
the stator.
satellite is transferred from its geostationary orbit to a
(c) The stator has a larger contribution to the total work
circular polar orbit of same radius. The relation between
extracted or work done.
the magnitude of required velocity (vr ) for orbit transfer
(d) The whirl components are same at the inlet of rotor
in terms of magnitude of escape velocity (ve ) is
and diffuser.
267. To obtain maximum sink rate, the glider must be oper- (c) I and IV are true.
ated at (d) II and IV are true.
CL
(a) maximum CD condition 273. The CREEP in a turbine blade is
CL
(b) minimum CD condition
CL (a) pronounced more at increased heats near melting
(c) minimum 3/2 condition
CD point.
CL (b) independent of exposure time to thermal stresses.
(d) minimum 1/2 condition
CD
(c) a temporary deformation.
268. In airplanes, a step input at the rudder excites which (d) a sudden plastic deformation.
dynamic mode of instability?
274. The construction of a fuselage in the modern aircrafts
(a) phugoid are based on
(b) dutch roll
(a) monocoque design
(c) spiral divergence
(b) semimonocoque design
(d) roll Subsidence
(c) truss structure
269. Bernoulli’s equation is based on the assumption that (d) geodesic construction
(a) There is no loss of energy of the liquid flowing. 275. The negative Poisson ratio in a circular rod shows that
(b) The velocity of flow is uniform across any cross the
section of the pipe.
(a) longitudinal strain increases when the lateral strain
(c) No force except gravity acts on the fluid.
decreases.
(b) lateral and longitudinal strains are equal.
270. Which of the following is the necessary criterion for (c) lateral strain increases when the longitudinal strain
stick fixed longitudinal balance and static stability? increases.
(d) longitudinal strain decreases when the lateral strain
(a) CM,cq at zero lift must be positive. increases.
∂CM,cq
(b) ∂αa must be positive (αa is absolute angle of
276. The relation between Bulk modulus and Young’s mod-
attack).
ulus is
(c) CM,cq at zero lift must be negative.
(d) Slope of CL versus αa must be negative. (a) E = (1−2ν)
K
271. The design characteristic that cannot improve the sta- (b) K = (2+ν)
K
bility of the roll subsidence is (c) E = (2−ν)

K
(d) K = (1−2ν)
E
(a) high and dihedral angles.
(b) swept wing. 277. In a canard wing configuration, the position of the C.G.
(c) low wing. is
(d) high and anhedral wings.
(a) far aft the wing and before the neutral point.
272. Which of the following are TRUE regarding service and (b) at 31rd of the chord.
absolute ceilings? (c) at 41th of the chord.
I. The rate of climb drops below a prescribed value in (d) at the neutral point.
absolute ceiling.
II. The rate of climb drops below a prescribed value in 278. The headwind or tailwind does not affect the
service ceiling.
III. The rate of climb drops to zero in absolute ceiling. (a) range of the aircraft.
IV. The rate of climb drops to zero in service ceiling. (b) the takeoff distance.
(c) endurance of the aircraft.
(a) I and II are true. (d) the angle of glide.
(b) II and III are true.
279. The area under the flight envelope of an aircraft is 285. Euler’s equation in the differential form for the motion
of liquids is given by
(a) larger for general aviation aircraft.
(b) larger for fighter aircraft. (a) ∂p
ρ − gdz + vdv = 0
(c) same for all the aircraft types. (b) ρ∂p + gdz + vdv = 0
(d) not dependent on the load factor of the aircraft. (c) ∂p
ρ + gdz + vdv = 0
(d) ρ∂p − gdz + vdv = 0
280. The value of bulk modulus of a fluid is required to deter-
mine 286. The square root of the ratio of inertia force to gravity
force is called
(a) Reynolds number
(b) Mach number (a) Reynolds number
(c) Froude number (b) Froude number
(d) Prandtl number (c) Mach number
(d) Euler number
281. For the maximum power transmission, what should be
the diameter of the nozzle (d)? (where D = Diameter 287. Which of the following statement holds correct?
of pipe, f = Darcy’s coefficient of friction for pipe, and
L = Length of pipe). (a) In laminar flow, Newton’s law of viscosity does not
apply.
1/2
(a) d = D5 (b) A forced vortex occurs when fluid particles rotates
8fL
1/3 about its own axis.
D5
(b) d = 8fL
(c) In turbulent flow, there are neither crosscurrents nor
1/4 eddies.
D5
(c) d = 8fL (d) A free vortex occurs, when fluid particles rotates
1/5
D5 about its own axis.
(d) d = 8fL
282. The velocity profile of a turbulent flow through a closed
I. Cruising phase of civil aviation flights takes place in
conduit is
TROPOSPHERE.
(a) parabolic II. Cruising phase of civil aviation flights takes place in
(b) hyperbolic STRATOSPHERE.
(c) linear III. Entropy remains constant across Prandtl–Meyer
(d) logarithmic expansion wave.
IV. Static temperature increases across Prandtl–Meyer
283. At high Reynolds number, expansion wave.
Which of the following is/are TRUE?
(a) the inertial forces control and viscous forces are
unimportant. (a) I, III, and IV
(b) viscous forces are predominant. (b) II, III, and IV
(c) inertial forces are unimportant and viscous forces (c) II and III
control. (d) II and IV
(d) both the inertial and viscous forces are predomi-
nant. 289. The Schlieren optical flow visualization technique is
sensitive to changes in
284. The power loss in the flow through an orifice is
the flow through a venturi tube having the same throat (a) fluid density.
diameter. (b) first derivative of the fluid density.
(c) second derivative of the fluid density.
(a) same as (d) third derivative of the fluid density.
(b) more than
(c) less than 290. An ideal Ramjet is based upon which of the following
(d) Insufficient data cycle?
(a) Brayton cycle (a) feed system

(b) Carnot cycle (b) conversion system
(c) Otto cycle (c) acceleration system
(d) Rankine cycle (d) separation system
291. High critical Mach number as a design criteria, the high- 297. Identify the monopropellant
speed subsonic aircraft are made of
(a) hydrogen peroxide
(a) thin airfoils (b) hydrazine
(b) thick airfoils (c) mono methyl hydrazine
(c) laminar airfoils (d) dimethyl hydrazine
(d) diamond airfoils
298. Eccentricity of an elliptical orbit determines
292. Consider the following statements about subsonic com-
pressible flow over an airfoil: (a) shape of the ellipse
I. Thick airfoils at small angle of attack. (b) size of the ellipse
II. Mach numbers do not approach too close to one. (c) both of these
III. Mach numbers should always be greater than one. (d) none of these
IV. Inviscid and irrotational flow.
The statements which are applicable to Prandtl–Glauert 299. A particle moving at uniform velocity in sea level stan-
rule? dard air creates two disturbance spheres at two different
times. If the later sphere is outside the initial sphere, the
(a) I, II, and IV particle Mach number is
(b) I, III, and IV
(c) II and IV (a) subsonic
(d) III and IV (b) transonic
(c) sonic
293. In the limit of a high Mach number flow across a shock (d) supersonic
wave, which of the following is TRUE?
300. Boundary layer thickness depends on

(a) pp21 = γ−12γ
M1 sin β
(a) wall roughness
(b) pp21 = γ−12γ
(M1 sin β)2 (b) freestream velocity
(c) Reynolds number
(c) ρρ21 = γ+1
γ−1 (d) all of the above
ρ2 γ
(d) ρ1 = γ−1
301. Ramjet and Scramjet engine differentiates themselves
294. In steady, level turning flight of an aircraft at a load in the concept of
factor n, the ratio of the horizontal component of lift
and aircraft weight is (a) compression
(b) combustion
√
(a) √n − 1 (c) expansion
(b) √n + 1 (d) afterburner
(c) √n2 − 1
(d) n2 + 1 302. When the freestream velocity in a subsonic wind tunnel
test section is increased by 3 times, the power require-
295. Which of the following is NOT a Duct jet engine? ments to run the tunnel will be
(a) turbojet (a) increased by 3 times

(b) turbofan (b) increased by 9 times
(c) pulse jet (c) increased by 27 times
(d) turboprop (d) increased by 81 times
296. Which of the following is NOT a part of the liquid pro- 303. In designing the Cryogenic wind tunnel, which of the
pellant rocket propulsion? following nondimensional parameters must match?
(a) Prandtl number and Mach number. 310. The maximum operating flow rate through a centrifugal
(b) Mach number and Strouhal number. compressor at a given RPM is limited by
(c) Reynolds number and Prandtl number.
(d) Reynolds number and Mach number. (a) impeller stall
(b) surge
304. A turbofan engine has a bypass ratio of 4 and a total mass (c) choking of diffuser throat
flow rate of 100 kgs−1 . The mass flow rate through the (d) inlet flow distortion
bypass duct is
311. For subsonic axial compressors, the inlet guide vanes
(a) 25 kgs−1 have been used to
(b) 80 kgs−1
(a) give the incoming air a TANGENTIAL velocity
(c) 100 kgs−1
component in the direction NORMAL to the blade
(d) 400 kgs−1
velocity.
305. An aircraft of mass 2500 kg in straight and level flight (b) reduce the relative Mach number of the flow.
at a constant speed of 100 ms−1 has available excess (c) increase the stage work, for the same rotor exit
power of W. The steady rate of climb it can attain at velocity.
that speed is (d) operate with a lower mass flow rate for a given rel-
ative Mach number of the flow.
(a) 100 ms−1
(b) 60 ms−1 312. Using 2-dimensional approximations in subsonic axial
(c) 40 ms−1 compressors, for the calculations of blade passage flow
(d) 20 ms−1 fields, which of the following methods can be used?
I. streamline curvature method
306. What are the four typical loads on an aircraft? II. stream function method
III. potential function method
(a) tension, torsion, creep, elongation. IV. Euler equation solvers
(b) elasticity, shear, compression, torsion. Select the best answer(s) given below.
(c) tension, compression, torsion, shear.
(d) compression, buckling, elasticity, shear. (a) I and II
(b) I, III, and IV
307. is where the materials natural inbuilt elasticity (c) II, III, and IV
enables it to stretch when under load. (d) all
(a) elasticity 313. A wing of taper ratio 0.2 has a planform area of 48 m2 .
(b) buckling If the total wing span is 16 m, the root and tip chords
(c) shear of the wing will be , respectively.
(d) creep
(a) 4 m, 0.8 m
308. Select one of the factors affecting CREEP? (b) 7.0 m, 1.4 m
(c) 3.0 m, 0.6 m
(a) strength of the material (d) 5.0 m, 1.0 m
(b) shear
(c) buckling of the material 314. The ideal static pressure coefficient of a diffuser with
(d) the duration of the load applied an area ratio of is
309. A model aircraft in a wind tunnel that is operating at (a) 0.25

50 ms−1 develops a minimum pressure coefficient of at (b) 0.5
some point on its upper surface. The local airspeed at (c) 0.75
that point is (d) 1.0
(a) 16.67 ms−1 315. An airfoil has following characteristics.

(b) 50 ms−1 I. Design lift coefficient of 0.3.
(c) 150 ms−1 II. Maximum camber located at 15 chord length from
(d) 200 ms−1 leading edge.
III. Maximum thickness of 12 chord length. 321. When the pressure drop across a converging–diverging
Which of the following airfoil is best represented by the nozzle is different from the design value for isentropic
above characteristics? flow, which of the following is possible?
(a) NACA 0012 (a) There is one normal shock in converging part and
(b) NACA 23012 one normal shock in diverging part.
(c) NACA 4412 (b) There is only one normal shock in converging part
(d) NACA 4512 and none in diverging part.
(c) There is only one normal shock in diverging part
316. Isentropic flow is and none in converging part.
(d) There are two or more normal shocks in diverging
(a) irreversible adiabatic flow. part, depending upon the pressure drop and none in
(b) reversible adiabatic flow. converging part.
(c) ideal fluid flow.
(d) frictionless reversible flow. 322. A spring used to absorb shocks and vibrations is
317. If the flow conditions satisfy Laplace equation, then (a) conical spring
flow (b) torsion spring
(c) disk spring
(a) is rotational. (d) leaf spring
(b) does not satisfy continuity equation.
(c) is irrotational but does not satisfy continuity equa- 323. Modulus of rigidity is defined as the ratio of
tion.
(d) is irrotational and satisfies continuity equation. (a) longitudinal stress and longitudinal strain.
(b) volumetric stress and volumetric strain.
318. A thin flat plate has been placed parallel to the flow (c) lateral stress and lateral strain.
direction. The relative magnitudes of skin friction and (d) shear stress and shear strain.
pressure drags will be
324. A concentrated mass m is attached at the center of a rod
(a) negligible skin friction as well as pressure drags. of length 2L as shown in Fig. E.5. The rod is kept in a
(b) negligible pressure drag and maximum skin friction horizontal equilibrium position by a spring of stiffness
drag. k. For very small amplitude of vibration, neglecting the
(c) maximum pressure drag and negligible skin friction weights of the rod and spring, the undamped natural
drag. frequency of the system is
(d) pressure drag equals skin friction drag.
(a) mk
319. An isentropic nozzle is discharging air through critical
pressure ratio. If the back pressure is further decreased (b) 4k
m
the discharge will (c) 2k
m
k
(a) decrease (d) 2m
(b) increase
(c) remain unaffected 325. The natural frequency of the system shown in Fig. E.6
(d) come to a dead stop due to shock waves is

320. The thickness of boundary layer in a turbulent flow is k
(a) 2m

(a) √5x (b) mk
Rex
5.835x
(b) √ (c) 2k
Rex
0.377x
m
(c) (d) 3k
(Rex )1/5 m
5.377x
(d)
(Rex )1/5
k
m
L L
Fig. E.5 Spring–mass system
K/2
K
K/2
326. Velocity potential

for an incompressible
fluid flow is (c) Uwδ
given as: φ = 2 x + 2y − y Assume the value of
2 2 (d) 23 Uwδ
stream function at the origin to be zero. The value of
stream function at the point (2, 2) is 328. For the given values of pressures and densities,
upstream and downstream of a normal shock as shown
(a) 8 in Fig. E.8, the velocity upstream of the normal shock
(b) 16 will be?
(c) 4
(d) 2 (a) 4 ms−1
(b) 5 ms−1
327. Consider a laminar flow over a flat plate of width w. At (c) 10 ms−1
section ‘1–1’, the velocity profile is uniform as shown (d) 15 ms−1
in Fig. E.7. The x−direction velocity profile at section
‘2-2’ is given by 329. Which of the following type of viscous damping will
give periodic motion to the vibrating body?
u y y 2 I. underdamping
=2 −
U δ δ II. critical damping
III. overdamping
where δ is the boundary layer thickness. Select the answer among the following:
The volume flow rate through section ‘2-2’ is given by (a) I

1
(b) II
(a) 2 Uwδ (c) III
1
(b) 3 Uwδ (d) I and II
U U
1 2
y
u(y) δ
x
1 2
Fig. E.7 Laminar flow over a flat plate
Upstream Downstream
5 5
P =1 x 10 Pa P2=21 x 10 Pa
1
ρ = 1 kg/m 3 ρ = 5 kg/m 3
1 2
v1 v2
Normal shock
Fig. E.8 Flow across a normal shock
330. The aircraft structural materials which show direction (c) steel
dependent properties are called (d) concrete
(a) homogeneous 333. Which of the following stress–strain response curve

(b) viscoelastic best represents linear elastic – perfectly plastic behav-
(c) isotropic ior?
(d) anisotropic
σ
331. Modulus of rigidity and bulk modulus of a material are
found to be 60 GPa and 140 GPa, respectively. Then
consider the following:
I. Elasticity modulus is nearly 200 GPa.
II. Poisson’s ratio is nearly 0.3.
III. Elasticity modulus is nearly 158 GPa.
IV. Poisson’s ratio is nearly 0.25.
(a) ε
(a) I and II
(b) I and IV σ
(c) II and III
(d) III and IV
332. Which of the following material has NEGATIVE value

of Poisson’s ratio?
(a) rubber
(b) novel Foam (b) ε
σ 337. In a hypersonic flow (M = 8) past a wedge (total vertex

angle, 30o ), the hypersonic similarity parameter (H) is
4π
(a) 3
2π
(b) 3
96
(c) π
48
(d) π
(c) ε 338. In the pitot-static probe (as shown in Fig. E.9) the static
pressure holes are generally kept at a distance
σ
(a) D = 2d
(b) D = 8d
(c) D = 25d
(d) D = 50d

339. If the Airy’s stresses function is φ = 3y2 + 2xy , the
(d) ε normal stresses σxx and σyy are , respectively.
(a) 6, 0
334. Two identical circular rods of same diameter and same
(b) 0, 6
length are subjected to same magnitude of the axial
(c) 0, 0
tensile force. One of the rods is made out of mild steel
(d) 6, 6
having the modulus of elasticity 206 GPa. The other rod
is made out of cast iron having the modulus of elasticity 340. The spacecraft reentry into the atmosphere takes place
of 100 GPa. Assume both the materials to be homoge- in the Mach number range, M 28–32. At such a high
neous and isotropic and axial force causes same amount Mach numbers, the gases present near the spacecraft
of uniform stress in both the rods. The stresses devel- nose are subjected to high temperature. To visualize this
oped are within the proportional limit of the respective phenomenon in the wind tunnel, which of the following
materials. Which of the following is correct? technique is best suited?
(a) Both rods elongate by the same amount. (a) oil flow visualization with color dyes
(b) Mild steel rod elongates more than the cast iron rod. (b) shadowgraph technique
(c) Cast iron rod elongates more than the mild steel (c) schlieren technique
rod. (d) interferometry technique
(d) As the stresses are equal strains are also equal in
both the rods. 341. In the wind tunnel testing of, the maximum lift coeffi-
cient is found to be 1.5. If the aircraft has an elliptical
335. In a Mohr’s circle, the radius of the circle is represented wing planform of gross area 25m2 and wing tip span of
as 10 m, then the maximum induced drag coefficient will
be?
σx −σy 2
(a) 2 + τxy
2
9
(a) 16π
(σx −σy )2
(b) + τxy
2 16
(b) 9π
2
2
σx −σy (c) 1.5
(c) 2 − τxy
2
(d) ∞

(d) (σx − σy )2 + τxy
2
342. A subsonic wind tunnel contains a circular Effuser (con-
336. Which of the following is a valid potential function (φ)? traction cone) with entry and exit diameters of 6 m and
2.5 m, respectively. The contraction ratio of effuser is
(a) φ = c ln x
(a) 0.17
(b) φ = c cos x
(b) 5.76
(c) φ = 3xy
(c) 2.4
(d) φ = c x 2 + y2
(d) 0.42
Static pressure
holes
Probe diameter = d
Fig. E.9 Pitot-static probe
343. The equation of free vibration of a system is 347. If p is the internal pressure in a thin-walled cylinder of
d2 x
dt2
+ 16π 2 x = 0. Its natural frequency would be diameter d and thickness t, the correct expression of
Hoop stress is
(a) 2π Hz
pd
(b) 4π Hz (a) t
(c) 16π 2 Hz (b) pd
2t
(d) 2 Hz pd
(c) 4t
pd
344. Consider an incompressible laminar boundary layer (d) 8t
flow past a flat plate at zero degree angle of attack.
If the direction velocity profile is given by 348. Consider the following statements. The efficiency of a
nozzle depends on
u y y 2
=2 − I. Size and shape of the nozzle.
U δ δ II. Flow Mach number.
III. Flow Reynolds number.
where δ is the boundary layer thickness, then the bound-
IV. Material of the nozzle.
ary layer shape factor will be
Which of the following is/are correct?
(a) 5
(a) I, II, and III
(b) 2.5
(b) I, III, and IV
(c) 0.4
(c) I and IV
(d) 1.0
(d) II and IV
345. The shape factor of a typical turbulent boundary layer
349. If vj is the exhaust jet velocity and vr is the rocket veloc-
lies in the range
ity, then the propulsive efficiency of the Rocket is given
(a) 0.1 - 0.2 by
(b) 0.2 - 0.5 v
j
2 vr
(c) 1.3 - 1.4 (a) ηp = v 2
(d) 5.5 - 5.6 1+ vrj
v
2 vrj
346. The velocity potential function in a two-dimensional (b) ηp = v 2
1− vrj
flow field is given by φ = 2x + y2 . The magnitude of
the velocity at a point (1, 3) is (c) ηp = 2 v
1+ vrj
(a) 0 (d) ηp = 2 v
1− vrj
(b) 4√
(c) 2 10 350. Consider the beam with pointed and distributed loads
(d) 40 shown in Fig. E.10.
400 N 800 N
1m 2m
w=50 N/m
Ra Ra
Fig. E.10 Beam with pointed and distributed loads
352. The impulse response of an initially relaxed linear sys-

tem is u (t) exp−2t , where u (t) is a unit step function.
To produce a response of tu (t) exp−2t , the input must
be equal to
(a) (a) 2u (t) exp−2t

(b) 21 u (t) exp−2t
(c) u (t) exp−2t
(d) u (t) exp−t
353. If M represents the Mach number, t is thickness-to-

chord ratio and γ is the ratio of specific heat. Then the
(b) transonic similarity parameter (K) will be
√
M2 −1
(a)
[t(γ+1)]1/3
M2 −1
(b) 2/3
[t(γ+1)]
√
M2 −1
(c)
(c) [t(γ+1)]−1/3
M2 −1
(d)
[t(γ+1)]−2/3
354. Closed-loop transfer function of a unity-feedback sys-

tem is given by Y(s)
R(s) = s+1 . Steady-state error to ramp
1
(d) input is
351. Consider the following statements: (a) ∞
I. At the point of contraflexure, shear force is either zero (b) τ
or changes its sign. (c) 1
II. For the flow past a circular cylinder, when the bound- (d) τ1
ary layer transits from laminar to turbulent then skin
friction drag increases. 355. If the roots of the characteristic equation are given by
III. The number of natural frequencies of an elastic s1,2 = −3 ± 2j the value of damping ξ and damped nat-
beam with cantilever boundary conditions is infinite. ural frequency are
Which of the above is/are TRUE? √
(a) √3 , 13
(a) I, only 13
(b) I & II (b) √1 , 2
13 √
(c) I & III (c) √1 , 13
(d) II & III 13
(d) √3 , 2
13
356. The condition that all the roots of the polynomial, 361. For the minimum power required for a level unac-
s = a0 s3 + a1 s2 + a2 s + a3 ; (ai > 0), have negative celerated flight, the aerodynamic
condition that holds
real parts is given by between
zero-lift
drag CD,0 , and induced drag due to
lift CD,i is given by
(a) a1 a3 > a0 a2
(b) a1 a0 > a2 a3 (a) CD,0 = 3CD,i
(c) a1 a2 > a0 a3 (b) CD,0 = 13 CD,i
(d) a2 a0 > a1 a3 (c) CD,0 = C2D,i
1
357. In a centrifugal pump, the liquid enters the pump (d) CD,0 = CD,i
2
(a) at the center 362. The maximum lift-to-drag ratio for the CP-1 aircraft is
(b) at the bottom 13.6. In the power-off mode, the minimum glide angle
(c) at the top will be
(d) from sides
(a) θmin = 4.2o
358. Which of the following is NOT correct about winglet? (b) θmin = 45o
(c) θmin = 60o
(a) increases the lift generated at the wingtip. (d) θmin =0
(b) increases the lift induced drag caused by wingtip
vortices. 363. In the previous question, if the CP-1 aircraft begins to
(c) increases the effective aspect ratio of a wing. glide in power-off mode at an altitude of 3048 m, the
(d) increases fuel efficiency. maximum range covered by the aircraft measured along
the ground will be approximately equal to
I. static pressure (a) 41500 m
II. total pressure (b) 20750 m
III. dynamic pressure (c) 83000 m
Which of the above affects airspeed measurement using (d) 10000 m
a pitot probe?
364. In a steady level flight, the difference in the rate of
(a) I climbs (R/C) between the service and absolute ceilings
(b) II will be (in ftmin−1 )
(c) II and III
(a) 0
(d) All
(b) 50
360. The characteristic equation of a closed-loop system as (c) 100
shown in the Fig. E.11, is (d) ∞
365. An object will escape from an Earth orbit, if the velocity

(a) s2 + 11s + 10 = 0
is:
(b) s2 + 11s + 130 = 0
(c) s2 + 10s + 120 = 0 (a) equal to 3.95 kmsec−1
(d) s2 + 10s + 12 = 0 (b) equal to 7.9 kmsec−1
R(s) + 3 Y(s)
4 s+1
−
10
s+10
Fig. E.11 Closed-loop system

(c) more than 7.9 kmsec−1 but must be less than 11.17 (d) I, IV, and V
kmsec−1
(d) equal to or more than 11.17 kmsec−1 370. Consider the following propellant combinations:
I. liquid O2 + liquid H2
366. On hot summer days, the air density is lesser than that II. H2 + F2
on cool winter days. For a given aircraft the ground III. monomethylhydrazine (MMH)+ N2 O4
liftoff distance on a summer day will be Which of the above forms a hypergolic mixture?
(a) shorter than the winter day (a) I

(b) equal to the winter day (b) I and II
(c) half than that on a winter day (c) II and III
(d) longer than the winter day (d) all the above
367. Consider the following statements: 371. Consider the following:

I. Static margin should be positive. I. concept of streamline
II. Static margin should be zero or negative. II. concept of streakline
III. Shorter the static margin implies more stable air- III. concept of pathline
craft. Which of the above concept(s) is/are IMAGINARY in
IV. Larger the static margin implies more stable aircraft. nature?
For static stability, which of the following are true?
(a) I only
(a) I and III (b) I and II
(b) II and III (c) II and III
(c) I and IV (d) I and III
(d) II and IV
372. Which of the following statement is NOT correct?
368. The heat transfer rate during the ballistic reentry phase
is an important parameter for designing a spacecraft. (a) A calorically perfect gas must be thermally perfect.
When the velocity of a given spacecraft is increased by (b) A thermal perfect gas must be calorically perfect.
3 times, the associated heat transfer rate will be? (c) A perfect gas must be both thermally and calorically
perfect.
(a) increased by 3 times (d) A real gas behaves like an ideal gas at higher tem-
(b) increased by 9 times perature and lower pressure.
(c) decreased by 9 times
(d) increased by 27 times 373. A typical rocket consumes fuel in the order of
369. Consider the following fuel–oxidizer combinations in (a) 8000 kgs−1

lightweight propellants: (b) 9000 kgs−1
I. Kerosene–Oxygen combination has the lowest adia- (c) 10000 kgs−1
batic flame temperature. (d) 11000 kgs−1
II. Kerosene–Oxygen combination has the highest spe-
cific impulse. 374. The standard rocket ejects burnt gases at speeds of over
III. Hydrogen–Oxygen combination has the highest adi-
(a) 2000 ms−1
abatic flame temperature.
(b) 3000 ms−1
IV. Hydrogen–Fluorine combination has the highest
(c) 3500 ms−1
adiabatic flame temperature.
(d) 4000 ms−1
V. Hydrogen–Fluorine combination has the lowest spe-
cific impulse. 375. Which of the following represents a turbojet engine?
Which of the following combinations are true?
(a) Olympus 593
(a) I, II, III, and V (b) Spey
(b) I, II, IV, and V (c) Tyne
(c) I and IV (d) Pegasus
376. The efficiency of a jet engine will be higher at 380. For a given aircraft flying level at steady state with sym-
metric load distribution, the circulation will be given
(a) high Altitudes by (Maximum circulation at root chord = 0 )
(b) low Altitudes
(c) high speeds (a) = 0 cos φ
0
(d) low speeds (b) = cos φ
(c) = 0 sin φ
0
377. For atmospheric ozone, which of the following state- (d) = sin φ
ments are NOT true?
381. The boundary layer shape factor (H) for a zero-pressure
(a) Ozone levels in the stratosphere have been danger- gradient boundary layer (Blasius Profile) is
ously reduced, due to release of CFCs containing
refrigerants and propellants. (a) 0
(b) Automobile exhaust can help replace ozone in the (b) 0.83
troposphere. (c) 2.59
(c) Ozone in the stratosphere protects the Earth from (d) 100
excessive UV radiation.
(d) Ozone in the troposphere has toxic effects on ani- 382. The turbulence number of a uniform horizontal flow at
mals and plants. 25 ms−1 is 6. If the turbulence is isotropic, determine
the mean square values of the fluctuations.
378. For an aircraft flying at supersonic speeds, consider the
following statements: (a) 3.37
I. The wave drag of wing with sharp leading edge will (b) 6.75
be lower. (c) 13.5
II. The wave drag of wing with rounded leading edge (d) 45.56
will be lower.
III. The wing with sharp leading edge will have well- 383. For a wing the root chord, tip chord, and span are 18 m,
defined point of boundary layer separation. 3.5 m and 25 m, respectively. The wing area will be
IV. The resultant wing loading with rounded leading
(a) 450 m2
edge will be steady.
(b) 87.5 m2
Choose the correct option among the following:
(c) 225 m2
(a) I and III (d) 268.75 m2
(b) I and IV
384. If the velocity of the flow is given by
(c) III and IV
v = c (x + y) î−c (x + y) ĵ, the vorticity will be
(d) II, III, and IV
(a) 0
379. For the aircraft which are operational, consider the fol-
(b) −c
lowing statements:
(c) c
I. Most of the cargo and transport aircraft use low aspect
(d) −2c
ratio wings.
II. Most of the cargo and transport aircraft use high 385. The elementary circulation at the midpoint of a flat plate
aspect ratio wings. kept at 2o to a freestream of speed 30 ms−1 will be
III. Fuel economy of low aspect ratio wings is higher approximately equal to
than high aspect ratio wings.
IV. Fuel economy of low aspect ratio wings is lower (a) 2.1
than high aspect ratio wings. (b) 1.1
Choose the correct option among the following: (c) 0
(d) 8
(a) I and III
(b) II and III 386. If the circulation distribution around a wing is given by
y 2
(c) II and IV
k = 16 1 − 10 . The downwash will be
(d) I and IV
(a) 0.2 ms−1 (c) same

(b) 0.4 ms−1 (d) unpredictable
(c) 1 ms−1
(d) 1.2 ms−1 393. When the camber of an airfoil is increased then the
induced lift will
387. A glider of aspect ratio 6 has a drag polar of,
CD = 0.02 + 0.06C2L . For minimum glide angle, the (a) increase
lift coefficient CL will be approximately equal to (b) decrease
(c) remains Same
(a) 1 (d) cannot say
(b) 0.6
(c) 1.7 394. In an isothermal atmosphere, the pressure
(d) data insufficient
(a) decreases linearly with elevation.
388. The aerodynamic efficiency of a sail plane of weight (b) increases exponentially with elevation.
3600 N andwing area 10 m2 is 30. If it is in level flight (c) varies in the same as density
at sea level ρ = 1.2kgm−3 with a speed of 36 kmh−1 , (d) remains constant
the drag coefficient will be
(a) 0.4 I. Viscosity of liquids increases with increase of tem-
(b) 0.8 perature.
(c) 0.2 II. Viscosity of gases decreases with increase of tem-
(d) 0.9 perature.
III. Viscosity of liquids increases with increase of pres-
389. Which of the following visualization technique cannot sure.
be used in supersonic flow? Which of the above statements is/are correct?
(a) schlieren (a) I

(b) shadowgraph (b) II
(c) tuft (c) III
(d) interferometry (d) all
390. The Bernoulli’s equation can be applied between any 396. In flow fields with rapidly varying density gradient,
two points located on two different streamlines, only if which of the following visualization technique is best
the flow is suited?
(a) viscous (a) schlieren

(b) supersonic (b) shadowgraph
(c) steady (c) interferometry
(d) irrotational (d) smoke
391. The Bernoulli’s equation is valid for which of the fol- 397. Consider the following statements:
lowing Mach number range? I. Fighter bombers use turboprop engine.
II. Adding ammonia and water vapor in the compressor
(a) 0≤M≤∞ decreases the effective power output of turbine.
(b) 0.8 ≤ M ≤ 1.2 III. The jet velocity in turbofan engine is less than that
(c) 0 ≤ M ≤ 0.5 in turbojet engine.
(d) 0 ≤ M ≤ 0.3 Which of the above is/are correct?
392. The stall angle in the symmetrical airfoil is than the (a) I
cambered airfoil. (b) I and II
(c) II and III
(a) higher (d) III
(b) lower
Normal shock
M 1= 2.5 M 1= 0.5
p = 100 kpa p = 690 kpa

2 2
Fig. E.12 Flow across normal shock wave
p , T , v0
0 0
ρ
0
v
Fig. E.13 Air discharge from a large reservoir
398. Choose the correct one among the following statements. (c) 0.85
(d) 5
(a) The pressure and temperature of air at the suction
of compressor are slightly more than atmospheric. 401. If the ratio of back pressure (pb ) to reservoir stagnation
(b) The pressure and temperature of air at the suction pressure (p0 ) is less than 0.528, then the flow exiting a
of compressor are slightly less than atmospheric. convergent nozzle will be
(c) The pressure is slightly more than atmospheric but
the temperature is slightly less than atmospheric. (a) underexpanded
(d) The pressure is slightly less than atmospheric but (b) overexpanded
the temperature is slightly more than atmospheric. (c) supersonic
(d) subsonic
399. Choose the INCORRECT option among the following:
402. Consider the high pressure air discharge coming out
(a) In an elliptical orbit, the velocity of a satellite at of a small opening from a large reservoir with given
apogee is more than that at perigee. stagnation conditions as shown in Fig. E.13.
(b) The velocity of a satellite in a circular orbit does not
depend on the altitude of the satellite from Earth’s The error involved in treating the air as an incompress-
surface. ible medium is
(c) Eccentricity of an elliptical orbit determines the
shape of the ellipse. (a) 20
(d) Geosynchronous satellite has same orbital time (b) 60
period as that of Earth. (c) 90
(d) data insufficient
400. Consider the flow across a normal shock with given
flow conditions as shown in Fig. E.12. 403. Consider a hypersonic flow past a wedge (vertex angle,
2θ), which of the following condition holds?
The strength of shock will be
(a) Mθ ≥ 1
(a) 6.9 (b) Mθ ≥ 5
(b) 5.9
(c) Mθ ≤ 1 increase of height.

(d) Mθ ≤ 1 III. In mesosphere, the temperature decreases with
increase of height.
404. A supersonic flow encounters a normal shock wave and IV. In thermosphere, the temperature increases with
brings down to a subsonic flow. Keeping the flow condi- increase of height.
tions same as of previous case the flow is reversed across Which of the above is/are correct?
the shock (i.e., from subsonic to supersonic). Which of
the following remains valid? (a) I
I. state equation (b) II and III
II. continuity equation (c) I, III, and IV
III. first law of thermodynamics (d) all are correct
IV. increase of entropy principle
Choose the correct option among the following: 409. The Ornithopter is
(a) I, III, and IV (a) fixed wing aircraft.

(b) I, II, and III (b) rotary wing aircraft.
(c) I, II, and IV (c) flapping wing aircraft.
(d) all the above (d) wingless aircraft.
405. Consider the following: 410. If the Young’s modulus, the shear modulus, and the
I. Creep bulk modulus have positive values, then the range of
II. Elongation Poisson’s ratio (ν) of a stable, isotropic, linear elastic
III. Tension material will be
IV. torsion
Which of the above are typical loads acting over an (a) −1 < ν < 0.5
aircraft? (b) −1.5 < ν < 1
(c) 0.5 < ν < 1
(a) I and II (d) 0 < ν < 1.5
(b) I and III
(c) III and IV 411. In two-dimensional state of stress and strain, it is given
(d) all the above that σx − σy = 2τxy . The principle angle will be
406. For inspection and maintenance purposes, which of the (a) 0o

following entire section of an aircraft is called empen- (b) 22.5o
nage? (c) 45o
(d) 90o
(a) entire wing section
(b) entire cockpit section 412. Generally, the aircraft instrument panels are shock
(c) entire landing gear section mounted to absorb
(d) entire tail section
(a) low frequency and high amplitude shocks.
407. Which of the following is used for the same purpose as (b) low frequency and low amplitude shocks.
hydraulic actuators? (c) high frequency and high amplitude shocks.
(d) all the vibrations.
(a) flaps
(b) ailerons 413. During flight, sometimes the altitude is measured by a
(c) trim tabs radar altimeter. It measures altitude by
(d) spring tabs
(a) transmitting a signal to the ground and receiving
408. Consider the following statements about standard back the reflected signal.
atmosphere: (b) means of transponder interrogation.
I. In troposphere, the temperature decreases with (c) receiving signals transmitted from ground radar sta-
increase of height. tions.
II. In stratosphere, the temperature increases with (d) receiving signals from the satellite.
414. An aircraft is flying at an altitude where the ambient (c) 4

temperature is found to be 200 K. If the stagnation tem- (d) 0.8
perature on the surface of the aircraft is 360 K, then the
aircraft is flying at Mach 420. At supersonic Mach numbers the drag coefficient varies
approximately as
(a) 1.5
(b) 2 (a) CD = 1
(M∞2 −1)
(c) 2.5 (b) CD = √ 12

(d) 3 (M∞ −1)
(c) CD = (M∞2 − 1)
415. Consider the flow past a flat plate. The pressure inside (d) CD = 2 1 3/2
(M∞ −1)
the boundary layer will be
(a) less than that of outside inviscid flow. I. The lift coefficient for a finite wing is more than that
(b) more than that of outside inviscid flow. for its airfoil section.
(c) same as of outside inviscid flow. II. The drag coefficient for a finite wing is greater than
(d) cannot be predicted. that for its airfoil section.
Which of the above is/are TRUE?
416. The normal stresses σx and σy for the Airy’s stress func-
tion φ = Ay2 + Bxy are , respectively. (a) I
(b) II
(a) 2A, 0 (c) all
(b) 0, 2A (d) none
(c) 0, −2A
(d) −2A, 0 422. For longitudinal static stability of an aircraft, the posi-
tion of center of gravity must be
417. Moon revolves around the Earth in an elliptical orbit.
The velocity of moon at perigee is (a) coinciding with the neutral point
(b) behind the neutral point
(a) same as the apogee. (c) forward of the neutral point
(b) higher than the apogee. (d) close enough to the nose of fuselage
(c) lower than the apogee.
(d) depends on the center of mass of the system. 423. The ratio of thrust-specific fuel consumption (TSFC) of
a typical turbojet to a typical turbofan is approximately
418. In continuity and momentum equations the air is equal to
replaced with water as fluid medium. The com-
plete analogy between these two sets of equations (a) 0.06
corresponds to (b) 0.7
(c) 1
(a) γ = 1.33 (d) 1.67
(b) γ = 1.4
(c) γ = 1.67 424. For a cantilever with point load at the free end, the
(d) γ =2 bending moment diagram will be
419. The drag measurement of turbulence sphere in sub- (a) parabola

sonic wind tunnel testing, the effective and measured (b) triangle with maximum height under free end
Reynolds numbers are found to be 400,000 and 80,000, (c) triangle with maximum height under fixed end
respectively. The turbulence factor will be (d) an arc
(a) 5 425. A beam subjected to the loading is shown in Fig. E.14.

(b) 0.2 The shear force diagram will look like
5 kN 12 kN (b) At smaller sizes, the axial blading at the rear of the

(a)
compressor is easy to manufacture accurately.
(c) A single-stage axial compressor increases the pres-
−6 kN sure higher than single-stage, centrifugal compres-
−7 kN sor.
6 kN
7 kN (d) It is easy to link them to obtain multistage compres-
(b) sor.
429. Which of the following statements concerning centrifu-

−12 kN
−5 kN gal compressors is/are TRUE?
6 kN 6 kN
(c)
(a) Centrifugal compressors with vane-less diffusers
are compact compared to vaned diffusers.
(b) In multistage centrifugal compressors, the width of
the blades reduces progressively in the direction of
−12kN
flow.
−7 kN (c) In multistage centrifugal compressors, the width of
(d) the blades increases progressively in the direction
of flow.
(d) The multistage centrifugal compressors are com-
monly used for high refrigerant capacity applica-
−6 kN −12 kN
tions.
426. Consider the following statements about centrifugal
pumps: 430. If is the Mach number and is the characteristic Mach
I. high suction pressure number then which of the following is TRUE?
II. low delivery pressure ∗2
III. reduction of flow rate at pump section M (γ−1)
(a) M = 2+M ∗2
Which of the above helps in avoiding cavitation in ∗2 (γ−1)
M (γ−1)
pumps? (b) M = 2+M∗2 (γ+1)
∗2
M (γ+1)
(c) M = 2+M ∗2
(a) I 2 (γ−1)
(b) II and III (d) M∗ = 2+M(γ−1)
M
2 (γ−1)
(c) III
(d) All 431. During the flight in troposphere layer of standard atmo-
sphere, when an aircraft goes to higher altitudes main-
427. The overall efficiency of a centrifugal pump is defined taining same speed its Mach number will be
as the ratio of
(a) increasing
(a) energy available at the impeller to the energy sup- (b) decreasing
plied to the pump by the prime mover. (c) same
(b) actual work done by the pump to the energy sup- (d) cannot say
plied to the pump by the prime mover.
(c) energy supplied to the pump to the energy available 432. When the flowing air is heated in a nozzle, which of the
at the impeller. following changes will NOT occur?
(d) manometric head to the energy supplied by the
impeller per Newton of water. (a) Velocity of air will decrease.
(b) Pressure increases.
428. The axial compressors are preferred over centrifugal (c) Increase in enthalpy.
compressors in gas turbine engines because (d) Increase in entropy.
(a) Because of the axial flow direction, it has high 433. Consider the following statements about propellants:
cross-sectional area than the corresponding cen- I. hydrogen peroxide
trifugal compressor. II. hydrazine
III. nitroglycerin
2 kN/m
6 kN 12 kN
A D
B C
1m 6m 1m
Fig. E.14 A beam subjected to the loading
C
A
TSFC
Fig. E.15 Plot between TSFC and the Mach number
IV. nitromethane isotropic material, the Bulk modulus of elasticity will

Which of the above is/are the example of monopropel- be
lant?
(a) 44 GPa
(a) I (b) 53 GPa
(b) II (c) 89 GPa
(c) III and IV (d) 133 GPa
(d) All
437. Consider the following graph (Fig. E.15), plotted
434. Dive flaps are used in the aircraft because between thrust-specific fuel consumption (TSFC) and
Mach number.
(a) It provides extra lift to the aircraft.
(b) It controls the flow separation over wings. Which of the above curves represent ideal ramjet
(c) It reduces the speed of aircraft. engine?
(d) It is used for longitudinal stability of aircraft.
(a) A
435. The hot-wire anemometer (HWA) is a device used to (b) B
measure flow speed. Which of the following is NOT a (c) C
type of HWA? (d) D
(a) constant resistance anemometer 438. Consider a 4-seater Cessna-172 propeller aircraft, with
(b) constant voltage anemometer propeller of 2 m in diameter spinning at 1200 RPM. If
(c) constant temperature anemometer the aircraft is advancing at a speed of 20 ms−1 , then the
(d) pulse width modulation anemometer advance ratio will be
436. For steel the Young’s modulus of elasticity is 200 GPa 1

(a) 120
and Shear modulus is 80 GPa. Assuming steel to be an (b) 1
30
(c) 21 444. Consider the following statements about Hohmann

(d) 2 transfer orbits:
I. These trajectories consume the maximum amount of
439. Consider a rocket engine burning hydrogen and oxygen propellant.
mixture at a rate of 125 kgs−1 and generating the thrust II. The orbit to Mars is achieved by accelerating the
of 4.75 × 105 N. The area of rocket nozzle throat is 0.1 spacecraft in the direction of Earth’s revolution to
m2 . The area of the exit is designed so that the exit pres- place in the Sun’s orbit with a perihelion equal to the
sure exactly equals ambient pressure at flying altitude orbit of Mars.
of 30 km. For the gas mixture, assume that γ = 1.4 and III. The orbit to Venus is achieved by accelerating
the molecular weight M = 16. At standard altitude of the spacecraft in the direction opposite of Earth’s
30 km, the specific impulse will be revolution to place in the Sun’s orbit with an aphelion
equal to the orbit of Venus.
(a) data insufficient Which of the above statements is/are TRUE?
(b) 388 s
(c) 475 s (a) I
(d) 125 s (b) II and III
(c) I, II and III
440. A body traveling in a hyperbolic orbit around a planet (d) none
will have
445. Assume that a planet is revolving around the Sun in an
(a) will always be at finite distance from planet. elliptical orbit, with a speed of 25 kms−1 at its average
(b) zero speed at infinite distance from planet. distance from the Sun. What could be the most likely
(c) infinite speed at infinite distance from planet. speed of the planet when it is closer to the Sun?
(d) finite speed at infinite distance from planet.
(a) 15 kms−1
441. Consider the following x−momentum Eulerian equa- (b) 20 kms−1
tion, (c) 25 kms−1
(d) 30 kms−1
∂u ∂u ∂u ∂u 1 ∂p
+u +v +w =−
∂t ∂x ∂y ∂z ρ ∂x 446. Which of the following should be applied to limit the
wing bending?
The above equation is valid if and only if the flow is
(a) High strength material as skin material of the wing.
(a) laminar
(b) High strength material as web plate material for the
(b) inviscid
spars.
(c) steady
(c) High stiffness material as skin material of the wing.
(d) unsteady
(d) High stiffness material as web material for the ribs.
442. The necessary and sufficient condition to define the
447. A total temperature probe measures the temperature of
potential function for a flow is:
a Mach 2 air flow as 520 K. If the probe has a recovery
(a) laminar factor of 0.9, then the stream static temperature will be
(b) incompressible (approximately)?
(c) steady
(d) irrotational (a) 520 K
(b) 350 K
443. Consider the flow past a flat plate as shown in Fig. E.16. (c) 300 K
At the point of separation which of the following con- (d) 250 K
dition is TRUE?
448. The oblique shock diffuser is more frequently used
∂u than the normal shock one in a supersonic wind tun-
(a) ∂x
∂u nel, because it
(b) ∂y
∂2u
(c) ∂x2 (a) rapidly accelerates the flow.
∂2u
(d) ∂y2 (b) increases total pressure loss.
y
Ua
Fig. E.16 Flow past a flat plate
(c) reduces the flow speed more rapidly. 454. Consider an aircraft which has the lift-to-drag ratio of
(d) reduces total pressure loss. 7. In PHUGOID mode, the damping ratio will be?
449. Consider a turbojet engine, which is operating with (a) 0.01

afterburner in switched-off mode. When the afterburner (b) 0.04
starts functioning then? (c) 0.1
(d) 0.5
(a) Both thrust and specific fuel consumption increase.
(b) The thrust increases but specific fuel consumption 455. In a turbojet engine nozzle, the temperature and Mach
decreases. number at a particular cross section are T and M, respec-
(c) The thrust decreases but specific fuel consumption tively. If T∗ is the temperature at the throat then
increases.
(d) Both thrust and specific fuel consumption decrease. T∗
= a + bM2
T
450. Consider a rocket engine with liquid propellant. The
exhaust velocity (ve ) will vary with molecular weight The constants a and b are , respectively.
(M) of the propellant as 1 2
(a) (γ+1) and (γ+1)
2 (γ−1)
(a) M10.5 (b) (γ−1) and (γ+1)
2 (γ−1)
(b) M11.5 (c) (γ+1) and (γ+1)
(c) M1.5 2 (1−γ)
(d) (γ+1) and (γ+1)
(d) M12
456. The volumetric strain experienced by a body with coef-
451. To achieve minimum time for orbital rendezvous by ficient of elasticity E and Poisson ratio ν = 0.3, when
Hohmann transfer, the value of phase error (θ) is . subjected
to auniform hydrostatic pressure of magni-
E×10−4
tude 3 is
(a) 180o
(b) 360o (a) −20 × 10−6
(c) 0o (b) −40 × 10−6
(d) 90o (c) −25.5 × 10−6
(d) −30 × 10−6
452. In which of the following parts of a turbojet engine the
maximum pressure loss will occur? 457. A spring–mass system is shown in Fig. E.17. below.
The natural frequency of the system is
(a) turbine
(b) compressor
3k
(c) inlet (a) 5m

(d) nozzle (b) 3m5k

8k
453. The damping ratio (ξ) for a critically damped system (c) 5m

will be? 5k
(d) 8m
(a) 0
458. Consider a rectangular beam. What is the nature of dis-
(b) >1
tribution of shear stress?
(c) −2
(d) 1
k m k
k
2k
(a) linear (c) 8.75 ms−1

(b) parabolic (d) 140 ms−1
(c) hyperbolic
(d) ellipse 463. Consider an aircraft is flying as shown in Fig. E.18.
The angle of attack is
459. If the diameter of a long column is reduced by 20%,
then the percentage reduction in Euler buckling load is (a) −A
(b) −B
(a) 4 (c) +B
(b) 40 (d) A+B
(c) 50
(d) 60 464. Which of the following phenomena, occurs because of
vortices?
I. An aircraft with a propeller mounted on the nose of (a) parasite drag
the fuselage will contribute negative to the longitudinal (b) skin friction drag
stability. (c) wave drag
II. For static stability of an aircraft: Cmα > 0. (d) lift induced drag
III. If the elevator size is decreased, keeping the tail area
same. The static longitudinal stability will increase. 465. The turbofan engine uses bypass
Which of the above is/are TRUE?
(a) to increase the mass flow rate through the exhaust
(a) I nozzle.
(b) II (b) to increase the inlet mass flow rate.
(c) II and III (c) to increase turbine inlet temperature.
(d) III (d) to increase the overall pressure ratio in the com-
pressor.
461. Consider the following statements about ground roll
distance during landing of an aircraft: 466. A thin cylinder of inner radius 250 mm and the thickness
I. Ground roll distance can be reduced by decreasing 10 mm is subjected to an internal pressure of 8 MPa.
wing loading and by increasing the lift coefficient. The maximum shear stress in the cylinder is
II. Ground roll distance will be reduced by increasing
thrust-to-weight ratio. (a) 25 MPa
Which of the above is/are TRUE? (b) 50 MPa
(c) 100 MPa
(a) I (d) 200 MPa
(b) II
(c) I and II 467. A two-dimensional flow field is given by φ = 3xy. The
(d) none stream function is represented by

462. An aircraft has a stall velocity of 35 ms−1 in the cruise (a) −3 x2 − y2
2
flight. During a maneuver, the load factor is found to be (b) −3
2 x + y
2
16. For this maneuver, the stall velocity will be? (c) −3
2 x − y
2 2
(d) −3 x + y2
2
(a) 35 ms−1
(b) 560 ms−1
horizon
B Chord line
Fig. E.18 A schematic diagram of a typical flying aircraft
468. In a convergent–divergent nozzle, generally a normal 472. Consider the following statements about layers in the
shock can occur Earth’s atmosphere:
(I) The troposphere is wider at the equator but narrower
(a) in the convergent portion. at the poles.
(b) in the divergent portion and throat. (II) The troposphere is narrower at the equator but wider
(c) near the inlet. at the poles.
(d) anywhere in the nozzle. (III) The ozone layer is predominantly located in the
lower segment of the stratosphere.
469. The efficiency of a centrifugal pump will be maximum (IV) The ozone layer is predominantly located in the
when it’s blades are upper segment of the stratosphere.
Choose the CORRECT options in the following:
(a) wave shaped
(b) bent forward (a) I and III
(c) bent backward (b) II and III
(d) straight (c) II and IV
(d) I and IV
470. For an orbit of a celestial body, the
ratio of radius of
apogee (ra ) to radius of perigee rp is 4. Then the eccen- 473. Consider the velocity field given by u = y
and
(x2 +y2 )
tricity of the elliptical orbit will be? −x
v= .
The equation of streamlines passing
(x2 +y2 )
(a) 1.67 through the point (0, 5) will be given by
(b) 0.6
(c) 0.78 (a) x+y=5
(d) 0.36 (b) x2 + y2 = 25
(c) xy = 25
471. Consider the following statements about Earth’s atmo- (d) x2 − y2 = 25
sphere:
(I) Longitude and latitude of the location on the Earth. 474. For the flow in a convergent–divergent nozzle, which
(II) Altitude above the sea level. of the following statement is CORRECT?
(III) Season in the year.
(a) The velocity is minimum at the throat.
(IV) Time in a day.
(b) The pressure is maximum at the throat.
(V) Sun-spot activities.
(c) The pressure is minimum at the throat.
The temperature and pressure in the Earth’s atmosphere
(d) When the throat is choked, the throat pressure
depends on
should be higher than the exit pressure to obtain
(a) I, II, and III subsonic flow in the divergent section.
(b) I, III, and IV
475. Consider an airfoil kept in a flow with a freestream
(c) II, III, and IV
velocity of 20 ms−1 . The velocity at a given point on
(d) all
the airfoil is 40 ms−1 . The pressure coefficient at this
point will be
(a) 2 481. Consider the following statements about Hohmann

(b) −3 Transfer:
(c) 3 (I) It is a two-impulse elliptical transfer between two
(d) − 13 coplanar circular orbits.
(II) The transfer itself consists of an elliptical orbit
476. For a doublet flow, the stream function in cylindrical with a perigee at the outer orbit and an apogee at the
coordinates will be given by inner orbit.
−κ sin θ
(III) The fundamental assumption behind it is that
(a) 2π r there is only one body which exerts a gravitational
−κ cos θ
(b) 2π r force on the body of interest, such as a satellite.
κ sin θ
(c) 2π r 2 (IV) Additional bodies can share the orbit which could
κ cos θ induce a gravitational attraction on the body of interest.
(d) 2π r 2
477. The purpose of fins on a rocket is to
(a) II and III
(a) reduce drag (b) I and III
(b) generate lift (c) II, III, and IV
(c) provide stability (d) I, II, and IV
(d) streamline shape
482. The force exerted by a jet impinging on a fixed plate
478. What is the main purpose of wing flaps? inclined at an angle θ with the jet is
(a) Decrease the angle of descent without increasing (a) (ρAv sin
4
2θ)
the airspeed. (b) ρAv

sin θ
(b) Increase the angle of descent without increasing the ρAv2 sin 2θ
(c) 2
airspeed. (d) ρAv2 sin 2θ
(c) Decreases the angle of descent by increasing the
airspeed. 483. Multistage centrifugal pumps are used to
(d) Decrease the drag.
(a) give high discharge
479. Consider the following statements about ground effect: (b) produce high heads
(I) It is the result of the interference of the surface of (c) pump viscous fluids
the Earth with the airflow patterns about an airplane. (d) all the above
(II) It decreases the induced drag, and therefore, any
excess speed at the point of flare may cause considerable 484. Consider the velocity field −
→
v = xî
floating.
(III) Becoming airborne before reaching recommended (a) For an incompressible fluid this field satisfies con-
takeoff speed. servation of mass.
Which of the above is/are CORRECT? (b) The acceleration of the particle in this field
decreases with x.
(a) I (c) The acceleration of the particle in this field
(b) I and II increases with x.
(c) II and III (d) The pressure of an incompressible and inviscid fluid
(d) all increase with x.
480. For an elliptical lift distribution over wing span, the 485. What is the primary control surface located on the wings
downwash that control the roll of the glider clockwise or counter-
clockwise?
(a) increases with angle of attack.
(b) decreases with angle of attack. (a) stabilizer
(c) is a constant. (b) rudder
(d) is equal to zero. (c) elevator
(d) aileron
CL
Plane 1
Plane 2
Fig. E.19 CL versus α curve
486. With the increase of camber of an airfoil, the induced (a) Plane 1.
lift will? (b) Plane 2.
(c) Both have same aspect ratio.
(a) increase. (d) Can’t be determined.
(b) decrease.
(c) no effect on induced lift. 491. The plot between powers required and true airspeed an
(d) no relationship between the camber and induced aircraft at two different altitudes is shown in Fig. E.20.
lift.
What is the correct order of altitude?
487. The aspect ratio shows the relationship between the
span and chord of a wing. Which of the following wing (a) H1 = H2
would create the maximum lift? (b) H1 < H2
(c) H1 > H2
(a) span = 10, chord = 5 (d) insufficient data
(b) span = 10, chord = 2
(c) span = 15, chord = 5 492. If the airfoil thickness increases, the critical Mach num-
(d) span = 15, chord = 2 ber will
488. The induced drag for airfoil is (a) Decrease.

(b) Increase.
(a) infinite (c) Remains constant.
(b) zero (d) Can’t be determined.
(c) half of finite wing.
(d) depends upon airfoil. 493. At half chord length from the leading edge, which of
the following airfoil will have location of the maximum
489. Consider the following combinations of area (S) and camber?
velocity
(v):

(I) √ , S
v
(a) NACA 5212
2 (b) NACA 2512
(II) v2 , 2S
(III) (v, (c) NACA 1225
2S) (d) NACA 2215
(IV) 2v, S2
For a certain angle of attack, at a given altitude the lift
494. A structural member supports loads, which produce at
will be doubled for which of the above combinations?
a particular point, a state of pure shear stress of 50
(a) I, II, and IV Nmm−2 . At what angles are the principal planes ori-
(b) II and III ented with respect to the plane of pure shear?
(c) I, III, and IV π
(a) 6 and 2π3
(d) II, III, and IV π
(b) 4 and 3π4
π
490. Consider the CL versus α curve as shown in Fig. E.19. (c) 4 and π2
π
Which plane has higher aspect ratio? (d) 2 and π
H1
Power required
H2
True air speed
Fig. E.20 Power required versus true airspeeds
495. Consider the following statements: (d) Heat and work are path functions.
(I) The combustion in gas turbine engines is an ideal
isochoric process. 499. Match the following criteria of material failure, under
(II) For a given chamber pressure, the thrust of a rocket biaxial stresses σ1 and σ2 and yield stress σy , with their
engine is highest when the rocket is operating at sea corresponding graphic representations (Fig. E.21).
level.
Which of the above statements is/are CORRECT? (a) P-M, Q-L, R-N
(b) P-N, Q-M, R-L
(a) I (c) P-M, Q-N, R-L
(b) II (d) P-N, Q-L, R-M
(c) I and II
(d) none 500. Consider Fig. E.22.
496. If the load factor of an aircraft turning at a constant If the surface is frictionless, the natural frequency will
altitude is 2, keeping the speed constant the required be
lift coefficient will be
(a) 32 Hz
(a) Same for turning as well as level flights. (b) 13 Hz
(b) Half for the turning flight as compared to level (c) 76 Hz
flight. (d) 51 Hz
(c) Double for the turning flight as compared to level
flight. 501. Consider the following statements:
(d) Four times for the turning flight as compared to (I) As compared to turboprop, the turbojet engine han-
level flight. dles low mass of air at high velocity.
(II) The modern fighter class engines are high-bypass
497. In a quasi-steady process, assuming the entropy turbofan engines.
increases for a substance. Then the rise in temperature Which of the above is/are CORRECT?
will be maximum for
(a) I
(a) process with constant enthalpy. (b) II
(b) isobaric process (c) I and II
(c) isothermal process (d) none
(d) isochoric process
502. Which of the following statements are NOT correct?
498. Which of the following statement is CORRECT?
(a) An airplane which has negative aerodynamic damp-
(a) Heat and work are intensive properties. ing will be dynamically unstable.
(b) Heat is intensive property but work is extensive (b) Forces and moments arising over the airplane due to
property. its motion provide negative aerodynamic damping.
(c) Heat is a point function and work is an extensive (c) Forces and moments arising over the airplane due to
property. its motion provide positive aerodynamic damping.
σ
2 σ2
σγ σγ
σ1 σ1
−σγ σγ −σγ σγ
−σ γ
−σγ
P. Maximum−normal−stress criterion Q. Maximum−distortion−energy criterion
σ
2
σγ
σ1
−σγ
σγ
−σγ
R. Maximum−shear−stress criterion
Fig. E.21 Schematic diagram of stresses and their graphical representation
1 Kg
Fig. E.22 Schematic diagram of spring–mass system
(d) Dynamic stability is usually specified by the time (a) flaps

it takes a disturbance to be damped to half of its (b) winglets
initial amplitude. (c) trim tabs
(d) high-lift devices
503. In order to have static longitudinal stability, through the
equilibrium point the aircraft pitching moment curve 505. A typical turbofan engines have the bypass ratio of
must have
(a) 2:1
(a) zero slope (b) 4:1
(b) negative slope (c) 8:1
(c) positive slope (d) 16:1
(d) can’t say
506. For the steady, fully developed flow inside a straight
504. The devices or modifications to the wing that increase pipe of diameter D, neglecting gravity effects, the pres-
the stall angle of attack are
K m
Fig. E.23 Schematic diagram of spring–mass system
sure drop p over a length L and the wall shear stress 510. Consider a disk of mass (m), which is attached to a
τw are related by spring of stiffness (k) as shown in the Fig. E.23. The
disk rolls without slipping on a horizontal surface. The
pD2
(a) τw = 4L2
natural frequency of vibration of the system will be
pD
(b) τw = 2L

pD 1 k
(c) τw = (a) 2π
4L m
4pD
(d) τw = L
1
(b) 2π 2k
m
1 2k
507. Consider the following statements for the turbulent flow (c) 2π
3m
of a fluid through a circular pipe of diameter (D): 1
(d) 2π 3k
2m
(I) The fluid is well mixed.
(II) The fluid is unmixed. 511. In an aircraft, constant roll rate can be produced using
(III) Reynolds number, ReD > 2300. ailerons by applying
(IV) Reynolds number, ReD < 2300.
Which of the above is/are TRUE? (a) a step input
(b) a ramp input
(a) I (c) a sinusoidal input
(b) I and III (d) an impulse input
(c) II and III
(d) I and IV 512. The Shadowgraph optical flow visualization technique
depends on the
508. Consider a simply supported beam of length, 50h, with a
rectangular cross section of depth h and width 2h. The (a) First derivative of density with respect to spatial
beam carries a vertical point load P, at its midpoint. coordinate.
Ratio of the maximum shear stress to the maximum (b) Second derivative of density with respect to spatial
bending stress in the beam will be coordinate.
(c) Third derivative of density with respect to spatial
(a) 0.02 coordinate.
(b) 0.1 (d) Fourth derivative of density with respect to spatial
(c) 0.05 coordinate.
(d) 0.01
513. A rocket is to be launched from the bottom of a very
509. The damping ratio of a single degree of freedom spring– deep crater on Mars for Earth return. The specific
mass–damper system with mass of 2 kg, stiffness 200 impulse of the rocket, measured in seconds, is to be
Nm−1 and viscous damping coefficient of 40 m−1 Ns is normalized by the acceleration due to gravity at
(a) 0.5 (a) The bottom of the crater on Mars.
(b) 1.0 (b) Mars standard sea level.
(c) 1.25 (c) Earth’s standard sea level.
(d) 2.0 (d) The same depth of the crater on Earth.
514. Which of the following is the CORRECT combination (a) I

of greenhouse gases? (b) I and II
(c) III and IV
(a) water vapor, oxygen, methane, nitrous oxide, and
(d) all
ozone.
(b) water vapor, carbon dioxide, methane, nitrous 520. The cast iron has the Poisson’s ratio (ν) in the range of
oxide, and ozone.
(c) water vapor, carbon dioxide, hydrogen, nitrous (a) 0.1 < ν < 0.2
oxide, and ozone. (b) 0.23 < ν < 0.27
(d) water vapor, carbon dioxide, methane, sulfur diox- (c) 0.25 < ν < 0.33
ide, and ozone. (d) 0.4 < ν < 0.6
515. The minimum period that any free flight object can 521. Consider a cross-sectional area over which the velocity
have in orbit around the Earth (also known as Schuler is zero in one-half and uniform over the rest half. The
period) is momentum correction factor will be
(a) 84.4 minutes
(b) 104.4 minutes (a) 1
(c) 60 minutes (b) 43
(d) 12 hours (c) 2
(d) 4
516. When the freestream velocity in a subsonic wind tunnel
test section is decreased by 2 times, the power require- 522. Consider a two-dimensional, steady, and incompress-
ments to run the tunnel will be ible flow over an airfoil. The freestream velocity suffi-
ciently far away from the airfoil is found to be 30 ms−1
(a) decreased by 2 times where the distances between streamlines are 2 cm. The
(b) decreased by 4 times velocity near the airfoil where the streamlines are 1.5
(c) decreased by 8 times cm apart, will be
(d) increased by 4 times
(a) 11.25 ms−1
517. If ratio of specific heats is γ, the Prandtl–Meyer super- (b) 22.5 ms−1
sonic expansion function (ν) can be written as (c) 33 ms−1
γ (d) 40 ms−1
(a) ν = γ−1
γ−1
(b) ν = γ+1 523. Which of the following statements are CORRECT for
γ−1
(c) ν = γ
turbulent flows?
γ+1
(d) ν = γ−1
(a) The eddy viscosity is a function of temperature
518. Assume that a planet is revolving around the Sun in an only.
elliptical orbit with eccentricity (e = 0.4). The ratio of (b) The eddy viscosity is a physical property of the
its velocities at perigee to apogee will be fluid.
(c) The eddy viscosity depends on the flow.
(a) 0.4 (d) The eddy viscosity is independent of the flow.
(b) 0.43
(c) 2.3 524. If the perturbation velocity is 2 ms−1 and freestream
(d) 2.5 velocity is 8 ms−1 . Using small perturbation theory
the pressure coefficient in two-dimensional planar flows
519. Consider the following statements: will be
(I) The volumetric change of the fluid caused by a resis-
tance is known as compressibility. (a) − 18
(II) The density of water is maximum at 4o C.
(b) − 41
(III) The bulk modulus of elasticity decreases with
increase in pressure. (c) −8
(IV) Viscosity of liquids is appreciably affected by (d) − 21
change in pressure.
Which of the above statements is/are TRUE?
525. Consider the following statements: (d) II and IV

(I) An orbit can be both in a Sun-synchronous orbit and
in a repeat orbit at the same time. 530. Consider the following statements:
(II) A geostationary orbit is geosynchronous and all (I) Airy stress function can be used only for two-
geosynchronous orbits are geostationary. dimensional problems.
Choose the CORRECT from following: (II) The duration of the load applied affects CREEP.
(III) In the constitutive equations of a generalized
(a) I anisotropic solid, the numbers of independent elastic
(b) II constants are 21.
(c) I and II Which of the above is/are CORRECT?
(d) none
(a) I only
526. An oblique shock wave with a wave angle of β = 60o (b) I and III
is generated from a wedge angle of θ = 30o . The ratio (c) II and III
of Mach number downstream of the shock to its normal (d) all
component will be
531. The roll stability of a glider can be improved if the wing
(a) √2 has mainly
3
(b) 0.87
(c) 0.5 (a) sweep angle
(d) 2 (b) anhedral
(c) dihedral
527. If an aircraft is in cruise motion at Mach 3, where the (d) winglets
outside air temperature is found to be 350 K. The stag-
nation temperature at the nose of the aircraft will be? 532. Consider a steady, level turning flight of an aircraft with
(For air, the specific heat ratio, γ = 1.4) the load factor (n = 3). The ratio of the horizontal com-
ponent of lift to the weight of aircraft will be
(a) 980 K √
(b) 1610 K (a) 2
(c) 350 K (b) √1
√
2
(d) insufficient data (c) 2 2
1
(d) √
2 2
528. Consider a thin-walled-closed and a thin-walled open
tubes with the radius, r = 10 mm and thickness, t = 1 533. Consider the fluid flow past a wooden wedge (semi-
mm in both cases. The ratio of torsional rigidity of thin- vertex angle, θ = 20o ) at Mach 10. The similarity
walled closed tube to thin-walled open tube will be? parameter for this flow will be?
(a) 100 10π

(a) 9
(b) 200 (b) 20π
9
(c) 300 (c) 90
π
(d) 400 45
(d) π
529. Consider the jet exhaust through an underexpanded noz-
534. If an open-loop unstable linear system is represented by
zle. 1
(I) normal shock wave (s−1)(s+2) . Its closed-loop characteristic equation will
be?
(II) expansion fans
(III) subsonic diffusion (a) s2 + s − 1 = 0
(IV) supersonic diffusion (b) s2 + 2s − 1 = 0
The pressure equalization takes place through which of (c) s2 + s + 1 = 0
the following combinations? (d) s2 + s − 2 = 0
(a) I and III 535. Consider the following statements about the control
(b) I and IV system:
(c) II and III
(I) The nature of bandwidth for a good control system 540. The classification of composite materials are based
should be small. upon
(II) The steady-state error is zero in closed-loop control
systems. (a) matrix type
(III) Gauss meter controls the speed of D.C. motor. (b) size and shape of reinforcement
(IV) A good control system should be sensitive to input (c) both (a) and (b)
signals (except noise). (d) none
541. Which of the following is not an example of laminar
(a) I and II composite?
(b) II and III
(c) II and IV (a) wood
(d) IV only (b) bimetallic
(c) coatings/paints
536. Which of the following statement is WRONG? (d) claddings
(a) In an open-loop control system the output is inde- 542. The oxidizing power is generally determined in terms
pendent of control input whereas in closed-loop of electronegativity. Which of the following substances
system, the control action is somehow dependent has the highest electronegativity?
on the output.
(a) hydrogen
(b) In a closed-loop control system the output is inde-
(b) fluorine
pendent of control input, whereas in an open-loop
(c) oxygen
system, the control action is somehow dependent
(d) chlorine
on the output.
(c) The positive value of feedback gain in a closed-loop 543. Consider the following beams:
control system will decrease the overall gain. (I) simply supported beams
(d) The closed-loop system has a tendency to oscillate. (II) cantilever beams
(III) overhanging beams
537. The satellite orbits are elliptical with a constantly vary-
(IV) fixed beams
ing radius. Since the satellite’s velocity depends on this
(V) continuous beams
varying radius, it changes as well. To resolve this prob-
Which of the above are statically determinate beams?
lem, an eccentric anomaly (E) is defined as: ( Take, ν
= True Anomaly) (a) I, II, and IV
e cos ν
(b) I, II, and III
(a) E= 1+e cos ν (c) II, III and IV
e−2 cos ν
(b) E= 1+e cos ν (d) III, IV and V
2e+cos ν
(c) E= 1−e cos ν
e+cos ν
(d) E= 1+e cos ν
544. Consider a cantilever beam with uniformly distributed
load starting from zero. The shear force diagram will be
538. Altimeter works on
(a) horizontal line parallel to x−axis
(a) differential pressure sensing. (b) line inclined to x−axis
(b) no air pressure sensing. (c) parabolic curve
(c) mono pressure sensing. (d) cubic curve
(d) radar sensing.
545. Consider an automobile axle with the loads as shown
539. The airspeed indicator (ASI) is the instrument that in Fig. E.24.
The maximum bending moment will be
(a) has both pitot and static ports.
(b) utilizes pitot port only. (a) Wl
(c) utilizes static port only. (b) W (l − a)
(d) does not operate on differential pressure sensing. (c) W (l + a)
(d) Wa
W W
a a
Fig. E.24 Schematic diagram of axle under pointed loads
5 kN 10 kN
A
B C
2m 2m
Fig. E.25 Schematic diagram of a beam under loads
546. Consider the beam with loads as shown in Fig. E.25. (c) III
(d) I and III
The slope of the bending moment diagram between B
and C will be 550. Consider the following statements:
(I) The conventional vertical tail of an aircraft con-
(a) 15 kN tributes toward longitudinal stability.
(b) 10 kN (II) The conventional vertical tail of an aircraft con-
(c) Zero tributes toward both lateral and directional stability.
(d) 20 kN (III) If the C.G. of an aircraft moves forward, the efforts
required trimming the aircraft will be increased.
547. The Bernoulli’s equation is valid for which of the fol- (IV) Keeping the tail area same, if the elevator size
lowing Mach number ranges? is decreased the static longitudinal stability will also
decrease.
(a) M ≤ 0.5 Which of the above statements is/are TRUE?
(b) 0.5 ≤ M ≤ 1.0
(c) M ≤ 0.3 (a) I
(d) 0.7 ≤ M ≤ 1.2 (b) II and III
(c) I and IV
548. As compared to symmetrical airfoil, the angle of stall (d) none
for a cambered airfoil is
551. The local skin friction coefficient for a compressible
(a) less fluid in laminar boundary layer will be
(b) more
f(Ma )
(c) same (a) Cfx = 1
(d) can’t say (Rex ) 5
(b) Cfx = √ a)
f(M
Rex
549. Consider the following statements: f(Ma )
(c) Cfx = 1
(I) Lift induced drag is caused by vortices. (Rex ) 7
(II) Aileron is primarily used for pitch. (d) independent of freestream Mach number
(III) Slots in flaps decrease the stall angle.
552. Which of the following statements about swept wings
in subsonic aircraft are INCORRECT?
(a) I
(a) Using swept wings the effective critical Mach num-
(b) II and III
ber is increased.
(b) By sweeping the wing, the drag divergence is 557. Consider a turbulent boundary layer over a flat plate.
delayed to higher Mach numbers. The approximate value of shape factor at which the
(c) Keeping all other parameters constant, the increase separation of boundary layer takes place, is
of the wing sweep reduces the lift coefficient.
(a) 1.4
(d) None.
(b) 2.4
553. Consider an airplane in level turn. (c) 0
(I) The highest possible load factor. (d) 3.5
(II) The lowest possible velocity.
558. Based on freestream velocity and momentum thick-
(III) The highest possible velocity.
To obtain both a small turn radius and a large turn rate, ness of a boundary layer, the typical value of critical
which of the above is/are TRUE? Reynolds number is
(a) 2300
(a) II (b) 5 × 105
(b) I and III (c) 350
(c) I and II (d) 200
(d) III
559.
If a two-dimensional
velocity field is given by −
→
v =
554. Consider the following statements for Blasius Bound- 3 2 2
2x yî−3x y ĵ
ary Layers:
(I) For laminar flows the typical value of shape factor (a) rotational
(H) is approximately 2.6. (b) incompressible
(II) For turbulent flows, the shape factor falls in the (c) irrotational
range of 3.3−3.4. (d) unsteady and compressible.
(III) The high value of shape factor reflects, weaker
adverse pressure gradient. 560. The circulation at the midpoint of a flat plate, at 6o to a
(IV) Higher adverse pressure gradient increases the freestream of speed 40 ms−1 , is
Reynolds number at which transition into turbulence
π
may occur. (a) 6
π
Which of the above statements is/are CORRECT? (b) 3
8π
(c) 3
(a) I (d) 4π
3
(b) I and II
(c) II, III, and IV 561. Consider a square ring vortex of side 2a. If each sides
(d) III and IV has the strength , the velocity induced at the center of
the ring is
555. In supersonic flows, which of the following waves can √
never be made isentropic? (a) 3 2
πa
√
(b) 2 2

(a) mach line √a
πa
2
(b) expansion wave (c)
√
a
2
(c) shock wave (d)
(d) none
562. The statement that “The airfoil generates sufficient cir-
556. If M1 and M2 are the upstream and downstream Mach culation to depress the rear stagnation point from its
numbers across a normal shock wave, then which of the position, in the absence of circulation downstream to
following is CORRECT? the sharp trailing edge” is known as
(I) Kutta condition
(a) (M2 )minimum = γ+1 (II) Joukowski postulation
γ−1 (III) Kutta- Joukowski theorem
(b) (M2 )minimum = γ+1 Which of the above statements is/are CORRECT?
2γ
(c) (M2 )minimum = γ+1 (a) I
2γ
(d) (M2 )minimum = γ−12γ
(b) I and II
(c) II and III (a) M≈2

(d) I and III (b) M ≈ 2.5
(c) M≈1
563. The semispan of a rectangular wing of planform area (d) M ≈ 1.5
8.4 m2 is 3.5 m. The aspect ratio of the wing is
569. Which of the following statements are CORRECT?
(a) 5.83 m2
(b) 11.66 m2 (a) In transonic flow, the density change is faster than
(c) 2.92 m2 the velocity change.
(d) 0.17 m2 (b) The density change in supersonic flow is slower
than the velocity change.
564. In comparison to combustion chamber, the temperature
(c) Mach number downstream of an oblique shock
rise in afterburner is
wave is always subsonic.
(a) low (d) In hot-wire anemometry, the hot-wire sensor is gen-
(b) equal erally made of Tungsten.
(c) high
(d) can’t say 570. In an aircraft wing if the incidence from root to tip is
decreased, it is known as
565. Consider the following statements about afterburners:
(I) Afterburners are not equipped with case and liners. (a) downwash
(II) Use of afterburners increases the efficiency. (b) washout
(III) Engines with afterburners consume lower amounts (c) slush
of fuel. (d) slosh
Which of the above statements is/are CORRECT?
571. A sailplane with a glide ratio of 12, flying 2400 m above
(a) I and II the ground. The greatest distance it can travel in still air
(b) II will be
(c) II and III
(d) none (a) 14,400 m
(b) 28,800 m
566. Mixture of liquid hydrogen and liquid oxygen may pro- (c) 200 m
duce the thrust up to (d) 100 m
(a) 1.5 kms−1 572. If the pitot tube becomes clogged, then which of the
(b) 4 kms−1 following parameters can’t be computed?
(c) 4.5 kms−1
(d) 5 Kms−1 (a) airspeed
(b) vertical speed
567. Consider the following statements about cryogenic
(c) altitude
rocket engines:
(d) outside air temperature
(I) Pyrotechnic initiators are used in cryogenic rocket
engines. 573. Which of the following atmospheric conditions will
(II) Cryogenic fuels are stored at room temperature lead to longer takeoff and lower rate of climb?
and pressure.
(III) Cryogenic rocket engines are also called as hybrid (a) high temperature, high relative humidity, and high
rocket engines. altitude.
Which of the above statements is/are CORRECT? (b) high temperature, low relative humidity, and low
(a) I and III altitude.
(b) I (c) low temperature, low relative humidity, and low
(c) II and III altitude.
(d) III (d) high temperature, low relative humidity, and high
altitude.
568. In centrifugal compressors, the flow sometimes leaves
the impeller at Mach number
574. For an airfoil, consider the following statements: (c) velocity variation in the flow field.
(I) Geometrical incidence is the angle between the chord (d) density gradient variation in the flow field.
of the profile and the direction of motion of the airfoil.
(II) Absolute incidence is the angle between the axis of 580. If the Reynolds number in a boundary layer flow
zero lift of the profile and the direction of motion of the decreases then
airfoil.
Which of the above statements is/are CORRECT? (a) Mach number increases.
(b) Pressure gradient normal to body surface decreases.
(a) I (c) Boundary layer thickness increases.
(b) II (d) Boundary layer thickness decreases.
(c) I and II
(d) none 581. For a supercritical airfoil, which of the following state-
ment is CORRECT?
575. For a given lift curve, decrease of aspect ratio increases
(a) Has higher wave drag.
(a) geometrical incidence only (b) Has higher critical Reynolds number.
(b) induced drag coefficient only (c) Greatly reduces shock-induced boundary layer sep-
(c) both geometrical incidence and induced drag coef- aration.
ficient (d) Has lower drag divergence Mach number.
(d) none
582. When a moving fluid is brought to rest adiabatically,
576. In a straight level flight, for a wing of elliptic loading, then which of the following statement is CORRECT?
the condition for minimum drag is
(a) Both stagnation pressure and stagnation tempera-
(a) CD0 = kC3L ture are conserved.
2 (b) Stagnation pressure is not conserved but stagnation
(b) CD0 = kCL3
3 temperature is conserved.
(c) CD0 = kCL2 (c) Stagnation pressure is conserved but stagnation
(d) CD0 = kC2L temperature is not conserved.
(d) Both stagnation pressure and stagnation tempera-
577. Air flows from a reservoir through a convergent– ture are not conserved.
divergent nozzle at low subsonic speed and is
exhausted into the atmosphere. A pitot tube is mounted 583. Which of the following statements are CORRECT
at the midsection of the nozzle and traversed along the about steady flow?
length of the nozzle from the exit to the reservoir end.
The pressure recorded by the pitot tube will (a) It occurs when pressure does not change along the
flow.
(a) increase during traverse. (b) It occurs when conditions do not change with time
(b) decrease during traverse. at any point.
(c) decrease up to the throat and then increase during (c) It occurs when velocity does not change.
traverse. (d) It occurs when conditions change gradually with
(d) remain constant during traverse. time.
578. For the same test-section speed and flow rate, the effi- 584. The mixing length model was first proposed by
ciency of a closed-circuit low-speed wind tunnel is
(a) Ludwig Prandtl
(a) greater than that of an open circuit wind tunnel. (b) Theodore von Karman
(b) less than that of an open circuit wind tunnel. (c) Albert Einstein
(c) equal to that of an open circuit wind tunnel. (d) Isaac Newton
(d) not comparable with that of an open circuit tunnel
because of design differences. 585. Consider the following statements about the boundary
layer flow:
579. The Schlieren technique works on the basis of (I) The flow field outside the boundary layer is rota-
tional.
(a) density variation in the flow field. (II) The pressure inside the boundary layer is equal to
(b) pressure variation in the flow field.
that of outside flow. (a) Both heat and work may cross the system boundary.
(III) The skin friction coefficient of laminar boundary (b) Both heat and work are path functions.
layer is more than the turbulent boundary layer. (c) Both heat and work are property of the system.
Which of the above statements is/are CORRECT? (d) Heat flows when the system and surrounding are
not in equilibrium which is not necessary for work.
(a) I and II
(b) II 591. A thin cylinder of inner radius 500 mm and thickness
(c) III 10 mm is subjected to an internal pressure of 5 MPa.
(d) all The average circumferential (hoop) stress in MPa is
586. Which of the following statements are CORRECT? (a) 100

(b) 250
(a) The jet velocity in turbofan engine is less than that (c) 500
of turbojet engine. (d) 1000
(b) Fighter bombers use turboprop engine.
(c) Adding ammonia and water vapor in compressor 592. The maximum gas flow rate that can be handled by
decreases the effective power output of turbine. a multistage axial compressors at a given rotational is
(d) Trim tabs are used for the same purpose as hydraulic dictated by
actuators.
(a) compressor surge
587. In centrifugal pumps, to obtain higher flow output the (b) rotating stall
impellers can be (c) choking
(d) optimum design pressure ratio
(a) connected in series
(b) connected in parallel 593. For a turbine stage, which one of the following losses
(c) connected either in series or in parallel occurs due to the turning of the wall boundary layer
(d) cannot be connected through an angle due to curved surface?
588. Consider the following statements: (a) profile loss

(I) Perfectly straight column and axial load apply. (b) annulus loss
(II) Length of column is large as compared to its cross- (c) tip clearance loss
sectional dimensions. (d) secondary flow loss
(III) The shortening of column due to direct compres-
sion is not neglected. 594. A column has a rectangular cross section of 10 mm ×
(IV) The failure of column occurs due to buckling alone. 20 mm and a length of 1 m. The slenderness ratio of the
Which of the above statements is/are taken as assump- column is close to
tion in Euler’s Column Theory?
(a) 200
(a) I, II, and III (b) 346
(b) I, II, and IV (c) 477
(c) II, III, and IV (d) 1000
(d) all
595. A streamline and an equipotential line in a flow field
589. Airy stress function satisfies which of the following
equation? (a) are parallel to each other.
(b) are perpendicular to each other.
(a) ∇2φ = 0 (c) intersect at an acute angle.
(b) ∇ 2 φ = g(x) (d) are identical.
(c) ∇ 3φ = 0
(d) ∇ 4φ = 0 596. Consider the following statements:
(I) Gases are considered incompressible when Mach
590. Which of the following is an INCORRECT statement? number is less than 0.2.
(II) A Newtonian fluid is incompressible and nonvis- (a) Potential function must satisfy the Laplace equa-
cous. tion, whereas stream function need not.
(III) An ideal fluid has negligible surface tension (b) Stream function must satisfy the Laplace Equation,
Which of these statements is/are correct? whereas potential function need not.
(c) Both stream function and potential function satisfy
(a) II and III the Laplace Equation.
(b) II (d) Neither the stream function nor the potential func-
(c) I tion need to satisfy the Laplace Equation.
(d) I and III
602. In thin airfoil theory, the lift curve slope dC
dα = 2π
L
597. Which one of the following statements is correct? Irro-
is valid for
tational flow is characterized as the one in which
(a) cambered airfoil.
(a) The fluid flows along a straight line.
(b) symmetric airfoil.
(b) The fluid does not rotate as it moves along.
(c) Joukowski airfoil.
(c) The net rotation of fluid particles about their mass
(d) any airfoil shape.
centers remains zero.
(d) The streamlines of flow are curved and closely 603. Which of the following stays constant for the flow
spaced. through a Prandtl–Meyer expansion wave?
598. In a two-dimensional incompressible steady flow, the (a) density
velocity component u = Aex is obtained. What is the (b) temperature
other velocity component, v? (c) mach Number
(d) entropy
(a) v = Aexy
(b) v = Aey 604. Downward deflection of the flap increases the lift coef-
(c) v = −Aex y + f (x) ficient of an airfoil by
(d) v = −Aey x + f (y)
(a) increasing the local airspeed near the trailing edge.
599. The continuity equation for a steady flow states that (b) increasing the effective camber of the airfoil.
(c) delaying the flow separation.
(a) Velocity field is continuous at all points in flow field.
(d) controlling the boundary layer growth.
(b) The velocity is tangential to the streamlines.
(c) The stream function exists for steady flows. 605. The maximum thickness-to-chord ratio for NACA
(d) The net efflux rate of mass through the control sur- 24012 airfoil is
faces is zero.
(a) 0.12
600. For a stream function to exist, which of the following (b) 0.24
conditions should hold? (c) 0.40
(I) The flow should always be irrotational. (d) 0.01
(II) Equation of continuity should be satisfied.
(III) The fluid should be incompressible. 606. The maximum possible value of the pressure coefficient
(IV) Equation of continuity and momentum should be in an incompressible flow is
satisfied.
Which of the following is/are CORRECT? (a) 0.25
(b) 0.5
(a) II (c) 0.75
(b) II and III (d) 1
(c) I, III, and IV
(d) all 607. An inviscid and irrotational flow becomes rotational on
passing through
601. For a two-dimensional incompressible and irrotational
flows (a) an oblique shock wave.
(b) a normal shock wave.
(c) a Mach wave. (c) 0

(d) a curved shock wave. (d) ∞
608. Laminar flow airfoil is used to reduce the 614. The geometrical features of a supercritical airfoil are
(a) pressure drag (a) sharp leading edge, flat suction surface, and no cam-
(b) induced drag ber at the rear.
(c) skin friction drag (b) rounded leading edge, curved suction surface, and
(d) wave drag no camber at the rear.
(c) sharp leading edge, curved suction surface, and high
609. Consider a steady inviscid flow in a convergent– camber at the rear.
divergent nozzle with a normal shock in the divergent (d) rounded leading edge, flat suction surface, and high
section. The static pressure downstream of the normal camber at the rear.
shock will
615. Which one of the following high-lift devices results in
(a) decrease isentropically to the static pressure at the higher stalling angle?
nozzle exit.
(b) increase isentropically to the static pressure at the (a) plain flap
nozzle exit. (b) fowler flap
(c) either increase or decrease depending on the mag- (c) split flap
nitude of the static pressure at the nozzle exit. (d) leading edge flap
(d) remain constant.
616. The total pressure at a point is defined as the pressure
610. If the Mach number in a turbulent boundary layer over a when the flow is brought to rest
flat plate is increased by keeping the Reynolds number
unchanged, then the skin friction coefficient will (a) isentropically
(b) adiabatically
(a) increase (c) isobarically
(b) decrease (d) isothermally
(c) remain constant
(d) becomes infinity 617. The drag divergence Mach number of an airfoil is
611. In supersonic wind tunnel design, an oblique shock dif- (a) always higher than the critical Mach number.
fuser is preferred over a normal shock diffuser because (b) equal to the critical Mach number at zero angle of
attack.
(a) it increases the total pressure loss. (c) a fixed value for a given airfoil.
(b) it reduces the total pressure loss. (d) the Mach number at which a shock wave first
(c) it rapidly accelerates the flow. appears on the airfoil.
(d) it rapidly decelerates the flow.
618. The Joukowski airfoil is studied because
612. The variation of downwash along the span of an
untwisted wing of elliptic planform is (a) it is used in many aircraft.
(b) it has a simple geometry.
(a) constant (c) it has the highest lift curve slope.
(b) parabolic (d) it is easily transformed into a circle, mathemati-
(c) sinusoidal cally.
(d) elliptic
619. Two airfoils of the same family are operating at the
613. The flow past an airfoil is modeled using the vortex same angle of attack. The dimensions of one airfoil are
sheet. The strength of vortex sheet at the trailing edge thrice as large as the other one. The ratio of the minimum
will be pressure coefficient of the larger airfoil to the minimum
pressure coefficient of the smaller airfoil is
(a) 0.5
(b) 1 (a) 1
(b) 3
(c) 5 (d) None of the above.

(d) 6
625. Which of the following is not true for a gouge flaps?
620. Wing A has a constant chord c and span b. Wing B is
identical to A, but has a span 4b. When both wings are (a) It increases the chord length.
operating at same geometric angle of attack at subsonic (b) It affects the trim.
speeds, then (c) It increases the wing area.
(d) None of the above.
(a) wing A produces a smaller lift coefficient than wing
B. 626. If the lift coefficient of a wing is increased by two times,
(b) wing A produces a greater lift coefficient than the then the induced drag becomes
wing B.
(c) wing A and B produce the same lift coefficients. (a) two times lower.
(d) freestream Mach number decides, which wing pro- (b) two times higher.
duces the greater lift coefficient. (c) three times higher.
(d) four times higher.
621. In comparison to a laminar boundary layer, the turbulent
boundary layer remains attached on the upper surface 627. The critical Mach number for a thick airfoil will be
of an airfoil over a longer distance, because
(a) lesser than a thin airfoil.
(a) the turbulent boundary layer is more energetic and (b) greater than a thin airfoil.
hence can overcome the adverse pressure gradient (c) equal to a thin airfoil.
better. (d) cannot be related to thin airfoil.
(b) the turbulent boundary layer is thicker, hence the
628. Consider the three-dimensional motion of fluid in the
velocity gradients in it are smaller, consequently
vicinity of a vortex filament. Which one of the following
viscous losses are less.
statement is not a Helmholtz’s theorem?
(c) the laminar boundary layer develops more skin fric-
tion and hence slows down more rapidly. (a) The strength of a vortex filament is constant along
(d) turbulence causes the effective coefficient of viscos- its length.
ity to reduce, resulting in lesser loss of momentum (b) A vortex filament cannot end in a fluid. It may
in the boundary layer. extend to the boundaries of the fluid.
(c) A vortex filament cannot form a closed path.
622. In a compressible flow over a flat plate, the boundary
(d) In the absence of rotational external forces, a fluid
layer thickness (δ) is given by
that is initially irrotational, remains irrotational
√5.0x
(a) Rex 629. What is the primary reason for an aircraft with delta
5.0x
(b) Rex wings having high stall angles?
5.0x2
(c) Rex
5.0x2 (a) The wing’s leading edge does not contact the shock
(d) √
Rex wave boundary formed at the nose of the fuselage
(b) The delta planform maximizes wing area with a
623. For a stream tube, the area–velocity relation is given by
very low wing per unit loading.
(c) Due to highly robust nature of the delta wings.
A = 1−M
(a) dA 2 dv
v
(d) Due to vortex formation at the leading edge which
(b) dA = (M − 1) dv
A 2 vdv energizes the flow.
A = M −1 v
(c) dA
A = (1 − M) v
(d) dA dv
630. An aircraft glides a distance of 5000 m, in which the
loss of altitude is 1000 m. If the freestream velocity is
624. An increase in the angle of attack has the following 100 ms−1 , the aircraft’s glide ratio is
effects.
(a) 0.5
(a) The center of pressure moves backward. (b) 5
(b) The center of pressure moves forward. (c) 50
(c) Both (a) and (b). (d) 500
631. The thrust developed in the jet engine is due to (a) 0.25
(b) 0.5
(a) balanced force (c) 0.75
(b) unbalanced force (d) 1
(c) both (a) and (b)
(d) none of the above 639. In an engine, the maximum pressure loss occurs across
the
632. The propulsive efficiency of the aircraft’s jet engine is
given by (a) inlet
(b) compressor
(a) energy input rate/propulsive power (c) turbine
(b) propulsive power/energy input rate (d) nozzle
(c) propulsive power/work done by engine
(d) work done by engine/propulsive power 640. Specific impulse will be maximum for
633. Flapper valves are present in the (a) liquid rocket

(b) solid rocket
(a) pulsejet engine (c) electric rocket
(b) ramjet engine (d) jet engine
(c) scramjet engine
(d) turbojet engine 641. The combustor efficiency of an aircraft engine
634. The engine with afterburner is also called as (a) increases slowly with the altitude.
(b) decreases with the altitude.
(a) augment engine (c) increase rapidly with the altitude.
(b) side engine (d) remains constant with change in the altitude.
(c) reheat engine
(d) additional engine 642. Premixed flame when compared to diffusion flame has a
635. The gas turbine engine was invented by (a) worse control on fuel–air ratio.
(b) better control on fuel–air ratio.
(a) Brayton (c) no control on fuel–air ratio.
(b) Otto (d) none of the above.
(c) Atkinson
(d) John Barber 643. Which of the following is not true for a Ramjet engine?
636. The compressor used for aircraft’s application must (a) It has a high thrust-to-weight ratio.
have (b) It works well at off-design Mach numbers.
(c) As compared to other jet engines, its fuel consump-
(a) low airflow capacity tion at subsonic speeds is very high.
(b) high frontal area (d) It has zero takeoff thrust.
(c) high pressure ratio per stage
(d) low volume flow rate 644. The relation between polytropic efficiency ηp and the
overall efficiency (ηo ) of a compressor is
637. In the rotor of an axial flow compressor, the absolute
velocity of the fluid will (a) ηp = ηo
(b) ηp > ηo
(a) decrease. (c) ηp < ηo
(b) increase. (d) no relation exists.
(c) initially increases and then decrease.
(d) remain constant. 645. The relation between polytropic efficiency ηp and the
overall efficiency (ηo ) of a turbine is
638. In the combustion chamber of a jet engine, the CO2
emission in the diffusion flame is maximum at the (a) ηp = ηo
equivalence ratio of (b) ηp > ηo
(c) ηp < ηo (d) combustion intensity will go down.

(d) no relation exists
652. Because of the diffusion in the inlet of a jet engine
646. In the constant area section, across the fan
(a) air velocity decreases and the pressure increases.
(a) both velocity and the static pressure increase. (b) both air velocity and the pressure decrease.
(b) velocity and static pressure are constant. (c) both air velocity and the pressure increase.
(c) velocity is constant and the static pressure (d) air velocity increases and the pressure decreases.
increases.
(d) velocity increases and the static pressure decreases. 653. The exhaust nozzle’s pressure ratio is a strong function
of
647. The specific speed of a centrifugal compressor is gen-
erally (a) Reynolds number
(b) Prandtl number
(a) higher than that of an axial compressor. (c) Euler number
(b) less than that of a reciprocating compressor. (d) Mach number
(c) independent of the type of the compressor but
depends only on the size of the compressor. 654. In a turbojet engine, the pressure thrust as compared to
(d) more than the specific speed of the reciprocating the momentum thrust is
compressor but less than that of axial compressor.
(a) almost equal
648. Consider the following statement: (b) quite high
(i). Almost all flow losses take place in the divergent (c) quite low
section of the nozzle. (d) cannot predict
(ii). Normal shocks are likely to occur in the convergent
part of the nozzle. 655. Which of the following is the lightest and most volatile
(iii). Efficiency of reaction turbines is higher than that liquid fuel?
of impulse turbine.
(a) kerosene
Out of these statements,
(b) gasoline
(a) All are correct. (c) fuel oil
(b) (ii) and (iii) are correct. (d) vegetable oil
(c) (i) and (ii) are correct.
656. The compression ratio for petrol engines is
(d) (i) and (iii) are correct.
(a) 3 to 6
649. An impulse turbine stage is characterized by the expan-
(b) 8 to 10
sion of the gases in
(c) 10 to 15
(a) stator nozzles. (d) 15 to 20
(b) rotor nozzles.
657. Propellants in the rocket engine should have
(c) both stator and rotor nozzles.
(d) neither stator nozzle nor rotor nozzle. (a) high calorific value.
(b) low calorific value.
650. Multistage reaction turbine are employed to achieve
(c) high viscosity.
(a) large pressure drop. (d) lower thermal conductivity.
(b) large mass flow rate.
658. In an axial flow compressor, the exit flow angle devia-
(c) large volume flow rate.
tion from the blade angle is a function of
(d) large pressure rise.
(a) space–chord ratio.
651. The maximum temperature from the combustor is lim-
(b) blade camber.
ited in jet engines, because
(c) blade camber and incidence angle.
(a) it is difficult to burn the fuel. (d) blade camber and space–chord ratio.
(b) the air–fuel ratio is too lean.
659. A double-throw crankshaft operates at
(c) turbine blades cannot accept very high temperature.
(a) 90o (c) K = 3E

(1−2ν)
(b) 180o (d) K = E
2(1+ν)
(c) 270o
(d) 360o 666. If Airy’s function is Φ, then which of the following is
correct?
660. Engine-specific weight is defined as the weight of the
engine per unit (a) ∇Φ = 0
(b) ∇2Φ = 0
(a) volume (c) ∇ 3Φ = 0
(b) mass (d) ∇ 4Φ = 0
(c) density
(d) power 667. Consider u is the displacement, x is the coordinate in
axial direction, M is the moment and V is shear force.
661. The intensity of stress which causes unit strain is termed In an elastic beam problem, boundary conditions at free
as end will be
(a) modulus of rigidity (a) M = 0 and V = 0

(b) bulk modulus (b) u = 0 and ∂u
∂x = 0
(c) young modulus (c) M = 0 and u = 0
(d) modulus of elasticity (d) V = 0 and ∂u
∂x = 0
662. Capability of the material in absorbing the large amount 668. Critical load for a fixed-free column of length l will be
of energy before the fracture is
π 2 EI
(a) Pcr = l2
(a) resilience π 2 EI
(b) ductility (b) Pcr = 2l2
π 2 EI
(c) stiffness (c) Pcr = 4l2
4π 2 EI
(d) toughness (d) Pcr = l2
663. Resilience is considered when the material is subjected 669. For a plane stress problem, the state of stress can be
to represented by Mohr’s circle. The equation of Mohr’s
circle is
(a) shock loading

(b) creep (a) σn − σX +σ Y 2
+ τ 2 = σY −σ X 2
(c) fatigue 2
2

(b) σn + σX +σ Y 2
+ τ 2 = σY −σ X 2
(d) fracture 2
2

(c) σn + σX −σ Y 2
+ τ 2 = σY +σ X 2
2
2

664. The highest load, which a spring can carry without per- (d) σn − σX −σ
2
Y 2
+ τ 2 = σY +σ
2
X 2
manent distortion, is known as

670. Due to presence of taper in the structure of an aircraft
(a) proof stress wing, which of the following remains unchanged com-
(b) proof load pared to the case with no taper and same applied loads?
(c) proof stiffness
(d) proof resilience (a) axial stress in longitudinal direction.
(b) shear flow due to torsional moment applied.
665. The relation between the Bulk modulus (K), Young (c) shear flow due to applied bending moment.
modulus (E) and Poisson’s ratio (ν) of the material is (d) none of the above.
given by
(a) K = E
3(1−2ν) E.1 Keys
E(1+ν)
(b) K = 2
(See Tables E.1 and E.2)
Table E.1 Questions 1–440

Q A Q A Q A Q A Q A Q A Q A Q A
1. b 56. a 111. c 166. b 221. c 276. d 331. c 386. b
2. b 57. d 112. d 167. b 222. b 277. a 332. b 387. b
3. c 58. c 113. a 168. c 223. b 278. c 333. c 388. c
4. c 59. b 114. b 169. c 224. d 279. b 334. c 389. c
5. b 60. a 115. a 170. d 225. a 280. b 335. a 390. d
6. c 61. a 116. b 171. a 226. d 281. c 336. c 391. d
7. d 62. c 117. c 172. c 227. d 282. d 337. b 392. a
8. c 63. d 118. c 173. c 228. b 283. a 338. b 393. a
9. d 64. c 119. b 174. b 229. b 284. b 339. a 394. c
10. c 65. b 120. c 175. c 230. a 285. c 340. d 395. c
11. d 66. c 121. a 176. b 231. c 286. b 341. a 396. b
12. c 67. c 122. d 177. d 232. b 287. b 342. b 397. d
13. c 68. a 123. a 178. b 233. b 288. c 343. d 398. d
14. a 69. b 124. b 179. d 234. d 289. b 344. b 399. b
15. a 70. d 125. a 180. a 235. d 290. a 345. c 400. b
16. d 71. a 126. c 181. a 236. b 291. a 346. c 401. a
17. a 72. b 127. a 182. d 237. b 292. c 347. b 402. c
18. b 73. a 128. b 183. b 238. c 293. c 348. d 403. a
19. b 74. c 129. c 184. b 239. b 294. c 349. a 404. b
20. d 75. a 130. c 185. c 240. d 295. d 350. a 405. c
21. c 76. d 131. d 186. a 241. c 296. d 351. d 406. d
22. c 77. b 132. c 187. c 242. b 297. b 352. c 407. d
23. c 78. a 133. c 188. b 243. d 298. a 353. b 408. d
24. b 79. c 134. a 189. a 244. d 299. d 354. b 409. c
25. b 80. b 135. b 190. a 245. c 300. d 355. d 410. a
26. a 81. c 136. b 191. c 246. c 301. b 356. c 411. b
27. c 82. b 137. a 192. a 247. b 302. c 357. a 412. a
28. a 83. a 138. c 193. b 248. a 303. d 358. b 413. a
29. d 84. b 139. c 194. b 249. b 304. b 359. d 414. b
30. d 85. a 140. d 195. c 250. b 305. c 360. b 415. c
31. a 86. b 141. b 196. c 251. c 306. c 361. b 416. a
32. d 87. d 142. c 197. b 252. a 307. d 362. a 417. b
33. c 88. a 143. c 198. d 253. b 308. d 363. a 418. d
34. c 89. c 144. b 199. b 254. a 309. c 364. c 419. a
35. a 90. a 145. c 200. d 255. c 310. c 365. d 420. b
36. b 91. c 146. d 201. c 256. b 311. b 366. d 421. b
37. b 92. d 147. a 202. a 257. c 312. d 367. c 422. c
38. a 93. c 148. b 203. a 258. d 313. d 368. d 423. d
39. a 94. b 149. a 204. c 259. b 314. c 369. c 424. c
40. d 95. c 150. b 205. d 260. a 315. b 370. c 425. a
41. b 96. d 151. c 206. b 261. b 316. b 371. a 426. a
42. c 97. c 152. a 207. d 262. b 317. d 372. b 427. c
43. a 98. b 153. a 208. a 263. d 318. b 373. c 428. d
44. c 99. a 154. d 209. b 264. c 319. c 374. d 429. b
45. a 100. c 155. c 210. d 265. a 320. c 375. a 430. d
46. b 101. c 156. b 211. c 266. c 321. c 376. a 431. a
47. c 102. a 157. c 212. c 267. c 322. d 377. b 432. a
48. b 103. d 158. b 213. d 268. b 323. d 378. a 433. d
49. c 104. b 159. a 214. d 269. d 324. b 379. b 434. c
50. d 105. c 160. c 215. a 270. a 325. a 380. c 435. a
51. c 106. c 161. b 216. d 271. d 326. a 381. c 436. d
52. a 107. b 162. d 217. c 272. b 327. d 382. b 437. c
53. b 108. a 163. a 218. c 273. a 328. b 383. d 438. d
54. d 109. a 164. b 219. b 274. b 329. a 384. d 439. b
55. d 110. d 165. c 220. b 275. c 330. d 385. a 440. d
Table E.2 Questions 441–670

Q A Q A Q A Q A Q A
441. b 496. c 551. b 606. d 661. d
442. d 497. d 552. d 607. d 662. d
443. b 498. d 553. c 608. c 663. a
444. d 499. c 554. a 609. b 664. b
445. d 500. a 555. d 610. b 665. a
446. c 501. a 556. b 611. b 666. d
447. c 502. b 557. b 612. a 667. a
448. d 503. b 558. c 613. c 668. c
449. a 504. d 559. a 614. d 669. a
450. a 505. c 560. d 615. c 670. b
451. b 506. c 561. c 616. a
452. b 507. b 562. b 617. a
453. d 508. d 563. a 618. d
454. c 509. b 564. c 619. a
455. c 510. c 565. d 620. c
456. b 511. d 566. c 621. a
457. d 512. b 567. a 622. a
458. b 513. c 568. c 623. c
459. d 514. b 569. d 624. b
460. a 515. a 570. b 625. b
461. c 516. c 571. b 626. d
462. d 517. d 572. a 627. a
463. c 518. c 573. a 628. c
464. d 519. b 574. c 629. d
465. a 520. b 575. c 630 b
466. b 521. c 576 d 631. b
467. c 522. d 577. d 632. b
468. b 523. c 578. a 633. a
469. d 524. d 579. d 634. c
470. b 525. a 580. c 635. d
471. d 526. b 581. c 636. c
472. a 527. a 582. b 637. b
473. b 528. c 583. b 638. c
474. c 529. d 584. a 639. b
475. b 530. d 585. b 640. c
476. a 531. c 586. a 641. b
477. c 532. c 587. b 642. b
478. b 533. a 588. b 643. b
479. d 534. a 589. d 644. b
480. c 535. c 590. c 645. c
481. b 536. b 591. b 646. c
482. c 537. d 592. c 647. c
483. b 538. c 593. d 648. d
484. c 539. a 594. b 649. a
485. d 540. c 595. b 650. a
486. a 541. a 596. d 651. c
487. d 542. b 597. c 652. a
488. b 543. b 598. c 653. d
489. c 544. c 599. c 654. c
490. a 545. d 600. a 655. b
491. c 546. d 601. c 656. b
492. a 547. c 602. d 657. a
493. b 548. a 603. d 658. c
494. b 549. a 604. b 659. b
495. a 550. b 605. a 660. d
Letter of Acknowledgment
F
The acknowledgment letter received from the honorable defense minister for the book “Essentials of Aircraft
Armaments” published by Springer (2016).

Index
A Anticyclone, 16
Absolute altitude, 7 Archimedes principle, 28
Absorption, 11 Area–Mach number relation, 211
Acoustic intensity, 347 Area–velocity relation, 207
Acoustic intensity level, 347 Armstrong Line, 7
Acoustic power, 347 Aspect ratio, 35
Acoustics, 347 Atmosphere, 4
Active control, 346 Fédération Aéronautique Internationale, 4
Active flow control, 346 Primary Layers in the Atmosphere, 4
Adiabatic flow ellipse, 213 Exosphere, 6
Adiabatic lapse rate, 14 Mesopause, 6
Adverse pressure gradient, 256 Mesosphere, 6
Aerodynamic center, 141 Stratopause, 6
Aerodynamic forces, 28 Stratosphere, 5
drag, 28 Thermosphere, 6
lift, 28 Tropopause, 4
Aerodynamic mixing enhancement, 344 Troposphere, 4
Aerodynamic moments, 29 Secondary Layers in the Atmosphere, 6
pitching moment, 29 Heterosphere, 6
rolling moment, 29 Homosphere, 6
yawing moment, 29 Ionosphere, 6
Aerodynamics, 3, 27 Ozone Layer, 6
aircraft aerodynamics, 4 Planetary Boundary Layer, 6
industrial aerodynamics, 4 Average kinetic energy, 31
Aerothermodynamics, 4 Axial force coefficient, 39
Ageostrophic wind, 18 Axis-switch, 334
Aileron, 28 Axisymmetric flows, 83
Airbreathing, 4
Air-Breathing Engine Intakes, 393
Aircraft, 27
Airfoil, 34 B
Airfoil thickness, 34 Back pressure, 223
Airspeed, 94 Baroclinic torque, 346
Calibrated Airspeed, 95 Barotropic fluid, 130
Equivalent Airspeed, 95 Bernoulli’s equation, 87
Ground Speed, 95 Bernoulli constant, 87
Indicated Airspeed, 95 Biot and Savart law, 154
Altitude, 7 Blackouts, 194
Angle of attack, 35, 37 Blasius solution, 263
Angle of incidence, 35 Boltzmann constant, 31
Angle of inclination, 35 Boundary layers, 48, 251
Angular velocity, 72 Momentum thickness, 252
Anhedral angle, 35 x - Momentum Equation, 259
y - Momentum Equation, 259
Boundary layer separation, 254

508 Index
no-slip condition, 255 Curl of a vector, 61

Boundary layer thickness, 251 Cyclone, 16
Boussinesq hypothesis, 323
Bricfielder, 20
Broadband shock-associated noise, 352 D
Bump, 374 Degree of turbulence, 288
Buoyant jet or forced plume, 321 Degrees of freedom, 193
De Laval nozzle, 222
Density, 31
C Differential analysis, 65
Calorical equation of state, 192, 193 Diffuser effectiveness, 293
Camber, 34 Diffuser efficiency, 292
Camber line, 133 Dihedral angle, 35
Cape Doctor, 20 Dimensional analysis, 41
Carnot cycle, 187 Buckingham pi theorem, 42
Cauchy–Riemann equations, 84 Rayleigh’s method, 42
Cavity covered with porous surface, 378 Dimensionless velocity, 206
Center of pressure, 39 Direct Numerical Simulation, 332
Characteristics decay, 322 Displacement thickness, 251
Characteristic velocity, 212 Divergence angle, 290
Chinook, 20 Divergence of a vector, 61
Choked, 223 Doldrums, 20
Chord, 34 Dot product, 60
Chord length, 34 Doublet, 113
Chord line, 34, 133 Downwash, 145
Circulation, 62, 123, 127 Drag coefficient, 39
clockwise circulation, 132 Dynamic pressure, 88
counterclockwise, 132 Dynamic similarity, 287
Clausius inequality, 188 Dynamic viscosity, 325
Clausius Statement, 186
Closed system, 181
Closed test section, 285 E
Closed-throat tunnel, 290 Eddy viscosity, 276, 325
Coefficient of pressure, 92 Edney classification, 365
Coefficient of thermal conductivity, 185 Effuser, 288
Coefficient of viscosity, 33 Ekman layer, 19
Newton’s law of viscosity, 33 Ekman number, 19
Sutherland formula, 33 Emission, 11
Co-flowing jet, 321 Energy equation, 99, 200
Compressibility, 47 Energy equation for an open system, 185
isentropic compressibility, 47 Energy integral equation, 262
isothermal compressibility, 47 Enthalpy, 183
Compressible flows, 199 stagnation enthalpy, 205
Compressible fluids, 181 static enthalpy, 205
Compressible jets, 322 Entropy, 181, 187
Compressible laminar flow over a flat plate, 266 Entropy layer, 238
Compression corner, 226 Entropy mode acoustics, 347
Concept of enthalpy, 183 Environmental lapse rate, 10
Constant pressure, 184 Equations of fluid motion, 57
Constant volume, 184 Equipotential lines, 85
Continuity equation, 64, 67, 200 Equivalent cone angle, 290
Continuum flow, 45 Eulerian or field description, 63
Knudsen number, 45 Euler number, 44, 287
Contraction cone, 288 Euler’s equation, 207
Control surface, 181 Expansion corner, 226
Control volume, 181 Expansion fan, 226
Convective derivative, 64
Convective Mach number, 346
Convective speed, 346 F
Convergent–divergent diffuser, 304 Fan efficiency, 293
Convergent–divergent nozzle, 222 Fanno flow, 215
Coriolis force, 14, 130 Favorable pressure gradient, 256
Coromuel, 21 Ffowcs Williams–Hawkings equation, 349
Correction factor, 299 Finite control mass approach, 65
Correctly expanded, 322 Finite control volume approach, 65
Cross product, 58 First law of thermodynamics, 64, 181
Index 509
Flaps, 28 hypersonic similarity parameter, 244

Flow control, 346 Hypersonic wind tunnel, 309
Flow deflection angle, 228, 230, 237
Flow similarity, 43
dynamic similarity, 43 I
geometric similarity, 43 Impinging jet, 321
kinematic similarity, 43 Incompressible jets, 322
scale factor, 43 Indraft tunnel, 297
Flow velocity, 31 Induced drag, 147
Flow work, 185 Infinitesimal fluid element approach, 65
Fluidity, 3 Infinite vortex, 155
Fluids, 3 Inlet, 394
Forced vortex, 149 Integral analysis, 65
Fourier’s law, 185 Intensity of turbulence, 288
Fourier’s law heat conduction, 185 Intensity of wave, 10
Free molecular flow, 45 Intermittent-blowdown wind tunnel, 296, 297
Free or submerged jet, 321 Intermittent-indraft wind tunnels, 297
Free vortex, 149 Internal energy, 181
Froude number, 44, 287 Inviscid flow, 48
Frozen, 223 Isentropic efficiency or diffuser effectiveness, 292
Fully developed region, 322 Isentropic process, 205
Fuselage, 28 Isolated system, 181
Isotropic, 325
G
Gauss divergence theorem, 63 J
Geometric altitude, 7 Jet centerline pitot pressure decay, 341
Geometric similarity, 287 Jet controls, 344
Geostrophic wind, 16 Jet streams, 20
Gibbs free energy, 189
Gliding angle, 28
Gradient of, 60 K
directional derivative, 61 Karman Line, 7
Gradient theorem, 63 Karman–Pohlhausen approximate solution method, 270
Green house effect, 4 Kelvin–Helmholtz instability, 344
Gun tunnel, 311 Kelvin–Helmholtz vortex rings, 328
Kelvin–Planck Statement, 186
Kelvin’s circulation theorem, 129
H Khamsin, 21
Haboob, 21 Kinematic similarity, 287
Half-saddle point, 255 Kinematic turbulent viscosity, 325
Harmattan, 21 Kinetic energy flux factor, 292
Hawk, 21 Kinetic energy thickness, 253
Heat sink, 186 Kinetic theory of gases, 30
Heat source, 186 Kronecker delta, 324
Heavier-than-air, 27 Kutta condition, 128
Helmholtz free energy, 189 Kutta–Joukowski theorem, 123
Helmholtz’s theorems, 150
Hess and Smith method, 169
Hiemenz flow, 267 L
Hodograph, 49 Lagrange stream function, 80
Homogeneous turbulence, 274 Lagrangian or the particle or the material description, 63
Honeycomb structures, 288 Laminar boundary layers, 263
Hooke’s law, 3 Laminar separation bubble, 257
Horizontal Buoyancy, 290 Laminar sub-layer, 277
Horizontal stabilizer, 28 Laminar–Turbulent transition, 271
Hurricanes, 16, 148 Land breeze, 19
Hydrostatic pressure, 31 Laplace’s equation, 117
Hypersonic, 4 Lapse rate, 9
Hypersonic flow, 49, 237 Large Eddy Simulation, 330
density ratio, 240 Leading edge, 34
θ − β − M relation, 240 L’ Hospital’s rule, 136
Prandtl–Meyer expansion fan, 242 Lift, 28
pressure ratio, 240 Lift coefficient, 39
temperature ratio, 240 Lift-to-drag ratio, 246
Hypersonic similarity, 243 Lighter-than-air, 27
510 Index
Lighthill stress tensor, 349 Static Pressure Ratio, 227

Line integral, 62 Static Temperature Ratio, 227
Line source distribution, 172 Oblique shock wave cancelation, 362
Line vortex, 114 Open system, 181
Local and Material Derivatives, 63 Open test section, 285
Logarithmic buffer layer, 278 Optical flow visualization, 344
Loo, 21 Ornithopter, 27
Losses in subsonic wind tunnels, 293 Orthogonal coordinate system, 58
Low-density tunnels, 286 Oswatitsch Relation, 369
Ludwieg tube, 311 Overexpanded, 322
Ludwig Prandtl, 251
P
M Panel method, 169
Mach angle, 203, 228, 230 Passive control, 346
Mach cone, 203 Pathline, 70
Mach number, 44, 203, 254, 287 Perfect gas, 193
Mach Number Independence, 242 Pitch, 28
Mach-disc, 344 Pitch angle, 35
Mass flow rate, 209 Pitching moment, 37
Matter, 3 Pitching moment coefficient, 141
gas, 3 Pitot probe, 340
intermolecular, 3 Pitot-static probe, 92
liquid, 3 Plasma wind tunnel, 311
solid, 3 Point of separation, 255
Maximum lift coefficient, 246 Polarizability, 11
Maxwell’s relations, 189 Poles, 14
Mayer’s relation, 191, 193 Potential flow, 109
Mean camber line, 34 Potential jet core, 322
Mesh factor, 294 Powered airships, 27
Method of characteristics, 223, 300 Prandtl–Glauret transformations, 348
Microscopic Approach, 65 Prandtl–Meyer expansion fan, 230
Microscopic point of view, 31 Prandtl–Meyer function, 231
Micro-Vortex Generators, 381 Prandtl–Meyer relation, 221
Milky Way, 148 Prandtl’s lifting line integro-differential equation, 160
Minimum wave angle, 228 Prandtl’s Mixing Length Hypothesis, 275
Mixing length, 325 Pressure, 10, 30
Mixing of two uniform laminar flows, 269 Pressure coefficient, 119, 241
Models, 43 Pressure mode acoustics, 347
Modified Newtonian theory, 247 Pressure profiles, 342
Moment coefficient, 39 Pressure regulating valve, 340
Momentum equation, 64, 200 Pressure surface, 34
θ − β − M Relation, 228 Prototypes, 43
N R
Navier–Stokes equations, 73 Ramjets, 393
Newtonian sine-squared law, 245 Rankine oval, 117, 169
Newton’s law of viscosity, 361 Rankine–Hugoniot equation, 221
Non-simple regions, 362 Rankine’s half-body, 116
Nonuniform flow, 46 Rarefied flow, 45
Normal force coefficient, 39 Rayleigh flow, 215
Normal shock, 216, 226 Rayleigh Pitot probe formula, 222
Density Ratio, 219 Reference area, 39
Entropy Change, 219 Refractive index, 11
Stagnation Pressure Ratio, 218 refractivity of the atmosphere, 11
Static Pressure Ratio, 218 Snell’s law, 11
Static Temperature Ratio, 218 Reichardt’s equation, 327
Nozzle Pressure Ratios, 341 Reichardt’s inductive theory of turbulent flows, 327
Reverse nozzle diffuser, 304
Reynolds Averaged Navier–Stokes, 322
O Reynolds number, 44, 239, 254, 287
Oblique shock, 216, 226 Reynolds Rules of Averaging, 273
Density Ratio, 227 Reynolds stresses, 275
Entropy Change, 227 Reynolds Transport theorem, 95
Stagnation Pressure Ratio, 227 Richardson number, 321
Index 511
Roll, 28 Stokes theorem, 63, 128

Root chord, 35 Straight vortex line segment, 155
Rossby number, 19 Streakline, 70
Streamline, 68
Streamtube, 69, 150
S Strength of shock, 229
Scalar field, 59 Stress, 31
Scalar product, 57 Mohr’s circle, 32
Scattering angle, 11 Subsonic flow, 48
Scattering phenomena, 10 Subsonic inlet, 394
Rayleigh scattering, 10 Subsonic jet noise, 350
Schlieren, 344 Suction surface, 34
Scramjets, 393 Superposition, 115
Screech tone, 351 Supersonic combustion ramjet, 344
Screens, 289 Supersonic flow, 48
Screen solidity, 294 Supersonic jet core, 322
Sea breeze, 19 Supersonic jet noise, 350
Second law of thermodynamics, 64, 186 Surface forces, 31
Second throat, 296 Surface integral, 62
Self-similar region, 322 Surroundings, 181
Semi-infinite body, 116 Symmetric airfoil, 132
Semi-infinite vortex, 155 System, 181
Separation pressure, 226
Settling chamber, 299
Shadowgraph, 344 T
Shaft work, 185 Taper Ratio, 35
Shear strain rate, 73 Taylor’s vorticity transport theory, 326
Shear velocity, 277 Temperature, 10, 31
Shock-associated noise, 351 Test rhombus, 304
Shock–boundary layer interactions, 361 Test section, 290
Shock cell, 328 Theodore von Karman, 7
Shock–Shock interference, 364 Thermal equation of state, 192, 193
Shock tube, 309 Thermally perfect gas, 192
Shock tunnel, 311 Thin airfoil theory, 127
Similarity parameters, 254 Timeline, 68
Simple waves, 362 Tip chord, 35
Sink, 110 Tractive force, 28
line sink, 110 Trailing edge, 34, 131
point sink, 110 Trailing vortex, 145
Sir George Cayley, 27 Transformations for panel coordinates, 177
Sirocco, 21 Transition flow, 45
Slip flow, 45 Transonic flow, 48
Sound pressure level, 347 Transonic interactions, 361
Source, 110 Trim tabs, 28
line source, 110 Tropics, 14
point source, 110 True airspeed, 92
Source panels, 169 Turbulence model, 327
Spalart–Allmaras turbulence model, 330 Turbulent boundary layers, 263, 272
Span, 145 Turbulent kinetic energy, 324
semi-span, 147 Turbulent shear stress, 325
Specific heat ratio, 193 Typhoons, 16, 148
Speed of sound, 10, 201
Spinning tunnels, 286
Stability tunnels, 286 U
Stagnation point flow, 267 Uncertainty analysis, 407
Stagnation properties, 205 Underexpanded, 322
Starting vortex, 131 Uniform flow, 46, 109
Static pressure coefficient, 293 Universe, 181, 188
Static properties, 205 Unsteady flow, 46
Static stability, 12
neutrally stable, 12
statically stable, 12 V
statically unstable, 12 Vector field, 60
Steady flow, 46 Vector product, 60
Stefan–Boltzmann law, 11 Vertical stabilizer, 28
Stewartson Layer, 19 Viscous flows, 48
512 Index
Viscous-inviscid flow interaction, 239 Westerlies, 20

Volume flow rate, 84 Wind, 14
Volume integral, 63 wind shear, 21
Von Karman constant, 276, 326 Wind tunnel, 285
Von Karman momentum integral equation, 260 Air Supply System and Storage Tanks, 299
Von Karman’s rules, 204 energy ratio, 295
Von Karman’s similarity hypothesis, 326 Fleigner’s formula, 295
Vortex filament, 150 Wing area, 35
Vortex flow, 119 Wing root, 145
Vortex line, 149 Wings, 28
Vortex panel methods, 174 port-wing, 28
Vortex panels, 169 starboard wing, 28
Vortex tube, 150 Wing span, 35
Vorticity, 72 Wing sweep, 35
Vorticity mode acoustics, 347 Wing tip, 145
Work done, 183
W
Wall temperature, 240 Y
Wave angle, 228, 237 Yaw, 28
Weber number, 44
Wein’s-displacement law, 11

Theoretical and Experimental Aerodynamics by Dr. Mrinal Kaushik

Uploaded by

Copyright:

Available Formats

Theoretical and Experimental Aerodynamics by Dr. Mrinal Kaushik

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Theoretical and Experimental Aerodynamics by Dr. Mrinal Kaushik

Uploaded by

Copyright:

Available Formats

Mrinal Kaushik

Theoretical and Experimental

ISBN 978-981-13-1677-7 ISBN 978-981-13-1678-4 (eBook)

Library of Congress Control Number: 2018951404

© Springer Nature Singapore Pte Ltd. 2019

Kharagpur, India Mrinal Kaushik

Part I Theoretical Aerodynamics

2.3.6 Coefﬁcient of Viscosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

3.10 Angular Velocity, Vorticity, and the Shear Strain Rate . . . . . . . . . . . . . . . 70

5.5 Summary of the Lift Generation Mechanism by the Airfoil . . . . . . . . . . . 132

8.4 Entropy and the Second Law of Thermodynamics . . . . . . . . . . . . . . . . . . 186

9.10 Two-Dimensional Flow Across an Oblique Shock Wave . . . . . . . . . . . . . 226

11.11 Turbulent Boundary Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272

Part II Applied Aerodynamics

13.3.4 Direct Numerical Simulation (DNS) . . . . . . . . . . . . . . . . . . . . . 332

Mrinal Kaushik is Assistant Professor of aerospace engineering

aT Coefﬁcient of thermal expansion

© Springer Nature Singapore Pte Ltd. 2019 3

Aeronautical Science Space Technology

Fig. 1.1 Classification of aerospace engineering

Fig. 1.3 Absolute and geometric altitudes

air column of radius r and height dhG as shown in Fig. 1.4. It

where1 a is called the environmental lapse rate for gradient

From Eq. (1.10), we have

p = (const) ρ (1.13) −g0

p ρ RT atmospheric pressure. It is dependent upon the Earth’s gravi-

In troposphere, the environmental lapse rate a = 6.5 ◦ Ckm−1

p = kρ1.235 (1.19) At sea level conditions, the density of air is found to be

T = T1 + a (h − h1 ) (1.20) 1.7.4 Scattering Phenomena

The phenomena of light scattering is also known as Rayleigh

1.7.5 Absorption and Emission N = 106 (n − 1) (1.25)

λmax T = 2.898 × 10−3 (1.23) α0 < α1 < α2 < · · · < αn · · · .

Fig. 1.5 A schematic diagram of stratified layers in the atmosphere

The above two equations can be written in generalized form

Air parcel Statically unstable

Initial perturbation Statically stable Neutrally stable

Fig. 1.6 Schematic diagram of the different types of stability

Fig. 1.7 Schematic diagram of upward movement of air parcel

mechanical and chemical) equilibrium with its surroundings or

2. Further, if n > γ, then γ−n γn < 0 which implies ρb < ρb

Pole star ω Zenith

Cold air Direction of motion of center

Stormy cloudy weather Calm clear weather

Hot rising air Cold subsiding air

Converging surface Diverging surface

Cyclonic flow Anticyclonic flow

Fig. 1.12 Cyclonic and anticyclonic flows

We can see that the geostrophic wind is inversely proportional or

Pressure gradient Real wind

Friction Coriolis force

frictional force ν ν 1.12.1 Land and Sea Breeze

Centrifugal force Centrifugal force

Coriolis force Pressure gradient

Surface water currents

Cold air Warm air Warm air Cold air

Ocean Land Ocean Land

(a) Sea breeze (b) Land breeze

(a) cyclone (I) Hurricanes do not occur in a belt of 5◦ –8◦ on either

9. In the above question, if the axis of rotation is parallel

Fuselage Landing flap

2.3.5.2 State of Stress in Viscous Fluid

where su is the linear distance measured along the upper sur-