Proceedings of the World Congress on Engineering 2009 Vol II

WCE 2009, July 1 - 3, 2009, London, U.K.

Modeling & Simulation of Supersonic Flow Using

McCormack’s Technique
M. Saif Ullah Khalid*, Afzaal M. Malik**

Abstract—In this work, two-dimensional inviscid To employ, this numerical procedure, inviscid
supersonic flow around a wedge has been investigated supersonic flow may be described using following
using CFD scheme. The employed scheme is Euler equations. With no source involved,
McCormack’s predictor-corrector scheme. Our own code,
CFMOD was written using MATLAB environment.
Proper boundary conditions and the grid parameters were
decided on the basis of standard techniques. The results
are verified using classic theories for supersonic flows and where;
a standard CFD package; Fluent. The proper grid
generation around the airfoil was decided on the basis of
numerical experimentation. Various flow parameters were
found to be in good agreement with the results obtained
from theory and Fluent.
Index Terms—CFD, Supersonic flow, McCormack’s
Scheme, Inverse solution for Prandtl-Mayer function

I. INTRODUCTION
Computational Fluid Dynamics has been converted into
a specialized field to find out the solution for many real
complex physical problems. Especially, the problems
involving the shock waves are of great importance. This partial differential equation is to be solved
Now-a-days, many space- marching and time-marching numerically in the transformed space i.e. computational
schemes are in use. Some of these schemes capture the ξη-plane.
effects of shock waves themselves and for others; one
Physically, the flow field may be imagined as shown in
has to follow shock-fitting approach in the CFD
Figure 1.
schemes. The analytical and numerical calculation of
supersonic flow parameters around the wedge and
sharp-edged airfoils has been the topic of wide research,
in the fields of Mechanical engineering, Aerospace
engineering and Mathematics. In [1], the authors
studied local existence and the uniqueness of weak
shock solution in steady supersonic flow past a wedge.
The authors [2] constructed an analytical solution for
the generation of shock wave as a result of supersonic
flow around a wedge. In this work, McCormack’s Fig. 1: wedge airfoil
explicit predictor-corrector scheme has been employed
to find out the parameters of the flow-field around the
wedge.

Department of Mechanical Engineering

College of Electrical & Mechanical Engineering
National University of Science & Technology, Pakistan
Email: *[email protected],
**[email protected]

ISBN:978-988-18210-1-0 WCE 2009

Proceedings of the World Congress on Engineering 2009 Vol II
WCE 2009, July 1 - 3, 2009, London, U.K.

For simplicity, the grid for physical space is shown to II. GRID GENERATION
be;
As described in [3], to calculate the step size for the
downstream space marching, CFL (Courant-Freidrichs-
Lewy) criterion may be used.

where, = Mach angle

= flow deflection angle

III. INITIAL & UPPER BOUNDARY CONDITIONS

Fig. 2: Grid in Physical space
Initial data line is selected to be at while upper
We can use following relations to transform the xy boundary line is . Along these vertical and
coordinate system into generalized ξη- coordinate horizontal lines, at each grid point, the initial data, equal
system [3]. to the uniform upstream flow conditions for Mach, M =
1.8.

IV. NUMERICAL MODELING

To use McCormack’s scheme [3, 4], following relations

where; need to be evaluated.

= Wedge surface =

= Local height of the physical system

where, These relations can be used to represent Euler equations

in the computational space, i.e. in the form of
= maximum height of the physical system
generalized coordinates.
Now, the computational space will look like a
rectangular grid as shown in Figure 3.

This is the required partial differential equation, to be

solved numerically using McCormack’s scheme.

V. WALL BOUNDARY CONDITIONS

Only condition that can be imposed as the wall

boundary condition is the flow to be tangent to the wall.
As described in [3], Abbett’s work can be used to
implement the wall boundary conditions. Using this
condition, the parameters on the wall can be calculated
according to the following scheme.

1). Flow parameters have been calculated using normal

Fig. 3: Grid in Computational space approach of McCormack’s technique.

ISBN:978-988-18210-1-0 WCE 2009

Proceedings of the World Congress on Engineering 2009 Vol II
WCE 2009, July 1 - 3, 2009, London, U.K.

2). The velocities, calculated from the usual approach

may not be tangent to the wall, as shown in the Fig. 4.

Fig. 5: Comparison of Numerical & Analytical Results (for

air)
Fig. 4: Rotation of flow to fulfill tangency condition
6). From these actual values of Mach number, pressure,
3). The direction of this calculated flow velocity may be density and temperature may be calculated [1].
calculated using the Prandtl-Mayer function [5, 6].
VI. DISCUSSION & COMPARISON OF RESULTS

The results obtained from the numerical solution

without involvement of numerical dissipation are
discontinuous. If the x and y coordinates of the grid
4). The actual direction of the velocity may be points with unstable numerical results are used, it is
incorporated using following relation. obviously noted that this discontinuity arises, right after
the production of shock wave at the corner of the
wedge. So, to remove this discontinuity, numerical
dissipation has been employed. The relation used for
It means that the actually flow may be tangent to the this purpose is, as proposed in [3].
wall of the wedge, if the numerically calculated flow be
rotated to an angle of . For predictor step, numerical dissipation is;

5). Actual Mach number M, with tangent flow to the

wall, can be calculated from the following For corrector step, this expression may be used to
approximation that has been estimated using MATLAB calculate the numerical dissipation factor, using values
curve fitting tool [7]. It actually represents the inverse of the pressure vector P and A-vector, as estimated in
solution for the Prandtl-Mayer function. the predictor step.

Numerical experimentation with our code reveals a fact

that the grid spacing is not that much important as the
choice of Courant number is, as described in Fig. 6.
This plot shows the graph of Courant number versus
Mach number for first point on the boundary wall after
the shock wave for different values of step size along
vertical direction of physical space.

ISBN:978-988-18210-1-0 WCE 2009

Proceedings of the World Congress on Engineering 2009 Vol II
WCE 2009, July 1 - 3, 2009, London, U.K.

Fig. 6: Plot of Courant number against Mach for

different values of step size

In this figure, represents the grid spacing in the Fig. 7: Mach number Contour plot along wedge front
upper surface from CFMOD
physical plane along vertical direction. From the classic
theory of supersonic flow, the calculated results for
flow parameters, behind the shock wave are shown in
the following table.

Table 1: Theoretical Values for flow parameters behind

shock wave

Flow Parameter Value

Pressure 131661.705
Temperature 321.81 K
Mach Number 1.623
Density 1.23
Figure 8: Mach number Contour plot along
wedge front upper surface from FLUENT
As shown in Table 1, the value of Mach number is
1.623, so according to the Fig. 6, the appropriate value
of the Courant number is 0.1.

Contour plots for Pressure, Temperature, Density and

Mach number, obtained from CFMOD and FLUENT
[8] for front upper surface of the wedge foil are shown
in Figures 7-14. These plots show that results obtained
from CFMOD are in good agreement with those from
FLUENT [8] and classic theories of supersonic flow.

Fig. 9: Density Contour plot along wedge front

upper surface using CFMOD

ISBN:978-988-18210-1-0 WCE 2009

Proceedings of the World Congress on Engineering 2009 Vol II
WCE 2009, July 1 - 3, 2009, London, U.K.

Fig. 10: Density Contour plot along wedge front upper Fig. 14: Temperature Contour plot along wedge front
surface from FLUENT upper surface using FLUENT

ACKNOWLEDGEMENTS

The authors are indebted to College of Electrical &

Mechanical Engineering, National University of
Sciences & Technology and Higher Education
Commission, Pakistan for having made this work
possible. The authors also appreciate the technical
support of Mr. Raja Amer Azim and Mr. Ameer Asim
Khan, College of Electrical & Mechanical Engineering,
National University of Science & Technology, Pakistan.

Figure 11: Pressure Contour plot along wedge front upper

surface using CFMOD
REFERENCES

[1] Sxuxing Chen, Jian Zhong Min and Yongqian Zhang, “Weak
Shock Solution in Supersonic Flow Past a Wedge”, pp. 115-
132, Discrete and Continuous Dynamical Systems, Volume 23,
Number 1 & 2, January & February 2009.

[2] Volker Elling and Tai-Ping Liu, “Supersonic Flow onto a Solid
Wedge”, Communication on Pure and Applied Mathematics,
Volume 61, Number 10, 2008.

[3] John. D. Anderson Jr., “Computational Fluid Dynamics; The

Fig. 12: Pressure Contour plot along wedge front upper Basics with Applications”, McGraw Hill Series in Mechanical
surface using FLUENT Engineering.

[4] John C. Tennehill, Dale, A. Anderson, Richard H. Pletcher,

“Computational Fluid Mechanics & Heat Transfer”, 2nd
Edition, Series in Computational Physical Processes in
Mechanics and Thermal Sciences, 1997.

[5] John. D. Anderson Jr., “Modern Compressible Flow; With

Historical Perspective”, 2nd Edition, McGraw Hill Series in
Aeronautical and Aerospace Engineering.

[6] John. D. Anderson Jr., “Fundamentals of Aerodynamics”, 4th

Edition, McGraw Hill Series in Aeronautical and Aerospace
Engineering.

[7] “MATLAB R2007b”, The MathWorks, Inc.

Fig. 13: Temperature Contour plot along wedge front
[8] “FLUENT 2ddp solver” Fluent Inc.
upper surface using CFMOD

ISBN:978-988-18210-1-0 WCE 2009

View publication stats