0% found this document useful (0 votes)
16 views77 pages

CMTE Classes 9, 10 & 11

The document provides an overview of the Computational Fluid Dynamics (CFD) analysis process, outlining steps from problem selection to reporting findings. It includes resources such as websites, books, and journals related to CFD, as well as essential governing equations and assumptions for modeling fluid dynamics. The objective is to ensure that CFD results yield accurate and credible information for real-world applications.

Uploaded by

Pratap
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
16 views77 pages

CMTE Classes 9, 10 & 11

The document provides an overview of the Computational Fluid Dynamics (CFD) analysis process, outlining steps from problem selection to reporting findings. It includes resources such as websites, books, and journals related to CFD, as well as essential governing equations and assumptions for modeling fluid dynamics. The objective is to ensure that CFD results yield accurate and credible information for real-world applications.

Uploaded by

Pratap
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 77

ME16002 – CMTE

Classes 9, 10 & 11 - CFD Overview

G.R.K.Gupta, NITW
09-09-2024, 11:00 hours

9/9/2024 GRK Gupta, MED, NITW 1


OVERVIEW OF C.F.D

9/9/2024 GRK Gupta, MED, NITW 2


CFD Analysis Process
• Select a real world Problem
• Understand the Physics of the problem
• Check the feasibility and Resources for the problem
• Formulate the Flow Problem (Governing Equations)
• Simplify the Model without Loss of Physics (Assumptions)
• AFD or EFD or CFD
• Decide about the Computational Domain
• Discretize the Geometry (Generate the Grid)
• Discretize the Equations (FDM, FVM or FEM)
• Specify the Initial and Boundary Conditions
• Set up the CFD Simulation
• Perform and Monitor the CFD Simulation
• Examine and Process the CFD Results (Post-Processing)
• Further Analysis? Validation and Verification of the code
• Report the Findings
• The objective is confidence
9/9/2024
that the CFD results provide accurate,
GRK Gupta, MED, NITW 3
credible, and useful information.
SELECT A REAL WORLD PROBLEM

9/9/2024 GRK Gupta, MED, NITW 4


USEFUL WEBSITES
• http://www.cfd-online.com/ leads to the following links
CFD Resources Online
A comprehensive collection of annotated
links to information related to CFD
CFD Jobs Database
A forum where companies and universities
can advertise open positions in CFD
CFD News and Announcements
Use this forum to announce CFD news. You
can subscribe to a weekly email news digest.
CFD Books Guide
A guide to CFD literature. Includes book-
descriptions and reader-reviews.
CFD Online Discussion Forums
Meet and discuss with fellow CFD colleagues
from all over the world.
CFD Events Calendar
Search for CFD conferences, courses, ...
Anyone can submit new events.
9/9/2024 GRK Gupta, MED, NITW 5
USEFUL WEBSITES - II
1.ansys.com
2.fluent.com
3.flow3d.com
4.cd-adapco.com
5.cfd review.com
6.qnet-ercoftac.cfms.org.uk
.......
9/9/2024 GRK Gupta, MED, NITW 6
BOOKS ON C.F.D-I
• Computational Fluid Dynamics: The Basics with
Applications
by John David Anderson (For FDM)
• Computational Methods for Fluid Dynamics
by Joel H. Ferziger, et al (For FVM, Unstructured Grids)
• Computational Fluid Dynamics
by Versteeg and Malalasekera (For FVM, Structured Grids)
Computational Fluid Dynamics: Principles and
Applications
by Blazek, J (For FVM, Unstructured Grids)
9/9/2024 GRK Gupta, MED, NITW 7
BOOKS ON C.F.D-II
• Computational Fluid Dynamics – 3 Volumes
by Hoffmann, K.A., and CHIANG, S.T. (FDM and FVM)
• Computational Fluid Dynamics
by Chung, T.J. (Covers FDM, FVM and FEM)
• Computational Fluid Dynamics
by Date, A.W., IIT, Bombay (FDM)
• Computational Fluid Dynamics
by Sengupta, IIT, Kanpur (FDM)

9/9/2024 GRK Gupta, MED, NITW 8


BOOKS ON C.F.D-III
• Computational Fluid Flow and Heat
Transfer
by Muralidhar, K (IITK) and Sundarajan, T (IITM)
• Computational Fluid Mechanics and Heat
Transfer
by Tannehill, Anderson and Pletcher (FDM)

9/9/2024 GRK Gupta, MED, NITW 9


JOURNALS ON C.F.D - I
• International Journal of Computational Fluid
Dynamics
• International Journal for Numerical Methods in
Fluids
• Theoretical and Computational Fluid Dynamics
• Computers & Fluids
• Journal of Computational Physics
Numerical Heat Transfer
Journal of Turbulence

9/9/2024 GRK Gupta, MED, NITW 10


JOURNALS ON C.F.D - II
• International Journal of Numerical
Methods for Heat & Fluid Flow
• International Journal of Computational
Methods
• International Journal of Fluid Dynamics
• Journal of Fluids Engineering, ASME
• VKI Links to Fluid Dynamics Related
Journals etc.
9/9/2024 GRK Gupta, MED, NITW 11
COMMERCIAL PACKAGES ON C.F.D
https://www.taygeta.com/CFD/CFD_codes_c.html
gives you a list of commercial softwares as given below

Fluent Inc. (Fluent/V4, Fluent/UNS, Rampant, Nekton)


AEA Technology (CFX: 3D fluid flow/heat transfer code)
Computational Dynamics Ltd. (STAR-CD)
Engineering Mechanics Research Corp. (NISA)
FLUIDYNE (Transoft International)
Flowtech Int. AB (SHIPFLOW: analysis of flow around ships)
Fluid Dynamics International, Inc. (FIDAP)
ICEM CFD (ICEM CFD, Icepak)
CFD Research Corporation (ACE: reactive flows)
ARSoftware (TEP: a combustion analysis tool for windows)
PORFLOW
FLOW3D
9/9/2024 GRK Gupta, MED, NITW 12
UNDERSTAND THE PHYSICS OF
THE PROBLEM

9/9/2024 GRK Gupta, MED, NITW 13


Check the feasibility and
Resources for the problem

9/9/2024 GRK Gupta, MED, NITW 14


Dimensionless Variables for
Experimental Methods
Knudsen Number
Strouhal Number
Reynolds Number
Euler Number
Mach Number
Froude Number
Galilio Number
Weber Number
9/9/2024 GRK Gupta, MED, NITW 15
Prandtl Number
Biot Number
Fourier Number
Peclet Number
Nusselt Number
Stanton Number
Eckert Number
Grashoff Number
Rayleigh Number
Lewis Number
Schmidt Number
Sherwood Number
9/9/2024 GRK Gupta, MED, NITW 16
MATHEMATICAL MODELLING
Obtain the Governing Equations for the
problem
– The conservation equations
• Mass Balance (Continuity)
• Momentum Balance (Newton’s Second
Law)
• Energy Balance (First Law of
Thermodynamics)
– Equations of property variations
– Equations of chemical kinetics etc.
9/9/2024 GRK Gupta, MED, NITW 17
Summary of equations in
conservation form
  ( u )  ( v )  ( w)
Mass : + + + =0
t x y z

( u ) ( uu) ( uv ) ( uw) p  xx  yx  zx


x − Momentum : + + + = + + + + f x
t x y z x x y z

( v ) ( vu ) ( vv ) ( vw) p  xy  yy  zy


y − Momentum : + + + = + + + + f y
t x y z x x y z

( u ) ( wu ) ( wv ) ( ww) p  xz  yz  zz


z − Momentum : + + + = + + + + f z
t x y z x x y z

Energy :
 
  (e + V 2 / 2 )
 
+ .  (e + V 2 / 2 )V =
(up ) (vp ) ( wp )
+ + + f .V + q
t x y z
Equations of state : p = p (  , T ) and i = i (  , T )
e.g . for perfect gas : p =  RT and i = CvT
9/9/2024 GRK Gupta, MED, NITW 18
EQUATIONS IN GENERIC (STRONG
CONSERVATION) FORM
U F G H
+ + + =J
t x y z

    u 
    u 2 + p −  xx 
u  

 
 
 vu −  xy 

U = v  F =
wu −  
   
  (e + V 2 ) + pu −
xz
w 2

   
  (e + V 2 )
 
 2 
 k T  x − u  − v  − w 
 xx xy xz

9/9/2024 GRK Gupta, MED, NITW 19


 v   w 
   wu −  xz 

vu −  xy   

 v 2 + p −  yy  H =
 vu −  yz 
G=   

wv −  yz   w 2 + p −  zz 
  (e + V 2 2 ) + pv −  ( )
  e + V 2 2 + pw − 
  



 k T z − u xz − v zy − w zz
 
 k T y − u yx − v yy − w yz  
  

 0 
  fx 

 

J = f y 
 f z 
 

  ( uf x + vf y + wf z ) +  
q 

Where U – Solution Vector


F,G,H - Flux Vectors
9/9/2024 j – SourceGRK
Vector
Gupta, MED, NITW 20
General transport equations
The system of equations is closed, with seven equations for seven
variables: pressure, three velocity components, enthalpy, temperature,
and density.
There are significant commonalities between the various equations. Using
a general variable , the conservative form of all fluid flow equations can
usefully be written in the following form:

 (  )
+ div(  u ) = div( grad  ) + S
t

Or, in words:

Net rate of flow


Rate of increase Rate of increase Rate of increase
of  out of
of  of fluid + = of  due to + of  due to
fluid element
element diffusion sources
(convection)

9/9/2024 GRK Gupta, MED, NITW 21


Integral form
• The key step of the finite volume method is to integrate the differential
equation shown in the previous slide, and then to apply Gauss’ divergence
theorem, which for a vector a states:

 div a dV =  n  a dA
CV A

•This then leads to the following general conservation equation in integral form:

 
   dV  +  n  (  u) dA =  n  ( grad  ) dA +  S dV
t  CV  A A CV

Net rate of Net rate of


Rate of Net rate of
decrease of  due increase of  due
increase + to convection = + creation
to diffusion
of  of 
across boundaries across boundaries

This is the actual form of the conservation equations solved by finite volume
9/9/2024 GRK Gupta, MED, NITW
based CFD programs to calculate the flow pattern and associated scalar fields.22
ASSUMPTIONS
Make suitable assumptions without loss of physics to simplify the
problem
• Steady or Unsteady
• Compressible or Incompressible
• Uniform or Non-Uniform
• Viscous or Inviscid
• 1-D, 2-D or 3-D
• Laminar or Turbulent
• Subsonic or Supersonic
• Forces dominant or negligible
• Single- or multi-phase (Ca)
• Thermal/density effects (Pr, g, Gr, Ec)
• Free-surface flow (Fr) and surface tension (We)
• Chemical reactions and combustion (Pe, Da) etc…
9/9/2024 GRK Gupta, MED, NITW 23
DECIDE THE GEOMETRY OR
THE BOUNDARY

9/9/2024 GRK Gupta, MED, NITW 24


Types of Domains

Simply connected: No holes; any loop can shrink to a


point. (e.g., a square region with no obstacles).

Doubly connected: One hole; there is one obstruction


in the domain. (e.g., flow around a cylinder).

Multiply connected: Multiple holes or obstructions.


(e.g., flow through a region with several obstacles).

9/9/2024 GRK Gupta, MED, NITW 25


Flow configurations - external flow
Fluid flows over an object in an unconfined domain.
Viscous effects are important only in the vicinity of the
object.
Away from the object, the flow is essentially inviscid.
Examples: flows over aircraft, projectiles, ground vehicles.

9/9/2024 GRK Gupta, MED, NITW 26


EXTERNAL FLOW

9/9/2024 GRK Gupta, MED, NITW 27


External Flow II

9/9/2024 GRK Gupta, MED, NITW 28


Flow configurations - internal flow
• Fluid flow is confined by walls, partitions, and other
boundaries.
• Viscous effects extend across the entire domain.
• Examples: flows in pipes, ducts, diffusers, enclosures,
nozzles.
car interior

Flow through duct

airflow
9/9/2024 GRK Gupta, MED, NITW 29
9/9/2024 GRK Gupta, MED, NITW 30
converging-diverging nozzle with a distributed inlet ports

Physical Domain

Computational Domain
Choosing the appropriate boundary conditions can reduce the
computer effort. In this example the slice shown in Figure 1 is
repeated to produce the whole physical domain. Using the periodic
boundary condition at the imaginary planes shown in Figure 2 can
reduce the computational domain to a much smaller area.
9/9/2024 GRK Gupta, MED, NITW 31
BOUNDARY CONDITIONS - I

• Dirichlet Conditions (I Kind) – If the dependent variable


along the boundary is prescribed – ϕ = f(x,y,z)
• Neumann Conditions (II Kind) – If the normal gradient of
the dependent variable is specified - dϕ/dn = f(x,y,z)
• Robin Conditions ( III Kind) – If the imposed boundary
condition is a linear combination of the Dirichlet and
Neumann types – aϕ+bdϕ/dn = f(x,y,z)
• Cauchy Conditions – If the dependent variable and its
derivative are specified. ϕ = f(x), dϕ/dn = g(x)
• Mixed Conditions – If a portion of the boundary is the
Dirichlet type and another portion is of Neumann type

9/9/2024 GRK Gupta, MED, NITW 32


Boundary Conditions 2
1. Solid Walls
2. Inlets
3. Outflows
4. Symmetry
5. Cyclic or Periodic
6. Free surface or Interfaces
7. Far field

9/9/2024 GRK Gupta, MED, NITW 33


Boundary Conditions 3

9/9/2024 GRK Gupta, MED, NITW 34


Boundary Conditions 4
Typical Boundary Conditions
No-slip(Wall), Axisymmetric, Inlet, Outlet

No-slip walls: u=0,v=0

Inlet ,u=c,v=0 Outlet, du/dx=0


dv/dy=0,dp/dx=0
r
v=0, dp/dr=0,du/dr=0
o x
Axisymmetric

9/9/2024 GRK Gupta, MED, NITW 35


Boundary Conditions 5
Flow through a Turbine Cascade

9/9/2024 GRK Gupta, MED, NITW 36


DISCRETIZE THE GEOMETRY
STRUCTURED GRIDS
- i,j,k indexing to locate neighboring cells.
- Grid lines must pass all through domain.
- Obviously can’t be used for very complicated geometries.
UNSTRUCTURED GRIDS
– The cells are arranged in an arbitrary fashion.
– No i,j,k grid index, no constraints on cell layout.
- There is some memory and CPU overhead for
unstructured referencing.

9/9/2024 GRK Gupta, MED, NITW 37


UNSTRUCTURED GRIDS

9/9/2024 GRK Gupta, MED, NITW 38


Grid types: hybrid
• Hybrid grid.
– Use the most appropriate cell type in any combination.
• Triangles and quadrilaterals in 2D.
• Tetrahedra, prisms and pyramids in 3D.
– Can be non-conformal: grids lines don’t need to match at
block boundaries. prism layer efficiently
tetrahedral resolves boundary
volume mesh layer
is generated
automatically

triangular surface
mesh on car body is non-conformal
quick and easy to interface
create
9/9/2024 GRK Gupta, MED, NITW 39
ADAPTIVE GRIDS

9/9/2024 GRK Gupta, MED, NITW 40


Adaption example: final grid and
solution

2D planar shell - final 2D planar shell - contours of pressure


grid final grid

9/9/2024 GRK Gupta, MED, NITW 41


Methods of Structured Grid Generation

1. Algebraic Methods
2. PDE Methods
a. Elliptic Methods
b. Hyperbolic Methods
c. Parabolic Methods
3. Conformal Mapping methods

9/9/2024 GRK Gupta, MED, NITW 42


Methods of Unstructured Grid Generation

Steps:
1. Domain Nodalization
2. Domain Triangulation
A. Advancing Method
B. Delaunay Triangulation

9/9/2024 GRK Gupta, MED, NITW 43


DISCRETIZE THE EQUATIONS
• F.D.M – Differential Approach – Only for
Structured Grids
• F.E.M – Integral Approach – Highly
successful for Structural Analysis
Problems – not well suited for large
problems and turbulent flows
• F.V.M – Integral Approach – Used in all
commercial softwares – Best suited for
any kind of fluid dynamic problem
9/9/2024 GRK Gupta, MED, NITW 44
Explicit versus Implicit Methods

9/9/2024 GRK Gupta, MED, NITW 45


Explicit Methods
In explicit methods, the solution at the next time step is directly computed using
known values from the current time step. This means the equations are solved in a
straightforward, step-by-step manner without the need for solving large systems of
equations.
Key Characteristics:
•Time-stepping: The future state of the system (e.g., velocity, pressure) is calculated
explicitly from the known current state.
•Stability: Explicit methods are usually conditionally stable, meaning that the time
step size (Δt) must satisfy certain criteria for the solution to remain stable. The most
common condition is the Courant-Friedrichs-Lewy (CFL) condition, which
restricts Δt based on the flow velocity and grid spacing (Δx).
•Computational Cost: Each time step is relatively cheap to compute because no
large systems of equations are solved. However, due to small time step sizes required
for stability, many iterations may be necessary, making the overall simulation time
longer.
•Ease of Implementation: Explicit methods are easier to program and implement
since they don’t involve solving complex systems of equations at each time step.
9/9/2024 GRK Gupta, MED, NITW 46
Implicit Methods
In explicit methods, the solution at the next time step is directly computed using
known values from the current time step. This means the equations are solved in a
straightforward, step-by-step manner without the need for solving large systems of
equations.
Key Characteristics:
•Time-stepping: The future state of the system (e.g., velocity, pressure) is calculated
explicitly from the known current state.
•Stability: Explicit methods are usually conditionally stable, meaning that the time
step size (Δt) must satisfy certain criteria for the solution to remain stable. The most
common condition is the Courant-Friedrichs-Lewy (CFL) condition, which
restricts Δt based on the flow velocity and grid spacing (Δx).
•Computational Cost: Each time step is relatively cheap to compute because no
large systems of equations are solved. However, due to small time step sizes required
for stability, many iterations may be necessary, making the overall simulation time
longer.
•Ease of Implementation: Explicit methods are easier to program and implement
since they don’t involve solving complex systems of equations at each time step.
9/9/2024 GRK Gupta, MED, NITW 47
9/9/2024 GRK Gupta, MED, NITW 48
Comparison Between
Explicit and Implicit Methods
Aspect Explicit Method Implicit Method
Small (limited by CFL Large (unconditionally
Time Step Size
condition) stable)
Stability Conditionally stable Unconditionally stable
Computational Cost per High (due to solving
Low
Step systems)
Large number (due to Fewer steps (due to large
Number of Time Steps
small Δt) Δt)
More complex (requires
Ease of Implementation Simple
solvers)
For non-stiff problems or
For stiff problems or when
when high precision is
Preferred Use large time scales are
needed in capturing fast
needed.
phenomena.
9/9/2024 GRK Gupta, MED, NITW 49
9/9/2024 GRK Gupta, MED, NITW 50
Iterative Matrix Solvers used in CFD
a. Jacobi Method
•An iterative method where each variable is updated independently in each iteration
using the values from the previous iteration.
•Advantages: Simple to implement and parallelize.
•Disadvantages: Convergence can be slow, especially for ill-conditioned matrices.
b. Gauss-Seidel Method
•Similar to the Jacobi method, but updates each variable as soon as a new value is
computed, which often leads to faster convergence.
•Advantages: Faster convergence than Jacobi for many problems.
•Disadvantages: May not converge for all types of systems, particularly for non-
diagonally dominant matrices.
c. Successive Over-Relaxation (SOR)
•An extension of the Gauss-Seidel method that introduces a relaxation factor to
accelerate convergence.
•Advantages: Faster convergence compared to both Jacobi and Gauss-Seidel for
certain problems.
•Disadvantages: Requires tuning the relaxation factor for optimal performance, and
it might not work well for all systems.
9/9/2024 GRK Gupta, MED, NITW 51
Thomas Algorithm (TDMA)
TDMA (Tri-Diagonal Matrix Algorithm), also known as the Thomas
algorithm, is a specialized solver for tri-diagonal matrices, which frequently
arise in finite difference methods when solving 1D problems (or 2D/3D
problems that can be reduced to a 1D form, such as when solving along one
direction at a time).
A tri-diagonal matrix is a matrix where only the main diagonal, the
diagonal above it, and the diagonal below it contain non-zero elements. This form
of matrix appears commonly when discretizing differential equations, especially
when using finite difference approximations of diffusion or heat conduction
equations.
TDMA is a simple and highly efficient algorithm for solving tri-diagonal systems,
commonly used in CFD when dealing with finite difference approximations of 1D
or line-based 2D/3D problems. Its speed and low memory usage make it an ideal
choice for solving diffusion-type equations in structured grid systems.

9/9/2024 GRK Gupta, MED, NITW 52


POST-PROCESSING OF RESULTS
• Most Important step in Project Execution
• Gives insights of system you are studying
• Needs good understanding of problem for
interpreting the results
• Different post-processing tools can be used
– xy Plots
– Streamline Plots
– Colour Plots
– Scatter Plots
– Contour Plots
– Flooded Contour Plots or Gray Scale Colour Maps
– Vector Plots
– Mesh Plots
9/9/2024
– Animations etc. GRK Gupta, MED, NITW 53
STREAMLINE PLOTS

9/9/2024 GRK Gupta, MED, NITW 54


XY PLOTS
LIFT AND DRAG Vs ANGLE OF
ATTACK

9/9/2024 GRK Gupta, MED, NITW 55


Velocity vectors around a dinosaur

9/9/2024 GRK Gupta, MED, NITW 56


Velocity – Flooded Contours or
Gray Scale Colour Maps

9/9/2024 GRK Gupta, MED, NITW 57


COLOUR PLOT

9/9/2024 GRK Gupta, MED, NITW 58


Velocity Vector Plot

9/9/2024 GRK Gupta, MED, NITW 59


VELOCITY VECTOR PLOT

9/9/2024 GRK Gupta, MED, NITW 60


PRESSURE CONTOUR PLOTS

Transonic aerofoil M6 WING

9/9/2024 GRK Gupta, MED, NITW 61


SCATTER PLOTS

9/9/2024 GRK Gupta, MED, NITW 62


MESH PLOTS

9/9/2024 GRK Gupta, MED, NITW 63


9/9/2024 GRK Gupta, MED, NITW 64
COMPOSITE PLOTS

9/9/2024 GRK Gupta, MED, NITW 65


ANIMATIONS – FLOW OVER CYLINDER

9/9/2024 GRK Gupta, MED, NITW 66


ANIMATIONS-TWO CYLINDERS -
STREAMWISE

9/9/2024 GRK Gupta, MED, NITW 67


ANIMATIONS – FOUR CYLINDERS

9/9/2024 GRK Gupta, MED, NITW 68


ANIMATIONS – DIFFUSER

9/9/2024 GRK Gupta, MED, NITW 69


ANIMATIONS – FLOW OVER STEP

9/9/2024 GRK Gupta, MED, NITW 70


Smoke ring

9/9/2024 GRK Gupta, MED, NITW 71


Fluid Structure Interaction
Flow over a cable

9/9/2024 GRK Gupta, MED, NITW 72


DamBreak Box

9/9/2024 GRK Gupta, MED, NITW 73


Fluidized Bed

DPM stands for Discrete Particle Method - Lagrangean


9/9/2024 GRK Gupta, MED, NITW 74
ANIMATIONS – CABLE
FLUID STRUCTURE INTERACTION

9/9/2024 GRK Gupta, MED, NITW 75


Some Bench Mark Problems in CFD
1. Lid-Driven Cavity Flow – Incompressible Boundary Layers
2. Flow Over a Backward Facing Step – Turbulent flow with separation and
reattachment
3. NACA 0012 Airfoil Simulation – External flows (subsonic, transonic)
4. Flow Over a Cylinder - Vortex shedding and drag
5. Rayleigh-Bénard Convection – Natural convection
6. Taylor-Couette Flow – Transition flow
7. Taylor-Green Vortex – Compressible turbulent flow
8. Channel Flow (Plane Poiseuille Flow) – Turbulence models with wall
conditions
9. Shock Tube Problem (Sod’s Shock Tube) – Compressible flows with
shocks
10.Karman Vortex Street – Unsteady flows
11.Driven Oscillating Flow in a Pipe – Unsteady periodic flow
12.2D Kelvin-Helmholtz Instability – Shear layer instabilities
9/9/2024 GRK Gupta, MED, NITW 76

You might also like