COMPUTATIONAL

FLUID DYNAMICS

Dr. Sana Zulfiqar

COPY RIGHT

• The material presented here is taken from various books,

research papers, and other online resources. The Copyright of
such material belong to the original authors and publishers.

2
COURSE LEARNING OUTCOMES (CLO)

• An underlying understanding of the theoretical basis of CFD.

• The ability to develop CFD model for various engineering
problems.
• The technical ability to address complex problems using CFD
with the specific focus on developing practical skills in using a
commercial CFD package, ANSYS FLUENT.
• The ability to interpret computational results and to write a
report conveying the result of the computational analysis.

3
 COURSE OUTLINE
3. Discretization Methods
1. Introduction
• Finite Difference Methods
• What is CFD?
• Finite Volume Method
• Comparison with other approaches
• Concepts of consistency, stability and
• CFD Solution Procedure convergence
2. Governing Equations • Application of Discretization Methods to Model
• Derivation of Mass, Momentum and Equations
Energy equations 4. Mesh Generation
• Conservative and Non-conservative • Types of meshes
Forms – General Transport Equation
• Generation methods
• Boundary conditions
5. Turbulence Modeling
• Mathematical behavior of PDEs and its
impact on CFD 6. Solution of Navier-Stokes Equation, Temporal
Discretization, Iterative Methods, Acceleration
• Physical Behavior, Marching, and Techniques
Equilibrium Problems
7. Applications to various types of fluid flow problems –
ANSYS FLUENT 4
GRADING POLICY
Assessment Weightage
Assignment/Project 10%
Quiz 15%
Mid Term Exam 30%
End Term Exam 45%

5
 REFERENCE/TEXT BOOKS
• Anderson, J. D. Computational Fluid Dynamics—The Basics with Applications.
McGraw-Hill.

• Versteeg, H.K. and Malalasekera, W. An Introduction to Computational Fluid

Dynamics: The Finite Volume Method. Longman Scientific & Technical.

• Jiyuan Tu, Guan-Heng Yeoh, Chaoqun Liu, Computational Fluid Dynamics: A

Practical Approach

• Tannehill, J.C., Anderson, D.A., and Pletcher, R.H. Computational Fluid Mechanics
and Heat Transfer. Second edition. Taylor & Francis.

• Ferziger, J.H. and Peri´c, M. Computational Methods for Fluid Dynamics. Springer.

6
PRE-REQUISITES
• Fluid Mechanics
• Numerical Methods
• MATLAB – Programming
• FEM Analysis

7
INTRODUCTION TO
COMPUTATIONAL
FLUID DYNAMICS
CFD Analysis
WHAT IS CFD?
• Computational fluid dynamics or CFD is the analysis of systems
involving fluid flow, heat transfer and associated phenomena such
as chemical reactions by means of computer-based simulation.
OR
• CFD is a branch of fluid mechanics that uses numerical methods
and algorithms to solve and analyze problems involving fluid flows.
• It is a powerful tool used in engineering and scientific research to
simulate fluid behavior in various applications, from aerospace and
automotive engineering to biomedical and environmental sciences.

9
WHAT IS CFD?
• Computational fluid dynamics or CFD is the analysis of systems
involving fluid flow, heat transfer and associated phenomena such
as chemical reactions by means of computer-based simulation.
OR
• CFD is a branch of fluid mechanics that uses numerical methods
and algorithms to solve and analyze problems involving fluid flows.
• It is a powerful tool used in engineering and scientific research to
simulate fluid behavior in various applications, from aerospace and
automotive engineering to biomedical and environmental sciences.

10
Different disciplines contained in CFD
WHAT IS CFD?
• Fluid Mechanics is the study of fluids either in motion (fluid in dynamic state) or
at rest (fluid in stationary mode).
• CFD is particularly dedicated to the fluids that are in motion, and how the fluid
flow behavior effects the process that may include heat transfer or chemical
reactions in the combusting flow – this is related to the fluid dynamics.
• The physical characteristics of the fluid motion are described through the
fundamental mathematical equations (usually PDEs) which govern a process of
interest and are called as governing equations.
• To solve these equations, high-level computer programming or software
packages are used.
• The computational part shows the study of the fluid flow using numerical
simulations that require several software packages on high-speed digital
computers.

12
WHY CFD IS IMPORTANT?
• Predict Fluid Behavior: CFD allows engineers to analyze fluid flow, heat
transfer, and turbulence without physical experiments.

• Reduce Costs & Improve Design: By simulating different conditions,

CFD helps optimize designs and reduce the need for expensive
prototypes.

• Apply to Real-World Problems: CFD is used in aerodynamics, weather

forecasting, industrial processes, biomedical applications, and more.

13
Methods to Solve Problems in Fluid Dynamics and Heat Transfer

Computational Fluid Dynamics (CFD)

problems can be solved using three
primary approaches: Experimental,
Analytical, and Numerical Methods.

Each method has its own strengths and

limitations, and the choice depends on
factors such as problem complexity,
required accuracy, computational
resources, and practical constraints.
COMPARISON OF THREE APPROACHES
• Experimental methods provide the most realistic results but are
expensive and time-consuming.
• Analytical solutions are useful for simple cases but fail for complex
geometries and turbulent flows.
• Numerical methods (CFD simulations) offer a balance between
accuracy and efficiency, making them the most widely used in
engineering applications.
• For most real-world CFD problems, a combination of all three approaches
is ideal: Analytical models provide theoretical insight, numerical
simulations allow for detailed analysis, and experimental data is used
for validation and accuracy assessment.
15
COMPARISON OF THREE APPROACHES
• Sometimes it is difficult to distinguish between the different methods.
• For example, when numerically computing turbulent flows, the eddy
viscosity models that are frequently used are obtained from experiments.
• Likewise, many theoretical techniques that employ numerical calculations
could be classified as computational approaches.

16
 Approach Basic Principle Advantages Disadvantages Common Applications
• Expensive and time-
• Provides real-world validation. • Wind tunnel testing
Uses laboratory consuming.
• Accounts for all physical for aerodynamics.
experiments to • Limited control over
Experimental phenomena, including • Hydraulic modeling
measure fluid all parameters.
(Physical turbulence and multiphase of water flow.
behavior using • Difficult to measure
Testing) flows. • Heat exchanger
techniques like wind some flow
• No need for complex performance
tunnels. properties in
assumptions. testing.
complex conditions.
• Limited to simple
• Provides exact solutions for
Solves fluid flow geometries and • Laminar flow over
simple cases.
equations (e.g., idealized flat plates.-
Analytical • Fast computations without
Navier-Stokes) using conditions. Poiseuille flow in
(Mathematical needing simulations.
exact or approximate • Cannot handle pipes.
Models) • Useful for fundamental
mathematical complex turbulence • Boundary layer
understanding and
methods. or real-world approximations.
benchmarking.
irregularities.
• Requires
computational
• Can model complex resources.
Uses numerical • Aircraft and car
geometries and boundary • Accuracy depends
methods (FDM, FVM, aerodynamics.
Numerical conditions. on discretization,
FEM) to approximate • Blood flow in
(Computation • Handles turbulence, multiphase meshing, and solver
and solve fluid flow arteries.- Heat
al Simulations) flow, and transient effects. settings.
equations on a transfer in industrial
• Cost-effective compared to • Needs validation
computer. systems.
experiments. against
experimental or
analytical results.
17
EXAMPLE (AERODYNAMIC
ANALYSIS OF AN AIRFOIL)
1. A real-world example
that involves
experimental, analytical,
and numerical
approaches is the
aerodynamic analysis of
an airfoil (e.g., an
aircraft wing section).
The goal is to
determine lift, drag, and
flow behavior around
the airfoil at different
angles of attack.

18
EXPERIMENTAL APPROACH
• Method: Conduct wind tunnel tests to obtain real-
world lift, drag, and pressure distribution data.
• Setup: Place a scaled airfoil model in a wind tunnel.
Measure lift and drag forces using force sensors. Use
smoke flow visualization or Particle Image Velocimetry
(PIV) to study airflow. Measure pressure distribution
using pressure taps along the airfoil surface.
• Advantages: Provides real-world validation for
analytical and numerical models. Captures flow
separation, turbulence, and boundary layer effects.
Essential for certification and performance validation.
• Limitations: Expensive and time-consuming (requires
specialized equipment). Results may be affected by
scaling effects if the test model is not full-size.

19
ANALYTICAL APPROACH
• Method: Solve simplified potential flow equations
(e.g., Bernoulli’s equation, Kutta-Joukowski theorem) to
estimate lift and circulation around the airfoil.
• Procedure: Use thin airfoil theory to estimate lift
coefficient.
CL=2πα where CL is the lift coefficient and α is the angle of
attack (in radians).

Assumes inviscid, irrotational flow, meaning effects like

viscosity and turbulence are ignored.
• Limitations: Works well for thin, symmetric airfoils at
low angles of attack. Fails at high angles of attack
where turbulence and flow separation occur (e.g., stall).

20
NUMERICAL APPROACH
• Method: Use Computational Fluid Dynamics (CFD) to
solve the Navier-Stokes equations numerically for real-
world conditions.
• Process: Create a 2D or 3D model of the airfoil. Generate
a computational mesh around the airfoil. Define boundary
conditions (e.g., freestream velocity, turbulence
model).Use CFD solvers (e.g., ANSYS Fluent, OpenFOAM)
to compute flow properties. Analyze results using velocity
contours, pressure distribution, and lift/drag coefficients.
• Advantages: Captures viscous effects like boundary layers
and turbulence. Provides detailed flow visualization (e.g.,
vortex shedding, wake structures). Allows parametric
studies without physical modifications.
• Limitations: Requires high computational power,
especially for turbulent flows. Needs validation against
experiments for reliability.
21
INTEGRATION OF ALL THREE APPROACHES
Steps Approach Used Purpose
Validate CFD and analytical
1 Experimental Testing (Wind Tunnel) results, capturing real-
world effects.
Quick estimation of lift and
Analytical Model (Thin Airfoil Theory,
2 pressure distribution for
Bernoulli’s Equation)
initial understanding.
Predict detailed flow
Numerical Simulation (CFD using
3 behavior, turbulence, and
FVM, FEM methods)
aerodynamic forces.

• The analytical approach gives a quick and simple estimation.

• The numerical approach (CFD) provides detailed insights into complex
flow behaviors.
• The experimental approach ensures real-world validation and improves
accuracy.
• Combining all three methods leads to a more reliable and
comprehensive aerodynamic analysis.
22
COMPUTATIONAL COST AND
PERFORMANCE OF CFD

23
COMPUTATIONAL COST AND PERFORMANCE

• Computational Fluid Dynamics (CFD) simulations require

significant computational resources, and their cost and
performance depend on multiple factors such as problem
complexity, numerical methods, hardware specifications,
and solver settings.

24
FACTORS AFFECTING CFD COMPUTING
COST & PERFORMANCE
• (A) Problem Complexity
• Geometry Complexity
• Simple 2D problems (e.g., flow over a flat plate) require less computation.
• Complex 3D geometries (e.g., airfoil, human arteries) increase computational cost.
• Flow Regime
• Laminar flow is computationally cheaper.
• Turbulent flow (using RANS, LES, or DNS) is computationally expensive.
• Time Dependency
• Steady-state simulations (e.g., using SIMPLE algorithm) are faster.
• Transient (time-dependent) simulations (e.g., LES or DNS) require high computational power.

25
FACTORS AFFECTING CFD COMPUTING COST
& PERFORMANCE
• (B) Numerical Approach & Solver Type
• Finite Volume Method (FVM)
• Most widely used in commercial CFD solvers (ANSYS Fluent, OpenFOAM).
• Efficient and cost-effective for industrial applications.
• Finite Element Method (FEM)
• Common in structural-fluid interaction problems (COMSOL, Abaqus).
• Computationally expensive but highly accurate.
• Finite Difference Method (FDM)
• Used in academic research but less flexible for complex geometries.
• Turbulence Modeling
• RANS (Reynolds-Averaged Navier-Stokes): Low cost, fast, but lower accuracy.
• LES (Large Eddy Simulation): Moderate cost, better accuracy.
• DNS (Direct Numerical Simulation): Extremely expensive but most accurate.
26
FACTORS AFFECTING CFD COMPUTING COST
& PERFORMANCE
• (C) Hardware Requirements
• Processors (CPUs vs. GPUs)
• CFD solvers require multi-core CPUs (e.g., Intel Xeon, AMD EPYC) for parallel
computing.
• GPUs (e.g., NVIDIA A100, RTX 4090) accelerate simulations (especially in AI-assisted
solvers like TensorFlow-based CFD).
• Memory (RAM)
• Large simulations need 64GB–1TB RAM for high mesh resolutions.
• Storage (SSD vs. HDD)
• High-speed SSDs reduce data read/write times.
• Cluster Computing (HPC Systems)
• Supercomputers & cloud computing (AWS, Google Cloud) reduce runtime for large
simulations. 27
STRATEGIES TO REDUCE CFD COST &
IMPROVE PERFORMANCE
• Choose the Right Turbulence Model.
• Optimize the Mesh Size.
• Distribute simulations across multiple cores (HPC clusters) for
faster results.
• Use Cloud Computing Instead of buying expensive hardware,
rent HPC time on AWS, Google Cloud, or Azure.
• Use Open-Source Software OpenFOAM or SU2 eliminates
licensing costs.

28
APPLICATIONS OF
CFD
Research Tool

Education Tool

Design Tool

29
CFD AS A RESEARCH TOOL

30
CFD AS A DESIGN TOOL (AEROSPACE)

31
CFD AS A DESIGN TOOL (AUTOMOTIVE)

32
CFD AS A DESIGN TOOL (BIOMEDICAL)

33
CFD AS A DESIGN TOOL (CHEMICALS &
MINERALS)

Simulation of gas-sparged stirred tank reactor 34

CFD AS A DESIGN TOOL (CIVIL &
ENVIRONMENT)

35
CFD AS A DESIGN TOOL (METALLURGY)

CFD application to predict molten iron flow (left, timeline) and carbon
dissolution (right, concentration) in the blast furnace hearth.
36
CFD AS A DESIGN TOOL (NUCLEAR)

Prediction of turbulent mixing in the

ROCOM test facility during boron dilution
transients (start-up of the first coolant
pump). 37
CFD AS A DESIGN TOOL (POWER
GENERATION)

Simulation of bubbles in fluidized coal bed.

38
CFD AS A DESIGN TOOL (SPORTS)

39
CFD SOLUTION
PROCEDURE

40
STEPS
Pre-processing
Solving
Post processing

41
PRE-PROCESSING
Creation of Geometry

43
PRE-PROCESSING
Mesh Generation

44
PRE-PROCESSING
Mesh Generation

45
PRE-PROCESSING

Selection of physics and fluid properties

46
PRE-PROCESSING
Specification of boundary conditions

47
NUMERICAL SOLUTION

• Initialization and Solution Control

• Good initial conditions are crucial to
the iterative procedure
• Solution control usually entails the
specification of appropriate
discretization (interpolation) schemes
and selection of suitable iterative
solvers

48
NUMERICAL SOLUTION
• Monitoring Convergence
• monitoring solution, CFD calculation,
and checking for convergence and grid
independence

49
NUMERICAL SOLUTION
• Aspects such as
• convergence,
• convergence criteria or tolerance values,
• residuals,
• stability,
• errors,
• underrelaxation factors, and
• grid independence
greatly underpin the many numerical considerations
for the simulation of a CFD problem.

50
POST PROCESSING

• CFD has a reputation of generating vivid graphical images.

• The ability to present the computational results effectively is an
invaluable design tool.
• Integrated and independent postprocessors are available.

51
POST PROCESSING
• X-Y Plot

52
POST PROCESSING
• Vector Plot

53
POST PROCESSING
• Contour Plot

54
POST PROCESSING
• Animation

55
COMMERCIAL CFD CODES
• What to expect and when to use commercial package:

• The current generation of CFD packages generally is capable of producing accurate

solutions of simple flows.

• The codes are designed to be able to handle very complex geometries and complex
industrial problems. When used with care by a knowledgeable user CFD codes are an
enormously valuable design tool.

• Commercial CFD codes are rarely useful for state-of-the art research due to accuracy
limitations, the limited access that the user has to the solution methodology, and
• the limited opportunities to change the code if needed
COMMERCIAL CFD CODES
Major current players include

Ansys (Fluent and other codes)

http://www.ansys.com/

Simens: (starCD and other codes)

http://mdx.plm.automation.siemens.com

Others

CHAM: http://www.cham.co.uk/

CFD2000: http://www.adaptive-research.com/

COMSOLE: https://www.comsol.com/cfd-module

OpenFOAM: http://www.openfoam.com

XFLOW: https://www.3ds.com/products-services/simulia/products/xflow/
ONLINE ANSYS
FLUENT RESOURCES
https://www.youtube.com/playlist?list=PL0lZXwHtV6OmedTRo0x
U5AFxCrtsSsHda

https://www.youtube.com/watch?v=1FazGZwiN_w&list=PLEpES_
WTfjdzrqcIv_QS1mrSF3VcUqyBO

https://www.youtube.com/watch?v=8rqvYTkXj7Q&list=PLykANdm
ho34zcPPHJ8NSDl8XeUB70jd1A

https://www.youtube.com/watch?v=slJOVbHST8w&list=PLykANd
mho34xaJPdplVkAgv2lL7PLNisk

https://www.youtube.com/watch?v=ttHmvYuVHWU&list=PLQMtm
0_chcLyEB1EJLUoouugMYZNQUslG
58