Scribd
Upload
117 views

NIT, Hamirpur: CFD: Developments, Applications & Analysis

This document outlines the topics that will be covered in a series of invited lectures on computational fluid dynamics (CFD) at the National Institute of Technology in Hamirpur. The lectures will cover: (1) an introduction to CFD including definitions, roles, and major aspects; (2) how CFD works, including grid generation, the finite volume method, and solution methodology; and (3) why CFD is important. Some of the key points that will be discussed are the role of CFD software as a virtual video camera, the three conservation laws that CFD solves, and the increasing importance of CFD in both industry and academics.

Uploaded by

Ankesh Kapil
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
117 views

NIT, Hamirpur: CFD: Developments, Applications & Analysis

This document outlines the topics that will be covered in a series of invited lectures on computational fluid dynamics (CFD) at the National Institute of Technology in Hamirpur. The lectures will cover: (1) an introduction to CFD including definitions, roles, and major aspects; (2) how CFD works, including grid generation, the finite volume method, and solution methodology; and (3) why CFD is important. Some of the key points that will be discussed are the role of CFD software as a virtual video camera, the three conservation laws that CFD solves, and the increasing importance of CFD in both industry and academics.

Uploaded by

Ankesh Kapil
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 10

1/23/2014

1
FUNDAMENTS OF CFD
1.
33:45 PM
Introduction
2.
3:45 AM 5 PM
Computational
Heat Transfer
Development
Seri es of Invi ted Lecture on
CFD: Developments, Applications & Analysis
NI T, Hami rpur
Topi c I : I ntroducti on
1.
CFD:
What is it?
3.
CFD:
Why?
2.
CFD:
How it is ?
Seri es of Invi ted Lecture on
CFD: Developments, Applications & Analysis
NI T, Hami rpur
Contents
01. CFD: What is it?
Definition
Role
Major Aspects
03. CFD: Why
02. CFD: How is it?
Grid Generation
Finite Volume Method
Solution Methodology
1/23/2014
2
01. CFD: What is it?
Definition
Role
Major Aspects
1
CFD: Defination
Computational Fluid Dynamics (CFD) is a theoretical
method of scientific and engineering investigation,
concerned with the development and application of a
video-camera like tool - a software - which is used to
analyze a fluid dynamics as well as heat and mass
transfer problemin a unified way.
Here, the software is like a virtual video-camera and
results in a movie where each picture gives a fluid-
dynamics information, i.e., flow-properties.
Vortex Shedding Fluid Flow
across a Circular Cylinder at Re=100
Vortex Shedding Fluid Flow
across a Circular Cylinder at Re=100
1/23/2014
3
Vorticity Contours at Re=100 Group of Fish Like Locomotion
Dam Break Simulation
2D Film-Boiling Over a Superheated Plate
1/23/2014
4
Granular Flow CFD: Role
Analogy between a video-camera and a CFD
software.
Analytical (CFD) solution are like a virtual video
camera of infinite (finite) spatial as well as temporal
resolution.
The role of CFD software is not limited to creating
scientifically exciting fluid dynamic movie of flow
properties but also to create engineering relevant
movie of engineering parameters, for a unified
cause-and-effect study of various heat and fluid
flow situations.
CFD: Major Aspects
There are three major aspects of CFD
Development
Application
Analysis
With a continuous development and a wider
application of CFD, the definition of Navier-
Stokes equations and CFD has broaden.
02. CFD: How is it?
Grid Generation
Finite Volume Method
Solution Methodology
2
1/23/2014
5
CFD: How is it?
Solution
Methodology
It consists of
1. Solution Method: A
method to solve the
set of algebraic
equations
Explicit and Implicit Method
for unsteady state formulation
Iterative method for steady
state formulation.
2. Implementation Details
3. Solution Algorithm
Grid-Generation
A method to convert the complete
domain into certain fixed number
of Control Volumes (CVs). The
points located at the centroid of
the CVs are called as grid points.
They are of two types:
Cartesian, Cylindrical and Spherical
Grid: used for simple geometry.
Curvilinear-Structured and
Unstructured Grid: used for complex
geometry.
Finite Volume Method
A method to obtain a system
of algebraic equations, with
unknowns as flow properties
at the grid points, obtained by
Physics based Approach:
Applying conservation laws to the CVs
and using certain approximations.
Mathematics based Approach:
Applying volume integral to the
governing partial differential equations,
using gauss divergence theoremfor
certain terms and using certain
approximations.
L
2
L
1
i
j
1 imax
1
jmax
Boundary CVs:
i=1 & imax;
j=2 to jmax-1
j=1 & jmax;
i=2 to imax-1
P
N
S
E
W
Interior CVs:
i=2 & imax-1;
j=2 to jmax-1
j=2 & jmax-1;
i=2 to imax-1
Outer Square Boundary
Plate
Inner Square Boundary
Square Hole
Both the Boundaries are
aligned along direction of
Cartesian Coordinates
Heat Transfer in a Square
Plate with a Square Hole
1/23/2014
6
Heat Transfer in a Square
plate with a Circular Hole
The Outer Square
Domain Boundary is
aligned along the
direction of Cartesian
Coordinates
Inner Circular Domain
Boundary is aligned
along the direction of
cylindrical coordinates
Complex Domain & Unstructured Grids
x
y
Pressure
Force
Pressure
Force in
X-direction
Pressure
Force in
Y-direction
ZERO
Body Force
or Heat
Generation
Body Force
in
X-direction
Body Force
in
Y-direction
Heat
Gained by
Heat
Generation
Viscous
Force or
Heat
Conduction
Viscous
Force in
X-direction
Viscous
Force in
Y-direction
Heat
Gained by
Conduction
ACROSS
the CV
X-
Momentum
Y-
Momentum
Energy
Conservation Laws: Mass, Momentum & Energy
INSIDE the
CV
Mass
X-
Momentum
Y-
Momentum
Energy
Rate of Change
Unsteady Advection Diffusion Source
Mass Zero
X-Mom.
Energy
Y-Mom.
Mass
1/23/2014
7
CFD: How is it?
Fi ni te Volume Method for Computational Fluid Dynamics
A Novel
Physics
based
Control
Volume
The
Traditional
Mathematics
based
Two Levels of
Approximations
Governing
PDEs
Gauss Div. Theorem
Approximations
V A }}}
Discretized LAEs
(Linear Algebraic Equations)
Fluid
Mechanics
& Heat
Transfer
Course
limV 0
N
a
v
i
e
r
-
S
t
o
k
e
s
E
q
u
a
t
i
o
n
s
Mass
Momentum
Energy
C
o
n
s
e
r
v
a
t
i
o
n

L
a
w
s Continuity
Momentum
Energy
T
nb
=400
0
C T
i,jmax
=400
0
C
Grid Points
Internal:
at 250
0
C
Boundary:
at 100/200/
300/400
0
C
T
wb
=100
0
C
T
1,j
=100
0
C
jmax=7
6
5
4
3
2
j=1
x
y
T
eb
=300
0
C
T
imax,j
=300
0
C
T
sb
=200
0
C T
i,1
=200
0
C
i=1 2 3 4 5 6 imax=7
n+1
n
Expli ci t Method
Unsteady State Heat Conduction: Computational Stencil
1
, , ,
n n
P P nb nb
nb E W N S
a T a T b
+
=
= +

CFD: How is it?


1/23/2014
8
Soluti on Methodology for
Unsteady Computati onal heat Conducti on
A Novel Physics based
The Traditional
Mathematics
based
S i n g l e S t e p
( )
( ) ( )
( ) ( )
P P
E E W W
N N S S
a T t t
a T t a T t
a T t a T t
b
+ A
~ +
+ +
+
( )
P
T t t + A
x x
q y
+A
A
y y
q x
+A
A
x F i r s t S t e p :
( ) ( ) ( ); ( ) ( ) ( )
( ) ( ) ( ); ( ) ( ) ( )
x x E E P P x P P W W
y y N N P P y P P S W
q t c T t c T t q t d T t d T t
q t e T t e T t q t f T t f T t
+A
+A
~ + ~ +
~ + ~ +
S e c o n d S t e p :
( ) | |
( )
( ) ( ) ( ) ( )
( ) ( )
conduction x x x y y y
P P P conduction
Q t q t q t y q t q t x
T t t T t g Q t
+A +A
( ~ A + A

+ A ~ +
x
q y A
y
q x A
N
P E W
S
y
03. CFD: Why
3
CFD: Why?
Now a days, a computer simulation and analysis have
become an integral part of a design and optimization study.
CFD is more commonly used as a powerful analysis, than a
design and optimization, tool by scientists as well as
engineers - dealing with fluid dynamics and heat-transfer
problems; in various industry such as aerospace,
automobile, turbo-machinery, chemical, electronics
cooling, biomedical, etc.
CFD: Why?
In academics, CFD is taught as an under-graduate
elective and post-graduate course, in different
branches of engineering.
The increasing importance of CFD development,
application and analysis, in the industry as well as
research organizations, along with the lack of trained
manpower in this area has greatly increased the
importance of this course.
1/23/2014
9
Brief History & Modern Trends
2
Brief History of Developments of CFD
Difficulties in Solving for Pressure
Discretization Method, Geometry and Grid
Modern Trends in CFD
Cartesian Grid Method
Multilevel Cartesian Grid Method
Parallel Computing
Industrial Applications
Difficulties in Solving for Pressure
Three conservation laws based equations (mass,
momentum and energy) and three dependent variables
(pressure, velocity and temperature).
Law of conservation of momentum and energy have a
term as rate of change of momentum and energy
No rate of change of pressure term in the mass
conservation law
No explicit equation for pressure and continuity equation
needs to be converted into an equation for pressure.
Difficulties in Solving for Pressure
This resulted in a stream-function vorticity method and was one
of the first popular method in CFD - limited to 2D and steady-
state problems.
To circumvent these limitations, a pressure-velocity formulation
called as pressure correction approach was proposed where the
continuity equation was employed as a constraint to derive
correct pressure field - through a detailed iterative solution
procedure.
However, there was an issue of pressure velocity decoupling
and the first remedy suggested was solution on a staggered
grid.
1/23/2014
10
Discretization Method, Geometry & Grid
Early development of CFD started with FDM for problems - called as
simple geometry problem.
The Cartesian/Cylindrical/Spherical geometry problems were solved in
a uniform or nonuniformgrid, where one of the grid line fits to the
boundary of the computational domain -called as body fitted grid.
With advances in CFD and more application to industrial problems,
there was a need to develop method for computing flows in complex
geometry.
Initially, this was attempted with FDM on a body-fitted (a) Cartesian
and (b) curvilinear grid
Discretization Method, Geometry & Grid
This led to development of FVM where the governing PDEs
are solved directly in the physical complex/curved domain, on
body-fitted curvilinear structured grid.
However, implementation of staggered grid was found almost
impossible for complex geometry and another remedy to
obtain physical solution in co-located grid was proposed as
usage of momentum interpolation.
This continued for quite some time with further development
from the FVM based solution in structured to multi-block
structured to unstructured body-fitted grid system.
Thank You for Your attention
Welcome for any Questions,
Comments, and Suggestions

You might also like