Computational Fluid Dynamics

Session – II
Prof. C. M. Sewatkar

Department of Mechanical Engineering

College of Engineering, Pune
 Fluid Kinematics

 Study of fluid motion without consideration of forces

Method of Study

Lagrangian Eulerian

Lagrangian – individual fluid particle is selected

Eulerian – any point occupied by fluid is selected

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Derivation of Governing Equations
 Differential approach

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Derivation of Governing Equations

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Integral Analysis
Integral Analysis
Integral Analysis
Integral Analysis
Integral Analysis
Integral Analysis
Integral Analysis
Integral Analysis
Integral Analysis
Integral Analysis
Integral Analysis
Integral Analysis
Integral Analysis
Integral Analysis
Integral Analysis
Integral Analysis
Integral Analysis
Integral Analysis
Integral Analysis
Integral Analysis
Derivation of Continuity Equation

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Derivation of Continuity Equation

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Derivation of Momentum Equation

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Derivation of Momentum Equation

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Derivation of Momentum Equation

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Derivation of Momentum Equation

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Derivation of Momentum Equation

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Derivation of Momentum Equation

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Derivation of Momentum Equation

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Derivation of Momentum Equation

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Derivation of Momentum Equation

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Derivation of Momentum Equation

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Derivation of Momentum Equation

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

 Derivation of Momentum Equation
 Flow Kinematics (Continued)

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

 Derivation of Momentum Equation
Expanded form of   u relationship

For details: Journal of Fluids Engineering 117, no. 1, pp. 3–5, 1995
C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune
 Stokes Hypothesis
Extension Newton's idea from simple 1-D flows to multidimensional
flows.
Sum of the normal stresses is zero

The above equation, and the

requirement that the three normal
stresses add up to zero is
called Stokes hypothesis.
Derivation of Momentum Equation

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Derivation of Energy Equation
•LAW OF CONSERVATION OF ENERGY:

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Derivation of Energy Equation

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Derivation of Energy Equation

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Governing Equations: 2-D Navier-Stokes

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Governing Equations: 2-D Navier-Stokes

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Governing Equations: 2-D Navier-Stokes

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Mass Fluxes (Fm)

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Convection Fluxes (Fc=Fm×φ)

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune
Diffusion Fluxes

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Diffusion Fluxes in X-Mom. Eq.
• Viscous Forces

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Diffusion Fluxes in Y-Mom. Eq.
• Viscous Forces

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Diffusion Fluxes in Energy Eq.

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Diffusion Fluxes (Fd)

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Non-Dimensional Governing Equation

• Dimensional G.E

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Non-Dimensional Governing Equation

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Initial and Boundary Conditions

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Initial and Boundary Conditions

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Initial and Boundary Conditions
Upstream boundary Y-axis
Downstream boundary
Uniform flow lx
Top boundary
profile
on upstream C9 CBC
s
boundary C8

Point for bifurcation f i f

C7 d  Uc i  0
Point for bifurcation analysis
analysis
 n
C6
Uo
X-axis ly
C5 Uc - spatially
f i f Centreline through cylinder
averaged
 U 0 i C4
 n streamwise
C3 velocity
Cetreline between two cylinders
No slip at the
C2
surfaces of
cylinder C1 Bottom boundary
Lu Ld

Periodic boundary condition at lateral

boundaries [ fi (1,x) = fi (ly,x) i = 2,5,6; 60
fi (ly,x) = fi (1,x) i = 4,7,8] C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune
Properties

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Non-dimensional Parameters for
Flow and Heat Transfer

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Integral Flow and Heat Transfer
Parameters

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Integral Quantities

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Mathematical Character of PDE
• Second order

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

Classification of Governing
Equations

C. M. Sewatkar, Faculty, Mech. Engg. Dept., COE, Pune

 Conservative and Non-conservative
form of equations

The equations obtained from the The equations obtained from the
finite control volume fixed in space finite control volume moving with
in either integral or partial the fluid in either integral or partial
differential form, are called the differential form, are called the non-
conservation form of the governing conservation form of the governing
equations. equations.
 The Substantial Derivative


u  ui  v j  wk
V  f ( x, y , z , t )
u  f ( x, y , z , t )
v  f ( x, y , z , t )
u  f ( x, y , z , t )
w  f ( x, y, z , t ),
v  f ( x, y , z , t )
w  f ( x, y , z , t )

   ( x, y , z , t )

1  1 ( x1 , y1 , z1 , t1 )  2   2 ( x2 , y2 , z2 , t2 )
 The Substantial Derivative

Taylor series expansion

Equation A

Equation B

LHS is the average time-rate-of change in density of the fluid element as it

moves from point 1 to point 2.
 The Substantial Derivative
In the limit as t2 approaches t1

Dρ/Dt is a symbol for the instantaneous time rate of

change of density of the fluid element as it moves
through point 1. By definition, this symbol is called the
substantial derivative and is further given as,
Approaches to Fluid Dynamics
Equation of fluid mechanics - known for over a
century
Solutions – limited but extremely useful
However these equations are rarely used in direct
design and analysis
Traditionally we are forced to use other methods
like DA
 Application of such methods to complex problems
is ruled out
 Non-dimensionalization is another approach
(Reynolds number) – dangerous to use with large
number of parameters
 Experimental methods – possible for only global
parameters
Approaches to Fluid Dynamics
An alternative - or at least a complementary method - came
with the birth of electronic computers.

The performance-to-cost ratio of computers increased at a

spectacular rate and shows no sign of slowing down.

Further increases in both computing speed and memory of

affordable computers are certain.

Solution of the equations of fluid mechanics on computers

has become so important and it occupies the attention of
perhaps a third of all researchers in fluid mechanics and the
proportion is still increasing. This field is known as
computational fluid dynamics (CFD).
 What is CFD
flows and related phenomena can be described by
partial differential – can not be solved analytically

To obtain numerical solution one has to discretize

these equations (obtain algebraic and approximate
form)

 Solving these discretized form of equations using

numerical methods is CFD

 Accuracy of experiments – tools of measurement

Accuracy of CFD – quality of dicretization

 Components of a Numerical
Solution Method
Mathematical Model
Discretization Method
Coordinate and Basis Vector Systems
Numerical Grid
Finite Approximations
Solution Method
 Convergence Criteria
Properties of Numerical Solution
Methods
Consistency
Stability
Convergence
Conservation
Boundedness
Realizability
 Accuracy
 Components of a Numerical Solution Method:
Mathematical Model

Starting point of any numerical method

May include simplification of exact

conservation laws/equations

Solution method is specific to model

General method is impractical

 Components of a Numerical Solution Method:
Discretization Method
Method of approximating the differential
equations by a system of algebraic equations
for the variables at some set of discrete
locations in space and time

 Important methods are:

Finite Difference Method
Finite Volume Method
Finite Element Method
Other Methods
Spectral Scheme; lattice Boltzmann
method (LBM); boundary element method;
Cellular automata – limited to special class
 Components of a Numerical Solution Method:
Coordinate and Basis Vector Systems

Cartesian, cylindrical, spherical, curvilinear

orthogonal or non-orthogonal coordinate
systems, which may be fixed or moving

Choice depends on the target flow, and may

influence the discretization method and grid
type to be used.
 Components of a Numerical Solution Method:
Numerical Grid
The discrete locations at which the variables
are to be calculated are defined by the numerical
grid

Numerical grid divides the solution domain

into a finite number of sub-domains (elements,
control volumes, grid points etc.).
 Components of a Numerical Solution Method:
Numerical Grid: Structured Grid
Structured grids consist of families of grid
lines with the property that members of a single
family do not cross each other and cross each
member of the other families only once.

1.Simplest grid structure,

since it is logically equivalent
to a Cartesian grid.
2.Used only for geometrically
simple solution domains.
3. difficult to control the
Figure: An example of a structured 2D grid
distribution of the grid points
 Components of a Numerical Solution Method:
Numerical Grid: Types of Structured Grid
H-Type grids: H-type grid which, when
mapped onto a rectangle, has distinct east,
west, north, and south boundaries

The position of any grid point (or control volume) within the domain is
uniquely identified by a set of two (in 2D) or three (in 3D) running
indices, e.g. (i, j, k).
 Components of a Numerical Solution Method:
Numerical Grid: Types of Structured Grid

 O-Type grids: one set of grid lines is "endless"; if the

grid lines are treated as coordinate lines and we follow
the coordinate around the cylinder, it will continuously
increase and, to avoid a problem, one must introduce an
artificial "cut" at which the coordinate jumps from a
finite value to zero.
 Components of a Numerical Solution Method:
Numerical Grid: Types of Structured Grid

 C-Type grids: Points on portions of one grid line

coincide, requiring the introduction of a cut similar to
the ones found in O-type grids. This type of grid is often
used for bodies with sharp edges for which they are
capable of good grid quality.
 Why Structured Grid?

Lead to very efficient numerical methods

High quality for sufficiently simple geometries

Larger grid control when high anisotropy is required

Multi-block approach allows for realistic geometries

 Components of a Numerical Solution Method:
Numerical Grid: Unstructured Grid
• For very complex geometries, the most flexible type of grid
is one which can fit an arbitrary solution domain boundary.
• Could be used with any discretization scheme, but are
best adapted to the finite volume and finite element
approaches.

• May have any shape; nor is there a restriction on the

number of neighbor elements or nodes

• In practice, grids made of triangles or quadrilaterals in 2D,

and tetrahedra or hexahedra in 3D are most often used.
 Components of a Numerical Solution Method:
Numerical Grid: Unstructured Grid
• For very complex geometries, the most flexible type of grid
is one which can fit an arbitrary solution domain boundary.
• Could be used with any discretization scheme, but are
best adapted to the finite volume and finite element
approaches.

• May have any shape; nor is there a restriction on the

number of neighbor elements or nodes

• In practice, grids made of triangles or quadrilaterals in 2D,

and tetrahedra or hexahedra in 3D are most often used.
 Components of a Numerical Solution Method:
Numerical Grid: Unstructured Grid
Disadvantages:
• Irregularity of the data structure.
• Node locations and neighbor connections need be
specified explicitly.
• The matrix of the algebraic equation system no longer has
regular, diagonal structure
• The band width needs to be reduced by reordering of the
points.
• The solvers for the algebraic equation systems are usually
slower than those for regular grids.
 Components of a Numerical Solution Method:
Numerical Grid: Unstructured Grid

Structured Grid Unstructured Grid

Components of a Numerical Solution Method:
Numerical Grid: Unstructured Grid

Figure: Example of a 2D unstructured grid

 Components of a Numerical Solution Method:
Numerical Grid: Block Structured Grid

Two (or more) level subdivision of solution domain

On the coarse level, there are blocks which are

relatively large segments of the domain; their structure
may be irregular and they may or may not overlap.

On the fine level (within each block) a structured grid is

defined. Special treatment is necessary at block
interfaces.
 Components of a Numerical Solution Method:
Numerical Grid: Block Structured Grid

100

80 5

(b)

60 3

2 4
Y

(a)
40
1

20
(c)

0
0 20 40 60 X 80 100 120

Figure: Multi Block Grid for Flow across a Cylinder

 Grid Generation
Grid generation is not be covered here

There is a vast literature devoted to grid generation and

interested reader is referred to books:

Thompson, J.F, Warsi, Z.U.A., Mastin, C.W. (1985):

Numerical grid generation - foundations and applications.
Elsevier, New York

Arcilla, A.S., Hauser, J., Eiseman, P.R., Thompson, J.F.

(eds.) (1991): Numerical grid generation in computational
fluid dynamics and related fields. North Holland,
Amsterdam
 Components of a Numerical Solution Method:
Solution Method
For unsteady flows, methods based on those used for
initial value problems for ordinary differential equations
(marching in time) are used. At each time step an elliptic
problem has to be solved.

Steady flow problems are usually solved by pseudo-time

march- ing or an equivalent iteration scheme. Since the
equations are non-linear, an iteration scheme is used to
solve them. These methods use successive linearization of
the equations and the resulting linear systems are almost
always solved by iterative techniques.

The choice of solver depends on the grid type and the

number of nodes involved in each algebraic equation.
 Components of a Numerical Solution Method:
Convergence Criterion
Finally, one needs to set the convergence criteria for the
iterative method.

Two levels of iterations:

Inner iterations: the linear equation are solved, and

 Outer iterations, that deal with the non-linearity and

coupling of the equations.

Deciding when to stop the iterative process on each level

is important, from both the accuracy and efficiency points
of view.
Properties of Numerical Solution
Methods
Consistency
Stability
Convergence
Conservation
Boundedness
Realizability
 Accuracy
 Properties of Numerical Solution Methods:
Consistency
The discretization should become exact as the grid spacing
tends to zero.
The difference between the discretized equation and the exact
one is called the truncation error.
 It is usually estimated by replacing all the nodal values in the
discrete approximation by a Taylor series expansion about a
single point.
 As a result one recovers the original differential equation plus a
remainder, which represents the truncation error.
 For a method to be consistent, the truncation error must
become zero when the mesh spacing ∆x→ 0 and/or ∆t→ 0 .
Truncation error is usually proportional to a power of the grid
spacing ∆x and/or the time step ∆t.
 If the most important term is proportional to (∆x) n or (∆t)n we call
the method an nth-order approximation;
 n > 0 is required for consistency
 Properties of Numerical Solution Methods:
Stability
A numerical solution method is said to be stable if it does not
magnify the errors that appear in the course of numerical solution
process.

 For temporal problems, stability guarantees that the method

produces a bounded solution whenever the solution of the exact
equation is bounded.

For iterative methods, a stable method is one that does not

diverge.

Stability can be difficult to investigate, especially when boundary

conditions and non-linearities are present.

For this reason, it is common to investigate the stability of a

method for linear problems with constant coefficients without
boundary conditions.
 Properties of Numerical Solution Methods:
Convergence
A numerical method is said to be convergent if the solution of
the discretized equations tends to the exact solution of the
differential equation as the grid spacing tends to zero.

A consistent scheme is useless unless the solution method

converges.

For non-linear problems which are strongly influenced by

boundary conditions, the stability and convergence of a method
are difficult to demonstrate.

Therefore convergence is usually checked using numerical

experiments, i.e. repeating the calculation on a series of
successively refined grids. If the method is stable and if all
approximations used in the discretization process are consistent,
we usually find that the solution does converge to a grid-
independent solution..
 Properties of Numerical Solution Methods:
Conservation
Since the equations to be solved are conservation laws, the
numerical scheme should also - on both a local and a global basis
respect these laws.

This means that, at steady state and in the absence of sources,

the amount of a conserved quantity leaving a closed volume is
equal to the amount entering that volume.

If the strong conservation form of equations and a finite volume

method are used, this is guaranteed for each individual control
volume and for the solution domain as a whole.
 Properties of Numerical Solution Methods:
Accuracy
Numerical solutions always include three kinds of systematic
errors:

1.Modeling errors: difference between the actual flow and the

exact solution of the mathematical model

1.Discretization errors: difference between the exact solution of

the conservation equations and the exact solution of the algebraic
system of equations obtained by discretizing these equations

2.Iteration errors: difference between the iterative and exact

solutions of the algebraic equations systems.
 Discretization Approaches
Finite Difference Method :

Oldest method for numerical solution of PDE's, believed

to have been introduced by Euler in the 18th century.
The starting point is the conservation equation in
differential form.

 The solution domain is covered by a grid.

At each grid point, the differential equation is

approximated by replacing the partial derivatives by
approximations in terms of the nodal values of the
functions.
The result is one algebraic equation per grid node, in
which the variable value at that and a certain number of
neighbor nodes appear as unknowns.
 Discretization Approaches
Finite Difference Method :

Oldest method for numerical solution of PDE's, believed

to have been introduced by Euler in the 18th century.
The starting point is the conservation equation in
differential form.

 The solution domain is covered by a grid.

At each grid point, the differential equation is

approximated by replacing the partial derivatives by
approximations in terms of the nodal values of the
functions.
The result is one algebraic equation per grid node, in
which the variable value at that and a certain number of
neighbor nodes appear as unknowns.
 Discretization Approaches
Finite Difference Method :

Advantages: On structured grids, the FD method is very

simple and effective. It is especially easy to obtain higher-
order schemes on regular grids;

Disadvantage: conservation is not enforced unless special

care is taken. Also, the restriction to simple geometries is
a significant disadvantage in complex flows.
 Discretization Approaches
Finite Volume Method :

Uses the integral form of the conservation equations as its

starting point.

The solution domain is subdivided into a finite number of

contiguous control volumes (CVs), and the conservation
equations are applied to each CV.

At the centroid of each CV lies a computational node at

which the variable values are to be calculated.
 Discretization Approaches
Finite Volume Method :

Interpolation is used to express variable values at the CV

surface in terms of the nodal (CV-center) values.

Surface and volume integrals are approximated using

suitable quadrature formulae.

As a result, one obtains an algebraic equation for each

CV, in which a number of neighbor nodal values appear.
 Discretization Approaches
Finite Volume Method :
Advantages:
The FV method can accommodate any type of grid, so it
is suitable for complex geometries.
The grid defines only the control volume boundaries and
need not be related to a coordinate system.
The FV approach is perhaps the simplest to understand
and to program. All terms that need be approximated have
physical meaning which is why it is popular with
engineers.

Disadvantages:
Methods of order higher than second are more difficult
to develop in 3D
FV approach requires three levels of approximation:
interpolation, differentiation, and integration.