Computational flow modeling

Partha S. Goswami

Department of Chemical Engineering

I.I.T. Bombay

Importance
Fluids related to life

Human body 65% water

Earths surface is 2/3 water
Atmosphere extends 17km above the earths surface
Without fluids !!
Weather & climate
Vehicles: automobiles, trains, ships, and planes, etc.
Environment
Physiology and medicine
Sports & recreation
Many more and more !!!!

Weather and Climate

Thunderstorm

Tornadoes

Hurricanes

Global Climate

K-H in the cloud

Environment

Air pollution

Transportation
Aircraft

High-speed rail

Surface ships

Submarines

Physiology and Medicine

Blood pump

Sports & Recreation

Water sports

Auto racing

Cycling

Offshore racing

Surfing

A brief history of .
Who made the life complicated ?

Archimedes

(C. 287-212 BC)

Navier

(1785-1836)

Newton

(1642-1727)

Stokes

(1819-1903)

Leibniz

(1646-1716)

Reynolds

(1842-1912)

Bernoulli

Euler

(1667-1748)

(1707-1783)

Prandtl

Taylor

(1875-1953)

(1886-1975)

First part of the 20th century

Theories of boundary layers and turbulence

Ludwig Prandtl (1875-1953)
Boundary layer theory, the mixing
length concept, compressible flows,
the Prandtl number, and more
Theodore von Karman (1881-1963)
Analyzed and described von Karman vortex
street
Geoffrey Ingram Taylor (1886-1975)
Statistical theory of turbulence and
Taylor micro-scale
Andrey Nikolaevich Kolmogorov (1903-1987)
The Kolmogorov scales and the universal
energy spectrum
George Keith Batchelor (1920-2000)
Contributions to the theory of homogeneous
turbulence

Analytical Fluid Dynamics

The theory of mathematical physics problem formulation

Control volume & differential analysis
Exact solutions only exist for simple geometry and
conditions
Approximate solutions for practical applications
Linear
Empirical relations using EFD data

Analytical Fluid Dynamics

Example: laminar pipe flow
UD
2000
Assumptions: Fully developed, Low Re

Approach: Simplify momentum equation,

integrate, apply boundary conditions to
determine integration constants and use
energy equation to calculate
head loss
0

Schematic

2u 2u
Du
p

2 2 gx
Dt
x
y
x

Exact solution :

u(r) 1 ( p)(R2 r 2)
4 x
8 du

8 w dy w 64
f

Friction factor:
V 2 V 2 Re
11

Analytical Fluid Dynamics

Example: turbulent flow in smooth pipe
Three layer concept (using dimensional analysis)

u u u
1.

u* w

0 y 5

Overlap layer (viscous and turbulent shear important)

u

3.

Laminar sub-layer (viscous shear dominates)

u y
2.

y yu

ln y B

20 y 105

(k=0.41, B=5.5)

Outer layer (turbulent shear dominates)

Assume log-law is valid across entire pipe:

U u
r
5

f
1

y 10
*
u
r0

u r
u*

r0 r u *

ln
B

Integration for average velocity and using EFD data to adjust constants:

Experimental Fluid Dynamics (EFD)

Definition:
Use of experimental methodology and procedures
Full and model scales, large and table top facilities,
Measurement systems (instrumentation, data acquisition and data
reduction), uncertainty analysis, and dimensional analysis and
similarity.

What to do?

Design of set up
test design
determination of error sources
estimation of uncertainty
documentation of the results

EFD: requirements
Instrumentation
Load cell to measure forces and moments
Pressure transducers
Pitot tubes
Hotwire anemometry
PIV, LDV
Data acquisition
Serial port devices
Desktop PCs
Plug-in data acquisition boards
Data Acquisition software Labview

Data analysis and data reduction

Data reduction equations
Spectral analysis

In 1970s : Journey of computational fluid dynamics starts

During the 1960s the theoretical division at Los Alamos contributed many
numerical methods that are still in use today, such as the following methods
Particle-in-Cell (PIC).
Marker-and-Cell (MAC).
Vorticity stream function methods.
Arbitrary Lagrangian Eulerian (ALE).
k- turbulence model

Overview of computational (fluid dynamics/ flow modeling)

What is computational flow modeling
Science of predicting fluid flow, heat transfer, mass transfer, chemical
reactions, and related phenomena
by
solving the mathematical equations which describe these processes
using
concept of numerical methods.
Why CFM is required?

Conceptual studies of new designs.

Detailed product development.
Troubleshooting of existing design and redesigning.
Computational flow analysis complements testing and experimentation
- reduces the total cost and effort required for experiments

How does computational flow modeling work ?

1. Complete mathematical model of a physical problem
Conservation of matter, momentum, and energy must be satisfied
throughout the region of interest.
Fluid properties are modeled empirically.
Simplifying assumptions are made: makes the problem tractable
(steady-state, incompressible, inviscid, two-dimensional).
Provide appropriate initial and boundary conditions for the problem.

2. Numerical methods (called discretization) to develop approximations of the

governing equations in the fluid region of interest.
Domain is discretized into a finite set of control volumes or cells :grid, mesh

General conservation (transport) equations for mass, momentum,

energy, etc., are discretized into algebraic equations
All equations are solved to render flow field

3. The set of algebraic equations are solved numerically for the flow field
variables at each node or cell.
4. The solution is post-processed to extract quantities of interest (e.g. lift,
drag, torque, heat transfer, separation, pressure loss, etc.).

CFD Examples

3D vortex shedding behind a circular

cylinder (Re=100,DNS,J.Dijkstra)

LES of a turbulent jet. Back wall shows a slice of

the dissipation rate and the bottom wall shows a
carpet plot of the mixture fraction in a slice
through the jet centerline, Re=21,000 (D. Glaze).

CFD simulation of a Reactor

Few more examples

Transport Processes
Momentum transfer (Fluid Mechanics)
Mass Transfer (Purification, Separation)

Heat Transfer
Chemical reaction

Reactors:
Stirred tank
Tubular
Packed bed and fluidized bed
reactors

Separation units
Distillation
Filtration
Cyclone Separator
Filtration

Complexities
Simple Equation but difficult to solve

Geometry of the equipment

Spatial variation of the properties
Random behavior

Simplification by Mathematical models

Simple physical models

Good physical insight

Procedures in CFD
Identification of right approximation: Viscous/Inviscid, Laminar/Turbulent,
Incompressible / compressible, Single-phase/multi-phase
Identification of right solution method: Finite Element / Difference/Volume,
Structured/Unstructured mesh, Order of accuracy
Pre-processing: Generate computational grid, assign boundary conditions,
set initial conditions, compile code, prepare input parameters
Solution: Run the code, monitor the solution
Post-processing: Collect and organize data, analyze results
Verification: Do the results make sense? Are the trends right? Does it agree
with previous calculations on similar configurations?
Validation: Does the result (or an aspect of the result) agree with
theory/experiment?
At every step, good understanding of theoretical fluid dynamics is
essential!!!

Basic CFD strategies

Finite difference
Use differential form of the equations
The computational is covered using grids
DE is solved at each grid points
Mainly used in structured grids
Conservation is not always guaranteed
Restricted to simple geometries

Finite volume

Uses integral form of governing equation

Domain is covered by control volume
Conservation equations are applied to each CV
Suitable for complex geometry
Method is conservative

o Most widely used

o Most difficult to implement higher that 2nd order

Finite Element
Domain is broken into elements
Equations are multiplied by weight function
before they are integrated
Useful for complex geometry
Not often used un CFD

Mesh generation
CFD Road Map

Specify the problem

Select turbulence model

Generation of Mesh/grid
Discretize equation
Solve discretize equations
Post Processing

Why do we care?
50% time spent on mesh
generation
Convergence depends on mesh
Accuracy depends on mesh
Main topics
Structured/unstructured mesh
body fitted
Adaptive mesh generation

Computational domain and mesh structure

Careful selection of computational domain

The mesh needs

to be able to resolve the boundary layer

to be appropriate for the Reynolds number
to suit the turbulence models selected
to be able to model the complex geometry

Structure/unstructured mesh
Structured grid
The volume elements (quadrilateral in 2D) are
well ordered and a simple scheme (e.g., i-j-k
indices) can be used to label elements and
identify neighbours.

Unstructured grid
Volume elements (triangular or quadrilateral in
2D) can be joined in any manner, and special
lists must be kept to identify neighbouring
elements

Adaptive mesh generation

Adaptive mesh generation
The mesh is modified according
to the solution of the flow

Two types of adaptive methods

Local mesh refinement
Mesh re-distribution

Dynamic adaptive method

Mesh refinement/redistribution
are automatically carried out
during iterations
Example: flow past a cylinder

Equation discretization
Specify the problem
Select turbulence model
Generation of Mesh/grid
Discretize equation
Solve discretize equations
Post Processing

Relevant issues
Convergence strongly depends
on numerical methods used.
Accuracy discretization
errors
Topics
Staggered/collocated variable
arrangement
Convection schemes
Accuracy
Artificial diffusion
Boundedness
Choice of many schemes

Pressure-velocity link
Linearization of source terms
Boundary conditions

Staggered/collocated variable arrangement

Collocated variable arrangement

All variables are defined at nodes

Staggered variable arrangement
Velocities are defined at the faces and scalars are
defined as the nodes
V

U,V,P,T
P,T

Collocated Arrangement

Staggered/collocated variable arrangement

The problem:

Unless special measures are taken, the collocated arrangement

often results in oscillations
The reason is the weak coupling between velocity & pressure

Staggered variable arrangement

Almost always been used between 60s and early 80s

Still most often used method for Cartesian grids
Disadvantage: difficult to treat complex geometry

Collocated variable arrangement

Methods have been developed to over-come oscillations in the 80s

and such methods are often being used since.
Used for non-orthogonal, unstructured grids, or, for multigrid
solution methods

Convection schemes
Examples of convection schemes

1st order schemes:

Upwind scheme (UW): most often used scheme!

Power law scheme
Skewed upwind

Higher order schemes

Central differencing scheme (CDS) 2nd order

Quadratic Upwind Interpolation for Convective Kinematics
(QUICK) 3rd order and very often used scheme

Pressure-velocity coupling
The problem
Pressure appears in the momentum equation as the
driving force for the flow. But for incompressible flows,
there is no transport equation for the pressure.
In stead, the continuity equation will be satisfied if the
appropriate pressure field is used in the momentum
equations
The non-linear nature of and the coupling between, the
various equations also pose problems that need care.
The remedy
Iterative guess-and-correct methods have been proposed

Pressure-velocity coupling
Most widely used methods

SIMPLE (Semi-implicit method for pressure-linked equations)

A basic guess-and-correct procedure

SIMPLER (SIMPLE-Revised):

used as default in many commercial code

Solve an extra equation for pressure correction (30% more effort
than SIMPLE). This is normally better than SIMPLE.

SIMPLEC (SIMPLE-Consistent): Generally better than SIMPLE.

PISO (Pressure Implicit with Splitting of Operators)
Initially developed for unsteady flow
Involves two correction stages

Boundary conditions

Specification of boundary conditions (BC) is a very

important part of CFD modelling
In most cases, this is straightforward but, in some
cases, it can be very difficult ...,

Typical boundary conditions:

Inlet boundary conditions

Outlet boundary conditions
Wall boundary conditions
Symmetry boundary conditions
Periodic boundary conditions

Solvers
Direct methods

Gauss elimination:
Tridiagonal Matrix Algorithm (TDMA):

Indirect methods

Basic methods:
Jacobi
Gauss-Seidel
Successive over-relaxation (SOR)
ADI-TDMA
Strongly implicit procedure (SIP)
Conjugate Gradient Methods (CGM)
Multigrid Methods

Convergence criteria

Basic methods:

Changes between any two iterations are less than

a given level
Residuals in the transport equations are less than a
given value

Criteria can be specified using absolute or relative

values

Under-relaxation
Under almost all circumstances, iterations will not converge
unless under-relaxation is used, because
The governing equations are very non-linear
And the equations are closely coupled

Under-relaxation ():
Different variables often require different levels of under-relaxation

Iteration diverged? Relaxation is the first thing to look at

Unsteady flow solvers

Explicit method

use only the values of the variable from last time step.
Conditionally stable, first order

Implicit method

Mainly use the values of the variable from the current

time step
Unconditionally stable, first order

Crank-Nicolson method

Use a mixture of values of the variable at the last and

current steps
Unconditionally stable, second order

Predictor-Corrector method

Predictor: Explicit method

Corrector: (Pseudo-) Crank-Nicolson method

Validation of CFD modeling

Different types of errors
Discretization errors
Depends on the order of the scheme used. Higher order scheme
normally help to reduce such error
Effect of mesh size. Reducing mesh size will normally help to
reduce such error.
Iteration errors:
For converged solutions, such errors are small
Turbulence modeling
Some turbulence models are better than others under certain flow
conditions! No turbulence model claim to work for all types of flows.
Physical problem vs mathematical problems
Approximation of boundary conditions
Use of a 2D or axisymmetric models
Simplification of the treatment of properties

Validation of CFD results

contd..

CFD results always need validation.

Compared with experiments
Compared with analytical solutions (for simple cases)
Checked by intuition/common sense
Compared with other codes (Only for coding validation)

How to use a CFD software?

Mesh independence check
Selection of an appropriate turbulence model
Validation of the solution with simplified problem (which contains
important features similar to the problem under consideration)
Careful interpretation of the results

Syllabus
Introduction and principles of conservation
Classifications of partial differential equation and physical behavior

Fundamentals of discretization and boundary conditions:

Finite difference and finite volume method

Dealing with unsteady state problems: consequences of discretization

Overview of solution of algebraic equation

Solution of convection-diffusion equation: different schemes
Solution of N-S equation
Introduction to turbulence modeling
Overview of modeling multiphase flow

References:
1. S. V. Patankar, Numerical heat transfer and fluid flow
2. Versteeg and Malalasekera, An introduction to computational
fluid dynamics
3. Ferziger and Peric, Computational methods for fluid dynamics
4.
5.

Tennehill, Anderson, and Pletcher, Computation fluid

mechanics and heat transfer
J. D. Anderson, Computational fluid dynamics

Assessments:
Mode

% weight

Assignments

20

Mid term exam

20

Grades
AA

: Absolute

(AB, BB, -----) : Relative

Project

30

End-sem Exam

30

Assignments: You have to write your own codes to solve the

problem (Fortran, C/C++, Matlab)
Project: You can use ANSYS/ COMSOL/ OpenFOAM
Requirements for AU: Assignments AND project

A theory is something nobody believes,

except the person who made it.
An experiment is something everybody
believes, except the person who made it.