You are on page 1of 18

# 1.

## Governing equations -- mass, momentum, energy, species

General form of the scalar transport equation
Elliptic, parabolic and hyperbolic equations
Behavior of the scalar transport equation with respect to these equation
types

## Mesh terminology and types

Methods for deriving discretization equations
Methods for solving discretization equations
Accuracy, consistency, stability and convergence

3. Diffusion Equation

## Discretization of conduction equation

Source terms, non-linearity
Boundary conditions
Under-relaxation
Solution of linear equations (preliminary)
Other coordinate systems
Two-dimensional conduction on unstructured meshes
Accuracy, stability and convergence revisited

## Steady one-dimensional convection and diffusion

Upwind scheme
Two-dimensional convection-diffusion
Accuracy of upwind scheme; false diffusion
Higher-order schemes
Boundary conditions
Generalization to unstructured meshes

5. Linear Solvers

Gauss-Seidel scheme
Tri-diagonal matrix algorithm (TDMA), line-iterative solvers
Multi-grid methods and application to unstructured meshes
6. Flow Field Calculation

## Incompressibility issues and pressure-velocity coupling

Primitive variables versus other methods
Staggered grid for structured meshes
SIMPLE family of algorithms
Non-staggered mesh methods
Body-fitted formulations
Unstructured mesh formulation

7. Projects

Assignments

## Assignment 1 (Due January 24, 2011)

Solution to Assignment 1

## Assignment 2 (Due February 7, 2011)

Solution to Assignment 2

## Assignment 3 (Due February 21, 2011)

Solution to Assignment 3

## Assignment 4 (Due March 7, 2011)

Solution to Assignment 4

## Assignment 5 (Due April 4, 2011)

Solution to Assignment 5

Final Exam

## Final Exam Solution

Mid-Term Exam
Mid-Term Exam Solution

## Mid-Term Exam Statistics

Average: 76.6/100

Max: 99/100

Min: 43/100

Week 1 (1/10/11)

1/10/2011 Lecture 1

1/12/2011 Lecture 2

Week 2 (1/17/11)

## 1/17/2011 Lecture 4: No class (Martin Luther King Day)

1/19/2011 Lecture 5

1/21/2011 Lecture 6

Week 3 (1/24/11)

1/24/2011 Lecture 7

## 1/26/2011 Lecture 8 Source Linearization

1/28/2011 Lecture 9

Week 4 (1/31/11)
1/31/2011 Lecture 10 Lecture 10 (narrated)

## 2/2/2011 Lecture 11 Lecture 11 (narrated)

2/4/2011 Lecture 12

Week 5 (2/7/11)

2/7/2011 Lecture 13

2/9/2011 Lecture 14

2/11/2011 Lecture 15

Week 6 (2/14/11)

2/14/2011 Lecture 16

2/16/2011 Lecture 17

2/18/2011 Lecture 18

Week 7 (2/21/11)

2/21/2011 Lecture 19

2/23/2011 Lecture 20

2/25/2011 Lecture 21

Week 8 (2/28/11)

## 3/4/2011 Lecture 24 Lecture 24 narrated

Week 9 (3/7/11)

3/7/2011 Lecture 25

## 3/11/2011 Lecture 27: Class cancelled because of evening exam

Week 10 (3/14/11)

## 3/18/2011 Lecture 30: No Lecture (Mid-term break)

Week 11 (3/21/11)

## 3/21/2011 Lecture 31 Lecture 31 (narrated)

3/23/2011 Lecture 32

3/25/2011 Lecture 33

Week 12 (3/28/11)

## 3/30/2011 Lecture 35 Problem 6.7-code Problem 6.7-results

4/1/2011 Lecture 36

Week 13 (4/4/11)

## 4/4/2011 Lecture 37 Lecture 37 (narrated)

4/6/2011 Lecture 38

## 4/8/2011 Lecture 39 co-located code co-located output

Week 14 (4/11/11)

4/11/2011 Lecture 40

## 4/13/2011 Lecture 41 Lecture 41 (narrated)

4/15/2011 Lecture 42

Week 15 (4/18/11)

4/18/2011 Lecture 43
4/20/2011 Lecture 44 Example AMG problem

4/22/2011 Lecture 45

Week 16 (4/25/11)

solution

## 4/29/2011 Project Presentations (Morning and Evening) and Closure

Instructor Information

## Prof. Jayathi Y. Murthy

Room 169, Mechanical Engineering, or BRK 2027B
Purdue University
West Lafayette, IN 47907-1288

## (765) 494-5701 (phone)

(765) 494-0539 (fax)

jmurthy@ecn.purdue.edu

Lectures

Assignments

## Bi-weekly assignments , typically consisting of 3-4 problems, will be given

every other Monday. These will be due on Monday two weeks later. You are
free to discuss assignments with your classmates and with me, but the final
submission must be your own work.

Examinations

## One mid-term examination will be held 6-8:00 pm, March 9, 2011 in ME

161. The examination will be open book, open notes. There will be a final
examination, the schedule for which will be announced in due course.

Project

Students will be expected to complete a CFD project during the course. The
project will typically require 3-4 weeks of effort and will involve groups of
two students working together. Suggested project topics will be discussed
by the instructor in class and will be finalized during the middle of March to
allow enough time for project completion. A project report will be required,
and classroom presentations will be scheduled for the end of the semester.

## We will be using extensive course notes provided by the instructor. These

are posted in PDF format on the class webpage. You are free to print and
browse them. A useful reference book for the class is:

1980.

Office Hours

## Office hours will be held on Mondays 10:30 am-11:30 am in ME 169

Mechanical Engineering. Please make an appointment if this time is not
convenient or if you need more time.

Assignments : 30 %
Mid Term Examination: 25 %
Final 25%
Project: 20 %
Total: 100 %

A mixture of absolute and relative grading will determine the final grade for
the class.

Project Topics

## A list of suggested project topics is given here.

Important Dates

Project reports are due on the last day of class, April 29, 2011.

Project Presentations

TBA

## The project report format should follow ASME conference publication

guidelines. A sample paper is shown below.

## Reports should be restricted to 8-12 single-spaced pages.

Please submit a paper copy to Prof. Murthy, along with an annotated copy of

ME 608

## 1. Multimode Heat Transfer in a Cavity

The project focuses on developing a solver combining surface-to-surface radiation and
conduction. Consider a square cavity containing a transparent gas such as air. The walls of the
cavity are cooled on the outside by convection.The gas itself is quiescent, and heat transfer
through it is by conduction alone. The walls, however, are hot enough that they radiate to each
other through the gas. Develop a finite volume solver to compute enclosure radiation heat
transfer by writing the radiosity equations for all wall control surfaces, finding view factors, and
solving the resulting linear system to find the wall temperatures, which are then used as
boundary conditions for the conduction problem in the gas. Explore the influence of enclosure
geometry and Planck number on the ratio of conduction to radiation heat transfer from the walls.

## 2. Algebraic Multigrid Scheme for Structured and/or Unstructured Meshes

In class, we will develop an algebraic multigrid scheme for unstructured meshes. This project
involves implementing the multigrid idea either using a Gauss-Seidel relaxation sweep or, for
structured meshes, an LBL-TDMA, and tying this to a multigrid cycle. Explore the performance
of the scheme for problems with large anisotropies in coefficients due to properties, mesh aspect
ratios and other complexities.

## 3. Phonon Boltzmann Transport Equation Solver

In recent years there has been a great deal of interest in solving the phonon Boltzmann trasnport
equation to model micro-scale conduction heat transfer. In this project, take the convection
schemes we developed in class and apply them to solve the phonon BTE. Show that you can
recover Fourier conduction in macro domains, and explore the numerical properties of your
solver over a range of acoustic thicknesses.

## 4. Solver for Chemically Reacting Systems

Given a flow field, write a solver to solve for species transport and chemical reaction in a
reacting system, taking care to properly linearize and couple your species transport equations.
Explore the performance of your numerical algorithm for a range of governing parameters.

Develop a solver for coupled thermal transport and participating radiation. Participating radiation
is described by the radiative transfer equation (RTE), which is amenable to solution using the
ideas about convection-diffusion equations developed in class. The RTE is coupled to the energy
equation through radiative source terms. Develop a solver for the coupled system, and test your
solver against a variety of published solutions.

## 6. Combined Lagrangian-Eulerian Solver for Particle Transport Through a Gas

Solvers for spray combustion and particle transport through gases and liquids sometimes employ
a coupled Lagrangian-Eulerian method for low solid/droplet volume fraction. In this approach,
droplets or particles traversing the fluid are tracked individually in a Lagrangian frame of
reference. Particle tracks are then located in a background mesh on which the gas phase
governing equations are solved. In general, the momentum, mass and energy loss by the particles
is that gained by the gas. The project involves implementing a coupled Lagrangian-Eulerian
solver within the framework of a finite volume scheme.

## 7. SIMPLE Solver for Flow in a Driven Cavity

Implement the SIMPLE algorithm for sequential solution of the incompressible continuity and
momentum equations within a finite volume framework. Establish that your code works against a
variety of published solutions and examine the convergence properties of your scheme using the
driven cavity problem as a test problem. This is a substantial project and will require good
programming skills, but you will learn a great deal of CFD by doing it.

## 8. Two-Temperature Model for Porous Media

Thermal transport in porous media is sometimes treated using a "two temperature" formulation.
If the thermal properties of the solid and fluid media are very different from each other, the two
media cannot be assumed to be in thermal equilibrium. One approach to modeling this type of
situation is to pretend that each point in the medium is described by two temperatures, one for
the solid and one for the fluid, and write two separate energy equations, accounting for the
volume occupied by each medium, and coupling them through energy exchange terms. The
objective of the project is to write a solver for this class of problem using a finite volume
framework and testing the properties of the scheme against analytical and published solutions.

## 9. Effective Conductivity of Foams

The objective of this project is to develop a solver for computing the effective thermal
conductivity of metallic foams impregnated with either a solid such as paraffin, or immersed in a
fluid such as air or water. For the purposes of this project, you may assume that the fluid is
stationary, and further, assume that the foam can be represented by a regular structure that can be
modeled in Cartesian coordinates. Develop a solver to address this coupled problem and present
the effective thermal conductivity of the medium as a function of geometry and metal-substrate
conductivity ratio.

## 10. Effective Conductivity of Particle Composites

The objective of this project is to develop a solver for computing the effective thermal
conductivity of particulate composites, where particles are embedded in a substrate such as
polymer. The substrate is stationary. Assume that the bed can be represented by a regular or
random matrix of cuboidal particles that can be modeled in Cartesian coordinates. Develop a
solver to address this coupled problem and present the effective thermal conductivity of the
medium as a function of geometry and particle-substrate conductivity ratio.

## 11. Convection-Diffusion Solver Using Control Volume-Based Finite Element Method

In this class, we have looked mainly at cell-based schemes. An alternative is to develop node-
based schemes. During the 1980's, a variety of node-based schemes using the so-called control-
volume finite element methods (CVFEM) were developed which sought to combine the
conservative property of finite volume schemes with the geometric flexibility of finite element
schemes. The project involves developing a CVFEM solver for the convection-diffusion
equation and to examine its properties.

## 12. Unstructured Mesh Solver for the Convection Diffusion Equation

The objective of this project is to implement the scheme that we have developed in this class,
working out all the issues with respect to data structures and solution algorithms. Test your
solver against published solutions and examine the numerical properties of the underlying
schemes.

## 1) Partial Differential Equations by W. E. Schiesser, very good

http://www.lehigh.edu/~wes1/apci/28apr00.pdf

## 2) Computational Fluid Dynamics in Hydraulic and Sedimentation Engineering by Nils

Reidar B. Olsen http://www.bygg.ntnu.no/~nilsol/

## 5) Applied Computational Fluid Dynamics ME 448/548: Lecture Materials

http://www.me.pdx.edu/~gerry/class/ME448/

## 6) Applied Computational Fluid Dynamics 5C1213 by KTH

http://www.mech.kth.se/new/list_courses.asp

## 7) Computational Fluid Dynamics of Turbulent Flow by Lars Davidson

http://www.tfd.chalmers.se/gr-kurs/MTF071

## 8) Fire Safety Engineering http://http-server.carleton.ca/~ghadjiso/index.html

10) ME 608 Numerical Methods in Heat, Mass and Momentum Transfer by Jayathi Y.
Murthy very good!!!!! http://widget.ecn.purdue.edu/~jmurthy/me608/

## 11) Numerical Computing with MATHLAB by Cleve Moler

http://www.mathworks.com/moler/

## 12) CVEN 302-503 Computer Applications in Engineering and Construction by Hamn-

Ching Chen

13) CVEN 680-600 Advanced Computation Methods for Fluid Flow by Hamn-Ching
Chen

## 14) CVEN 688-600 Computational Fluid Dynamics by Hamn-Ching Chen

http://ceprofs.tamu.edu/hchen/

## 15) MAE672-1 Computational Fluid Dynamics by University of Virginia

https://toolkit.itc.virginia.edu/cgi...ring_MAE672-1/

## 16) Advanced Topics in Aerodynamics – Computational Aerodynamics

http://www.aerodyn.org/CFD/

## 19) courses by David D. Apsley (Dr) all are very good

http://personalpages.umist.ac.uk/staff/david.d.apsley/

## 20) Numerical Solutions Applied to Heat Transfer and Fluid Mechanics

http://www.personal.psu.edu/faculty/j/h/jhm/ME540/

## 21) 58:160 Intermediate Mechanics of Fluids, College of Engineering, The University of

Iowa http://css.engineering.uiowa.edu/~me_160/

## 24) Computational Fluid Dynamics by Dr.Eng.Reima Iwatsu - very good

http://www.fml.m.dendai.ac.jp/iwatsu.../Computational%20Fluid%20Dynamics.pdf

## 25) Turbulent Thermal-Fluid Dynamics - Institute of Energy Technology, Department of

Mechanical Engineering, ETH Zurich http://www.lkt.iet.ethz.ch/~lakehal/...4/material.htm

## 26) MTF112 Convective heat transfer by Chalmers http://www.tfd.chalmers.se/gr-

kurs/MTF112/index.html

## 27) SOLUTION METHODS IN COMPUTATIONAL FLUID DYNAMICS by Thomas

H. Pulliam http://people.nas.nasa.gov/~pulliam/...vki_notes.html

## 29) NUMERICAL METHODS FOR PARTIAL DIFFERENTIAL EQUATIONS -

CAAM 452 Lecturer: Dr Tim Warburton http://www.caam.rice.edu/~caam452/

30) METR 5344 CFD Course Home Page (Spring 2000) Instructor: Dr. Ming Xue
http://twister.caps.ou.edu/cfd2000/

31) CSI 756 - Numerical Methods for Climate Dynamics by Professor Paul Schopf
http://www.scs.gmu.edu/climate/courses/csi756/
32) ACM 210a Numerical Methods for PDEs
http://www.acm.caltech.edu/~acm210/2005/WINTER/

## 34) AA510 CFD II by Washington University

http://www.aa.washington.edu/courses/aa510/

## 36) Strömningsmekanik/Fluid Mechanics 5C1214

Strömningsmekaniskaberäkningar/Computational Fluid Dynamics 5C1212 Dan
Henningson, Professor in Fluid Mechanics KTH http://www2.mech.kth.se/~henning/

37) courses by Division of Thermo and Fluid Dynamics at the Department of Mechanical
Engineering, Chalmers University of Technology in Göteborg http://www.tfd.chalmers.se/

REPORTS

A few sample project reports submitted for the final project in ME 608 in
Spring 2006 are posted below. The projects were approximately one month's
effort. Students could select from a list provided by the instructor, or work
on a topic of their own choosing. The use of commercial CFD codes was not
allowed, and each project required code development based on the material
covered in the class.

## Radiative transport in participating media

Boltzmann solver for phonon transport

## Course Organization This course focuses on the finite volume method

Syllabus and its use in the analysis of fluid dynamics, heat
Assignments and mass transfer problems. A
Examinations general framework for the discretization of
Class Notes convection-diffusion equations is developed for
Fluent/Gambit Materials both structured and unstructured meshes. Issues
Announcements of accuracy, stability and convergence are
Project Requirements addressed. Techniques for the solution of linear
Project Reports algebraic systems are presented for both
Lectures structured and unstructured discretizations. The
computation of fluid flow and issues of pressure-
velocity coupling for incompressible flows are
addressed next, with an emphasis on sequential
techniques. The course involves hands-on code
development as well as use of a commercial code
for solving a variety of practical engineering
problems.

Thomas H. Pulliam
NASA Ames Research Center
Moffett Field, CA 94035
email:Thomas.H.Pulliam@nasa.gov

## Days/Times/Classroom: MW 03:15PM-04:30PM BLD: 540 Room

108
Office Hours: Wednesdays 2PM-3:05PM, Location: Lobby of
Durand 2nd Floor
Announcements
General Info:

##  Week 1: Read Chaps 1-2, Homework #1 assigned

 Notes AA214A
Intro Slide AA214A: Lecture 1 [pdf 214Kb]
Euler to Wave Equation Notes : Lecture 2a [pdf 51Kb]
Wave Equation Exact Solution Notes : Lecture 2b [pdf 51Kb]

## Notes from Chapter 3: Finite-Differences / Taylor Tables: Lecture 3 [pdf 114Kb]

Notes from Chapter 3: Compact Schemes [pdf 75Kb]
Taylor table for 6th Order Accurate Generalized Pade Differenceing schemes: Lecture 3 [pdf
5.8Kb]
Notes from Chapter 3: Matrix Form For Difference Schemes : Lecture 4 [pdf 20.1Kb]
Notes from Chapter 3: Modified Wave Number Analysis : Lecture 4 [pdf 6.6Kb]

Homework #2 Assigned

## Continue Lecture 4 (Updated):

Notes from Chapter 3: Modified Wave Number Analysis : Lecture 5 [pdf 6.6Kb]
Chapter 4 Notes: Lecture 5,6 [pdf 917Kb]

## Notes from Chapter 6 [pdf 92Kb]

Homework #3 Assigned

## Notes from Chapter 7 [pdf 138Kb]

Notes from Chapter 8 [pdf 90Kb]

##  The Mid Term from AA214 Fall 2008, Stanford U.

 MIDTERM 2008 (PDF) MID TERM Answers (PDF)
 Week 6 Read Chap 11

## Chap 11 Notes AA214A

Notes from Chapter 11 [pdf 107Kb] {New Notes}

## Chap 12 Notes AA214A

Notes from Chapter 12 [pdf 150Kb] {New Notes}

## Chap 13 Notes AA214A

Notes from Chapter 13 [pdf 62Kb] {New Notes}

Homework #4 Assigned

## FUNDAMENTALS OF COMPUTATIONAL FLUID DYNAMICS

Thomas H. Pulliam
NASA Ames Research Center
Moffett Field, CA 94035

## Notes AA214A Latest Update Sept. 28 2009

Intro Slide AA214A [pdf 15Kb]
Euler to Wave Equation Notes: Lecture [pdf 7.8Kb]
Wave Equation Exact Solution Notes : Lecture pdf 6.9Kb]
Notes from Chapter 3: Finite-Differences / Taylor Tables: Lecture 3 [pdf 114Kb]
Notes from Chapter 3: Compact Schemes [pdf 75Kb]
Taylor table for 4th Order Accurate Pade Differenceing schemes [pdf 3.9Kb]
Taylor table for 6th Order Accurate Generalized Pade Differenceing schemes [pdf 5.8Kb]
Notes from Chapter 3: Modified Wave Number Analysis [pdf 6.6Kb]
Notes from Chapter 3: Matrix Form For Difference Schemes [pdf 20.1Kb]
Chapter 4 Notes: Semi-Discrete Approach [pdf 11.6Kb]
Notes from Chapter 5: Finitie-Volume Methods [pdf 12.1Kb]
Notes from Chapter 6: Time-Marching Methids for ODE's [pdf 36.6Kb]
Notes from Chapter 7: Stability of Linear Systems [pdf 18Kb]
Notes from Chapter 8: Choosing a Time-marching Method[pdf 11.1Kb]
Notes from Chapter 11: Numerical Dissipation [pdf 23Kb]
Notes from Chapter 12: Split and Factored Forms [pdf 23.9Kb]
Notes from Chapter 13: Analysis of Split and Factored Forms [pdf 12.7Kb]
Final Project Notes
Slides for Final Project Discussion [pdf 32Kb]

##  New Matlab Examples

Stability in the Lambda Plane for Various O-Delta-E's
Lambda Eigenvalues/Modified Wave Number for Difference Schemes
Shows various Matrix forms of differencing schemes and their eigenvalues
Matrix forms for various differencing schemes
Example of some time differencing schemes for wave propagation
Driver Matlab code for plotting various Modified Wave Numbers
Matlab code for computing Modified Wave Numbers for Spatial Differencing Schemes

##  Appendices A & B from notes [postscript ] ....[pdf format ]

 Supplemental Handout #1 [postscript 61Kb] ....[pdf format 289Kb]
 Notes on useful series expansions [postscript 21Kb] ....[pdf format 71Kb]
 Notes Compact Pade Differenceing schemes [postscript 468Kb] ....[pdf format 259Kb]

##  Notes on the Euler Equations [postscript 10Mb] ....[pdf format 9.5Mb]

 Numerical Chaos and Iterative Equations [postscript 7.2Mb] ....[pdf format 7.9Mb]
!! Note Page 4 Takes a Long Time to Display

##  Tar file of Notes on the Euler Equations

 Tar file of Numerical Chaos and Iterative Equations

##  Implicit Methods in CFD [postscript 376Kb] ....[pdf format 2.9Mb]

Check out the Matlab codes for plotting Stability in the Sigma-Lambda Planes
A Standalone Matlab code plotting Various O-Delta-E'sin the Complex sigma plane: See Chapter
7 of the notes
A Standalone Matlab code plotting Various O-Delta-E'sin the Complex lambda plane: See
Chapter 7 of the notes
I have also included a GUI version of the some code: grab the next 4 files and put into a single
directory and run "Stabilty" at the matlab prompt
A GUI Matlab code plotting Various O-Delta-E's: See Chapter 7 of the notes
Needed for GUI
Same as sigma_plot_alone.m for use with Stabilty.m
Same as lambda_plot_alone.m for use with Stabilty.m
title matlab file
unpack the unix tar file and make the file accessible to matlab
or