CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   Book (http://www.cfd-online.com/Forums/main/1141-book.html)

Jan Ramboer August 6, 1999 06:20

Book
 
Hello again,

I am looking for a book which deals about boundary conditions in more depth. Since most of the CFD books just devote a few lines about them. Any advises?

Greetings,

Jan Ramboer

John C. Chien August 6, 1999 09:40

Re: Book
 
(1). The boundary conditions are additional conditions applied to the governing equations. (2). So, you must know the governing equations first. If it is a wall boundary condition, for a stationary wall, u,v,w can be zero for stationary wall, it can be u=0,v=0,w=Ut, for a rotating wall, it can be u=U1,v=U2, w=U3 for a moving wall. With a porous wall, u=0, v=Vnormal, w=0, can be used. These additional conditions are then applied to the governing equations. So, as you see, it depends highly on the problem you are trying to solve, or to model. (3). Basically, you need to apply the additional constraints on the flow parameters ( or flow variables, such as u,v,w,rho,T,p,k,e,zmu,time,...etc) at the boundary of the computational domain. It is totally up to you to do so. (4). Whether your boundary conditions will give you the desired soultions or realistic answers, depends on the particular problem and the computational domain you defined. This part is art. It is impossible to define the boundary conditions without knowing the problem you are trying to solve and the computational domain you have selected. You must develop an intimate relationship with the governing equations and the problem first before you can specify what conditions are most suitable in this case. ( when someone says far-field condition, it simply says the boundary location is far away from the point of interest. It is up to you to locate it for your problem. ) (5). The boundary conditions are additional constraints placed on the computational domain you defined for the problem. In order to simulate the solution of the problem more closely to reality, the boundary conditions must also change with the computational domain boundary. In other words, the boundary conditions for a small computational domain will be different from that for a large one in order to obtain the same solution. (6). In short, to define your CFD problem properly, you need to define the boundary conditions. Without the boundary conditions, your problem is not yet defined. Before you can properly define the boundary conditions for your problem, you need to know your problem. (the boundary condition is used to define your problem.) It is hard to write a book without knowing what problem you are trying to solve.

Jan Ramboer August 6, 1999 11:10

Re: Book
 
First of all thanks for the explanation. Secondly, as you say that it is hard to write a book without knowing the problem, isn't there than some literature which talks about how to develop your own B.C. for a specific case?

Greetings,

Jan Ramboer

John C. Chien August 6, 1999 11:56

Re: Book
 
(1). Well, it is possible to find such information. So, you may want to do some journal search when you have the time. (2). On the other hand, when you publish a paper, you don't want the reviewer to ask the questions such as," are you sure that your boundary conditions are not affecting your solutions?", " is your wall function boundary condition implemented correctly?" , or .....etc. (3). Most of the time, the author will keep the information about the boundary conditions to a minimum in their paper to avoid un-necessary questions from the reviewers. (4). My suggestion is: pick a paper, read the formulation, study the boundary conditions related to the problem, step-by-step. This will tell you how each problem is properly defined. (Each CFD paper should have the boundary conditions section included.)

Duane Baker August 8, 1999 22:47

Re: Book
 
Hi,

you want to have a look first at Ferziger and Peric' (see 1 below) and particularly the codes of Peric's which are available via download. Many of the issues are covered there!

For staggered grid CV formulations, you will benefit from reading Versteeg and Malalasekera (see 2. below) as it has lots of diagrams and discussion for various b.c.'s but no code!

There are also some good tips in the chapter by Raithby and Schneider in The Handbook of Numerical Heat Transfer.

Best of luck.....................................Duane

1. Computational Methods for fluid dynamics Joel H. Ferziger and Milovan Peric springer verlag, 1997 corrected 2nd printing isbn 3-540-59434-5

2. "An Introduction to Computational Fluid Dynamics: The Finite Volume Method", H.K. Versteeg and W. Malalasekera, Addison-Wesley Longman Limited, 1995, Reprinted in 1996. ISBN 0-582-218845.

Praveen Chandrashekar August 9, 1999 02:01

Re: Book
 
Like most other things in CFD, there is no definite answer for the choice of boundary conditions. The best way is to read some papers, and try out various options, and choose the one that you find to be the best. You can have look at Charles Hirsch's book, which I am sure contains a lot of information and references. The book by Godlewki and Raviart is very good, but treats only hyperbolic systems.

(1) Hirsch: NUMERICAL COMPUTATION OF INTERNAL AND EXTERNAL FLOWS. John Wiley

(2) Godlewski and Raviart: NUMERICAL SOLUTION OF HYPERBOLIC SYSTEMS OF CONSERVATION LAWS. Springer.


All times are GMT -4. The time now is 09:08.