CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   questions about SIMPLE algorithm (by Patankar) (http://www.cfd-online.com/Forums/main/11568-questions-about-simple-algorithm-patankar.html)

 Nutty June 2, 2006 14:53

questions about SIMPLE algorithm (by Patankar)

I try to do some computer codes for SIMPLE algorithm by using equations in "Numerical Heat Transfer and Fluid Flow" by Suhas V. Patankar. But I'm confused with terms in this book. My questions are as following

1. From page 122, I need to solve eqs.(6.8) to (6.10). But I don't understand term "nb". Did they mean neighbors on nodes or CV faces? For an example, in equation (6.8), I have to submit u*nb to obtain u*e, did u*nb mean "uP and uE" or "up and ue" or "uP, uE, and ..."?

2. From page 101, I need to find coefficients in equation (5.62), do I have to use all parameter on the nodes or CV faces? For an example, to calculate aE, I have to use all parameter on "E"(node) or "e"(east CV face)?

I hope there are some kind people feel free to help one stupid student. Thank you

 ztdep June 3, 2006 13:52

Re: questions about SIMPLE algorithm (by Patankar)

Hi:

you can see the pre-coded code instead of develope it by your self!

I can help you if you need! ztdep@yahoo.com.cn

 diaw June 4, 2006 05:12

Re: questions about SIMPLE algorithm (by Patankar)

Looking in Patankar, my takes is as follows:

2. Pg 101 - scalar properties evaluated at cell-centre & cell-centres of neighbouring cells.

1. Pg 122 onwards refers to his 'staggered grid' concept for velocity-determination. Refers to neighbours for staggered cells grid.

To be honest, this method can make anyone become more than a little cross-eyed. I have always seen the staggered-grid technique as being a rather clumsy cludge. You may want to look into 'collocated methods' & alternative more direct procedures.

In the end, the staggered grid tries to overcome a 'checkerboard' pressure grid, which is said to be a non-physical solution. I don't currently happen to agree with Patankar & friends on that one. :)

diaw...

 pipenx January 22, 2015 03:08

did you have the answer ?

I have the same questions about the neighbors coefficients with a staggered grid, did you get the answer?:(

 ztdep January 22, 2015 09:05

Quote:
 Originally Posted by pipenx (Post 528538) I have the same questions about the neighbors coefficients with a staggered grid, did you get the answer?:(
The value of variable is interpolated by its neighbor nodes no matter it is staggered grid or collocated grid. for example:

=0.5*(+)

further questions are wellcomed

 pipenx January 22, 2015 13:22

?

Thank you for your answer. So, for the N neighbor the diffusion coefficient would be Gamma = (Gamma N + Gamma P)*0.5? I ask this because the middle of the face North (and south) for the staggered cell for e is not associated with the cell not staggered.
[IMG] http://thumbs.subefotos.com/4124e187...933985a90o.jpg[/IMG]
In the image, yellow control volume is for v staggered grid and red for u. Diffusion terms are to be calculated in ew=P, ee=E, en=?, es=?