CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Siemens (
-   -   Moderately porous media (

Jing March 7, 2002 09:33

Moderately porous media
According to the suggestion of the StarCD manual, I call the user subroutine SORMOM.f and solve the full momentum equations in order to simulate a moderately porous media.

I got the expected velocity value in the middle of the porous media. But the velocity at the inlet and the outlet of the porous media are smaller than the value in the middle. The flow field before and after the poros media has higher velocity again. I have hoped that I should get a gradually decreased velocity. Not like currently results: higher-lower-higher-lower-higher

Have some people had the same problem? Your opinions are appreciated.


john March 8, 2002 15:03

Re: Moderately porous media
The pressure gradient is very different at the two sides of the porous-nonporous interface. With mometum source, this "discontinuity" in pressure gradient is not considered properly, resulting in the "decoupling" problem you described. If you use STAR's internal porous media treatment, the problem will not exist.

Jing March 11, 2002 05:32

Re: Moderately porous media
Thanks, john! Your opinions are always very helpful for me. In my calculation, alpha= 70000 and Beta = 0. Could I use the internal porous treatment of StarCD for the moderately porous media?

john March 11, 2002 05:50

Re: Moderately porous media
I think when you see noticeable decoupling, then the mometum source treatment is not no longer good. Of course, the Darcy's law only applies when the following condition is valid:

(ALPHA*U+BETA)*L/(density*U) >>1 where L and U are the characteristic length and velocity scales of the medium. It is also advisable to NOT use BETA=0. because ALPHA*U could go to zero in reverse flow, leading to numerical instability.

july March 12, 2002 04:00

Re: Moderately porous media
i'm not a great user of star cd but setiing beta = 0 leads to instability... refer to the methodology manual

ml March 12, 2002 09:25

Re: Moderately porous media
I once set beta=1*e-4 and alpha=~1000 and the run did converge and the result seems reasonable. So I would suggest you define beta as a very small value and alpha=70000, see it it works.

My point is although it is not recommended to set beta=0 which may lead instability, it doesn't hurt to have a try.

Jing March 13, 2002 05:51

Re: Moderately porous media
Thanks for all the suggestions! I will try the internal porous method of StarCD in my case.

Joern Beilke March 13, 2002 07:58

Re: Moderately porous media
You can try "switch 149" if you run into problems.

july April 8, 2002 09:28

Re: Moderately porous media
I'm interested also in porous media cause i'm working in Fuel Cell modelling. My question is: what kind of problem have you modelled? In my case i've a straight channel with a very thin porous layer above.

All times are GMT -4. The time now is 18:41.