CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM (https://www.cfd-online.com/Forums/openfoam/)
-   -   missing p.relax() in pisoFOAM (https://www.cfd-online.com/Forums/openfoam/82987-missing-p-relax-pisofoam.html)

vgalindo December 10, 2010 12:00

missing p.relax() in pisoFOAM
 
One can find at the end of simpleFOAM the code line

/ / Explicitly relax pressure for momentum corrector
p.relax ();

but this line is missing in pisoFoam.C.

Why?

best regards,
V. G.

nimasam December 11, 2010 01:59

piso Foam is a transient solution so relaxation can not be true for a transient solution;) look CFD book you can find more about difference between piso and simple algorithm

akidess December 13, 2010 06:04

You can also use relaxation with transient solutions, see eg. PIMPLE or transient SIMPLE. However, the whole point about the PISO algorithm is to eliminate the need for underrelaxation. I do agree with you VG. should have a look into a CFD book on this though.

Quote:

Originally Posted by nimasam (Post 286984)
piso Foam is a transient solution so relaxation can not be true for a transient solution;)


sharonyue September 3, 2013 00:47

Quote:

Originally Posted by akidess (Post 287191)
You can also use relaxation with transient solutions, see eg. PIMPLE or transient SIMPLE. However, the whole point about the PISO algorithm is to eliminate the need for underrelaxation. I do agree with you VG. should have a look into a CFD book on this though.

But in pisoFoam.C, we have UEqn.relax(), why UEqn is relaxed?
and whats the difference between p.relax() with PEqn.relax()?

BTW, in Verssteeg'book page196, it said:
Quote:

As before, under - relaxation is required with the above procedure to stabilise the caculation process.

akidess September 3, 2013 05:02

It doesn't really matter that much if you relax UEqn. In fact, you can throw out the solution UEqn completely by deactivating the momentum predictor.

As for your question about the two types of relaxation, this has been answered multiple times in the past, e.g. here.

Quote:

BTW, in Verssteeg'book page196, it said
Page numbers vary between editions of a book, and the quote you give is so isolated it is of limited use to someone without access to the book.

sharonyue June 1, 2015 06:50

Quote:

Originally Posted by sharonyue (Post 449464)
But in pisoFoam.C, we have UEqn.relax(), why UEqn is relaxed?
and whats the difference between p.relax() with PEqn.relax()?

BTW, in Verssteeg'book page196, it said:

Okay, after a long time, I got some documents(Chinese) to talk about this, this is what they said, I tried to translate it:


Code:

In SIMPLE, cuz there is an omission of velocity correction of neighbor cells, using this u^' to correct velocity is moderate. But for pressure, using this u^' to correct pressure is exaggerated. Cause we need to do explicit relaxation in pressure field to make pressure correction to be moderate.

For velocity, we need to use relaxation to make matrix solver robuster. Then we use Ueqn.relax.

Looks interesting, shed a light on this but not very strightforward. :D


All times are GMT -4. The time now is 09:38.