CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   FLUENT (https://www.cfd-online.com/Forums/fluent/)
-   -   divergence in AMG solver (https://www.cfd-online.com/Forums/fluent/45878-divergence-amg-solver.html)

Louisa August 29, 2007 12:51

divergence in AMG solver
 
hey all,

I am trying to model flow through porous media in fluent but when I iterate I get the followng error message:

"Error: Internal error at line 842 in file 'rp_mstage.c'. divergence detected in AMG solver Error Object: ()"

Can anyone explain the meaning of this and suggest what I should do to resolve it?

Best, Louisa

Tim August 30, 2007 07:23

Re: divergence in AMG solver
 
It basically means that your model is numerically unstable.

Have you tried solving the flow field without the porous zone activated and then switching it on later when the solution has converged somewhat?

Also, double check the values that you've used for the viscous and inertial resistance coefficients to make sure that they are realistic.

Hassan August 30, 2007 11:53

Re: divergence in AMG solver
 
in addition to what Tim said, ur Mesh type also plays an important rule....what kind of mesh r u using for porous zone?


Aimara September 5, 2007 17:07

Re: divergence in AMG solver
 
I had this message once and worked through it changing the Courant number as well as initial guessing in the solution initialization panel, if you are using coupled solution try changing the Courant number if not try the other way...

Louisa September 6, 2007 09:15

Re: divergence in AMG solver
 
thanks so much aimara.. yes im using coupled equations-i found out about the courant number a few days ago and i am so excited you said just now that your problem was solved changing the c. number as well.. i was just going to make a new post asking if it is sufficient to just change the courant number to overcome this problem!

do you know y fluent has as default courant number 5 whereas the courant number can have values between 0 and 1, though?

Louisa

Aimara September 9, 2007 16:35

Re: divergence in AMG solver
 
You can use a lower Courant number (CFL) for the initial iterations when changes in the solution are highly nonlinear...and then you can increase it....

Regards

Louisa September 9, 2007 17:17

Re: divergence in AMG solver
 
thanks a lot!


All times are GMT -4. The time now is 20:53.