CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   amg solver (http://www.cfd-online.com/Forums/fluent/28564-amg-solver.html)

rafidah May 29, 2001 00:30

amg solver
 
i'm doing multiphase flow in fluidized bed using air and prticle. during iteration i got the problem with AMG solver. i'm not very familiar with it. can anybaody help me ??

rafi friend May 30, 2001 12:46

Re: amg solver
 
dont bother about amg solver. the error is from yours, maybe it comes from your setting or your grids. check grid first. the grid cannot be negative volume

regards,
:)

rafidah May 31, 2001 00:23

Re: amg solver
 
i already check my grid status but there is nothing problem and the volume is not negative. during my iteration, after 2 iteration there is an error. internal error at line 562 in file amgif.c. divergence detected in amg solver. i'm not understand what does it's mean.

rafi friend May 31, 2001 10:27

Re: amg solver
 
have you check your problem setting?. tell me about your problem.

Alain May 31, 2001 10:30

Re: amg solver
 
This message means that you have a divergence problem which is quite frequent in euler multiphase simulation due to high ratio of density between phases.

1. Check out that you have choosen the right velocity/pressure coupling method (body force weight) and try to lower your relaxation factors.

2. Try to converg with first order discretisation schemes first before switching to second order.

3. You can also make transient itération with very small time steps at the begining and try to reach steady state by gradually increasing your time step

4. Check out for turbulent values a few iteration before divergence. If you have non physical values (a high turbulent viscosity ratio for example) converge on laminar solution and use it as initialisation for turbulent flow calculation

Best regards

Alain

rafidah May 31, 2001 23:03

Re: amg solver
 
i already ask some one that i can used another multiphase model instead of algebraic slip model.so i no need to used amg solver. is it volume of fluid model more suitable to simulate the fluidized bed column


All times are GMT -4. The time now is 00:30.