CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Pre-Processing (http://www.cfd-online.com/Forums/openfoam-pre-processing/)
-   -   SimpleFoam as laminar solver (http://www.cfd-online.com/Forums/openfoam-pre-processing/62142-simplefoam-laminar-solver.html)

anja March 24, 2006 05:48

Hi, as simpleFoam is a stea
 
Hi,

as simpleFoam is a steady-state solver, the defined "timestep" is an iteration counter. Does that mean, that the "endTime" is the max number of iterations? And why is then within each timestep an iteration of p (in my case only in the beginning of the calculation, problem follows)

I want to calculate u and p but:
Time = 999
AMG: Solving for p, Initial residual = 1.02579e-06, Final residual = 2.76786e-07, No Iterations 1
time step continuity errors : sum local = 6.13966e-06, global = 5.96414e-08, cumulative = 0.000117451
ExecutionTime = 9411.37 s

Time = 1000
time step continuity errors : sum local = 1.01213e-05, global = -4.50328e-08, cumulative = 0.000117406
ExecutionTime = 9419.22 s


Thanks a lot for your help!
Anja

hjasak March 24, 2006 05:59

Yes, endTime is the max iterat
 
Yes, endTime is the max iteration count. However, based on the snippet of the log file, you run has converged a long time ago. What you are seeing are the oscillations in the pressure due to round-off.

Have a look at the point where the momentum equation was solved last - this is where you should have stopped the solver. Note: the convergence checks in simpleFoam may be somewhat different to what you're used to: for example, if you want to converge the global residual to 1e-6, you should specify the linear solver convergence to be lower. Otherwise, as you approach the global convergence tolerance, the solver stops doing the work and nothing much happens. I personally always check the global convergence by hand (well, eyes) :-)

Enjoy,

Hrv

anja March 24, 2006 07:20

Well, this is what I was using
 
Well, this is what I was using:
p AMG 1e-06 0 100;
U BICCG 1e-05 0.1;

What would you change?

How do you check the global convergence?

THANKS!

lam April 5, 2007 09:28

Hi world, I am in the same
 
Hi world,

I am in the same situation like Anja:
I want my computation to stop when it reachs the criterium of convergence that I precise (the solver tolerance in system/fvSolution file, isn't it?), and not in function an arbitrary number of iterations (endtime, in system/controlDict in file).

Can somebody help me to do this, please?

Thanks in advance,

Lam


All times are GMT -4. The time now is 01:37.