[TBERGE]

Hi,

I am conducting a simple CFD analysis of a t- junction with pimpleFoam.

At first, I ran a simple mesh (40.000 cells) with the pimpleFoam solver, and it worked perfectly.

Then, running a larger mesh (400.000 cells), the value for epsilon reaches the sky and after some time the iteration aborts with the following error message:
- Floating point exception (core dumped) ; Aborted

I know that for the same epsilon stability problem with simpleFoam, it is possible to run potentialFoam to initialize the iterations and then run simpleFoam with more stability. Is there anything similar for pimpleFoam?

The t- junction have to inlets, and these are the codes written for the patches;
Inlets: type turbulentMixingLengthDissipationRateInlet
mixinglength is set to 10% of the diameter of the inlets
uniform 1
Outlet: type inletOutlet
inletValue uniform 1
value uniform 1
Shell: epsilonWallFunction
uniform 0

Any good advice on this matter?

Regards

[jherb]

Have you tried to use the solution of a simpleFoam run on the same geometry as an initial solution for pimpleFoam (by copying the last "time step" directory of the simpleFoam case into your pimpleFoam case as directory 0)?

You could also try to set the underrelaxation factors to smaller values in pimpleFoam (for a first guess of the solution) or use first order schemes.

[TBERGE]

Originally Posted by jherb

Have you tried to use the solution of a simpleFoam run on the same geometry as an initial solution for pimpleFoam (by copying the last "time step" directory of the simpleFoam case into your pimpleFoam case as directory 0)?

You could also try to set the underrelaxation factors to smaller values in pimpleFoam (for a first guess of the solution) or use first order schemes.

The program will not accept the new directory 0 (last time step folder from the previous case).

However, I tried to start running the program with turbulence off, and after 1 second;sol.time stop the iteration, put turbulence on, and then continue to run the solution. This resulted in the same error message.

Then I tried to alter the time steps, with no results.