I have a fatal overflow error in the FINMES subroutine. Someone can help me? I try to change the resolution method, but don´t work.

Hi,

Try reducing the physical time step size.

Regards James

Thanks James! I reduce the phisical time step size to 10-5. SO I put to run and did the same error. May I change the size of the mesh grid? Do you know the best relation between the phisical time step and the mesh size? Thanks again. Regards Henry

Hi,

The appropraite physical time step size is largely independant of the grid. You should use a timescale representative of your problem.

If you are still getting FINMES errors in the first iteration then you almost certainly need to reduce the timestep further.

Glenn

Henry,

Many things could be wrong:

- Physical timescale may only be part of the problem. Typical rule of thumb is to use a maximum of 1/3 of the fluid residence time in through your device.

- Startup difficulties are typically initial guess related. How are you setting the initial guess for velocity and pressure for example??

- Perhaps you are running with turbulence on? Maybe you need to use the zero equation model to get a rough solution, and then restart with k-epsilon??

- Maybe your boundary conditions are too much at startup so you might need to ramp up your inlet total pressure or velocity as a function of iteration

As always, many things to go wrong, many things to try...

Neale

Thank you for the help Neale,I will try some changes and see what happen. Thank you again.