[philippose]

Hello,
I have been trying to simulate a temporal variation in inlet pressure in a model containing around 400,000 cells using turbFoam.

In order to start with a steady solution with an initial pressure of 15bar (normalised to 1724 m^2/s^2), I ran a steadystate simulation using simpleFoam for around 1600 iterations (till convergence).

After the steady state simulation, I copied the whole directory and renamed the last iteration of the simpleFoam simulation to "0", and added the "timeVaryingUniformFixedValue" inlet pressure type, along with a data file containing the new pressure and time values I wanted to simulate.

For fvSchemes and fvSolution, I used parameters similar to those in the turbFoam tutorial case (cavity) in OpenFOAM 1.3.

When I start the new simulation (dT = 1e-4) with turbFoam, for the first iteration, the solver runs through over 100 iterations for each momentum solver (Ux, Uy and Uz), over 450 iterations for the Pressure solver, and from then on, the courant numbers and the residuals for all the solved variables keep increasing.

Within a few iterations, the average epsilon value goes up to over 1e+8.

What could be the problem in this case? Is it ok to use the result of a simpleFoam simulation as the starting point for a turbFoam simulation? And, should I use the same solver schemes as I used during the simpleFoam simulation also for the turbFoam case?

Have a nice day!

Philippose