[retech]

I just recently started using FOAM (v1.4). After exercising and reviewing the tutorials, I'm running my own problem. It's a channel flow problem, with a forward facing step entry. It's ultimately an LES problem, but I'm running simpleFoam on it to generate a RANS field as an initial condition for the LES.

I'm using the PCG solver for pressure, with the DIC pre-conditioner. My concern is that at each time step the pressure iterations max out (I presume) at 1001. My relTol is 0.01. I've tried backing off on relTol to 0.1, but the p iterations still max out. Use of paraFoam to examine the solution indicates the velocity and pressure fields are reasonable. I'm using an under-relaxation factor of 0.3 for p.

Anyone have any thoughts on what might be causing the excessive number of p iterations?

Thanks,
Joe

[msrinath80]

If your mesh is at least a million cells, I have found that GAMG (combined with the DICGaussSeidel smoother) as opposed to PCG helps, particularly during the initial stages when iterations max out on PCG. Use nCellsInCoarsestLevel 100; for a start.

Alternately, it could also be that your mesh is bad.