UWO validation cases with GEOM take a long time to finish

[mcgratta]

Take a look at the cases running in /home/firebot/FireModels_val/fds/Validation/UWO_Wind_Tunnel/Current_Results. The ones that use GEOM seem to be taking alot of time, mainly PRES and COMM.

[marcosvanella]

From start to finish it is taking 50 pressure iterations per step. The threshold velocity error in mesh boundaries is not reached. I see the same without GEOM/ULMAT and FFT solver for this grid, which tells me the case is a hard one for the pressure iteration scheme.
I'll look at the imbalance from solving some meshes with PARDISO and others with FFT when using ULMAT.

[mcgratta]

Seems odd that we specify 50 iterations for only two cases. Do you recall why? If not, I'm going to remove the restriction and run again.

[marcosvanella]

I don't recall but I took it out and see a pressure/velocity wave bouncing back and forth in the tunnel at startup. It takes a long time to dampen. I'll try starting the problem with average initial velocity in x (now 0).

[marcosvanella]

Kevin, let's rerun these. I took out MAX_PRESSURE_ITERATIONS + other optimization changes.