Learning Forum

[Garrett Oehlerking]

Greetings,

I am trying to run a 29M polyhedral cell mesh using the pressure-based Coupled solver with k-omega SST. Previously, it would initialize using fmg, but now it stalls on the first step, so I am using hybrid initialization. I am trying to get the first real iteration to solve.

Computer specs: Intel(R) Xeon(R) Gold 6254 CPU @ 3.10GHz, 3100 Mhz, 18 Cores and 384 GB RAM. We are able to use 4 out of the 36 cores with the current license.

After looking at the initialized solution contours, the distributions look normal, so I am assuming the setup is correct.

So far, I have tried moving from workbench to standalone, switching coupled to SIMPLE, and changing fluid properties to constants.

When solving, the computer uses ~8% CPU and 144 GB RAM. I was expecting the CPU usage to be higher, but I am not sure.

[Rob]

Check the load on each core, you'll have 4 running flat out & the rest doing very little. With only 4 cores available due to the licence it's going to take a while to solve. There's a reason people use parallel/HPC packs. Note, if that's 18 real and 18 virtual cores do NOT try running on all 36, the solver won't share the resource well between the real & virtual core and may even slow down overall.