[renegade_eng]

Hi all,

OF newbie here. This is not my first case, but most of my previous OF sims were external aerodynamics, not CHT, and there aren't that many

I've been trying to simulate the heat dissipation from a custom heatsink for a set of radio amplifiers. At first I was using a forced convection setup, but was having sigfpe problems during the first time step, so I decided to make a simpler case with just the heatsink and amps sitting in a cubic meter of air to see if I could isolate the problem to my air inflow/outflow conditions.

It didn't work.

Now my natural convection case solves most of one time step, then throws a sigfpe while running the isothermalFluid::correctBuoyantPressure() function...the stacktrace looks like division by zero occurs in "operator/<Foam::fvsPatchField, Foam::surfaceMesh>(). I have no idea what that means, or what part of my setup is causing it. I wondered if maybe it was a zero area/face patch, but I can't find any.

I've tinkered around with boundary and initial conditions, and fvSchemes and Solution for the "air" region, mostly trying to match the CHT and then the HotRoom tutorials. I've also adjusted the mesh resolution/refinement a few times. Those got my to solving most of the first time step

Any help would be greatly apprciated

I'm attaching the output from foamMultiRun and here is a Dropbox link to the tar.gz of my case (it was too big to attach).

---------------------------

PS: I've tried searching the forum, the internet, and looking at the manual. Nothing I've seen so far seemed related, but if I just missed it, please point me in the right direction. Thanks!

[renegade_eng]

Hi FOAMers,

I found the issue. For some reason (seems likely because this is a compressible case), setting the initial value of p_rgh to 0 caused the division by zero FPE. Seems pretty obvious now