[Jelle Lagerweij]

Dear CFD friends,
I am currently working on a gas-gas turbulent advection simulation. Our goal is to model the safety hazard of leaking Hydrogen in a possible production facility (40ft container). After some searching, I found that I probably need reactingFoam with the reactions turned off. I started by adjusting a tutorial case.

I was quite happy with my progress, but I do have some problems. As I have not too much compute power available, and it is a large room with turbulent flow, I try to cheap out as much as possible on the physics of the problem. As our container has only small deviations in temperature, I would just like to set fixed properties (density, viscosity etc) instead of these more specific models like Janaf and Sutherland. If possible, I would turn off temperature effect (and calculations) altogether, but then I can't work with reactingFoam anymore. Writing a solver myself is above my expertise.

Therefore, I at least wanted to use the most simple thermophysical models I could find: const and hConst. However, if I switch Janaf to hConst my code breaks. It runs for a bit (always 41 time steps, independent of the time step size) and then crashes (maybe it is a division by 0?). Can someone explain why and how this happens? I attached the current versions of my adjusted tutorial code as a zip, containing both the working and the crashing version.

With kind regards,
Jelle Lagerweij

PS: I would be really happy if someone knows a better alternative for single phase (gas) two species flow modelling, without reactions but with interest in the mixing. For example steady state and/or temperature independent models would fit my problem well. I know that this exists in Ansys Fluent, but I want to stay in open source software (that is one of my principles ).