[pizzaice]

hi everyone!

I am simulating a room and have produced a model including a prismatic BL with Salome and Engrid. I simulated with buoyantPisoFoam before without major issues but with my current model the simulation stops after a few timesteps and obviously there is a jump in the residuals. My question is - could it be due to the grid that I created with Engrid? I think it might be due to the fact that I have massive cells in the model (1.5m length) and tiny ones (0.05m) which might somehow not fit together.. Have you got suggestions what I could do?
I already set the k and epsilon values to Gauss linear limited 0.7 and tried to turn turbulence off - but no effect. Basically the solver is crashing because the adaptive timestepping (Co nr. 0.7) is getting too narrow. My initial conditions are equal for the entire room apart from the fact that T is 30K lower in one part (initialised with funkySetFields) than in the other...
I attached a screenshot that shows BL part of the grid and the solver error I am experiencing.

Thanks and have a nice day ,

Chris