[zytra]

Hi

I've got a problem with one of my models.
It's composed of one fluid body and 2 solids.
Both interfaces are coupled.

This simulates heat transfer between an aluminum cold plate subject to convection of free air at various temperatures (parameter). The cold plate is in contact with liquid water at various mass flow rates (parameter).

This is the third iteration of the maze/cold plate design, which only differs from the other ones with a slightly different maze, in this case the liquid flow divides in a cross flow pattern - which should yield better temperature gradient across the surface that is seeing convection from free air. I had no issue whatsoever with the other 2 models. And I am using the same workflow (in terms of interfaces, BC, and such).

With this one I can converge without the energy equation (I suppose the 2 solids are ignored in this case). But with the energy equation active, residus are going down normally until at some point the energy residus start rising (with everything else still heading down fine). After hundreds of iterations the residu plots is constant around E-2 or so and everything else E-6 or so.

I tried playing with the URF for energy, too low and it won't calculate the solution, too high and it introduces fluctuations (and no convergence). Anyhow I can't seem to find a way to converge that problem.

Any idea?

Best,
Steve