[nvalencia]

Hi all,

I've been trying to generate a cooling plate conjugate heat transfer model (for battery cell cooling). It consists of a plate which has two cooling channels in it (water), as seen in the mesh photo attached. I generated the grid using Pointwise, with an average element size of 0.5 mm (the plate is 1 mm thick). My input is a constant heat flux at the bottom wall of the plate (820 W/m^2).
When I run the simulations, after a few iterations I get the message of temperature limited to 5000 K in the solid region. I don't think the issue is the mesh as it is well refined. Also, it seems as if heat is not being transferred into the cooling channels, as their temperature is not changed.

If someone has an idea of what the issue can be I would appreciate it, or if someone has used Pointwise to mesh these type of problems, as it could have to do with the way the BCs are assigned there.

Thank you.

[vinerm]

If the plate is 1 mm thick, then the maximum cell size should be no more than 0.25 mm. You need minimum four cells across the thickness. And it appears that you are solving only conduction. So, you don't need such a fine mesh. A coarser mesh along the width and length will do, but along the thickness, maintain at least four cells.

[nvalencia]

Hi Vinerm,

Thank you very much. I tried that and it fixed the problem.

Regards,

Nicolas

[vinerm]

Nice to know that. It is the requirement of FVM to have at least three cells between any two boundaries.