I am simulating 1 cell of PEM with an area of 50 cm². The mesh has an orthogonality of 0.2 for the channels and bipoar plates (polyhedra) and a quality of 1 for the active zones (hexagonal).

The aspect ratio is 740. (I used to have 320 before but with the sme divergence problems).

In fluent I proceed with the following in fluent:

1st: I turn with only Butler-Volmer Rate+ multicomponent diffusion+ Half-cell potentials to ensure fluide ciruclation in channels for 100iter

2nd: I either turn Joule heatings & Reaction heatings for 200iter and THEN I turn on Electrochemistry source. Or I turn on the three at the same time.

The result is the same everytime. As soon as I turn on the Electrochemistry sources on in the pem model, it diverges. Mostly because of uds3, sometimes I get Temperature as the reason, but in the residuals it's the energy lot that skyrocket.

I changed (lowed and upped) urfs as ansys suggest or as other suggest in different articles yet no luck.

I have reached my limits and I have no more ideas. Could you suggest me some solutions.