Slab- Discontinuity in Reinforcement

Hi,

I modeled a slab with some recessed areas. I modeled these areas as an attribute area function, and defined the thickness which is thinner than the slab itself.
The reinforcement distribution seems to be odd because there are these jumps instead of having a uniform distribution.
Does anyone know what this could be ? Any references on the subject ?
Thanks !

I’m guessing that the option in the screenshot might help.

Uploading a simplified data set for the problem would usually make it easier to answer questions in the forum.

Best regards,
Jost