Can EMAG submodeling be done with the DSP/GSP formulations?

Possibly not due to defect 15582, which reports that DSP/GSP solutions are incorrect for a boundary condition utilizing non-zero MAG values, which would be the case in a submodel.

A possible workaround is to specify the permeability of air regions as MURX = 1.0001 instead of 1. However, this would probably lose the advantages of the DSP formulation because a very poor "guess" solution would be obtained based on the whole model being given a very high permeability.

The source of the problem is the inconsistency between the source (guess) field and applied boundary conditions. According to the DSP and GSP philosophy the guess field should be a good approximation of the target solution. A non-zero MAG BC creates a significant difference between the guess and final solutions and hence violates the philosophy.

Show Form
No comments yet. Be the first to add a comment!