Hello Richard,
There are at least a couple of options here:
1) The problem presented to the AMS solver is bad in some sense. For example by the coil current calculation steps yielding an unphysical source current distribution or by the linearization of the HB-curve going wrong. I would suggest plotting the current directions as arrows after the coil current calculation step(s) to see that they make sense. The HB-curve has to be monotonously increasing (local maxima/minima are forbidden) and extrapolate to a finite slope. Check your data and try using linear for both interpolation and extrapolation.
2) There is a bug in the AMS solver.
Especially the second case would be good if we could investigate so please consider sending in the model via COMSOL support.
Best regards,
--
Magnus
There are at least a couple of options here:
1) The problem presented to the AMS solver is bad in some sense. For example by the coil current calculation steps yielding an unphysical source current distribution or by the linearization of the HB-curve going wrong. I would suggest plotting the current directions as arrows after the coil current calculation step(s) to see that they make sense. The HB-curve has to be monotonously increasing (local maxima/minima are forbidden) and extrapolate to a finite slope. Check your data and try using linear for both interpolation and extrapolation.
2) There is a bug in the AMS solver.
Especially the second case would be good if we could investigate so please consider sending in the model via COMSOL support.
Best regards,
--
Magnus
