Issue with radObjDivMag (radia, Mathematica)

[msigrist]

I'm using Radia 4.32 and Mathematica 12.0

I've come across a problem with an issue with subdivisions using radObjDivMag.

It appears to occur when specifying parallel planes which aren't in the X, Y, Z directions and also have a larger number of subdivisions.

I think the following code should demonstrate this problem.

This works for me as I expect:

material = radMatLin[{0.06, 0.17}, {1, 0, 1}];
tstmag = radObjRecMag[{15, 5, 15}, {30, 10, 30}];
radMatApl[tstmag, material];

SDtype = {"pln", {1, 0, -1}, {0, 1, 0}, {1, 0, 1}};
SDnum = {2, 1, 2};
radObjDivMag[tstmag, SDnum, SDtype];

RadSolve[tstmag, 0.01, 10];

ListVectorPlot[radObjM[tstmag][[All, All, 1 ;; 3 ;; 2]],
VectorPoints -> All]
Graphics3D[radObjDrw[tstmag]]'

I get a square block cut on diagonals into 4 triangles and each subdivision is magnetised on the diagonal in the x-z plane.

If I change the number of subdivisions and keep the rest the same

SDnum = {10, 3, 10};
the magnetization of the subdivisions is no longer all in the same direction. Some pieces are correct and others have some rotation to the applied magnetization.

If I change the number of subdivisions and subdivide along the X, Y, Z planes I don't get the same problem.

SDnum = {10, 3, 10};

and

SDtype = "pln";

or

SDtype = {"pln", {1, 0, 0}, {0, 1, 0}, {0, 0, 1}};

I don't find any problems with the "cyl" form of subdivisions, and don't find any differences by changing the Frame option. Changing the magnetization direction doesn't seem to effect the issue.

I think this is a bug, but would welcome any usage advice if I am using the options for this function incorrectly.