Hello the dedicated developers of scuff-em,
I am testing scuff-caspol in a simplistic situation where there is only a PECPlate and a Rb atom. Instead of invoking the --PECPlate option, however, I am inputing a geometry file to fully verify its calculation of Green's tensor from finite element geometry files. The mesh file UnitPlate_52.msh (with a redundant .txt allowing to upload) looks like
whose dimension is 0.01x0.01 and normal vector is along -z . Correspondingly the geometry file
Plane_52.scuffgeo is
LATTICE
VECTOR 0.01 0.00
VECTOR 0.00 0.01
ENDLATTICE
REGION PlateVol MATERIAL PEC
SURFACE Plane
MESHFILE UnitPlate_52.msh
REGIONS Exterior PlateVol
ENDOBJECT
where I took special care to the ordering of REGIONS according to doc on scuff-em geometry so that the z>0 region will be in vacuum and I can evaluate the potential along that half axis, see EPFile.
It turns out to have no meaningful outcome because every output value in .byXi or .byXikBloch is nan. My command was scuff-caspol --geometry Plane_52.scuffgeo --EPFile EPFile --Atom Rb --BZIMethod TC --BZIOrder 9 --BZSymmetryFactor 8. Although the sampling in the BZ is really coarse, I was expecting it to at least offer some less accurate results rather than nan. I am practically running the patched version #222. The log file is scuff-caspol_52.log.
I preliminarily identify the issue to be with the Green’s tensor calculation from the geometry file, because when I replace it with the analytical form of the Green’s tensor by invoking the option –PECPlate the results are correct.
Any suggestion or help will be highly appreciated!
Hello the dedicated developers of scuff-em,
I am testing scuff-caspol in a simplistic situation where there is only a PECPlate and a Rb atom. Instead of invoking the
--PECPlateoption, however, I am inputing a geometry file to fully verify its calculation of Green's tensor from finite element geometry files. The mesh file UnitPlate_52.msh (with a redundant .txt allowing to upload) looks likewhose dimension is 0.01x0.01 and normal vector is along -z . Correspondingly the geometry file
Plane_52.scuffgeo is
where I took special care to the ordering of
REGIONSaccording to doc on scuff-em geometry so that the z>0 region will be in vacuum and I can evaluate the potential along that half axis, see EPFile.It turns out to have no meaningful outcome because every output value in .byXi or .byXikBloch is nan. My command was
scuff-caspol --geometry Plane_52.scuffgeo --EPFile EPFile --Atom Rb --BZIMethod TC --BZIOrder 9 --BZSymmetryFactor 8. Although the sampling in the BZ is really coarse, I was expecting it to at least offer some less accurate results rather than nan. I am practically running the patched version #222. The log file is scuff-caspol_52.log.I preliminarily identify the issue to be with the Green’s tensor calculation from the geometry file, because when I replace it with the analytical form of the Green’s tensor by invoking the option –PECPlate the results are correct.
Any suggestion or help will be highly appreciated!