This video is taken from the EME Learning Track on Ansys Innovation Courses.
Transcript
In cell group regions where the cross section of the structure is changing smoothly, such
as a taper, the continuously varying cross-sectional, or CVCS, subcell method should be used, as
discussed earlier in the Solver Region section of the course.
The CVCS subcell method increases accuracy by avoiding the staircasing effect that you
would see with the standard EME method, but it relies on the assumption that the modes
don't change substantially from one cell to the next.
This means that there needs to be enough cells over the length of the region where the CVCS
subcell method is applied for this assumption to be true.
To test whether enough cells are being used, convergence testing by increasing the number
of cells used in each cell group region where the cross section of the structure is continuously
varying can be performed.
A script file can be used to sweep the number of cells in a given cell group region and
collect the user s matrix or S parameter result of interest.
For example, here is the absolute value squared of S52 from the polarization converter structure
that we looked at in the previous unit, plotted against the number of cells used in the taper
region of the device.
The script file used to perform the sweep and plot the results can be found on the page
linked below this video.