Field projection result has strange plateaus

[tomflexcompute]

I'm using this notebook to simulate metalens with a certain radius. A FieldProjectionCartesianMonitor is used to project the near field to areas around the focus. What I found strange is that when the lens radius is large, the projected field shows strange discontinuity.

For radius = 50 lambda, the projected field is smooth as expected.

For radius = 200 lambda, the projected field has some strange "plateaus" at the center.

For radius = 250 lambda, more "plateaus" are observed.

In all simulations, the projected x and y in the FieldProjectionCartesianMonitor are the same so the "plateaus" are not due to under-sampling. @shashwat-sh could you take a look and see if you see anything wrong? I also attach the notebook here if you want to try. Running the notebook at a large radius does take a significant amount of time tho.

MidIRMetalens.zip

[momchil-flex]

Yeah zooming in on the 2D plot, it definitely looks strange. @shashwat-sh note that the projection distance increases as the diameter of the metalens increases - if I had to guess, it's probably again something related to overflow/numerical precision as the distance becomes large?

[tylerflex]

what's the status of this?