Replies: 1 comment 1 reply
-
Maybe you have a very sharp resonance? I would try looking at the transmission spectrum. |
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hi, I'm simulating meta-len with pymeep 1.27.0(conda-forge). The illustration of experiment is exactly same as #2578 , the meta-len is placed at Design Region in figure, and a monitor plane gives field distribution after gaussian source passing the len.
It's highly stable when simulate with same metalen design and SAME resolution, however, when resolution changes, result becomes inconsistent even with very high resolution:
wavelength is 0.98 unit, and the size of pixels in design surface(left) is 0.05
here I demonstrate Ez's real and imag part(middle and right) under different simulation resolutions 170 136 150 135, about 150 pixel per wavelenth unit, 7 pixel per design unit(a pixel of design surface). As you can see, huge difference between different resolution makes results look unreliable. BTW I have tried using sim.run(until=..time..) and set a very long stop time and did not see any improvement.
Is this expected behavior or just I missed something?
here's full simulate result
07-27-09-14-15_resolution_test.1.mp4
here's a result with transmission spectrum
08-20-05-58-13.mp4
here's code for simulation and visualization
Beta Was this translation helpful? Give feedback.
All reactions