Learning Forum

[bernhard1]

Hello,

I want to simulate a taper using EME with CVCS. How many subcells do I need for such a taper using CVCS? Or is there a rule of thumb regarding the change of width within one cell? At the moment I have one cell group and 50 subcells for a maximum width of 15 microns and a minimum width of 4 microns of the taper.

[Guilin Sun]

It can be difficult to have a rule of thumb, since different people may have different expectations for the result and accuracy. My suggestion is to check the overlap of the first few propagation modes. If the overlap is close to 1, it is better. If it is small, such as 0.8 or smaller, it may indicate that more subcells are needed. Of course there is no clear criterion for the smallness, and we do not know how the overlap directly affect the results. You also need to check the current result. If abs(S21) has reached above 0.99, the transmission is quite high in my opinion.

If you want to investigate more, please refer those links:
EME Convergence Testing - An Intuitive Approach
Convergence testing process for EME simulations