inhomogeneous materials in hfss

Dear all members,

I'd like to learn some software to simulate metamaterials and scattering problems, but there are so many kinds of softwares, like HFSS, CST. So I need your recommendation, thanks!

I know there are many people using HFSS and CST to simulate metamaterials. Which one is better in this field?

As for scattering problems, maybe I need the ability to model prescribed anisotropic and inhomogeneous materials, PMLs around the computational domain. Furthermore, I want to use an obliquely incident plane wave as excitation. Actually these things are what I need for one of my recent projects. I'm using COMSOL Multiphysics, where an obliquely incident plane wave is impossible (at least no solution up to now), then I must turn to other softwares.

Thanks for any reply!

Best regards!

i have little experience on CST, more on HFSS.

your choice should depend on the type of metamaterial unit cells you are using. as far as I know and remember, CST is using rectangular grids and subgriding. It becomes a little bit difficult for rounded geometries to use these types of rectangular grids. and, from my experinece, adaptive gridding was not that easy in CST. you have to be careful in seeding your grids, make fine grids, and run it for once.

there are some differences between CST and HFSS in assigning materials, especially conducting things. it becomes very easy in HFSS if you just use 2D surfaces and just apply PEC boundary condition.

if you will need to put negative epsilon/mu with losses for verification etc., CTS 5.0 was only allowing esp>1, mu>1. in HFSS it is piece of cake if you use frequency dependent tables for eps, mu, and loss tangents.

i simulated a metamaterial with rectangular SRR and wire in both HFSS and CST, the results were close. but anyway, i trust and feel confident in hfss more. :)

and about oblique incidence, i guess you can use master/slave periodic boundary conditions in HFSS by selecting known phase difference between planes. You can put some air region, send a plane wave and because of phase matching, you will know the phase difference between planes. For a frequency sweep i think you should be able to enter the phase differences as a function of frequency. by the way, i am not that sure about this whole paragraph.

Thanks very much!

What about introducing obliquely incident plane wave together with PMLs?