I am trying to simulate a relatively simple parallel plate structure using HFSS. The structure is several microns in size. What I'm trying to see is estimate the complex impedance I get at a few THz.
Now, I get good results as long as I am not asking for the 'solve inside' option. In other words, as long as I do not ask HFSS to solve the fields INSIDE the metal plates, everything is fine. HOWEVER, whenever I use the 'solve inside' option, I get rejected for lack of memory.
I need the 'solve inside' because in THz, Drude model must be taken into account.
My questions to the forum:
1. Has anyone ever used HFSS to simulate metallic structure INCLUDING drude model?
2. Would it help if I install more RAM on my computer (currently, I have 2GB).
Hey, we also use the HFSS to do some simulations with frequency up to 20GHz.
The device we are simulating now are in micrometer range. Based on our experiences, the smaller the device, the higher density of the mesh, then, the more rams you may need. Originally, we have 4g ram (4years ago), later on, it was upgraded to 8g, but now we are using 24g (dell server). But sometimes, it still takes us a long time to get the result.