size of HFSS model
I working on patch antennas on HFSS. I designed a antenna inside a dox of air so to be able to simulate VSWR and radiation diagrams.
I have two questions!
1. My model converges with aprox. 113000 tetrahedra, which is to much for my PC. I tried on two different PC, one with 1G RAM and other with 2G RAM, and they both stop at about 90000 tetrahedra. Does the size of mash HFSS can manage depend on RAM or not? Is there something I can do except make the model smaler?
2. How big should the airbox around antenna be? I 1 lambda in all directions enough?
Thanks
1. Try installing 3 GB of RAM. HFSS v10 is compiled to utilize a PC with 3 GB of RAM, although there will be an upper limit. The other solution is to use a 64-bit OS such as Linux or Unix. This circumvents the entire memory addressing limitation of a 32-bit OS.
2. Radiation boundaries can be placed up to lambda/4 away from the antenna. PMLs can be placed even closer, but the penalty is that the PML itself must be meshed since it is a volume instead of a surface boundary condition. One wavelength away is too big.
3. For planar antennas such as p_atches, you should consider using a planar EM tool such as IE3D or Designer. The solve times will be much faster since MoM codes only need to mesh the surfaces and not the entire volume. I wouldn't recommend using a fields-based, 3D method (FEM) when a source-based, 2.5D method (MoM) will suffice.
Good luck.
Thanks
You said HFSS v10 is compiled to utilize a PC with 3 GB of RAM! What about v9.1? I have v9.1.
I know that for patch antenna MoM works better, but I have a reason for using HFSS.
Thanks again!
Added after 3 minutes:
I mean "I know that for pach antenna MoM works better..."
sorry :)