HFSS Simulation of an Antenna in Lossy Medium
I performed a simulation of the antenna in a radiation box composed of free space and the results (e.g. S11, radiation patterns, gain vs. freq) obtained are correct. I used radiation boundary conditions and the size of my radiation box is at least lambda/4 from the radiating structure.
For the lossy case simulation, I assigned the lossy material properties to the radiation box. Again, I'm using radiation boundary conditions and the lambda/4 dimension spec. I checked for convergence. However, the results that I obtain do not make sense. For example, the gain vs. freq plot depends on the size of the radiation box. The imaginary input impedance plot seems to be off by some orders of magnitude from what it should be.
Has anyone had problems in the past with HFSS and radiation into lossy media? Are there some settings that I should be aware of and am missing when it comes to simulating in lossy materials instead of free space radiation boxes?
Regards!
My experience is that HFSS is doing its job. Check your HFSS model once again! Lossy dielectric, I suppose the real part of the dielectric c. is also high? Or? Are you sure that your patch antenna works at all under these conditions? Have you found any surface propagating waves?
WillEd
for doing such type of analysis , i think u must make slab, sphere or any thype 3D object in direct contact with antenna and define the slab or sphere as lossy medium and then see the performance of antenna........i think it will provide better results..................
We have used HFSS to model antenna's within the body.
WillEd's method is the way we have used when designing antennas for use on implants. We found keeping the structure simple shortens the simulation time without degrading the results.
However we also found with "body tissue" dielectrics with relatively high conductivities patches don't work very well. They are very sensitive to detuning and the body tissue tends to "short out" the patch which I believe is due to them having a strong electric near field.
We have found "loop" works much better in a body application presumably due to it being predominantly magnetic in the near field.