I simulate a lot of lossy fluids using frequency-dependent materials in HFSS. Lately, I've discovered that whenever the dielectric loss tangent is set to 1.0 (unity) or greater, the port solver gives me nonsensical impedance and gamma values for my suspended microstrip design. It seems to happen anytime loss tangent gets to unity... which is the case for things like DI water, ethanol, etc. at microwave frequencies.
Anybody have a clue how to fix this problem?