Discussion Forum

In the EM wave-transient module the magnetic vector potential A is calculated, which according to the manual can give both E and H fields (see the top of page 136 in people.ee.ethz.ch/~fieldcom/pps-comsol/documents/User%20Guide/WaveOpticsModuleUsersGuide.pdf) . But as far as I know one needs to calculate the 4-potential (A, plus the electric potential Phi) to get all the fields. There are some gauge freedom, but normally one can't just ignore Phi all together.

So is there any implicit gauge constraint that COMSOL is implementing under the hood, or the module is only suitable for certain approximations?

I started to notice this while trying to solve the transient EM radiation from a single-pulse surface current. I know the radiation should be transient but the EM-transient module always gives me oscillations that persist way after the surface current pulse. On the other hand EM wave-Time explicit gives totally bizarre results. In fact, what's the major difference between the two, assuming magnetic potential A alone is sufficient to describe all the fields?

Link to the .mph file (somehow can't attach it directly due to "file extension error"): www.dropbox.com/s/yhu7s8gohdxfus2/2DTransientCurrentRadiation_ModelOnly.mph?dl=0

Thanks!