Discussion Forum

I cannot give a complete answer, nor am I fully confident in this response. I'm also curious to learn more about this. I've found very little documentation on it, but I've usually found the plots to behave in a reasonable manner.

Your answer could depend on the study type you solved, the plot type, and settings buried within results and physics nodes. For example, if you solved a frequency domain study, then the computed result at each point is the amplitude and phase. That amplitude and phase, combined with the frequency you set for the study, completely specify the sinusoidal signal at each point. If you proceed to plot this (for example, on a slice through your 3D domain), then I think it only plots the real part by default. This real part is essentially the real field value at a given snapshot in time. The "time" of that snapshot can be adjusted by going to Data Sets -> Solution and changing the "Solution at angle (phase)" value. By default, this is set to 0 degrees, which is often when your driving sources will be at their maximum... unless you changed the phase of your driving sources up in the physics nodes for your model.

If you solved a stationary study, then I assume the results are straight-forward with no time variation. I don't have enough experience yet with the time or eigen studies in COMSOL to comment on how those results are handled.

Rui,

field components are a snapshot in time (or phase). You can change the phase in the solution. The normalized values are amplitudes.
Cheers,
Edgar

--
Edgar J. Kaiser
emPhys Physical Technology
www.emphys.com

On a related topic, do only some of the derived values update based on the value you specify for the phase of the solution? And, tangentially to that, is there any way to get at the time average of quantities in a frequency domain solution?

For example, I solve a model in the Magnetic and Electric Fields interface. My model has a terminal present, so two of the values I can analyze in the "Derived Values" section of the results are Terminal Voltage and Terminal Current. Terminal Voltage varies properly when I modify the desired phase for my solution. Terminal Current does not vary at all.

At phase = 0 degrees:
Terminal Voltage = 1 V
Terminal Current = 0.29 - 58i A

At phase = 45 degrees:
Terminal Voltage = 0.7 + 0.7i V
Terminal Current = 0.29 - 58i A

At phase = 90 degrees:
Terminal Voltage = 1i V
Terminal Current = 0.29 - 58i A

It appears that if you are interested in your solution at non-zero phase, then you need to be careful about which values you analyze. Some values report their zero-phase solution regardless of what phase you request.

At the moment, I am interested in computing the time-averaged resistive losses in part of my model, but I'm finding that to be a difficult number to get at in COMSOL. I think the number reported to me by mef.Qrh is the instantaneous resistive loss at phase = 0 degrees. The reported value does not change as I vary the phase of my solution (just like Terminal Current in the example above), and I see no way to compute a time average.