Discussion Forum

If your physics is entirely linear (and Isuspectit is), then you can combine the two separately computed results just as you described, working in the frequency domain. If there are any non-linear effects of significance, then you need to do this in time domain. (Note: I amassumingyou are presently attempting to do this in the frequency domain.) Now, assuming linearity holds, if the outcomes in your model "seem incorrect" to you, then: (1) there may be one or more errors somewhere in the setup of the problem, involving one or both of the two frequencies, and/or in how you endeavored to combine the results, either in post-processing or elsewhere; or (2) another possibility is that the result that concerns you may actually be correct, but is also different from what you expected. To get this resolved, I suggest you post your .mph file to the forum so that other Comsol users reading this can take a look at it and offer you suggestions specific to your model and physics.

Thank you so much for your support.

First, you need to make sure that you actually get the mathematics of what you want to do. In the attached picture it is seemingly a 'superposition of phasors that have two slightly different frequencies', and so that suggests a sum, in which case you can do what Robert says, and run each frequency on its own and then combine studies to get the full solution. But the attached picture also mentions E_AM, suggesting Amplitude Modulation, which in turn would imply a multiplication, in which case you need a time domain solution.

Try setting up two analytical solutions in COMSOL (no FEM), and compare sin(100[Hz]t)+sin(101[Hz]t) and sin(100[Hz]t)sin(1[Hz]*t)

and look at the frequency content of each of them. I sounds like you want to do the former of the two, but make sure.

Thank you so much @René Christensen, PhD.