Discussion Forum

For a linear system both loads can be simulated separately the system response can be appropriately summed.

Thank you for your reply. The problem is that I have to keep the frequency of one load fixed to 100kHz. So if I do the two studies separately when I combine them the range from 35 to 4000Hz remains unchanged (this is the one due to the other load which varies in this range). How can I evaluate how the Transmission Loss changes in this range of interest (35-4000 Hz) due to the presence of the load at 100kHz together with the one with frequency from 35 to 4000Hz? Thank you,

Tommaso

Tommaso,

the straigthforward move is to do a time dependent study. But that comes at a price because you need to resolve the 100 kHz wave in time and space and you need to run it long enough to see the dynamic steady state. Can the problem be reduced to 2D?

Cheers Edgar

Unless there is a nonlinearity that leads to a failure of superposition, a time dependent analysis will lead to the same result as summing the two solutions- but at a considerable computational cost.

Now if superposition holds- summing the frequency dependent solutions is the way to go.

Just wondering how you would do that practically in this case. The two signals are applied on different boundaries leading to totally different excitation modes. We need to assume that the 100 kHz excitation affects the lower frequency signal. That is actually what Tommaso wants to find out as I understand it.

The question is HOW the two waves interact. Suppose a lower-frequency wave has high amplitude and it deforms the plate signfiicantly. Then propagation of the higher-frequency wave will be different (and modulated by the low-frequency wave). But this probably requires a LOT of displacement, far more than is likely in most physical situations.

In the non-destructive testing world there are some researchers who use harmonic or mixing wave generation to detect cracks. In these case the cracks provide the nonlinear behavior that results in wave mixing.

In the absence of a nonlinear interaction (caused by large displacements, cracks, or other phenomena) the waves do not interact.

Well, I think Tommaso got a few hints to consider. It really depends on what this is about in detail.

First of all thank you for your replies. I'm trying to do the superposition of effects by doing: -a frequency domain study at 100kHz with the edge load only - a frequency domain study in the range 35-4000Hz with only the boundary load My question is how can I sum the responses properly in Comsol?