Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Displacement probe is occasionally NaN using nonlinear solver

Please login with a confirmed email address before reporting spam

Hi,

I am doing a mechanical, nonlinear analysis on a microphone's diaphragm. The goal is to perform a time-dependent analysis and see any nonlinearities in the diaphragm's deflection.

Below is a summary.

  • 2D axisymmetric model
  • Solid Mechanics physics
  • Time dependent study
  • Sinusoidal pressure input on the diaphragm (diaphragm is stressed)
  • Probe at the center of the diaphragm to see the displacement as a function of time

The issue:

  • When looking at the sinusoidal deflection as a function of time during the simulation (from the probe), everything looks fine until the deflections reach a certain positive value. For example, I see the bottom half of the sine wave (negative deflections, real numbers) and then the sine wave will stop producing. My probe is showing deflection values of NaN during this period. It will then resume producing the sine wave.

Troubleshooting:

  • The problem goes away when I turn off the geometric nonlinearities.
  • When I reduce the sound pressure level, the problem also goes away (but I'm unable to generate the nonlinearities I'm looking for).
  • I have tried making the mesh finer and I don't believe any major differences occurred.
  • I have changed the diaphragm geometry, which changes the point where the NaN's begin, but it doesn't get rid of them.

Does anyone have any advice for how to solve this? The fact that it is able to run and produce portions of the sine wave is something I haven't seen before. Thank you.


1 Reply Last Post 15 déc. 2022, 02:33 UTC−5
Henrik Sönnerlind COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 2 years ago 15 déc. 2022, 02:33 UTC−5

One thing that I would check (assuming that you are using a linear elastic material) is that the strains do not have unreasonable magnitudes (not even locally). One problem which is sometimes encountered is that there are compressive strains <-0.4. The material model will fail somewhere in that region.

This is explained in detail in

https://www.comsol.com/blogs/modeling-linear-elastic-materials-how-difficult-can-it-be

-------------------
Henrik Sönnerlind
COMSOL
One thing that I would check (assuming that you are using a linear elastic material) is that the strains do not have unreasonable magnitudes (not even locally). One problem which is sometimes encountered is that there are compressive strains

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.