Adjusting the model size for a specific modal test



We are working on experimental modal analysis of a vehicle acoustic cavity. I want to extract the correct natural frequencies and mode shapes of cavity. One of my problems is deciding about the model size for calculating the stability diagram. By increasing the model size, the software finds more stable modes. How can we adjust this parameter for obtaining the right number of modes?


Re: Adjusting the model size for a specific modal test

Siemens Valued Contributor Siemens Valued Contributor
Siemens Valued Contributor

Hi, and thanks for posting this question!

The problem you are trying to solve is indeed a difficult one, as the modes of the cavity are generally highly damped, and normally coupled with the one of the surronding structure as well making the modal density quite high.

My colleagues have recently posted and article explaining the basics of the stabilization diagram and how the poles shoudl be selected (see: Getting-Started-with-Modal-Curvefitting). This can be used in general to select the modes in standard scenario, but to better approach your problem i would suggest you to take some time in reading also this other article about the MLMM solver (see: Maximum-Likelihood-estimation-of-a-Modal-Model-MLMM)


The MLMM solver was specifically developed with the acoustic cavity modal analysis scenario in mind. As the selection of the poles in the stabilization diagram is difficult, and often the estimates end up to be be not fully reliable, MLMM will take your initially selected poles and optimize them to better represent the data that you measured. 


So in your case, my suggestion would be:

  1. perform a regular modal analysis using polymax, possibly relying on automatic modal parameter selection 9obviously the settings will depend on the frequency range you are interested in, the number of references you are using, etc)
  2. calculate your shapes
  3. apply MLMM with a number of iteration sufficient for the modal model to converge (it will either stop at the number of iterations you set, or if the convergence criterion is met)
  4. review your modal model using the FRF synthesis, MAC and complexity criteria to make sure your fit is good, the modes are orthogonal and your mode set is consistent.


I hope this can help you


If you want to discuss more in details the applications or have still some question, feel free to reach out to us. 


Re: Adjusting the model size for a specific modal test

Thanks for giving the answer for the above query.