Question - Identification of critical modal freq.

[Mat_Aero]

Hello everyone,

I have a question regarding modal / harmonic analysis. Is there any general rule or recommendation to determine which modal shapes can be critical for structure from stress point of view? I know that partially this information can be determined from modal effective mass/participation factors, but that’s not always true mainly if subject of analysis is large assembly with both large and small mass components and if a structure has some symmetric modes. Or am I wrong?

For example, let’s have assembly which contains several parts, some of them has substantially higher mass than others. Let say I am interested in frequency range of 0-2kHz in which there are 500 modal shapes. By looking at modal effective mass, I will only identify those modes, which are mostly linked to high mass components. And also, if some mode has zero modal effective mass let say in all directions, it doesn’t necessarily means that this mode will be insignificant from stress point of view. It can be easily some symmetric modal shape which produces zero modal effective mass.

My question is, if there is a way how to determine critical modes for other “small” components without examining all 500 modes individually or without including all 500 modes into harmonic/random analysis which would cause high comp. time? There is also practice that modes with very low modal effective mass are neglected and not included into list of modes on which harmonic or random response calculation is based on for speeding up the calculation, general rule I believe is at least 90% of total mass excited in all directions. But how can I be sure that those ignored modes with low modal mass are not critical for those small components from stress point of view? My main goal is to speed up harmonic/random vibration calculation by not including all of the modes but only the selection which does not cause significant error in stress results.

I hope my explanation and questions makes at least a bit sense and I would be very glad for any opinions on this topic. Thank you!

Mat

 

[FEA way]

You explained why modal (effective) mass won't help you but what about participation factors ? They seem to be the closest to what you want to get. Why aren't they sufficient for you ?

 

[271828]

Are you working from EMA results or calculated predictions from FEA?

If the latter, then you could probably set up an FRF with stress, bending moment, or other output per force input and look at the FRF peaks.

 

[Mat_Aero]

Thank you for responses,

FEA way - well, as I understand it, the participation factor only tells you how each mode contributes to the overall response of the system, therefore which modes can be critical for overall response, but not which modes are critical for particular component. The modes of the small, lightweight component probably will not significantly contribute to the overall response of the whole assembly, but those modes can be critical for that component and participation factors won't tell you that. Also, modal effective mass and participation factors are directly linked together since MEM=MPF^2.

271828 - I am working with predictions from FEA. I thought about that approach, to look on FRF but that would mean that I would have to look on FRF of each element of the mesh since each mode would stress structure in different way, therefore produce high stresses in different areas/elements. But perhaps to overcome/reduce this, it would be possible to take envelope of modal stresses of all 500 modes and then pick the most critical elements/areas...