modal analysis for location of "weak spots"

Paul_K wrote: >

That will only be useful if the part/structure is loaded so that mode(s) gets excited. Also, you couldn't compare a hot spot in one mode to a hot spot in another mode.

Jeff out ...

You are 100% right, thank you for replying. Could you direct me to some published references to prove this point?

thanks

Paul

In an eigenvalue solution (modal analysis), the eigenvectors for each mode are normalized, so the eigenvectors and subsequent results from each mode can't be compared. You are speaking of a type of transient analysis, which is different.

He is not 100% right.

Fatigue analysis of structures is done successfully using FEA. This implies examining the stresses due to many modes at any particular point on the structure, since the waterfall plot used to excite a fatigue model is based on the total signal(s) exciting the structure, not one frequency at a time.

Cheers

Greg Locock

I am intrigued to hear that we cannot compare stresses between different modes. Fatigue analysis using FEA models is a well known technique, and of course the rainfall signals used to excite the model are based on real signals with many frequency components in them.

Cheers

Greg Locock

Again, you are referring to a transient analysis, not an eigenvalue (modal analysis) one that is in question here.

The eigenvectors themselves say nothing about stresses. They are a matrix of nodal displacements. Post processing is required to derive elemental stresses. Further postprocessing is required to be able to compare stresses between modes.

It seems an arbitrary distinction to me to allow one level of p-p to get to stresses, yet forbid the next stage that will allow useful comparisons between modes. Just because the software pumps out the stresses as a matter of course does not affect the maths.

Sorry about the other post in the other sub thread, I had tried to delete it.

Cheers

Greg Locock

Stresses calculated in modal analysis can be meaningfully compared only within the same mode to obtain the ratio between stress values in different locations in structure vibrating in this given mode. Comparison of stress levels between different modes would be meaningless.

And to get back to my original question, would you know of some published references about usefulness of stresses obtained from modal analysis for locating inherent weaknesses of structure?

Thanks

Paul

A textbook on dynamics and vibrations seems like a good place to start. Maybe one on continuous systems. That's all I can tell you.

Your first paragraph is wrong. Both experimental and analytical modal analysis results can be used to compare results at different frequencies and modes.

Structural Testing Part 1: Mechanical Mobility Measurements A comprehensive introduction to the theoretical background to modal analysis and structural dynamics.

Structural Testing Part 2: Modal Analysis and Simulation A comprehensive introduction to the theoretical background to modal analysis and structural dynamics. Both available from the link below

Forgot to post this reply.

But the analytical results need to be correlated to the experimental data first.

There are only a few slides relating to modal analysis. I don't see anything that really covers to the subject here.