FEA Modal Survey, what does it mean?
An FEA Modal Survey
I was asked this question. For me a modal survey means establishing confidence in and understanding modal content in preparation for any subsequent response analysis (especially if using a direct method which does not reveal the modes).
Here is my check list (in any order):
1. Identifying and labelling modes (first bend, second bend, first torsion, in-phase leg motion etc.).
2. Looking for ‘problem’ modes from bad modelling.
3. Identifying dominant modes, using Modal Effective Mass (MEM).
4. Reconciling mode shapes with MEM values.
5. Understanding physically if low MEM directional totals occur.
6. Looking for modes which will be sympathetic to loading directions in a subsequent response analysis using MEM and mode shapes.
7. Looking for simple order mode shapes higher up the frequency range (e.g. axial or extensional modes in plates, bars etc.)
8. Looking for widget or noisy modes at low frequencies which will get in the way of a response analysis.
9. Identifying modes with similar characteristics and understanding why.
10. Looking at clusters of similar frequencies and jumps between clusters and asking why.
11. Identifying very closely spaced modes or repeated modes – asking why they are there.
12. Picking off modes to spot check using Single Degree of Freedom (SDOF) hand calculations.
13. Using static FEA unit loading to estimate SDOF k and MEM to estimate M for spot checking (note k will pick up any modelling errors).
14. Comparing against previous analysis or test.
15. Check Rigid Body Modes (RBM) – unconstrain the model, check they are clean and strain free.
16. Check mesh fidelity for highest order mode SHAPE of interest (about 5 elements per half wave).
17. Estimate the number of modes needed in a subsequent response analysis.
There may be more!
Regards,
Tony
Do not forget to check carefully: - low order modes with low MEM, could the low MEM come from symmetric motion such that they cancel out and total MEM remains low ? If so do NOT neglect these modes solely on MEM values !
FEA Evangelist , consultant and mentor
7 年Yes the modes are bend 1- in plane. bend 2 - out of plane, Torsion 1. The arms are in phase. The higher modes migrate to out of phase motion of the two arms.
Senior Stress Engineer at Héroux-Devtek
7 年Looks like bending in 2 planes and torsion. First and second modes are flaping, third mode of vibration is torsion.
Application Development Engineer - CAE at Syensqo
7 年It would be great if you can help us with an example explaining the same. Thank you.
Technology Simulation- R&D Principal Engineer at NKT || Subsea Engineering || Engineering Leadership || Offshore Renewable || Green Energy
9 年Tony, Is there anyway can we simulate a Shaker table test when the unit is just rest on the table without being held by any support. We see a bouncing effect of the unit on the Shaker table when it is excited with PSD input. I have tried this in FEA , assuming that the base of the unit is rigidly fixed to the table but the FEA result is not yielding good and not correlating well with the test. I carried out Modal analysis to extract frequencies upto the range of PSD frequency and followed by Random analysis in FEA with an assumption that the Unit is rigidly fixed to the table. Please help me is there any other alternate way to do Shaker simulation when the unit is not fixed to the table