A Problem with natural frequency for modal analysis

[amrhmz]

I've a problem with my analysis with modal. I made a model of locomotive car body, and i want to find out it mode shape and natural frequency. But i got a very weird result, this is my result

***** INDEX OF DATA SETS ON RESULTS FILE *****

SET TIME/FREQ LOAD STEP SUBSTEP CUMULATIVE
1 0.21429 1 1 1
2 0.23288 1 2 2
3 0.28656 1 3 3
4 0.35503 1 4 4
5 0.36421 1 5 5

I'm sure that for my car body, its must have a natural frequency that more than 1, may be 9 or more, but from the result its very small frequency. May be someone have an answer for my problem. Did i make some mistake from making model?
Thanks for your help before.

 

[CornishPasty]

"Did i make some mistake from making model?"

Yes, this is the likley cause of the problem assuming your results are not as expected...

Hope this helps.

 

[Drej]

A good place to start would be to check the mode shape for each frequency. Which parts/assemblies/structure/s are related to the modal frequencies, for example?


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[amrhmz]

i've remake my model with the different scale. when i got a small natural frequency, i use milimeter. but in my new model i use meters. and the result natural frequency was increase. Is that true that scale of model affect the result for natural frequency?

And for additional information, my model was not just a structure and car body, but it was fill with mass from locomotive component such as engine, transmission, and another component. I know that mass will affect the natural frequency result, so i added all mass of the component to my model with MASS21.

 

[Drej]

Ask yourself: would the modes of the structure change if you measured it in mm, m, inches, feet or any other length unit? No. Your model has changed or your units are likely inconsistent.


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[Elabbasi]

As Drej mentioned either units are inconsistent or model has changed in some other way.

Units can be a little tricky. If you stick with SI units it's easy. If you want to use mm instead of m, you have to specify your other primary units and convert all derived quantities accordingly. Two common choices are (N, sec) and (kg, sec). So, as an example, if you select (N, sec) that makes mass a derived unit (based on mm, N, sec). This will result in concentrated mass values scaled down by a factor of 1e3 and density scaled down by a factor of 1e12. You seem to have both in your model. Don't forget of course to also scale Young's Modulus and any concentrated springs.

I know that some FEA programs can do the conversions automatically for you if you request it, but I prefer to do them manually and use self-consistent units all along.

Nagi Elabbasi
Veryst Engineering