Guidelines during vibration testing

[elogesh]

Hai,

We design and manufacture automotive components.

We used to carryout accelerated testing of the components for specific vibration level(+ or - 25 'g') at particular frequency(70 Hz).The vibration shaker machine is from MTS.

During one of the component testing, the fixture has failed and tested component also failed. During the testing vibration the vibration level on the component was found to be 29 'g'.
Hence new vibration fixture was designed and component has passed the vibration test with this new fixture.
Then failed fixture and new fixture is given for finite element based dynamic stress analysis.

The findings were,

1) The failed fixture alone had a natural frequency of 400 Hz. When the component assembled with the fixture, the frequency is shifted to 120 Hz.
2) The new fixture alone had a natural frequency of 700 Hz. When the component assembled with the fixture, the frequency is shifted to 350 Hz.
3) In failed fixture the natural frequency of 100 Hz is closer to base exciation frequency of 70 Hz. This has resulted in dynamic amplification of applied force.
4) It is desired to have natural frequency more than 5 times of exciation frequency.
5)Culprit is fixture design.

Based on this experience, we planned for avoiding any future acquirance of this kind of failures,

"if the vibration level during testing is higher or lower than the specified level.Then stop the test and investigate the problem"

Whether the above instruction can be included as a guideline during vibration testing? Is there any other controversy will arise with this guideline.I am looking forward for your suggestions/opinions.

In our case vibration level lower than specified level is also not desired,since the component has frequency lower than 70 Hz and operates in mass controlled region.

Our sitaution is that we can't do stress analysis for every vibration fixture, which goes for testing.Hence we are in the process of formulating guidelines for the design of vibration fixture.

Thanks for reading this thread.

Regards,
Logesh.E

 

[vanstoja]

Your description of this case history leads me to believe that you have been running meaningless and worthless vibration endurance tests of various vibration fixture designs at a fixed frequency and vibration acceleration amplitude that possibly have no relationship to the vibration conditions that must be endured by your components in their actual mounting conditions in the automotive application in which they are used. The 5 times excitation frequency criterion you propose for the fixture design is about 4 times higher than the roughly 1.25 component vibration design recommendation for resonance avoidance and seems ridiculously high. Your failed component/fixture case only had an input amplification factor of 1.16 (29/25g's)at an excitation to response frequency ratio of 1.71 (120/70Hz). Your apparent belief that your component, fixtures and combinations of both could have a single natural frequency is not viable. Even a single plate fixture would have several modal vibration frequencies that would have to be considered when tested by itself. Your components,depending on their complexity, could have many other modal response frequencies that might be excited by the test input frequency. If you don't know the rationale for the present test frequency and amplitude of 70Hz and 25 g's, I would suggest that you find out what it is and reconsider your endurance testing practices in the context of the actual vibration conditions to which your components are subjected in service.

 

[elogesh]

Hai,

Thanks for your suggestions.

I hope that I understood your suggestions.

"vibration endurance tests of various vibration fixture designs at a fixed frequency and vibration acceleration amplitude that possibly have no relationship to the vibration conditions that must be endured by your components in their actual mounting conditions in the automotive application in which they are used."

We are shifting our testing conditions from single frequency to frequency sweep of 0-400Hz.We are carrying out frequency sweep test using Dactron machine.Due to certain limitations,we are unable to carryout all the vibration testing in frequency sweep.We are in the process of resolving it.But with our experience single frequency testing seems to be more severe than frequency sweep,eventhough it depends upon the case. We follow testing at single frequency based on JIS standard.

"Your apparent belief that your component, fixtures and combinations of both could have a single natural frequency is not viable. Even a single plate fixture would have several modal vibration frequencies that would have to be considered when tested by itself".

We evaluated sufficiently finite number of modes.We found out the contibution of first bending mode is quite high when excited at 70 HZ compared to other modes.Hence I discussed mostly about first mode.

We are testing our components based on estabilished standards.

once again thanks for your suggestions.

Regards,
Logesh.E

 

[elogesh]

Hai,

The estabilished reference, we use for single frequency testing is JIS D 1601.

Regards,
E.Logesh