Learning Forum

[ChrisD]

Hello, I have a shaft part with threads that is highly stressed in tension due to bolt preload.  It seems over-conservative to add vibration or shock stress to the preload stress, as I would imagine that for the shock/vibration to have any effect, the stress from it (in the axial direction) would have to be higher than the preload stress.  Any thoughts?

Thanks!

[dlooman]

I agree with the idea behind what you say.  The nominal stress in the bolt doesn't change until the preload is exceeded.  Some people do a random vibration analysis just to see if the bolt will loosen, but typically vibration stresses in the bolt are very small.  It's not a bad idea to do a prestressed random vibration analysis after first loading and locking the bolt pretension forces though.  The contact definition is more accurate that way.

[ChrisD]

OK,thank you!

[Dennis Chen]

 

I had dealt with this same issue at work many months ago.   Just as Dave Looman up there said (Hi Dave), nominal stress on the bolt does not change until preload is exceeded.   in addition, shock/random vibe are linear dynamic analysis which means we cannot model contact separations (contact nonlineraity). 

The key to simulation is to have a clear objective and understand the limitations of the method with respect to that objective.   It’s hardly “conservative” to add shock induced stress on top of preload induced stress to calculate max load, I think it’s straight-up incorrect.   that’s my opinion at least.   

 

[ChrisD]

Thank you, Dennis.