vibration loosening

[SimonJ]

I have just read a report that says-
If a fastener (bolt/screw) has an stress applied to it that is greater than 65% of the yield strength of the fastener, then no matter how severe the vibration the fastener will not come loose.
Does anyone have any opinions on the accuracy of this statement?

 

[timadoma]

Hi SimonJ,
I think you are referring to the "torqued to yeild" process.
If so then I recently came accross a report wherein some description was given regarding your query:
Some excerpt about the report:

Q. If a con-rod bolt is "torqued to yeild", how does it survive the additional loading of operation without stretch and permanent set resulting in loss of preload? Or is "torque to yeild" a loosely based term that actually leaves a margin for operational loading before yeild point is encountered?
Ans. Torque to yield techniques use an electronic control system that identifies when the yield point is reached by measuring the gradient of the curve of the nut rotation angle and the applied tightening torque. When the yield is reached the gradient falls and the torque is cut-off automatically. The yield point of the bolt is reached by the combined effects of torsional as well as tensile stresses. The yield point will be slightly exceeded in order for the control system to be able to measure it. A new elastic limit at this point will be created (the Bauschinger effect) with some permanent set being caused. If a bolt is repeatably tightened using this technique it will ultimately fail during the tightening process because of the tensile limit being exceeded.

Because a degree of embedding occurs in the threads and under the nut face in all bolted joints, the stress is relieved in the bolt because of this and in particular the torsional stress is significantly reduced. (Typically the combined stress at yield - as determined by the Von-Mises criteria - is represented by 75% direct stress and 25% torsional stress.) Thus when a service load is applied there will be ample reserve for the bolt before it goes into yield again. Since typically the bolt only sees usually about 5% of the service load - the joint will frequently fail by other mechanisms prior to the bolt being overloaded.

 

[Rob45]

Simon:
A more accurate statement would be that, as long as the tensile load applied to a fastener does not exceed the pre-load, it won't come loose. That is, torquing a bolt applies a tensile stress to it, and as long as the applied load does not exceed that value, the bolt doesn't "see" any fluctuation in its' load. Therefore, it can't fail in fatigue, nor can it come loose.