We could do with a bit of advice about bolts. Need to do a lot better than we are doing.
Are fatiguing our fatigue-testing machine rather than fatiguing the weld samples it is supposed to be testing.
Overall view
formatting link
Problem is fatiguing and/or breaking the 8 bolts holding that flange (blackened) to the hydraulic piston (white) - see the lower "dynamic" grip region.
Bolts are 1/2inch UNF so 20TPI, hex. socket head.
Not the most common size over here in UK.
Last lot only lasted two weeks - done about 4million cycles to
7~1/2tonnes (75kN) load.
What sort of bolts should we specify and where could we get them?
Any other tips?
We assume that there must be some supreme quality bolts for say aircraft applications which could endure much longer. Is that so and where could we get them - especially in the UK?
Fatigue life is directly related to proper bolt pre-tensioning...
"...a 180,000 psi tension bolt with a shank diameter of 3/8 in. that was pre-stressed to forty percent of UTS (ultimate tensile strength) and subjected to a cyclic tension load of 12,000 lb. failed after 4,900 cycles. An identical bolt, pre-stressed to a level of 108,000 psi (60% of UTS) and subjected to the same cyclic load, went more than 6,000,000 cycles before failure." - Carroll Smith's "Nuts, Bolts, Fasteners and Plumbing Handbook"
No - we realised this is going to get any bolts destroyed. We need to make something to lock the hydraulic piston from turning while we torque-off the bolts.
We need to torque enough that the flange surfaces do not pull apart at maximum loads and "work" the bolts free, yes?
Two places
- the head of the socket head cap screw (SHCS) can fatigue off - sometimes classic burnished fatigue surface
- the bolts can fatigue (early fail) or tear (final fail) at the flange interface between the grips-flange (black) and the piston-flange
The bolts don't protrude through the piston-flange - nor does the finish on the underside look like that was ever intended. ...
The jaws are supposed to be shimmed by that fibre-reinforced thermoset (?) material. The lower grips lacks them underneath the jaws.
Anyway - no problem because
- I "take-up" the jaws against the body of the grips by lightly nipping the jaws then gently ramping up the static force until the jaws slide-up against the grips body. Then I lock-off the grips read to go for the test.
- the machine "clanks" hideously anyway if you go to true zero load - the ideal "pulsating tension" condition (and R=0). Have to leave some tension on at the lowest stress and the way to have a like-for-like condition is to use R=0.1. Didn't know why some many works in literature use R=0.1 until I experienced that "clanking" and I realised then without being told. So the jaws don't experience a reversal-of-stress and don't "float" in the grips-body.
Not sure I am getting you with that. The clamps work well - though for high stresses you cannot take a sample out and then re-clamp it to continue - well if you try the sample "frets" in the jaws.
Agreed. The OP will need to toque the bolts well up into the UTS curve. Not sure how far I'd push it but right near the SAE proof load would be a good start.
I don't explain it very well but any bolt that is properly torqued will not see any of the cyclic stresses as long as the total load is less than the torque load.
Did the broken screws come from the same lot? Besides others' advice re installation torque, don't discount the possibility that the screws were improperly processed. Happened to me recently with a batch of alloy socket set screws that were apparently not tempered after hardening. They were US made by a reputable mfr, who replaced them promptly.
====================== Installation torque has a considerable effect on fastener fatigue life. The idea is that there enough tensional load put on the bolt or cap screw such that it is above the working load, thus eliminating any flexing. This is probably more important that a super-grade (12) bolt or capscrew, which ain't cheap. More than likely grade 8 will be adequate. for some US suppliers see
The original bolts lasted a long time since before anyone working there's time.
What might have changed is my structural steel weld samples need a lot more force than the small thin Aluminum / Aluminium aerospace samples that have been the mainstay for some time. Also I'm doing a lot of simple cycles - say 2million - where others are testing instruments detecting acoustic emission from crack propagation and the like.
Anyway, the new bolts must be from a different source.
I don't easily endorse products yet alone proclaim them to be "MAGIC". They have solved a number of problems I have had over the years with breaking fasteners...BUT, they aren't cheap!
Fatigue testing machine is back in action. Have replaced bolts and made sure torqued-down well.
Spacered bolt-heads to surface, therefore slightly longer bolts, so can see if bolt-heads fatigue off - makes it obvious.
Thanks ever so much for all these very helpful replies.
Richard Smith
BTW - getting interesting results. Hybrid laser/arc welds (HLAW) - laser and MIG - are performing better than traditional fillet weld, for instance. So all promising.
No - we realised this is going to get any bolts destroyed. We need to make something to lock the hydraulic piston from turning while we torque-off the bolts.
We need to torque enough that the flange surfaces do not pull apart at maximum loads and "work" the bolts free, yes?
Two places
- the head of the socket head cap screw (SHCS) can fatigue off - sometimes classic burnished fatigue surface
- the bolts can fatigue (early fail) or tear (final fail) at the flange interface between the grips-flange (black) and the piston-flange
The bolts don't protrude through the piston-flange - nor does the finish on the underside look like that was ever intended. ...
The jaws are supposed to be shimmed by that fibre-reinforced thermoset (?) material. The lower grips lacks them underneath the jaws.
Anyway - no problem because
- I "take-up" the jaws against the body of the grips by lightly nipping the jaws then gently ramping up the static force until the jaws slide-up against the grips body. Then I lock-off the grips read to go for the test.
- the machine "clanks" hideously anyway if you go to true zero load - the ideal "pulsating tension" condition (and R=0). Have to leave some tension on at the lowest stress and the way to have a like-for-like condition is to use R=0.1. Didn't know why some many works in literature use R=0.1 until I experienced that "clanking" and I realised then without being told. So the jaws don't experience a reversal-of-stress and don't "float" in the grips-body.
Not sure I am getting you with that. The clamps work well - though for high stresses you cannot take a sample out and then re-clamp it to continue - well if you try the sample "frets" in the jaws.
Go with a lower grade bolt. The Grade 8, which is most cap screws are stronger, but do not have the stretch length of a lower grade bolt before going in to the plastic region. Get to the plastic region and the bolt will be deformed, then the bolt will loosen and you will most likely snap it then. Was a problem about 40 years ago on a Hopto excavator, where the bolts on the bucket would not hold.
PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.