Relationship - Break over (click type) torque wrench and short term joint relaxation

[Charles Harvey]

Has anyone seen any studies of the relationship between break over, ‘click’ type torque wrenches and short term joint relaxation? I was recently told that the break over action creates a reaction, which increases the variation of short-term relaxation. I am wondering if anyone has seen data to support that idea.

 

[Graeme]

No data, but -

Charles,

I don't have any data that would address this topic (short-term relaxation). However --

• I do recall seeing a study about 15-20 (?) years ago, probably done by the Navy, which concluded that the break-over torque wrenches were more repeatable than the torsion bar type. However I am relying on fuzzy memory because I don't have a copy of the report. So, I don't know if it addressed the relaxation issue.
• I have frequently used break-over torque wrenches in the range 7 to 12 inch-pound (0.79 to 1.36 N m) over the last 20+ years, for correct tightening of RF and microwave coaxial cable connectors. My empirical experience with those has been that the stability of the connection is very high. Based on measurements of the electrical characterisitcs, there is no measurable change in the mated connection which could be reasonably be attributed to relaxation of the torque over time periods of up to an hour or so. (For that type of measurement, temperature changes and the inherent uncertainty of the network analyzer are much more significant.) In addition, the repeatability of making the connection is one or two orders of magnitude better than hand-tightening.
• When I was a tracked vehicle mechanic in the Marine Corps (several decades ago) I know I preferred the break-over type of torque wrench primarily because I didn't have to be able to see the scale in order to use it. That's real important when you are upside down and on your back in the engine compartment of an amtrac trying to get it fixed "now" because you're out in the field. Of course, with a big engine like that the vibration would cause a lot more force than strain relaxation does to loosen the bolt. There may have been an effect but as a practical matter it was surely insignificant compared to all other factors.

Again, I have no data to show ...

 

[Jerry Eldred]

As this isn't my forte, I forwarded to our mechanical expert. He said in mechanical break-over torque wrenches (I always (ignorantly) call them click-over or click-type) there is a dynamic related to slight heating in the lubricant used inside. He couldn't quantify numerically, however, as these are not critical tools. Our facilities people use them for things like torquing screws in high current electrical panels. We don't use on product.

He said when a break-over torque wrench is used repeatedly in a short time, the friction slightly heats the lubricant, which slightly shifts the actual breakover point. If this is a critical measurement, he suggested using an electrical torque standard and do what amounts to a mini-MSA or GRR study to see how much variability exists.

Perhaps ANSI or NIST or an OEM for break-over torque wrenches has a white paper addressing this. If I can get back into my account, I'll try logging into GIDEP to see if I find any studies there.

P.S. - I will be offline/out of the country from April 13th to April 27th.

 

[Graeme]

I have not yet been able to find anything direcly addressing the question posed by Charles. But while looking at documents available on GIDEP (thanks for the memory jog, Jerry!) I did find a couple of reports with data that is new to me, at least. (Of course, there's lots of stuff that will be new to me when I discover it!)

• NASA Technical memorandum 106943, December 1995. Preloaded joint analysis methodology for space flight systems, J. A. Chambers.
• Army Research Laboratory technical report 1952, May 1999. The performance of three alternative coatings to electroplated cadmium for corrosion protection in fastener applications, S. Peterson, G. Wechsler & M. Pepi.

The interesting and relevant points out of these reports are:

• The purpose of torqueing a connector is to preload the joint so that the other parts and not the bolt carry most of the load. The bolt is in tension (typically 60-90% of yield strength) and the joined parts are in compression.
• Many factors affect torque and preload final values. A guideline is to assume +/- 25% uncertainty in the final preload value when a hand-operated torque wrench is used with a lubricated fastener. Using an instrumented bolt confused: I've not heard of that before!) can reduce the uncertainty to +/-5%.
• A major cause of relaxation is "embedding" of the metals, particularly with new parts. Because of the microscopic peaks and valleys, initial contact us made by the raised portions. The metals deform under the pressure, which relaxes some of the forces. This can cause a loss of 5% to 10% [NASA] or 15% [Army] or more of the torque force.
• Thermal cycling is another important contributor to relaxation.

I suppose the big question is: if the snap action does affect the torque value, then how significant is it compared to these other effects? But that's a question I'm not able to answer at this time.

 

[bpritts]

Graeme,

Unfortunately I can't speak to your critical question about the torque wrenches,
but I will share my $.02 that the prevailing wisdom from GM is that most of the
relaxation that will be found will appear in the first 24 hours. Many of their
specs have a re-test 24 hours after torquing. (This is in safety critical,
high torque chassis bolts/nuts on brake systems.)

By the way, we have gone to electronic monitors on torques, which we find
are much more repeatable than torque wrench measurements... although
there are some chicken and egg problems if you use the torque wrench
to audit the electronic system!

Brad