Calculating bolt torque from measured change in bolt length 1

[BSukolsky]

Doing some forensics work on an assembled mechanical joint. I took measurements of bolt length in the assembled state, and after disassembly. I'm attempting to calculate the torque load based on this change in bolt length. Any ideas how I could do this?

 

[TheTick]

What I find interesting about this is that torque has always been a sloppy analog for strain and tension, which are what really count.

 

[desertfox]

BSukolsky

Here is a link that might help:-

https://www.engineersedge.com/calculators/bolt_elongation_14608.htm

If you have the bolt elongation that will be enough to calculate bolt tension so you don’t need to calculate the torque, torque is only a means of obtaining bolt tension and it’s not very accurate, bolt elongation is much more reliable.

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[BSukolsky]

We suspect inattention during assembly. Manufacturers installation instructions are given in terms of foot-pounds. This is a backwards way of trying to approximate the torque applied to check on whether the installers followed suggested procedure or not.

 

[TugboatEng]

Why don't you just measure the break-loose torque?

 

[hydtools]

Can not determine from bolt elongation the torque applied to achieve that elongation.

Ted

 

[BSukolsky]

I did that too, but there was significant corrosion to the nuts/bolts and so I don't trust the break-loose torque to be accurate--values were all over the place. I was planning to compare the measured break-loose torque to the torque calculated by bolt stretch/relaxation just as a point of reference though...

 

[JohnRBaker]

I like to watch 'Engine Power' on PowerNation TV, which can be seen on NBC Sports Network as well as some other channels. This show covers the building of high performance automobile and truck engines, and most of the time they use your typical torque wrench when assembling an engine. However, there is one application that for the last year or so I've noticed that they've been measuring the length of the rod bolts, used to hold the rod cap to the connecting rod...

...and using this to determine the applied torque. While they've never really explained how to convert these measurements into a torque setting, they use something that looks like an over-designed micrometer with a dial-gauge on it, like the one shown below:

Here's some more info about this tool:

https://www.summitracing.com/parts/arp-100-9942

John R. Baker, P.E. (ret)
EX-'Product Evangelist'
Irvine, CA
Siemens PLM:
UG/NX Museum:

The secret of life is not finding someone to live with
It's finding someone you can't live without

 

[dvd]

There is some discussion of "stretch control" in this article -

https://www.fastenal.com/en/69/bolted-joint-design.

Still agree, that torque is not a good indicator of bolt tension.

 

[JohnRBaker]

And here's an article talking about measuring the stretch of rod bolts to determine the applied torque:

https://www.enginelabs.com/engine-t...easuring-rod-bolt-stretch-vs-torque-with-arp/

John R. Baker, P.E. (ret)
EX-'Product Evangelist'
Irvine, CA
Siemens PLM:
UG/NX Museum:

The secret of life is not finding someone to live with
It's finding someone you can't live without

 

[TugboatEng]

I think OP needs to take a new assembly (to better simulate conditions) measure fasteners, torque to spec, and measure fasteners. Use the information to generate an expected range of elongation and apply this to the failed assemblies. No sense doing this mathematically.

 

[BSukolsky]

I think OP needs to take a new assembly (to better simulate conditions) measure fasteners, torque to spec, and measure fasteners. Use the information to generate an expected range of elongation and apply this to the failed assemblies

I like this idea, thanks. There are 20 bolts on these, I think I may try a range of over and under target torques so I get a few samples of each.

 

[MintJulep]

The equations to relate elongation to torque are simple.

However one of them includes "k", the "nut factor" or "friction factor". The mother of all fudge factors.

The uncertainty of "k" will likely be 2 or 3 orders of magnitude greater than the location of the significant digits of measured bolt elongation.

Rendering the results of the equation pretty useless.

 

[mfgenggear]

while bolt tension is more accurate, it does have challenges as precision assemblers following it correctly. so one must take ownership and over see the process.
edit: as always training is very important.

 

[Tmoose]

I believe Breakaway torque is probably not very good at all for estimating installation torque.
Might report high due to corrosion or static friction effects. My experience is that it is pretty common for fasteners to "break free" with a bit of a snap.
Might report low due to embedment or other creatures of the night that conspire to steal preload.

What happened that triggered the forensics ?

 

[Strong]

You could rent or purchase an ultrasound bolt tension meter. I am sure the formulas and calculated result are provided by the meter vendor.

https://www.bing.com/search?q=ultra...i57j0l5.11932j0j1&pglt=43&FORM=ANNTA1&PC=U531

Walt

 

[rb1957]

If you can measure the bolts on a new assembly, then compare these with your assembly to see if reasonable. Be aware that torque control on preload (like using a torque wrench) is very inaccurate, as much as +- 30% ... so there could be a difference (to the new assembly) but still no obvious malpractice in assembly.

You could also ask the manufacturer what the torque is on their drawings (and how they control it in manufacture).

another day in paradise, or is paradise one day closer ?