Clamping force over time....A LONG time. 3

[Ryan351]

Coworker and I were discussing this yesterday.
We needed to do a quick 24 study on the clamp force, or rather the holding ability (not sure what term to use, sorry) of a bolt/nut.
We torqued the nut on a pancake load cell and recorded the results over a 24hr period.
The torque applied was 90Nm and the clamp force was 17,663N. After an hour it has relaxed to 17,542N and at 24 hours 17,471N
So-
Initial 17,663N
1 hour 17,542N; Loss of 121N
24 hour 17,471N; Loss of 192N

This was to be expected. An initial higher loss with lower loss over time as the bolt stretched and relaxed.
The discussion then went to this hypothesis, "if the system was left alone forever in a vacuum, assuming no corrosion or outside reactionary forces, the nut and bolt clamp force will relax to 0".
My coworker cited entropy. But I'm under the impression that is relocated for thermal. But I've been wrong many times!
The other hypothesis is that eventually the opposition force of the part being clamped with reach equilibrium with the opposing clamping force of the nut and bolt and stop there. In short, there will always and forever be a clamp force, albeit lower than the initial.

So the question is, if a nut and bolt were torqued and left alone in a vacuum, would it EVENTUALLY reach 0 clamp force.
And I mean indefinite amount of time.
I'm sorry of I didn't use correct terms here, I've read a lot on these forums and there are A LOT of smart people on here!
Thank you!
Ryan

 

[RVAmeche]

In this hypothetical vacuum scenario, is it bolted in the vacuum and at ambient temps? Or is bolted "inside" and then released to vacuum?

If there's no outside forces or corrosion I wouldn't think the bolt would continue to relax/stretch indefinitely, but I'm also curious

 

[Ryan351]

Not sure how I quote you..... but good question!
Lets say, held at constant temp. Torqued in a vacuum, left in said vacuum.
Just looking for straight material science.

 

[handleman]

Any consideration of drift from the load cell? Is it possible for you to load your load cell with a load you are confident doesn't change (like dead weight) and repeat to see if there is any drift effect on your measurement?

 

[handleman]

This may be of some value:

https://www.sciencedirect.com/science/article/pii/S0261306900000741

 

[Ryan351]

No, we didn't consider drift. That's a good point and we could easily check that.
However the initial relaxing of the bolt is expected, which prompted the question posed in the opening.
That article looks promising. Thank you for the link. However, I can't say for sure if it's worth 40$ for the PDF.

 

[IRstuff]

https://www.researchgate.net/publication/248464620_Room_temperature_creep_of_a_high_strength_steel

would it EVENTUALLY reach 0 clamp force

The little guy with the cartoon voice bubble in the tool bar -- select the text to be quoted and click him.
Nevertheless, your notional conclusion is illogical, since that would mean that all such joints relax to zero over time, which they don't

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[hydtools]

Should there be consideration given to pent up twist unwinding resulting in the reduction of load? That could contribute to presumed relaxation of stress. I suppose that would be influenced by friction, more or less. I, the absence of friction, it would unwind.

Ted

 

[Ryan351]

Ok Thank you.
My thinking was the 2nd one, that it would never reach zero. That the joint would indeed reach an equilibrium and retain a clamping force, forever.

And thank you for the link.

 

[SnTMan]

...that would mean that all such joints relax to zero over time, which they don't

Agree 100%. They do not.

Regards,

Mike

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[btrueblood]

More fun, see links.

Before reading these papers I would have argued that below the yield strength, typical metals would not display room temperature creep, and thus a bolted joint would be expected to relax until all areas of the joint had stresses below yield. Now I'm not so sure, see the data for dead-loaded 304L wires and copper wires.

summary viewgraphs -

https://www.osti.gov/servlets/purl/1107903

full tech. paper with data

https://prod-ng.sandia.gov/techlib-noauth/access-control.cgi/2014/1417935.pdf

Still, there are data points consisting of bridges and other structures which have stood for nearly 100 years without noticeable deflections. But if nobody is measuring, how do we know how much the bridge members have deflected, and do we have any real idea of the magnitude of stress in the members? "It didn't fall down" is a pretty crude measurement.

It's also interesting that the deadweight test data gets noisy as the experiment progressed, the authors suspect movement of the camera used to log deflections was the culprit. Regardless of the source of the noise, how would one design a long term experiment that would yield better data?

 

[btrueblood]

"that would mean that all such joints relax to zero over time, which they don't"

"Agree 100%. They do not."

Not in humanly measurable time limits, I would agree. But over geologic time scales, we can see the effects. Brittle rock species like granite, schist and serpentine can be found in astonishingly folded configurations, implying that plastic deformation occurred below a rupture limit. Wondering if meteorite iron/nickel can be examined for residual stresses? Hm, it can but not sure what the papers I could find actually imply, other than some residual stresses appear in the samples, but their magnitude depends on details of the meteorites history (accretion, collisions, reentry heating).

 

[EdStainless]

It is stress and temp dependent.
We had a device with some springs in it. They were coil springs about 1" tall. They were under load all of the time and the resulting stress was about 1/3 of yield. They were removed and their height was measured every 5 years. In 95 years they had shortened by 0.0005" (they had always been measured on optical comparitors). The data said that all of the change had happened in the first 20 years.
Based on what I know of high strength steels I would first question the stability of the microstructure under long term loads. At the stress in well designed fasteners I would expect significant relaxation over the first few decades.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy

 

[Ryan351]

Still, there are data points consisting of bridges and other structures which have stood for nearly 100 years without noticeable deflections

While that is true, there are many factors aside from the material in question. Environmental factors which would assist in the reduction of clamping force. Corrosion for one big one.

That is very interesting....
"The results of the successive loading dead-weight creep tests showed that creep was
occurring at room temperature in the 304L and pure copper.... In both the 304L and
Cu curves displayed, the sample failed at the end of the final creep curve plotted."

So if I'm reading this correctly, btrueblood is correct-

Not in humanly measurable time limits, I would agree. But over geologic time scales, we can see the effects.

Which is what I was asking, over an indeterminate amount of time. It seems that the clamping force will indeed reach 0 if left alone "long enough".

Still not sure, however.
EDIT:

Brittle rock species like granite, schist and serpentine can be found in astonishingly folded configurations, implying that plastic deformation occurred below a rupture limit

This does go against the initial hypothesis though. Which you did call out, outside forces acting upon it.

 

[Ryan351]

It is stress and temp dependent.
We had a device with some springs in it. They were coil springs about 1" tall. They were under load all of the time and the resulting stress was about 1/3 of yield. They were removed and their height was measured every 5 years. In 95 years they had shortened by 0.0005" (they had always been measured on optical comparitors). The data said that all of the change had happened in the first 20 years.
Based on what I know of high strength steels I would first question the stability of the microstructure under long term loads. At the stress in well designed fasteners I would expect significant relaxation over the first few decades.

So after the first 20, was there ANY relaxation? If you can measure to the 4th decimal, I'd assume you can be confident that it didn't compress the next 75 years.

Also, thank you to everyone! This has been very interesting!

 

[moon161]

In a vacuum its gonna weld after it outgasses sufficiently. Won't come undone in your time or mine.

 

[Ryan351]

In a vacuum its gonna weld after it outgasses sufficiently. Won't come undone in your time or mine.

The vacuum was simply to illustrate "no outside forces" such as environmental.
I know I'm asking something odd. And to the second point, I know it won't in any of our time.
But when Jesus shows back up, is He going to have to bring a torque wrench? (please don't get into a religion discussion-using that as another example)

 

[IRstuff]

If HE did come back, I don't think anyone would care about THAT...

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[Ryan351]

If HE did come back, I don't think anyone would care about THAT...

HAHA I agree.

 

[IRstuff]

I think the answer is still no. We know that the Earth's magnetic pole orientation is determinable from various rocks on the surface as well as ocean, and they've been detectable up to 250 million years ago. So, at least to that extent, the magnetic domains in ferromagnetic materials are completely stable.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm