Tightening bolts to yield - is this bad? 3

[just_some_shmuck]

Over the years, I have seen a few sources talking about tightening bolts until they yield. (

https://en.wikipedia.org/wiki/Torque-to-yield_fastener)

The motivation is keeping the bolts tight and yield is easy to accomplish and yield is relatively predictable (compared to friction and torque). But if I torque a bolt all the way to its yield point, won't the bolt yield even more the first time the joint is loaded in tension? And if the bolt yields even more, then the bolt will lose some (perhaps all) of its preload (which is really, really bad).

Do I count on relaxation of the joint to keep it from yielding farther? Am I missing something?

 

[desertfox]

Hi

Maybe this paper might help:-

https://www.appliedbolting.com/pdf/yielding.pdf

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

 

[jimstructures]

For structural bolts, you don't tighten them until they yield, you tighten them until they are under tension but haven't yielded. The bolt material is elastic and stretches under load, that stretch is what keeps the bolt tight. However if you tighten the bolt until it yields then the bolt is failing already.

Maybe I am stating the obvious, but I think it is an important distinction.

Jim

 

[TimSchrader2]

Hello

90% of yield is normal. Not to yield. The load actually taken by the bolt is a function of the stiffness of the joint or plates. Rule of thumb is 10% of applied load is taken by bolt. There are two load paths. One thru the plates and one thru the bolt. As the compressed plates un-compress they handle alot of the applied load

http://www.boltscience.com/pages/basics5.htm

 

[drawoh]

just_some_shmuck,

Torquing bolts to 90% of yield seems to be standard procedure. I have seen all sorts of torque tables based on this, and most of the screw torques I have worked out have been based on this. Think about this. Torque is not an accurate method of working out screw tension. If I torque down a hundred screws to 90% of yield, a bunch of them will have been stretched plastically. This is good design and assembly practice if screws are systematically discarded and replaced upon disassembly and re-assembly. Most of my screws have been 303[ ]stainless, which is ductile, and very tolerant of being over-stretched.

If I screw a bracket to a base, and I apply 1000lb tension, the clamping force on the base will be 1000lb. If I apply 200lb force on the bracket, the clamping force will drop to 800lb. The bolts tension still will be 1000lb. If I apply 1200lb to the bracket, it will separate from the base, and the tension on the screw will be 1200lb. Good design practise is to ensure that the bolt clamping force is well in excess of the external applied force. If the force approaches or exceeds the clamping force, the bolt will rotate out, and will be subject to metal fatigue.

If your system is to be taken apart and your screws re-used, consider specifying 75% of yield/proof stress, and call out high strength screws. Work out the effect of metal fatigue on a screw cycled from 0 to 75% of yield. I don't have a good feel for this.

--
JHG

 

[just_some_shmuck]

Desertfox, the link you posted is very helpful. Thank you. But there is one part of the article that I do not understand:

"This would require a pushing load, which would have to be maintained at levels along the line XU. But the joint members cannot exert loads at these levels. They can only exert loads at levels along the lower line XF. This is because when their compressive strain decreases, their compressive load falls from J towards 0 and XF is parallel to JO" (

https://www.appliedbolting.com/pdf/yielding.pdf

page 1, column 2, halfway down)

I don't understand what Mr. Gill is saying happens in the bolt after it is loaded past point X. Can anyone enlighten me?

Thanks!

 

[TimSchrader2]

Hello. OP. The two graphs are similiar to stress strain curves that you have seen. Note that in this case the bolt is tensioned beyound yield, beyound where it should be tensioned. It is describing what is happening in the bolt after yield. The slope of stess/strain is the stiffeness of the bolt after X along slope line X-U. Apparently, the external loads are not directly shown on these graphs but inffered by the displacement along the lines. I did not see that at first.
I like to think of it as two springs in parralel. One with large compressive stiffness (the Plates)and the other spring with small extension stiffeness (the bolt, about 1/10 or so). The actual free body is more complicated


The formula tells us that

F bolt= F (pre tension) + ((K bolt/(K bolt+ K plates)) x Force applied to joint (usually at plates).

F Bolt =90% yield + about 10% of applied load

I think this is one case where the math makes more sense.

 

[desertfox]

Hi

I will try to explain in words what is happening to the joint after the bolt has gone beyond yield.

Firstly on the graphs for both the compressive load on the plates and the tension load in the bolt are the same upto the yield point 'Y' on the graph. All that is different is the stiffness of the plates and bolt, so the bolt elongation is slightly greater than the compressive reduction of the plates thickness under the original bolt force.
Now the joint below the yield point Y assuming an external load on the plates causes the clamp force to be reduced and the plates thicken back towards their original thickness and in turn the bolt stretches to accommodate this change, all this takes place elastically.
When the bolt load is further increased and exceeds the yield point 'Y' the stiffness of the clamped plates has remained constant I.e. Line 'X' 'f' which is parallel to the original line 'O' 'J', however the stiffness of the bolt as changed and follows the line 'X' 'U', if you now consider the clamped plates trying to return towards their original thickness then the bolt should elongate proportionally to allow this change to take place, however it cannot do so because a given increase in plate thickness from their compressed position would mean a bolt elongation greater than the increased thickness of the plates which would be impossible task.
So what I am saying is that if the plates were to increase their thickness by .003" the bolt elongation would need to be .004" due to it following the 'x' 'u' curve, this would not be possible.




















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

 

[just_some_shmuck]

Thanks, desertfox!

 

[desertfox]

Hi justvsome shmuck

You're very welcome

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

 

[EngrPaper]

I do not have much personal experience with this as I am new in the field, but I can cite Shigley's Mechanical Engineering Design (10th Ed). In short, 75% proof is for non-permanent connections while 90% proof is for permanent. Similarly, it states:

"Good-quality bolts can be preloaded into the plastic range to develop more strength. Some of the bolt torque used in tightening produces torsion, which increases the principal tensile stress. However, this torsion is held only by the friction of the bolt head and nut; in time it relaxes and lowers the bolt tension slightly. Thus, as a rule, a bolt will either fracture during tightening or not at all." (433)

 

I don't allow vehicle tire techs to 'tighten' my car's wheel nuts without personal supervision. Partly to prevent air tools and partly to prevent the application of 'lubricants'. If I cannot be there I remove them at home and torque them up myself.

Having to replace 20 wheels studs because one stretched to failure is as expensive as it is unnecessary.

"Everyone is entitled to their own opinions, but they are not entitled to their own facts."

 

[weldstan]

In slip critical designs, bolting is often tensioned above yield.

 

[CWB1]

In most instances where high cycle fatigue is a concern TTY has been common for a couple decades now. The downside of this naturally is that TTY fasteners and IMHO any fastener thats been torqued to >80% yield is one-time use only.

 

[itsmoked]

Along these lines there was a company across the street from where I worked called StressTel. They had an instrument for tightening bolts accurately regardless of washers, oil, contaminants, compressing clamp materials etc.

StressTel bolt stress measurement instruments ultrasonically measure the elongation of installed bolts and fasteners. Ultrasonic measurement of bolt elongation determines the true bolt stress and is often used when traditional methods are limited by unpredictable operational variables such as friction and lubricant.

StressTel offers two types of bolt stress measurement instruments: the full featured BoltMike SMII and the handheld Bolt Monitor with easy-to-interpret display. Both units are portable, easy-to-use, battery operated, and self-calibrating.

I always thought something like this would allow one to never over-torque a fastener.


Keith Cress
kcress -

http://www.flaminsystems.com