Bolt Vector Diagram?

That's interesting: that first calculator reveals something which is quite frightening, though on reflection obvious if the bolts have the same pitch:- double the diameter of the bolt and you halve the "clamp" force for a given torque. ... No wait a minute (sorry thinking as I'm typing) I'm not sure that is correct. Surely if you double the diameter, the slope of the thread is halved: so if there's no friction wouldn't that double the clamping force. I'm struggling to get my head round this now, but, if frictional forces dominate (probably the case) wouldn't that make the diameter insignificant.

Let's hope someone who knows what they're talking about comments.

Did you follow this earlier thread, which started as a discussion about exposed threads.

First, are "Fastening" force and "Clamping" force the same thing?

That diagram seems screwy to me! Clamping force cannot be greater than the ultimate. But this shows that increased friction leads to a larger clamping force--can't be.

I have never seen those bolt-head symbols. What I have seen is radial marks indicating the Grade of the bolt.

Take a look at Boltscience.com for a vector diagram. The challenge this problem poses is the inelastic effects of friction. Friction at the thread and friction under the head are different and typically guestimated. When an answer needs to be more precise then extensive testing is done under controlled conditions (aircraft industry) or special measuring is done using special fasteners etc. which typically measure actual bolt stretch.

Your chart is also missing another important variable - bolt diameter.

e.g. - A 1" bolt will take a lot more extension force (torque) than a 1/4" bolt.

I don't have much time to comment (I'm on a project in sub-Saharan Africa) but, I'll quickly throw this out:

Friction required to tighten a fastener? Bollocks! - You can develop all of the clamp force that you want/need without any friction whatsoever. Friction is required to maintain clamp load but not in its application.

A few more comments:

Bolt selection is a bit more complex than many of us would want it to be. One must first look at the requirements of the application in terms of temperature, pressure, environment, flange geometry, gasket characteristics (if applicable), materials and operating characteristics. From this, the required clamp load necessary for joint reliability under all conditions would be calculated.

Next, the number of and the nominal diameter of the fasteners (of a suitable material) would be determined. This would be done by reviewing the bolt material specifications (in a table or chart form, if you would like) to determine if each bolt could provide its share of the necessary clamp load under the operating conditions. This is where strength of the bolt material comes into play. Note that in some applications, tightening to a point beyond yield is acceptable. The selection could be optimized a bit through judicious consideration of longer bolts, use of spacers, live loading etc. if the process is quite severe.

At this point, the designer would take a wild a** guess at what the friction might be during initial assembly, final assembly and re-assembly. This factor would be applied to the torque calculation in an effort to provide a meaningful torque figure (if you can read between the lines, you'll see that this is an engineering oxymoron). If you can't, please re-read the first sentence of this paragraph

There's much more to write but, I have no time: I must get back out into the blowing sand and 50 degree heat...

friction: I didn't say anything about friction, so I assume you were talking to Randall. If you wish to respond to a particular post, you should hit that particular reply-to button... otherwise if you are speaking to the group, but making a point to a particular person, it would be appropriate to name them.

Do you think a version or variation of my diagram could help the selection? can you think of a graphic system that could codify the basic parameters?

Thats what I'm after.

Thanks,

Chris

Correct, you didn't say anything about friction, but friction is implicit in the torque value. I read somewhere recently that friction can make up 90% of the torque with only 10% going to axial/clamping load.

Changes to the diagram? I'm not an expert, but perhaps exchange the bottom 2 axes--load and ultimate I think they were. (The diagram is out of sight now!)

thank you.

I'm off site today, about to explore and experience the culture in the capital city of a relatively safe African nation . But before I head out, a couple of comments about your graph:

• On the Y axis you mention "system fail" (by "fail" I assume that you mean joint separation). A system comprises more than just the bolt. Your diagram makes no attempt to depict the bolted joint in its entirety; It only refers to the bolt. The flange and gasket characteristics play a large role in whether the system would "fail" at a certain load.
• The label of the y axis includes "torque". For a couple of reasons, this isn't correct in terms of what you're trying to achieve. As somebody had already correctly pointed out, the torque target is dependent upon the diameter of the fastener. And, of course, it's dependent upon the friction encountered when turning the nut. Neither are considered in the graph. Furthermore, following from the latter point, one can apply "maximum" torque yet there may not be any "fastening force applied".

The graph that you're trying to create is quite similar to the ubiquitous stress/strain curves for certain materials that we've all seen. In essence, that particular part of the wheel has already been invented. As has the Joint Diagram, which by referring to the characteristics of all components, does indeed do a fairly good job of predicting when a system would fail.

If you're simply trying to choose a bolt based on when it would fail ("fail" meaning break?) and you know how much bolt load is required, you'll initially have to make sure that it has enough capacity to carry this preload. Then, you'll want to make sure that this load is close to the fastener's proof load so that adequate "spring" is available. If you can ensure very close accuracy in the application of preload, you can even choose to go beyond yield (up to a certain point). As mentioned earlier, you also have to ensure that the bolt material is compatible with the conditions of its working environment. A caveat: Although this is the method that, unfortunately, the vast majority of people use to choose bolts, it is not correct because it does not consider the entire system (An example of a double-whammy no-no: Asking bolt suppliers for a 'torque spec'!).

Sorry: If one wants to properly choose a bolt, there are no short cuts

thank you BoltIntegrity

by fail, I mean when the bolt/nut/thread/bolted object lets go.

I put only the bolts in the diagram but I mean that it is representative of the system, and that different threading and diameters and materials would be more easily compared if there was one common graphic system for comparing all bolting arrangments.

I put torque from the point of view of a torque wrench as a means of measuring the turning force applied, which then creates the 'sandwich' force applied. of course it varies based on the bolt dia/thread.. again, I'm after a common system of comparison, but should relate directly to the tools used by mechanics. (torquewrench) so anyone can use and understand the results practically, where stress-strain curves are no so applicable.

everyone can understand torque wrench. (I hope ) and everyone can understand something like 850 pounds clamping force. (ie put 850 pounds on top of your hand on the table.. thats the force clamping your hand to the table.. sandwich)

I really appreciate your input. great stuff. but I still think there must be a way to make it easier to compare bolting systems. I would have to disagree about stress-strain curves being ubiquitous... and don't agree that everyone is going to understand them or be able to read them without having been through an engineering class to understand even what the difference between stress and strain are, technically.

but I will agree with you about the similarity of the Joint Diagram.. which is new to me.. again, I wouldn't really call it ubiquitous. (maybe for a bolt scientist such as yourself ) http://www.boltscience.com/pages/basics4.htm

Chris

To start a bolted joint design

- Joint purpose

- Joint configuration (flange type, with gaskets? type of gasket, gasket material, properties,...)

- stress on the joint, constant and variable portion.

- based on all these and a few other parameters, you should calculate the minimum stiffness required.

- Based on the stiffness required the flange loading.

- based on the required flange preload and the purpose- sealing? torque transmission (shear)? compressive? tensile? compound?.... calculate fastener spacing. (keep the pressure cone in mind)

- With bolt spacing and (hence) number of bolts calculate the preload on each fastener.

- Check the availabuility of fasteners that can take the load, adjust to nearest, iterate through the spacing and finally freeze these aspecs.

Now you have the number of fasteners, grade, pitch, hole locations and the required preload.

_{(Upto this is off topic)}

Now to answer your original question The bolt can fail on

a) fatique

ib) Tensile

c) Shear (torsional stress on the joint eg shaft coupling)

d) Thread stripping (shear)

or the joint can fail and that will sooner or later cause the bolt to fail.

all these are mainly due to

i) improper stiffness calculation and hence improper preload.

ii) Improper selection of fastener- including material.

So final result is for joint (including the fastener) not to fail, you should properly

select the fasteners (again stress including material and process, because we did have some bitter experience- most of the requirements say 8.8 or 12.9 can be achieved in inferior material by heat treatment- and they failed- not all at once, not making the equipment fail, but the number of failures were a bit more than alarming- then we specified the material and heat treatment along with all the mechanical tests, thread forming method etc - it is a bit too detailed to list everything here- and then company secret )

Prestress (preload) the fasteners, that includes the sequence and steps of pre-loading (these we have discussed in detail about 1 or 1.5 years back?)

Now is the crunch

How to prestress?

- Torque- most common and most deceptive tool. variation between assumed and actual is too high and this itself is enough to make the otherwise perfectly calculated joint to fail.

- Pre-Tension- better - only a bit relaxation is observed as the jacks are released. However we as a practice for our critical joints do it in steps to ensure at least nearest.

- Pretension with Length measurement - accurate enough but not always possible, we do it with certain assumption in most critical cases, if direct measurement is not possible. (Now a number of length measuring Ultrasonic equipments are available, however that is not accurate enough in a number of cases)

- Strain gauges or other measuring instruments - not always economical or practical.

(The Pre-tension is done by the heating method too in certain fasteners)

The whole moral of the story is there is no simple equation (or graph) that will, with reasonable probability calculate the torque to the stress. And for that with all the modern calculations, we see the number of fasteners and the stress far exceed the calculated, the redundancy will be 30-40% (optimistic, that I am )

"Now a number of length measuring Ultrasonic equipments are available, however that is not accurate enough in a number of cases"

You are both correct and incorrect in your assertion.

Those cases in which this method is not accurate enough is when it is used in the hands of the inexperienced and untrained. This is the unfortunate result of having lower-priced yet functionally-capable devices become widely available (granted, since we provide the equipment, we even contribute to dilemma). There have been numerous cases where we've been called to sort out the mess left behind by such users. Unfortunately, when that happens, the job usually has to be re-worked.

On the other hand, bolt elongation measurement accomplished by capable specialists is very accurate. The best that we've done is +/- 0.25% variation between indicated and actual. A range of 2-4% is realistic. Not bad when you consider that the torque-to-bolt-load variation can be anywhere between 40 and 100%

I appreciate that you have made a wonderfully simple rendition of the science of bolted joint design...

I agonize over still how complex it is.. (not your fault)

perhaps I'm engaging in wishful thinking.

but ga anyway.

Chris