Bolt Length

One of you is only thinking of the static assembled configuration. The one who believes this aids assembly, I think is correct.

During assembly there will be a gap between your two flanges. So during assembly when initially connecting the flanged assembly the first bolt, nut, and all washers will be asked to briefly hold all of the components together with this added gap between the flanges. During this brief period, you should have as many threads of the nut as possible in contact with the bolt to handle the loads that will occur during alignment and inclusion of the remaining hardware.

In some cases I think it is just a way of dealing with the issue of "is the bolt too short"? I have heard 1.5 threads exposed and others have told me 3 threads exposed. This was in equipment where the hardware was holding the electronics enclosure together. The extra threads served nothing beyond making the workmanship inspectable.

Not only, nut and bolt have chamfers so that if the bolt would have only the length to be flush with the nut surface at least one thread would be not engaged so that the loading capability in shear would be less required. From an other point of view this over length considers all possible deviations in the thickness of components so that in the worse case all nut threads will be engaged.

it can be argued that the last threads in the nut really don't do much either since it's the first few threads that take most of the load.

Explain that in terms of an unbroken inclined plain

Bolts stretch.

I wonder if there is an argument for a slightly variable (pre-tightened) pitch for units in very precise assemblies?

Bolts stretch doesn't explain the first three threads taking more load than the others.

I think the variable pre-tightened concept is covered by the various classes of thread.

The tension in the bolt at A is constant; is decreasing through B towards zero at C.

Therefore the slope of the thread in the bolt is steeper at A decreasing through B to C.

The effect in the nut is the opposite but not as pronounced and not shown here.

I hope you can just see that the slope of the thread in both nut and bolt is parallel in the upper section, but curves so that it is steeper in the bolt in the lower section.

Dear Randal,

As a retention feature this has merits (such as tri-lobial bolts or uneven pitch bolts).

My concern with all these retention features however is that they tend to increase friction significantly, reducing even further the energy available for clamp-force. Then you need to increase the bolt material grade, then the cost goes up. Also, it is virtually impossible to calculate and predict the behaviour of such joints, so joint development relies exclusively on physical testing which is also expensive and time-consuming.

Good points, the number of time I've encounter the philosophy that tight threads equal a preferred engagement eludes me. Cheater bars aren't needed to thread a nut onto a bolt

You have a wrong view angle of the needs for such anti loosening features.

Without them the bolt/nut pair is hold in place only by the preloading generated friction which has to be bigger than the active moment generated by preloading and thread angle. If for any reason the preload decreases then holding capacity decreases , it is an unstable process.

The different approaches (trilobs or similar) have the advantage that the friction forces which hold the bolt/nut together are NOT any more depending on the preload magnitude !

Of course friction is higher and torque has to be increased according to the higher losses at tightening but this is via tests possible to define and adjust.

With respect to material quality the trend to miniaturisation leads anyway to smaller bolts and the stronger material allows a higher ration L/d which is better for fatigue. Since Young modulus is the same independent of elastic limit, for same load, a high strength bolt will be more elastic and reduce the variation of bolt forces under load thus being more resistant to fatigue.

Thank you for taking time for this and yes I understand the point you're making but it seems odd especially in this day when even wood screws are having rolled thread that the threaded portion of bolt encapsulated within the nut should be deemed of less integrity than the remainder/shoulder or unthreaded length of the bolt.

I ran out of time to add

I would expect stretch between A and B to reflect the stretch between B and the bolt head

I strongly disagree with this. In the circumferential direction the bolt is in compression and the nut is in tension. When the bolt is just as long as the nut, some of that compression deforms the end of the bolt and makes it get smaller in diameter. The thread engagement is reduced. I understood that this is the reason for having threads protruding from the end of the nut from a calculation point of view.

The radial forces are the same for bolt and nut. But the radial stiffness of bolt is a lot higher than the nut stiffness so that the major deformation will be the nut one and this leads to an increase of mean diameter of the nut and a reduction of the shear area.

This is the basis for the ratio between nut hexagon and bolt diameter: the nut has to be stiff enough in order to limit this radial deformation which penalizes the carrying capacity.

If for any reason you should use nuts made from a material having a young modulus smaller than the bolt the hexagon has to be increased to compensate the higher radial deformability.

However, when the bolt is cut short (say flush with the nut) then the end of the bolt will dimple under the compression forces which in turn reduces its external radius and so the engagement of threads with the nut (which itself has expanded and so also reduced its engagment).

So the final capacity is not as per the calculated value.

Effectively, the different codes that I have managed to find that reference the number of additional threads beyond the nut have a varied view on the number of threads between 1.5 to 3.

Of course this is all to do with bolts in tension and not shear.

A couple of remarks:

1- the load is transmitted by ≈3 threads at the OPPOSITE of the bolt free end!

2- the nut diameter increase and the shear force reduction CAN be calculated thus it is possible to know from the start how much the bolt will hold in tension!

3- the standard nut dimensions (as mentioned) were defined so that their radial stiffness will be big enough to avoid such portance losses. Never the less the bolt stiffness is SOO big in comparison with the nut that its own compression can be TOTALLY neglected.

So that from the beginning the probable load carrying capacity

I have several FEA results I can send you, I have not any more the original jpg files (are embbeded in power point and word files and I do not know how to record them again as pictures and integrate to a message on CR4) so that I can send only if you give an email address but not here only on the private channel.

I have read through this forum and a lot of excellent points have been made. Your 2 posts which also include the imaging make the most sense. Although referred to by many as "standard engineering practices", would you have a reference to this quote some where? I have used the practice of 3 threads minimum since doing drafting in high school but now doing inspections, unless there is written reference to refer to it can be tossed out by the Canadian courts in the case of an incident or at bare minimum a very rough ride for the person testifying. I prefer to have all my ducks in a row.

We have 2 hard fast sayings which hold true;

1 - Confucius is a wise man, a picture is worth a thousand words;

This you imaging JPG which is excellent

2 - If it is not written, it does not exist.

I believe #2 is written somewhere but I am struggling to find it. Any help would be much appreciated.

Best regards,

Thank you for the comment. As you would notice the comment was not so appreciated and it got only 1 positive note so that it is not considered as soooo good!

How ever the rule of the 3 threads carrying almost all load comes from the stiffnesses play between bolt and nut (or female thread). I shall try to find a picture I saw where the distributions under "normal" conditions and when a helicoil is used are compared with of course an advantage for the helicoil since its torsional compliance allows a better distribution, more progressive.

I will post it as soon as I find it.

Much appreciated. Do you mind if I use the pictures to present a change to code within our code committee? I see you have a separate post to which you need an email to send a better copy of the stress pictures. If you allow it, would I be able to get a better copy of those pictures?

Best regards,

Yes provided it is mentioned the origin.

Send the mail and I shall see which are the best I have

Thanks to recommendation I recovered the pictures as JPG file and can display them.

The results come from different FEA with different software and threads done under different contact conditions, but all of them are consistent in following directions:

- Load is transmitted by the first threads

- End threads are NOT for load transmission but for all other mentioned reasons or at least for some of them.

- Nut radial expansion (which visible on the 2 pictures at top) is very small and the bolt has almost no radial deformation.

- The loading can be from start correctly estimated.

The picture top-right shows also how the nut thread is deformed under load (the small red circles indicate the region where the nut is supported by the parts considered as stiff).

It is also interesting how the bolt is loaded and how the stress expands within it.

The bottom picture shows also how every bolt thread root is a stress concentrator.

If from a qualitative point of view all comments are valid it is different if the quantitative aspect is considered.

Hope it will make some aspects clear.

Thanks again for the software indication it can be a real help.

Nick Name

Great pictures. Thanks for posting them.

I guess that this is a standard height coarse thread with about about 5 threads. Even with a reduced height lock-nut we would still have 4 threads.

I'm happy to see that the bolt does dimple slightly at the top. This had been reported to me by a nut-bolt specialist and it is good to see the confirmation from your pictures.

So if the bolt head was flush with the nut, the other remaining engaged threads would have a higher load ? I imagine that for a standard height bolt with 5 threads, this would be reduced to 4 threads (more or less).

There is no need for it though it may help in assembly. Bolts though are manufactured in standard lengths. Better to chose one that has a few threads sticking out then one that you can't get a full nut on. Failure to have all the threads in the nut when torque it down may cause the threads to fail usually on the bolt but have seen them pull out of the nut. To obtain the nut flush with the end of the bolt means you would have to start your tightening before you had a nut full of threads.

Thanks for all your replies and Best Regards ,

Hello Guest,

What a great question

The extra length is generally an allowance as bolt lenghts come in 5mm or 10mm increments pending on the dia. it may be 25mm increments therefore a bolt lenght would be rounded up to the nearest length in the ideal world. In many cases to exert clamping pressure you may only need one thread but this would be dependant on the dia. and clamp pressure required.

hope this helps

Anthony@alno

http://www.alnoproductservices.com.au

The following reasonings are offered from my end : -

1. Bolt length should always be more than the measured length of parts to be assembled because many a time the parts being loose or difficult to access, shall require pre-holding of parts on bolt length so that bole end is available clearly for engaging nut.

2. Actually bolt length required is = measured length of parts + one nut width and any extra length does not serve any useful purpose. Infact extra projection shall be an obstruction while dismantling at a later stage due to corrosion , dust particles sticking and jamming the clearance between bolt & nut threads. Use of end caps are advised in this regard.

3. Sometimes exact required bolt length may not be available and instead of cutting and using, it is better to use bolt of nearest length available.

ANSI piping was designed for outdoor Oil Industry piping. The bolts had to do the same service. But contractors try to cut costs so bolting standards were included as part of the Piping Specification. For maintenance purposes it was required that bolts not be too long, the exposed thread would rust. And also long enough for Assembly, allowing for gasket compression. The requirement was settled by half the diameter of bolt maximum, and the bolt end to show. Single plain washers were standard under bolt head and nut. Lock Nuts (thinner than full nuts) were required where vibration from machinery would affect the piping.

Are you suggesting a longer bolt will cause rust?

And a short bolt there will be not rust?

Anthony@ALNO

www.alnoproductservixces.com.au

I think that it's quite clear and obvious:

Doesn't matter if a bolt is "long" or "short", any exposed threads will indeed be susceptible to corrosion. Not only that but, also mechanical damage. Thus, whenever longer extensions beyond the nut are required (as in the case of tensioned rather than torqued bolts), it's always prudent to use thread protectors.

Do you mean jam nuts? as lock nuts typically designate a nut with some form of mechanical self locking feature. Ie, Nylock, center lock, distorted thread, serrated flange, integral lock washer, etc...

This is a question of regional terminology. A secondary nut (which may or may not be thinner than the primary nut) added to lock (or perhaps jam) the primary is commonly known as a lock nut (or locknut) in the UK. Can't speak for any other regions.

In my experience Nylocks etc. are normally referred to in the UK either by the proprietary name, or as "self-lockers" (with a type qualifier where necessary).

The thread at the start of a bolt is usually incomplete, i.e. it does not meet the geometric requirements for an ISO thread and cannot be used to take torque.

This is because of the method of manufacturing threads which leaves a short length incomplete, and is made worse by normal piece-to-piece variation.

Therefore, it is usually recommended that bolts are chosen to slightly protrude at the back of the nut so as to ensure that the incomplete portion of the thread is not engaged in the nut. The recommended protrusion in my company is at least 1.5x thread pitch. In practice of course we do not design every bolt from scratch, we select a bolt from a list of standard sizes. Therefore sometimes the only available size happens to be even longer than 1.5 threads more than the clamped piece, and so you see some joints where the bolt protrudes a lot at the back.

In addition to this, you may find bolts which have a short, unthreaded, cylindrical portion attached at the tip and this make the bolt stick out even more. This is called a "pilot point" or I think "dog point" in the US. It is there to guide the bolt into the nut and prevent cross-threading during fast assembly

bolt drawing

Hope this helps

The people that answered this have some good opinions and a good set of handy guidelines. There is no use for extra threads unless you may want the option to put on a "jam" nut or safety wire. The exposed thraeds are sometimes sharp enough to pose a safety risk in which case you would opt for an acorn nut or plastic cap. If extra threads aren't used for drawing components together. Design the bolt/assembled unit without them.

Loctite makes good products for securing theaded items.

Good assembly techniques include removal of sharp edges to avoid safety hazards...

Other than using the bolt projection as a visual inspection method, as a safety precaution, I'd rather have extra bolt thread projection especially where there are vibrations or load cycles present in a structure or assembly. No matter how hard we try to torque the little buggers down they always loosen up!

Making my 2 Cents known! Hmmmmm I wonder how magnificent a Moose quarter would look like if the US Treasury ever gets around to making 'em!!! LOL Anyone interested in signing my petition????

Dear Guest,

What holds the joint together is CLAMP FORCE, not FRICTION. Clamp force comes from bolt extension. This comes from torque on the bolt head against the thread gripping the nut side. There is probably some theoretical merit in the previous comments about only a few threads actually taking the majority of the load, however when designing a new joint for mass production you need to design it in such a way as to achieve the full required thread engagement or else you may have some threads stripping under the torque or others coming lose in operation.

The correct design thread engagement varies from joint to joint and does need good calculation and/or testing (e.g., torque-angle or torque-clamp force testing) before it can be decided. As a rule of thumb start with 1.5 bolt diameters for bolt into steel pieces or nuts and 2 bolt diameters for aluminium.

PS1.: Loctite may be a good solution for one-off fixes on my classic car but is not going to hold joints together in production. At best it may prevent a bolt from coming off after it has already gone lose, i.e. the joint already failed.

PS2: On a typical joint, friction takes up about 80%-90% of the energy you put in through torque! So only about 10-20% of your torque is left to extend the bolt and create clamp load. Do not waste this energy further by bad joint design. Follow the rules, there is good reason for them.

"What holds the joint together is CLAMP FORCE, not FRICTION."

The clamping force without friction will not maintain the bolt/nut in place !

There are 2 aspects:

- holding together the parts against transverse forces → the clamping force has to be so big that the friction force it generates on the potential sliding surface will be more important than the transverse maximal force. This is the reason this force has to be maintained all time long and which ever will be the loading of the pair.

- holding together the pair bolt/nut → this is more difficult since any even minute sliding on the thread surface will get the bolt/nut loose. Axial forces having a neglectable effect on it. This is the reason the machines used to check the performance of an anti moose principle are based on a transverse sliding system. Here again the combination between axial force (clamping) and friction on the thread contact surfaces will keep the pair together or not. For THIS is a bonder as LOCTITE or similar is good since it maintains the threads after preloading together and does not allow a loss of preload.

It is a bit more complex than it appears at first look.

Lots of good answers here.

The standard in aviation is minimum 1.5 threads through the nut. You'll hear lots of design reasons for this but I can say from experience that anybody that has to do pre-flight or post-flight inspections really appreciates the excess.

From my experience in helicopters it is really nice to be able to glance at an area and quickly visually verify that everything is in place. If a bolt did not significantly protrude through a nut the inspection process would be much more tedious.

Also, on an assembly where a nut is not visible it is nice to be able to run a finger behind and verify everything is intact. If I couldn't feel the bolt protrusion, that would prod me into a closer inspection of the area.

Hooker

Hooker,

It's good that you point out the importance of inspection. Indeed, physical inspection is a great way to ensure that the bolts are still there. The next most important issue is, "is the bolt still tight enough".

To address this, special bolts are available which have an optical sensor that changes colour based on the degree of clamp load. In this example, if a bolt has loosened, the sensor shows red. When re-tightened to the proper load (irrespective of torque) the sensor becomes black again.

Proper elongation measurement is a more definitive way of measuring clamp load but, as a simple "go, no-go" indication, this technology seems to be quite interesting.

Wow, love those optical bolts. I can think of a couple of helicopter accidents where having those in critical systems may have prevented the mishaps.

"When re-tightened to the proper load (irrespective of torque) the sensor becomes black again."

So, if not by torque, how is the bolt installed to the proper clamp load? I'm sure I'm probably missing something outlandishly simple here.

Hooker

Think about what happens with a bolt that binds from galling or debris in the thread. You can bring the wrench up to the specified torque but not have the bolt's head touching the flange.

Agreed, to a point.

If I ever found a licensed mechanic that allowed that to happen on an aircraft, I'd make sure his license was pulled and that he'd never work in the field again.

There is no excuse for that to happen, and even if inadvertent, would "probably" be caught by the inspector at sign-off. Unless, of course, there was corrupt maintenance management shortcutting the practices and processes.

I am speaking exclusively of the aviation field.

Hooker

Exactly. Even if the bolt's head does touch the flange, variations in friction factors arising from lack of lubricant, excess lubricant, lubrication contamination, dirt in the threads, damaged spotfaces, washer inconsistencies, thread damage (and on and on and on....) the "proper" torque will likely not produce the desired (required!) clamp load.

Interestingly, torque wrenches are usually required to be calibrated to +- 2%, or so. This is absolutely useless when the use of these "calibrated" torque wrenches can still result in preload variations of +- 20, 30, 60, 80 or even over 100%!

So load would be the clamping force under the head?

Hello Hooker,

I'm sorry for the confusion. Indeed, the bolt is tightened by "torquing". This can be done with a torque wrench, a hand spanner or even by a hammer and flogging spanner. This is because it doesn't matter what resistance is encountered when trying to turn the bolt/nut: the sensor only changes colour once the appropriate bolt load has been achieved .

As noted in a previous post, one may have the "proper torque" yet the bolt may still be loose. Actually, if less friction is encountered, a "properly torqued" bolt can even be too tight. Load verification such as elongation measurement or by visual inspection utilizing this technology eliminates these insipient risks.

Thanks for the clarification. I certainly understand the risks.

I really like the load verification by elongation, and will certainly keep these in mind for future design tasks. There's always some worrisome place where we can't use normal retention practices (safety wire, loc-tite, etc)

Hooker

Actually, there are two additionnal concerns that may be particular to the aviation field :

1) Showing a MINIMUM of 1.5 threads with nylon insert locking nuts insures that any shear force on the pieces assembled with the bolt is effectively applied against the shank of the bolt (unthreaded part).

2) Showing a MAXIMUM of 3 threads with nylon insert locking nuts insures that the thread of the nut have NOT started to make their way into the untheaded part of the bolt (shank), hereby damaging both nut and bolt.

Interesting discussion !

Are you aware that there are special structural bolts available that are torqued to a point where the threads deform, creating a permanent lock. These bolts cannot be reused if undone and must be replaced with new bolts/nuts. They are used in bridge construction and in buildings steel joints where they are hidden from sight and must be relied upon to not loosen.

Indicate the source, manufacturer, address, catalogue, .... Thanks Nick Name

Are you aware that your description apply to all properly tensioned bolt joints?

No, bolts can be used under conditions several times.

What the guy said (in an improper way which could be considered as arrogant or at least impolite) is that some bolts when torqued over a threshold deform the threads so that a mechanical lock appears. I was informed of such bolts but I was not able to find sources this is the reason for my question. In fact as far as I know it is the nut which is deformed so that a zone is plasticized and generates in the thread a friction torque INDEPENDENT of the preload. This is a guaranty for a reduced risk to have the bolt/nut pair going loose. Anyway such solutions cannot apply for machines and are used in structures where the loads have not same dynamics as in mechanical systems.

I've encountered "torque-to-yield" in automotive applications but not in civil. Perhaps the writer refers to the colloquially-termed "twist-off" bolts in which a necked portion of the bolt's shank breaks loose upon reaching a certain torque. The nominal threaded portion remains and is locked by the nut. Notwithstanding the need for specialized assembly tools, the obvious problem with these is that the nebulous torque/tension relationship still applies. As such, actual bolt load when the spline breaks off is still an unknown value

The component is called "TC Bolt" from tension controlled and is intensively used in structures for civil engineering for many reasons related to cost and quality. Here are a few infos about component and applications:

The right part is used as reaction for the torque and will shear at the torque limit given by the grove at right end of thread.

The shaft is ONLY under tension the torque loop closes at the right side between thread nut and right end.

The tool has a special design with a holder inthe middle for closing the torque loop.

An example of civil connection between big profiles.

Use in railway connections.

A tpical assembly for TC bolt: a bridge structure.

An other one in London near to Big Ben

Other structure same fastener.

The grove is such that the torque at shear is near to the one needed to stress the bolt in the right strain range.

The only problem with those bolts is that in most cases they are short and thus cannot be used where transverse loads have a high frequency. It is but possible to use them in calibrated holes and then the side movement which is the most dangerous for loosening is maintained at a very low level.

The biggest problem is:

" ... the nebulous torque/tension relationship still applies. As such, actual bolt load when the spline breaks off is still an unknown value..."

Granted, there is much meat left in the safety factor of civil bolting applications. Thus, there may not be as much of an issue here as there is in critical industrial bolting applications. However, it's misleading when the description suggests "... used... for...reasons related to...quality".

Quality is knowing what you've got. This is still, unfortunately, "guessing".

Point well made BoltIntegrity. These bolts show what torque had been applied which need not be the actual bolt load. I would expect that the bolt load will change on these bolts as normal load conditions change on these civil engineering projects.

But I wish to compliment you, Nickname, TVP45, Trevor Walden, and Daedalus for such an informative discussion. This type of engineering insight is precisely why I linger on this blog.

Ok. All this is good discussion, but does not answer the original answer.

The reason for the extra thread (exposed past the outer face of the nut) is to prevent yielding of the threads under the nut at the end of the bolt. The normal standard is to assure that 3 threads are exposed. With fewer threads exposed, there could be a risk of the bolt threads yielding in shear under the nut. Exposing any more threads is not necessary and could be wastful. Note here that the bolt yields before the nut (in correctly specified connections).

Also note that, for many applications, this characteristic would preclude reusing the same bolt for the same application. However, since yielding usually does not occur, this normally is not a problem for applications for which this limitation is not specified.

Some comments on bolt tension: See my next post, which I consider to be off topic somewhat.

Where can i get a copy of the normal standard?

Anthony@ALNO

www.alnoproductservices.com.au

You are right but in this particular case the dispersion is a lot less important than in other cases for following reasons:

Geometry: the grove is manufactured with a tolerance of about 0.1mm which represents 1% of the diameter even for the smallest type. surface. So that the torsional modulus will have about 3% dispersion.

Material properties: elastic limit has a dispersion of about 4% since the bolt are made from a high quality steel and heat treated.

Friction: nut and washer come from same supplier and have same surface conditions, in general the washer is large and thick (to assure a quasi plane contact since the structural steel is weaker than the bolt or nut). This guaranties a narrower dispersion of friction coefficients and contact conditions between nut and sliding surface (washer only!). Same conditions for the friction between threads. U estimate the dispersion at about 4..6% based on measurements.

All together the dispersion of the applied tension will be (3^2+4^2+6^2)^0.5=7.8%.

Due to the very very high redundancy of structural joints (you saw the number of bolts in one example) the result will be very near to the average so that it is not so bad.

Of course this fastener is not of use for applications with only a few bolts as in automotive, but could be of interest in airborne assemblies where the redundancy is also high imposed by the thin walls with limited local carnying capacity. To make a comparison, for a car in general an assembly consists of about 3 to 4 bolts working together (one side of suspension), for a small plane one wing is fastened with over 150 bolts to the main body and for A380 there are ≈2500 bolts on one wing side.

Lug nuts etc. are torqued and re-torqued many times but many applications require replacement of the fasteners each time.

Hi There

There is indeed need for that extra threaded length, because there could be vibrations in the assembly, Vibrations can b caused due to thermal contactions and expansions due to climatic changes. That extra length can prvent the nut from falling off and thus save your assembly from the failure.

From a retired Union Steamfitter's perspective, you guys are straining at gnats. This applies to the 1/2 to 1 inch or so B7 studs and Heavy hex nuts that hold together most of the flanges in US refineries. Only at nuclear sites have I seen torque wrenches used for nuts smaller than about 3 inches. Elongation measurement is obviously best. One obvious reason to leave extra length is to allow for a slip blind and its extra gasket when isolating parts of systems. On 150# to 600 # series flanges, the flange nuts get pulled up in stages until final tightening with two combination wrenches linked together for leverage. If it leaks on hydrotest, go round and round with a knocker wrench (what one poster called a thrashing spanner) until it tightens up. Most of the time this extends to the largest pipe sizes. Make the contractor see that the studs are buttered with appropriate compound, Coppr cote, Ni antisieze or whatever from one end to the other. Demolition of old worn out units often reveals studs with intact threads near the nuts, and rotted away in the area that didn't get any bolt dope.

"you guys are straining at gnats ..." - that's as maybe, but it doesn't wash with a customer's inspector when you're trying to get a new piece of gear passed for payment.