Can someone explain how tension and torque relates to bolted connections?

We'll try our best. The relationship between tension and torque should be looked at cautiously, since it is very difficult to indicate the range of conditions expected to be experienced by a fastener. Torque is simply a measure of the twisting force required to spin the nut up along the threads of a bolt, whereas tension is the stretch or elongation of a bolt that provides the clamping force of a joint. Bolts are designed to stretch just a tiny bit, and this elongation is what clamps the joint together. Torque is a very indirect indication of tension, as many factors can affect this relationship, such as surface texture, rust, oil, debris, thread series and material type just to name a few. Virtually all the torque/tension tables that have been developed, including ours, are based on the following formula:

T = (K D P)/12

  • T = Torque (ft-lbs)
  • D = Nominal Diameter (inches)
  • P = Desired Clamp Load Tension (lbs)
  • K = Torque Coefficient (dimensionless)

The value of K is a dimensionless torque coefficient that encompasses variables such as those listed above, as well as the most significant variable, friction. The value of K can range from 0.10 for a well lubricated/waxed assembly, to over 0.30 for one that is dirty or rusty. The values we used when calculating our values are:

  • 0.10 = Waxed/Lubricated
  • 0.20 = Plain, as received condition, slightly oily
  • 0.25 = Hot-Dip Galvanized

The appropriate torque value to use in a specific application is best obtained by using a calibrated torque wrench and a Skidmore-Wilhelm load indicating device to equate actual torque to the desired tension. For ASTM A325 and A490 structural bolts The Research Council on Structural Connections (RCSC) recommends:

The pre-installation verification procedures specified in Section 7 shall be performed daily for the calibration of the installation wrench. Torque values determined from tables or from equations that claim to relate torque to pretension without verification shall not be used.
RCSC Specifications, June 2004, pg. 62, 8.2.2

An alternative and more accurate method for assuring proper tension would be to use a direct tension indicator or DTI. These are available for use with ASTM A325 and A490 structural bolts and are engineered to compress at the proper tension, assuring the installer that the proper clamp load is achieved. Hopefully, this short introduction to bolt connections helps address some of the confusion surrounding this issue.

See Estimated Torque Values

Written ,

65 comments

    Dan, I was taught, just what you are saying, that its really impossible to pin down, and all these reasons specified plus many more come into play. For instance the values assume there is a bolt and a nut, which has a standard size. If it was into a drilled threaded hole, it would be entirely different as the depth of penetration was greater, since, obviously — more threads and increasing friction.

    For newcomers I’d tell them what i was told. Assume a very rust, pitted, and gnarled up threads. Assuming you could get the nut to go on through that messiness, and no threads were ruined (i.e. gone) – well, that can’t be determined except that its likely a helluva lot worse then great/perfect threads that are well oiled, right? And I have a fabricator that makes screw parts for a government special application and the grooves, and the peaks of his threads have to be much better meshed than most standard bolts, and they’d likely have different torque values since a lot more surface area is touching on each groove/and thread. When i was told that, it makes a lot more sense, and some common sense gets thrown in when we use simple formulas and standard configurations. (Some say galvanized is similar to the boogered up threads i
    described above – and it sometimes is that bad..).

    Thanks for a great rule of thumb, though.

    My one question is, why doesn’t any chart ever have an ‘oiled’ condition. I realize wax is a lube, but its mostly to get over the rough texture of the galv, i feel. Its good, but its not the same as light oil on the threads, is it? clearly the oil, regardless of its quality is squeezed to a molecular layer, or completely gone when a bolt is fully torqued (on the friction surfaces), but i’d think it’d be a lesser K value. right? Why don’t they speak to that?

    My second question for today is, if a bucket of bolts is rusted (and obviously that’s not good, and also they’d all be rusted imperfectly/differently). And, lets assume its surface and not pitting rust, so the cross sections of all parts are all the same. but obviously there is no clean smooth factory surface. And lets say with good intentions they brush them lightly and oil them. Comments, and if you had to extrapolate, would you use a ‘waxed galv’ type K factor, or what?

    @Barry- The ‘plain’ finish assumes some residual oil. Of course there will be differences as some will be more heavily oiled than others, that is why these torque values and K factors are estimates only. For your second question, I’d say those bolts would fall into the ‘plain’ K factor still, but that doesn’t mean that the K factor will be exactly 0.20. There will always be variations that will be dependent on surface condition, oil type and weight, etc.

    WE HAVE TO CALCULATE THE CHAIN NUMBER FROM TORQUE.

    SO AFTER SOME RESEARCH WE HAVE SEEN THAT IF WE HAVE A TENSILE STRENGTH, WITH THE HELP OF IT WE CAN FIND THE CHAIN NUMBER. BUT WE ARE NOT GETTING THE RELATION BETWEEN TENSILE STRENGTH AND TORQUE.

    CAN WE HAVE A GENERAL FORMULA TO CALCULATE THE CHAIN NUMBER FROM TORQUE?

    PLEASE SUGGEST.

    @Himanshu- We are not certain what you mean by the chain number? The normal torque formula is T=KDP/12 which is further explained in this article. If you have some of the variables, you could solve for the one you are missing.

    Can you explain why the helix angle of the threads doesn’t play into the relationship between the tension and torque? If it’s encompassed in K does that mean that K changes with thread type? Or is it that the majority of torque is absorbed by friction (which is proportional to clamping force) and not reaction force from the thread ramp?

    @Nick- I think you can spend an eternity going down the rabbit hole of the torque-tension relationship. Many things affect the K factor, the thread type would certainly be a part of that. However, after much research, the IFI and many engineering organizations have decided that the T=KDP/12 formula is an easy and reasonably accurate approximation. I am certain that if you wanted to break it down further and dissect it from several directions, there are mathematical ways to do that. For ease of use, and general purpose questions, we’ve chosen to use the simple IFI formula.

    During actual process of bolt tightening esp. by using impact wrench, how to control the possibility of over-tightened bolt (more than the tension capacity of bolt) whereby, only at 70% of bolt tension are the most workable torque condition.

    @Leo- The best way to to it is to do some testing on a skidmore with the actual lot of bolts, and determine the proper torque. Using a calibrated DTI may also help in zeroing in on your desired tension.

    Hi gens,
    I’m a marine engineer by profession, love learning in details… I think I need to equip myself with this torque issue, how can I start?
    Also would love to have an idea on bolts making

    @Fego- Thank you for your interest. Besides the library of information available on our website, you can also find technical information through the Industrial Fasteners Institute. We also have several videos on our website showing the various manufacturing processes.

    Great work, however a simple manufacturing method is fail 10 bolts and take the average and reduce by 10% of the failure point.

    hi i am robert
    manufacturer high tensile bolts
    also torquing and tensioning tools
    whenever i go to on site for torquing i use tensioning tool so how calculate torquing value in tensioning tool

    @Robert- We are sorry, but we don’t have any experience using tensioning tools, so are unsure how to use them to calculate torque.

    @RD- Torque requirements will vary by application type, so we can’t make any specific recommendations. Additionally, socket head screws usually thread into tapped holes, whereas grade 8 bolts may be used with a nut, so it isn’t an apples to apples comparison.

    Dan,

    We have tested few Grade 8 Bolts & nuts Plain with slightly oil finish of 1.1/2-6 .
    And results are torque is approx. 3146 Nm – 3164 Nm, Pretension force is approx. 455 kN – 497 kN & Torque co-efficient is approx. 0.166-0.183.

    So, let me know your feedback.

    @Jugal- We are not surprise that your values vary a bit from ours. Torque is a tricky and unpredictable measure of tension and can vary from bolt to bolt depending on the conditions.

    Hi, I used 5/8″ bolt with ASTM A36 material and calculated Torque value by above formula. Through calculations I got 70 N-m torque value but in actual practise when I measured, it came around 30 N-m until there was a bent in Bolt clamp bar. I used Friction factor of 0.15 and calculated. Not able to understand why am I getting this huge difference.

    @Asmita- Torque is a tricky subject, and there are many variables that can come into play. Variations are common, although yours seems excessive. Without having all the information, we are unable to help, but I would start by double checking your values to make sure an error was not made. That said, 70N-m seems quite high for 5/8″ A36, I would have expected 60 N-m or lower with a K factor of 0.15.

    dane,
    Just curious about something I learned – one design called for 1 1/2″ Gr 8 bolt with 2200 ft-lb tightening torque. now the engineer claimed that by specifying a 1000 ft-lb tightening torque will be adequate based his calculation. my comment to him was that then don’t use 1 1/2″ size bolt and go to a 7/8″ or whatever, so that I can save some money. the other aspect of the bolt property for 1 1/2″ size, in order for this bolt to be effective was that ~ 2200 ft-lb is to apply. please advise and thank you in advance.

    @Steve- Torque is a crude method of measuring tension, so more information would be required before a determination could be made. I would discuss more in depth with your engineer to see what alternatives might be possible.

    What’s amusing is that this equation doesn’t even slightly resemble the ideal model of the simple circular wedge machine that a bolt actually is. It turns out that this is because friction dominates the demand for input torque. Even though it is the thread pitch term (which cannot be seen here) that actually physically generates the tension, it becomes almost irrelevant in calculating it!

    I have a question,

    Someone told me that if I use an helical spring in a particular connection, I shoul decrease the torque applied to the bolt by maybe one half?

    I am not a specialist, but I think that it is not right! We should use the same torque??

    Am I right?

    Thank you very much.

    Y. Lavoie

    Specifying bolts sizes and torque value requires you first, analyse the force interaction at the joint you’re connecting, then calculate the tensile stress area based on 75% of your selected material proof strength. Then specify a bolt size. These days, people just specify over and above what is required without any engineering!

    This question is a bit off the beaten path. I am reinstalling a handhole cover on a tank. The drawing gives a torque spec of 27 +/-3 ft/lbs. Original gasket was Grade 1 (65 durometer) rubber. Only gasket material I can get is Grade 2 at (80 durometer). Does the increased hardness of the gasket material affect my torque value?

    Due to stocking of hardware and varying applications, we use a majority of ASTM A574 Screws. If we are torquing ASTM A574 Screws to a Grade 5 torque(due to lesser application requirements), is it feasible that we are more likely to have screws working loose? My thought is that we are not putting enough tension into the higher strength screw.

    @Jeremy – It would all depend on the application. Static loads may be OK, but if there is movement or vibration, you may have an issue. We don’t have any engineers on staff, so we can’t really make a determination. Apologies.

    Hi,

    I am looking for bolts that have rounded/dome ends. Do you know where I could find some? Specifically, I need 1/4-20 stainless steel bolts with rounded ends. I don’t even know if there is a special name for such, but the closest I’ve come to finding them is this website, and I searched for “round end bolt”

    Thanks!

    @Steve – We use 0.10 in our torque calculations, but I have read anywhere from 0.10 to 0.15 is realistic. It will likely vary in that range from manufacturer to manufacturer.

    Can you explain to me what will happen if we apply more torque on the installed bolt? Example, the required torque is only 200 ft-lb… then we apply 250 ft- lb…. what will be the effect of this on the structure and on the bolt itself?

    @Mel – Without more information I cannot be certain, but the possibilities include snapping the bolt, crushing whatever you are bolting, or maybe nothing. It will depend on the capacity of the bolt in question, what you are bolting, and the amount of friction in the bolted assembly.

    I used M10 10.9 allen head screw on the aluminium rail. for assembly of bus seats..under rail there is sheet. ı searched a lot. everybody is saying aluminium reduce the torque value.
    ı found 33.9 Nm and 20Nm. difference is too big.
    whats your idea?

    @Ersin – Thank you for your question, but our area of expertise is with carbon and alloy steel structural applications, we don’t have any information regarding aluminum Apologies.

    @Jared – These torque charts are calculated for use in steel to steel connections. Anchor bolts are not typically torqued to a specific requirement, as the AISC says that snug tight is acceptable.

    Any guidance for 3/4″ lag bolts in SYP timber? Specifically a railing base plate with lags into stringer below (perpendicular to grain), with concerns about torque loading that twists off the lag and how much to reduce this torque to avoid lag damage? Do you have maximum allowable torques for lags?

    @Duncan – Apologies, but we do not have any engineers on staff, so cannot comment on application-specific questions like this. I am not aware of any allowable torque values for lag bolts.

    P = Desired Clamp Load Tension (lbs), is this the value given in kips on the RCSC code, if so how to convert to lbs

    Hi, I need the Torque report format, Could you provide format(sample)? Until we dont have the format.Thanks

    For 5/8″ Step bolts, of SAE grade 2, what torque value should we be specifying during installation. This is for a transmission line tower.

    @Lily – That is a difficult question to answer. We have torque values on our website, but every application is a bit different, so there is no one all encompassing torque value that works for all applications. Controlled lab testing is the only sure fire way of determining your specific torque values. /technical/bolt-torque-chart/

    The same above mentioned formulae are used for calculating torque for countersunk screws or any other multiplication factor available…???

    @Manoj – The torque formula would differ from the one listed because screwing into tapped holes usually creates a longer thread engagement which can affect the friction coefficient. An engineer would be the best person to consult regarding this.

    Hevii Guy,

    There is no such thing as an overtightened connection, until you break the bolt in tension. RCSC – Sec. 9.2.4, states: “A pretension that is greater than that specified in Table 8.1 shall not be cause for rejection”

    @Mike Girard – You are absolutely right. The content of the post has been updated. Thanks for catching that.

    Direct Tension Indicationg washers provide a false sense of security: They cannot indicate whether the joint has been overtightended. Neither do they provide a means of checking load migration after initial installation.
    Skidmore-Wilhelm machines provide an accurate indication of torque to load but ONLY on the exact fastener being measured under the the conditions experienced in the device AT THAT TIME. Once the tested wrench is used in the field, myriad friction factors and load interactions mean that all bets are off; one has no way of knowing what the actual load will be – it’s still a “guessing” exercise unless load is verified AFTER the bolt has been tightened and then again, after all of the other bolts have been tightned (to compensate for load transfer)

    That’s why head gaskets, and heads, on race cars stay in place, since the bolt tightening sequence, pattern, and steps upwardly in the force is set so everything goes down evenly and then is tested to a rather precise torque. And those are commonly almost sterilized clean holes and very well lubricated threads.

    If you are tightening the lid, or butt end, to a high pressure vessel, it wouldn’t hurt to do a pattern either..

    But the ironworker buttoning up a connection, and the usual connections of steel to steel in structural, is overbuilt, and is not subjected to the kinds of scrutiny that you are speaking about.

    When we have failures of structures, there is usually a GROSS mistake, in loads, design assumptions, vibrations/movements/harmonics or something extreme, and usually not the one or two bolts (of 6 or 8 or 20….) are slightly less tight than the others.

    Galliou » It is hard to say if these calculations would be very accurate for anchor bolts. The biggest uncontrolled variable when determining torque is friction. Since the surface between the bolt assembly and the joint surface in the field can vary greatly it is difficult, if not impossible to accurately estimate torque for anchor bolts. One option would be to consider load indicating washers. This washer will give consistent tension values regardless of environment.

    Bryan Carr » The formula from Industrial Fastener Institute (IFI) “Design of Bolted Connections” (M-64), used to determine values for our toque chart states it is meant for, “… steel bolts in their as-received condition.” Although, IFI goes on to state there many factors (surface texture, material hardness, thread series) that only have a “modest effect” on the torque-tension relationship. The primary influence on torque calculations is coefficient of friction. Again, the only “correct” way to establish an accurate torque value is to test the fastener in their actual joint application.

Leave a Reply

Your email address will not be published. Required fields are marked *