As the construction industry is outside my scope of knowledge I seek advice for a colleague. When a building or warehouse is erected these days there seems to be a steel skeleton frame work erected first. My question is : 1) Where the steel sections are bolted together, is there a set known torque to which the bolts are tightened? 2) Is there a British Standard for the erection of such frame work available. As ever, any and all assistance or guidance in this matter is gratefully appreciated. Badger

achrn So, effectively then there is NO set torque value only reliance on squashing a washer of sorts. whilst waiting for a reply I came across a piece on SteelConstruction.info under 'Torque Method', that implied that wrenches had to be calibrated weekly surely that implies a known torque value? Badger

Thanks Jane So would you be able to give an example of a torque value either as an average or from a job you've been on? As an construction outsider, from a pure engineering background I'm surprised there is no standard for bolting a steel framework together and not at all surprised at the earlier comment from achrn saying ...but I've come across an installation where the bolts weren't properly tight. It's a wonder many structures haven't fallen down! Badger

Eh? Where did that statement come from? There absolutely is a standard - I gave you the reference and name of the standard - BS EN 1090:2 "Execution of steel structures and aluminium structures - Part 2: Technical requirements for steel structures". The parameter that matters in a friction-grip connection is the bolt tension, which is directly indicated by load-indicating washers, but is a bit hit-and-miss if you rely solely on torque because it's affected by the friction in the threads and the friction under the nut / bolt-head. Simply relying on nut torque is potentially less reliable than something that directly measures the tension in the bolt shank. The tension on the bolt shank is a much more precise definition of the parameter that matters than specifying a tightening torque. The method I described is the most reliable method out of the several methods permitted in the code. If you want to work to a torque you can (it is another one of the methods permitted in the code), but it's a less reliable method because it's only an indirect indication of the bolt tension. Notwithstanding which, BS EN 1090, in talking about calibration of torque wrenches, says: "Torque wrenches used in all steps of the torque method shall be capable of an accuracy of ± 4 % according to EN ISO 6789. Each wrench shall be maintained in accordance with EN ISO 6789, and in case of pneumatic wrenches checked every time the hose length is changed. For torque wrenches used in the first step of the combined method these requirements are modified to ± 10 % for the accuracy and yearly for the periodicity. "Checking shall be carried out after any incident occurring during use (significant impact, fall, overloading etc.)and affecting the wrench." I don't know if EN ISO 6789 specifies a calibration interval, however - this stuff would normally be self-certified by the contractor, and I'm a designer not contractor, so I'd just have a requirement on the drawings that they erect in accordance with 1090.

ACHRN I believe there is a breakdown in communication in that I accept you have given "Torque wrenches used in all steps of the torque method shall be capable of an accuracy of ± 4 % according to EN ISO 6789". What I'm getting at is, + / - 4% of what? 2 N.m.; 75 N.m.; 150 N.m. or 1500 N.m. torque force applied to tighten a bolt? Also what I should have typed but only thought .. (my fingers cannot keep up with my thoughts) I'm surprised there is no standard torque figure for bolting a steel framework together. To me as a construction outsider, from a pure engineering background it looks like meccano to me and a simple bolt up. But with my engineering head on I wish to know what value torque (number in Newton metres) is applied? Squashed washers and a % tolerance of fresh air tells me nothing! No intent to feather ruffle, but at the moment its something plucked from nowhere to me. Badger

The problem is that a given torque on the spanner does not reliably translate into a given strain in the bolt, because of friction in the threads. And the strain that you want, which will determine the clamping force, will also depend on the friction that exists in the surfaces clamped together. Either way, the direct measurement of strain, via compressible washers etc is a more consistent way of checking that the parts have been assembled correctly.

It's more complicated than that. The specified torque is a function of the reliability class declared by the manufacturer and the particular type of nut and bolt combination, and is still horribly variable so you'd probably do a test in a load-cell, and even after all that you'd expect at least 30% variability.

As you have repeatedly been told, the safety critical aspect is the bolt shank tension. That is what is controlled. That is not inappropriately inconsistent. For some reason you persist in averring that torque should be controlled. Nut torque is a poor indicator of the parameter that matters. If the joints were controlled in the manner you seem to want there would be inconsistency and poor predictability of joint performance. However, when done in accordance with the standards that are in place, this is not the case.

With reference to my earlier comment on building an engine (don't forget I'm non construction) that is how we would control a build and have documented evidence that engine number 1234567 was locked up to known values. So for a different field of work such practices seem a little strange to me. BUT I am learning and apart from passing the information on to a colleague of mine who raised the original question, I think I'll write everyone's comments up for my cpd as I've learnt something over the last couple of days. Thanks again. Badger