Tightening Torque of High-Strength Bolts

The tightening torque required will depend on size of bolt and material. The torque should be sufficient to deform the bolt.

Heavy Hex Head Bolt size is M20 and material is AISI 4140

I presume that the material was heat treated so only giving the type is not enough you should give the mechanical properties as elastic limit and ultimate stress for a correct answer.

More it is important to give two other informations:

- what material and surface are present where the bolt head or the nut will work on

- how are the bolt conditioned, is any anti rust protection used?

A last question is the bolt all over of same OD or does it have a thinner part between thread and head?

It could be also good to know the length of the bolt shaft.

The torque value cannot be used as it is in the tables and for different reasons so that better give answers ans you get a good answer too.

http://www.portlandbolt.com/technicalinformation/bolt-torque-chart.html

Not enough information to answer your question!!!

An awful lot depends on the surface condition of the internal and external threads, lubrication or lack of it, condition of the surface under the head or nut, etc. Torque is a poor measure of the tension, because a large percentage of the torque does NOT go into tensioning the fastener. You also have to know whether it will be loaded in tension or shear. Consider using load indicating washers.

And there is probably a good bit of voodoo involved too!

Indeed, there is quite a bit of voodoo involved, even when following some of the guidelines in the spec linked by PMoon.

Most of the procedures in the RCSC specification unfortunately require assumptions to be made rather than prescribing the use of actual data.

Load indicating washers being one of the better options to eliminate the nebulous variables from the torque/tension relationship also have issues: Whereas they usually provide an acceptable indication of sufficient clamp load, they can't show whether too much load has been applied. A joint which is too tight can be just as dangerous as one which is too loose.

I still use a theory passed down to me from an old fitter many years ago...

"Tighten it till it strips, then back it off half a turn!" :)

It still makes me smile to this day.

Tony

In pretensioning bolts the only correct way is with the proper use a Skidmore to measure the tension on the bolt. The tension requirements for the specific bolt diameter and grade can be found in the AISC "Steel Construction Manual"-Thirteenth Edition, or 2004 RCSC Specification for Structural Joints Using ASTM A325 or A490 Bolts, these tensions are also marked on the face of the Skidmore device. Usually TC-bolts or LIWs (load indicating washers) can be used, or the old "Turn-of-the-nut" Method. The turn-of-the-nut method again requires the use of the skidmore device. A skidmore device measures the tension on the bolt, not the torque. Basically you use it to test a number of bolts from each lot to determine the required torque or turn-of-the-nut to achieve the tension required, you then apply that torque or "Turn" to each bolt installed. To do this correctly you test bolts daily prior to installing bolts in your structure. There is no formula to accurately calculate torque to tension relationships, because you cannot account for all the variables involved. These variables include moisture, temperature, surface condition of the bolts and washers, heat treatment, galvanizing, and many others.

With all due respect, the use of a Skidmore-Wilhelm still forces one to predict, guess and hope what the final pre-load of any individual fastener will be it's been torqued. Sure, the device will establish a correlation of torque to load for the specific fastener within the rig at the time. However, when applied in the field and extrapolated for all similar fasteners, many other factors including sprung joints, uncontrolled lubrication, damaged threads, damaged spotfaces ad infinitum, can throw and will throw this correlation out the window .

The only way to properly determine residual preload is by not predicting, guessing or hoping but rather by measuring after load has been applied. Measuring before really isn't in keeping with good (or even logical) engineering practice because this clearly disregards a number of critical factors which can impact the safety of the joint.

I agree with Guest. While there are assumptions made in the RCSC spec. it is based on years of constuction work and practices that have had satisfactory results. Perform the required tests and the proper QA/QC work as the bolting is done. Codes and Standards are written with a certain amount of tolerance in them to allow for the POD (Probability of Detection) factor.

Good point, Inspector. However, if the joint design is marginal it may only take one wrong guess or assumption to cause a significant problem leading to risk of safety, loss of production or, loss of property.

I suppose it's all about Risk Analysis and an indidual's level of risk tolerance.