How to establish margin for Torque test 1

[engtiuser2]

I am setting up a torque test to verify an insert nut/joint strength for a steel structure. I am wondering if there is any specification/guideline for how much over torque should be used to take into account for the uncertainties (surface condition, lubrication used, workmenship...etc). As an example, 1/2"-20 bolt calls for 46 ft-lb, I normally would arbitrary add 10% (i.e. 50.6 ft-lb) to account for the uncertainty for the torque test. Any comment on the validity of this?

 

[BridgeSmith]

When greater accuracy is required in the tension on a bolt than the rough approximation provided by specifying torque, use tension indicating washers, turn of the nut, or another more accurate method.

 

[drawoh]

engtiuser2,

What does your test determine? What measurements are you making? I am punching in materials here to get your 46lb.ft, and I think you have a soft, ductile bolt. What are the consequences of exceeding your calculated torque by 10%? My guess is that you may or may not get some plastic deformation.

--
JHG

 

[engtiuser2]

The objective is to confirm the nut can withstand the applied torque in assembly. We want to make sure the nut will not be damaged. If the nut failed before reaching the targeted torque, that means, the nut/joint could be failed during assembly.

 

[Agent666]

Test it to failure, what you actually really want to know is how far away from failure your nut is, or that it's not the limiting factor. Otherwise you just have a tightened bolt with no idea how close to failure it might be which tells you absolutely nothing in my honest opinion. 99% sure your bolt will fail before the nut if you tighten it until something fails, but you don't note type/material/etc to add any further.

I believe AISC tightening procedures using torque are required to be calibrated each day using 1.05 times the target minimum tension and a minimum number of bolts, this allows for changing conditions like some of the items you listed in the initial post. So maybe read up on that procedure and see if it suits your need to validate a torque tightening methodology. Key is you must calibrate the torque required to tension the bolts (unsure if you want tensioned bolts or not?).

 

[MintJulep]

If the specified proof stress of the nut is equal to or higher than the specified minimum tensile strength of the bolt then **almost always** the bolt will fail before the nut threads strip.

There are some fringe cases if you are unlucky enough to pair a "weakest within tolerance" nut with a "strongest within tolerance" bolt.

If you are looking for assurance against such fringe cases then you should not be using torque controlled tightening in the first place.

So my comment on the validity of your test is that is is not valid.

 

[engtiuser2]

The nut is a insert nut type. I do not anticipate the thread or the bolt will fail. I am more worry about the insert (swage) side where it detached from the base parent metal.

 

[drawoh]

engtiuser2,

You have manufacturer specifications for the nut, which should include its pull-through strength. Probably, you need a high strength screw to test and verify what the manufacturer is telling you.

--
JHG

 

[SnTMan]

I suspect the OP is interested in torque-out of the insert. I further suspect the OP is looking for a "proof-test" or process control test of production-installed inserts.

Some inserts do not relay entirely on the crimp to resist torque, but install into keyed or non-circular holes.

The foregoing is guess-work on my part.

Regards,

Mike

The problem with sloppy work is that the supply FAR EXCEEDS the demand