[John in YVR]

I have a pretty good torque wrench but would like to calibrate it or at last verify its accuracy.

Anyone here have any suggestions if this is possible, and if so how?

(I thought of using two to torque against each other, but what if they are BOTH wrong or is one if out how would you know which one?)

 

[NovellRed]

Yes torque wrenches can be calibrated. When I was in the Navy part of my job was to calibrate instruments including torque wrenches. It can be done but I haven't seen the tools outside of the military. They also need calibrated against a standard every so often, kept at the proper temperature, flexed before testing and so on.
Some torque values are being replaced by stretch values. Instead of torquing to a specified value in foot pounds you tighten until the bolt stretched to a particular value.
You may be able to use a micrometer to measure a bolt then using your torque wrench to tighten to a specified # of foot pounds then check to see how much the bolt stretched. That might be the easier way to calibrate your torque wrench.
The stretch values we used had to come from somewhere, there must be a table of bolt/stretch/torque values about.

 

[Tater]

Most torque wrenches are calibrated in lb/ft. That being said, calibration is set as if the handle was 1 foot long.

Lock the square drive firmly in a vice with the handle horizontal. Find something of known weight and attach a loop of heavy string to it. Set the wrench to that weight. Place the loop of string over the handle near the head of the wrench. Slowly slide the loop away from the wrench head until "break away". This should occur 1 foot from the center of the square drive but likely sooner due to the weight of the handle. Not very accurate but should be OK for home use. For lb/in wrenches, simply multiply the wrench setting by 12.

This is the procedure as told to me. I have never tried it myself.

 

[Rastoff]

I'll try to keep this short...

Calibrating your torque wrench for home use is not worth the money; if you are talking about the click type. Over time the spring will compress and the linearity of the wrench will be compromised. A new wrench can be had for about $25 and it will cost you at least that to get one calibrated. Replacing your torque wrench every year or two is good cheap insurance.

I'm sorry if this agitates anyone, but, when it comes to click type torque wrenches, they're all garbage. It doesn't matter if it's a Snap-On or made in Taiwan, they're all mediocre at best. I've worked in a calibration lab for 22 years and seen it over and over. Don't waste your money on an expensive click type torque wrench.

If you just have to have a good wrench get one of these:

They only cost about $325, but they will hold their accuracy much better than a click type.

Tater,
Your method will not work because the wrenches are designed to work with force applied to the handle only. Force applied to any other part of the shaft will yeild false readings. Also, as the wrench handle moves while the head stays fixed you will encounter cosign error and that will also throw off your measurement. Further the speed at which you apply the weight has to be uniform. In the end it's just not worth the hasssle.

NovellRed,
The idea of measureing how much a bolt stretches is truly innovative, but it won't work either. The bolt would have to stretch and recover the same amount and that just won't happen. Also, you would have to have a very stable material to get accurate readings. Then you would need a micrometer with at least .0001" resolution, but .00005" would be better. They are easy to come by, but not easy to use and it takes a lot of practice to get repeatable readings out to that last digit. It took me a while to learn under laboratory conditions.

 

[Dream Catcher]

I'll try to keep this short...

Calibrating your torque wrench for home use is not worth the money; if you are talking about the click type. Over time the spring will compress and the linearity of the wrench will be compromised. A new wrench can be had for about $25 and it will cost you at least that to get one calibrated. Replacing your torque wrench every year or two is good cheap insurance.

I'm sorry if this agitates anyone, but, when it comes to click type torque wrenches, they're all garbage. It doesn't matter if it's a Snap-On or made in Taiwan, they're all mediocre at best. I've worked in a calibration lab for 22 years and seen it over and over. Don't waste your money on an expensive click type torque wrench.

If you just have to have a good wrench get one of these:

They only cost about $325, but they will hold their accuracy much better than a click type.

Tater,
Your method will not work because the wrenches are designed to work with force applied to the handle only. Force applied to any other part of the shaft will yeild false readings. Also, as the wrench handle moves while the head stays fixed you will encounter cosign error and that will also throw off your measurement. Further the speed at which you apply the weight has to be uniform. In the end it's just not worth the hasssle.

NovellRed,
The idea of measureing how much a bolt stretches is truly innovative, but it won't work either. The bolt would have to stretch and recover the same amount and that just won't happen. Also, you would have to have a very stable material to get accurate readings. Then you would need a micrometer with at least .0001" resolution, but .00005" would be better. They are easy to come by, but not easy to use and it takes a lot of practice to get repeatable readings out to that last digit. It took me a while to learn under laboratory conditions.

Pretty interesting stuff from a guy "in the field" Rastoff.
I, for one, thank you for the info. I was involved with some pretty sensitive instrumentation over the years and because of that I have always known there was no way the average torque wrench could be accurate in all the varying conditions. I don't figure the stuff I mess with requires that I be "dead nuts" on torque, so I just use the wrenches I've got and give it my best shot. I would like your opinion on what is a good "rule of thumb" for adjusting a torque spec. due to using anti-seize. An engineer friend of mine says he always reduces the torque setting a few pounds when tightening something he's put anti-seize compound on the threads of.
He says it offsets the reduced friction that would effect the actual "tightness" or "stretch" obtained at a given torque reading. Makes sense to me.
Do you have an input on that?
Thanks
DC

 

[Tater]

I'll try to keep this short...

Tater,
Your method will not work because the wrenches are designed to work with force applied to the handle only. Force applied to any other part of the shaft will yeild false readings. Also, as the wrench handle moves while the head stays fixed you will encounter cosign error and that will also throw off your measurement. Further the speed at which you apply the weight has to be uniform. In the end it's just not worth the hasssle.

Well, I did say this was secondary information. The guy that told me this was a PMEL tech in the late '70s. He then went officer. Now I understand! You are 100% correct on one point. I only paid $20 for my 1/2 drive wrench. It isn't worth calibrating or even checking.

 

[Silverback]

Rastoff- I appreciate your input on this subject. In general, once a torque wrench begins to lose it's accuracy, is it a safe assumption that it would always read either high or low? More specifically, would it be wise to torque a bolt to a slightly higher or lower reading if you have suspicions about the accuracy of the wrench itself?

 

[John in YVR]

Calibrating your torque wrench for home use is not worth the money; if you are talking about the click type. Over time the spring will compress and the linearity of the wrench will be compromised. .....

Thanks Rastoff - very informative post. (Note to Hal - this is what make your forum worth the effort.)

I have a 'click' type and always back off the spring after using plus I have a heated garage so cold temps are not a factor therefore I hope that it is still sort of accurate.

This winter I'll be installing new bearings and just wanted to get an opinioin on torque wrenches before doing so.

 

[Rastoff]

I would like your opinion on what is a good "rule of thumb" for adjusting a torque spec. due to using anti-seize. An engineer friend of mine says he always reduces the torque setting a few pounds when tightening something he's put anti-seize compound on the threads of.
He says it offsets the reduced friction that would effect the actual "tightness" or "stretch" obtained at a given torque reading. Makes sense to me.
Do you have an input on that?
Thanks
DC

When using anti-sieze all bets are off. That stuff never truly allows the bolt/nut to set. Your friend is right about reducing the torque setting. I don't know enough about different metals to give an informed opinion aobut weather or not you should use anti-sieze. If you do though, I would reduce the torque setting by at laest a third if not in half. Because of the reduced friction you run the
risk of stripping threads or breaking a bolt.

In general, once a torque wrench begins to lose it's accuracy, is it a safe assumption that it would always read either high or low? More specifically, would it be wise to torque a bolt to a slightly higher or lower reading if you have suspicions about the accuracy of the wrench itself?

Unfortunately there are no absolutes. If you suspect your wrench is no longer accurate I would just retire it and get a new one.

There are many reasons the manufacturer gives recommended torque specs. One is to ensure that the fastener will remain fastened while it's in service. Another is to ensure that a fastener isn't overtightened. On the Wing we have many bolts that thread into aluminum. Since aluminum is soft, it's easy to over tighten an strip out. I try to always follow the recommended torque on these fasteners.

The average click type wrench is spec'd to +/-4%. Most will pass at +/-10% if checked. The 10% spec is good enough for all the uses on the Wing.

This is the wrench I use:

It costs about $22. I've calibrated several of these and all but one have met their advertised specs when new. The one that was out was +/-6%. I get it at Harbor Freight.

 

[John in YVR]

This is the wrench I use:

That's the one I have as well.

I have one petty irritant in that when I twist the handle to a "0" it does not line up exactly with the horizontal mark for a given torque setting.

I have to use about +4 lbs to get it to show exactly ‘on the mark’. (e.g. I have to show 24 ft lbs to be on the 20 ft lbs mark.)

 

[Rastoff]

[quote="Dream Catcher":2t852ciz]This is the wrench I use:

That's the one I have as well.

I have one petty irritant in that when I twist the handle to a "0" it does not line up exactly with the horizontal mark for a given torque setting.

I have to use about +4 lbs to get it to show exactly ‘on the mark’. (e.g. I have to show 24 ft lbs to be on the 20 ft lbs mark.)

[/quote:2t852ciz]
Use the "0" on the handle rather than trying to match the line on the shaft.

 

[Clamper]

Rastoff, I'm glad to see you respond to this thread. I have looked for the old thread where you discussed the how and what you used to calibrate torque wrenches. If you have that site could you or someone else post it again on here or bring it to the top.

Thanks

 

[praline]

Get a 2x4, put a long bolt through the middle, making sure it doesn't turn in the wood. Rest the ends of the bolt on anything (axle supports?), and make sure the 2x4 is balanced. Then hang a known weight at a certain distance from the bolt - e.g. 20 lb. at 3 ft. gives you 60 ft/lb. Put your wrench on the bolt.

 

[Rastoff]

Rastoff, I'm glad to see you respond to this thread. I have looked for the old thread where you discussed the how and what you used to calibrate torque wrenches. If you have that site could you or someone else post it again on here or bring it to the top.

Thanks

This is that makes our machine. http://www.cditorque.com/main.html Of course they make newer machines now.
Is that what you're looking for?

 

[Clamper]

No, I was looking for the thread that was on here earlier this year that talked about Torque wrenches. You had check a couple of different wrenches for someone and did a report on it. I felt it was worth keeping, but that was before I had my hard drive crash.

 

[Jack D]

The idea of measureing how much a bolt stretches is truly innovative, but it won't work either.

WARNING Long boring strength of materials dialog follows that has nothing to do with calibrating torque wrenches.

Bolt stretch is a very accurate way of determing if a bolt is tightened to the correct amount. Unfortunately most bolts cannot be checked this way and a torque wrench is used to approximate the stress in the bolt. While I agree that measuring the bolt stretch is not an accurate way of determining torque one must remember that torque is not what the real issue is. A bolt is supposed to exert pressure on the pieces being held together that is greater than the forces trying to push them apart if we are talking a dynamic environment. If we are talking fender mounting bolts, etc. then we want to make sure the bolt does not come loose and is not overstressed. In a cylinder head held in place by 4 bolts we want to ensure that if there is 1000 lbs of force pushing the head up during combustion then each bolt must have at least 250 pounds of force holding the head down. This is determined by the modulus of elasticity of the bolt's material. If the bolt is stretched a given amount the modulus of elasticity will allow one to determine how much clamping force the bolt is providing. Of course the stretch of a head bolt cannot easily be determined so a torque wrench is used to determine the approximate amount of stretch the bolt is undergoing which is dependent upon thread pitch and of course the modulus of elasticity. In the case of connecting rod bolts it is possible to measure the bolt stretch and many engine specs, particularly diesels use this as their spec. The other consideration is too much stretch which takes the bolt out of it's elastic range and permanently stretches the bolt (or worse) which now does not allow it to exert the correct clamping force.

Sorry if this put you to sleep but there was a warning at the beginning.

 

[gpick]

Thanks for sharing this, Jack. All of a sudden as I read what you wrote I think it clicked -- the bolts stretch and are elastic within certain limits. Makes sense. Too little tension at either limit reduces the bolts effectiveness.

So if I follow you, the desired outcome is to always stretch the bolt to its optimum length which is determined by lab testing of the bolt's elasticity at certain torque settings. That way when the bolt is being used in a location where it would be impossible to measure, like a cylinder head bolt, the torque setting is used to approximate the proper stretch. But, the stretching is the objective and the torque measurement is the means.

I can now see why it is important to torque the bolts, because with the varying bolt sizes and especially different materials (like aluminum) feel must be pretty deceiving.

 

[Dream Catcher]

The idea of measureing how much a bolt stretches is truly innovative, but it won't work either.

WARNING Long boring strength of materials dialog follows that has nothing to do with calibrating torque wrenches.

Bolt stretch is a very accurate way of determing if a bolt is tightened to the correct amount. Unfortunately most bolts cannot be checked this way and a torque wrench is used to approximate the stress in the bolt. While I agree that measuring the bolt stretch is not an accurate way of determining torque one must remember that torque is not what the real issue is. A bolt is supposed to exert pressure on the pieces being held together that is greater than the forces trying to push them apart if we are talking a dynamic environment. If we are talking fender mounting bolts, etc. then we want to make sure the bolt does not come loose and is not overstressed. In a cylinder head held in place by 4 bolts we want to ensure that if there is 1000 lbs of force pushing the head up during combustion then each bolt must have at least 250 pounds of force holding the head down. This is determined by the modulus of elasticity of the bolt's material. If the bolt is stretched a given amount the modulus of elasticity will allow one to determine how much clamping force the bolt is providing. Of course the stretch of a head bolt cannot easily be determined so a torque wrench is used to determine the approximate amount of stretch the bolt is undergoing which is dependent upon thread pitch and of course the modulus of elasticity. In the case of connecting rod bolts it is possible to measure the bolt stretch and many engine specs, particularly diesels use this as their spec. The other consideration is too much stretch which takes the bolt out of it's elastic range and permanently stretches the bolt (or worse) which now does not allow it to exert the correct clamping force.

Sorry if this put you to sleep but there was a warning at the beginning.

DANG! After all these years, you finally cleared it up for me Jack. Now I know what that old fart in the factory wanted when he told me..."Hey kid, run down to the tool crib and tell em to give you a bolt stretcher. This bolt's too short." :lol: :lol:
DC

 

[NovellRed]

Jack D,
Thanks for your input, it broke some rust loose from the old grey matter. My experience comes from main propulsion diesel engines. GM 645s and 567s. Head bolts were torqued to 1,200 ft.# using a torque multiplier and a big, big ratchet handle. Connecting rod bearings were stretched to match a "snap gauge" supplied for the purpose. I think we also stretched bolts on the Fairbanks Morse ND 8 1/8. I retired over 16 years ago and the memory is a bit fuzzy. Now I do IEEE 802.11 wireless stuff and nothing is more than finger tight.

 

[Quincy]

I think you guys may have missed the point about bolt stretch and torq wrenches. Head bolts on modern engines are supposed to stretch and stay stretched. That is why you can not reuse them. The reason for torqing all bolts is so you can get them all at the same tightness and not deform or warp the item that they are fastening and by following the torq spec you can do this and get maximum tightness without breaking the bolt or stripping the threads. If one had the ability to tighten all bolts evenly and stop just before the bolt failed one would not need a wrench.


Q