Calibrating cheap torque wrenches for cheap

[Zung]

After reading all the horror stories here about over-torqued bolt, I decided it’s time to invest, so I got 2 torque wrenches off eBay, for $20 each: a 2-20 Nm (20-200 in-lbs) for the swivel bolt, and a 10-100 Nm (5-80 ft-lbs) for the yoke.

Now how trustworthy are they for 20 bucks? At least one member reported he beheaded a MK5 with a brand new torque wrench…

I decided they need a calibration, at least a check before use, and the Internet provides: no need for any fancy and expensive machinery, just clamp the square drive, and hang some known weight off the handle until it clicks, and read the torque.

If L is the distance between the center of the square drive and where you hang the weight,
W is the weight of the weight, and
T is the resulting torque, then

T = L x W

In whatever unit you please, as long as it’s consistent.

Procedure: calculate the torque, set the wrench where it should be, put the weight on; if it clicks, try a bit more, if it doesn’t, back out a bit. Ease the weight up every time you adjust.

This gentleman improved upon the above with an intriguing idea I call “self-torque”: the weight of the wrench itself contributes to the torque. To derive this value, he proposed to weight the wrench and find its center of gravity by balancing it on your finger, and the “self-torque” is the weight of the wrench times the distance from the square drive to the center of gravity.

So now if L2 is the distance to the center of gravity and W2 the weight of the wrench:

T = (L x W) + (L2 x W2)

Now it’s time to do some preliminary calibrations; I took my wrenches to 2 different supermarkets and weighted them. Here’s what I got:

Small wrench:
Supermarket 1: 0.608 Kg
Supermarket 2: 0.608 Kg
My $15 digital kitchen scale: 0.610 Kg (+0.33%)

The center of gravity is approx. 0.143 m, so the “self-torque” is 0.85 Nm, far from negligible at the lower end of the range. But the accuracy of the measurement of the cg is negligible.

On to the calibration, using combinations of bottles of water carefully weighted with my now calibrated scale. Results are given in in-lbs:

Read: 31.8, actual: 35.9, error: +11.5%
Read: 49, actual: 45.6, error: -7.5%
Read: 78.8, actual: 73.9, error: -6.6%
Read: 93.5, actual: 90.9, error: -2.8%

Not perfect, but what do you expect for 20 bucks? Most importantly, the 1st value above will let me deal with the brass bolt, and the 3rd the stainless steel bolt, and that’s what I bought the wrench for.

I’m still trying to figure out the weights to deal with the big wrench. I’ll report back when I’m there.

 

[redacted]

You could set up a small pulley and take a reading pulling upwards with the weight hanging off the pulley. Average the up and down reading and don't be concerned with the force of the weight of the tool.

 

[Zung]

Cool idea!
I'll talk to the wife see what she thinks of this setup in **HER** dining room

 

[NAM001]

This gentleman improved upon the above with an intriguing idea I call “self-torque”: the weight of the wrench itself contributes to the torque. To derive this value, he proposed to weight the wrench and find its center of gravity by balancing it on your finger, and the “self-torque” is the weight of the wrench times the distance from the square drive to the center of gravity.

So now if L2 is the distance to the center of gravity and W2 the weight of the wrench:

T = (L x W) + (L2 x W2)


why would you do this? dont you think that this weight thing has been accounted for by the manufacturer. eliminate the weight and pull horizontal on a virtical axis and not the other way. now the weight has no effect on the result. retest and try your results. use a spring gage to pull on the handle. better yet a line through a pully with a bucket on the end add weight (sand) till the ratchet brakes, and do your calc. if you need to much weight use a cheater bar to increase the wrench handle length and you will use less weight.

 

[Tortuga68]

The weight compensation seems flawed to me but I can't be bothered working it out

You can also use an accurate bathroom/kitchen scale for the calibration

In fact you can just use a scale and a regular wrench/socket & bar and not bother with a torque wrench at all

Personally I hand torque everything but that's just me

 

[Zung]

dont you think that this weight thing has been accounted for by the manufacturer

Yes & no: if the wrench is operated in a horizontal plane, the influence of the gravity is nil. However, to do the weight trick, we have to use gravity and operate in the vertical plane; in this case, the influence of the weight of the wrench CAN be a factor, relatively: the "self-torque" above is about 9 in-lbs, not a lot per se, and 25% of the target of 35 in-lbs, which is not too bad, whever or not you use it.

 

[Tim.]

You are right to consider the weight of the wrench in the calibration procedure if you have the bar horizontal but after that it doesn't matter which attitude you use it in. The torque contributed by the handle is included in that seen by the break out device. Using it pushing up, down or horizontal simply affects the amount of force the user has to apply to achieve a given torque.

I would question the CG = L/2 approach. Much of the mass tends to be concentrated at the business end. Simply balancing it on your finger will find the real CG.

Ah, EDIT. You said that about the handle CG, the rest of my post is right. - Makes note to read OPs before replying

 

[Tortuga68]

Hi Tim

to do the weight trick, we have to use gravity and operate in the vertical plane

 

[redacted]

Yes, working in the horizontal plane solves the problem.

BTW, I usually "calibrate" my torque wrenches by comparing the performance of one to another (believed to be good). It may not be very precise, but neither are most of our specs in scuba.

 

[DA Aquamaster]

If you are using a pully, you can use the pully to convert the force of the gravity on the weight 90 degrees to the wrench in the horizontal plane and avoid all that weight compensation BS.

Pullys do not need to bend the rope a whole 180 degrees.

Of course the whole weight idea is only as accurate as the known weight and the accuracy of the lever arm measurement.

Personally, I think you really need more to do. Just use the wrench as is out of the box and torque it to the middle of the specified range and call it good. It's not like it is a component for the space shuttle or anything.