Mechanics: what are your most common tools?

That's not really and accurate way to "calibrate" or check a torque wrench.
Do you know for a fact that weight is exactly the amount it says it is?
Did you count the weight of the rope? Is it in the exact spot?
Torque wrenches are different length, I don't believe they are calibrated on their length.
A regular wrench you could do that with.

 

You can hang the load at a measured point on the handle, and set the torque wrench accordingly. For a 30 pound weight hung 15" along the handle, the optimum setting on the torque wrench would be 30x15/12 pounds, or 37.5.

Also, good to just set it a little high first, verify it doesn't click, then work down. Or alternately: set the weight a little more up the handle, then work up to that 15" point, see when it clicks.

I know, you'll be back to explain different, I might not hear it lol.

 

Hearing and listening.

I suffer from that as well.

 

Maybe the better way of posing these questions is:

What does the rope weigh, as a percentage of the total weight you are hanging on the wrench? How did you measure the weight, and within what percent plus or minus? How did you measure the distance from the drive center to where the rope hangs, and within what percent plus or minus?

That's exactly how the scientific measurement game is played. There will always be some statement of an error margin, and one doesn't say "that's bogus because there's an error margin", one says "is the error margin narrow enough for the work I need to do or the regulations I need to meet?"

A professional calibration certificate will also state an allowable % error (often narrower for clockwise than counterclockwise, and different for different fractions of the wrench's full scale).

The only method that would be really bogus would be one claiming to be exact.

 

Too, how much accuracy is needed, considering the other vagaries, like condition of bolt threads for example. Yeah, I'd rather consider methods to see for myself if the torque wrench is "in the ball park", and revisit every so often.

 

Nice job. I did a similar testing of mine last year. I used two scales to measure the actual weight of the dumbell and the rope, and took the average. My wrench was just over 12", so I was able to put it exactly at 1'.

 

Your method is just fine, Mendel, with the understanding of accuracy limitations as discussed by ChapmanF.

And how accurate does a torque wrench need to be? 10% should be fine, I expect, and should be easily achieved with a home brewed calibration setup such as that. Even better if you do some things to improve accuracy. A torque spec is basically aimed at achieving a clamping spec (maybe a minimum, in some cases a target range) without breaking the fastener. There is enough uncertainty due to surface conditions of the fastener and dynamics of the torquing process, that safety factor is designed in, and 10% will be fine. Beyond accuracy, repeatability is important as well when you have multiple fasteners clamping a single object, where uniformity is important.

These torque wrenches measure torque at the head, so you are right that it is the attachment point to the wrench that matters in the calculation. And more specifically, it is the length of the moment arm of the gravity force - so for example, if your wrench were not perfectly horizontal, there would a cosine term in there to fix that.

When getting to higher levels of accuracy, the weight of the wrench itself imparts some torque in your experimental setup there.

An alternative that removes the weight of the wrench from the measurement would be to mount the wrench so it rotates horizontally, pulls the string horizontally, and then use a pulley to send that string downwards to your bucket or whatever. If doing that, you could then also put a pipe/broomstick/whatever to extend the wrench length. This would allow you to measure much higher torques without going to much higher weights - just using a longer arm. Vertically supporting this extended arm, as long as it does not affect torque as measured at the wrench head, would be fine.

An idea you might try, if you want to measure the click-point very carefully, is to put a weight on there that is below click torque. And then pour water into a container in the bucket until it clicks, and measure the amount of water when it clicked. Repeatability tests would also be interesting if measuring with this fine resolution.

 

No, my way of "posing" the questions is correct.
Everyone is just making TOO big of a deal with torque wrenches.

 

The ANSI standard (seen back in #175) calls for 4% or better, within the 20% to 100% of range region. Even that doesn't sound unachievable for a home rig with a bit of care.

Kinda sounded like #201 was saying "that's not an accurate way to calibrate/check a torque wrench" because of questions like weight of the rope or position on the handle. I looked again just now and it still kinda sounds like that.

I didn't say that way of posing the questions was 'incorrect', only that it could be improved, which it can. What matters is: what's the max difference those questions could make in the result?

If the bag and rope are half a pound, as Mendel estimates, next to 30 pounds of dumbbells, that's a max error of 1.7% even if Mendel completely forgot to add it in to the total weight. Position on the handle, if measuring 15 inches, ±1% gives you a little better than an eighth of an inch to play in, achievable with tools around the house as long as you're not completely slapdash about it. A handle extension as 2012 suggested would also make the positioning less critical; extend the handle to 30 inches and your ±1% is now around a quarter inch rather than an eighth.

 

Just found this puppy too, could hook it up directly, see what numbers it clicks at.


That worked a charm: hang it off the handle at a measured distance, (calculate the factor if that distance is other than 12"). Then dial in a number on the wrench (with the factor), turn on the scale and it zeros itself, then pull and watch the display till it clicks. Pulling straight down doesn't account for the arm weight, but I think that's pretty trivial. Pulling straight down is most controllable.

Verdict, all three were very close, at low to medium test values.

 

... and as a bonus you used your torque wrenches to calibrate that luggage scale.

 

I disagree.
I’ve twice stripped threads when using torque wrenches which didn’t click at the right point.

Of course, perfect accuracy isn’t necessary, but the damn thing has to work.

 

Opinions vary....

 

Off-topic, a make-shift kitchen tool. My wife likes getting these blocks of dark cooking chocolate, and we just whittle away at them, for a chocolate snack. Anyway, the usual implement, quite effective at calving-off chunks without a lot of slivers, is an old fondue fork, and found beach stone that acts as a hammer.

Who says we've left the stone age...

 

My Snap-on ratchets have 80 teeth, but I guess maybe I’ve got to the point in life that I’m the “other half”. (Not really I just have nice tools still drive a bunch of hooptys)

 

It's funny, I wrote that question originally without even checking what the typical tooth counts in real ratchet drivers are, as it was intended more as a think-about-it question. But I guess I wasn't out of the ballpark for the "low-tooth" category, according to this article:

So, we can also offer a high-tooth version of the question:

You are packing for a maintenance job on a space station. You get to bring a collection of six-point, square-drive sockets and your choice of high-tooth ratchet. It can have any number of teeth you choose, from 72 to 160. There are bolts on all kinds of cramped-access spacey things, and you can't call for backup if you find one you can't get your ratchet and socket onto. What number of teeth do you choose?

Or, a related question: there are 89 choices there (72 teeth, up to 160 teeth, inclusive). How many of the 89 choices are better than the 160-tooth? ... in the sense of giving you the most possible rest fits on some cramped-access nut or bolt.

 

You're very lucky, most people don't get more teeth as they get older!

 

I would use the one with the highest reliability, with lubricant that won’t boil off in space, which has a quick release I can operate with my space glove, and a handle I can hold with my space glove...

IOW, high tooth count is nice but not the only factor.

I saw a documentary about how they developed and tested tools for the Hubble telescope repair. It’s very difficult to perform seemingly simple tasks such as removing bolts, not loosing bolts, threading bolts, and tightening them. When you drop a bolt on Earth, at least it goes down! They had to develop special fixtures to capture the bolts and align them for re-installation.

 

Not digital accuracy, is it?

 

Recently swapped a Honda drivetrain: 14mm & 17mm on an electric impact wrench and a BFH were the most common used.

Outside the box: a 31/32's inch shower valve tool removed the upper O2 sensor perfectly!