Date:         Tue, 9 May 95 21:18:10 EDT
Sender:       Vanagon Mailing List <vanagon@vanagon.com>
From:         ja@decws3.coe.wvu.edu (John Anderson)
Subject:      Re: Don't Use: Click type torque wrench?

> 
> For some reason, throughout my repair microfiche, my manual urges again and
> again something like: 
> 
> Do not use inaccurate click type torque wrench
> 
> Every time I read that I go "Uh!" and then I wonder how inaccurate these
> wrenches are likely to be.
> 
> ___________________________________
> Derek Drew
> drew@interport.net (main address for e-mail)
> derekdrew@aol.com (alternate/backup address, checked infrequently) 
> 
> 
Oddly I got some experience here as well, when calibrating a torque
load cell for an experimental stress analysis class I was TA'ing, I
calibrated using a torque arm and dead weight to a level of 300 ft-lbs.
The cell had a 1/2" drill chuck mounted on it to grip specimens for
torsion testing and I had a brand new Sears click style wrench I
had been using in the lab (one of the new ones with plastic windows
and red and green numbers in a black handle).  For the hell of it I
put the torque wrench in the chuck through a 1/2" to 3/8" adapter so
it would fit the chuck set it at some different levels, set up the data
aquisition at 10 Hz and recorded nice continuous curves of the torque
reading versus time.  The results were that if one let off the wrench
immediately, and I mean immediately when it clicked, you usually got
a peak reading within 20 percent of the set value, a fairly even spread
statisticly around the set amount, push a little longer than the click
though and you could really go over.  The whole thing was pretty erratic
and random though and sometimes could be way under as well.  And this
is a $80 Sears wrench not a cheapo $20 Taiwan job like I usually use
at home so it concerns me a bit, however I'm no good with those older
style wrenches that measure twist angle of the end and convert it to
torque, so I'm stickin with the clicker, my best advice is, don't
overdo it though, stop immediately when you get the click.  I note
the experimental setup was further verified by torsion tests of aluminum
that yielded perfect shear moduli, there could have been some lag
in the adapter I had to use to put the wrench in the chuck though.
20 percent is fairly major error BTW obviously, probably the reason
pros measure the elongation (stretch) of components like rod bolts
when building an engine.

John Anderson
ja@coe.wvu.edu
'71 Westy, '90 Corrado G60