Torque Wrench 101...

[George M]

I asked a question about using a torque wrench with an extension below. I found a thread on the web that answers this question and others...as in using a torque wrench with a universal for example.
Have a look if interested...a pretty good article:
http://home.jtan.com/~joe/KIAT/kiat_3.htm

George

[Hector]

for the hell of it. Here's a formula that relates the original torque to the new torque when an extension is used:

NT = OT/(1 + (b/a)cos(180 - theta)) with constraint a > b.

OT is the original torque, a is the torque wrench length from center-to-center, b is the extension length from center-to-center and NT is the new torque.

Using the diagram of torque wrench from the web site, when theta = 90, NT = OT. So their claim is true regardless of extension length.

For the case when theta = 180 degrees, i.e., when the extension is in line with the length of torque wrench then,

NT = OT/(1 + b/a)

As an example, when I do valve adjustment, the excentric locknut need be torqued 7 lb ft. So using an open ended 10 mm wrench of 0.35 ft long and torque wrench length of 1.125 ft, the torque wrench needs to read 5.32 lb ft to get a torque of 7 lb ft at the locknut.

So when the 90 degree angle method becomes awkward for those hard to reach places, then you can use the simple expression above to figure what torque is when the ext is in line with torque wrench.

Just trying to help a bit more.

[George M]

pretty much a given as to how to modify torque wrench readings when adding to the length of the torque wrench. This wasn't the spirit of my post Hector. Is was when using an extension from the socket to the wrench...not adding an extension on the torque wrench handle. The study concludes that there is no change in torque setting on the wrench when using a socket extension. Things change however when using a universal because the torque imparted in no longer orthogonal to the socket/bolt head and efficiencies are lost due to the internal friction of a cardin joint universal which is not a constant velocity joint.
George

[Hector]

I meant adding an extension to the 3/8" drive end of the torque wrench. I'm aware of internal frictions and I agree with you... The equation was meant to apply to 2 cases: (1) proving the web site's right-angle point of use and adding an in-line ext to the 3/8" drive end of the wrench .

[George M]

excuse me Hector, my mistake...you were proving that torque remains unchanged by introducing an extension between the wrench and the socket. I should have looked a bit closer at the math....derivation at a glance looked like changing moment arm length and thought you were mixing apples and oranges.
Your last statement though digging a bit deeper is circumspect...
"As an example, when I do valve adjustment, the eccentric locknut need be torqued 7 lb ft. So using an open ended 10 mm wrench of 0.35 ft long and torque wrench length of 1.125 ft, the torque wrench needs to read 5.32 lb ft to get a torque of 7 lb ft at the locknut." You may want to add some clarity as what you wrote does not make sense...there should be no calibration change to the torque wrench as its moment arm length is part of its calibration...set value shouldn't be changed. Also because of torque wrench calibration accuracy, I personally would never use a torque wrench on an eccentric. Instead I rely on the feel of my hand with a 10mm box end wrench to tighten the eccentrics. They strip quite easily I might add.
George

[Hector]

I'm gonna have to disagree about the calibration part. The calibration is as accurate as the moment arm measurements. George, you are telling me that if I added (using the valve adjustment example) a 1 meter extension to the torque wrench that I would still have to apply enough force to read 7 lb ft on the torque wrench? I don't think so. You would have more leverage with a longer extension (moment arm) so the longer the arm, the smaller the applied force which translates to lower beam bending of the torque indicator. Your torque wrench would need to read a lower value. Remember that the moment arm of the torque wrench when bent and the moment arm of the extension converge at their point of intersection.

As to your personal preference for tighening the excentric locknut, if it works by feel, then fine. I on the other hand don't trust myself for feel so I use the torque wrench/ext technique. I've stripped way too many nuts/bolts. I've been using this technique for 6 yrs and still haven't stripped an eccentric.

[George M]

1. You don't need an extension when tightening an eccentric with a torque wrench.
2. You would never use a meter long extension for an eccentric.
3. If you do not add any length to a torque wrench handle, the calibration does not change. If the spec calls for 7 ft-lbs, that is what you set your torque wrench to.
4. You method of course doesn't strip any eccentrics because your erroneous torque conversion is biasing torque lower than the spec not higher inducing stripping.
5. Independent of adding length to a torque wrench handle or using an extension between the wrench and the socket, all torque wrenches are subject to calibration error. The smaller the torque setting, the higher the net measurement error by percentage.

George

[Hector]

1. disagree to a point
2. just an example
3. by not increasing length; I agree
4. disagree
5. agree

Let's just leave this as is.


1. You don't need an extension when tightening an eccentric with a torque wrench.
2. You would never use a meter long extension for an eccentric.
3. If you do not add any length to a torque wrench handle, the calibration does not change. If the spec calls for 7 ft-lbs, that is what you set your torque wrench to.
4. You method of course doesn't strip any eccentrics because your erroneous torque conversion is biasing torque lower than the spec not higher inducing stripping.
5. Independent of adding length to a torque wrench handle or using an extension between the wrench and the socket, all torque wrenches are subject to calibration error. The smaller the torque setting, the higher the net measurement error by percentage.

George

[George M]

Ok