Torque Experts - What's The Equation?

[goldgunner]

Perhaps this is well known - but not to me:

Assumptions: 3/8" standard high quality socket and torque wrench and extension dimensions and materials.

For each inch of extension - between torque wrench head and driven socket or bit - how much torque is 'lost' from the torque wrench set amount? It can't be the same I'm pretty sure: there has to be a cumulative twist effect losing torque 'over distance'.

This one is probably harder: When using a universal joint adjacent to a driven socket or bit, at 15 degrees offset how much torque is 'lost' from the torque wrench set amount?

Finally, a combination: From a driven socket or bit, with a universal joint attached driven at 15 degrees by a 6" extension, how much torque is 'lost' from the torque wrench set amount?

Esoteric? Can't be - real world stuff - doncha see...

Thanks,
Jim

[viejopatron]

From the centerline of the item being torqued, you'd measure laterally (away from) that centerline. Visualize 2 perpendicular lines. The lateral distance multiplied by the applied force (again perpendicular to the lateral line) is the applied work (work = force X distance).

The setup can include universals, extenstions, pipes, mother-in-law, whatever.

However, be aware of applied work losses due to inefficiencies (uni’s) and running torque (friction).

Piece of cake.

[patkeefe]

In theory, for a straight extension, you lose zero torque. For the universal joint questions, I would have to get out my physics books....do a web search for "double cardan joint" (that's what a u-joint is). There is an easy formula for it which escapes me at the moment.

Foat Wuth. That's great, Just how the natives spoke it when I was stationed down that way.

[mattr_914]

torque = radius crossed with the force

a cross product basically multiplies values in perpendicular axes, so when a force is applied in a parallel axes, there is no torque. So, like patkeefe said, theoretically there is no torque loss with an extension. The equation for a u joint is very approximate because of the losses in the u joint itself. My craftsman u joints have plastic, so there is an associated deformation of the plastic which doesnt help an accurate torque value.

My opinion: use a torque wrench on things that call for it (in the engine, for instance). Everything else should just be by feel. One of the techs I worked with wouldnt let me use a torque wrench on cup cars. He wanted me to get a feel for it and he would go back and check me with a torque wrench. You learn your hand position and force applied and with some practice you can get within a few N/m. Again, on the engine side of things I always use a torque wrench on the important stuff and a strain gauge on rod bolts.

[goldgunner]

All good info - thanks all.

Although this is a recurring thought and quandry for me, what brought it up this time was torquing the half axel cv joints at the 915 transmission sides. Impossible to get a straight shot with the hex bit on a couple of the screws, and requiring a variety of extensions to get a good elevated torque installation - whether with a torque wrench or without it.

I'm also pretty good at gauging torque by feel, however when you have only a very small radius to operate within, and/or a very awkward position that does not allow anything close to normal body or arm positioning, it becomes very difficult to gauge torque applied by feel.

I note that 2 responders above said that 'in theory' a straight extension would lose no torque. Yes, in theory our tools and Archimede's lever are solid, rigid and inflexible to present understandable physical relations. However, we KNOW that ALL things flex and twist cumulatively over distance. I would guess that even a 3/8 extension that is 20" long will lose at least 5 ft/lbs of torque against what is set on the torque wrench. That is, the torsional flex absorbs at least 5 ft/lbs effectively fooling the torque wrench. Maybe it's more, maybe it's less: I was thinking someone had done controlled tests to make objective determinations.

Thanks again,
Jim

[mattr_914]

To assemble case halves and heads on GT3 based motors, you cant rely on torque. We have a torque wrench that measures angle of rotation of the wrench. Like I said before, the rod bolts are measured with a rod bolt stretch gauge. I would say the rest of the torque values on an engine arent too important. A few things like the front pully and flywheel have such high torques, a +/- 5 N/m tolerance isnt significant. I adressed the strain with a universal joint, but I havent done the calculation. I was making the point that a torque wrench is nice, but not really critical for most applications.

[Eagledriver]

It's not just in theory that there is no loss in an extension. There is actually no loss. When you apply the force the wrench may twist but the force still makes it all the way through. Think about a balance beam. If you load each side with a large mass the beam will bend. There is still no loss since each side still has the same mass and the beam still balances. The only time there is a loss is when you shorten the lever by an angle or if you deformed the tool so much with your torque that you shortened the lever length.

In theory a beam type torque wrench has it's lever shortened slightly when it takes on it's bend but the amount is not significant and could be calculated. It could in fact be taken into account on the scale you read the torque from.

-Andy

[id911T]

There won't be any loss. Think of it this way. Suppose you have a torque wrench on a bolt with no extension, and you tighten the bolt to 40 Nm. As you do this, you note the anglular movement of the wrench, say 30 degrees. Then you put an extension on the wrench, and tighten an identical bolt to 40 Nm. If you turn the wrench the same angular displacement (30 degrees) as the first bolt, the torque will be less, due to the torsional "give" of the extension. However, that's not how you use a torque wrench. If you continue to the proper value, the same torque will be applied, and the wrench will still be accurate, you will just have to move the wrench slightly further to get it.