John Struthers

Good post!
Great responses!
Might as well take care of the boots while Zack's 81' is off the ground with the exhaust completely off.
As to myth-busting test criteria: If its standardized and repeatable, you know the rest.

Now then...
If there is a full moon and you can see it during the day... how does one;
1. Draw a flat-plane, planetary map depicting the moon in relation to the earth/sun.
2. A planetary shadow (cone) map of same.
3. A ground view mapping depiction of the seqential movement -daily overlap mapping with degree cone from horizon- of the same event.
4. A method of explaining all of the above to a hard -headed wife.

SharkSkin

John, you might start with the fact that you will never see a "true" full moon during the daytime, unless you're in the arctic circle. The exception is that you will see the full moon rise near sunset, but that's more "evening" than "daytime" IMHO.

Back on the subject of socket extensions for a sec... key factors if you try to repeat the experiment: (1) use enough torque and thin enough extensions(and sufficient length) to get meaningful flex in the extensions, (2) use a bolt that will not stretch significantly at that torque.

jeff jackson

"The angle of the Dangle"... is directly proportional to the .....

SharkSkin

Bob, no further thoughts on this?

TopDownL79

Have you ever tried this with a calibrated dial indicator torque wrench?

This topic came up during a dynamics class many years ago. While the phenomenon was indeed repeatable using different "click" type torque wrenches, using the dial type showed no difference dependant on number of extensions (aside from angular alignment). We later repeated this using an electronic torque meter.
What does this mean? Who knows! But it sure wasted a Friday afternoon class...

dr bob

I couldn't duplicate your results, Dave. ;(

One might postulate from your proposal then that there is torque loss in the central driveshaft due to length, and similar losses between the drive flanges on the differential and the wheel hubs. Especially since there is virtually no motion involved, the "friction loss" is nil. Does using a deep socket make a difference compared to a shallow socket and a 1.5" hard extension? I need both the deep socket and the mini extension when tightening the C4 wheels on my car.

=====

Important point: So long as the drive head on the wrench and the bolt are coaxial, the torque will be exactly the same, no matter how long or how springy the extension is. Breakaway torque will be the same. Move the wrench head off axis, and you start to have effects similar to moving the wrench head up and down the handle. The amount of torque required for breakaway on the head will be the same, though.

j.kenzie@sbcglobal.net

Hi,
The torque question is pretty confusing. The best way I can explain Dave's results are as follows:
Think of the torque wrench as a springy wrench with a gauge. The more torque, the higher the gauge reads because the handle and the output (socket) have deflection or movement between them. If you add an extension that is especially springy itself such as small diameter, long length or multiple slip joints them this introduces more movement between the input and output of the extension, behaving like a torque wrench itself but without a gauge. The only trouble is the torque wrench measures only its own deflection and not that of the extension(s). The bolt sees the total torque transmitted from wrench through extensions to socket, but the torque wrench only measures part of this pathway. A similar effect occurs if you put a spring type bathroom scale on top of a (truly) resilient floor; you will not get accurate weights until the floor is fully compressing and no longer resilient.
In actual practice none of this probably matters much.
YMMV
Dave McK.

jeff jackson

I look at the "retorque" like this...if using the same equipment used to "break loose" the nut in question...use what you used to "break it loose"...tight IS tight...and 336#/ft being an OEM specified torque figure,... does NOT change this... whatever power was exerted to break loose this nut....should be applied in the opposite direction. If 336#/ft. was required to break it loose, then lean that hard on the same setup when retightening, and you'll definitely be "in the ballpark"...you WON'T have to worry about your wheel falling off while driving... This "to the knats ***" torque banter is rediculous, when you are talking 336#/ft. if you ask me...

SharkSkin

Jeff, we were talking 60# on flange bolts, and we were talking three feet of extensions. Those flange bolts love to loosen up anyway, so undertorquing on install is contraindicated. I'd say use a short extension on final torque-down and forget about it.

But this has become more interesting... torque loss is something that I've taken for granted as I've seen the effects. I'm not talking about the well-known torque loss one gets with extensions on air tools, that's an effect caused by the fact that hammering on a spring is less effective than hammering on something solid.

Bob, I just went out and reproduced these results twice and did some other experiments on half a dozen bolts, washer stacks, etc. I wonder what we're doing that's so different? Physics should be the same in my garage as yours... My old 6" Wright extensions have narrow shanks, .353" so they are a bit springy. If you have beefy, high-quality extensions or if you are using a grade-10 bolt your results will be less noticeable. The bolt needs to be long enough to stretch some and the extensions need to give you some flex -- 60° or more.

Each time I did the exercise as above, I saw at least 4-5° rotation of the bolt head when I switched to the short socket. I did some other experiments, where I would torque to 70 with the extensions then back the wrench off to 60 ft/lbs and increase the setting until I saw the bolt move. It seemed that for my setup, the bolt began to move when I reached 66-67 ft/lbs again confirming that there is a difference.

I imagine that there is some loss in the drive shaft as well, for the same reasons. Twist a piece of metal, and some of the energy that you put in converts to heat. Where does that heat come from if the torque going in is the same as the torque coming out? Take a coathanger and bend it sharply back and forth... does the metal get hot? there's an example of internal friction.

Surprisingly, I've found very little info on the web about this. Nothing authoritative so far, but judging by this link I think it's safe to say we're not the first ones to kick this idea around.

SharkSkin

A follow-up from a Tim Gross at teamtorque.com, a torque wrench calibration business:

"I brought enough extensions from home to make 2 footer, adjusted a torque wrench to 70 ft/lb, and put it onto a digital tester. I pulled this combination three times with an average reading of 69.89 ft/lb. I then took the extensions off. Using only the torque wrench onto the tester, I again pulled three readings and the average was 70.08."

So, his conclusion, there is measurable(but not significant) torque loss through extensions. He went further to say that my setup must have relaxed it's tension somewhat, maybe due to soft washers. I thought I accounted for that by waiting a while, then retorquing with the extension, then torquing without extensions after the bolt/washers had plenty of time to relax. Also, a couple of the extensions that I had to use to get this much length are relatively slender and flexy, which may have exaggerated the results.