Torque Wrench with Extension

When you put an extension on a torque wrench, like a crow foot or a dog bone, the torque applied to the fastener increases, since the lever arm increases.

Comes up after I googled: Torque wrench with extension

I don't think you are off topic.

I do think you are mistaken.

"You will need less force to establish the required torque" is true,and the torque wrench can't take that into account. It will tell you the total force you are applying to the wrench, which with a lever arm will be less than the force applied to the fastener.

The extension goes between the torque wrench and the fastener. Not between the torque wrench and the mechanic. If a wobble joint exists so that the axis of rotation for the torque wrench and the fastener are not identical then less torque will be placed on the fastener. This is probably a moot point. I found out here at CR4 that setting the torque applied to a bolt or nut is not the most accurate method to tighten. Measuring the stretch of the fastener is better. (I wish I could locate those threads.)

You might try searching on "Bolt Integrity" or "HeviiGuy" (not sure of the spelling).

I'll attest to that. The only way to gauge the actual force a fastener is exerting on the two mated flanges/plates/"fastenees" is bolt stretch. That cancels out any friction resistance to turning of the threads of the fasteners during tightening.

Yes, correct! I hinted that little google search as my back up.

But as it turns out the extension is on the nut end. Which is a complete different story and answers for this are given below!

It just depends on how the torque wrench is built and how and where you extend it. If it is based on a spring loaded head, where the handle does not bend and you torque slowly enough (speed can ruin your torque point) an extension like a pipe can be used on the handle. When you extend on the axis and wrench in line, the torque stays equal.

When you start using extensions that shift the axis, like a fork wrench, your torque will be different.

Other responders are reading the question differently than I.

This extension... the driven component is not accessible with the tool alone, and an extension is required in-line with the driver socket and the torque wrench, thus:

Maybe I didn't pay my dues this month, but I just see an empty space.

No, if by "extension" you mean what Doorman's picture shows.

Yes, I meant the pictured extensions. Why or why not is the reading affected? Assume we are real world, so the extension can twist, etc.

The torque applied "in" one end of the extension is the same as the torque felt "out" the other end. There is no loss or gain anywhere even though twisting may occur.

I am NOT an engineer but it seems that if a long extension (any length) is used, then the force applied to the handle must be given to both the bolt and also to the extension. IE part of the applied force is absorbed in the extension by twisting it. Admittedly, this amount is negligible and may be totally ignored.......But it does happen.

No change as long as the torque wrench and its extension are all on-axis to the fastener.

If the extension has a flexible joint, all bets may be off.

Hmmm...why would a flexible joint make a difference?

This compliments of Wikipedia:

The magnitude of torque depends on three quantities: the force applied, the length of the lever arm connecting the axis to the point of force application, and the angle between the force vector and the lever arm.

In our case, I think the angle mentioned is between the wrench (lever arm) and the extension, which would be constant at 90 degrees.

I'm fairly sure that the extension and any joints won't affect the reading, but don't have a proof of that yet. Maybe this: the sum of the torques on the extension must be zero, or it would rotate. Therefore the torque at the wrench end and the torque at the fastener end must be equal and opposite.

What kind of extension ? of the lever arm or between square of TW and bolt head ?

If the first yes if second no.

Dear Mr.SSCpal,

The answer for (your question) is Yes.

DHAYANANDHAN.S

try this

Let's put this to rest quickly. There are three types of extension, lengthening the handle, inline with the center of rotation, and offset from the center of rotation.

Lengthening the handle increases the length of the lever arm thereby reducing the amount of force required at the far end of the lever to get the same torque on the inline fastener. Conversely, if the same amount of force applied at the end of the shorter handle is applied to the end of the longer handle then the force on the inline fastener increases.

An extension inline with the center of rotation causes no (appreciable) change in applied force. An exception are the individual extensions used on torque limiting nut drivers such as found in auto wheel shops. The thickness and materials are chosen so that when the desired torque is exceed the shaft distorts so that no further tightening is possible.

An offset extension placed on the drive end of the torque wrench has multiple cases, some are geared for torque multiplication and usually have their own readout dial to save the calculations that are required on a solid offset type. I'm only an EE so I'll defer to this site:

...that explains the calculations better than I could.

SScpal-

If the extension is between a torque wrench and a socket, which is on the fastener, There will be no change in the torque applied to the fastener IF---> the extension is in the same planes as the bolt, i.e. the bolt and extension are directly aligned. The wrench must also be at a right-angle to the extension. The extension is simple transferring the torque to the fastener in line with the direction without the extension. Anything that is set up differently such as a universal, a universal socket, flex extension will effect the readings and will not be correct..

This will not work with a crows-foot attachment unless you make some calculations to the wrench readings.

Good Luck, Old Salt

How would a universal or other joining affect the reading?

Up to 3° almost no effect over proportional to the cos of the angle between axis and variable according to rotation angle.

How did you get that?

What do you think of post 13?

SSCpal-

A universal joint in the line of components could lengthen or shorten the effective length between the socket and the wrench. If the universal is at an angle the effective length of the torque wrench is only the distance between a straight line going up the axis of the socket and the grip point of the torque wrench. Don't have time now to show the math but if the head of the torque wrench is 4" to one side of this line and the handle is 8" to the other side (assuming a 12" wrench) then the handle is only 8" from this line on the other side. Therefore your reading will be 1-1/2 greater than it really is.

Good Luck, Old Salt

I think I might have an answer. Consider this setup: two torque wrenches are connected via extensions to a U-joint set at some angle. Torque applied by wrench 1 will cause a reading on wrench 2 when it is held stationary. But the designations "wrench 1" and "wrench 2" are interchangeable and arbitrary; therefore the readings on the wrenches must be equal, and QED the torque is not affected by the U-joint.

Comments?

<Sigh>

You are correct that if the sum of the net forces and torques on an object in an inertial reference frame are zero then no change in motion happens. You are forgetting that a torque can easily become a force.

Lets take your example. Lay this assembly down on a table and make the angle of your U-Joint a 90° angle. Applying torque with one wrench will lift the other off of the table if enough torque is applied to exceed the gravitational force of wrench and extension bar or drive a force into the table. The table will produce an equal but opposite force to prevent motion. No torque will happen at the other wrench. The amount of force at the table will depend on the lever arm length from the U-joint to the point of contact with the table.

If you refuse to believe my analysis, then take your lazy.... (excuse me) perform the experiment yourself.

Redfred, simple real world U-joints only work for angles less than 90 degrees, but it's a mechanical limitation, not a logical one. Your point is correct if we tried to use such a U-joint at 90 degrees.

A double U-joint (I use one in my hot rod steering) can handle angles exceeding 90 degrees, and I think that validates my analysis.

How do you get that degree notation in your post instead of writing it out as I do?

If your browser supports it, the CR4 composition window has a menu bar above it. Selecting the symbol Ω pulls down a further menu of special characters.

Thanks, Tornado! 90°, 90€, 90¢...love it!

A loose or wobble end socket extension is a simple U-joint. Now you wish to change the mechanics of this analysis by claiming multiple U-joints are acceptable. Why don't you just include a differential gear design to an acceptable analysis for a socket extension kinematics. That brings up an interesting point, if two U-joints could make a 90° translation of torque without any loss of torque, why use a differential gear to translate axis of rotation in a rear wheel drive vehicle? A differential can change the gearing but it is not a requirement.

Again, the 90° limitation is a mechanical one, not a logical one. And we can put a black box in for the U-joint, and it could have 1:1 gears or whatever in it, and that doesn't change the analysis. The question is is the torque transmitted from wrench 1 to the socket/bolt/wrench 2 the same regardless of angle?

And of course car rear ends use gears because they are simple, cheap, sturdy and can change speed and torque.

SSCpal-

Good Answer! It is good for, for example, if the upper torque wrench head (the end with the male "socket") is held tightly in its location so that it does not move other than rotationally down the extension-universal-extension-lower wrench.

Not so good, but only a little not so good, if the torque wrench is used the way most people use them. Most people will hold the handle in one hand and hold the head in the other hand. When this happens the reading is incorrect since by holding the head most people will push somewhat with the head hand. This has the effect of lengthening the effective length of the top torque wrench.

On the opposite side of things, if your illustration showed the same parts but the upper torque wrench was shifted so that its head was on the left of the centerline and the handle still to the right of it, the reading would be lower if the hands were in the same places as the "not so good" example. This is because the effective length of the leverage arm is shortened.

Some people will say they always use a torque wrench in the correct way but less realize what they are doing incorrectly.

Really good answer! lots of common sense in it. The thing that makes engineering possible. I gave you a GA.

Good Luck, Old Salt

Thanks, Old Salt, for the GA and the considered response.

Restraining the heads of the wrenches is of course required, but I don't see a difference between doing it by hand or otherwise.

On your second point re the left/right location of the upper wrench, consider that the picture could be flopped about the vertical axis and that would give the "left" example, but nothing would have changed. And the handle location is arbitrary, since it rotates about it's axis and can be anywhere in the 360 degrees without changing anything.

This is really an interesting question. I've been googling it and have come up with lots of opinions, but not much proof. I have just one torque wrench, or I'd go try it in the garage and see if I'm right.

Too late to edit my post, but re the handle location: if I was tightening a bolt and the handle was to the right of the wrench head I would pull on the handle. If the handle was to the left of the head I would push on it. Either way the torque applied to the bolt (and extension if used) is the same.

SSCpal-

Based on my torque wrench experience and too many physics and geometry lessons I don't agree about the handle position but that is not the end of the earth. If possible I will try it tonight to see. If you have the time also, try it and we can compare notes.

Good Luck, Old Salt

I think I'll go buy another wrench so I can try it. These thought experiments make my head hurt (imagine what Einstein went thru, he couldn't test his theories and didn't live to see proof!)

SSCpal-

Ironic you mentioned Einstein. As a young boy my father took me to see him while he had some business meeting with him. Very "different guy" from most but now I realize he was very interesting. I wish I was old enough then to appreciate that experience. Perhaps it would have changed my pursuits.

Good Luck, Old Salt

SSCpal-

Very interesting question, probably one that will have as many solutions as there are submitters.

My left example is if the lower wrench and lower extension are kept in the same position and not rotated at all as you illustrated. The upper wrench and upper extension are then tilted to the left at the same angle as you illustrated it to the right. Instead of the upper head being higher than the handle it is now lower than the handle and at the same radius. This puts the handle closer to the axis of the lower extension. If the user holds and twist the head of the upper it will not read a correct torque. The upper wrench will read less than upper because two hands are (unconsciously) pushing the upper at a smaller radius without any addition of force.

The difference would be due to people using the wrench in a convenient way vs. the correct way. Also very few people use a torque wrench single handed or without any force from the hand on the head.

The Chinese say "a picture is worth a 1,000 words" but they never tried to illustrate something technical with words. If there are still questions, and I would be surprised if there aren't, I will make up an illustration. Don't have time right now, the "honey do list" has a couple of things that have to be done today. Will try to make time later if needed. She feeds me so I must comply.

Thanks for the question and making me think some more.

Good Luck, Old Salt

If the centre of the bolt to which the torque is being applied remains in line with the centre of the touque wrench coupling the torque measurement will be accurate unless a part of the coupling between the torque wrench and bolt has exceeded its yield strength.

I found another torque wrench and cobbled together the connection. I did the test with 30 foot-pounds, as much as I could handle freestyle alone. At several angles from 0 to 45°, the readings on the wrenches match. I might do it again tomorrow with help and try some other torques and angles, but I think it's gonna be the same.

Went to Northern Tool and got a bunch of U-joints and adapters. One of my wrenches is a conventional "click" unit, the other is a digital job with readout and beeps and lights. I used a 3/8" U-joint and several lengths of extensions. With 10 and 20 ft-lbs at angles of 0-60° the readings were the same on the wrenches for all torque/angle combinations and positions of wrench handles.

Old Salt, hope you can try it and confirm my results. This has been a good discussion and I appreciate all the posts from everyone. And I'm so envious of you meeting Einstein. I was just a kid when he died, but I had a picture of him on my wall. He was one reason I became an engineer.