For the engineers in the crowd

[SGKent]

The question is on torque.

Let's say I have a 20' rod that draws two steel boxes together. There is a nut on each end. one nut I anchor by welding to one of the boxes and the other I torque to 100 ft lbs.

In another situation I have a 1' rod that draws two steel boxes together. There is a nut on each end. One nut I anchor by welding to the boxes and the other end I torque to 100 ft lbs.

Next scenario same as 1 but I use a 100' linear extension on the torque wrench so I can stand 100' above the nut.

Next scenario same as 2 but I use a 1000' linear extension on the wrench so I can stand 1000' above the nut.

Q. Assuming the same wrench is used in all cases, do all 4 nuts actually get the same 100 ft pounds when the wrench clicks, or is that load spread out across different rods and extensions. I have a suspicion that the load the nut applies on the thread is affected by the twist absorbed in the distance of the rod and extension.

The reason for this question pertains to a pinion bearing nut that needs a specific torque applied to it, and whether the fixture that hold the transmission affects that torque compared to the factory appliance that hold the transmission in the exact plane of the pinion nut vs way out at the end of the transmission. I am thinking inertia eats away a little of the torque. Hence a 10,000' rod with 100 ft pounds of torque at one end would not apply 100 ft pounds of torque to the other end. In the transmission situation the question becomes does the deformation of the housing affect the end value on the pinion nut.

 

[cjp4627]

That's really complex description. A longer rod with torque applied at one end could in theory change the true torque applied to the nut. This would be based on rotational deflection of the rod. This deflection would be a factor of material, rod size, and shape. I don't have much more input that that with really wracking my brain.

 

[Coasterbuilder]

Torque is the rotational force you apply to the nut. It won't be affected by the length of the rod, or the length of the extension. What may be affected is the clamp load between the nuts. Torque is a means to approximate clamp load based on calculated bolt (or in your case, rod) stretch based on that applied rotational force and the material properties of the bolt (rod).

In the real world- i.e. not using 1000' foot extensions, measuring torque to derive clamp load is not accurate enough to make a difference.

Let the games begin.

 

[MEngineer]

So, you would think about it from two different directions...

The length of your torque wrench handle should not matter because the spring tensioned component is located in the ratcheting mechanism.

If you are talking about applying a given force/load to a static wrench (like the various cases that you described) then you have to keep in mind that there will inevitably be some level of deflection of the wrench handle over its length.

To specifically answer your question:

Yes, the length of a static wrench can/will have some level of an effect on your applied torque values, but I would be fairly confident in saying that in less the bending/flexing of the wrench is noticeable, then you don't need to worry about it. If the wrench is noticeably deflecting when you apply the torque load to it then you will have less accuracy...

The inaccuracy comes from the elastic absorption of the metal in the wrench. The more it flexes, the worse

Hope this helps some with the thought process.

FYI - the magnitudes of error we are talking about are most likely for any relatively normal wrench length and torque load probably less than or equal to 1%. If you are building airplanes or rockets, get the right tools and do it right because the FAA will get upset.

 

[niferous]

 

[ssdave]

Within the normal range of sockets, extensions, etc that a person has at their disposal, any combination of such parts will impart essentially the same torque to the nut. If you had a long enough, thin enough section, some torque might be lost to inertia and internal stresses. I don't know if you could measure the loss it would be so small, however.

 

[larry_g]

The length of the extention has no affect on the torque applied. If you have nothing to resist rotation at the end of the extension then you would see not torque reading on the wrench. The torque wrench measure the RESISTANCE to rotation provided by the nut cinching up on the stud or rod as you say. The long extension may wind up a bit and give you lost motion but it will not add or absorb torque applied. make sense?

Coasterbuilder explained the stretch vs torque.

A couple more threads on the subject:
http://www.garagejournal.com/forum/showthread.php?t=4751

http://www.garagejournal.com/forum/showthread.php?t=239376&highlight=torque

http://www.garagejournal.com/forum/showthread.php?t=94743&highlight=torque&page=4

lg
no neat sig line

 

[DenisG]

The torque wrench is a dumb machine. If you set it for 100 ft-lbs, it will click at 100 ft-lbs when that torque is reached, regardless of the circumstances at each end.

If you apply 100 ft-lbs to a 1000 ft rod, you need to satisfy Newton's third law if you want the body to be in equilibrium. That applies to translational and rotational forces.

Rod of different length will behave differently depending upon their length, but the torsional constant is determined by the properties of the cross-section (just like bending for a beam):
http://www.projectengineer.net/what-is-the-torsion-constant/
(i.e. - Using the same torque, it's easier to twist a long rod than a shorter one if the cross-section and material properties are the same. See the formula.)

 

[kbeefy]

I think it would be more appropriate to compare your holding fixture to the factory one, rather than pose the 1000' question.

Keep it Simple.


KISS

 

[SGKent]

The torque wrench is a dumb machine. If you set it for 100 ft-lbs, it will click at 100 ft-lbs when that torque is reached, regardless of the circumstances at each end.

If you apply 100 ft-lbs to a 1000 ft rod, you need to satisfy Newton's third law if you want the body to be in equilibrium. That applies to translational and rotational forces.

Rod of different length will behave differently depending upon their length, but the torsional constant is determined by the properties of the cross-section (just like bending for a beam):
http://www.projectengineer.net/what-is-the-torsion-constant/
(i.e. - Using the same torque, it's easier to twist a long rod than a shorter one if the cross-section and material properties are the same. See the formula.)

Thank you. Will look at the formula in the morning but I figured out tonight the solution requires a formula to solve,

Examole: if I torque a nut to 100 Ft lbs then attach a free standing extension above that torque wrench it would add nothing to the nut. If I add a second torque wrench above that 1st wrench and add 10 additional pounds to the first wrench while holding the force steady on the first wrench I would have 110 ft lbs on the nut. Or to hold 100 ft lbs I could reduce my pressure to 90 ft lbs on the first wrench from my hand and 10 from the second wrench above it to have a combined 100 ft lbs. From the nuts perspective it is exerting 100 ft lbs back against the two wrenches in the amounts of say 90 and 10 or 100 and 10 depending on which example is used. So I realized that this is more complex than many threads make it out to be, It is like trying to answer a question about parallel and series circuits with one answer. It takes a formula to decide what the force is at a specific point. The bottom of the lower torque wrench in the example has say 100 ft pounds on it but the top has 10 on it with the difference spread thru it and the handle. Very complex.

The reason for this question was I have a pinion nut to torque to a very specific torque. VW used a special trans holder that grabbed the trans in the plane of the pinion nut. The other solutions suggested included ones that would allow for some unmeasurable distortion in the transmission case. I was trying to decide if there is a difference between methods of holding the case while torqueing the pinion nut. The answer is yes there is a difference but it would take a NASA study to determine what it is. The question now is do I want to buy the needed tools for about $1000 or simply come close. I'll look at the formula that has been linked in the morning and decide.

In answer to those who say Loctite it and come close - yes but it was pointed out to me by a long time trans builder of this type trans that trying to remove the pinion nut after that will likely destroy the case. I feel like I am tiptoeing between land mines.

 

[Schurkey]

When using a torque wrench or even a ratchet wrench with SMOOTH MOTION, the length of the extension only matters in terms of wrench swing--the longer the "torsion bar", the more wrench swing is needed to achieve the torque.

HOWEVER, an extension on an impact wrench will absolutely reduce the torque delivered to the fastener. The impact wrench does not apply smooth motion; each impact blow loses force when it "winds-up" the torsion bar/extension, only to lose that force between impacts as the extension/torsion bar "unwinds". This is an example of "wasted motion", all it does is work-harden and heat the extension (and beat-up the square-drive.)

In your particular instance, I wouldn't be too concerned about transmission case flex affecting torque or damaging the case, although that might change depending on what YOU engineer to provide torque reaction vs. the official Kraut Can tools.

I'd be very unlikely to drop a thousand bucks on a special tool I was only going to use once or twice. How often do you torque these nuts? How difficult would it be to weld-up a suitable substitute that accomplishes the same thing?

 

[matt_i]

Probably they are trying to prevent the pinion shaft from getting bent by the side load on the shaft, or some compliance element in the transmission from taking up all the torque as sort of a pseudo "soft joint" or something they found was getting damaged (?) You could spend years trying to figure out the reason, they have already figured the practical reason out. Trying to solve a theoretical problem of esoterica is just a mental exercise. You just need to duplicate the tool.

There is no way Loctite will destroy anything. If you heat it with a torch up to around 250-300 degrees it becomes a liquid with no ability to stick to anything. Even the red "permanent" Loctite...

If it were me, I'd try to get a picture of the thing, take some dimensions, and reverse engineer it as suggested. Having some metal fab capabilities would absolutely help.

Torquing a collapsible spacer in a pinion is fairly serious, you risk the preload on tapered roller bearings being wrong, either side of this is something that's going to blow up on you in spectacular fashion. The collapsible spacer is designed to have a predictable linear torque value vs the preload, vs. something "hard" like a machine tool spindle.

 

[MoparTrucks]

My head hurts.


Sent from my iPhone using Tapatalk

 

[APEowner]

The reason for this question was I have a pinion nut to torque to a very specific torque. VW used a special trans holder that grabbed the trans in the plane of the pinion nut. The other solutions suggested included ones that would allow for some unmeasurable distortion in the transmission case. I was trying to decide if there is a difference between methods of holding the case while torqueing the pinion nut. The answer is yes there is a difference but it would take a NASA study to determine what it is. The question now is do I want to buy the needed tools for about $1000 or simply come close. I'll look at the formula that has been linked in the morning and decide.

In answer to those who say Loctite it and come close - yes but it was pointed out to me by a long time trans builder of this type trans that trying to remove the pinion nut after that will likely destroy the case. I feel like I am tiptoeing between land mines.
__________________

You're over thinking this. Figure out a way to hold the pinion or lock the pinion and hold the case that doesn't damage anything and torque it slowly to spec. Yes, there's a minute difference in torque while you're turning the fastener but once you reach the point that the wrench is measuring the specified torque with no fastener movement the torque at the nut is going to be the same as the torque at the wrench. You'll have balanced equal and opposite forces. If they're not balanced then you'll still have motion.

 

[rsanter]

The basic math side of it says there is no difference because the materials are treated as ideal.
The more complex math says that there is a difference due to material stretch, friction, and torque adsorption (kind of a form of stretch).
You have not given enough information to do anything but the basic math.

You need to supply material type, hardness, with or without lube on the threads and under the nut....etc to be able to calculate the complex math.

For most materials that would be used in that scenario ( steel) there will be a very very very small difference in the torque readings. If on the other hand you are using plastic or soft aluminum there would be a bit more

Bob

 

[SGKent]

How difficult would it be to weld-up a suitable substitute that accomplishes the same thing?

unfortunately I am not a welder and my attempts to find help here in Sacramento have not been very successful. I have been unable to get my hands on a tool to duplicate, or a spare case to find someone to build the device. People in the community who build this type trans are conflicted on alternate solutions - and unfortunately sometimes the nuts do come loose when they are under or over torqued, destroying the pinion and ring.

trans carrier

gear carrier stand

One alternate solution for the trans carrier when torquing the pinion. The concern by some is that the distance between the two pins is quite finite. They are used to maintain the width of the trans case rather than location of the bell housing. Most likely this is the solution I will go with.

Here is a photo of a bridge between the pins holding the case width when setting the ring.

another is a 4' 2x4 thru the axle flange area.

 

[joe_padavano]

Unless this is a beam-style torque wrench, the torque sensing of the wrench is all locally in the wrench head. The wrench head neither knows nor cares how long the wrench handle is. ***-U-MING that the wrenches are calibrated the same, the friction in the fasteners is the same, the force on the wrench handle is applied with equal care, etc, etc. the torque measured AT THE WRENCH HEAD is the same.

Of course, this does not mean that the clamping force in the rods is the same.

The longer rod stretches more, so it will act like a softer spring. The short rod is much stiffer and will impart a higher clamping load at a given torque. This is why they sell rod bolt stretch gauges and why modern engines require torque-angle meters on the critical fasteners.

 

[matt_i]

If you want to hold the case width to a dimension you need to use the existing dowel pins, not a lashup of hex bolts as shown. To create the spacing of holes that enforces the width, (equivalent to the cover I would guess) you need a milling machine and a boring head. I would personally make the plate wider, lets say 3/8" x 2" steel flat bar and then bore a larger clearance hole for the socket extension.

Also you'd need to precisely measure the cover dowel bores, and their centerline distance. Again, a job for the milling machine and a dial test indicator.

On mission critical bolts like that I would use the red loctite in addition to the torque spec provided, making sure to replace any fasteners purposefully driven into yield (torque + angle). While you have it that far apart, it takes about 10 seconds longer per fastener to apply the threadlocker.

 

[Schurkey]

One alternate solution for the trans carrier when torquing the pinion. The concern by some is that the distance between the two pins is quite finite. They are used to maintain the width of the trans case

HOW does torquing the pinion affect case diameter? I don't see that as something worth considering. You need a method of countering the torque reaction, and the counter should be something that causes no damage to the case.

Case diameter may be affected by carrier bearing preload--if the case is poorly designed. I'll accept that there may be a need to consider case diameter when loading the carrier.

But not for torquing the pinion nut.

 

[Skidonenko]

If the job is that sensitive why not buy a flange bearing to fit the extension diameter. Make a plate that will bolt to multiple holes/fit over the dowels. Oversize the hole in the plate above the pinion nut. Assemble it with the 2 extensions pictured and the flanged bearing on the outer extension (could be substituted for one longer one. Center the flange bearing as needed and tighten it to the plate.

Just thinking out loud, your mileage may vary

If the desired bearing inner diameter a larger one with a bronze bushing around the extension could be used. A set of transfer punches may be your friend here. Lots of ways to accomplish what you are looking for. Are the diagrams of the housing pattern avaliable? One could be made fairly easily out of a plate that would in turn make that thing rigid enough without turning it into rocket science.

 

[Coasterbuilder]

I simply don't believe that the difference will exceed the margin of error (tolerance) of the torque wrench you are using. Torque is only a proxy measure of clamp load- it's just easy,convenient, and works well in 99% of the applications we, as mechanics typically utilize,

If you are convinced that the clamp load required for this application is that critical- I would find another way to measure it than a torque wrench.

If the factory used a torque wrench, and your wrench is calibrated- you are probably fine

 

[Skidonenko]

You could support the rear of the case many ways too. Find some pipe with the correct Id toatch the dowels. Cut a section of sizeable square bar or thick square tube. Drill holes in the square small enough to press the round tube into. Should be rigid enough.

Or make a template to make a plate large enough to cover the whole rear area. I'd install studs into the case and press it into a piece of cardboard. Bolt the plate to the case using spacers to ouch it way from the ring gear.

It could also be drilled to install a jacking bolt to press against one of the ring gear bolts to secure the assembly for torquing

 

[Skidonenko]

I guess this would be akin to supporting an old online engine from just the bell housing g and not being able to rotate the crank due to flex. Yes that is more weight/mass involved but this is smaller and the aluminum may not be that stout.

 

[sanddan]

Torque is the rotational force you apply to the nut. It won't be affected by the length of the rod, or the length of the extension. What may be affected is the clamp load between the nuts. Torque is a means to approximate clamp load based on calculated bolt (or in your case, rod) stretch based on that applied rotational force and the material properties of the bolt (rod).

In the real world- i.e. not using 1000' foot extensions, measuring torque to derive clamp load is not accurate enough to make a difference.

Let the games begin.

You nailed it!

 

[iceman510]

Also, what is the nominal torque value? I have worked in automotive for 27 years, and the usual allowable torque range on a fastener nominal value when audited was nominal +/- 10 %. I don't think your variation will be outside of that range with the factors you are considering.