Extension changes torque?

[Kevin54]

Correct and correct again.

Not true. The image above will provide two different torque ratings.

In the interest of ending the argument:

I'm going to have to agree with DSL. Although a wrench at a right angle to the torque wrench does not change the length of the torque distance, doing it that way allows for all sorts of errors. I would say that if you had three people torquing three bolts of equal length, you will find that all three readings will vary. For one, you are pulling on the torque wrench, but if not pulled the same way by all three people and all standing in the same place, they will pull on it differently. One could torque a bolt and be nowhere near close. If person number two pulls on the torque wrench, and person number three pushes on the torque wrench. It would be possible to click the wrench and not even move the bolt.

If I get some time today, I'll set up an example and take some pics. Here's a quick sketch as to what I'm talking about. I would thing that a wrench straight out of the end of a torque wrench would give a more consistent reading. Not accurate, but consistent. over one being at a right angle. One would have to figure out the difference in torque because of the length difference.

 

[mjozefow]

If you think of the torque wrench and adapter as two sides of a rectangle then it is easier to envision. Imagine a fat torque wrench with the socket at one corner.

As long as the force applied continues to be perpendicular to the adapter, the lever arm length remains unchanged despite the offset. Forces applied purely in the Y-axis do not affect forces applied purely in the X-axis.

BTW, Pratt and Whitney wasn't and isn't a company that produces products with tolerances that can be left to chance or approximation.

 

[Steinmetz]

I love watching engineers proving a point.

Next week we will debate properties of a vacuum

Sent from my SAMSUNG-SGH-I847 using Tapatalk 2

I used to work in high vacuum engineering. It's not as simple as you think.

 

[Steinmetz]

So to conclude, as long as whatever it is I'm using is at a right angle to my wrench, it's not changing the end value of torque. As soon as I add something in line with the wrench which adds a length off the head horizontally, it needs calculation.

That's correct.

 

[Steinmetz]

"...BTW, Pratt and Whitney wasn't and isn't a company that produces products with tolerances that can be left to chance or approximation…"

P&W also isn't a company that propagates bogus engineering analysis.

 

[A_Pmech]

Do you really think companies would have people calculate the actual torque?
As you saw in my example, the actual torque differed by only a few ft-lbs.

To understand why I posted a page out of the P&W Standard Practices Manual, you would have to understand the company culture. P&W has been synonymous with "Accuracy" for over 150 years. In fact, it's their motto. "Close Enough" is not something the people at P&W consider acceptable - it would kill thousands of people in their line of work.

P&W made the original U.S. standards of length which lie in the vault at the NIST. They are accurate to better than one part in one million. The measurement systems company, which comprises the remains of the original tool company, still builds the most accurate micrometer in the world. The aircraft division provides it's mechanics with a torque correction slide rule, an example of which is here:

http://www.worthpoint.com/worthopedia/collectible-pratt-whitney-torque-wrench-slide

If P&W says "no correction required" you can take that answer as coming from people at the pinnacle of engineering in America.

This is where I bow out. As long as the conversation stays technical in nature and doesn't degenerate into name-calling and drama I'll keep the topic open.

 

[NotStock]

As long as the force applied continues to be perpendicular to the adapter, the lever arm length remains unchanged despite the offset. Forces applied purely in the Y-axis do not affect forces applied purely in the X-axis.

^This.

Apply your force normal to the torque wrench and you will not change the torque applied to the fastener. Start applying it at an angle and the X component perpendicular to the adapter lever arm starts to apply a torque to the fastener not measured by the torque wrench. Use it right and you'll be just fine.

To over simplify this, think of pushing on the end of the torque wrench with the adapter. It'll spin the fastener but the wrench will never click.

 

[tyndall]

P&W has been synonymous with "Accuracy" for over 150 years. In fact, it's their motto. "Close Enough" is not something the people at P&W consider acceptable - it would kill thousands of people in their line of work.

I build Wasps. I'm gonna have to respectfully disagree with you there.

 

[Steinmetz]

Boeing Torquing Manual (Doc. No. 6M54-400) provides in pertinent part, for the perpendicular adapter configuration: "…No correction necessary. Torque wrench indicator reading needed equals required torque…".

My copy is more than a few years old, but I don't think that this fundamental fact has changed.

 

[OxJaw]

If I didn't feel this topic was this important, I wouldn't go this far to prove my point.

This is important to know, especially when torquing head studs where access is limited and these extensions are frequently used.
As I've said before, the value the torque wrench is set at will not indicate the applied torque to the fastener. The longer the extension, as well as the angle it makes to the normal of the torque wrench will affect the actual torque applied.

I understand the point you are trying to make, although I believe you are wrong.

But what you have proven, and I think most people are forgetting here, is that even if you are correct you well within the tolerances on any torque wrench that your average Joe is going to buy.

Last I looked Snapon/CDI was +/- 2%

 

[arms1970]

I love watching engineers proving a point.

Next week we will debate properties of a vacuum

Sent from my SAMSUNG-SGH-I847 using Tapatalk 2

They sure got more time then I do

 

[Techie1961]

I like to take things to extremes when dealing with principles like this so here goes.

If the adapter that is on the torque wrench is one mile long and you have a 24" wrench, you will get an accurate result AS LONG AS the wrench is at 90 degrees to the adapter AND you continue to apply your pulling force parallel to the mile long adapter. This means that you will need to travel a lot as the fastener is tightened as you will be at the end of an adapter that is a mile long. As long as you continue to pull parallel to the adapter, the reading is correct. If you pulled the wrench in a direction that is tangent to a circle around the fastener, you would get an incorrect reading.

This is how I proved it in my head anyways. Does this make sense?

 

[Chuck122]

I'm going to have to agree with DSL. Although a wrench at a right angle to the torque wrench does not change the length of the torque distance, doing it that way allows for all sorts of errors. I would say that if you had three people torquing three bolts of equal length, you will find that all three readings will vary. For one, you are pulling on the torque wrench, but if not pulled the same way by all three people and all standing in the same place, they will pull on it differently. One could torque a bolt and be nowhere near close. If person number two pulls on the torque wrench, and person number three pushes on the torque wrench. It would be possible to click the wrench and not even move the bolt.



If I get some time today, I'll set up an example and take some pics. Here's a quick sketch as to what I'm talking about. I would thing that a wrench straight out of the end of a torque wrench would give a more consistent reading. Not accurate, but consistent. over one being at a right angle. One would have to figure out the difference in torque because of the length difference.

I agree that using the extension makes all kind of errors possible but I think those errors would be due to the extension not being perpendicular to the wrench or not applying the force normal to the wrench. I do not think that there is a correction for that. In other words, if you are going to break out the protractor to verify your angle and give the right correction, you may as well just put it square and not have to correct. The correction that DSL gave us assumes that the force vector is not applied normal to the wrench (ie: the wrench is used improperly)
In any case it is an interesting topic. Keep'em posts coming!

 

[SuzukiGS750EZ]

Realize today I can torque with a deep socket and be fine. Close enough for me.

 

[jvitez]

I came across this exact issue when I had to torque a large nut using a crows-foot wrench installing a steering box stabilizer on my Dodge Cummins. I turned the crows-foot 90° to the long axis of the torque wrench and all was well. I had to think about for a while initially, but I have enough of a science background I was able to figure it out.

This is a very interesting thread. I much appreciate the scientific detail. (I guess that makes me a science geek, Mea Culpa. ) Many thanks to all the erudite, technically trained gents who've posted.

 

[Outlawmws]

All the math, and all the theories in the world won't prove a point as fast as a simple test.

Someone on GJ will have a torque tester (or access to one). Test a torque wrench on the tester, it will register "x"; add the right angle adapter and try it again...

Thread end.

 

[Cadillac fan]

Does using an adapter change the reading? I mean putting something like a 1/4 male and 3/8 female or 3/8 female and 1/2 male.

I was at the market and saw some 1/4 inch and 3/8 torque wrenches but wasn't sure if using an adapter would do anything.

 

[zkling]

But what you have proven, and I think most people are forgetting here, is that even if you are correct you well within the tolerances on any torque wrench that your average Joe is going to buy.

Last I looked Snapon/CDI was +/- 2%

Winner winner chicken dinner.

Mathematically DSLTRK is correct in that a 90° extension can and does change the applied torque value. HOWEVER with that said, the common extension at 90° is so short compared to the overall length of the torque wrench that the change in torque value is very small. Typically smaller than the accuracy and resolution of the torque wrench itself. I forget exactly where but IIRC Ford, Honda and a military issue service manual have a foot note about 90° extensions, limiting them to a specific % of the effective length of the torque wrench. Something around 2-3" for a standard 1/2" torque wrench.

All the math, and all the theories in the world won't prove a point as fast as a simple test.

Someone on GJ will have a torque tester (or access to one). Test a torque wrench on the tester, it will register "x"; add the right angle adapter and try it again...

Thread end.

That is the difference between knowing what and knowing why though. You as an engineer should appreciate that

Does using an adapter change the reading? I mean putting something like a 1/4 male and 3/8 female or 3/8 female and 1/2 male.

I was at the market and saw some 1/4 inch and 3/8 torque wrenches but wasn't sure if using an adapter would do anything.

As long as you stay coaxial to the drive square the torque value will not be changed. Just be careful using the extreme ranges of a torque wrench for a given size. Like pressure gauges most torque wrenches have a sweet spot between the min and max of their range.

 

[Steinmetz]

Winner winner chicken dinner.

Mathematically DSLTRK is correct in that a 90° extension can and does change the applied torque value. HOWEVER with that said, the common extension at 90° is so short compared to the overall length of the torque wrench that the change in torque value is very small. Typically smaller than the accuracy and resolution of the torque wrench itself. I forget exactly where but IIRC Ford, Honda and a military issue service manual have a foot note about 90° extensions, limiting them to a specific % of the effective length of the torque wrench. Something around 2-3" for a standard 1/2" torque wrench.




That is the difference between knowing what and knowing why though. You as an engineer should appreciate that



As long as you stay coaxial to the drive square the torque value will not be changed. Just be careful using the extreme ranges of a torque wrench for a given size. Like pressure gauges most torque wrenches have a sweet spot between the min and max of their range.

The aforementioned Boeing Manual specifies that the force applied to the wrench handle must be within +/- 3 degrees of 90 degrees, thus rendering the analysis of DSLTRK moot in any case.

 

[zkling]

The aforementioned Boeing Manual specifies that the force applied to the wrench handle must be within +/- 3 degrees of 90 degrees, thus rendering the analysis of DSLTRK moot in any case.

Do you have a copy of Beer & Johnston handy? DSLTRK is not wrong from a purely mathematical sense.

 

[Outlawmws]

Yes ZK, I know; but these ******* matches get tiresome after a while, and when it's splitting hairs, that simple test would end the endless speculation...

As an engineer, I do rely on well know and well established standards and references. I don't go out of my way to "disprove" them unless there is an actual issue. I'm far too pragmatic, and I have a practical engineering mindset, not one that gets all wrapped up in analysis and theory unless it's new territory or a unique application.

I'ts been pointed out the error can't really be measured easily. Then it's a moot point. 99% of all torquing isn't about the accuracy of the torque being applied anyway; i.e Head bolts: + or - 5 lbs ain't going to break the bank, as long as the bolts on a particular head are within the setting used: all at 65 lbs, fine. all at 75 lbs? also fine.

 

[crane_guy]

It will change it since an extension can twist. The effect is small enough that it wont matter

Here's the correct reply. Its good practice to use chrome sockets and extensions when torquing a bolt. If you must use an extension, use a high-quality extension and the shortest one that works, to keep the twist as low as possible. Also use it with all the bolts in the connection involved.

 

[Westly]

Here's the correct reply.

Actually that was one of the incorrect repiles. The correct replies are in the thread too..

Some extensions can twist quite a bit, but all of them will be applying the same torque to the fastener once it's twisted as much as it's going to twist and it all becomes stabilized.

But twisting extensions are something I hate anyway. They feel to me like something is stripping. In my experience some of the cheapest, flea market type, extensions are the hardest so they're what I use. Never broke one yet.

 

[zkling]

Yes ZK, I know; but these ******* matches get tiresome after a while, and when it's splitting hairs, that simple test would end the endless speculation...

As an engineer, I do rely on well know and well established standards and references. I don't go out of my way to "disprove" them unless there is an actual issue. I'm far too pragmatic, and I have a practical engineering mindset, not one that gets all wrapped up in analysis and theory unless it's new territory or a unique application.

I'ts been pointed out the error can't really be measured easily. Then it's a moot point. 99% of all torquing isn't about the accuracy of the torque being applied anyway; i.e Head bolts: + or - 5 lbs ain't going to break the bank, as long as the bolts on a particular head are within the setting used: all at 65 lbs, fine. all at 75 lbs? also fine.

I completely agree, it just seemed that majority thought DSLTRK was out in left field. While to some it may be considered over analyzing he is correct from a mathematical standpoint. Cases like this is how that shop hearsay and miss informed facts get started. Saying a 90° ext doesn't change the torque value, while usually correct in an everyday in the common shop application, is missing the underlying fundamental why. Which is the reason I questioned him in the first place. I have a feeling you would agree all too often there is a lack of fundamental principles in a shop environment.

Case in point.

Here's the correct reply. Its good practice to use chrome sockets and extensions when torquing a bolt. If you must use an extension, use a high-quality extension and the shortest one that works, to keep the twist as low as possible. Also use it with all the bolts in the connection involved.

 

[Outlawmws]

Not only in the shop environment; I've seen it lacking all too often in engineering, and in so-called scientific experiments. 98% of the time it really doesn't matter in the shop environment, so long as good shop practices are known and applied.

With that I'm out. Until someone posts a test, the discussion is still