Troque Using Crowsfoot vs Socket Not The Same

Try this site:

http://www.engineersedge.com/manufacturing_spec/torque_wrench_1.htm

It depends upon the alignment of the handle with the opening of the foot.

I have proven to myself and others here that the orientation of the crowsfoot does not matter. I have tried the crowsfoot at 90 degrees, straight out, and at 45 degrees. There is no difference. Using the crowsfoot and torqueing until I get a dial reading of 105 in.lbs. (which according to all documentation should give me more than 105 at the nut) - the nut still moves at approximately 25 in lbs when using the same torque wrench with a socket.

GA. Wished we had those plug in calculators when I was in school! Not knowing the length of the torque wrench, who knows? In reality he would have over torqued the nut if he didn't correct for the extension of the crow's foot.

I prefer not to use a crow's foot for any torque sensitive fasteners, and if I do have a situation where I need a crow's foot, I prefer to use one designed for tube fittings. They have a little extra material to encompass the fitting.

With the extension inline with the wrench (as shown on your diagram)

T(E) = T(W)* (L+E)/L

E - Effective length of extension, stated as 2.3"
L - Lever length of the wrench - center of grip to center of drive
T(W) - Torque set on the wrench
T(E)- Torque applied by the extension to the fastener.

This is the math... this doesn't explain the huge disparity in the readings you are reporting.

This is what everyone has been taught. I believed it until we ran into this problem. All evidence so far shows that using extensions such as a crowsfoot does not work like you have been taught. If you look at my original post and see if you can answer the question - How does rotating (torque wrench) around point A give enough force (the question mark) to apply a final torque of 105 in. lbs on nut centered at point B. I don't see that happening.

Don't use the crows foot.

Rumor has it that Snap-on Tools has a chart that covers this.

A cursory look produced no results

That is part of the corrective action we are going to put in place. The problem is we have to explain to our customer as to why we had a problem and why we are going to change the way we are going to torque. To answer this I need to explain why using a crowsfoot does not give results as expected. I believe in our case the problem is exaggerated due to the size and torque value. The nut is a 1 - 11/16 nut. Experimenting with smaller extensions and nuts (Ex. 3/8" nut and 20 in lbs) you can't see anything wrong. So far in this post, I haven't seen anybody take a shot at answering my original question. I appreciate all inputs. No answer indicates nobody has run into this.

How's this. Torque is force/length, or distance.

The torque wrench is calibrated for the length from the handle to the center of the drive shaft, thus the center of the resistive force, the nut. Extending that length increases the effective amount of torque applied to the nut while the wrench is still only measuring the torque applied over the distance from the handle to the center of the drive shaft.

Depending on the resolution of the wrench smaller errors may not be noticed.

In other words a wrench that reads from 0-150 in lb may not record an error of only 5 in lb on the lower end of the scale.

The problem is the information you posted is the standard information that has been around forever. I was trained to this. I am very familiar with this information. At my company I am certified to perform torqueing. We have discovered that in reality it does not work. (See my original post and my previous responses.) I believe that everything we have been taught about using extensions and using the torque correction formula is not true. on small fasteners you don't see the error the values are close enough (between using a crowsfoot or a socket). I have (8) connectors (four different size D38999 with Jam nuts) They get torqued to 33, 57, 72 and 105 in lbs. based on their size. Our current process is to use torque wrench with crowsfoot to torque. We do this because we use adaptors to hold the connectors so they don't move while we torque. All connectors show they were under-torqued significantly when torqued with crowsfoot and checked with socket.

Then have somebody independently test your theory....without arms length independent confirmation, you can't be taken seriously.....sometimes we're so close to something we don't see the obvious....

Would you tell us the make and model of torque wrench you are using?

I'll ask the question: you have made certain there is no mechanical interference? I know, I know... I'm just askin'.

These 'adapters' you speak of, they don't come into play at all? Do they clamp to the workpieces in a way that might be contributing to this problem?

Actually, to take the two extremes, on your torque wrench you can either apply pure torque (you need TWO hands for that, and that torque is transmitted EXACTLY the same whatever the angle and the shape of the wrench is, as long as ALL are on the same plane) OR you can apply pure force, some distance from the square socket holding pin, and the torque the levering will translate it to, is relative, and WILL vary by the socket angle AND the distance you apply the force. Take an extreme case to understand what I'm saying: The socket to be the same length as the distance that you apply the force, but being towards you, not the other side. The force you apply to the torque wrench will give ZERO torque to your bolt, but your torque wrench will click, will "see" torque. Now anything in-between the two extremes can be your case, meaning you can apply BOTH some pure torque AND some pure force, so it's easy to get you confused. S.M.

Why would you engineer something requiring specific torque and leave yourself no way to actually, accurately torque it?

Just sayin....

Nice one Fred! All puns intended, huh?

Take two teaspoons of this

and call us in the morning.

Redfred - I gave you a good answer because unknowingly you're comment about galling actually came close to the root cause. I attempted in my original post and my replies to provide all relevant information so that people could possibly help. Example I stated it was a 0-150 in lb dial torque wrench The manufacturer of this wrench would not add anything to solving the problem. For some you're comments about how did we get so far - you have no clue where we are at or how we operate. I can tell you this - we are not in production yet. We are at the beginning of the program.

I could go on but I won't - For those who might be interested in the ROOT CAUSE -

I mentioned in one of my responses that what we were torqueing was D38999 connector Jam Nuts. It end up that this particular one which is the largest (1-11/16)

being aluminum and very thin is very easily distorted. While attempting to torque the nut was actually getting distorted which caused it to lock up on the threads. Lesson learned you can not use a regular crowsfoot to torque D38999 jam nuts as they will distort and give you false torque values. We now are looking for a box style crowsfoot

so that the nut is grabbed on all sides. For all those who tried to help - thanks. I will ignore those that felt the need to make assumptions about how we do business.

At NO POINT, until now did you ever state that these nuts were aluminum.

This whole thread is an insult to the people who tried to help you!

Lyn - you were too fast for me. I was just getting ready to submit a correction to my last reply. The nuts are not aluminum. Sorry. I believe they are a nickel alloy (I would have to look at the datasheet. Bottom line is that the nuts for D38999 connecotrs are pretty flimsy and can be distorted if wrenched with an open end wrench. This will cause erroneous torqueing.

Ok, so given your own statement: "can be distorted if wrenched with an open end wrench."

Why are you attempting this procedure?

See my response #31

Thanks Larry.

My current assignment is Mechanical Reliability. I am still not figuring this out from that standpoint unless you are hoping this will never need to be replaced/removed.

In your diagram you note that you can use a straight socket and torque wrench. That would be the right tool for something this size and stated delicacy. Using open end anything and offset torque wrenches would just add complexity and unreliability. Plus potential for damage beyond the part and then down time, unexpected expense, higher labor costs, etc.

If you have to do this in this fashion go for it. I tend to look at things from a warranty or future repair in the field, planned maint. or customer happy standpoint.

All The Best!

R

Flimsy isn't the right word, they don't engage continuous threads and are drilled for locking wire!:

Glad you liked my "tongue in cheek" answer. We all need to laugh at all of our problems from time to time. Sometimes the joke itself can provide some insight. I'm glad you found the actual root cause of your problem.

Now, one of the biggest problems I see with your situation is, if your connector has an elastomer seal, you will not get a true torque on the jam nut until the o-ring is completely crushed.

I have modified some of my snap on sockets for situations just like your or used a fitting crows foot like I said in #14 reply.

And Larry, it was me who questioned your process not Redfred, so don't blame him for your discomfort of being questioned about it. You made the statement, "The problem is we have to explain to our customer as to why we had a problem and why we are going to change the way we are going to torque."

If your dealing with a Government project, I feel for you.

Explain to us, how you guys got so far along in your process practices that now your having to explain to your customer that your process is all F***ed up?

Where's your R&D and what are they saying?

R&D? How about the QA department.

Them too, but it shouldn't have got out of R&D to begin with. When I was working manufacturing QA, our R&D dept. made sure all their ducks were in a row before they ever brought to QA for their buy off.

Seems they have a failed process chain

Ok, now I feel really puzzled. Looking closely at the drawing, which is nice I might add, I can't help but wonder. If you can get a straight socket and torque wrench on it, why do you want to use a crows foot?

Homework?

Been dealing with "dumb" and "Dumber" all day, should have seen this right off. Must be getting old.

The difficult thing to comprehend at first pass, is that applied TORQUE on the torque wrench, pure torque that is, (and THE WRENCH IS MEASURING THAT, in that case), will be transfered UNCHANGED, even if the crowfoot is a mile long, as long as all elements are ON THE SAME PLANE (geometrical plane of course, perpendicular to the torque vector). I understand it is not that obvious. Please try to... decrypt my badly(?) written last post. S.M.

Something strikes me as odd -

If you are able to check the torque with a socket, why aren't you installing it with a socket?

D38999 connectors go into a hole in the shape of the letter D (a D hole) there is clearances between the hole and the connector. This allows the connector to twist when torqueing the jam nut. The connectors we use have pins that solder into a circuit board. Any twisting of the connector causes stress to the solder joints. During thermal or vibe this stress could cause solder joints to facture. There are tools that allow you to hold the connector while you torque the nut. This is the case here.

I think you have answered this question yourself. When you use a crow's foot, you are putting moments (such as shear) on your nut in addition to pure torque, and will therefore read higher than the pure moment that a on-axis reading using a socket will provide. Trust the torque wrench attached to a socket; it will provide a more accurate reading.

A lot of people here appear to have missed the concept that a flexible joint has been introduced to the lever by the connection to the crows foot.
In the ordinary torque wrench the body of the wrench experiences bending forces along its length; these are inconsequential because the body of the wrench is capable of far greater bending strength than the maximum torque of the wrench.
However once you place the end of that torque wrench into the crows socket, you create a flexible point at which the torque of the wrench will be loaded. The nut being tightened in whatever configuration will have a resistance to the tightening process, and that resistance opposes the torque being applied AT THE crow socket.
Therefore the crow will never tighten the nut to the indicated wrench setting, but always some value less.
Resistances are not restricted to the friction of the nut against the panel but also include the quality of the threads, and any retaining hardware in between.
Therefore it is difficult, perhaps impossible to consistently predict the ACTUAL torque that will remain after the wrench signals its satisfaction because each location and thread / nut pair will have its unique friction coefficients.

GA! I was wondering when someone was going to bring all that extra "play" into consideration. I've used flexible extensions, crow's feet, etc. and have always wondered how to accommodate all that "slop" into the readings.

It's also good to see ME's fighting over seemingly basic concepts such as torque, just like us EE's get twisted over things like reactive power!

Thanks ! If you consider some of the comments about the deformation of the nut, it can in some cases be a red herring. Deformation of the nut will absorb applied torque, but unless the nut itself is deformed around the threads, it does not really change the resistance to torque unless the threads themselves strip. If there is NOT sufficient torque to strip the nut external surfaces to the point at which they do not bind against the wrench surfaces ( external strip ), there will come a point at which the deformation stops and the expectations of torque applied conform to the previous post.

The "extra play" is not a factor in this case, or any other case using a torque wrench. The extra play is absorbed into the device and only the ultimate torque applied to the nut is measured. The wrench can be deformed to any degree, but that deformation is not registered as torque until the unit, as a whole, reaches the resistance dialed into the wrench.

The torque applied to the nut is different from the reading because the length of the arm is different. There is never any "extra play".

Of note:

It is a poor practice to use a crow foot to tighten a nut/bolt except when access is restricted. A crows foot, just as an open-end wrench only contacts the hex of the nut at two places, one at the inner most place of the slot and the other at the outer most place on the slot of the wrench. This can tend to deform the points of the hex of the nut/bolt. If a 6-point socket is used all points of the nut are in the fullest contact of each point of the hex that can be obtained. With a 12-point socket the contact area is reduced since the depth of the 6-points of contact are not as deep.

When this used to be critical we would cut a box-end wrench in two, cut an appropriate slot in the open end (6 or 12 point portion) of the socket to accommodate the handle of the wrench. The shortened and cut box-end wrench was then welded into the slot of the socket to create a crow's foot style hex socket extension. This way the torque wrench square post could be placed in the square hole of the socket and the hex was around the hex of the nut/bolt. An appropriate adjustment factor was used when making torque calculations.

This can also be made by welding the socket to the handle flat side of a striking box wrench.

Good Luck, Old Salt

It is possible to get too fixated with respect to the precision of torque measurements. Normally it is not torque but rather clamping force that is the desired measurement. The use of a torque wrench assumes a correlation between torque and clamping force. I think that the correlation is not close because of the variables discussed above. I have tested this correlation in large bolts and have concluded that the range is around 15%. This based on a 3" bolt, hydraulic torque wrench and a load cell.