Proper way to retorque a connection

[fredelect]

What is the best way to check an electrical connection for proper torque tightness? Should I just back it off and retorque or cut and redo?

 

[Jraef]

Good question, I'd be interested in hearing too. When I was an apprentice I was taught to back off and re-tighten. But later I was told to NEVER back off a bolt, just put on the torque wrench and if it clicks, you're done. I've never been comfortable with the inconsistency.

 

[gar]

110202-1531 EST

My comment is not an answer to your question.

1. Most materials have a static friction coefficient that is higher and different than the sliding friction coefficient.

2. When tightening screws, nuts, or bolts one really is not interested in the torque. What you really want to know is the tension in the fastener. But, in most cases there is no easy economical means to measure this tension. Thus, torque is used as an indirect method to measure tension. But this is dependent upon the coefficient of friction. The coefficient of friction can be highly variable, and therefore when torque is used to measure tension the accuracy of the measurement is dependent upon this variation.

3. On automotive axles the pinion nut was previously made with a cadmium plating. This coefficient was reasonably constant. Cadmium has been banned. The new coatings are not as consistent and thus in some cases torque limits have been widened. In other applications the tension variation will be greater.

4. If we assume that the tightening specification for torque is based upon a measurement while the fastener is in sliding friction, then just using a clicker type torque wrench, and not rotating the fastener before the click, will provide erroneous information.

I would favor using a dial type wrench and actually initiating slight rotation and use the reading just after breakaway.

Just changing the speed of rotation of the fastener may change the sliding coefficient of friction.

5. Any change in lubrication of the fastener components changes the coefficient of friction. Never change the lubrication on any fastener component from that which was used to determine the torque specification.

Some time in past history someone at Cadillac decided to put grease on the wheel lug nuts. Since constant torque drivers, impact wrenches, were used to install the nuts the tension in the studs was much greater than the design values and studs were breaking off. This was unlikely someone from engineering, but rather probably a line worker idea.

You need to go back to the original source of a torque specification and get their definition of how the measurement should be made. And keep in mind they may be wrong because they do not really understand the fastening system.

I am not suggesting a method, just some comments.

.

 

[ptonsparky]

If these guys have questions about the process...

 

[dbuckley]

I'm also of the "never back off a bolt" school.

My preferred approach is when the bolt is initially tightened and torqued, use a sharpie and put a line from the bolt to something stationary. Then visual inspection is all that is required to know that the bolt (or nut!) hasn't undone itself.

 

[zbang]

Adding to gar's excellent notes:

My school (of thought) says if you think you have to retorque a bolt*, in most cases you should probably back it off and replace it, then bring the new one up to the proper tension/torque. Otherwise see if it's actually loose, then leave it alone. If it is loose, find out why... Because of stiction, all a torque wrench will tell you is that the breakaway torque is higher than the wrench's setting. Not very useful.

* applies to steel, other materials behave differently. Since brass and bronze can stretch, you actually might have to re-set the tension. This is also what belleville washers are for. Also does not apply to cable connectors, those can easily loosen on their own.

In any case, see the manufacturer's recommendations.

 

[wptski]

There is a practice that's been around for years that I never heard of that's used in some automotive assembly operations. A certain amount of torque is applied to a fastner and then a addtional amount measured in degrees of a turn. Engine head bolts is one place where this is used.

One can purchase an adapter for this that can be used with any standard torque wrench or a torque wrench with that feature included.

 

[fredelect]

I am not engineer but understand a little bit of what you gentlemen are saying. As far as electrical connections are concerned....

It is odvious if the connection has burned we cut back the wire and even replace the lug if needed. I notice when a conductor has been tightened, not necessarly torqued, as I'll bet most electricans do not use a torque wrench, the connector leaves a "dent" in the conductor. I think this would leave a week spot in the wire. Just as a torqued bolt could stretch a little making it weak. It has always been my thought, correct or not, that a properly torqued connection will not loosen up. Just going around and tightening even with a torqueing dosen't seen right as the connection maybe too tight already, squeezing the wire too much and streaching the fasenter too much. Typicaly, I go around and just tighten connections and find many loose but that is not using a torque driver. maybe the best process is to use a torque driver when I di this and call it "the best I can do."

 

[LarryFine]

My preferred approach is when the bolt is initially tightened and torqued, use a sharpie and put a line from the bolt to something stationary. Then visual inspection is all that is required to know that the bolt (or nut!) hasn't undone itself.

And, if the conductor is loose and the bolt has not turned, does it just stay loose?

 

[suemarkp]

I'm not so sure all these automotive analogies apply to torquing lugs for wire. You're using the bolt to crush the wire -- it isn't under tension but compression, and you're not really compressing the steel but the conductor. But thread friction probably is going to have an effect too, but hopefully no one is oiling lugs threads.

I have noticed most connections to copper wire that I do with a torque wrench don't loosen up. If I did a #6 on a breaker with just my hand, I'll often find later that I can tighten it a bit more (to probably didn't do it tight enough the first time). I've never used a torque wrench on smaller wires -- just the larger Service ones where it isn't so easy to go back and retorque.

I wonder if aluminum -vs- copper conductors also behave differently to torquing?

 

[dbuckley]

And, if the conductor is loose and the bolt has not turned, does it just stay loose?

It would. Thus marking the bolt position is only useful when you know the arrangement is correctly torqued. This can preferably be done either at installation, or some time later when it gets re-torqued. re-torquing = bad, thus doing it once and then marking up saves future re-torquing.

 

[wptski]

This topic has come up before and although I don't recall which branch of military service they were in but the member stated that retorquing was part of routine maintanence.

 

[cschmid]

I myself am a loosen and re-tighten type of man and I also like the dial better then the clicker as the clicker is less consistent. now just because the bolt is torqued properly does not mean it is not seized in place. So I believe it is loosen and re-tighten and for me and I only use torque wrench on aluminum wires and incoming wires and bolts on cabinet equipment. wires under screws are just tight as I feel necessary with screw driver. I normally use a clicker as well because it is more compact and travel better in its nice fancy little plastic case. I also agree just because it is torqued to a specific does mean it is the correct tension on wire.

 

[weressl]

I am not engineer but understand a little bit of what you gentlemen are saying. As far as electrical connections are concerned....

It is odvious if the connection has burned we cut back the wire and even replace the lug if needed. I notice when a conductor has been tightened, not necessarly torqued, as I'll bet most electricans do not use a torque wrench, the connector leaves a "dent" in the conductor. I think this would leave a week spot in the wire. Just as a torqued bolt could stretch a little making it weak. It has always been my thought, correct or not, that a properly torqued connection will not loosen up. Just going around and tightening even with a torqueing dosen't seen right as the connection maybe too tight already, squeezing the wire too much and streaching the fasenter too much. Typicaly, I go around and just tighten connections and find many loose but that is not using a torque driver. maybe the best process is to use a torque driver when I di this and call it "the best I can do."

Electrical connections are subject to continuous thermal cycling as current/heat goes up and down. Since all materials expand/contract with temperature changes, electrical connections are designed to compensate for it. In some cases the material itself is capable to 'follow' and 'spring back' to its original form like bronze and other copper alloys do but as it was found out aluminum does that poorly. Belleville washers were introduced as the device that can provide the longest compensation capability, but of course it is necessary that the proper initial tesion is provided, hence torqueing requirement. Simple spring washers can also provide the same tensioning, returning tension, but depending on the material and the temperature differences and the number of cycles it is subjected to (10^n) it is subjected to, it will eventualy loose its modulus and the heat runaway occurs. The method is called thermal compensation.

 

[izak]

on larger aluminum lugs -2/0 and up- I really like to back them most of the way out and put a little oxide inhibitor on the threads. this makes them Much easier to back out later, and it doesnt feel like the aluminum threads are galling themselves when tightened.

I love finding the older meter cans around here with the two screw lay-in lugs - they Always have NoAlOx on them and they are a breeze to remove and reinstall

 

[don_resqcapt19]

on larger aluminum lugs -2/0 and up- I really like to back them most of the way out and put a little oxide inhibitor on the threads. this makes them Much easier to back out later, and it doesnt feel like the aluminum threads are galling themselves when tightened.

I love finding the older meter cans around here with the two screw lay-in lugs - they Always have NoAlOx on them and they are a breeze to remove and reinstall

Adding the oxide inhibitor where it was not installed by the lug manufacturer will increase the pressure on the conductor if you tighten the set screw to the manufacturer's listed torque. This may result in a poor connection. Too much pressure on the conductor is just as bad as too little in terms of the long time performance of the termination.

 

[weressl]

Adding the oxide inhibitor where it was not installed by the lug manufacturer will increase the pressure on the conductor if you tighten the set screw to the manufacturer's listed torque. This may result in a poor connection. Too much pressure on the conductor is just as bad as too little in terms of the long time performance of the termination.

The pressure increase is hardly perceptible when compared to the accuracy of the torgue setting. The oxide inhibitor will decrease the exposed voids' exposure to atmospheric oxygen and other oxydants and will decrease the natural oxidation that has higher resistance than the metal itself. Indeed too much pressure is bad, no argument there.

 

[don_resqcapt19]

Laszlo,
If the oxide inhibitor is placed on the threads of the set screw, (my understanding of what the poster was doing), it will act as a lubricant and will increase the crushing pressure on the conductors, assuming that the manufacturers torque value was based on dry threads. I would think that in the absence of factory applied lubricant to the set screw threads that the torque value would be based on dry threads. The same amount of torque applied to lubricated threads would result in increased crushing pressure on the conductors.

 

[weressl]

Laszlo,
If the oxide inhibitor is placed on the threads of the set screw, (my understanding of what the poster was doing), it will act as a lubricant and will increase the crushing pressure on the conductors, assuming that the manufacturers torque value was based on dry threads. I would think that in the absence of factory applied lubricant to the set screw threads that the torque value would be based on dry threads. The same amount of torque applied to lubricated threads would result in increased crushing pressure on the conductors.

Don,

I follow your logic and it is clear how you get to the conclusion. However I think that there may be some clarification needed in terms of physics. Lubrication aids the friction resistance that results in less torque need WHILE the thread is in motion that disappears when the stationary position is reached, eg. when the torque setting 'clicks off' and results in the remaining pure static pressure. So while moving the bolt you are working against friction resistance and modulus created static pressure AFTER the two surfaces come in contact and at the resting point only the static pressure remains.

In the case of extremely heat damaged threads the friction resistance could even be so great that it exceeds the recommended torque setting, yet the contacting surfaces be apart or being under less than required static pressure. (This would be a re-torquing issue, although the heat damage would most likely be visible.)

 

[Cow]

Laszlo,
If the oxide inhibitor is placed on the threads of the set screw, (my understanding of what the poster was doing), it will act as a lubricant and will increase the crushing pressure on the conductors, assuming that the manufacturers torque value was based on dry threads. I would think that in the absence of factory applied lubricant to the set screw threads that the torque value would be based on dry threads. The same amount of torque applied to lubricated threads would result in increased crushing pressure on the conductors.

That makes sense.

Though, in the grand scheme of things I'd rather take my chances with potentially overtightening a connection due to Noalox on the threads than have it sieze a few years down the road when I need to take it apart again.

I always try to lube set screws when I'm in an environment where I've had trouble with siezing before.