Torque Talk

[Marvo]

Long time no see and I hope you're all doing well.

I see that torque screwdrivers have become the height of fashion in the UK over the last couple of years. I wondered what general practice has been adopted regarding their use.

Are electricians expected to;
Have a torque screwdriver on site at all times?
Check torque of all terminations in CU's / DB's they open or just any new terminations they make?
Check the torque of all terminations in any accessories they open?

Also;
Is acceptable to do torque testing live in a CU (assuming a VDE screwdriver)? Is a risk assessment required for live torque testing?
If torque testing is performed in a DB/CU do you just check screws for tightness or do you loosen off all terminations first then re-torque to the correct figure to eliminate overtightened connections as well as loose ones?

 

[Rockingit]

It's a bit like plumbing, you get a feel for it.

I use the same method, tighten up gently making sure no cross threading is taking place, then give it another tweak, and a final tweak before fixing in place.

Usually the third tweak produces nothing in the tightening procedure but gives peace of mind.

I wouldn't trust torque screwdrivers, you might think it's tight because it's torqued to the correct reading, but everything can be faulty.

Good old fashioned feeling can never be wrong, unless you get arthritis.

"When it's white, it's tight"...... (knuckle, for anyone not familiar with the saying!)

 

[SJD]

A related comment about stranded cables.

If you tighten a 6 or 10mm2 conductor as found in twin & earth, if you leave it a while, likely as the strands settle, it will become a bit loose. I'd expect someone used to working with these cables would be aware of that, and recheck tightness after (for example) 10 or 15 mins. But I often see loose connections in shower or cooker isolators, either on an EICR, or after one has failed, and suspect the connections were never checked properly, someone in too much of a hurry to finish.

 

[nicebutdim]

A related comment about stranded cables.

If you tighten a 6 or 10mm2 conductor as found in twin & earth, if you leave it a while, likely as the strands settle, it will become a bit loose. I'd expect someone used to working with these cables would be aware of that, and recheck tightness after (for example) 10 or 15 mins. But I often see loose connections in shower or cooker isolators, either on an EICR, or after one has failed, and suspect the connections were never checked properly, someone in too much of a hurry to finish.

I like to hasten the bedding of stranded conductors, by giving them a wiggle and retightening a few times until there's no more movement.

 

[davesparks]

Here's an example... tyre fitters.

They run up the wheel nuts on your very expensive alloy wheels... Then finish them with a torque wrench, set at the correct value.
This ensures that its not too loose, that the vibration of driving loosens them more, and they fall out... nor too tight, as they would damage your alloys.

Its the same type of forces on the screw terminals within a consumer unit. Not too loose, not too tight...You dont want the cable wiggling out... nor do you want the circuit breaker cracking apart because someone turned the driver too much.
Luckily the manufacturers helpfully give you the optimum values.

I would rather see a tyre fitter use a torque wrench on my wheels, that im going to drive away in.... rather than him doing it by 'feel' alone.... using a manual wrench, and him saying "Aye, that'll do"



Don't tell me.... you change your own tyres.....
As we tend to say around here: Get a professional to do it!

Wheel nuts and electrical terminals are different situations.

The torque value for the wheel nuts is set knowing exactly what is being fitted, what its made of etc etc.

Electrical terminals could have a range of types and sizes of conductor put in them, there may even be multiple conductors in there. I can't see how one universal torque setting can be correct for all sizes and types of conductor.

 

[davesparks]

A related comment about stranded cables.

If you tighten a 6 or 10mm2 conductor as found in twin & earth, if you leave it a while, likely as the strands settle, it will become a bit loose. I'd expect someone used to working with these cables would be aware of that, and recheck tightness after (for example) 10 or 15 mins. But I often see loose connections in shower or cooker isolators, either on an EICR, or after one has failed, and suspect the connections were never checked properly, someone in too much of a hurry to finish.

The same is true for any fixing which is being tightened to a specified torque setting, it should be torqued, allowed time to relax, and then re-torqued.

For example when tightening down a cylinder head on an engine it's normal practice to tighten to the specified torque, go and have a tea break and then re-torqued to that same setting.

 

[littlespark]

Electrical terminals could have a range of types and sizes of conductor put in them, there may even be multiple conductors in there. I can't see how one universal torque setting can be correct for all sizes and types of conductor.

So the torque values published by the manufacturers mean nothing?

I’m sure they’ve done their research and whatever value they state will cover any size, quantity or material of conductor.

 

[nicebutdim]

So the torque values published by the manufacturers mean nothing?

I’m sure they’ve done their research and whatever value they state will cover any size, quantity or material of conductor.

I don't see any claim that torque settings suppled by manufacturers mean nothing.

What has been stated, several times by people who are familiar with the phenomena, is that there are variables which may render those recommendations less than ideal. I'd be amazed if you were to tell me that you'd never had such an experience, that you never check how secure a termination is or that you'd never rechecked a termination and found that it was no longer at the required torque.

Edit: To clarify; my background is originally mechanical and involved regular torquing of components and assemblies which were subjected to fairly extreme pressures. The main variable in those situations were bolts and tie rods, which tended to stretch a little. When clamping several, soft strands of copper it comes as no surprise to me that torquing once and walking away is not an option and that clamping soft copper by various different means may yield variable results.

 

[Dustydazzler]

I'm with DS on this one when it come s to torquing up and there is an element of twist back / slackening when you first torque up something, and it is good practice to re-torque especially with bigger cores

 

[davesparks]

So the torque values published by the manufacturers mean nothing?

No, and that's not what I said!

I said I can't see that one universal torque value could be right for all possible situations.

I’m sure they’ve done their research and whatever value they state will cover any size, quantity or material of conductor.

I doubt that it will, and I suspect that if you asked the manufacturer for further information on this you may find that those torque values are intended for a single copper conductor not doubled back or a fine stranded conductor fitted with a ferrule.

 

[SparkySy]

25mm tails are a prime example of the torque, tea break, torque, tug test and torque again scenario!
Manufacturer's should have tested all manor of types and sizes of conductor in their device, weather that is the case or not, the entry position and any pressure or bends in the cable will to some degree all have an effect on the final termination, especially with larger cables.
I would rather have some idea of how tight it should be than either trashing the device / terminal or coming loose.
As with most of our work a lot of it is still down to knowledge and experience. If something doesn't feel tight at the required manufacturers setting would you give it a bit more or say that is it that's how tight it should be so f-it?

 

[Dustydazzler]

Perhaps there needs to be several ranges based on cable type , bare , ferruled , bent , multi or single etc

 

[davesparks]

Perhaps there needs to be several ranges based on cable type , bare , ferruled , bent , multi or single etc

That won't happen, if pushed on the issue manufacturers will start declaring that terminals are only suitable for one conductor in their instructions.

We'd end up like the USA where they are not allowed to put more than one conductor in a terminal unless the manufacturer specifically states it is OK.

 

[slartybartfast]

Wheel nuts and electrical terminals are different situations.

The torque value for the wheel nuts is set knowing exactly what is being fitted, what its made of etc etc.

Electrical terminals could have a range of types and sizes of conductor put in them, there may even be multiple conductors in there. I can't see how one universal torque setting can be correct for all sizes and types of conductor.

Spot on.

Also wheel nuts are subject to significant forces and vibration making torque settings important. Whereas Domestic Electrical installations are (generally) not subject to these forces & vibrations, and so terminal screws, if nipped up, are highly unlikely to work loose.

I guess the intent to using torque screwdrivers is a catch all to ensure that "nipping up" is done, which explains the vagueness in the regs and manufacturer instructions.

You simply cannot have a universal torque setting that caters for all cable types.

 

[Dustydazzler]

Also to mention some electricians torque screwdrivers only range from 1-3 NM like the Wera , which is a bit naff from a £175 tool when some makes of main switches ask for 3.2-3.4NM

 

[slartybartfast]

Also to mention some electricians torque screwdrivers only range from 1-3 NM like the Wera , which is a bit naff from a £175 tool when some makes of main switches ask for 3.2-3.4NM

There are plenty of mini-torque wrenches out that that can take screwdriver bits, but fully accept space constraints in some cases.

 

[Dustydazzler]

I used to have a torque wrench , I be danged if I know what happened to it thou

One a bit like this

https://www.amazon.co.uk/Draper-30357-Square-Drive-22-1-154-9lb-ft/dp/B0001K9S52/ref=asc_df_B0001K9S52?tag=bingshoppinga-21&linkCode=df0&hvadid=79920808416455&hvnetw=o&hvqmt=e&hvbmt=be&hvdev=c&hvlocint=&hvlocphy=&hvtargid=pla-4583520383274666&psc=1

 

[pc1966]

I think most of the key points have been said.

Terminals should be correctly tightened but the regs don't specify how that is done, however, nobody can argue that using the manufacturer's torque setting is not a good way of doing so and thus meeting the fundamental requirement of a sound connection. To avoid liability then proving you have the tool to do so than comes along.

As an aside, wheel nuts and terminals both have forces making them come apart, in the case of electrical stuff though it is usually thermal cycling and not heavy vibration. What is very different is the general condition of them with terminals generally clean & dry and wheel nut varying a lot in terms of surface condition (rust through to well-greased).

Tightening by torque is not that precise as while the tool may be accurate to a few percent, the thread and fastener head friction can vary a lot depending on its finish (machining / plating quality, etc) and state of lubrication. From memory to around 30% or so. That is why high stress bolts like car cylinder heads, etc, now have other methods of final tightening such as going to a set torque to bed them down and then a given final turn angle so the bolt/stud has a certain degree of elongation and thus preload.

Another approach that is good is to observe the torque versus angle do determine how much yield there is, oddly enough that is basically the traditional craftsman's method found out by years of experience (and often a couple of failures along the way)! More summary here:

Methods of Tightening Threaded Fasteners

Personally I use a torque tool for anything critical, but for small DIN screws, etc, usually just hand feel of experience. I have a VDE screwdriver for doing MCBs, etc, and also a few non-insulated tools left over from DIY on car's etc over my lifetime that get used for the occasional big lug bolts, etc.

In my own very limited experience, Hager, Wylex, and Schneider stuff survived fine at their stated torque, one CPN device started to come apart at the seams so lesson learned - don't use them or take it slightly easier!

 

[nicebutdim]

Spot on.

Also wheel nuts are subject to significant forces and vibration making torque settings important. Whereas Domestic Electrical installations are (generally) not subject to these forces & vibrations, and so terminal screws, if nipped up, are highly unlikely to work loose.

Screw terminals are subject to thermal creep, through which they can become loose, although that's not going to be much of an issue at the time of installation.

 

[slartybartfast]

Screw terminals are subject to thermal creep, through which they can become loose, although that's not going to be much of an issue at the time of installation.

hmmm....if temperatures are getting high enough to that then cable is likely undersized

 

[DPG]

hmmm....if temperatures are getting high enough to that then cable is likely undersized

Happens at moderate temperatures, especially when stranded cables have been tinned with solder. That is why they shouldn't be tinned.

Doesn't necessarily need such high temperatures for it occur.

 

[slartybartfast]

Happens at moderate temperatures, especially when stranded cables have been tinned with solder. That is why they shouldn't be tinned.

Doesn't necessarily need such high temperatures for it occur.

Ferrules on stranded cables will be mandated in regs before long for sure !

 

[DPG]

Ferrules on stranded cables will be mandated in regs before long for sure !

I think they are mentioned in the regs for fine stranded wires aren't they?

 

[pc1966]

I think they are mentioned in the regs for fine stranded wires aren't they?

Not exactly, but 526.9.1 says the ends have to be "suitably treated" and ferrules are one way to do that. The other common way is the ultrasonic weld you see on some RCBO neutral tails that are fine-stranded.

Also that section (526.9.2/3) say no to tinning conductors for screw terminals, etc.

Soldering is great though if it is actually the means of connection!

 

[pc1966]

hmmm....if temperatures are getting high enough to that then cable is likely undersized

It is more the cycling of hot/cold you get with time-varying loads or, to a lesser degree, just from day/night temperature changes, etc.

 

[Rockingit]

Not exactly, but 526.9.1 says the ends have to be "suitably treated" and ferrules are one way to do that. The other common way is the ultrasonic weld you see on some RCBO neutral tails that are fine-stranded.

Also that section (526.9.2/3) say no to tinning conductors for screw terminals, etc.

Soldering is great though if it is actually the means of connection!

Indeed, but who in their right mind gets out the soldering iron station when it takes 2s to put a ferrule on?!

 

[Lucien Nunes]

In the days before ferrules were widely known outside the panel-building industry, it was quite a popular albeit misguided pastime. For example, we saw a number of small theatre lighting installations with a plug patch at the dimmers, where the flexible patch cords were tinned before connecting to the conduit cables via normal terminal block. Someone obviously tinned their way along the line of them making each and every connection a loose one.

 

[pc1966]

As above, was quite common and obviously enough for it to be added to the regs as a no-no.

Ferrules are quite a neat way of helping fine stranded wire, and occasionally I have put them on coarse stranded if it is a big screw terminal instead of cage-clamp, etc, as I have seen wires removed and looking worryingly tin and flat as a result.

 

[Other_Tucker]

I'm all for the correct tightness in theory, it winds me up when colleagues use impact drivers to tighten (permanently weld) screw connections in CU's. I'd bet the correct, recommended torque, is less than normal wrist driver tightness.

If it becomes mandatory bureaucracy, how long until you need a calibration certificate for your torque screwdriver.

 

[Rockingit]

As above, was quite common and obviously enough for it to be added to the regs as a no-no.

Ferrules are quite a neat way of helping fine stranded wire, and occasionally I have put them on coarse stranded if it is a big screw terminal instead of cage-clamp, etc, as I have seen wires removed and looking worryingly tin and flat as a result.

You only have to fix one and then take it out again to see where the screw clamped down and all the strands it missed!

 

[DPG]

I'm all for the correct tightness in theory, it winds me up when colleagues use impact drivers to tighten (permanently weld) screw connections in CU's. I'd bet the correct, recommended torque, is less than normal wrist driver tightness.

If it becomes mandatory bureaucracy, how long until you need a calibration certificate for your torque screwdriver.

You should have it calibrated regularly already.