Applying torque

When tightening a nut or bolt to a specific torque, should the thread be lubricated?

If so should one use grease or oil?

If it makes a difference from a dry situation, what sort of difference one should expect?

What sort of accuracy should one expect from the typical torque wrench purchased over the counter? (yes I know, how long is a piece of string? But I expect that you get the idea)

I ask because it has been suggested to me that lubrication causes up to 40% error & a typical wrench 15- 20% inaccuracy. The 2 combined could make a mockery of torque figures if they were meant to be dry fastened.

But if so how would one control friction in the thread?

The torque specification should state if the thread is lubricated or not.

Your other comments are why the actual bolt stretch is measured for critical applications - where possible - like bearing caps, and angular tightening has become common

Edited By Nicholas Wheeler 1 on 23/11/2022 07:59:46

Both statements are true. Lubrication of threads makes a huge difference. Not just oil, different platings have an effect.
There are two common reasons for stating torque figures:
To prevent damage
To provide pre-load on an assembly, often to reduce fatique failures.

The first case is less critical. Measuring a angular movment of the fastener after the joint is clamped (rough torque setting) is a much more accurate means of setting pre-load and is commonly used in automotive applications.
As Nicholas says measuring fastener length is used for some applications and is even more accurate.

At work our recommendations regarding torque were that the figures given were for natural finish or lightly oiled parts and for joints using nuts, bolts or studs only. When a lubricant such as molybdenum or graphite is used the torque figure should be reduced by 20%.

With heavily galvanized parts the zinc could squeeze out over time so hot dip galvanized fasteners were never used for critical bolting requirements.

Martin C

I have been putting motorcycle engines together for many years with common store-bought torque wrenches, with the threads on the fasteners always having either copper slip or Loctite of varying grades on them and haven't had a failure yet. I don't think it's as critical as all that. Main thing is that all the nuts are tightened down the same. If they are 20 per cent over due to lubrication, it does not seem to have any detrimental effect. Never stripped one yet, or had one fail.

As for calibration of torque wrenches, aircraft engineers do it. I have never bothered and never had a problem in the home workshop. The plain old beam type torque wrench with a direct reading pointer and scale is most reliable as there are no moving parts to jam or wear etc. But I have never had any problem with my "click" type torque wrenches either, which are used at home for hobby use. Might be a different story if you were using them all day every day at work etc.

I posted comparative numbers for the various methods on the forum a while back, and now I can't find the post.

From memory, although torque wrenches are considerably better than man-with-spanner, they're only about ±30% accurate. That's good enough for most ordinary torquing, but not high-tech applications. Accuracy matters when a posh bolt is highly torqued close to yield point for maximum holding.

When a manual says replace undone bolts with new ones, it may be because the old bolt was deliberately torqued close to the limit and isn't trustworthy any more.

Dave

The torque to be applied, should specify wherher lubricated or not..

Toyota specified "75 lb ft, DRY" for the zinc plated wheel bolts on the Yaris.

At Perkins when we went to yield tightening, we specified lubrication with soluble oil.

We found that anything between 5 and 20 % produced no different results, on 1/2 UNF W range steel bolts into cast iron.

In general, we found that lubrication reduced the spread of torque required to produce the same clamping force, (Which is why torques are specified, with the object of delivering a required level of clamping )

Howard

Interesting Excel spreadsheet for calculating bolt stretch and clamping force can be down loaded here :-

**LINK**

https://www.securitylocknut.com/how-much-does-a-bolt-stretch/

Slightly scary how much change the differential expansion rate of an aluminium engine block and head relative to steel bolt or studs makes to clamp loads and how close you can get to yield point. Around 30% difference between hot and cold for my Rover V8. No wonder they went to stretch bolts.

Current torque wrench standard is BS EN ISO 6789:2017, basically ± 4 % doing up, ± 6 % undoing over 20% to 80% of nominal range for normal right hand thread. Most half decent ones easily meet that mechanically but accuracy of scale setting and operator skill have considerable effect on what you get. Same guy doing same job is generally surprisingly repeatable. Lots of different guys ..... um .....

Good introduction to the engineering behind bolted joints here :-

**LINK**

https://www.fastenal.com/content/feds/pdf/Article%20-%20Bolted%20Joint%20Design.pdf

Clive

Edited By Clive Foster on 23/11/2022 12:07:16

In aerospace it is normal to check torque wrenches against a calibrated gauge e.g. Acratork, before use.

I personally have a small acratork at home and use it to check my small (5/18" 8mm ~1N-m) preset torque wrenches used for precision coaxial connectors on RF test equipment. Using a torque wrench here produces repeatable measurments and prevents damage. Even if the connector is field replaceable they are expensive. The one on my 26.5 GHz spectrum analyser is about £500.

Torque setting a fastener is actually a poor way of getting an accurate load due to variations in friction between threads and abutment faces. If you want accurate use a tightening angle, heated bolts, stretch bolts etc.

To answer the OP torque is usually quoted for dry fasteners, but there are exceptions.

Martin

Torque is the least accurate measured method of tightening fasteners, but is unavoidable in most cases. One thing that must me adhered to is that the fastener must be torqued up the final 20% in one movement. If you stop even slightly short, possibly due to difficulties with swinging the torque wrench, the torque will not be correct. As soom as the tightening is interrupted, the subsequent torque needed to start things moving again will be much higher, this is called stiction. Backing off and re applying the torque without a pause right up to the full torque is the answer. Special requirements for stretch bolts must be adhered to exactly.

Whatever ... DON'T lubricate the wheel-nuts on cars. Quite apart from its effect on the stud, you will be in grave danger of cracking the wheel where the nut seats.

As has been stated several times, torque is pretty inaccurate, a seating torque followed by angular increases is far better and as others have said there are other more accurate ways of doing it.

In the general and regular automotive world, on a spread of fasteners that have the same tightening spec, the fact that they are all the same is more desirable. The accuracy of the torque wrench isn't that much of an issue as long as it's used and they are the all same and close to the spec.

And just to highlight what Old Mart said regarding the final tightening in 1 movement is a must. Getting close then having another bite at it so to speak will usually give a 'false' tight click or beep as the breakaway force required to make the fastener rotate again can be significantly higher than the required setting, especially on something like a flange nut that has a large surface area against the static face.

Personally, all the performance / race bike engines I've built / prep'd I'm with Hopper, generally loctite or a very small amount of light oil to prevent any galling or binding on pre cleaned fasteners and threads.

And yes, if it says replace the fastener then you really should as they will be 'stretch' bolts.

Can you expand on that? Is it because a lubricated thread can increase the force applied for a given torque value, and put too much force on the wheel?

Edited By John Doe 2 on 24/11/2022 09:15:22

Had not occurred to me. Thanks

I always put a dab of coppaslip on wheelnuts, and have never had a problem! Cracking of the metal behind the wheelnut is caused by over tightening, usually by idiots at tyre centres who fall asleep with the windygun trigger pressed! I watch these people as they hammer each nut tight and then pick up a torque wrench and go round the nuts to make sure it clicks. This is completely wrong, the torque settings on both steel and aluminium wheels are NOT a minimum, and the torque wrench will click even though the wheel has been tightened way beyond the specified torque! If you look at a steel wheel you will see that between the wheelnuts there is a pressing which faces the drum/disc, and you will see on the wheel and the drum/disc telltale marks where these pads contact. the metal directly behind the wheelnuts MUST NOT contact the drum/disc or the wheels will work loose and the metal will crack! The wheelnut seats in the pressed taper on the wheel and the metal below it acts as a spring in compression to keep the nut tight, but if the wheel is over tightened this spring is crushed back to the drum/disc,, and the wheel comes loose, and cracks develop as the wheel "works" around the wheelstuds. It is not caused by lubrication, it is caused by gross over tightening! There is more to a wheel design than most people realise!

Phil

In the 1970s There were instances of Bedford CF ambulance wheels coming of whilst leaving the wheel nuts in place. Two explanations were offered one was metal fatigue due to over flexing when cornering, the other was put down to overtightening of the wheel nuts when changing tyres.

One case where lubrication can be critical is when using stainless steel fasteners. Clean stainless parts will gall - the threads will weld together under fairly gentle tightening. I have a vague memory that other materials also show this effect. Just one more thing not to forget ...

Cheers, Tim

I guess both! Overtightening wheel nuts is a common pastime, and doing it to a vehicle that does high-speed cornering is begging to open up micro-cracks.

Anyone know about ambulance performance? Possibly an urban myth but I was told at school they were amongst the highest performing vehicles on UK roads. Were they souped up or just a big engine in a lightly loaded van? Judging by the way we were left at the lights by one, they certainly had much better acceleration than my mate's trusty Morris Marina!

Dave

My recollection is that ambulances were (and likely still are) exceptionally heavy. All that resussitating kit, etc.

I suspect the extra weight had more to do with wheel failures than extra engine power.

Cheers, Tim