Jump to content

Greases, Friction Factors & Torque


Fatboy

Recommended Posts

Ladies & Gents,

We've just done some trials at work where we took various thread compounds and made up some connections to 30klbf.ft to see how they behaved. We made some very interesting observations and I thought it may be worth starting off a bit of a debate here:

1. In most cases, torque is a means to an end. What you are actually looking for is clamping force. In critical cases, it is better to measure elongation of bolts to verify this force but this is often impractical or "over the top" therefore torque in the nut/bolt is measured instead.

If you imagine a bad case, you could be misled by a high torque into thinking you have actually "clamped" a gasket but have actually expended the energy into damaged threads. The gasket will then possibly leak.

2. Greases have different friction factors and in critical cases, this should be taken into account when making fastners up. In most cases, you want a friction factor of 1.0. That will mean that if you need to apply a certain force which equates to 30klbf.ft on the bolt, you only have to apply that torque.

If you use a grease with a FF of 0.5, you will get the clamping force you need with half the torque on the bolt. The danger is that if you go to the "original" torque of 30lbf.ft with a 0.5 FF, you may well yield the fastner or damage the equipment as you will double the clamping force.

3. Applying grease to the contact faces of the fastner / washer makes a significant difference to the energy which gets to the thread.

4. The break-out force on a torqued fastener bears little resemblance to the make-up torque, especially if it was made-up a while ago. ie: If you undo the bolts holding the calipers on and it takes 100lbf.ft to do it, it does not mean that you should apply that figure to re-attain the same contact force. Often, break-out force bears no relevance.

I am sure there will be wiser comments from cleverer peeps than me but that is generally why debates are so valuable :D

Link to comment
Share on other sites

Mark,

Koprkote was actually very good and to be honest, the best of the products we tested with a FF which we believe to be 1.0 (This was our control compound)..

We saw a FF variation from 0.6 to 1.0 which is fairly significant on some of the new products.

One of the drivers for our tests were to check on the more recent "environmentally friendly" compounds available. Historically, Copper, Lead, Zinc etc have been used to provide bearing capacity to prevent galling of the threads. However, discharge of these metals are not good for the environement.

Hydrocarbon based greases are also not the most friendly... so you will notice more use of graphite and water based bases which although they can do the job, are not as good as some of the more established products.

Link to comment
Share on other sites

Thinking about it there have been times when using copper slipped bolts that I have felt I was over tightening the joint before the required torque was reached. Maybe should start applying a torque reduction factor of say 25-30% to compensate.

Link to comment
Share on other sites

Mark,

Re-reading my last post, it was misleading so I've edited it. As far as we could tell, Koprkote has a friction factor of 1.0. It has been used for years so was our control compound that we compared against..

If you feel that you have gone to far with a bolt, you may have either begun to yield the shank to strip the threads.. That is usually the point that I develop a cold sweat and dread. "It will be okay" I lie to myself and retreat praying. In reality, that is probably the best time to try and get it out as the holding force is similar to make-up. If you leave it, it gets harder to remove as it freezes in.... :(

Link to comment
Share on other sites

Thinking about it there have been times when using copper slipped bolts that I have felt I was over tightening the joint before the required torque was reached. Maybe should start applying a torque reduction factor of say 25-30% to compensate.

I have found this a problem to with the racers with the wheel nuts being massively overtightened due to having copper slip on them.

Link to comment
Share on other sites

I wonder if it is one reason why they say wheel studs should not be lubricated. I give a squirt of WD40 on mine, not that keen on greasing them as some people do...

Other threads get either copper grease or graphite grease.

Link to comment
Share on other sites

... In most cases, you want a friction factor of 1.0. That will mean that if you need to apply a certain force which equates to 30klbf.ft on the bolt, you only have to apply that torque...

I agree with most of your observations, but find your discussion of friction factor for lubricants is misleading, and I suspect totally wrong. How do you define friction factor?

By definition the value for coefficient of friction can't exceed 1 - heated soft compound tyres can produce results that indicate higher values, but is more to do with adhesion combined with friction. For brake/clutch linings it is generally about 0.5 or less. For lubricated steel parts in contact, it is generally less than 0.1

Tabulated values for bolt tightening torques are based on lightly oiled threads (normally as supplied). In this case, a good rule of thumb for calculating the tightening torque is Torque = 0.2 x bolt dia x pre-tension.

If bolt dia is in mm and required pre-tension is in kilo Newtons, torque will be Newton metres. If bolt dia is in inches and pre-tension is lbf, the torque will be lbf inch.

Link to comment
Share on other sites

I wonder if it is one reason why they say wheel studs should not be lubricated. I give a squirt of WD40 on mine, not that keen on greasing them as some people do...

Other threads get either copper grease or graphite grease.

I've often heard it quoted, yet never seen a convincing reason not to lube wheel studs.

FWIW, I've done it seemingly forever and never had a problem, on both road and race cars. (always using a tension wrench )

Rod and main bolts are generally oiled then tensioned and they never come loose. (or at least shouldn't ;) )

On this theme, I had a report yeras ago compiled by Ford Racing in the US on rod bolt tension and fatigue cycles to failure. The sample groups were to correct stretch, under-tensioned and over (plastic deformation ahieved) tensioned. The under-tensioned group failed on average consistantly and significantly earlier than the other two groups, who, IIRC, were remakably close.

Care to weigh in John ?

Link to comment
Share on other sites

By definition the value for coefficient of friction can't exceed 1

Surely it can if (as in this case) 1.0 is actually a semi-arbitrary lubricated reference (ie. all greases will lubricate the bolt and therefore reduce resistance to turning, but some may reduce it less than the reference grease)?

Link to comment
Share on other sites

I always thought the coefficient of friction was the ratio between the force required to move an object at constant speed and the normal force between the two surfaces.

F= uN

F = Force required to move object

u = (mu) Coefficient of friction

N = Normal (perpendicular) force between the two planes

I don't see how 1 or any other number can be a maximum value for mu

mu will depend on the two surfaces and any lubricant applied.

Link to comment
Share on other sites

I always thought the coefficient of friction was the ratio between the force required to move an object at constant speed and the normal force between the two surfaces.

F= uN

F = Force required to move object

u = (mu) Coefficient of friction

N = Normal (perpendicular) force between the two planes

I don't see how 1 or any other number can be a maximum value for mu

mu will depend on the two surfaces and any lubricant applied.

Sorry, should have read your post more carefully - I was talking about the 'friction factor' discussed by Fatboy, not coefficient of friction at all (which I don't pretend to know anything about... :( ). You were in fact stating what 'friction factor' couldn't be...so I was talking completely at cross purposes :rolleyes:

Link to comment
Share on other sites

I always thought the coefficient of friction was the ratio between the force required to move an object at constant speed and the normal force between the two surfaces.

F= uN

F = Force required to move object

u = (mu) Coefficient of friction

N = Normal (perpendicular) force between the two planes

I don't see how 1 or any other number can be a maximum value for mu

mu will depend on the two surfaces and any lubricant applied.

Yes, that is the coefficient of kinetic friction.

The coefficient of static friction is the ratio of the maximum force, when motion is impending, and the normal force between the two surfaces.

As for my statement about the coefficient of friction having a limit of 1. It is something that I remember from my student days (long ago). Thinking about it just now, like you I don't see how it is valid.

However, there is no friction material that I am aware of, which has a published coefficient of friction greater than 1 (rubber on concrete is close).

Link to comment
Share on other sites

I've often heard it quoted, yet never seen a convincing reason not to lube wheel studs.

FWIW, I've done it seemingly forever and never had a problem, on both road and race cars. (always using a tension wrench )

Rod and main bolts are generally oiled then tensioned and they never come loose. (or at least shouldn't ;) )

On this theme, I had a report yeras ago compiled by Ford Racing in the US on rod bolt tension and fatigue cycles to failure. The sample groups were to correct stretch, under-tensioned and over (plastic deformation ahieved) tensioned. The under-tensioned group failed on average consistantly and significantly earlier than the other two groups, who, IIRC, were remakably close.

Care to weigh in John ?

I always lubricate wheel studs, but I don't have alloy wheels.

Neglecting mechanical locking methods, friction is relied upon to prevent nuts and screws/bolts from working loose.

Some use inserts or distorted threads to increase the friction, when the tension in the screw/bolt/stud is to low.

Steel wheels and nuts use mating conical surfaces to increase the friction between the wheel and the nut, so don't depend so much on friction between the mating threads. To my knowledge, alloy wheels don't have the conical surface, so thread friction is more important

Usually, when bolts subject to cyclic loads (conrod bolts are a good example) fail, it is due to fatigue.

A good rule of thumb to avoid this fatigue, is to pretension the bolts to at least twice the applied load. For heavy vibration/shock loads up to 5 times may be used.

For machine parts, which are regularly dis-assembled and bolts re-used, the pretension used is approximately 65 to 70% of the bolt proof load.

For steel structures subject to dynamic loads, it has long been a practice to pretension bolts to their proof load (it does not matter that the bolts may yield slightly). In these cases bolts should not be re-used. In some cases a bolt may be re-used once, so long at it is used in the same position.

It is becoming more common to pretension some engine bolts (eg. head bolts) to their proof load.

Torque control (unless using specialised equipment) is not permissible for pretensioning to proof load. This is because the friction between the mating threads and surfaces is too variable. Common methods of pretensioning are part turn, or measurement of elongation.

With fatigue of bolts in tension, the variation between minimum and maximum applied tension is more important the peak tension.

When a bolt is pretensioned, the bolt is stretched and the joint compressed.

With a well designed bolted joint, the bolt is not as stiff as the joint (often about 1/10). This will change if the joint has a flexible gasket or the modulus of elasticity of the joint material is less than for the bolt (eg aluminium).

When a load is applied parallel to the bolt, the bolt tension will increase and the joint compression will reduce. The change in bolt tension plus the change in joint compression is equal to the applied load.

The change in bolt tension compared to change in joint compression is the same ratio as the bolt and joint stiffness.

With the well designed joint, the variation in bolt tension, which affects fatigue strength, is small compared to the applied load.

If the pretension of the bolt is insufficient, the compression of the joint will reduce to zero (joint separation)under the applied load, and the variation in the bolt tension will then equal the applied load. This is what usually causes bolts to fail. The other common cause is a change in the joint stiffness, due to dirt, warped surfaces, or a gasket in a joint designed for metal to metal contact.

Increasing bolt pretension, increases the fatigue strength.

Reducing bolt stiffness and increasing joint stiffness by increasing thickness of joint and length of bolt (or necking the bolt body down) will increase the fatigue strength.

Link to comment
Share on other sites

Just to wade in, quote out of context and try to shoot you down on the only part I know anything about ;) :

Steel wheels and nuts use mating conical surfaces to increase the friction between the wheel and the nut, so don't depend so much on friction between the mating threads. To my knowledge, alloy wheels don't have the conical surface, so thread friction is more important

The conical surfaces on the steel wheel and normal nut are simply to locate and centralise the wheel. Since the friction is independent of contact area, concentrating the force into a small area doesn't make any difference to the holding torque or clamping force, although the stress concentration would excessively deform an alloy wheel so LR alloys use a spigot mounting and plain washers under the wheel nuts (coincidentally, just like a bus wheel). The LR nuts also have a cone machined on so you can use the same nuts with your steel spare though.

The rest of your post looks spot on though, as far as I understand it! :ninja:

Link to comment
Share on other sites

I agree with most of your observations, but find your discussion of friction factor for lubricants is misleading, and I suspect totally wrong. How do you define friction factor?

John,

For our application, a control compound is used with a defined nut and bolt. This is regarded to have a FF of 1.0. The other compounds are then compared against it. I didn't mention coefficients of friction anywhere. For our application, this calibration is pretty "rough" actually.

The whole point of the thread was to raise the topic for discussion because I made some interesting observations and felt they were worth sharing to hopefully save anyone making expensive mistakes... Specifically, I wanted to highlight:

Break-out torque has nothing to do with what the fastner was made up to.

There has been discussion on what compound should be put on head bolts... I wanted to point out that you can have radically different results depending on what grease you use.

Link to comment
Share on other sites

Just to wade in, quote out of context and try to shoot you down on the only part I know anything about ;) :

The conical surfaces on the steel wheel and normal nut are simply to locate and centralise the wheel. Since the friction is independent of contact area, concentrating the force into a small area doesn't make any difference to the holding torque or clamping force, although the stress concentration would excessively deform an alloy wheel so LR alloys use a spigot mounting and plain washers under the wheel nuts (coincidentally, just like a bus wheel). The LR nuts also have a cone machined on so you can use the same nuts with your steel spare though.

The rest of your post looks spot on though, as far as I understand it! :ninja:

The way that the conical surface increases the friction between the nut and the wheel has nothing to do with the area, but everything to do with the cone angle.

As rtbarton correctly stated, the limiting friction force = uN where u is coefficient of friction and N is the normal force/reaction.

The angle of the conical surface increases the normal force, thus increasing the friction force. Same way that vee grooves increase the friction between belts and pulleys.

Link to comment
Share on other sites

The angle of the conical surface increases the normal force, thus increasing the friction force. Same way that vee grooves increase the friction between belts and pulleys.

I would have thought for a given tension in the bolt a conical surface would reduce the friction. You will only get a component of the tension in the stud providing the normal force, something like N = COS(90-A) * T, where N is the normal force, A the angle between the stud and the cone and T the tension in the stud.

Link to comment
Share on other sites

I would have thought for a given tension in the bolt a conical surface would reduce the friction. You will only get a component of the tension in the stud providing the normal force, something like N = COS(90-A) * T, where N is the normal force, A the angle between the stud and the cone and T the tension in the stud.

COS(90-A) * T would be the value of the radial component.

The normal force would be T / COS A

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...

Important Information

We use cookies to ensure you get the best experience. By using our website you agree to our Cookie Policy