Friction material on centerlock wheel mating surfaces?

[stownsen914]

I have a tube frame GT car that uses March 87C Indycar centerlock uprights and hubs. I think I've mostly solved an issue I had at first with the centerlock wheel nuts coming loose, thanks to help here on the forum (thread here). I am still seeing evidence of some back and forth rotational movement on the rear wheels due to the braking and acceleration cycles, even with the wheel nuts holding their torque. One thing I've noticed is there appears to be a lightly textured material applied to the mating surface on the hub side. Maybe friction material to resist rotational movement?

See pics below. The first one shows the material on one of the front hubs (actually rotor hats) - see the dark circle in line with the drive/locating pins. It's pretty much intact on the front still. The second pic shows the rear where it's mostly worn off, presumably due to the repeated braking and acceleration cycles.

I'm wondering if I should look into re-applying the friction material to the rear mating surfaces. Anyone know what this stuff is? Plasma spray coating?

Thanks for any input!

[DaveW]

Working back and forth due to unavoidable slop in the drive pins is, IMO, a combination of 4 things:
o Acceleration torque (exacerbated by losing traction over bumps)
o Braking torque (exacerbated by losing traction over bumps)
o Friction coefficient between wheel and hub (what you are asking about)
o Centerlock wheel nut clamping tension

So while your question is, IMO, valid, you may be able to improve the situation by reducing friction coefficient (lube or low-friction plating) between the nut either on its mating face (1st choice) and/or the hub threads (not as good because nut might work loose), thus increasing clamping tension.

Just something more to think about.

[Jerry Kehoe]

You could always either insert bushings into the wheels and ream them to size or create slightly larger drive pins if you believe there is too much clearance, being aware of course that everything grows when it gets hot so since the pins are likely steel fairly snug bushes may work better. Most wheels/hubs I have ever seen show some galling from even very slight movement.

[DaveW]

You could always either insert bushings into the wheels and ream them to size or create slightly larger drive pins if you believe there is too much clearance, being aware of course that everything grows when it gets hot so since the pins are likely steel fairly snug bushes may work better. Most wheels/hubs I have ever seen show some galling from even very slight movement.

The problem with trying to make things tight enough so they won't slip is that one needs some minor amount of clearance to account for tolerance buildup between the wheel and the hub. Too tight and getting the wheels on and off becomes an issue.

In practice, IMO, there is no way to eliminate 100% of the slop with a pin-drive setup, so friction between wheel and hub needs to be high enough to keep things from moving once the slop is initially used up in the highest axle/hub torque direction.

[reidhazelton]

How much is the "rocking back and forth?" The tolerance between the drive pins and the mating holes should be a slip fit. The wheel nut doesn't keep the wheel from rotating as much as it just keeps it on the pins. If there has been enough wobble over time the pins or the mating holes on the wheel could be deformed. I'd snug the nut down just enough to have the wheel seated on the pins, have someone step on the brake, and rotate the wheel by hand clockwise and counterclockwise and see if there is any slop in the pins/wheel.

If a wheel has come loose and left to wiggle around on the pin drives it can open the mating holes up on the wheel too.

I had a friend in high school that had problems with lug nuts coming loose on his Dodge Shadow. He used a hole bottle of red Loctite and that solved it. Turns out metric lug nuts don't work so hot on standard threads, but the Loctite sure did the trick.

[stownsen914]

Thanks guys for sharing your thoughts. To respond to a few suggestions/questions:

• Re: putting lubricant on the face of the wheel nut - I've understood that this isn't usually recommended. When I asked about lubing the threads, I got all kinds of responses. I wound up putting a WD40 anti-corrosion product on the threads, but leaving the contact face of the wheel nut dry. Lubing it would increase the clamping load, but it seems like no one does that?
• Re: slop in the drive pin slots in the wheels, there is some. I had a machinist make me some thin wall sleeves to fit over the pins to take up most of the slop. I measured everything carefully before having them made. I think I got the slop down to .010" or so with the sleeves.

[DaveW]

Thanks guys for sharing your thoughts. To respond to a few suggestions/questions:

• Re: putting lubricant on the face of the wheel nut - I've understood that this isn't usually recommended. When I asked about lubing the threads, I got all kinds of responses. I wound up putting a WD40 anti-corrosion product on the threads, but leaving the contact face of the wheel nut dry. Lubing it would increase the clamping load, but it seems like no one does that?
• Re: slop in the drive pin slots in the wheels, there is some. I had a machinist make me some thin wall sleeves to fit over the pins to take up most of the slop. I measured everything carefully before having them made. I think I got the slop down to .010" or so with the sleeves.

1. Since clamping force and CF between wheel and hub are the dominant factors in keeping the wheel from "working" back and forth, I would concentrate most on improving those, even if lubing the nut on all contact surfaces is required.
2. The mechanism for centerlock nut loosening is as follows:
ANY amount of slop (a small amount is unavoidable if there is enough clearance to assemble the wheel to the hub) between the pins and holes in the wheels, no matter how little, will allow the wheel to rotate back & forth if the CF x clamping force between the wheel and hub is too small. This "working" can rotate the nut a tiny amount in the loosening direction each time the wheel "works" in that direction. When (if) the wheel rotates back in the opposite direction, the nut will not tighten since it would have to "go uphill" to do that. So the nut eventually will loosen.

Bottom line: the wheel MUST NOT be allowed to "work" back and forth on the hub, or loosening will likely occur. If the wheel cannot repeatedly rotate in alternating directions on the hub (sufficient CF x clamping force), the nut WILL NOT LOOSEN no matter how it was lubed.

[tlracer]

There is always a level of free movement between drive pins and the mating holes in the wheels, simply so they don't bind during fitting/removal (which, when they're being changed mid-race, needs to be inherently a fast process).

The problem with this is that repeated use, movement on acceleration/braking whilst fitted, etc. enlarges the holes - made worse since the drive pins are harder material (steel) than the wheels (alloy).

It's common practice with endurance teams to check wheels and remachine/sleeve when the tolerance goes out of acceptable limits. Eventually they will simply discard/sell worn wheels - which is possibly why there are so many that have been converted into coffee tables!

It's not unusual for wheel nuts to be aluminium to save weight and also promotes a degree of galling ( = increased friction ) between the wheel and nut. The latter is fine initially, as it helps to prevent the nut from loosening. However over time the nut/wheel mating faces become worn/rough, so the nut only touches on the high points then - given the small amount of rotation of the wheel on the pins - wears and loosens, further increasing the potential for relative movement.

Again, once tolerances are exceeded, the wheels are usually discarded.

Turning to the material on the disc bell, it could be a friction coating; but this would only work effectively if the same material was on the mating face of the wheel. Otherwise all it would serve to do is wear away that face and - again - potentially increase relative movement. I would guess this is going to be similar to the coating sometimes found on the bead seat, to assist tire retention/reduce rotation;

The trouble is all these interact together to contribute to a downward spiral in stability of the wheel/hub interface. Solving this involves many things:

- Measure and note the diameter of all drive pins. Any that are significantly smaller should be replaced so they are all within an acceptable tolerance range;
- Measure and note the diameter of all wheel location holes. Machine and sleeve based on drive pin diameter, but only once the latter have been addressed (as above);
- Check condition on mating faces of nuts and wheels. To an extent these can be remachined but if this means removing a lot of material they are best discarded;
- Consider light lubrication of the threads on the nut but not the nut/wheel mating faces.

[t walgamuth]

I'd try a drop or two of red loctite on it.

[Charles Warner]

This argument/subject comes up once a year, usually in the winter. There are two sides.

One side considers the pins as "drive pins" and feels that these impart the turning force between hub and wheel.

The other side, to which I fully subscribe, is that these are "locating pins" which center the wheel on the hub. In a properly designed application, they should not impart any driving force to the wheel at all. Case in point: my 1987 March F1 cars had massive 18" wide rear tires. Yet, there were only three (yes, THREE) pins on the each hub (front and rear) that were about 14mm in diameter.

As Dave W. points out, it is the clamping force between wheel nut and wheel that keeps the wheel from turning. While, during installation, the wheel may rotate that little bit to take up the minimal difference between pin radius and hole radius, the wheel is kept on by the clamping force of the nut, whether a conical one (i.e. current F1/Indy Cars/F3000) or a flat one (i.e. RT-4/F2000, et al.)

Any galling of wheel or damage to the pins indicates the wheel has worked loose in the past due to a loss of clamping force.

YMMV.

By the way, I have a ton of new locating pins left over from the Ralt/March factory.

Overall length 1.375"
Diameter .550"
Pin height .700"
Thread 3/8-24
Thread length .375"

$5.00 each shipped.

[DaveW]

... it is the clamping force between wheel nut and wheel that keeps the wheel from turning. While, during installation or 1st/largest driving/braking torque application, the wheel may rotate that little bit to take up the minimal difference between pin radius and hole radius, the wheel is kept on and kept from further rotation with respect to the hub by the clamping force of the nut, whether a conical one (i.e. current F1/Indy Cars/F3000) or a flat one (i.e. RT-4/F2000, et al.)

Any galling of wheel or damage to the pins indicates the wheel has worked loose in the past due to a loss of clamping force.

YMMV...

Added text in bold red above...

[tlracer]

This argument/subject comes up once a year, usually in the winter. There are two sides.

One side considers the pins as "drive pins" and feels that these impart the turning force between hub and wheel.

The other side, to which I fully subscribe, is that these are "locating pins" which center the wheel on the hub. In a properly designed application, they should not impart any driving force to the wheel at all.

I'm not sure I agree with this, for the following reason:

If we accept the clamping force between the wheel and hub, provided by a central nut, is sufficient to prevent relative movement between wheel and hub, and the pins are simply there as a locating device, then it should be perfectly possible to delete the pins entirely since the conical nut face and corresponding seat on the wheel would provide the required centering of the wheel on the hub.

A simple way to ascertain whether the pins provide driving force or not would perhaps be to measure the diameter of the holes in a number of old/worn wheels. If they are still of uniform diameter then the pins provide no driving force; if, however, the width along the pins' PCD radius is smaller than the width perpendicular to that line, then clearly there must be some driving force through the pins.

[DaveW]

Charles and tlracer, IMO, the pins definitely provide driving force. In my view, the clamping friction keeps that force from being applied in alternating directions during acceleration/braking cycles once a set is taken, preventing loosening of the nut.

Theoretically one COULD make it work w/o the pins, but I sure wouldn't be the one to try it.

[t walgamuth]

Added text in bold red above...

I disagree. The pins are there to put the wheel where it belongs. The nut clamps it on and the friction should keep it in place. Any slippage I suspect is from stretching of the fasteners while being loaded.

[tlracer]

Well, I looked out an old Corvette GT1 wheel I have and photographed the holes:

IMG_20210420_184808.jpg

They are noticeably oval...

[t walgamuth]

Charles and tlracer, IMO, the pins definitely provide driving force. In my view, the clamping friction keeps that force from being applied in alternating directions during acceleration/braking cycles once a set is taken, preventing loosening of the nut.

Theoretically one COULD make it work w/o the pins, but I sure wouldn't be the one to try it.

If it slips one direction why would it not slip the other?

In my structures class we were taught the friction is what keeps it all together We provide bolts to create the friction. The bolted connection should be stronger than the members it connects, so if there is a failure the beam or column would fail and the bolted connection would remain.

So you should be able to tear the center out of the rim before the bolts fail.

[tlracer]

So you should be able to tear the center out of the rim before the bolts fail.

As well as the Corvette wheel I have one from a Roock Porsche. Very little wear and perfectly round, but all the Y-shape spokes are bent radially where they meet the rim, suggesting the torque load has been more than it could cope with ie. nearly tearing the centre out of the wheel.

[DaveW]

If it slips one direction why would it not slip the other...

Because the torque is never the same in both directions. The fronts see only brake torque. The rears see acceleration and brake torque, but they are never exactly equal. So if the clamping load is high enough, the wheel will tend to rotate (if it rotates at all) in the most heavily loaded direction, and if the clamping load is high enough it will stay there. Everything still depends on the clamping load even if the wheel has shifted against the pins.

Remember in the rear that there are shock loads caused by wheel breakaway and sudden re-gripping during acceleration or braking over curbs, etc. But the accel loads are almost always higher due to weight transfer during accel v braking.

[stownsen914]

Thanks guys for all the responses, I appreciate it. So a additional thoughts:

• I'm wondering if the "friction surface" applied to mating surface where the wheel contact it, may actually be a protective/sacrificial layer to prevent or reduce galling of the rotor hat and wheel mating surfaces. It's mostly worn off on the rear hats. (Looks new on the fronts where there isn't much/any back and forth movement due to only braking cycles, no acceleration.) Or maybe it serves two purposes - friction + protective?
• I'm torqueing my wheel nuts to 450 ft-lbs, and still seeing evidence of the rotation/rocking on the rears. I lightly lube the wheel nut threads with a WD40 anti-corrosion product and make sure all the contact surfaces (hub, wheel, nut) are clean. Should I be torqueing to more than 450 ft-lbs?

[DaveW]

Thanks guys for all the responses, I appreciate it. So a additional thoughts:

• I'm wondering if the "friction surface" applied to mating surface where the wheel contact it, may actually be a protective/sacrificial layer to prevent or reduce galling of the rotor hat and wheel mating surfaces. It's mostly worn off on the rear hats. (Looks new on the fronts where there isn't much/any back and forth movement due to only braking cycles, no acceleration.) Or maybe it serves two purposes - friction + protective?
• I'm torqueing my wheel nuts to 450 ft-lbs, and still seeing evidence of the rotation/rocking on the rears. I lightly lube the wheel nut threads with a WD40 anti-corrosion product and make sure all the contact surfaces (hub, wheel, nut) are clean. Should I be torqueing to more than 450 ft-lbs?

To have a better idea about what does what, IMO, you need to do an analysis of what clamping load (and thus possible friction holding the wheel from working back and forth) results from a given nut torque. A large % of that torque is dissipated in friction on the threads and the nut-wheel interface and does not result in clamping load. Larger thread angle and diameter eat up a lot of the torque. I did that analysis a few times on various fasteners and it was surprising to me how inefficient screw threads can be. Reducing friction coefficient on the threads and nut face makes a huge difference.

BTW, WD40 is a lousy lubricant for friction reduction. From your comments, I assume you know this. About all it's good for is reducing corrosion.

[tlracer]

Some other points to consider:

a) Are the pins correctly torqued and into good threads with no excess play?
b) Be sure to separate hub/wheel play from any other driveline play (jack up the car, get someone to press the brake pedal then, with the centrelock nut fitted just tight enough to prevent the wheel from wobbling, try rotating it to determine the pin/wheel free play)

[t walgamuth]

I was thinking I can lock the brakes easily but only can spin the tires in the lower gears but I don't know how that all fits in. I'm not sure about the rears getting more abuse. Of course the tires are wider ....yeah probably you are right about them getting more stress.

But unless something (threads?) is worn a lot and not getting correct torque the wheel should not move on the hub.

[DaveW]

I was thinking I can lock the brakes easily but only can spin the tires in the lower gears but I don't know how that all fits in. I'm not sure about the rears getting more abuse. Of course the tires are wider ....yeah probably you are right about them getting more stress.

But unless something (threads?) is worn a lot and not getting correct torque the wheel should not move on the hub.

Tires are wider and more weight on the rear. 1st gear acceleration over bumps/gators is likely the worst scenario.

[stownsen914]

To answer respond to a few comments/questions:

• I'll check the drive/locating pins for tightness. I've not noticed them being loose, but will check. I checked my notes, and they are indeed about .550" as noted a few posts above.
• Re: lubricating qualities of WD40, Roger that. I've gotten feedback all over the place about lubing vs. not. I chose a light lubricant as a compromise. Maybe I need to try a more substantial lubricant to increase clamp load? I understand that different lubricants can cause reduction in required torque 10-50% to achieve a given clamping force. In my case, perhaps I should keep the same torque value for the additional clamping force.
• I noticed a comment about testing clamp load. I have no idea how to test that. Ideas?

Still wondering if anyone knows about this coating on the mating surface on the hub side. It seems strategically placed - wondering how important it is.

[Purple Frog]

I always used anti-seize on the nut threads, as did every pro team I ever worked for. Just saying. You may think you are getting 450#, but like DaveW says, thread friction can eat up a lot of torque. WD-40 is not anti-seize.

[DaveW]

...Maybe I need to try a more substantial lubricant to increase clamp load? I understand that different lubricants can cause reduction in required torque 10-50% to achieve a given clamping force. In my case, perhaps I should keep the same torque value for the additional clamping force.[*]I noticed a comment about testing clamp load. I have no idea how to test that. Ideas?...

I think lubing with something like NeverSeez or similar but using the same 450 lb-ft torque is a good way to go at least until you have more knowledge about what you need.

I think there are load cells made in a donut shape that could measure that load - they are not cheap, though, IIRC. A calculation might ultimately be easier and definitely less expensive. There are probably formulas for that calculation on the web that could at least get you more understanding of your situation.

[jchracer]

Preload is what keeps the nut from loosening. Preload varies wildly with friction. A good rule of thumb is to try to get to ~85% of the yield strength of the stud with the preload. In the jet engine arena where I work, we do analysis to determine the spring rates of the tension and compression stacks and apply an angle-of-turn rather than a torque spec when assembling critical bolted joints. This removes the variability of the friction. Obviously, this is not practical for a race car wheel stud. In the absence good experience or better specs, there are sophisticated torque wrenches that will data log the angle of turn and torque reading simultaneously. From the generated angle of turn vs. torque graph you could creep up on a torque value (using a consistent anti-seize lube) that would get you close to the yield strength of the stud without exceeding it.

[Jerry Kehoe]

In the 40 plus years of racing Ralt and March cars and never having a wheel nut loosen up although I had a stud break on the Ralt, which was not that uncommon on the stock studs until higher grade steel was used for replacements. All of these studs used a 3/4 fine thread and steel nut with a reasonably thick locating washer. They were always torqued to 145-150 lbs with anti seize. Perhaps the much larger diameter nuts with their much coarser threads and usually being made of aluminum makes it harder to torque them properly as well as the expansion of the aluminum as it is heated up. It is interesting that my old TR3 uses left hand/right hand knock offs, for all of you engineers could this be a factor?

[DaveW]

... It is interesting that my old TR3 uses left hand/right hand knock offs, for all of you engineers could this be a factor?

That fits with the torque making the wheels want to shift on their hubs being directional. In the old days, IIRC, braking was almost always the predominant torque. So I'm guessing on that basis that the RH knock-offs would have been RH thread, and the LH would have been LH. That would tend to tighten the knock-offs when the wheels rotated slightly on the hubs under hard braking.

I had a TR4 with knock-off wire wheels for a while, and the knockoffs were directional, but I don't remember which way. I converted to steel wheels after a short time because it was almost impossible to keep the wire wheels as straight as I wanted.

[jrh3]

How far do the drive pins extend into the wheels?
possible longer pins?

[Jerry Kehoe]

How far do the drive pins extend into the wheels?
possible longer pins?

pins stick out about a half to 3/4 inch

[Jerry Kehoe]

That fits with the torque making the wheels want to shift on their hubs being directional. In the old days, IIRC, braking was almost always the predominant torque. So I'm guessing on that basis that the RH knock-offs would have been RH thread, and the LH would have been LH. That would tend to tighten the knock-offs when the wheels rotated slightly on the hubs under hard braking.

I had a TR4 with knock-off wire wheels for a while, and the knockoffs were directional, but I don't remember which way. I converted to steel wheels after a short time because it was almost impossible to keep the wire wheels as straight as I wanted.

mine is a 56 that I drive on the street and since they ride so crappy, have tons of bump steer, I probably couldn’t feel if the wire wheels are straight

[DaveW]

mine is a 56 that I drive on the street and since they ride so crappy, have tons of bump steer, I probably couldn’t feel if the wire wheels are straight

Even our TR4 rode like a truck - leaf springs in the rear sure didn't help. Good thing we were young at the time.

[stownsen914]

Sounds like using antiseize on the wheel nut threads would be a good next step. I'll keep the torque at 450 ft-lbs.

As for the stuff that's applied to the wheel mating contact surface on the rotor hat (friction material, anti-galling?) ... not important?

[tlracer]

I might be way off with this, but is the coating actually remains of anodising on the bell?

[stownsen914]

I might be way off with this, but is the coating actually remains of anodising on the bell?

I don't think it's anodizing. See the first pics at the top of the thread. It's the black ring. The stuff has a texture to it. I'm pulling the car out of storage hopefully this weekend. I'll post a closeup of it.

[tlracer]

I don't think it's anodizing. See the first pics at the top of the thread. It's the black ring. The stuff has a texture to it. I'm pulling the car out of storage hopefully this weekend. I'll post a closeup of it.

It did look a bit too regular for the remains of coloured anodising, but worth asking just to discount the idea.

Another possibility might be something like one of the variants of Apticoat - anti-friction, anti-wear, etc.

[jchracer]

I've always wondered why "price is no object" racing teams like F1 didn't use these couplings to mate the wheels to the hubs:

http://moderncnc.com/v-tooth/hirthco...xoCNfsQAvD_BwE

They self center and transmit the torque with zero slop. Expensive but pretty much eliminate all the problems discussed here.

[DaveW]

I've always wondered why "price is no object" racing teams like F1 didn't use these couplings to mate the wheels to the hubs:

http://moderncnc.com/v-tooth/hirthco...xoCNfsQAvD_BwE

They self center and transmit the torque with zero slop. Expensive but pretty much eliminate all the problems discussed here.

Probably weigh a few ounces more or needs to be slightly larger than what they do use... Or it's somehow prohibited by the regulations.

[t walgamuth]

I've always wondered why "price is no object" racing teams like F1 didn't use these couplings to mate the wheels to the hubs:

http://moderncnc.com/v-tooth/hirthco...xoCNfsQAvD_BwE

They self center and transmit the torque with zero slop. Expensive but pretty much eliminate all the problems discussed here.

So its basically a lock washer?