Self-tightening wheel spinners?

(Paul Wigton) #43

Did not know THAT!!!

(Paul Wigton) #44

WAG: did they have the center toothed locks, like a PIII Rolls?

(Robert Wilkinson) #45

Nuts tend to tighten or loosen under conditions of high angular acceleration. As the wheel accelerates, it transmits a torque to the lug nut, whose inertia resists the acceleration.

So a grinder or buffer has left hand threads on the left, and right hand threads on the right, such that the nut tightens when the machine is turned on. Many cars a few decades ago adopted the same convention, for the same reason.

Convinced? Don’t be yet, because buffers and grinders rotate in a direction that corresponds to a car going backwards. But note that, particularly in those days, angular acceleration was far greater during braking (deceleration) than with pedal to the metal. So to believe my hypothesis, you have to believe that the manufacturers knew this.

The physics behind centre-lock tightening (as someone mentioned, the same principle as rotation of a hula hoop) is well-established and easy to look up. I just googled and could find nothing to support what I just proposed for lug nuts. But, evidently, it was determined experimentally that the wheels most likely to loosen in service were on the left–and this was improved by LH lugs on that side. Clearly IMHO.

(Chris Fell) #46

Not on Dad’s car, a TurboR from 1986.

(PeterCrespin) #47

Epicyclic action and deflection under load do it for me David. The RW system relies on that pretty simple concept, exemplified by the hula hoop or spinning a roll of Sellotape around your index finger. I.e. that two circular parts of an assembly may ‘rotate the same direction’ whilst still having a degree of contra-rotation ( in this case mostly ‘creep’) with respect to one another.

This is not serious, presumably? It’s a bit like “If the universe is infinite how come it’s expanding? Or “If God is almighty can she make a jam jar lid so tight she can’t get it off?” A constant designed-in tightening epicyclic action will maintain the assembly snug and free of frank loosening, despite service conditions (reverse loading via imperfect splines etc.) that may generate loosening forces. The wheel hub and spinner don’t oscillate between loose and tight on a macro level.

Same with your oxymoron statement, questioning the logic of a tight spinner fully in contact on all appropriate surfaces being able to get any tighter. Your torque wrench and spinner tool demonstrate that a spinner can be tight at 50 lbs/ft, 100lbs/ft or double that figure. They all qualify as tight, just gradations of tight.

Being slightly mischievous with respect to Lotus, their system IS very different to my Dunlop peg drives. I have a high regard for thr UK judiciary overall, but unless the Judge was a trained Mech Eng I would not put too much store by what was fundamentally a judgement in law rather than applied physics and engineering. You could probably find a lawyer to act for the architects of Giza against Suchard because Toblerone looks like a pyramid…

(David Jones) #48

Decided to take up Ray’s suggestion and run a practical test.

I jacked up the rear of my FHC and removed the nearside spinner and wheel. Followed the Jaguar instructions regarding making sure there was no grease on the hub cone and the Dunlop instructions on refitting the wheel and spinner. The spinners are genuine NOS Jaguar made by Albion Brockhurst, wheels and hubs are new and only covered about 1,500 miles since. Spinner tightened up by hand with no binding (as repro spinners tend to). I used a rubber mallet to give the spinner a single whack to ensure it was seated and I could not remove it by hand. I used the Wiggles specified Sharpie to mark the position of the spinner on the wheel. I found there was slight movement in the wheel when I rocked it but less than I expected.

I took the car out through our small village at 30 mph, out on to a country road and felt confident enough to take it up to 60 mph. Moved on to a long straight dual carriagway and looped back at 50 mph (speed cameras!). No knocking noises or other nasties but I did avoid harsh acceleration, braking and cornering. Total distance traveled was about 5 miles and I have to admit I would not have suspected a loose spinner so the self tightening theory began to make sense.

Back in the garage I checked the spinner and found it had not moved at all from the start position; it was exactly as when I tightened it. I jacked the rear up and used a lead mallet to see if there was any possibility of tightening and indeed the spinner moved about 3/4" from the mark. Using a spinner removal tool and torque wrench it tested as above 220lb/ft. So the RW system is indeed ‘self locking’ as Dunlop described in that it will hold the spinner where you left it. It is fail safe in that a loose spinner will not loosen further but neither will it tighten further it seems.

In the interests of experiment I decided to do a second run but this time with a front wheel spinner ‘loose’. Same procedure and route as above but a far more disconcerting experience. Any major application of steering produced a knocking sound from the front. A hard left turn resulted in ‘graunching’ but I was confident the spinner would not loosen further so finished the route. Examination of the Sharpie marks again showed no movement of the spinner relative to the wheel. Three whacks with a lead hammer moved the spinner about 1".

Maybe with more or harsher driving there would be some movement but doing those tests was not an enjoyable experience or indeed really necessary considering the potential danger and spline damage. The spinners are self-locking and provided the Jaguar/Dunlop instructions are followed there should be no reason to expect self-tightening beyond what was set.

Let us not forget that RW made these claims in the 1920’s and 1930’s (when smoking was considered good for your health) and subsequently went bust. Jaguar/Dunlop bought the patent out of receivership, did their own tests and dropped all references to ‘self-tightening’ replacing it with ‘self-locking’.

Anyway, I will get my mallet and leave!


PS In the original Rudge-Whitworth design, the male taper was on the wheel and the female taper on the nut, so the car’s left-hand wheels required a right-hand thread and vice versa. On Lotus Elan S/E wheels the male taper is on the spinner and the female on the wheel so the thread sense is reversed. Same principle in action and only used by Chapman to try and avoid royalties which he failed to do because of what was obvious to everyone, even the Judge.

(PeterCrespin) #49

Rudge motorbikes used LH nuts on one side of the bike and RH on the other. I never owned one but one of my bikes had a LH nut on the left side footrest. I think it was a mid fifties BSA Gold Flash.

The footrests were a short forged arm with the footrest bar on the forward end and a female taper on the other end that mounted over a male taper brazed to the frame. The idea was that you could adjust the footrest through an arc to suit your preferred riding position, but unlike a square bar or splines you could lose control if the peg suddenly moved and jammed on the brake.

The LH nut on the left and RH nut on the right meant that if the footrest moved down it got tighter and jammed, provided the nut and locking washer were reasonably tight to start with.

And of course bicycle pedal threads are handed.

(Mitchell Andrus) #50

Thanks David.


(Philip Hurlston) #51

Self tightening spinner threads are almost as bad as “which oil should I use” threads…

(69 FHC ) #52

Both do get beat to death. :grinning:

(Paul Wigton) #53


Who mentioned K&Ns?


(Mark Gordon) #54

They’re a major cause of global warming.:grin:

(Geo Hahn 1969 Series 2 OTS) #55

Those are not the conditions under which I observed the ‘self-tightening’ feature so I do not know what results could be expected.

When I have observed it the knock-offs were well-tightened (several whacks with a 6# lead hammer until they moved no more), marked with a Sharpie and then regular driving (fast-slow, stop-start) for a couple of days showed further tightening.

(Paul Wigton) #56

Those are also the conditions under which I observed said tightening, on my Jag, and a few other smaller (42 mm) RW Brit cars.

(Mitchell Andrus) #58

From the MWS website.

MWS wesite

  • A spinner should be fitted with the wheel jacked up off the ground so that the hub, wheel and spinner are all centralised.
  • Apply the spinner by hand and tighten 3 to 3½ turns; give the spinner a couple of taps with the hammer; lower the wheel to the ground and give the spinner a couple more taps.
  • As long as the hubs are fitted correctly then the spinners will self-tighten and lock as the car is driving.

(67 OTS S1) #59

Point in fact- I went around the block slowly after years of restoration. I heard something and saw something silver in my side mirror. It was the spinner! I put it back on and hit all the spinners. The next day I did the same loop to show off for my daughter and the other side fell off. I then figured out what was wrong, I swapped the hubs.

(Ben E) #60

I once bought an MGB that had two RH hubs on the rear. I noticed it while I was inspecting the car, and asked the seller how long it had been that way. He said he’d owned it 5 years, and it must have been that way when he bought it.

The spinner came off normally, so it wasn’t as if a gorilla had pounded it on…I’ve never understood how he didn’t have a problem with it.

(Robin O'Connor) #61

I believe it was a Mazda that I had in my shop that had left hand wheel studs on the LH side, evidenced when on of the workers sheared one off :frowning:

(PeterCrespin) #62

“Robert is correct about the Lotus system which reverses the way the RW hubs/cones work.”

Cone singular. The Lotus system is pure mechanical peg drive IMO, with a single taper on the clamping nut/spinner.

“On Lotus Elan S/E wheels the male taper is on the spinner and the female on the wheel so the thread sense is reversed. Same principle in action"

I was with you David until the last four words :slight_smile:

This is fresh meat in the old spinner debate so I looked it all up and learned something new to me if not others.

Therefore, being not the ‘same-old, same-old’, I hope we can get a little more mileage out of this Lotus learning opportunity? My take follows, and is not aimed at contradicting David so much as raising questions that qualified engineers may help us answer.

This thread title and topic refer to the Rudge-Whitworth (R-W) principle whereby a wheel is sandwiched between a male and female taper. These jam the wheel tightly, holding it concentric with hub splines, without relying on them totally for torque transfer in normal use. I’m not a Mech Eng but there seems to be a fundamental difference between that R-W system and the Lotus system.

The Lotus system holds the wheel not jammed between tapers but pressed against the edge of a flat hub flange, with four perimeter pegs that engage holes in the wheel and provide almost all of the drive. There must be a small friction component at the narrow ring of contact between the pegs, but I suspect it is less than self-tightening R-W tapers. There
are no splines and NO taper anywhere on the Lotus hub. The latter is a fundamental difference from the RW system, with no sandwich between two tapers.

Lotus used a single male-tapered nut to hold the wheel against the flat peg hub, rather than the usual 4 or 5 male-tapered lug nuts.

For reference, the Dunlop peg drive hubs DID use the R-W two-taper system. Yet despite this commonality, the Dunlop system was filed under a separate patent. whereas Lotus’s very different system was judged to have infringed R-W’s rights.

Maybe Chapman’s lawyers weren’t as nimble as his cars? All IMO, as Robert would say…

(Robert Wilkinson) #63

I think it might be true (or at least I’m aware of no counterexamples) that the taper of the outboard contact, between wheel centre and knock-off determines the handedness of the knock-off’s threads.

Kelsey-Hayes Corvair wire wheels have direct contact inboard, but with a casting that resembles a tapered gear–similar to peg drive. Corvette alloy wheels (1962) uses the same knock-offs but have actual peg drive. Dayton wheels also seat flat against a casting that looks like a bicycle chain ring (“gear drive”). All of these variants have no inner taper, only an outer one whose polarity is like R-W, and all use RH threads on the left and LH threads on the right. Conversely, the Elan has reversed outer taper and consequently reversed thread direction to achieve epicyclic tightening.

My conjecture is that an inner taper–a ball joint bearing in effect that allows for a more robust epiclycic tightening–is optimal but isn’t necessary. The weakness of this (original R-W) design is spine wear; pegs or gears are more robust. But an outer taper, either male or female, is required. Moreover, given the effect of braking on lug nuts, it would seem to me that the Elan’s outer taper is superior. Both phenomena–epiclycic tightening and the knock-off’s moment of inertia–are in concord, favouring LH thread on left, RH on right.