Talk Morgan Forums Morgan Sports Cars Morgan Three Wheeler & Super 3 Centa rubbers

Sorry - when I said "no cushioning" I meant as in any material that would absorb sharp blows such as may occur when the Centa rollers are undersize. I understand how the HD works (have a newish one on bench in front of me) and how you have managed to keep going with this is great, but perhaps your nut issue is the same as the PB (harder) roller equipped Centa - there is no compliant material to absorb 'sharp' rotational blows?

Yes they do vary from car to car. When I took my car to Phil to be Bleazeyed, he took great care to select a crank adapter to fit my particular crank spline. Even then it needed a little fettling to achieve a perfect fit.

Just took Centa out of a b128cu/in engined m3w - 10,000 miles - rollers shredded to pieces.
Nut on tight.
Pulled Centa housing off by hand.
Splines are perfect condition.

So although the splines might be a factor in the issue with the nut undoing - personally I think the splines are not as super critical as made out to be - the nut tightness is as important or more than likely more important.
I still think the rollers are the main culprit as Laurens explained earlier.

I know PB doesn't like threadlock etc, I agree if properly engineered it should all stay together with the correct torque. However if the nut is coming loose for whatever reason, rollers / splines and causing roadside failures which was one if not the main benefit of the Centa Kit to be able to change the rollers reasonably easily. Would physically locking the nut in place actually cause any harm? I think Ian Brett mentioned in a post that he had done this.

I would say the fit of the splines is critical. As the loading in the coupling reverses, any movement in the splines will potentially contribute to loosening the nut. It's the same principal as fitting a loose key in a keyed shaft. As Bunny has now confirmed the fit between coupling and shaft does vary from car to car. Not good.

Rather like the bevel box saga it seem to be a combination of factors that decide whether an individual car suffers a problem. The common denominator being inadequate control of tolerances and assembly.

I'm also not a massive fan of threadlock etc. but there are several mechanical means of locking a nut in place, one of which may be suitable.

One of Mr. Bleazeys recommendations is to lube the rollers. I had never done it until I was able to do so with the Bleazey conversion and I must say that it certainly makes a difference in the way how smooth it is now on take off. Really glad I tried doing it and even if you put to much lube on the rollers I personally do not see how that can hurt, other than throwing lube inside the bellhousing.

As we all know, in the beginning, our M3W’s had the HD compensator, which didn’t work well due to the lack of lubrication as this construction is designed to run in an oil bath.
MMC needed to find a solution to this problem.
Probably they contacted the engineers at Centa Couplings Germany (the leading company in the world for transmission couplings (see their website www.centa.info). Centaflex-R looks similar to ours.
But Centa was limited in the shape of the solution due to the form of the bell housing-splined crankshaft-clutch etc.
On top of that came the heat inside where the coupling needed to do its job
(we all know what the heat can do because the cam belt also couldn’t stand the heat, and after the ventilation slot in the cover, the belt problems were 99% solved).
The original Centa rubber rollers probably can’t stand the heat as well and break up after a while due to the high temperature.
MMC and Centa didn’t foresee the problem of rubbers breaking up.
And it is a real problem as it’s impossible to change the rubbers without taking the engine out.

Phil Bleazey came up with a perfect solution to easily change the rubbers by turning the Centa and making a giant hole in the bell housing.
He also designed new rollers which wouldn’t break up as quickly.
This is what we now discuss in this treat: why does the big crankshaft nut come undone?
Is that because of the stiffer / smaller rubbers or not?

Now I’m going to give you some more thoughts.
Centa designed the unit in a way that it’s locked on the crankshaft by a HALF NUT - A STAR WASHER and a FLAT WASHER.

When watching the videos from Phil, he stated clearly that the flat washer combined with the star washer is stupid. As it ONLY holds the nut in place against the flat washer. This all looks correct what he is saying.
But that’s only correct if the underlying piece absolutely can’t move! Not a thousand of a millimetre.

Let’s assume the tight fit of the splines is never 100% possible (in that case, you would need to put the part on with a hydraulic press).
So there will always be an extremely slight movement (thousands of a millimetre) from left to right (turning wise) from the Centa over the crankshaft.
This flat washer allows the Centa to do that, as this doesn’t bring this movement over to the star washer and the nut.
That there is movement, you can see on the picture, which shows the imprint of the Centa splines (you see the light polished footprint of the splines, which means it was rubbing as this gives a shiny surface).

As we take this flat washer out, this slight movement will be transferred from the star washer to the nut. And undo the nut.
With only the flat washer, this slight movement will also be directly transferred to the nut.
Because Phil delivered his upgraded Centa with a FULL nut, we don’t have enough place to use the star washer and the flat washer.

On top of that, if we use now as well stiffer rubber rollers, the shocks between the splines will be more violent and have more impact on the star washer and nut.
This will result in the nut coming undone more easily.

I’m not an engineer, and I don’t say all the above is correct, but in my opinion, changing the rubbers and missing one of the washers could be the culprit of the problem we created.

Food for thoughts?

I doubt that Centa designed the splined join from Centa to crank on the M3W - I would expect that to have been Morgan's responsibility (although they may have outsourced that ... ). I would also suspect that any slight rotational movement that undoes a full nut would also undo the nut via a star washer and flat washer... (but I'm also not an engineer!).
Intrigued by the theory that the different (PB) roller types are causing the shock effect to either undo the bolt or mash the spline. But if that was the case I would expect Phil's own car(s) to have been affected!

If you mean the clients cars that Phil has upgraded, then yes they have been affected as well as other cars upgraded by various 'others'.
It seems Its all a matter of mileage since upgrade and whether you take any notice of yet another mechanical noise from your M3W...

M3W comes back tomorrow from M3W services in Montignac tomorrow with a new PB compensator fitted but with the softer standard rubber. Before putting away for the winter, I will probably treat them to some silicone grease.

I wonder if the solution to the loosening nut, might be to mount the PB compensator on the spline using Loctite 660. This product is specially designed to take up any minor mismatches between the mating surfaces of male and female splines. Many years ago, I had a similar problem on my Ralt-Judd F2 hill climb car where the rear drive shaft nuts kept coming loose. This became much worse, after I fitted 14" wide F1 OZ wheels for traction, in place of the original 12" wide plus very sticky Hoosier tyres. Loctite on the nut threads did not help and the locking wire just sheared. I used the forerunner of Loctite 660, called something like Mounting Compound between the male splines of the drive shafts and female splines of the rear peg drive hubs. Problem totally solved. It was exactly the same problem, very large torque reversals shaking the splines in the hubs and loosening the clamping nuts.

Wilson