It occured to me when I last had my bevel box off, that the centre line of the universal joint is on the centre line of the swingarm (and, of course, that's where you would expect BMW to put it).

This would mean that the rear splines do not slide as the suspension moves, which is the commonly quoted reason for the splines to wear.

To confirm this hypothesis, I set up a dial gauge yesterday and moved the swing arm as far as it would go in each direction and, sure enough, there was no flicker on the needle. The end of the shaft is nicely supported in my oil seal mod. I'm sorry, I can't do videos....too complicated.



Now I've only done about 2K miles since I last greased the splines and although they were still quite 'wet', most of the grease was spread around the inside of the housing. As I now know that the two parts of the drive don't move with respect to each other, then it would be feasible to put an 'O' ring over the male spline which would seal the joint and keep the grease in the right place.

If my oil bath mod doesn't work for any reason, then I'll try this next.

Do you think I need to see a shrink?

Paul.

No, but I think you possibly are overseeing something

The shaft movement on the FD spline is not an effect of the swing arm bearing/universal joint alignment and therefore you don't measure a real difference in the way you did it. If you would turn the shaft around I think you would notice that there is axial movement at the rear end. I think that the forward end is fixed on the splines with a ring to ensure that the universal joint is (as much as possible; its not a watch ) aligned with the pivot point of the swing arm. Therefore any axial movement as a result of the turning of the universal joint will result in slight axial movements on the FD shaft splines. Therefore you rarely see wear on the forward end, but always on the rear.

But that's what I'm thinking :cyclops:

Sorry Rene, I don't agree with that. The UJ pivots about the centre point of the 'X', which is on the centre line of the swing arm, so it doesn't matter if it's stationary or rotating, it still doesn't slide on the bevel box spline.
The only movement could be from the play at the circlip in the gearbox end of the shaft.

I reckon the wear is caused by the variation in speed between the input and output of the shaft, which is magnified as the angle of the shaft increases.

With two UJ's, as the paralever, this variation should be cancelled out - provided that they are set 'in phase'.
Do paralever splines wear as much as the monolevers?

I guess it's all academic really, but it's fun to sumise on what's happening while we're bowling along at 80ish.

Oh, and a K is better than a watch, you can ride as well as telling the time!

Cheers,
Paul.

Sidecar Paul,

You just opened a can of worms and I like it.
but....
The pivot for the swing arm is not located at the same position of the universal joint and consequently, Mr. Pythagoras should be the one giving us the real answer.
I personally believe that he will prove you wrong, but it was very nice to see you installing an indicator to measure the result.
Now, you and ReneZ get me interested, and I will take measurements and replicate the setup on CAD to measure the reality against the theory.

Yeah! The grey cells will be smoking again!

Cheers and thank you for sharing your experience.

CF

Hi there, I had been thinking on the subject since learn of the spline problem a few years ago and was of Paul´s opinion, but always thought that was missing something.

After reading your opinions and learn that the ujoint is not at the same position of the swing arm, I could grasp the answer: the shaft and the swing arm have different rotation centers, that will cause the axial movement at the spline, otherwise the swing arm and the shaft would create and indeformable triangle. Even if you try to put the joint at the arm´s rotation center any small fabrication error would cause and unbearable stress.

On the other hand the cyclic variations between angular speeds of the shafts concurring at the ujoint are a mechanical fact and that is why usually you´ll find two as stated by Paul, that also explains (at least in part) why our shafts have two concentric elements joined by rubber filling the anular space.

All said I think that the cause of the spline wear is a combination of the axial displacement, the torsional stresses caused by the use of a single ujoint and a poor design regarding sizing and lubrication.

As far as I know paralever system has a better performance and other monolevers are less prone to failure due to a better design.

So Paul may succeed in improving the original design and I´m sure that we´ll be attent to his progress, willing to learn and help if needed.

Best regards

JAP

Interesting discussion!

Paul, I don't see how one end of the shaft can have a different rotational speed than the other end as both are mechanically fixed through the UJ.

My reasoning: the wear on the splines comes from rubbing, which would mean axial movement in that area. If 'backlash' would be the cause you would see more prove of material displacement by 'hammering'. That rubbing can only come from axial movement of the shaft. The only variable point in the shaft is the UJ, so I feel the movement has to be generated there. Experience has shown that there is no real wear on the gearbox end (quite often the shaft is almost rotten tight there) so the circlips is working.

Please let me know your thoughts?

BTW, the paralever has 'no' wear on the splines we are discussing as both ends are fixed with a circlip. The shaft is a two piece arrangement with a substantially bigger spline-type transfer between the two UJ's. However alignment of the two UJ's seems a serious concern with those I understand.

Regarding the cyclic variations between angular speeds of the shafts concurring at the ujoint, there is an explanation here.

Best regards

JAP

I thought it was the difference in materials the F/D being unable to expand as much as the tube it fits into. That bloke who adds another half length to the F/D splines has it sorted though I am wondering about movement back there now that you've all mentioned the difference up front, ever so minute as it is, flex in the rubber bit cant be enough can it??

I quickly looked at a spare gear box but don't have time to hunt for the spare shaft (it's 6:30am) I may do that tonight and will take measurements to see the relation of the center of the UJ and the pivot of the swing arm.

CF

Hi JAP, had a look at the link but my knowledge of English is not sufficient to understand "the instantaneous angular displacement of the first shaft is the same as that of the second shaft only at the end of each 90° of shaft rotation. Thus, only at four positions during each revolution is angular velocity of both shafts the same."

However, wear occurs when surfaces rub, not when they push harder or not against each other. For me that proves that there must be axial movement in the shaft.

Hi Rene, certanly the explanation can be a little confusing, but the result is the one you state ( your english is good enough), the surfaces are pushing against each other, but the pushing is cyclical and so are the resulting stresses (nornal stress), this hammering effect can reach high values and cause plastic deformation (more gap between the surfaces) that leads to more intense hammering in a vicious circle.

But not only the spline has to deal with that, as you state, at the same time the axial displacement causes the surfaces rubbing and this sum shear stress to the normal stress, increasing the wearing.

Of course you can design knowing this and give more area and better lubrication to the splines to keep the stress low.

Best regards

JAP

Sorry to get you up so early CF, is there really a 6.30 in the morning?
I'll look forward to your results, but my dial gauge didn't flicker, so how do we explain that?

"Buy a new dial gauge"......no that's not the right answer!

japuentes, thanks for your input, the hammering effect is, IMHO, the most likely cause of wear.

(I do love this forum)

Paul.

Hi Paul, if your dial gauge did not show any displacement, your shaft has practically the same rotation center than your swing arm, so the eccentricity effect being minor than the dial gauge resolution.

Best regards
JAP

Paul and JAP, I appreciate the consideration of the hammering, however when you look at worn splines they show little plastic deformation which I would expect to see as a result of the hammering. Hammering displaces material, it doesn't grind it away.
Paul, if you don't measure axial movement with the shaft at angle I'm surprised, as the force in the UJ 'wants it to break out' and therefore I would expect movement, even with the rotational axis in line. Maybe it materialises more at speed and not (so much?) at a stationary situation? On the other hand, why would they fit a rubber vibration damper in the shaft, which is more suited for radial vibration than axial.
Interesting, reverse engineering! This is the reason why I still like my job (and like tinkering with older bikes!).

Hi Rene,
My engineering is based on what I can measure and what I can see is happening rather than theory. Maybe there is some movement at speed but I can't see where it would come from.

I think the rubber damper is supposed to minimise the hammering caused by the shaft speed variations.

CF,
Did you imagine, when you started this forum, that years later, you would be seaching for K parts so early in the morning to satisfy the whims of someone thousands of miles away?

Best wishes to all, I'll be back on Monday,
Paul.

Difference between the co-efficient of expansion between the steel shaft and the aluminum housing??

Moto Smith wrote:Difference between the co-efficient of expansion between the steel shaft and the aluminum housing??

Aluminium´s thermal expansion coefficient is almost twice than steel, but this differential elongation is taken by the spline with no stress.
Best regards
JAP

japuentes wrote:

Moto Smith wrote:Difference between the co-efficient of expansion between the steel shaft and the aluminum housing??

Aluminium´s thermal expansion coefficient is almost twice than steel, but this differential elongation is taken by the spline with no stress.
Best regards
JAP

But would it effect the proposed oil seal? Or at least the style of seal required.

Moto Smith wrote:

japuentes wrote:

Moto Smith wrote:Difference between the co-efficient of expansion between the steel shaft and the aluminum housing??

Aluminium´s thermal expansion coefficient is almost twice than steel, but this differential elongation is taken by the spline with no stress.
Best regards
JAP

But would it effect the proposed oil seal? Or at least the style of seal required.

I Don't think so is a difference of less than 0,7mm for a temperature change of 100ºC