More about spline failure and a machinists solution
 

Way over my head, but some of you might understand it.

www.bmwsporttouring.com
Forums » 'RENCHING » Oilheads » 2002 1150RT CLUTCH SPLINE REPAIR

Easy enough to understand the theory of it but tough for the majority of riders to accomplish. And completely outrageous that anyone should need to do this.

However there does seem to be a fundamental difference in the RT design. I have had the issue on my S but it doesn't seem to be as rampant a problem for us.

I admire the guy for his skill and effort, but it's a shame the customer is doing what should have been done in the factory.BMW knows the splines are a potential problem regardless of model.That's why I'm looking at the F800.

I could be hopelessly wrong, God knows I have been in the past, but I'd be hard pressed to believe a 0.01" or 0.2mm misalignment is enough to destroy the clutch splines. That's not a whole lot, and if it is that critical I'd say the entire clutch/gearbox design is faulty.

that's ten thou, not very precise if you ask me.
they index chebby v8 bellhousings to .0015, so that's a LOT closer over a much farther distance.
I believe it's a combo of things; a mis-indexed bell housing and marginal heattreat on the shaft combined with gruel for spline lube.
Ya should'a seen what BSA used for generator rotor bushings (a strip of leather).

BSA got worse than leather bushings, much worse. On the triples, a Borg & Beck dry clutch was used, with the chaincase cover holding a thrust bearing that wouldn't last 2k miles before turning to powder. Talk about misalignment and the stupidity of using the chaincase cover as a stressed drivetrain component. That, combined with a crappy clutch pull mechanism, meant you'd be lucky to get 1k miles on a clutch cable. The cable was also a real pain in the butt to change out. Anyone who rode a triple any distance developed a left forearm like Popeye because the lever pull was so stiff.

Thanks for the URL. Something similar with great pictures appeared here too.

The measurements and the reasoning once seemed very cogent to me. But I am not as sure now.

The nature of this "joint" and a rinky-dink part like a clutch disk are pretty unfriendly to high accuracy. In many car clutch plates, you have two parts separated by springs and that makes so-so accuracy more understandable and perhaps more tolerable. Not to mention the mess of loose pieces in a wet clutch.

There has to be some "orbital" movement. When lube goes bye-bye, then there is wear. But that wear might be only to the extent of the misalignment leading to the orbital motion and/or already the same order of looseness as the factory spec'd sliding fit.

I'm not sure.

Yes, good thread. Where I work we disassemble/reassemble components made in Germany. We see a lot of sloppy work from the other side of the Pond (in CNC components, no less). Mostly I think it is using indicators in mm not inches. At .001 inch you can easily split it to .0005 by eyeball. In mm .001inch is about .0254mm. So 1/2 of that is .0127mm, a weird value to get on a dial. In theory the metric system should give you more accuracy because the dials should go out to .01mm. In reality I don't think the European people use/exercise the accuracy available.

now this makes sense.

Onr in Simple English;
There ain't no mo' Old World Craftsmanship..it's all a guest worker world now:)

What Jony said. I'll bet a large fraction of BMW workforce in Germany are of arabic or african origin. But what the heck, BMW cage has a factory in South Carolina!