Spline Design Question

[pmdave]

It's pretty obvious that BMW's K bike spline design has some problems. Here's a shaft drive machine at extra cost--so we can (it is hoped) ride forever without worrying about drive failures. Yet, the only way to achieve halfway decent reliability is to disassemble and relube the driveshaft every 10K - 30k miles (depending on whether you use the BMW lube or Honda moly lube, or your own secret blend of chemicals and bats wings)

When I'm driving a car, I never worry about driveshaft problems. I just drive it. If something needs regular lubing, there's a grease fitting and they give it a squirt during a lube job. It's only at the higher mileage intervals that even the 4WD systems need disassembly and relubing.

So, why do the K splines wear out? I'm thinking it's related to the machining tolerances. Dig out the January 2005 issue of ON, flip to page 25, and notice that both the male and female splines in the photos (Paul Glaves article) are worn halfway down the flank of each spline. Doesn't that indicate that the actual contact between spline faces is only 50%? Of course there must be enough clearance to get the parts together, but wouldn't it be better to have say, 80% or 90%?

And, since the shaft is free to move around, the splines are free to move radially within the available clearance. Seems to me that would add some additional wear from the radial sliding--at least until the splines wear into that step shape. Then they would be popping and growling as the shaft jumped around from deceleration/acceleration.

So, when Bruno does his machine work on a spline, does he resize the splines for better contact?

And, if the new K12 bikes have the same sort of spline tolerances, won't they wear out just as fast from radial sliding as from the longitudinal sliding the new design is supposed to eliminate?

If this subject has already been covered, say so, and we can avoid another tedious repetition. It's just that those photos bother me.

pmdave :type

[manicmechanic]

I've only pulled a couple automotive driveshafts, but it seems the splined end is usually running in a seal at the tail of the transmission, and whenever I've removed the shaft some fluid came out. So would it be a far step to conclude that the splines are lubricated in this instance? Granted, not a lot of fluid, maybe a cup or less, but some fluid. Even on vehicles with standard trannies there was usually a little fluid evident.

[deilenberger]

It's pretty obvious that BMW's K bike spline design has some problems. Here's a shaft drive machine at extra cost--so we can (it is hoped) ride forever without worrying about drive failures. Yet, the only way to achieve halfway decent reliability is to disassemble and relube the driveshaft every 10K - 30k miles (depending on whether you use the BMW lube or Honda moly lube, or your own secret blend of chemicals and bats wings)


Very true, it does seem a paradox on paying more and not getting more.



When I'm driving a car, I never worry about driveshaft problems. I just drive it. If something needs regular lubing, there's a grease fitting and they give it a squirt during a lube job. It's only at the higher mileage intervals that even the 4WD systems need disassembly and relubing.

So, why do the K splines wear out? I'm thinking it's related to the machining tolerances. Dig out the January 2005 issue of ON, flip to page 25, and notice that both the male and female splines in the photos (Paul Glaves article) are worn halfway down the flank of each spline. Doesn't that indicate that the actual contact between spline faces is only 50%? Of course there must be enough clearance to get the parts together, but wouldn't it be better to have say, 80% or 90%?


I don't think it's quite as simple as that. The rear drive wear is caused by several factors on the early monolever K's:

1. Sliding engagement - so one end of the spline will always be in contact, and the other end only intermittently, if you think about it - the end always in contact will wear more than the end that isn't always in contact.

2. Lack of a universal or CV joint near the rear-drive end. Wear of the rear-drive splines on the paralever equipped bikes (which have two universal joints in the shaft) is a non-issue.



And, since the shaft is free to move around, the splines are free to move radially within the available clearance. Seems to me that would add some additional wear from the radial sliding--at least until the splines wear into that step shape. Then they would be popping and growling as the shaft jumped around from deceleration/acceleration.

So, when Bruno does his machine work on a spline, does he resize the splines for better contact?

AFAIK - and I haven't had Bruno do anything - he is using the basic stock BMW spline shape, but he makes the driveshaft female portion longer so there is more "engagement" between the two.



And, if the new K12 bikes have the same sort of spline tolerances, won't they wear out just as fast from radial sliding as from the longitudinal sliding the new design is supposed to eliminate?



One change they made with the new R bike designs is they put both universal joints on the shaft at the exact point where the pivot points for the swingarm housing are located. By doing this - they've completely eliminated any splines... so I'd have to guess there won't be a spline wear problem. I would expect that the new K bike will follow this design.



If this subject has already been covered, say so, and we can avoid another tedious repetition. It's just that those photos bother me.

pmdave :type

Best,

[nrpetersen]

As a mech engr, I suspect the spline life problems are due to 1) radial misalignments OR 2) radial loading.

Misalignments could be caused by a machining error.

Radial loading would be caused by the adjacent gear and bearing design, wherein the overhung gear is mounted to a shaft supported by a non-moment resisting bearing concept, such that the radial guidance of the spline is necessary to provide a second radial support.

Either way, the spline is having to work in a marginal or almost non-existant lubrication situation. Hardness of the surfaces would define the most important wear limit. The possibilities of one element being soft should be considered.

A single single-row ball bearing, for example, should not be used to resist overturning moments, and will allow the shaft alignment to shift around a fraction of a degree. Some times you want this to avoid redundant loading in a shaft system, sometimes you don't.

Someone above mentioned that K bikes with two U-joints don't have spline problems. That is what I would expect, in that each end of the adjacent shafts would have to be designed have their own moment support, rather than relying on the splines for radial guidance.

NRP