Clutch Spline Failures

[Boxer]

With close to 45,000 miles on my Roadster, I have started to worry about all the tales of clutch spline failure. Over at the SportTouring site someone just has his crap-out (possibly-no report back yet) at 17K. Someone reported that as early as 10K or 8K these things have failed.

Could some people in the know mechanically, please address this possibility, and what causes it, preventive maintenance, etc.

[riceburner]

oddly - this just happened to my Rockster. The Gearbox input shaft splines on the clutch side went. Just as I was exiting a roundabout and changing up a gear, so not under any real stress (compared to usual. )

Was replaced within 48 hours.

Mileage was at 44751. Seeing as I hope to be riding this bike for at least 200,000 miles (well, 30K a year and I hope to keep it for a few years) I hope it doesn't become a regular thing....

[boxermania]

Hmmmm.....if that is a common occurrence, someone needs to look at the mettalurgy of the input shaft.......it appears to be soft....soft...soft.....

Just my 2 cents.......

[riceburner]

Hmmmm.....if that is a common occurrence, someone needs to look at the mettalurgy of the input shaft.......it appears to be soft....soft...soft.....

Just my 2 cents.......

well - so far we've heard of 4 (I think) - out of how many units produced??? That's going to be less than 0.001 - it's not worth BMW worrying about unless it more than 1% of units produced. That's business.

[Martyn]

Hmmmm.....if that is a common occurrence, someone needs to look at the mettalurgy of the input shaft.......it appears to be soft....soft...soft.....

Just my 2 cents.......

well - so far we've heard of 4 (I think) - out of how many units produced??? That's going to be less than 0.001 - it's not worth BMW worrying about unless it more than 1% of units produced. That's business.

I know 7 such failures in my local area, although 5 were on RT's. My dealer also had all the required parts on the shelf ready to go, & turned the job around in the next 3 working days, which mollified me a little - but only a little.

[MikeCam]

RT's have a different angle of the shaft than Roadsters. Not sure about Rocksters. No R1100S's reported abnormal spline failures (Pelican...).

In 2003 when this was a big concern across the boards, there were several likely candidate reasons for failures. One was the rider and weight combinations were too taxing for the bike (RT/KRS) given the higher state of tune (torque) and the usage profiles of those with failures. In short, riders overloading and/or over riding their bikes.

Not to say that is the culprit...just to point out it is not necessarily a BMW engineering denial.

As pointed out above, overall the numbers are still less than 1/2% of the line.

[jm1515]

With close to 45,000 miles on my Roadster, I have started to worry about all the tales of clutch spline failure. Over at the SportTouring site someone just has his crap-out (possibly-no report back yet) at 17K. Someone reported that as early as 10K or 8K these things have failed.

Could some people in the know mechanically, please address this possibility, and what causes it, preventive maintenance, etc.

The very knowledgable crew at
http://bmwsporttouring.com/index.html
BMWSportTouring.com
discuss this all the time. Basically on the RT, of course, but it sorta scares me.
'Preventative maintenance' (basically the same work as a clutch etc replacement) looks to require tearing so much off the bike I don't know if 'maintenance' is the right word...more like 'surgery'

I've only got 9000mi on my 04, but I have serious thoughts about maybe having the dealer check it out before the warrantee expires. Been stuck waiting for a tow twice...it can be a real PITA.

[Paul Mihalka]

I bought my '02 R1150R this year in March, with 6K miles on it. The clutch/input shaft spline failed at 30K miles in between cotton fields in eastern Arkansas, this September. It cost me $900 in UHaul and $400 gas to get us home (MD). General concensus is that if the splines fail early, like under 50K miles, it is some alignment problem, most of the times in the gearbox housing. My bike got a new gearbox and a new clutch assembly. Bike was out of warranty, but BMWNA helped some with "Goodwill" warranty. I hope it is fixed for the long term. I plan to take it apart every 30/40K miles to inspect and lube. For me that's about once a year

[12bar1]

Are there any symptoms prior to failure or does it just take a dump

[r1150rider]

I'm almost at 51,000 miles on my '02 "R", and this topic is in my mind every time I go on a ride. My plan is somewhere around 55k-60k to have it and the clutch serviced/replaced and the "tall" 6th gear swapped for the "short" one.

The service manage at my dealer says that's about all I need to worry about at that kind of mileage, I figure it'll then be "good to go" for another 30k-40k or so, (hopefully). By then perhaps the R1200R will have materialized and be past it's "first year" issues!

[Paul Mihalka]

Are there any symptoms prior to failure or does it just take a dump

Usually difficult down-shifting is a sign of worn or too dry clutch splines. I still think that at +/- 40K miles a clutch spline lube/inspection is a good idea.

[Biff's R]

Are there any symptoms prior to failure or does it just take a dump

Usually difficult down-shifting is a sign of worn or too dry clutch splines. I still think that at +/- 40K miles a clutch spline lube/inspection is a good idea.

My Dad had trouble downshifting on his '96 RT at about 40k miles, and trouble downshifting seems to be one of the symptoms of spline troubles. But This was on a bike that was almost 9 years old at the time.

[CycleRob]

Initially, I would have said it's an input shaft sloppy heat treating problem, but gearbox misalignment makes more sense. The fact that mostly heavyweight Beemers are failing supports that. Niot so much the wild riders, but the sustained high speed cruisers that the splines are failing. With even a slight misalignment of only .050" (1.27mm), the male/female engagement splines will be sliding axially in and then out of their slots, once per turn. At 4,000 RPM, that steel is effectively sliding 400 inches (10.16 meters) a minute. That taxes the factory grease until there's only burnt residues of it left. Then the engagemment surfaces get really really hot until the shaft is annealed. The expensive, perfect factory heat treat is now history. It's that day, that ride, that the splines "fail" from overheating. I bet the failed parts are blue/black/gold from the overheating.

This type of misalignment is no stranger to Jeep vehicles where the clutch chatters and grabs on an easy take-off. If your bike does that . . . . it may be a clue of things to come . . . . .

[cc]

Initially, I would have said it's an input shaft sloppy heat treating problem, but gearbox misalignment makes more sense.
. .

I'm not sure I understand what you are saying here. Maybe because its late.

Are you talking about the tranny input shaft not being co-axial with the crankshaft? If so, is that an angular mis-alignment or is the tranny moved (translated) in a plane perpendicular to the axis of the crank?

You may remember that BMW had a bunch of tranny input spline failures 25-26 years ago on airheads. I'm not sure anybody really figured out the cause of those failures, but that's when BMW decided that we should all pull the gearbox and lube the splines every 20K miles. Later they changed it to 40K miles. On my 1980 I did it about every 60-80K and it went 190K with no failure, then I sold it.

[CycleRob]

cc,

Angular misalignment is possible. Translated misalignment is more likely given there are at least two manufacturers involved in the mating parts. For the visual perception impaired, a translated misalignment would happen if you loosened the transmission mount bolts, let the transmission slide downward a small amount (without using the alignment pins), and then tightened the bolts. The rotating axis of the clutch splines, mounted on the crankshaft, would still be parrallel to but above the input shaft axis and the splines would eventually eat each other from the axial offset.

Maybe it is a crappy heat treat because they are failing at low miles too.

This is an easy problem for a concerned mfgr to solve.

[boxermania]

OK, this is an interesting topic, let's do some troubleshooting

A poster in the BMW Sport-Touring site listed the following as potential causes.

[quote}
1) Inadequate lubrication of the clutch/input shaft splines at the factory and/or poor quality lubricant.
2) Faulty parts - poor heat-treating of the input shaft or clutch disk on some machines.
3) A mis-alignment of the clutch housing and transmission such that the clutch plate does not make even contact with the input shaft splines.
4) Bad design. The input shaft and/or clutch plate are not supported adequately, and depend on perfect alignment of the cases to do their job properly[/quote]

Based on my experience with these type transmissions I can venture the following in response to the above.

1) Typically input shaft splines do not require any lubrication, the splines must be clean however. There is very lttle axial motion and ample radial clearance between the shaft and the female splines on the disk. Unlikely for this to be the mode of failure.

2) With the limited failure information available, this would be a likely culprit, however I would guess that as soon as the failures started croping up a responsible OEM would have gone and checked the metallurgy of the shaft and made an on the fly correction if required, Since the failures continue over several model years this is also an unlikely culprit.

3) A misalignment of the transmision to the engine case, also mentioned by CYCLEROB is multifaceted and requires additional investigation. Assuming there is misalignment, is it axial in nature (the centerline of the engine crank and the input shaft are offset side to side) or angular (the vertical plane of the transmission housing is not parallel to the vertical plane of the transmission mounting on the engine side). I'll have to place this item in the same category as #2 above since it would have been noticed early on and a change made to correct it.

Additionally if there is some flexing going on at the juncture between the shaft and the clutch splines I would have expected the clutch disk to fail first, since it would be operating in a bind and the metal between the clutch spline and the friction material is way thinner than the transmission input shaft.

4) Bad design. In response to the poster, the end of the input shaft is machined down to fit into a pilot bushing or a needle bearing in the back of the crank. There is a clearance in this area as well, they need to be able to turn freely from each other.

So, we have come around a full circle. At this point the best thing to do is to look at some of the failed pieces.....can anyone out there have one or can make arrangements for taking pictures of one. The pictures should be taken perpendicular to the area of failure to look at the way the metal separated.

In closing, my heart still tells me there is a metallurgical issue that BMW has chosen to ignore.......

[jm1515]

OK, this is an interesting topic, let's do some troubleshooting

......So, we have come around a full circle. At this point the best thing to do is to look at some of the failed pieces.....can anyone out there have one or can make arrangements for taking pictures of one. The pictures should be taken perpendicular to the area of failure to look at the way the metal separated.

In closing, my heart still tells me there is a metallurgical issue that BMW has chosen to ignore.......

http://bmwsporttouring.com/ubbthreads/s ... art/1/vc/1

http://bmwsporttouring.com/ubbthreads/s ... art/1/vc/1

http://bmwsporttouring.com/ubbthreads/s ... art/1/vc/1

Link to some pics and a good discussion...

[wncbmw]

I have only had one failure in five BMWs that prevented me from making it home and that was clutch/spline failure in a K75. There were some symptoms beforehand but not enough for my feeble mind to notice until after the fact. Fortunately, I was fairly close to home with a group that included friends with trucks and trailers willing to go home and come back for me.

If there was maintainenace items that could be done to prevent the same thing happening while alone and far from home, I would certainly look into it.

[CycleRob]

Here are the best bmwsporttouring.com responses to the problem:

I hope you have better luck than me. My bike has a similar problem that BMW has tried to fix 7 times over pretty much the entire life of the bike. The input splines are destroyed in as little as a few hundred miles of riding. Once it destroyed the clutch disk.
In all but one case BMW fixed it but under duress. The last case they replaced the gearbox and the symptoms returned after a few months. I had to disassemble the bike 2 years ago and diagnose the problem for them: The engine and gearbox are not aligned and the bike was assembled at the factory with one of two engine-gearbox alignment pins missing!
It is pretty obvious to me at this point that BMW is not capable of solving this problem once and for all. Yesterday I started yet another disassembly (looks just like your photo) and I should have the gearbox off the bike later this morning.
If your bike gets fixed let us know if they have a true diagnosis and also if it continues to work for a long time. As I said before mine failed in as little as a few hundred miles and in as much as 30000 miles.

When my shaft/clutch went, I thought it a bit of a clue that my dealer had all the parts, including the input shaft, "on the shelf" - I asked if they had similar spares in stock for K-Series or air-heads, to which the reply was "no"

I've modified my gearbox cover and engine/gearbox alignment pins to bring the crankshaft and gearbox input shaft into alignment. When I started they were off by about 0.005" and now they are less than 0.002". Tolerance limit for DIN 5480 splines is 0.003". I'm presently re-assembling the bike and will report back on results.
Seeing how far out my alignment was makes me suspect that this is why some bikes have early spline problems. Mine started at 14000 miles and has persisted to 70000 after 7 repairs. (Yes, I could probably write a book on this saga). If mine was off by 0.005" it is hard to believe that is the only bike ever built that was not in spec.

As a mechanical engineer with driveline experience, I've been mulling over the last year of posted reports and pictures of oilhead transmission input spline problems. I don't want to take apart my 2000 R1100RT (yet) but I have a theory that needs your observations to backup or refute this theory:
The oilhead is unique in that unlike every other manual clutch vehicle I'm aware of, it has no pilot bearing in the flywheel to support the cantilevered transmission input shaft and the clutch disk.
Thinking as an engineer, maybe someone thought the pilot bearing is redundant - thinking the input shaft and clutch disk can be supported on just the transmission bearings only.
BUT thinking more as an engineer having had more than a few design screwups in my career - what may have been missed is that there is now nothing keeping the pre-engaged clutch disc centered on the flywheel axis except the transmission bearings. As long as the clutch takeup torque is uniformly distributed around the flywheel face, nothing should be needed after engagement.
However, if there is a non-uniformity in the clutch disk friction characteristics across the flywheel face, there will be a radial load locked into the clutch assembly on engagement and imposed on the input shaft which is reacted by the crankshaft rear main bearing. The transmission input shaft will then flex and the spline would encounter one-per-rev angular fretting, causing the spline to wear in a barrel shape.
If that same non-uniform clutch assembly included a pilot bearing mounted in the flywheel, the pilot bearing would stiffly react any radial force component from the clutch disk and maintain alignment. Once engagement is complete, the pilot bearing would see this only as static radial force since both the pilot bearing's inner and outer races rotate together with the clutch assembly. The spline sees little of that radial load – even though the rotating assembly has a locked in rotating static load. The spline is not being flexed. The pilot bearing sees the radial load once only per engagement. All are happy.
On the other hand, if as in our case, there is no pilot bearing, any non-uniform clutch disk engagement puts a radial load on the spline, and causes the input shaft to flex once per rev, and fret the spline surfaces. The fretting continues until the clutch is disengaged, which remember, may be a long ways down the road. Each time the clutch is re-engaged, a new alignment problem is presented to the transmission input shaft bearings, and a new one-per rev angulation motion is applied to the input spline - until the next time the clutch is disengaged.
What might cause the clutch disk to be non uniform in its torque transmission? 1) non uniform thickness of the clutch disc, 2) non uniform springs in the pressure plate, or 3) non uniform friction on the flywheel face or clutch disk surface, which might be caused by trace amounts of oil, grease, or corrosion. Without a pilot bearing, any non-uniform characteristics on this assembly could cascade into a spline wear problem. This could be caused by something as simple as excessive grease or oil in the area being dropped on the flywheel or clutch disk on assembly.
At first the barrel shaped wear pattern fooled me into thinking there must be an alignment problem between the engine and the transmission. Now, I doubt that, as excessive production tolerances for this type of thing would be quickly identified in most quality manufacturing operations.
As riders, what can be done to minimize spline wear? Maybe reclutching and a floating re-engagment when speed is reached before any long run. This is certainly not practical around town, and not real convenient, but it could minimize the radial force applied to the spline by the clutch disc to the input shaft.

I like to take your theory one step further: No only uneven friction characteristic but also a not properly balanced friction plate could cause input shaft flexing: When disengaging the clutch at high (RPM) the friction plate is only supported by the input shaft. If there is an unbalance the shaft will flex and this condition might become frozen once the clutch is engaged again. As the splines wear some play between the clutch and the input shaft will develop and aggravate the situation.

I have seen two "problem bikes" that were fixed by replacing the front section of the transmission housing (along with the clutch friction plate and input shaft and bearings). They were both having the recurring "stripped input shaft spline" problem at 29K intervals.
Knowing this, I still believe it is an angular mis-alignment at the tranny to engine interface causing this particular barrel-shaped wear and the eventual spline failures.

It seems hard to believe that a part as basic as an engine-transmission housing would be manufactured wrong (especially in Germany and especially over a period of years). Also, it is weird that one of the posters had a definite skewed wear pattern on his flywheel. I have also experienced the different rattle in a neutral clutch re-engagement that was noted.

[boxermania]

Drawing upon my experience in the failure analysis arena and upon review of the posted material, along with the pictures I can draw the following mode of failure.

First the input shaft does not follow typical automobile transmission practice, of having a pilot bushing to engage at the end of the crank. That being said picture not only the weight of the clutch disk, but also the centrifugal forces due to rotation when the clutch is disengaged, along with the torque of the engine upon demand. We have an overhung moment......

Second, pay close attention at the wear pattern on the input shaft. It's very consistent from tooth to toth but inconsistent from front to rear. Notice that at the rear of the shaft yspline there is a depression in the tooth surface, that means that the tooth is a highly loaded area. Why there one asks.....

Also notice, when looking at the splines on the disk, that the disk doesn't ride full up on the shaft splines. This indicates a potential problem, the designed torque carrying capacity is being compromised since there isn't full spline contact. However of more significnce is the ability of the disk, let's say, to load up against the back of the spline on the input shaft.

What that tells me is that the disk, during it's operation, is "rocking" towards the rear of the spline. The only way that I can see this happening is: Excesive clearance between the clutch splines and the input shaft splines, throw a pinch for lack of full spline contact and then add the canteliever operation of the shaft and you have a textbook case of a failure in the making.

Now, why does it happen across all the models but is predominant on the RT. Well it's simple, the RT is heavier and most will carry additional loads, two up and luggage. Additionally, to ofset the added weight, the RT has been fitted with a shorter rear ratio, numerically higher, which increses the torque load on the rear and the tranny. To this you can factor a smidgeon for potential abuse......and there you are.

In my line of business we have a motto, fix it right the first time, that is because if you don't the issue is not if is going to fail again, but when......I feel for those of you that have to endure this experience.

When everything is said an done, the input shaft design does not follow common industry practice and certainly it has proven it's lack of reliability. The thing that chaffes me is the blatant disregard by the OEM in recognizing the problem......something that we the owners have experienced more than once with the marque.