Splines ready to go - need advice on cure

[BucksTherapy]

I have an 02 with 30,000 miles. At 16,000 I lubed the splines and the infamous wear pattern was already showing. It was just out of warranty so I lubed it and put it back together knowing I would have to repair at 30,000 or so. It is getting hard to downshift into 2nd and I figure its time.

 

I'm getting ready to pull it apart and am debating on what cure to use. Economics are important but I want to do the right thing. I need advice. I've read a great number of very informative postings on this site and I think these are the options:

1 Replace clutch, input shaft and bearings.

2 Same as 1 but replace complete clutch assembly.

3 Same as above but replace transmission cover as well(assume a misalignment in the cover)

4 Replace transmission and clutcu assembly with a used one with a good input shaft.

5 Same as 4 but use a new transmission with a short 6th.

 

Has anyone done one of these and put significant miles on without another spline failure?

 

Any advice or thoughts?

[russell_bynum]

I would think that the accelerated wear you're seeing indicates some sort of mis-alignment. Maybe the transmission case?

 

Because of that, I'd lean towards using a different transmission. If I had an R1150RT, I'd look for one with the short 6th. You'll probably have more luck finding a good used transmission with the normal ratios, though...so if the tall 6th doesn't bother you, that's probably your easiest/best route.

[nrp]

Misalignment could be either/and/or the engine, the clutch bell housing, or the transmission front casting. I suggest you measure the runnout of the front transmission bearing housing bore with respect to the crankshaft rotation axis. Another posting a few weeks past showed a way to do it. He found about .005 inch runnout.

[Jim Moore]

As you are taking it apart, could you check the play in the clutch disc IAW this thread:

 

http://bmwsporttouring.com/ubbthreads/showflat.php/Cat/0/Number/617804/an/0/page/2#617804

 

It only takes a second and I'd really appreciate it. I'm trying to accumulate some data on spline wear.

 

Thanks,

[BucksTherapy]

I am just collecting parts now and will be pulling it apart in 6 weeks or so. I really like your idea and think is has great potential. I will measure the run out and then compare that to the actual spline wear. I will send you the info.

 

The simplest ideas are most often the best. Well done.

[BucksTherapy]

Another posting a few weeks past showed a way to do it. He found about .005 inch runnout.

 

I have looked for this posting but couldn't find it. When I lubed the splines at 16,000 I used the combination of a straight edge and a square to visually see if the input shaft was square to the tramsmision cover and that the clutch assembly was square to the engine. They appeared to be right on at any position across the housing I could measure.

This did not tell me if the transmission housing and engine block were square to each other and I could come up with no way to measure. In other words if the transmission cover and the input shaft were both off then they could be square to each other but not to the block.

 

In addition I did not measure to see if the flywheel was square to the block.

 

Intuitively, it would be more likely that the transmission cover would be out then the block/flywheel but there is no wat to know for sure. I can see why there are so many recurrances.

 

Thanks to all for the info. I now know more about what I don't know and can't find out and this knowledge leaves me more confused about what to replace. Ideally I think the best chance of correcting this is to find a used transmission and clutch assemply with no spline wear, replace the clutch and hope for the best.

 

Any thoughts would be appreciated.

[nrp]

Try this one

 

http://bmwsporttouring.com/ubbthreads/sh...3&fpart=all

[BucksTherapy]

Try this one

 

http://bmwsporttouring.com/ubbthreads/sh...3&fpart=all

 

Wow, machining the crankshaft extender is beyond my talent and tools. I wonder if the transmission cover or block was out of alignment. If I knew the fault was with the transmission cover I would replace it as well.

[Jim Moore]

I am just collecting parts now and will be pulling it apart in 6 weeks or so. I really like your idea and think is has great potential. I will measure the run out and then compare that to the actual spline wear. I will send you the info.

 

The simplest ideas are most often the best. Well done.

Thank Stan "The Man" Walker and the Cali Kid. I'm merely a mouthpiece.

[Stan Walker]

Perhaps the person who made the crankshaft extender would lend it to you?

 

Or post the dimensions so others could duplicate his tool?

 

It wouldn't hurt to ask.

 

Stan

[nrp]

The refernced post made a dummy crank extender & measured the clearance to the bearing bore with a telescoping gage. That's a hard way to do this & has its own potential for errors unless it is done VERY carefully.

 

What's really needed is a dial indicator setup that is jury rigged to mount onto the crankshaft to indicate to the transmission input bearing bore while the crank is rotated. It would be neat if this group could make one, & I'd even be willing to cobble it together as a kit, but I don't want to tear apart my RT (yet)just to figure one out.

 

Any machinists out there with or near this problem? Or can someone sketch what the cross section looks likes? I do have metal working capability & a dial indicator.

 

Bear in mind that a wholesale replacement of parts will also have to include the engine block as the locating pins could be located in error there too. This is a nasty manufacturing problem for BMW when you look at the potential in this assembly for cumulative errors.

[BucksTherapy]

Bear in mind that a wholesale replacement of parts will also have to include the engine block as the locating pins could be located in error there too. This is a nasty manufacturing problem for BMW when you look at the potential in this assembly for cumulative errors.

 

What I really need are stats on which part is out of line. This would give us folks who find machining and measuring run out beyond their tools and ability to know the odds that each peice they replace is going to cure their alignment issue.

Only BMW might have these stats and they can not provide them without taking on significant liability for out of warranty failures.

Intutitvely I would think the transmission cover is the more likely culprit. What do you think? I'm down to gambling and I am leaning towards the following:

Preference- Find a used transmission that has no sign of spline wear and replace complete clutch assembly.

Second choice - Replace complete clutch assembly, input shaft, bearing and seal and transmission cover.

 

Do you agree?

[Boffin]

The fact that BMW have allegedly replaced the transmission on several low-mileage failed splines suggests that this is the more likely culprit. But they may just be throwing out sops to keep customers happy.

 

Cya, Andy

[nrp]

This is strictly my opinion -

 

Since the errors are cumulative and subject to random scatter, many of the potential errors would have to approach the limits of production accuracy to get the necessary overall accuracy. For example, ask a machinist if he can repeatedly locate a dowel pins within .001 inch over a large dimension without extraordinary production efforts.

 

Over say a 12 inch distance in steel, 14 degrees F will cause that much dimensional shift simply from expansion. In aluminum, it would take only 7 degrees. That sort of errors accumulating wrongly, coupled with the short stiff transmission input shaft extension, is what radially loads the spline. Cars have larger splines, & longer extension shafts, so they can flex, and are not so likely to bind the spline into an impossible situation.

 

It is even possible that BMW is match assembling parts for maximum error cancellation. It is a nasty problem to hold this sort of tolerance for any production process.

 

I would guess though that chances are much of the error is in the clutch housing. I have no basis on how to make that though, other than speculate on the nature of the manufacturing process and possible deflections in the parts as they are mounted into a machine.

 

FWIW - I just took the starter off my 2000 R1100RT @ 18000 miles & found the clutch disc to float very little (maybe .050 inch tangentially), and there was no debris inside the clutch housing adjacent to the spline. Therefore I'm not going to do a premature spline lube. Thanks to whoever suggested that a week or two ago.

[BucksTherapy]

I am always impressed with the depth of insight I receive from the good folks on this site. In this particular case there is no easy answer but the insights offered are a great help. I have concluded I probably have as much information as I can get so I am at least going to make an informed gamble.

 

I would like to comment on measuring the slack at the outer edge of the clutch to indicate spline wear. I think this is a great idea and am very interested in seeing the data bear this out. I don't know if I would depend on it if the wear is not yet extreme as I fould little slack(didn't measure it) when the wear pattern was well involved but there was still flats on the tops of the splines. The wear pattern would indicate that deflection occurs when under pressure but might not result in much slack when not under pressure. My observation from the wear pattern is the problem that results from misalignment results in too small a tolerance on some portions of the mating surface and too large at others. Where the gap is tighter there are high pressure areas and uneven wear ensues.

 

My conclusion from all of this is if you are unlucky enough to get a bike with a misalignmentthen correcting this problem is beyond what most of us or most dealers can resolve with any certainty.

 

If BMW have drawn the same conclusion then they have no practical way to resolve these problems. The only for certain cure would be to replace block, clutch assembly and transmission with a matched unit from the factory and that certainly isn't practical.

 

I love the RT. Best bike for me that I've owned in 30 years of riding. With that said, I don't think I'm prepared to replace clutch assembly and transmission every 30,000 miles. I think it's time to spend a little time in the K-bike forum and see what systemic issues the GT might have. Apparently its supposed to be a little more touring oriented this year.

 

On the other hand I could keep the RT for another 20,000. I have to fix it either way.

 

Again, thanks for all of your help.

[nrp]

A couple of thoughts -

 

1) I looked at the bell housing interior by removing the starter. In my case it was completely clean inside. I suspect if there was spline distress there would be debris scattered around. The other posted pictures of various spline failures show a lot of wear debris around the interior of the bell housing.

 

2) If you have to do a teardown to fix the spline situation, some way of measuring and then modifying the engine transmission interface would be in order for us self respecting gearheads. It would be possible to slightly shift the locating holes for the dowel pins by swaging the hole shut slightly and then filing them to get a better alignment.

 

It would be laborious as hell, but like here in Minnesota, what else is there to do in the winter?

[BucksTherapy]

Good thought... I think

 

I'm no novice, but this would be well beyond my tools and talents.

 

Thanks though!

[selyab]

My '03 RT failed on the freeway with 20k miles on it last weekend. Portland Motorcycle has diagnosed a clutch spline failure. They mentioned seeing this in earlier RT models and mentioned the alignment issue. They are getting a list of parts they think need replacing including ones that would prevent it from reoccurring. They said they have a call in to the BMW rep to approve the parts they want to order. I guess they need the additional approval since they want to replace parts beyond damaged ones. My question is what should I be asking for in terms of parts to try and identify the source of a misalignment? Remember this happened in only 20k miles. At least I have 3 months left on my warranty.

[Thatman]

Damn. My '04 RT with 17.3k miles failed a couple of weeks ago while I was in Tacoma.

 

For me South Sound BMW replaced my input shaft, clutch disc and seal for the input shaft.

[mtwillyman]

Thatman,

What happened to your bike? I was following your post a couple weeks ago. Did South Sound BMW treat you well?

[Ralph]

BMW replaced the trans housing on my 03 along with the spline shaft ect.

[mbohn]

Here is an update on my dealings with this spline issue:

 

http://bmwsporttouring.com/ubbthreads/showflat.php/Cat/0/Number/628029/an/0/page/0/gonew/1#UNREAD

 

In this earlier thread I described how I made the misalignment measurement:

 

http://bmwsporttouring.com/ubbthreads/sh...3&fpart=all

 

it is buried near the bottom of the first page, look for my username mbohn. I don't see any shortcuts to this procedure, you are trying to measure a misalignment of just a few thousands of an inch and your work has to be spot on.

 

Also, replacing parts may not fix this. BMW tried that ad nauseum on my bike to no avail.

 

It is my opinion that BMW is not capable/qualified of making this repair, that is why I gave up on them and undertook it myself. The sad news is that for a normal person (one not capable/interested in making the repair), I suggest getting a spline lube and selling the bike, if your conscience will allow you to do so. It will go a few hundred miles on a spline lube, maybe further.

[wilsons]

should I be worried about any of this in my ST? cuz by the end of next year i'll probably have 20K+ on my bike. (or is this worthy enough to ask in hexheads?).

[Paul Mihalka]

should I be worried about any of this in my ST? cuz by the end of next year i'll probably have 20K+ on my bike. (or is this worthy enough to ask in hexheads?).

All the parts in the hexheads are different from the oilheads, so oilhead history does not apply. There are not that many hexhead high mileage bikes to show any frequent failure tendencies. May be some early R1200GS. Any regular participants in ADVRIDER? That should be the place where problems show up first.

[nrp]

I'm almost certain my 1975 R90/6 has a spline alignment problem. It failed at about 20,000 miles. I replaced the clutch & input shaft & greased the whole thing with molygrease thinking that the factory had somehow missed any assembly lube.

 

I then checked my 1977 R100/7 with about 20,000 miles & it was still perfect. (Guess they must have figured out how to lube it eh?).

 

The '77 was sold later - but the '75 now has 40,000 miles and is going to come apart again since the whole issue now looks like alignment and manufacturing errors.

 

I don't think any bike is necessarily completely free of having this problem. I suggest that any bike torn down for this should have the engine-transmission axis runnout checked - and do whatever it takes to get it less than .001 inch TIR.