Clutch and input shaft spline - need advice please

AntonLargiader said:

Those input shafts always wear like that on the Boxers, because of the instability of the clutch hub on the shaft (due to the insufficient engagement). It's not indicative of any particular problem, just that there is a problem that is causing wear.

The same problems cause wear on other designs, but the hubs are stable on the shaft and the wear tends to be even. I've replaced input shafts on just about every BMW model since 1970.

Click to expand...

Are you saying that this may not be a case of the often talked about misalignment between cases?

This IS a classic case of misalignment.

If you look at the pictures, most of the spline tooth wear on the shaft is closest to the transmission, and most of the internal spline wear on the clutch hub is closest to the engine. This suggests the surfaces are still conjugate, were and will always be in full contact (until failure!).

The only wear source for this type of failure is the engine/clutch housing spline ID was not rotating on the same center as the transmission spline OD. The only way for this situation to resolve itself is for the transmission shaft to drag or orbit the clutch ever so slightly around the face of the flywheel every revolution - but it does this every revolution and with a lot of radial force (like maybe ~1000 lbs?) being reacted by the spline surfaces. That is also why your clutch face wear surfaces appear a little burned.

The spline surfaces are being fretted every revolution. The spline teeth see a fluctuating load with a slight motion every revolution, rather than a fixed load and no motion. Soon the lube is gone and the products of fretting appear - the red/gray dust you found in the clutch housing.

Very possibly this is being caused by a slight manufacturing error, but BMW's feet have been held to the fire for many years without a quality solution.

I am starting to think that in many cases this just might also be caused by an assembly technique where the clutch disk is initially very slightly off centered and the transmission-engine bolts unintentionally used to pull everything together. I know there are two alignment spools that are supposed to keep things in alignment under such a procedure, but the clutch housing on the oilheads have thin sections and a large cutout for the starter, which compromises the stiffness of of the housing far more than is obvious to anyone not familiar with structure principles (BTDT).

You should also check to see if the crankshaft rear main bearings are worn to be egg shaped in any direction. Prying on the flywheel in a couple of directions will show this. If the flywheel can be rattled around much more than a couple thousands, you should replace the rear main bearing if you want to keep the bike. Otherwise be sure to release the clutch frequently during the transmission-engine bolt up so that the clutch can seek its happy point and not force the transmission offset.

My guess is you could get by with the present shaft and clutch disk, but I'd remove the starter every 5K miles to re lube each of the spline surfaces, see how the debris field looks, and note the tangential backlash on the clutch disk outside diameter.

MNMechEngr (retired)

Since we are on the topic I have to ask about my recently purchased 1150RT. Bike now has 13K and is a 2003. Am I good to wait the 40K before a spline lube or is it possible the grease is old and dried up by now?

My 2000 R1100RT at 22K was pretty dry based on a view thru the starter opening, but unworn. I disassembled for a full spline lube. YMMV!

You could take off the starter & look inside. You could even re lube the spline thru the starter opening with a modified (i. e. a heated & bent tip) insulin hypodermic needle & thinned lube. You would have to be sure to get all spline teeth individually with about ~ 0.1 cc or less. You don't want to allow any extra to get on the clutch disk. In hindsight I wish I had attempted this rather than a full tear down.

If nothing else, if there is no debris in the area and the tangential clearance at the clutch disk is OK, the alignment is probably OK, and best left alone. But as long as you are in there, try the hypodermic lube as it would only take a few minutes. Write down the tangential clearance somewhere for the next time you open it up. I didn't, dang it!

nrpetersen said:

My 2000 R1100RT at 22K was pretty dry based on a view thru the starter opening, but unworn. I disassembled for a full spline lube. YMMV!

You could take off the starter & look inside. You could even re lube the spline thru the starter opening with a modified (i. e. a heated & bent tip) insulin hypodermic needle & thinned lube. You would have to be sure to get all spline teeth individually with about ~ 0.1 cc or less. You don't want to allow any extra to get on the clutch disk. In hindsight I wish I had attempted this rather than a full tear down.

If nothing else, if there is no debris in the area and the tangential clearance at the clutch disk is OK, the alignment is probably OK, and best left alone. But as long as you are in there, try the hypodermic lube as it would only take a few minutes. Write down the tangential clearance somewhere for the next time you open it up. I didn't, dang it!

Click to expand...

I'm at 28K miles and thinking of trying it. I've had the starter out twice in the past month and know that's easy. What are your thoughts on type of lube?

nrpetersen said:

My 2000 R1100RT at 22K was pretty dry based on a view thru the starter opening, but unworn. I disassembled for a full spline lube. YMMV!

You could take off the starter & look inside. You could even re lube the spline thru the starter opening with a modified (i. e. a heated & bent tip) insulin hypodermic needle & thinned lube. You would have to be sure to get all spline teeth individually with about ~ 0.1 cc or less. You don't want to allow any extra to get on the clutch disk. In hindsight I wish I had attempted this rather than a full tear down.

If nothing else, if there is no debris in the area and the tangential clearance at the clutch disk is OK, the alignment is probably OK, and best left alone. But as long as you are in there, try the hypodermic lube as it would only take a few minutes. Write down the tangential clearance somewhere for the next time you open it up. I didn't, dang it!

Click to expand...

With the starter removed, exactly what are you able to lube? Wouldnt the "usable part of the spline" be inside the clutch hub? If so your not lubing the section that counts.
Hope I'm wrong!

I've poked around from the starter hole a couple of times. I finally reached the conclusion that if a gap in the spline is large enough to be able to squeeze a meaningful amount of lubricant into it, the spline is already rather severely worn. Maybe my 90 degree needle was too large. In any event, I concluded I was on a fools errand and just took it apart.

nrpetersen said:

...... I am starting to think that in many cases this just might also be caused by an assembly technique where the clutch disk is initially very slightly off centered and the transmission-engine bolts unintentionally used to pull everything together. I know there are two alignment spools that are supposed to keep things in alignment under such a procedure, but the clutch housing on the oilheads have thin sections and a large cutout for the starter, which compromises the stiffness of of the housing far more than is obvious to anyone not familiar with structure principles (BTDT).

You should also check to see if the crankshaft rear main bearings are worn to be egg shaped in any direction. Prying on the flywheel in a couple of directions will show this. If the flywheel can be rattled around much more than a couple thousands, you should replace the rear main bearing if you want to keep the bike. Otherwise be sure to release the clutch frequently during the transmission-engine bolt up so that the clutch can seek its happy point and not force the transmission offset.

My guess is you could get by with the present shaft and clutch disk, but I'd remove the starter every 5K miles to re lube each of the spline surfaces, see how the debris field looks, and note the tangential backlash on the clutch disk outside diameter.

MNMechEngr (retired)

Click to expand...

Very interesting. Thank you for your explanation and advice, nrpetersen.

I'll check the flywheel (as much as I can) and follow your advice during transmission-engine bolt up. However, I don't fully understand how the clutch would force the transmission offset considering that those dowel pins seem to keep the cases pretty snug. But you did question the stiffness of the housing.

I don't think the splines can be lubed through the starter opening; unless there is a technique that I am not aware of.

Bohan said:

Are you saying that this may not be a case of the often talked about misalignment between cases?

Click to expand...

Yes. I think that sort of misalignment, to a degree that matters here, is rare.

I'm not saying that such misalignment isn't possible, but it would likely cause a repeat failure in most cases after a similar number of miles (unless fixed). OTOH replacing the entire clutch has transformed several multi-time losers into nice reliable bikes, so that's what I do.

DIN specs say the splines should be aligned within .003 inch TIR. I don't know if that is even a valid number for this particular application (there is a DIN spec for everything in Germany). But given the thin aluminum sections in this clutch housing, getting that much misalignment even with the pins in place wouldn't take a lot of force. In addition once the main bearings wear and the oil thins out, the reciprocating loads of the pistons will also hammer the spline.

Ideally a spline is an inside-outside gear set, with equal number of teeth, pressure angles, and generated profiles etc. Ideally the tooth loads should change only with torque/throttle if everything is aligned. But if the crankshaft bearings get loose, or there is even a slight misalignment, the cyclic loading of the spline teeth multiplies enormously to engine rpm.

I measured the diameter of wire that could be inserted between the spline tooth tips and the spline housing major diameter on my RT. As I recall it was about .030 inch. There are cheap insulin hypodermics available over-the-counter at Walgreens that are quite a bit smaller. The hypodermic was also calibrated on the side so that equal amounts could be put into each tooth cavity. I never actually did it though.

The next trick would be to find a good thinner for the spline lube so that it can be hypo'd. Maybe a little bit of lacquer thinner (but not too much!)?

I used straight Honda Moly 60 stuff on my R1100 last winter.

About 20 years ago my R90/6 stripped its spline at ~30K miles. It seemed dry on dis assembly. That was the days before Internet forums etc so I just rebuilt it using some wild nickel based stuff from work called "Neva Seize". I did take it apart a couple of years ago & was impressed how that sticky stuff had stayed in there. It took a week to clean my hands of it.

I never checked alignment. Did it correct itself? Dunno! It may come apart again as this year's Minnesota winter project.

Bohan said:

Very interesting. Thank you for your explanation and advice, nrpetersen.

I'll check the flywheel (as much as I can) and follow your advice during transmission-engine bolt up. However, I don't fully understand how the clutch would force the transmission offset considering that those dowel pins seem to keep the cases pretty snug. But you did question the stiffness of the housing.

I don't think the splines can be lubed through the starter opening; unless there is a technique that I am not aware of.

Click to expand...

If the disk is slightly off-center when the transmission is installed it makes aligning the splines and the clutch rod entry into the pilot bearing in the crankshaft difficult. But I seriously doubt it would allow you to bend the transmission case or bellhousing. Those are stiff enough to constitute the main frame of the motorcycle. And, the clutch is going to be relieved of any lateral pressure on the clutch disk the very first time you pull the lever and release the spring pressure so any lateral pressure will not exist through any heat cycles. As an aside, as soon as there is appreciable wear on the splines gravity will slightly off center the clutch every time it is fully released in a stopped position. Then when you engage the clutch it engages in that slightly off centered position. I suspect it tends to recenter when you shift to 2nd, etc. But once worn a little the splines wear faster thereafter at an ever increasing rate.

PGlaves said:

If the disk is slightly off-center when the transmission is installed it makes aligning the splines and the clutch rod entry into the pilot bearing in the crankshaft difficult. But I seriously doubt it would allow you to bend the transmission case or bellhousing. Those are stiff enough to constitute the main frame of the motorcycle. And, the clutch is going to be relieved of any lateral pressure on the clutch disk the very first time you pull the lever and release the spring pressure so any lateral pressure will not exist through any heat cycles.

Click to expand...

My contention is that it never recenters. If there was an initial misalignment, that misalignment will reappear and the resulting shear force across the spline will be present every engine revolution. The spline shear force vector will be stationary, as the spline rotates. If I was to build a spline wear fatigue test system (my pre-retirement job) I would use a rotating spline set and radial misalignment control.

PGlaves said:

As an aside, as soon as there is appreciable wear on the splines gravity will slightly off center the clutch every time it is fully released in a stopped position. Then when you engage the clutch it engages in that slightly off centered position. I suspect it tends to recenter when you shift to 2nd, etc. But once worn a little the splines wear faster thereafter at an ever increasing rate.

Click to expand...

Again - it never recenters except for the trivial case where the clutch is disengaged. The radial stiffness of the aluminum housing with its thin and open sections could be checked with a dial indicator & a screwdriver to pry things. I am suspicious of the R series in that they are the only ones that have the open starter port.

Later - OK, if the clutch disk - engine transmission assembly is elastically sprung/dragged together with the engine transmission bolts, and assuming the alignment dowels have no clearance and are in the correct position, and the crank main bearing, and the spline clearance is near zero, and there is no housing deflection, then the clutch, when first released, will allow everything to center. But heat cycles can't overcome the faying surface friction under typical bolt preloads.

There's a lot of chance for assembly errors to compound alignment problems despite an accurately machined clutch housing.

A radical screwball idea -

Maybe leave the engine-transmission bolts a half turn loose, start the engine, and repeatedly engage the clutch in neutral until a happy point is reached by the assembly where there is no relative motion between the two cases. Then slowly torque everything up.

My concern though is that there is nothing in such a lash up except chance that will ever lock the transmission to the engine with no run out. A finger bridging the case joint would be able to feel any motion though.

Over many years, I have had to replace the clutch disc, flywheels, pressure plates, etc., on a variety of cars, almost all foreign made. In all that time I have never seen an input shaft with worn splines. Nor a clutch disc with worn splines for that matter. Many of these vehicles had high mileage on them. No one I know who buys a used car with say, 50,000 miles on it, immediately sends it to the shop to have the transmission dropped and the clutch assembly re-lubed. Any car company either suggesting, or requiring that, would be quickly out of business.

Taking these bikes apart, essentially splitting them in the middle on the oildeads-on, is no walk in the park. And often, it seems, the input shaft is worn. Well, there's $1700 for a transmission rebuild. And the worst part is that all you're doing is just re-setting the clock till it goes again. I'm just really curious what it is about these relatively little engines, and even littler clutch assemblies, that makes them so fragile, and such a maintenance problem to own? Like the one guy says, should you expect to split the bike in half every year or so just to make sure your "splines" are not on the verge of failure? This has got to be a moneymaker for the BMW dealerships.

There is no way to know from outside that these things are failing, or have proper lube, etc. As someone looking for a new bike, it's depressing to realize that the first thing I will have to do after driving it home is to disassemble half the machine, with a deep sense of dread that inside that shiny transmission case I will find a worn or shot input shaft. Yes, I know that Japanese bikes have their own problems, but taking them apart to lube/replace transmission shafts on a fairly regular basis doesn't seem to be one of them. Jeeeze, I would think BMW could come up with a solution to this seemingly unending and major problem/headache on their bikes. Apparently I'd be wrong, though.

jconway607 said:

Over many years, I have had to replace the clutch disc, flywheels, pressure plates, etc., on a variety of cars, almost all foreign made. In all that time I have never seen an input shaft with worn splines. Nor a clutch disc with worn splines for that matter. Many of these vehicles had high mileage on them. No one I know who buys a used car with say, 50,000 miles on it, immediately sends it to the shop to have the transmission dropped and the clutch assembly re-lubed. Any car company either suggesting, or requiring that, would be quickly out of business.

Taking these bikes apart, essentially splitting them in the middle on the oildeads-on, is no walk in the park. And often, it seems, the input shaft is worn. Well, there's $1700 for a transmission rebuild. And the worst part is that all you're doing is just re-setting the clock till it goes again. I'm just really curious what it is about these relatively little engines, and even littler clutch assemblies, that makes them so fragile, and such a maintenance problem to own? Like the one guy says, should you expect to split the bike in half every year or so just to make sure your "splines" are not on the verge of failure? This has got to be a moneymaker for the BMW dealerships.

There is no way to know from outside that these things are failing, or have proper lube, etc. As someone looking for a new bike, it's depressing to realize that the first thing I will have to do after driving it home is to disassemble half the machine, with a deep sense of dread that inside that shiny transmission case I will find a worn or shot input shaft. Yes, I know that Japanese bikes have their own problems, but taking them apart to lube/replace transmission shafts on a fairly regular basis doesn't seem to be one of them. Jeeeze, I would think BMW could come up with a solution to this seemingly unending and major problem/headache on their bikes. Apparently I'd be wrong, though.

Click to expand...

I agree, but then again, you can always buy a bike that needs to have the engine removed to replace the stator every so often. Pick your poison.

p.s. I have an R1100RS in the awkward tail up position, with the transmission sent off to Pennsylvania as I type, so I do feel the pain.

PGlaves said:

p.s. I have an R1100RS in the awkward tail up position, with the transmission sent off to Pennsylvania as I type, so I do feel the pain.

Click to expand...

Paul,
I, for one, would love to know just how many miles that R1100RS of yours that has suffered terminal spline failure has on it? Thanks,

John

PGlaves said:

I agree, but then again, you can always buy a bike that needs to have the engine removed to replace the stator every so often. Pick your poison.

p.s. I have an R1100RS in the awkward tail up position, with the transmission sent off to Pennsylvania as I type, so I do feel the pain.

Click to expand...

Seems I read or heard somewhere that BMW is soon to do away with the dry clutch all together and go all wet clutch just like the made for Hondakawsusyammi.

nrpetersen said:

My contention is that it never recenters.

Click to expand...

Can you support that? When the clutch disk is positioned solely by the spinning input shaft, what is going to throw it off-center again when the clutch engages? And in that case, why wouldn't it be thrown off-center even if originally installed perfectly?

nrpetersen said:

A radical screwball idea -

Maybe leave the engine-transmission bolts a half turn loose...

Click to expand...

It must have been in a different thread where that came up recently. You're not familiar with the system in question: not only is there simply no movement between the engine and transmission due to the dowels, but the upper bolts are heavily obscured in later reassembly.

jconway607 said:

...... Taking these bikes apart, essentially splitting them in the middle on the oildeads-on, is no walk in the park. And often, it seems, the input shaft is worn. Well, there's $1700 for a transmission rebuild. And the worst part is that all you're doing is just re-setting the clock till it goes again.

..... As someone looking for a new bike, it's depressing to realize that the first thing I will have to do after driving it home is to disassemble half the machine, with a deep sense of dread that inside that shiny transmission case I will find a worn or shot input shaft.

Click to expand...

My thoughts exactly. Needless to say, right now I am quite disappointed with BMW, R bikes in particular. If I knew going in that I'd have to replace the transmission input shaft and clutch assembly at 54k miles I would have considered another model or manufacturer.

jconway607 said:

Paul,
I, for one, would love to know just how many miles that R1100RS of yours that has suffered terminal spline failure has on it? Thanks,

John

Click to expand...

357,658 miles