Checking for spline wear 1999 R1100S

R100RTurbo said:

I concur with the opinion that gaining full spline engagement (especially when it runs out so close to disc center line on a rather flexible plate) would be beneficial, just another opinion however.

Click to expand...

No. Spline wear is due to high cycle radial loading induced by radial (added 2-10-14) axes misalignment.

Adding to the spline engagement by shifting the hub or adding to the hub making its center even further towards the engine would simply increase the local loading. Ideally the clutch flex plate should be centered (longitudinally) on the hub to more equally spread the radial load caused by misalignment.

It would be better though to simply have less misalignment through better manufacturing and assembly control - something many trouble free bikes seem to already have.

Error-compensating alignment spools would be quite easy to make. The problem is documenting the misalignment in the field, then communicating & getting the correct spools oriented correctly on reassembly.

jammess said:

Does anyone know why BMW went to the hydraulic clutch on the oil head? What advantage does the hydraulic clutch have over cable operated in this application? I certainly can't see how the hydraulic clutch benefits we the end users.

Click to expand...

I would suspect the hydraulic clutch, like the iABS power brakes looked like a great idea from a marketing perspective! Reliability and longevity certainly wasn't a consideration.

nrpetersen said:

I think I understand what yuo meant here. You are suggesting that a warped flywheel or weird worn clutch disk would moment load the spline across the flex plate (spider).

I don't think this is so as the number of loaded cycles (transmitting torque while slipping only) is so small that we would not be seeing the kind of high-cycle fretting wear products we are encountering. The flex plate should be soft enough in generating overturning moment into the spline hub, that the wear component would be miniscule. I doubt the flywheel face etc would have that much run out.

The flex plate however is very stiff radially, and even a minor axes misalignment would generate a lot of spline wear dragging the clutch disc around. I figure the rotating radial force vector being applied to the spline in a misalignment is on the order of 1000 lbs. And it continues always being that full ~1,000 lb force vector every revolution whenever the engine is running and the clutch is released.

Remember many bikes don't show spline wear at all. Unfortunately, not enough though.

Click to expand...

In steady riding the engine torque is about 8 lb-ft at 40 mph and 15 lb-ft at 60 mph. Peak torque during acceleration can reach 65-70 lb-ft. Isn't the wear you're speaking if far mor likely during hard acceleration? After all, between 40-60 mph the shaft load is pretty low.

That being the case, can any part of the clutch system be deformed by acceleration? E.g. Arms of the clutch disk twisting and in turn affecting the planarity of the clutch hub?

I know the popular theory is misalignment but if that were the case Anton's replacing of the clutch pack wouldn't be a fix. I think we have to give a high weight to his experience since he has a lot of data. We need a theory of cause that matches the wear pattern and his data.

More Questions

I'd like to visualize the actual fit of the 6 speed tranny shaft in relation to the clutch disc in standard form, a bit better.
Based on the images provided of the worn shafts, it seems clear the spline engagement ends roughly 2/3'rds of the way onto the tranny shaft. On the other end, where is the tranny shaft "tip" in relationship to the engine side of the clutch disc splines? If I understand GSAddict's modification, he has had a shaft manufactured that is slightly longer towards the front of the engine (would reside closer then to the release rod plate that is part of the spring dish). If that is true, is then the stock tranny shaft not making it all the way through to the front side of the "clutch disc" splines? (that would seem to be the reason for the longer shaft?) If this explanation doesn't deliver, I can offer a sketch to clarify the basis of the questions, but in essence a cross sectional side view of an assembled unit would answer it.
Based on the above, perhaps other conclusions can be drawn.

roger 04 rt said:

In steady riding the engine torque is about 8 lb-ft at 40 mph and 15 lb-ft at 60 mph. Peak torque during acceleration can reach 65-70 lb-ft. Isn't the wear you're speaking if far mor likely during hard acceleration? After all, between 40-60 mph the shaft load is pretty low.

Click to expand...

I don't have dimensions but approximating for 18 spline teeth, the 8 ft lbs would be about 190 lbs total tangential load, or about 10 lbs simple compression load for each tooth face. But if there is radial misalignment dragging the clutch disk around the flywheel with a radial force of 1000 lbs, the load on each tooth will be cyclic (and fully reversing every engine rev and never going away even at part throttle) of about 110 lbs of fully reversing fretting/sliding on each loaded tooth (only about half of the teeth are carrying this radial load at any point in time). This is WAG engineering but it suggests that radial loading wear at the spline is potentially much more of the problem. Various schemes that apply the load off-hub-center would compound the problem.

roger 04 rt said:

That being the case, can any part of the clutch system be deformed by acceleration? E.g. Arms of the clutch disk twisting and in turn affecting the planarity of the clutch hub?

Click to expand...

If things are fully planar and aligned, the only deformation will be torsional and that will be mostly in the shaft. It can be shown to be very minor. The flex plate should remain planar.

roger 04 rt said:

I know the popular theory is misalignment but if that were the case Anton's replacing of the clutch pack wouldn't be a fix. I think we have to give a high weight to his experience since he has a lot of data. We need a theory of cause that matches the wear pattern and his data.

Click to expand...

A very true statement!

I can't explain why Anton's total clutch pack fixes the problem unless it is because he is careful not to drag the clutch disk during bolt-up. He hasn't said he does this, but I'll bet he does. But I would think by now the BMW factory would be doing the same thing yet these alignment problems seem to persist.

Another factor though is two specific instances of considerable measured misalignment are in the various forums. I question the accuracy of one of those sets of measurements because the setup didn't seem stiff enough to eliminate the effects of gravity on the dial indicator, and didn't measure the main bearing clearance effects. In the case I tried to document, the loose crankshaft main bearings would cause spline wear just due to crankshaft radial motion. Neither solution has been proven but I think the initiating problem has been identified.

BTW - If anyone has a spline failure repair under way & wants the fixture etc to measure the run out between an engine crankshaft and a transmission input bearing housing bore, let me know.

Questions answered

R100RTurbo said:

I'd like to visualize the actual fit of the 6 speed tranny shaft in relation to the clutch disc in standard form, a bit better.
Based on the images provided of the worn shafts, it seems clear the spline engagement ends roughly 2/3'rds of the way onto the tranny shaft. On the other end, where is the tranny shaft "tip" in relationship to the engine side of the clutch disc splines? If I understand GSAddict's modification, he has had a shaft manufactured that is slightly longer towards the front of the engine (would reside closer then to the release rod plate that is part of the spring dish). If that is true, is then the stock tranny shaft not making it all the way through to the front side of the "clutch disc" splines? (that would seem to be the reason for the longer shaft?) If this explanation doesn't deliver, I can offer a sketch to clarify the basis of the questions, but in essence a cross sectional side view of an assembled unit would answer it.
Based on the above, perhaps other conclusions can be drawn.

Click to expand...

No need, looked through some earlier posts and gained the answers. The input shaft ending where it does, well short of the forward clutch disc splines lends the comment, "that could be a lot better".
Regardless, lots of this already gone over many times before so no need to rehash.

Some interesting comments on splined shaft wear: http://www.franklin-electric.com.au/media/8551/Spline wear pge SEM 38.pdf.

And some more food for thought: http://www.eng-tips.com/viewthread.cfm?qid=318697

Check these loctite products, page 12-13. http://www.henkelna.com/us/content_data/192137_LT5338_Shaft_Repair_final.pdf

A couple possible issues:
-under-designed spline coupling leading to fretting and abrasive wear.
-possible under-design caused by insufficient engagement
-possible improper factory lunrication leading surface bond fretting
-possible accelerated fretting wear due to oxidized fretting particles embedded in hub

I'm not sure if this adds anything to the discussion. If I heard him correctly, he thinks the 6th gear use at lower RPM's is the problem but his fix is an extended improved clutch disk. Did I miss something?
http://www.youtube.com/watch?v=Jx3e0gqzylY

Mis-alignment?

To the extent that shaft mis-alignment could be causing spline wear, it seems to me this could be checked out.

First, there must be a factory tolerance on some drawing, for engine case and transmission case:
-Concentricity: of crankshaft, or trans main shaft, bore to an ideal hole center fixed by the two dowel pins in the trans mounting flange
-Perpendicularity: of crankshaft bore to the transmission mounting flange.

A simple gage could be made, such as:

-a 1/2? thick gage plate, mounted to trans mounting flange on engine block, located by the dowel pins
-it is precision-bored at ideal crank centerline, with say 3? dia bore
-a dial indicator base is mounted to flywheel or crank face, extending through the gage plate bore, where the dial indicator is attached.

Then the dial indicator could sweep the gage bore ID and the gage plate face as the crank is rotated, to determine the runout deviation from spec. Using the locating dowel pins, the gage plate could be clamped to engine or trans to determine trueness.

I wonder if such a gage exists.

And for those who have relocated the dowel pins to correct misalignment, how did they measure the error? I suppose you could mount a dial indicator on the flywheel, bolt up the transmission, and read the radial runout directly on the trans input shaft, peering through the starter motor cavity. Sounds hard to do.

DougRaymond said:

..............And for those who have relocated the dowel pins to correct misalignment, how did they measure the error? I suppose you could mount a dial indicator on the flywheel, bolt up the transmission, and read the radial runout directly on the trans input shaft, peering through the starter motor cavity. Sounds hard to do.

Click to expand...

It isn't that hard to do. See my post 65 at:

http://forums.bmwmoa.org/showthread...put-shaft-spline-ÔÇô-need-advice-please/page5

I still have that fixture.

Making a match plate would tend to compound errors, and requires a lot of very accurate machining - if you even knew the actual desired dimensions, which you will never get.

I made my fixture with my band saw. The dial indicator in the transmission bearing bore measures what you want directly. It also quickly measures any possible rear main bearing clearance contribution from the engine. The only thing critical is that there not be any significant deflection in the flywheel extension fixture. I used a piece of conduit & a couple of magnets from an old hard drive assembly. It held like all fury on the flywheel bolts.

roger 04 rt said:

Some interesting comments on splined shaft wear: http://www.franklin-electric.com.au/media/8551/Spline wear pge SEM 38.pdf.

http://www.eng-tips.com/viewthread.cfm?qid=318697

Check these loctite products, page 12-13. http://www.henkelna.com/us/content_data/192137_LT5338_Shaft_Repair_final.pdf

Click to expand...

It is interesting that none of these outside references identifies high speed splines as a 1:1 inside-outside gear set. They all assume splines are square teeth into square grooves - which they are not except maybe for crude farm implement stuff!

The reason splines are involute (gear-like rather than square) are strength and fatigue durability (especially when properly lubricated), and ease of manufacture. Shaft splines are created on a hobbing machine which generates the involute profile the same as gears. The same machine is used to create a broach which is a tool with multiple teeth pressed through the hub to create the proper internal matching spline tooth profile. That's why the hub is probably softer than the shaft.

PAS said:

I'm not sure if this adds anything to the discussion. If I heard him correctly, he thinks the 6th gear use at lower RPM's is the problem but his fix is an extended improved clutch disk. Did I miss something?
http://www.youtube.com/watch?v=Jx3e0gqzylY

Click to expand...

This is an interesting video, at least it is to me as it pretty much states my humble opinion on this subject. That said I won't pay for a modified clutch disc but I will continue to ride my 1150 in 5th gear and strive to make every shift smoothly and at 4K RPM or higher (4K - 5K). Don't really have an opinion on trans/engine axial alignment issues as I think it would be difficult at best to measure an axial error and anyway, what are the specs? You will never find out!

PAS said:

I'm not sure if this adds anything to the discussion. If I heard him correctly, he thinks the 6th gear use at lower RPM's is the problem but his fix is an extended improved clutch disk. Did I miss something?
http://www.youtube.com/watch?v=Jx3e0gqzylY

Click to expand...

Something's wrong. I never heard Chris swear during that video!

What he says makes sense to me. I've remarked before that having splines that do not fully engage the hub does not seem logical. And maybe it is a design error that was carried over several models. The cynic would say it saves on steel and sells a lot of expensive parts so why fix it?

But other things mentioned also come to mind. I used to work on a lot of VWs. The clutches had springs to absorb some of the shock of engagement. Oilhead cluthes are rigid. That can't help the situation. Other non hydraulic clutches I've seen in automotive applications have nose bearings to keep things where they belong under stress. Oilhead input shafts have no such aids.

I've also always thought that the surging issues Chris mentions and low RPM operation are contributing factors. Both my 1100 RTs were nasty surgers and you're paying attention you can feel that this cannot be good for the drive train. That on again off again that Chris describes is smacking every driveline component to and fro a LOT.

I don't think there is a holy grail here. There is no single factor that is the definitive reason for spline issues. It seems more than likely it is the sum of many if not all of the things we've gone over many times in the forum.

jammess said:

Don't know if anyone has pointed to this sight yet: www.rbracing-rsr.com I found this sight several years ago and saved it for future reference just in case, scroll down to about midway in this link to the section on BMW clutches. At about 36K miles I did a spline lube/inspection on my '04RT and found splines to be pristine with absolutely no wear. It is interesting what Anton says about spline issues mainly having to do with the 6 speed trans models. The way I have ridden my bike since new is to hardly ever use 6th gear and I wonder if this mode of operation has contributed to negligible spline wear. Since installing a wideband O2 sensor with LC-1 which resulted in a much improved low end power output I have found 6th gear much more usable but I can't help but wonder what increased 6th gear use will over time mean for the trans input shaft splines. Hmmmmm...

Click to expand...

Hi HW,
Check out the above link and scroll down to the BMW clutch section.

Feedback from Jacks re Torsional Whip

Ref: Jacks Transmissions article on Torsional Whip.
I submitted some pictures including some of my own.
Asked if this was Torsional Whip.
Here is his reply.

"Yes, I would say so. The fact the splines do not fit tightly would also contribute as you then have them taking more of a beating too. I wonder why they didn?t make the shaft a little longer."

When I had my modified shafts made, the spline looseness and projection were concerns.
Hardness was the 3rd.

jammess said:

Hi HW,
Check out the above link and scroll down to the BMW clutch section.

Click to expand...

Been there and read the whole package they put together on clutches two years ago. Turbo 1100s too! Wow.
But if you do some googling on those clutches you will also find some not so good reviews about them and the business. Their focus seems to be racing. I decided back then it was not a solution I was interested in.

Bruno's extended clutch hub however... that is an idea that sounds like a good possible solution to long life splines but man it's spendy. I should call him and see what they go for shipped direct to a customer like me and not to a shop like Affordable Beemer to be marked up and installed.

My brother is picking up a 1999 RT with 42000 miles, and we are curious what would a spline lube/service cost at the dealer? The bike otherwise has been taken care of it just has not had the spline service. So what is the cost for this type of maintenance?

Other aspects of the clutch

GSAddict said:

Ref: Jacks Transmissions article on Torsional Whip.
I submitted some pictures including some of my own.
Asked if this was Torsional Whip.
Here is his reply.

"Yes, I would say so. The fact the splines do not fit tightly would also contribute as you then have them taking more of a beating too. I wonder why they didn?t make the shaft a little longer."

When I had my modified shafts made, the spline looseness and projection were concerns.
Hardness was the 3rd.

Click to expand...

Just when it seems the BMW clutch design has taken about as many hits possible, perhaps add this to the collection of factors - weakness of the plates adjacent to the clutch disc allowing those to "pucker" from their initial flat condition during hours of usage. Earlier airhead versions warp easily (the non heavy flywheel design that is similar to later oilheads of course) and I've documented distortion where both steel plates have angularity spreading away from flat when moving towards the inner circle. That is to say, the metal plates squeezing against the disc pucker away from it on each side at the central contact areas. I suspect oilhead assemblies warp in the same manner, perhaps more (higher hp, offset splines, and the further driving habit aspects offered on the video by "Chris") The disc might wear in similar fashion, but the potential for the ultra flexible clutch disc sheet metal to "wobble" then is large (portion riveted to hub and supporting organic material on circumference) Of course then if all new parts are not used during a refit then reasonably quick repeat failure might be expected.

I would go for a longer shaft if I were going to seriously attempt to improve this set up (and had the resources), 2nd would be longer spline engagement on the clutch disc (I have access to longer hubs for the airhead 31 spline nomenclature, which don't require increased engagement but could see that source helping offset the apparent availability and cost challenge of the product mentioned by "Chris" which he clearly offered with similar angst as the RB Racing people do lol.) 3rd idea (and this is totally back yard rigging) would be to take advantage of the unused spline zones on the disc and shaft and have some sealing devices made to hold your generous application of spline lubricant in place more effectively, perhaps for significantly longer time periods. Couple of random thoughts offered:

-For the front of the clutch assembly (where the input shaft is shy of engagement) the clutch hub might have a smallish splined "welch" type plug pressed into place and that having a small seal to allow clutch release rod to transition through center without leakage of precious lubricant.
-For the rear of the assembly (where the input shaft has roughly 1/3 of its exposed splines unused) the clutch disc might have a smallish seal fashioned - but a tricky solution admittedly. Ideas might be to broach the spline form (slightly tight) to a neoprene washer, in turn held to the disc by a collar or adhesive. Or a sleeve with some compression that would locate tightly at the inner end of splines just clear of input seal, that pressed against the hub.
-Lubricant then filled generously to assembly without concern for facing contamination.
Sure, kind of a dumb, fiddly solution but I'd venture to say possible if you had some time with the required machine shop tools. The above idea probably has some "con's" not obvious at first glance, but that's how the process of solutions develop where dumb ideas are tossed around to the point of solving previously felt to be closed issues.

Wow, I can't believe all of the info in this thread! I never realized what a huge question I was asking. Anyway, just a thought, if seems like most everyone agrees that if the splines were more thoroughly engaged (i.e. the plate would sit completely on the splines) then the wear wouldn't be as bad. I don't have any experience with replacing an oilhead clutch (yet) but I've done a bunch of Porsche and VW clutches. Would it be possible to use a spacer behind the flywheel and "push" the entire assembly further towards the transmission? I assume that would require a little rework on the slave cylinder push rod and I'm not sure about the engagement length on the starter but it might be easier than a custom clutch plate. I've also researched clutch sizes. It looks like a VW friction plate might possibly fit?

happy wanderer said:

Been there and read the whole package they put together on clutches two years ago. Turbo 1100s too! Wow.
But if you do some googling on those clutches you will also find some not so good reviews about them and the business. Their focus seems to be racing. I decided back then it was not a solution I was interested in.

Bruno's extended clutch hub however... that is an idea that sounds like a good possible solution to long life splines but man it's spendy. I should call him and see what they go for shipped direct to a customer like me and not to a shop like Affordable Beemer to be marked up and installed.

Click to expand...

Hi again HW,
Yeah, now that you mention it I do remember reading something negative about these clutches several years ago, seems I recall something about clutch chatter or slippage on engagement. Still, I saved the link in case I ever get to the point of clutch replacement. i also gave some thought to trying to install a cable operated clutch in the 1150 but quickly decided to not go there. After installation of the LC-1 with wideband O2 sensor which improved low end power output, I began using 6th gear much more than ever and began having thoughts about what this might be doing to the input shaft splines which got me thinking about these rsr shock absorbing clutches. Oh what the hey, I think I'll just ride the beast.