Jump to content
IGNORED

Chocolate Gear Oil?


smiller

Recommended Posts

Was helping a friend change his transmission oil (on a '97 1100RT, 35k miles) and even though Mobil synthetic 75W-90 went in there 12k miles ago what came out had a dark chocolate color not unlike very used motor oil. No metal bits of significance on the magnet and the viscosity and odor seemed normal, but the color sure wasn't right. I've seen metal-contaminated gear oil before but nothing like this. I don't see how it could be contaminated by water since the vent on an RT is so well protected.

 

He's planning on a 3500 mile trip shortly. Cause for concern..?

Link to comment

Smiller, that really doesn't sound like water as water usually leaves a milky or dark gray to even a coffee color appearance.. Usually that dark of a color is from moly type content, or running hot.. With the trans sitting right on top of the catalytic converter I guess trans heat isn't out of the question..

That Mobil 1 75W90is a newer entry synthetic gear oil built for modern light duty energy applications (not as hardy of a base stock as the xxW140 gear oil's).. I suppose it could contain a moly light EP additive.. I have the Mobil spec sheet on that gear oil & most entry's are listed as proprietary so that doesn’t shed much light on the content..

 

Two suggestions: if you have any of that used gear oil left that isn’t contaminated by the drain pan or the removal process send it to the Blackstone lab for analysis..

Or call Mobil Oil’s tec line tomorrow & ask them what that color means at 1200 miles of use..

 

Twisty

Link to comment

I've seen this on my 97 rt as well - it appears to be a result of the synthetic fluid not mixing with the original mineral based fluid and in my experience it will not go away with a few fluid changed. My original trans failed when the spur gear ate into the adjacent bearing, but I do not believe that was a result of the fluid change...just the design where the spur gear can put too much pressure on a ball bearing. The bottom line is, have your old fluid checked at a lab and watch the iron content. If you have a problem you'll see it climb fairly quick so change it when you return from your 3500 mile tril and see if it's different from the last change. If you didn't see any metal in the fluid you should be ok for this trip, but I'd watch it and change the fluid more frequently than recommended until you see the chocolate color disappear.

 

Dave

Link to comment
I've seen this on my 97 rt as well - it appears to be a result of the synthetic fluid not mixing with the original mineral based fluid and in my experience it will not go away with a few fluid changed.
Interesting. But then again I've swapped back & forth between standard and synthetic gear oil many times in my own bike and have never seen any kind of reaction.
Link to comment

Mu experience was with BMW synth gear lube if that matters. I also saw it on two trans and the lab that did the analysis could not explain it other than the two fluids...that doesn't mean it's gospel though. The lab did say that it wasn't moisture and it wasn't iron.

Link to comment

He's planning on a 3500 mile trip shortly. Cause for concern..?

 

 

Maybe..if I remember correctly,there were some threads several years ago,where the trans fluid had turned that color and the trans went out shortly thereafter.

 

You might try doing a search using "transmission fluid color" or maybe someone with a better memory than mine will chime in. tongue.gif

 

 

I believe the color change was due to the transmission seals melting eek.gif

Link to comment

On my R1100-RT, a change in the fluid color preceded the failure of my transmission. The discoloration I experienced was similar to milk chocolate. The failure I experienced was the helical gear on the input shaft contacting the ball bearing and eventually chewing throught the entire outer race leading tho catastrophic (to the tranny) failure.

What the best guess as to the oil discoloration cause was the release of the lubriction from the sealed bearing into the gear oil after the seal was ruptured. This was noticed approximately 700-1000 miles of hard riding before the failure occurred.

Attached is a photo showing the failed shaft next to a new shaft.

741632-R1100-RTTransmission014.jpg.d6a6173fc47e22982e7b86169930191a.jpg

Link to comment
  • 5 weeks later...

I am the unfortunate friend with the chocolate gear oil and did complete the 4300 mile trip on my '97 RT without incident. During my tour, downshifting gradually became less smooth. Towards the end I would occasionally have to release the clutch and try downshifting again due to the resistance on the shift lever.

Upon my return, I gathered a sample for analysis which also appeared similar to used motor oil (slightly more translucent.) Blackstone Labs replied promptly with the attached results.

It would appear that my transmission may very well have the same type of failure as SageRider. I will be pulling my tranny in the near future and will advise on the outcome.

753433-97RT-BlackstoneTransResults.jpg.d5818fb392b56b87d604e05c2264c254.jpg

Link to comment

OK, we have the above transmission apart and fortunately (or unfortunately?) there doesn't seem to be any obvious signs of a serious problem, although one would assume that something must be wrong based on the oil sample report. The only thing we can find that might be a problem is that the rear bearing on the input shaft feels a little rough compared to the other five bearings, which all turn a smoothly as a baby's bottom. So perhaps that one bearing is on the way out (although it would surprise me if it was making enough metal to contaminate the gear oil.)

 

There also does seem what might be the earliest stages of the problem that SageRider described in that there appears to be some slight scraping on the side of the rear input shaft bearing outer race. On close inspection it appears that this is not due to the drive gear on the input shaft rubbing against it (if you look very closely you can see that the side of this gear is relieved so that it can't possibly touch the outer race of the bearing), but rather the driven gear on the intermediate shaft contacting the rear input shaft bearing. Not sure why this would occur, perhaps a case of excessive end float in either the input or intremediate shaft. But in our case this contact looks very slight and we're not sure that it really represents a problem.

 

Other than the above everything looks great inside, shift forks OK, etc. Our quandary is whether to just replace the one bearing that seems bad (the transmisison has only 30k on it so it seems like overkill to replace all of the other bearings, especially since they feel perfect) and reinstall, or send it to a pro for evaluation?

Link to comment

While you have the tranny apart, you might want to consider replacing the oil seals. I have a 97 RT and had an oil seal leak at about 60K on my way back from Torrey XII. Found this out by having a slipping clutch that had to be replaced. Also the replacement seal was slightly different (so says the BMW mechanic) than the original. Goo news was that the splines were ok. Hopefully better. Just my .02

Link to comment

Ouch... looks like the only think left is the inner race! But also note that the helical gear next to the bearing (or in your case what once was the bearing) has a relief cut in it such that it really can't contact the outer race of the adjacent bearing. If I had to guess I would say that your transmission had a much worse case than what we have, i.e. the driven gear in the intermediate shaft in your transmission must have been really rubbing on the rear input shaft bearing. When you have the tranny apart it's easy to see just how possible this is if the shafts have excessive end float.

 

In our case, we're still wondering where all the metal might have come from unless the single slightly rough bearing on the input shaft is in worse shape than it seems. Also curious why only one bearing would start to go while the others seem fine, and it's the same bearing that disintigrated in Michael's transmission. There does seem to be something funny going on with that rear input shaft bearing on some transmissions.

 

BTW we also can't find any reason to explain the difficult downshifting. Current theory is that it was simply due to the very dry (but otherwise undamaged) splines we found.

Link to comment

On my bike, when we inspected the input shaft , it appeared that the beveling of the gear had been a result of contact with the outer race of the bearing. Very definite abrasion marks on the edge of the gear. All other parts were fine, including the intermediate shaft.

Enough heat was generated by the failing bearing that it melted the throwout bearing. This resulted in the pushrod snapping and no clutch disengagement.

 

{edit}

Also as shown in picture, there was zero clearance between the bearing inner race and the gear.

 

{edit2}

I also had difficulty in shifting prior to the failure. Splines on my bike was not an issue. Slop in the input shaft position due to the failing bearing would be my guess.

Link to comment
  • 2 weeks later...

Michael,

 

It looks like you are right, after removing the bearing it is apparent that the helical spur gear was pressing against (or had been pounding against) the inner race of the bearing so hard that the inner race metal had started to deform (and that's some hard metal!) The deformation had caused the bearing oil seal to fail and the leak was contaminating the rest of the gear oil. If left as it was for much longer the bearing itself would have eventually failed leading to a similar situation as what you faced. Luckily it was caught early in this case and all that seems to be needed is a new bearing. Lesson here for all is to not ignore chocolate-colored gear oil...

 

Also as shown in picture, there was zero clearance between the bearing inner race and the gear.
As far as I can tell, it's supposed to be that way... or at least there doesn't seem to be anything that would prevent these two from touching each other(?)

 

So I wonder why this failure is occuring in just a small number of transmissions..?

Link to comment
Lesson here for all is to not ignore chocolate-colored gear oil...

Yep! I was a little slow learning that lesson blush.gif !

 

 

So I wonder why this failure is occuring in just a small number of transmissions..?

It is a small number, but seems to be the most common 1100 transmission failure after splines. It would be interesting to see the shape of the bearing / gear in bikes that are being torn down for input shaft replacement due to spline issues. It might be (pure speculation here) that many of the bikes show some wear, just not to the point of failure.

Link to comment
  • 2 weeks later...

I am now sure that I would have ended up with the same results as the dreaded images of SageRider's input shaft depict. Although I have discovered and repaired the failed component in my transmission, I am still unsure of the cause. So, I put together a website detailing the results of my transmission failure.

See them at:

Input Shaft Rear Bearing Failure

I could really use some help so anyone with ideas as to the real problem here please chime in.

Link to comment
  • 6 months later...

Gonna light this thread up again because we have a lot of new info... and it's not good...

 

After repairing the failure described above and inspecting several oilhead transmissions of various vintages we (James and I) came up with a theory as to why this is happening, and perhaps how to rectify it. I'll get to that in a minute, but one nagging question we had is 'if our theory concerning the failure is correct then why don't all M97 transmissions exhibit this problem?

 

Flash forward to my transmission oil change last Saturday (on an M97 transmission with approx. 16k miles on it.) What do I see? Chocolate-colored gear oil! Crap! Not as bad as James's bike but definitely discolored in the same way. I first thought about sending an oil sample to Blackstone but then decided not to waste my money... I knew what was wrong...

 

So Sunday the transmission came out and we tore into it today. The attached pictures tell the story... (hint: the step on the gear face and bearing inner race shouldn't be there, both should be perfectly flat)

 

gear.jpg

bearing.jpg

 

 

The exact same problem as we, and several others in the thread, and countless unknown, have experienced. It is in the very early stages so I think the $175 helical gear is salvageable, but clearly this would have eventually degraded into a failure if the gear oil anomaly was ignored.

 

So bottom line is that we now believe this is an inherent design flaw and that all M97 transmissions will likely exhibit this problem sooner or later. Here's why:

 

On the M97 there is a mysterious wave washer (the subject of a few previous threads) on the input shaft, in addition to the large torque compensator spring. There has been a lot of speculation as to the purpose of the wave washer and after comparing various vintage transmissions we think we have the answer, and it has to do with oilhead transmission evolution and the various 'fixes' BMW made to try to remedy some various problems. One of these was the tendency of the M93 gearsets to rattle under no-load conditions, especially audible at idle. The fix for the M94 was the insertion of o-rings under the gears to prevent the rattle, but apparently BMW felt that this was lame and with the M97 model the o-rings were removed and the light wave washer added to the input shaft (the wave washer is not present on pre-M97 transmissions.) The wave washer is very light and is compressed instantly as any significant power is applied to the input shaft (due to the action of the torque compensator cam), however under 'no load' conditions the power variations applied to the input shaft (by the large twin-cylinder engine connected to it) are absorbed the the wave washer via the same mechanism that large power variations are absorbed by the large torque compensator spring. There are a couple of problems with this, and one big one. First, the wave washer sees a lot of compression duty and has been known to fracture and fail. Second, and more serious, is the fact that the action of the wave washer allows the helical drive gear to constantly slam into the rear input shaft bearing. Normally the rear input shaft bearing turns at the same speed as the helical gear so no wear should occur due to any circular sliding action, but the axial hammering allowed by the wave washer eventually pounds the rear bearing to death. This is evidenced by the wear pattern in which you can see the clearly-defined marks (180 degrees apart) caused by the oil-relief channels on the helical gear. These marks would not be evident if the two parts were wearing due to a circular sliding motion but of course would be present if the wear was caused by hammering in an axial plane. Another confirming bit of evidence is that an input shaft from a pre-M97 transmission (without the wave washer) showed absolutely no wear at the gear/bearing interface, even at 60k miles. As I mentioned previously, I'm pretty certain that all input shafts with the wave washer will develop this problem sooner or later, and apparently it doesn't take a lot of miles for the problem to show (although it may take many miles for the bearing to actually hard fail.) Note that in the early the early stages of failure the damage isn't visible until you disassemble the input shaft, other than the dark-colored gear oil caused by failure of the rear bearing oil seal (which allows the bearing grease into the gear oil, giving it a darker color.)

 

How to remedy? Simply remove the wave washer and substitute an approx. 3mm thick flat washer in its place. This will be the functional duplicate of a pre-M97 input shaft and should hopefully prevent a recurrence of the problem. This is exactly what we have done with two transmissions with no apparent problem (although there is a very slight rattle in neutral in one of them, as might be expected due to removal of the wave washer.) Here are some pre- and post-mod pics.

 

There is also one other possibility that would explain the problem existing in the M97 but not prior, and that is the switch to caged roller bearings in the M97 vs. tapered roller bearings in previous versions. It could be that the tapered roller bearings provide a bit of a cushion, i.e. the hammering action causes small changes in the preload that absorb the energy, saving the bearing inner race from the punishment. The caged roller bearings used in the M97 would allow no such play, placing the entire stress on the bearing inner race. I think the wave washer explanation is more likely though as the heavy pre-M97 input shaft torque compensator spring places a lot of tension on the gear/bearing interface, preventing that from being the source of the hammering action (or so I would think.) OTOH the wave washer would allow an almost constant hammering action at idle or neutral-throttle conditions.

 

Anyway, we hate to come up with such a dark conclusion regarding this issue of M97 design, but we don't really know what else to think. We'll know for sure regarding the fix in a little while... as soon as we get maybe another 20k on one of the modified transmissions we're going to tear it down again and see if removing the wave washer cures the problem. In the meantime, if you have a 1997 or later 1100 (I don't know if this issue is present in the 1150 transmission, maybe) then carefully examine the oil drained after a transmission oil change and make sure that it looks like the new oil going in. If it is noticeably darker then you likely have trouble brewing.

 

 

Link to comment
I've seen this on my 97 rt as well - it appears to be a result of the synthetic fluid not mixing with the original mineral based fluid and in my experience it will not go away with a few fluid changed.

I doubt that this is the problem. Synthetic oil MUST be completely intermixable with Dino oil according to basic oil industry requirements. In order to be offered for sale, all oils must be completely miscible with each other.

 

Besides, most so-called "synthetics" are in fact made mostly from highly refined Group 3 "Dino" oil stock.

Link to comment
I doubt that this is the problem.
Correct, it isn't, or at least not in this case. Once you disassemble the input shaft it's clear that the contaminant is coming from the grease pack in the damaged rear input shaft bearing.
Link to comment

Another one ! 69000mls.

Just bought this @#$#@$@ '98 and I'm pissed bncry.gif

Changed the tranny oil after reading this thread and wondering if this bearing failure is the reason for all the noise I've been hearing & YEP - CHOCOLATE !!

Not a lot of metal though, so maybe just the seals and Smillers wave washer replacement? The oil does contain fine black particles that I assume to be the seal coming apart.

How many miles do you have on that trans with the washer change Smiller? I need to know what I'm in for here before I tear it apart.

Damn - I was intending a Colorado trip in August.

Link to comment
Another one ! 69000mls.
Unfortunately I think we will see more and more of these now that we're beginning to understand what is going on. As I said I can't think of a single reason why all M97 transmissions will not exhibit this failure sooner or later.

 

Luckily if caught early there shouldn't be any serious damage done. Hopefully your drive gear will be salvageable, or if not you're looking at $175-$200 for a new one. The bearing is only about $20 or so at your nearest bearing supplier, plus I'd replace all the seals while the tranny is open. If you're capable of doing the work yourself then you should be good to go well before August.

 

We have built two transmissions without the wave washer so far. One has only about 5k on it so too early to tear it back apart for a follow-up analysis. I just put the other into my bike last week. Both are running without any problems. And I'm putting a third together right now. I believe that I've found a source for the required spacer washer, will know for sure when it shows up later this week. Maybe I should put together some repair kits... smirk.gif

Link to comment

(I believe that I've found a source for the required spacer washer, will know for sure when it shows up later this week. Maybe I should put together some repair kits... )

 

I thought the washer was a duplicate of the one that is already there?

Are the seals common units too?

A kit you say - Hmmmm ! I'd be interested. $?

My closest bmw dealer is 85 miles away so I don't want to make that trip if I don't have to.

My tranny is coming out tomorrow, I gave myself today to cool down !!!

Link to comment
I thought the washer was a duplicate of the one that is already there?
Yes, that's correct, but for some reason known only to BMW the individual parts for an M97 input shaft are not shown on the fiche... only the entire shaft assembly at mucho $$.... so I don't know if that washer is available separately, or at least I don't know how to get a p/n for it. But individual parts are shown for the M93/M94 input shaft, and a washer used on that shaft (28x44x3, p/n 23211340794) would seem to fill the bill. The ID (28mm) and thickness (3mm) are right but the diameter is a few mm less... but still OK for this application since the only function it is performing is a spacer. Also note that if your helical gear is munged beyond redemption you can order the 20-tooth gear from the M93/M94 input shaft (p/n 23212345502) as it is the same part used on the M97 shaft.

 

Are the seals common units too?
Yes, but rather than hunt them down I just ordered them from BMW since I needed the spacer washer anyway.
Link to comment

Thanks for the part numbers Smiller - makes me feel a little better knowing guys like you have done this before, and share discovery. Kudos to you and James !!

Link to comment

Looking at the gear, I would like to ask why you think the gear is salvageable? Wouldn't the reduced contact area between the gear and the bearing be a concern?

Just askin'.

Link to comment
Looking at the gear, I would like to ask why you think the gear is salvageable? Wouldn't the reduced contact area between the gear and the bearing be a concern?
The depth of the dished area on the face of the gear was only a few thousandths deep and I was able to mill the gear face flat. That will move the gear rearward a few thousands when installed on the shaft but I don't believe that this will be harmful. Regarding reconditioning the gear though, keep in mind that this transmission had relatively low (under 20k) miles. James's bike was in the 40k range and his gear was was past the point that it could be reconditioned. Unfortunately I think that Mark's bike (at almost 70k) will probably need a new gear as well.

 

BTW, there would be no change in contact area either way. The bearing inner race only contacts about half of the gear face! That's part of the problem as any hammering force is concentrated over a relatively small area rather than the entire gear face, making it easier to reform the metal. I cleaned up the gear face only so that I could see whether (during a subsequent teardown) removing the wave washer prevented any further damage (if I had left the dished area then I wouldn't be able to tell very easily.) But the lack of a complete mating surface between the bearing and gear wouldn't change either way.

 

We looked at how we might remedy this with some machine work, but in the end decided that it wasn't necessary... if we could eliminate the hammering action then that should eliminate the problem as there should be no other wear between these two surfaces (since they both rotate at the same speed on the shaft.)

Link to comment
Bart Anderson

Seth, I gotta say that I'm not sold on your theory that the wave washer causes this wear/damage. It just seems too far away from the gear and bearing, and too puny anyway. It's on the other side of the cush link from where the damage is occuring, and just seems too small to affect the big steel pieces on other end in such a dramatic fashion.

 

I have an alternate theory: I think the root cause is the wide spec BMW allows for the end float of all three shafts, especially the input shaft. With that helical gear, the forward-and-back force on that shaft with acceleration and engine braking has to be tremendous. If the shaft floats axially at all within the case, all that pressure and momentum will be constantly hammering the gear and bearing into the back of the case, and I think that's the primary cause of the wear you're seeing on the gear and inner bearing race.

 

BMW calls for 0.05-0.15mm of float...that's too much. (that's a valve gap spec!) My version of the manual doesn't give the shaft float spec for the pre-M97 tranny, since they want everyone to do the "clean bearing" conversion. Given that they were tapered bearings in the pre-M97s I'm going to guess that it was much tighter than the current spec.

 

It would be interesting to know how much end float existed on these failed trannys. Were there any shims on the input shaft? I could see how over time the float would increase as the gear eats into the bearing, leading to earlier failure.

 

Assuming you shimmed the test case tightly, I'm going to predict that I will not show the wear you're looking for. I'm just not sure the removal of the wave washer will be the reason.

 

...Normally the rear input shaft bearing turns at the same speed as the helical gear so no wear should occur due to any circular sliding action...

 

I think the gear floats on the shaft, being on the output side of the cush spring/cam assembly, so its motion would not (should not) be exactly the same as the inner race of the pressed-on bearing. As I understand it, it's the input side of the cush drive that's splined to the shaft and would move in sync with the bearing. Those pictures of yours are very interesting indeed, if that's the case.

 

Maybe the axial shaft jacking only occurs under an extreme load situation, with the cush drive's cam already at its limit, so the gear ends up in the same spot relative to the bearing race when impact occurs?

 

Your thoughts?

Link to comment

Hi Bart, how ya doing?

 

My thoughts follow...

 

 

It just seems too far away from the gear and bearing, and too puny anyway. It's on the other side of the cush link from where the damage is occuring, and just seems too small to affect the big steel pieces on other end in such a dramatic fashion.

 

The distance from the wave washer to the gear doesn't matter as the entire spring/cam assembly floats freely on the shaft.

 

If the shaft floats axially at all within the case, all that pressure and momentum will be constantly hammering the gear and bearing into the back of the case, and I think that's the primary cause of the wear you're seeing on the gear and inner bearing race.

 

I doubt very much that 0.15mm of float could allow enough inertia to build to cause this sort if problem. But even if it could it wouldn't have the effect you describe because both front and rear bearings are installed against steps machined into the input shaft, so forces generated by end play would only be applied against the bearing inner race and the step on the the input shaft. This junction shows no wear problem at either end of the shaft. Lastly, we found nothing really amiss in the end float measurements taken using the stock shim pack.

 

I could see how over time the float would increase as the gear eats into the bearing, leading to earlier failure.

 

Due the the above-mentioned steps on the input shaft wear between the gear and the inner face of the bearing won't affect end float at all.

 

I think the gear floats on the shaft, being on the output side of the cush spring/cam assembly, so its motion would not (should not) be exactly the same as the inner race of the pressed-on bearing.

 

Correct, not always exactly, as a slight difference in rotation is possible due to the torque compensator cam. But it does not appear that this slight motion is the source of the wear because if it were we would not see the clearly defined oil-relief channel on the gear so perfectly transferred to the bearing. Plus, no such wear was evident on the pre-M97 transmission (before the wave washer was installed.)

 

I dunno what to say... it's kind of tough to visualize and ever harder to explain, but if you had the parts laid out in front of you (as I do at this moment) I think you'd agree that excessive end float is not a very likely cause of this particular problem.

Link to comment

Hi Seth,

Have you given any thought to the installation of your 3mm washer between the offending helical gear and the bearing. I don't know if this is even feasable, or if everything can be slid over after removing the wave washer. I'm close to getting my tranny out and was thinking about this hammering theory. It seems to me that for the outer teeth of the gear to contact the bearings outer race, as some of these pics clearly show, the bearing would have to cave tremendously. Any clearance that can be given in this area should be a plus. Surely the ball bearing would be wasted in fairly short order if it is being side loaded like this?

A big WHY, is why doesn't Getrag or BMW have an opinion or "fix"? Do you believe they are aware of this, or care?

After all, some of these problematic transmissions have 20K or less !

Going back to basic mechanical bearing theory learned 30 years ago, ball bearings support top load, and tapered rollers can support top and side loads. Helical gears generate side load in both directions with power on/off, so another big WHY is why are the ball bearings there in the first place? Or why is it a helical gear? I know they are quieter than straight cut, but I don't think lowering noise levels was ever in the minds of an RT engineer lmao.gif

 

You know, I always thought if I bought a BMW, this kind of BS would not exist.

Link to comment

Have you given any thought to the installation of your 3mm washer between the offending helical gear and the bearing.

 

Not possible because that would move the gear itself 3mm which would mess up the alignment between it and the mating gear on the intermediate shaft. We did consider the ideal of machining 3mm off off the face of the gear and inserting the washer (which would maintain alignment and provide full support for the bearing face, thus spreading out the load and perhaps preventing the problem), but it just seemed unnecessary... preventing the hammering action in the first place ought to solve the issue, as evidenced by our seeing no trace of it in the pre-M97 tranny. But if removing the wave washer doesn't cure the problem then we may have to consider a step like this.

 

It seems to me that for the outer teeth of the gear to contact the bearings outer race, as some of these pics clearly show, the bearing would have to cave tremendously.

 

Which it does. eek.gif

 

Surely the ball bearing would be wasted in fairly short order if it is being side loaded like this?

 

Yes and no. As the inner race is distorted the seal fails rapidly, but that in itself isn't much of a problem because the gear oil that enters is a pretty good lubricant. And while the external face of the inner race gets pretty munged the ball bearings themselves tend to keep it from totally collapsing. From the examples I've seen it appears that the bearing could run quite a while in this condition before a hard failure occurs (although it will certainly fail eventually, ask Michael.) That is why I think this problem may be a lot more prevalent than it appears.

 

A big WHY, is why doesn't Getrag or BMW have an opinion or "fix"? Do you believe they are aware of this, or care?

 

I would think that they would have to be aware but of course we can't know for sure. Even if so I doubt BMW would rush to offer a recall that 1) would be very expensive for them to implement, and 2) in most cases only affects bikes out of warranty. So it could be either of 'don't know' or 'don't care'... but either way it kind of sucks.

 

so another big WHY is why are the ball bearings there in the first place?

 

Yes, tapered bearings would be a better choice, and that is what the M93/M94 transmissions used. Why did they 'upgrade' (BMW's words) to roller bearings? Who knows. My suspicion is that it has to do with cost of assembly... the procedure for setting the preload of the tapered roller bearings is kind of time consuming, or at least much more so than selecting shims for the roller bearings. It could have been nothing more than a cost-saving measure on the assembly line, or there could have been a good engineering reason (although I'm at a loss to come up with an example of the latter, especially since AFAIK there wasn't any particular problem with failure of the original tapered bearings.) Either way what they ended up with is a pretty poor design, at least in this one specific area. To be fair though I can't say that I would ever have anticipated this problem merely by looking at the new parts and I could see how the potential could slip by the engineers.

 

On the bright side though, with the other M97 improvements (and the exception of this problem, which we will come up with an answer for, one way or the other) the BMW gearbox is rather stout for a motorcycle transmission and should be good for many miles. Or at least here's to hoping... wink.gif

Link to comment

I originally thought this was perhaps a case of bearings that were improperly hardened for the application. Say BMW had a sound idea (in using the wave washer to correct some noise problem or whatever), but then a batch of bearings comes through that can't take the stress, thereby allowing the face to get torn up, thereby allowing the gear to come in contact with the bearing outer race, thereby releasing metal into the oil, thereby....But then I can't imagine any wave washer being strong enough to resist the sidewards motion of the assembly that you describe, which could have prevented the bearing-gear contact. Isn't the the bearing press fit onto the shaft, with a lip determining its location? Why didn't they put a lip for the gear to ride up against also?

I guess these problems could be a result of Getrag not being used to building these transmissions for the additional horsepower of the oilheads. Or maybe the new engineers were originally from the old Peoples Democratic Republik, where 45 hp was a powerful motor.

Link to comment
Why didn't they put a lip for the gear to ride up against also?
Ya got me, who knows. That would certainly be a better design.

 

There are several (manufacturing related) fixes one could employ to solve the problem, I'm just trying to propose something that is practical and easy for the home mechanic to implement.

Link to comment

Seth, James bike has 5K on it now - have you changed the tranny oil? I'm assuming you are looking to see if you get chocolate, or shavings, at some point? How many miles do you anticipate letting it go before you dive into it again?

I have my tranny sitting on my workbench, and I'm working up the gumption to tear it apart - perhaps you can help me in that regard as this will be the first motorcycle trans. for me - I've done several manual cars, but many years ago!

Can I remove the 15 bolts holding the front case half and just separate the halves. The manual wants me to remove the neutral switch, retaining ring and clutch lever assy from the other end - is this necessary?

Any procedure tips once apart would be appreciated.

Hey, you two guys should make a procedure movie like "Teds spline lube dvd" I have this and it was helpful !

Link to comment
Seth, James bike has 5K on it now - have you changed the tranny oil? I'm assuming you are looking to see if you get chocolate, or shavings, at some point? How many miles do you anticipate letting it go before you dive into it again?

I have my tranny sitting on my workbench, and I'm working up the gumption to tear it apart - perhaps you can help me in that regard as this will be the first motorcycle trans. for me - I've done several manual cars, but many years ago!

Can I remove the 15 bolts holding the front case half and just separate the halves. The manual wants me to remove the neutral switch, retaining ring and clutch lever assy from the other end - is this necessary?

Any procedure tips once apart would be appreciated.

The oil has been changed twice, looks good (no coloration) so far. We probably won't go back into the transmission until at least 20k miles are on it.

 

Re: opening the transmission, remove the cover bolts (all of them, there are a few that like to hide) and the large bolt that holds the neutral detent spring and ball. Invert the transmission and remove these so you don't lose them when you pull the cover off. The cover should come off simply by tapping it in the appropriate places with a rubber mallet, or if it is sticky a little heat will help release it. If it won't budge then stop and check for any hiding bolts. And yes, you will need to remove everything from the back of the transmission. This is not necessary just to remove the cover but it will have to be done to allow disassembly of the internals, so you might as well get it out of the way early while your hands are still clean (they won't be that way for long... grin.gif)

 

Everything comes apart as detailed in the manual, just remember heat is your friend. The gearshafts won't come out of their bores very easily unless you heat the case first, and once you do they should lift out easily. The fun part is getting them back in.

 

Make sure you keep track of what shims go where when you pull the cover. Since all you will be doing (we hope) is replacing the rear input shaft bearing (and maybe the gear) then you should be able to re-use the existing shim packs. If you feel a burning desire to check the shimming there are methods to do so without the BMW test jig but they are a little too complex to go into here, plus you will need some special tools such as a depth micrometer set.

 

When you have the input shaft in your hand the bearing damage may or may not be evident until you pull the rear bearing. This is easily done with a gear puller (pull on the helical gear and use it to force off the press-fit bearing.) Let me know when you've gotten that far and we'll see where we need to go from there.

 

Cardinal rules: Go slowly, pay attention to how all the parts fit together, and don't force anything!

Link to comment
Bart Anderson

Ok, I'm not married to the idea of the shaft float being the cause, but I'm still not ready to head for the altar with the wave washer theory. I must say that I'm really enjoying the discussion. It's an interesting mystery.

 

...both front and rear bearings are installed against steps machined into the input shaft, so forces generated by end play would only be applied against the bearing inner race and the step on the the input shaft....

 

I didn't mean to imply that the bearings are moving around on the shaft; quite the contrary: I was proposing that the shaft and inner race are slamming against the gear as a unit.

 

Your fantastic pictures tell the tale. The perfectly transferred image of the gear's oil relief channel on the inner race of the bearing tells us a couple of things. First, that the wear is not rotational. If it's not rotational, it must be axial. Second, it tells us that the gear is in the exact same position relative to the shaft/bearing race when this axial impact occurs.

 

Since the gear floats on the shaft, and the inner race is fixed on the shaft, and the impact point on both pieces seems fixed, then it would seem that the impact damage is going to occur at one of three likely positions relative to the cush cam: in the center, or at one of the two extremes. In the center implies either a no-load (or relative twist) condition or something that happens in the instant of transition between loads, as the cam moves between the two ends of its groove. As the cam climbs the groove in either direction, it must reach a point where the spring is fully compressed; the cam cannot move any further up the groove because of the compressed spring, but it will want to. At some point, as the evidence shows, the gear is then axially slamming into the inner race of the bearing suddenly, rather than sliding/spinning against it like it does when it's climbing the cam.

 

The cam groove does get shallow very quickly toward the top, but we know that it can't actually reach the top because the complete compression of the spring will stop it. At this point, what would cause the gear to slam into the bearing race with enough force to deform them both? (or is it the bearing race slamming into the gear?)

 

The wave washer theory seems to imply that the axial hammering is happening with the cam in the center of its groove. I have a complete input shaft in my hands, and I can move the shaft and cam freely through several degrees of rotation just by hand; one must assume that when the engine is idling that cam is moving back and forth through several degrees of twist with every revolution. Unless the wave washer causes some sort of slingshot action at the bottom of the cam groove, I just don't see what would be causing enough axial motion and force to deform the inner bearing race like that.

 

Is there any way to know, with the trannys you've looked at, where the indentation on the bearing race tells us the gear is when the impact occurs (relative to the cam)? Is in in the center, or at one of the ends of travel?

 

Again, this is a fascinating disucssion.

Link to comment

Unless the wave washer causes some sort of slingshot action at the bottom of the cam groove

 

That's essentially what I think is happening. The power input variations (at idle) may not be enough to actuate the cam very much, but still enough to move the (relatively heavy) mass of the entire spring/cam/gear assembly against the weak wave washer and cause the hammering action.

 

Is there any way to know, with the trannys you've looked at, where the indentation on the bearing race tells us the gear is when the impact occurs (relative to the cam)? Is in in the center, or at one of the ends of travel?

 

Nothing that would be certain. If my supposition above is incorrect then the cam should allow the gear to rattle back and forth equally over the entire +/- 2 (or whatever) degree range allowed by the wave washer, and the fact that the damage seems limited to one spot (evidenced by the fact that the bump in the bearing race is about the same size as the oil relief cutout on the gear, vs. being elongated) makes me suspicious. I believe that this spot is the center, but I don't know that for sure. It's possible that it is at one full limit of wave washer compression or the other, but if so I don't know what would cause that in a consistent fashion, and even if so... the wave washer is still the culprit.

 

FWIW I'm not wedded to this explanation either, it just seems like the most likely cause to me, or at least I'm having trouble visualizing how such a small amount of end float could cause it. Not that my inability to visualize something means it isn't true... wink.gif

 

In any event it sure is a fascinating mystery (if you're into this stuff... I'm sure there a lot of readers with their eyes glazing over as they read this too... grin.gif)

 

One thing is for certain... we need to get to the bottom of this and find a cure, whatever it may be. I'm not going to be replacing my rear input shaft bearing every 20-40 k miles!

Link to comment

OK, got it apart and I am not seeing anything wrong! confused.gifconfused.gif There is no ground worn edge to the helical gear teeth and no apparent contact with the outer race. Now admittedly I have not separated the bearing from the shaft yet, but from all appearances ???

I have pics I'd love to post, but do not know the best way to do so. The detail will require the max resolution my camera can give, but the site will not accept it (too large). How are you guys posting those great pics?

Link to comment
OK, got it apart and I am not seeing anything wrong! There is no ground worn edge to the helical gear teeth and no apparent contact with the outer race.
The gear won't contact the outer race until it gets pretty bad, otherwise the damage may not be visible until the bearing is removed. For instance, this was the case for the pictures I posed earlier in the thread. Everything looked perfect until I removed the bearing and had a look directly at the face of the gear and bearing inner race.
Link to comment

You are exactly right. I have it separated and it looks no worse than your most recent trans pics. (see attachment.)

I expected that with almost 70K, this trans would show a lot of wear !

One thing I did notice was that the wave washer, even though it is fairly weak, puts a constant tension on the wearing surfaces, allowing them to grind together as the cam moves through its action. I'm assuming your replacement flat washer somewhat eliminates this tensioned contact?

Another thing, are we assuming the heat being generated by this friction is causing the bearings seal to fail, or the physical distortion of the bearing?

Whatever is going on here, it is obvious Getrag underestimated the Strength of these metal surfaces.

Seth, what was the depth of the wear groove on your gear that made you decide to save it by refacing the surface?

Mine is .006". What do you think ?

864178-100_2660.jpg.43baca59d62a6b80ad13038acea673ea.jpg

Link to comment

AHA, another realization ! If I push hard enough to about completely collapse the wave washer, I can create at least .010" gap between the gear and bearing. Definitely enough to create some hammer damage ...yes ??

Link to comment

Yep, there was no doubt. frown.gif

 

Your result is interesting Mark in there appears to be relatively minimal damage for 70k miles. Not sure why but it may help explain why we aren't seeing M97 trannies dropping like flies, i.e. in some (many? most?) examples it apparently takes a while for the problem to show, and if you ignored the darker gear oil (as many people may be doing) then who knows, the transmission might have gone to 100k before it failed. So that kind of reaffirms my belief that this issue is present in all M97 models to one extent or another, although for an unknown reason some have problems at much lower mileages than others.

 

My guess (emphasis on guess) would be that your gear is salvageable. Have it milled or cut flat on a lathe (or just oil up some 400 grit carbide sandpaper and spend an hour of elbow grease, that's what I did. Rotate the gear frequently as you go) and you should be OK. Don't take off any more metal than you absolutely have to, though.

 

Check all your other bearings by rotating them by hand and feeling for any roughness. If they seem OK then they probably are. Some would advise you to replace them all while you're in there and that's a judgment call.

 

Otherwise you will just need a new set of seals and a replacement bearing, and a washer if you choose to buy my wave washer theory. No guarantees there either but the way I look at it, couldn't hurt. Or shim it up tight as well if you follow Bart's view. We're all test pilots here.

 

AHA, another realization ! If I push hard enough to about completely collapse the wave washer, I can create at least .010" gap between the gear and bearing. Definitely enough to create some hammer damage ...yes ??
Bingo.

 

BTW the 3mm spacer (in place of the wave washer) doesn't decrease the tension between the gear and bearing, rather it increases it, which is the whole point... with increased tension the gear and bearing cannot separate during operation. There is enough lubrication to handle the slight rotational movement caused by the cam action of the large torque compensator spring. The M93/M94 shaft is under a lot of tension but showed no wear at all.

Link to comment
From the examples I've seen it appears that the bearing could run quite a while in this condition before a hard failure occurs (although it will certainly fail eventually, ask Michael.)

At least 1K miles between chocolate gear oil discovery during service of my tranny to failure. Theoretically, there could be as much as 6K miles additional between seal rupture and failure. Tranny was run hard for the last 1K miles with a lot of triple digit speed / high load involved (to/from Torrey!)

 

Your result is interesting Mark in there appears to be relatively minimal damage for 70k miles. Not sure why but it may help explain why we aren't seeing M97 trannies dropping like flies, i.e. in some (many? most?) examples it apparently takes a while for the problem to show, and if you ignored the darker gear oil (as many people may be doing) then who knows, the transmission might have gone to 100k before it failed.

I suspect this is directly related to how hard the bike was used. In all the cases I am personally familiar with, the bike was used very hard with lots of high speed and aggressive acceleration / deceleration.

Link to comment

I'd like to buy the wave washer theory as it makes sense to me, but I'm having a hard time believing that Seths trans. at 16K, and mine @ 70K, have essentially the same wear marks. I'm thinking you're both right and sloppy assembly with regards to quality control in shimming clearances, explains the low mileage failures.

I dunno - its been a long day & my head hurts ! dopeslap.gif

Link to comment

but I'm having a hard time believing that Seths trans. at 16K, and mine @ 70K, have essentially the same wear marks.

 

I should note that the '16k' transmission was a used unit and that was mileage was reported to me, but I don't know for certain that it is correct. (It appeared to be a low-mileage unit though as the shift forks showed virtually no wear.) Regardless, there seems to be quite a variation as James's transmission had around 40k and it was much worse than yours.

 

I'm thinking you're both right and sloppy assembly with regards to quality control in shimming clearances, explains the low mileage failures.

 

I'll be measuring the end play clearance of my transmission (the reportedly low-mileage one) this weekend and will find out if it is excessive.

 

I dunno - its been a long day & my head hurts !

 

Yeah, but you just saved yourself a lot of money...

Link to comment

I'm now wondering if there is a stronger version of this NTN6304CU bearing available. I'll ask my local experts today. It's doubtful there is a difference in the metal, but maybe a stronger ball/race/cage version is available. Or maybe a better seal in the bearing? If so, this must help withstand whatever is going on !!

I think I'll get the seals while I'm there, so all BMW can get me for is this washer/spacer !

Have you heard back from your possible supplier on that Seth?

One thing that has me thinking - How is the relatively weak alloy case putting up with this hammering? Some energy must be getting transferred to the outer race, surely Sageriders disaster ruined the case - yes?

Link to comment

There seems to be a bit of a delay in getting the washer. No surprise there as it's probably not a very commonly-ordered part.

 

In cases where the bearing actually fails (as in Michael's case) there may be some damage to the input shaft itself or case (in which case the transmission would be toast in terms of a cost-effective repair) but short of this I haven't seen any damage whatsoever beyond the gear and bearing so I don't think that you have much to worry about there, especially with the relatively mild damage in your unit.

Link to comment

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...