[CharlieVT]

Pictures of recent "preemptive" rebuild posted for the FD's owner and anyone else who might be interested.

This FD is off a 2001 with 56K miles. The owner had the FD off the bike to do other things, and decided to have me do a preemptive rebuild. Owner reported no metal flakes in the drive lube and no problems with the drive.

He is lucky, this drive was going to fail. And given the condition of the shims, I am very suspicious that this FD had a catatrophic failure sometime in the past and a new bearing was installed.

The bearing looked fine, but after removal from the crowngear a slight roughness could be felt when rotating by hand. Upon cutting the bearing open two significant areas of pitting on the inner race were readily visible. The outer race and balls looked okay, but my inspection of the balls was rather cursory; given the condition of the inner race, it was just a matter of time before total failure.

I don't know if the current owner is the original owner and I have no additional information regarding the FD history at this time. But I am curious. The shims looked like shims that have been through catastrophic failure which the current bearing did not experience. This makes me think that the FD was rebuilt at one time using the existing shims. The 0.60mm shim has the scoring around its circumference consistent with the bearing spinning in its seat, and there is a pitted area the 0.60mm shim where it got torn during spinning.

Likewise, the 0.15mm shim has an area that is shredded, again consistent with a spun bearing:

 

[DanMartin]

Was it overshimmed?

Dan

 

[CharlieVT]

Was it overshimmed?

Dan

Original shims were 0.60mm + 0.15mm = 0.75mm

My calculations gave 0.60mm shim.

Therefore, my conclusion is it was probably overshimmed by 0.15mm.

Thanks Dan :thumb:

 

[deputy5211]

Curtis... Could have used you on my forensics unit back in the day!

Nice analysis, and a great save from a possible catastrophic failure.

 

[K100Dennis]

On the money again Curtis, typical inner raceway localised spalling. The inner raceway tends to exhibit break up before the outer when the arrangement has a stationary outer ring. Later on though, both raceways become toast from the contamination getting rolled through the bearing.
You must have quite a data base of failure modes by now. The managenment at the Bavarian establishment need to seek out that rascall overshimmer and deaL with him, or could that possibly even be a her, who knows.

 

[lesn]

Great picks and great write up.
For all practical purposes--this FD has already failed--the owner just didn't know it yet.
Thanks Curtis !

 

[edmann]

Hi guys, I'm the owner of that final drive. I bought this LT in the crate from the BMW dealer in Pittsburgh in 2000. I have never had any trouble with my final or had any work done to it. I had it off because I was replacing the slave cylinder and I thought this would be a good time to do a pre-emptive repair. And obviously, I am very happy with that decision. Bike has 57000 miles. Thanks Curtis..Ed

 

[CharlieVT]

Given that we now know that this FD was never worked on since factory new, I have to conclude that the bearing outer race must have spun in its seat. That's the only way I can explain the condition of the shims. The is some slight scoring on the outer aspect of the outer race consistent with it having spun in its seat.

I wouldn't have guessed that those pitted areas on the inner race would be enough to cause the outer race to spin.

Any of you mechanical engineers confirm or dismiss this theory?

Thanks.

 

[hschisler]

This is interesting info.

Is there a mileage at which you would say it is too early to do a preemptive rebuild?
For example, what about an '09 LT with 12k miles? Is it "worth it" to check it this early...? It's still under warranty for 6 months.

 

[CharlieVT]

This is interesting info.

Is there a mileage at which you would say it is too early to do a preemptive rebuild?
For example, what about an '09 LT with 12k miles? Is it "worth it" to check it this early...? It's still under warranty for 6 months.

I wouldn't suggest a preemptive rebuild on any year post 2005.

2005 models are worth looking into because of the incidence of the tapered roller bearing issue.

99-02 seem to be the bikes at greatest risk for the "classic" crownwheel bearing problem shown in this thread. Failures on these years has happened mostly between 20 and 40K miles.

But that's just my Scientific Wild Guess.

Nobody knows the answer to the question: "when to do a preemptive rebuild?". It is a personal decision, like deciding how much insurance coverage you want to pay for.

 

[K100Dennis]

On the issue of the outer raceway creep, the condition normally comes down to housing fit and bearing loads. In our final drives the steel outer raceway is fitted into the cast aluminium alloy gear casing. Both these items have vastly different thermal expansion properties so it is important that the initial fit specification is correct for purpose. (BMW would never release that technical data) This (purpose) needs to take into account a number of things such as housing design, cooling, operational speeds, surrounding structures effect on bearing internal clearanace and lubrication. So as you can see, this is a complex issue, but in this case I would think housing fit, whether initial or after a rebuild (when some tolerance may be lost) would be a major contributor. Are there scores in the housing ? If so, the creep has been excessive and a rebuild may not be 100% reliable unless some mitigating action can be taken during assembly. If the bearing does not need to move axially to compensate for expansion then it may be acceptable to use a Loctite retaining compound to address the issue. Hope this helps rather than confuses. Dennis.

 

[11862]

I wouldn't have guessed that those pitted areas on the inner race would be enough to cause the outer race to spin.

Any of you mechanical engineers confirm or dismiss this theory.

My guess is the outer race may have excess clearance between it and the spindle housing bore. The fluid would then make a hydrodynamic wedge of oil between the bearing outer race OD and the housing bore.as the housing rotates. This wouldn't have happened if the shaft rotated instead.

Maybe you could do a final assembly with some locktite between the outer race & the housing.

I notice that one of the shims (the thin one?) could look like it was thinking of buckling. Was it next to the face gap in the housing by chance?

Failures by others are so easy to analyze after the fact ................. :histerica

 

[tips727]

This is interesting info.

Is there a mileage at which you would say it is too early to do a preemptive rebuild?
For example, what about an '09 LT with 12k miles? Is it "worth it" to check it this early...? It's still under warranty for 6 months.

I am wondering what kind of warranty you have that would pay for a preemptive rebuild. I was thinking that anything preemptive would be at the owners expense regardless of where the work is done. And would you not risk voiding any subsequent warranty if you do the work yourself? It's been a long time since I owned anything with a warranty and that was a Ford - they didn't want to fix much of anything until the warranty was expired. Maybe warranties are a lot better these days.

 

[hschisler]

I am wondering what kind of warranty you have that would pay for a preemptive rebuild. I was thinking that anything preemptive would be at the owners expense regardless of where the work is done. And would you not risk voiding any subsequent warranty if you do the work yourself? It's been a long time since I owned anything with a warranty and that was a Ford - they didn't want to fix much of anything until the warranty was expired. Maybe warranties are a lot better these days.

A pre-emptive rebuild wouldn't be covered under the warranty; I confused the issue by adding that last statement. What I was trying to say was: If I did a rebuild on an '09 LT's FD I would wait until after the warranty expired.

 

[K100Dennis]

I'm going to appease my paranoia of a FD failure by continuing to check the magnetic plug for anything other than slight fuzz and checking the rear wheel for signs of movement. There must be some lead time to failure (pF curve in reliability speak), the trick is in understanding how far down that curve I am before I intervene. Oil sampling and analysis would give good information but each test is around $30 here in Oz.
Then I'll box it up and send it to Curtis 'cause he has all the jigs, shims and tools as well as the knowledge & experience, this being his subject of passion. Sounds to me like most of the over-shimmed drives need about 0.15mm removed (previous posts and discussion).

 

[CharlieVT]

Dennis,

Not confusing, rather a most helpful post, thank you.
The whole concept of the outer race creeping was one I hadn't really entertained, but now makes a lot of sense. I was thinking more in terms of an abrupt spinning of the race.

Excess axial loading on the bearing due to overshimming, combined with the pitting of races, during hot running temps, coupled with a possible interference fit on the light side, resulting in the bearing creeping in its seat makes sense to me.

Neil, to answer your question: "I notice that one of the shims (the thin one?) could look like it was thinking of buckling. Was it next to the face gap in the housing by chance?"
Yes, it was. I've seen this on many failed drives where the shim gets buckled at the oil drain cutout at the bottom of the bearing seat in the housing. BMW usually put the thin shim in first (where there were 2 shims) and the thin shim gets torn or folded up at the oil port. In this case it seemed that the buckled thin shim then dug into the thicker shim creating the gouge I pictured in the original post of this thread.

On the issue of the outer raceway creep, the condition normally comes down to housing fit and bearing loads. In our final drives the steel outer raceway is fitted into the cast aluminium alloy gear casing. Both these items have vastly different thermal expansion properties so it is important that the initial fit specification is correct for purpose. (BMW would never release that technical data) This (purpose) needs to take into account a number of things such as housing design, cooling, operational speeds, surrounding structures effect on bearing internal clearanace and lubrication. So as you can see, this is a complex issue, but in this case I would think housing fit, whether initial or after a rebuild (when some tolerance may be lost) would be a major contributor. Are there scores in the housing ? If so, the creep has been excessive and a rebuild may not be 100% reliable unless some mitigating action can be taken during assembly. If the bearing does not need to move axially to compensate for expansion then it may be acceptable to use a Loctite retaining compound to address the issue. Hope this helps rather than confuses. Dennis.

 

[Pedro]

You don't always get a warning,I changed the oil,it was as new,small amount of fuzz on the magnet,no play anywhere in preparation a 3 week trip,700 klms later on the first day oil everywhere and the bearing totally gone,I wish I had replaced the bearing ,replacing the bearing is easy for anyone with mechanical knowledge as I know now ,but just as well the gearbox has gone since and would have gone on the trip,but that allowed me to put a Siebenrock clutch plate in,which I did'nt have when oil ended up on the clutch not long ago,love the bike but my patience is wearing very thin.

 

[CharlieVT]

You don't always get a warning.....replacing the bearing is easy....

Agreed, warning signs are lacking; much has been inquired and discussed about this. I tried to summarize in this thread:
http://www.bmwlt.com/forums/showthread.php?t=63109

And, while beating the proverbial dead horse, replacing the bearing IS easy. Replacing the bearing with attention to proper bearing preload setup is also relatively easy but more involved than just swapping out bearings. I reiterate this because many folks coming to this board read the most current posts and miss lots of information that is deeper in the archives. It would be unfortunate to leave someone with the impression that just swapping out the bearing is a good, long term fix.

That said, carrying a spare crownwheel bearing and hub seal on trips is a good alternative to a preemptive rebuild as road insurance. With the help of any local machine shop, a new crownwheel bearing can be installed without removing the final drive from the bike. You won't have a proper preload setup, but you'll be back on the road.

 

[CharlieVT]

.....The managenment at the Bavarian establishment need to seek out that rascall overshimmer and deaL with him, or could that possibly even be a her, who knows......

Dennis,
here's one of my theories about how these drives get over shimmed:

The BMW Service Manual method of measuring for the crowngear assembly preload uses a BMW special tool:

I fabricated a bearing stabilizer using a cut up final drive cover to substitute for the BMW special tool:

The bearing designation C3 identifies it as a "sloppy" fit between inner an outer races. If you push down on one side of the outer race, the other side 180 degrees opposite will come up. If you measure on only one side of the bearing, and don't stabilize the outer race to prevent tilting of the race you'll get an inaccurate (bigger) measurement.

Using my bearing stabilizer, I found it necessary to measure at opposite sides of the outer race and make sure I got the same measurement at both points:
Measuring one side:

And the other side:

Very often, when setting up my homebrew special tool, I have to make adjustments to get the same reading on opposite sides of the outer race. It is very easy for the bearing or whole crowngear assembly to have tilted a little.

I've never seen or used the BMW tool, but from looking at the diagram of it pictured above, I wonder if it doesn't stabilize the bearing and allows a tilting during measurement that leads to a larger measurement and over shimming. And I wonder if this is the method they were using on the factory assembly line. BTW, in my experiments, the amount of tilt that occurs does increase the measurement in the range of over shimming that I have seen on many drives.

 

[11862]

C VT - I somehow assumed you measured both sides at the same side & averaged them to get bearing clearance info.

Others - As a more sensitive test, I suggest that the drain plug be simply rinsed in thin solvent BEFORE wiping the metal accumulations off the magnet tip. Using this, here is what I found in my K1200 before Charlie VT's rebuild - although there was essentially no bearing distress except the usual over-preloading.