Update on my final drive rebuild

[OoPEZoO]

Well.....I got my final drive back from the dealership last night. Final bill came to $425.87 for a complete rebuild.

 

$185.76 (parts) + $216 (labor $60 X 3.6 Hrs) + 6% sales tax

 

They said that my previous bearing was WAY out of alignment due to improper shimming a the last rebuild (10k miles ago). I'll take full responsibility for that one. I just slapped new bearings in without measuring last time. I removed the final drive and took it to them. They replaced and reshimmed both bearings, put in a new o-ring, and a new seal. They didn't charge me for the shims. I'm not thrilled about the cost, but I think it is fair. I'm heading down to the garage now to start the reinstallation.

 

I'll have a full day in the garage. I'm also

 

Installing rubber chicken racing paralever bushings

a Stebel air horn

a set of grip puppies

new front brake pads

and bleeding the brakes

 

Should be fun, wish me luck

[Joe Frickin' Friday]

I'm not thrilled about the cost, but I think it is fair. I'm heading down to the garage now to start the reinstallation.

 

$425 hurts alright, but as you note, they actually did the work and installed the parts; such is life. FWIW, this is still cheaper than a new FD, which I believe is around $800 now; so all in all, I'd say you did alright.

[OoPEZoO]

All done except for connecting the horn relay wires to the battery (I bought the wrong size connectors), but all the wires are run and the horn works great.

 

A quick stop at the auto parts store tomorrow and a few more minutes to slap the tupperware back on and I will be back on the road. It was a good day to work on the bike today, it was certainly too damn hot out to ride it.

 

I was really surprised how easy it was to get the driveshaft back together. I read so many horror stories about how many hours it took to get the shaft splines lined up properly. I had to do it twice (I forgot the rubber boot the first time), and found it to be pretty easy. Oh well.....glad that is over with. Hopefully I can be failure free for a little while now.

[smiller]

They said that my previous bearing was WAY out of alignment due to improper shimming a the last rebuild (10k miles ago). I'll take full responsibility for that one. I just slapped new bearings in without measuring last time.

Hi Keith,

 

Just wondering, do you feel confident that you installed the bearing properly when you did the work 10k miles ago (other than shimming of course, for example did you seat the bearing all the way, etc.?) If so then your experience would make for a pretty strong argument that one shouldn't just slap a new crown bearing in there and go... but you also replaced the tapered bearing, right? Maybe that is what moved things so far out? Do you think it might have been OK if you replaced the crown bearing only? I have more than a little interest in your story as given the miles on my bike I expect to have to go into the final drive sooner or later...

 

If they did a proper job then I don't think the $425 is a bad deal... as Mitch said, 1/2 the cost of a new unit and probably set up better...

[Dr Klawn]

They said that my previous bearing was WAY out of alignment due to improper shimming a the last rebuild (10k miles ago). I'll take full responsibility for that one. I just slapped new bearings in without measuring last time.

Hi Keith,

 

Just wondering, do you feel confident that you installed the bearing properly when you did the work 10k miles ago (other than shimming of course, for example did you seat the bearing all the way, etc.?) If so then your experience would make for a pretty strong argument that one shouldn't just slap a new crown bearing in there and go... but you also replaced the tapered bearing, right? Maybe that is what moved things so far out? Do you think it might have been OK if you replaced the crown bearing only? I have more than a little interest in your story as given the miles on my bike I expect to have to go into the final drive sooner or later...

 

 

 

 

 

If they did a proper job then I don't think the $425 is a bad deal... as Mitch said, 1/2 the cost of a new unit and probably set up better...

 

 

 

I'd be interested too. Every other thread I've seen on this seems to point to just replacing the bearings using the existing shims....

[KMG_365]

I replaced just the crown bearing after measuring both the old and new and surmising that they were well within the tolerances for re-using the same shims. I've got 40K miles on the new bearing now (126+K miles on the bike total after the original 17-ball crown bearing failed at 86K miles on the way to UnRally-IV in Staunton, VA! ). I did NOT mess with the tapered rollered bearing so I think that would account for the luck I've had so far with not having to change the shims.

[Stan Walker]

Every other thread I've seen on this seems to point to just replacing the bearings using the existing shims....

 

There have been threads by people who understand ball bearing tolerances that have indicated the need to check this after replacement of the crown gear/wheel bearing. There have also been other final drive rebuilds that didn't last or didn't feel correct where this measurement wasn't done.

 

My own opinion is that you are taking a needless chance by not doing this measurement. Maybe you get lucky, maybe not. It's your bike and your choice of course.

 

Stan

[OoPEZoO]

Hi Keith,

 

Just wondering, do you feel confident that you installed the bearing properly when you did the work 10k miles ago (other than shimming of course, for example did you seat the bearing all the way, etc.?) If so then your experience would make for a pretty strong argument that one shouldn't just slap a new crown bearing in there and go... but you also replaced the tapered bearing, right? Maybe that is what moved things so far out? Do you think it might have been OK if you replaced the crown bearing only? I have more than a little interest in your story as given the miles on my bike I expect to have to go into the final drive sooner or later...

 

If they did a proper job then I don't think the $425 is a bad deal... as Mitch said, 1/2 the cost of a new unit and probably set up better...

 

I'm confident that I seated the crown bearing properly. The bearing that I used had the same dimensions (measured with calipers) as the old one as well. The tech at the dealership told me that if I wouldn't have messed with the conical bearing, I most likely would have been fine. I wasn't planning on replacing the conical bearing when I did it the first time, but I accidently bent the carrier that holds the needle bearings so I had to replace it. It is my opinion (and supported by the tech at the dealership) that replacing the conical bearing without measuring or reshimming is what caused the second crown bearing failure. I think everything would have been fine if I hadn't damaged and replaced the conical bearing.

[notacop]

A buddy paid about 2X that plus the uhaul. Consider it good 'specially since you 'fixed it' the first time.

[smiller]

A buddy paid about 2X that

Oh I'm sure that it's easy to pay more than Keith paid. He did the final drive R/R himself, plus was pretty educated about the work he needed done. OTOH, just drop the bike off at a dealer and say 'fix it'... and you're gonna pay...

[smootz]

I just finished buttoning my RT final back together. This was my first BMW final but I do own a machine shop and have years of experience with transmissions, differentials and pumps. The repair was pretty straight forward but I was a little upset that the first 3 dealers I approached for a pre-load gaging tool did not own one. They all told me to put the original shims back in. I was not comfortable with that so I made a measuring tool. The setup would have been .006" too tight! After measuring it several different ways, I called a 4th dealer who told me he finds tight ones all the time. They rarely make it past 20,000 miles. I was extremely lucky to get 87,000 out of mine.

 

I am a little jaded with BMW at the moment. I am still wondering what goofous engineer decided to put lateral pre-load on a BALL bearing. Two taper bearings would have run forever.

 

SCOTT

[cgdR]

Can you post a picture of your measuring tool?

I have been wondering if the preload gap could more easily be measured with a thickness gauge from the outside of the housing, as shown in the picture. To test the concept, I first measured the initial or "closed" gap. Then I untorqued the bolts and, with the bolts finger-tight, I measured the "unloaded" gap. The difference between the two measurements would be the required preload. I am assuming that in the "untorqued-but-finger-tight" state, the housing cover is in light contact with the bearing. In the "closed" state, the cover is obviously pressing into the bearing. Unless there are deformations that I am not accounting for, the difference between these two gaps should be the amount of preload. Mine came out to be in the middle of the spec range. What you think?

[nrp]

Thickness gages are a hard way to measure preload to the necessary accuracy. It could be done supposedly, but I'd rather use a dial indicator or a micrometer.

 

I have not seen the cross section of the FD, but might it be practical to remove the axle seals & measure the axle drag torque of the assembled FD? Then one more disassembly to reinstall the seal? To me that would be the most accurate way. I suspect the drag torque would be be only a few inch pounds. A standard lube would have to be used like WD-40.

 

It is interesting that some are seeing preloads of several thousanths. Ouch! No wonder these are failing.

[Slartidbartfast]

I am a little jaded with BMW at the moment. I am still wondering what goofous engineer decided to put lateral pre-load on a BALL bearing. Two taper bearings would have run forever.

 

SCOTT

I'm guessing that, as with many other things, the FD design is a compromise between engineering requirements and cost. Firstly, a tapered bearing in that application would have required the FD to be either wider or 'taller' and consequently heavier amongst other disadvantages. Secondly, I doubt a taper bearing in the correct size is readily available and would probably be considerably more expensive than the bearing that is used.

 

As has been noted, with correct preload, this bearing set-up while less than ideal, should work well. The question of how the factory properly measures/sets preload at the factory is one that I would bet BMW engineers have spent a lot of time fretting over. Presumably they thought they had the issue fixed with the 1200's FD.

[russell_bynum]

I am still wondering what goofous engineer decided to put lateral pre-load on a BALL bearing. Two taper bearings would have run forever.

 

You're gonna love the bearing setup in the transmission.

 

[ShovelStrokeEd]

Do note that the bearing in question is an angular contact bearing, not a conventional ball. It is designed to accept some side thrust.

[BMW_Ken]

Tom Cutter just finished the Final Drive on my 2004 R1150RT(25,000 miles). In the phone call yesterday he mentioned that it had had been shimmed to tightly by the factory.

 

With the proper preload, hopefully it will last longer than 25k miles this time.

[T__]

I just finished buttoning my RT final back together. This was my first BMW final but I do own a machine shop and have years of experience with transmissions, differentials and pumps. The repair was pretty straight forward but I was a little upset that the first 3 dealers I approached for a pre-load gaging tool did not own one. They all told me to put the original shims back in. I was not comfortable with that so I made a measuring tool. The setup would have been .006" too tight! After measuring it several different ways, I called a 4th dealer who told me he finds tight ones all the time. They rarely make it past 20,000 miles. I was extremely lucky to get 87,000 out of mine.

 

I am a little jaded with BMW at the moment. I am still wondering what goofous engineer decided to put lateral pre-load on a BALL bearing. Two taper bearings would have run forever.

 

SCOTT

 

Scott, I guess it depends on how you look at it.. Obviously .006” too tight is out of specs.. But personally I would rather have that bearing .006” too tight than .003” too loose.. A loose thrust type ball bearing is destined to failure.. Too tight probably won’t have much effect if it isn’t overheated until broken in for a while..

 

You also have to look at the most of the bearing failure modes reported (My bearing failure wasn’t from a too tight bearing but was from the ball separator failing).. Most BMW failures don’t show a ball or race wear issue but usually show a ball separator issue..

 

Back when I was racing trucks I always set the bearing preload to the very tight side as that was far better than the loose side of specs as far as failures went..

 

I usually check the bearing preload on the BMW’s using my surface plate & digital height gauge using a couple of gauge blocks & a mico screw jack to level up the housing cover then a gauge block between the housing surface & the bearing bore bottom,, same with the main housing & gear carrier large bearing..

 

A quick & dirty I have found is to press the bearing home on the carrier then press the side cover on using the original shim.. Then setting the gear & cover back in place without the “O” ring in place.. That should leave the side cover proud of the main housing to the amount of the bearing preload after you even up the cover clearance out all the way around.. Then it is a simple check of the cover to housing clearance with a set of feelers.. Probably not perfect but darn close when checked against my surface plate & height gauge..

 

Twisty

[smootz]

My shop made tool was a replica of the BMW tool shown in the manual. I checked the distances to the bearing seats with a depth micrometer. Finding it tight led me to use an old foolproof method. I left out the seal and shim pack and loosely assembled the case. Just before contact I slipped 3 equally spaced pieces of electrical solder under the bearing and finished snugging the bolts. I dissassembled and checked the squeeched solder with a mic. It verified my previous suspicions. Without any shims the solder measured .023. That would allow .027 as MAX amount of shim. My original shim pack would have been .033.

 

BTW - The new bearing and the old one mic'ed dead nuts with each other and at the advice of every mechanic I did not change the taper bearing.

 

Also of note - my "bad" bearing had 3 balls that were scarred, grooved, striped. I am not sure would cause that kind of damage except maybe the plating failed? Fluid was changed every 6K religiously.

 

Thanks for the feedback. I feel a little better knowing that the bearing is designed for some pre-load.

 

SCOTTIE

[cgdR]

Without any shims the solder measured .023. That would allow .027 as MAX amount of shim.

That's assuming that the solder is so soft that it does not preload the bearing. Is that a valid assumption?

My original shim pack would have been .033.

Are you saying that the factory shim was 0.033?

my "bad" bearing had 3 balls that were scarred, grooved, striped. I am not sure would cause that kind of damage except maybe the plating failed?

Like this?

Original bearing on '02 R1150R w/ 40k miles. Bearing had been rough for 1,000 miles when removed.

[smiller]

That appears to be spalling as a result of normal wear (as a bearing rotates the contact surfaces flex microscopically, and after some number of rotations the metal will begin to fatigue and flake.) Excessive load can cause a similar effect although I believe the damage from such a condition is usually more extensive. Your picture looks just like a wheel bearing that I took out of service after about 75k miles and was beginning to get rough because of normal aging/wear. This document describes bearing failure patterns and your pictues seem to match 'normal wear' the most closely.

 

But then again I would think that a final drive bearing should have a much longer life than what you experienced so perhaps excessive load or improper preload is indeed a factor in your case. Excessive preload would probably lead to excessive deformation so perhaps the results of excessive preload would look like a normal end-of-life wear pattern, but accelerated in time.

 

[T__]

My shop made tool was a replica of the BMW tool shown in the manual. I checked the distances to the bearing seats with a depth micrometer. Finding it tight led me to use an old foolproof method. I left out the seal and shim pack and loosely assembled the case. Just before contact I slipped 3 equally spaced pieces of electrical solder under the bearing and finished snugging the bolts. I dissassembled and checked the squeeched solder with a mic. It verified my previous suspicions. Without any shims the solder measured .023. That would allow .027 as MAX amount of shim. My original shim pack would have been .033.

 

BTW - The new bearing and the old one mic'ed dead nuts with each other and at the advice of every mechanic I did not change the taper bearing.

 

Also of note - my "bad" bearing had 3 balls that were scarred, grooved, striped. I am not sure would cause that kind of damage except maybe the plating failed? Fluid was changed every 6K religiously.

 

Thanks for the feedback. I feel a little better knowing that the bearing is designed for some pre-load.

 

SCOTTIE

 

Scottie,, any signs that the ball separator was failing & allowing some metal particles to enter the bearing race or ball area?

 

If not, then there is always the possibility of a bearing defect,, OR pitting due to moisture or contamination during storage (does the bike get power washed?),, or possible damage from tailoring the bike over bumpy roads & brinelling the bearing due to vibration without rotation..

 

Bearing failure analysis is difficult without seeing repeated failures due to the same causes..

 

Twisty

[smootz]

Indeed, factory shim pack was .033

 

Solder method is very accurate. The solder used was about .050 diameter and very soft. The method is widely used in high speed pump repair. It is similar to using "plastigage" on crankshaft bearings. It gives you a very good real measurement of the gap before the shim pack is calculated. Also the result was spot on with my previous readings with the shop-made depth gauge.

 

When I get to the shop after while I will try to take some photos. Unfortunately I didn't take any while the drive was disassembled.

 

SCOTTIE

[nrp]

That (above) is a classic ball bearing spalling failure of both the ball and the race. It is caused by too much load, too many times/revolutions.

 

Cage failures could be expected if the balls don't release from a race each revolution where the ball contact points make for a slightly different effective rolling radius around the bearing. The accumulated difference eventually can wear out the cage since it is the cage that has to force the balls to all roll with the same spacing.

 

And none of the parts of a ball bearing are plated. It only makes sense to replace any bearing assembly if there is any spalling on any parts.

 

Technically these are deep groove ball bearings, not angular contact bearings. Angular contacts take load in only one direction. In this application, the balls are always rolling on one or the other side faces of those deep race grooves, and it is the combination of thrust and radial loads and preloads that keep changing the effective contact rolling radii. That is why the cage seems to be involved in many of these failures.

[smiller]

Would excessive preload (over a period of time) simulate the wear effects of excessive load?

[T__]

Would excessive preload (over a period of time) simulate the wear effects of excessive load?

 

Seth, I would think it would depend on the amount of over preload.. W-a-y too tight to begin with would produce excess heat at high speeds & probably burn it up quick like.. Final drive would probably smell hot (hot oil smell) after a long hi-way run..

 

Moderately too tight would still make a little more heat & chance damaging the bearing short term but if it survived many miles it would eventually loosen up to normal operating preload..

 

Worse case for a purposely speced out preloaded bearing is too loose as that allows the bearing to pound under load/unload over about every sharp bump..

 

The side cover on the BMW final drive is none too robust so probably can allow a little more preload on the bearing.. Problem is that deep groove ball bearing is probably not the best choice for a side loaded application.. A tapered roller bearing would more than likely be a better choice but those things are large & very wide so real-estate would be a real issue..

 

Possibly way too tight preload with thick gear oil would hinder the gear oil film from staying in place as the ball rolls along..

 

Bearing failure modes are very difficult to figure out as the bearing usually isn’t pulled & viewed until a failure of some sort is noted..

 

The motor company I work for has been fighting front wheel bearing failures for a few years now & there must be 50 fixes or bearing design changes by now & we still can’t fail them in the lab but the customer sure can (some repeatedly).. Even something as simple as static electricity from a rotating tire on the road can cause a bearing internal electrical arc & take a bearing out over time & mileage..

 

You find me someone that can look at a bearing & tell EXACTLY what killed it & how to prevent it from happening again under a multitude of usages,, conditions,, & drivers,, I have a $180,000 a year job for them..

 

Twisty

[nrp]

Of course the force between each rolling ball and its race is increased as the preload is increased. If the bearing assembly is called upon to react an externally applied force, that force has to be distributed between the balls too. On a micro basis of course only some of the balls can carry that added load. Complicating in this application, is the balls are already running on the side of the race's deep groove due to the axial preload.

 

Herzian (compressive) stresses between the ball and the race can be very high (like ~200,000 psi or more) for bearing steels, but even so,at some point the material fatigues and a small piece pops out. This tends to give a cascading failure as the presence of this very hard material in the bearing assembly soon trashes it.

 

My contention is that since many of the failures are ball and race related, the stresses are not determined by the lubrication, but by the applied external force plus preload. Our field force profiles are probably not that different, so probably it is the preload that is highly variable.

 

On the other hand cage failure is probably due to the ball spacing wanting to change with respect to each other, as they all roll around the bearing assembly. All it would take for this to happen is if the contact points start moving around under the combined loading.

 

Another point - yes a tapered roller bearing might be appropriate, but they are more speed limited in larger diameters.

 

And - The reason they want a preloaded assembly is to provide a precise axis of rotation to each of the gears to reduce gear noise and wear.

 

At this point I see preload to be the only variable that is practically available to us for field fixes. Some FDs go a long way without failure. It would be interesting to see what is the actual preload in one of those long-lasting units.

 

We also might consider alternate ways to measure the preload which might be more repeatable, such as remove all seals, and then see what the breakout torque is with say WD-40 lube, then reassemble with seals.

[smiller]

That sounds logical.

 

My personal view is that the oilhead final drive failure issues have more to do with sloppy original assembly fitting than anything else, and all of the frets over specific oil grades or 17/19 ball bearing assemblies, etc. are off the mark. Hans feeling good that day, your final drive lasts a long time. Argument with his wife, you're out of luck... whether you use 80W-90, 85W-140, whale oil, or whatever.

 

I do like the idea of using rotational torque to set preload on these things... hopefully we can come up with a baseline.

[smootz]

Forgive me but I am just becoming kind of jaded. Maybe I feel this way because the problems came to "my house".

 

I love the handling characteristics of the paralever and the telelever but I am not sure that the single sided swingarm is really worth it all. Sure it is nice to take the rear wheel off in less than a minute but the maintenance/repair issues are mounting. I saw an article in BMWMOA that says BMW recommends replacing airhead driveshafts every 40,000 miles. At $650- a pop that is unacceptable. I can replace chains and sprockets much faster and cheaper, even if more often. I see the belt drives on the H-D's consistently going over 100K. Maybe wheels belong on a shaft with a little taper bearing on each side. It sure was a simpler world.

 

Forgive all this blasphemy, I really do love my Beemers but I am a little frustrated at the moment. I have done the best rebuild I could on a final drive and there is still no guarantee that it will survive my first long trip. On top of that it seems that the BMW trained mechanics don't know much more than us.

 

SCOTTIE

[smiller]

I have done the best rebuild I could on a final drive and there is still no guarantee that it will survive my first long trip.

I understand your frustration but if you did a careful job on the rebuild I think you can be fairly confident in the result. Most final drive units last a long time and yours should now as well.

 

I was a little upset that the first 3 dealers I approached for a pre-load gaging tool did not own one.

There's more likely the real problem...

[nrp]

My personal view is that the oilhead final drive failure issues have more to do with sloppy original assembly fitting than anything else, and all of the frets over specific oil grades or 17/19 ball bearing assemblies, etc. are off the mark. Hans feeling good that day, your final drive lasts a long time. Argument with his wife, you're out of luck... whether you use 80W-90, 85W-140, whale oil, or whatever.

I agree. Anyone replacing bearings in a FD without using a dial indicator etc (or equivalent) to measure clearance is just asking for another failure.

 

But I wonder what the clearance should really be?

[T__]

~clip~ I agree. Anyone replacing bearings in a FD without using a dial indicator etc (or equivalent) to measure clearance is just asking for another failure.

 

But I wonder what the clearance should really be?

 

NRP, no clearance as it’s preloaded to “0” lateral play then additional latteral preload on the bearing ..

 

Twisty

[Paul Mihalka]

"I saw an article in BMWMOA that says BMW recommends replacing airhead driveshafts every 40,000 miles."

 

I'm in the BMW bike business for 25 years and never saw a recommendation like that. I sold my '99 R1100RT with 175K miles with the original driveshaft. Now with about 200K it still has it. The final drive went south at 150K.

[Motodan]

Me thinks 40,000 miles refers to lubing of splines, not replacing the driveshaft. Don't you just love the Internet knowledge base...information for the picking.

[smootz]

Me thinks 40,000 miles refers to lubing of splines, not replacing the driveshaft. Don't you just love the Internet knowledge base...information for the picking.

 

BMWON - February 2008 - page 22 - "Keep'emflying"

Matthew Parkhouse - QUOTE " These driveshafts have achieved a certain amount of notoriety for their being classified as a consumable item by BMWNA." - farther down the page - "The BMWNA recommendation is that the shaft be changed every 40,000 miles or so to prevent sudden failure" - still farher - "The GS's seem to have the most problems with driveshaft failures; although they are far from unknown in other models".

 

The above refers to the earlier driveshafts as used on R100's

 

SCOTTIE

[nrp]

NRP, no clearance as it’s preloaded to “0” lateral play then additional latteral preload on the bearing ..

My bad, I should have said preload, but it should still be measured with a dial indicator & then the known shim removed. What is it spec'd at now, .001 to .002 inch preload? That's quite a range.

[smootz]

Although it is a common practice to use a dial indicator on most differentials, transmissions, and pumps to determine pre-load the BMW final drive is a different animal. Because it combines a taper bearing with a ball bearing and a "push fit" cover it really is impractical to use the dial indicator. BMW makes a setup tool that measures the height of the bearing while partially assembled and then references it to the offset in the cover. You need to look at a manual to understand this.

 

As stated earlier, I machined my own gaging fixture and later confirmed my results with "solder" gages.

 

I have no problem with BMW's method of measurement. I do however think it is inexcusable that some dealers do not own or use the set-up tool.

 

BTW - I rode the bike today. YIPPEE!

 

SCOTTIE