Designing a different mousetrap - Final Drive Bearing Tool for 1150's

[bcostell]

Having read through the BMW service manual, viewed you-tube vids and looked at various articles written on the topic I've had a go at putting together a tool that I can use in the field to accurately measure and correctly set up the final drive main bearing.

To 'correctly' set up the bearing pre-load you need to know the size of the gap between the face of the bearing outer race and the inside face of the final drive cover - all when it's assembled. Once you have that you select a shim that is that size plus a bit for pre-load.

But how to measure?

Well, it looks to me like you have two choices at the moment. Both of which involve the use of either a set of BMW measuring tools or a heat gun, dial gauge and a fridge. The common wisdom is if you experience bearing failure on the road you just pop a replacement bearing on and carry on until you're someplace where you can do the job properly. Common wisdom also has it that the dimensions of the bearing races are held to a pretty tight tolerance so what you put in should be the same as what you take out. There are two problems with this logic however. 1). Was the shimming correct to start with and 2). Are the dimensions really the same?

I don't doubt that the race dimensions are always going to be within 1 tenth of a gnats cock of each other, but I'm not so convinced about the axial play in the bearings.

If you just pop another bearing on you then are stuck with one of the currently accepted methods of setting things up properly at a later date. My thinking is to find a way of figuring out what the correct shimming should be in advance of a trip, or once a failure has happened be able to determine what the shims need to be so that you can 'quickly' add the correct shims later. By quickly I mean with the final drive still assembled to the bike. The approach is to be able to take a bunch of measurements using a feeler gauge and a simple tool. The key measurements you need are:

1). Dimension from the outer mating face of the crown-wheel (where the big bearing inner race mates) to the mounting face for the big bearing inside the cover.

2). The bearing race dimensions including the axial play. The axial play can be measured by laying the bearing on a flat surface and pushing feeler guages under the inner race while holding the outer race against the flat surface.

By making up a bearing plug, with known dimensions, that is a close, but sliding fit you can assemble the drive without a bearing and measure the

play. By then doing the calculation:
Assembly play + thickness of the plug + axial play
And then subtracting the thickness of the bearing race
you then have the gap between the bearing outer race and the final drive cover mating face.

My thinking is if I carry a spare bearing with me I will of course pre-measure the bearing dimensions and axial play, all I need to worry about in the field is the calculated play of the assembly. If I prefer I can pull the drive apart before I set out and do all of the measurements in advance and if there is some difference in bearing axial play I can just take the required shims with me. Confused yet? Here's some pictures which 'might' help.

The tool

Fitting the bearing blank to the crown-wheel

Fitting the cover and measurement plate

Measuring the gap with the crown wheel in the lower position

Measuring the gap with the crown wheel snug up against the outer cover

Subtracting one from the other gives the total play of the assembly - used below.

Measuring the axial play in the new bearing

All of the above takes about 5-10 minutes. And it's done without keeping parts hot or cold.

Here's my calculations:

Play measurement (in thou) 0.032" 0.8128mm
Blank thickness (in thou) 0.695" 17.653mm
Bearing thickness (in thou) 0.7075" 17.9705mm
Bearing axial play (in thou) 0.007" 0.1778mm

Total gap of assembly 0.0265" 0.6731mm

Pre-Load (mm) - Low Side 0.05mm
Pre-Load (mm) - High Side 0.1mm

Shim Min 0.7231mm
Shim Max 0.7731mm

Required shim 0.75mm

As I hinted earlier, all of these measurements with the exception of the bearing axial play, can be taken with the final drive still mounted to the bike, and in the field. As the tool is the approx the same size as the final drive bearing and made of alloy you should be able to find a home for it in your luggage if you want the insurance. Btw, I'm not selling the tool, it's just something I made up for my personal use, thought I'd share and take comments

 

[MattW]

Nice job. I like your reasoning and the way you've come up with a solution to the 'problem'. Like you, I've thought that just reusing the existing shims depends on both the quality of the original shimming operation and lack of variation in bearings (both of which are by no means certain).

The only possible issue that occurs to me is the technique for measuring axial play in the new bearing...

Will the inner race 'rock' when you insert the feeler gauge on one side and give a false measurement? It might be a non issue but would it be better to use two matched feeler blades from opposing sides?

As I say, it might not be significant or indeed an issue at all

Would you mind if I shamelessly pinched your ideas / tool designs?

I must admit that for a repair on the road, I'd probably just use the original shim and trust that it lasts long enough for me to complete my trip but I'd definitely prefer to measure things up properly for a 'workshop' bearing replacement.

 

[bcostell]

Yeah, I thought about the 'rocking' potential also. It makes sense to use two gauges.

Part of my thinking would be to pre-measure the spare bearing before a trip and also the dummy bearing, maybe just write the dimensions in permanent marker on a face. Go ahead and do what you want with the idea if it helps.

For reference I made the blank a couple of thou bigger than the bearing ID and a couple of thou smaller for the OD and something like 10 thou thinner for thickness. Make sure to break all machined corners or add a small chamfer to make sure all faces sit square. The spacers were made out of 3/4 stainless stock and 21mm in length, they didn't need to be turned down - 3/4" seemed just perfect. The bushes fit snug in to the counter-bored holes in the cover plate and allow a small bit of clearance when the crown-wheel is draw up. I used 10mm screws to mount the measuring plate and I think the smaller ones are 8mm that screw in to the crown-wheel. The measuring plate is 1/4" alloy stock - quite thick so it doesn't flex when things are tightened up.

I think that common consensus has it that the taper roller bearing doesn't fail all that often, but it's such a small item I figure keeping that as a spare along with the tool and some shims and about the only thing I'll have need for is a way of removing the old bearings.

 

[Timolgra]

Did you also measure the depth of the counter bored holes in the cover plate for any variation before using your measuring spacers and plate?

 

[MikeP]

I'd have thought that any bearing that has the sort of variation requiring the shims isn't worth fitting.

I believe that the shims are for variations in the housing. As such, once it's shimmed correctly you can just fit new bearings without any need to measure anything.

 

[bcostell]

Didn't feel the need to measure the difference in counter-bore depths. As long as you measure the gap in the same place it shouldn't make any difference.

 

[MattW]

As such, once it's shimmed correctly you can just fit new bearings without any need to measure anything.

But are they shimmed correctly from the factory (and are the shimming tolerances over lenient in order to speed up production)?

I would suggest that there is significant variation in the quality of the gearbox / FD assembly from the factory as evidenced by huge variation in their lifespan's?

I don't think you can necessarily depend upon 'correct' shimming from new...

 

[Timolgra]

I'd have thought that any bearing that has the sort of variation requiring the shims isn't worth fitting.

I believe that the shims are for variations in the housing. As such, once it's shimmed correctly you can just fit new bearings without any need to measure anything.

I think that's what bcostell said and it's what I've always held as true, I think he's suggesting that 1) the shimming may be wrong in the first place and 2) there may be variation in the axial play of the bearings.

I'm not so sure about the latter, but it would be interesting to know if there is.

 

[MikeP]

I think that's what bcostell said and it's what I've always held as true, I think he's suggesting that 1) the shimming may be wrong in the first place and 2) there may be variation in the axial play of the bearings.

I'm not so sure about the latter, but it would be interesting to know if there is.

But his tool is for use in the "field" and I take that to mean it's for if a bearing fails at the roadside. If there's incorrect shimming suspected, I'd have thought the best solution is to sort it out on the bench before setting forth for the "field".

 

[bcostell]

Part of my motivation is the move to angular contact bearings away from the std ball bearings. Different characteristics and measurements. Now, if we just consider the std bearings.....

The shims are to allow for variations in the housing and cover as you mentioned...also for the position of the crown-wheel once it is set up correctly relative to the pinion. There is a shim at the other end of the crown-wheel between it and the taper roller bearing. This affects the shim size required - smaller shim under the taper roller means larger shim in the drive cover. Then there is the small wear that may occur with the taper roller bearing. So numerous areas of dimensional variation. I guess BMW had to anticipate how these could all stack up, then leave a little more space just to be sure.

I'd have thought that any bearing that has the sort of variation requiring the shims isn't worth fitting.

I believe that the shims are for variations in the housing. As such, once it's shimmed correctly you can just fit new bearings without any need to measure anything.

 

[MattW]

But his tool is for use in the "field" and I take that to mean it's for if a bearing fails at the roadside. If there's incorrect shimming suspected, I'd have thought the best solution is to sort it out on the bench before setting forth for the "field".

I don't think anyone 'suspects' incorrect shimming until they're at the side of the road with a puddle of oil next to their back wheel

 

[bcostell]

Absolutely, prevention is better than cure. In coming up with the design I figured it could be used with the drive in situ without a dial gauge, using just tools that we all have around - with the possible exception of a micrometer to accurately measure the bearing blank first time round (or get the machine shop to measure it when the part is made). Removing the final drive is not a massive job, but why have the hassle of removing it if you can do the job by only having to unbolt the cover?

But his tool is for use in the "field" and I take that to mean it's for if a bearing fails at the roadside. If there's incorrect shimming suspected, I'd have thought the best solution is to sort it out on the bench before setting forth for the "field".

 

[Timolgra]

But his tool is for use in the "field" and I take that to mean it's for if a bearing fails at the roadside. If there's incorrect shimming suspected, I'd have thought the best solution is to sort it out on the bench before setting forth for the "field".

I agree with you

Perhaps I've just been lucky, in 14 continuous years of 1150 ownership. I've never had one fail although I have changed a few as 'preventative maintenance'.

I guess that having an engineering background, I'd assumed BMW shimmed them correctly in the first place and that there's no discernible variations in the bearings.

So, to my mind, I'm either correct in that assumption or as I said, just lucky

 

[Timolgra]

I don't think anyone 'suspects' incorrect shimming until they're at the side of the road with a puddle of oil next to their back wheel

Ah, but now your suggesting the bearing fails because the shimming was incorrect ....perhaps for the last 60k miles.....

 

[MikeP]

I don't think anyone 'suspects' incorrect shimming until they're at the side of the road with a puddle of oil next to their back wheel

Given that there cannot be any of these bikes with zero miles on them, if the "incorrect" shimming lasted several thousand miles before it spat its guts, why drag around a tool to measure it at the roadside? Just do what has been done often enough in the past, stick a new bearing in and carry on.

Beaten to it by Tim.

Like Tim, I've been careful to check for play in all the BMW's I've owned that use the same FD set-up. The only one I've ever had to change (on my own bikes) was at 98,000 miles on a K75 and that was changed because there was very slight play not because it had failed.

 

[MattW]

Ah, but now your suggesting the bearing fails because the shimming was incorrect ....perhaps for the last 60k miles.....

No.
I'm suggesting that the 'possibility' of incorrect shimming might cross one's mind as one stares at the oil puddle
There are of course lots of potential causes for the failure...

 

[bcostell]

I guess it all depends on IF you're the sort of person who carries a spare bearing with them on a trip. If you are, then in the rare circumstance that a bearing fails you can at least determine if the shimming is correct or if you need to change it when you get home or someplace where you can source the correct shims. If you don't make that determination before you fit the bearing then you will need to use one of the other methods to determine the correct shimming requirements. IF you use the tool above you can figure out what the shim needs to be just before you fit the new bearing and if you don't have the correct one all you should need to do later is warm up the cover, push out the crown-wheel, add the correct shim and re-assemble - all with the drive in situ.

Ah, but now your suggesting the bearing fails because the shimming was incorrect ....perhaps for the last 60k miles.....

 

[bcostell]

Agree, taking tools and spares is a pain in the arse. Doing the quick and dirty is an easy solution and no doubt gets you home and on your way quickly. I guess there are only a relatively low percentage of riders who might consider it worthwhile: the anal (me), those who might be in the 'at-risk' category (high load, rough surface), those doing a lot of distance away from home base or BMW parts who want to do the bearing swap 'right' and not as a get you home repair. I guess the trade off is about 3 oz of weight and maybe 10mins of measuring time - all about personal choice I suppose

Given that there cannot be any of these bikes with zero miles on them, if the "incorrect" shimming lasted several thousand miles before it spat its guts, why drag around a tool to measure it at the roadside? Just do what has been done often enough in the past, stick a new bearing in and carry on.

Beaten to it by Tim.

Like Tim, I've been careful to check for play in all the BMW's I've owned that use the same FD set-up. The only one I've ever had to change (on my own bikes) was at 98,000 miles on a K75 and that was changed because there was very slight play not because it had failed.

 

[Smug]

That is a knackered ball race with .007" of axial play

 

[MikeP]

No.
I'm suggesting that the 'possibility' of incorrect shimming might cross one's mind as one stares at the oil puddle
There are of course lots of potential causes for the failure...

My pet theory is that play in the Paralever pivots is the prime cause of FD bearing failure in the 11XX.

The same set-up has been used in K-Series Bricks for years without a problem. It was also used in the 16v K-Series with Paralever and the R80/100GS without the failure rates of the later R-bikes.

The big difference between the applications is the location of the spring-strut. On the early K-Series there's no Paralever, on the 16v K and the R80/100GS the spring-strut locates on the FD itself. The spring-strut acts as a brace, the third arm of triangular location.

On the later R-bikes the Paralever pivots and the Paralever bar (which itself is located on metalastic bushes) are loaded with all the lateral forces when cornering. If the pivot pins are worn or badly adjusted then lateral 'snap' forces are sent through a bearing intended for radial not lateral strength. It probably explains why the cage is usually in pieces when you open one that has failed catastrophically.