What's new
What's new

Newbie measuring motorcycle transmission parts...

spacewrench

Plastic
Joined
Apr 2, 2020
Hi, I'm a patent lawyer and motorcyclist, and in no way a machinist or even mechanic. Unfortunately, due to a stripped clutch spline, I'm obliged to dive into both; I need to open a three-shaft, six-speed transmission and replace an input shaft that has a gear, a couple of bearings, and some other gadgets pressed or slid onto it.

The service manual procedure includes a section to measure and shim (if necessary) the transmission shafts to within +0-0.002" over bearing-to-bearing distances of about 8". The procedure is to set up a dial indicator against a reference cylinder, then measure/shim the shafts. (Of course, I don't have the reference cylinder, so I guess I'll have to measure against the two shafts I don't have to fool with.)

I am comfortable working with 0.002-0.012 over short distances (such as intake & exhaust valves) but have never had to measure a large, freestanding object to 1 part in 4000; it seems implausible that you can do this reliably outside of a pretty controlled environment.

I'm hoping that my fear is unfounded, and that with a few simple things you learn the first day at machinists' school, even a hamfisted fool can measure a part like this successfully.

Thus my question: is there a good reference for beginner stuff like this? Is it 2-3 tips that you can put in a reply post? Is this whole measurement thing likely a snipe hunt that the BMW guys are having a laugh over?

Thanks for any help -- I haven't disassembled all the way into the transmission yet, but hope to get there in the next week or so.
 
I'm a bit surprised you didn't get a reply pointing you towards a more home shop oriented page, and I'll briefly chastise you as a patent lawyer for missing the common methods in research.

1) The first answer falls under "If there is already a readily available expert, pay them to do what they are good at so that you can use the time to do what you are good at". That said, if your experience with motorcycle mechanics is similar to what I've seen you don't want them anywhere near your bike.

2) As a reminder, be careful which race surface (and sometimes what assembly order to allow this) you are pushing those bearings on with. Transferring force through the rolling surfaces is a big no-no.

3) The item you're seeking is a calibration rod, or standard. You can get an 8" one such as a Mitutoyo 167-148, new with calibration slip, for under $50 online.

4) Once you have your calibration rod you still need something to do the measuring. If you're good enough with your shims you might just pull this off with a pair of calipers because you are near the calibration point. If not and you have enough space to get one in, a micrometer (preferably with a ratchet thimble) is a good tool for this. Be aware that if you get an inch model (say 7-8") they tend to have only bit more space on the ends. IE, an 8" to 9" mic. might be able to measure a bit over 9", but not so much under 8". A 7" to 8" might get you just far enough past 8", YMMV. These are only a bit more than $200 new from reputable manufacturers, and I'm sure used ones in varying conditions and qualities abound. More importantly, they also come with the above mentioned standard.

Plenty of other ways to skin this cat, but hard to know more without seeing the part.

Add an occasional re-greasing of your clutch spline to your longer term maintenance list.
(Edit, with a high moly paste, not some generic grease. I'm also aware that putting/not putting moly paste on this spline might be a matter of religious preference).
 
Thanks very much!

I'm a bit surprised you didn't get a reply pointing you towards a more home shop oriented page, and I'll briefly chastise you as a patent lawyer for missing the common methods in research.

I've definitely hit the motorcycle boards and have found a bunch of pictures and descriptions, but few people have done the whole transmission teardown -- it's more common to buy a used transmission from a wreck and swap it out. There's several theories on the cause of these frequent (at the time) failures, but it's certain that the OEM input shaft splines were shorter than there was room for, that the splines did not extend past the load point for the clutch disc to which the shaft connects, and that both the internal and external splines tended to wear -- failure just happens to be on the clutch hub side. So a used-transmission replacement doesn't solve the whole problem. I have a longer-spline input shaft from a lot manufactured by an enthusiast, but I also have to perform a more-invasive transmission surgery to install it.
OEM Clutch-Shaft.jpgOEM-Clutch Load.jpg

1) The first answer falls under "If there is already a readily available expert, pay them to do what they are good at so that you can use the time to do what you are good at". That said, if your experience with motorcycle mechanics is similar to what I've seen you don't want them anywhere near your bike.

There are a few well-respected transmission surgeons who will do this repair, but the cost exceeds the value of the bike by a substantial margin. I'm only doing this because I like the bike, and because I like to learn how to do new things. The rational economic decision would be to part out the bike and buy a new one.

2) As a reminder, be careful which race surface (and sometimes what assembly order to allow this) you are pushing those bearings on with. Transferring force through the rolling surfaces is a big no-no.

Understood, thanks.

3) The item you're seeking is a calibration rod, or standard. You can get an 8" one such as a Mitutoyo 167-148, new with calibration slip, for under $50 online.

Neat! The service manual calls out a BMW tool number, but it could very well be a standard-length reference. Their procedure also uses a calibrated shim that you're supposed to stack under the bearing (before pressing it on); you calculate the thickness of the shim that goes on the shaft from the overall measured length. I don't think I can use the procedure anyway because my input shaft is longer than OEM by the length of the extended splines, so I'll have to measure everything differently anyway.

4) Once you have your calibration rod you still need something to do the measuring.[...]
Plenty of other ways to skin this cat, but hard to know more without seeing the part.

Since you didn't mention my 1-part-in-4000 concern, I guess that's understood to be trivial. I'll get things as clean, square and accurate as I can, measure a few ways, and keep my fingers crossed. Measure twice, assemble once.

Add an occasional re-greasing of your clutch spline to your longer term maintenance list.

Now you're getting into biker religious-war territory! You can find plenty of opinions either way, but it's certain that just getting to the splines involves major disassembly. I do have a tube of the official moly paste that BMW calls for, so I'll be lubing them as specified, but I hope I never have to venture into this portion of the bike again.

Thanks again for the reply!
 
...but it's certain that the OEM input shaft splines were shorter than there was room for, that the splines did not extend past the load point for the clutch disc to which the shaft connects, and that both the internal and external splines tended to wear -- failure just happens to be on the clutch hub side. So a used-transmission replacement doesn't solve the whole problem. I have a longer-spline input shaft from a lot manufactured by an enthusiast, but I also have to perform a more-invasive transmission surgery to install it.
View attachment 284318View attachment 284319


There are a few well-respected transmission surgeons who will do this repair, but the cost exceeds the value of the bike by a substantial margin. I'm only doing this because I like the bike, and because I like to learn how to do new things. The rational economic decision would be to part out the bike and buy a new one.


Neat! The service manual calls out a BMW tool number, but it could very well be a standard-length reference. Their procedure also uses a calibrated shim

I have zero expertise in motorcycle engineering, so no subject-specific ideas to offer, but the short-shaft problem you describe may simply be a manufacturing tactic to allow a reasonable assembly method in the factory. It surprises me that BMW would do something like that without really pressure-testing the design concept, but our global pool of experienced engineers is getting much shallower (and sometimes stupider) on a daily basis.

I assume you have a procedure mapped out for installing the extra-length shaft in the existing assembly, acquired from your aftermarket source. Do you know what the BMW alignment tool set actually looks like? It is possible that with a minimal investment in measuring tools from eBay (maybe a height gage and a V-block) you may be able to get the capability you want. As jccaclimber mentioned, pictures might illuminate the problem and create more opportunity for suggestions.
 
I'm not a motorcycle engineer either but perhaps my .02 will be of some help. I've dug well into the inner workings of several BMW automobiles so I'm a bit familiar with some of the engineering choices from Bavaria. I haven't worked on one of their bikes but worked on many other bikes from different makers. For any repair/modification/maintenance procedure I always scan the manual (or on-line) for the "Special Tool" involved. Most often it eases the procedure but sometimes it's an already existing tool that can be purchased from other sources (for less cost). Sometimes it's not. BMW (autos) has had a habit over the last few decades of employing "special tools" with little or no thought to the owners. Sometimes it almost seems as though any thought to the owners involved consideration for the service department at the dealership. The engineering capability of BMW does not always seem to arrive at the showroom floor with each passing year.

To your questions:
Is the "calibrated shim" a standard that could be purchased? Quite likely yes or it could be duplicated with a standard set of feeler gauges. A shim is usually relatively thin and most feeler gauges will cover the thickness with one or more leaves.

The "reference cylinder" is another matter. Will it be nominal standard size? Maybe/maybe not. BMW service tools are often NOT a standard nominal size. Is it absolutely required for the procedure? Maybe/maybe not. I've made several service tools in various levels of simplicity/complexity and on occasion purchased a service tool to use in conjunction with things I had or made. It's very hard to give an answer without seeing the application.

What is important here is to have ANY and ALL numbers from factory specs. Even if your modified shaft negates using the standard procedure it may still be critical to use the dimensions as a base. Without seeing the application and the procedure it's really difficult to give a better answer. Is this beyond your capability? That will depend upon what tooling/instruments you have and how much time/money you're willing to spend in pursuing this. I can tell by your questions that it will take quite a bit of your time since this is all an unknown procedure to you. I contend that this could still be within your abilities if learning is the primary goal.

Is this axial clearance (float/play) of a shaft? Centerline distance or location between components? Does the maker of this revised shaft offer any guidance/advice for setting the new components properly? How much does it cost to have one of the "surgeons" do the job? That could determine whether the cost outlay for tooling is worth doing it yourself.

I don't mean to be vague but more details, photos, drawings, dimensions are required for a better answer.
 
No further information/input from the OP in just over 2 weeks, guess this means the he/she either found an answer or gave up the project, or it's on the back burners. A more detailed answer with a possible solution might have been possible IF more information and/or details had been supplied. I can grasp that not everybody is capable of describing technical procedures in the appropriate nomenclature but even a scanned napkin sketch or crude rendering in "Paint" or Corel could have been something more to work with than a copy/paste diagram with no further explanation of what was needed.

To the OP: Thanks for stopping by, hope you found something that helped.
 
" I need to open a three-shaft, six-speed transmission and replace an input shaft that has a gear, a couple of bearings, and some other gadgets pressed or slid onto it."

1) you don't say what make this motorbike is.

2) Youtube is your friend. Somebody's done this before if it's a known bug.

3) Mostly transmission shafts are contstrained in an aluminum housing. The bearings are a press or heat fit into the housing, and
the shafts are set up to have minimal end float between the hard stops in the bores in the housing. This is done by measuring the
effective space between the bearing bores in the housing and comparing that with the actual shaft length.

In practice this means a bit of math and some work with a depth micrometer. Then select the shim(s) that give slight clearance.

Because you are replacing one shaft and both bearings, one thing you could consider is that the setup was correct from the factory.
If the bearing OD race distances are the same in the old and the new assemblies, you might be done. Bearings are manufactured
in thickness tolerances that are sub 0.001 inch and the motorbike manufacturer probably has the shafts manufactured to similar
tight tolerances.

(oh yeah - put the shims back *exactly* where they came from..)
 
Here's a status report & wrap-up for everybody who was curious and everyone who offered helpful advice. I'll try to answer the questions as well.

The bike is a 2000 BMW R1100S. It has about 77k miles. Many bikes from this time frame that used the same engine and transmission had the same problem, but it did not affect every bike. (On the other hand, many bikes are never ridden 30k miles over their entire life, so it's possible that many more would have failed if they'd been ridden longer.) The failure usually appeared around 30k miles, so it's possible that this was the second failure on my bike.

Here are a couple of pictures of the approach to the transmission. Getting all of the back end off is a nasty, time-consuming job that probably accounts for a lot of the cost of having a shop do this repair.
1.jpg2.jpg

Here are pictures of the clutch (at the back of the engine) showing the stripped splines; and the front end of the transmission showing the damaged splines on the input shaft
3.jpg4.jpg5.jpg

(The system only allows five images per post, so I'll split this response in two.)
 
(Continued...)

This is the cleaned-up transmission, and the view with the front cover removed.
6.jpg7.jpg

Finally, this is the length I wanted to measure.
8.jpg

Unfortunately, the service manual's instructions for extracting the transmission shafts (heat & pull, basically) did not work at the levels of heating and pulling I was comfortable subjecting the transmission to. Therefore, I took the advice of an early respondent: have a professional deal with it. (As mentioned above, getting this far in and being able to deliver the cleaned transmission to the shop significantly reduced the cost. It would have been nice to be able to do everything myself, but this way gets me back on the road quicker.

For completeness, here is the service manual page for assembling the input shaft. The actual lengths are not provided; you have to use BMW's special tools and shims. So even if I had been able to get the input shaft out, probably the best I could have done was to match the length of the original shaft (which, I suppose, might have been altered by the extraction process).

Manual.jpg

Thanks again to everybody who commented!
 
That's the resulting carnage I half expected from the diagram you included in post #3. This isn't the first time or the first design for BMW where something didn't go well in the field, hard to second guess otherwise good engineers when the entire approval process isn't known.

The question(s) is(are) did you enact a repair to your bike? OEM shaft (that follows the same life/destruction cycle) or was an improved solution installed?

Surprised that this happens on this model, even occasionally. I did a test ride on a similar model in 2000 while at the dealer for parts on a 7 series (auto) and was disappointed with the low speed throttle response. I thought you were describing the problem for an R32 or a K series.

If you still want help/advice with measuring someone here can likely advise. If not, thanks for posting a follow-up. Many people start a post asking for advice and we never hear from them again.
 
" instructions for extracting the transmission shafts (heat & pull, basically) did not work at the levels of heating and pulling I was comfortable "

Spit has to sizzle for this. The trick is to rinse off as much of the 90 wt as you can, then put the thing in a reynolds brown-in-bag, and pop
it in the overn. 200, 250 F until things loosen up. The aluminum expands more than the steel. One trick is the shafts are intlocked and
they all have to come out, also don't forget the shift forks.

It's possible this bike may have benefitted from a spline lube before things went pear-shaped. The good news is you can buy brand-new parts
for this gearbox. Gearboxes like this one come apart basically the same way, but it's a labor of love finding NOS stuff to put in them:

Bmw.jpg


I have two bikes that use the same 4 speed gearbox. Note that the input shaft splines are about three time coarser than the ones on your
bike.
 
Your second half of the follow up wasn't there when I posted. I can't make out the page from the manual you included but it looks like what's important is the distance for the farthest bearing, which would be expected. The overall length of the shaft is probably not important, provided there's nothing that would cause interference. Setting the bearing distance would be important and probably could have been a matter of duplicating factory settings by taking a measurement off the toasted transmission after cleaning it up. This might have been possible with standard measuring instruments by using a comparative (as opposed to actual) dimension, duplicating factory settings. Improved components from owners of the bikes and autos are disappointingly commonplace on the owner boards. Even the factory trained techs have harsh criticism for some of the design deficiencies they see come in with common problems.

The use of a reference cylinder with no published dimensions for that cylinder is typical of repair specs/procedures from the folks in Bavaria. Were the dimensions known a cylinder of any length could be fabricated/used (with corresponding math) and sales for "Special Tool #_ _ _ _ _" would be affected. It's a purpose driven decision just like the software employed in the electonics (EMS) that wasn't available for years. That's part of why there's an improved component available when the factory plays hide-n-seek with repair specs.

Why am I stating all this? In case this happens again (2 out of 3 so far...) you might be either better prepared or have an option other than OEM replacement. Good luck.
 
The question(s) is(are) did you enact a repair to your bike? OEM shaft (that follows the same life/destruction cycle) or was an improved solution installed?

A bike enthusiast with a manufacturing background produced a run of replacement input shafts with proper dimensions, hardening & etc. However, the input splines were extended by about 6mm. In conjunction with an extended-spline clutch hub that's available on the aftermarket, you can get an extra 12mm of spline engagement. If that turns out to be enough to solve this problem, it'll be worth the increased cost of the improved shaft.
 
Glad to hear all is sorted, and particularly glad that the pictures followed. Unfortunately the instruction sheet loaded at very low resolution, perhaps a link to the original would help people in the future.

While I've seen all sorts of awfully hard to fix things on BMW cars, the people I've known with their motorcycles have typically been pretty happy with their serviceability. Might have something to do with those all being 70's/80's bikes, but something incredibly involved in a car like changing a wrist pin was a nice Saturday project on the bikes. For his bike some of the parts (say the alternator) are oddly expensive, but at the same time they provide the wiring diagram and the ability to buy subcomponents like individual diodes for pennies if you are willing to repair the existing item instead of just replace it.

I second the above advice that "heat" is best applied via an oven in many of these cases, although a recommended temperature sure would be nice. Torching locally would quite likely reach very unreasonable temperatures and do bad things, but a nice uniform 250°F in an oven might work wonders. Sometimes also required to reinstall depending on how much press fit is used.
Things like valve guides and seats need to not get loose at engine operating temps, and therefore are occasionally so much of an interference at room temp that they'll damage something on installation if you press them in cold. Put them in the oven at the right temp on the other hand and I've seen guides fall out simply by flipping upside down. Pop the new ones in, cool, ream, Serdi on the new seat, and you're good to go.

All of that said, it's nice to see that this looks to turned in to the mechanic's job being faster/easier due to some good prep. on the part of the customer. Much better than "Here are a pile of parts I messed up, can you fix it?".
 
A bike enthusiast with a manufacturing background produced a run of replacement input shafts with proper dimensions, hardening & etc. However, the input splines were extended by about 6mm. In conjunction with an extended-spline clutch hub that's available on the aftermarket, you can get an extra 12mm of spline engagement. If that turns out to be enough to solve this problem, it'll be worth the increased cost of the improved shaft.

-I wouldn't know the cost of the improved components but I would think it would be worth the cost of another tear down and replacement in terms of parts+labor for this job. When I do a repair/upgrade/fabrication I always consider how much confidence I have in a known component that's problematic. This holds my attention particularly when I'm hundreds of miles from home on a road trip and it's 2AM in unfamiliar territory. The peace of mind alone, to me, is always worth the extra trouble and expense. Fixation on odd noises tends to ruin the journey, roadside repairs tend to ruin the wallet.
 








 
Back
Top