Jump to content
IGNORED

Measuring Oilhead Engine-Transmission alignment


nrp

Recommended Posts

 

Let's try this and see how it goes..............

 

The Airhead, R1100, and R1150 Oilheads have a dry clutch with no pilot bearing in the clutch assembly. Rather they rely on the engine's main bearings to establish the rotating centerline of the clutch pack, while the transmission mainshaft bearings establishes the rotation axis of the transmission input shaft. This requires a very close control of the rotating axes between the engine crankcase, and the transmission housing. The two axes meet in the dry area using a spline connection between the clutch disk and the transmission input shaft. A spline can be thought of as an internal/external 1:1 ratio gear set. Although it may seem like a released (locked) clutch disk would remove any locked-in misalignment, this is only true when tyhe clutch is disengaged. When the two parts lock together (i. e. clutch engaged) and rotate, at every 180 degrees, any disagreement between the rotyating axes will create indeterminate bending loads in the spline/hub assembly because all the other parts are so stiff. These radial loads cycle for every engine revolution as long as the clutch remains engaged. The radial forces on the spline teeth may well be orders of magnitude greater than the combined tangential forces required to drive the bike. The result is hi-cycle once-per-rev metal fretting in the spline area. It can be seen as dark gray dust in the spline area on oilheads. Fretting wear debris can be seen by removing the starter to peer into the clutch housing. Careful wiggling of the clutch disk can give a qualitative feel for the tangential clearance in the splin teeth. Because there is only initial assembly lubrication in this otherwise dry area, substantial spline wear appears to occur if the alignment of the two axes is more than about .003 inch out of radial alignment. There are several possible resons for mis-alignment wear debris but they all appear to be fixed by the initial manufacturing or assembly errors, and only possibly by subsequent events. Spline failures seem to be randomly distributed in the R11XX Oilhead fleet over a period of many years. Explanations for dimensional discrepancies in this area have not been made by the manufacturer. Speculations by others include 1) a random machining error in one of the components, 2) the clutch housing being deformed from shipping damage, 3) inadequate (partial) spline engagement of certain models, 4) heat treats, 5) incorrect factory lubes and no ongoing lubrication provisions, 6) assembly techniques, and various other anomalies. I don't have an answer and apparently neither does the manufacturer as spline failures have existed for about 25 years of manufacture. I even found my '75 R90/6 (yes, an Airhead) has an inherent alignment problem that explained a spline failure years ago. Even yet, if the engine-transmission bolts are loosened at neutral-idle, the transmission still wiggles noticeably.

 

Some oilhead and most airhead R bikes last forever, yet others have failed repeatibly in only a few thousand miles. Failure is something built in when the bike is new. The manufacturer's warranty approach for spline failure has been to quietly replace the clutch housing and transmission as a unit suggesting initial manufacturing error. Post warranty failures have been explained as unique to a riders driving profile - whatever that may be. On the other hand some of the original model bikes runup impressive mileages with little or even no wear.

 

The clutch area was redesigned in the R1200 Oilheads and the now-lubricated clutch & spline seems to have eliminated failures on these more recent machines. This posting does not apply to those models. But we are faced with a legacy of Oilhead models that may experience that sudden catastrophic failure, which gives little or no warning. The clutch disk hub spline is virtually always stripped out completely, and the transmission input shaft splines are usually so worn they are also unsalvageable. The owner options are to junk the now-unuseable bike, replace the transmission and clutch housing with a used one, or repair the old transmission input shaft and clutch with about $2K of new parts, otherwise to only to wait and see if the spline problem resurfaces.

 

I first wrote my analysis on this about June 2011 with a series of pictures and discussion at http://bmwsporttouring.com/ubbthreads/ubbthreads.php?ubb=showflat&Number=704272&page3 which shows a rigorous way to document any radial misalignment. Unfortunately the hosted pictures have been removed from that thread nor can I even find them on my computer.

 

 

A schematic cross section of this area is sketched below:

 

alignc009.jpg?dl=0

 

The spline coupling includes a flex spider plate that forms the connection between the clutch friction material and the spline OD hub. This thin flex plate allows any small angular axes discrepancies to be accommodated with out adding unnecessary loads to the spline elements. However, the transmission input shaft is supported on a pair of ball bearings, and the crankshaft on a pair of journal bearings. This dual bearing set for each elementch means there is no accommodation in either rotating part for radial errors.

 

Alignment error measurements require a small dial indicator to be mounted on the engine flywheel with its probe sweeping on an accurate measuring surface of the transmission. This surface is best assumed as the transmission input bearing housing bore. Alignment is verified by slowly rotating the crankshaft and reading any dial indicator variation The indicator can be read from the transmission housing rear, with rotation. Ideally such alignment should be within .003 inch total indicator runout (TIR).

 

Many dial indicator sets have a large 1 inch range, and some even have a big magnetic base and miscellaneous rods, clamps etc for use in a machine shop. They are absolutely NOT suitable for this alignment check. The pieces are so large and so springy that they will not fit within the clutch housing, and will deflect to show an incorrect rotation center, if they could even be fit into the clutch housing. However, a cheap common dial test indicator (available new on EBay for less than $20 is ideal for this since it is small, lightweight, and sensitive.

 

I made a simple fixture pictured below to mount the dial test indicator on three of the engine flywheel bolts. I used a pair of rare earth magnets from an old computer hard drive, epoxying them to a brazed extension tube made of 1/2 inch electrical conduit, which had a "precision" (well it was done with a hacksaw) dovetail cut into it to mount the indicator. A clamping cross screw, and a relief cut allowed the rig to clamp the dial test indicator's dovetail. The rare earth magnets hold on the flywheel bolts like all fury, and a pair of them can span three flywheel bolts (two adjacent and one opposite) to give a rigid tripod support. Note that care must be used around these extremely small but powerful magnets including removing your watch and having a iron keeper plate in place for shipping.

 

align5.jpg?dl=0

 

The dial test indicator has an indicator range of only +/- .015 inch so everything has to be set fairly carefully so that the indicator doesn't over-range. The picture shows how the fixture fits into a clutch housing and the surface that it reads.

 

alignmark.jpg?dl=0

 

The alignment measurement fixture fits into the clutch housing area and reads on the input bearing housing bore.

 

The run out measurement requires all the rotating stuff, including the input bearing, to be removed from the transmission housing. This hasn't seemed to be a major problem for others, as most of this has to be done to rebuild the clutch and transmission anyway. The indicator dial is read looking into the transmission case from the rear. I read the indicator about every 45 degrees of crank rotation to average the results. Then optionally in a second test, at each angle of indicator rotation, I note the flywheel's radial motion under modest prying to measure the backlash amount and direction of the engine main bearing system. I have found that bikes with alignment problems also have engine main bearing wear as those spline radial forces are also reacting into the engines main bearing system in a consistent direction.

 

The easiest way to correct a measured radial alignment error is to make up new slightly offset alignment pins to offset the transmission and housing from the engine the amount measured from above. Since the measurements of the engine bearing clearances are made statically, if there is a substantial egg-shaped main bearing wear is found, either the engine main bearing should be assume that the crankshaft will run at the center of the optional measured egg rather than simply at the static position. Fortunately there is enough clearance in the engine-transmisson bolt-holes to accommodate any necessary shift. I have noted in a couple of repairs that the transmission axis was about .008 below and slightly to the right of the engine center. This is too small a sample amount to be consistent, but it is interesting that those errors seem to be similar.

 

Once the offset has been established, with my old lathe and some raw material, I can make up a set of tow offset alignment pins which are easily replaced, and these are marked so they can oriented properly on assembly. To date I have not been charging for new offset alignment pins. There are also others that have made new pins

 

The alignment test fixture is easily mailed. I have not been charging for loaning it out except for nominal postage and if the indicator is damaged. Not happened yet......!

 

 

 

 

 

 

Link to comment

Morning nrp

 

That pretty well covers it.

 

When making your corrections--

 

The clutch area was redesigned in the R1200 Oilheads-- The 1200 OILHEAD uses the same dry clutch type set up with no pilot bearing. The 1200 hexhead uses the same basic dry clutch set up with no pilot bearing but a deeper input shaft to clutch disk engagement . The 1200 camhead uses the same basic dry clutch set up with no pilot bearing but a deeper input shaft to clutch disk engagement . The 1200 wethead is the BMW with the new design wet clutch & multi-plate clutch pack running fully lubricated.

 

The 1200 hexhead & 1200 camhead do use a full input shaft spline engagement through the clutch plate splines so that is different than on the 1100, 1150, 1200 oilheads.

 

The easiest way to correct a measured radial alignment error is to make up new slightly offset alignment pins to offset the transmission and housing from the engine the amount measured from above.-- I would think you would want to move the transmission ½ the distance measured above. You are measuring using a 360° sweep not a 180° sweep so .008” through 360° would need a .004” movement to correct to center.

 

You kind of discounted angular misalignment but I'm not sure that I totally agree with that as the crankshaft mounted clutch parts are fairly rigid (too rigid to flex & forgive angular misalignment). The clutch disk itself has a somewhat flexible center but that flexing is over a very short distance & stiff enough to put & keep a significant & continuous bending motion on the spline joint. The problem with an angular misalignment part is the difficulty in measuring it so it usually gets overlooked in normal radial alignment checks.

 

If the spline wear causing misalignment is due only to OEM mis-machining of locating holes, or incorrect trans front cover hole lay-out & machining then, yes, we are probably looking at only a straight-offset type alignment problem. On the other hand if we are looking at shipping-damage type misalignment (like a dropped bike in shipping) then more than likely we are looking at an angular type of misalignment or both angular & some straight offset due to trans front cover bending (I can't imagine how dropping a bike could cause a straight offset type misalignment, but easy to imagine a bending type misalignment).

Link to comment

My early R1100RSL with 5 speed trans also has full clutch disc spline engagement. Might help the wear issue but i doubt it totally fixes the issue caused by no pilot bearing and misalignment. When I first obtained the bike it had zero miles on the odometer and before i ever started it I removed the trans and cleaned/lubed the input shaft splines. Made me feel better but I don't think it solves much.

 

 

Link to comment
Any thoughts on why k-bikes with the same clutch arrangement do not fail to the degree (or maybe at all) oil-heads do?

 

Morning Craig

 

They did have a number of spline failures on the old K bikes-- I can't say how many failures compared to the R bikes but enough that BMW sent out a service bulletin on the old K bikes.

 

Here is a section of the BMW bulletin that pertains to the K bike-- "Even a slight misalignment (above 0.25mm) can lead to spline wear between clutch and input shaft"

 

Unfortunately, while the bulletin mentions the K bike misalignment/wear issue it doesn't address on how to measure or correct it.

Link to comment

A quarter (0.25) mm is a massive alignment error. Would be about ,020" total indicated run out (TIR).

 

I apologize about starting this forum thread with improper picture hosting, but have been unable to get the attention of the moderator to have the thread deleted. I can't even edit it. What's the policy/time allowed here on editing?

I plan to repost it when I get this all figured out but it may take a few days. The last Windows I understood was XP.... :dopeslap:

 

Comments - Spline engagement on some 1150s doesn't seem to be responsible for failures. Some last, and some don't. The only failures I been exposed to have all been on R1100s (or 20 years ago on my R90/6). I suspect the deflection of the clutch housing during manufacture has been the source of all this.

 

I'm not familiar with the LTs inside (thank goodness as I have a 2005 K1200LT) but I'll bet the clutch housing flanges are more substantial - especially at the engine end.

Link to comment

I recently had my 2K LT apart and the flywheel is cast aluminum with machined mating surfaces for the housing.... nice looking part.

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...