Update on 2006 Spline Repair and Realignment

[2wheelterry]

An update on my spline repair from 2006... My RT now has 58K, my initial failure and repair occurred at 31K. Today I checked the transmission input shaft spline to disk slop and found it to be the same as when I did the repair. No additional lube has been applied since the repair at 31K. I am a firm believer of checking the play or slop as an indicator of the health of the transmission input shaft splines. Folks from this site turned me onto the check. Pull the left tubberware, starter, use a zip tie to hold the clutch in the disengaged position, put the trans in gear, use a screwdriver to rotate the clutch disk and see how much movement the disk has on the input shaft (watch that the input shaft is not moving). With a new input shaft and disk I had about 1/8” to 3/16” of play. As near as I can tell I have the same now. This suggests to me that I have no appreciable wear of the disk/splines since the repair.

 

Below is a copy of the post that I wrote in December 2006. I’ve removed the links to other posts, I don’t think they work anymore. Also I can figure out how to link to the original photos if anyone wants. The old location no longer has the photos, so I also pulled that link.

 

If anyone has the bad luck of a spline failure and would like to borrow my alignment tool, please send me an email. I’m not on the site too often these days so a PM may not get read for a while. Thanks again to those of you who helped me in 2006.

 

----------------------------------------------

 

Alignment Measurements from a Failed Input Shaft - December 2006

 

It is a long write-up, so be forewarned. I had knee surgery earlier this week, so I had some idle time. This could also be called my failed Saddle Sore 1000. I was 310 miles into my first attempt at a SS1000 and found a permanent neutral. After gliding to the side of I-5, a tow truck took me to the nearest U-Haul in Medford, Oregon. As suspected, my 2002 R1150RT with nearly 31,000 miles had a failed splined connection between the clutch disk and transmission input shaft. I had noticed that the low gear downshifts were very sticky over the last several thousand miles and planned to perform a spline lube during the coming winter.

 

After reading a bunch of old posts, I began to get a general feel for the thinking on these failures. Combining my thoughts with what I gathered in the posts, I came to the following conclusions on each of the failure modes for the transmission input shaft.

 

Insufficient lubrication or wrong lubricant: The splines aft of the transmission had sufficient lubrication and appeared to be serviceable. No evidence of lubricant in the input shaft was present. I do not believe insufficient lubrication alone will take this connection to failure at only 31K miles. Insufficient lubrication appears to me to be a contributor of the failure at most.

 

Material: My failure clearly showed a much more significant failure of the splines on the clutch disk versus those of the input shaft. I do not think that a material issue is a contributor.

 

Angular misalignment between the engine crankshaft and transmission input shaft: This assumes the crankshaft and input shafts have misalignment at an angle from each other. This may be in addition to a radial offset alignment. The way I think this through, the clutch plates and disk would show abnormal wear. It seems like more wear would be present on the outer diameter of the clutch disk. My disk and plate wear was minimal with some heat checking present but nothing that suggests a heavier loading on the outer diameter which should be present if there was an angular misalignment problem. Additionally, the distance of this coupling is so short, I think that an angular misalignment would need to be large to be a significant contributor to spline loading. I can not discount this angular misalignment may be a part of this failure, but I cannot convince myself of it. Furthermore, I cannot see a decent way to measure this.

 

Radial offset misalignment between the engine crankshaft and transmission input shaft: This type of alignment has the crankshaft and input shaft parallel to each other but offset in a radial direction by a few thousands of an inch. Others have suspected this as the main cause of these failures and a couple folks have measured an offset misalignment. At first I did not get this, but after a couple of you responded to a post with an explanation, it became clear to me. None of the other failure modes seem to fit, this one does. This failure mode has been validated by some others. So I decided to follow their lead and perform some type of radial measurement.

 

Once I had the transmission out, I checked the face run-out of the clutch housing cover, the clutch housing (similar to a flywheel) and the end of the crankshaft. The clutch housing cover had a 0.007” run-out on the face, for a non-machined (stamped?) part this did not seem out of the ordinary. The clutch housing had a 0.001” run-out on the face which is fine and I had less than 0.001” for the end of the crankshaft, also fine.

 

The next step was to follow the methods of the two posts referenced above. To do this I needed the front cover removed from the transmission. I was not comfortable disassembling the transmission. I spoke with a couple folks that do this work and decided to go with a local independent guy. Some of you may know him, Ted Porter, owner of the BeamerShop. During our brief discussion, his thoughts as to the cause of this failure mirrored mine. His shop disassembled my transmission and mailed me the front case.

 

With the case in hand, I fabricated a bracket to hold a “dial test indicator”. I purchased a cheap version of the indicator on the net for under $40 shipped. The bracket mounts to the crankshaft, extends through the transmission case. Neil Petersen (NRP) made this same recommendation about a year ago. What I have done, pretty much follows his recommendations. The dial indicator mounts on the end of the bracket and measures the centerline radial misalignment of the bore that holds the bearing on the input shaft. Alignment of the bracket and dial indicator relative to the crankshaft is not critical; it will rotate around the crankshaft centerline. The only requirement is that it penetrates though the case to allow proper measurement of the transmission case bore.

 

There are two locations in the bore (that holds the input shaft bearing) that are notched or cutout. This prevents making a 360 degree sweep of the dial indicator. My friend Joe F. suggested putting some tape over this area so the dial indicator could run over this area and disregarding the readings while on the tape. Also, after taking the cover on and off a few times, I added tape across the bore that holds the seal for the input shaft, this was getting buggered up from hitting the bracket when putting the case on. The key to these readings is that they must be repeatable. I initially had a couple things funky and once resolved, generated repeatable readings. To ensure repeatable readings, I disassembled the cover, dial indicator and bracket and reinstalled. I expected and got readings within 0.001’ of the initial readings.

 

I measured the centerline by rotating the dial indicator 360 degrees and finding the highest and lowest readings, (which were 180 degrees apart). At the lowest reading, I zero’d the dial indicator and found 180 degrees away the high reading of 0.010” total indicated reading, (TIR). This is a 0.005” centerline offset. To the splines, this causes a maximum offset of 0.010” (TIR) at one point of each revolution. Thanks to Twisty1 and Joe_Frickin_Friday who helped me understand this.

 

This centerline offset of 0.005” or a maximum of a 0.010” displacement absorbed by the spline connection seemed way high to me. I discussed it with a couple friends and all agreed this was excessive. I also found a chart in a book titled “Shaft Alignment Handbook” by John Piotrowski, (Figure 5-4) that shows misalignment tolerance of flexibly coupled rotating machinery. The way I interpret this chart, the allowable misalignment is something below 0.0015” (or 0.003” TIR). Per a post from Mark Bohn (mbohn), the “tolerance limit for DIN 5480 splines is 0.003".

 

The next step is to move the transmission relative to the engine. The gearbox is located to the engine using two “tubular dowels”. It was suggested to make offset dowels. A web search for “offset dowels” had many hits and suggested this is a common solution for auto/racing engines. Luckily, a friend, Jim G., also an engineer, has an impressive machine shop in his backyard. Making the offset dowels has the added advantage of a “no regrets” solution. It allows me to re-measure everything and ensure I get near zero offset, without making any modifications to the transmission case. Lastly, I can rotate the dowels if needed to refine my measurements. Once done, I can mark the orientation of the dowels to ensure they do not get moved.

 

January 2009 Edit: Making the dowels at this precision was very difficult. Find a good machinist, who can do precision work on a small scale. This was the most difficult part of the effort. I marked the end of the dowels with a notch (using a hacksaw blade) one notch for the left dowel and two for the right, (or visa-versa?). The notch also helped me define the orientation. I found that I had to do a lot of on and off of the trans case and rotate the dowels until I got as close to a zero offset as possible. I also found that removing and reinstalling the trans case without moving/rotating the dowels could give me variations in readings. The goal is to get readings as close to a zero offset and per the above shooting for a limit of 0.003” TIR. In summary, my thinking is that if you are going to go through the pain of a spline repair, do your best to figure out the root cause and fix that too. With some luck and effort, it will be your last spline repair and maybe also not require the periodic spline lube.

 

[SageRider]

Thanks for the update, Terry

[StuGotz]

Great read Terry... any observations on your trans after the rebuild?

 

MB>

[ProductUser]

A good read and thanks for the information.

 

PU

 

 

[nrp]

Thanks Terry - I find it interesting that you had trouble getting good radial alignment repeatibility on disassembly. I was also unaware of the gap that had to be bridged for the dial indicator measurements. I hope I never find out those details from personal experience on my R1100RT!

 

 

[2wheelterry]

MB:

 

I don't seem to notice a change in the trans. Close to failure I had trouble downshifting to 1st after running at high revs., but that was the disk sticking on the splines. Specific to the trans, it feels the same as before.

[WillieJ]

Were you ever able to determine if the factory machining misalignment that you corrected was on the engine or the transmission?

[WillieJ]

Some additional thoughts and questions on this spline failure issue:

 

I initiated this thread on Advrider. http://www.advrider.com/forums/showthread.php?t=207014

In reading many of the posts surrounding this, the misalignment issue keeps emerging. In fact that is why I installed a low mileage used transmission rather than repairing the original, reasoning that if there was a machining alignment issue, there was at least a 50/50 chance the bad machining would be on the transmission.

 

 

 

So, a few additional questions.

 

1. If there is a machining alignment issue, has it been isolated to the transmission, or could it be the rear of the engine?

 

2. Since I did not replace the clutch carrier, only the disk, what is the chance that it is carrier misalignment causing the problem?

 

3. Have any other potential causes emerged that I should address when I pull the bike apart for an inspection? I've put about 8000 miles on the repair. Since the original failure occurred at about 13000 miles some wear pattern should be evident at this point.

[awagnon]

Thanks, Terry, for the followup. It would be interesting if you actually looked at the splines rather than just checking the play which may not detect early wear.

 

The splines on my 2002 RT failed at 43,000 miles. I took the transmission off to lube the splines, and found the splnes were worn to just before failure. I replaced the input shaft and bearing and clutch plate. I did not have the tools or expertise to determine if there was an axial mal-alignment. I lubed the new splines well with Honda Moly-60 and planned to take a look after 10,000 miles. Yesterday I pulled the tranny after 10,000 miles on the repaired splines and the splines were dry. There was evidence of fretting corrosion, and early spline wear in the same pattern as before. So, obviously I must have an alignment problem. Now to measure and fix it......

[awagnon]

Terry. Would you be interested in renting out your measuring setup? Also, would your friend with a machine shop be for hire in making new locating pins with an offset? Now that he's done it once, he may be able to do it second time easier. My options are limited.

 

Thanks,

Al