Thanks for posting this. Your sensing a "very subtle vibration" which led you to check the drain plug and find "a few very small, very shiny metallic flakes" is probably the best early detection method available to us short of final drive disassembly or very frequent lube changes.
Also, thanks for the link to bearing failure analysis. I've read that and similar descriptions with interest before. I think our final drive is best represented by figure #10 showing both axial and radial loading. However, the loading of our final drives, being on a motorcycle, is going to be much more dynamic than would be a similar axial and radial loaded setup in a static machine. I have tried to correlate the spalling pattern with the final drive setup with limited success. Spalling of the both the inner (the rotating race in the FD) and the outer race (the non-rotating race in the FD) has always seemed to be randomly located. There is just more of it as the bearing degrades. I suspect that once the pitting process starts, and metal flakes are being generated, the spalling will continue randomly, where ever the metal flakes happen to get caught between the balls and races. I think that's why the neat diagrams of bearing analysis shown in the document you linked to don't correlate well with failed bearings from failed drives that I have seen.
I've cut open quite a few of these bearings in all stages of failure and your speculation that bearing quality is a factor stands to reason. More recently I have seen bearings where the races were severely worn with out pitting. In these cases there is an increase of fine metal particles (grey fuzz) on the drain magnet but not the shiny metal flakes. These bearings also had excess preload, and the excess wear of the groove was the side which correlated with the axial loading. I have speculated that these bearing that are showing severe wear in the grooves but no pitting were made of better quality metal and were standing up to the stress of excess preload better than earlier bearings.
Consequent to these observations, I agree with you that bearing quality may be a factor in these failures, but I still maintain that excess preload is a significant factor. Among other arguments, perhaps the most salient fact is that if it were purely a bearing quality issue the 90% or more of bikes that have not failed would have failed. Many of those bikes have gone 100K miles without problems and they are running the same bearings as those that failed.
No doubt BMW went to their bearing manufacturer and talked about the issue, and very likely the bearing manufacturer looked at their metallurgy and maybe improved the metal going into the bearings. If we are getting better bearings now, that's really good news. Maybe we are. But unless properly setup they are going to fail.
Based on my observations, I think your statement: "This case does not indicated [sic] a preload issue to me as much as perhaps a brg quality issue ...." reaches the wrong conclusion. I think the evidence shows that it is more a preload issue than a bearing quality issue, although I agree bearing quality may play a role.
This has been much discussed over the years on this site by members who have training in mechanical engineering and this post is a revisitation of an old discussion.
Anyone who has a failed final drive should be cautioned not to just rebuild their drive with a new bearing, thinking that the new bearing will be of better quality. Proper assembly is required.
Originally Posted by giarcg
Performed the 60K scheduled service on my 2000 LT and the final drive drain plug was clean except for the usual "paste" thats been found at every 12K service since I bought the bike at 19K miles.
Hard to say when I started sensing a very subtle vibration from the rear (maybe well before the 60K service) but it was enough to make me pull the drain plug 300 miles after the 60k service and I found just a few very small, very shiny metallic flakes. Refilled the final drive and did another 300 miles. Checked it again and sure enough, a few more flakes. That was all I needed to see. Wheel play was present.
Drop check on the new bearing showed .28mm clearance. Replaced the .61mm factory installed shims with .35mm shims.
I cut the suspect brg apart and sure enough....see pics. Classic case of flaking/spalling in just one spot in the center of the raceway. This case does not indicated a preload issue to me as much as perhaps a brg quality issue (I believe that is BMW's position on the early model failures).
Good reading if interested... http://www.vibanalysis.co.uk/technical/contents.html
Anyway, I'm back on the road and the bike is noticeably smoother....