Thanks for posting this, I have not seen the rust and pitting on a compression ring like that before. Interesting, I wonder what caused that?
Regarding measurement of axial movement of the inner race with respect to the outer race in the DMAN method, you don't need to push the outer race down to close the axial play "gap" to get an accurate measurement, you just need to push the hub down to make sure the tapered roller bearing is fully seated. If you were directly measuring the outer race during the DMAN method you would need to push down the outer race of the crownwheel bearing, but measurments are taken at the hub. The downward movement of the crown gear assembly is limited by the tapered roller bearing fully seated in its race, not by the outer race of the crownwheel bearing. As long as the outer race is sliding freely in the cover and the tapered roller bearing is full seated, it doesn't matter that the outer race of the crownwheel bearing isn't pushed down to take out the axial play. I had to spend some time thinking about this to "get my head wrapped around it", but I think you'll reach the same conclusion once you sort it out. This was discussed in a thread quite some time ago by someone else (I've forgotten who, but it was a knowledgable poster) but they expressed the same conclusion, i.e. the outer race of the crownwheel bearing does not need to be pushed fully down, only the tapered roller bearing needs to be fully seated.
It is on the "upstroke" that the axial play of the crown wheel bearing must be taken up and it occurs easily. Once the axial play is taken up on the upstroke (a low resistance, and the first movement you'll see on the dial indicator) and the outer race of the crownwheel bearing is pulled up to fully seat in the cover (higher resistance because the race has to slide in the cover, and the second movement you'll see on the dial indicator), you get can get a direct measurement at the hub of the total movement. You'll generally see two distinct movements on the upstroke on the dial indicator. The first, the small low resistance movement is the uptake of axial play. The second, typically a larger movement and requiring more force, is the movement of the outer race in the cover.
You are correct in that the first movement isn't a measure of the total axial play in the bearing, you would have to push the outer race fully down to get that, but it doesn't matter for the purpose of measuring for shim thickness.
In my experience, some bearings will move in the cover easily with the cover at 250F. Others will not, and I have had to heat the cover more. I also grease the bearing seat to make the bearing race move in the cover more easily. But the key is to make sure the tapered roller bearing is fully seated, it doesn't matter that the outer race of the crownwheel bearing isn't fully down on the "downstroke" for the DMAN method.
I have discovered that the crownwheel bearing is "sloppy". Engineers tell the that the "C" designation on the bearing relates to its play. This play makes measuring the axial movement prone to error. I spend some time trying figure out why sometimes the DMAN method and the Service Manual (static) method gave slight different results. I realised that pressure has to be applied evenly all around the bearing or one race will "tilt" with respect to the other. I earlier posted pics of my bearing stabilizer (homebrew BMW special tool) made of a modifed FD cover. Even with that stabilizer in place, care must be made to measure the bearing on opposite sides because the outer race maybe tilted. Only when the same measurement is obtained at opposite sides of the bearing is the measurement meaningful.
In my experience and experimentation, it is the tendency of the outer race to tilt with respect to the inner race that introduces differences between the DMAN and static (service manual) methods. Careful elimination of the tilting by stabilization of the bearing is needed to eliminate the "tilt error".
You don't really need to measure the axial play of the bearing to get accurate measurements using either the DMAN or static methods, so measurement of axial play is sort of academic interest. But if you wanted to get a reasonably accurate measurement of axial play, I think that there'll be errors from the method shown in your pics. You've supported the outer race in two locations for the "up" measurement but your dial indiator is measuring a point 90 degrees to the support points; I'll bet the outer race is tilted, and if you could measure 180 degrees from your current measurement point WITHOUT moving anything you'd get a different reading. You wouldn't be about to do this with the setup pictured because just the pressure of the dial indicator will change things.
In your pic measuring the outer race in the "down" position, the outer race isn't stabilized at all that I can see, again I suspect that the force of the dial indicator may be introducing some "tilt error".
While measuring axial play isn't necessary, discussion of the "tilt error" is important because the tilt error affects measurements in the service manual (static) method. I was not until I was careful to stabilize the outer race and take multiple measurement 180 degrees opposite eachother that I was able to get consistent results between the DMAN and static methods. I've never seen the BMW special tool, only pics of it in the service manual, but I have speculated that it may not eliminate the "tilt error" and this may have contributed to so many FDs being overshimmed at the factory. If the outer race isn't stabilized, downward pressure from a dial indicator or depth micormeter will case the outer race to tilt down, resulting in a larger measurement. This would result in a thicker shim. Think that one over for a while...
Sorry for the verbosity.... or, This post was brought to you by the department of redundancy department, which was responsible for this post.
PS Did you get that ABS computer I sent?
Originally Posted by jzeiler
I just finished up a first failure at 15K for Wes on an 02 LT. Classic big bearing cage failure. There was some oil leaking at the pinion area and I found the compression ring severely pitted and rusty. The inner and outer races were spalling but the balls were in good shape. Drive was over shimmed by 0.15 mm. There was no creep on the pinion pilot bearing
I did notice that the outer race of the bearing would move up about 0.008 inches so I decided to take that into consideration when doing the Dman method of measuring. I would only do one pull up at a time because I could not push the outer race back down that 0.008. I feel you miss that distance if you just push the crown wheel back down because the outer race is displaced by 0.008. By doing one pull up and then removing the cover and doing another pull up I still got consistent readings. They were also matched to my static measurement method.