Rear wheel bearings, my take

[Rocket Scientist]

I recently removed my back wheel for tire replacement and low and behold my wheels bearing are bound up again. I've been reading about people shortening the spacer. I don't agree. I've attached a drawing from All Balls on bearing installation with spacer. In the drawing the "fully seated bearing" would be the one on brake disc side, that's held in place with a circlip. The bearing on the cush drive side, should only go in until it contacts the spacer, and not all the way until the outer race seats. As an experiment, I used the axle and some spacers to simulate the wheel being in the bike. As soon as I tightened up the axle the bearings moved freely. What does this all mean ? I believe because the center spacer is aluminum, it's soft enough to squeeze a little bit under pressure. When the wheel is put in the bike and tightened down, the center spacer compresses a little bit and lets the cush drive side bearing move in a little farther. Everything works fine until you loosen the axle and the spacer becomes uncompressed. It then pushes outward on the bearings enough to bind them, but not enough to move the bearing back out. This is my opinion and may not reflect the general view of R3 owners.net or any of it's subsidiary's

 

[TURBO200R4]

adding to your name
Rocket Scientist/Genius

 

[Gregger]

As an experiment, I used the axle and some spacers to simulate the wheel being in the bike. As soon as I tightened up the axle the bearings moved freely.

That's interesting. I understand the concept you're proposing but have a bit of trouble visualizing the experiment. Did you take a wheel complete with one binding bearing (brake side), the long spacer and cush side bearing all mounted, then place the axle through all this, complete with outboard spacers so you could tighten the assembly, simulating it being mounted on the swing arm? Then, in this position you noticed no binding as you spun the axle in the bearings? If so, it makes sense and would answer a lot or questions but it seems hard to fathom that the aluminum spacer is that flexible. It would be nice if someone could fab up a steel spacer to replace it.

I do agree with you about not shortening the spacer though. So long as the bearings are installed as per "All Balls" directions, things should be fine even with a slightly longer spacer. Just don't bottom out both outer races.

 

[Rocket Scientist]

That's interesting. I understand the concept you're proposing but have a bit of trouble visualizing the experiment. Did you take a wheel complete with one binding bearing (brake side), the long spacer and cush side bearing all mounted, then place the axle through all this, complete with outboard spacers so you could tighten the assembly, simulating it being mounted on the swing arm? Then, in this position you noticed no binding as you spun the axle in the bearings? If so, it makes sense and would answer a lot or questions but it seems hard to fathom that the aluminum spacer is that flexible. It would be nice if someone could fab up a steel spacer to replace it.

I do agree with you about not shortening the spacer though. So long as the bearings are installed as per "All Balls" directions, things should be fine even with a slightly longer spacer. Just don't bottom out both outer races.

I didn't put the cush drive on because that bearing always spins free. Then put the axle through with some spacers I found in my junk box to simulate the bearings on each side of the spacer being clamped like they would be in between the swing arm. Bearings then turned smoothly. Before that neither bearing or the spacer would turn.

 

[TURBO200R4]

jmo
the side with the clip/snap ring sets the position the other bearing seeks its own position unless u shorten the space to much then it will bind the bearing or break the wheel.

 

[Dean0]

Makes sense... cheers for that, Fred

 

[Journeyman]

That does sound like a possible explanation, but what is it about the Rocket that causes this symptom when other bikes have the same design?

Those who have shaved off a fraction from the spacer claim to have solved the problem, but the majority here seem to be saying that's crazy talk, as it flies in the face of Triumph's engineers. On the other hand we (me included) deviate from their superior knowledge by tightening the axle well below their recommended setting.

My wheel is off now (another tire change yesterday) and I can barely turn the bearing with the added friction of a nitrile glove. I replaced that bearing with the fancy Japanese bearings about 4,000 miles ago.

Whenever I do change the bearings I have to whack the spacer pretty darn hard to shift it enough to be able to start working the bearing out with a drift. This tells me that it's jammed in there and, yes, when installing I tap that bearing in just until it touches lightly on the spacer. Once torqued down I think the bearing gets pushed in further until it jams down on the spacer.

Shortening the spacer, (but exactly how much?) if it is actually too long, does seem to make sense. But, I understand if it's too short then that will also cause problems.

Watching this thread with great interest.....

 

[Rocket Scientist]

My experiment has convinced me that cause of the problem is that the center spacer is somewhat compressible. Bikes that I've had in past, had spacers with much heavier wall thickness. They were also only three inches or so long, due to much narrower tires. The R3 spacer must be in the neighborhood of 7 inches long. If you double or tripled the wall thickness you would add even more weight to the already heavy wheel assembly. Steel might work, but rust would be a problem. Stainless steel might be the answer ? For now, knowing that when I install and tighten (also less than recommend torque) the wheel in the bike, the bearings are not binding, kind of makes it a none issue for me. Again I wouldn't recommend shortening the spacer. If the cush drive side bearing makes it to the landing before it contacts the spacer you are going to have the opposite problem. Bearings turn easy until you tighten the axle. You won't know they are binding because the weight and leverage of the wheel will hide it until it gets really bad.

 

[Speedy]

Fred, you have described the problem well. But there is even more to the story. Normally, ball bearings are mounted with one race a press fit and the other race a slip fit, with the rotating race pressed or fixed. So, in an electric motor, for example, the bearings are pressed on the shaft and the bearing outer race is a slip fit in the housings. This arrangement lets the bearing align themselves so that there is no side load on the bearing assembly.

In our Rockets, the inner race slips over the axle but is clamped in place when tightening the axle nut as if it was a press fit. Then the outer race is also a press fit, with only one side retained. Thus if the unretained side bearing outer race doesn't align with the inner race during installation and axle bolt tightening, the bearing(s) run with an unintended side load. Add to this the difference in thermal expansion rates between the steel axle and the aluminum hub and you get heat expansion created side loads. Of course any manufacturing tolerances introduce variability from one bike to the next. Thus some bikes have problems and some don't.

In my opinion, the bore in the hub opposite the side with the retainer should be a slip fit so that the bearing can align itself. It is a design issue that was imposed by Triumph's selection of that wheel bearing likely due to cost controls. A wider bearing race can run in a slip fit bore carrying a load that a narrower bearing can carry only with a zero or press fit.

The Rocket has the same problem in the front wheel. I will be changing to my third set of front bearings in less than 36,000 miles. Some other owners have had a similar experience. I may "take care" of the problem this time as I am tired of changing them. (I am on the original rear wheel bearings though. See above.)

 

[Rocket Scientist]

Fred, you have described the problem well. But there is even more to the story. Normally, ball bearings are mounted with one race a press fit and the other race a slip fit, with the rotating race pressed or fixed. So, in an electric motor, for example, the bearings are pressed on the shaft and the bearing outer race is a slip fit in the housings. This arrangement lets the bearing align themselves so that there is no side load on the bearing assembly.

In our Rockets, the inner race slips over the axle but is clamped in place when tightening the axle nut as if it was a press fit. Then the outer race is also a press fit, with only one side retained. Thus if the unretained side bearing outer race doesn't align with the inner race during installation and axle bolt tightening, the bearing(s) run with an unintended side load. Add to this the difference in thermal expansion rates between the steel axle and the aluminum hub and you get heat expansion created side loads. Of course any manufacturing tolerances introduce variability from one bike to the next. Thus some bikes have problems and some don't.

In my opinion, the bore in the hub opposite the side with the retainer should be a slip fit so that the bearing can align itself. It is a design issue that was imposed by Triumph's selection of that wheel bearing likely due to cost controls. A wider bearing race can run in a slip fit bore carrying a load that a narrower bearing can carry only with a zero or press fit.

The Rocket has the same problem in the front wheel. I will be changing to my third set of front bearings in less than 36,000 miles. Some other owners have had a similar experience. I may "take care" of the problem this time as I am tired of changing them. (I am on the original rear wheel bearings though. See above.)

Hi Blain ! Good to hear from you ! The none retained side is a double row ball bearing. It seems based off what you are saying, it could more of a slip fit on the outer race.