Wheel bearing question for the experts

For the sake of this forum, let's say this is about an XJ. I'm sure it applies to many bikes anyway.

On bike wheels that have a machined shoulder on one side of the wheel for a bearing to seat and no shoulder on the other side with a spacer in between, how does one seat the non shouldered bearing so that there is no side load? Too tight against the spacer and there is a constant load from the center outward, too loose and there is a constant load inward toward the center when the axle is tightened up. How does one get it just right?

 

I'm no bearing expert, but it seems nobody else wants to play. :lol:

One side at a time...

So this machined surface on one side of the wheel... It contacts just the outer race of the bearing on that of the wheel? You press the bearing into that side of the wheel and this machined surface establishes the depth of the bearing into the wheel on that side, right?

Tough to do without pics or sketches, and I've got to admit that I've never had the pleasure of messing with the bearings on my bike!

 

Thats just not right what bike are you working on

 

My thoughts too, that's why I haven't piped in sooner.

I don't think I've ever been into one that didn't have a shoulder on both sides to locate the bearing race.

 

Are you sure you have the right bearings? All the ones ive seen are nothing like that.

 

All right! I got some responses. I guess I need to clarify.

The bearings in question are the front wheel bearings on my Kaw. See pic below. One side of the wheel has a shoulder for the bearing outer race to seat against (upper 601). The other side does not. There is a steel spacer tube (92027) that fits between the bearings that sets up the seating depth for the bearing on the unshouldered side (lower 601). The problem is that the process of seating that bearing presses it against the spacer causing it to have constant outward pressure on the inner races and if it isn't seated enough, there is inward pressure when the axle is tightened. How do I seat it so that there is minimal or no side pressure?

 

92033 and 92033A are clips that keep bearing from coming out of the wheel?

 

Moving thread... (NOT XJ related but interesting just the same).

 

"lower" #601 should be able to move slightly laterally within the wheel so that it seeks its own final position when the axle nut is tightened, shouldn't it?

 

Hey Fitz. Yes, you would think so but it's a press fit just like the other side.

I got hold of a 30 year motorcycle mechanic and he told me the bearing on the shoulder free side needs to be driven in so that it just touches the spacer - not hard against it and not so far away that the spacer is loose. He said it's an acquired feel. It must be positioned so that both inner races touch the spacer and yet still can turn. If it isn't just right, the bearing life is shortened. Tricky, to say the least.

 

When pressing in bearings, its important to support the outer and inner races equally. First press in the (shoulder) side bearing. Flip wheel over with something flat against the bearing (across both races)for support, insert spacer, press other bearing the same way slowly till it seats. As long as both races are supported equally they will be in plane with each other and not lead to premature failure.

 

Wow. That design obviously works, but... it's really cheezy.

The bottom of the counterbored recess establishes not only mounting depth, but it also establishes alignment of the outer race parallel to the direction of wheel rotation. Without that shoulder, even if it's a press fit, you run the risk of having the outer race be in misalignment. Cocked in the bore a little if you will.

Also, you're trusting a press fit, and only a press fit, to retain that outer race where you want it. Forever. No matter what bumps you hit? No matter what temperature changes it see? No matter hot far you lean over in a turn... Forever?.

I smell bean counter influence...

The input from your 30 mechanic is good advice, but there's more to it than that. What you REALLY want is for the balls to be centered in all the race grooves when the alxe is tightened up. Even if you press against both the inner and outer races when you insert the second bearing, you run the risk of having the balls slapped all the way to the outside edges of the groove in the outer race (and obviously the inner edges of the inner races.)

I bet that if you insert it too tight and then go for a short ride, the balls will heat up, expand, and push that semi-floating outer race back a bit. Might start out tight, but would end up with some slack once cooled. Haha!! Might cook the bearings in the process, but alignment will be achieved!

 

92033 and 92033A are clips that keep bearing from coming out of the wheel?

Actually the clips would keep the wheel from walking off the bearing so insert the bearing on the non shouldered side just enough to get the clip in. At least thats what it looks like to me

 

Thanks, guys. To your points:

Orange and Waldo: Agreed.

Rice:

Maybe, but they didn't change it for 20 years.

Agreed, although I can't see how the bearing can be cocked much when the outer race is in full contact with the bore. At least not cocked any more than the play built into the bearing itself.

The snap rings will keep them from going too far.

I'm not sure if that design is really any cheaper. It still has to be precision bored with a snap ring groove cut. The only difference being that it's not bored to a particular depth.

Right. That was his point. Having no side load on the bearings when all is torqued up.

That is a good point - and maybe the reason for the design - allowing the bearing to "find its happy place".

Excellent feedback! I never thought installing wheel bearings would churn up the gray matter so much.

 

Yeah, I don't mean to give the impression that I expect the bearing to come flying out. My point is that I just don't see the advantage of not simply counterboring both sides like other applications. You're right about the cost. Can't be much different and it sure makes assembly more finicky!

Maybe they didn't think they could hold the length accuracy on the spacer between the two inner races? :?:

Two precision depth counterbores vs. one side counterbored and the other side bored through.... All with snap ring grooves.

So when you put this thing together and tweak the preload so you've got the balls centered in the races, is there a (intentional) gap between the snap ring and the semi-floating outer race?

 

Sounds like you've done some metal machining.

I originally seated the bearing too tight so I had to fab a tool in order to tap it back out a bit from the other side. The way I have it set now is with no gaps. The inner races just touch the spacer so they can still be turned, whereas before I had them so tight I couldn't turn them. It makes me wonder how they do it at the factory. I have the Kawasaki service manual but it doesn't go into any detail.

Now if I can just get used to handling sub freezing metal, I think I could finish this project.

 

Yes, I do some metal machining. Doesn't pay the bills, but it keeps me out of trouble (for the most part). :lol:

So what's your take on the "why"? Ease of machinability? Alignment accuracy? Tolerance stack-up? Sure as heck sounds like it's not "for ease of assembly"!!

Whatever the reason, it was good enough for them to keep it like that for 30 years, right?

Out of curiosity... Do you know what happens if you press in the semi-floating side first. A little past the snap ring groove. Put in that side's snap ring and then (from the counterbored side) press the first side back out a little until it's snug against the snap-ring. Then put in your spacer tube and press the second bearing into the counterbore. Is everything too loose then?

PS - Haha!! Don't let the bearing purists hear how tapped that bearing back out and re-used it...

 

Some engine bearings don't have a shoulder,snap ring or bearing retainer to hold them in place. The only thing that holds the bearing in place is the press fit. If a bearing moves its because of wear in the bore. In the case of the rim the bearings not going to move in because of the steel bushing and its not going to move out because of the retainer ring. I know it seems to be better to have a shoulder on each side, but if you've ever had to remove a destroyed bearing before,you would be thankful for not having a shoulder on each side. As far as cocking the bearing goes, that will always cause a problem to bearing and bore. So its important to make sure its straight to start with and then press, you can think about it but don't use the hammer. :lol:

 

Rice, I made that special tool so I could tap the outer race, not the inner one, so the bearing should be fine. It didn't take much to get it just right. I think I was pretty close already. Good idea about the assembly procedure. I might try that on the rear - it's the same design. I might email their training facility ask them why. If I studied mechanical engineering instead of electrical, I might know the answer.

Orange, the Kawasaki manual describes cocking the spacer slightly so the inner race can contacted with a long drift and a hammer to drive it out. Then the spacer is removed and the shouldered bearing is driven out the same way. Installation is with a special tool (looks like a large socket so it only contacts the outer race) and a hammer. With the wheel hub heated and the bearings from the freezer (my garage), they tap in pretty easy.

PS: Rice, I took a metal machining course at the local college just because I always had an interest in it. I learned a LOT and it was great fun.

 

I know it's academic at this point, but I wish I knew what the advantage is... If I knew why, maybe I could use it in the future in some application. It's those curious little tips and tricks that can really become valuable once you understand the reasoning.

I'm an EE too. Probably should have been an ME but could never have made it through the thermo and fluid flow classes without going insane!! 8O

Hey Orange-n-Black, do you have any theories as to the why Kawasaki would make it this way instead of two counterbores?