Why does increased Torque on axle change Preload....Duane?

[mattcfish]

So after over 200,000 miles I'm changing the bearing on my /6. I didn't need to, but I powder coated my wheels and mistakenly thought the heat would not harm the races. The bearing were removed and everything was well masked, but it turns out that heat over 250F (powder coating uses 400F) softens the races and could contribute to early failure.
I don't want to worry about my bearing when I'm riding to Billings this summer, so the can of worms is opened.

Anyway, I want to use Duane's method for adjusting preload, but was wondering if anyone could tell me why it works.

This is the section of Duane's article that actually explains his procedure. Funny thing is, I don't think he explains why tightening the axle would introduce more preload. All the torque goes to the inner spacers with this system, not the bearing. My guess is that the spacers must pucker out slightly with torque, which...in effect, shortens them. That causes the bearing to move slightly closer to the outer race and preload increases. Like I said, that's my guess.


"Checking the spacing of the /5 and the /2

The BMW method requires removal of the stack to check or change the spacing. This is not only time consuming, but allows for an inaccurate result, especially important with the /5. A common failure (mentioned above) of the /5 is the bearings becoming loose in the hub. The wheel is then loose on the axle and allows for funny handling. It is easy to test for this fault and prevent a failure.
Warning, use only the very short 22 mm wrench from the tool kit, or a torque wrench. Using the short wrench makes it very hard to over tighten the nut.
I have a very easy test that you can perform to determine the spacing. I call it the "shake the wheel" test. This takes two persons. Have your helper, on the right side, with the 22 mm wrench at the axle nut. Have the pinch bolt loose.
"Tightener"
My helper is taking the picture, so I must do both for purposes of taking this picture. I am in position to adjust the axle nut for testing.
"Shaker"
You are on the left side with the wheel in one hand and the frame in the other hand. Try to feel side play, you should feel nothing with the axle nut tight. Loosen the axle nut a turn, or two and feel the play. Once the "shaker" fells the play, the "tightener" slowly tightens the axle. The shaker can feel the play go away as the tightener slowly tightens up the axle nut. The shaker just orders with "more" until the play is gone. Do the test again for accuracy. Now go around and loosen the axle nut with a torque wrench to see how tight the axle nut was. If it happens under 15-20 lbs, then the preload is too tight. If it happens over 25 lbs. then the preload is too loose.
The simple solution
The easy fix for the /5

To change the spacing means changing the thickness of the inner spacer, or wedding band. By installing a thicker one, it will space the bearings looser. This will fix the problem. This can be done by one of two methods. I will first describe a method that doesn't even require removal of the wheel from the bike.
Pull the axle. Remove the 5 bolts holding the chrome hub cap onto the wheel. Carefully remove the cap and the seal holder. Be careful, the cone of the outer bearing can fall on the ground, or it may stick to the seal holder. The wedding band may even stick to the bearing too. Reach in and remove the wedding band. Now, try to pull the outer bearing race, or cup, from the hub. If it comes out, then it is already spun in the hub and should have been felt, during the test, as play that didn't go away. The bearing will probably still be ok, but the hub is not. See "fixing or replacing the hub" below.
Replace the wedding band with a thicker one or add a "shim" to the existing one. Shims are available from me and wedding bands are available from your BMW dealer. Which size of wedding band do you need to buy? You probably don't even have a micrometer, much less a metric one. With shims it doesn't matter. More on shims below.
For a quick test, reinstall everything, but the chrome hub cap. Don't forget to put the top hat spacer into the seal holder. Those 5 bolts don't really do anything in this test so just use one to keep the parts from falling on the ground. Reinstall the axle and perform the "shake the wheel" test again. Be sure to reinstall the axle correctly. Keep adding spacers or shims until the play just goes away when the nut is about 15-20 lbs of torque. It is better to be too loose than to be too tight. If one wedding band produces a spacing that is 10 lbs and another one produces 25 lbs, go for the 25 lbs. You could lap the wedding ring down a bit and get 15-20 lbs too. Shims make the spacing easy. Install all 5 bolts and go riding.
This "easy fix" only fixes the spacing. This method assumes that both bearings, seals and lubrication are ok. If your bike has junk bearings, no lube or shot seals, then you are wasting time with this type of procedure. You really should remove the stack and inspect all parts.
Full service of the /5 bearings

For both the /2 and /5, BMW recommends about the same maintenance schedule for wheel bearings. Here you will learn how to get more reliability and with less service. BMW wants us to remove the "stack" every 8 or 16 k miles and fully lube and check the spacing. This is not needed and on the /5, and it can harm the hub. Each removal of the stack could damage the hub.
The reason for renewing the lube is twofold. One, it assures lubrication. Two, it cleans out dirt. A type of greasing tool can be bought, or made, that will lube both bearings without removing anything but the axle. With that tool, one can give it a shot of grease every time a tire is changed. Not a lot of grease, but a pump or two. The first I knew of this tool was when my ace mechanic, Brian Hilton, suggested it. He made one and it worked beyond our expectations. It just slides into the axle hole and has O-rings to seal off the bearings. A hole through the center allows new grease to be pumped into each bearing from one side. The old grease is pushed out the other side. The hub isn't threatened by being heated, or the hammering on the bearing stack.
If one rides in the rain for a few days, then it is a good idea to insure that there is no water in the wheel bearings by adding some grease to shove out any water. The water would only get in the outside bearing, so here is the simple way. Just remove the axle and you will have room to insert this tool and give it a shot or two of grease. If the bike was riding through water so deep that it temporarily got water in the rear drum, then it is possible that water could have gotten in both bearings. I would completely remove the wheel and give both bearings a shot or two."

[Duane Ausherman]

Your guess is a factor and I will deal with it first.

The axle nut being tightened up on the /5 and later should give a solid feel, but doesn't. My article should say somewhere that the stack parts are very poorly machined. The ends are fuzzy and will compress slightly. That is a minor affect.

The major one is that the surfaced are sometimes not parallel. That means that the bearings aren't held squarely and the cone doesn't seat squarely in the cup.

Mike too is asking me for more clarity in my article.

If you don't have a shim set, or are not willing to lap the parts to perfection, then just perform the shake the wheel test to see where the stack is pulled together to zero play. Personally, I would just leave the axle nut at that torque. That assures that you don't over heat the hub.

That article has been edited at least 50 times over the years and nobody has asked about more detail for some 5-8 years, so I thought it was fully explained.

Keep in mind, you aren't protecting the bearings, but the hub.

[barryh]

When I was setting up my bearings I noticed this effect of some play being taken up with torque but it didn't amount to much. Maybe 0.0005" to 0.001" at most. If the spacers or shims were too thick I could torque as tight as I liked and the last bit of play would still not go away. Changing to the next smallest spacer then makes a huge difference and all play goes away at a level of torque which is too low. This is the flaw in the BMW system in that the increments between spacer sizes of 0.004" is simply too big to get the pre-load spot on. They really should be available in 0.001" increments. I finished up making and lapping my own shims and used Duane's method to set an acceptable torque level. When I did this the first time on bearings that hadn't been touch since new I also found the factory settings to be grossly too tight as per Duane's article.

Taking a perfectionist approach to wheel bearing setup will have you lapping to the nearest half thou but it's really not that critical and I found it very easy to improve on the factory set up.

[mattcfish]

Thanks Duane and Barry, that clears it up a bit. Imperfections and metal distortion. I'm doing this on Lester's, so there are no wedding bands. Will need to machine down the main spacers (if necessary) and insert shims behind the bearing. Like I said before, these bearings have over 200,000 miles on them, despite what Snow Bum says, somehow, Lester got the preload right.
I was wondering if mounting a dial indicator to the swingarm or fork tube to measure side to side wheel movement might be handy during the "shake the wheel"? Or is that overkill?
Waiting on my bearings and shims, so I have time to over think this.

[Duane Ausherman]

It is way overkill. Plan to do the "shake the wheel" test several times and trade positions with your helper. Soon you both will be able to feel the play go away. I prefer to stop when the play goes away and leave the axle nut there. I prefer that to happen with around 15-20 lbs of torque. Don't tighten further.

That whole thing about preload is flawed. One bearings like this, it is well known that "some" preload will allow the bearings to last longer.

On the /5 and later, they are calculated to last well over 2 BILLION miles. That is under the assumption that the bike is laden to 1000 lbs and going 100 mph all of the time. Further, that failure point is when only 10% will fail.

Do you load it to 1000 lbs and drive 100 mph forever? These bearings are super over sized for the axle size. Suppose that you don't get the spacing correctly set and they only last 1/10th of one percent. That would be 2 MILLION miles. Is that enough miles for you? See, there is no reason in the world to bother about setting some preload to get max miles.

It is the hubs that fail, not the bearings. Setting for preload only increases the risk of hub damage due to heat.

Follow my article, protect your hub.

[Ken in Oklahoma]

. . . Follow my article, protect your hub.

I do.

It's actually satisfying to have a mental picture of what is going on and the dynamics and engineering decisions that go into the design and making of things. I also enjoy the mental imagery of other things at work, like forks and shocks at work, engines and transmissions, carbs and camshafts, and the list goes on.

Would that I had similar mental models of what goes on with less tangible things such as people, including myself.

Ken, waxing philosophic in Oklahoma

[enigmaT120]

Would that I had similar mental models of what goes on with less tangible things such as people, including myself.

Ken, waxing philosophic in Oklahoma

People are tangible. Some, like me, don't like to be tanged by just anybody, but even I'm tangible if you can catch me.

[Major Softie]

Would that I had similar mental models of what goes on with less tangible things such as people, including myself.

Ken, waxing philosophic in Oklahoma

People are tangible. Some, like me, don't like to be tanged by just anybody, but even I'm tangible if you can catch me.

I can certainly understand you not wanting me to touch you with my tang...

[Chuey]

People are tangible. Some, like me, don't like to be tanged by just anybody, but even I'm tangible if you can catch me.

I thought you could only be tanged by an astronaut.

But wait! I forgot Pootie Tang! http://www.zimbio.com/watch/ofa1V2sgzE0 ... anda+Sykes

Chuey