Wheel bearings / Back wheel alignment

[Mike V]

It was harder for me to understand the mechanical principle behind the design than understanding integral calculus, but once I got it, the design is elegantly simple.

 8-) Well said Julio!

[arvo92]

I managed to take the job on today. No new bearing yet but I drifted the old ones out. Even oakwood was too brittle to drive them out. Finally I used a proper screwdriver and wacked them out with care. A few odd marks for the first try but just cosmetic damage.
In the long run I will go for a puller still. I need these wheels for years to come.

Still a bit confused to properly pre-load the wheels. With front wheel on the bike grease seals off I torque the axle nut 7 lbs and I can not move the wheel from side to side nor force move the metal spacer outside the hat on the left hand side of the bike if sitting on it. Reading info about pre-load the metal spacer should still be moving under 15-20 lbs force, am I right?

Wheel bearing pre-load must count as a right of passage to Airhead ownership.

I think you are right, Barry. I am definitely not there yet I will need to order the spring scale as I just do not know other way to achieve the proper pre-load.

[Barry]

Still a bit confused to properly pre-load the wheels. With front wheel on the bike grease seals off I torque the axle nut 7 lbs and I can not move the wheel from side to side nor force move the metal spacer outside the hat on the left hand side of the bike if sitting on it. Reading info about pre-load the metal spacer should still be moving under 15-20 lbs force, am I right?
 

Yes if you leave out the bit about being able to move the spacer outside of the bearing stack. Rocking the wheel is the only test needed to determine the correct wedding band size at least in this test method.

And what you have found is not so unusual as even if a previous owner has not messed with bearing pre-load they tended to come from the factory like that as described in my experience here http://www.bmwr65.org/cgi-bin/yabb2/YaBB.pl?num=1382821010/5#5  

In my view there is some scope to stop short of the full specified torque value and that's what I have done with my front wheel but not at 7 ftlbs which is much too low. You will need to put in a wider wedding band or add a shim.  The wedding bands only come in 2 thou increments so it will be interesting to see if that allows  elimination of all play at the specified torque of 32 ftlbs. The reason I settled for a lower than 32ftlb torque value is because when I added the thinnest shim the free play was not quite eliminated at 32 ftbs. I decided I was better off without the shim. Maybe one reason why the factory shipped the bikes with too much pre-load is because the 2 thou increment in wedding band size is simply too big and they thought too much pre-load was better than none at all.

[arvo92]

I guess the only way to determine the effect of "jump" from one increment up with 0.002 is to try it out with the wedding band ordered from one of the suppliers. I will order two wedding bands 2 steps up (0.002 and 0.004) and will give them a go. I will try to see what torque value for each of them will torque them to the same result as I get with current wedding band and 7 ftlbs. It will take some time for me since I will be going away for roughly 2 weeks but I will let you all know when I get back.

Even though 7 lbs got my front axle fully torqued, the bearings were good after being torqued to proper values (I assume 30 ftlbs). They are from the same set as the rear ones which failed. I assume they were replaced at the same time or might be even originals (30203J2/VT108 - SKF/Italy). So I guess reasonable over-torquing is fine but there definitely is the limit which destroys them.

[Mike V]

Arvo,

I'll try to explain the best I can the procedure I use for preload. I like to have the wheel off of the bike and captured in a vice or in my Work Mate. The important part of this exercise is to have a take-up collar or spacer to substitute the space the swingarm and final drive takes when the wheel is on the bike. This collar can be purchased from Cycle Works or you can fabricate one, or use a 3/4" pipe at appropriate length. The important part about the collar is that the sides are as close to perpendicular as possible so equal force is applied to both bearings when the stack is torqued.  

Clean new bearings and races, lightly oiled, no grease seals or top hats.
Assemble axle, and take-up collar, torque axle nut to 20 Ft-Lb.  If using the string pull test - take readings and note your target in-oz of pull rate.  If the preload is too high (stiff axle) you need to increase the width or thickness of your wedding band.  If the preload is too light, you need to reduce the width or thickness of the wedding band.  The wedding band is a simple spacer inside the stack of fixed spaced races and bearings. The outer races are not moving, they are captured and fixed inside the wheel hub.  The rollers on the other hand are being adjusted (in width) to match the correct width of the outer races by the way of the wedding band.  Do some mock-ups, experimentation to understand what's really going on. Do some trials with different torque loads on the axle nut in increments of 5 Ft-Lbs and notice the difference in axle rotation resistance.  You are looking for the correct preload to diminish and stay constant beyond approximately 25 Ft-Lb of torque.  Re-read Barry's post and even go to Snowbum's web site and study that too.

You are looking for a consistent preload amount with the axle nut torqued to ±25 to 30 Ft-Lb.  NEVER exceed 35 Ft-Lb of torque on the axle nut.  The axle should turn in your hands smoothly but with a slight amount of drag.  But it should never wobble (loose) or move when grabbed and pushed and pulled alternately. (wiggle test)

When you have determined proper preload, clean your bearings and pack well with bearing grease, install top hats and new grease seals, torque to 30 Ft-Lbs.

That's the "general" explanation of my procedure... 

[arvo92]

Thank you very much, Mike, for this illustrated summary. It pretty much says it all. I think you are right about taking the wheel off and using a spacing collar. That is the only way to measure the drag you have.

So you are not using the top hats for the measurement? That is something I have missed so far while digging into different sources of information. Do You just subsitute them with a longer spacing collar?

My main problem is getting a proper spacing collar for this trick as I do not have an access to a lathe. As I understand the exact length of the spacing collar is not so important as long as it replaces the width of the missing swingarm and final drive and you are able to torque the nut freely. But the perpendicular sides seem vital, that is why I am a bit worried to buy a 3/4 ince tube from a hardware store.

[Mike V]

Arvo,

The top hats have no affect on the preload determination. Think about where they're positioned ... 'outside" the rollers while the wedding band is positioned "inside" the rollers which will directly affect the spacing width of the rollers.

I think you're getting it now.  Yes, the collar is important to be as close to perpendicular as possible. Some here have used pipe sections from the Hardware Store.  If the collar you decide on is too raw at the ends what you'll experience is "radial stepping" when turning the axle.  You'll feel it spin-stop-spin-stop, if you get what I'm saying. You will have to determine on your own what to use. The closer the inside diameter of the collar to 17mm that will fit nicely on the axle will help also so it self-aligns on the inner races of the rollers.  Maybe those that have used hardware supplies can help you with material.

[georgesgiralt]

Hello !
I use a spacer to replace the front fork thickness or swing arm/transmission thickness when doing my wheel balance after changing tires.
I've found 18mm dia plastic tubing used in home electricity to be both cheap easy to cut and sturdy enough. In order to press properly on the inner race of the bearings, I put the bearing stop (number 3 and 12 on the image below). This is made for perpendicular and correct dia on the bearing...

[livingdeadhead]

a screwdriver is much too hard to use as a drift ! a rod about a foot long,  circular , and about 1/2 " thick , plain mild steel is fine , i've yet to mark a wheel . anything softer seems to deform too easily and/or absorb the impact which defeats the object , besides a screwdriver blade can break , seen it happen , hope you wore eye protection , you only get one set!