Rear wheel rotation stiffens up when tightening axle problem.

[K2-K6]

Good work Ted.

Its a skill that used to be genuinely labelled "Mechanical Fitter" in which the person assembling etc had to be adaptive with approach to competently build assembled machinery.

What could be "fettled" aspect of structural, mechanical and metallurgy, all part of training to cope with various components from disparate scource being built into one machine.

Quite a valued occupation, when we used to make things in this country 

[Bryanj]

Ted, you have just completed something a younger mechanic would fail at mate.
By the way the chain whilst new was not an expensive one

[McCabe-Thiele (Ted)]

Ted, you have just completed something a younger mechanic would fail at mate.
By the way the chain whilst new was not an expensive one

It's marked DID on the links so I'm happy with it to go back on - it's not going to see much mileage in my ownership.

My target date was to have it running for my birthday - I've missed that by two days already. As I have my 400 to use I'm in no hurry to get it finished this month - it's still not very warm in my garage either.

[K2-K6]

It's 23C "up the smoke" in the great metropolis today.

Sweltering darn sarf 

[davidcumbria]

Good fettling Ted. Sounds like you’re getting close just as the weather is picking up👍

I’ve previously been confused by the interaction of the spindle clamping and bearing fitment on other bikes. . Inner races spin easily after bearing  fitment  but then are pressed against the spacer tube by tightening the spindle. I’ve changed bearings thinking they were tight only to find they were fine once removed. Am I missing something?

[K2-K6]

In concept, the inner race of the bearings form one complete tube along with the other bits mounted onto the axle, tightening should clamp all those elements tight together in one continuous load path.

In production reality and tolerancing, the bearing "seats" in the wheel casting and the bearing spacer length would have to be very accurate OR individually shimmed to get an exact match once torque is applied at assembly.

What the common design does is to to use one reference bearing (usually the chain side on rear, often left side on front) to fully fix the outer race hard against it's hub shoulder, then retainer ring to keep it from moving.

The opposite side is usually left with some space between outer track and wheel shoulder, this to effectively "float" in the wheel axialy. That can then be used for tolerance of the assembly comprising of primary bearing, spacers etc. Not too critical in other words, the primary bearing controlling lateral location.

It SHOULD turn freely when all tightened, but will accommodate some bearing preload if need be (that would mean spacer total was less than the two hub shoulder in dimensions) with any significant drag being cause for concern and a look at just what components and their placing is doing in there.

If the rear wheel was tightened with no brake drag and no chain fitted, then likely you'd feel some small shift in rotation with anything significant causing more concern. They should spin reasonably easily, but not like a bicycle wheel does 

It's odd to describe and sounds like a "fudge" to say so, but unlike setting a bearing arrangement for preloaded by measuring turning torque etc, these are just meant to be assembled, correctly,  with no real methods except assuring the right parts are used and in the right place.

Judgement would give that any significant changes as it's tightened should raise questions, as Ted's done, then investigating exactly how the various parts are installed to try and make absolutely sure all is in the right orientation.

Bit long winded, but difficult to describe in a different way,, possibly 

[Bryanj]

Only thing to add to that is that in design they take into effect that a ball bearing hates sideways load and will fail quickly if this is applied unless specificaly designed for this with different, expensive, races

[McCabe-Thiele (Ted)]

It's different but the same in some aspects as the old Mini Cooper S had taper roller bearings in the front hubs that were not adjustable you just torqued them up. The other Minis all  had plain ball bearings.

[Johnny4428]

Good work there Ted! Had to do similar work on my replacement shoes on the 550 except they weren’t binding when fitted.👍

[Mike_Berkshire]

I have this task coming up in a month or two. There I was thinking I just screw everything back together and job’s a goodun! Another learning opportunity looms. I suspect I will be re-reading this thread carefully. Thanks for starting it Ted.

[McCabe-Thiele (Ted)]

I had no issues Mike  when I did my 400 rear wheel so I was well miffed when the 500 did not work out first fit.

After chatting to Ken on reflection I obviously did not fit the right side bearing fully home. The 500 has that short extra spacer that fits into the left bearing as well as the long center spacer. Make sure you fit the short spacer fully home  into the fixed left side bearing before you fit it.

[davidcumbria]

In concept, the inner race of the bearings form one complete tube along with the other bits mounted onto the axle, tightening should clamp all those elements tight together in one continuous load path.

In production reality and tolerancing, the bearing "seats" in the wheel casting and the bearing spacer length would have to be very accurate OR individually shimmed to get an exact match once torque is applied at assembly.

What the common design does is to to use one reference bearing (usually the chain side on rear, often left side on front) to fully fix the outer race hard against it's hub shoulder, then retainer ring to keep it from moving.

The opposite side is usually left with some space between outer track and wheel shoulder, this to effectively "float" in the wheel axialy. That can then be used for tolerance of the assembly comprising of primary bearing, spacers etc. Not too critical in other words, the primary bearing controlling lateral location.

It SHOULD turn freely when all tightened, but will accommodate some bearing preload if need be (that would mean spacer total was less than the two hub shoulder in dimensions) with any significant drag being cause for concern and a look at just what components and their placing is doing in there.

If the rear wheel was tightened with no brake drag and no chain fitted, then likely you'd feel some small shift in rotation with anything significant causing more concern. They should spin reasonably easily, but not like a bicycle wheel does 

It's odd to describe and sounds like a "fudge" to say so, but unlike setting a bearing arrangement for preloaded by measuring turning torque etc, these are just meant to be assembled, correctly,  with no real methods except assuring the right parts are used and in the right place.

Judgement would give that any significant changes as it's tightened should raise questions, as Ted's done, then investigating exactly how the various parts are installed to try and make absolutely sure all is in the right orientation.

Bit long winded, but difficult to describe in a different way,, possibly 

Thanks for that helpful clarification. I remember now what the issue was on a front wheel. After shouldering the first bearing continued driving inward  of the the outer race of the non shouldering second bearing using a socket  jams the the inner race against the separating tube and introduces side load by pushing the outer race further inward. It’s a matter of feel to decide when to stop since there is no shouldering of the outer race. So to get a free spinning bearing I had to pull it out slightly again. When tightening the spindle with its spacers  fixed length of the inner race assembly prevents excessive inward movement of the outer race.

[McCabe-Thiele (Ted)]

It's probably  between my ears rather than reality - when I did my rectification fit of the right bearing instead of using my lump hammer as previously to get it fully home -  I used my Orange Plastic shot filled hammer for the final strike - thinking it might stop a rebound.