A practical machining/engineering problem. Please look in. Tah.

[CC12345678910]

Alright peoples, i've got a quick question for you as all them answers the seach engines are returning go flying straight over my practically-capable-but-never-done-A levels-or-any-of-that-guff  brain. I'm sure there are people on here who know this sort of backwards so, if you please;

What would be the (perhaps ballpark) rate of contraction in aluminium (in this case a 15mmOD, 2mm wall hub axle) when frozen to minus 16C (the coldest the house freezer goes to on fast freeze)?

And, how much would the 15mmOD 2mm wall Alu sleeve i am going to make expand under heat from a butane/propane mix plumbers torch? I don't know the temperature the torch is capable of but suffice to say I plan to heat the snot of it.

All of this is in aid of making a good hot/cold interferance fit, but not so tight that the sleeve cannot be removed in the future, In case anyone was wondering.

I have the option of using 16mmOD 1mm wall tubing for the sleeve but fear that would have to be epoxied to the axle; A process that has given me mixed results in the past and would be a right b*tch to remove later down the line.

Thanks as ever.

 

[forteh]

Make the hub axle steel and you'll have much better results. Once you've contracted the aluminum axle and sleeve together it's going to be very difficult to apply heat to just the sleeve for removal.

Also consider only heating the sleeve, if you have to freeze the axle as well in order to shrink it enough to fit then you're not going to be able to simultaneously freeze and heat the assembly for future.

Work on the differential rates of thermal expansion for different materials and you should be good, as to what machining tolerance you need I wouldn't know off hand.

My only real world experience of shrink fitting is dropping the main bearings into the crankcases on my rotax motor, they will not push on cold but drop in nicely with the cases heated to 80°c.

edit: what's the application and what sort of loads are you applying to the sleeve? Why not try a loctite bearing retainer? Something like loctite 641 instantly springs to mind, get it apart with relatively low temperatures.

Edited November 26, 2017 by forteh

[CC12345678910]

The application is/was increasing the axle diameter to fit a later model freehub body with larger I.D. bearings (the only difference to the fitment), so I can make up another "bitsa" hub. If the sleeve can be fixed to the axle then as far as the bearings are concerned they are running on the correct axle and will work as designed.

However if I could buy the correct late model axle or early model freehub body later down the line I would remove the jerry riggery and make everything proper again. I mean if I wanted be real bodgey I could PVC tape the snot out of the axle and slap everything together, but I'm trying to be better than that...

In retrospect I was massively over thinking the job. I have yet to put theory into practice as the hub I was buying fell through but I definately think a combination of the 16x1 tube and @forteh's 641 bearing fit, or more acurately whatever brand I have in the draw when it comes time, ought to work just fine.

[forteh]

So fitting a pro2 freehub to an xc shell and axle? My only reservation with spacing out the axle for the freehub bearings would be to ensure concentricity; easily solved by making it a snug sliding fit and a spot of 641 just to retain it rather than to fill a gap.

[RobinJI]

I might be confused, or missing something, but surely the 16 by 1 tube (14mm I.D.) will need boring out to fit over a 15mm axle? Or is that the plan all along and I'm just misinterpreting it? 

Personally I'd just be looking to replicate the fit of the bearings onto the axle, as the sleeve's basically doing the same thing as the inner bearing race in terms of loading. So a press fit with everything at room temp, and in most hub's cases not even a hugely tight press fit. As forteh says, a snug sliding fit and some loctite should do it. No need to mess around with heating and contracting things, certainly not both or as he says, getting it off will be near impossible. 

[CC12345678910]

I'm making a right bollocks of this aren't I?

@forteh Close. 9spd pro2 old has 15mm od axle, the 9spd pro2 EVO driver I have has bigger bearings with a 17mm Id (factory spec). The aim of the game is to step down the driver to fit a pro2 old, in such a way that the sleeve bodge acts like the proper axle would, and make up a hub from odds and ends. The sleeve spinning and making a gauled up mess was my main concern.

 

21 hours ago, RobinJI said:

but surely the 16 by 1 tube (14mm I.D.) will need boring out to fit over a 15mm axle?

Glad you caught that, what an absolute pillock I am for thinking 15x2mm would fit; what a f**king moron...

Or is that the plan all along and I'm just misinterpreting it?

 16x1mm tube, a Blacksmiths drill, dremel and/or reamer springs to mind, but I'll sleep on it.

Personally I'd just be looking to replicate the fit of the bearings onto the axle, as the sleeve's basically doing the same thing as the inner bearing race in terms of loading. So a press fit with everything at room temp, and in most hub's cases not even a hugely tight press fit. As forteh says, a snug sliding fit and some loctite should do it. No need to mess around with heating and contracting things, certainly not both or as he says, getting it off will be near impossible. 

As I said above, massively over thought this one. It's not one of my better traits. I think watching project binky has gone to my head.

Cheers boys.

[RobinJI]

We all do it occationally!

Speaking of over-thinking, might it be possible to get bearings with a 15mm ID that fit the other dimensions? It might be worth checking somewhere like www.simplybearings.co.uk in case you can save your self a whole lot of hassle that way.