tapping CIM axles?

[Oldbikerider]

Whilst not addressing the exact question of the OP, I'll add here that heat shrinking a pulley/hub/gear to a motor shaft will work, but it is tricky.

The heat from the hub will soak back into the motor shaft which can damage the rotor, and will certainly bleed oil from the front motor bearing if it is sintered bronze.

Also, heat shrinking is a one-shot process, you need to get it just right first time. If you don't get the hub fully seated down to the correct position before it starts to grip on the shaft, you then have to bash or press the hub into place which will risk bending the shaft. If it won't budge, you'll have to turn the hub off the pulley on a lathe, which is now difficult to do because you have the front motor end cap sandwiched in there, making steadying the shaft all but impossible.

I can warn anyone attempting to heat shrink a hub or gear on this way that when it grips and the hub starts to loose heat to the shaft and the shaft starts to warm up, the whole assembly locks up real quick.

A much easier way to do this job is to use a retaining compound such as Loctite 680. I have assembled thousands of gears on motor shafts with 680 and never had a failure. It will fill gaps up to a few thousandths of an inch IIRC. 680 works very well on "reactive" metals such as steel, but will still grip with aluminum.

Graham.

[Ether]

Some suggestions to address Graham's concerns for anyone who wants to try shrink fitting:

1) put the CIM in the freezer

2) make sure the mass of the piece you want to shrink onto the CIM shaft is less than the CIM rotor

3) place a heat sink (e.g. a simple slotted piece of steel plate) snugly around the base of the CIM shaft

4) If at first you don't succeed, use a press to push the CIM shaft out

[Nate Laverdure]

I was interested in how well the shrink-fit was able to transmit torque from the shaft to the hub, so I made a quick calculator to evaluate a sample design.

An aluminum hub with arbitrary length, shrunk over the 8 mm steel CIM output shaft, is able to transmit a maximum of 4.2 N*m for every 1 mm of hub length (or 78 ft*lbf every inch). It requires a total radial interference of .009 mm (or .0004 in). Assuming linear thermal expansion, it takes about 190 F of heating to achieve this radial displacement.

Greater amounts of shrinkage can be achieved, but they'll force you to cut into the safety factor you're using for the maximum circumferential stress in the hub.

[Oldbikerider]

Interesting Nate. There's nothing like some hard calculations to assess an engineering problem! Your numbers agree with calculations I've done in the past.

However the situation is more complex. With heat shrinking, not only do you need the half thou or so interference to achieve the required torque capacity, but you need some clearance to be able to get the hub over the shaft in the first place. So your 190F is not going to be sufficient. You'll probably need more than double that temperature, and then the heat soak problems into the motor start to be an issue. For a data point, my experience is that red heat is needed for a steel gear onto a room temperature shaft.

Ether's heat sink idea might help, but usually in this situation you're applying a bending load to the motor shaft in operation and you don't want excessive distance from the motor bearing to the hub. This is especially important when you're dealing with the relatively small motor shaft of 8mm on the CIM motor. So typically, there won't be room for a decent heat sink, or for that matter for a plate to facilitate removal of a poorly located hub by using a press.

Heating may also alter the heat treatment of the hub or gear, plus damage the surface finish.

Don't get me wrong, heat shrinking can be used, and is a standard production method, and works very well in tightly controlled production conditions. However, for school students, using often less than adequate workshop equipment, the Loctite 680 method is quicker, less messy, safer, less risky in terms of certainty of getting it right, probably stronger, doesn't risk over-stressing the hub, and probably cheaper.

But heating stuff up to red heat with a blow-torch is always a good way to have fun

Graham.