Press fitting steel into cast iron?

[loply]

Hi folks,

I have a 4" cast iron wheel which I'm using as a drive wheel for something. It's originally a castor and so is quite imprecise, it features a needle roller bearing which takes a 20mm shaft that is at least 5 degrees off axis. As it happens I want it to be fixed to the shaft too, so...

I hammered the needle bearing out and I'm left with a bore of about 30mm.

I want to bore this out a few more mm to ensure it is aligned properly, then press in a new steel 'hub' which I can bore out to accept a 20mm shaft again.

Because this is going to be a drive wheel and under a fair bit of force at times I'm concerned about ensuring this hub remains put.

My guess is to mechanically link them by drilling and tapping 4 holes around the interface (ie half the thread in the iron, half in the steel) but this will take a while to do and if I'm not accurate will contribute to the already bad balance (which I'll try to rectify later).

If anybody has any better suggestions I'd be interested to hear em 

[John Hill]

Your drilling and tapping will work and I think is a recognised procedure.  My suggestion however would be to make a snug fit and use a chemical bond.  Another way would be to heat the wheel and shrink it on.

[Bernd]

I would bore out the hole as large as I could, say 40 or 50 mm. Then turn a plug that matches the bore size. Then put the plug in the freezer for 24hrs. when ready to assemble heat the wheel and put the plug in the bore. Let it cool and you should have a shrink fit. You can still pin the the plug if you feel the need.

Bernd

[Lew_Merrick_PE]

loply -- There is a section in Machinery's Handbook about how to calculate the interference you need to carry a given radial (torsional) load.  It has been several years since I last ran such a calculation, but I believe it is couched as clamping load.  Basically, the clamping load X shaft radius is the allowable torque (in consistent units) you can place on the joint before it starts slipping.  Remember that you need a large safety margin if there is any impact to the applied torque.

Joseph Shigley's text Machine Design also has this information.

[RichardShute]

I would go with a shrink fit, not a pressed-in one, unless you have a damned big press. Iron and steel have near enough the same coefficient of expansion and it's a convenient to remember 10um/K/m or thereabouts. So if you want a 001" interference (ie 25um) on a 30mm diameter you need:

25 / (10 * 0.03) = 83.3 deg (K or C) temperature difference between the two parts.

The bush in the freezer isn't going to be enough on its own, you would need to cook the wheel somewhat and I'd probably want (a bit) more than 001". As Lew says, if you want to do teh calculation you can, but in an arm-waving way, at 002" interference, it ain't goin' anywhere if you did it right.

Richard

[kvom]

I'd use green Loctite, and then drill for a round key afterwards if necessary.

[AR1911]

Don't try to press it in, the cast iron will crack.
I relearned that once more last week. 

[bp]

A press fit at room temperature will probably crack the cast iron, the shrink fit will almost certainly crack or break the cast iron.  Surprisingly the easiest method is probably the best, turn a centre plug or bush, bore the cast iron to a bit over the plug size (details on the Loctite web page) likely to be something like +0.1mm.  Clean plug and bore thoroughly (read instructions!!) apply Loctite and push together.  Done.
cheers
bp