Not DIY at all really, but knowing the group's expertise in mechanical engineering.......
If you have a keyway, why would you also want an interference or transition fit on a pulley? The drive suppliers we use are offering gearboxes with m6 tolerance output shafts. This gives a tolerance that has an upper limit larger than the nominal bore size. This implies to me that a pulley fitted to this shaft will tend to have at least a possible interference on the shaft rather than being a free fit (because the standard hole tolerances start at the nominal size and go larger).
Am I missing something here - is this standard practice to maybe have to lump hammer a pulley on?
It can be. If ANY wobble exists it can wear things badly. The aim is to just about be able to assemble and disassemble, but otherwise have the parts essentially locked together.
A key can shear under a jerk..
Normal practice would be to heat the pulley and drift it into place carefully, or use a press.
Both. You'd want to choose the fit according to your needs, _then_ worry about whether there's a keyway to lock it against rotation. Whether there's going to be a keyway or not doesn't change the fit needed.
Typically you'd use a transitional fit. An interference fit would only be needed if there's some reason to lock the pulley with better precision, such as a need to avoid wobble with a high accuracy.
Tolerance isn't interference. Certainly don't just ask for a sloppy tolerance if you really do need a reliable interference fit.
_Possible_ interference.
You deal with this by first checking the comparable tolerance sizes for the pulley (maybe it doesn't go that small - read your Machinery's Handbook (and go get one if you haven't already - eBay's cheap enough). You can even take a statistical approach and calculate the likelihood of a non-fit actually happening [1].
If there is a risk of parts not fitting by hand pressure, then you deal with it. You choose to either check fit before assembly (bin the rare non-fitters), to play mix-and-match with paired parts (slow), hammer them in (the British Way) or hand-ream when needed (even slower). It's _essential_ to train your line operators to do the right thing. If throwing it away is the right thing (for the overall process), then don't force it to fit! You'll make a duff assembly that may have a short lifetime, or might not be field serviceable.
The engineering way is to correctly spec the tolerances on things that need to fit together. That's what this stuff is _for_, all those standard tables of acceptable limits. Those tolerances are well worked out - things do still fit together (i.e. tolerances are deliberately asymmetrical, so holes are always bigger than shafts), so long as everyone does what they ought to. (I don't have my MH to hand though, so I can't quote figures)
[1] This used to be a good technique, but as manufacturing gets better it doesn't actually work any more. Modern machining no longer gives an imprecise size with a Gaussian distribution, it now gives highly precise sizes that may be systematically inaccurate. If you had to reject 1 in a 1000 and that's really how rare it was, then you could live with it. These days though your SPC-using manufacturer probably makes parts that are highly repeatable, but they might _all_ be wrong!
Force fit on a lawnmower disc. Got loose, hit a stone, and hey presto. Disk spinning freely on crankshaft, and key sheared off. Better than a new crankshaft I suppose.
Production engineering you would slam the thing on so tight on it would never come apart.
It all depends on what its for.
Anyone who has had exposure to cars for long enough including such joys as non force fit wire wheels on splined hubs, and other wheel bearings and parts that ned all you can throw at them to torque up enough to stand high performance use, will tell you that the tighter the fit the better.
Thanks for comprehensive response. A bit of clarification - we are not talking about a mass production environment here, these are 1-off or maybe 6-off max per project. The two applications I am considering currently are both chain drives. The pulleys wouldn't need to be particularly accurately located on the shaft I wouldn't have thought as the chain would take up a few thou' of misalignment? The shaft sizes are 60mm and 200mm for the different applications and (obviously) carrying some serious torque.
You do not quote torque/hp but at these shaft sizes it`s obviously considerable.In addition to the torque you have to allow for the service factor in deciding what sort of fit you require.As you appear to be out your depth in this you should ask the transmission people who are supplying the sprockets for advice.At the very least you need to use taperlock bushes.A sliding fit relying on the key to hold it won`t last very long before the shaft,key and sprocket are worn and needing replaced. The reason the gearbox manufacturer works to that tolerance is that he has to tie his sizes down somewhere.The fit of the bore to his shaft is up to the assembler who will decide whether to specify a press fit or use one of the many products available to obtain the same result as a press fit. Remember also that a 200mm bore will not be a sliding fit on a 200mm shaft,in fact it will take a lot of heat or pressure to make it fit at all. Mark.
BTW, you should also be aware that the key is not usually there to drive the pulley but to locate it in a fixed orientation such as for ignition timing purposes on an engine crank pulley. The vast majority of the drive force in most applications is provided by friction between the pulley and the flange on the shaft it abuts that is applied by the nut or bolt holding the pulley on. However the key provides a useful backup for light use in case the nut comes undone. If there is no purpose in the pulley being in a fixed orientation there is no need for a key.
When very high drive loads are encountered, such as the axles on a car, then the shaft is usually splined. Providing drive force via an interference fit is very unusual and not sensible unless the axle is a very large proportion of the pulley diameter. No matter how tight the fit on a small axle it won't apply very much rotational torque and as soon as it slips once the system is buggered due to wear.
However sometimes keys or pins are indeed meant to provide drive force and in such cases are often sacrificial so they will break before something else more critical does.
Same as before- check MH for details, but a looser tolerance on a shaft still fits into the matching loose tolerance on a hole. AFAIR, You might need an M5 fit class rather than an M6 if there's appreciable side-load, but you didn't mention the diameter of the shaft anyway. M6 is the "usual" fit for motor shafts an inch or two diameter.
I don't think you'd need any better location here than a transitional fit and a key. If the shaft is appropriately big, then taperlocks are another answer, assuming that the sprocket supplier offers a suitable version. if both parts already have keyways though, I'd just use a key.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.