Coupling interference fit

OK; here goes --- seems a lot of different ways out there. Anyone care to post or share their stuff?

What shrink fit do you use for --- list them all!

3600 RPM, 100-250 HP - keyed

1800 RPM 500-1200 HP - Keyed

And so on:

Or, is rule of thumb the order of the day?

Or, does anyone use some other info or chart?

Cordially, Sam Pickens
Original Post
I would use the same standards that Falk uses for their products.

The only reason is that I learned how to install couplings at a Falk distributor and if you used their specs for shaft diameter and coupling bore you could mark any coupling on the top with a 275 deg. f grease pencil, heat it from the bottom and when the greases melted, slide it onto the shaft with ease.

If not, I would look it up in the Machinery's Handbook.
Thanks Danny,

And as you know metal above 275 starts changing its temper and/or metallurgy - especially bad for bearings. Even with the couplings, care must be taken to ensure even heat.

Please keep 'em coming everyone. This should be out in the open. I've seen a lot of stuff.

Myself; I've seen most everything over the last, well, over 40 years.

Worst one I saw was: 9 mils interference with two guys beating it on with 12# hammers. The coupling was about cherry red and the shaft was thickly painted with neversize - did I hear niv...vvvvvv....eeee-rrrrrrrrrrrrrrrrrrrrr. When I questioned this I was told, we always do these at 9 mils interference, we have to. I left it at that.

On the other hand: I came across an old MW installing a big project job. He had many couplings to install (as in ~50). They arrived at 2 mils shrink. He took them all to and exact, "you could slip them on by hand". He pain-stakingly sanded them to fit-for-fit to slip on. He said, I've never had one to fail or a key to shear like this. In his application I could belive that.

I went to do an alignment job - 200 HP AC mtr to compressor. Noticed the O/B brg and observed its setting - wasn't seated. I wouldn't beat it on. They called the motor shop. The guy came out and beat it on with an 8# hammer. I made mention to the company and complained but they let the guy go on. So, I aligned it and we started it - I got an RTA and took vibration readings. It was surprisingly smooth with almost no signs of the rough treatment. The only one I ever saw as such may I point out. I took many readings as I couldn't believe it.

We used a standard 0.0012" per inch of shaft diameter for all shrink fit couplings in the motor shops that I grew up in. I don't know where this number came from, but it was what the boss required in my first shop and it always seemed to work. We always installed the couplings by measuring the bore as we heated the coupling to insure there was adequate clearance to slide the coupling on the shaft by hand and not by hammer. We did use a little never sieze and as I remember we heated the couplings to around 300 degrees F to get the clearance we wanted to allow the coupling to slide on. I did it that way for about 30 years and never had a problem.

Since you're telling stories, I once watched a couple of mechanics install a coupling on a 1500 HP motor by running into it with the back of a fork truck. Eeker
A delta of about 170 def F above ambient will give you about 1 mil/inch of change over nominal. So, if your bore is 5.000 inches and you heat it to roughly 250 deg F the bore will expand to 5.005. Compare that to the shaft and you will get a feel for how easy it will be to mount. In the case of the 3600 RPM application there might be some undersizing of the bore to allow for bore growth at speed; I doubt thi is the case for the 1800 RPM application. Check the bores in advance cold and you can determine what if any interference was incorporated into the design.

A couple of other considerations. A properly mounted coupling transmits torque through its fit, not through the key. In theory you should able to remove the key and transmit the torque. Before mounting, you should also measure over the top of the key when mounted to the shaft and the "inside" dimension of the hub bore to the keyway to make sure that when finally mounted there is a slight clearance over the top of the key. This is to avoid topping which is esentially a coupling hub held off center due to insufficient clearance over the key.

As others have pointed out 275 deg F is about an upper limit but most coupling manufacturers recommend to not heat with a torch, although this practice is common. Just keep the torch moving and heat slowly to avoid changing the metallurgy of the coupling. Check with the coupling manufacturer for more specific guidelines.

John from PA
I aim for the coupling hub to be removable by pullers without using heat. I found that the bore should be tight on the shaft by at least quarter of a thou but usually no more than half a thou for larger hubs. Size for size will work, but I prefer that there is slight interference. I consider the bore to be too tight if the hub won't go on easily at 100 deg centigrade.

I don't like to sand out bores, I prefer to set up the hub in a four jaw chuck and true the face and rim where the coupling spacer fits, then machine out the bore. If the lathe is in good nick, the bore will be round and parallel. Also it gives an opportunity to machine away the hammer marks from previous fits.

As far as running a machine without the key, this may be correct, but I won't be trying it.

Joe Mc Cormack
Joe, relative to your comment "As far as running a machine without the key, this may be correct, but I won't be trying it". I did not mean to infer you could or should do this. I did say "In theory you should able to remove the key and transmit the torque." That is true in that the fit of the coupling should carry the torque. If the fit is "loose" and the key carries the torque you can create essentially a rotating 1X vector caused by the torque now applied at a radius (the key) and a unique vibration situation.

John from PA
I wasn't critisising the idea. I suppose it is academic. I understand that bolted couplings are meant to transmit torque by the friction of the faces and not by the bolts, also, tapered bore, hydraulicaly mounted, coupling hubs transmit torque only by friction. Probably some of the tighter fits mentioned in the post could do the same, with the key removed, however, I prefer the fit to be less severe, but still interference.

Your comments on the fit of the key are crucial. I tend to follow the published standards for the fit of the key in the keyways.
Joe Mc Cormack
OK, hadn't thought of telescoping gauges... makes sense. But should I hesitate to use an inside mic at 300F? Or is there a technique I'm not thinking of.

I always say that if you can measure something, then you should, but I never thought of measuring the bore while heating.
I like John's approach, transmit the torque through friction with an interference fit.

It seems like one has a choice clearance or interference. If interference, wouldn't you want a non-working fit? This requires a fit sufficient to remain under rotation and thermal changes. A slight interference fit, may become a clearance fit under rotation.
And calipers or templet gage. There is also a gauge outmost to indicate internal opening. Gage internally measured by a gauge externally.

And yes; it is best to go with an engineered calculation.

I suspect there is a lot going on out there but mostly 'sit it on a fire brick and heat it with a troch'. One guy told me, "when it starts to turn slight brown, it's ready". And he said he was color blind. True story. I was at a shut-down in a coal fired boiler and at lunch he said he could tell red from green on the lights - jut knew which one 'go' was.

Generally, I've found brown to be associated with the gov-ment.

Add Reply