Jump to content

Photo

Hub interference fit on a keyed shaft


  • Please log in to reply
8 replies to this topic

#1
WCR

WCR

    Western Canadian Rockwell

  • Members
  • 158 posts
  • Joined 01-April 14
  • Gender:Male
  • Location:Red Deer Alberta
Wasn't really sure where to post this but I figured it was something interesting to share.

It seems like in our industry most driveline components are splined, however there is the odd time you run into a keyed shaft especially on homemade things like drop boxes.

I've heard some different opinions on running an interference fit where the hub slides into the keyed shaft. Most people understand a key will hold a lot of torque and I agree, but I did run into a bit different falure the other day that i found fairly unusual

This is the main input shaft off a rotary screw compressor, it's driven by a 3406 CAT engine, in this application it's common practice to install the driven hub with an interference fit, the manufacturer in this example did not and basically let the key do all the work.



This is what the new installed shaft and hub look like
image.jpeg image.jpeg

And this is what happened to the old shaft, by the looks of it the driven side of the key way work hardend and eventually started to crack, the entire shaft was so brittle in the spot where the hub was mounted was all full of cracks. the weird part was the key it self was in perfect shape,

image.jpeg
image.jpeg image.jpeg image.jpeg
Rockwell - Axletech - Meritor, axles and parts for sale
http://westerncanadi...ca/index.html?m

#2
Cody Ford

Cody Ford
  • Members
  • 1,001 posts
  • Joined 05-September 06
  • Gender:Male
  • Location:Calgary/coutts
  • Interests:90' C350 AKA 4 door 1 ton bronco
    96' Ranger: kings, EMF parts, 800hp SBF
That's a fatigue failure which started at the root of the machined keyway (no work hardening going on there) as the keyway like you say isn't designed to take the full load on its own in that application.

All of the big crushers in the oilsands use a shrink disc coupling, which is essentially a n interference fit as well.
90' C350 AKA 4 door 1 ton Bronco
96' Ranger: kings, EMF parts, 800hp SBF

#3
Twister

Twister

    OWNER of EMF Rod Ends

  • Root Admin
  • 7,371 posts
  • Joined 03-August 06
  • Gender:Male

  Thats a great failure. I've actually had similar happen and it was because my key groove was a c hair bigger than the than the key stock and best I could figure was it acted like a jack hammer on the shaft. When the pump sees pressure the torque would drive the key to the one lip and on de accelertation/ slow down/whatever it would shift to the opposite side like a worn u-joint.  The hub holds the shaft together and gets the crap pounded out of it over a period of use. As it fractures the pieces have no where to go so the key just keeps hammering the parts until it makes it's way all the way around or until you see movement between the shaft and hub. All I did was make a new shaft and a tighter key slot.

 

  So what's the correct fit? .001 tolerance and use heat to enlarge the bore for assembly? Puller for removal? The other thing that causes failures is improper key slot depth. I can't remember but I think the key is supposed to be 75% or more into the shaft. Not sure if that's a strength thing or difficulty of broaching the key groove into the hub. Usually the broach comes with the appropriate shims for ultimate depth if a company uses a shaper / Shafer? they can go to deep. That one looks a little light on the shaft and deep on the hub I think.

 

 This is actually more commonly used than you would think. I work with this on hydraulic pumps all the time as well as pivot irrigation pumps, generators etc... I still haven't fully figured out the tapered shaft key way combo



#4
Twister

Twister

    OWNER of EMF Rod Ends

  • Root Admin
  • 7,371 posts
  • Joined 03-August 06
  • Gender:Male

There are three main types of keys: Sunk, Saddle, and Tangent, Round, and Spline keys.

 

Parallel keys are the most widely used. They have a square or rectangular cross-section. Square keys are used for smaller shafts and rectangular faced keys are used for shaft diameters over 6.5 in (170 mm) or when the wall thickness of the mating hub is an issue. Set screws often accompany parallel keys to lock the mating parts into place.[3] The keyway is a longitudinal slot in both the shaft and mating part.

113px-Passfedernut.jpg
 

I'll see what I can dig up the last time I ran into this though I remember researching it for hours before I found what I wanted

 

 



#5
Twister

Twister

    OWNER of EMF Rod Ends

  • Root Admin
  • 7,371 posts
  • Joined 03-August 06
  • Gender:Male

I could be way out to lunch but if it was me the key depth definitely would be what I would be looking at for cause of failure.  That and the fit of the key. When I get failures like this I take days going over them. People slam them back together blame it on whatever and carry on. Only for a failure a few years down the road. Probably why I drive my customers nuts they just want it running right away.  A few years ago we had a pump throwing shafts. We made bigger ones up to 1450 u-joints which also  sheared in half . I don't know much about water pumps but I asked if it was possible that there was a surge that caused the pump to suck itself together. I was told no. 4 drive shafts later and a split  3 foot diameter impeller form italy we pulled the pump apart and the key was machined wrong as well as a none locking  nut (that was supposed to be that had backed off) allowing the impellar to suck itself into the housing trying to stall out the 3 or 400 horse diesel. I coulda saved two months down time and a ton of money but the pump company said it wasn't possible, the owner said the same thing and so did my shaft guy.



#6
WCR

WCR

    Western Canadian Rockwell

  • Members
  • 158 posts
  • Joined 01-April 14
  • Gender:Male
  • Location:Red Deer Alberta
It was a pretty interesting failure, the interference is .02 if I remember correctly, heat the hub to install. I've never seen anything like it before, in the Nat gas industry we deal with a lot of keyed shafts. the only styles that I can remember seeing are the parallel and the woodruff (half moon)

The other thing that I've seen that was pretty interesting is setups where someone has forgotten to install the key but with the proper interference fit there hasn't been any movement between the hub and the shaft, goes to show how much load is removed from the key with the proper fit.
Rockwell - Axletech - Meritor, axles and parts for sale
http://westerncanadi...ca/index.html?m

#7
Twister

Twister

    OWNER of EMF Rod Ends

  • Root Admin
  • 7,371 posts
  • Joined 03-August 06
  • Gender:Male

a press fit usually is a press fit. As good as a weld no key needed. On a shaft the size in your picture .01 I think for a sweat fit.  Odd they would use a key and a sweat. Guess there must be a shit ton of torque.



#8
A-rod

A-rod
  • Members
  • 27 posts
  • Joined 27-February 12
  • Gender:Male
Is that a mycom?

#9
WCR

WCR

    Western Canadian Rockwell

  • Members
  • 158 posts
  • Joined 01-April 14
  • Gender:Male
  • Location:Red Deer Alberta
There is quite a bit of torque going through that one, it's on a large sullair natural gas compressor.
Rockwell - Axletech - Meritor, axles and parts for sale
http://westerncanadi...ca/index.html?m




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users