Best way to positively lock a shrink fit gear

[BatmanMan]

Hi there,

I'm making a gear to go on the end of an engine that I am manufacturing. Before we start, no I cannot use a standard gear, or go and buy one. I need to shrink fit a gear onto a shaft. This gear drives a cam chain system for the engine. What is the best way to positively lock this gear considering that the gear and the shaft will have been case hardened. I originally had a key way in the shaft but I think the stress raisers will be too large. My first thoughts were once the gear has been shrunk into place, drill a small hole axially in the interface between the gear and the shaft and to press in a small pin to stop the two from twisting. My concerns with this are 1) If I drill say a 3mm hole for this pin and the case hardening is less than 1mm deep, will this discontinuity in the hardened layer screw with everything? 2) Drilling through hardened steel is a horrible experience, especially when the shaft and gear you are making are quoted at $1500.

Background info: I am making a V-Twin engine from two single cylinder engines. The gear drives a layshaft which drives the camchain on one cylinder as its direction needs reversing. The shaft I spoke of is not so much a shaft, it is more another gear. This gear has a shaft from one side for this other gear to fit onto. This mega gear then slides on a spline on the crankshaft.

 

[Tmoose]

Near the beginning of this millenium past some companies with reputations for innovation and motorsports competition success at the highest level had some reliability problems Driving valve trains with interference fits and hollow shafts. I suspect the problem involved under-estimating the load reversals, and the resulting microdeformation and micromotion eventually caused the drive sprocket to lose its grip on the cam shaft.

For my money, A key by itself, even a LARGE key, can not reliably transmit much power when the conditions are variable, like everything connected to an internal combustion engine. The joint must be locked by heavy-enough clamping forces provided by controlled interference fits or axial faces clamped by preloaded fasteners.

Here's an image of a cam sprocket failure that plagued at least 1201 owners of the FORD/Yamaha SHO V8.

http://i24.photobucket.com/albums/c44/pcross_53/Taurus SHO/spuncam1.jpg

Linked frm this webbsite

http://www.v8sho.com/SHO/96shohome.html

 

[gearcutter]

Can you post some drawings?

Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

http://www.aussieweb.com.au/email.aspx?id=1194181

 

[BatmanMan]

Course I can

http://www.freeimagehosting.net/image.php?e6570ffd09.jpg

In this picture I have the key way in it.

I've been doing a bit of thinking and I could possibly drill the hole for the pin before I harden the parts so I don't have that discontinuity. I might make the two parts a wee bit harder to align when shrink fitting, but I think it may be worth it.

 

[MikeHalloran]

Just eyeballing things, you don't have enough wall in either the gear or the hub to sustain a serious press/shrink fit.

Multiple "Dutch Pin"s, as you have described, might work. They are ordinarily made a little longer than the assembly and peened at both ends for retention.



Mike Halloran
Pembroke Pines, FL, USA

 

[desertfox]

Hi BatmanMan

If your concerned about the stress raisers in a keyway what about the stress raisers of drilling the holes?

desertfox

 

[BatmanMan]

Cheers for all your replies and help guys, I really appreciate it.

Desertfox, a key way will be square and will protrude into each part by about 2mm, a hole drilled has a much nicer shape for stress concentrations and will only protrude 1 or so millimetres into each part. Good question though

 

[williedawg]

keyways don't have to be "standard" depth; they can be much shallower and still provide the needed torque resistance, as the key is in direct shear.

A method I used one time to affix a chain sprocket that had a tapered hug to a hardened shaft that had an involute gear already on it, was to (firstly) measure where (2) dowels would be closest to the gear OD. Fortunately, N was an even number.
Then drilled/reamed these locations in the sprocket's tapered hub which had quite a bit undersized standard bore. Then bored out the tapered hub to fit the OD of the geared shaft.
This left ~ half-holes for the dowel pins. When all was assembled, it drew up tight and had no movement.

Reaming holes is much easier to control than cutting keyways for slight intereference fit.

This was on an output shaft of a hydrostatic transmission which saw a lot of reversals.

 

[williedawg]

hug = hub

also, just to clarify, the dowel pin centers were close to the OD.

 

[Tmoose]

How many HP and torques will this be transmitting? What is the added gear's bore and width ?

 

[BatmanMan]

I imagine the power will be around 70+ HP at 10,000 RPM. This is the gear for the cam chain drive.

The bore of the gear is 28mm and width is 6mm. Not much at all unfortunately...

 

[gearcutter]

I assume that that the gear will have to be timed relative to the shaft?

Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

http://www.aussieweb.com.au/email.aspx?id=1194181

 

[tensor47]

This is a test, do not reply.

 

[tensor47]

This is another test please do not reply.

this bit of logic refers to SPM trial.

This bit of logic refers to grinding profile

 

[BatmanMan]

Hey gearcutter,

The gear doesn't necessarily have to be timed to a certain position on the shaft. The spline on the shaft has 22 positions and the gear has 30 teeth so there should be around 22 combinations between the two to find the correct position. Plus the gear and shaft that it is driving has another set of combinations before the timing chain.

Cheers

 

[bigTomHanks]

Weld it on.

 

[BatmanMan]

If I weld it on I will potentially ruin any heat treatment that has been done and may cock up the axial alignment of it. I know a guy who welded gears together in his gearbox and it screwed it royally.

My own personal philosophy is to never weld a machined part.

 

[mfgenggear]

use combination of keyway & shrink fit, I done this for others.
it will not fail.

Mfgenggear

 

[unclesyd]

You might want to look at using a polygon type connection for your gear. It looks like it might be possible from your model
A properly developed polygon connection doesn't go any where under loads.

Look at the pictures in the banner on the Stoffel link.

http://www.stoffelpolygon.com/

http://www.generalpolygon.com/profiles.htm

 

[DesignerGuy16]

Have you tried a fine thread and keensert like tab to deform the threads to lock it into place? Seems like a simple way to go if they don't need them to be timed?

James Spisich
Design Engineer, CSWP