Donor Gearboxes

[F5 Dave]

Chest photo.

[el chupacabra]

Here it is spinning like it should.

Time to design the rest of the engine.

[jato]

i'm not familiar at all with 3D printers - can the printed cassette plate be used as a pattern at a foundry? or can it be printed as one? if so i'm sure we could put together a wish list ...

[el chupacabra]

If I add the draft angles it could 100% be used as a pattern for a casting. I had only planned on lost foam casting but plastic patterns would be suitable for multiple castings.

[el chupacabra]

I've started chipping away at the design again and I am contemplating whether to integrate the lay shaft into the cassette plate or not and what bearing i should use on the clutch side, ball bearing or cylindrical roller. . .
I am thinking with the lay shaft I will have a single key way for each gear and it will be secured with a single nut on each. The shaft is looking fairly chunky, but I don't want it to flex. Any feedback on keeping the lay shaft separate or integrated would be welcome.

Also I'm thinking long and hard about how to install said shaft as it won't pass straight into the cases. The only way I can see it working is if the non clutch side of the shaft bearing is installed last, maybe I can put it in a small housing like this.

[jato]

first things that comes to mind are you could have a single shouldered (NJ?) roller bearing on the left side of the layshaft and a ball bearing captivated in a pretty little housing 1mm bigger than the driven gear - then it would be easy to install as a unit. any thoughts on crank and layshaft gear diameters?

[el chupacabra]

They are module 2 44 teeth on the crank and 36 on the layshaft. I can’t really do much about the diameters as I need the 2.5:1 ratio. The 44 tooth gear is almost the same diameter as the cranks. This means the layshaft will need to go in on an angle till it clears the casting around the crank. Otherwise I can enlarge the gears to 50/41 teeth but the whole engine gets a whole lot longer.

Maybe I could make that pretty little bearing housing threaded and locate in the cassette from the inside, making assembly easier avoiding a press fit when placing the cassette plate on.

I’m just waffling now but there are a few challenges to work through.

[jato]

a 34T layshaft gear is not a possibility? you could then use a 72mm ball bearing with no pretty housing needed? the 2.5:1 is set in stone?

[el chupacabra]

I could muck around with the sprockets a tooth or two to get the same result. The little housing is a good way of stopping the bearing spinning while making it easy to install. I could always make the shaft a slip fit instead. I’ve thought up a way to install the shaft which will involve off setting the gears on the cranks to the LHS of the motor as far as possible so I can place the shaft in before sliding it into the NJ cylindrical roller.

I’ll model it and share it. Are there any drawbacks of a slip fit? I’ll probably have to do the same with the output shaft from the triumph on the sprocket side.

[Grumph]

I wouldn't do a slip fit on the output side bearing. Very high loads there and it could easily turn to custard.

[jato]

while not ideal from a tech point of view as greg mentions a slip fit (output shaft into output bearing) is a practical neccessity for a cassette style gearbox and for a relatively low use thing they appear to survive ok. as always its the minor detail thats important - an output shaft of a big enough diameter, ground to give something like 5 to 8 microns clearance in the bearing bore, and potentially the presence of the thin o-ring on the outboard end of the inner race. some 500 gp bikes i've worked on had done hundreds of hard hours and the output shaft was still happy enough with this arrangement.

[Grumph]

Fair enough. My experience is coloured by large heavy posties - aka racing couches. That bearing is probably the highest loaded on a chain drive bike.

[husaberg]

Lateral location of the gearbox shafts shouldn't be a big issue as it anchored on the mineshaft with the side play contained with the clutch.
plus on the countershaft with the drive sprocket and in the cassette box itself also with the inner housing.
About a week ago i posted a pic of the Honda RS750 countershaft its a separate piece than than the countershaft which i believe is a result of its original part of the shaft drive set up on the XLV750.







Note it actually has a Chain primary unlike the XLV750R gear primary
that gear isnt a kickstart gear its the final drive...

All or nearly all motorcycle gearboxs are simple sliding fits
bearings are cheap...

[el chupacabra]

Interesting, I thought slip fits wouldn’t cause too much grief. Are all cassettes a slip fit?

There is always option B which would entail no cassette and splitting the cases like any mx bike horizontally, essentially two halves and then having a third piece which holds the 4 cranks.

This might be far simpler to assemble but I’m unsure about the strength differences. I’ll model it with a cassette first and see how it goes although I don’t exactly have interchangeable gears so the cassette probably won’t come out anyways.

[el chupacabra]

So I went to a 55/40 tooth crank to lay shaft and 28/87 to the clutch 2.26:1 For a top speed of about 260kmh @13,000rpm.

Here's the latest design update, the cassette is now a lot larger. I am using the NJ roller bearing on the clutch side, between this and a ball bearing the shaft should stay centered as all of the gears will be straight cut. I will have to remove a lot of meat from the crank gears to make them lighter, they can likely also be thinner being such a large diameter.