[brian_]

if you make it rigid then your alignment needs to be absolutely perfect and you need to 'clock' the shafts to each other first by putting a DTI on one shaft and rotating it around the other in several places to ensure it is both parallel and concentric. This is not an easy task if your coupling lives hidden inside a bellhousing and even harder if you don't have a shimming/adjustment system.

As soon as you release the clutch pedal, a traditional clutch setup is also a rigid coupling. The alignment of the coupler is reset every time you push and release the pedal, and that's okay because the engine and transmission are properly aligned, and in a traditional setup the short transmission input shaft is supported and aligned on the engine end by a pilot bushing or bearing.

 

[brian_]

Ah, didn't realise he was using a clutch too so yea, those little metal tabs on the periphery of the disc inside of the lining are the flexible coupling. In fact having a flex coupling and a disc would not be a good idea. Otoh, two separate items with average alignment ability joined by a rigid coupling ends in damage eventually.

Sorry that wasn't clear, I didn't mean that there's a clutch in the proposed coupling, only that a clutch obviously works and isn't compliant. Yes, the springs in the clutch disk allow torsional compliance, but shouldn't allow radial or angular compliance.

But speaking of how things move...
just as with a normal clutch setup, something needs to have a sliding fit and remain free to float axially to make assembly practical. That's normally a splined connection, and the transmission has a male splined shaft, so the coupler rationally should be clamped to the motor shaft and have a matching female spline to slip over the transmission input shaft, with a pilot bore for the transmission shaft end if that's what the original transmission installation does.

 

[brian_]

This is why an automatic transmission uses a flex plate and why a manual trans uses a flex coupling in the friction disc. These are done by design for a very good reason- automotive parts aren't good enough tolerance in the worst case situation out of the factory to reliably use a solid coupling. Solution: chuck a flex in.

But every production EV in which the motor shaft is not also the transmission input shaft couples the motor to the transmission input (which is exactly what we're talking about here) with a splined shaft connection and nothing flexible.

 

[brian_]

Just send a dimensioned drawing to Rexnord and have them size and configure it for you.
View attachment 127792

That looks like a really nice coupling line, although not fundamentally different from other rubber flex joints (or giubos).
The rubber is available in a range of hardness, but any of them will be torsionally much softer than required; maybe that's harmless.
It can handle some misalignment, but every rotation of the shaft is a cycle of distorting the rubber, by an amount dependent on misalignment. I don't know how big one of these things would need to be to withstand highway travel, but I didn't look at the specs in detail.

I'm inclined to like these just on a sentimental basis - my father once worked for Rexnord.

 

[brian_]

...
2) the coupling split will cause a stress concentration where it intersects the weld.

I would have welded up the coupling split on the gearbox half to stop movement at the weld. At least shim the gap tight and install some screws to limit flexing of that half...

This is an excellent observation. The original coupler is designed to clamp to each shaft, but in this modification the transmission side uses a spline instead, so the clamping action won't be used at all... and is completely incompatible with the rigid splined insert welded all the way around.

 

[brian_]

An industrial coupling is a professional solution for the very-common motor to gearbox application. Properly sized these would last a long time.
...
Decoupling motor interia from the gearbox is never a bad thing as long as resonances are avoided.

Sure, but no one is engineering this system of interacting inertial objects, so torsional compliance in the coupling could create resonant problems, rather than solve them... and no production EV designer finds a need to decouple the motor inertia from the transmission. At least the hysteresis of the rubber will tend to damp out problems, unlike the steel coil springs of a typical clutch hub.

 

[brian_]

... transmissions that use a pilot bearing can be a bit wobbly at the input shaft when detached from the accompanying engine.

sifawangiaEV: Can you confirm if your crankshaft has a pilot bearing or not?

I don't think that it has a pilot bearing. I attached a picture of the clutch disc in my starting post so you can see that too for confirmation.

It is hard to determine if your gearbox used a pilot bearing based on pictures of the clutch disc. Gotta have pics of those shafts as KiwiME said, short of us doing research into your car's design.

Rather than hard, it's impossible to tell if the transmission input shaft is supported by a pilot bearing (or bushing) in the crankshaft from looking at the clutch disk. Fortunately...

The other end of the coupler that's hidden fits the motor shaft and part that's welded fits the transmission shaft. Here are the pictures:
View attachment 127822

... this image shows that the transmission input shaft is splined to the end, without a plain section to insert into a pilot bearing. A pilot bearing is normally needed in a traditional layshaft transmission, in which the input shaft is very short, carrying only a gear which transfers the drive to the layshaft and a coupling to the mainshaft for the direct "gear" (1:1 ratio). This Mini Cooper is typical of transverse transaxles which are all-indirect, meaning that the input shaft is long, carrying the input-side gear for every ratio; since the shaft is supported at two widely-separated bearings, the clutch disk can be safely cantilevered on the end so no pilot bearing is needed or used.

A 2006 Mini Cooper should be the last year of the first generation. If the Wikipedia article is correct, that means that - because it is the Cooper S version - it has a Getrag G285 6-speed transaxle (which was also used in the Ford Focus ST170 and SVT). It is a typical "3-shaft" transverse manual transaxle. This is a photo of the internals of a racing version (with straight-cut gears and dog ring shifting), showing that each of the three shafts (one input, two layshafts each driving the diff's ring gear) is supported by bearings are each extreme end:

The bearings visible (on top as it sits in this photo) are in the end opposite the engine, supported in the main case which has been removed; the bearings at the engine end (the bottom as it sits on the bench here) are in the engine-side case which is visible. The input shaft is the nearest one (without any shift elements on it); the differential housing area is just visible on the far side).

Edit note: replaced image with view from more useful perspective