I thought of another solution that would be very cheap and within my capabilities (no machining, heat treatment etc).

The bearing width is 11mm. It goes inside a 22mm deep cavity inside the crankshaft and is kept in place there by a cover ring. See picture below:



I could fit an 11 mm spacer inside the crankshaft (an outer race of a worn pilot bearing for ex.) and then fit the new pilot bearing. This would bring it backwards enough so that the inner race would be on a not damaged portion of the input shaft, while the outer race would still be fully inside the crankshaft. The problem is that I could not fit the cover ring to retain the bearing from sliding back and out of the crankshaft. I could then use one or a combination of both the following two solutions:

1. Glue the outer race of the pilot bearing to the crankshaft. I see Loctite makes products specifically for this application, but I don't know if this alone would be strong enough (though I don't see that cover ring in the original set up retain the pilot bearing with a lot of strength).

2. Fit a carefully sized spacer on the input shaft between the pilot bearing and the bigger spline section to prevent it from sliding back.

A variant is that I could use the 8mm thick cover ring as a spacer inside the crankshaft and add a 3mm spacer between the cover ring and the inner race of the pilot bearing. This would allow the cover ring to compensate for any variation due to heat dilatation (its shape would allow it to work as a "spring") and keep a low pressure on the input shaft bearing, transmission side, while keeping the pilot bearing tightly in place.

This is a long explanation for something very simple. It may not look very "mechanically correct" but would prevent the input shaft from going through various thermic treatments that may well weaken it if not done properly.

What do you think?