Kohler CV18 Flywheel Magnet

If you can 'mate' the broken pieces together closely(snugly), such that there is no gap, then do that. Magnetism will have a small field that 'leaks' if you were to sprinkle iron filings on a piece of paper over the joint, but the majority of the magnetic field will flow through. Along that line, if there are small 'chunks' that keep the two pieces separated, I'd remove the chunks if that would get a better meeting of the pieces.
If you add a gap, you are changing the timing of the pulses, adding two smaller ones in place of one larger one. The 'leakage' pulse will be minimal in comparison, IMO.
I have glued together a broken ferrite rod AM radio antenna successfully, and that is the same type of material in flywheel magnets. It likely does not perform as well as a new, undamaged rod, but still gets the stations.
tom

'holic, If you remember 8th grade physics {Longmeadow Jr High back in 1961-2}, when you put two magnets together {the way where they like each other} and put them under a piece of paper sprinkled with iron filings, the magnetic field developed more or less 'flows'(?) from tip-to-tail right on through, and doesn't 'leak out' as it would when they are separated. The closer the bits fit together, the better the field is contained. You do get a bit of leak, but the high majority will be kept in check.
Take a piece of magnet, such as the flexy one on the refrigerator door seal, and break a piece off. Put the two pieces together and do the paper test. I'd bet my paycheck {$0} that you'll see what I describe. But it's your refrigerator and The Boss might not like you snipping bits of it off for experimentation.
tom

You're right. It was science class. I made it to being on TV on WBZ as our schools 'science whiz' back then. Of course, I doofed it up on camera in the first show. But it was a long ride to Boston to the studio, so I likely would not have been able to continue anyway. Did get to see Quabbin(sp?) reservoir though.
I think a magneto needs to have its magnets placed in exact positions to fire at the proper time. That is why a sheared flywheel key will cause kickback in a lot of cases. The magnets are passing by the 'gizzards' a bit sooner than is desired, so the generated spark is earlier than desired. The flywheel in most cases continues to rotate a bit relative to the crankshaft after impact that shears the key. When you try to crank it over, the flywheel is there just a little bit too soon as the extra rotation excites the windings too soon, and then collapses too soon. Magnet done get there and gone before its spozta as the flywheel is spun past its proper timing spot.
tom