"That flange wasn't broken off, it was perfectly in place when I removed it"I hate to tell ya this Blake, but that flange was broken off (like completely broken off). Just because the remnants were captured in the fracture doesn't change the fact that it was broken.
I am a bit surprised that centrifugal force didn't throw the chunks out of there, but if you assess the entire design you can see why. There's a bellville washer in between the nut and primary drive sprocket. That concave washer continues to apply force even if the sprocket migrates to the left. That force from the washer kept the chunks of sprocket flange clamped between the nut/washer and the end of the crank. Unfortunately, there was no longer any force being applied to the bearing inner race, or the balancer drive gear.
Dave's evaluation makes perfect sense, and the photos he provided give almost a perfect visual description of the problem.
This design is a land mine. Yes, it is probably a very rare failure, but it has the potential to kill someone. As I see it, there are a number of design problems, and when the planets are in alignment they can all come together to be disastrous.
First, the nut is LH thread. Yes, it has to be LH thread but the designer should be taking into consideration that more than likely someone will screw up and try to turn it the wrong way. That almost always occurs on disassembly and the offending mechanic more than likely will really reef on that bad-boy trying to get the thing loose. Then there's the potential for rattle-wrench application. Ruttly has been harping about this sprocket for a long time. We should all pay attention. Thanks Ruttly, you definitely have my attention now. I think cracking that sprocket set the groundwork for your BIG failure. Whoever designed this should be thinking about maintenance too. If you must use a LH nut, design the system so the internal threads on the cheap nut will fail before more important stuff (like the sprocket), or make the more important stuff more robust so they can handle the over-torque.
Second, the flange on the sprocket is thin. You can see that the designer did a good job on the filet on the flange (it's generous), but the flange bears on the bearing inner race at a point that looks to be beyond, or right at the edge of, the fillet. Now when you tighten the nut it causes the sprocket flange to bear hard on the inner race at a point beyond the edge of the filet. It applies a bending moment on the flange and a BIG stress concentration right at the edge of the fillet. I guess that's fine as long as you don't overtighten the nut.
Third, the designer is relying solely on friction between the balancer drive gear, the bearing inner race, and the crankshaft. I know there is a drive pin, but the pin is not an interference fit. It has a little clearance so that you can assemble and disassemble the gear from the crank. I guess that's fine as long as the nut doesn't come loose. But guess what, stuff comes loose sometimes. Ya gotta be thinkin bout that stuff.
Fourth, the drive pin (the round one) is not captured. That hole in the crank should be stepped so that in the event that the drive pin starts to creep it can't come all the way out and let all hell break loose. Your pin clearly came out, and that was the nail in the coffin.
All of that mumbo-jumbo about the crank moving right and left and bustin up the sprocket. Malarkey.
I previously said we shouldn't be so quick to judge the mechanic. Now that all the other bits of info are on display (I'm lovin the pics, thanks again), I don't know who's more at fault, the person who designed it, or the person who worked on it last.