This crankshaft fractured with 186 flight hours on it. You can see the crack at 7:30 where the fracture goes through a very slow fatigue phase that may have occurred over decades (until about 8:30), then several shorter lived fatigue phases (perhaps only a few minutes or seconds seconds), until total failure at about 1:30.

This is the view inside the forward nose of the crank, showing the shear plane crack propagation in the area adjacent to the end of the safety shaft (at the bottom of the photo). The circle shows three small cracks perpendicular to the original crack, caused by tension as the last of the material separated from the rest of the crank. Note that these three little cracks originated at stress risers created by drill marks. Another take-away from this photo is the incomplete threads, clearly caused by the tap not being concentric with the drilled hole. On the other side of the hole, NO threads are visible at all!

The circle shows the "beach marks" of the slowly progressing crack. The "centers" of the arcs point to the origin, the outside surface. The brown coloration is likely from baked-on motor oil. You can also barely see the crack leads up to the corner of the keyway, where the crack probably originated due to a stress concentration, something keyways are notorious for, due to the stress concentrations in the thinner wall and more highly stressed surface. It is worth noting that the gear was not removed when the crank was Maganfluxed. Removing the gear first may have exposed the crack, if it were present at that time, although that's an unknown.

This is a good view of the failure zone, and the red mark shows how the safety shaft is not even in contact with the hole opposite of where the crack is, which means there's likely higher stress on the crack side of the hole, which unfortunately also aligns with the thinner shaft thickness at the keyway location.

Here you can see where the crack originated at the keyway, and how the crank gear and oil sealing surface held the crack tightly together until it propagated out into the bearing journal, where it was free to grow unchecked.

This side view shows how badly out of line the safety shaft was with the crank nose, and how the shaft could have increased stress on the crank nose on the side adjacent to the keyway. This is likely a contributor to the eventual crack propagation, and the off-center safety shaft didn't help. This crank was drilled and tapped by a local "crankshaft machine shop" in Minnesota. Apparently it was not done on a lathe, and with an off-center tap.

Morals of this story are to Magnaflux (and remove the gear first), inspect all work done by folks who do machine work (it's not their butt on the line), and take a step back and contemplate when something that's not quite right to consider the possible outcome.

Although not a factor in this failure, checking for proper and consistent main and connecting rod fillet radii using radius gauges and a powerful magnifying glass (or loupe) is also an absolute requirement, no matter where or who machined or polished your crankshaft.

This crank nose was supported by a Sport Performance Aviation "BTA" ("Better Than Air") front bearing, which was not a factor. Without the front bearing, it probably would have failed a LOT sooner, and in an even more catastrophic manner. At least the prop, spinner, etc were retained in this case. Patrick Hoyt was taking off in his 601 at the time, but got it on the ground successfully.

Fortunately, crank failures on CorvAircraft engines are becoming increasingly rare, with the advent of nitriding and front bearings. We've become a lot smarter in many ways, and Corvairs are racking up a lot of hours in the air without incident.

Return to www.Corvaircraft.org