Well, not having a beveled edge sounds like the main problem. Knife makers often go with a 30 to 40 degree included edge angle, but I'm not sure what the geometry was on the failed section. A 2" wide double edged blade that's 3/16" thick will only naturally have about a 10 degree angle, which is too thin to support hard impact against steel. You have the compression forces from the impact, plus the rest of the blade is rapidly decelerating, which creates a bending moment, putting the back edge in tension and the lead edge in longitudinal compression. The only plane that isn't having compressive forces is the side-to-side plane in the area that hasn't started biting into the steel target yet. These forces can buckle and shatter a thin edge in several ways, which is why tools that strike into steel are never thin. For very hard and repeated impacts against steel, with a tool made of HSS, angles less 70 degrees can fail.
As the sectional area gets wider, moving back from the edge, the forces are of course getting lower, thus producing lower pressures, and the surface angle can be shallower without risking a fracture. So a multi-surface bevel, or a convex bevel, is probably the optimal structural shape from a failure standpoint.
On the other side there is a long area with some noticeable damage. When I made the comment that it sounded like a burr had snapped off, I'd say that was fairly accurate. It almost looks like an "edge" still, at the bottom of the "damage".
I think we must be using two different definitions of the word 'edge' and 'burr'. Knifemakers refer to a burr as the incredibly thin piece of steel drug out past the intended edge during sharpening, which feels very sharp but is easily removed, and won't stand up to any serious use. A good sharpening job doesn't leave burrs. Removing a burr, which everyone does anyway, and having the edge of your blade shatter are very different things, not easily confused. The edge your refering to is no longer the carefully shaped cutting surface, but just the jagged, tortured line where the steel now ends. True, the word edge is still an accurate description in one sense, but not in another.
Flipping the sword so that the body is facing me, it becomes apparent on inspection that the blade isn't perfectly symmetrical. The way the blade is roughed out on the machine, the fuller is dead center. The finished blade has the fuller offcenter by .03 inch. This led to a "tighter" edge on the one side than on the other. The tighter side failed, and the other did not.
If getting the fuller 0.03 inches off center results in a breakable edge, you're probably producing something very different from what a medieval smith made.. Very likely having the stock canted 0.01 inch or so shifted your final fuller location that far, with around a 0.6 degree difference in the angle of the two edges. Nobody sharpens to that degree of precision, so I wouldn't worry about the slight change in that angle instead of working to get it into the range of what won't shatter in normal strikes against steel targets.
I could find no evidence of cracks or any significant damage to the body of the blade.
I'd have my doubts about its integrity behind the fracture zone, since that area was obviously stressed far beyond its rupture point.
I took the sword with me to Tinker's place last nite, so the two of us could go over it and compare notes. The sword as is, is still a very functional sword. A sword with some edge damage, but the damage is "inside" the cop, not really a factor in the function of the sword, as a sword.....
Well, the "new knurled edge" certainly might make a nice marketing angle. The only problem with it is that it will still function pretty well until his hits the next hard object.
Its also easy to repair the damage. Not to make light of the damage, as it is definite, but from a mechanical point of view the damage is slight. Looking at the bottom of the damage, the "new" edge could be as "tight" as .015 inch if that's what I intended to do.
Um, my intention would be to throw that one in a bog, or toss it on the pile of unpleasant learning experiences. It you fix the one edge, you'd need to fix the other, since they really don't vary by very much.
However, over the last three months, I've started doing edges on heavy swords like this, around .025 inch. This gives a lot more support to the edge and still makes for a fine cutting edge.
But you thought your "expendable edge" design was fine, too. All we're talking about is a couple extra machining passes to produce an edge with a bit more strength. It might be easier to take a close look at the geometry of some authentic sword edges, instead of exhausting all the possibilities by trial and error.
