Beautiful Failure

Learning often hurts.  I learned a lot with this blade, and it was incredibly painful.

This seax was going to be a beauty.  19" long blade out of 1075.  I'd cut two fullers down each side.  The tapers were not quite what I wanted but were in the acceptable range.  I had a nice handle planned for it.  Then the trouble began...
After heat treat the blade still  had a little more up-sweep than I wanted.  Single edge blades curl downward when quenched in oil, up when quenched in water.  Katanas are forged straight and only get their characteristic curve when quenched in water.  I had forged this blade with a bit of a curve upwards, knowing it would curl down a bit.
The bigger problem was that it wasn't  uniformly hard on the edge.  Some spots were nice and hard, others were quite soft.  So I decided to heat treat again.  The blade curled down to exactly where I wanted it (the shape in the photo), but again it had weird splotchy hardening that didn't really make sense.  Back into the heat treat oven.
I let it soak for 7 minutes at temp this time, even though 1075 doesn't need it.  My theory was that was such a big mass of steel it wasn't completely getting up to temp.  Wrong - same result.  And now the blade had a slight downward curve-ok, but not ideal.  So one more time.  This time upped the temp 50 degrees and soaked again.  The results were better, but still not ideal.  And now it was definitively curved down.  I pondered...
My next theory was that the problem was me.  Simple carbon steel needs to go from it's critical temperature to below 900 degrees in less than a second.  Because this blade was so big, I think I was simply missing the window because it took me a second or more to get it our of the kiln and into the oil.  Because I was missing the window, pearlite (a softer form of steel) was starting to form in spots.  This would explain the splotchy pattern of hardness, which was very different from what I'd see if the blade were inadequately heated.  It didn't help that I was using McMaster-Carr Fastquench oil, which is an 11 second oil.  Simple carbon steel would harden better in a faster oil (5-6 second) or water.
With the ugly downward curve, and the questionable hardness, I decided I couldn't leave the blade as it was.  Besides, I needed to know!  So I rolled the dice and did a final heat treat.  I heated it to 1550 (50 over normal) set up so I could move from the kiln to the quench as fast as possible, and used 120 degree brine as a quench.  Brine is dangerous because it's so fast a quenchant it can crack blades.  But it's virtue is speed, and being mostly water it would cause the blade to curl upwards.  This would solve both my problems at the same time.
The result?  The picture above is the result.  Edge hardness was decent (I'm still not fast enough), and the blade curled upward to exactly where i had wanted it, but...

This is what makes bladesmiths cry.  I heard the ping of failure about a second and a half into the quench.  I actually heard 3 pings - each a big crack.

So, as painful as it was I learned a few things:
1)  I'm too slow with blades over 12" or so to use simple steels with my current setup.
2)  MC Fastquech is borderline too slow for simple steels.
3)  I hate water & brine as quenchants (kinda knew that already).

The orders went out a couple days ago for some 80CRV2 steel and Parks 50 quenchant.  Hopefully I'll be posting happy picture of a blade similar to this in a couple months (life gets in the way of smithing).