There are several factors involved here. If side of the blade makes pearlite at 1000F it will expand and pull the other side, if that other side is still ductile it may accommodate this, if it is not it will act as an anchor and pull the blade into a curve. The same is true for the martensitic conversion, but much more prone to the anchoring effect. You also have normal thermal contraction vs. expansion going on.
First of all, I want to say thanks, everyone, for your replies.
Now, I'm a bit confused by this statement, because due to what you just described, i would expect an *upwards* curve. That is, assuming the edge reaches Ms first (it is much thinner, thus cools more quickly), I would have guessed that it would expand, causing compression in the spine.
Well, now that i try to picture it in my mind, i can see that this compression/expansion phenomenon must be very complex...I can sort of picture (but not really put in to words) how varying rates of expansion, varying cross-sections, varying degrees of ductility, and varying thermal change can combine to cause all sorts of unintuitive phenomenon...
I guess I better just take your word for it, and try a spine-down quench!
Thanks so much, everyone who replied...gives me lots to think about.
Dustin