I've done a number of chisel grinds, don't go too thin if you haven't already ground them, I stop at 0.04" on the 1080 I'm using. When I HT I normalize three times at a reducing temp each time but don't let it cool particularly before starting the heat treat after the last step. I bring it up to temp, moving it around a bit in the forge a bit and judging by eye that all of it is evenly heated and then out and STRAIGHT down into the oil. I do an up and down motion, stabbing into the oil (staying under) to agitate at first, then change it to a more edge/spine agitation after a few seconds. Of course, this is also tailored to the fact that I use a compressed gas cylinder a couple feet tall with the top cut off as a quench tank. Up and down takes advantage of more of the oil than any other motion would.
I have yet to get a noticeable curl since I started keeping it to around 0.04" and doing a better job of getting a nice even heat. On the other hand, I'm also not doing particularly long blades and at least half had significant spines relative to the bevel, as much as 1/3 of the blade height. I have a feeling a longer blade with less of a spine might show increased curling than what I've been working with.
