I have found this body of knowledge from Randal Graham very helpful. If you know Randal, tell him I said thank you.
From a post by Randal Graham on the Swordforum's Bladesmith's Cafe
Just some opinion and observation, based on what I've done myself, seen done by others, and problem solving sessions with some of my friends. I use satanite as the base for all my clay-coat heat-treating, specifically to differentially harden blades. Sometimes I'll add small amounts of crushed and powdered soft firebrick to adjust the insulation ability... from about 1/4 to 1/3 rd volume. The powdered brick will hold heat longer and insulate more.
I forge to as close to a finished shape as possible and generally remove very little material to get it set-up for the clay and hardening cycle. I do the hardening with the edges no more than 1/16th thick.
I personally feel the most important thing in this shop, to make it go smoothly, is that normalizing has been done carefully and complete before going to hardening. I have found sloppy normalizing is one of the major causes of edge cracking and warping in the quench.
Second to normalizing is hardening temperatures... most every time I hear someone talk about difficulties with edges cracking in the quench, it's simply because the person is not taking their time and getting the blades WAY to hot at the hardening stage. Most who have troubles with water the most usually harden in oil and expect those temps to work arbitrarily for water quenching as well. It won't.
For example, when hardening 5160 in oil in my shop I look for a temp of around 1525-1550f... quench in water from this temp and the edges crack just about every time. Drop the temp to 1440-1460F with a decent soak time, like 2-5 minutes for example, and the same 5160 blade will get screaming full hard without cracking. I have found that this lower-temp/longer soak rule has worked for every steel I've used, and also in every case solved most all of my cracking/excessive warpage problems.
A big problem I had was sloppy post-quench practices... I tended to wait and fiddle and look at the blades too long after they came out of the water...and some will crack some time after the quench has been done, as stresses load up in the blade. To resolve this, I now go directly from the quench to the tempering tank...no fooling around, just put it there and leave it alone.
I use temps between 350F and 450F for most things. The first dunk in the tempering bath I consider a simple stress-relief bath, and nothing more. 1/2 hour min for small blades, anything over 12 inches I like an hour stress-relief soak. After that I'll do straightening if needed, and follow that with at least one more tempering soak, and most times I do three separate tempering cycles to make myself feel confident about the blade, more than anything else.
Sloppy grinds, lines, shapes, and burred edges will cause at least warpage and at most cracking. You cannot quench something in a fast quench like water with grinder burrs and really sharp corners all over the place and expect it to survive the shock. Knock off the edges, make sure everything is even and smooth, make sure ridgelines are straight and consistent. What comes out of the quench is directly connected to what went in.
I think water is a very good quench medium for some things, but not all. I try to harden most everything in water, because in my opinion it hardens more thoroughly and finer, and more consistently, but some blades are not conducive to water-hardening. Double-edged symmetrical blades are very difficult to harden in water straight enough to make it worthwhile...most times I'll go to oil when it's a longer dagger or double-edged sword. Rapiers and fencing blades are almost always a disaster in water.
In my opinion, if there are cracking problems or severe warpage in your blades when trying water quench, I personally think the problems are going to lie with the maker and techniques, not the steel. Is water really better than anything else? I think probably not. It most certainly is different however. To say " steel x cannot be hardened in water" is untrue and unproven. To use temps and techniques based on oil hardening in a water quench is also doomed to fail. I don't profess to be an expert in anything, or particularly great at anything either. I know little about metallurgy and as I get older I don't WANT to know any more. I simply try and try again, and test in practical ways.
The list of steels I have successfully and repeatedly hardened in water, and in blade form are:
most all 10XX
W1 and W2
L-6 / 15N20
D-2 / D-3
440C / 440V
All my full-size differentially-hardened katana have been in 10XX, W-1 and a couple in 1086M,and pattern-welded with various steels, some lately with 15N20 and 1095, with rare problems always tracked down to mistakes in temp or technique on my part. Consistency and technique are the two most important things to me, since I've gone to water almost exclusiv