Heat Treating your Blade
A bad knife is one that will dull with little use, or break with little pressure. A good knife is the inverse; it will stay sharp for a long time, and stay straight and true with a lot of stress. This is a large part of the difference between a Walmart knife and a custom handmade knife.
With different heating and cooling, steel changes physical properties. Your goal is to manipulate these properties to best fit the use of a knife. Much of this you have likely heard before, but it is worth it to get a good knowledge of the works.
The two Physical Properties
When dealing with knives, there are two desired properties. The first is Hardness/brittleness. A very hard knife is one that will not dull; the knife can be used a lot on a variety of materials and the cutting edge will be unaffected. However, a very hard knife can also snap easily. Basically, the atoms are formed in a rigid pattern; the atoms would rather break bonds completely than move slightly in relation to each other.
The second is Toughness/softness. A very tough knife can be used even for prying, and will not snap. The only problem is the cutting edge is too soft to stay sharp very long, and you'll end up with a piece of metal in a wedge shape. On the molecular level, the atoms would rather move slightly in relation to each other than break bonds completely, meaning the cutting edge microscopically folds over on itself.
Think of it as a sliding scale; one end you have hardness, the other you have toughness. Your goal is to get the right balance of the two; as hard as possible, to stay sharp for as long as possible, but not too hard that it breaks under the reasonable pressure you would expect a knife to be used in.
So now for the actual process. I'll go step-by step in the heat treating, and also provide a molecular explanation for what's really going on. If you understand the science behind it, it helps with trouble shooting when you run into problems at some point down the road. Normalizing I explained in the previous article, so I'll go right ahead with the quench.
Begin by heating up the blade in your forge, evenly, to critical temperature. Critical Temperature (or CT), is the point at which the steel no longer sticks to a magnet. Your blade will likely be longer than the hotspot in your forge, so it'll take a little bit of sliding back and forth. Take it out when it's a good orange, and test it with the magnet. If it doesn't stick, put back in the forge and let it get a little hotter before testing again. If it does stick, remember the original color and replace in the forge until it gains that color again (it'll lose too much heat during the testing).
When hot enough, remove quickly and smoothly from the forge, and plunge it in oil (quenching), tip first. You can use water, but water can cool the blade too quickly, and the thermal shock can crack the blade, at which point you have to throw it away and forge another. Using water gives you a 50/50 chance of it cracking. Motor oil smoke will be poisonous to breath in, so just use canola oil or vegetable oil. It should be plenty to fully submerge the blade, as well as enough room to stir it back and forth.
Now, be warned, there will be a lot of smoke, some popping, hot oil, and quite possibly flame. The flame is not a concern and the oil will not blow up; you'll just get it directly around, above, or on the blade. Just make sure there's nothing else that it will set on fire (such as you). Wear gloves and glasses. Stir back and forth (not side to side). The stirring keep the blade constantly cooling so it's not surrounded by already hot oil. If you go side to side, it will cool unevenly and will warp, which is a nightmare to fix. Keep stirring until the smoke stops, then cool the rest of the way in water.
So what does this do?
Before you apply any heat to the steel, the iron atoms in that steel are built in a cubic pattern, with an iron atom at each corner of the "cube" and one in the center. The carbon atoms are floating between these cubes. This formation is called Body Centered Cubic (BCC).
You then begin to heat up the steel in the forge, until the blade is an orange-red. As the steel gets to CT, the iron atoms switch formation from BCC into Face Centered Cubic (FCC), where there no longer is an atom at the center of the cube like in BCC, but now has an iron atom at the center of the face of every cube. Because there is more room between the atoms, and inside the cube itself, carbon atoms float from outside the cubes to the inside. Still with me?
You then take the blade out of the forge, and plunge it in oil. Because of the cooling, the steel turns from FCC back into the original BCC. If the steel were to air cool, the change would be slow enough that the carbon atoms slip outside the cubes as it cools. However, the oil shock cools it, switching cube formations almost immediately. There is no time for the carbon atoms to slip out and so they are trapped in a box too small for them. This puts a lot of stress across the steel, keeping the atoms rigid in relation to each other, which translates to hardness, and brittleness.
To test if the quench worked, get a file and try to mark the edge. If it bites in and marks the steel, there's a problem. Troubleshoot and retry. There's a variety of possibilities, such as uneven heating, not hot enough, the oil didn't cool fast enough, or the steel was of insufficient carbon content. If the quench worked, the file will simply skate over the steel and not bite. This shows that the steel has become harder than a file, which is pretty high itself on the hardness scale.
With the quench complete, it's time to temper it. The steel is as hard as possible, and will not dull for a long time of use, but it's way too brittle to be an effective blade; I've snapped a blade at this stage with my fingers. Tempering is the slight softening of the steel. Imagine the hardness-toughness scale. With quenching, you brought the mark all the way to the hard end, with tempering, you bring it slowly down to the center.
Tempering is quite simple; put the blade in the oven at around 400 F for three hours, then cool in water.
The exact temperature will vary according to the type of steel, but 375-400 is generally a good mark. In general, the higher the carbon, the higher the temperature. Google around a bit to find the best temperature for the steel you're using. For a file, do 400, for a spring of sorts, go 375.
The molecular action behind tempering is very simple. With the slight increase of temperature, the BCC begins to open into the FCC, releasing a few carbon atoms and easing up the pressure. This softens the steel and makes it tougher, but not so tough that it's too soft.
From time to time, during quenching you will warp a blade. If a blade is not straight, it's ugly and flawed, and must be fixed or thrown away.
Fixing is scary, I'll just say that outright. The time to fix is during tempering; before that, anything you might try will crack it. After that, and it'll just spring back to original shape. About half of the blades I've broken were from trying to fix a warp. The best method I've found, is to (very gently) put the bent blade between two straight pieces of steel, the same length as the blade, and use a C-clamp to sandwich them around the blade, just enough pressure to hold them together, yet not straighten it at all. Put it in the oven for tempering, then a half hour later, tighten the clamp just a little bit. Replace, keep tempering, and repeat every half hour, tightening every time. About an hour or so before the tempering will be finished, tighten fully so the blade is straight, and leave in there to finish tempering. If all goes well, when you take it out and cool, then take off the clamp, it should be straight, having taken a set from the heat.
With the normalizing, hardening, and tempering done, your blade is just about finished. Next will come a bit more grinding, with a lot of hand sanding, which I'll describe sometime in the next week.