Generally speaking, the forging process adds a lot of internal stresses to steel. These stresses can cause problems during heat treat such as warps and cracks. Normalizing a blade removes much of this stress and reduces the chance of problems during HT. To normalize a blade just heat it to critical temp (when a magnet will no longer stick to blade) then allow it to cool all the way until you can handle. The norm here is to do that three times.
Our resident metalurgist has suggested that this also benefits stock removal blades due to stress in the steel from the manufacturing process.
Normalizing also evenly distributes carbides in the steel. This takes a higher heat than just non-magnetic, usually 1600 degrees or higher depending on the steel being used. Next, the steel is cooled to black and then you would start your thermal cycling, a series of reduced heats that is used to shrink the grain in the steel.
Both Chris and Gene describe it well. I'll throw in the way I describe it to students.....
Image you've just driven in a sub-compact car for 10 hours straight....... and think about how tight, twisted, and "stove up" you would be. The same happens to steel when forged. Then, you get to your destination, get out of the car, and stretch/relaxed those twisted up muscles....that's what happens to steel when normalized.
The only caution I would give you with not only normalizing, but with all operations with your steel...... DO NOT believe what you've read on the net about SOAKING! This is a serious pet-peeve of mine. I have no idea where people get off posting to soak this or that carbon/alloy steel for 10, 15, 20, or even 30 minutes! I don't often say this in relationship to steels and Bladesmithing, but to do so is WRONG, and because the misinformation on soaking has become so prolific, I take every chance I can to try to steer new makers away from falling for the misinformation. Do it, and all you've done is totally wreck a likely otherwise good piece of steel.
It all depends on what you are looking for in a definition, what knifemakers refer to as “normalizing” and what the steel industry has developed as the heat treatment known as normalizing. Knifemakers will refer to any number of heats, from 1800°F + down to less than 1400°F as “normalizing”, but it would be more accurate to describe most of these treatments as “thermal cycles” used for grain refinement, stress relief, or annealing effects. Proper normalizing is a homogenizing process that involves heating to full solution, which places it above most other heat treatments in temperature. To homogenize the internal structures and phases of the steel it is necessary to put both the ferrite (iron), and carbide into solution. This, in turn, will allow the total reorganization of the grain framework on recrystallization. So, both the carbide condition and the grains are completely reset in normalizing. This does not necessarily result in finer grain, but instead a more uniform grain size.
Normalizing is most often employed to restore normalcy to a piece of steel that has been heavily deformed, such as in forging operations. The process of forging results in an accumulation of strain defects within the crystalline lattice of the steel, but could also result in uneven carbide segregations due to the localized introduced energy of the hammer or the cooling between it and the anvil. These defects could lead to distortion later on, but the larger, segregated, carbides will also complicate later heat treatments and effect edge fineness and stability.
The lowest actual normalizing temperatures would be those for the simplest carbon steel, and even these begin at around 1550°F and go up from there. So, to normalize, you get it hot, but the real key is that the steel must be heated evenly and cooled evenly. And by definition the cooling must be in still air. So you heat the steel to a temperature, and time, which will give you full solution and then remove it from the heat and allow it to cool in still air, that is normalizing.
About soaking- soaking is STANDARD practice in every steel industry, even with carbon steels. I know my friend Ed is very passionate about this, and so I normally simply refrain from countering how he prefers for do things. But I feel his passion has gotten the best of him enough this time that an alternate viewpoint is necessary to avoid a great disservice to our community. I have been at this for close to 40 years, I have put more hours in looking at soaked steel in a laboratory than many makers have ever heating it, and I cannot say that I have seen anything, ever, to support the idea that soaking any steel will totally wreck it, or that a very standard, and proven, practice is misinformation.
Once again Ed is passionate about this issue and has probably had one of those days where he has lost patience with a practice he doesn't use in his shop. I believe Ed and I are on the same page that soaking is necessary in alloy steels, and I will concede that it is not necessary in simple carbon steels, but that is a far cry from totally wrecking it. Other than decarburization, there is nothing that could result from an extended soak in carbon steel that could not be quickly undone. I am a bladesmith, but I am also a man of science, so I would need some solid data to make me rethink all that I, and the rest of steel-working world, observe daily. I will be happy to provide micrographs of 10XX steels soaked everywhere from 1 minute out to 30 minutes, and explain, in detail what they show, as I have hard drives full of them. Misinformation is a tricky thing, and I found taking on what I see as misinformation even trickier, solid data is very important.
For what it is worth, normalizing can be so powerful a tool that with careful normalizing one can shorten and almost eliminate the need for soaking even in some alloy steels. In the end it is all just really cool.
Your correct Kevin! We are on the same page when it comes to soaking being a necessity on some alloy steels.
On the issue of soaking being a standard practice in every steel industry......that may be so, but that doesn't necessarily mean it should be a boilerplate standard practice in Bladesmithing.....so that's an area that we'll have to agree to partially disagree..... and remain friends. We've spoken on issues like this before.....and I've always admired/appreciated your ability to differ on subjects, without taking it personal.
I'd just finished responding to a frustrating email, where an individual asked for my help/advice........ who was soaking 52100 blades for "45-60 mins" when hardening, but not under controlled temps (he was doing it by eye) and couldn't figure out why the blades were coming out so poorly..... and it was an individual whom I'd previously tried to assist........so the soaking subject was like an oozing boil on my behind!
That being said, I don't know where it started, or why it's blown up so much, particularly among newer makers.....but again, soaking situations as I described above are all too common with newer makers these days. Maybe I am too passionate about it......maybe I should just let the dog lie, become a cynical old man.....and tell people to figure it out on their own..... but that's just not me.
For what it is worth, normalizing can be so powerful a tool that with careful normalizing one can shorten and almost eliminate the need for soaking even in some alloy steels.
I figured there was something going on there. I am now thoroughly tired of 52100, as I have stared at more of it through the microscope than any other steel trying to map all the possibilities. But what I can say is that more than 30 minutes, even with complete temperature control, is insane. It is possible to adjust that steel, with normalizing and similar treatments, to respond with surprisingly short hold times. My standard soak for alloy steels is 10 minutes, if I have to go to 15 or more something is wrong. 10XX series is a different matter and as you go higher in the range from 1400°F to 1500°F the shorter the time you want it at temperature, with no soak over 1475°F.
Once again, normalizing is actually a key factor in the soaking discussion. By preparing the carbide for easier solution you can greatly adjust the need for time at temperature. Forging a piece of steel is essentially multiple, massive normalizations followed, hopefully by another normalization, in this condition it is very receptive to full solution with very little time or effort. One of the things I have learned about 52100 is if you are going to heat treat it in a forge, it is best to prepare it with a forge; if you are going to use an oven it may be better to stick to the oven.