Normalizing

Should this thread be indexed?


  • Total voters
    43

Kevin R. Cashen

Super Moderator
What is the difference between normalizing and annealing? It is one of the most common questions we often see and while on the surface the two treatments appear quite similar in the scope of their purpose they are actually very different. In the thread on annealing we covered how the purpose of that treatment was to soften and relieve stress and the resulting internal structure may be fine or coarse with the carbon widely separated. In normalizing the intended goal is a homogenous and uniform internal structure regardless of its form, or even perhaps size. A successful normalization is not defined by level of softness, hardness or stress, but, more importantly, that it is the same throughout.


The most important heat treatment?

In its possible effects on the final outcome of a blade, the power of proper normalization cannot be overstated. Hardening and tempering may seem like the most important operations to the fledgling heat treater but their success and approach is profoundly determined by conditions established in normalization. Behaviors in grinding and machining also indicate the importance of normalizing in preparation for annealing and shaping operations.

Normalizing involves the heating of the steel to above recrystallization temperatures (around 1335F minimum), and beyond, to a full solution in order to create as uniform internal conditions as possible. We often use the term “critical temperature” but in normalizing it is important to know that there are actually three “critical temperatures” to be concerned with. The first is the previously mentioned 1335F which is when the carbon first begins to go into solution and the new crystalline structure begins to form. The next is the temperature at which steel with less than .8% carbon will have its extra iron entirely filled with carbon, and it varies according to carbon content, the lower the carbon the higher this temperature, but it will always be higher than 1335F. The third is the temperature at which all of the extra carbon in a steel with more than .80% will be dissolved into the iron; this time the inverse, the greater the carbon, the higher the temperature above 1335F. With steels that contain around .80% carbon the upper critical temperature is the same as the lower, so 1335F is it.

Actual, full normalization requires exceeding the upper critical temperature in order to dissolve everything evenly and then allow it to create uniform structures as it cools. To this end the main operative word is “evenly”, heating must be done as evenly as possible and cooling must be done as evenly as possible. Thus setting a blade on any surface while cooling can interfere with a proper normalization, as well as most attempts to insulate it, so overtly the most obvious difference from annealing to normalizing is the still air cooling rather than any attempt to retard it. In most simple steels this will result in a relatively fine pearlite but this is all right since the goal is not to soften but to normalize the internal structures. Industry would say that you cannot normalize air hardening steels, as technically the operation would be identical to hardening, however the two can be separated even with these alloys based on the desired effects and wherein the heat treating line-up we choose to put them. Even an air hardening steel has not been officially quench hardened if it has yet to be annealed.


The three main targets of normalizing

1. Grain size. Normalization doesn’t really care about what size things are so long as they are the same size, in steel the kiss of death for many important properties is inconsistency. Odd and non-uniform places in the structure will create points of higher energy which will profoundly affect how things go into solution on subsequent heats as well as how the steel will handle stress later on. A “stress riser” is a point of higher energy where the steel will reach its limits much quicker when loaded, the more uniform you can have things the more the steel, as a whole, can handle loads. So while fine grain size is good, a uniform grain size is critical in reaping the benefits of it being fine. If there are huge grains surrounded by fine ones this discrepancy will only increase on subsequent heating and create points of greater inconsistency and ultimately weakness. The even heating to full solution, and even cooling, helps ensure a uniform grain size.

2. Carbide size. While most knifemakers focus on grain size, too often we ignore the much more important factor of carbide size when it comes to edge quality. Large and unevenly distributed carbon levels can create very detrimental segregations and concentrations in the uniformity of the steel. Large carbides make very fine and stable edges almost impossible as they are very hard and tend to pull out of the surrounding steel during sharpening, leaving microscopic flat spot on what should have been a sharp edge. Some large carbide forming steels have gotten the reputation of “taking a terrible edge and holding it forever”, and this is exactly why. When all of this is considered the ability to create very fine and evenly distributed carbides is invaluable in creating a quality blade, and normalizing is the operation that makes that possible. New grains are formed in almost every other heat treating operation but only normalizing gives us the opportunity to put the carbides exactly where they need to be for those subsequent operations to work with. Carbide requires serious temperature to be manipulated and only normalizing provides that at the proper time in the knifemaking sequence to set things up for later success.

3. Internal energy, or what normal folks call stress, must be uniform in order to keep a blade from warping. This is accomplished by keeping the internal makeup of the blade as uniform as possible and all of the points covered in the first two areas should make it abundantly clear how critical proper normalization can be in keeping a blade straight now and in subsequent heat treatments.

treatment%20temperatures.jpg



How to do a full, traditional normalization

To perform what industry defines and a true normalization you need to exceed all the critical temperatures we have discussed and achieve total recrystallization and solution. To accomplish this, the target temperature is much higher that you would expect, from 1550F to 1700F, with the lower carbon contents requiring the higher end of that scale. But don’t be too shocked, this is actually a bit lower than proper forging temperatures for most of the simple steels we typically work with. Once again the real trick is to be certain that the heating to that temperature is as even as possible followed by a simple cooling in still air.

This is a traditional normalization that has been utilized throughout industry for years, and it seems simple enough, as indeed it is, but knifemakers often incorporate some deviations on this to accomplish some other goals. After the full normalization cycle you have all of your grains a nice and even size but now you may want to focus on making that size smaller, for this you can follow the initial normalization with some other heats that do not exactly fit the true definition of normalizing but will work hand in hand with it.

The first cycle can be followed with subsequent heats involving decreasing maximum temperatures before cooling. The next could fall in line with the appropriate hardening temperature for that steel to leave carbide untouched but reduce grain size. This in turn could be followed by an even lower heat to initiate yet another even finer grain set with no chance the grain enlargement; it is worth noting however that eventually there is a point of diminishing returns as the grain coarsening temperature drops in conjunction with size. Eventually you are on the lower end of the scale where temperatures conveniently resemble the spheroidizing treatments discussed in the thread on annealing.


Recommended reading on this topic: “Principles of Heat Treatment” by M.A. Grossman and E.C. Bain



If this post, and the resulting thread, has information that you feel is useful enough to be linked to in the sticky index at the top of the page you may indicate that by voting in the pole. If you feel the information, or resulting thread, is not helpful, productive, or positive enough for all visitors to benefit from, do feel free to indicate that with a “no” vote in the pole. Both opinions will be regarded with equal value in ensuring only the best information is highlighted in this forum. I have changed the rule regarding two pages to be considered for the sticky, and will participate with answering any questions to better see if the thread is of interest.
 
Last edited:
Thank you so much for posting this because grain reduction is what I have been confused about. If I read this correctly the first heat depends on the amount of carbon in the steel. I'm guessing that hypoeutectoids steels should be heated to ~1700, eutectoid steels should reach ~1625, & hypereutectoids should reach ~1550. The second heat should be the quench temp for a given steel. And the final heat should be somewhere between 1335 and the curie point, say 1375. Are these assumptions correct? I hope I am understanding this correctly. Thanks again.
 
You've been busy, Kevin, and it's much appreciated. A couple of questions. From my texts I have learned that the critical points shift slightly higher or lower than what is usually listed in relationship to whether the temperature in on the increase of decrease, respectively. What we usually see listed as A1, A3, and Acm are sort of an average between those points. How significant is that? Am I assuming that on the second and third heats to normalize that you heat to above lower critical? I have read posts where the maker says that he did a primary heat of 1500 degrees, a second heat of 1400 degrees, and the third heat of 1300 degrees. In a case like that wouldn't the final heat, if that was an accurate temperature, produce no grain refinement as it's not hot enough to trigger a phase change in the iron crystals.

Doug
 
Thank you so much for posting this because grain reduction is what I have been confused about. If I read this correctly the first heat depends on the amount of carbon in the steel. I'm guessing that hypoeutectoids steels should be heated to ~1700, eutectoid steels should reach ~1625, & hypereutectoids should reach ~1550. The second heat should be the quench temp for a given steel. And the final heat should be somewhere between 1335 and the curie point, say 1375. Are these assumptions correct? I hope I am understanding this correctly. Thanks again.

Yes Darrin, that would be a good general sequence. Do be aware that it can change a bit for each steel and what it is you hope to accomplish. For this writing I had to try to come up with a very general sequence that would work with most steels. The idea is to dissolve any large carbide structures down a fine size in hypereutectoids, and totally wipe the slate clean and evenly distribute carbon in steels having .8% C or less. It is critical to get the grain size inform or subsequent heating, even to make things finer, will result in uneven grain coarsening. It has also been observed that the finer the grain the lower can become their coarsening temperature at which they will grow. So while it is not essential to refine grain, if one is playing extreme grain reduction methods it may be helpful to reduce that heat in each sequence. Of course once you drop below Ac1 you will quickly start losing any recrystallization.
 
You've been busy, Kevin, and it's much appreciated. A couple of questions. From my texts I have learned that the critical points shift slightly higher or lower than what is usually listed in relationship to whether the temperature in on the increase of decrease, respectively. What we usually see listed as A1, A3, and Acm are sort of an average between those points. How significant is that? Am I assuming that on the second and third heats to normalize that you heat to above lower critical? I have read posts where the maker says that he did a primary heat of 1500 degrees, a second heat of 1400 degrees, and the third heat of 1300 degrees. In a case like that wouldn't the final heat, if that was an accurate temperature, produce no grain refinement as it's not hot enough to trigger a phase change in the iron crystals.

Doug

I also have a few other pieces wrote up already, but need folks to show that the ones that are here are what is needed before I post them.

Since I was asked to do “basic” heat treating chats here I have purposely avoided references to the iron/carbon equilibrium diagram. But the hysteresis effects that you mention are partly responsible for there being three versions of the three critical temperatures (yep folks, as Doug knows there are actually as many as seven or more critical temperature designations on heating and cooling), and the faster you heat, or cool, the greater will be the difference between Ac1, Ac3, Accm and Ar1, Ar3, Arcm. As great and handy as the Fe/Fe3C equilibrium, or phase, diagram can be, it is wholly inadequate in many real world heating and cooling operations. Its biggest drawback is that it assumes equilibrium conditions and heating or cooling is far from it, once again the faster the heating or cooling the farther you will get from the parameters of the diagram. Next, it is just iron and carbon, so the slightest bit of alloying and all bets are off concerning the iron/carbon phase diagram.

Actual normalizing involves exceeding Ac3, and Accm to some extent, but the other cycles than knifemakers will use, that they call “normalizing”, will fall between Ac1 and the two upper limits. In the strictest sense of the term, these other heat cycles are not actually normalizing but we all refer to them as that for convenience. Temperatures below Ac3 and Accm will do little to effect carbide size and distribution but will do wonders for grain refinement.

Your next question covers what many makers have asked me about regarding very low temperature cycles, at or below Ac1. Of course since the definition of Ac1 is basically the point at which recrystallization begins, you obviously should not get any full austenite grain formation, but there are steps leading up to full recrystallization that can effect things. In the heating process there will be recovery of any previous strain effects within the current crystalline system and this is what stress relieving is all about. It is well below Ac1 and does not alter grain size at all. Next there will be nucleation, where the pinpoint seeds, or embryonic grains will initiate at points of high energy in the existing grain boundaries, and this is the thing that is responsible for the effects at or just below Ac1. In this case the existing grains are still entirely intact but will have a new set of fresh grains initiating formation within. Since multiple phases could be involved it is often referred to as “duplexing”. A steel in such a condition on cooling may have fine pearlite or even upper bainite subgrains within a predominantly spheroidized pearlitic grain network. If it goes awry you get uneven growth again, but if all goes right you could halve the grain size in one heat.

But here I would like to give my usual reminder that most bladesmiths are entirely too preoccupied with grain size and are too often obsessed with the smallest possible size without even asking why. For impact toughness fine grains are great but there are some down sides as well. Each steel has its own comfortable natural range for grain size and it is possible to have overkill. Of much greater importance to edge quality and stability is carbide size and unfortunately it is often ignored in the tunnel vision we have about grain size. A steel with an ASTM grain size of 9 with very fine evenly distributed carbides is going to outperform one with a much finer grain size but out of control carbides.
 
Last edited:
Thanks, Kevin. It's always great to correspond with you. Again, as you have pointed out before, text books are frequently incomplete in their explanations, especially those that are basically Cliff Notes on metallurgy like Metallurgy Fundamentals and Steel Metallurgy for the Non-Metallurgist. They had statements about the shift in the critical points depending on whether the steel is heating or cooling but nothing about how the rate of temperature change effects it. I should have already realized it from what I've seen alloying do to IT diagrams but I didn't realize that it effected the iron/carbon phase diagram as much. I hope that I didn't try to make your explanations too complicated. I can see where you want to keep things simple enough to keep from losing the beginner but still give information to the more advanced worker.

Doug
 
Yes, thanks again Kevin. The more I learn, the more I realize how little I actually know. LOL To me, heat treating is by far, the most interesting part of knifemaking. Keep the info. coming.
 
Just thought of another question. When doing stock removal using annealed stock would grain reduction cycles be beneficial or would it be better to just do a stress relieve then go directly to Austenization?
 
Very nice reading :D .

I have red about putting the blade in different media after heated it, but i get much slover cooling in the oven standing on the back.

Seved
 
Thanks, Kevin. It's always great to correspond with you...


And Likewise, it is good to see others such as yourself helping out on several of the forums across the internet. As I mentioned, just chatting about these things helps me wrap my mind around them in new ways, and the great thing about open discussion is that topics, from basic to advanced, can be covered and allows folks to get exposure to all levels, then they can take what they want from it.
 
Just thought of another question. When doing stock removal using annealed stock would grain reduction cycles be beneficial or would it be better to just do a stress relieve then go directly to Austenization?

This is one of those prime areas where grain size can get attention at the expense of carbide condition. Grain size can be altered at any time in the process, but carbides need normalization. Grain size will be pretty darned uniform in stock directly from the mill. But carbide distribution, particularly in the simple hypereutectoid (.80% C and above) steels we work with, can be an outright mess from the mill. 1095 and 52100 are particularly prone to serious segregation that will lead to all kinds of issues ranging from edge destabilization to cracking. Good normalization will break this stuff up and make it more uniform. So I would say that unless you are messing with grain size just for the sake of grain size that there are ho huge reasons to get concerned about it, but for uniformity of carbide and other segregated structures normalizing is of benefit to the stock remover just as much as the forger. It is probably this type of effect in simple high carbon steels that led to many of the misconceptions and quirky claims being made about forging, but lose the hammer and the heats will still do their job.
 
Thanks Kevin. That gives rise to a couple more questions from a stock removal viewpoint:

Is there any disadvantage to just wrapping the steel in an HT foil package and leaving it there through the
whole process? The foil will slow cooling a little of course.

Would it be better to normalize and get the carbides set up before grinding the blade or wait until the
grinding stresses can be dealt with at the same time?
 
Thanks Kevin. That gives rise to a couple more questions from a stock removal viewpoint:

Is there any disadvantage to just wrapping the steel in an HT foil package and leaving it there through the
whole process? The foil will slow cooling a little of course.

Would it be better to normalize and get the carbides set up before grinding the blade or wait until the
grinding stresses can be dealt with at the same time?

The foil shouldn't give too many issues and may cut back on scaling an decarb. However it is for these reasons that I would do my normalizing before annealing or grinding, a simple flat bar of steel would be far easier to normalize than a blade and things would be in great condition for subsequent annealing. After grinding the best way to deal with any stress/strain issues from that operation would be a quick stress relieve. i am still debating on whether we need a separate "stress relieving" thread.
 
Very nice reading :D .

I have red about putting the blade in different media after heated it, but i get much slover cooling in the oven standing on the back.

Seved

Seved, this was my concern in posting too many of these threads at a time, especially with something as similar as normalizing and annealing. I was going to help you out an move your post to the annealing thread but wanted to make sure that you didn't actually intend it for the normalizing thread. In normalizing slow cooling in insulation, or even an oven should be avoided, but it your idea would be great for annealing, except with steels having more than .8%C.
 
.....i am still debating on whether we need a separate "stress relieving" thread.

The various subjects and pointers are very much appreciated. I couldn't help but come back to this comment and think it would be a helpful topic or subtopic to consider, at least for me.

Take care, Craig
 
Back
Top