Normalizing - how far to cool

Knifemaker.ca

Dealer - Purveyor
My Google Fu must be suffering from the long shifts. I hope this is a simple question. In the past when I normalize, I hang the blade to air cool between cycles and leave it till it's cool enough to handle. Someone recently told me all I have to do is air cool it to black - and back in for the next cycle. Have I been wasting my time cooling the blade so far between cycles?
 
I only take it to black which somewhere around 900 degrees. By that time all of the new grain boundries have formed and not much can be gained by waiting until you can touch it.
 
Yep, you only have to change the iron phase that the steel is in to normalize it and something like 900° is actually lower than you need to go to achieve this. You're not trying to form martensite which will happen at a lower temperature and over a widerrange of temperature, maybe 200°, depending on the alloy.

Doug
 
My Google Fu must be suffering from the long shifts. I hope this is a simple question. In the past when I normalize, I hang the blade to air cool between cycles and leave it till it's cool enough to handle. Someone recently told me all I have to do is air cool it to black - and back in for the next cycle. Have I been wasting my time cooling the blade so far between cycles?

If you are looking to perform an actual total and proper normalizing you should probably cool it all the way. If you are merely cycling for grain refinement you need to at least cool to where the magnet will stick again. I use the magnet because it will cover all steels, in fact the magnet is more useful here than it is on the way up. Some steels, like L6 and O1 will have a suppressed tendency to form pearlite, so while a quick cool to black, 1,000F to 1,100F, will work just dandy for something like 1084, it won't accomplish much at all with L6 which will have to go to around 750F to make upper bainite instead. The magnet doesn't give rip what phase we call it all it knows is that another phase has been made from the austenite so that you can now make all new austenite grains from that new phase.

There are other things to "normalize" than phases and structures. Stored energy, while it cannot be seen, even with microscopes, needs to be evened out as well if you want your blades to stay straight, so full normalization includes full cooling, but always gently in air.
 
Yep. The magnet is a good call and full cooling to room temp., on the stresses. I usually let it cool all the way down to see if there is any warping. If it does warp, I straighten and renormalize until it sets up straight.
 
I normally ( no pun intended) heat To just below critical and cool to room temp. To normalize ,to get steel like it was before forging just like it comes from factory nice and soft . Although I have purchased steel that was supposely annealed and normalized , but I normalize anyway , just to be sure. the Magnet will tell you when temperature is critical for quenching as will color. This is how I do larger blades Bubba
 
Magnetic temp - OK - I'm confused - I thought steel went non-magnetic at in the 1420-1675f range regardless of the phase the steel was in - due to some arcane issue with Fe atom's electron's state above that temp. Am I wrong? Is magnetism based on steel phase?
 
Yes and no. Steel will become non magnetic just before it changes phases as it is heated and the change in magnetism occurs at the Curie point, A1 which is 1420° regardless of the steel alloy. That's why people are told not to heat the steel to just where it goes non-magnetic but to heat it a little brighter (hotter) than that. The point at which the austinite becomes stable, A3, varies with the alloy. Also the steel does not regain it's magnetism at that same point as it cools. My experience is that it will have to be a a dark heat first. You will find that most of those points shown on the IT diagrams actually depend on whether the steel is being heated or cooled.

Doug
 
Last edited:
Thanks - I plan to play with a magnet and careful temp control sometime this week - !**! it's nice to have an actual pyrometer in the new forge - to get a better feel of the heat-up and cool-down transition between non-magnetic and magnetic.

Doug...
Kevin...
Tia Goo...
Bubba-San...
Thanks guys - a little knowledge is dangerous and you are helping me be a little LESS dangerous!
 
Last edited:
I did an interesting study of the Currie point of several popular blade steels two years ago and presented it at the ABS Piney Woods hammer in as part of my “Advanced Heat Treating Concepts” lecture. The first thing one must remember about the Fe-Fe3C phase diagram is that it deals with conditions of equilibrium in pure iron-carbon systems, and virtually no modern steel is a pure iron-carbon system.

The Currie point is the temperature at which iron gains or loses ferromagnetism at 1414F. This is due to effects of the electron spin that is beyond the scope of metallurgy and into physics, suffice it to say that this change in the electron spin accompanies the atomic shift from a body centered cubic unit cell to a face centered unit cell in the stacking. But the rate of this allotropic shift, as is with any atomic movement in the matrix, is affected by other atoms present, which takes us back to the Fe-Fe3C diagram…

There are four main arrest points on the Fe-Fe3C diagram – A1, A2, A3, and Acm. A1 (1335F) is the point at which the atoms will have enough energy to begin the shift from bcc to fcc in any steel, regardless of the carbon content. A2 is the Currie point. A3 is the point at which the phase shift is complete in steels with less than .8% carbon and you will have all austenite with no leftover ferrite. Acm is the point at which steels with more than .8% carbon will have full solution of austenite with no leftover carbon (cementite).

Now is where things get complicated- there are two sets of arrest temperatures because the temperatures are different for cooling than they are in heating. So when the temperature is from heating a “c” is added to the arrest point designation- Ac1, Ac2, Ac3 and Accm, and when the temperature is upon cooling an “r” is added to the arrest point designation- Ar1, Ar2, Ar3 and Arcm. Don’t ask why these letters because it is a long story why it is from French words.

Since the diagram is for equilibrium conditions it is heavily skewed by rate of cooling or heating. Due to hysteresis in diffusion not only will the temperatures all be lower on cooling, they can be VERY much lower due to the rate of cooling. Under normal circumstances there will be around 300 degrees F difference between Ac2 and Ar2 but if rate of cooling outpaces rate of diffusion (hysteresis) enough the difference can be around 1000F when the return of ferromagnetism coincides with the martensitic phase at 400F and below.

Now back to my study- My testing actually revealed a small range between 1414F and around 1428F based upon alloying and rate of heating in the typical gradual heat for hardening in a group including 1084, O-1, 52100, 1095, 5160, 1075, and L6. So due to heating methods and thermal transfer rates it is still safe to call 1414F “close enough”. But cooling can be accomplished so much more easily that all the Ar numbers are significantly different. Under normal circumstances one can expect Ar1 to be at least 300F lower than Ac1 with air cooling, but with alloying that can suppress the transformation it can go many hundreds of degrees lower. This is why merely cooling L6 to black will only get you larger grains as you are simply reheating the same austenite over again, but cooling 1084 to black will work out just dandy.

This is just one more example of how due to even the slightest alloying there can be no “one size fits all” heat treatment, and because of alloying the major portion of successful contemporary knifemaking is knowing your steel and how it works on the inside.

Michael, I am not sure if a pyrometer will make your life any easier, I just spend half of last week calibrating all of my thermocouples, three controllers and two pyrometers. After 25 tests at 5 different temperature ranges I finally had them all reading less than half a degree from each other, only to have one of my ornery portable units start to read 4 degrees higher when I commenced with the next heat treatment.:sad:
 
I think another problem with just cooling to a black heat is that the color/temp., can vary a lot just due to the ambient light conditions. It's lower in the dark and higher in direct sunlight. It's another good reason to use a magnet with any steel. Just to play it safe, I usually check it with a magnet and then give it a minute or two more to make sure the shift is complete,... I only do this occasionally, when I'm not too concerned with warpage, (on shorter/thicker blades), or if time is of the essence as in doing a demo or teaching a class.

Otherwise, I think cooling to room temp., is the safest bet and least problematic.
 
Last edited:
.....Now is where things get complicated-

:s12137:

Sorry - couldn't resist. Kevin, you know how much I respect your advice and I think you know I'm not knocking it - but it already had my simple undergrad head spinning long before this. :)

What I'm reading here - from all of our KD expertise - is that black may not be enough for a whole lot of reasons. Having said that, the ability to hold it in my bare hand without screaming in pain may not be quite required either - especially with some steels like 1084. I have read this thread over and over. I'm thinking if I can hold the blade without morphine, it's good for the next cycle. The difference between too hot to handle with bare hands (maybe 300F?) and room temperature can be an hour. I think the advice here will help me save some time, but I'm trying not to read too much into it. ALL the advice is much appreciated.
 
I'm thinking that there should be no harm in letting it cool to black, then oil quenching when normalizing. At least I really hope not, as this is how I've been doing it.... :)
 
Sometimes blades will warp between a black heat and room temp., If they do that while air cooling chances are even greater that they will warp (worse) in the quench. If the blade stays straight through the normalizing cycles, the chances that it won't warp in the quench are higher. Of course, you can always straighten the blade out of the quench and/or out of the temper, so... either way should be O.K. I just think it's better to avoid warping and do whatever is possible or practical to have it come out straight. Every now and then a blade gets snapped trying to straighten it. It doesn't happen often, but when it does,... it ruins your day.

For me, part of the normalizing process involves getting the blade to set up straight, (down to room temp.).

If a blade warps in the normalizing cycle/cycles and you simply bend it straight pre-quench, the chances that it will warp in the quench cycle are higher. I think it's by far better to straighten it and re-normalize it until it sets up straight through the normalizing cycle/cycles.

I don't like warped and/or broken blades...
 
Last edited:
:s12137:

Sorry - couldn't resist. Kevin, you know how much I respect your advice and I think you know I'm not knocking it - but it already had my simple undergrad head spinning long before this. :)

What I'm reading here - from all of our KD expertise - is that black may not be enough for a whole lot of reasons. Having said that, the ability to hold it in my bare hand without screaming in pain may not be quite required either - especially with some steels like 1084. I have read this thread over and over. I'm thinking if I can hold the blade without morphine, it's good for the next cycle. The difference between too hot to handle with bare hands (maybe 300F?) and room temperature can be an hour. I think the advice here will help me save some time, but I'm trying not to read too much into it. ALL the advice is much appreciated.

The problem with simple answers is that they always come with a whole line of “excepts”, i.e. “just do it like this… except when you are using this steel, and except when you want to machine it, or except when you want to lamellar anneal it, but not if it is a hypereutectoid etc… etc….”

That being said, the simpler answer is that if you are just looking to refine grain wait until the magnet sticks again and then go back to the heat, if you are looking to homogenize everything, including any residual strain energy that could cause warping the cool until you can hold it in your hands. And while you are holding it in your hands have a look, as Tai advises, to see if any warping did occur, if so, fix it and normalize again.

Sorry about that, I just refuse to give “just do this and don’t ask why” answers as I feel they limit you more than help you. The reason so many flock to knifemaking is the challenge of accomplishing something that is not simple or easy, otherwise, well, it would not be much of an accomplishment. When I was six mastering tic tac toe was great but now I prefer the rewards of more substantial challenges, so the true complexity of this craft still excites me.

Now here is another complication that I didn’t add, but some others have mentioned, and it does complicate things to be sure. If you quench during normalizing, which would mean you couldn’t actually call it normalizing anymore, you must then be certain to go all the way to hand cold. By quenching you will be austenitic all the way down to warm. If you quench to just below black and then reheat you are just reheating the same austenite again, even with 10XX series. Think about it, the whole point of quenching is to avoid making pearlite, if you don’t make pearlite all you have is austenite until another phase is made, so you can wait on the possibility of upper bainite taking up the slack, or you can cool all the way to room temp and let martensite do the job. So the quench method needs to be all in or all out, otherwise you are in limbo. Not that there is anything wrong with using a quench cycle for faster refinement, I often use it to keep carbides where I want them and really get on the grain size, but it does complicate things.
 
I thought the part about the electron spin was especially interesting. I've always wondered about that. :)
 
Last edited:
Kevin - the pyrometer may not hit the exact degree F - but it's a monster step forward for me - my previous forge had huge temp variation w/in the forge - the new one is much more even. To give you a reference - up to now my temp control has consisted of setting a ceramics firing cone or two in the corner of the forge and seeing when they droop!

Today I took an hour or two to play with Currie points... grabbed a scrap that (from it's cross section) was a piece of some of my D2 stock. Heated gradually and presto - by 1420f it went non-magnetic...
cooled back to 1400f - magnetic
heated back to 1420f - non-magnetic (I bet you could do that all day)
but after heating over 1600f and soaking for 5 minutes then
cooled to 1400f - non-magnetic...
1000f - non-magnetic
900f - non-magnetic
800f - non-magnetic (at this point letting it cool at 2 or 3 seconds per degree F)
775f - a hint of magnetism
750f - magnetic
... so "cool to black" is nowhere close to the cool-down magnetic point!
 
Last edited:
Did you take note of what the temp., was went it looked "black" and what the ambient light condition was?

Out here in Tucson, in direct sunlight on a hot summer day, subjectively (depending what you call a "black heat") it can be as high as the Currie point.

Might be interesting to try the same thing with 1084...
 
Last edited:
Kevin - the pyrometer may not hit the exact degree F - but it's a monster step forward for me - my previous forge had huge temp variation w/in the forge - the new one is much more even. To give you a reference - up to now my temp control has consisted of setting a ceramics firing cone or two in the corner of the forge and seeing when they droop!...

I am certain that the pyrometer will open up a new world for you and was a sound investment. My whining goes along the same lines of how I was told that a computer would make my life so much easier :les:, For certain, my life is much more complicated but I need to carefully assess the areas that have become "easier":3:.
 
Back
Top