Hardening Part 1

Should this thread be indexed?

  • No

    Votes: 0 0.0%
  • Needs editing or more information

    Votes: 0 0.0%

  • Total voters
    40

Kevin R. Cashen

Super Moderator
Hardening I
The Soak

As we covered in the initial thread Basics-of-heat-treating, heat treatment of steel is merely a matter of rearranging carbon in iron in order to achieve our desired effects by where it is placed and temperature is our tool in doing that. But temperatures partner is time, in fact while not entirely in proportion in the power of its effects (temperature is infinitely more powerful) time can often take the place of temperature. One can move carbon a given distance it a few seconds with a higher temperature or in many minutes with a lower temperature. In hardening, the goal is to move enough carbon into solution to saturate the iron and then trap it there for the maximum wheel chocking effect that it will have on the movement of iron atoms, thus greatly strengthening the steel.

To accomplish this, carbon will need to be freed from its bonded carbide form separating it from the iron. It will then need to be diffused through the iron to create the solid solution known as austenite. The breaking of the carbide bonds will require temperature but the even distribution of it throughout the iron will require time. Over this time at temperature more carbon atoms will be stripped from the carbide and diffused into the iron, making the carbide groupings smaller. With simple carbon steels like 1084 or 1095 the carbide is simple iron carbide (known as cementite) with bonds that are rather easy to break at 1335F and so temperatures and times are not as extreme. But other alloying elements form much tighter bonds and much harder carbides, requiring more time or temperature to free and move the carbon into position. So the more complex the steel is that you are using, the more time or temperature will need to be applied to achieve the appropriate solution for a successful hardening.

Carbon can determine how hard a steel can get but it cannot change how easily a quench can bring this hardness about, that is up to the alloying and conditions within the steel. As one moves up from .5% Carbon to the magic number of .8%, maximum hardness (which we will describe with the Rockwell scale) increases on a very steady climb from perhaps 55HRC to around 62HRC where it begins to level out at .60% carbon and the hardness begins to approach a point of diminishing returns from carbon content alone. At .8% the saturation level results in a maximum of 65 to 66 HRC. This hardness limit is the result of the austenite becoming fortified by too much carbon and resisting transformation into the hard stuff that we will want in the quench, a condition known as retained austenite. Thus it should be realized that any carbon over .8% need not, and indeed should not, be brought into solution, but kept in finely dispersed and extremely hard carbides . 1080 can use all of its carbon in the solution but 1095 should not, or it could actually loose hardness. This is why you will notice that the recommended soak temperature for 1080 is as high as 1500F but 1095 is limited to 1475F.

Alloying has an even more profound effect on the outcome by locking up carbon in the carbide form and then giving us nasty surprises should we overheat and release it. O-1 with all of its carbides will behave quite similarly to 1084 if not soaked properly to release its hidden potential, but will suddenly give miserable results if overheated to the point that all the carbon is unleashed, it may be easy to quench but in the hands of one ill prepared for this it will not be much better than 1084 but at three times the price. Thus are the intricacies of temperature and the importance of time in the soak.

How long to soak

We can easily obtain information on recommended soak temperatures for each steel, but soak times seem to be more elusive, this is because it is rather relative. The size and distribution of carbides before the soak will determine how long it will take to dissolve them and move the carbon into the space between them. Heavy and widely spaced carbides are going need time, while fine pearlite will be rather quick to go into solution. Regardless of the steel, what you did to it in normalizing and annealing will have the most effect on how long it needs to soak. While some steel, such as 1084 do not make a soak as critical, any steel can benefit from time at temperature just in assuring an even and thorough solution. You can soak too little but as long as you have a tight control on the temperature it takes quite a bit to soak too long. The prior grain boundaries are often defined by the undissolved material that rests within them, thus their size will not change much until that material is gone at what can be described as the grain coarsening temperature. So long as the soak is held below this temperature steel can be soaked for hours with no change in grain size, the long dreaded grain growth in soaking is only an imagined bogeyman so long as you have control of temperature. With more alloying there can be more time added, so 1084 may be good to go with only five minutes but O-1 may need twice as much if fully spheroidized.

In time considerations it must also be considered that the steel will require a given amount of time to simply assume its surrounding temperature. Steel heated in uncontrolled heat sources may come to critical temp in just minute because the surrounding temperature is hundreds of degrees hotter, but in a controlled oven or forge, held at precisely 1475F, the blade will be observed to assume the same temperature over many minutes. The soak time should be counted at the point that the blade reaches the proper temperature, so one can see how soaking can involve much more time in the heat than originally expected.

Problems in soaking

Just as the desired effects of carbon in solution are a matter of time at temperature so, unfortunately, are some undesirable issues that can arise. The two most common are often confused or lumped together when in fact they are quite different and even, at times, in opposition.

Scaling:

The most obvious sign of holding steel at high temperatures is the annoying layer of thin black scale that forms on the outside of the piece. This scale can be a pain to remove and leaves the surface underneath mottled, further complicating the cleaning and polishing process. Scale can also interfere with an even and consistent quench. This obnoxious scale is iron oxide and is the result of the 21% oxygen that is available in the air readily combining with the iron in the steel. Think of it as superheated rust, and indeed of rust could form as fast without the heat it would be thicker layers as well.

The most obvious way to avoid scale it is to limit the amount of exposure the steel has to the atmospheric oxygen; this is done in many ways ranging from coating the steel to eliminating the atmosphere from the oven. Coating the blade is accomplished by many of the readily available anti-scale compounds which are sprinkled onto, and melt over, the surface of the blade. A similar effect can be had by wrapping the blade in stainless steel foil to shield it from the atmosphere; a small piece of paper added to the envelope can further eliminate any residual oxygen.

Industrial methods of dealing with scale involve purging the atmosphere of the oven itself. This can be done by either replacing the air with an inert gas or by vacuum procedure. We will assume the latter is beyond the means of the average home knifemaker and focus instead on replacing the troublesome oxygen. The easiest way for the DIY heat treater to purge excess oxygen is to use a gas fired oven or forge and simply change the fuel to air mixture to a carburizing or reducing flame in which all excess oxygen is consumed. Kilns can be outfitted with purge systems that can pump argon or similar gasses into the heating chamber but the aforementioned gas fire adjustment is much more simple and economical.

Decarb:

By far the more insidious and ubiquitous problem from atmospheric effects on hot steel is decarburization. This is the actual loss of carbon from the steel to the atmosphere and while it is all but invisible to the naked eye when compared to scale it is a far greater threat to the finished blade because it is actually robbing you of the chemistry you so carefully chose for your steel. It is a very common mistake to confuse scale with decarb but, as previously mentioned, they are two entirely different issues, in fact there are times when scale can counter decarb by removing iron at an equal or greater rate than the carbon loss.

decarb400x.jpg

decarb400x2.jpg

The above images show a band of total decarb (ferrite) as the pure white line of grains just under the surface. The darker line of particles on the outer surface is oxide. The diffuse area inside the total decarb is a gradient with a lack of overall carbon. Notice the deeper, three dimensional penetration shown the corner piece, a condition worth remembering when dealing with a knife edge.

Decarb is a much more complicated process than the rather straight forward scaling and oxidizing of iron, and has many more atmospheric considerations that make the mere reduction of oxygen an over simplification. Things such as atmospheric moisture and the chemistry of the steel itself can have profound effects on the issue regardless of the oxygen levels. Of more concern than total decarburization is partial loss of carbon. With total decarburization one can immediately identify the problem and deal with it, but with partial decarburization the very edge of your knife may only be working with a percentage of the carbon that you so carefully chose your steel for, while still exhibiting signs of successful hardening. Perhaps the best way to deal with decarb is to be aware of the threat and plan for it in your process. Carbon is removed at the interface of the steel and the atmosphere and thus creates a decarburized skin that requires more time to penetrate deeper into your blade. This skin is typically between .003” and .005” and can most easily be dealt with by simply removing that amount during post heat treat cleanup. This is the most practical way to deal with it rather than the average knifemaker attempting to deal with the myriad of atmospheric considerations. Do be aware that this is a three dimensional problem so the effects will be amplified at outside corners and the edge itself.


Recommended reading on this topic: “Heat Treater’s Guide” by ASM, “Tool Steels Simplified” by Palmer and Luerssen (this book has the most detailed information regarding furnace atmospheres and the consequences)




For information on processing of some of the most common simple steels used by knifemakers I also have these pages at my website:

http://www.cashenblades.com/heattreatment.html you will find a description of the various heat treating processes, but at the bottom of the page you will see a columnar list of common steels, clicking on those steels will give a page with all the compiled information on the various operations and temperatures along with charts and other information. I hope this is helpful to anybody who can use it.[/i]


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.
 
Last edited:
Thanks Kevin for posting, As always, Im learning from every post you make.
Much Respect.

God Bless
Randy
 
Does it seem strange that I copy and save all the info I can find from Kevin? LOL I sorta feel like a stalker or something. What Im finding is this.. Even if I dont get it now, or cant totally understand the entirety of most of the info, Im slowly learning and the pieces are starting to fit together, in this great big puzzle!!

Thanks again Kevin for taking the time to share you knowledge.

Much Respect!
 
Ok, let me see if I understand you. If one puts too much carbon into solution in the austinite with hypereutectic steels then it may physically prevent the face centered cube iron crystal from converting to a body centered cube iron crystal with entrapped carbon to form martensite when rapidly cooled. Do I have that right?

I and others really appreciate the time and effort that you put into these articles. I hope you don't end up publishing you book here one article at a time. Again, thanks.

Doug
 
Ok, let me see if I understand you. If one puts too much carbon into solution in the austinite with hypereutectic steels then it may physically prevent the face centered cube iron crystal from converting to a body centered cube iron crystal with entrapped carbon to form martensite when rapidly cooled. Do I have that right?

I and others really appreciate the time and effort that you put into these articles. I hope you don't end up publishing you book here one article at a time. Again, thanks.

Doug


Excellent interpretation Doug, that is exactly it. The formation of martensite, i.e. the actual hardening of the steel, is a duffusionless transformation, in other words it does not involve the movement of carbon atoms like all the other changes we get in heat treating. In this case the carbon stays put and the surrounding iron based matrix must deform to accommodate it. That surrounding matrix is in a FCC (face centered cubic) atomic arrangement that cannot resume the natural room temp configuration of BCC (body centered cubic) due to carbon atoms blocking that action. Instead the only option left is a distorted BCT (body centered tetragonal) arrangement (here's a pic to help folks not familiar with all this technobabble):

bct.jpg



Since you can't just squeeze irons atomic configurations tighter (despite many bladesmiths fantasies to the contrary), they must slide past each other in order for this distortion to occur, this is why martensite formation is known as a "shear type" transformation. What will happen is an entire plain of atoms will tilt out of alignment with the others with a sliding action at the interface; this is referred to as the “habit plane” in the textbooks. This action, once initiated, will progress like a wave across the austenite field, leaving martensite in its wake, and it is incredibly fast, like speed of sound type fast. Rick Marchand and I were just discussing hearing steel “cry” or “scream”, when quenched WAY too fast, at my smelt last weekend due to this whole speed of sound thing.

Now, in order for all this tilting and shearing to occur unabated you need austenite that is weak enough to permit it. If that austenite is loaded with carbon or other alloying elements it is strengthened and fortified, thus every percentage more of carbon will lower the temperature at which martensite will begin to form and when it will be done. And once you are over .6%, a certain percentage of austenite will tend to remain regardless, and limit the maximum hardness to the 65-66HRC range. But if you are heat treating 1095 you have .15% more carbon than you need for maximum hardness and if you over heat it you will only succeed in helping the austenite resist transformation and thus working counter to your goals. I hate to think of all the poorly heat treated 1095 that has been unfairly blamed for the knifemakers mistakes. It is a good steel but its underserved reputation is a testament to the error of the one heat treat fits all approach.

It can be frustrating, because one can increase that depth of hardening of any steel by bumping up the heating temperature, but one will then lose the maximum hardness obtainable if too much carbon is added to the mix.
 
Last edited:
Thanks, Kevin. So here's my next question. I have a book of IT diagrams and Jominy end quench graphs that lists 1095 austinized at 1625° with a maximum as quenched hardness of HRc of 60. What would be the difference in the end product between a blade austinized at this temperature and another austinized at 1475° for an as quenched hardness of 65 HRc if both are then tempered to a hardness of 58 HRc? All other things being equal and the steel subjected to three 2 hour cycles at appropriate temperatures to achieve the end hardness.

Doug
 
Thanks, Kevin. So here's my next question. I have a book of IT diagrams and Jominy end quench graphs that lists 1095 austinized at 1625° with a maximum as quenched hardness of HRC of 60. What would be the difference in the end product between a blade austinized at this temperature and another austinized at 1475° for an as quenched hardness of 65 HRC if both are then tempered to a hardness of 58 HRC? All other things being equal and the steel subjected to three 2 hour cycles at appropriate temperatures to achieve the end hardness.

Doug

Edge holding would suffer in the overheated blade. Martensite is the strongest phase of steel (barring carbide within the phases), nothing beats it for stable cutting edges. Austenite is hands down the softest form of steel and is the antitheses of what you want for a cutting edge, it is also not stable and so is bound to give you “surprises” down the road. I have heard makers say that they didn’t care about the as-quench hardness so long as the final hardness is what they want. This is a disturbing, it is no different than a farmer saying that since his crops will only be 6” high after the combine or picker is done, he need not wait for them to grow any more than 6” before harvesting. Every speck of pearlite formed during a quench is a fraction less strength performance you can get from the blade due to martensite, and austenite is even worse, but with the added possibility of it not remaining austenite; at least pearlite can be trusted to behave the same over the long haul.

Austenite is at best metastable at temps below Ar1 (approx. 1300F), thus many things can cause to convert to other phases when you may not want it to. This conversion will result in changes in Rockwell, distortion or even cracking issues. One of the most often noted benefits of freezing steel at extreme subzero temperatures is dimensional stability, that is because it converts the austenite and eliminates the threat later on. The overheating stabilizes the austenite to the point that the temperature at which martensite is done forming is well below room temp., so it stands to reason that super cooling it is the way to get it by continuing the quenching action. This extra step is often unavoidable in stainless alloys that have austenite so stubborn that normal quenching will never get it all, but if it is necessary with 1075, 1080, 1084, 1095, 5160, 52100, W1, W2 or any of the simple alloys we often work with, all we are doing is propping up a poor initial heat treatment. If one sees any significant gain on Rockwell with these steels they need to get control over their soak temperatures before buying liquid nitrogen.
 
So basically, you don't trust multiple tempering cycles to convert the retained austinite to untempered martensite then temper the martensite? Or is it just not creating a problem that you will have to correct later?

Doug
 
So basically, you don't trust multiple tempering cycles to convert the retained austinite to untempered martensite then temper the martensite? Or is it just not creating a problem that you will have to correct later?

Doug

I file the whole thing under “insurance”. Nothing in this world is perfect and you won’t get 100% especially with steel, so even with everything done right you can expect as much as 6%-7% retained austenite even in simpler steels. To somebody wound as tight as I am this is maddening, so yes I do multiple tempering cycles as well as quench from the temper, just in case. But one should definitely not rely on these measures to save a steel that wasn’t hardened properly to begin with. Especially since the jury is still out on exactly what the effects and mechanisms are regarding austenite conversion during tempering; there are a lot of disagreements, or at least contradictions among the experts. I have my theories but still need to find access time to X-Ray diffraction equipment to prove them; I am applying increasing pressure on my son to make friends with somebody in the material science dept. at his university, but he has this silly idea that he is there for his studies:001_rolleyes:.
 
I file the whole thing under “insurance”. Nothing in this world is perfect and you won’t get 100% especially with steel, so even with everything done right you can expect as much as 6%-7% retained austenite even in simpler steels. To somebody wound as tight as I am this is maddening, so yes I do multiple tempering cycles as well as quench from the temper, just in case. But one should definitely not rely on these measures to save a steel that wasn’t hardened properly to begin with. Especially since the jury is still out on exactly what the effects and mechanisms are regarding austenite conversion during tempering; there are a lot of disagreements, or at least contradictions among the experts. I have my theories but still need to find access time to X-Ray diffraction equipment to prove them; I am applying increasing pressure on my son to make friends with somebody in the material science dept. at his university, but he has this silly idea that he is there for his studies:001_rolleyes:.

Kevin, you said something in this post (okay a few things : ) ) that I didn't really understand. You said you "quench from the temper." Are you quenching the blade after you pull it from the tempering oven as you do when you initially harden the blade?
 
Kevin, you said something in this post (okay a few things : ) ) that I didn't really understand. You said you "quench from the temper." Are you quenching the blade after you pull it from the tempering oven as you do when you initially harden the blade?

I would need to see the conversation you are quoting to know the exact context, but I believe I was answering a question dealing with quick cooling a blade when it comes out of the tempering oven. Many folks fear perceived dire consequences of swishing a 400F in water to get back to work. There really is no danger in this and while the effects are borderline to minimal there my be some benefit in conversion of certain phases. I wouldn't expect to see a noticeable change in knife behavior, by quenching from the temper (unless things went VERY wrong elsewhere) but if it can't hurt, it does allow you to handle the blade and get back to work sooner.
 
Sorry about that. I thought I quoted the post... I'll try it again. Here's your post that prompted my question. It's #9 in this thread.

Originally Posted by Kevin R. Cashen
I file the whole thing under “insurance”. Nothing in this world is perfect and you won’t get 100% especially with steel, so even with everything done right you can expect as much as 6%-7% retained austenite even in simpler steels. To somebody wound as tight as I am this is maddening, so yes I do multiple tempering cycles as well as quench from the temper, just in case. But one should definitely not rely on these measures to save a steel that wasn’t hardened properly to begin with. Especially since the jury is still out on exactly what the effects and mechanisms are regarding austenite conversion during tempering; there are a lot of disagreements, or at least contradictions among the experts. I have my theories but still need to find access time to X-Ray diffraction equipment to prove them; I am applying increasing pressure on my son to make friends with somebody in the material science dept. at his university, but he has this silly idea that he is there for his studies.

By the way, I greatly appreciate all the writing you've done on heat treating here on the forum and what you've done over on your website. I've been reading it in my spare time and it has really helped me understand more of what happens during the heat treating process and why certain steps are taken. I never really understood this whole idea of soaking a blade until I read your posts and all the "technobabble" just confused me more. Thanks a million for your explanations!

Peter
 
Last edited:
So what in the chemical make up of the steel dictates a range of temps that allow for correct hardening? Was it that the non magnetic temp is similar for the high carbon steels like O-1 and 1095; was it determined by trial and error at first?
 
That is a fairly complex question. Many elements are ferrite stabilizers. These elements, such as chromium, raise the temperature at which austenite begins to form from ferrite. This is one of the reasons stainless steels have austenizing temperatures higher than carbon and low alloy steels. Some elements lower the temperature at which austenite forms. Nickel and manganese can push the austenizing temperature and if enough is present, can make austenite stable below room temperature, though this takes a lot of Ni or Mn. Carbon is a potent austenite stabilizer in plain carbon and low alloy steels like 1095 and 5160. 1060 will require a higher temperature than 1080 to reach full hardness. 1080ish steels will harden below the temperature where it turns nonmagnetic. Nonmagnetic is really just a happy coincidence.
 
Kevin, the books that you suggested, how do I go about finding them? The library or is there an online section.
 
Last edited:
Kevin, the books that you suggested, how do I go about finding them? The library or is there an online section.

A library would be wonderful, I wasn’t sure of people actually went to libraries anymore. My favorite place on earth use to be the library of the State of Michigan, I spend entire days there, sometimes just enjoying the feeling of being surrounded by thousands of books. I would often sit there and say that this was something I was happy to pay taxes for! But it was one of the first things they dismantled when Michigan hit hard times. Now I just try to add to my own personal library, and in similar fashion I have books I have never read, it just feels good to be surrounded by them.

These are the sites I use most for my book acquisitions, (and, ironically enough, my deals are library discards):
http://www.abebooks.com/
http://www.alibris.com/
http://www.amazon.com/
http://www.oxbowbooks.com/

Becoming a member of ASM will get you their books a little cheaper but they are already so high end that for me it is not enough, so I get them used as well. I think I prefer used books. It is like the ethical choice to go to an animal shelter rather than a puppy mill, if you know what I mean:3:.
 
If it's ok, I'd add in ebay for a place to look for book bargains. I have a vague wish list and every now and then good ones will show up for very moderate prices. I agree, there's some stuff I'd prefer in nice condition, but if the price is right there can be gold between the covers of an overlooked mutt.

Happy New Year, Craig
 
Back
Top