Trying to learn more about carbides....

[deker]

So, in the past years we've seen a lot of high grade steels bandied about as the new hot thing, and I've noticed one thing that many have in common is a good shot of carbide forming elements. So, in trying to continue to broaden my understanding of the science behind why certain things work well, I figured it would be a good topic for a post.

So, here's what I think I know so far, can some of you smarter folks confirm or deny any of this, and help me generally learn more about carbides, carbide formers, what's good, what's bad, and what's just savvy marketing?

1. Carbides are harder than martensite which means they help enhance edge holding ability
2. Elements that form carbides in steel: V, Cr, W, Nb
3. Cr carbides grow rapidly when overheated and add to the problem of embrittlement
4. Nb, V, and W carbides are small and thus contribute to good edge holding on a very fine (<5 micron) edges
5. Too much of a carbide former lessens available carbon for maternsite formation. How can you tell how much though?

Anybody else with questions on this, please pile on, but please let's try to stick with the science and metallurgy behind it. I think overall that we could generate a good corpus of information on this topic if we can get some input from folks who are in the know on how and why carbides work like they do.

-d

 

[Sam Salvati]

Roman's information on Carbides in steels where they are large and steels where they are small and the types of edges that go with each was very interesting. Where large carbides as with high chromium steels are good for a push cut, because they fracture in large (still microscopic) chunks, and the small carbides are good for a slicing cut because they have more structural integrity to support a finer edge.

 

[sunshadow]

vanadium carbides are good for grain size control since they form nucleation sites that pin grain boundaries.

-Page

 

[Kevin R. Cashen]

The subject of carbides is a tough one because while they are quite ubiquitous in alloy steel and have profound influences on what we do, there is a frustratingly small amount of texts devoted to it. One that has been around for a very long time that I recommend is &#8220;Alloying Elements in Steel&#8221; By Edgar C. Bain.

Carbides are just different from the general things about steel that we work with. Most of the things we work with in softening or hardening steel involves solid solutions not compounds. mete and others will point out that steel is not made of molecules, but carbides in the steel bond in different ways and that is why they so often require separate consideration.

The most common carbide we deal with is iron carbide (fe3c), or cementite. It is the tamest and easiest to deal with. Heating to Ac1 (around 1335F) will dissolve any free carbon into solution but carbides are not free carbon, they are carbon locked up in a bond with another element. This tighter bond does a few things. The first thing it does is make the carbide considerably harder than the surrounding solid solution, so while martensite is the hardest form of the steel matrix it is still nowhere near as hard as a carbide would be. Carbon has a greater affinity to certain elements than it has even with iron, and it is this that makes increasingly stronger carbides and carbide bonds. The greater the bond, the more energy it takes to break it, so if you want to free the carbon in a vanadium carbide you will have to heat the steel much hotter than if all you had was iron carbide. This is why steels with more alloys require different austenitizing temperatures. Since they are so blasted hard, they are also very brittle so it is best to keep carbides as small and as scattered as you can, if you let them bunch up or grow large they will inflict their brittleness on the steel itself.

Some common carbide formers in steel are chromium, molybdenum, tungsten, and vanadium there are others but these are the ones we encounter most often. Columbium (niobium) and titanium are very strong carbide formers but not seen as often in common alloy steels. Some carbide formers prefer doing their thing in the grain boundaries where they contribute to brittleness, while others like columbium from fine carbide within the grains, something to think about.

Since carbide formers will indeed hog up carbon they can be a problem but this is controllable with proper application of heat. This is one reason that proper normalizing uses higher temperatures, it breaks carbon free from the carbides so it can be used in the matrix at large instead of making large problematic carbides. The ideal condition would be from .6-.8% carbon in solution for maximum martensite hardness, with the remainder left in the finest and widely scattered carbides possible. This will give good overall metal strength with the abrasion resistance contributed by all those little carbides. That adds up to an edge that does not chip out or deforms, but laughs at the things that would wear it to cause dulling.

 

[Matthew Gregory]

Something that was only touched upon here is the importance of proper carbide distribution through out the steel. If the steel doesn't have a consistent (or homogenous) distribution of carbide (big stringers or clumps of it randomly spaced through a piece, or bands of it between), it dramatically weakens the matrix. Proper thermal cycling and heat treat process reduces or eliminates this issue.
I have a love-hate relationship with vanadium. I love what it does in steels -retarding grain growth being one, as Page pointed out- but it also forms wonderful wear-resistant carbides when used in quantity. Makes the stuff nasty to hand finish, but holds a fabulous edge.

 

[Troop]

Roman's information on Carbides in steels where they are large and steels where they are small and the types of edges that go with each was very interesting. Where large carbides as with high chromium steels are good for a push cut, because they fracture in large (still microscopic) chunks, and the small carbides are good for a slicing cut because they have more structural integrity to support a finer edge.

Sam, I think you got it backwards. I'm pretty sure He said that the larger carbide steels, like D2, were better for slicing because of the "saw tooth-like" edge, and that the finer carbide-containing steels, like 52100, were better for push cutting, because of the better edge stability.

 

[sunshadow]

The problem as I understand it with large carbides (carbides with a mean diameter of over 1.5 micron) is that they are brittle, and harder to abrade than the surrounding steel matrix (and not that well bonded to the surrounding steel) so when you try to sharpen a steel with large chunky carbides, the carbides will keep you from attaining a critically sharp(defined as <1 micron thick) edge until they fracture and pull out leaving a jagged pit that is several microns across and deep which will tear through soft things like meat and stringy things like rope, but will not technically be sharp.
Personally I like a simple steel with just a pinch of vanadium to form fine carbides (.75 micron and smaller) as for the difficulty hand finishing, I just use better abrasives. I use ceramic belts, and when it comes to lapping and sharpening I use diamond with a water/dish soap coolant

-Page

 

[Mike Krall]

The ideal condition would be from .6-.8% carbon in solution for maximum martensite hardness, with the remainder left in the finest and widely scattered carbides possible. This will give good overall metal strength with the abrasion resistance contributed by all those little carbides. That adds up to an edge that does not chip out or deforms, but laughs at the things that would wear it to cause dulling.

How does this relate to O1... say Carpenter O1 with approx 0.50% Cr and W with a touch of vanadium and 0.90% C ?

Mike

 

[Kevin R. Cashen]

Good question, as usual, Mike, and that is why I was careful to say “ideal”. Not all carbide forming elements that are in alloys are there for that purpose. Quite a few are added for other effects they can have either in carbide form or not, chromium is one of those. Chromium is not one of the superstars of carbide formation and so in most low alloy steels the main goal for its addition is deeper hardenability. Even in richer alloys it often is used for this purpose and eventually as a corrosion inhibitor, which is worth noting that it cannot effectively do if it is locked up in carbide form. When it does form carbides they tend to be large, chunky and unruly, it really is not my favorite element added to modern steels but is often a necessary evil.

Since the chromium carbides can be defeated with lesser heat than the others, the two to really look at for carbide in O1 is tungsten and vanadium. The vanadium is normally under .3% and this is something to look for in alloying to understand what it is being used for. When added in less than .3% amounts vanadium is not meant to add significant abrasion resistance, but to act as a grain refiner. “Grain refiner” is a bit of a misnomer as its very presence does not shrink grains, but rather stabilizes the grains from getting larger. Vanadium carbides gather in the grain boundaries and keep them from moving to accommodate grain growth, and since on needs to get well above 1900F to break vanadium carbides, it makes most of our working ranges safe from grain growth.

The tungsten, is added for some abrasion resistance but it effect on the steels ability to resist softening in heat is even more significant. I know I am not the only one who has noticed how blasted hot you have to temper O1 such as Ketos, which includes W, if you really nail the heat treatment and those W carbides are bought into play; surprisingly hotter than other steel for the equivalent HRC. Often one will see O1 with either V or W and not both due to the amount of carbon they my use up when forming carbide.

So this brings me back to my point, even with O1. The ideal would be to put just enough carbon into solution to obtain maximum martensite hardness (once you get it adding more carbon only makes the strain greater and increases brittleness) and use the leftovers in carbides that will either refine grain, improve another property or add to abrasion resistance. If you allow carbides to control your objectives you can rob the steel of the carbon it needs to do the primary functions of martensitic steel. This is why the addition of V above .3% without compensating with additions of carbon can result in drastic reductions in hardness. But not so much with something like chromium which can reside in much lesser carbide formation, or even with the ferrite, with less effort. Thus you can get steel like 5160 that has very little carbon to spare yet containing chromium that doesn’t have to rob the carbon if you don’t let it.

 

[S.Knowles]

Not all carbide forming elements that are in alloys are there for that purpose. Quite a few are added for other effects they can have either in carbide form or not, chromium is one of those. Chromium is not one of the superstars of carbide formation and so in most low alloy steels the main goal for its addition is deeper hardenability. Even in richer alloys it often is used for this purpose and eventually as a corrosion inhibitor, which is worth noting that it cannot effectively do if it is locked up in carbide form. When it does form carbides they tend to be large, chunky and unruly, it really is not my favorite element added to modern steels but is often a necessary evil.

Since the chromium carbides can be defeated with lesser heat than the others, the two to really look at for carbide in O1 is tungsten and vanadium. The vanadium is normally under .3% and this is something to look for in alloying to understand what it is being used for. When added in less than .3% amounts vanadium is not meant to add significant abrasion resistance, but to act as a grain refiner. “Grain refiner” is a bit of a misnomer as its very presence does not shrink grains, but rather stabilizes the grains from getting larger. Vanadium carbides gather in the grain boundaries and keep them from moving to accommodate grain growth, and since on needs to get well above 1900F to break vanadium carbides, it makes most of our working ranges safe from grain growth.

The tungsten, is added for some abrasion resistance but it effect on the steels ability to resist softening in heat is even more significant. I know I am not the only one who has noticed how blasted hot you have to temper O1 such as Ketos, which includes W, if you really nail the heat treatment and those W carbides are bought into play; surprisingly hotter than other steel for the equivalent HRC. Often one will see O1 with either V or W and not both due to the amount of carbon they my use up when forming carbide.

So this brings me back to my point, even with O1. The ideal would be to put just enough carbon into solution to obtain maximum martensite hardness (once you get it adding more carbon only makes the strain greater and increases brittleness) and use the leftovers in carbides that will either refine grain, improve another property or add to abrasion resistance. If you allow carbides to control your objectives you can rob the steel of the carbon it needs to do the primary functions of martensitic steel. This is why the addition of V above .3% without compensating with additions of carbon can result in drastic reductions in hardness. But not so much with something like chromium which can reside in much lesser carbide formation, or even with the ferrite, with less effort. Thus you can get steel like 5160 that has very little carbon to spare yet containing chromium that doesn’t have to rob the carbon if you don’t let it.

Kevin, forgive my ignorance... In the case of O1 and 5160 it would seem to me that in the amounts of chromium added it is surely not there for rust inhibiting, And in the case of O1 there are better carbide formers, and with the manganese levels what they are, I would deduct the Mang. is adding to to hardenability. Is the Chromium (for these 2 steels) what adds to their toughness?

Shawn

 

[sunshadow]

now here is where I'm in need of more info,
In superalloy we use solid solution precipitate phases to pin grain boundaries as carbides tend to have too low solution temperatures for our needs, and we choose operation temps based on melting/ formation temps of those phases. Does anyone know what the critical temperatures of the various carbides is? Vanadium carbides are stable to 1900, what about the other carbide formers?
Also, is there a table somewhere in which the relationship of carbon sequestering by the various carbide formers is discussed as it relates to net retained carbon available for martensite?

-Page

 

[Kevin R. Cashen]

Kevin, forgive my ignorance... In the case of O1 and 5160 it would seem to me that in the amounts of chromium added it is surely not there for rust inhibiting, And in the case of O1 there are better carbide formers, and with the manganese levels what they are, I would deduct the Mang. is adding to to hardenability. Is the Chromium (for these 2 steels) what adds to their toughness?

Shawn

Absolutely nothing to forgive, Shawn, the manganese thing was something that was overlooked and needed consideration. I am uncomfortable with anybody having any reservations about asking questions. No matter how many years I study the metallurgy of knifemaking, or teach classes on the craft, I still learn the most when dealing with questions from everybody from veterans to the guy making his first knife.

Indeed chrome would not be added to the steels you mention for any rust inhibiting considerations, since one really doesn’t see much effect in this area until you exceed around 12%, and these steels seem to rust every bit as quickly as any straight carbon steel.

Manganese started out in our steel supply as a desulfurizing addition, however due to its effects on the matrix and the inhibiting effect it has on diffusion it also gave the natural side effect of increasing hardenability. Since it is already there and it is more available it is the easiest way to increase hardenability but requires greater amounts than many more powerful substutional, like chromium. Due to how Mn interacts with other elements and effects the matrix, chromium can indeed increase hardenability while maintaining toughness, if left disengaged from carbide formation.

With many 10XX series steels having as much as .9% Mn to assist with hardenability and still remaining classified as “shallow hardening”, one can see that the modest .1% boost in Mn for O1 or 5160 (.01% at best fro 5160) offers very little hardenability over steels that can be called water hardening. So the much more powerful chromium brings the hardenability of these steels well into the oil hardening range, while simply relying on Mn alone something like O2 requires as much as 1.5% in order to achieve the same results, but with no benefits of carbide or the possibility of the same ferritic influences. The manganese may only be there because they start with a simple carbon steel base with it already there or it is added to at lest lend some reasonable hardenability should the heat treating consumer not have the savvy to manipulate the chromium and carbide interaction.