I just had an in-depth discussion on this very topic on another forum, I will see if I can pull some of the visuals from that chat to here.

There has always been a lot of confusion on this topic due to the interchangeable use of words. Bladesmiths are fascinated with “grains” in steel, this is not a bad thing but sometimes it can lead to focusing on the wrong issues. Steel has both a directional property to it as well as being made up of individual crystalline grains, these are two distinct and separate concepts.

We should be very careful to distinguish between the anisotropic flow lines of the steel (like the wood has grain) and anything having to do with the crystalline (e.g. austenite) grain, which is much less directional and can be altered with basic heat treatment. If the steel was made in the traditional melting, pouring and casting methods then it will be anisotropic in nature. When the steel ingot was cast the voids, segregations and inclusions were isotropic in nature, i.e. no direction differed much from any other. The rolling process mashed it all down and drew it out in the direction if rolling and made it anisotropic in nature. CPM and similar process can overcome anisotropy to some extent. The anisotropy will only really show itself heavily in certain properties. Impact toughness, or tensile strength are two that will be heavily affected and the condition can be visualized in how it shapes the alloy banding we often see.

These directional properties are best observed in extreme cold working, which is very anisotropic. Wrought iron is like an exaggerated caricature of steel with directional issues and when you cold-work it frays almost like a cable; if you score it and bend it like you are going to break it, it will look like string cheese in the inside surface. If you cold-work steel until it fails the lines of failure will be lengthwise, in the direction of working, to a lesser degree but similar to the wrought iron.

Here is a clearer example:

These are images of O-1 tool steel, and a good batch at that, but I have done some treatments to it to accentuate the flow lines and then polished and etched it to reveal them. The image on the left is an overall picture of how the material consists of lines of flow that run in the direction of the mill rolling operation. This “grain” direction has nothing to do with the actual grains that make up the phases of the steel, which is illustrated by the image on the right. The image on the right is the same steel at over 400X where you can see that within the larger flow lines are the phases (martensite, carbide, ferrite, austenite, etc…) made of what the actual crystalline grains of the steel. The crystalline grains are more isotropic, in that with heat treatments we can remake them and alter their orientations. But you can also see how the larger anisotropic flow lines of the steel can have an influence on the directional properties of the phase grains as well. This is the bases for that obnoxious alloy banding we often see.

Now for the other “grains” let’s go back to cold working. Metal is made up of individual grains in which the atoms all line up in the same crystalline orientation. These are, for the most part isotropic (not directional in nature), but if we cold-work the metal the plains of atoms will tilt and slip as they deform in the direction of working and are elongated and drawn out to become more anisotropic (directional). When the metal is reheated recrystallization will occur and all of the effects are wiped out as an entirely new set of grains are formed in a new isotropic condition.

When we forge, rather than cold-work, we are above the recrystallization temperature and so the directional properties are not imparted due to what is known as dynamic recrystallization, in which the strain energy of forging causes new grains to form and refresh things. And even if things did get anisotropic, it would all be wiped out on the next heat, be it forging, normalizing, hardening etc… since most of our heat treatments involve recrystallization.

I have heard folks try to sell their forged blades on the idea that they are somehow aligning the grains with the edge, but they are mistaken and should avoid propagating bad information. Now as for the other directional property inherent in the steel, yes, we do tend to align that with the shape of the blade, and if our blade had a shape like a crankshaft with sharp 90 degree changes in direction then this would have a noticeable effect but we really haven’t effected much on a shape as simple as a knife which is not much different than a flat bar with a sharp edge.

In a nutshell, grain flow "direction" is created when steel is rolled, which pretty much all of it is. Through multiple passes through the rollers, when being rolled to whatever the final size/shape might be, the grain(s) are made smaller and elongated, creating the "grain flow" in a given bar/piece of steel. I think it's important to know/understand that there is a huge difference between "grain flow" and "grain size". How large or small the "grain size" ends up is determined by the heating and cooling processes we induce. Whenever a piece of steel is forged to shape, the "grain flow" remains in the as forged "shape" (for lack of a better way to say it).

Based on my experiments over the years, I believe the "grain flow" is always there, it's just altered when heated, forged, etc., and remains in that as forge state unless altered through grinding or other methods of removing stock.

Does the orientation of the "grain flow" have any impact as it applies to a knife blade? I suppose that's debatable, but personally I have come to realize/understand it does.

The way I started out understanding, was to work with wrought iron.... it's "grain flow" is visible with the naked eye, and is greatly enhanced with etching.... and that same type of "grain flow" exists in just about any non-cast material we'd use in forging.

Since a picture is worth a thousand words, here are a few that support what it took me years to understand:

As I was reviewing this before posting it, I saw that Kevin was able to provide a more scientific explanation then I, so hopefully between us we can invoke some thought on the issue. I think in the end, at least from my perspective, each individual has to take the information for what it's worth to them..... and figure out for themselves how, or if they can realize any impact, positive or negative in their own blades.

I'd love to see someone make like 5 tests knives out of stock removal and forge 5 of the same knives and do some bend, impact, and edge retention tests on them. Intuitively a forged object could be stronger than a machined object. But I'd like to see if that is true in practice with a knife. Maybe that theory becomes more pronounced when you start getting into sword sized objects that need to hold up against severe shock and vibrations caused at impact?

Good idea, maybe we could find somebody who has done such testing that could share some insights on the matter, that would be cool <img src=' http://www.americanbladesmith.com/ipboard/public/style_emoticons//wink.gi f' class='bbc_emoticon' alt=';)' />. But somehow I doubt it would settle many of the questions. <img src=' http://www.americanbladesmith.com/ipboard/public/style_emoticons//biggrin.gi f' class='bbc_emoticon' alt=':D' />

Wow. This is great stuff! While studying Verhoeven's book a few weeks ago I remember reading about the control of grain size and that soap bubbles were used as an example to model the properties of the grains of steel, but I don't recall reading about "grain flow".

Thanks for the pictures and explanation of "grain flow" direction. Being new to this world, pictures seem to make little light bulbs start glimmering in my head as this is something I had a hard time picturing.

I thought about mentioning stock removal versus forging in my previous post, but thought it might get me a lot of jeering and maybe even some hostility based on the age old "Stock Removal versus Forging" argument..... but what got me started on my particular road was that I noticed some subtle difference between the performance of stock removal blades and forged blades I had made of the same materials. That lead me to experiment and test blades of the same steel types, head to head, with a stock removal versus forged mentality, in an attempt to better understand how and why.

In the majority of my experiments, forged blades always seemed to out perform counterparts made via stock removal. Now, that being said, some of the comparison tests revealed subtle differences depending on specific steel types, but others, such as 52100 and 5160 exhibited profound differences between the stock removal and forged versions. Because 52100 usually comes in a highly spherodized state, I believe the forging/heat and cooling cycles played a major role in the differences seen between stock removal and forged blades, and the same, although to a lesser degree, with 5160. Plain carbon steels exhibited the least differences, but as alloys increased, it seemed the differences in performance widened.

In short, the experiments I did, have lead me to the choices of steels I primarily use, namely 52100, 5160, and 1084. I of course use others, but because of what I learned, I tend to use more of the aforementioned steels then anything else.

What did all my experimenting prove..... in the overall scheme of knifemaking..... nothing.

I wasn't out to prove anybody else right or wrong, but rather to determine whether there were any facts in the theories of knifemaking that I had built for myself..... For me, and what I do, and how I do it, it proved much, and it gave me a confidence in my forged blades that I otherwise would not have.

In a world were it seems everyone thinks if its on the internet its gotta be true, or rather then being lazy and just accepting another's word, my advice is to build theories about why your knives are as they are, or why they are capable of doing what they do. Then, prove or disprove those theories to yourself. Its not a matter of proving anyone right, or anyone wrong... it's about understanding your craft, seeking improvement, and having the knowledge and confidence to do what you do, and understand why you do it.

So.......what's the "grain flow" in twist Damascus?

Thanks, Kevin & Ed for taking the time to explain this. I enjoyed it thoroughly.

Gary

So.......what's the "grain flow" in twist Damascus?

Always gotta be one in the crowd. <img src=' http://www.americanbladesmith.com/ipboard/public/style_emoticons//smile.gi f' class='bbc_emoticon' alt=':)' />

|quoted:

Always gotta be one in the crowd. <img src=' http://www.americanbladesmith.com/ipboard/public/style_emoticons//smile.gi f' class='bbc_emoticon' alt=':)' />

I was serious!

But I don't mind being the "one".

Very informative thread, Thanks very much!

|quoted:

I was serious!

But I don't mind being the "one".

There is so many topics attached to this one that an entire forum could be devoted to the directions it could take, so I will first try to keep to those closest to the initial topic. Which brings me to Karl’s question which does address the topic. Making damascus does indeed play havoc with the initial anisotropy of the steel, but… it introduces all new, and sometimes more extreme, directional artifacts via the weld zones. First I should touch on the differences in what we refer to as “toughness” Bending or flexing blades is a very nebulous concept and has some many more variables attached that it is very difficult to glean useable data from it. How a material, especially steel, behaves in gradual loading is entirely different than in sudden loading where the slip mechanism within the steel does not have the time to compensate for the load. One of the best illustrations of this effect is what I have observes in something like vinyl siding. You can flex and bend and even stretch vinyl siding, but whack it quickly and it shatters in a brittle fashion, steel is pretty much the same. And this is why whenever I refer to “toughness” I am referring to a property that can be isolated and measured and that is impact toughness. And that is why years ago I invested in this toy:

It is a Riehle Charpy/Izod impact tester and its sole function is to break steel and tell you how much energy it took to do it. Now I found a funny thing, you can flex and bend twist pattern Damascus all day and not notice any real difference, but when you throw 240ft lbs. at it all in one shot you start to see noticeable trends in the numbers. Flat patterns, like random and others that do not have any weld zones going transverse across the piece will show higher impact resistance than damascus that does.

Now please let me qualify that- 240 ft. lbs. concentrated on a striking edge, that resembles a dull axe edge, applied to a piece of steel that is 10mm X 10mm X 75mm is more force that any human being is ever going to exert on a knife blade. The forces involved are almost incomprehensible when you consider a knife. So while there is a difference, before the guy with the ladder pattern hunter gets cocky with the guy with the twist pattern hunter, they need to get real and realize they would have to practically shoot the blade with a cannon to notice the difference. And then they also need to explore the other properties…

Other tests for other properties were even more interesting. Many years ago Tim Zowada and I spent about a month testing a variety of damascus steels and single alloys in several key properties for a presentation we did at the Ashokan seminar. Included in those tests were a series of cutting measurements done by a tester somewhat like the CATRA tester that we made. We found something very noticeable in the way that steel which had directional properties transverse to the edge effected edge stability and cutting aggressiveness. Twist patterns tended to wear into a more toothy condition that made them much more aggressive on soft things like paper, meat or cloth, than patterns running parallel to the edge or single alloy blades.

I could go on, and on, but suffice it to say that just like everting else in knifemaking there are always trade-offs and no single thing can give you the ultimate knife. In-fact you can even get into trouble saying that this one, or that one, is better than the other, because as soon as you do you find another property that changes things; not to mention that it is the end user that determines if the knife is great for what they are doing with it.

|quoted:

...I remember reading about the control of grain size and that soap bubbles were used as an example to model the properties of the grains of steel, but I don't recall reading about "grain flow"...

Yes, and this is exactly where we need to be careful to keep two very different things separated. Whenever we discuss "grain size", we are dealing with the crystalline grains that make up the many phases of the steel, e.g. austenite grain size. This grain generally is not as directional and is only really made that way by extensive cold working, but will reform in a more uniform, non-directional, condition when heated to recrystallization. The flow lines,(you will notice that I try like crazy to avoid even using the term "grain" when discussing the directional properties of the steel overall), on the other hand, will survive the heating and reheating, and can be manipulated by forging. These lines are artifacts left in the steel from when it was cast and then rolled.

A cast ingot will be filled with alloy and carbide segregations, voids from gas bubbles and even some non-metallic inclusions. These are the sort of things that are very bad to have isotropic and large, the steel would simply fail in almost every strength/toughness category*. The ingot is then ran through rolling mills at temperature to break the carbides and inclusions up and weld the voids shut. Rolling an ingot many feet across down into a bar of steel that is mere inches, results in all of these point defects being drawn out lengthwise throughout the bar and making the steel much stronger in that direction. But if you slice a section off the end the steel is still weaker in that transverse direction due to the defects being more of a problem across the direction of drawing. It is these lengthwise inconsistencies that you see in the illustrations that Ed posted above, not the crystalline grains in the steel which form, and reform, within that framework, as you can see in my micrographs above.

On the side topic of trying to test these concepts. The reason why I went to ridiculous lengths in equipment to simply test and analyze steel was frustration in trying to isolate and retrieve accurate data. To make any study, or test, mean anything, we first have to isolate what it is we are testing, and quite often we really are not sure about that. “Is this knife better than that knife” is a classic example of testing that is doomed to fail or, worse yet produce misleading or bad data. This is mainly because we are starting from a completely subjective standard, I’ve seen knives that somebody raved about that I thought were terrible, there are no objective answers here because we all have different definitions and equally valid opinions on performance. We need to isolate the properties we wish to explore to get a clear picture of what we are observing, and even then we still need to eliminate countless variables, or determine which one is responsible for the effects observed.

Two good examples- edge stability and flexibility, a couple of characteristics that we focus on a lot in the ABS. But if we simply go to testing these qualities by picking up a knife and going at it, we have already compromised our chances of retrieving objective data. The power of human bias is astronomical, ubiquitous and all but impossible to recognize. From doing the test, to crunching the final numbers, we will try to make the data fit any beliefs we had going into it, it is unavoidable, even if we try not to.

Next comes the challenge of isolating the variables. Once again considering edge stability and flexibility, more often than not we associate these with proper heat treatment or steel condition/quality, but a more profound influence is simple geometries. An edge with a terrible heat treatment but a geometry that is just right for its task and condition will outperform a blade with a great heat treatment but an incompatible geometry. Flexing is even worse, if all one does is flex the blade, we may think we have learned something about the heat treatment but simple flexing is entirely a function of geometry. It may seem crazy to say it within the ABS but blade bending and flexing is one of the most nebulous tests I can think of, in fact I would prefer to call it a “demonstration” rather than an actual test. But it is useful in demonstrating a smith’s skill in controlling heat within zones on a blade with a conducive geometry.

Another great example that involves this thread topic was evident in a series of test that I did with forged steel compared to non-forged. In higher carbon alloyed steels (O-1, 52100 etc..) I tested two samples, one was forged and heat treated the other simply ground and heat treated, and I saw noticeable differences in the properties. First of all was Rockwell hardness and consistency of the same. The hammered steel had better numbers working with this concept of manipulating the flow lines and refining with the hammering. But...

Then I did another series of tests that involved the same process but this time every time I put the forging in the heat, I also placed the unforged bar beside it, I just didn’t hammer it but let it air cool on the anvil while I forged the other. Suddenly the forged one lost its advantage as the numbers equaled out again. I was so focused on the hammering that I ignored the effects of repeated homogenizing heats on the condition the steel had from the mill. If I had simply cried eureka with the first results because, as a bladesmith, that is what I wanted to believe, I would have inflicted some seriously flawed information on our craft, that would have carried undeserved weight because it was the result of what looked like serious testing. That is the real danger of testing if we are not careful.

This all ties into the topic of flow lines, as well as grain size, because it is very easy to put an awful lot of effort into testing and not really answer any questions, because of all the variables. Then when you isolate the specific properties you find you may lose on some and then gain in others. The universe likes balance, and tradeoffs, it abhors the concept of unobtainium.

*compression strength might not be too bad.

Excellent discussion!

While I wouldn't be able to offer such a detailed and technically descriptive rundown of grain flow and grain size/structure, I am happy that I my mental image of things is essentially the same as what Kevin has laid out.

Ed, those graphics are great! I'm saving them to show my students.

Thank you Gary for prompting this discussion!