Sunday, July 8, 2012

An Approach to Choosing Blade Steels, Part 2

The previous article in this series dealt with a little about why blade steel matters and a little with metallurgy with a good rant thrown in for measure.  It can be found here.  This article is going to lay out a little about traits steel possesses and how they work in tension with each other, a bit about heat treating and blade geometry, and then a word on the Rockwell scale and whether or not it matters. 

The Power of Three

All of the elements and processes discussed in the previous article combine to make steels with different attributes.  In cutlery steels there are three broad traits that exist in equipoise with each other--usually increasing performance in one will decrease performance in another or both.  Additionally, within these three traits there are sub-traits that again exist in a balance with other sub-traits.  The elements used and the methods of heat treating determine how well a steel performs, both on the big trait level and in each sub-traits.

The big three are:

1.  Corrosion resistance
2.  Hardness
3.  Toughness

Corrosion resistance generally means resistance to rust, though other things can and do cause corrosion.  Generally speaking lower carbon content equals less rust and higher carbon content equals more rust.  Other things make a difference too.  Chromium is the usual answer to rust problems, but recently some makers have taken a different approach to the rust problem.  The idea is interesting, instead of increasing the chromium content, these makers have REMOVED the carbon and added nitrogen.  This has traditionally been a huge no-no because the thing that makes (or made) steel better than straight iron was the hardness carbon lent the mixture.  Nitrogen can be used to harden iron, though until recently it was difficult to do.  Now that we can do it we are getting very corrosion resistance blade steels, like H1, X15TN, and Elmax that can be hardened quite high as well.

Hardness is usually represented by a Rockwell score, more on that later as well.  Generally it related to how resistant a steel is to wear damage and how long, in knife steels, it can retain an edge.  Carbon in large amounts typically adds to hardness.  ZDP-189, for example, has a HUGE amount of carbon proportionally speaking at 3.00%.  Many non-stainless steels don't even have that much carbon.  But with lots of carbon comes lots of potential for rust (hence the 20% chromium content in ZDP-189).  Many people have found that ZDP-189 can tarnish or get off colored even without actual rusting because of the high carbon content.  I am lucky in that I have not had that happen, but lots of folks have.  One caution about hardness--the harder something is the more brittle it is.  Diamonds for example are VERY hard and can shatter.  Ceramic is likewise very hard but can can shatter if impacted even moderately.

Toughness is a steel's ability to absorb shock and force and still retain its shape.  If hardness is best seen in something like ceramic, toughness is best seen in something like plastic or taffy.  You can smash a ceramic cup with a baseball bat, but you can't smash a piece of taffy (unless you freeze it, which is really cool to do by the way).  Iron is naturally tough and toughness is seen in steels with low Rockwell scores.  But sometimes toughness is really important.  In a survival blade that has to withstand tremendous forces, such as those seen in batonning wood, a hard steel would just shatter.

The trick with steels is that if hardness is increased, usually toughness and corrosion resistance suffer, and so on with each attribute.  It is not possible for a steel to be everything for everyone.  The question is how do you balance these traits with your intended use.

Finally a word about other traits of steel.  You might see someone say that they want a steel with good wear resistance, chip resistance, edge retention, or ease of sharpening.  Most of these different, more specific traits are related to one of the three major traits.  Wear resistance is highly correlated, but not the exact same as hardness.  Chip resistance is highly correlated but not the exact same as toughness.  So once you find a specific task and you have figured out how your steel to performance in the three major areas, further research might help you figure out which sub-traits to emphasize.

Just about any steel chart worth its salt will tell you what elements are included and at what percent.  They will also tell you about which elements promote one of the sub traits over another.

Heat Treat and Blade Geometry

If there is any part of metallurgy that is still shrouded in mystery, at least for the general knife-buying public, it is heat treating.  There are all sorts of methods and myths.   For example a lot of folks believe that hammer blows near the edge of a blade can "pack" the edge.  This is 100% horseshit.  The density of steel is controlled by the chemical structures and unlike sand, steel is not meaningfully compressible by mere hammer blows.  If anyone tells you that they can pack an edge or wants to sell you a knife with a packed edge walk away.  They are simply a more sophisticated pick pocket.  It IS possible to differentially harden steel, which is what proponents of packed edges seem to claim happens, but hammer blows from a human being will not do it. There is an old fashioned method of differential hardening used in Far East bladesmithing involving clay and selective quenching and there is a vastly more high tech method, seen in these blades, but Thor himself could not do what the charlatans claim they can do when making a packed edge.

But there are folks that can get better performance out of steel through finishing properly.  Paul Bos, a long time metallurgist associated with Buck Knives, has an amazing reputation for his Bos heat treat.  It is proprietary and a well guarded trade secret, and unlike the packed edge bullshit, it really is better.  A Buck Bos-treated blade will perform substantially better than the same steel treated by someone else.  My experience has proven this to be true as has the experience of millions of Buck knife owners.  Bos is for real.  It makes a difference and a substantial one.

Let's cut through all of the murky claims and reputations and look at what heat treating really does.  Heat treating steel is design to do one thing--control the chemical (crystal or grain) structure of the steel alloy.  When steel is made, the microscopic crystal structures are set into the strongest configuration.  As the steel cools, things shift around.  Think of the steel like instant pudding--as it cools things shift around and the pudding develops different textures in different places.  Steel acts the same way if it is not cooled in a very controlled way--the helpful properties of the steel are distributed unevenly.

This is because steel is an allotrope.  That is, it can have the SAME EXACT ingredients but produce very different materials.  Diamonds and graphite are good examples of carbon allotropes and a perfect example of why heat treating is important.  Diamonds contain a vastly more sturdy configuration of carbon atoms than graphite does and that is why, even though they are made of the same thing, carbon, diamonds are harder than graphite.  As the steel cools the chemical composition remains the same, but the chemical structure does not.  As the temperature changes, a different allotrope of the steel emerges, one with inferior properties.

Typically heat treating heats the steel up again, redoing the structure as this happens, and then cools in a very, very controlled and precise way.  This precision cooling limits the restructuring of the chemical components of the steel and forces them to remain in a closer to ideal configuration.  Conversely, some heat treat with cold temperatures (SOG does this).  In essence they flash freeze the steel, sealing the more idealized structures in their original positions through SUPER cold temperatures.

Heat treating is important, but it is really hard to judge the effectiveness of one heat treat over another.  I don't really like the SOG cryo treating, but that comes from lots of experience.  In contrast, I really like the Buck heat treating, but again this is from a lot of experience.  Because testing a heat treat requires multiple samples and lots of time, it is hard to convey one method's effectiveness in a two week long testing period (this is one reason I like to update the reviews a year later).  Good heat treat is really helpful, a cheap and easy way of improving lesser steels, but it is hard to figure out which ones work and which don't.

Blade geometry, however, can be readily subjected to immediate analysis.  Blade geometry is really, really important.  How important?  I think it is more important than the steel choice itself.  In fact, what knife makers call blade geometry I split into two categories in my knife scoring system--blade shape and grind.  So a knife with great blade geometry can max out a score of 4, while a knife with great steel can only get a 2.  In fact, I think if you know what you are doing you can make up for a subpar steel with good blade geometry.  Additionally, if you want to really take advantage of high hardness steels, you can play a little with the blade geometry and get truly spectacular results.  Blade geometry is a general category for two aspects of a blade, as I mentioned above--grind and blade shape, and it also interplays with how you sharpen the blade.

Grind is the cutting profile of a knife.  There are a ton of different names and grind types, but they all break down into three main groups--concave, convex, and flat grinds.  Concave grinds are called hollow grinds.  Convex grinds, which typically have no secondary grind or bevel (the cutting edge itself), are known by many names--appleseed grind, Moran grind (after Bill Moran).  There are some differences between appleseed grinds and Moran grinds, but in general they are all slightly bowed out if you looked at them from the tip.  These grinds are the most difficult to do because they essentially require the grinder to work the blade at all points, never simply resting the steel on the belt.  They are also very strong and generally more expensive (because of all the labor involved).  Flat grinds are just like they sound--flat.  Typically they have a secondary grind as well.  Finally, some companies, like Emerson, sharpen their knives only on one side.  They may have a typical grind and a secondary bevel on one side or they may grind only one side and leave the other completely flat.  This makes for easier field sharpening and a thinner profile, but it is a different experience using these "chisel" ground blades to cut with.  Additionally because of the thinner profile they can chip more often than traditional V grinds. 

There is a fourth grind, one that takes some three dimensional visualization to understand, called a Besh Wedge.  Here is a video of the Besh Wedge in action, as I have found it impossible to describe adequately in words:

It is like no other grind out there and has ENORMOUS tip strength comparatively speaking.  I could see this grind becoming more popular in the future, but for now it is relegated to a few Buck blades, a few Boker knives, Meyerco blades, Blackhawk knives and Beshera's own custom knives.  It is pretty freaking amazing though, and proof that even after 10,000 years of working with edged tools human beings are still smart enough to innovate.

In terms of grind, I like simple grinds.  The multifaceted grinds or the complex grinds can work in specific applications, but for EDC/utility tasks simpler is better.  I like Spyderco's full flat grind.  I also like SOG's full flat grind.  I think the high hollow grind on the Sebenza is great for slicing, but when you have to do heavy duty cutting it can bind.  This has not happened to be, but I imagine if you were cutting material that tended to bunch, like lots of cardboard, it could happen.  As with most things, grinds follow the simpler the better rule.  

In addition to the grind, there is also the blade shape itself.  There are literally dozens of blade shapes out there, but for me there are a few that stand out.  The classic SOG drop point blade is probably my favorite overall shape, seen in my Flash I review.   The long primary edge coupled with a nice belly provides a lot of cutting power.  I also like the leaf shape blade from Spyderco as well as their saber type grind seen on the Delica, Endura, Military, and Paramilitary.  All are excellent cutters and easy to maintain.

I dislike recurves of all kinds.  They are unnecessarily hard to sharpen and though they obviously give you a longer cutting edge in a shorter distance, it is not worth the hassle in my opinion.   I also dislike tanto points as they do not typically perform roll cuts well.  A reverse tanto, seen in the Benchmade 940, for example, is a very good design, incorporating the tip strength of a tanto and the belly of a drop point.

Again, as with most things, the simpler the better. 

Rockwell Scale

The Rockwell Scale is not so much a scale as it is a method and a set of different scales. There are four primary Rockwell Scales and the one used in the knife industry, the Rockwell Hardness C scale, is for medium hardness materials, such as hard steel.  Rockwell A is used for tungsten based materials, i.e. VERY hard stuff.  Rockwell B is for softer stuff, like brass.  All of the scales work the same way.  They are measurement of a material's resistance to deformation.  A tip or cone of a known hardness material, usually steel, is attached to a device that looks like a drill press.  The material being tested is laid in a bed beneath the tip and the tip is pressed down into the test material.  The depth of penetration into the test material is compared to other known hardness materials and the test material is then given a score.

Most cutlery steels score somewhere between 45-70.  A 45 would be very soft steel, used in very high impact operations like a ax or a maul.  Some of the hardest utility steels score closer to the 70, with ZDP-189 hitting 66 on a regular basis.  Typically the scores are given as a range, usually two points, so ZDP-189 will be listed as a HRC of 64-66.  This range reflects the variation between steels of the same make and it also reflects variations in test results.

Steels of a particular type have a "recommended" hardness.  That is, the maker of the steel has a target range in mind for the hardness of their steels and they believe that this range gets the best performance from the steel.  Crucible, for example, usually recommends between 58-62 HRC on S30V.  Knife makers can increase or decrease the hardness of a steel based on their heat treat method.  Bob Dozier, for example, gets very high hardness out of D2, more so than others, based on his high treat method, which is, of course secret.

I think that utility blades benefit from a high hardness because it means less maintenance.  I like ZDP-189, but I have seen some blades with micro chipping on the edge.  It can be fixed, but if you prefer no edge chipping, S30V or S35VN might be the way to go.  They are softer but still capable of getting hard enough for really steep cutting angles.  I also like D2, but it requires anti rusting maintenance, a tradeoff for the harder steel.  In a survival blade, I like 1095 from ESEE which hits in the mid 50s on the HRC.    

Hopefully, this second piece has been interesting and useful.  Next up, a list of steels and explanations of their properties plus a peek at my personal favorites.  


  1. Very interesting and useful. Thanks.

  2. This series is developing into a significant accomplishment that promises to be the best compendium of such information on the web. It is beautifully organized and written and should be more readily accessible on your site, perhaps under the "Special Series" tab. Thanks for your continuing focus, generosity and efforts.

  3. Very impressive and informative. Ideally there would be a more standardized system for comparison. Unfortunately, the subtle changes makers can do to the steels make it impossible. As you have mentioned in previous reviews, AUS-8 has performed relatively poorly in your tests but you were surprised by Cold Steel's (or was it SOG?) version. There are just too many variables for that system from cutting material, frequency of use, heat treat, etc. One can hope though.

  4. Wonderful reference -- thank you.

  5. This post answered many questions I had and gave me a better overall understanding of knife steel. Thanks!

  6. Clearly a well-researched and well-worded series - thank you so much!

  7. I enjoyed this read, as usual. Gonna need to start calling you professor T!

  8. Most of the info is good, however, you need to dig a little deeper. The details are far more complicated and ambiguous than what you've outlined here.

    I.E: You state that hammer forging knives is a load of "horseshit" and that packing isn't real. Controversially, a technique used in powder metallurgy does exactly that. It is called HIP (Hot Isostatic Pressing).

    The reason why people hammer forge knives (repeated folds or flattening rod) is to rid the steel of inclusions that are unavoidable during the processing of metal. PM techniques have gone a long way to try and reduce inclusions which adversely affect mechanical properties, producing a "clean" material. Even with PM, alloy processing isn't perfect (inclusions still form) but it isn't far off from ideas used hundreds of years ago.

    Also, there really isn't such a thing as an "inferior" allotrope. The different carbon matrix phases (if all the carbon is used in forming matrix during heating) all have a purpose depending on application. Quenching and tempering are done for a reason. You don't want a blade that is all martensite, you don't want a blade that has no martensite. This is oversimplified, of course, when you start throwing in other elements (V, W, Cr, etc.). Well, it is still oversimplified even with simple carbon steels.

    1. Your right about the details being more complex than what I have presented here. In fact, I made a point of stating that in the first article. Metallurgy is very complicated, but quite frankly a lot of it is irrelevant to choosing a blade.

      As for the edge packing thing, it is going to take some very compelling evidence to show that this is true. There is, to my knowledge, no definitive proof that this can occur. Why, if it could, would we spend all this money on new and better methods of making steel if a little tap, tap from a hammer can do the magic that edge packing proponents claim it can? I asked three people--a metal and materials engineer for a defense contractor, my wife a chemistry professor, and one of her collegues, about this and they all said this is not possible. Hammer forging can get rid of impurities in traditionally made steel and it does have a role in knife making. What doesn't is the notion of a packed edge. The idea that somehow hammering the edge produces a superior steel is ludicrious. The forces and energy involved in HIP and all of PM for that matter so outstrip what a hammer can produce that comparisons between the two make no sense. Hammer non-PM steel does have an effect but it cannot make something better than the steel is minus impurities. Packing an edge is 100% horseshit, noting that most people see it as different from merely hammer forging. Its MAGICALLY hammer forging.

      I also agree that there is no such thing as an inferior allotrope. As I referenced in the first article there are sorts of configurations of steels many of which are designed to do things entirely different than cutlery steel. I referenced COR-TEN for example, a steel designed to rust. The point of heat treating is to keep the steel (or return it to) the intended structure and keep it that way as long as possible. It is not superior in the sense that it is better in every way, but superior in the sense that it is closer to the intended structure.

    2. I wasn't asserting that edge packing is real (which, I agree, is BS). I was merely stating that hammer forging does make a difference for certain non-PM steels and that modern metallurgy still follows certain old school thought (even if it isn't inherently explicit).

      Also, S30V/S35VN are not "tough" steels. They microchip/chip quite easily (lack of ability to hold a keen edge after cutting a semi-hard material, thick plastic, aluminum). Even at 58RC, max impact resistance if probably only going to be 20ft lbs. Compared to 3V at 58, which can reach 100+ ft lbs. From my experience (I've used quite a numer knives with S30V, custom and production) it is and feels just too brittle. Steels like M390 and even Elmax fair so much better than S30V. They are tougher/stronger/more wear resistance even at the same RC hardness (which partly has to do with composition and partly due to better PM processing [and HTing, but you get my point])

      IMO, I think you should also mention that while RC hardness is a good indicator of heat treat, it shouldn't be used as a scale to measure the other properties of a steel. S90V vs. S30V vs. Elmax both at 60RC have enormously (okay, maybe not that much) different properties.

    3. I agree with the hammer forging point. I hope I did not come off as disagreeing.

      "Tough steel" is relative. Compared to ZDP-189 S30V is tougher. Compared to 15V even ZDP-189 is comparatively tough. CPM's 15V is a high speed tool steel that will regularly hit 70 HRC and has very high wear resistance properties as well, so it is a relative term. Yes, S30V/S35VN is not typically seen as a tough steel. That is absolutely correct.

      I also agree 100% with the HRC point. It is a very limited utility measurement.

  9. No, not at all. You've clearly done your research, I just wanted to point out somethings that might not be obvious to the casual reader.

    Though, I'm not sure you're correct about 15V reaching 70 RC regularly. Are you sure you aren't confusing it with CPM REX-121?

    As for anything, it's all relative.

    There is also another characteristic of steel that you didn't really mention and that most people don't think about: edge stability/ductility. I would actually consider that toughness falls under this category and not vice versa. This characteristic is particularly important for thin edges. However, since most companies make knives with a fair amount of material behind the edge, again, most people really don't think about it. Once you start making your edges at <15 degrees and have less than 0.01" behind the edge, you'll realize that a tough steel means nothing if can't actually hold itself together under use. I believe Mo is on element that is utilized to help form a very tight crystalline matrix which M4 has a relatively high amount of Mo. I've seen many makers make excellent Competition Choppers out of it. Many custom makers grind their choppers extremely thin at the edge. Even then, the edge can still withstand the stress induced from impacting and cutting through a hard oak dowel.

  10. Are you talking about this Elmax:

    I would assume yes, because this is the registered trademark producer.

    If so, I think you are mistaken about the nitrogen replacing carbon. The data sheet clearly shows ~1.7% carbon. It doesn't show any nitrogen in the composition. Do they perhaps use a nitrogen environment in production to prevent any oxidation while producing/normalizing the steel?

    1. Scott, good catch. I had a bunch of steel white papers on my desk and I must have mixed up two B-U steels. Vanax is the the carbon-poor steel that uses nitrogen as a hardened element, not Elmax.

  11. Wow this approach for selecting steel blade and thanks for share this.
    Godaddy Coupon and greengeeks Coupons