Head to Head Test

Oct 19, 1998
I think it would be interesting to test some of the "better" steels that are available in folding knives against each other. How about testing BG-42 in a full size Sebenza, An AFCK with M2, a Spyderco Military with 440V, and maybe an AFCK with ATS-34 as a benchmark. It would be interesting if Cliff Stamp could test these knives and steel types head to head. If he could start with new knives, he could first test their factory edge. Then maybe a sharpening with a Sharpmaker since all of the makers of the above knives reccommend this sharpner. This could test them with the edge that most of us would be putting on the blades. After that, Mr Stamp could use his expertise and change the sharpening angles and test the potential of the knives.
I think that the only way to truly test which steels are "better" than others is to make identical knives out of the different steels. Same shape, grind, edge geometry, etc... I think that using different knives throws way too many variables into the equation. But, that is JMHO...


AKTI Member #A000289

Deo Vindice

I agree with you Senator. Compare steel for steel only. But don't we get a little overboard with this supersteel business? In this day and age do we not have a certain steel and a certain knife for a certain purpose. I own several knives, like many on this forum. I also like obtaining new knives with new steels. I do not toss out my previous knives. Test steel for steel, but remember the purpose of each. Only my opinion. I also like to read these tests BTW because I obtain information. Keep on testing! Do not reject a knife or a steel based on single results. Many knifemakers make good knives with "old" steels. Everyone has an opinion, as shown on these threads.
I predict Cliff's answer, when he sees this thread, will be, "Send me the knives and I'll test them."

Cliff does a fantastic service for us all with his knife tests ... he does a great job testing knives and a great job writing up his results, too ... it'd be a little much to expect him to not only go to all that trouble, but also buy every knife we want him to test....

He seems to be very willing to test any knife you want to send him, though, and since you're asking about edge retention rather than toughness, you'll get them back in one piece. If you have all those knives, that is ... of course you don't. That's the problem -- everybody wants to see tests of the knives we're *considering* buying. Oh, well -- maybe somebody else will send some of those knives to Cliff to test, or test them themselves ... these dream comparison test proposals seldom seem to work out as envisioned, though.

-Cougar Allen :{)
Of course Cliff is eager to test any knife, he having somebody send him free knives to beat on stuff with; who wouldn't be eager?
There's a web page running around which has a listing of a byunch of different steels tested by Wayne Goddar, who made identical knives with each steel...It's interesting stuff, but I lost the damn link...ugh...I'll try to find it...

When Cliff did the testing of the Uluchet we, P.J. Turner Knife Mfg, Inc., supplied a piece for this process. Cliff DID NOT keep the knife, nor did he ask to. He did however provide the time for testing, writing the review and answering a bazillion questions from forum members who read the review. When Cliff returned the Uluchet he did so at his own expense.
Cliff is NOT in this for the free knives and whatever else. Not everyone in this world is out for #1, some people do things for the pure enjoyment it provides themselves and others.

P.J. Turner Knife Mfg, Inc.
Uluchet, What's an Uluchet? Find out at...

Well, I have done side by side testing of a Benchmade M-2 AFCK, an ATS-34 AFCK, and a Spyderco Military in CPM440V. For toughness, I cut through a lot of comm cable, electical cords, and tin cans. All three blades performed pretty much equally, dulling about the same amount, no visible chips from the edges. So I got mad and decided to dig my way through a 2X8 with the two AFCKs, stabbing them into the board and prying out chips. I managed to break about an eigth of an inch off the tip of the ATS-34 blade. I only managed to bend the tip of the M-2 blade, and easily bent it back. I didn't do the prying with the Military because I wanted to sell it rather than damage it.

For edge holding, I whittled away at a big piece of particle board with all three blades, taking ten similar cuts with each blade in rotation, waiting for one blade to differentiate itself as the winner, and one the loser. It never happened. I whittled away for hours, hundreds of cuts, and all three continued to perform equally until I gave up.

I liked sharpeneing the M-2 AFCK the best. It has a very low tendency to form a burr, and produced a nice fine edge with medium effort. The ATS blade was less abrasion resistant to the stone, but forms a significant burr that takes some extra effort to polish off, making it just as much trouble to sharpen as the M-2 blade in my opinion. The 440C blade took the most effort of the three.

Sorry, I didn't have a Sebenza at the time, so there was no BG-42 in the test, but my big impression from the exercise was that the differences in practical performance between premium cutlery steels are small rather than large. Reeve BG-42 is probably finer grained and tougher than production ATS-34 blades, but the actual difference in cutting performance is likely to be pretty hard to demonstrate, even harder if you compare BG-42 to ATS-34 that has been custom heat treated.


[This message has been edited by Steve Harvey (edited 30 August 1999).]
Danny, I don't have a sharpmaker, and generally freehand grind to a 15 degree or lower bevel depending on the steel. It would not be overly difficult to set this at whatever the Sharpmaker sets though if a comparison was of interest.

Kelly :

I think that using different knives throws way too many variables into the equation.

While the more you restrict the conditions the more precision you get, it is also the case that the less knowledge you gain. Precision can always be refined by later work.

Cougar :

He seems to be very willing to test any knife you want to send him, though, and since you're asking about edge retention rather than toughness, you'll get them back in one piece.

Likely, but not guaranteed. Accidents happen. Even if I can hold back natural curiosity blades seem to want to all by themselves experiences great stresses while I am handling them. Often times I can block slips by interjecting myself infront of the blade but sometimes I am a little too slow.

Anyway, I don't have much of a problem with looking at much of anything. About the only restriction is that I won't do biased testing. This basically means that I will not showcase a strength if it is at the expense of a lack of performance in another area. If it is a personally owned item, I will discuss limits but they have to be reasonable because of this. If the piece if from a maker/dealer then I will stress the knife to failure or try to anyway, some knives resist rather strongly.

For example, when I looked at the Trailmaster awhile back I noticed upon initial inspection that the tip looked weak and with a little work I concluded it would break rather easily because of the profile. Now it of course had excellent penetration and would easily win over many similar knives in any sort of stabbing test. However would it be fair to just compare them in this area and ignore the weakness? I don't think so which is why I broke about 7 cm off of the tip.

Dano, Steve B. has this as well as a host of other information at :

Concerning PJ, I have nothing but good to say about him and his product. His word is as strong as the D2 in their blades. After I finished looking at the Uluchet a forum member asked me why I didn't do extensive durability testing on it. I replied that I saw it as more of a hunting tool that survival and didn't see the purpose in evaluating it in that area. The forum member contacted PJ, who then dropped me an email asking me to actually go ahead as he was interested in how it would handle heavy stress. I then did what I considered would seriously harm the Uluchet, but which did nothing except shear off a stop pin (which I could just easily epoxy back on) and to achive this I had to stress the Uluchet to a level that snaps many fixed blades. Of course when I did this PJ sent out a new set of handle slabs.


[This message has been edited by Cliff Stamp (edited 31 August 1999).]
Cliff, you broke 2 3/4 inches off the tip of a Trailmaster?! Geez, it's fragile, but I didn't think it would be THAT fragile!

Work hard, play hard, live long.

Outlaw :


I didn't think it would be THAT fragile!

It has to be noted what I was doing. I put the tip in a stump and pulled on it with both hands on the handle. It did snap rather easily though I was nowhere near full effort. It was taking my weight but not much else. It was unexpected and I went down in a pile to the amusement of my brother standing nearby.

You can read the full details here :


[This message has been edited by Cliff Stamp (edited 31 August 1999).]
The first problem I see with the head-to-head test outlined above has been mentioned already, namely that all those knives have different blade geometries.

But even if you have one maker make identical knives out of each steel, how do you prove that the maker is equally (and the most) skilled at heat treating all the steels involved (e.g. what if Chris Reeve's years of experience with BG-42 have taught him some subtle tricks of which your chosen maker is unaware).

What if the varying steels are best suited to differing blade geometries? How then do you choose a single blade design that brings out the best qualities of every steel in the test?

What if some of the steels work best with a thicker, coarser edge, while others work best with a thinner polished edge?

What if some of the steels are better suited for differing applications (e.g. one is best for heavy duty camp knives, chopping wood, while another shows superior edge retention when cutting abrasive cardboard)?

The overall problem I see, though, is that likely all these different steels excel in different things, but once some numbers come out that say, e.g., "BG-42 scored a 9.8, but CPM440V only scored a 9.73", some people will skip over the details of the test and abandon all but the highest scoring steel for their future knife selections.

Just my $0.02,
-- Carl
Carl :

What if the varying steels are best suited to differing blade geometries?

Of course they. Since they are different materials they will have different material properties which means they will want different geometries to perform at optimal levels.

How then do you choose a single blade design that brings out the best qualities of every steel in the test?

If you chose a uniform blade design the material best suited to that design will win soundly. This can then be used by you to pick a best material for the blade design chosen.


You can let each blade take the optimal geometry based on the material properties of the steel and thus again see which one has the best performance for you as the different geometries will be each suited to different things.

Both are easily quantitatively possible. And are done regularly by tool makers. You don't see a lot of drill bits made out of 440C do you? No, because M2 makes a better bit. It is that simple.

You can just as easily make the same determinations regarding blade material you just need to define a performance range for the materials to be evaluated over.

As for different makers and such. While a lot of people claim all kinds of things like "I can really bring out true ability of XXX steel". Few if none will post any hard numbers on just how much better their treatment of XXX steel is compared to the norm. If you really have a significantly better way to heat treat a well known material then this is easily proven by walking into a engr. lab and paying them to do a strict materials comparison. This cannot be sanely argued with.

I would make a standard simple test knife design. Start with blanks that are 1/8" thick by 1" wide by 12" long. On one end grind the steel to a 45-degree point (leave the back edge straight). Smooth the transition between this region and the body of blade to about a 2" radius. Create a fully ground blade that tapers from spine to edge 7.5" long. Move down another 1/4" and create a reduced width handle region that is 3/4" wide and 4" long (ending 1/4" before butt end of the steel). Smooth the contours and edges in the handle region, it will be wrapped later in leather cord.

Have each maker create one of these that he heat treats for cutting durability and one for toughness. Begin all tests with no secondary edge bevels and as tests proceed do uniform sharpening on each blade to achieve equivalent secondary bevels. Compare sharpening effort as well as the sharpness and durability of each edge at each step.
An excellent way to test steels "head-to head" in a scientific way would be to machine many identical test blades, apply a variety of heat-treatments, put a uniform edge on each, and then put them in a machine that would measure multiple variables such as initial cutting performance, edge-retention at set intervals, etc. against a known and uniform cutting medium. Oh, wait, one company already does this...

A lot of good points were made about not being able to compare steels because of differences in blade geometry, grinds, etc. My original post was not very detailed nor clear. I understand that these factors would contribute just as much as the differences in steel. I guess the purpose of my test would be to give a head to head test of some of the “better” production knives that utilized these great steels. I thought it would be interesting to see how these high end steels performed in knives that are popular and readily available at a reasonable price (the Sebenza is definitely border line in this category for me). All of these knives get great reviews and are probably separated in performance by very little as Steve pointed out. I guess my main interest was to see how the Sebenza performed compared to the M2 AFCK and 440V Military.

Oh, OK. Well what I said about all the premium knife steels being really close in real world performance applies to the Benza. But the Benza beats the AFCK and Military by a significant margin in terms of cutting efficiency. Reeve puts a very thin hollow ground edge on the Benza, and he sharpens the edge at a fine angle as well. It cuts much more efficiently. A piece of rope that the Benza could cut through in one cut would take two or three cuts from either of the other two you mentioned. You can thin out the edge on an AFCK so that it will cut almost as good, but it never quite gets there.

Dannyc :

A lot of good points were made about not being able to compare steels because of differences in blade geometry, grinds, etc.

This is exactly how they should be compared. Using the same design with different materials is not sensible. The whole reason of having better materials is so that you can build better tools. When I had Phil Wilson make me my blade out of CPM-10V we pushed hard on the geometry. If I had used a design suited to 440A I would have just wasted the abilities of the steel. Same thing with a large machete/chopper that I am thinking of getting made out of 3V (probably ground by Mel Sorg). I will use a geometry that takes advantage of the significant increase in toughness, yeilding and wear resistance of 3V.

Recently on the Spyderco forum (I think it was there anyway), someone commented that the Moran needed a better sheath and Kydex was mentioned. Was the design going to be something identical to the leather one just improved because it was Kydex? No, of course not, that would be pretty silly wouldn't it. Kydex has radically different material properties than leather and thus very different sheaths can be made because of it. Look at Jim March's Outsider sheath as an excellent example and further read some of Jim's comments on how it was developed. He thought about the properties of the material and how they could be best taken advantage of - this is the way it has to be done.

And, you can judge inherent material abilities even if the geometries are not identical, just factor out whatever is different. This does mean you have to understand what effect the differences make of course. As a really simple example - take a piece of pine thick enough so that you cannot break it and now take a piece of black spruce of similar abilties. You will note that the black spruce is much thinner. You can then make the judgement that black spruce has a much higher fault stress for a given cross section, and even further you can judge how much higher it is over pine by using the cross section ratio as an estimator.

Drew :

[blade testing]

put them in a machine

This would be excellent if the blades are going to be used by the ELU's that are machines. You cannot simulate human use by trying to model it in a machine unless you can make a machine that acts and feels like a person. For example one of the most important factors that will be ignored are all biofeedback responses. These strongly determine cutting ability, edge retention and durability.


[This message has been edited by Cliff Stamp (edited 02 September 1999).]
I understand, Cliff, but don't sell short the raw physical properties of the material. A machine is all I need to tell me I'll never make a good blade out of zinc, I don't need "biofeedback responses" in order to see the basic properties of the material.

Obviously I was speaking of Spyderco. When they see exceptional performance from a steel in their testing process, they release a model or two to see how this holds up "in the field." The Moran and Calypso were testbeds for VG-10. The Military was a testbed for 440V. Only after they get ELU response do these steels go into widespread use.

I'm not advocating testing only with machines. But if you want to answer simple questions like "what's its original cutting performance in medium X" or "how long does it hold an edge cutting medium X," specialized machines using identical blades give the most scientific answers.

Drew :

A machine is all I need to tell me I'll never make a good blade out of zinc

What machines can do very well is give you precise estimates of very specific material properties. You could do this without the machines of course, but it would just take much, much more work. They are great for this. My starting interest in the CPM steels was initiated by reading the specification sheets from Crucible.

What is most important though is that you have to realize that the answers are *extremely* specific. This is where the real problems start. You have to know exactly how the experiement is being done to determine what the results mean. Most materials testing done is extremely simplified for exactly this reason.

if you want to answer simple questions like "what's its original cutting performance in medium X" or "how long does it hold an edge cutting medium X

Those are not simple questions and they will not be well handled by a machine.

What a machine could do well is answer the following :

"Determine the the force required to push blade XXX through card stock of grade YYY as a function of length of card stock pushed through. Blade XXX is to be held in a fixed position in the vertical and horizontal and at a set angle to the stock. No motion is allowed except straight through the card stock. The experiment is terminated when the force exceeds ZZZ."

Now just think about this and compare it to actual utility work with a blade.

Regarding that test, if the blades used are of equal geometry you will not be measuring real life edge retention, too many very necessary factors are being left out. What you are determining mainly is wear resistance and while this is very useful info it has to be realized that is all you are obtaining.

specialized machines using identical blades give the most scientific answers.

No they do not. The quality of science is independent of the equipment used. It has nothing to do with it all all. It is how you use the equipment you have. I have much better equipment than Newton had, my work is not much more scientific.

As a simple example of the independence of equipment, one of the first things I do when teaching a class how to do an experiment correctly is have them estimate the acceleration due to gravity with just a piece of lead. No watches or devices for measuring time are even used. It is easily possible with a little care to get an estimate that within the tolerances of the experiement (which are also estimated) matches the known value. The bigger the class size the more precise the result will be it does not get any more scientific however.


[This message has been edited by Cliff Stamp (edited 02 September 1999).]