Blade Materials - General

This is the main page I have on blade materials, I just updated the 1084 section : [www.cliffstamp.com] .

Any information on blade steels would be appreciated if posted here. It should be peer reviewed, no manufacturer data sheets. Commentary from the lay community is also appreciated but strictly only if they are willing to discuss their results, no soapboxing/ranting desired.



Edited 1 time(s). Last edit at 02/16/2015 06:53AM by CliffStamp.

52100 seems to come and go as the non-stainless steel of the hour. Roman Landes has noted that 52100 is widely used world wide as a non-stainless razor blade/box cutter. here are some papers i have found over the years. an interesting one is by Nortrop in 1960.

[digitalcommons.lmu.edu] compares 52100 with 440C and two CPM steels.

[www.hindawi.com] compares different quench/temper schemes for 52100

[www.google.com]
the Northrop tests

[link.springer.com] carbide refining heat treatment for 52100

scott

Quote
oldsailorsknives

[digitalcommons.lmu.edu] compares 52100 with 440C and two CPM steels.

The very interesting result here is that the compressive yield strength of CRU80 (S90V) has a lower compressive strength than 52100 at the same HRC reading, the very high carbide volume weakens the steel.

Quote

[www.hindawi.com] compares different quench/temper schemes for 52100

This is interesting if you look at the difference between martensite vs bainite as the toughness/wear resistance while following the common known behavior (martensite has higher wear, bainite is tougher) is far lower than you might expect.

Quote

http://link.springer.com/article/10.1007/BF02643140 carbide refining heat treatment for 52100

This actually provides evidence for multiple quenching for refinement of carbide size though the ideal way to do it isn't the traditional full quench, it is to form pearlite/bainite and then use that as the starting point to requench and form martensite vs the common spheroidzed starting point.



Edited 1 time(s). Last edit at 02/06/2015 10:03PM by CliffStamp.

I haven't read these directly, but they are similar to the patent by Stickels and some others that recommended multiple quenching and very rapid (salt bath) austenization for optimal carbide refinement and grain size finer than 10, though it was 2 steps insteadof 3. This was the basis of a heat treatment I discussed with Joe Calton for 1095 which involved a normaliing heat followed by the unusual practice of plate quenching, at least for 1095. I don't know if he tried it. If he has, maybe he can provide some details.

it was pointed out on HF when i posted some of these, that the findings are related to the use of 52100 for ball bearings and bearing races. what is good for a ball bearing might not be good for a fine edged blade.
scott

While that is generally true, the papers/patents I have read recently have concentrated on refinement of carbide size, which should be beneficial to fine edges. There are other issues, as I understand bearings are hardened with the desire for significant retained austenite. These types of procedures are, as you say, not good for blades. The ones I've read above don't address this. Now granted, I've only read one of them.

Here is some work by a maker which appears to give the opposite results :

-[www.britishblades.com]

vs

-[drive.google.com]

Reference link : [www.bestexcompany.com]

Abstract :

Quote
Verhoeven, et. al.
A study is presented on the relative wear rates of two carbon steels, a Damascus steel and a stainless
steel, using the Cutlery and Allied Trades Research Association (CATRA) of Sheffield England cutting test
machine. The carbon steels and stainless steel were heat treated to produce a fine array of carbides in a
martensite matrix. Tests were done at hardness values of HRC = 41 and 61. At HRC = 61 the stainless steel
had slightly superior cutting performance over the carbon steels, while at HRC = 41 the Damascus steel
had slightly superior cutting performance.

Aside from the obvious results as stated in the abstract there are a number of other interesting points in the work, one of which is that the CATRA tests are very sensitive to geometry and thus when a relief grind is used, the performance is so sensitive that the depth of the relief grind has to be set by a jig. Just consider what this means in general as to how much of cutting performance is steel vs geometry, basically unless you are really optomizing the geometry to the steel the steel is not even a factor.

There are also a number of well argued myths such as the cutting ability of Damascus which are not bound out by the data unless the martensitic steels are vastly under hardened (and no one uses 52100, AEB-L or 1086 at 41 HRC), and as well that stainless is in general out cut or out lasted by similar carbon steels as the AEB-L clearly out performs 52100, and yes the 52100 was forged. Two of the authors are also ABS Mastersmiths, so the argument that the forging was not "correct" is a fairly weak one. As well the grain sized for both of the forged steels are >> 10 .

Conclusions :

Quote
Verhoeven, et. al.
(1) Use of the CATRA machine with their standard board to evaluate
cutting performance of steel blades is best carried out at small
stroke numbers.
(2) These experiments support the view that martensitic stainless
steels optimized for hardness and fine carbide distributions
have slightly superior cutting performance than high carbon
steels.
(3) At a hardness of HRC = 61, 52100 steel has a better cutting performance
than 1086 steel and both are generally better than
Damascus steel. However, at HRC levels of 41 obtained with
fine pearlite or quenched and tempered conditions the Damascus
steel has slightly superior cutting performance than high
carbon steels.

Since Cliff's forum is so young, I'm not sure necro-threads are really possible, but I'll be the first to try. A fairly diverse response from posting this on BF. That is a necro-thread there, so please discuss, but don't reopen w/o some good information.

[www.bladeforums.com]

Reference Paper : [www.cliffstamp.com]

There are lots of interesting points made in this article such as :

-tempered martensite has a low wear resistance than normalized steel at similar hardness

The main reason why I was interested is that it clearly shows the wear resistance of two low alloy steels 4140 and 1090 at different hardness levels so it is possible to clearly see the influence of hardness and carbide volume (carbon content) on wear resistance. Here are some sample results :

0.040 C : 50 HRC : 140 WF
0.090 C : 50 HRC : 130 WF

So at the same hardness level the WF (rate of metal loss) is about 10% higher in a 0.40 C steel vs a 0.90 C carbon steel. The 10% holds roughly at other hardness comparisons between the two. Also informative is what happens to the same steel at different hardness levels :


0.090 C : 60 HRC : 107 WF
0.090 C : 50 HRC : 130 WF

Note a 10 point change in HRC, which is huge for knives only causes a 20% gain in rate of wear. To put this in perspective, a common 1050 machete at about 50 HRC would be expected to have a wear rate only about 30% faster than a 1095 blade at the common 60 HRC.

Now of course there is more to edge retention in knives than wear resistance, but this data shows that the difference in wear resistance is no where hear what is often implied, or outright stated when people are talking about for example machete vs the larger cutlery blades.

Nice article. I'll have to sit down and read the whole thing some time. Seems to back Verhoeven's work on wootz when compared to modern steels at the same low hardness (40's HRc). IMS, those wootz blades were pearlitic.

me2 Wrote:
-------------------------------------------------------
> IMS, those wootz blades were pearlitic.

Tempered martensite basically turns into ferrite + cementite and the cementite is spheroized which isn't ideal for wear resistance. Of course this is mainly an academic point, it isn't like someone wants to make 30 HRC 5160 and would go to the trouble of austenizing, quenching and then temper extremely high to draw the hardness down. The main thing that I was interested in was the wear rates at a given hardness and at a given carbon content (in similar alloy), these are a lot lower than you would expect based on internet comments about the performance of various steels.

Here is my 2 SQ 52100 blades baton & reverse-chop toughness tests video. 62rc & 63rc; 10dps; 0.01 & 0.012 behind edge thickness. I gather ht params for my next chopper.
video: [youtu.be]


edit: wow, just finished reading Me2's excellent thread on BF. 2012 is pre-dated my knife making rabbit hole.



Edited 1 time(s). Last edit at 02/07/2015 11:05AM by bluntcut.

Quote
bluntcut
Here is my 2 SQ 52100 blades baton & reverse-chop toughness tests video. 62rc & 63rc; 10dps; 0.01 & 0.012 behind edge thickness. I gather ht params for my next chopper.
video: [youtu.be]


edit: wow, just finished reading Me2's excellent thread on BF. 2012 is pre-dated my knife making rabbit hole.

I feel as though I keep asking this, but what did I say? I have a few on there, but few would be described as excellent, at least on BF and particularly in the Hinderer forum.

"Properties of High Strength Steels by B. Johansson, et. al." refers to the following :

-1095
-420
-UHB SS 716 (420HC)
-ASI 301 (non-heat treatable steel, usually used cold rolled)
-17-7 PH (this is a very low carbon, precipitation hardening stainless)

This is the hardness that they were compared at (converted from HV) :

-1095 : 53
-420 : 51
-716 : 54
-301 : 43
-17-7 PH : 46

Now there are some interesting points such as UHB SS 716 has by far the smallest number of inclusions, followed by 1095 and then there is a big jump to the rest. This is an indication of the quality of the steel and is set not by the type specifically but by the tolerances on it by the maker/manufacturer

-the tensile strength is very similar between 1095, 420 and SS 716, the other two are much weaker

-at a given tensile strength, the two martensite stainless steels are tougher than 1095

-the other austenitic steels are tougher again, but can not reach the same level of strength

-420J2 is significantly tougher than UHB SS 716 (again similar to 420HC)

The reason for this is "...due to the larger amounts of chromium carbides in the UHB SS 716. When the material deforms plastically, voids are initiated in the carbide-matrix interface giving rise to a more favorable path for crack propagation."

Of course they also look at wear resistance (abrasive, the steels are exposed to aluminum oxide and SiC3) and the ranking (from least to best) :

-17-7 PH
-AISI 301
-1095
-420
-SS 716

The differences get smaller as you move down the list. The paper also looks at things like fatigue, creep and other properties of the steels such as blankability.

Interestingly, it might not be the obvious choice that for example for a rough use type of knife that making it from 1095 and under hardening it makes a superior choice to a 420 or 420HC type steel. It certainly isn't true that in general that 1095 would have superior material properties.

Now care has to be taken to not infer from this paper that 1095 could not be a better knife steel than 420. 1095 is capable of reaching 66/67 HRC, of having ultra-fine austenite grain and thus would have much higher strength, wear resistance and apex stability than it was shown in that paper. The same change can not be said of 420 or even 420HC.

Papers :

-Verhoeven et. al. : [www.bestexcompany.com]

-B. Johansson, et. al. : [docs.lib.purdue.edu]



Edited 2 time(s). Last edit at 02/10/2015 08:48AM by CliffStamp.

This is an interesting thesis on AISI 420, the first few chapters are available for download, the rest can be obtained at any library as all thesis documents once submitted are public reference :

[repository.up.ac.za]

There is a lot of interesting data as he explores the HT in full open detail, showing how for example 420 HC easily reached an as-quenched hardness of 61 HRC without a cryogenic treatment (austenization temperature of 1120C).

It is somewhat interesting to read how Buck (or Bos) talks about HT of 420 compared to the full and open detail given in this thesis. Of course there is much more detail here, but mainly the completely lack of anything but data. It is very simple - do this, get that. Not "Bos Treatment" or similar which implies that Bos himself is somehow influencing the result and it isn't the simple rate and points of heating and cooling.

It also give the data of the very low primary carbide volume, and the micrographs in the as-quenched condition showing no large primary aggregates. The carbides appear visually to be 1-2 microns in size.

In short, it shows that AISI 420 (with the higher carbon content usually called 420HC) is an excellent steel for knives if treated properly and can give above 60 HRC in hardness, very fine aus-grain, very low primary carbide, and of course high corrosion resistance and toughness for a stainless steel.



Edited 1 time(s). Last edit at 02/10/2015 09:39AM by CliffStamp.

Terrific, valuable work. Thanks for linking to it, Cliff.

She ended up fulfilling the brief and giving a good, clear workable protocol for the target hardness and finely distributed carbide structure. I know every thesis, if not every project, leaves avenues for further work since they all come to an end at some time, but I do have to wish that she had tried some lower tempering temperatures.

At the given austenizing temps the ASTM grain size had already started to blow out from over 9 to between 5 to 7. The reason given for not pursuing any of the lower austenizing temperatures was that the hardness was too low after tempering at 550 degrees. Otherwise somewhere around 1100 degrees looked pretty promising, particulary if given cryo to drive the retained austenite down as the later runs were.

In the end she settled on the lowest austenizing temperature and tempering temperature of the final runs, limited by the onset of grain boundary carbide precipitation at higher ones. While the end result would I am sure do a very fine job and probably be well infront of some widely promoted and less suitable materials, I can't help but wonder if there is some more optimizing left in there yet.

Wilson used to use this in his kitchen knives, I have one of them and bought it for a friend. This is a very similar steel to the Sandvik cutlery grades, same properties, high corrosion resistance, high toughness, very low carbide volume, high edge stability, etc. . The main problem with it is that because it is so inexpensive it gets used in very cheap knives which get very cheap treatments. It also gets lumped in with AISI 420 as the HC version is much different in behavior due to the much higher hardness possible. But a lot of people pick up a AISI 420 stainless fantasy knife which can be 45-50 HRC and think that a 420HC blade from Wilson or Buck would be the same thing, but as noted this can be hardened up to 61 HRC with very small amounts of retained austenite (<5%), little to no primary carbide volume, no aggregate, etc. . It is basically the stainless equivalent of 52100 .

This is not likely to every catch on because of technical difficulty, however it does allow for some interesting applications in zone-hardening stainless :

-[journal.issiran.com]

It shows how you can laser harden 420 steel (very low carbon version) to ~60 HRC and provide high abrasion resistance (vs not doing it). The potential for marketing laser-hardened is also amusing.

a quiet evening here at the office, so i have wandering around in internet land and found:
a discussion of all tool steels. good explanation of normalizing. has reference charts with normalizing and hardening temperatures
[www.isoflama.com.br]
role of vanadium in steel
[hsla-v.org]
good overview of all elements used in making alloy steel
[www.metallurgvanadium.com]
a recipe for super quench
[www.artmetal.com]



Edited 3 time(s). Last edit at 02/12/2015 02:25AM by oldsailorsknives.

Recently, for some odd reason, a lot of metallurgy has been ignored in the promotion of certain steels, m390 and elmax are the worst examples. These are mold steels, designed for high wear on abrasive materials where corrosion resistance is required. They are not designed for cutting tools, and more importantly they are not optimized for abrasion resistance and toughness as is often implied.

A little bit of history, when Crucible came out with S60V (440V) it set a very high standard for abrasion resistance for a stainless mold steel. However after a significant amount of research Crucible (and others) noted that if the carbide volume was maximized for MC type and the chromium rich carbides were minimized then a similar wear resistance could be obtained at a higher toughness.

This is because toughness is correlated to total carbide volume, but wear resistance is correlated to the hardness of the carbides. and thus MC carbides can produce a higher wear resistance and increase the toughness at the same time.

Now m390 and elmax do not do this, they instead use a very high amount of chromium vs the very high vanadium in S90v, what is the effect of this, they have a much LOWER wear resistance to toughness ratio. To be specific, at the same toughness at S90V they have about half the wear resistance, they are not in the same class.

This would be expected because m390 and elmax are similar S60V modified to lower the vanadium content. The only advantage they have over S90V (for high carbide steels) is that they have higher corrosion resistance - otherwise they are simply easier to manufacture and have an inferior wear/toughness ratio - again which they have to have by basic metallurgy.

Note as well, in regards to the low toughness of S60V, the toughness of elmax, m390, s90v etc. is not a tremendous amount increased over S60V. for perspective, the difference is LESS than the difference between D2 and A2. It is hardly the case that one of these can be an extremely tough steel and the other extremely brittle.

Thus if you want a steel which has a much lower wear resistance with slightly increased toughness over S60V, then m390 and elmax can be a solution -or- if you want a steel which has similar toughness, much less wear resistance but higher corrosion resistance than s90v, they can also be a solution.

A few references :

S60V sheet : [www.alphaknifesupply.com] and [www.crucibleservice.com]

S90V sheet : [www.crucible.com]

3V sheet : [www.zapp.com]

Research on plastic mold steels : [www.kau.se]

And I rather liked s60v, though I havent tried m390 or elmax.

The history on S60V has seen a lot of revisionist alteration as well, claims now that it was difficult to harden - it has the same response as 440C, which is even stated in the original data sheet. A soak at 1950, oil/salt, gives 58 HRC after a low temper without extending the quench to cold, doing so gains 1-2 HRC points, again the same as it does in 440C.

The advantage of S60V due to the PM nature and more importantly the very high vanadium is that it can tolerate a higher soak and thus a 2050 soak, +oil/cold can give 62 HRC and an extremely high wear, very high corrosion resistance. However that requires a lot higher QC than most would have because as you increase temperature you have to have much less slop in the temperature/time and you absolutely have to use a forced quench and cold to prevent issues with precipitation in cooling and retained austenite.

The problem is that there is too much metallurgy=kryptonite in the knife industry. For those that used it as it was designed to be used, the performance was extremely positive, the most extreme would be Tom Mayo who used it in skinners for hunting guides and it was the highest performing blade he seen, quoting blades commonly lasting a season and literally dozens of animals. Now to be clear, the blades were extremely dull at the edge of the season, this isn't shaving, not even paper cutting, they were worn smooth.

I thought I posted this here, but I guess I didn't....




Been wanting to do this for awhile, a direct straight edge retention comparison between these two steels. They are very close in performance, closer than some would believe. Both knives are CF Sprint run Military's sharpened to 400 grit Congress Moldmaster finish at 15 degrees per side. This leaves IMO the optimal edge finish for maximum edge retention and cutting efficiency.

The Test:

Slicing 5/8" Manila rope checking every 20 cuts for down force and cutting was continued until 20 LBS was reached. Starting force was 11 LBS for both knives and after 20 cuts both knives were at 14 LBS.

Hardness as tested.

M390 - 61 HRC

S90V - 60 HRC

The Steels:

M390

C - 1.9%
Mo - 1.0%
Si - .70%
W - .60%
Cr - 20%
Mn - .30%
V - 4.0%

CPM S90V

C - 2.30%
Mo - 1.0%
Cr - 14.0%
V - 9.0%

M390

Cut aggressive with the edge finish and took some edge damage as expected as all steels have, it was still sharp enough to slice printer paper after testing.







CPM S90V

Very Aggressive with the edge finish and as with M390 it took some edge damage, but slightly less. Sharp enough to slice printer paper after testing.








Knives together.








Results:

Both knives performed extremely well as expected with S90V over M390 in the end by 21% or 460 to 380 cuts.

Very close as I expected it to be, under normal use most wouldn't be able to tell them apart depending on actual use unless they were pushed to the limit of edge retention. In that case the added Vanadium content of S90V at 9% over M390 at 4% would take over.

Somehow I have a feeling that only a limited audience would accept this, even with the test results provided, on a larger forum.
It is interesting however that most often Elmax is seen in the rank of S30V, S35VN and 154CM

Quote
styx
It is interesting however that most often Elmax is seen in the rank of S30V, S35VN and 154CM

That is because that is where it does rank according to the actual materials properties thus in general you would expect the majority of people to have that experience with the normal variance expected.

It isn't surprising that they would not be significantly different in Jim's work because of the precision of his measurements. If you are measuring sharpness on a scale by cutting the hemp then the most you can get is about +/- a pound. Even if the scale is more precise, the force can change by that much due to the hemp itself and even during a single cut it will vary by a lot more than a lbs. Normally you would take the peak force during a cut as the force measurement.

Given that your precision is thus a pound or so this means that your final stopping point of 20 lbs is actually the range from when the knife first started to hit 20 lbs until the moment all cuts were over 20 lbs. Because the cutting is done to a very low sharpness, about 10%, this is in the plateau region and would be much > 100 cuts, i.e. the steels would have to be VERY different to separate with such work.

Even this is a over estimate as you will also get variance from the hemp itself, the initial sharpness, the speed of the cutting, etc. and other issues, so a more realistic estimate is about twice that. If you do repeated trials and use a centering statistic you can reduce that, but given you are in the plateau your precision will always be very low.

To give you some perspective, even CATRA machines can not tell apart very small difference in steels because even the tolerance on the initial sharpness from the jig/machine alone can make a 10% difference.

Quote
styx
Somehow I have a feeling that only a limited audience would accept this, even with the test results provided, on a larger forum.
It is interesting however that most often Elmax is seen in the rank of S30V, S35VN and 154CM

That's likely due to the finial hardness being kept in the 59-60 range so yes ELMAX, S30V, and S35VN will be close enough to each other given that.

The advantage of ELMAX really starts once it's taken over 60 HRC getting into that 61-62 HRC range, that's speaking about edge retention. There is a rather large difference in performance from that 59-60 range to 61-62 range, that's having tested knives in ELMAX from 58.5 HRC to 62 HRC... Tested hardness.

Also look at MOST of the knives available in ELMAX, they aren't exactly efficient cutters as thick as they tend to be at this point anyway, that also hurts the performance.



Edited 3 time(s). Last edit at 12/05/2012 02:03PM by Ankerson.

Quote
Ankerson

The advantage of ELMAX really starts once it's taken over 60 HRC getting into that 61-62 HRC range, that's speaking about edge retention. There is a rather large difference in performance from that 59-60 range to 61-62 range, that's having tested knives in ELMAX from 58.5 HRC to 62 HRC.

Jim there is no significant difference in materials properties in Elmax in that range, neither in resistance to deformation/rolling, nor in wear, there could be no significant difference in edge retention which is thus dependent on those properties. This is born out by the fact that CATRA data is available on exactly that and the difference is so small that a CATRA machine can barely detect a difference. Are you really saying that any normal person using two knives can detect a difference which of a range comparable to the limits of a CATRA machine and further that is a "rather large difference". Note as well that the CATRA results are only able to be detected again when the sharpness is extremely low and the force and/or length of blade being used to make the cut is 10X that of a sharp blade.

Quote
CliffStamp
Jim there is no significant difference in materials properties in Elmax in that range, neither in resistance to deformation/rolling, nor in wear, there could be no significant difference in edge retention which is thus dependent on those properties.

What are your sources?
(honest question, I've done a quick search and haven't found much except the Uddeholm fact sheet which has limited data regarding 60+ hardness).

Cliff

Would you take into consideration buying yourself a Blue M390 Paramilitary2 from BentoBoxShop ? as its coming out again. I think you are wrong, and that M390 wouldnt have half the wear. I know its less wear resistant, but I would think only by 25-15 % . I reckon the steel will surprise you. It takes a kickass aggressive edge, its a really nice performer.

I have found m390 to hold an edge longer than my S30v and Superblue. Now elmax compares to s30v at lower hardnesses of 59-60 and should be similar to Superblue at 61-62 , looking at Jim's data.

I haven't got any S90v yet, but once I do, ill be comparing the steels, though it might be an unfair test,what ever S90V knife I get would have to be thin behind the edge or reground as thin as my reground Para to be a good comparison. Id be testing a 10/side edge with a 600 grit microbevel a 15 degrees. I should be able to afford something in S90V when I have a job. (Making full time pay , not paying rent- ill have a fair amount of spare money)