I Tested the Edge Retention of 48 Steels

A lot of data but are this data reliable?
I have VG10(Chroma) and Aogami Super (Eden) chef knives. Same sharpening angles. 20 vs 23cm. AS holds the edge at least 50% more.

Than I have small knives VG10, 1095, 1086(opinel) and D2(lionsteel). Same sharpening angle but different blade thickness. I do not see too much difference in edge retention, D2 is a little better but if it touches bone or metal it needs resharpening on diamond stone. the others just need few passes on steel.

I don't know what to say about AEBL. I see it compared to victorinox. But victorinox is significantly worse than anything else that I have. It is close to 420 from my experience. Few times worse edge retention than D2/AS.

Toughness I would say 1086, AS, 1095, VG10, D2 in this order, with D2 far from the rest.
 
p.george said:
A lot of data but are this data reliable?
I have VG10(Chroma) and Aogami Super (Eden) chef knives. Same sharpening angles. 20 vs 23cm. AS holds the edge at least 50% more.

Than I have small knives VG10, 1095, 1086(opinel) and D2(lionsteel). Same sharpening angle but different blade thickness. I do not see too much difference in edge retention, D2 is a little better but if it touches bone or metal it needs resharpening on diamond stone. the others just need few passes on steel.

I don't know what to say about AEBL. I see it compared to victorinox. But victorinox is significantly worse than anything else that I have. It is close to 420 from my experience. Few times worse edge retention than D2/AS.

Toughness I would say 1086, AS, 1095, VG10, D2 in this order, with D2 far from the rest.
Click to expand...
The data is reliable. It is a test of edge wear. Whether it directly predicts edge retention depends on the dominant dulling mechanism in your use case. For example, perhaps the higher hardness of the Blue Super in your knife is leading to less deformation. Or it could be your VG10 knife has poor toughness and is micro-chipping. Maybe your sharpening angle is the same but thickness behind the edge is different. Maybe your knives aren’t 50% different but your bias is making you think so without carefully controlled comparisons. Could be lots of things. As noted in the article, I think Blue Super tests lower than it could because of the abrasive used. But if we assume that hypothesis is correct, Blue Super would be similar to VG-10 in wear resistance, not 50% better. So it must be some other factor leading to a difference.
 
Excellent! Last year, I did a much smaller practical only (no precision measurement - just feet of planing and weight of shavings planed) test with plane irons, testing:
O1
A2
V11 (XHP as far as I can tell)
M2
CPM M4
Japanese Blue Steel
Ward laminated (this is not going to interest knife people, but it's probably a lot like 1084-1095)
CPM 3V

My results largely parallel yours. I found this site today because I'm going to make an iron or two out of 52100 to address a question that the woodworking community asked "why don't we see 52100 in plane blades if it's used by knife makers? Would it be better?" (O1 is common, and by alloying, I expect a same-hardness test will put 52100 somewhere near O1 - 63Rc is where my test O1 iron is, probably harder than it would normally be found in a knife).

The only differences I can think of in terms of results is that we can't use a plane to the same level of dullness - my point of dullness was when the plane iron was too dull for the plane to stay in the cut under its own weight, which happens rather abruptly, and how ideal planing is for edge wear vs. toughness (in continuous wood, it's not demanding at all and the blade is stabilized and supported almost all the way to the edge).

What I found was more or less that hardness and alloying controlled (I'm not a metallurgist) edge lifetime. Outside of the test in rougher work, some of the comparisons can break down.

I can imagine (Without reading this entire thread) that one of the most common responses is "good test, but X and Y in your test don't compare the way they do in my experience so the whole thing is called into question". It's always flattering to have someone say that. I used a single piece of wood for each test and rotated plane irons every 200 feet of planing, and ultimately planed about 50,000 feet (I guess that makes me the testing machine), and then took pictures of the edges with a metallurgical scope.

When the japanese plane iron didn't win (the one I got, made commercially, had bad behavior, but there was nothing about it that suggested it would've outperformed A2 by much whereas XHP, M4 and 3V (and M2) outran the others considerably. That triggered "your test results aren't reliable, nothing lasts longer than a hand forged blade", and then various other fans of various things attempted to poke holes.

However, what I found was almost in lock step with what you found in terms of relative performance. XHP landed slightly below your 204p ratio vs. more plain steels (near M4).

I enjoyed reading through your data. I did my test in isolation without knowing what to expect from any of the irons other than O1 vs. A2 more or less. Some of the results were surprising, but I didn't want to know what the results would be by reading ahead from anyone else's work. I'm glad you did your test through card stock - it makes for a good comparison to planing wood.

(i understand the importance of microstructure in tougher work with knives, but for planing wood in a hand plane, rolled material with the grain oriented properly generally will work through a test without showing anything other than abrasive wear).
 
(none of the people who questioned the data despite seeing the data trail and methods could be motivated to go repeat the test and show me a subset of data that differed substantially, though. One, they would get out of their chair to do it, and 2, if they had, they would've gotten the same results that I did. I also tested abrasives from 5 micron diamond down to 1 - we can't really use coarse edges -and found that we're probably using only what knife people would refer to as a fine edge and sharpening before spending long in a "working edge". 5 micron diamond edges fail at about 65% of the feet planed vs. 1 micron, but if a dull edge was able to be used much longer, I'd guess the data would end up looking more like yours. Any substantial dullness in a plane results in a user having to push down on the plane and things go downhill quickly at that point - a "dull" plane edge would still easily cut someone and the bevel is only really rounded on one side and depending on setup, cupped on the other (the shaving is directed into the back of the blade and the edge goes from wedge to half moon.

This is the top side of O1 (where it's "Scooped" at the edge vs rounded out. The rounding out of the opposite side of the cutting bevel due to wear is what prevents a plane from staying in a cut on its own). The very tip of the edge is obviously still crisp. (150x optical).

https://i.imgur.com/J2OMH4M.jpg
 
Always happy to hear experiences for other people doing testing. I agree it doesn’t make sense to ignore results just because we don’t like them. Discovering why something behaved differently than expected is how we learn new things. But discovery how or why is of course very challenging. Much more challenging than finding results only. I am happy to hear that push cutting had similar trends, as I still don’t know clearly how behavior changes between slicing and push cutting. I would imagine that the wearing of edges would be similar as long as rolling or chipping isn’t the dominant mechanism instead, but there may be surprises in behavior that we don’t know about.
 
Exactly - what we found in wood to some extent is that chromium carbides in large quantities (which aren't particularly prized by the folks rubbing metal on metal vs. large vanadium carbides) are pretty valuable. I didn't see any real difference in wear for PM vs. non-pm, but edge uniformity with PM (thus surface finish on wood) was more uniform.

Carbon steel in more general work has been more resistant to significant damage at the fine edge. Most of the time, though, it's as you say - just wear. No rolling or chipping (either of those in planing wood are something to eliminate as we sharpen a lot more often than "knife people", perhaps once every 20 minutes from fully dull to sharp. It needs to be quick, and uniform wear is quick to refresh. Damage, not so much. The sharpness level that we like also means no burr, and some of the very tough steels have an extremely persistent burr.

It's far more interesting to get a surprise from data and learn why than it is to dig in and argue against it for no reason (chromium carbides in large amounts but small size also lead to slickness on surface to surface contact - I haven't confirmed that anywhere in any published information, but it can be felt very easily. This may be different than knife cutting because wood has a certain level of stiffness and thus rubs the blade on the way by under some mechanical strength. If it rubs a slicker surface, that's great - especially if the slicker surface wears just as long as one less slick (3V and M4 created quite a bit of felt resistance).

Thanks for the response.
 
D-weaver said:
The only differences I can think of in terms of results is that we can't use a plane to the same level of dullness - my point of dullness was when the plane iron was too dull for the plane to stay in the cut under its own weight, which happens rather abruptly, and how ideal planing is for edge wear vs. toughness (in continuous wood, it's not demanding at all and the blade is stabilized and supported almost all the way to the edge).

.
Click to expand...
I forgot to respond to this point before. It actually doesn’t matter at which point the test ends, the trends stay the same. I looked at this before: https://knifesteelnerds.com/2018/11/26/steel-edge-retention2/
 
Larrin said:
I forgot to respond to this point before. It actually doesn’t matter at which point the test ends, the trends stay the same. I looked at this before: https://knifesteelnerds.com/2018/11/26/steel-edge-retention2/
Click to expand...

I agree. I should've identified better what I was talking about (though I might get a few points crossed as it's hours later). I was thinking about (but didn't say) your testing where you showed the relative difference in edge life for 200 grit sharpening vs. a very fine edge. In your test, whatever the sharpening method, all of the relatively differences pretty much stay the same.

What I meant to illustrate is that something as fine as 5 micron diamonds will wear much less long in my test (testing the same steel with the media - if I switched steels with any media, I'd expect the relative differences to be similar between the steels).

But, I saw a 35% drop in edge life just for starting with 5 micron diamonds vs. 1 micron (both compared on the same steel), solely because my dullness point is a lot sooner. At some point, probably around 400 grit with a strop, a plane is almost impossible to use under my test parameters because it's too dull at the state.

If I'd have tested heavy planing, I think all of the relative steel comparisons would've been similar, but heavy planing can tolerate a duller edge - there's more wood on the top side of the cut to keep the plane in the cut and the difference between 5 microns and 1 would've been minimal. I think if you were able to use your machine to call "dullness" a point 3 times sooner, the disparity between sharpening fineness would be greater in terms of feet cut coarse vs. fine edge.

I was hoping to it appearing nonsensical to say that a difference between 5 microns and 1 is really that large in edge life by suggesting it's different solely due to the dullness point in planing being a lot less dull.

I didn't test the steels against each other with 5 micron diamonds, though - just 1 micron. As it was, it took about 40-50k feet of planing, and I took pictures for most if it every 2 feet under a microscope and weighed the shavings. One of the woodworkers involved wanted to know if he could go slightly coarser with finish honing at 5 microns and hone all in one step instead of two after the grind, and the answer was no. I'd suspect a stropped 5 micron diamond edge on a knife in heavier work to feel quite fine, but in smooth planing, it's coarse and leaves a dull surface from the start.

------

The issue with fineness of the edge is that it imparts surface finish for the (very) few who want to apply finish off of a hand plane. There's probably no parallel in knives other than carving. This is a reflection off of an unfinished board where the surface is this reflective just as a matter of the quality of the cut from the plane.

https://i.imgur.com/ivsb5X8.jpg

One of the things that I ended up having to do while planing is defend the results vs. knife tests (which are often much more abusive and where the edge quality matters a little bit less. Something like S90V would make a horrible plane blade. I would expect the carbides to leave the matrix and the effect is thisL
https://i.imgur.com/kCmy8TH.jpg

That doesn't look that bad, but that's blue steel. Something is wrong. It should look like this
https://i.imgur.com/NV3aIgu.jpg

That is water hardening steel (an old Ward and Payne plane iron - those bevels are not honed bevels, they are wear - the surface is flat before them).

There were a few little nits like that (the relative edge life in the same steel. vs initial edge refinement vs. yours where the relative life is much closer) that are different than most knife tests, but as you say, once you observe them, it's a matter of explaining them. If you can repeat the same thing over and over under controlled conditions, then you can get stuck having to explain it to the people who know what they're talking about. We were a bit shocked about the edge life being that sensitive to refinement, but it's an easy test to do (and partially explained by what we call jack planing, which is much heavier and deeper into the wood seemingly going on for an hour or two before sharpening is needed. On a smoothing plane, it may be less than 20 minutes).

I think testing woodworking tools is probably easier than knives- we all use them on relatively similar media and we tend to set them up similarly (generally close to flat bevels with an edge somewhere between 30 and 35 degrees at the tip).
 
Yes a finer polish is better for push cutting. Coarser is better for slicing.

S90V would be no worse than XHP or 204P. They have a similar carbide volume but S90V has smaller carbides. S90V is more wear resistant because of vanadium carbides.
 
Slicing is something we don't do much, and if we do, it's with carving tools (where the finish needs to be good).

I'll admit that I don't know that much about S30-S125V steels - we tend to have some bias against them in woodworking, not because they wouldn't make a long wearing blade, but because there are regular occurrences in planing less ideal than my test where things like this occur.

https://i.imgur.com/vuyz0Ho.jpg

This is damage to V11/XHP from hitting an inclusion in maple. The plainer steels have two advantages with this:
1) they don't seem to chip quite as deeply
2) vs. XHP, they're twice as fast to hone and get on with the work after dealing with whatever the inclusion (or surface contaminant), etc, may be

Anything with a lot of vanadium hones more slowly when you get into needing to remove several thousandths of the edge (and grinds more slowly) - those notches are several thousandths deep. Even 3V is kind of a nuisance - it breaks rhythm in the cycle of work because even on diamonds, I found it takes about twice the number of strokes to get a fine edge and remove the same amount of damage vs. carbon steel. After testing, I immediately switched the bulk of my plane irons to XHP (it wears twice as long as O1), but ended up switching a couple of those back out due to the same issue as above - the practical plaining situation involves intermittent damage and XHP is enough slower to hone that the balance only works out for smoothing planes (where other planes have done most of the work where defects would be encountered).

But, I think in a standardized test where hitting an inclusion like this is thrown out, whatever has the most carbides and abrasion resistance to abrasives is always going to win the controlled planing test. Lee Valley did most of their testing in MDF as far as I'm aware - that provides no real challenge to edge toughness (it's softer but more abrasive). Everything else in our tests showed pretty much proportional results between resistance to sharpening abrasion and edge longevity.

One of the people who goaded me into testing is Bill Tindall. You may not remember his name, but I forwarded your study to him just to say "look, another data point that ties in with our results". He mentioned that you helped him pick 3V perhaps a dozen years ago as a good steel to use in woodworking. He and I are at odds sometimes about what kind of edge failures actually matter in the cycle of work woodworking, but I use hand tools a lot more. Bill came to mind when you said something above more or less along the lines of finding something and then having to figure out the why. I can come across something and try to fire a diversion at the start "i'm just interested in the what, I don't want to go down the rabbit hole of figuring out why right now", but bill will never stop at that.

I may be inclined to get an S90V blade made at some point in the future. I make my own irons mostly by hand, and I'm guessing it's intolerable for that. I have the ability to work something like XHP mostly by hand, but I don't have the heat or the ability to temper S90V. My setup is at it's limits to get to 1900F for XHP.

Thanks for the discussion. It gives me something to think about.
 
Separately, I've found for chisels (no clue what the parallel is for knives there -it's still less demanding than knifing) that forging older files (an older heller file, for example) makes superb chisels. Better than almost anything commercially made (chisels don't get much abrasive wear). I've also made chisels out of 1095, but the files are better - even if 1095 makes a very good chisel. Have you ever had the opportunity to XRF any steel from an older file? I'm only mildly curious, the what proves out with them, so I don't really need to know the why, but I suspect there's something in them (maybe small amounts of molybdenum and vanadium) that makes them reharden better (harder) in a quench oil than 1095, and by working them, I can tell they have more carbon (the spark while grinding them can be intense enough to create an area near the origin that looks almost like solid combustion and will instantly light card stock on fire). I'm not sure that newer files are made from the same steel (basic files, vs. the high hardness files made to file hardened ferrous materials).
 
Other people are more familiar with file steels than me. They are generally low alloy high carbon steels. I know at least some of them are ~1.25% carbon steels.
 
ErocSD said:
Impressive, I'm amazed that cruwear faired so low, below Elmax even
Click to expand...
CPM-CruWear is a great steel with an excellent combination of toughness and wear resistance. That doesn’t mean it is the most wear resistant steel however. Elmax has higher carbon, chromium, and vanadium than CruWear and therefore has more carbide and wear resistance.
 
Larrin,

That study/test is fascinating to me. Thanks again so much for providing us with so much information.

But I'm afraid you left me wondering one thing... why did I spend so much money on a Randall made of 01? :eek: :oops: :( :confused:

Stay healthy,
Bob
 
desmobob said:
... why did I spend so much money on a Randall made of 01? :eek: :oops: :( :confused:

Stay healthy,
Bob
Click to expand...

If I didn't make my own tools and knives, I'd probably prefer an O1 knife vs. most anything that's got a much higher % of carbides (unless stainless is a need). It's amenable to a very fast sharpening cycle and has a stable edge that's easy to keep in shape.
 
AEB-L is where it’s at if you like stable edges and easy to sharpen.
 
You must log in or register to reply here.
Back
Top