Episode 89 of GGL had Cedric and Ada, Peter, on to discuss lots of stuff, but his steel test insights need to be put into writing for future reference. There are two parts to this, the method and the results. Peter wrote the bulk of this post. I wrote the Editor's Take.
My edge testing shows how various steels retain absolute sharpness at the fine cutting edge vs man made materials, in this case, fibrous rope. The test produces a number for each steel of rope cuts made, so steels can be placed in a hierarchy where the interested person can get a decent idea of where one steel might sit against another.
I take the knife and put a 20 degree edge on each side of the bevel using the Worksharp (basic model) 'kitchen knife' guide. This makes a 40 degree inclusive microconvex edge on the blade. I use the brown (mid grit) and then the blue (fine/honing) belts. I get the blade hair popping sharp and ensure it can easily slice a held piece of paper.
I use the same 'Grunt' brand of rope every time. Its called 10mm twisted sisal rope. I buy it in 10 meter hanks. I've probably used 300 meters or it so far (the Maxamet cut one and a half on its own if I remember correctly!)
I then use the forward ("belly") portion of whatever blade I am using and cut the rope in small pieces, stopping to test if it still slices the paper at increments that generally feel right for the steel. For example, on a non PM steel I generally check every 20 cuts or so. With Maxamet it became pretty apparent I had something special, and only had to check every 100 cuts. As the knife feels like it is slowing down I check again and then incrementally cut and check. Once the knife no longer slices held paper I stop and record how many cuts it took.
The constants in these tests are:
*The edge angle
*The medium cut
*The board under the cutting medium
*The person doing the cutting (less important)
The controlled variable (the thing being tested) is the blade steel
The uncontrolled variable which in my opinion keep these test in the realm of more 'supporting data' rather than 'ironclad stats' is Rockwell hardness (the harder the steel the higher the edge retention will generally be. 420HC at 58 will last longer than 420HC at 55, for example.
This is due to two things - often makers wont publish the Rockwell or some steels don't go as low and others don't go as high.
Ideally, every steel would be at 60, but alas this isnt the case!
I think the testing in a general sense lines up enough with peoples anecdotal experiences as well as other accepted means of testing (CATRA Testing and other cardboard cutting tests as seen from Michael Christys work as well as that of mysterious and now defunct Youtuber with the handle 'BLUNTTRUTH4U') that the numbers I get are fairly reliable. Therefore, when there is a notable anomaly such as with my pair of ZTs with lower than expected performance on s35vn, it is worth extra investigation.
Overall I would definitely note that as long as your steel has a purpose to its composition it still has value, despite what the edge retention is. I would rather CTS-XHP with its moderate stainless properties, manageable sharpenability, and great edge retention over Maxamet with its rustprone nature and sharpening difficulty despite its huge edge retention. It is only when steels are exceptionally low in retention without 'winning' in other attributes that I would steer clear, or at least not pay too much for. For example, 8cr13mov will neither hold and edge for long or resist rust particulrly well, leaving only decent toughness and low cost as its benefits. I would therefore not hold its value as high as even humble Victorinox stainless which has low edge retention, but has high stainlessness, toughness and sharpenability. As we said, you can generally either pick two of three attributes for a decent price, or pay a lot for all three to be covered adequately.
Another thing I would note is that importance of edge retention to the end user is generally linked with how confident a sharpener they are. Personally, I used to hate sharpening before I got the means of doing it easily via a Tormek T4 system. I was never good at freehand stone sharpening so the high edge retaining steels were desirable to me as I would generally use the service of a professional, and high edge retention meant longer time between appointments! Now I have the ability to do decent to very good edges at home (although they still look ugly sometimes) I am happier to settle further down the line for more balanced steels.
I also note that steels like M4 and 3V that maybe scored lower than some folks would expect, are directly in the centre of the 'being great on multiple fronts' venn diagram of sorts. Frankly, steels like M4 and 3V, steels that can be so tough as well as have very good edge retention, are as impressive to me as Maxamet. M4 is pretty much the bladesports steel of choice for chopping 2x4s and toilet rolls. LC200N, also, with its great edge and supreme stainlessness. As well as Paul Bos getting such seemingly large improvements from basic steels just by knowing his stuff. That's where the real wizardry is! But then, to quote the Nick Shabazz - I am just some jackass on the internet!
Results (in Descending Order):
CPM 10V: 389
CTS 40CP: 285
CPM S110V: 281
CTS XHP: 242
CPM M4: 224
CPM S35VN: Kizer Gemini: 198, Chris Reeve Sebenza: 187
CPM 3V: 198
Laminated CoS: 176
CPM S30V: 175
CTS BD1: 112
CPM S35VN: 109/113 (ZT)
1095 Cro Van: 87
1095: 75 (ESEE), 69 (Schrade)
420HC (Paul Bos Treated): 67
Junk Steel: 4-6
Butter Knife Steel: 3
A2 (LT Wright Knives Genesis): 216
Mora Carbon: 133
Mora Stainless 12c27: 88
Average (excluding scandi grinds): 149
Steel Closest to Average: Niolox at 135
33% Score (below this is a steel that scores a zero): 187 (33% of Maxamet's 563); closest is 3V at 198.
66% Score (below this but above 33% is a steel that scores a 1): 372 (66% of Maxamet's 563); CPM 10V at 389
This shows how good Maxamet is at this one particular thing. To normalize the numbers to what's readily available on the market, I averaged the top five steels (average is 381) and ran the numbers again.
33% Score of Average of Top 5: 126; closest steel is 154CM at 124.
66% Score of Average of Top 5: 251; closest steel is CTS-XHP at 242
I think even these numbers are skewing a bit high because of Maxamet. I am comfortable with CTS-XHP representing the threshold for a steel that scores a 2, but 154CM feels a bit too good for the upper threshold of a score of 1. I still think 154CM, when done correctly, is a 2 steel as are other steels here that score relatively low. That is because this is a measurement of only on of the three major aspects of steel--hardness/edge retention. 154CM does well because it is relatively hard, relatively tough, and corrosion resistant. Either way, though, these data indicate that I need to be more harsh on my steel scores going forward. The market has changed, as Maxamet shows.
I really detest VG10. I think it is one of the most overrated steels in the knife business and these tests bear that out. They also bear out the intuitions I had about Bos-treated 420HC. I am pleasantly surprised at how good ZDP-189 does especially compared to its relatively lower score on the CATRA tests, which are more scientific but less of a test of real world performance (as the well-know problem the CATRA test has with titanium nitride coated blades proves). Notice how variable steels can be based on manufacturers--the aforementioned 420HC steel compared to regular 420HC and all of different scores for S35VN. A couple of steels look like a good buy--CTS-XHP, BD-1, A2, M4 (of course), and ELMAX. Take a look at that Elmax haters. It does well, surprising isn't it? No, not really. I am also surprised at how relatively low S110V scored.
It bears special mention--A2 outperformed 3V. Now I know that 3V is a newer steel, the grinds tested were different, and 3V is amazingly tough, but the results here are surprising. The results show you A2's potential more than 3V's weakness because the grinds are different. I thought for sure that 3V would lap something like A2, but A2 has proven over the years to be a damn good steel and the results bear that out. It would be very interesting to test apples to apples and see how 3V would do in a Scandi grind.
I'd love to see two things with Peter's tests going forward--new steels and a standardized main grind. I'd love to see how AEB-L would do. I'd also like to see results for SM100, Sandrin's Tungsten Carbide, TM Hunt's O1, INFI, Cru Forge V, W2, SG2 and 20CV. The other thing that would be extra fun would be to control for the main grind as well as the cutting edge. It would be really fascinating to see how a bunch of Manix2, PM2s, Mules, or Kershaw Blurs do with different steel.
Thanks to Peter for this information and coming on the podcast. I love numbers like these.