I Tested the Edge Retention of 48 Steels

I have this book in my Library, frankly, I forgot I had it. I am used to eBooks and Audiobooks, so hard copies slip my mind. I'm definitely going to move it to the sharpening bench.
 
kasumi said:
You are right. I did not say that Geometry is the only factor making a good knife . I only wanted to say that Geometry for me is the leading factor. Of course when all contributing factors to a good knife for you come together it is always better!

When the steel is in such bad condition as you describe only the waste bin remains as the last resort.🥴.

By the way ( speaking about Victorinox) I give to anyone on a tight budget who is asking me what kitchen knives I would suggest the Victorinox FIBROX serie. A good example of cheap good stuff if you do your part in the geometry/ sharpening/stropping area. Even a much better suggestion for many people than a 400 dollar Japanese “ high quality , high hardness” steel knife.
Click to expand...
Another issue with prioritizing geometry is that the end user can much more easily change the geometry of their knife than they can to heat treat. It is relatively straightforward thinning a knife or adjusting the bevel to make it cut better, however trying to reheat treat the blade would be significantly more challenging and out of the realm of most people's capability.

So while both are obviously important it could be argued that heat treat is the more important quality simply due to the fact that it's the one that is harder to change after the fact.
 
Last edited:
Been interesting reading the results. Bob Dozier D2 proved heat treat make a difference. I have experienced D2 out of China vs EU Manly at 60ish whom had probably the best value knives till they stopped. I'd say better value than even Chinese whom stop at 58hrc with their native D2. Points snap and chipping is the thing here. Geometry.

10Cr15CoMov often sold as VG10 and all sorts of others has some ductility. I made a hollow internal tang but due to surprising flex at 60 used the solid tang instead. Has D2 beating edge retention in my layman hunting experience and sisal rope. At least better than vs Chinese D2 at 58.
Had possession of a Chinese D2 sample and it was fragile brittle. Honey Badger out of China that has 58 D2, it's point can snap.

Now the Manly point. A clip point 3 mm spine could bend. But also many had blade snap but leverage force is contextual. Saying that though it could cut and was very popular for those use to .5 carbon blades, a revelation.

D2 really wasn't meant as a blade steel. It was well priced and Dozier managed to get it to work. From there got noticed.

Besides edge retention, balancing steel flex so point that won't easily snap is as important for hunting use. You buy a profile, you want to keep that profile. Protruding drop or spear point for penetration with belly for slicing needs a balance of edge retention and blade toughness.
 
Last edited:
ALLHONE NZ said:
Been interesting reading the results. Bob Dozier D2 proved heat treat make a difference. I have experienced D2 out of China vs EU Manly at 60ish whom had probably the best value knives till they stopped. I'd say better value than even Chinese whom stop at 58hrc with their native D2. Points snap and chipping is the thing here. Geometry.

10Cr15CoMov often sold as VG10 and all sorts of others has some ductility. I made a hollow internal tang but due to surprising flex at 60 used the solid tang instead. Has D2 beating edge retention in my layman hunting experience and sisal rope. At least better than vs Chinese D2 at 58.
Had possession of a Chinese D2 sample and it was fragile brittle. Honey Badger out of China that has 58 D2, it's point can snap.

Now the Manly point. A clip point 3 mm spine could bend. But also many had blade snap but leverage force is contextual. Saying that though it could cut and was very popular for those use to .5 carbon blades, a revelation.

D2 really wasn't meant as a blade steel. It was well priced and Dozier managed to get it to work. From there got noticed.

Besides edge retention, balancing steel flex so point that won't easily snap is as important for hunting use. You buy a profile, you want to keep that profile. Protruding drop or spear point for penetration with belly for slicing needs a balance of edge retention and blade toughness.
Click to expand...

I'm not generally a fan of Chinese D2. Its edge retention really shines when used as a replacement for 8Cr13MoV in budget knives. Some companies do a better heat treatment on it than others but still... The answer to "why bother" is what you mentioned here. "D2" had cachet, a popular reputation that still drives sales. "VG-10" might not be as hot as "D2" but 10Cr15CoMoV is definitely the better steel. The edge retention is more consistent and often better. 10Cr15CoMoV is also tougher and a whole lot more stainless.

Heck, even 9Cr18MoV with a good heat treatment can hold a noticeably better edge. It's less tough than 10Cr15CoMoV but has even better corrosion resistance. The older Civivi knives in 9Cr18MoV were awesome for the price, and the Sencut knives running 9Cr18MoV still are. Spyderco has finally started using 9Cr18MoV in their Chinese-made knives but they're charging shockingly high prices for the knives that use it.
 
Probably Taiwan made. But Taiwanese companies are deep into Chinese factories even their own there. The CNC machine like Colt 45 made everyone equal so long as tool change and quality control is performed.

America is always going to have the new item for hunting because Charlie don't surf or hunt to quote Apocalypse now. With robotics making robotics its the end of China as being the lowest cost provider. And that's good, initiative and inventive culture is American.

e.g. Bob Dozier's D2 and Bob Loveless drop point. Knife changing events.
 
Last edited:
Is there going to be an addition of Magnamax ratings? I see Spyderco is rolling it out now on their manix knife.
 
stillshot2 said:
Is there going to be an addition of Magnamax ratings? I see Spyderco is rolling it out now on their manix knife.
Click to expand...
From the significant amount of data Larrin has posted, I would say

Tougness: 6
Edge Retention: 8
Corrosion Resistance: 9.5
 
Been reading this thread on and off for years — it's still the reference I point people to when someone asks whether a steel is "good."

I built a steel comparison matrix a while back and this thread is a big part of why I kept second-guessing it. Mine covers 83 steels across four axes — edge retention, toughness, corrosion resistance, sharpenability — scored 0–10 from published manufacturer figures and HRC ranges. It's free on the homepage of choil.app, no login, no signup.

It's a fundamentally different thing from what Larrin did here, and I want to be clear about which is which. This thread is measurement — he ran the CATRA machine, heat treated the steels, made the knives, and produced numbers that are actually true about actual blades. Mine is orientation — a rough map of where 83 steels sit relative to each other, so someone who's never heard of Vanax can see the neighbourhood it lives in. Narrow and empirical versus broad and approximate. His tells you what happened; mine tells you where to look. If you want the real depth, Knife Engineering is the answer and nothing I've built competes with it.

Where I'd like the criticism: the matrix collapses each steel into one number per axis, but heat treat moves the target enormously. S30V at 58 and at 61 aren't the same steel, and this thread makes that point better than anything else I've read. So — does a single score per axis actively mislead people? Should the axes show a range instead of a point? Or is a four-axis chart just the wrong shape for steel data?

I'd rather hear "this is misleading, fix it" from you lot than keep guessing.


1785545249626.png
 
Pat@Choil said:
Been reading this thread on and off for years — it's still the reference I point people to when someone asks whether a steel is "good."

I built a steel comparison matrix a while back and this thread is a big part of why I kept second-guessing it. Mine covers 83 steels across four axes — edge retention, toughness, corrosion resistance, sharpenability — scored 0–10 from published manufacturer figures and HRC ranges. It's free on the homepage of choil.app, no login, no signup.

It's a fundamentally different thing from what Larrin did here, and I want to be clear about which is which. This thread is measurement — he ran the CATRA machine, heat treated the steels, made the knives, and produced numbers that are actually true about actual blades. Mine is orientation — a rough map of where 83 steels sit relative to each other, so someone who's never heard of Vanax can see the neighbourhood it lives in. Narrow and empirical versus broad and approximate. His tells you what happened; mine tells you where to look. If you want the real depth, Knife Engineering is the answer and nothing I've built competes with it.

Where I'd like the criticism: the matrix collapses each steel into one number per axis, but heat treat moves the target enormously. S30V at 58 and at 61 aren't the same steel, and this thread makes that point better than anything else I've read. So — does a single score per axis actively mislead people? Should the axes show a range instead of a point? Or is a four-axis chart just the wrong shape for steel data?

I'd rather hear "this is misleading, fix it" from you lot than keep guessing.


View attachment 3243798
Click to expand...
Looks interesting. Bookmarked for later when my eyes aren't used up!
 
Pat@Choil said:
Been reading this thread on and off for years — it's still the reference I point people to when someone asks whether a steel is "good."

I built a steel comparison matrix a while back and this thread is a big part of why I kept second-guessing it. Mine covers 83 steels across four axes — edge retention, toughness, corrosion resistance, sharpenability — scored 0–10 from published manufacturer figures and HRC ranges. It's free on the homepage of choil.app, no login, no signup.

It's a fundamentally different thing from what Larrin did here, and I want to be clear about which is which. This thread is measurement — he ran the CATRA machine, heat treated the steels, made the knives, and produced numbers that are actually true about actual blades. Mine is orientation — a rough map of where 83 steels sit relative to each other, so someone who's never heard of Vanax can see the neighbourhood it lives in. Narrow and empirical versus broad and approximate. His tells you what happened; mine tells you where to look. If you want the real depth, Knife Engineering is the answer and nothing I've built competes with it.

Where I'd like the criticism: the matrix collapses each steel into one number per axis, but heat treat moves the target enormously. S30V at 58 and at 61 aren't the same steel, and this thread makes that point better than anything else I've read. So — does a single score per axis actively mislead people? Should the axes show a range instead of a point? Or is a four-axis chart just the wrong shape for steel data?

I'd rather hear "this is misleading, fix it" from you lot than keep guessing.


View attachment 3243798
Click to expand...
My main question is where do the ratings come from? It shows Magnamax as only having slightly more edge retention than Magnacut, which is wrong. Were these scraped from other ratings sites? Because they're mostly made up.

Also, I never understood the sharpenability ratings that everyone does. It should essentially be the same as edge retention.

Taking into account the heat treatment is important but difficult to do with respect to toughness. You ideally always want to compare steels at the same hardness. I made a simple model that predicts the characteristics of a steel as the hardness and edge angle changes, but it's just a guess.
 
Synov said:
My main question is where do the ratings come from? It shows Magnamax as only having slightly more edge retention than Magnacut, which is wrong. Were these scraped from other ratings sites? Because they're mostly made up.

Also, I never understood the sharpenability ratings that everyone does. It should essentially be the same as edge retention.

Taking into account the heat treatment is important but difficult to do with respect to toughness. You ideally always want to compare steels at the same hardness. I made a simple model that predicts the characteristics of a steel as the hardness and edge angle changes, but it's just a guess.
Click to expand...
 
Fair

On where the ratings come from: nothing is scraped from other ratings sites. Every entry is hand-written, normalized 0–10 from maker datasheets (CPM/Böhler), published HRC ranges, and third-party CATRA results where they exist. But you found the weak spot in that method. MagnaMax has no Crucible datasheet yet — the composition isn't even public — so the original entry leaned on announcement copy and under-credited the wear resistance. You were right.

It's corrected as of tonight, re-derived from the designer's published testing and calibrated against the matrix's own anchors rather than anyone else's charts: edge retention now sits level with 10V and just under S110V on my scale (up from barely-above-MagnaCut), toughness came down to the S35VN/Vanax tier where the published comparisons actually put it, and ease-of-sharpening dropped to reflect that carbide load. Live on the site now if you want to re-check my work — genuinely, please do.

On sharpenability: you're mostly right there too. As scored it's largely inverse wear resistance, since carbide volume and hardness dominate how a steel behaves on stones. I've kept it because "how annoying is this to maintain" is the question casual collectors actually ask, but I agree it carries little independent signal, and I'm considering replacing it with an axis that earns its spot.

On hardness: agreed, and that's where this thread has pushed me — pin scores to a reference hardness and show the HRC range explicitly, rather than smearing one number across whatever hardness different makers run. Your model is a good example of the measurement side of that line I was trying to draw.

This is exactly the "this is misleading, fix it" I was asking for. Keep it coming.
 
You must log in or register to reply here.
Back
Top