Anyone ever try doing edge retention testing on their own?

Has anyone ever tried doing something like this for yourself? I'd be curious to hear your thoughts.

Quite a bit, actually.
https://www.bladeforums.com/threads...-alloys-using-a-visual-inspection-tec.641279/

https://www.bladeforums.com/threads...-vs-vg10-vs-buck-s30v-vs-benchmade-d2.769447/

https://www.bladeforums.com/threads...-zdp-189-m390-and-s30v-edge-retention.792540/

There's a couple other threads that I started if you want to look for them.

Three issues:
The largest issue is that geometry has more impact on edge retention than blade alloy.
♦ You must sharpen all the blades at exactly the same angle.
♦ The overall profile can also change your results, so you need to structure your test to reduce the impact of that.
But a lot of testers pay no attention to geometry, then wonder why they get weird results.

Hardness is also significant. If you don't know the hardness of the blade you are testing, then you really have no idea what you are testing. A lot of folks ignore this, and again, they get weird results.

Last one: make certain you don't dull the edge on the surface that supports what you are cutting.
 
Quite a bit, actually.
https://www.bladeforums.com/threads...-alloys-using-a-visual-inspection-tec.641279/

https://www.bladeforums.com/threads...-vs-vg10-vs-buck-s30v-vs-benchmade-d2.769447/

https://www.bladeforums.com/threads...-zdp-189-m390-and-s30v-edge-retention.792540/

There's a couple other threads that I started if you want to look for them.

Three issues:
The largest issue is that geometry has more impact on edge retention than blade alloy.
♦ You must sharpen all the blades at exactly the same angle.
♦ The overall profile can also change your results, so you need to structure your test to reduce the impact of that.
But a lot of testers pay no attention to geometry, then wonder why they get weird results.

Hardness is also significant. If you don't know the hardness of the blade you are testing, then you really have no idea what you are testing. A lot of folks ignore this, and again, they get weird results.

Last one: make certain you don't dull the edge on the surface that supports what you are cutting.
Interesting. So you're not looking at how the edges continue to cut after the 20 cuts, but instead at how the edges look under a lens and how much visible damage there is? That definitely is one way to do it.

In the comparisons I've been doing, I'm trying to keep everything at a measured 17 dps (using a digital angle finder on my phone) on a KME with a 600 grit finish, and I cut the same lengths of media a recorded number of times while using a cutting board (in my case, stacks of cardboard under the cutting test piece) to ensure no edge damage takes place inadvertently. I'm also taping off a section to ensure I'm only looking at 1" of the blade at a time.

I'm also not looking at cutting performance but simply the absolute edge, so I'm not concerned about the differences in overall blade geometry between the tested knives. I don't personally believe that one knife being thinner at the edge than another will make it cut paper or hair any better, nor affect how hard the very edge is working in the tests. It should only affect how hard the knife is to push through a material, though that is definitely noticeable for some materials other than cardboard.

With all that being said, my results have been rather difficult to quantify so far. Running more tests today on Spyderco M390 and 4V, they seemed to do quite a bit better than what I posted yesterday, but not at the level to where I think I'd see a difference in normal use. I want to test Maxamet with a measured and consistent angle to see if I get noticeably better results.
 
I recently posted this elsewhere:

I have some preliminary results. I tested my Para3 LW SPY-27 against a Benchmade Bugout in S30V and did two different tests, with one variable changed for each pair of tests.

Before testing, as previously mentioned, I did a heavy sharpening and removed any steel near the edge that theoretically may have been burnt or fatigued, as is common among production folders. The Bugout is easily on its 20th edge.

I used the Para3 to cut down a lot of cardboard boxes in preparation for the test. Anecdotally, SPY-27 seemed to outperform S30V. After stropping and using it more, I sharpened the knife again.

The cardboard is all brand new, 32lb edge crush rating, purchased from Home Depot in a pack of 25 large boxes. They've been in my basement for 2 weeks.

Because cardboard is corrugated, to cancel the variable of which direction it's being cut, I alternate directions every 30 or every 15 cuts, depending on if im getting close to the end of the test.

For the purpose of this test, I was making 9" cuts since the box started out with 18" dimensions.

There's no cutting board being used that's dulling the knife. I'm hanging each section over the end of a table, cutting strait down through the cardboard, with cardboard strips falling into a box underneath. From start to end of the cut, the knife is parallel with the ground and makes contact with nothing except for what's being cut.

Sharpening progression used: DMT DiaSharp 220, Atoma 400, Atoma 600
Strop: 10um 5 passes per side, 3um 5pps, 1um 10pps, and 0.25um 5pps.

With each, I verified the edge by first whittling a free hanging hair, then cut magazine paper in all directions.

After that, the blade gets taped off, exposing only a 1" section.

The stopping point to the test is when the knife no longer will easily slice copy paper in both directions. With a slow slice from end to end of the exposed edge, if it hangs up or fails to start, it's done. I check every 15 or 30 cuts. It should be able to make 5 consecutive cuts in a row without hanging up.

With a cardboard cut test, what's being tested isnt entirely Wear Resistance, which means results often do not directly correlate with CATRA tests. Clean cardboard is almost entirely cellulose, not sillica. Because the primary abradant (cellulose) is significantly softer than the test media (steel), what's being tested more than anything is the durability of the fine apex to the dynamic compressive forces that it endures while cutting cardboard.

For those reasons, the pressure and angle need to be controlled by the tester, which is relatively easy since we're testing a knife that's sharp from start to end of the test.

The angle of approach relative to the cardboard must be maintained. If you cut cardboard while holding the knife at 90° to the cardboard, its substantially more difficult to push cut than if you hold at about 45°. The harder you push the knife, the quicker it dulls.

So I did two pairs of tests, one at 90° to the cardboard and the other at 45°. Here are the results:

Holding at 90° to the cardboard
Benchmade Bugout S30V - 157.5 feet
Spyderco Para3 LW SPY-27 - 303 feet

Its important to note, this result does not mean SPY-27 has double the edge retention of S30V. It's also important to note that my S30V result is within the range typically found with other published cardboard cut tests. It also approximately matches the result I got with Spyderco S30V with a different brand cardboard of the same specifications.

Test #2

Holding at 45° to the cardboard
Benchmade Bugout S30V - 315 feet
Spyderco Para3 LW SPY-27 - 690 feet

This is a significant gain. I actually conducted the 45° degree tests before the 90° tests.

The reason the 45° test set is about double of what the 90° tests are because it takes about twice as much pressure to push cut at 90° versus 45°.

Cardboard has to spread out of the way of the blade. This is much more difficult if the cardboard has to spread from side to side, like it does when the knife is held at 90°. When the knife is held at 45°, one side of the cut bends in front of the blade and the other in back of it, which is much easier. Pressure matters.

Edit: I ran a 3rd knife on both angles. I tested a Civivi Shredder in D2 steel. This is a $60 knife with an inferior steel and a big geometry advantage. Its significantly thinner behind the edge. I need to just buy a caliper but the thickness difference is obvious since all of these have about a 16 degree per side edge.

When tested at both angles, the dynamics of the Civivi versus the other two give an interesting insight into the cardboard cut test itself.

With the Civivi Shredder being thinner behind the edge, and thinner overall, plus having a hollow grind, make it act much differently than the others with this test.

First I'll say the hollow grind makes the knife significantly more unstable during the cutting. I'm trying to stay focused on doing strait thin cuts and it wants to wobble alot when held at 90 degrees. The full flat grind of the Para3 and high flat grind of the Bugout make them much more stable.

When held at 45 degrees, the Shredder glides through like a dream. Being thinner behind the edge, and thinner overall, it produces less friction because the cardboard doesn't have to spread as wide as the blade passes through. This means less compressive forces at the apex which means better edge retention, at least for this test anyway.

Here's the results:

Civivi Shredder
Held at 90° to the cardboard it cut 117 feet

It should be expected that Civivi's D2 will fall below Benchmade's S30V. I suspect without the geometry advantage, the D2 would have quit sooner.

#2 Civivi Shredder
Held at 45° to the cardboard it cut 348 feet.

That beats the Bugout result at 45° which was 315 feet. Again the Para3 SPY-27 did 690 feet at 45°

For purposes of this test the Bugout and Para3 did not have a significant geometry advantage over one another. They both required similar pressure. The Civivi on the other hand was noticeably easier when held at 45 degrees.

My belief so far is that holding the knife at 90 degrees is a better test of the steel and holding it at 45 degrees is a better test of the knife overall. In the case of the Para3 SPY-27, it just plowed through with superior edge stability.

When holding at 90 degrees, the pressures of the cardboard against the whole blade seem to cancel out most of the geometry advantages regarding the thickness behind the edge. This would be true only up to a point, and then geometry will always win. Wear resistance and edge stability only go so far.

It would appear that the 45 degree angle allows geometry advantages of the blade to shine through more easily, which allowed D2 to beat S30V.

With all of this cutting, and a sore hand, I can say that the Spyderco Para3 LW definitely has an ergonomic advantage over the Benchmade Bugout.

Because SPY-27 is run a bit harder than S30V typically is, it will have better edge stability than S30V thus giving it better edge retention in many real world scenarios. The difference will be most noticeable by those who keep their knives very sharp.

We need more tests of SPY-27. My numbers are surprisingly high. I'll do some more when possible.
 
At my last job, I cut the same stuff day in and day out. I sharpened my knives to the same angle using the same technique. I only sharpened them up when they wouldn’t cut worth a sh!t.

They all behaved differently. Some steels lose razor sharpness pretty quickly, but retain a good working edge for a long time. Some steels go outright dull quickly.

Just one comparison that stuck out in my mind from two knives I used extensively - I only had to sharpen my Sere 2K (VG10) a coupe of times a year, but I had to sharpen my Cold Steel mini Recon 1 (AUS8) every month. This was with both knives sharpened to the same angle, using the same technique, to the same level of sharpness, then seeing near constant use until they wouldn’t cut worth a damn anymore.

Daily use will tell you a lot about how different steels function.
Have you tried to do a controlled test? I'm not doubting your experiences, but at least from what I'm seeing with controlled cardboard testing it can be difficult to see a difference even between extremely different steels or knives with different edge geometries. This feels contrary to what I think I notice when I just carry and use a pocket knife for normal tasks.
 
I recently posted this elsewhere:
Very interesting. I'm glad to see someone testing SPY27.

One interesting topic worth that you discussed how much better the edge retention is at 45° vs. 90°. I can definitely see that having a significant effect if the cardboard at the very edge is applying a load to either side of the apex, which seems accurate in my experience (this is why I and many others cut cardboard at 45°). Repeated stress on the apex = apex damage, no question about it.

What is harder for me to understand is why a knife with thinner edge thickness would exhibit significantly better edge retention rather than simply greater cutting ability. The way I see it, the edge cuts a material and the blade forces its way through it, and I would think those would represent the total sum of forces on the blade during a cut. Hence, I wouldn't think that a knife with a thinner edge would have an apex that wears much slower than that of a thicker knife, as both edges still do the exact same amount of cutting and the apexes of each still see the same amount of material passing by, at the same angle, at similar forces at that very apex, etc, and the additional force noted by the user is a direct function of the geometry above the apex. I would agree there's more friction for a thicker edge, but I can't figure out how that friction would be at the very apex rather than where the additional interference is (at the shoulders of the edge).

Have you tested any knives with identical steel, heat treat, edge angle, etc. where the only difference between tests was the edge thickness? I ask because I would assume it to be more likely that, in your test, your particular D2 Civivi was just better heat treated than your S30V Benchmade, or that the differences might be sure to statistical variance or even just a difference in sharpening angle between tests. I have some knives where the edge geometry changes quite dramatically from tip to heel, so I'll try one of those in an upcoming test.
 
Last edited:
I am thinking of doing similar tests myself but have a few issues to resolve.

(1) Sharpness measurement. I built a DIY sharpness tester similar to Edge-on-up and have been testing it if it reliably reports the level of edge sharpness. At the moment, sharp to dull resolution is not wide enough (from 30 g to 175 g) presumably because I use very thin fish line. If this works, I should be able to set the initial sharpness comparable and follow how that degrades.

(2) Cut medium. It seems that some cardboard leaves glue on the edge, which would affect the sharpness measurement. I am considering to use something like thick name card paper instead.

(3) Cutting force and speed control. A knife with poor geometry would need more force to cut the medium than one with superior geometry at a fixed cutting speed. This may dull the poor geometry knife faster. I am not sure if I can compensate this.
 
Why not just use the can opener on the SAK?
The old Tinker I use to carry was so worn the shoulder on the “hook” part of the can opener was rounded off and didn’t want to stay under the can rim. When you would try to bear down to force the blade part of the opener into the can the hook part would pop off the rim. It could be done but was much easier to use the blade to start the hole.
 
One thing you’ve learned is trying to do a quantitative comparison with a qualitative measurement is extremely difficult. Also learning to test is sometimes ignored. If you watch the YouTube testers from the first video to the last you’ll see big differences. Though almost none of them discuss statistical variation, in part because almost none of them test for it. Without multiple runs you might just be testing randomness. And as knarfeng mentioned there is sometimes a misunderstanding of what was actually tested, such as claims it was steel differences but edge angle differences drown that out, or sharpening differences (such as inconsistency or comparing factory edges), etc.
 
Marren your results match my anecdotal experience in use. I've never noticed a steel holding that hair popping edge longer than another. I don't use my edges down to a "working edge" or anything that would even struggle to cleanly cut paper. I'm one who will always opt to save a few bucks if there's an option between say 20CV and n690 on the same knife. I'll always take the less expensive steel option.

BTE geometry has always been the biggest function of edge retention in my personal use. Even more so than blade stock thickness. I would fully expect a Medford Marauder ground aggressively thin BTE in D2 steel to cut longer (at my preferred sharpness) than a ZT 0450 in 20CV.
 
I recently decided to start carrying a knife in 8cr13mov to see how it would do for me as an edc. I don't live a hard life but have a kid so there are lots of boxes to break down and I do woodworking but that's about it. I find for day to day use it didn't really matter what the blade steel is. I also loose that shaving edge quickly no matter the blade steel. The biggest difference is that I can strop the 8cr13mov back to shaving quite easily where the supper steels take a bit more work. Also the simple steels tend to roll and not chip. So to keep a sharp edge the simple steels actually require less sharpening depending on how you maintain your edges.
 
Marren your results match my anecdotal experience in use. I've never noticed a steel holding that hair popping edge longer than another. I don't use my edges down to a "working edge" or anything that would even struggle to cleanly cut paper. I'm one who will always opt to save a few bucks if there's an option between say 20CV and n690 on the same knife. I'll always take the less expensive steel option.

BTE geometry has always been the biggest function of edge retention in my personal use. Even more so than blade stock thickness. I would fully expect a Medford Marauder ground aggressively thin BTE in D2 steel to cut longer (at my preferred sharpness) than a ZT 0450 in 20CV.
I've never believed that edge thickness affects edge retention, but rather cutting ability. This is a long video but this individual tests exactly this assertion.
Tl;Dw: edge thickness being varied between two otherwise identical knives resulted in the same edge retention.
 
I think one of the issues faced is the differences in edge thickness, angle, and etc. The ZT is like a brick compared to the Spyderco in 8Cr. That is why there was negligiable difference. Thickness behind the edge and edge angle are of paramount importance. If you had two knives, one in M390 and another in 8Cr and they were identical, you would see a huge difference in edge holding. I think Cliff Stamp has shown conclusively that thickness behind the edge and edge angle are the two most important features. In this video
you can see the lowly SAK out perform super steels because of this.
 
One thing you’ve learned is trying to do a quantitative comparison with a qualitative measurement is extremely difficult. Also learning to test is sometimes ignored. If you watch the YouTube testers from the first video to the last you’ll see big differences. Though almost none of them discuss statistical variation, in part because almost none of them test for it. Without multiple runs you might just be testing randomness. And as knarfeng mentioned there is sometimes a misunderstanding of what was actually tested, such as claims it was steel differences but edge angle differences drown that out, or sharpening differences (such as inconsistency or comparing factory edges), etc.
Yeah, the biggest thing I've seen so far is that paper cutting/hair shaving isn't objective in the slightest. A better system, like a sharpness tester, might be a good idea.

What's very surprising to me is that, even intentionally changing edge angles to try to make one knife outperform another, I still saw little difference.
 
Regarding edge thickness, its something I've certainly been trying to wrap my head around.

Steel is pretty hard. Most of the things that most people cut are soft and not abrasive.

Dirty cardboard is full of sand and grit which are extremely abrasive.

Clean new cardboard is not abrasive, well it is slightly, but cardboard basically is cellulose, and it's not doing much to dull a knife without pressure.

On a 600 grit edge, almost any knife steel can cut in excess of 1000 feet of clean new cardboard if using the blade from heel to tip, back and forth, slicing your way through. I've personally tested this with 420HC on a Kershaw Dividend. After 1000 feet it was still easily slicing through copy paper.

That test demonstrates that cardboard isnt doing a whole lot to wear down steel.

When you tape off the blade, only exposing an inch of it, and push cut your way through the cardboard, everything changes. You're primarily testing the effects of the compressive forces of the cardboard against the apex.
 
Have you tried to do a controlled test? I'm not doubting your experiences, but at least from what I'm seeing with controlled cardboard testing it can be difficult to see a difference even between extremely different steels or knives with different edge geometries. This feels contrary to what I think I notice when I just carry and use a pocket knife for normal tasks.

I’ve done a couple of head to head tests with different steels, but they weren’t what you could call “controlled”.

I have definitely seen different steels behave differently, but I can’t be sure that it’s the actual STEEL, or some other variable.

I have seen some dull faster, some chip or deform easier, but this may have more to do with heat treatment and edge geometry rather than steel designation. There are too many variables and everything when it comes to steels is a trade-off.

Some things have surprised me, like the things that will dull some steels faster or put chips or dings in them (or not), etc.
 
I have done some of my own testing. Generally I saw the expected trends with some steels lasting much longer than others, but there were a few surprises. And I have never been able to get hundreds of feet of cardboard cut and still have a shaving edge.
 
Another fun thread!

Has anyone ever tried doing something like this for yourself? I'd be curious to hear your thoughts.

Not in a controlled apples to apples manner, but yes. I took my plain edge Pacific Salt down to a zero edge.

Pacific Salt b.jpg

Then I used it alongside a 204P Millie that I had also brought down to a zero edge.

IMG_6965.JPG

I collect cardboard from random sources to use for packaging my custom knives to ship to customers, so the only consistent criteria for it is that it needs to be clean. Aside from that there will be a lot of variability. That being said, I used both knives to cut up some of this cardboard I had in excess, but I took care to cut each box down half with one knife, and half with the other. I found little discernible difference between the two knives ability to cut the cardboard, and was quite surprised to find the H1 retained it hair shaving ability for what felt like forever. I also got some rolling in the 204P during turning cuts, whereas the H1 did not roll in the same type of cut. I was cutting cardboard for (IIRC) half an hour between the two knives.

I had this experience, it's a bit of a reality check- I have cut down my supersteel knife purchase by a bunch..I have really grown to love 14c28n to me I really like the feel of a super refined edge over a knife that feels dull but can supposedly cut forever

Larrin's conclusion was that 14C28N is basically "the best" budget steel one can ask for, with AEB-L being a close second. I have found that AEB-L, hardened to 63 RC holds a great edge in my (admittedly limited) testing with it so far, and is anything but brittle.

What's very surprising to me is that, even intentionally changing edge angles to try to make one knife outperform another, I still saw little difference.

Surely the lower the angle, though, the better the knives performed overall. That's what I've always found.
 
Also, I forgot to mention that my most commonly carried and used folding knives are my Resiliences in 8Cr13MoV because once I brought them down to under .012 BTE, I found them to perform incredibly well. So well that I have tailored a Resilience for every type of job.

Spyderco Resiliences.JPG
 
Back
Top