Carbides on Cutting Edge

[Cyrano]

... It is difficult to photograph the cutting edge because the shallow depth of field of light microscope ...

To overcome the shallow DoF, have you tried focus stacking?

Single exposure, near focus:

Single exposure, far focus:

23 exposures (near to far focus) stacked:

 

[Mo2]

Skip to 9min

Too course a stone ripped the edges up. I assume the steel is worse than 8cr13mov. Looks like an mtech. But anyway since we're showing cool photos

 

[Juha Perttula]

So this may support Juha’s OP a bit. I have two puukkos, one made by a knife maker Lauri who Juha collaborates with. It uses 80CrV2 steel — a relatively simple carbon steel that is heat treated to Juha's specifications to produce very fine grain and high hardness. The blade’s edge is hardened to 63 Rc. The edge is sharpened about 35 degrees inclusive, but it's difficult to measure with my laser protractor because the edge bevel is so narrow. The edge shoulders are about 0.011 inches.


The other knife is a special project by Deadboxhero, using Vanadis 4 Extra steel handed by Peters to 64 Rc. The edge is about 40 degrees inclusive (but hard to measure because the edge bevel is so narrow), with edge shoulders about 0.006 inches. It’s difficult to measure these edge geometries when the edge bevel is all but invisible. The knife was made by top knifemaker Malanika.


I did two tests. One is my standard test for edge stability: chopping a piece of bailing wire in two on a large flat block of Doug fir. In this case, I used a rubber mallet and tried to chop the wire in two. Only the Malanika could chop the wire in two. No other knife I have can do that. But the edge rolled badly, which seems odd for a steel this hard.


The Juha/Lauri puukko’s edge flattened somewhat, but came out almost intact, which is extremely good edge stability (toughness and strength). It didn’t cut the wire in half.


The other test was to jab the tips of the knives into a piece of Doug fir and pry them out. This was a relatively gentle test meant replicate the most abuse that a pukka would see in normal whittling. Both knives passed that test easily, and both have very pointy tips.


I can’t really say too much about the Vandals 4E blade because it was much thinner than the Juha/Lauri pukka. I’ve had other knives suffer this much damage on the bailing wire test, but none of them were even close to the geometry of the Malanika.

But for sure, Juha's heat treat of a simple steel performed extremely well for edge stability on the bailing wire test. Few of my knives can match that.

This is very interesting comparison. I hope we will hear more your test results.

Malanika blade has classic Scandi grind with micro size secondary bevel. Lauri’s blade has slightly larger secondary bevel which makes the edge stronger, but it is not so sharp. I myself like a Scandi grind which have no secondary bevel at all, but this kind of edge rolls quite easily and an average customer may not be happy with it. I love old skool Scandi sharpening tradition, but nowadays many Scandinavian puukkos have quite large clearly noticeable secondary bevels (larger than the bevel of Lauri's blade). I suppose it is influence of American knives.

 

[BluntCut MetalWorks]

Thanks for putting really nice knives through an abusive test Hopefully, you didn't fixed these edges yet because it would be informative to see how their apexes after cut to about 1/3 diameter (then pull out) of the bailing wire. Essentially testing apex stability - and of course, macro pics of edges afterward. Please don't try to peen 4V ripple back, just sharpen it out, where affected area will be weaker and still better than a large chip

Cutting edge geometries (inline green bold) - 4V has about 49% [(40/35)^3] more strength than 80CrV2.

However grind bevel strength played a bigger role. I used 'ripple' because it is more of bent than roll (more than 90* bend). From pic - my estimate grind bevel for 4V (0.14" thick, 0.8" grind bevel face) to about 5 dps; 80CrV2 (0.14" thick, 0.65" face) ~ 6.5dps. Using (x/y)^3, 80CrV2 grind bevel strength is 120% stronger than 4V. In consequence, 4V resulted with a big ripple in grind bevel.

...

It uses 80CrV2 steel — a relatively simple carbon steel that is heat treated to Juha's specifications to produce very fine grain and high hardness. The blade’s edge is hardened to 63 Rc. The edge is sharpened about 35 degrees inclusive, but it's difficult to measure with my laser protractor because the edge bevel is so narrow. The edge shoulders are about 0.011 inches.


The other knife is a special project by Deadboxhero, using Vanadis 4 Extra steel handed by Peters to 64 Rc. The edge is about 40 degrees inclusive (but hard to measure because the edge bevel is so narrow), with edge shoulders about 0.006 inches. It’s difficult to measure these edge geometries when the edge bevel is all but invisible. The knife was made by top knifemaker Malanika.

I did two tests. One is my standard test for edge stability: chopping a piece of bailing wire in two on a large flat block of Doug fir. In this case, I used a rubber mallet and tried to chop the wire in two. Only the Malanika could chop the wire in two. No other knife I have can do that. But the edge rolled badly, which seems odd for a steel this hard.

The Juha/Lauri puukko’s edge flattened somewhat, but came out almost intact, which is extremely good edge stability (toughness and strength). It didn’t cut the wire in half.

The other test was to jab the tips of the knives into a piece of Doug fir and pry them out. This was a relatively gentle test meant replicate the most abuse that a pukka would see in normal whittling. Both knives passed that test easily, and both have very pointy tips.

I can’t really say too much about the Vandals 4E blade because it was much thinner than the Juha/Lauri pukka. I’ve had other knives suffer this much damage on the bailing wire test, but none of them were even close to the geometry of the Malanika.
...

...

 

[Juha Perttula]

This post is useless with a lot of unknowns
Like what kind of testing have you done.?What sharpening medium you used? The obvious questions that arise.

I have made two kind of cutting tests: leather cutting tests by leather knives, and wood and bone working tests by puukko knives. I sharpened the blades by aluoxide "stone", silicone carbide grinding papers, or diamond file. I occasionally polished the edge by aluoxide or diamond particles embedded in leather. I tested included angles about 20 degree. I have mostly studied carbon steels with 0.7-1.6%carbon, D2, and Vanadis 4. I did not found that carbides were useful, but I found that hardness was very useful. According to my test difference of 2 HRC unit can be noticed.

In my severe wood or bone working tests edge generally rolled or chipped. So, carbides and abrasive wear resistance were not interesting. In my leather cutting tests I expected great results for carbide-rich steels, but they were not better than simple carbon steels. The hazelnut chocolate demonstration try to explain the reason.

 

[John Stubley]

The "The hazelnut chocolate demonstration" would have not have last the hour never mind the day before it was eaten.

I watched a lot of knife youtube video`s and cutting rope seems to blunt the blades very quickly, like between 60-150 cuts.

John.

 

[Wowbagger]

No other knife I have can do that. But the edge rolled badly, which seems odd for a steel this hard.

Sounds to me like the edge got annealed while grinding to final shape.
Shame.
The other thought I had was maybe the rubber mallet caused some loss of control.
??? hmmmm
good test though.

PS: I have seen knife makers cut nails using a blade hardness test sample. They used a steel hammer. They DID cut the nail and there was NO edge damage.
They may have annealed the nail and not let on . . . but still . . . impressive.

PPS: I have come to wonder also (besides thinness behind edge being critical to nail cutting) . . . if the surface we are cutting nails / bailing wire against isn't important. I have used thick blocks of aluminum backed by a large wood stump on concrete and I repeatedly hit the blades with a large hammer (small sledge) and repeatedly the nail folded up some and then (for the most part) on the third hit shot out from under the knife uncut. I repeated the same using a sharp smallish steel cutting chisel. Same result.

What I am getting at is it may be best to use a thick mild steel surface to cut against rather than aluminum or wood.
The other things that seem to be key are thinness of blade behind the edge and perfect heat treat for the task at hand.

 

[Wowbagger]

They may have annealed the nail and not let on

My apologies to Blunt Cut.
He pulled a random nail out of a fresh box so obviously not annealed.
My mistake
I had forgotten about that.
I hope it is OK that I post this here

 

[Juha Perttula]

My apologies to Blunt Cut.
He pulled a random nail out of a fresh box so obviously not annealed.
My mistake
I had forgotten about that.
I hope it is OK that I post this here

Great test. It inspired me also to test my puukko. I put a nail dia 0.12" on anvil and cut it by the puukko by hammering on the spine of the blade. The blade (80CrV2, edge 63 HRC) made by Lauri chop the nail easily. The edge remained sharp but bent slightly (red circle). Twindog in earlier post has measured the geometry of Lauri's blade: "The edge is sharpened about 35 degrees inclusive, but it's difficult to measure with my laser protractor because the edge bevel is so narrow. The edge shoulders are about 0.011 inches."

And notice, alloy carbides in steel do not help cutting nails. Simple steel can be amazing.

 

[BluntCut MetalWorks]

Juha - 80crv2 63rc 17.5 dps 0.011" bet (behind edge thick) edge passed 12d nail baton test very well

* my brain is running backward this morning

I don't do this test much any more, although still actively tests whittle and chop nail because those are more useful to me. Nevertheless, I did a hurry-reprofile a D6 test blade... Here is a macro pic of edge afterward. *6d surface area (cross section) relative ((dia6/diaX)^2) to 12d and 16d are about 65% and 53% respectively.

btw - I chose D6 at 67.5rc steel because (might as well) for my record/data, *note - rolled/smushed edge from impact to anvil.

 

[Alberta Ed]

Interesting, but I'm not sure it's a valid test for those who use their knives for cutting purposes.

 

[John Stubley]

Juha - 80crv2 63rc 17.5 dps 0.011" bet (behind edge thick) edge passed 12d nail baton test very well

* my brain is running backward this morning

I don't do this test much any more, although still actively tests whittle and chop nail because those are more useful to me. Nevertheless, I did a hurry-reprofile a D6 test blade... Here is a macro pic of edge afterward. *6d surface area (cross section) relative ((dia6/diaX)^2) to 12d and 16d are about 65% and 53% respectively.

btw - I chose D6 at 67.5rc steel because (might as well) for my record/data, *note - rolled/smushed edge from impact to anvil.

Try an SK-5 blade for your nail chopping test next.

John.

 

[BluntCut MetalWorks]

Here is a niolox 65rc test blade, I batoned rack-of-lamb to shortened the bone (when edge was at 15dps). As I mentioned in my prev post - 'not doing this test much any more'. Well because result varying so much depend on how test is conducted - such as angle and number of hammer strike; edge finished & steering. I agree with Ed about invalidity of test in real-use. While I see whittle & chop nail still useful for testing lateral strength and global toughness.

For 6d nail - edge failure started as rolled then lead to crack (probably from rebounce from anvil). As usual, my haptic test conduction is more wild rather than high precision. Obviously niolox at 65rc at 17.5dps isn't strong enough. otoh, lower hardness will help with bigger nail, since edge would smush-in and rest of edge would impact anvil, which will avoid this big chip - (hahaha I kept hammer until cut through, chip bits are still being push down for cutting). Key take away for me - no catastrophic failure from hammer strike nor sheer force pivoted on nail.

John - I don't have sk-5 but have cut many nails with 52100, w2, etc.. before. After this test above - I am done with baton nail, of course until curiosity or need to confirm/refresh my perspective (this test lacked of usefulness) again.

 

[BluntCut MetalWorks]

D6 100+micron dia carbides are much easier to detect and understand what happen at the edge. *in order to extrapolation for smaller dia carbide one must take into account psi differences at carbide interface depend on surface area - i.e. not linear but involve load volume and proportion of matrix vs carbide interface.

Just to be clear - edge *as picture* still fast-slice phonebook paper and cut cardboard, meat, etc like a buzz saw - just forget about shaving with it. Also if this edge has maybe a 20-40um tall 20dps micro bevel, it would has much less carbides tear damage. So mitigate with geometry but liabilities of carbides on edge still there, hence choose steels per usage accordingly.

** pics with corrected scale bar **

 

[Twindog]

Thanks, Luong, for that amazing photography and testing. (And all you share with us on the forum.)

I was wondering why my 4V (basically Vanadis 4 Extra) blade rippled (or rolled) when I chopped the bailing wire. It seems that it should have chipped, more like what you're showing in your photos below. Juha's heat treat of 80CrV2 is clearly excellent, but to my thinking the 4V blade with its big load of carbides should have shown tear outs and chipping before rolling or rippling. But instead it showed far more toughness (resistance to chipping) than strength (resistance to rolling and bending). Yet, that steel was heat treated to 64 Rc, and hardness is a good proxy for strength. So even on a super high-strength blade, it was the toughness that prevailed. I can't quite wrap my head around that result.

By the way, the ripple mostly went away with hard, edge-trailing strokes on a coarse diamond stone. The edge is almost back to it's original shape. Good steel is fun to work with.

D6 100+micron dia carbides are much easier to detect and understand what happen at the edge. *in order to extrapolation for smaller dia carbide one must take into account psi differences at carbide interface depend on surface area - i.e. not linear but involve load volume and proportion of matrix vs carbide interface.

Just to be clear - edge *as picture* still fast-slice phonebook paper and cut cardboard, meat, etc like a buzz saw - just forget about shaving with it. Also if this edge has maybe a 20-40um tall 20dps micro bevel, it would has much less carbides tear damage. So mitigate with geometry but liabilities of carbides on edge still there, hence choose steels per usage accordingly.

** pics with corrected scale bar **

 

[BluntCut MetalWorks]

Umm seem odd to me too. So luckily I found a small 4v test blade close by.

Cutting wire damaged consists of ripples and micro-chips. In order to pass the bone whittling (do 4 sides) test - minimum edge strength is about 62rc. This one barely passed that test. So perhaps try whittle bone with your 4V.

Here is a 100x closeup on combination of ripple+smush+chip

Edit to OT:
4V didn't responded well with HT 1.0, performance was below my expectation, so I wrote it off along with s30v. Recently my S30V 65rc with HT 2.5 perform very well, which is inline with other pm steels with same ht. I expect 4V does well with HT 2.5 too. I just did 2 cuts same wire with D6 67.5rc (against rosewood backing) - ~half size of ripples and smaller micro-chip than 4V 63rc. When time permits - there are a couple 4V blanks avail for experiment with ht 2.5.

Thanks, Luong, for that amazing photography and testing. (And all you share with us on the forum.)

I was wondering why my 4V (basically Vanadis 4 Extra) blade rippled (or rolled) when I chopped the bailing wire. It seems that it should have chipped, more like what you're showing in your photos below. Juha's heat treat of 80CrV2 is clearly excellent, but to my thinking the 4V blade with its big load of carbides should have shown tear outs and chipping before rolling or rippling. But instead it showed far more toughness (resistance to chipping) than strength (resistance to rolling and bending). Yet, that steel was heat treated to 64 Rc, and hardness is a good proxy for strength. So even on a super high-strength blade, it was the toughness that prevailed. I can't quite wrap my head around that result.

By the way, the ripple mostly went away with hard, edge-trailing strokes on a coarse diamond stone. The edge is almost back to it's original shape. Good steel is fun to work with.

 

[Juha Perttula]

Umm seem odd to me too. So luckily I found a small 4v test blade close by.

Cutting wire damaged consists of ripples and micro-chips. In order to pass the bone whittling (do 4 sides) test - minimum edge strength is about 62rc. This one barely passed that test. So perhaps try whittle bone with your 4V.

Here is a 100x closeup on combination of ripple+smush+chip

Edit to OT:
4V didn't responded well with HT 1.0, performance was below my expectation, so I wrote it off along with s30v. Recently my S30V 65rc with HT 2.5 perform very well, which is inline with other pm steels with same ht. I expect 4V does well with HT 2.5 too. I just did 2 cuts same wire with D6 67.5rc (against rosewood backing) - ~half size of ripples and smaller micro-chip than 4V 63rc. When time permits - there are a couple 4V blanks avail for experiment with ht 2.5.

Yes, it is good idea to make bone whittling test, it reveals is hardness OK. Vanadis 4 have very small carbides and according to my experience it can attain excellent toughness. So, it could be possible that Vanadis 4 Extra blade did not chip because exceptional toughness.

 

[BluntCut MetalWorks]

First of my understanding & perception on compressive and bend strength, might has major flaws.

I think. bend strength contributes about 80% performance of normal keen edge usage, except vertical/perpendicular-to-surface cuts: e.g. Catra, punch, press, etc... From chart below - Vanadis4e (extrapolate to 4v) edge yielded (plastic deform) on very low deflection. Also plasticity range (between deformation and fracture), esp at high hardness, indicate high ductility/ferrite.

If my suspicion is correct, this ferrite% improve impact load and plasticity BUT compromised the edge with this % of ferrite. Although overall hardness of blade is 64rc, that average includes sub 20rc from ferrite/iron in matrix. So a lot of 64rc might benefits more to the 20% component of performance, while ferrite degrades the 80% aspect of perf. Question maybe - are we sure, this type of toughness (ductility) is what we seek for keen edge tools? It's a 'No' for me - when ht according to mfg spec. Maybe (yup plenty of maybe), Van4E/4V would shine at 66rc via BCMW HT 2.5

Yes, it is good idea to make bone whittling test, it reveals is hardness OK. Vanadis 4 have very small carbides and according to my experience it can attain excellent toughness. So, it could be possible that Vanadis 4 Extra blade did not chip because exceptional toughness.

 

[chiral.grolim]

... I have tested extremely sharp blades and in that case there is no room for carbides on the thin cutting edge.

...the hazelnut chocolate demonstration shows it clearly; when the target is ultimate sharpness, carbide-rich alloy steels are not necessarily better than plain carbon steels.

While I agree that carbide-rich alloys may not be preferred to low-alloy steels for achieving high sharpness (depending on HT), I am often struck by the need to re-iterate that "when the target is ultimate sharpness" you want ONLY carbide in the apex.

As Luong correctly points out, bending strength accounts for the vast majority of edge performance in cutting tasks - i.e. the stiffer apex-material takes the keener edge. It is precisely for this reason that the finest-edged cutting tools are made from carbide hard-metals, not from steel at all!!! The finest steel apices on fine-grained steel are commonly ~0.5 um, while the working apex diameter of WC-Co and diamond and obsidian edges are commonly 0.05 - 0.005 um, some 10x - 100x sharper! The "pure" carbide materials can achieve this because of their strength (hardness). Increasing Rockwell hardness in steel brings it closer and closer to the performance of these materials. However, where these carbide hard-metals fail is macro-toughness - once you overcome that resistance to bending, they fracture readily, and the thinner the bevel-angle (geometry) BEHIND the apex, the weaker (cubic) the material is with regard to resisting lateral stress.

As Luong previously pointed out, there is a complex relationship between the carbides and the steel matrix "binder" in which they are embedded. The more carbide that you have embedded in the very apex of the cutting edge, the SHARPER your blade can be made!!! But the ability to achieve that edge, and the durability of that edge in your specific use, as well as the utility of even having an edge that is 10x - 100x sharper than steel can achieve, coupled with the expense of replacing the blade when it fails, all these factors contribute to the impracticality of such tools in everyday life.

I just want to be sure to avoid confusion in these conversations. When we are talking about a blade as "sharp", it is helpful to give an actual measurement, e.g. apex diameter, to put the concept into context. A 12C27 blade sharpened to 0.5 um apex diameter is "sharp" for a steel, but it is flat "dull" compared to blades achieving 0.05 and 0.005 um diameters. 12C27 (one of Sanvik's very fine-grained steels) cannot achieve 0.005 um with any level of stability because it is far too soft. And just to play with one's head, recognize that appropriately sharpened D2 with massive carbides in the apex could achieve that sharper edge specifically where those carbides reside.

Carbides are not only able to achieve superior sharpness, but they are also less susceptible to heat (and generate less heat during cutting) and, as has been mentioned and often comes up in knife discussions, they are more resistance to abrasion. These features make them ideal for drill-bits and saw-teeth and milling-tools, stamps, lathes, they make them ideal for the tools you use to cut and sharpen steel blades.

The trouble with carbide is, again, that most people could neither achieve nor appreciate the level of sharpness attainable with such materials when what they REALLY want is a TOUGH tool that is very thin behind the edge while simply acceptably sharp at the apex. I am happy with a knife that can whittle a hair though even that is excessive for shaving, and of what use is such a tool to most people if it is sharp enough to trace out individual cells from a slice of plant stem but is easily fractured if made into a blade 4" long with a 5-dps primary bevel?

I really appreciate what Luong does because he is pushing for the HIGHEST possible Rc hardness, to bring his blades as close to the strength of carbide hard-metals as he can, but while simultaneously demanding that they endure severe impact force and edge-steering. The steel industry has been working toward more and more carbide-rich steel in order to improve tool-performance while trying to maintain the toughness that steel is known for. If they didn't care about toughness, they wouldn't be working on refining carbides. While on the other end, hard-metals are almost entirely carbide but the industry is working on making them tougher to be more like steel without sacrificing the strength that carbide is known for

When it comes to razor-blades and shaving, applications where a fine edge of very low edge-diameter is required but relatively stout behind-the-edge geometry may be preferred (e.g. 0.05 um diameter with a convex edge starting from around 45-dps) and impact-toughness is unimportant, then they very best material to use is one with the near-uniform superior hardness of carbide. Steel is acceptable but inferior.
When it comes to saw-blades which require moderate impact toughness and a very fine edge of low diameter resistant to heat and abrasion, but where again a very thin behind-the-edge geometry is not really required, again the carbide is preferred.

But when it comes to a long knife subject to bending and impacts and irregular attempts at re-profiling or changing the geometry, where a fine edge of relatively thick diameter is all that is required, THEN we prefer steel because it is stronger than plain old iron and more ductile than plain carbon - mix them together and Tah-Dah!! Magic

To get back to the OP, a "carbide rich" steel suffers from rough edges NOT because of the carbides but rather because of the binder, i.e. the steel around them. Improve the bind between carbide and matrix, and you improve your edge. Carbides "tear-out" is not the fault of the carbides, it is the fault of the binder not holding them. This can be mitigated in large part by geometry, but it can also be mitigated by increasing carbide content WELL above the <20% level found in steels. When you have no or fewer or smaller carbides in the apex, you can make a finely finished apex but of inferior diameter and longevity if your ultimate goal was the sharpest possible edge.

 

[the possum]

I struggle with this issue.
And we have lots of evidence that high-carbide steels out cut simple steels by a wide margin, including rope cutting, cardboard cutting and well-controlled, scientific CATRA tests. What besides the presence of carbides on the apex could explain that performance advantage over simple steels at the same sharpness?

The big difference is where we want to draw the line at "dull". If ya just gotta sharpen your blade when it starts to pull at yer arm hair instead of cleanly "wiping" every hair off effortlessly, you'd be better served by a low/no carbide steel. If you want a "working sharp" edge that may or may not take any hair off your arm, but will continue to saw through rope all day, you're better off with lots of carbides.