Got the scope out again.

I think you're underestimating the significance of metallic bonding.


New Mystery hone. The sawmarks, coloring, feel and weight all say LG thuri, but it has speckles like a Lynn Id and seems a bit too hard for a Thuri. First pic is slurry, second is water.

That may be, but on the other hand, I think you're underestimating the force applied at such small scale. Flexing a blade a few mm with the load distributed along a comparatively huge area is much different than a concentrated load like a diamond point that exceeds the yield strength of the steel in an area in the micron size range.

Exceeding yield strength is the problem, not just flexing the steel. If you flexed that same singing razor edge to the point that it took a permanent set, wouldn't you consider it damaged? To get steel to take a permanent set, the yield strength needs to be exceeded - and the force it takes to do that is dependent on area - yield strength is measured in PSI. At such a small scale, it really doesn't take a lot of force to exceed the yield strength of hardened steel. A spring steel shim that is .001" thick will bend as easily as a piece of paper. A razor edge (and even behind the edge) at tenths of a micron is orders of magnitude thinner and weaker.

Repeatedly exceeding the yield strength (like repetitive honing on the coarse diamond plate pushing the steel beyond the plane of the bevel with every opposite stroke) is what fatigues the steel to the point that it fails, merely flexing it will do hardly anything. Think of the old coat hanger example everyone uses - repeated bending of the hanger ends up breaking it. If you merely flexed the hanger so that it returned to its original starting point on its own, it would probably take a hundred years to fatigue enough to break.

eKretz said:

That may be, but on the other hand, I think you're underestimating the force applied at such small scale. Flexing a blade a few mm with the load distributed along a comparatively huge area is much different than a concentrated load like a diamond point that exceeds the yield strength of the steel in an area in the micron size range.

Click to expand...

Are we talking about high grit diamond points or broad faces of coarse grit diamond plates? I really doubt a 120grit diamond plate with 120micron diamonds is applying any significant amount of force in a 1 micron area.

Repeatedly exceeding the yield strength (like repetitive honing on the coarse diamond plate pushing the steel beyond the plane of the bevel with every opposite stroke) is what fatigues the steel to the point that it fails, merely flexing it will do hardly anything.

Click to expand...

This would be a concern if you were not removing steel and recessing the edge while doing this. You aren't repeatedly pushing a micron-thick edge back and forth, because it doesn't even make a single pass on a low grit plate before that micron thick edge is removed and replaced with an edge formed of steel that has not been flexed as such. On exceedingly slow finishing stones, this could be an issue, which is why overly slow hones are not ideal for finishing, regardless of how fine they are. If you have a beveled razor, you're not going to make any significant number of passes on a coarse DMT before the extent of the edge that sees the level of flex you're concerned over is entirely removed.

I'm just describing the theory arrived at by observing what was seen under the electron microscope. Your description of the yielded steel being removed means you aren't quite understanding the theory I think. The steel that yields away will never be abraded away by the diamond plate with alternating strokes - it's constantly being replaced as the edge is worked back. As steel is removed, the steel behind is still yielding away. The worn diamond particles push it back because they are dull - they aren't cutting much at the apex, which is why the edge convexes and why it's so much straighter and finer than what it would be if it shared the scratch depth and finish the rest of the bevel has.

The coarse diamond plate with 120 micron diamond is definitely applying significant force at the tips of the diamonds - because they are so tiny, a very small force works out to a huge PSI. These diamond particles aren't cutting at full width, they are only cutting with the very tips of the particles. Furthermore, with the very wide spacing of the diamonds on these coarser plates (they're nowhere near as densely loaded as the finer plates) the pressure per unit of area is magnified even more.

If you are convinced that this is not correct, how would you explain what is going on at the apex? What else do you think would cause the convexing in the last so many microns behind the apex, smoother finish and straighter edge than what should be possible taking into account the deeper scratches and raggedness of the bevel further back? I am open to hear other ideas, but this is the best I've heard so far that explains everything that's going on to my satisfaction.

If this theory were correct, it would lead to a concave bevel, not a convex one.


Convexing is not demonstrated. A broad edge is demonstrated. A broad edge is explained by the diamond plate having little to no variance in cut depth, as the substrate has zero yield. Take a potato and a two graters. One with teeth aligned, one with teeth at varying depths, the difference in edge produced will be very obvious. Now an edge can be forced to convex on a DMT by decreasing pressure during a pass on one (effectively "rolling" the edge into the plate at the end of the pass by flexing the bevel and then allowing it to correct itself). I think viewing SEM images people are misinterpreting the natural appearance of an edge with a larger than average edge radius as a "convexed" bevel.

because they are so tiny, a very small force works out to a huge PSI

Click to expand...

For this theory to work, the force must be exerted exclusively at these fine points, thus the diamonds would have to be incapable of penetrating the steel, and distributing the force. In other words, they must not be harder than steel. In fact, this may actually be an issue with abrasives that almost or perfectly matched the hardness of the steel, diamonds are the least likely abrasive for this issue to arise. It sounds like you applying a macroscopic concept and presuming that exceeding the yield strength of the steel causes microscopic separation at the structural level (this tendency to chipping you describe), rather than a sheer cutting when the force is applied by a harder than steel, microscopic object. What are you basing this on? Thousands of years of history would say this theory is mistaken.

Furthermore I'm pretty sure that PSI ratings vs Yield strength become almost irrelevant to the ends you're seeking once your area is as significantly below the steel structure you're interacting with as we see here. At that point, the force is simply being applied by something too sharp for the yield strength to hold any relevance, as the steel will be penetrated long before it will be deflected, which returns to your earlier supposition that "broad, worn" diamonds can deflect steel. You seem to be convoluting these ideas, using one to justify one side of your theory and the other to justify the other side, when the theory doesn't work without both.

Thinking it through, it seems to me that more material is removed where there is more psi. With the edge lifting up because it has less resistance to flexing, and therefore less psi than the rear of the bevel, it wears less. So the fastest removal is where the bevel begins, and there is progressively less until you get to the least erosion at the very flexible edge. This forms a convex bevel. I might be missing something here. If I am, what is it?

A burr is created in front of the bevel if this process is continued, until it breaks off, but the formation of the burr is a result of the hysteresis inherent to the material being exceeded. Or, in other words, the super thin edge bends back and forth, avoiding abrasion until it snaps off, but never constitutes a concavity, really.

@Ian, what do you mean by "edge deflection"?

His theory, as I understand it is that on a diamond plate, rather than removing steel, the diamonds push the edge away via plastic deformation due to the amount of pressure exerted on a <1um area exceeding the yield strength of the steel. Then upon reversing the razor, the steel is again deflected in the opposite direction, creating an issue of metal fatigue where the steel fails and breaks upon the crystalline structure of the steel (ballpark 15-30um and seemingly random when viewed in 2d). Furthermore, his reasoning is based upon an oversized "convexed" edge radius, which he theorizes comes from this deflection resulting in an folded in angle at the beginning of the reverse pass, immediately followed by this reversing of the plastic deformation, (presuming a finite amount of cutting is done before the steel yields/deforms/deflects again). The problem is this theory relies on constantly changing behavior from the steel, which is why sometimes it's based on broad 100+um radius diamond faces and others on tiny <1um diamond points. The faces could cause this deflection... provided the force was applied at the area of the points, but the points would easily cut the steel long before the amount of force the faces are capable of delivering was achieved. It's a catch 22.

spoken36 said:

Thinking it through, it seems to me that more material is removed where there is more psi. With the edge lifting up because it has less resistance to flexing, and therefore less psi than the rear of the bevel, it wears less. So the fastest removal is where the bevel begins, and there is progressively less until you get to the least erosion at the very flexible edge. This forms a convex bevel. I might be missing something here. If I am, what is it?

A burr is created in front of the bevel if this process is continued, until it breaks off, but the formation of the burr is a result of the hysteresis inherent to the material being exceeded. Or, in other words, the super thin edge bends back and forth, avoiding abrasion until it snaps off, but never constitutes a concavity, really.

Click to expand...

This is a bit more interesting, and can hold somewhat true, but generally the increased wear is not at the back of the bevel, but favoring the second half of the middle. It depends on the razor, grind, steel, bevel size, and user. There is flex away from the diamond plate. There is always flex away from the honing surface, to a varying degree. Flex to the extent that we see plastic deformation would become brutally obvious and require either absurd pressure from the user (possible on a diamond plate if you WANTED to, you'd never do this by mistake), or a ludicrously slow, or even nonabrasive hone. You could probably do it on a sheet of highly polished glass with a sufficiently fine edge.


To your second point. In general, Diamond plates are sufficiently aggressive cutters that this "wire edge" doesn't happen. It can on the 8k though (and on others, again if that's your intent). And I suspect it may be a reason why finer Diamond plates aren't being made. It seems this becomes an issue only as you move up in grit (finer edge, less aggressive material removal). This doesn't seem to be the result of the material failing due to fatigue, however, but rather of excessive force (We are not applying force to the edge remember, but rather to a point behind the edge... you're pushing a tent into the ground by pressing down at the top) creating a case where the bevel is quite simply honed away behind the edge. It's why an understanding of pressure becomes somewhat necessary on the 8k on lower grind angle razors.

Sensitivity to, and control of pressure: that's where the magic is.

Remember, the diamonds are fixed. They can penetrate the steel so far and no further. Yes, you can bend an edge on a diamond plate. You're bending thin steel on thick steel once the diamonds have bottomed out. The diamonds cut steel. That's what they do. They're diamonds. And if you press a thin edge into a thick plate, the plate will do what a plate does, deflect the steel, yes, until the steel experiences plastic deformation. Blaming this on the diamonds fails to see the forest for the trees.

Hmm, I can see you aren't understanding. I had the same problem at first. The edge doesn't flex away and flex back. It's being bent permanently. When the first side is facing the diamond plate, the apex gets pushed up beyond the plane of the opposite bevel and stays there. When the razor is flipped for the next stroke on the opposite side, that little bit at the apex that is standing proud of the plane of the bevel gets cut away by the beginning of the stroke before the apex again gets pushed away and stands proud of the plane of the opposite bevel. After a number of strokes a convex is formed only at the very apex which has a much finer finish and straighter edge than could possibly occur if the apex hadn't been flexing away.

No, I understand that's what you believe is happening. I doubt that is the case, as anything which is plastically deformed is removed on the opposite stroke, rather than being returned and pushed in the opposite direction. In order for this not to happen, the fold would have to be happening so far back from the edge that the edge would fold back before the diamonds could remove it; essentially the burr would be significant enough the hone failed to remove it. This requires an extremely slow hone (which diamond plates are not) or an insane amount of flex in the bevel due to extremely excessive pressure from the user (certainly possible, but not really the fault of the plate).

What you describe can certainly be achieved; but it is not demonstrated in most use, and when done is very evident during use. You can wire and break off edges by doing this as much as you want on a diamond plate, but you won't fail to notice it. You can feel the wires coming off. Now does burring and cutting the burr convex an edge as you suspect? Arguably it must. Does it do so on a diamond plate any more than any other hone, or even enough to compensate for the loss of convexing due to no free abrasive or yield in the substrate? It doesn't seem to.

Also, viewed under the scope the end of the Diamond plate edges are far too steep to be described as convexing. They're a full stop. As the other user observed, they are best mimic'ed by a joined edge. The diamonds simply do not penetrate deeper at the edge than they do as you move further up the bevel, which is strictly necessary to create a convexed edge, whether this function is due to the properties of the plate, or bending in the steel. In either case, it would leave patterns that demonstrated the depth of cut reducing as you moved away from the edge. This is not the case. Rather, as you'll notice in this 8k DMT image, the cuts at the edge are if anything shallower, not deeper as you approach the edge. This demonstrates that there is perhaps elastic flexion in the edge, reducing the depth of cuts at the edge, and creating a thicker edge.

I'm a fool. I could have done this much earlier but it didn't occur to me. Visual aids.


One is a convexed bevel I freehand convexed to about 40*... any more than that and it pretty much wont shave arm hair anymore. The other is a joined edge. It's flat on the edge, not convexed, bullet straight to the edge then a dead stop. Same razor both times.

Which does the DMT edge look like?

For the DMT edges to be reflecting as they are due to a convex in the edge, they would need to be convexed to 85+ degree angles in exclusively the last few microns of bevel leading up to the edge. It's simply unbelievable. Even if you choose to believe that to be the case, in function they will surely perform exactly identically to an unconvexed edge of much greater edge thickness in all cases. And certainly a convexing that occurs over less than the edge radius can't even be considered a convexing.

Nope, I don't agree there. First of all, most of my discussion really revolves around this happening on a coarse diamond plate, where the pressure will be inherently higher because of the larger particles, causing more flexing and pushing away of the steel at and near the apex. The finer stones may still exhibit the same behavior but it will be on an even smaller scale, and even harder to see. Then again, they may be able to cut cleanly that much better just due to the smaller surface area of the particles in contact with the steel.

There really is a lot of flex happening near and behind the apex. The steel there is extremely pliant. Having worked with steel for a couple decades doing both machining and grinding of both heat treated and untreated steel, I have gained quite an appreciation for just how flexible it really is. I know that most people think of steel as an extremely stiff material, but at the scale of a razor apex it is anything but.

I think you are overestimating the stiffness of the steel near the apex by quite a lot. And I don't mean any disrespect, but I don't think the photos from a visible light optical microscope can even discern a meaningful amount of information at the scale necessary to learn much about this subject. The electron microscope photos tell the tale pretty plainly and paint the whole picture for me. All the detail in your image is obscured by the flare of light near the apex.

There are some great images at Todd's site here:

https://scienceofsharp.wordpress.com/2015/03/01/the-diamond-plate-progression/

The point is that the "convexing" you describe, if it exists, on a DMT edge occurs in a micron or less of bevel depth. Any more is obviously demonstrated by the increased light reflexion. Even a near 45 degree angle on this razor extends several microns deep into the bevel. The convexing you believe to exist quite simply is not there. If it is there it is to such an extreme angle that either A: it occurs at such a point of edge thinness to be entirely irrelevant; meaning it becomes character of the edge radius or B: it renders the tool unable to cut.

This is pretty clearly demonstrable by these images.

He needs to stop varying pressure on the DMT's. They are not the kind of abrasives that you Grind and then ease off of to erase on. That's why he's running into problems on the finer plates. I expect your theory is exactly the issue he's running into on both the 1200 and 8000 plates. It's nothing to do with the nature of the diamonds as he suspects. He's simply warping the steel against the plate, which the lower grits are capable of overcoming through sheer muscle. The sloppiness of his resulting edges renders his presumptions about convexing irrelevant. I can see areas on his 1200 (and 8000) edge that are convex as well as area's that are concave. The edge is broken up and does look like it's been bent back and forth repeatedly. That edge is crap. The way the edge became wildly jagged when viewed in profile as soon as he left the 600 should have been a big red flag for him.

It looks like your theories are exactly what's happening in his case, but even his description of how he is testing tells me it's user error.

I'm not claiming the theory as my own, just what I think after talking to Todd and seeing his work. I believe this happens due to dull diamonds creating excess pressure, not honing error. The larger diamond particles can't cut as deep due to the larger surface area not being able to exceed the yield strength of the steel across the entire width of the particle. The smaller particles can do that, so they cut deeper creating a more "jagged" edge. Keep in mind those "jagged" edges are at 5,000x magnification or higher - and that the DMT 325 achieved an edge width of somewhere around 1/10 of a micron.

If you have some old throw-away razors why don't you hone three of them up on worn-in diamond plates - one on 325, one on 1200 and one on 8k. They need to be something you don't want back though as he has to cut the blades to image them. I bet I could talk Todd into imaging them and we can compare his vs. yours. I'd surmise that they will look pretty similar.

They won't. Scale his 5k images of the 8k and 1.2k edges down 10x, and they still look like crap. His 8k edge has gouges in it with a 15-25 micron radius. That's visible at 40x, much less 400x. His 1200 and 8000 edges are absolute trash. You could see that with a loupe.

And diamonds aren't presenting their "worn in" surface to the steel. They're presenting an edge. It would defy three dimensional space for the worn in face of the diamond to be presented flush against the steel unless you are pushing the razor into the plate without moving it. The smaller particles don't cut deeper by their nature, nor do the larger particles fail to cut the steel for any reason. The finer grit plates very likely DO allow the razor to yield more to excessive pressure, allowing him to bottom out on the 1200 grit plate much easier than on a 320 grit plate, as there is less surface area of the razor in contact with diamonds to distribute the force he's applying. He's wildly overthinking the fact that he's just using too much pressure.