I’ve titled this article “carbides in K390” with the idea that it will be part of a series that aims to help understand the role of carbides in knife performance and sharpening. The first article in this series, Carbides in Maxamet, dispelled the myth that carbides are weakly bonded to the surrounding steel matrix. Of particular interest in this article is the poorly understood interaction between relatively coarse (40-50 micron) grit stones and the 1-2 micron carbides found in powder metallurgy steels. As usual, we should be cautious about generalizing these observations to all examples of K390 or similar steels, or jumping to conclusions about how these microscopic observations translate to macroscopic performance. However, the observations I show here are consistent with those I have made in other samples – while they may not be the rule, they are definitely not the exception.
The particular knife used in this article is a Spyderco Endela in K390 that I purchased from Blades Canada. K390 is one of the currently popular high (9%) vanadium powder metallurgy steels. With a carbide volume fraction of around 17%, carbides in or at the apex play a major roll in how a knife made from this steel will perform and will affect the sharpening protocols required to maximize that carbide-enhanced performance.
The factory edge shows the typical morphology of carbides exposed by buffing. Grinding lines are visible in the bevel 20 microns away from the apex. Exposed carbides with matrix shadow trails dominate the last 20 microns of the bevel.
The factory edge performed adequately, slicing paper cleanly.
High vanadium steels like K390 can be challenging to sharpen as vanadium carbides are harder than typical aluminum oxide abrasive and grind/wear relatively slowly. In particular, this wear-resistance may lead to difficulty with burr formation as the apex is more likely to bend away from the stone rather than being cleanly abraded. This repeated bending of the apex may also weaken the steel, leading to micro-chipping. It is sometimes claimed that these micron-scale carbides have minimal impact on coarse grinding, with unsubstantiated claims that coarse grit stones can simply scoop out or “pop-out” the carbides. The observations I show here suggest that diamond and silicon carbide are able to abrade the vanadium carbides and remove steel sufficiently well to form a new apex. The cut depth is less than the typical carbide diameter – there is no evidence that carbides are scooped or popped out during sharpening.
Shown below is the K390 blade sharpened on a DMT Coarse diamond plate, where a moderate burr was formed. This the default result where minimal effort is made to avoid/remove the burr. This particular burr can be felt when brushing fingers off the edge. Cracked carbides are visible near the end of the burr, presumably broken during repeated flexing of the apex.
Another interesting observation is that the carbide near the surface appears to be pulverized and partially smeared out in the direction of sharpening while carbides one micron below the surface are intact.
The knife was sharpened again on the DMT Coarse, but this time with alternating, edge leading strokes, which still results in a triangular burr, although not one that can be felt or perceived in any way. Again, the part of the burr that was flexing on the diamond plate displays cracked carbides. The burnishing action of the coarse diamond plate can produce a keen edge, however the cracked carbides will likely reduce its wear resistance. This is also likely one mechanism by which microchips form.
The knife was sharpened again, this time on a Shapton Pro 320, freehand at about 33 degrees inclusive. A much smaller burr results, as compared with the diamond plate, with cleanly abraded carbides at the bevel surface. This small burr could not be detected by traditional methods.
The knife was sharpened again, this time on a Sigma Power Select II 240 grit stone at around 31 degrees inclusive angle. Once again, a relatively small burr is formed, and one that is not detectable by traditional methods. The sharper abrasives and the presence of loose particles in the mud likely minimize the burr formation as compared to the diamond plate.
Fractured carbides at the surface of the bevel suggest that the abrasion mechanism involves fracturing the carbides rather than simply abrading them or “removing them whole.”
After sharpening on a Shapton Pro 320, the knife was micro-bevelled using a translucent Arkansas stone. This type of stone generally removes steel by adhesive wear and its silicon oxide composition is much softer than the vanadium carbides in the K390 steel. Not surprisingly, it is relatively ineffective, but it does produce some interesting results. First, as with the factory edge, carbides are exposed as the matrix is wears faster than the carbides (if the carbides wear at all). Based on the amount of metal removed, this procedure has likely just exposed the carbides previously shaped by the coarse stone.
Close examination of the carbides on the micro-bevel surface show some evidence of wear (or flattening) of the carbides. There is no evidence of carbides being dislodged or “popped out.”
For comparison, the Spyderco Native in Maxamet steel was sharpened in the same way and micro-bevelled with the translucent Arkensas. Again, I was not able to find a hole or pit that might indicate a “popped out” carbide, but the one shown below did appear mostly excavated.
To investigate, I cross-sectioned through this particular vanadium-rich carbide and the tungsten-rich carbide on its right.
Contrary to popular belief, coarse stones do not cut deep enough to “scoop out” 1-2 micron diameter carbides in these very hard steels. Instead, the mechanism appears to be that the carbides are abraded/worn in place, flattening and thinning them until they are thin enough to shatter and be removed along with the metal swarf.