Why a Knife Edge Fails on Geometry, Not Hardness

This page contains affiliate links. If you buy through one we may earn a commission, at no cost to you.

Yes, but not in the way the question means, and the distinction is the whole answer. Two different things get called sharpness. One is the radius of the very tip, which decides how little force it takes to start a cut. The other is the angle between the bevels, which decides how much sideways load the tip can survive. They are set independently, they are not traded against each other, and only the second one has a penalty.

That penalty is a square law, and it is large enough to settle an argument. Going from 20 degrees per side to 15 removes 45.8 percent of the edge's lateral strength. The entire hardness range the kitchen knife category offers, from about 50 Rockwell C up to 61, buys back 40.4 percent. Five degrees of geometry outweighs every steel in the shop.

Sharp and acute are not the same specification

Picture the last micron of an edge. Two ground faces come together, and where they meet there is a rounding, because no two surfaces meet at a mathematical line. That rounding is the apex radius, and on a well finished edge it is a few tenths of a micron, while on a tired one it is tens.

The apex radius is what you feel. The force to start a cut in a soft solid scales with it, which is why a freshly finished knife bites a tomato skin and a dull one slides. And it has nothing to do with the angle: you can bring a 25 degree edge to the same apex radius as a 10 degree one. It takes the same abrasives and the same care.

So a knife cannot be made too sharp in the sense of the radius. What it can be is too acute, and that is a separate dial on the same tool. Our own guide to honing versus sharpening is careful and correct about the related point: the very tip is "thin and slightly flexible", it "rolls or leans to one side" in use, and a rod straightens it without removing metal. This page is about how much load that tip can take before it does that, and the answer is almost entirely geometric.

The square law

Treat the apex as a wedge. At a depth z behind the tip, the thickness of steel across the wedge is twice z times the tangent of the per side angle. At 20 degrees per side that is 0.728 of z. At 15 degrees it is 0.536. So the finer edge has 26 percent less material at every depth, which sounds modest.

It is not modest, because what resists a sideways load is a bending section, and a bending section's capacity goes as the square of its thickness. Square the tangent and you have the whole relationship: the lateral load an apex can carry is proportional to the square of the tangent of its angle.

Run the angles knives are actually ground to, all relative to 20 degrees per side. Ten degrees carries 23.5 percent. Fifteen carries 54.2. Sixteen carries 62.1. Seventeen carries 70.6. Eighteen carries 79.7. Twenty five carries 164.

A NOVKSO 7-inch santoku knife publishes both numbers that matter, which almost nothing in this category does: a hardness of HRC 60 to 61 and a "15-degree per side edge", on a 2 mm blade. Its own record says the edge is "geared toward precise, effortless cuts", which is exactly right and exactly what the geometry delivers. It also publishes, among its drawbacks, that the "Thin blade not for chopping bone", which is the same arithmetic stated as a warning.

Hardness is linear, and the range is narrower than it looks

Now the other variable, worked the same way so the two can be compared.

The yield strength of hardened steel tracks its hardness closely. Convert Rockwell C to Vickers with the standard table and then use the Tabor relation, that indentation hardness is roughly three times yield. Fifty Rockwell is 513 Vickers and about 1,677 megapascals. Fifty six is 613 and 2,004. Fifty eight is 653 and 2,135. Sixty one is 720 and 2,354.

From the softest kitchen knife steel to the hardest, that is a gain of 40.4 percent. Not nothing. But it is less than the 45.8 percent you give away by taking five degrees off the bevel, and it is a long way short of the 76.5 percent you give away going from 20 degrees to 10.

Which means one line on our own guide to knife steel and hardness needs correcting. It says of the Rockwell figure that "That single figure drives the most important trade-off in any knife". It does not. It drives the second most important one. The same page is right about everything else, including the part people get wrong: that hard steel chips while soft steel rolls, and that "Sharper is not automatically better". The correction is one of rank, not of fact, and the rest of that page stands.

It is worth being precise about what hardness actually decides, because it is not the load. It decides the failure mode. A softer, tougher apex yields plastically and bends over, which is the rolled edge a honing rod puts back. A harder apex has less capacity to deform and fractures instead, which leaves a chip that honing cannot touch. So hardness decides whether a failure is recoverable, and angle decides at what load it happens. Two jobs, routinely described as one.

Our own catalogue proves it, and gets the credit backwards

Multiply the two together and you get a comparative index for the apex: the square of the tangent of the angle, times the yield strength. It has no units and only its ratios mean anything, but the ranking it produces is instructive.

A Dalstrong 8 inch chef knife in forged German steel publishes "56+ HRC" and a "hand-polished 16-18° edge per side". Take the midpoint, 17 degrees, and its index is 187.3. That is the highest of the three knives in our catalogue that publish both figures, and it is the one with the softest steel of the three.

The santoku above, at 15 degrees and HRC 60 to 61, scores 166.1, which is 11 percent below it. And a TILIBEIKA 6-piece knife set publishes "15-degree factory edges" in 3Cr13Mov, which sits around 50 Rockwell, giving an index of 120.4: 36 percent below the Dalstrong despite having exactly the same bevel angle as the knife that is eleven points harder.

So the softest published steel on our shelf has the toughest edge on our shelf, by a margin the hardness figures would never predict. And here is where the records get it backwards. The Dalstrong's own drawbacks list "16-18° edge angle is less acute than some Japanese knives" as a weakness. On this arithmetic it is the single strongest thing about the knife, and the reason its edge survives the rock chopping its own performance text describes it as built for. The listing has correctly identified its most durable feature and filed it under problems.

The knife that leaves out the bigger variable

A WÜSTHOF Classic 8 inch chef's knife publishes a hardness, "tempered to 58 HRC", and a sharpness claim, that its edge technology "is claimed to make the blade 20% sharper than previous models and hold its [edge longer]". What it does not publish anywhere in prose is an angle.

Put that through the index. At 58 Rockwell and 15 degrees per side it scores 153.3. At 58 Rockwell and 20 degrees it scores 282.8. The unpublished figure moves the answer by 84 percent, which is more than twice the span of the entire hardness range this page started with. The number the listing does give you resolves the smaller half of the question.

That is not a criticism of the knife, which by reputation is ground somewhere in the middle of that range. It is a criticism of what a specification sheet chooses to carry. A hardness without an angle is like a wire gauge without a length: it is one of the two numbers you need and it is not the dominant one.

Two more records concede the gap outright rather than papering over it. A Victorinox Fibrox Pro 8 inch chef's knife lists among its drawbacks that the "Exact blade steel not specified", and a Gidli 8-inch chef knife set with diamond sharpener says plainly that "Blade thickness and hardness are not stated". Both of those are honest, and both are more use to a reader than a confident number with no provenance. Across seven knives in this catalogue, two publish an edge angle, and one of those two publishes it as a band spanning 22 percent of its own strength.

Why going finer always buys less than it costs

The reason there is an optimum, rather than a direction, is that the benefit and the cost have different exponents.

The force to drive a wedge into a soft material has a component proportional to how far sideways the material must be pushed per unit of depth, which is the tangent of the angle. So the cutting benefit of a finer edge goes as the tangent, while the strength cost goes as the tangent squared.

Put numbers on the usual step. Going from 20 degrees per side to 15 reduces the wedging force by 26.4 percent and the apex strength by 45.8. Going on from 15 to 10 buys another 34.2 percent off the force and costs another 56.7 percent of what is left. The ratio of what you gain to what you pay is one over the tangent of the angle, so it gets steadily worse the finer you go, and there is no angle at which it turns around.

Which gives the real answer to whether a knife can be too sharp. Yes, and the threshold is not a number of degrees. It is wherever the lateral loads of your actual cutting exceed what the apex carries. If you push sideways through a squash, twist out of a cut, or scrape the board to gather what you chopped, you are applying exactly the load this arithmetic is about, and no steel in the category supplies enough margin to make a 15 degree edge behave like a 20 degree one.

The micro bevel, and what it is actually doing

There is a way to have both, and the arithmetic explains why it works rather than sounding like a compromise.

Grind the blade at 15 degrees per side, then put two or three passes at 20 degrees on the very apex. The result is a tiny 20 degree facet a few microns deep sitting on a 15 degree primary. Because the strength of the apex is governed by the material in the first few microns, the apex now behaves like a 20 degree edge: full strength restored. And because the wedging work of the cut happens over tens and hundreds of microns of depth, the 15 degree primary still governs how easily the blade passes through the food.

So a micro bevel is not a dulling step and it is not a hedge. It is the only move available that separates the two variables this page has been separating all along, and it is why it keeps turning up in sharpening practice without a stated reason. Our page on keeping scissors and garden tools sharp is right that "each blade has a specific bevel" and that matching it matters, which is the same principle in a tool where the bevel is set by the manufacturer and not by you.

That last point has a practical consequence on our own shelf. The TILIBEIKA set publishes a 15 degree factory edge and, as a feature, a storage block with a built in pull-through sharpener, and its record is honest that the "Built-in sharpener is pull-through, for upkeep not precision". A pull-through imposes its own fixed angle on whatever you draw through it. If that angle is wider than 15 degrees, and on general purpose devices it usually is, then the set's published factory geometry survives a few passes and then becomes the block's geometry instead. That is not a fault. By the index above it raises the apex strength by up to 85 percent, which on a mid range steel is probably the right trade. But the two features are being sold as complementary when one of them overwrites the other, and nobody says so.

The honest limit

The wedge model is first order. A real apex does not fail cleanly in bending: it rolls, microfractures, or both, and the balance shifts with hardness in a way a single section modulus cannot capture. Treat the index as a ranking rather than a prediction of when something will fail.

The Tabor relation between hardness and yield is good to perhaps ten percent for hardened steel, and the Rockwell to Vickers conversion is a standard table rather than a measurement of any particular blade. The apex radii quoted, tenths of a micron for a good finish and tens for a worn edge, come from edge metrology rather than from any product record, and no knife listing we carry publishes one. Nor does any publish a micro bevel. Those are the two figures that actually determine how sharp a knife is, and the category publishes neither.

What survives all of that is the exponents. Strength goes as the square of the angle, cutting benefit goes as the angle, hardness enters linearly and spans only 40 percent across the whole category. Change every constant on this page and a knife ground at 15 degrees still has an edge that fails at about half the sideways load of one ground at 20, and no steel will rescue it.