Field record
The Smallest Apex Angle
that survives the work.
A record of how knife steels actually cut, measured at the smallest micro-bevel angle each steel still holds to 600 push cuts. Every number rides with its test count, so you can see how far to trust it.
- 58
- steels tested
- 201
- recorded tests
- 13°
- smallest passing angle
- 45,228
- push cuts documented
- 45
- brands documented
Last test 2026-07-25 · 52 of 201 reached 600 · HRC recorded on 9 rows
Most tested, never passed
D2
14 tests · 0 passes
Best attempt: 419 cuts, still 181 short of the pass mark.
Most tested, passed
MAGNACUT
14 tests · 8 passes · ties on tests: S35VN
Smallest pass: 13° per side, across 2 makers.
Near misses: these steels came closest without passing
How the test runs
One wood, one push cut, one pass mark, all held steady. Across tests, two things move on purpose: the steel and the micro-bevel angle.
Cut in the secondary bevel
Behind the maker's primary grind, we sharpen in a secondary bevel at about 11 degrees per side, so the apex sits as acute as the steel can take.
Set the micro-bevel
A small micro-bevel apex finishes the edge. Its angle is the number we track: lower angle, more acute, more capable but more fragile.
Push-cut spruce
The knife push-cuts A-grade FAS spruce, the same wood every test, so results stay comparable.
600 cuts to pass
Reaching 600 push cuts is a pass, and the count caps there. A lower number is where the edge gave out.
The headline number: the smallest feasible micro-bevel angle
For each steel we report the smallest feasible micro-bevel angle: the most acute, lowest-degree micro-bevel angle at which the steel still passed 600 cuts. A lower angle cuts better but holds up less, so the smallest angle that survives is the steel's feasible edge.
It is a synthesis across an uneven sample. Some steels carry many tests, some carry one. That is why the count of supporting tests sits next to every number, and why a steel with one test is marked as resting on one test. We never show a bare angle as if it were settled.
Two things that still shift the result
We sweep the steel and the angle on purpose. Two more things shift what an edge holds without being part of that sweep: how it was hardened, and damage left in the edge from earlier grinding.
Heat treat by brand
The same steel behaves differently depending on who hardened it. At the same micro-bevel angle, one maker's heat treat can clear 600 cuts where another's gives out early.
Edge heat damage
Grinding can leave a heat-damaged layer in the steel near the apex. An edge taken down into that layer fails early, and clearing it can take repeated sharpenings: the one such case in our data took four before the edge came back. We show those results as context, not as a limit of the steel itself, so a heat-damage failure never lowers a steel's feasible angle.