Steel matchup
8670 and 15N20, compared on the five things that decide a blade: toughness, edge retention, corrosion, heat treat and how each one grinds.
Band-saw steel turned knife steel: ~0.70% carbon with a nickel addition that buys elite impact toughness at working hardness, a cheap price per bar, and one of the most forgiving heat treats in the shop. Low wear resistance is the honest tradeoff — this is a chopper's steel, not a slicer's.
The bright layer in damascus that is secretly an excellent standalone knife steel: a 2% nickel bandsaw alloy with outstanding toughness, very fine grain, and the hardenability to run 61–62 HRC in thin blades — easy to heat treat and easy on belts.
Dimension by dimension
Each steel's own note, its score out of five, and what separates the two in that one column. The higher score in a row is marked in purple.
| Dimension | 8670 | 15N20 | What separates them |
|---|---|---|---|
| Toughness | The reason to buy it. Measured at roughly 27–50 ft-lbs around 59–60 HRC (Knife Steel Nerds — batch and machining dependent), which only the 5160/15N20 class matches among common knife steels. The nickel lets it keep that toughness at a hardness where 5160 is usually run softer. | Excellent, and the reason the steel is worth taking seriously on its own. Roughly 2% nickel refines grain and improves impact resistance, while the near-total absence of alloy carbides removes the usual crack-initiation sites. This is what lets you run 15N20 at 61–62 HRC and still grind it thin without chipping. | A family tie at the top of the charts — both ride nickel to the elite class alongside 5160, and testing puts them within noise of each other at working hardness. Nobody has ever broken a knife because they picked the wrong one of these two. |
| Edge retention | Low wear resistance and no vanadium carbides — a fine-grained edge that takes a keen apex, dulls faster than any alloyed steel, and comes back with a few passes on a cheap stone. For its intended hard-use knives that fast maintenance loop is a feature, not a flaw. | Better than its plain chemistry suggests, because in low-carbide steels edge retention tracks hardness — and 15N20's toughness lets it live at 61–62 HRC where 1084 typically runs 58–60. It will not out-slice a carbide-rich steel like D2, but against the simple carbon steels it holds a keen edge longer than expected. | Effectively even: both are fine-grained, low-wear steels with no vanadium carbides, keen off the stone and quick to dull by alloy-steel standards. 15N20's slightly higher carbon-in-solution at times gives it a hair more bite; call it a tie in use. |
| Corrosion resistance | Carbon-steel rust behavior; the sub-1% nickel does not meaningfully slow it (15N20 carries twice the nickel and still only manages a shade better). Oil it, dry it, patina it. | Slightly better than plain carbon steel, entirely thanks to the nickel — the same trait that keeps it bright in a ferric chloride etch slows ordinary staining a little too. It is not stainless and it is not close. Dry it, oil it, and expect a patina on a working blade. | 15N20 by a shade — its ~2% nickel slows rust just enough to notice, which is also why it stays bright in an etched damascus billet while 1084 goes dark. 8670 behaves like plain carbon steel. Both need oil. |
| Heat treat | Austenitize 1500°F, 10–15 min soak (1475–1550°F window; avoid 1450°F and 1600°F+). Parks 50 or AAA both reach full hardness. Temper 2×2h anywhere in 300–450°F — hardness/toughness trade is linear with no tempered-martensite embrittlement in that range. ~58–60 HRC at a 400°F temper; forge heat treating works well thanks to decent hardenability. | Normalize and thermal cycle if forged or damascus-welded. Austenitize at 1475–1500°F with a 10 minute soak, then quench in a fast oil (Parks 50 class) — hardenability is moderate, so speed matters, especially in thin sections. Temper twice for two hours at 300–400°F: 300–350°F holds 61–62 HRC for thin kitchen blades, 400°F backs down to about 59–60 for harder-use work. | Nearly identical, by shared ancestry: austenitize around 1475–1500°F, fast or medium oil, temper 300–450°F, generous margins, forge-friendly. Knife Steel Nerds literally covers them in the same article. 8670's tested recipe is the more documented; 15N20's thin stock air-cools fast enough to demand slightly more quench attention. |
| Grinding | A pleasure. Annealed 8670 grinds, drills and files like the band-saw stock it is, forges smoothly, and its thin common bar sizes (often ~0.080–0.125 inch) mean less hogging to start with. Belts last, files bite, nothing tears. | Easy on belts and files in the annealed state, and forgiving under heat because there are no coarse carbides to tear out. Most commonly sold in thin stock (0.060–0.125"), which grinds fast but also heats fast — keep it moving and dip often. | Both grind beautifully. The practical difference is stock: 15N20 overwhelmingly ships as thin damascus-ready sheet (often 0.065–0.095 inch), which limits standalone designs; 8670 shows up in more knife-usable thicknesses. For a mono-steel blade that alone often decides it. |
Best use
8670
standalone camp knives, choppers and first knives in proper knife thicknesses, at the friendliest price in the tough-steel class.
Open the 8670 reference15N20
the bright damascus layer first and forever, plus excellent thin standalone blades — kitchen knives, fillet knives, working slicers.
Open the 15N20 reference
The long answer
Cousins from the same industrial family: both are band-saw steels toughened with nickel, both forgive forge heat treating, and both sit in the elite toughness class. 15N20 carries twice the nickel and a damascus day job; 8670 carries a lower price and thicker availability. Picking between them is less about performance than about what the blade is for.
Keep comparing
Keep reading
15N20 is a pleasure to grind and very easy to overheat, for the same reason: it is almost always sold thin. Stock in the 0.060 to 0.125 inch range has…
8670 might be the most cooperative steel a 2×72 ever meets. It arrives thin, grinds clean, tolerates heat-treat improvisation, and sharpens on anything…
Every knife steel is iron plus a specific recipe of other elements, and each of those elements pulls the steel’s behavior in a particular direction…
Same bench
Whichever steel wins, the grind is only as flat as the platen behind it. I machine the parts, hand-assemble, track and test Torus grinders in Sterling, Illinois.
Current dual-axis platform with a fiber-laser-cut 6061 chassis and single-knob tracking.
Current dual-axis platform with a steel chassis, more mass and less vibration.
Prior-design aluminum chassis, 90-degree tilt and 1.5-inch tooling arms.
One-man shop
Tell me what the blade is for and how you heat treat, and I will tell you which of these I would put on the bench — including when the answer is neither.
