Knowledge

The steel behind the edge

A sword is a materials problem before it is a smithing problem. Hardness and springiness pull in opposite directions, and a blade needs both at once. Here is how that was handled for two thousand years, and how it is handled now.

It all comes down to carbon

Pure iron is soft and useless as an edge. Add a few tenths of a percent of carbon and the material starts responding to heat treatment — it can be hardened, locked into a fine, hard structure.

Below about 0.3% carbon, steel barely hardens at all. Above roughly 1% it turns brittle and cracks rather than bends. Sword steels therefore live in a narrow window of 0.5–1.0%, and every grade inside it is a different answer to one question: does this blade need to hold an edge, or to survive a bad hit?

The rule

A harder edge cuts for longer. A springier blade comes back straight. You cannot max out both — a sword is always a compromise.

The medieval smith solved it structurally: a soft, springy core with hard edges forge-welded on. A modern workshop solves it with alloy content and precise heat treatment. The goal has not changed.

Material through time

Four ways to good steel

Bloomery iron

to the 14th–15th c.

Ore reduced in a shaft furnace at a temperature too low to melt the iron. What comes out is a bloom — a spongy lump of iron mixed with slag. The smith has to forge it, fold it and forge it again, over and over, to squeeze the impurities out. It stopped being the main method only once the blast furnace spread through Europe; in out-of-the-way places bloomeries kept working for centuries after.

Carbon content in such material is uneven: part of the bloom is soft iron, part is hard steel. The whole skill of the medieval smith lay in knowing which part belonged where.

Pattern welding

2nd–11th c.

Attested on Roman blades as early as the 2nd and 3rd centuries, it peaks in the Migration period and the Viking age. Rods of differing carbon content twisted, forge-welded and drawn into a single billet, with hard cutting edges added on top. The result is a springy core and a hard edge — a way around the problem of uneven material.

The side effect is the pattern on the surface: herringbone, braid, wood grain. By the Viking age it had become a mark of prestige in itself.

Crucible steel (wootz)

from the 3rd c. BC

Indian and Persian technology: iron melted in a crucible with a carbon source gives a uniform high-carbon steel. Slow cooling forms bands of carbides that produce the famous pattern once etched.

This is not the same thing as pattern welding. The wootz pattern comes from the crystal structure rather than from stacked layers — which is why nobody could reproduce it for centuries.

Industrial steel

from the 19th c.

The Bessemer process and the open-hearth furnace give steel of controlled composition in industrial quantities. For the first time a smith can order exactly the material the job needs.

A modern replica is therefore not “worse” than an original — it is made of material a medieval smith could only dream about. The difficulty moved from the metal to the geometry and the balance.

Comparison

Eight steels, eight compromises

Open a row to see what a grade does well and what not to expect from it. The bar values are relative — they compare the grades with each other and are not figures from a standard. The HRC column gives the range typical of the grade, not a limit — a given blade's hardness comes from its own heat treatment, so a sword's data sheet may state a figure outside this table.

GradeCarbonHRCTypical use

Plain carbon steel with no alloying additions. Cheap to work, hardens well in oil, forgives small mistakes at the forge. It deforms permanently under serious contact — which makes it a steel for display pieces and light reenactment.

Resistance to bending: 62
Edge retention: 58
Corrosion resistance: 45

More carbon means more hardness and better edge retention, but less tolerance for error in the quench. The classic choice for blades with real mass behind them: claymores, Zweihänders.

Resistance to bending: 68
Edge retention: 68
Corrosion resistance: 45

The Polish chromium-vanadium spring steel, equivalent to 51CrV4 in EN 10089 — a standardised composition, available with a mill certificate. Chromium (~1%) lets even thicker sections harden right through, and vanadium (~0.15%) refines the grain: the blade copes well with impacts and repeated flexing, and the edge chips less. It gives a little ground to 9260 in resistance to bending, but hardens more evenly in thick sections.

Resistance to bending: 92
Edge retention: 70
Corrosion resistance: 60

A spring steel with silicon and manganese. The blade comes back straight after a bend — which matters most on replicas that end up in inexperienced hands.

Resistance to bending: 85
Edge retention: 60
Corrosion resistance: 55

Chromium added to a spring steel. The most widely used training-sword material in the world: the trade-off between hardness and impact resistance is about as good as it gets.

Resistance to bending: 90
Edge retention: 66
Corrosion resistance: 55

Around 2% silicon buys the springiness this grade is known for: it takes bends that would leave plain carbon steel crooked. The showroom “90° and back” is a seller's line rather than a measured standard — but in its class 9260 really is the most forgiving.

Resistance to bending: 96
Edge retention: 72
Corrosion resistance: 60

A Chinese carbon tool steel of around 1% carbon. Despite what sellers often write, it contains no tungsten — the name gets confused with the American AISI T series. Fine grain, high hardness and a clear hamon, but prone to corrosion.

Resistance to bending: 52
Edge retention: 90
Corrosion resistance: 30

Heat treatment

The curve that decides everything

The same bar of steel can leave a workshop as a springy blade or as a glass rod that shatters on the first hit. The difference is what the temperature did over time.

Hardening

Above the transformation temperature — roughly 780–830 °C for carbon steels — the structure turns to austenite. A fast quench in oil freezes it as martensite: very hard and very brittle.

Tempering

Reheating to 180–250 °C releases part of the stress locked into the martensite. You give up a few points of HRC and gain resistance to cracking. Without this step the blade is useless.

Selective hardening

The spine is coated in clay and the edge left bare. They cool at different rates, giving a hard edge, a springy spine and a visible hardening line — the hamon.

Maintenance

Steel does not forgive

None of the steels used for swords is stainless, and that is deliberate: stainless alloys are too brittle at these cross-sections. The price of toughness is a few minutes of attention once a month.

How often should I oil the blade?: Every four to eight weeks in a dry room, more often if the sword hangs on an external wall or in a garage. A thin film of mineral or camellia oil wiped on with a soft cloth is enough.
How do I remove the first spots of rust?: Light surface rust comes off with oil and a cork, with a rust eraser, or with 0000 steel wool — always oiled, worked along the blade, without pressure. Do not reach for sandpaper or a coarser grade on a polished blade: they leave scratches deeper than the rust was.
Should I keep the sword in its scabbard?: Not for storage. Leather and wood absorb moisture and hold it against the steel. For display, use a stand and keep the scabbard beside it.
Do fingerprints really matter?: Yes. Sweat is slightly acidic and salty — a fingerprint can etch a dull mark into a polish within days. Wipe the blade down and pass some oil over it after every handling.
What about the grip and scabbard?: Treat the leather once a season with a wax-based dressing. Avoid silicone products: they stiffen leather, and after a year it starts cracking at the folds.

You know what it is made of. Now see how.

Five stages from bar stock to a finished blade — and why anyone still forges swords at all.

See the craft