Katana Steel Metallurgy: Comparing Tamahagane, T10, 1095, and 1060 High-Carbon Blades

A date cut into steel

Among the Japanese blades at the Metropolitan Museum of Art is a katana dated 1526. The catalogue gives a length for the blade edge, a different length for the object as a whole, and a material description consisting of one word, steel. The date is much more specific than the material description. Nothing has gone wrong. Evidence adequate to associate the piece with a year is different from evidence that could identify the chemical composition of its metal. An inscription may be legible to the unaided eye. A distinction between two constituents of steel may require an instrument almost as long as the gallery.

A comparison of tamahagane, T10, 1095 and 1060 starts with this difference in the kinds of evidence that names preserve. Tamahagane carries the history of a Japanese raw material and the people who produced and selected it. The modern designations belong to industrial systems for specifying materials. Folded steel refers to an operation or a resulting construction. Martensite names a constituent that a materials scientist may identify in an appropriate examination. None of the terms performs the work of all the others. This is why a single list of supposed steel types often puts answers to several different questions in the same column.

The four collection photographs that accompany this account show distinct Japanese blades. They allow a viewer to compare the evidence available from a museum catalogue with what can be seen on a screen. The scientific studies discussed below examined other, specifically identified objects. The distinction matters. A researcher may establish a fact about the interior of one sword without authorizing the same statement about a celebrated sword whose image happens to illustrate the article. Keeping those two objects separate makes it possible to enjoy the photograph and appreciate the experiment without making either do a job for which it was never intended.

Mikazuki Munechika, tachi, Tokyo National Museum, via ColBase
Mikazuki Munechika, tachi, Tokyo National Museum, via ColBase. Tokyo National Museum, via ColBase.

What tamahagane tells us

A modern consignment of engineering steel can travel with paperwork linking it to a production batch. The name tamahagane takes the historian in another direction. It leads to iron sand, charcoal, the Japanese tatara tradition, the judgment involved in selecting suitable metal, and a vocabulary formed over generations of work. The Society for Preservation of Japanese Art Swords maintains educational material on that tradition. For an art historian the institutional story is important in its own right. What survives includes knowledge maintained by people at work, and not simply antique pieces preserved behind glass. The relationship between the present practice and earlier examples is a question for evidence in each case.

Traditional smelting could produce metal of differing character within a single product of the furnace. This makes the idea of one chemical percentage for all tamahagane misleading from the start. A single numerical result, if one were obtained, would need a sampling location and an account of what material was sampled. A fragment of raw material at a demonstration, a modern reference specimen prepared for a teaching collection and material from an antique blade have different histories. All three might legitimately enter a discussion of tamahagane. A percentage determined for the first cannot be transferred to the third merely because both appear under the same heading.

The distinction has consequences for collecting histories. A catalogue description written before scientific analysis may preserve an attribution made by a respected specialist. A later exhibition can repeat it, perhaps in translation. A laboratory report may subsequently supply a different kind of information. The three records do not become independent laboratory confirmations simply because a search engine returns them as three results. A reader should ask who first made the attribution and whether anyone has actually examined the material since. A carefully qualified traditional attribution is valuable. Silently turning it into an assay deprives the reader of the chance to understand what sort of knowledge it was.

First-grade tamahagane specimen with its Japanese display label
A piece of tamahagane with its first-grade display label. Photo by さぱしあ, 2015. Original photograph · CC BY-SA 4.0.

What the industrial numbers mean

The familiar designations 1060 and 1095 illustrate a different way of naming carbon steel. In the SAE family of designations the final two figures in these examples point to approximate carbon content in hundredths of a percent by mass. That gives nominal values of about 0.60 percent and 0.95 percent respectively. A nominal value is a convenient centre for a discussion of a grade. An actual specification defines limits, and an actual piece of metal has its own batch history. Saying that 1095 contains more carbon nominally than 1060 is an intelligible comparison. Claiming to know the exact assay of an unidentified object from that comparison would require a considerable leap.

Five steel labels, read as materials
Material Carbon by mass Treatment in context What the hamon shows Historical setting
Tamahagane First grade ≈1.2%; second grade 0.8–1.5% at Nittōho Tatara Traditional differential hardening A boundary may appear in a polished blade; the raw fragment has none. Tatara bloom steel; still produced in Japan.
T10 tool steel 0.95–1.04% Hardening and tempering in toolmaking A grade stamp gives no hamon shape. Modern Chinese carbon-tool grade.
1095 high-carbon steel 0.90–1.03% Hardened and tempered in industrial use Any visible boundary needs an object-specific record. Modern SAE plain-carbon steel.
1060 carbon steel 0.55–0.65% May be supplied normalized or quenched and tempered The number alone establishes no visible pattern. Modern SAE plain-carbon steel.
9260 spring steel 0.56–0.64% Commonly through-hardened and tempered for springs Through-hardening produces no distinct edge-hardening line by itself. Modern silicon–manganese spring grade.

Carbon percentages describe starting materials. The Nittōho grades refer to sorted pieces of a bloom. A finished blade needs its own analytical record.

Sources · Tatara grading · 1060 and 1095 · T10 · 9260

T10 in the Chinese carbon tool steel nomenclature belongs to another system. There the designation indicates a nominal carbon content around one percent. The letter and number must travel together with the naming system. Across industrial history similar combinations of letters have been used for quite different materials. A translated document that silently substitutes the nearest familiar American name can lose a meaningful difference in specification. An editor assembling a historical comparison should retain the original designation in a separate field, alongside any cautious modern description. Equivalence is a claim to demonstrate. Resemblance between two nominal percentages is only a starting point for that inquiry.

The designation 9260 illustrates why reading only the final two digits is insufficient. The grade belongs to a different alloy family from 1060 and is associated with silicon-bearing spring steel. Both names lead the reader toward a similar nominal amount of carbon. They do not therefore identify the same material. Equally, the phrase spring steel can describe a broad family in commercial or historical prose. A source that says spring steel but supplies neither a standard nor a grade has given less specific information than one that supplies both. A historian should preserve that difference instead of completing the missing digits from expectation.

Blade for a Sword (Katana), The Metropolitan Museum of Art
Blade for a Sword (Katana), The Metropolitan Museum of Art. The Metropolitan Museum of Art.

A number such as 1095 has another advantage for modern manufacturers that can become a trap for historians. It is compact enough to survive translation and to look perfectly at home beside a date in a caption. Imagine a label describing a Japanese blade as sixteenth century, high carbon steel. It is easy to rearrange those two pieces of information into a modern grade and an approximate date of manufacture. The modern grade would be new information introduced by the editor. Even finding a local carbon measurement near a nominal value would not demonstrate that the historical material met every requirement of a specification defined long after the blade was made.

Looking below the surface

At room scale steel looks continuous. At the scale of a metallographic examination it may contain regions with quite different arrangements of atoms and constituents. Ferrite is an iron-rich phase. Cementite is an iron carbide. Pearlite is the name for an arrangement of ferrite and cementite, traditionally recognizable in section by its fine alternating structure. This way of speaking matters because one often finds all three words presented as if they named interchangeable families. A pearlitic region can contain both ferrite and cementite. Seeing those two names in a diffraction report does not automatically contradict a description made with an appropriate microscope.

Martensite is associated with a different transformation of austenite, the parent phase. In this transformation the arrangement changes without the long-range redistribution of carbon required for the familiar pearlitic structure. Carbon held in the resulting arrangement and the defects within it influence what an investigator measures. Some austenite may remain. Once again the words describe the material at a certain scale and state. A visitor looking at the bright outline of a polished blade is seeing reflected light from the surface. A scientist identifying a crystal structure is reporting the interpretation of a measurement. The two observations may relate closely while remaining distinct.

The visible hamon gives this distinction particular force. An experienced observer can learn a great deal from the form, texture and continuity of the boundary visible in a good polish. A photograph may convey some of that evidence and lose much of the rest. A change in the angle of illumination can make a subtle feature fade or suddenly dominate the image. If a viewer assigns a precise hardness number to a bright band in a single photograph, the number comes from somewhere outside the picture. No amount of enlarging the pixels can recover a mechanical test that was never carried out.

What a neutron experiment can answer

A study by Stefanus Harjo and colleagues used neutron diffraction to investigate an intact Japanese sword dated to the Keichō era. The object could be examined at multiple positions without cutting a slice out of it. The authors interpreted differences in the measured diffraction patterns in terms of constituent phases and the internal structure of the steel. This is a particularly instructive case for a public collection. A familiar historical claim, that the material varies within a single blade, became a question that could be investigated on a documented object at identifiable locations. The technique gave the researchers access to information beyond the polished exterior.

To understand the achievement, imagine being given only a list of chemical elements. Iron and carbon would be fundamental entries. The list would say little about how the atoms are arranged or whether the same arrangement occurs across the whole object. Diffraction addresses such questions by exploiting the orderly arrangement of atoms in crystalline material. A set of measured signals has to be interpreted with a model, and some signals can be easier to separate than others. This is why a research paper describes its apparatus and analysis alongside its findings. The measurement and the reasoning that turns it into an identification belong together.

Kikkō Sadamune, Tokyo National Museum, via ColBase
Kikkō Sadamune, Tokyo National Museum, via ColBase. Tokyo National Museum, via ColBase.

A map made at several positions also creates an obligation to describe scale. Each measurement represents a finite volume, and the path through an irregular object can affect the result. A published image with neatly coloured regions may give a stronger impression of continuous coverage than the underlying set of measurements supports. The careful reader looks for the position diagram, the scale bar and the authors' account of uncertainty. Those details do more than decorate a scientific paper. They decide whether a finding about one local region can sensibly be compared with a photograph of the entire blade or with a measurement made in another study.

Heat treatment, in the language of a catalogue

Differential hardening describes an intended difference in metallurgical response across parts of an object. Through-hardening describes an intention to obtain a hardened structure through a section. The English phrase clay tempering, commonly encountered in writing about Japanese swords, can cause confusion. In metallurgy tempering is a separate term for a thermal treatment of previously hardened material. A reference to a clay coating in an account of Japanese craft does not, by itself, specify a tempering treatment in this technical sense. Readers comparing a museum translation with an engineering textbook can therefore encounter apparently contradictory descriptions that turn out to be using one English word in different ways.

The historical question concerns what happened to an identifiable object. A museum may provide a visual attribution, a description of the practice associated with a tradition, or evidence from a scientific examination. Each bears differently on the answer. A general account of differential treatment can explain why an expert looks for certain kinds of variation. It does not establish that every point in a particular blade has been examined. Conversely, measuring an internal difference can raise fresh questions about the object's history without recovering every event in the workshop. Scientific investigation often narrows the range of plausible accounts while leaving more than one possibility open.

Folding belongs on a separate line of that description. It concerns the working and joining of material. The expression folded steel therefore does not supply a competing carbon percentage alongside 1060 or 1095. Nor does it give a date or country of production. Patterns in worked metal have long attracted descriptive names and artistic attention. A polished surface may preserve significant evidence of the material's working history, but a visible pattern cannot independently identify all of the constituents beneath it. The appropriate comparison is between a surface observation, an account of construction and any measurements that can be securely tied to the same piece.

Hardness and impact belong to different experiments

Rockwell hardness is method dependent. A result reflects an indentation made under a defined sequence of loads with a specified indenter. The named scale matters, as do a suitable surface and the conditions of measurement. NIST's work on hardness standards explains why reference materials and consistent procedures are needed even in laboratories equipped for that purpose. This is very different from looking up the name of an alloy. A grade describes a material family. A Rockwell C result describes a specified test on a particular specimen. A paper reporting one cannot silently be made to certify the other.

An impact test asks a different question about how a specimen behaves under defined conditions. Specimen shape, the presence and form of a notch, temperature and the reporting convention all affect what a result means. A common classroom analogy is to compare reading a mark left by a pointed load with recording the energy absorbed by a test specimen in a machine. Both observations can be informative. Neither is a substitute for the other. In an intact historical object even an indentation can be unacceptable, and sacrificing a specimen for a destructive test would pose a rather different curatorial question.

For that reason a museum collection offers no ordinary basis for a league table of impact toughness or HRC values across its named treasures. The absence of those numbers is compatible with excellent scholarship. Some inquiries can be answered by examining a historically documented fragment, another by studying a modern reference sample, and yet another by nondestructive examination of an intact work. A researcher who transfers a result between those categories should explain why the comparison is justified. A catalogue that leaves a field blank may be preserving a distinction that a more impressive-looking table would erase.

One further distinction concerns the ownership of the data. A museum may publish an image under an open licence and encourage viewers to consult the record. That licence permits appropriate reuse of the photograph. It does not imply that the institution has endorsed a new identification proposed by someone reusing it. Giving the accession number makes an interpretation testable. Giving a separate reference to an experimental paper makes it possible to see exactly which object the scientists studied. In this field, two well-placed references can prevent an error that would otherwise be carried forward into dozens of later accounts.

What to take away from a collection photograph

Look again at one of the four photographs, this time with the object record open beside it. Write down just three claims. One might describe a visible feature of the photograph, one might record an attribution or date from the catalogue, and one might identify a question that could in principle be answered by scientific examination. Now ask what would cause each of those claims to change. A better photograph might overturn the first. A newly identified document might alter the second. Only a suitable investigation tied to the particular object could settle the third. Distinguishing their sources makes the comparison much more informative.

There is considerable pleasure in leaving those distinctions intact. A visitor can appreciate the way a polished surface records and reflects light, follow a historical account of tatara steel and understand why an industrial metallurgist needs exact designations. None of those pleasures requires an invented measurement. The dated blade of 1526 can remain exactly what the museum knows it to be, while the Keichō-era sword studied with neutrons supplies a different and unusually intimate kind of evidence. Together they show how a seemingly simple question about katana steel can bring a workshop tradition, a system of modern standards and the interior of one old object into the same conversation.

Frequently asked questions

What is tamahagane?

Tamahagane is steel associated with the Japanese tatara tradition. It names a historically specific material whose selected pieces can vary in composition.

What is T10 steel?

In Chinese carbon tool steel nomenclature, T10 indicates a grade with nominal carbon content around one percent. The naming system matters when comparing it with an American designation.

What is 1060 carbon steel?

In the SAE system 1060 is a plain carbon steel designation with a nominal carbon content around 0.60 percent by mass. The name describes a grade, while an actual batch has a measured composition.

What is folded steel?

The phrase describes a working or construction history. By itself it supplies neither a unique chemical grade nor an assay of a particular antique object.

What is pearlite?

Pearlite is a microstructure composed of ferrite and cementite. On an appropriately examined section it can show a fine layered arrangement.

What is Rockwell hardness?

It is a scale-specific result of a defined indentation test. An alloy name or a photograph of a polished surface cannot establish an HRC result for a historical blade.

Films and talks

The Met — Art of the Samurai—An Introduction

Birmingham Museums Trust — Hoard Conservation Blog: X-Ray Fluorescence (XRF) Machine

Collection records and further reading

Related Articles