Paper by Prof. Omura: Was tamahagane really used for Japanese swords?

[OceanoNox]

The title seems like clickbait, here is the original title and reference info:

 

日本刀の素材は本当に玉鋼か？刀剣類の分析から素材鉄の変遷を考える, 大村 紀征, 金属 Vol.94 (2024) No.7

 

The conclusion of the paper is that tamahagane was mainly used for Japanese swords merely during the Bakumatsu (specifically the last 70 years of the Tokugawa Bakufu). Arguments are:

1. Little local production of steel, that did not match consumption and demand;

2. Large imports (either official or contraband) of iron and steel from the Chinese mainland and the Korean peninsula, with added metal from Europe later;

3. Steel from the Chinese mainland contains more copper than the iron sands in Japan, and several extant swords that were analyzed show a copper content congruent with Chinese steel.

4. The whole "hard steel for edge and outside, soft steel for the back and inside" construction of nihonto is relatively new, and older swords (in particular koto) were made with mixing soft and hard steel, folding/twisting them together and forging the sword from the block (same as several swords found in Europe too).

5. Analysis of some iron/steel items imply they were made by mixing non-Japanese pig iron and native iron sands.

 

Personal impression: The author does not appear to mention export of swords from Japan to China, nor the pervasiveness of recycling and oroshi. He doesn't cite other studies that show virtually no copper in Japanese swords from Kamakura to Edo (Prof. Kitada, 日本刀の材料科学), i.e. the use of native iron ores to make Japanese swords. I am yet to understand how they obtained (maybe from another study?) the numbers for production and demand of iron/steel items during the history of Japan. The paper itself is certainly trying to shake the established idea, relatively new according to the author, that nihonto = tamahagane, but as of yet, I lack the time to review the sources and studies cited in the paper, and the knowledge to agree/disagree outright.

 

[Bugyotsuji]

Interesting ideas there. 

First thought is to wonder if kazu-uchi-mono may have short-circuited the laborious tamahagane method for the sake of larger production volumes.

[Rivkin]

I don't have the article, but the first question I would ask is how they measured content. If its XRF I would take it with a largish grain of salt since its a fit, and most companies producing testers worry a lot more about expensive soft metals than inexpensive iron. Its very common for it to shown considerably more soft metal in iron then there actually is.

[Robert S]

11 hours ago, OceanoNox said:

4. The whole "hard steel for edge and outside, soft steel for the back and inside" construction of nihonto is relatively new, and older swords (in particular koto) were made with mixing soft and hard steel, folding/twisting them together and forging the sword from the block (same as several swords found in Europe too).

 

 

I wonder what the evidence is for this claim in particular.  I would expect that there would be few blades out there which have been polished through to the core steel if that were true... but it seems that they are not uncommon.

[ROKUJURO]

13 hours ago, OceanoNox said:

......4. The whole "hard steel for edge and outside, soft steel for the back and inside" construction of NIHONTO is relatively new, and older swords (in particular KOTO) were made with mixing soft and hard steel, folding/twisting them together and forging the sword from the block (same as several swords found in Europe too).
 

5. Analysis of some iron/steel items imply they were made by mixing non-Japanese pig iron and native iron sands....

4)  This "method" would make no sense, and every smith would have known that. Even if you realized a proper homogenization of both materials, you would just lower the overall carbon content. Useless.

5)  Pig iron is the lowest quality cast iron with many impurities and detrimental content. It cannot be forged because of its high carbon content and requires a lot of refining work to be usable. Iron sand is not iron but a form of iron ore! It has to go through a direct reduction process (like TATARA) to make iron out of it.

The text does not sound very competent in parts.

[OceanoNox]

1 hour ago, ROKUJURO said:

4)  This "method" would make no sense, and every smith would have known that. Even if you realized a proper homogenization of both materials, you would just lower the overall carbon content. Useless.

I cannot say whether it's useless or not, but the paper, and others that deal with Western swords made in the same way, shows a banded structure in the cross section. I think the idea is less to have a uniform carbon content and more to have a composite structure "throughout".

There are also swords that seem to be made of one type of steel.

 

To be honest, my feeling is that the author worked from the conclusion backwards to find fitting evidence, instead of gathering evidence to come to a conclusion. I don't think they misrepresented the sources they cited, but I was left wondering how representative their examples were.

[Rivkin]

My interpretation of the old swordsmithing process is its basically adaptability to the fact that any bloomery, tartara in particular, is going to produce a very large range of material properties. Small portion of the load is going to be relatively perfect sword-ready steel. So you either go into piece-by-piece refinement techniques like they did in the Middle East or your forging and heat treatment has to account for the fact. It can be banded structure, it can be multi-layers. There is analysis which clearly shown multi-layers in more or less all Asian swords from the 10th century, but then it also shows that some forged from just iron, some had an opportunity to select specifically steel, sometimes you find some weird admixtures etc..  And each one of them adapted heat treatment - old multi-layers will always gravitate to narrow differential heat treatment of suguha variety, but with other compositions you have other options.

 

Problem of Japanese scholarship is they don't read nothing of this, they follow Japanese publications based on testing like two or 10 blades from the entire history of nihonto and make a big conclusion out of it.

[OceanoNox]

I have to admit I haven't read much outside of academic work on Japanese swords. I am trying to remedy that by reading stuff written by Alan Williams, and get to other stuff, the goal being to understand the similarities and differences in methods and goals when making swords (I am still not sure why the Japanese apparently aimed for such high hardness).

That being said, so far, the body of academic work on Japanese swords, in particular material science and mechanical properties, is quite consistent, whether it's about old swords or modern ones.

[Hoshi]

Truth is we don't know. 

 

The author has an axe to grind about the treatment of Gunto and the requirements that art swords are made according to "traditional methods" which are defined as using Tamahagane, etc. 

 

With this caveat in place, he's right to point at the massive gap in scholarship when it comes to the nature of the iron used prior to the Muromachi era. 

 

The 'state of the art' metallurgical studies usually analyze a few low value blade from the Muromachi period. You can't draw conclusions on such tiny samples with non-representative blades and extend this to the golden age of Nihonto, the Kamakura period. 

 

Quote

I don't have the article, but the first question I would ask is how they measured content. If its XRF I would take it with a largish grain of salt since its a fit, and most companies producing testers worry a lot more about expensive soft metals than inexpensive iron. Its very common for it to shown considerably more soft metal in iron then there actually is.

 

Grain of salt yes, but would personally love a properly calibrated XRF-bases study, with multipoint measurements, over a vast collection of treasures.

 

For now, we just lack in-depth scholarship in the topic. 

 

[ROKUJURO]

There is some more recent information available

mechanical-properties-of-samurai-swords-carbon-steel-made-using-atraditional-steelmaking-technology-tatara-2169-0022-1000162.pdf

[The Forest Ninja]

I'm not sure how many of You know this, but the Tatara is broken up and metal sorted.

1. The pieces with low carbon can be reused to add more carbon in a reducing environment.
2. The pieces with too much carbon (aka cast iron) can be reused in an oxidizing environment to reduce its carbon content.
3. The pieces that are good from the intial run are used as is.

[OceanoNox]

Some of the results come from a study made by Nippon Steel (新日本製鐵第一研究所) using CMA (Computer-aided Micro Analyzer or Comprehensive Multi Analyzer). It seems to be like EPMA but faster and with a larger analysis area. Unfortunately, the paper does not give the details of the research report...

[Rivkin]

Destructive testing. Probably the only serious way to go, but then ofcourse there is nothing to destroy between the end of Kofun and late Heian, so one should rely on Asian data instead. 

Availability of items for destructive testing is what limiting them - couple of Muromachi pieces, one or two before but plenty of Edo items. I liked Kimura's data but that's pretty much how many pieces he worked with.

Kamakura there will be very few opportunities also, probably ko Naminohira is the only thing that can be haid that is not unreasonable pricewise... Maybe some late Kamakura pieces.