A short presentation on the metallurgy in Japanese swords

[OceanoNox]

Nihonto-Steel.pdf

Hello all,

I have been recently quite miffed at the apparent return of the "Japanese SteEL was Baaaad" cliche on the internet. I've yet to seen anyone back their claim and decided to gather some of the most excellent data that has been published throughout the years. A lot is in Japanese, which means it remains inaccessible to many.

At any rate, feel free to take a look at it. I have focused on the metallurgy, i.e. the chemical composition, microstructure, and mechanical properties.

 

Cheers

[Brian]

Will add this to the downloads section. Thanks!

[OceanoNox]

Thank you Brian!

[Pierre F]

It's fascinating. I didn't know, I'm pretty sure I didn't understand everything. I would read it again.

 

thanks OceanoNox.

[Tohagi]

That's à piece of science and knowledge!

 

Thank you for sharing,

 

Éric VD 

[Jussi Ekholm]

Wonderful Arnaud! Will have to do better reading with time and some of the info is too advanced for me to grasp. Great article!

[DoTanuki yokai]

Im not cool with the generalization of what water and steel temperatures are used for yakiire. 
Otherwise I think it is a good summary of the available information. 

[OceanoNox]

On 8/30/2024 at 3:53 PM, DoTanuki yokai said:

Im not cool with the generalization of what water and steel temperatures are used for yakiire. 

I used the data available. I'd like to keep updating this file, so if you have other sources, please let me know.

EDIT: I wonder if the smiths themselves measure the temperature or keep doing it by eye and experience.

Edited August 31 by OceanoNox

[Robert S]

Nice work!

 

A few comments:

1. The "Vickers hardness in cross section" page at the beginning of the Mechanical section is in there twice, once with an added "antique" label, and once without.

2. The data presented on trace element concentration (Phosphorus and Sulphur) in iron sands indicates that these elements tended to be high enough to increase brittleness, and there seems to be limited effect of forging/folding on the concentration of these elements.  Despite this, the antique blades tested seemed to have achieved lower trace element levels... so either they were using better iron sands, or eliminating some of these elements through the bloomery process, or they were able to reduce the concentration of these elements through forging or other processes.

3. The relative consistency of carbon content in the edges of the antique blades is pretty remarkable, given the variation on the iron sands.  I'd love to know how they determined carbon percentage, since they clearly had pretty good ability to control it.  Maybe it was just trial and error with a given iron sand/bloomery process and resulting grades of tamahagane, resulting in a rule of thumb, but maybe there were some other indicators in heating, color or forging properties.  I've done some knife making, but of course with modern steels I was starting with tightly controlled steel grades, so never worries about this.

4.  Your summary of the forging and folding process on page 5 sort of implies that forging and folding were exclusively associated with the forge welding process of different steel grades toward the end of the forging process.  But of course there was a whole bunch of forging and folding of the individual steel grades prior to the process of forge welding and folding the different steel grades together.

 

Thanks for putting this together!  I'm getting awfully tempted to try to put together a miniature bloomery furnace and see what I can get with local iron sands :-)

[ROKUJURO]

Robert,

many different sizes of bloomery kilns have been tried. It seems there is an efficiency threshold which makes too small kilns useless. The insulation of the kiln wall seems to play a role, and an inner height of about 120 cm, a top inner diameter of 25 cm, and a botton diameter of the reduction zone of 30 cm are seen as minimum. The efficiency factor can reach up to 30% or even a bit more if the dimensions and construction design are well balanced.

In the end, operating a bloomery kiln successfully requires a lot of experiences. Different kilns, materials or conditions will yield different results; many details will have an impact on the function.

[Emil]

Adding another article on the topic, use ChatGTP for translation.

 

In short, Japanese swords were made of tamahagane mostly in the later years and during Edo. Earlier, imported steel from China and Korea was used to make katanas in Japan.

 

This is part of a wider topic that Noriyuki Omura is advocating on his website. Who I had the pleasure of exchanging emails with. See: http://ohmura-study.net/003.html

 

His main thesis is that the Japanese sword community is snobbishly discriminsting Showa-to from being considered Nihonto, on false grounds. Japanese swords were never exclusively made of tamahagane and the process of making a sword was never unified anyway. All sorts of methods and steels have been used throughout history and by different smiths at the same time. Special ordered or Koto massproductions, they are all considered Nihonto.

 

The evolution of the forging process took place in parallelly inside the country for centuries, but the last evolution which led to the revival of Japanese swords and possibly saved the art form is neglected.

 

Mainly due to politics and lack of historical knowledge, the current legal definition of Nihonto is restricted to the very narrow Edo style production methods of both the steel and forging, which unfortunately excludes Showa-to.

 

His father laid down his life in the last stand, defending Japan from the American occupation. He asks why the same respect is not granted that sword, which was carried by the modern Japanese warrior.

 

He claims despite the Japanese sword community's attempt to discredit the Showa-to, many of them were better suited as weapons than Edo period swords which could be too brittle due to overemphasized focus on esthetics, which led to too wide hamon, etc. Especially Manchurian and Spring steel forged swords had an excellent reputation for their quality and durability in combat.

 

I like this quote by Mr Omura:

 

"as a result of the occupation policy, the art sword world has made the purpose of Japanese swords the beauty of the blade.
This has led to a misunderstanding of traditional Japanese swords.
As a result, the irrational notion that military swords made of superior steel are not Japanese swords has spread throughout the world. This is a big mistake."

 

 

再校刷-2.pdf

[Bazza]

Thank you Emilll, it looks a very interesting article.  I have saved it in case I ever get to the stage of reading Japanese fluently, but even so, it would indeed be nice to see an English translation...

 

BaZZa.

[OceanoNox]

14 hours ago, Robert S said:

Nice work!

Thank you Robert!

 

Thank you for the comments too. I'll answer below, but I'll use your comments to upgrade the file. If I can get close to a "definite" version (HA!), I might upload it.

I realized later I had forgotten to erase the duplicated Vickers hardness data. About the concentration of P and S, I have no idea where the sands/ore came from for the old swords (or rather if it was in Prof. Kitada's book, I skimmed over it...). About the folding part, it's the section I am not happy with; I wanted to get over it in one slide, and the nuance is difficult to get. I did want to explain that each type of steel was separately worked before being joined into the final block for the sword. And for the carbon content, I was amazed too. From what I understand, they usually look at the fractured areas when they break the initial metal before sorting into high or low carbon steel. It seems the folding tends to give a pretty consistent 0.5~0.6 mass% carbon, so maybe they realized that a certain number of folding cycles worked. I know that the people from older times were likely just as smart as us, but knowing how little technology they had, I am still amazed at what they did.

[OceanoNox]

4 hours ago, Emilll said:

In short, Japanese swords were made of tamahagane mostly in the later years and during Edo. Earlier, imported steel from China and Korea was used to make katanas in Japan.

To add, also steel from the Netherlands (anywhere, really). I did not know that, but it seems the "nihonto = tamahagane" narrative started, according to Prof. Omura, after the sword ban of Meiji. He adds in the conclusion that nihonto were not made with tamahagane during the bakumatsu because it was good steel, but because it's what the smiths could get their hands on. So tamahagane was mainly used for 70 years out of 1000 years of history for nihonto.

This is a tough paper to digest (a lot of content, and it challenges my own ideas on the topic), but very interesting.

[OceanoNox]

To add about Prof. Omura's research paper, it is part of a larger research project, funded by JSPS (Japanese Society for the Promotion of Science, the main government agency for basic research funding, relatively competitive and attribution depends on both the application and the applicant's previous research results and achievements). The topic was "The iron that changed Asia " (アジアを変えた鉄). There was a symposium at Kyushu National Museum (https://www.kyuhaku..../event-220128-2.html), and Prof. Omura's own website gives more details (http://ohmura-study.net/410.html).

[Robert S]

8 hours ago, OceanoNox said:

To add, also steel from the Netherlands (anywhere, really). I did not know that, but it seems the "nihonto = tamahagane" narrative started, according to Prof. Omura, after the sword ban of Meiji.

 

I have a bit of a question about this theory, since it seems that for most of the time that Japanese swords were produced, production was clustered around streams bearing quality iron sands.  I don't know why this would be the case unless they were utilizing the iron sands to make tamahagane, and using that to make the swords.  I'm sure that they did integrate steel from other sources when available, but the idea that tamahagane was only used as a primary material for 70 years seems pretty unlikely.

[OceanoNox]

3 hours ago, Robert S said:

I don't know why this would be the case unless they were utilizing the iron sands to make tamahagane, and using that to make the swords.

Yes. I know an experimental archaeologist in Japan working on the evolution of tatara, so I might ask him about that. He told me that Japan exported a lot of swords to China, and I think I read that China decreased the price per sword they were willing to pay, so the Japanese side increased the number of swords sent for the next shipment. It does seem weird that Japan would get Chinese material and send it back, since a lot of the iron technology came from China and Korea. I will reread Prof. Omura's paper with more focus.

[Emil]

On 9/11/2024 at 12:21 AM, Robert S said:

I'm sure that they did integrate steel from other sources when available, but the idea that tamahagane was only used as a primary material for 70 years seems pretty unlikely.

 

Correct, in the Japanese research paper I shared "Is the material of Japanese Swords really tamahagane" the conclusion is that foreign steel has always been used to make Japanese swords due to limited domestic supply of tatara. It doesn't mean tamahagane wasn't used earlier, but it was not the primary source of steel for most of the history.

 

Here is the conclusion part of the reasarch paper translated:

 

"Tatara steel production was not solely for making Japanese swords. It also met the demands for agricultural tools, construction materials, household goods, and weapons. However, the production of domestically made steel was very limited, and the majority of the demand was met by imported iron. Domestically produced iron (pig iron and tamahagane) only dominated the domestic iron market for a short period at the end of the Edo period. The history of Japanese swords came to an end with the 1876 edict banning the wearing of swords (Haitōrei). After that, the appreciation of Japanese swords became a topic discussed mainly by enthusiasts who regretted the decline of swords. Additionally, in the post-Meiji era, with the rise of nationalist sentiments, the narrative that Japanese swords were made from tamahagane emerged. This trend has persisted to the present day. The truth is that during the late Edo period, it wasn't because tamahagane was superior that it was used for making Japanese swords, but rather because most of the steel available at the time was inevitably domestic Japanese steel. Out of the 1,000-year history of Japanese swords, tamahagane was primarily used for about 70 years during the late Edo period."

 

 

[Emil]

Prof Ohmura just replied to my email about his view on Japanese steel used in katanas. Read the translation below:

 

"Mr. Emil,

 

That is a very difficult question.

 

The methods of making swords and the materials used vary greatly from era to era, and there are also differences in the skill levels of swordsmiths. Therefore, I think it's dangerous to strictly compare the differences between periods.

 

However, if we roughly look at the trends, there are a few perspectives:

 

1. One view is that old swords (kotō) (best) > new swords (shintō) > modern swords (art swords).

 

 

2. Blade structure (construction):

 

The mixing of hard and soft steel (steel with uneven carbon content, common in the old sword period) -> The sanmai structure where hard steel is enveloped by soft steel, seen in new swords, newer swords (shinshintō), and modern swords.

 

Even among newer swords (shinshintō) and modern swords, there are strong swords made from mixed hard and soft steel materials.

 

Examples include newer swords (Mito Clan, Katsumura Tokukatsu) and modern swords (Kobayashi Yasuhiro), among others.

 

3. Materials used:

 

Spring steel (best) > alloy steel > army sword steel > hydrogen-reduced steel > Japanese steel (wakō)

 

In addition, there are also strong swords made by the American company Cold Steel using regular carbon steel. It seems that the secret lies in the sword-making method, including the quenching process.

 

Even swords with the sanmai structure seen after the new sword period are stronger than new swords when it comes to Mantetsu swords. This is likely due to the iron from Manchuria and the sword-making methods used.

 

Given the complexity of these various factors, it is difficult to simply rank the quality based on the period or the materials used.

 

However, one thing that can be said is that the swords from the Edo period, which became decorative swords for samurai, and the modern swords that followed them are the worst as weapons. At the very least, the serious military swords made before the war were better than those."

[OceanoNox]

1 hour ago, Emilll said:

Spring steel (best) > alloy steel > army sword steel > hydrogen-reduced steel > Japanese steel (wakō)

I an not sure what is meant precisely with "alloy steel" or "army sword steel", but the hagane and kawagane with high carbon in older nihonto are spring steels, essentially. The main reasons why Japanese swords bend less and when they do, they remain bent are:

(1) Blade cross section: the blades are thick enough that they are resistant to bending. This resistance depends mostly on the Young modulus and the size of the cross section (i.e. it doesn't depend on the type of steel);

(2) The use of low carbon steel (sometimes close to iron) for the core, that has such low yield stress that it will deform permanently when bent. The hagane and kawagane may not be permanently deformed but the core is, and the sword stays bent.

 

I have had discussions with friends, and we can continue here, but in my view (and from what I have seen so far), there is little difference between the steels used for swords in Japan and Europe. The main differences seem to be the type of non-metallic inclusions, which form because of elements contained in the original iron ore and the walls of the bloomery furnace, and the hardness. Maybe the cleanliness too (sulphur and phosphorus contents), but I need to gather more data and reports.

So far in my review of published data, the Japanese swords' edges appear quite harder than European swords in general (regardless of composition, even some crucible steel swords). I see it as a choice to aim for strong cutting edges, at the expense of impact resistance. I am not sure why this choice was made. Many will say that Japanese martial arts with swords teach to deflect and evade rather than block the enemy's blade, but my own school at least does very hard edge to edge blocks (and I was told to consider the sword a consumable, in a fight).

 

 

[Rivkin]

I am going to be critical. The issue with all publications on the subject is that each deals with a particular selection, makes the most general conclusions, and is conducted by a person who either understands the modern steel, or the old steel or the sword history, but never all three combined. As a result you can find opposite statements on the same subject, in each case supported by graphs and observations.

 

First, in regards to Ohmura, I feel its a case of liberal contrarian arguing against mainstream patriotic. It has been long postulated that the transition from Chinese to Japanese steel is observable by changes in titanium and to lesser extent tungsten concentration, which distinguish continental and Japanese sources. Accordingly Kamakura period blades are seen as at least 90% Japanese sources, Kofun blades are at most 10% Japanese sourced. If the conclusions being contested, this observation (different characteristic inclusions in Japanese versus Chinese sources) needs to be addressed first. In many ways the conclusion also mirrors similar one for bronze, though in this case the argument is much stronger since its based on the ratio of Pb isotopes which are very characteristic of specific locations.

Second, Kitada makes an important point that what distinguishes Japanese sword steel by period is mostly the grain size, which is related to forging practice. Accordingly in his analysis shinto swords come out on top because their grain structure is exceptionally well controlled. Accordingly his position that changes in swords have much to do with forging, like more foldings for shinto, and less with sources. Also a good example of contradictory data - in his publications P and S in Japanese swords are generally low. He makes a good point however that in European steel native presence of Mn can negate even high concentrations of P and S, therefore for Japanese steel presence of such inclusions would have been more problematic. But again his statement is that its very low, which is in contradiction with some other measurements.

Third, still no definitive analysis of composition versus period. Shinto steel does look different in many aspects, but are we certain of this, and why it is different remains unanswered. Is it better selection (composition varies a lot within tatara), is it better sourcing (specific region), is it better forging, is it imported etc. etc.. 

 

In regards to hardness - european technology includes a final operation of taking a blade and heating it a bit for day/days to reduce its hardness but greatly reduce stress and dissolve large martensite. European blades do not hagire, and frankly non-Japanese blades do not hagire, though its achieved by different methods.

[Lewis B]

Quite a nice, simple to digest, overview of the Japanese Tatara process and an account of the history. Seems the process of smelting various iron containing raw materials to create high quality steel began in the Middle East and the technology moved East and West, through China and Korea in the case of the former. Japan started to produce its own steel for tools and swords in the 6-7th centuries. 

 

https://www.Japan-in...JapaneseBlade_01.pdf

[OceanoNox]

17 hours ago, Lewis B said:

uite a nice, simple to digest, overview of the Japanese Tatara

Nice one! Funny to see I used the same source for images.

 

@Rivkin, I agree with you, and even would add that we are limited by survivor bias. There is little to say that the blades we have are even representative examples for each era. This is even further limited by the impossibility of conducting destructive analysis in many cases. From some recent articles, the use of X-ray diffraction or neutron diffraction gives good information but is clearly limited. A nice section cut and polished to see the microstructure, get the hardness, chemical composition (melting etc.).

To me, Prof. Kitada does manage to have al: knowledge of old steel, modern steel, and history (although maybe more as an amateur than dedicated scholar).

Control of the grain size is indeed important, first uniformity, and then small grains, those give the best mechanical properties. For Mn, P, and S, it's also why we add Mn to steel even if we don't need it for the final properties of the steel, it's essentially there to capture the sulphur into MnS inclusions. These are actually not detrimental per se, unless they are exceptionally large, at which point they may become the initiation points of cracks.

I want to ask you about the long heating of European swords. This sounds like aging, but in carbon steel, that would spheroidize the steel (form large spherical carbides). Unless it's simply some low-temperature tempering, that might indeed recover some martensite, and at least reduce the internal stresses (to be clear, it was done in Japan too, but for some reason, this is almost never described in documents for the layman). Do you have more info on the topic?