Iron and Steel Technology in Japanese Arms & Armors - Part 2: Smelting Process

Iron and Steel Technology in Japanese Arms & Armors - Part 2:  Smelting Process


Men using foot bellows to operate a Tatara furnace in the Japanese mountains, from 
日本山海名物図会 5巻. Note the box like shape and the big dimensions of the furnace in the background


Last time
I've covered in my "introduction" to Japanese Iron and steel technology the iron sources used at the beginning of this long journey to make arms and armors, with a perspective on quality and quantity, two features that are often misunderstood when people talks about Japanese traditional iron technology.
If you haven't read it, I suggest you to do it before reading this article which is directly linked to the previous one:

Iron and Steel Technology in Japanese Arms & Armors - Part 1: Iron Sources

Before I start, here is a nice disclaimer I want to write again;

I might have missed something here and there, and I might be wrong since I don't have neither the experience nor the academic background to be precisely accurate. Please take these information with a "pinch of salt".
Despite this limitation, I'm using academic references that you can find throughout the article to back up my thesis.


If you remember from last time, we talked about the raw iron sources, which were chemically bounded to other elements; in this article I will try to explain how these other elements were separated from Iron ( and sometimes how other elements were added to make steel). This process is called smelting.


Tatara (
) - The Japanese furnace

To efficiently smelt the iron ores into iron and steel a huge amount of heat is required to allow important reducing reactions to take place, and to have that energy a furnace  is required.
For the sake of this series we will cover the  Late Muromachi and Azuchi-Momoyama period, which means from the end of the 15th century to the beginning of the 17th.

The furnace used by the Japanese is called Tatara; there are a lot of theories about its various names, its origin and its development over time which I will left uncovered, but what is important to know is that the Tatara is not a single type of furnace; we will get to this later.

Iron and steel technology arrived in Japan probably from China through Korea although this topic is still debated.
What we know is that by the 16th century, the Tatara furnace was made of clay and it had the shape of a big box.
Its dimensions varied a lot especially from regions and through out the periods, but generally speaking they were around 2-5 meters in length, and 1-2 meters in width. The height could be around 1.1-2 meters as well as we can see from artistic depictions.
These furnaces were much bigger than their predecessors, and so could reach an higher temperature compared to the past.

To increase the temperature inside the furnace, and thus allowing better reducing conditions, bellows (fuigo - 
 ) used to feed air to the furnace were required.
By the period discussed in this article, there were mainly two types of bellows used; foot bellows fumifuigo - 踏み鞴 where the air was pushed by the action of several men, and the more sophisticated fukisashi bellow (吹差鞴), which was a wooden structure that worked as a double - acting piston bellows. This device was able to generate a constant and thus stronger air blast which highly increased the temperature and required less men to be operated. Several fukisashi bellows were used for one Tatara.



A fukisashi fuigo on the far left used to blow air to decarbonize cast iron, a process that will be explained below, from 日本山海名物図会 5巻. By the time this book was published in the 18th century, more advanced tembin fuigo were in use for smelting.


The Tatara of this period had also an underground structure, which was used to remove the moisture and stabilize the temperature, although the real effects of this aspect is still debated.
The Tatara were operated into a factory form called Takadono-tatara (
高殿たたら
) and thanks to this organization the productivity of iron and steel in this period rose; this is confirmed by the fact that the prices of iron went down by the end of the 16th century, despite the huge demand.

One of the critical aspect of this furnace is that after every round of smelting it was destroyed to extract the final product; this was necessary mainly because the walls of the furnace were burned by the chemical reactions since they acted as a flux to form slag.



The structure of a late Tatara; in the 16th century, those wooden tembin bellows on the sides didn't exist and the underground structure was less developed, but this scheme offers a good overview of the furnace.
(References:
"Origin and development of iron and steel technology in Japan" by "Kenichi Iida"

"鉄の道" - Tetsu no Michi

"Iron Prices in Ancient Japan and the International Comparison" by "Hiroshi Arai"


"Hitachi Metals Co."

"Wakou Museum" )


Kera-oshi Tatara (鉧押し) : Direct steel making process

As I've written above, the Tatara is not a single type of furnace; according on how it was operated, the final output and the "type" of furnace is established.
The majority of people are familiar with this type of process known as Kera oshi method, which produced a Kera (
), a massive conglomerate of iron, steel and slag
This is a direct steel making process, and in this case the Tatara is operated as a bloomery.

Bloomery furnaces have different advantages and disadvantages; the Kera Oshi method is also known as low temperature furnace: the temperature in this case don't reach the melting point of iron and so the material is not completely melted.
However in this case the temperature could reach 1500° C which means that the majority of the inclusions would melt and be separated by the final product, but not all of them.

The iron produced by this method was still full of slag (non metallic impurities harmful to iron) and need a process of consolidation to be used, hence the name of wrought iron.
However, the temperature in the Tatara was not uniform; towards the bottom, where it was higher (and thus the reducing conditions were increased ), the carbon atoms in the charcoal react whit the iron atoms to form steel, with various amount of carbon content. If steel is produced in the bloomery, the volume of slag will become less and the product will be a bloom of steel containing less impurities.


The risk of directly producing steel in a bloomery, given the medieval technology is that is impossible to control the carbon content with precision, so you might end with undesired "cast iron" or "pig iron" (steel with more than 2% of carbon content) which is melted steel: this condition is possible due to the fact that an higher amount of carbon in the steel lower the melting point.
Cast iron has less impurities because its liquid state allow the slag to be completely separated, but its carbon content makes it too brittle to be used for edged weapons or armor and need a further process of refining. This is why in this direct method was undesired.

The Kera oshi required 3 days to be completed and after this time the furnace is destroyed and the Kera is extracted.





A piece of bloom being consolidated by hammering, known as Kera in Japanese; the picture is taken from a small furnace compared to the Japanese ones. However the look would be the same, since the Tatara operated in Kera-oshi is just a large bloomery.


Process description:



The furnace is charge with iron sand or iron ores and then charcoal is added; in this phase the first reduction reactions take place and iron slag start to form. The temperature slowly rise, and the slag which is in its liquid state is "tapped off" from the furnace through an hole at the bottom.
At this point, iron ores/sand are added again together with charcoal, and while the Kera grows in size, the wall of the furnaces start to thin out.
After three days, the clay walls couldn't bear any more reduction and the furnace is broken to extract the massive "bloom".
This process is called hitoyo (
一代) which means one generation.

To make 2.8 tons of Kera, and 0.8 tons of cast iron known as Zuku in Japanese (
), 8 tons of iron sand and 13 tons of charcoal are used in 3 days.
It's quite easy to see that is clearly not an efficient way of smelting iron.

As I said before, the Kera is a mixture of iron, steel and impurities; only 1 ton circa of these 2.8 tons is high carbon steel, known as Tamahagane (
玉鋼), with a carbon content in between 0.5-1.5%.

And a little reminder, Tamahagane IS NOT cast iron (or the more commonly known, "pig iron").
This idea is been around for so long, and is even erroneously written in the English Wikipedia page!
First of all, the Japanese knew what was pig iron, they even have a name for it (Zuku)
, second when pig iron is formed inside the Tatara, being in its liquid state, is easily separated from the Kera which is solid, third the Japanese knew how to deal with undesired pig/cast iron (this process would be explained later) and finally Tamahagane means "jewel steel"; not because was some sort of precious steel (although being produced in small number, it was valued as precious) but because being  high carbon "sponge iron" it looks like conglomerate of "shiny things" like jewels. The same is not true for pig iron ingots.



This is Tamahagane; it's quite silvery, isn't it? Hence the name, "Jewel steel".



The rest is mainly low carbon steel and iron which is generally referred as Bugera () with a carbon content in between 0.08% to 0.5 % more or less.
To sort out which piece of Kera was Tamahagane and which not, they broken the block into several pieces and performed several test to classify them.
If the small block analyzed was harder, breaks instead of bending and was generally located in the outer zone of the Kera (hence why it was limited compared to the amount of Bugera) it was Tamahagane.

The other grades of steel ( low carbon  and pure iron) react in different ways as well, so it was easier with the right amount of experience figuring out how much carbon was inside.

Before sending the various pieces of steel and iron to blacksmiths, they were usually consolidated by hammering; the pieces were heated until they were red and then hammered. This process allow to further improve the purity of the iron (or the steel) since the slag trapped inside is "squeezed out".

You can see the full process here in this video.



(References:

"Hitachi metals & co."


"Origin and development of iron and steel technology in Japan" by "Kenichi Iida"

"鉄の道" - Tetsu no Michi

"The knights and the Blast furnace" by "Alan Williams"

"Ancient and historic steel in Japan, India and Europe, a non-invasive comparative study using thermal neutron diffraction" by "F. Grazzi & F. Civita & A. Williams & A. Scherillo & E. Barzagli & L. Bartoli & D. Edge & M. Zoppi"

"Technology and the Culture of Progress in Meiji Japan" by "David G. Wittner")




Zuku-oshi (
押し) Tatara: Indirect steel making process

It took me some time to realize that this process existed as well. Despite being extremely well documented, the major english sources dealing with traditional Japanese iron & steel smelting seems to avoid mentioning it in their works, which translate into a distorted and misinformed picture of traditional Japanese technology.
I will leave some thoughts and conclusion in this topic, but first I'll try to explain how steel and iron were obtained with an indirect process.

As I've written above, one of the drawbacks of having a bloomery is that you end up with an heterogeneous mix of iron, slag and steel mainly because the iron and the steel formed inside the Tatara is in its solid state, due to the whole process being a low temperature ones.

On the other hand, the Zuku-oshi Tatara is an high temperature furnace which main output is indeed melted steel known as cast iron or pig iron.
The Zuku Oshi Tatara slightly differs from its counterpart, the Kera oshi; usually the starting material is either iron ore or a particular iron sand type called akome which is easily reduced in the furnace, the air is supplied throughout the whole of the lower part of the furnace, the angle of the lower part of the furnace is larger and iron sand is added after charcoal.
Although the Tatara could reach about 1500°C which is still below the melting point of iron, in this process the iron ores are left 4 days inside the furnace, so the temperature is even in the whole furnaces, unlike in the Kera version.
This allow the carbon atoms to spread evenly and react with the iron ones, forming
cast iron which melting point is around 1150°C. When the steel is in its liquid state, the slag is fully separated and the final product would be much purer compared to the ones of a bloomery.
However, cast iron is produced and not steel; this means that a further process called "fining" is needed to lower down the carbon content of the material and make steel and wrought iron.
In Japanese the whole process is called "Ookaji" (
大鍛冶) and involves Sage (下げ) which turn cast iron into medium/high carbon steel called Sagegane (下げ) and Honba (本場) which lower the carbon content even further and produced Oroshigane (下ろし金) with the same properties of wrought iron.




Molten steel being cast in a mould from a modern process.


Process description:

Unlike the Kera oshi, to produce steel with the Zuku method two phases are needed (hence the name "Indirect").
The first one is the smelting of liquid steel, pig iron or cast iron.
The furnace is heated, charcoal is added and later on iron sand/ores; the process last 4 days, in which the slag is fully separated from the steel.
The final product would be cast into moulds and then is brought into the Japanese "finery" which were closer to the furnace and decarburized into the required material; this is where the fining phase start.

In the firs passage called Sage, the pig iron blocks were set in a tunnel in the front of tuyere, covered with charcoal and then fired with blowing air, and after 1 hour the blocks were melted down and their carbon content was lowered in between 0.7 - 1%. Humidity was carefully checked and eventually water was added since high temperatures caused to dry the hearth and make the steel too brittle.
After 2 hours, Sagegane was obtained and there was no material loss.

The second passage called Honba, Sagegane blocks were mixed with Bugera blocks in 2:1 proportions; the process is similar, charcoal was added and with air blast the first sagegane blocks started to melt down. Then they were agglomerated and decarburized further.
After this process was ended, the output was forged by hammering and heating, and transformed into plates; in this case there was a loss of material of 30-40% and the final carbon content was around 0.1%

However different proportions of carbon were obtained based on the time spent in the Sage and Honba process, so those values shouldn't be used as specific and unique.


You can see the smelting process here in this video.





A cast iron ingot obtained through Zuku Oshi Tatara; source here.
Is quite different from Tamahagane!



This whole indirect process is similar to the one used in Europe and China, and since the main output of the Zuku Oshi tatara is pig iron we could call this type of furnace a blast furnace, opposed to the Kera ones which is a bloomery. This is why I made my previous statement about the Tatara being multiple variations of furnaces.

Last but not least I would like to pointed out that despite both methods did produce pig iron, it was NEVER directly used to make edged weapons or armor. Its carbon content is too high to be effective and since the Japanese knew how to lower it, there isn't much sense in claiming that pig iron was ever used.

(References:
"Hitachi metals & co."

"Home-made steel: A week at Manabe Sumihira’s zuku-oshi tatara" by "Pierre Nadeau"

"Manabe Shumihira official website: http://www.eonet.ne.jp/~sumihira/19sageba/00sage/00sage%20.html "

"The Knight and the Blast furnace" by "Alan Williams"

"Pre-modern Refining Process of "Okaji" without Deoxidation" by "Kazuhiro Nagata and Takashi Watanabe"

"Wakou Museum official website")


Final thoughts: Tamahagane or Sagegane?


After this reading I bet that nobody has ever heard of Sagegane but a fair amount of people would be familiar with Tamahagane.  It is quite a mystery explain why one is more popular than the other; although is written practically everywhere, Tamahagane was not the only high carbon steel used for weapons and armor.In fact, according to some researches done in the past, the Zuku Oshi method which was established since the Kamakura period (1192-1333), not only was older than the Kera Oshi method, which dated back to the Tenbun Era (1532-155) but it was even more used throughout Japan.And is not an impressive feat; although blast furnaces are more sophisticated and advanced than bloomery ones, the Japanese probably obtained their technology from China which was already using blast furnaces, and until recently it was assumed that in China bloomery were not used at all (this thesis has been debunked by recent evidences, but blast furnaces remains the main source for steel and iron in ancient China).

Although it is assumed that both methods started to be practiced since the 6th century in Japan, the Zuku Oshi was the one used the most especially in the Edo period.
So the majority of iron and steel objects would have been made from Zuku Tatara.

So why is the Tamahagane so important nowdays? There are few considerations to be made, and is better to underline that we are under the realm of speculation from now on.
Back in the early 20th century, the Tatara was obsoleted; modern western style blast furnaces were incredibly better than these old and traditional clay smelters.
After WW2 a lot of history and knowledge was lost in Japan, and few gentlemen started to study sword's history in depth with modern tools.
There was actually a debate if the swordsmiths of the old days used tamahagane or sagegane, and in the '70s when the Nittoho Tatara started to be operated they choose a Kera-oshi Tatara; the emphasis on this method is likely to be expressed by the fact that the bloomery might have been a Japanese native development and is quite unique in the sense that uses iron sand instead of iron ores (although not always!). This is where the inflation of Tamahagane came from.
Apart from that, the Kera oshi tatara is just a bloomery, nothing special about it, and Tamahagane is just high carbon "sponge steel".
However there are still modern historian and swordsmith like Manabe Sumihara that stress on the importance of the Zuku oshi tatara and still use decarburized cast iron to make sword blades.

But what's the difference?



Sagegane could be made with the same amount of carbon of Tamahagane, which in both cases is not suitable for weapons and armor; in fact is lowered down during the forging process  (which would be explained in the next article) and modern analysis have shown that the carbon in sword's edge was never higher than 0.8%.Both would still be rather heterogeneous due to the technology used.

One of the key differences is in the amount of non metallic inclusions or slag; artifacts made in the bloomery always contain trapped slags from the extraction process since the iron doesn't liquefied and also smithing slags, inclusions introduced during the forging process.
On the other hands, artifacts made through  refined cast iron only contain smithing slags, which means that Sagegane is purer and better than Tamahagane, even if by a small amount.
This also true for the whole output of the two Tatara compared, the Zuku Oshi method produced and higher quality final product.
However it is fair to highlight and underline that in a pre-industrial world, slag would have always been present inside metal objects, be it in Europe or Japan.

Despite its flaws, the Kera oshi method offers other advantages; it was faster and could directly produce high carbon and low carbon steel, without going into the lengthy process of decarburization, an asset quite important during the Sengoku period where the demand of iron and steel artifacts was incredibly high.

(References:
"The Knight and the Blast furnaces" by "Alan Williams"
"Analysis of the Products of Ancient and Medieval Low Shaft Furnaces in Japan" by "Minoru Sasaki"

"The Investigation of Establishing Time of Zuku-Oshi and Kera-Oshi with Data of Iron Image of Buddha Making Age and Old Document "Kokon-Kajibiko" " by Takuo Suzuki)

Conclusions

The first time I started to make researches in this topic, the final outcome was that the Japanese didn't have the technology required to produce high quality steel and iron through an indirect process, but after digging enough into scientific articles the reality is far from that, since the Japanese had their own indirect steel making process which allow them to create artifacts on par with European and Chinese ones in terms of the quality of the steel ( incredibly low amount of impurities ).
Even across English scientific papers, there was barely any mention of the Zuku-oshi tatara, which would have explained why some artifacts didn't had smelting inclusions in them and the overall slag content of some of these items was closer to zero!

So to sum it up, not only the Japanese didn't had bad starting material, but they also had the technology required to efficiently smelting iron ores into iron and steel of extremely high quality for their time periods.

The third part of this series that covers bladesmithing is available here.

Until then, I hope that this super-technical article wasn't too boring! If you have experience in the field, suggestions are welcomed, and if you liked this article please feel free to share it or ask any questions in the comment.

Gunbai



Further References:

I decided to add scientific papers that are available on google scholar which could be useful for the most experienced and curious ones! Unfortunately some of them are in Japanese...

"Metallurgical Research on Japanese Swords- Focusing on Swords for Practical Use- " by "Hideo Hoshi and Minoru Sasaki"

"Analysis of crystallographic structure of a Japanese sword by the pulsed neutron transmission method"  by "K. Kino, N. Ayukawa, Y. Kiyanagi, T. Uchida, S. Uno, F. Grazzi, A. Scherillo"

"Neutron diffraction characterization of Japanese armour components" by "A. Fedrigo, F. Grazzi, A. Williams, A. Scherillo, F. Civita and M. Zoppi"

"From Koto age to modern times: Quantitative characterization of Japanese swords with Time of Flight Neutron Diffraction" by  "F. Grazzi, L. Bartoli, F. Civita, R. Franci, A. Paradowska, A. Scherilloa and M. Zoppi"

"Phase composition mapping of a 17th century Japanese helmet  by "A. Fedrigo, F. Grazzi, A. Williams, S. Kabra and M. Zoppi"

" Neutron diffraction characterization of Japanese artworks of Tokugawa age" by " F. Grazzi & L. Bartoli & F. Civita & M. Zoppi"

Comments

  1. thanks! that's very informative
    cant wait to read about smithing...

    ReplyDelete
  2. i think foot bellows were much more powerful than box bellows because much more force could be applied to them, with all those workers stepping with their full weight. box bellows were portable though and more compact.
    had water wheels been used to power bellows in Japan?

    ReplyDelete
    Replies
    1. I'm glad you liked it!! Unfortunately I have been very busy theese months so I still have to collect all the right sources to write about bladesmithing, but I'll do my best to post the article asap!

      Those box bellows however were able to create a costant and continuos blast of air, which is the main reasons why they were used. They could also be "stacked" rather easily to obtain a stronger blast, whereas foot bellows are too big to be efficiently stacked!

      As far as I am aware, although water wheels were widely used in Japan, they were never used to power bellows until the end of the Edo period.

      Delete
  3. What method did the Japanese use in the Kofun Period? Did they use Chinese and Korean method of steel working?

    ReplyDelete
    Replies
    1. I haven't look into that specific period but I'm quite confident that they used primarily Korean method of steel working, which is reflected in the Japanese arms & armors of that period; they were extremely similar to the Korean ones.

      Delete
  4. This article was an absolute paradigm shift for what I thought I knew about Japanese metallurgical technology. I had ZERO idea that Japan had blast furnaces of any type. I knew the Japanese tatara bloomeries were plenty capable of producing high-quality steel, but knowing that they also had this level of technology...totally blows my mind. In a good way!

    ReplyDelete
    Replies
    1. Thank you!

      I know, it's amazing how few people are aware of the fact that an indirect method of steel making was known and used in Japan - despite the fact that even the Japanese wikipedia acknowledge that!!
      And while the Nittoho tatara is the main and most famous one ( which is a kera oshi tatara), bladesmiths like Manabe Sumihira still use the zuku oshi tatara and decarburized cast iron to make high quality steel for his blades.

      Delete

Post a Comment

Popular posts from this blog

Masakari (鉞) - The Samurai's War Axe

Tate & Tedate (盾 & 手盾) - Japanese Shields

Yumi (弓) - The Japanese Bow

Sengoku Period Warfare: Part 1 - Army and Battle Formations

Tosei Gusoku (当世具足) - Body Coverage Explained

Cagayan Battles of 1582: Debunking the Hoax

Wantō (湾刀): Early Curved Japanese Swords