This article is a revised version of a talk I gave at the one-day Symposium on Cast Iron in Ancient China, in Beijing, 20 July 2009. Many thanks to the organizers and to several participants who made useful comments.
Click on any illustration to see it enlarged.
Participants in the Symposium were all familiar with the Iron–Carbon Equilibrium Diagram, which gives the atomic state of iron–carbon alloys in relation to temperature. But it seemed worthwhile to remind them of the implications of the Liquidus Line, emphasized here. This can also be called, somewhat imprecisely, the “melting point” of an alloy. Practical casting temperatures lie 50–100°C higher than this.
We can see that casting steel, with up to about 1.5% carbon, requires temperatures well over 1500°C. Such temperatures were often reached in early times in various parts of the world, but the refractory materials necessary to manipulate molten steel and cast it into useful artefacts were not developed until the late 19th century, in Britain.
With a higher carbon content, 3–4%, practical casting is much easier, requiring temperatures around 1300°C. Iron with this carbon content is called “cast iron” because it is easy to form by casting. It lends itself well to large-scale production, which has always been important in ancient China. It was undoubtedly the use of cast iron in ancient China that made it possible for every peasant to have iron implements.Some Ancient Cast Iron Artefacts
The use of cast iron makes it possible to mass-produce implements cheaply, but for most implement types it has inferior mechanical properties. All pre-modern Chinese cast iron has a very low silicon content, usually under 0.5% Si, and therefore solidifies as “white cast iron”, in which the carbon is present as cementite, Fe3C, which is extremely hard, harder than quartz, and this makes the iron brittle.
Etched with Nital, scale bar 100 µm.
An example is this mattock-head. The micrograph shows that it is white cast iron. Its hardness may be an advantage with such an implement, for it would be very abrasion-resistant. But the worker would have had to be careful, for if it hit a rock it might shatter.
| 2nd–1st cent. BCE | 1102–1106 CE | 1851–1861 CE |
| 1208–1224 CE, X-ray | AD 1211, Etched with Nital, scale bar 50 µm |
(Mike Wayman & Helen Wang, Historical Metallurgy, 2003, 37.1, pp. 13, 14)
Coins were usually made of copper alloys, but sometimes shortages of copper made it necessary to cast them of iron. Here are three iron coins, from the Western Han, Song, and Qing periods.
Mike Wayman and Helen Wang studied 37 iron coins of the Song period in the British Museum, London. They are all of white cast iron, and their article has a great deal of detail on the types of white cast iron found here.
An interesting aspect is revealed by X-rays of some of the coins: these show that there are sometimes quite large casting bubbles in the coins, making them feel distinctly light in the hand. It seems possible that the coin-founders intended these bubbles, in order to save material.
Diameter 13 cm.
This is a cast-iron mirror – probably Han-dynasty (206 BC – AD 220), probably white cast iron.
Ding 鼎 from a grave excavated at Yutaishan in Jiangling, Hubei, dated to the 4th century BCE.
An interesting characteristic of traditional Chinese metallurgy is that cast iron is often combined with other materials. The legs of this vessel are cast iron, while the body is bronze. I have not seen a metallurgical investigation of a vessel of this type, but the legs are most probably of white cast iron.
Image taken from ebay.com
These crossbow-bolts have bronze tips and iron shafts – presumably because bronze has better casting properties than iron, while iron is cheaper than bronze.
Lian Haiping 廉海萍 in Shanghai has studied some examples of this type of artefact and found that the shafts are of cast iron which has been decarburized in the solid state – this type of iron is called “whiteheart malleable cast iron”. She reported this at the BUMA conference in Beijing in 2006.
Hoes and spades of wood with iron caps were common in ancient times and continued in use well into modern times, as can be seen here:
Mawangdui, 2nd century BCE
|
|
 |
 |
| Ancient well, 3rd century BCE | Han Dynasty, Sichuan |
Drawing by William Alexander, 1793 | |
| From a grave in Changsha, Hunan, 3rd–4th century BCE | Etched, scale bar 250 µm |
This implement-cap is “blackheart malleable cast iron” – it was first cast, then annealed at a high temperature, probably around 950°C, for a period of days. This treatment caused the carbon in the iron to precipitate as graphite, making a material which has much better mechanical properties than ordinary cast iron (white or grey).
| Tonglüshan 铜绿山 copper-mine site, 4th–3rd century BC | Etched, scale bar 50 µm |
(Hua Jueming 华觉明, 自然 科学史研究, 1982, 1.1, pl. 1)
These gads from a copper-mine site show how tough the ancient Chinese cast iron could be. They are obviously subject to very hard punishment, being hammered into cracks in the rock.
But they are of cast iron – again annealed at a high temperature for a period of days, this time decarburizing the iron at the surface and precipitating graphite farther in.
| Scissors, 1st–2nd century CE | Etched, scale bar 40 µm |
(Hua Jueming 华觉明, 自然科学史研究, 1982, 1.1, pl. 1)
These scissors are also of cast iron.
They were cast, then annealed in an oxidizing atmosphere to decarburize to a uniform carbon content around 1%, with a few very small grapite nodules, then bent into shape by a smith.
Cast Iron Sword, 2nd–1st century BCE
Ethanolic picral etch (David A. Scott & Qinglin Ma, Historical Metallurgy, 2006, 40.2, p. 109)
This sword from Gansu appears to have been made by casting it and then decarburizing to a uniform 0.6% carbon.
|
 |
| Cast iron mould for casting an iron mattock-head, 3rd century BC | Etched with Nital, scale bar 200µm. 2.31% C, 0.21% Si (文物, 1976.8, p. 54) |
Here is a cast-iron mould for casting an iron implement.
That in itself is surprising enough, but note that the iron is very low in both carbon and silicon, and nevertheless the structure is grey cast iron.
Normally such an alloy would solidify as white castiron – only very slow cooling would allow it to solidify as grey cast iron.
This must have been cast in a massive heated ceramic mould and allowed to cool very slowly, over a period of days.
The only ancient grey-cast iron artefacts I know of are moulds, and I believe the reason is that the moulds must be very tough to tolerate the thermal shock when molten iron is poured in. They could not be made of malleable cast iron, because the annealing process tends to warp castings slightly, so that it would be difficult to fit the parts of the mould together.
(满城汉墓, 1980, vol. 1, pp. 216, 218, 220; vol. 2, pl. 154, 155.1)
Finally here is another amazing use of cast iron in ancient China. This is the entrance to a tomb cut into rock in Mancheng, Hebei, dated shortly before 104 BCE.
Two brick walls were constructed, after which molten iron was cast in between them to seal up the tomb.
(Torben Witt, Aalborg og Fabrikkerne, 1980, p. 21
The cupola furnace was the usual furnace for casting iron from ancient times to the middle of the 20th century.
Here is a small iron foundry in Denmark in 1890. The cupola is a shaft furnace. Iron and fuel (in this case coke) are charged in the top, and air is blown in near the bottom. There is a steam engine inside the building at the left.
The fuel burns in contact with the iron, the iron melts, and molten iron is tapped out at the bottom.
Notice the workers. The one whose job is to open and close the taphole is protected by a screen from the high radiant heat from the molten iron.
(Gottwald, Technische Mitteilungen Krupp, 1938, 6.4: 110–111)
 |
 |
 |
| Courtesy of Martyn Gregory Gallery, London | Courtesy of British Museum, Oriental and India Office Collections (see here) | Hommel, China at work, 1937, p. 27 |
A 50-second film-clip by Yang Ruidong 杨瑞栋 of the casting of iron in Huize 会泽, Yunnan.
(考古学报, 1978.1: 12)
This is a reconstruction of a large cupola furnace at a Han-dynasty ironworks site at Wafangzhuang in Nanyang, Henan 南陽瓦房莊. Height 3–4 m.
Note that the blast goes through a pipe over the top of the furnace, so that the air is somewhat heated before it enters the furnace.
(Li Jinghua 李京华, 中原古代冶金技 术研究, 1992, p. 147)
This is a reconstruction of a cupola furnace for melting bronze, from the Chunqiu period (777–476 BC).
The furnace is composed of three sections. It is charged copper, tin, and fuel, air is blown in through the top (or sometimes through a hole in the side), and the bronze melts and settles in the bottom. The two upper parts are then removed, and the bronze is poured from the bottom part.
This type of furnace has had a long history . . .
(René Antoine Ferchault de Réaumur, [雷 歐姆], L’Art de Convertir le Fer . . ., 1722)
. . . Here is the same sort of furnace, used by Gypsy ironfounders in Europe. The illustration is from the 18th century.
The Gypsies are a minority people in Europe who traditionally have no fixed domicile, but travel around. They are also called Romanies, Travellers, Zigeuner, Tsiganes, and many other names. They seem to have come to Europe from India in the 13th century, but they certainly had this type of furnace from China. It has sometimes been suggested that they were the first ironfounders in Europe.
Here again we see a cupola furnace in three parts. When the iron has melted, the upper sections are removed and the iron is poured from the bottom section.
A description of a slightly different Gypsy cupola furnace can be seen here (quotation in English in a Danish article).
At Huangkiao we witnessed the operations of a Chinese foundry.
. . .
The blast was stopped, the bellows disconnected, and the upper and middle sections of the furnace taken off and laid aside. The surface of the molten iron being skimmed of its slag, it was well covered with rice husk ashes. This protected the face of the man who next had to handle it from the intense heat that would otherwise have radiated from the molten iron. This man’s duty was to clasp the crucible in his arms, literally hugging it to himself, and to fill the molds arranged around.
. . .
(American Manufacturer, 1899, 64: 125)
I have not found an illustration of this type of furnace in China, but one very interesting description, from 1899. Here on the left is a brief extract; the whole description can be seen here.
It is amazing to imagine the worker clasping the crucible in his arms, literally hugging it to himself.
Alfred Raquez, Laotian pages: A classic account of travel in Upper, Middle and Lower Laos, Copenhagen 2019, p. 140. Translated from Pages laotiennes, Hanoi 1902, p. 133.
Here is a poor photograph of this type of furnace in Laos in 1900, with a description of its operation.
Some Tai Neua people [tribesmen from Northern Thailand] have settled at the entrance to the village to make plowshares.
They place old cooking pots on a small charcoal stove. Bellows similar to those we saw so often in China blow in air. While the iron is being melted, a Tai Neua man prepares the mold of hardened clay and smears it with soot mixed with water. This layer will prevent the content from sticking. In a few minutes, everything is ready. The molten metal is poured into the mold, and the perfectly made plowshare is sold for six salung, or about 60 cents, to the potential client who oversaw the operation.
Diameter 65 cm, thickness 3.5 mm
Here is a wok – that is the usual English word, which comes from the Cantonese pronunciation of 镬, which in Mandarin is pronounced huo.
It is broken, so it was possible to measure the thickness at its thinnest part.
How they were cast in the 19th century can be seen here. They are still being cast today, but probably by different methods.
Here are a few illustrations related to woks.
 |
 |
 |
 |
| Guangdong, ca. 1860? | Shanxi, ca. 1910 (E. T. Nyström) | Kaifeng, 1987 | Denmark, 2009 |
It is interesting to note that cast-iron woks are now being imported to Europe in large numbers and used as portable fireplaces.
Gouache by an anonymous Chinese artist in Guangzhou, mid-19th century. (The Royal Library, Copenhagen, Denmark)
Woks were expensive, and they could easily break, as can be seen in a photograph further above.
This is an itinerant tinker, who mends broken woks using molten cast iron. There is more on how the tinkers worked here.
Photo by Tang Huancheng 唐寰澂
Another major use for cast iron was for statues and other large monuments. These are the famous bridge anchors at Pujin 蒲津, Shanxi, cast in AD 724.
Zhongyue Temple, Dengfeng, Henan (登封中岳庙), 1987
And another quite famous cast iron statue, one of four at the Zhongyue Temple in Dengfeng, cast in AD 1024.
By studying the mould-seams on statues like this one and the next we can learn a lot about how these statues were cast.
Iron Rhinoceros (铁犀), cast AD 1446, village of Tieniu 铁牛, near Kaifeng 开封, Henan
Here on the left we have a very interesting phenomenon. The photograph shows what looks like a repair to the statue, but in fact it represents a repair to the mould.
Something happened to the mould – it broke and had to be repaired. The bumps which can be seen are probably the heads of spikes that were used to hold the repair in place.
Shaolin Monastery, Dengfeng, Henan (登 封少林寺), 1987
I think it is only in China that bells have been made of cast iron.
This broken bell is interesting because we can see in the fracture that something has gone wrong. Those air bubbles at the surface would (I believe) have had a bad effect on the tone of the bell.
Probably the founders added sulphur to the iron to assure that it would solidify as white cast iron, but added too much, so that the bubbles formed.
Nanjing Museum (南京博物院), 1987
Here is again an example of a combination of materials – this anchor is made of several pieces of wrought iron, joined together by casting iron at the join instead of by forge welding, which also was practised in China.
| Thomas T. Read, Mining and Metallurgy, August 1937, p. 383. Photographed ca. 1910. | Inside the statue, photographed from below, 1987. Note modern repairs. |
And here is the famous Iron Lion of Cangzhou, which also uses a combination of materials.
The lotus seat, and a bronze Buddha which originally sat on the seat, were too heavy to be supported by cast iron alone, so wrought-iron supports were incorporated into the casting. Part of the wrought-iron reinforcement can be seen extending from the centre of the rightmost photograph to the lower right corner.
There is more on Chinese monumental iron castings in my article in Journal of East Asian Archaeology, 2000, 2.3/4, pp. 199-224. There is also a Chinese translation.
Cast Iron Cannon, Ming Dynasty (1368–1644).
Military Museum, Beijing (中国人民革命军事博物馆), 2006.
The ancient founders’ experience in casting statues and such came in very handy when they began casting iron cannons.
The first iron cannons in China come at the very beginning of the Ming Dynasty, in the 1370’s.
George Banks, Illustrated London News, 6 April 1861, p. 325
The cannon above is probably a combination of materials – wrought iron and cast iron – as we can see here on the right: two cannons with wrought iron inside and cast iron outside. George Banks sketched these at the Dagu Forts, in modern Tianjin, in 1860. He believed that these were ‘evidently very old’, from the +17th century or before; while this dating is quite plausible it is not clear what evidence he could have had for it. He does not mention any inscriptions.
He described the guns as follows:
No. 1 had a piece broken from the muzzle, which enabled me to see how it was made. The inner part or bore was made of longitudinal bars, one inch wide and half an inch thick [2.5, 1.25 cm], welded together, and forming a lip where they terminated at the muzzle. Round these, and binding them together, were rings, one inch thick and three inches wide [2.5, 4.5 cm], also welded. Outside these, again, is a layer of cast iron, two inches and three-quarters [7 cm] thick at the muzzle, and of course much thicker at the breech, giving shape to the gun. The faint lines on the surface are caused by the crevices between the bricks of which the mould was built in which the casting took place. This piece is 9 feet 6 inches long, 23 inches diameter at the breech, and 15 inches diameter at the mouth [291, 60, 39 cm]. No. 2 is a similar gun, but with only rings welded together and encased in cast iron. It is very singular that both these guns should be broken in the same way. It is 9 feet 7 inches long, 2 feet 1 inch diameter at the breech, and 16 inches at the mouth [respectively 292, 64, and 41 cm]. (Illustrated London News, 6 April 1861, p. 325)
Here on the right is part of a description of iron casting, in a German manuscript dated 1454.
15th-century German is not easy to read, but it says roughly that if you want to cast iron, you should use the same sort of furnace as is used to cast bronze bells, and add certain materials which seem to have the effect of adding phosphorus to the iron, thereby lowering its melting point.
This seems to be the earliest technical description of iron casting in the world.
The original manuscript was in the Zeughausmuseum, Berlin. It was one of a large number carried off as war booty by Soviet troops at the end of World War II, and is now presumably lost.
David E. Potter, The Bells and Bellringers of York Minster, 1987
To show the type of furnace which the German manuscript refers to, here is a picture of bell-founding from around that time, the famous Bellfounders’ Window in York Cathedral, England.
With a little imagination one can see the cupola furnace, the bellows, and molten bronze flowing out.
Cast Iron Boundary Post, Trzemeszno Lubuskie, Poland, said to be dated AD 1345–1364
http://www.brandenburg1260.
de/die_balley_
brandenburg.html
This boundary post in the Märkisches Museum, Berlin, has been considered to be the earliest extant iron casting in Europe, dated by Otto Johannsen to 1345–1364 and by some as early as 1251. It turns out to be somewhat later, probably 15th century. This is still very early, and the artefact remains interesting, but it is no longer the earliest in Europe.
Analysis of the boundary post:
| Carbon | 2.31% |
| Manganese | 0.43% |
| Silicon | 0.23% |
| Phosphorus | 6.20% |
| Sulphur | 0.061% |
| Arsenic | 0.03% |
| Melting point ca. 950°C. | |
| (Otto Niezoldi, Die Giesserei, 1942, 29.8, pp. 136–137) | |
It can be seen that a large amount of phosphorus has brought down the melting point considerably, just as the 15th-century manuscript suggests.
From the museum's inventory journal:
566. 15.8.1876. Grenz-Pfahl von Eisen, 84 cm hoch, wovon die zugespitzte Länge von 54 cm in der Erde gewesen, während die übrige Länge von 30 cm, welche 4eckig und 14 cm im Quadrat stark ist, außerhalb der Erde (frei) stand; die horizontale Fläche zeigt ein vertieftes Ordenskreuz. Geschenk des Gutsbesitzers Landsky in Tempel bei Schermeisel. Solche Pfähle wurden im Jahre 1251 in grösserer Anzahl zur Bezeichnung der Grenze zwischen Polen und der Neumark gesetzt. cfr. Wedekind, Geschichte der Neumark, Anhang fol. IX. X, sowie Acta M II. 1508 C.B. 1876. Diese Pfahl wurde gefunden an der alten Straße von Schwiebus nach Landsberg, auf der Grenze des Märkischen Dorfes Tempel und des Posen'schen Dorfes Neudorf. Er ist wahrscheinlich identisch mit dem von Wedekind, pag IX unten, bezeichneten.
The boundary post is mentioned in a document dated 1364, a renewal of a boundary settlement of 1251. The document has been published in Codex diplomaticus Brandenburgensis, Reihe A, Bd. 24, 1876, pp. 71-76, and also, together with a German translation of 1613, in Johann Carl Conrad Oelrichs Beyträge zur Brandenburgischen Geschichte, 1761, pp. 57-75. The editor of the Codex, A. Fr. Riedel, states that the document is clearly not genuine, but that it is an important source for the boundary controversies of the 15th century. This information was kindly supplied to me by Dr Uwe Winkler of the Stiftung Stadtmuseum Berlin.
Otto Johannsen, Geschichte des Eisens, 1953, p. 203
Cast iron objects were always very simple until Europeans began casting cannons and other artillery.
This German bombard, AD ca. 1400, seems to be the earliest extant cast-iron gun in Europe.
It seems that in the course of the 14th century European iron-founding quickly became quite advanced.
Joseph Needham, Science and civilisation in China, vol. 5, part 7, p. 286, from Tushu jicheng 圖書集成.
In China, on the other hand, cannon-casting had a long background in the casting of other large objects, and I am sure that the European development owed a great deal to China.
Here is a Chinese bombard of much the same type around the same time.
The drawing is from the 18th century, but it is a copy of a copy of a copy, and Joseph Needham believes the original would have been from the 14th or 15th century.
Ebbe Johannsen, Danske Antikviteter af Støbejern, 1982, p. 59
Later, cast iron was used for all manner of useful products, like this stove in Denmark, ca. 1900 . . .
Ebbe Johannsen, Danske Antikviteter af Støbejern, 1982, p. 59
. . . and these kitchen pots, flatirons, etc. (Denmark, ca. 1850).
Cast-iron manhole covers are found in every modern city in the world. These are in Beijng. The one on the left is dated 1953; the one on the right may be considerably older, since the text reads right-to-left rather than left-to-right.
Here is a flatiron (England, 19th century?), which is interesting because it is one of the very few examples I have seen in Europe of the use of a combination of cast iron and wrought iron. The bottom part is cast iron, the handle wrought iron.
Anchor Chain, Wrought Iron with Cast Iron Reinforcement, Bristol, 19th Century?
And here is one more example. These chain links are of wrought iron or steel, but the reinforcements are of cast iron.
Williams’ Perry Foundry Co., Catalogue, ca. 1875, p. 7
Cast iron became very important in the Industrial Revolution of the 19th century. This enormous steam engine (England, 1840) could not have been built without cast iron. In particular, the tilt arm is 11 metres long, and at this time could not possibly have been made by smithy methods.
Prince Rupert (Rupprecht von der Pfalz), 1619–1682
I have talked about malleable cast iron in ancient China – the first malleable cast iron in Europe seems to come more than 2000 years later, in the 17th or 18th century.
Prince Rupert doesn’t look like a scientist or a soldier, but he was in fact both.
He patented a process for softening cast iron, but the patent was kept secret, and it now seems to be lost, so we do not know what the process was. But undoubtedly it was a kind of malleablizing annealing.
René Antoine Ferchault de Réaumur [雷歐姆], 1683–1757
Réaumur was the first person to make a scientific study of cast iron.
His book was published in 1722. It has been translated into English, and it would be a good idea to translate it into Chinese. It is still useful as an introduction to the early European techniques of iron casting.
He describes very clearly the process of making malleable cast iron, and says he learned of it from artisans in Paris.
Réaumur, L’art de convertir . . ., opposite p. 560.
Here are the sorts of things he wanted to make of malleable cast iron – mostly decorative objects which would otherwise have been made by a smith.
Illustrated catalogue of malleable iron castings made by the California Iron and Steel Company, 1884.
Later we find other objects of malleable cast iron which earlier had been made by smiths and now are being made cheaper of malleable cast iron.
Today malleable cast iron is used for a wide variety of fittings of various complex shapes which need to be cast but don’t need the strength of steel.
Shijiazhuang Jizhong Malleable Iron Co., Ltd. (石 家莊冀中瑪鋼有限公司)
These on the left happen to be made in China, but I could have shown similar products from almost anywhere in the world.
There is also a tradition that the Chinese and the Japanese know an art of bringing [cast] iron to a high degree of softness, so that it can receive impressions of figures as easily as lead can; . . .
(translated from the Swedish translation of H. Sjögren, 1923)
Their cast-iron wares appear light and neat, and are annealed in heated ovens, to take off somewhat of their brittleness.
It is very curious that the latest known Chinese artefacts of malleable cast iron seem to be from the 9th century AD, but European travellers as late as the 18th century told of Chinese and Japanese malleable cast iron.
J. J. Rein, Japan nach Reisen und Studien, vol. 2, 1886, pl. VII; translation,The industries of Japan, 1899, p. 433.
And here are some 19th-century Japanese vessels of cast iron which have been surface-decarburized so that they can be engraved.
J. J. Rein in 1886 gave a very interesting description of the Japanese process.
There is more to read in some of my publications:
Toward the reconstruction of ancient Chinese techniques for the production of malleable cast iron (East Asian Institute occasional papers, 4). Copenhagen: East Asian Institute, University of Copenhagen, 1989.
Iron and steel in ancient China (Handbuch der Orientalistik, IV:9). Leiden: E. J. Brill, 1993.
Science and civilisation in China. Vol. 5: Chemistry and chemical technology. Part 11: Ferrous metallurgy. Cambridge University Press, 2008. 512 pp.
“The cast iron lion of Ts'ang-chou”, Needham Research Institute newsletter, no. 10, June 1991, pp. 2-3.
“Chinese monumental iron castings”, Journal of East Asian archaeology, 2000, 2.3/4: 199-224. | Colour Illustrations.
“Zhongguo gudai de daxing zhutieqi”, tr. by Li Yuan, in Wenwu keji yanjiu (“Scientific and technical research on cultural heritage”), vol. 5, Beijing: Kexue Chubanshe, 2007, pp. 68-82 + colour plates 10-15. Tr. of “Chinese monumental iron castings”, above.
“The casting of iron woks in Guangdong, China, in 1840”. Poster presentation, Founders, smiths and platers: International Conference on metal forming and finishing from the earliest times, Oxford, 20-24 September, 1999.
“Støbejerns metallurgi og lidt om kinesisk støbejern”, 53 pp. in Jern: Fremstilling, nedbrydning og bevaring. Fortryk af forelæsninger til Nordisk Videreuddannelse af Konservatorer, København, 17-28 august 1987. København: Nationalmuseet, Bevaringssektionen, 1987. (“The metallurgy of cast iron, with some notes on Chinese cast iron”, two lectures for museum conservators). The Google translation can be useful.
This page was last updated 5 November 2023.
