The development of ancient Indian wootz steel is reviewed. Wootz is the anglicized version of ukku in the languages of the states of Karnataka, and Andhra Pradesh, a term denoting steel. Literary accounts suggest that the steel from the southern part of the Indian subcontinent was exported to Europe, China, the Arab world and the Middle East.
Though an ancient material, wootz steel also fulfills the description of an advanced material, since it is an ultra-high carbon steel exhibiting properties such as superplasticity and high impact hardness and held sway over a millennium in three continents- a feat unlikely to be surpassed by advanced materials of the current era.
Wootz deserves a place in the annals of western science due to the stimulus provided by the study of this material in the 18th and 19th centuries to modern metallurgical advances, not only in the metallurgy of iron and steel, but also to the development of physical metallurgy in general and metallography in particular.
Some of the recent experiments in studying wootz by re-constructing composition, microstructure and mechanical behaviour, along with some recent archaeological evidence, are described.
Wootz, High-carbon Steel, South India, Superplasticity, Crucibles, Analyses
India has been reputed for its iron and steel since ancient times. Literary accounts indicate that steel from southern India was rated as some of the finest in the world and was traded over ancient Europe, China, the Arab world and the Middle East. Studies on wootz indicate that it was an ultra-high carbon steel with 1-2% carbon and was believed to have been used to fashion the Damascus blades with a watered steel pattern. Wootz steel also spurred developments in modern metallographic studies and also qualifies as an advanced material in modern terminology since such steels are shown to exhibit super-plastic properties. This paper reviews some of these developments.
2. History of wootz steel
There are numerous early literary references to steel from India from Mediterranean sources including one from the time of Alexander (3rd c. BC) who was said to have been presented with 100 talents of Indian steel, mentioned by Pant . Bronson  has summarised several accounts of the reputation of Indian iron and steel in Greek and Roman sources which suggest the export of high quality iron and steel from ancient India. Srinivasan , Biswas  and Srinivasan and Griffiths  have pointed out that the archaeological evidence from the region of Tamil Nadu suggests that the Indian crucible steel process is likely to have started before the Christian era from that region. Zaky  pointed out that it was the Arabs who took ingots of wootz steel to Damascus following which a thriving industry developed there for making weapons and armour of this steel, the renown of which has given the steel its name. In the 12th century the Arab Edrisi mentioned that the Hindus excelled in the manufacture of iron and that it was impossible to find anything to surpass the edge from Indian steel, and he also mentioned that the Indians had workshops where the most famous sabres in the world were forged, while other Arab records mention the excellence of Hinduwani or Indian steel as discussed by Egerton .
Several European travellers including Francis Buchanan  and Voysey  from the 17th century onwards observed the manufacture of steel in south India by a crucible process at several locales including Mysore, Malabar and Golconda. By the late 1600s shipments running into tens of thousands of wootz ingots were traded from the Coromandel coast to Persia. This indicates that the production of wootz steel was almost on an industrial scale in what was still an activity predating the Industrial Revolution in Europe.
Indeed the word wootz is a corruption of the word for steel ukku in many south Indian languages. Indian wootz ingots are believed to have been used to forge Oriental Damascus swords which were reputed to cut even gauze kerchiefs and were found to be of a very high carbon content of 1.5-2.0% and the best of these were believed to have been made from Indian steel in Persia (Figure 1) and Damascus according to Smith . Some of the finest swords and artefacts of Damascus steel seen in museums today are from the Ottoman region i.e. Turkey.
In India till the 19th century swords and daggers of wootz steel were made at centres including Lahore, Amritsar, Agra, Jaipur, Gwalior, Tanjore, Mysore, Golconda etc. although none of these centres survive today. Different types of Damascus sword
Figure 1. Detail of 17th century Persian blade of Damascus steel or Wootz steel showing typical etched crystalline structure of high-carbon steel (Smith )
patterns have been identified, described in some depth by Pant , who also identified a new design from blades kept in the collection of the Salar Jung Museum in Hyderabad.
It may be mentioned however that the term Damascus steel can refer to two different types of artefacts, one of which is the true Damascus steel which is a high carbon alloy with a texture originating from the etched crystalline structure, and the other is a composite structure made by welding together iron and steel to give a visible pattern on the surface. Although both were referred to as Damascus steels, Smith  has clarified that the true Damascus steels were not replicated in Europe until 1821.
3. Role of wootz steel in the development of modern metallurgy
The legends associated with the excellent properties of the wootz steel and the beautiful patterns on Damascus blades caught the imagination of European scientists in the 17th-19th centuries since the use of high-carbon iron alloys was not really known previously in Europe and hence played an important role in the development of modern metallurgy. British, French and Russian metallography developed largely due to the quest to document this structure. Similarly the textured Damascus steel was one of the earliest materials to be examined by the microstructure. Smith [10, 11] has fascinatingly elucidated this early historiography of the interest in the study of wootz steel and its significance to the growth of metallurgy.
Although iron and steel had been used for thousands of years the role of carbon in steel as the dominant element was found only in 1774 by the Swedish chemist Tobern Bergman, and was due to the efforts of Europeans to unravel the mysteries of wootz. Tobern Bergman was able to determine that the compositions of cast iron, steel and wrought iron varied due to the composition of plumbago i.e. graphite or carbon. As suggested by Smith  the Swedish studies received an impetus following the setting up of a factory to make gun barrels of welded Damascus steels, and it was on observation of the black and white etching of the steel and iron parts that a Swede metallurgist guessed that there was carbon in steel, and interest in replicating true Damascus steels followed.
In the early 1800s, following the descriptions of crucible steel making in south India by the European travellers, there was a spurt in interest in Europe in investigating south Indian wootz steel, from which the fabled Damascus blades were known to be made, with the aim of reproducing it on an industrial scale. Mushets  studies in 1804 were one of the first to correctly conclude that there was more carbon in wootz than in steel from England, although this idea did not gain currency until later. Michael Faraday , the inventor of electricity and one of the greatest of the early experimenters and material scientists, as pointed out by Peter Day , was also fascinated by wootz steel and enthusiastically studied it. Along with the cutler Stodart, Faraday attempted to study how to make Damascus steel and they incorrectly concluded that aluminium oxide and silica additions contributed to the properties of the steel and their studies were published in 1820 . They also attempted to make steel by alloying nickel and noble metals like platinum and silver and indeed Faradays studies did show that that the addition of noble metals hardens steel. Stodart  reported that wootz steel had a very fine cutting edge.
Following this the interest in Damascus steel moved to France. Wadsworth and Sherby  have pointed out that Faradays research made a big impact in France where steel research on weapons thrived in the Napoleonic period. The struggle to characterize the nature of wootz steel is well reflected in the efforts of Breant  in the 1820s from the Paris mint who conducted an astonishing number of about 300 experiments adding a range of elements ranging from platinum, gold. silver, copper, tin, zinc, lead, bismuth, manganese, arsenic, boron and even uranium, before he finally also came to the conclusion that the properties of Damascus steel were due to carburetted steel. Smith  has indicated that the analysis of ingots of wootz steel made in the 1800s showed them to have over 1.3% carbon. The Russian Anasoff  also studied the process of manufacturing wootz steel and succeeded in making blades of Damascus steel by the early 1800s.
In the early 1900s wootz steel continued to be studied as a special material and its properties were better understood as discussed further in the next section. Belaiew  reported that blades of such steel to cut a gauze handkerchief in midair. In 1912, Robert Hadfield  who studied crucible steel from Sri Lanka recorded that Indian wootz steel was far superior to that previously produced in Europe. Indeed in the 18th-19th century special steels were produced in Europe as crucible steels, as discussed by Barraclough .
4. Investigations of superplasticity and other mechanical properties of wootz steel
Some European scientists were successful in replicating and forging wootz and Stodart who used it in his cutlery business found that wootz steel had a superior cutting edge to any other, while Zschokke in 1924 found that with heat treatment this steel had special properties such as higher hardness, strength and ductility, mentioned by Smith . By 1918 an important finding concerning Damascus steel was made by Belaiew  who was probably the first to attribute the malleability of Damascus steel to the globulitic (i.e. spheroidised) nature of the forged steel and to recognize that this occurs during forging at a temperature of red heat (i.e. 700-800 0 C).
Panseri  in the 1960s was one of the first to point out that Damascus steel was a hypereutectoid ferrocarbon alloy with spheroidised carbides and carbon content between 1.2-1.8%. Recent studies have indicated that ultra-high carbon steels exhibit superplastic properties. As pointed out by Wadsworth and Sherby , by 1975 Stanford University had found that steels with 1-2.1% C i.e. ultrahigh carbon steels could be both superplastic at warm temperatures and strong and ductile at room temperatures. It was only subsequently that it came to the authors notice that these steels were in fact similar in carbon content to the Damascus steels.
Superplasticity is a phenomenon whereby an elongation of several hundred percent can be observed in certain alloys in tension, with neck free elongations and without fracture. By contrast most crystalline materials can be stretched to no more than 50-100 per cent. Superplasticity occurs at high temperatures and superplastic materials can be formed into complex shapes. For superplastic materials the index of strain rate sensitivity (m) is high, being around 0.5. At ideal m=1 flow stress is proportional to strain rate and the material behaves like a Newtonian viscous fluid such as hot glass. Superplasticity occurs only above 0.3-0.4 Tm K where Tm is the melting point. Another feature is that once super-plastic flow is initiated the flow stress required to maintain it is very low. Superplastic material essentially comprises of a two-phase material of spherical grains of extremely fine grain size of not more than 5 microns at the working temperature. Such ultrafine grained materials exhibit grain boundary sliding yielding superplastic properties.
Contemporary studies by Wadsworth and Sherby  and Sherby  indicated that UHCS (i.e. ultra-high carbon steels) with 1.8% C showed a strain-rate sensitivity exponent nearing 0.5 at around 7500 C (Figure 2) suggesting that Damascus steel could well have exhibited superplastic properties and a patent was awarded for the manufacture of such UHCS.
The explanation of the superplasticity of the steel is that the typical microstructure of ultra-high carbon steel with the coarse network of pro-eutectoid cementite forming along the grain boundaries of prior austenite (Figure 3 a, b), can lead to a fine uniform distribution of spheroidised cementite particles (0.1 m m diam.) in a fine grained ferrite matrix. This spheroidisation of cementite is described in Wadsworth and Sherby , Sherby  and Ghose et al. . Such steels are also found to have strength, hardness and wear resistance.
Figure 2. The flow stress-strain rate response of ultra-high carbon steel at 7500 C illustrates that the stress-strain rate curve has a slope showing a strain-rate sensitivity exponent of 0.43 indicating it is a superplastic material (Sherby )
Figure. 3 a) Photomicrograph of ultra-high carbon steel with 1.8% C, showing coarse pro- eutectoid carbide (cementite) network (Sherby )
b) Photomicrograph of same structure at high magnification shows iron grains with fine spheroidised carbides (Sherby )
Such steels had to be forged, however, in a narrow range of 850-6500 C and not at the white heat of 12000 C to get the desired fine grain structure and plasticity. In fact as pointed out in an appraisal of Indian crucible steel making by Rao , and in a review of ancient iron and steel in India by Biswas , the early European blacksmiths failed to duplicate Damascus blades because they were in the practice of forging only low carbon steels at white heat, which have a higher melting point. Biswas  mentions that the forging of wootz at high heat would have led to the dissolution of the cementite phase in austenite so that the steels were found to be brittle enough to crumble under the hammer.
Moreover, attractive combinations of strength and ductility were found to be achieved by Wadsworth and Sherby  and Sherby  when the ultra-high carbon steels were in spheroidised conditions with high yield strengths varying from 800 Mpa to 1500 Mpa with increasing fineness of spheroidised carbides, while the steel with coarsely spheroidised carbides was especially ductile with up to 23% tensile elongation.
While it is not yet known how fully the superplastic or superformable properties of this steel were exploited by the ancient blacksmiths of West Asia and India, accounts indicate that they were certainly able to manipulate the alloy with a skill that could not be easily replicated by the European experimenters of the 19th century. Indeed the swords of Damascus steel were reported to have high strength and ductility. Nevertheless, whereas the links between the patterns on the traditional Damascus blades and the crystalline structure of ultra-high carbon steels have been better established, the mechanical properties of the traditional Damascus blades and the degree of exploitation of the unique properties of the steel are less well understood.
Verhoeven  and Verhoeven et al. [28, 29] have attempted to re-invent the Damascus steel and blades as it were with replication experiments based on historical studies of Damascus blades and composition of wootz ingots. Verhoeven et al.  used two methods by which the ingots were made, one of which consisted of melting iron charge in a small sealed clay graphite crucible inside a gas-fired furnace with the ingot formed by furnace cooling. These were made by rapidly heating the charge and holding it for a period of 20-40 minutes between 14400 C-14800 C followed by cooling at furnace cooling rates or faster. The composition of the charge was chosen to match that of genuine Damascus blades of about 1.6% C and 0.1% P. However the fairly high level of phosphorus made the blades very hot short and difficult to forge. To overcome this problem the ingots were held at 12000 C in iron oxide to produce a protective rim of pure iron around the ingot which was ductile so that the ingot could be forged. Ingots were also made with the phosphorus levels reduced to the point where the ingots were not hot short which eliminated the need for the rim heat treatment. Verhoeven et al.  also made ingots by a process of vacuum-induced melting whereby the charge was melted by heating to around 10000 C, backfilling with nitrogen gas, heating to about 15800 C and then outgassing for around 5 minutes so that cooling rates at arrest temperature were around 5-100 C/minute.
It may be commented however, that although the structures of the ingots so produced do simulate those of Damascus blades, the methods used by Verhoeven et al.  are not strictly experimental re-constructions of the traditional processes, but rather laboratory simulations of the process, since the methods used do not really replicate conditions related to traditional or archaeological processes. For instance the charge is fired in both the methods described above in a very short time and the melt is cooled very rapidly under modern industrial conditions which could not have been achieved traditionally, while the 19th century descriptions of the wootz process suggest a very long firing cycle for the charge. In fact the eye witness descriptions of Voysey  and Buchanan  lay emphasis on the fact that the prolonged heating of the charge and its slow cooling were essential for obtaining the optimum results in the wootz process.
However the experimental simulations by Verhoeven et al.  served to monitor in detail the thermal cycles and cooling curves and composition so as to be able to arrive at a final product which matched that of Damascus blades and to understand the mechanism of formation of the pattern of aligned bands on the blades, which is reported by them to be produced by a carbide banding mechanism which was found to be assisted by the addition of P, S along with V, Cr, and Ti. Moreover their experiments are amongst the few comprensive studies on the general process of manufacture of the ingots themselves.
5. Archaeological and analytical evidence
Some of the archaeological and analytical evidence for crucible steel production is discussed covering the investigations of Rao , Rao et al. , Lowe [32, 33], Srinivasan  and Srinivasan and Griffiths . These indicate that the crucible processes for steel production were spread over large parts of south India. Lowes investigations have concentrated mainly on surveying and studying numerous sites from the Hyderabad region or the Deccani crucible steel process while pioneering investigations by Rao et al.  have covered other parts of south India such as the Mysore region and Salem district of Tamil Nadu. Field and analytical investigations were made by Srinivasan in 1990, whereby she was able to identify some hitherto unreported sites of crucible steel production in South Arcot, Tamil Nadu and from Gulbarga, Karnataka, reported in Srinivasan  and Srinivasan and Griffiths . Figure 4 gives a view of a dump for wootz crucible steel production from South Arcot, Tamil Nadu and Figure 5 of fragments of fired wootz crucibles from Gulbarga identified by Srinivasan.
Srinivasan  has pointed out that whereas the process documented by Lowe [32, 33], the Hyderabadi or Deccani process, involved the co-fusion of cast iron with wrought iron, the crucibles from sites reported by Srinivasan from Tamil Nadu and Karnataka pertained to the carburisation of wrought iron in crucibles by packing it with carbonaceous material. Analytical investigations made by Rao et al , Lowe [32, 33], Srinivasan , Craddock  and Srinivasan and Griffiths  on crucibles from production sites are briefly summarized.
The details of the furnace described and sketched by Buchanan  indicate that crucibles were packed in rows of about fifteen inside a sunken pit filled with ash to constitute the furnace which was operated by bellows of the buffalo hide, fixed into a perforated wall which separated them from the furnace probably to minimize fire hazards (Figure 6). The fire was stoked from a circular pit which was connected to the bottom of the ash pit. The crucibles themselves were conical and could contain up to 14 oz. of iron, along with stems and leaves. The wootz steel process in general refers to a closed crucible
Figure 4. View of newly identified old dump for high-carbon wootz crucible steel production from South Arcot, Tamil Nadu (photographed by S. Srinivasan)
Figure 5. Fragments of newly identified remains of fired
wootz crucibles from Gulbarga, Karnataka (photographed by S. Srinivasan)
Figure 6. Furnace for production of crucible steel production sketched by Buchanan (1807) during his travels, indicating that crucibles were packed in a pit with the furnace being operated by bellows of buffalo hide (reproduced from K. N. P. Rao, unpublished monograph)
process and Lowe  has remarked that the processing of plant and mineral materials in closed crucibles is often described in Indian alchemical Sanskrit texts of the 7th-13th c. AD.
Investigations by Craddock  indicated the wootz ingot itself had a dendritic cast structure. Lowe [32, 33] has investigated particularly well the refractory nature of the crucibles of the crucibles which indicate that they were robust enough refractories to withstand the long firing cycles of up to 24 hours for the process. The formation of mullite and cryistobalite was detected in the crucible fragments studied by Lowe [32, 33] suggesting they had been well fired to high temperatures of over 1300-14000 C, while Rao et al  also observed the formation of mullite and cryistobalite in crucibles.
However the microstructures investigated by Lowe  of the metal remnants within the particular Deccani crucibles studied by her from Konasamudram could only be related to a failed process of crucible steel production at that particular site or context since they related more to white cast iron, a brittle and not very malleable material formed by over-carburisation, rather than ultra-high carbon steel. In fact based on these findings Lowe  has preferred to cautiously aver that it was a white cast iron ingot that was produced by the Indian crucible process. Craddock  has also opined that the product of the Indian crucible steel process was probably a general homogenous steel rather than specifically a high-carbon steel.
On the other hand investigations by Srinivasan  and Srinivasan and Griffiths  indicated the presence of solidified metal droplets in the crucibles with a typical micro-structure and micro-hardness corresponding to a good quality hypereutectoid steel with the formation of hexagonal grains of prior austenite with fine lamellar pearlite within the grains, with the precipitation of pro-eutectoid cementite along the grain boundaries of prior austenite: which is in fact the classic structure of ultra-high carbon steels of about 1.5% C which were made under laboratory conditions by Wadsworth and Sherby [17}and Verhoeven et al. . The findings reported in Srinivasan  and Srinivasan and Griffiths  are hence significant in that they prove beyond doubt that high-carbon steels were indeed made by crucible processes in south India. Studies by Srinivasan and Griffiths  also indicated that temperatures of over 14000 C had indeed been reached inside the crucibles to melt the wrought iron and carburise it to get a molten high-carbon steel with the typical hypereutectoid structure on solidification.
The above review indicates that the reputation of wootz steel as an exceptional and novel material is one that has endured from early history right into the present day, with the story of the endeavours to study it in recent history being nearly as intriguing as the story of its past. The archaeological findings indicate that crucible steel does have an ancient history in the Indian subcontinent where it took roots as suggested by literary references, while the analytical investigations indicate that a high-grade ultra-high carbon steel was indeed produced by crucible processes in south India. Recent investigations on the properties of the ultra-high carbon wootz steel such as superplasticity justify it being called an advanced material of the ancient world with not merely a past but also perhaps a future.
The authors would like to acknowledge the Indian National Academy of Engineering. Srinivasan would like to acknowledge the support of British Council, New Delhi for a British Chevening Scholarship for doctoral research, and the interest of Dr. D. Griffiths, Institute of Archaeology, University College London, Dr. J. A. Charles, Cambridge University, late Dr. C. V. Seshadri, founder-President, Congress of Traditional Science and Technology, and Hutti Gold Mines Ltd. for assistance with fieldwork and the support of the Homi Bhabha Research Council.