MATERIALS CHARACIERZAION ELSEVIER Materials Characterization 45(2000)289-313 Ancient and modern laminated composites from the Great Pyramid of Gizeh to Y2K* Jeffrey Wadsworth ", Donald R. Lesuer awrence Livermore National Laboratory, University of California, PO Box 808, L-001, Livermore, CA 94551, USA Received 20 March 2000; accepted 1 June 2000 Abstract Laminated metal composites(LMCs) have been cited in antiquity; for example, an iron laminate that may date as far back as 2750 BC was found in the great Pyramid in gizeh in 1837. A laminated shield containing bronze, tin, and gold layers is described in detail by Homer. Well-known examples of steel laminates, such as an Adze blade, dating to 400 BC can be found in the literature. The Japanese sword is a laminated composite at several different levels and Merovingian blades were composed of laminated steels. Other examples are also available including composites from China, Thailand, Indonesia, Germany, Britain, Belgium, France, and Persia. The concept of lamination to provide improved properties has also found expression in modern materials. Of particular interest is the development of laminates including high-carbon and low-carbon layers. These materials have unusual properties that are of engineering interest; they are similar to ancient welded Damascus steels. The manufacture of collectable knives, labeled"welded Damascus, "has also been a focus of contemporary knife akers. Additionally, in the former Soviet Union, laminated composite designs have been used in engineering applications. Each of the above areas will be briefly reviewed, and some of the metallurgical principles will be described that underlie improvement in properties by lamination. Where appropriate, links are made between these property improvements and those that may have been present in ancient artifacts. C 2001 Elsevier Science Inc. All Keywords: Laminated metal composite; Low-carbon layer; High-carbon layer, Steel 1. Introduction metal composites(LMCs). LMCs consist of alternat ing metal or metal-containing layers that are bonded In a recent review he authors with"sharp"interfaces. These materials represent colleagues presented examples of historic unique laminated or composite form that is different nated composites and described in detail from graded materials, which have diffuse interfaces studies of the mechanical behavior of la or layered materials, in general, which can consist of altermating layers of a wide range of materials. LMCs can dramatically improve many properties including w Originally presented at the IMS Symposium held in fracture toughness, fatigue behavior, impact beha Cincinnati, OH, October 21-November 3. 1999 vior, wear, corrosion, and damping capacity: or Corresponding author. TeL: +1-925-423-2184; fax: provide enhanced formability or ductility for other +1-925-4243625 wise brittle materials. In many cases, through the E-mail address: wadsworth(lInl.gov(J. Wadsworth) choice of component materials, laminate architecture 1044-5803/00/S ont matter 2001 Elsevier Science Inc. All rights reserved. PI:S1044-5803(00)00077-2
Ancient and modern laminated composites Ð from the Great Pyramid of Gizeh to Y2K$ Jeffrey Wadsworth*, Donald R. Lesuer Lawrence Livermore National Laboratory, University of California, PO Box 808, L-001, Livermore, CA 94551, USA Received 20 March 2000; accepted 1 June 2000 Abstract Laminated metal composites (LMCs) have been cited in antiquity; for example, an iron laminate that may date as far back as 2750 BC was found in the Great Pyramid in Gizeh in 1837. A laminated shield containing bronze, tin, and gold layers is described in detail by Homer. Well-known examples of steel laminates, such as an Adze blade, dating to 400 BC can be found in the literature. The Japanese sword is a laminated composite at several different levels and Merovingian blades were composed of laminated steels. Other examples are also available, including composites from China, Thailand, Indonesia, Germany, Britain, Belgium, France, and Persia. The concept of lamination to provide improved properties has also found expression in modern materials. Of particular interest is the development of laminates including high-carbon and low-carbon layers. These materials have unusual properties that are of engineering interest; they are similar to ancient welded Damascus steels. The manufacture of collectable knives, labeled ``welded Damascus,'' has also been a focus of contemporary knife makers. Additionally, in the former Soviet Union, laminated composite designs have been used in engineering applications. Each of the above areas will be briefly reviewed, and some of the metallurgical principles will be described that underlie improvement in properties by lamination. Where appropriate, links are made between these property improvements and those that may have been present in ancient artifacts. D 2001 Elsevier Science Inc. All rights reserved. Keywords: Laminated metal composite; Low-carbon layer; High-carbon layer; Steel 1. Introduction In a recent review [1], the authors and their colleagues presented examples of historical laminated composites and described in detail modern studies of the mechanical behavior of laminated metal composites (LMCs). LMCs consist of alternating metal or metal-containing layers that are bonded with ``sharp'' interfaces. These materials represent a unique laminated or composite form that is different from graded materials, which have diffuse interfaces, or layered materials, in general, which can consist of alternating layers of a wide range of materials. LMCs can dramatically improve many properties including fracture toughness, fatigue behavior, impact behavior, wear, corrosion, and damping capacity; or provide enhanced formability or ductility for otherwise brittle materials. In many cases, through the choice of component materials, laminate architecture 1044-5803/00/$ ± see front matter D 2001 Elsevier Science Inc. All rights reserved. PII: S1044-5803(00)00077-2 $ Originally presented at the IMS Symposium held in Cincinnati, OH, October 21 ± November 3, 1999. * Corresponding author. Tel.: +1-925-423-2184; fax: +1-925-424-3625. E-mail address: wadsworth3@llnl.gov (J. Wadsworth). Materials Characterization 45 (2000) 289 ± 313
J. Wadsworth, D.R. Lesuer /Materials Characterization 45(2000)289-313 (such as volume percent of the con materials g) was discovered by an excavation team near an Ind layer thickness), and processin ry, LMCs ir passage(Southern side) in the great Pyramid at can be engineered to produce with pre- Gizeh, Egypt. The location of the plate was within scribed properties. an undisturbed section high up on the pyramid In the present paper, Section 2 deals with the The plate was removed to the British Museum and history of laminated composites. In particular, exam- was not examined for its structure until El Gayer ples of their existence, composition, and structure are nd Jones used modern metallographic techniques detailed. Extra emphasis is given to the Japanese on a small (1.7 g) sample from the plate and sword both because of its complexity as a laminate published their findings in 1989[]. A comment at several levels, and its relatively well-documented by Craddock and Lang [3 was included in the history and the technical details tha same issue of the journal story. Included at the end of the section are some The significance of the plate is twofold. First, if it comments on modern knives that duplicate, in part, can be shown to be contemporaneous with the build the ancient weapons. Section 3 first describes modern ing of the pyramid, then it is one of the oldest known engineering applications of LMCs. Scientific and plates of iron metal ever discovered and dates from engineering studies on laminated composites are then the 4th Dynasty, circa 2750 BC. Second, the metallo- presented- in some cases at the very thin-layer graphic study of El Gayer and Jones revealed that the level, and in others at layer thicknesses that were close to those found in ancient laminates. processing methods, as well as strength, durability, toughness and damping properties, are discussed. The mechan- hese laminates have been inexpertly wela ates of wrought iron and isms of ving toughn ness by lamination are de- gether by hammering. The various layers differ scribed in detail om each other in their grain sizes, carbon contents Where possible, the mechanisms leading to im- the nature of their non-metallic inclusions and in proved properties of modern engineered laminated their thicknesses composites are linked back to ancient artifacts It was further deduced from elongated non-metal- lic inclusions that the welding process had been 2. History carried out at modest temperatures( 800 C)allow- ing recrystallization of the iron matrix grains. The The idea of laminating similar or dissimilar metals bsence of metallic copper globules and only small or alloys to form a composite material has been traces of elemental copper suggested that the plate known from antiquity. The motivations for laminat had not been produced as a by-product of copper ing metals are varied. For example, in carburizing the smelting operations of iron-rich copper ores. Also, earliest forms of wrought iron, only thin layers could a chemical analysis reported in 1926 revealed only be carburized and so lamination was a way to create race levels of nickel, thereby confirming the plate ulk material. ( This could be the motivation to be of terrestrial (but not natural) origin rather most ancient laminates Another reason is that the than to be meteoric [2]. (It is noted that the above ard material, steel, was rare and it was expedient to view on lamination is not universally agreed upon sandwich it between more common materials (Thi An alternate view is that the heterogeneous nature motive is found in medieval knives )From a mechan- of te is a direct result of a heterogeneou ical viewpoint, optimizing the combination of starting piece [4]). trength, toughness, and sharpness is the basis for Summarizing, El Gayer and Jones concluded that lamination.(Examples include the Japanese sword, the iron pieces comprising the laminate were the Halberd, and modern laminates. Finally, there is a strong motivation based on decorative appeal .intentionally produced during small-scale(and, Many modern knives are made in laminated form ossibly, very primitive) operations primarily de- for this reason but it could have been a motive in signed for the production of iron metal(rather than ancient knives also. ) Some selected examples of copper metal). Furthermore, the presence of abur laminated materials follow ant inclusions of unreduced (or incompletely oxides in the metal 2.1. Laminated iron plate found at the great Pyramid laminations shows that the'smelting operations had of Gizeh rried out at low temperatures (probably between 1000"C and 1100C)and that the ron had been produced by the direct reduction In 1837, an iron plate (26 cm x 86 cm x a method maximum thickness of 0.4 cm and weighing 750 produced In which no molten iron is normally
(such as volume percent of the component materials and layer thickness), and processing history, LMCs can be engineered to produce a material with prescribed properties. In the present paper, Section 2 deals with the history of laminated composites. In particular, examples of their existence, composition, and structure are detailed. Extra emphasis is given to the Japanese sword both because of its complexity as a laminate at several levels, and its relatively well-documented history and the technical details that accompany that history. Included at the end of the section are some comments on modern knives that duplicate, in part, the ancient weapons. Section 3 first describes modern engineering applications of LMCs. Scientific and engineering studies on laminated composites are then presented Ð in some cases at the very thin-layer level, and in others at layer thicknesses that were close to those found in ancient laminates. Processing methods, as well as strength, durability, toughness, and damping properties, are discussed. The mechanisms of improving toughness by lamination are described in detail. Where possible, the mechanisms leading to improved properties of modern engineered laminated composites are linked back to ancient artifacts. 2. History The idea of laminating similar or dissimilar metals or alloys to form a composite material has been known from antiquity. The motivations for laminating metals are varied. For example, in carburizing the earliest forms of wrought iron, only thin layers could be carburized and so lamination was a way to create bulk material. (This could be the motivation for the most ancient laminates.) Another reason is that the hard material, steel, was rare and it was expedient to sandwich it between more common materials. (This motive is found in medieval knives.) From a mechanical viewpoint, optimizing the combination of strength, toughness, and sharpness is the basis for lamination. (Examples include the Japanese sword, the Halberd, and modern laminates.) Finally, there is a strong motivation based on decorative appeal. (Many modern knives are made in laminated form for this reason, but it could have been a motive in ancient knives also.) Some selected examples of laminated materials follow. 2.1. Laminated iron plate found at the Great Pyramid of Gizeh In 1837, an iron plate (26 cm 86 cm a maximum thickness of 0.4 cm and weighing 750 g) was discovered by an excavation team near an air passage (Southern side) in the Great Pyramid at Gizeh, Egypt. The location of the plate was within an undisturbed section high up on the pyramid. The plate was removed to the British Museum and was not examined for its structure until El Gayer and Jones used modern metallographic techniques on a small (1.7 g) sample from the plate and published their findings in 1989 [2]. A comment by Craddock and Lang [3] was included in the same issue of the journal. The significance of the plate is twofold. First, if it can be shown to be contemporaneous with the building of the pyramid, then it is one of the oldest known plates of iron metal ever discovered and dates from the 4th Dynasty, circa 2750 BC. Second, the metallographic study of El Gayer and Jones revealed that the plate consists of: ...numerous laminates of wrought iron and that these laminates have been inexpertly welded together by hammering. The various layers differ from each other in their grain sizes, carbon contents, the nature of their non-metallic inclusions, and in their thicknesses. It was further deduced from elongated non-metallic inclusions that the welding process had been carried out at modest temperatures ( 800°C) allowing recrystallization of the iron matrix grains. The absence of metallic copper globules and only small traces of elemental copper suggested that the plate had not been produced as a by-product of copper smelting operations of iron-rich copper ores. Also, a chemical analysis reported in 1926 revealed only trace levels of nickel, thereby confirming the plate to be of terrestrial (but not natural) origin rather than to be meteoric [2]. (It is noted that the above view on lamination is not universally agreed upon. An alternate view is that the heterogeneous nature of the plate is a direct result of a heterogeneous starting piece [4]). Summarizing, El Gayer and Jones concluded that the iron pieces comprising the laminate were: ...intentionally produced during small-scale (and, possibly, very primitive) operations primarily designed for the production of iron metal (rather than copper metal). Furthermore, the presence of abundant inclusions of unreduced (or incompletely reduced) fragments of iron oxides in the metal laminations shows that the `smelting' operations had been inexpertly carried out at low temperatures (probably between 1000°C and 1100°C) and that the iron had been produced by the `direct reduction' method Ð in which no molten iron is normally produced. 290 J. Wadsworth, D.R. Lesuer / Materials Characterization 45 (2000) 289±313
J. Wadsworth, D.R. Lesuer /Materials Characterization 45(2000) 289-313 And, most importantly, they also concluded Herbert Hoover (1928-1932), was a mining and metallurgical engineer (Stanford University, 1896) Furthermore, the metallurgical evidence supports the archaeological evidence, which suggests that the who became a famous and wealthy engineer before entering the political scene. He and his that structure was being bui wife translated the famous text de re metallica by Agricola, from the Latin to the English in 1912 Although accounts by the excavation teams em- [6, p. 421]. In that book, he footnoted his phasize the fact that the plate was found within the thoughts on the history of iron in Agricolas pyramid, and is therefore contemporaneous with the section on iron making. He considered that the amid. this view has not been generally a beginning of the Iron Age was in the prehistory period, that the egyptians knew iron 5000 to 6000 Subsequent to the paper by El Gayer and Jones, years ago, and used iron tools to carve the stones the only other investigation of the plate came in of the great pyramids. Thus, if the iron plate of 1993 by Craddock and Lang [5]. They agreed with Gizeh could be accurately dated, it would be a he El Gayer and Jones study that the structure was significant point in determining the evolution of similar to banded, wrought iron consisting of areas large, man-made, iron-based artifacts of varying carbon content. However, the absence o In order to resolve the issue of the date of the slag stringers and the presence of very large num- plate, it is possible to turn toC dating. Using this bers of other inclusions, containing unusually high technique, the dating of ancient steels has in fact levels of Ca and P, led Craddock and Lang to a been done successfully. In the last decade in parti- quite different conclusion regarding the method o cular, carbon dating on relati manufacture(and therefore the origin and likely weighing as little as a fraction of a gram to several age)of the plate. They believe the structure to be grams, has been accomplished by using accelerator one derived from "cast iron smelted with charcoal and then treated by the finery process to remove the carbon and produce a solid lump or bloom of wrought iron. "They go on to cite work proposing that this technique was the usual method of making iron in the post-medieval Islamic world. The Gizeh plate remains unusual, even in this scenario, be- ause of the very high level of inclusions that it contains. Craddock and Lang do not. on the basis of their 1993 analysis, believe the plate is con- temporaneous with the pyramid, concluding that: the plate of iron from the Great Pyramid is of no great antiquity. Nonetheless, these authors confirm that if its age were to be contemporaneous with the d that it would be the earliest substantial piece of iron known, a finding accepted by the famous scientist Petrie in 1883 Given these controversial and competing views, it is worth emphasizing the importance of the late of the plate. It is generally accepted that iron and steel were not made in this quantity until about 1500 BC. Certainly, examples exist in that me frame. some of them fame example, daggers of both gold and iron were found on Tutankhamun's mummy (Fig. 1)which known to be from 1350 BC. There are occasional claims that older pieces exist; for example, it is claimed that an iron knife blade in a museum in Turkey is from 2500 BC, but there is no supportin Fig. 1. Evidence of steel from 1350 BC. Daggers of iron and Nonetheless, some significant authors have pro- gold from Tutankhamun's grave, and their sheaths. Insert: posed a much older start to the Iron Age. For Position of iron dagger on Tutankhamun's mummy(after example, a former president of the United States, Sherby [7D
And, most importantly, they also concluded: Furthermore, the metallurgical evidence supports the archaeological evidence, which suggests that the plate was incorporated within the pyramid at the time that structure was being built. Although accounts by the excavation teams emphasize the fact that the plate was found within the pyramid, and is therefore contemporaneous with the pyramid, this view has not been generally accepted by archeologists. Subsequent to the paper by El Gayer and Jones, the only other investigation of the plate came in 1993 by Craddock and Lang [5]. They agreed with the El Gayer and Jones study that the structure was similar to banded, wrought iron consisting of areas of varying carbon content. However, the absence of slag stringers and the presence of very large numbers of other inclusions, containing unusually high levels of Ca and P, led Craddock and Lang to a quite different conclusion regarding the method of manufacture (and therefore the origin and likely age) of the plate. They believe the structure to be one derived from ``cast iron smelted with charcoal, and then treated by the finery process to remove the carbon and produce a solid lump or bloom of wrought iron.'' They go on to cite work proposing that this technique was the usual method of making iron in the post-medieval Islamic world. The Gizeh plate remains unusual, even in this scenario, because of the very high level of inclusions that it contains. Craddock and Lang do not, on the basis of their 1993 analysis, believe the plate is contemporaneous with the pyramid, concluding that: ``the plate of iron from the Great Pyramid is of no great antiquity.'' Nonetheless, these authors confirm that if its age were to be contemporaneous with the pyramid that it would be ``the earliest substantial piece of iron known,'' a finding accepted by the famous scientist Petrie in 1883. Given these controversial and competing views, it is worth emphasizing the importance of the date of the plate. It is generally accepted that iron and steel were not made in this quantity until about 1500 BC. Certainly, examples exist in that time frame, some of them famous ones. For example, daggers of both gold and iron were found on Tutankhamun's mummy (Fig. 1) which is known to be from 1350 BC. There are occasional claims that older pieces exist; for example, it is claimed that an iron knife blade in a museum in Turkey is from 2500 BC, but there is no supporting evidence presented. Nonetheless, some significant authors have proposed a much older start to the Iron Age. For example, a former president of the United States, Herbert Hoover (1928 ± 1932), was a mining and metallurgical engineer (Stanford University, 1896), who became a famous and wealthy engineer before entering the political scene. He and his wife translated the famous text De Re Metallica by Agricola, from the Latin to the English in 1912 [6, p. 421]. In that book, he footnoted his thoughts on the history of iron in Agricola's section on iron making. He considered that the beginning of the Iron Age was in the prehistory period, that the Egyptians knew iron 5000 to 6000 years ago, and used iron tools to carve the stones of the great pyramids. Thus, if the iron plate of Gizeh could be accurately dated, it would be a significant point in determining the evolution of large, man-made, iron-based artifacts. In order to resolve the issue of the date of the plate, it is possible to turn to 14C dating. Using this technique, the dating of ancient steels has in fact been done successfully. In the last decade in particular, carbon dating on relatively small samples, weighing as little as a fraction of a gram to several grams, has been accomplished by using accelerator Fig. 1. Evidence of steel from 1350 BC. Daggers of iron and gold from Tutankhamun's grave, and their sheaths. Insert: Position of iron dagger on Tutankhamun's mummy (after Sherby [7]). J. Wadsworth, D.R. Lesuer / Materials Characterization 45 (2000) 289±313 291
J. Wadsworth, D.R. Lesuer / Materials Characterization 45(2000)289-313 mass spectrometry (AMS). One of the best AMs carbon contents (up to 2.5% in both combined machines is at the Lawrence Livermore Laboratory. graphitic forms). Although these contributions The possibility of establishing a capability to confuse the dating of ancient artifacts, the presend steel and iron objects is being explored by one of of meteorite iron can be identified by its high Ni the authors and a colleague [8 content(4% to 7%)whereas terrestrial iron is rare Cresswell [9 in 1992 published a good sum- and is sufficiently well documented to not be mary of the history, issues, and limitations sur- problematical. Other complications arising from the rounding carbon dating of iron and steel artifacts. use of coal can sometimes be indirectly determined In summary, Turekian first conceived the use of Wootz steel from Sri Lanka was analyzed by C dating for artifacts and van der Merwe built a Cresswell who describes it as"ideal for C-dating system at Yale to do so in the early 1960s. The the sample size required was a mere 274 mg becaus ystem required 500 mg of carbon equivalent, of the high C content of the wootz(1.79%C) however, which corresponded to up to 1 kg o Subsequent to the work of Cresswell, other studies iron thereby severely restricting the use of the have appeared. For example, Nakamura et al. [11] technique. Developments in the 1970-1990 period dated a Japanese sword (2.27 g, 0.49%C), a planing level of tens of grams. The transition from propor Qlg.(0.93 g, 3.6%C), and an iron hook(4.53 g. succeeded in reducing sample size, but only to the adze C). Table 1 summarizes the dating of steel ional counters to the AMs technique of Cresswell artifacts using AMs. [10 in 1991 allowed reductions in carbon equiva- Thus, it is concluded that it should be possible lent to 5 mg. The sample size(which of necessity to date the iron plate at Gizeh using modern AMS depends upon the C content of the steel) ranges C techniqu from about 5 g for wrought iron (0.05% C) about 100 mg for cast iron of 2%C content. ForC dating to be meaningful, the source of C 2.2. Achilles shield used in the steel making has to be charcoal or freshly cut wood. Sources such as coal and coke are ex Perhaps the earliest known reference to improved hausted of"C. The dilution or contamination by lime properties in an LMC can be found in "The iliad of nd recycling of artifacts also has to be considered Homer"[12] in 800 BC which describes achilles Only in the 19th century did coke become a universal shield as having five layers- two bronze, two tin, fuel in the industrial world. In fact, up until the and one gold. The laminate was in the sequence Industrial Revolution, most smelting was carried out bronze/tin/gold/tin/bronze. During combat, the super- ior performance of the laminate was demonstrated by using charcoal-fired furnaces and historical records the fact that Aeneas'bronze spear penetrated the first ndicate that freshly cut wood was a fuel source. s extensive was the use of charcoal that vast deforest- two layers but stuck in the gold layer. Some details of the encounter between Aeneas and achilles can b tion took place in the US in Pennsylvania in the 17th found in translations of Homer as shown below and 18th century. In England, an act was passed by imber for fueling iron smelts. It is worth noting, ger therefor, sha thou not by words tum me till we have fought however, that the romans and the chinese from the with the bronze man to man; nay, come, let us 4th century Ad did use coaL. forthwith make tria ch of the other with Cresswell points out that meteoric iron, or even bronze-tipped spears. terrestrial iron, can also be incorporated in ancient He spake, and let drive his mighty spear against steel making and these forms can have quite high the other 's dread and wondrous shield. and loud summary of ancient iron and steel artifacts dated using theC technique by accelerator mass spectrometry %o c Age, BP 0.18 4.53 1330±110 Nakamura et al. [11 2.27 80±150 Nakamura et al. [11] Frobisher bloom 1340±70 Luriston dagger 0.30-1.0 940±60 Cresswell [9, 10] MIT dagger 0.30-10 2880±60 0.274 980±40 Cresswell [9, 1t Cast iron planing adze 3.6 0.93 1770±160 Nakamura et al. [11] Smallest sample dated
mass spectrometry (AMS). One of the best AMS machines is at the Lawrence Livermore Laboratory. The possibility of establishing a capability to age steel and iron objects is being explored by one of the authors and a colleague [8]. Cresswell [9] in 1992 published a good summary of the history, issues, and limitations surrounding carbon dating of iron and steel artifacts. In summary, Turekian first conceived the use of 14C dating for artifacts and van der Merwe built a system at Yale to do so in the early 1960s. The system required 500 mg of carbon equivalent, however, which corresponded to up to 1 kg of iron thereby severely restricting the use of the technique. Developments in the 1970 ± 1990 period succeeded in reducing sample size, but only to the level of tens of grams. The transition from proportional counters to the AMS technique of Cresswell [10] in 1991 allowed reductions in carbon equivalent to 5 mg. The sample size (which of necessity depends upon the C content of the steel) ranges from about 5 g for wrought iron (0.05% C) to about 100 mg for cast iron of 2% C content. For 14C dating to be meaningful, the source of C used in the steel making has to be charcoal or freshly cut wood. Sources such as coal and coke are exhausted of 14C. The dilution or contamination by lime and recycling of artifacts also has to be considered. Only in the 19th century did coke become a universal fuel in the industrial world. In fact, up until the Industrial Revolution, most smelting was carried out using charcoal-fired furnaces and historical records indicate that freshly cut wood was a fuel source. So extensive was the use of charcoal that vast deforestation took place in the US in Pennsylvania in the 17th and 18th century. In England, an act was passed by Queen Elizabeth I in 1558 restricting the use of timber for fueling iron smelts. It is worth noting, however, that the Romans and the Chinese from the 4th century AD did use coal. Cresswell points out that meteoric iron, or even terrestrial iron, can also be incorporated in ancient steel making and these forms can have quite high carbon contents (up to 2.5% in both combined and graphitic forms). Although these contributions could confuse the dating of ancient artifacts, the presence of meteorite iron can be identified by its high Ni content (4% to 7%) whereas terrestrial iron is rare and is sufficiently well documented to not be problematical. Other complications arising from the use of coal can sometimes be indirectly determined. Wootz steel from Sri Lanka was analyzed by Cresswell who describes it as ``ideal for 14C-dating''; the sample size required was a mere 274 mg because of the high C content of the wootz (1.79% C). Subsequent to the work of Cresswell, other studies have appeared. For example, Nakamura et al. [11] dated a Japanese sword (2.27 g, 0.49% C), a planing adze (0.93 g, 3.6% C), and an iron hook (4.53 g, 0.18% C). Table 1 summarizes the dating of steel artifacts using AMS. Thus, it is concluded that it should be possible to date the iron plate at Gizeh using modern AMS 14C techniques. 2.2. Achilles shield Perhaps the earliest known reference to improved properties in an LMC can be found in ``The Iliad of Homer'' [12] in 800 BC which describes Achilles' shield as having five layers Ð two bronze, two tin, and one gold. The laminate was in the sequence bronze/tin/gold/tin/bronze. During combat, the superior performance of the laminate was demonstrated by the fact that Aeneas' bronze spear penetrated the first two layers but stuck in the gold layer. Some details of the encounter between Aeneas and Achilles can be found in translations of Homer as shown below: But from battle, seeing I am eager therefor, shalt thou not by words turn me till we have fought with the bronze man to man; nay, come, let us forthwith make trial each of the other with bronze-tipped spears. He spake, and let drive his mighty spear against the other's dread and wondrous shield, and loud Table 1 A summary of ancient iron and steel artifacts dated using the 14C technique by accelerator mass spectrometry Object % C Grams Age, BP Reference Iron hook 0.18 4.53 1330 110 Nakamura et al. [11] Japanese sword 0.49 2.27 880 150 Nakamura et al. [11] Frobisher bloom 0.30 1.34 1340 70 Cresswell [9,10] Luriston dagger 0.30 ± 1.0 0.485 2940 60 Cresswell [9,10] MIT dagger 0.30 ± 1.0 1.44 2880 60 Cresswell [9,10] Sri Lankan wootz a 1.79 0.274 980 40 Cresswell [9,10] Cast iron planing adze 3.6 0.93 1770 160 Nakamura et al. [11] a Smallest sample dated. 292 J. Wadsworth, D.R. Lesuer / Materials Characterization 45 (2000) 289±313
J. Wadsworth, D.R. Lesuer /Materials Characterization 45(2000) 289-313 Peleus held the shield from him with his stout hand. eing seized with dread for he deemed that the far- shadowing spear of great-hearted Aeneas would lightly pierce it through- fool that he was, nor knew in his mind and heart that not easy are the they give place withal. Nor did the mighty wise-hearted Aeneas then break through the shield, for the gold stayed it, the gift of Howbeit through two folds he drave there still three, for five layers had the god welded, two of bronze, and two within of tin, nd one of gold, in which the spear of ash was Achilles then turns his spear upon Aeneas shie as made in about 400 BC. found at Al Mina on the but that shield also protects Aeneas. The contest tums realizes Achilles and Aeneas will kill each other, and thank Dr. R. Maddin for permission to publish the f ors to swords and stones. Poseidon intervenes when he and the backing piece is a low-carbon steel. The aut spirits Aeneas away while shedding a"mist over the eyes of Achilles. It is of interest to note that steel was known and described in the iliad but was not ed in the description of a tough laminated material For example, as the battle subsequently continues, tectoid. cementite stringers as a result of the Hector of the Trojans goes against Achilles but fears specialized processing of the ultrahigh carbon his fury as the flashing steel. "Later, a description o content(1.3% to 1.8% C)steels. However, it is the hardening of steel by quenching is given in quite likely that in some cases, laminated compo- the 33rd book of the iliad Achilles offer sites were developed in an attempt to duplicate the iron as a valuable prize at the funeral Damascus steel pattern. This is because such Patrochus, although there is the possibility that this pattems certainly could appear to be the result Is meteoric iron of the intimate mixing of two dissimilar metals Damascus steels were famous through the centu- ries. Their use in the Crusades by Saladin, the 23. Adze blade leader of the Saracen warriors, in a meeting with Richard the lionhearted, was immortalized by Sir The solid state joining of two dissimilar ferrous Walter Scott in his book(1871)"The Talism early as the first millennium BC [13]. The welded which was subsequently made into two movies and a BBC television mini-series, thereby bringing product often consisted of a steel and an iron. a the fame of these steels into the modern times photomicrograph is shown in Fig. 2, of an adze blade (a cutting tool used in farming), made by Greek 2.4. Chinese steel of hundred refining blacksmiths around 400 BC. The figure shows a fairly sharp interface between a carburized iron cut- Rubin describes early iron making in China in ing blade adjoining a low-carbon backing plate. The a recent paper [15]. He examined over 1000 iron blade was found at Al Mina, the ruins of a Greek artifacts from 60 iron-making sites and tombs. The trading colony on the coast of Turkey near Syria. The artifacts dated from between 900 bc and 1800 motive for using a sheet of carburized iron for the AD. Of interest to the present paper is his working face of the adze but soft iron for the other discussion of steel of hundred refinings the was an economical one based on the scarcity of phrase"hundred refinings make quality steel"is a carburized iron Chinese saying dating to the 2nd century AD It is worth noting that in about this era, it is In examination of a knife of "30 refinings from believed that Alexander the great was given a gift 112 AD. Rubin notes that the knife the India steel,“ wootz,” by the Indian King Peru. The wootz, which was contained in a gold seemed to be a box. was the starting material for the famous 6 layers. The su was proposed that the Damascus steels. The patterns on Damascus steel number of refinings ed the number of laye arise from the aggregation of spheroidized, proeu- after repeated doubling. Thus 30 and 100 refining
rang the shield about the spear-point. And the son of Peleus held the shield from him with his stout hand, being seized with dread; for he deemed that the farshadowing spear of great-hearted Aeneas would lightly pierce it through Ð fool that he was, nor knew in his mind and heart that not easy are the glorious gifts of the gods for mortal men to master or that they give place withal. Nor did the mighty spear of wise-hearted Aeneas then break through the shield, for the gold stayed it, the gift of the god. Howbeit through two folds he drave it, yet were there still three, for five layers had the crook-foot god welded, two of bronze, and two within of tin, and one of gold, in which the spear of ash was stayed [12]. Achilles then turns his spear upon Aeneas' shield, but that shield also protects Aeneas. The contest turns to swords and stones. Poseidon intervenes when he realizes Achilles and Aeneas will kill each other, and spirits Aeneas away while shedding a ``mist over the eyes of Achilles.'' It is of interest to note that steel was known and described in the Iliad, but was not used in the description of a tough laminated material. For example, as the battle subsequently continues, Hector of the Trojans goes against Achilles but fears ``his fury as the flashing steel.'' Later, a description of the hardening of steel by quenching is given. Also, in the 33rd book of the Iliad, Achilles offers a lump of iron as a valuable prize at the funeral games of Patrochus, although there is the possibility that this is meteoric iron. 2.3. Adze blade The solid state joining of two dissimilar ferrous materials is well documented as being practiced as early as the first millennium BC [13]. The welded product often consisted of a steel and an iron. A photomicrograph is shown in Fig. 2, of an adze blade (a cutting tool used in farming), made by Greek blacksmiths around 400 BC. The figure shows a fairly sharp interface between a carburized iron cutting blade adjoining a low-carbon backing plate. The blade was found at Al Mina, the ruins of a Greek trading colony on the coast of Turkey near Syria. The motive for using a sheet of carburized iron for the working face of the adze, but soft iron for the other face, was an economical one based on the scarcity of carburized iron. It is worth noting that in about this era, it is believed that Alexander the Great was given a gift of the India steel, ``wootz,'' by the Indian King, Peru. The wootz, which was contained in a gold box, was the starting material for the famous Damascus steels. The patterns on Damascus steel arise from the aggregation of spheroidized, proeutectoid, cementite stringers as a result of the specialized processing of the ultrahigh carbon content (1.3% to 1.8% C) steels. However, it is quite likely that in some cases, laminated composites were developed in an attempt to duplicate the Damascus steel pattern. This is because such patterns certainly could appear to be the result of the intimate mixing of two dissimilar metals. Damascus steels were famous through the centuries. Their use in the Crusades by Saladin, the leader of the Saracen warriors, in a meeting with Richard the Lionhearted, was immortalized by Sir Walter Scott in his book (1871) ``The Talisman,'' which was subsequently made into two movies and a BBC television mini-series, thereby bringing the fame of these steels into the modern times. 2.4. Chinese steel of hundred refinings Rubin describes early iron making in China in a recent paper [15]. He examined over 1000 iron artifacts from 60 iron-making sites and tombs. The artifacts dated from between 900 BC and 1800 AD. Of interest to the present paper is his discussion of ``steel of hundred refinings.'' The phrase ``hundred refinings make quality steel'' is a Chinese saying dating to the 2nd century AD. In examination of a knife of ``30 refinings,'' from 112 AD, Rubin notes that the knife: ...seemed to be a composite of approximately 30 ± 36 layers. The suggestion was proposed that the number of refinings specified the number of layers after repeated doubling. Thus 30 and 100 refining Fig. 2. Shown in the above figure is an Adze blade, which was made in about 400 BC, found at Al Mina on the coast of Turkey. The cutting edge is medium carbon steel and the backing piece is a low-carbon steel. The authors thank Dr. R. Maddin for permission to publish the above photograph [14]. J. Wadsworth, D.R. Lesuer / Materials Characterization 45 (2000) 289±313 293
J. Wadsworth, D.R. Lesuer /Materials Characterization 45(2000) 289-313 Also, in 1960, Sir Cyril Stanley Smith [1 6 wrote ely and sword[s]of 60 or 50 refinings(actually 64 that in his opinion: yers)should be expected. The Japanese sword blade is the supreme the prediction was made. A sword of marked metallurgical art. omb in Xuzhou, Jiang-su .. Examination showed Certainly, world wide, the Japanese sword (Nippon that . it consisted of alternative low- and high- to)is one of the most famous of all swords. The sword carbon layers, about 60 in total. The cost of the has always held a special place in the history and marked as worthy of 1500 coins, culture of Japan. Japanese legend, for example, tells us that Susanoo brother of the Sun goddess amater half years"living asu, slew an eight-headed dragon with a single stroke of a sword. Despite this, the blade was deemed inferior 2.5. Merovingian pattern-welded blade and Susanoo was given a magnificent sword from the dragon's tail. He gave the sword to his sister, the Sun An early European technique dating from about Goddess, who then handed to her grandson the Imper the end of the second century resulted in the ial Regalia that included three items-the jewel, the Merovingian patterm-welded blades. (Iron objects mirror, and the sword. manufactured prior to this date are frequently too From a metallurgical viewpoint, the Japanes severely corroded for their surfaces to be evaluated sword is of interest at several different levels. First, metallurgically. According to Smith [16], Merovin- their manufacture involved the solid-state bonding of blades were primarily manufactured on the steels to themselves and to steels of radically different Rhine although they were widespread through trade carbon contents. That is, in its simplest form, with the and war. The blades consisted of strips of pure iron exception of the very earliest blades, the sword is a and carbon steel (or iron strips that had been composite. An example of one of the composite de- carburized on one side) hammered or forged to- signs is that of a high-carbon external sheath surround gether in a manner involving folding or twisting The cutting edge consisted of the high-carbon-con tent steel, often inserted between plates of low- carbon material. Upon grinding to shape after heat yers become visible. In addition, it is quite likely that etching in fruit juice or sour beer was carried out to develop the patterns. An example is given in Fig 3, from a blade discovered in a Viking grave in South Finland [17] Thus, beginning in the period around 500 AD pattern-welded daggers and swords were made, ncluding Viking blades starting in about 600 AD. Smith [16 comments that the edges martensitic between plates of iron in similar to the Japanese sword. He also notes that it is" difficult to justify the particular pattern used in the center of the swords on any but aesthetic grounds. " Included in this group of materials is the carly Japanese sword. 2.6. Japanese sword The Japanese sword has universally been regarded by metallurgists as the ultimate metallurgical accomplishment. For example, in 1962 Edgar C. Bain [19]wrote attern- welded blade discover University [18]. Courtesy of The Gun Report
would probably indicate 32 and 128 layers respectively and sword[s] of 60 or 50 refinings (actually 64 layers) should be expected. This was confirmed by an excavation soon after the prediction was made. A sword of marked 50 refinings dated back to 77 AD was unearthed from a tomb in Xuzhou, Jiang-su...Examination showed that...it consisted of alternative low- and highcarbon layers, about 60 in total. The cost of the sword was also marked as worthy of 1500 coins, equivalent to grains enough for one man's two and half years' living. 2.5. Merovingian pattern-welded blade An early European technique dating from about the end of the second century resulted in the Merovingian pattern-welded blades. (Iron objects manufactured prior to this date are frequently too severely corroded for their surfaces to be evaluated metallurgically.) According to Smith [16], Merovingian blades were primarily manufactured on the Rhine although they were widespread through trade and war. The blades consisted of strips of pure iron and carbon steel (or iron strips that had been carburized on one side) hammered or forged together in a manner involving folding or twisting. The cutting edge consisted of the high-carbon-content steel, often inserted between plates of lowcarbon material. Upon grinding to shape after heat treating, patterns arising from the different layers become visible. In addition, it is quite likely that etching in fruit juice or sour beer was carried out to develop the patterns. An example is given in Fig. 3, from a blade discovered in a Viking grave in South Finland [17]. Thus, beginning in the period around 500 AD, pattern-welded daggers and swords were made, including Viking blades starting in about 600 AD. Smith [16] comments that the edges are martensitic between plates of iron in a manner similar to the Japanese sword. He also notes that it is ``difficult to justify the particular pattern used in the center of the swords on any but aesthetic grounds.'' Included in this group of materials is the early Japanese sword. 2.6. Japanese sword The Japanese sword has universally been regarded by eminent metallurgists as the ultimate expression of metallurgical accomplishment. For example, in 1962, Edgar C. Bain [19] wrote: The old swords of Japan are probably the best examples of the almost incredible pains taken to produce a superb implement. Also, in 1960, Sir Cyril Stanley Smith [16] wrote that in his opinion: The Japanese sword blade is the supreme metallurgical art. Certainly, world wide, the Japanese sword (Nipponto) is one of the most famous of all swords. The sword has always held a special place in the history and culture of Japan. Japanese legend, for example, tells us that Susanoo, brother of the Sun Goddess Amaterasu, slew an eight-headed dragon with a single stroke of a sword. Despite this, the blade was deemed inferior and Susanoo was given a magnificent sword from the dragon's tail. He gave the sword to his sister, the Sun Goddess, who then handed to her grandson the Imperial Regalia that included three items Ð the jewel, the mirror, and the sword. From a metallurgical viewpoint, the Japanese sword is of interest at several different levels. First, their manufacture involved the solid-state bonding of steels to themselves and to steels of radically different carbon contents. That is, in its simplest form, with the exception of the very earliest blades, the sword is a composite. An example of one of the composite designs is that of a high-carbon external sheath surroundFig. 3. Merovingian pattern-welded blade discovered in a Viking grave in the South of Finland. It was most likely made on the Rhine in the period 650 ± 700 AD. Helsinki University [18]. Courtesy of The Gun Report. 294 J. Wadsworth, D.R. Lesuer / Materials Characterization 45 (2000) 289±313
J. Wadsworth, D.R. Lesuer /Materials Characterization 45(2000) 289- 95 ng a low-carbon core. The procedure used to make such a blade is shown in Fig. 4. This composite pproach allowed the swordmaker to achieve the desirable properties of both hardness and ductility in a single weapon. Often, one of these properties is only developed at the expense of the other. Tool steel econd, to produce the high-carbon part of the blade, the steel was subjected to multiple folding perations; this has given rise to the erroneous con- 5 cept that the swords contain millions of discrete layers. Thi experimentally examined in next section on modern lmcs Third, the development of the complex and Hardened beautiful surface markings was a consequence a selective surface heat-treatment process(known as yaki-ire), achieved in part by covering the blade with different thicknesses of clay; this gave rise to an intriguing combination of transformation pro- ducts and surface patterns. This was not only visible and aesthetic example of the skill of the (ultrahigh carbon steel) swordmaker, but it was also evidence that the edge of the blade had been hardened Fig. 4. Procedure used by Japanese blacksmiths to make The blade must have a recognizable pattern, called laminated tools by solid-state bonding ultrahigh carbon the hamon, where the structure changes from the hard steel, known as kawagane, to soft iron; including a cross martensite at the edge, called akiba, to soft pearlite section of a blade coated with various thicknesses of clay to The hamon is perhaps the most important aesthetic control temperature, with minimal clay coat at the edge to allow proper hardening(adapted from Bain [19)[14] feature of a blade, and the first thing sword aficiona dos look for, as it is essentially the swordsmiths signature. Kapp et al. [21] cite the work of B w. stage of manufacture of the Japanese sword in which Robinsons classic book The Arts of the Japanese steel of about 1.8%C is used Sword [22], illustrating 53 different hamon, each with The high-carbon steel (called kawagane, but also its own name(from the descriptive"straight irreg sometimes called magane)used for blademaking wa lar"to the more suggestive "chrysanthemum and principally prepared using a reduction process meth- water") and the name of the smith or school with od, in which iron, sand, and charcoal produced tama- which it is identified. (See Fig. 5 for an illustration of hagane. A fixed amount of iron ore and charcoal was broad classes of hamon patterns mixed and heated in air to 1200C. resulting in The Japanese swords external sheath has a similar products of molten pig iron, slag, and unmelted carbon content (i.e, it can be hypereutectoid, about ultrahigh-carbon steel (UHCS). The iron ore came 0.8-1%C)to that of the Damascus sword (also from so-called black sands known as satetsu(iron hypereutectoid, but in the 1.5-1.8%C range). This oxide). The carbon was added to the black sand in a similarity of composition is especially so in the early smelter called a tatara. When the pig iron and sla were allowed to separate by pouring, the end product was lumps of UHCS containing about 1.7%C(tama i Tylecote[20] has pointed out a similarity in this hagane). This material was then repeatedly forged regard between the Japanese sword composite design and folded until the appropriate shape and reduction and the much later development of"shear steel"in in carbon content was achieved through decarburiza Western Europe following the Industrial Revolution. tion. (It should be noted that, depending upon the The similarity lies in the fact that comparatively few arbon additions, temperature, and time at tempera- pieces of steel of different carbon content wer welded together to make a composite, single-edged, ture, the result of such a repeated folding and forgin blade which was finally heat treated. In shear steel process could be low-carbon steel. mild steel outer strips encased a high-carbon center blade involved a number trip-a design currently available in handmade First, the tama-hagane was repeatedly forged and knives by contemporary artisans. In fact, Tylecot folded to produce the kawagane, which becomes the goes as far as to say, "There is essentially no sheath or jacket steel. Second, the shingane, or low- difference in principle between a scythe [made from carbon core steel, was formed, also by a repeated shear steel], and a Japanese sword. folding procedure. Third, the low-carbon core was
ing a low-carbon core. The procedure used to make such a blade is shown in Fig. 4. This composite approach allowed the swordmaker to achieve the desirable properties of both hardness and ductility in a single weapon.1 Often, one of these properties is only developed at the expense of the other. Second, to produce the high-carbon part of the blade, the steel was subjected to multiple folding operations; this has given rise to the erroneous concept that the swords contain millions of discrete layers. This point is experimentally examined in the next section on Modern LMCs. Third, the development of the complex and beautiful surface markings was a consequence of a selective surface heat-treatment process (known as yaki-ire), achieved in part by covering the blade with different thicknesses of clay; this gave rise to an intriguing combination of transformation products and surface patterns. This was not only a visible and aesthetic example of the skill of the swordmaker, but it was also evidence that the edge of the blade had been hardened. The blade must have a recognizable pattern, called the hamon, where the structure changes from the hard martensite at the edge, called yakiba, to soft pearlite. The hamon is perhaps the most important aesthetic feature of a blade, and the first thing sword aficionados look for, as it is essentially the swordsmith's signature. Kapp et al. [21] cite the work of B.W. Robinson's classic book The Arts of the Japanese Sword [22], illustrating 53 different hamon, each with its own name (from the descriptive ``straight irregular'' to the more suggestive ``chrysanthemum and water'') and the name of the smith or school with which it is identified. (See Fig. 5 for an illustration of broad classes of hamon patterns.) The Japanese sword's external sheath has a similar carbon content (i.e., it can be hypereutectoid, about 0.8 ± 1% C) to that of the Damascus sword (also hypereutectoid, but in the 1.5 ± 1.8% C range). This similarity of composition is especially so in the early stage of manufacture of the Japanese sword in which steel of about 1.8% C is used. The high-carbon steel (called kawagane, but also sometimes called uagane) used for blademaking was principally prepared using a reduction process method, in which iron, sand, and charcoal produced tamahagane. A fixed amount of iron ore and charcoal was mixed and heated in air to 1200°C, resulting in products of molten pig iron, slag, and unmelted ultrahigh-carbon steel (UHCS). The iron ore came from so-called black sands known as satetsu (iron oxide). The carbon was added to the black sand in a smelter called a tatara. When the pig iron and slag were allowed to separate by pouring, the end product was lumps of UHCS containing about 1.7% C (tamahagane). This material was then repeatedly forged and folded until the appropriate shape and reduction in carbon content was achieved through decarburization. (It should be noted that, depending upon the carbon additions, temperature, and time at temperature, the result of such a repeated folding and forging process could be low-carbon steel.) Forging the blade involved a number of steps. First, the tama-hagane was repeatedly forged and folded to produce the kawagane, which becomes the sheath or jacket steel. Second, the shingane, or lowcarbon core steel, was formed, also by a repeated folding procedure. Third, the low-carbon core was Fig. 4. Procedure used by Japanese blacksmiths to make laminated tools by solid-state bonding ultrahigh carbon steel, known as kawagane, to soft iron; including a cross section of a blade coated with various thicknesses of clay to control temperature, with minimal clay coat at the edge to allow proper hardening (adapted from Bain [19]) [14]. 1 Tylecote [20] has pointed out a similarity in this regard between the Japanese sword composite design and the much later development of ``shear steel'' in Western Europe following the Industrial Revolution. The similarity lies in the fact that comparatively few pieces of steel of different carbon content were welded together to make a composite, single-edged, blade which was finally heat treated. In shear steel, mild steel outer strips encased a high-carbon center stripÐ a design currently available in handmade knives by contemporary artisans. In fact, Tylecote goes as far as to say, ``There is essentially no difference in principle between a scythe [made from shear steel], and a Japanese sword.'' J. Wadsworth, D.R. Lesuer / Materials Characterization 45 (2000) 289±313 295
J. Wadsworth, D.R. Lesuer /Materials Characterization 45(2000) 289-313 Kanehira and the dojigiri by Yasutsuna made over 900 years ago. Photographs of the o-kanehira are shown in Fig.6. 2.7. Medieval damascened knive Piaskowski [24 has reviewed in detail pat sugita komidare welded damascened knives found in Poland dating from the Sth to 12th century. Their real origin is not clear. but Piaskowski believes that some of the knives may be examples of the work of early medieval Polish smiths. The knives all consisted of three regions: a steel cutting edge, adjacent to a complex, central, patterned layered region of carbur ized iron and iron, and a backing layer of iron or itatstrra steel. This is in contrast to other related techniques in which a steel central layer is sandwiched between pattemed layers of iron and steel. Evidence for heat Fig. 5. Types of Hamon(after Sato [23]). treatment to produce martensitic structures, and tempered structures, is described. Techniques invol ving from 3 to 17 initial layers are presented. In all inserted, by one of several methods, inside the high- cases, hammer welding and plastic deformation took carbon jacket steel. In the fourth step, the composite place during manufacture. The carbon contents of the was drawn out to the approximate length of the blade, composition, 0.5%C, is given for one steel in one of layers are poorly evaluated, however, and only one and the fifth step shaped the final blade. The end product is a kawagane steel with the knives excellent mechanical properties because the carbon content is both relatively low(about 0.6-1.0% C) and the carbides are distributed uniformly fine-grained matrix. As discussed later, no visible pattern-welded structure is obtained from this scale of folding, not only because the individual, 0.2-m layers are unresolvable to the naked eye, but also because the carbon content of each layer is iden- cal(carbon atoms diffuse a distance of 1. 4 um in 30 s at 1000C). An observable pattern-welded tructure. however often emerges from the final several folds Thus the method of manufacture and the of the surface patterns on the Japanese sword are quite different from those of Damascus swords Specifically, the principal surface pattem on a Japa nese sword is created as a result of the variou transformation products following heat treatment the blade. There are also surface pattems that consist of a gross texture from the final stages of piling folding, and forging. The earliest reference to sur- face patterns on a Japanese sword, referenced by Smith [161, is to 1065 AD. There are subtleties to these patterns that illustrate several intriguing me- There are an estimated 1 million swords now own to exist; 117 have been designated as Japa- 6.“O- Kanehira” achi by Ka nese national treasures. The most famous and rev- 9.2 cm. Mid-Heian period, approximately 1000 AD ered of the swords are identified with a name o National Museum. Signed Bizen no kuni Kanehira . Tokyo hira of Bizen providence").(After Sato [23])
inserted, by one of several methods, inside the highcarbon jacket steel. In the fourth step, the composite was drawn out to the approximate length of the blade; and the fifth step shaped the final blade. The end product is a kawagane steel with excellent mechanical properties because the carbon content is both relatively low (about 0.6 ± 1.0% C), and the carbides are distributed uniformly in a fine-grained matrix. As discussed later, no visible pattern-welded structure is obtained from this scale of folding, not only because the individual, 0.2-mm layers are unresolvable to the naked eye, but also because the carbon content of each layer is identical (carbon atoms diffuse a distance of 1.4 mm in 30 s at 1000°C). An observable pattern-welded structure, however, often emerges from the final several folds. Thus, the method of manufacture and the origins of the surface patterns on the Japanese sword are quite different from those of Damascus swords. Specifically, the principal surface pattern on a Japanese sword is created as a result of the various transformation products following heat treatment of the blade. There are also surface patterns that consist of a gross texture from the final stages of piling, folding, and forging. The earliest reference to surface patterns on a Japanese sword, referenced by Smith [16], is to 1065 AD. There are subtleties to these patterns that illustrate several intriguing metallurgical issues. There are an estimated 1 million swords now known to exist; 117 have been designated as Japanese national treasures. The most famous and revered of the swords are identified with a name or meito. Two such examples are the o-kanehira by Kanehira and the dojigiri by Yasutsuna made over 900 years ago. Photographs of the o-kanehira are shown in Fig. 6. 2.7. Medieval damascened knives Piaskowski [24] has reviewed in detail patternwelded damascened knives found in Poland dating from the 8th to 12th century. Their real origin is not clear, but Piaskowski believes that some of the knives may be examples of the work of early medieval Polish smiths. The knives all consisted of three regions: a steel cutting edge, adjacent to a complex, central, patterned layered region of carburized iron and iron, and a backing layer of iron or steel. This is in contrast to other related techniques in which a steel central layer is sandwiched between patterned layers of iron and steel. Evidence for heat treatment to produce martensitic structures, and even tempered structures, is described. Techniques involving from 3 to 17 initial layers are presented. In all cases, hammer welding and plastic deformation took place during manufacture. The carbon contents of the layers are poorly evaluated, however, and only one composition, 0.5% C, is given for one steel in one of the knives. Fig. 5. Types of Hamon (after Sato [23]). Fig. 6. ``O-Kanehira.'' Tachi by Kanehira. Steel. Nagasa 89.2 cm. Mid-Heian period, approximately 1000 AD. Tokyo National Museum. Signed Bizen no kuni Kanehira (``Kanehira of Bizen providence''). (After Sato [23]). 296 J. Wadsworth, D.R. Lesuer / Materials Characterization 45 (2000) 289±313
J. Wadsworth, D.R. Lesuer /Materials Characterization 45(2000) 289-313 2.8 Thailand axes and tools prepared by the smith Although less extensively developed than the by cementation, ie burying a thin strip of Indonesian metal kisses (next section). there are low-carbon iron about 1 mm thick in a charcoal fire interesting artifacts originating in Thailand. In a study with limited air access to give a reducing atmo- of such objects [25], four iron artifacts were exca sphere, rich in carbon monoxide. vated in Northeast Thailand and three were dated to 2.9 Indonesian kris the Late Iron Age(e.g, 300-400 AD). The fourth artifact, dated by association with the other three, involved welding an ultrahigh carbon content steel The Indonesians of Java and other Malayan onto a wrought iron core to form a high quality axe Islands made a number of knives known as kisses blade. In concluding that the artifacts date to the late Indonesian kisses usually are forged to have re. Iron Age, the authors note that iron was produced in petitive curves along their length. There are in fact Thailand from 500 BC or earlier [26] two classes. One is that containing long blades that were used as sabers, with a slashing motion. The The crescent-bladed axe was found a short dis- other class is that of short stabbing blades. All are tance from the Late Iron Age mound site of No double edged. It is believed that the undulating Phrik which is located near Ban Hua Na village, Phu curves might make for more efficient thrusts and Luong sub-district, Loei Province. Although the carbon content was estimated as 1. 8% and the steel recoveries of the weapon. Other theories are based as concluded to be hypereutectoid, this finding was on religion. For example, under the Hindu influ- based on the assumption that the grain boundary ence, the oriental snake gad, Naga, may be repre- material was all massive carbide. It is not clear to sented in the serpentine curves. Unlike Japanese the present authors that this is the case at all, and the swords, the kisses bear no names, dates, or places steel may in fact be hypoeutectoid. Nonetheless, the of manufacture. although there are over 30 different artifact is an example of an axe that has a relatively types that can be associated with different regions massive cutting head formed by welding a layer The blades are usually laminated; in fact, the name carbon steel onto a wrought iron core for the most popular ones is pamur, a Malay word In the microstructure of a second artifact, an iron for combination or mixture. Smith [16 included socketted chisel, a"laminated semi-circular pattem is excellent examples of surface patterns in his book readily visible. "Hogan and Rutnin[25] proposed that on historical metallurgy; and the detailed manufac the manufacturing process was"a simple procedure." ture of a relatively modern kris also has been described by the famous metallurgist Walter Rosen- The starting material was piled wrought iron, made hain [27] by hammering sponge from the smelting furnace A typical Indonesian kris is shown in Fig. 7(a), to thin sheets, then folding and re-folding while and an interesting example of a specialized execu- ot and hammering them together to form a sha tioners kris, with a straight blade is shown in Fig uired for sale to blacksmiths. When re-heated 7(b). In this case, as with others, one of the layers is he forge the surface of each sheet may be either meteoric iron containing Ni. According to Smith xidised or reduced. so that the carbon content kisses were made from about 1379 AD onward different in surface and centre of the sheets. when these sheets are welded together the laminated in Indonesia under Hindu influences. From the description by Rosenhain, the modern kris was made by solid-state ng of a tool steel Laminations were also evident in an iron sock- high-carbon steel' such as is commonly used for etted spearhead which tools and cutlery, "to quote Rosenhain) to welded yers of wrought iron. In addition, according to appears to have been made in a sandwich construction Rosenhain [27] carbon strip of the iron, lying parallel to the top The imperfection of the [solid-state) welds between surface.. This was sandwiched between strips of the wrought iron [layers]also play an important part soft, low-carbon iron and the sandwich forge welded in the formation of the damask pattem. Thu 2.10. Halberds of the spear can always be resharpened to give A halberd is a weapon that is both a spear and a elatively hard, sharp point, while the soft outer battle ax that was used in warfare in the 15th to 16th yers are more easily ground to remove the bulk of century. According to Meier [28], statements regard
2.8. Thailand axes and tools Although less extensively developed than the Indonesian metal krisses (next section), there are interesting artifacts originating in Thailand. In a study of such objects [25], four iron artifacts were excavated in Northeast Thailand and three were dated to the Late Iron Age (e.g., 300 ± 400 AD). The fourth artifact, dated by association with the other three, involved welding an ultrahigh carbon content steel onto a wrought iron core to form a high quality axe blade. In concluding that the artifacts date to the Late Iron Age, the authors note that iron was produced in Thailand from 500 BC or earlier [26]. The crescent-bladed axe was found a short distance from the Late Iron Age mound site of Non Phrik which is located near Ban Hua Na village, Phu Luong sub-district, Loei Province. Although the carbon content was estimated as 1.8% and the steel was concluded to be hypereutectoid, this finding was based on the assumption that the grain boundary material was all massive carbide. It is not clear to the present authors that this is the case at all, and the steel may in fact be hypoeutectoid. Nonetheless, the artifact is an example of an axe that has a relatively massive cutting head formed by welding a layer of carbon steel onto a wrought iron core. In the microstructure of a second artifact, an iron socketted chisel, a ``laminated semi-circular pattern is readily visible.'' Hogan and Rutnin [25] proposed that the manufacturing process was ``a simple procedure.'' The starting material was piled wrought iron, made by hammering sponge from the smelting furnace into thin sheets, then folding and re-folding while hot and hammering them together to form a shape required for sale to blacksmiths. When re-heated in the forge the surface of each sheet may be either oxidised or reduced, so that the carbon content is different in surface and centre of the sheets. When these sheets are welded together the laminated appearance results. Laminations were also evident in an iron socketted spearhead which: ...appears to have been made in a sandwich construction, commencing with a relatively highcarbon strip of the iron, lying parallel to the top surface.... This was sandwiched between strips of soft, low-carbon iron and the sandwich forge welded and shaped... Thus, the: ...use of a higher carbon core ensures that the point of the spear can always be resharpened to give a relatively hard, sharp point, while the soft outer layers are more easily ground to remove the bulk of the metal required for sharpening. The higher carbon central strip would have been prepared by the smith by cementation, ie [sic] by burying a thin strip of low-carbon iron about 1 mm thick in a charcoal fire with limited air access to give a reducing atmosphere, rich in carbon monoxide. 2.9. Indonesian kris The Indonesians of Java and other Malayan Islands made a number of knives known as krisses. Indonesian krisses usually are forged to have repetitive curves along their length. There are in fact two classes. One is that containing long blades that were used as sabers, with a slashing motion. The other class is that of short stabbing blades. All are double edged. It is believed that the undulating curves might make for more efficient thrusts and recoveries of the weapon. Other theories are based on religion. For example, under the Hindu influence, the oriental snake gad, Naga, may be represented in the serpentine curves. Unlike Japanese swords, the krisses bear no names, dates, or places of manufacture, although there are over 30 different types that can be associated with different regions. The blades are usually laminated; in fact, the name for the most popular ones is pamur, a Malay word for combination or mixture. Smith [16] included excellent examples of surface patterns in his book on historical metallurgy; and the detailed manufacture of a relatively modern kris also has been described by the famous metallurgist Walter Rosenhain [27]. A typical Indonesian kris is shown in Fig. 7(a), and an interesting example of a specialized executioner's kris, with a straight blade, is shown in Fig. 7(b). In this case, as with others, one of the layers is meteoric iron containing Ni. According to Smith, krisses were made from about 1379 AD onward in Indonesia under Hindu influences. From the description by Rosenhain, the modern kris was made by solid-state welding of a tool steel (``a `high-carbon steel' such as is commonly used for tools and cutlery,'' to quote Rosenhain) to welded layers of wrought iron. In addition, according to Rosenhain [27]: The imperfection of the [solid-state] welds between the wrought iron [layers] also play an important part in the formation of the damask pattern. 2.10. Halberds A halberd is a weapon that is both a spear and a battle ax that was used in warfare in the 15th to 16th century. According to Meier [28], statements regardJ. Wadsworth, D.R. Lesuer / Materials Characterization 45 (2000) 289±313 297
298 J. Wadsworth, D.R. Lesuer /Materials Characterization 45(2000)289-313 2. 12. Pattern-welded blades of more recent origin A more recent example, i.e., a welded Da- mascus steel dagger, made about 150 years ago, is shown in Fig. 9(a) and illustrates the multi layered composite nature of the weapon. A low Fig. 9(b), illustrates the unique surface markings These markings were created by forging many alternating layers of high-carbon steel and iron plates, followed by a complicated twisting and forging step. The micrograph in Fig. 9(c), taken Fig. 7.(a) Typical Indonesian kris. (b) Indonesian at high magnification, readily identifies the layers executioners kris is a composite of meteoric nickel as consisting of alternating high-carbon spheroi- iron and plain carbon steel. dized steel with graphite stringers and low-carbon Iron layers ing the distribution and origin of herd 2.13. European welded damascus rarely met with until the 15th century. Howey poems, songs, and chronicles provide written After its origination in the east in the 16th sions to the halberd from the 13th century. Also because of the troop and arms rolls records of Zurich from the time of the old zurich war. some firm was pursued in Et assertions can, in fact, be made with regard to the the 18th century. a typical welded Damascus barrel distribution of the halberd. For example, in 1442 AD, would have consisted of seven layers( four of low- of the 1591 short-arms counted, 856 were described carbon material or pure iron, and three of steel) These would have been forged together, and then The halberd of the 14th and 15th centuries was the resulting strip coiled to the desired barrel shape conceived principally as a cutting weapon; the hal- In more complex routes, the welded strip was The construction of the blade of the halberd was and reforged repeatedly to provide intricate final improved in the course of the 16th century, thanks to pattems. Some swords were made in this manner better raw materials and new forging-techniques (for example, the welded Chinese sword shown in Until about 1500. the blade of the halberd was generally composed of four pieces. In the 16th century the cut and thrust function of the halberd acquired importance. Henceforth the point of the weapon lay in the axis of the shaft. The centrally mounted halberd blade. which was attached to the staff with shaft-straps, comprises 10 pieces, as for example on a"Sempach-halberd"of the 17th cen- ury. It is of interest to note that the nature of the construction parallels that of other laminated weap- ons in that a high-carbon steel cutting edge surrounded by a low-carbon sheath. 211. Chinese blades Chinese steel of“ hundred refinings”from100 AD was mentioned earlier. Other relatively recent amples of we Chinese blacksmiths. In addition to perhaps being Fig. 8. A 17th century Chinese sword of the patterm-v from the originators of the Japanese sword in the difference in etching of the two dissimilar steels from century AD, they also made their own which the sword was manufactured, involving a sold-state welded blades as shown in a 17th century bonding process. Courtesy of the Metropolitan Museum of in Fig. 8 Art, New York [17]
ing the distribution and origin of the halberd are rarely met with until the 15th century. However, poems, songs, and chronicles provide written allusions to the halberd from the 13th century. Also, because of the troop and arms-rolls records of ZuÈrich from the time of the Old ZuÈrich War, some firm assertions can, in fact, be made with regard to the distribution of the halberd. For example, in 1442 AD, of the 1591 short-arms counted, 856 were described as halberds. The halberd of the 14th and 15th centuries was conceived principally as a cutting weapon; the halberd was fixed on the ash shaft with sockets (straps). The construction of the blade of the halberd was improved in the course of the 16th century, thanks to better raw materials and new forging-techniques. Until about 1500, the blade of the halberd was generally composed of four pieces. In the 16th century the cut and thrust function of the halberd acquired importance. Henceforth the point of the weapon lay in the axis of the shaft. The centrally mounted halberd blade, which was attached to the staff with shaft-straps, comprises 10 pieces, as for example on a ``Sempach-halberd'' of the 17th century. It is of interest to note that the nature of the construction parallels that of other laminated weapons in that a high-carbon steel cutting edge is surrounded by a low-carbon sheath. 2.11. Chinese blades Chinese steel of ``hundred refinings'' from 100 AD was mentioned earlier. Other relatively recent examples of welded knives and swords exist from Chinese blacksmiths. In addition to perhaps being the originators of the Japanese sword in the 5th century AD, they also made their own patternwelded blades as shown in a 17th century example in Fig. 8. 2.12. Pattern-welded blades of more recent origin A more recent example, i.e., a welded Damascus steel dagger, made about 150 years ago, is shown in Fig. 9(a) and illustrates the multilayered composite nature of the weapon. A low magnification picture of the dagger, shown in Fig. 9(b), illustrates the unique surface markings. These markings were created by forging many alternating layers of high-carbon steel and iron plates, followed by a complicated twisting and forging step. The micrograph in Fig. 9(c), taken at high magnification, readily identifies the layers as consisting of alternating high-carbon spheroidized steel with graphite stringers and low-carbon iron layers. 2.13. European welded Damascus After its origination in the East in the 16th century, a technique of welding steel to iron, in strips, to give strength and texture to guns and swords, was pursued in Europe from the end of the 18th century. A typical welded Damascus barrel would have consisted of seven layers (four of lowcarbon material or pure iron, and three of steel). These would have been forged together, and then the resulting strip coiled to the desired barrel shape. In more complex routes, the welded strip was twisted, and then several such strips were rewelded and reforged repeatedly to provide intricate final patterns. Some swords were made in this manner (for example, the welded Chinese sword shown in Fig. 7. (a) Typical Indonesian kris. (b) Indonesian executioner's kris is a composite of meteoric nickel ± iron and plain carbon steel. Fig. 8. A 17th century Chinese sword of the pattern-welded variety. The pattern shown on the macrograph arises from the difference in etching of the two dissimilar steels from which the sword was manufactured, involving a sold-state bonding process. Courtesy of the Metropolitan Museum of Art, New York [17]. 298 J. Wadsworth, D.R. Lesuer / Materials Characterization 45 (2000) 289±313