A treatise on the making of iron in a bloomery and on the quality of the armour made from such iron
Sylvia Leever BSc
Literature survey Supervision by B.J. Thijsse and J. Dik
Delft, Materials Science and Engineering October 2003
The exact time when people started to produce iron is not known. Also the exact way people found out how iron could be made is unknown. But it is known that in order to produce iron a new technology had to be developed. This development has been a continuous one, and in medieval times this development has led to a bloomery. A bloomery is a furnace and forge in which wrought iron in the form of blooms is made directly from the ore, or (more rarely) from cast iron. Crushed, washed and/or roasted chunks of ore of definite size were put in the furnace that was heated by burning pieces of charcoal. The quality of charcoal influences the maximum temperature that can be obtained in the furnace. The ratio charcoal to ore is very important to the amount of carbon in the bloom. The amount of carbon in the iron produced in a bloomery seldom exceeds 0.8% and is commonly in the order of a few points of a percent. In modern blast furnaces the raw iron is saturated with carbon (4.5%), but before the iron goes to the secondary steelmaking process, process, all (up to a few hundreds of a percent) the carbon will be removed, and the desired amount of carbon will be added later. This fact expresses a big difference between medieval and modern iron production. Not only is the scale of production in modern times much larger than in medieval times, and has the modern iron making process become more cost efficient, also the technical knowledge of steelmaking is much more advanced. In medieval times the practical knowledge was in general better developed. This is also seen in the skills of the smiths and armourers when some pieces of armour are examined. The precision with which some armourers and etchers and gilders worked, speaks of a very great talent. The quality of the iron and steel used for making armour is, compared to modern steel equivalents, not exquisite. But the quality was good enough to serve the purpose of protecting the wearer of the armour against most attacks. The fact that the armour business has flourished for over 300 years also tells us that wearing armour was thought to be useful and effective. When in the sixteenth century guns and cannons were used more extensively at battlefields, against which armour would not give enough protection, and the cost of equipping all soldiers with armour was becoming too large, the production of armour faltered and armour became more and more a fashion thing, instead of a functional item. This development can also be seen when the microstructures of various pieces of armour are examined. In specimens with late fifteenth century origin, when the production of armour had its peak, the microstructures show mostly hardened steels of relatively pure quality, to give the best protection. Microstructures of late sixteenth century armours show more and more a soft iron structure. In general the microstructures of pieces of armour made in different times and at various locations are as unlike as the people who made them.
i
After busying myself for more than a year with re-enactment of medieval life (i.e. portraying someone from a certain period in history and everything that comes with that (authentically made clothes, proper equipment, learning to fight with sword, pike, spear and bow and arrow, learning various crafts)), the idea came to me to combine my interests in history with my interests in materials science. At first I didn’t really know what to do, but soon enough I wanted to find out what the quality of medieval armour was, and how skilled the armourers in the fifteenth century were. Did they really know what their business was on a more scientific field or was it just the years of experience that made them masters in the working of iron and steel? In January 2003, my friend, Joram van Essen, pointed out to me that at the Wallace Collection in London a workshop on analysing historic steel and iron was going to be held on 30 March. I signed up for that and this workshop encouraged me to really go on with the history of metallurgy. After the workshop I had a talk with the organisers of the workshop, dr. Alan Williams and David Edge (both working at the Wallace Collection). In this conversation it became clear that still a lot of research on the topic of historical metallurgy had to be done. They said that what is really needed is that someone will make iron in an authentic medieval way, so that the produced iron can be researched, and the quality of (armour)plates made from it can be tested. To me it sounded very interesting, and I started having the idea of actually reproducing the iron making in a bloomery. This idea I have dropped by now, because it is too complex for someone with as little experience in this field as I, and because at least three people are needed fulltime for a couple of years to pull this off. So I settled instead with writing the theory about medieval iron production, and I added a part about the quality of medieval armour, to get back to my original idea. At this point I would like to thank the following people: dr. Alan Williams and David Edge for their time for the conversation in London, and Alan Williams for some advise on the matters through email and for supplying me with a piece of Japanese bloomiron. I’d like to thank Peter King for sending me a part of his reference list of his PhD thesis on medieval charcoal, which became very useful. Thanks to professor Rob Boom for some additional information on modern steelmaking. My supervisors, dr. Joris Dik and professor Barend Thijsse, were very helpful with their support and enthusiasm for my ideas. And lastly I’d like to thank Joram van Essen, for making me enthusiastic about medieval steelwork and helping me out on various things.
ii
Summary...................... Summary ............................................. .............................................. .............................................. .............................................. ..............................i .......i Preface...................................... Preface............................................................. .............................................. ............................................... ....................................... ............... ii Introduction...................................................... Introduction.............................. ................................................ ................................................ ......................................1 ..............1 Methodology..................... Methodology ............................................ .............................................. .............................................. ...........................................2 ....................2 Current research on medieval iron production.................................................. production...........................................................2 .........2 1. History of iron........................................................... iron.................................................................................. .............................................. ..........................4 ...4 2. Iron and how to extract it from the ore ............................................... ......................................................................5 .......................5 3. The Bloomery Process ............................................ .................................................................... ................................................ ...........................6 ...6 4. Construction of Furnaces ............................................. ..................................................................... ..............................................8 ......................8 5. Preparations necessary before smelting........................ smelting ................................................. ...........................................10 ..................10 5.1. Assaying the ore.................................................. ore.......................................................................... ..............................................10 ......................10 5.2. Making charcoal............................................ charcoal................................................................... ............................................... .............................11 .....11 5.3. Preparing the ore ............................................... ...................................................................... ............................................... .........................13 .13 6. Costs of running a bloomery....................... bloomery ............................................... ................................................ .....................................16 .............16 7. Modern steelmaking process................................... process........................................................... ............................................... .........................19 ..19 8. The quality of medieval armour.................................................................... armour...............................................................................22 ...........22 8.1. Protection by armour............................................ armour.................................................................... .............................................24 .....................24 8.2. Microstructure of some pieces of armour .............................................. .........................................................26 ...........26 8.3. The cost of armour ............................................. .................................................................... ...............................................31 ........................31 9. Discussion..................... Discussion ............................................ .............................................. .............................................. .............................................33 ......................33 10. Conclusion .............................................. ..................................................................... .............................................. ........................................37 .................37 References................................................ References....................................................................... .............................................. .............................................38 ......................38 Bibliography ............................................ .................................................................. ............................................ ............................................ ........................40 ..40
On the title page: picture from fighting manuscript by Hans Talhoffer, 1457
iii
The subject of this thesis is actually twofold. I will start by describing the process of making iron in a bloomery in late medieval times. This will be followed by a detailed description of the quality of some products made from this ‘bloomery-iron’. To start with the elemental process of making iron in a bloomery, first a modern interpretation of the process of reducing iron ore will be given and this will be followed by a description of the bloomery process. The medieval people probably did not know what exactly was going on in their furnaces, but they had enough practical knowledge to make good quality iron and steel. In the first century AD Plinius has written Natural History in which he describes amongst other thing the fundamentals of producing metals from their ores and the properties of various metals and alloys. His conceptions have been used throughout the Middle Ages and have not changed much. In the sixteenth century a couple of books were written which enhanced on that subject. The famous works from that century Pirotechnia and De Re Metallica by the Italian Biringuccio and the German Agricola respectively, have contributed a great deal to our understanding of the medieval world of metallurgy. They describe in great detail what various processes of making iron or other metals from their ores should be. The methods of building a furnace and making charcoal and the process of reducing the ore explained by Biringuccio and Agricola will be discussed and compared with each other and with modern equivalent methods. Another sixteenth century book, Lazarus Ercker’s Treatise on Ores and Assaying , gives additional information on ores, and this will also be compared with what Biringuccio and Agricola have written. With the help of an early fifteenth century account roll of the ironmaster of a bloomery an overview of the costs of running a bloomery will be given, as well as an overview of the people employed at the bloomery. This will be compared with the situation in a modern steelmaking factory. After this a short description of how modern steelmaking takes place will be given, followed by a comparison between the modern process and the late medieval process of making iron and/or steel. At this point one may become curious and wonder whether the iron or steel obtained in the bloomery process can ever contribute to a solid and impenetrable armour. What is the quality of medieval iron and steel? It appears that the quality of the final product depends on a lot of things, and not in the last place on the skills of the smiths and armourers. A description of the qualities of some pieces of armour produced at various locations will be given. And again, this will be compared with the properties of a modern steel equivalent. In a discussion where I will give my own interpretation on matters I will also go back on my original plan of reproducing the process of making iron in a bloomery in such way that I will give a short list of parameters one should pay attention to when reproducing the process. Finally, after the discussion, conclusions will be drawn: How effective is the iron production process in medieval times with regard to yield effectiveness and cost effectiveness? Is medieval steel, like most people tend to believe, always inferior to modern steel? Or are there exceptions and has a lot of practical knowledge been lost in the past few centuries, so we cannot know how certain pieces of high-quality armour were made? I hope to answer at least a few of these questions in this thesis.
1
Methodology In this thesis I used a few books of 16th century origin. When handling old sources it is necessary to be careful, because a modern reader will interpret what is written in the books differently than a contemporary reader. This is because the view of the world changes over the course of time; for example, the concept of ‘a day of hard work’ would today be a working day from 08:00 till 20:00 in the office, with lots of meetings and deadlines to meet. In medieval times this concept would more likely be a day of hard manual labour from dawn until dusk. There were regulations within some working organisations that work could only be done during daylight (as a quality control), so most medieval people would not work after the sun has set. And of course this interpretation will also vary between contemporaries from different social backgrounds. This illustrates the care that should be taken when using old sources. Often it is also the case that the original source is lost or not available to the reader. The reader has to settle with a copy (with most likely a few copying-mistakes) or with a translation of the original. I used four books (three originally from the 16 th century and one from the 1 st century AD) in translation. Most of them were translated from the original in the 20 th century (one in the 16 th century). But even then meanings can get lost in translation. The Pirotechnia and De Re Metallica describe in great detail how one should accomplish certain metallurgical things. But most of the time the authors do not go all the way in their details. Some details can be so vague that they are open to more than one interpretation, or some essential details may be omitted deliberately. This can be because the authors did not know themselves how things were supposed to be, so they kept their descriptions vague. If on the other hand the authors deliberately did not write down essential things, which were known to them, might find its reason in the fact that in medieval times there were a lot of trade secrets. A master armourer, for example, would never reveal exactly how he accomplished the sometimes difficult tasks of making armour. He might tell you bits and pieces, but never enough to reproduce his work. This is logical, for if he did tell everyone his trade secrets, he would soon be out of a job, or at least have a dozen more competitors. Another explanation for the omissions is a more trivial one. If the author was convinced that certain facts were obvious to his audience, he would not have to write them down. Similarly, if I told someone to tie his shoelaces, I would not have to explain how he has to do it, for it is common knowledge. But if I wrote it in a letter to someone, and 500 years later that letter was found, in a civilisation where everyone wears velcro fastened shoes, to them it would be a mystery what shoelaces are and above all else, how they should be tied. Or if I hadn’t specified by saying ‘tie your shoelaces’ but had instead written ‘put your shoes on decently’, the velcro shoe civilisation would put their velcro shoes on properly, but they would not come close to the original intention of the text, because they lack the knowledge that was common 500 years ago. I tried, by mixing quoted parts of the translated texts with my own comments, to give the reader of this thesis a view of how I interpret things in such way that the reader can also give his or her own interpretation to it.
Current research on medieval iron production In order to improve certain methods, it is important that you learn from the history of these methods and from the history of the predecessors of these methods. Nobody wants to invent the wheel again, so before something is researched, it is necessary to look for other researches on the same field. The field of historical 2
metallurgy is very broad, and late medieval iron production and everything that goes with that has been researched extensively. The interest in medieval times is not something of the last few decades, but it goes back a long time. In the beginning of the renaissance, people thought that in the Middle Ages nothing significant has been developed, and they weren’t very interested in anything medieval. A bit later in time though, people started to wonder whether the assumption that nothing happened in the Middle Ages was true. As for iron production, the basics of the blast furnace were developed in late medieval times, and the knowledge of making armour has also its origin in the Middle Ages. At the end of the nineteenth century Lapsley translated a medieval account roll of an early fifteenth century bloomery. In the twentieth century, when archaeology gained more interest, people also found remnants of bloomeries and they started to wonder how they were operated. In the sixties and seventies of the twentieth century curiosity rose so much that people wanted to reproduce the process of making iron in a bloomery. Nowadays this wish has not diminished, although some people were able to do some experiments with the reproduction of a bloomery process. Because the bloomery process is actually a rather complex thing to research thoroughly, often people research only an aspect of making iron in a bloomery. Also a lot of thought has been given on (the reconstruction of) the production of medieval armours and weapons. Recent and current researches on these various subjects are: •
•
•
The Iron Trade 1500-1815: the Charcoal Iron Industry and its Transition to Coke, by Peter King, PhD thesis at Wolverhampton University. Arms and Armour with a strong ferrous bias, by David Starley, Royal
Armouries Museum, Leeds, England: Investigation of Armour; (Including plate, brigandine jack and mail); weapons (current major research project is on the metallurgy of arrows but also early guns (wrought iron and cast iron), spearheads and swords). Also maintaining earlier interest in Architectural iron and the use of analytical techniques for distinguishing technological processes. Wrought Iron, by Steve Suff, Real Wrought Iron Group: The group is a means to co-ordinate the activities and provide a forum of those involved in studying, researching and working in the field of genuine wrought iron (the material of antiquity & the industrial revolution).
Research not really concerning medieval times, but earlier iron production: •
Archaeometric and Archaeological Analysis of Early Iron Age Smelting (and smithing) Remains from Tell Hammeh (az-Zarqa), Jordan, by Xander
Veldhuijzen, Institute of Archaeology, University College London: research on ironslag, tuyères, charcoal, furnace remnants etc. of a predominantly iron smelting site from 1000-800 BC in Jordan In the bibliography at the end of this thesis a few finished works with relevance to this subject are listed.
3
Long ago, at a time when no one had ever heard of iron, let alone of steel, it came to pass that a copper smelter put a piece of haematite iron ore, reddish and not so unlike roasted copper ore, into his copper smelting furnace. He did not find any copper in the slag on the bottom of the furnace after smelting, which was strange. He did find a couple of small pieces of porous iron between the charcoal. This gave him the brilliant idea that if he kept these pieces of iron hot enough, say between 1000ºC and 1100ºC, he would find them to be malleable. This is very difficult to accept, but something like this must have happened if the development of iron smelting is to be accounted for. The iron obtained from such a primitive smelting process is, apart from the slag inclusions, remarkably pure [1]. Tracing the developments of smelting in the Early Iron Age to the more advanced smelting techniques of the Romans is beyond the scope of this thesis. It should suffice to say that because of the larger demand for iron from the Roman Legions the iron smelting techniques had to become more advanced and yield more iron per quantity of ore. When the Roman Empire fell, the smelting techniques for large amounts of iron ore disappeared with the Roman armies. Because there was no longer a military need for the iron, the large-scale operations of the Romans were replaced by the older techniques of the Early Iron Age [1]. Another factor was that the local smiths and smelters were not familiar with the Roman smelting techniques and so, when the Romans left, they re-introduced the old techniques. These techniques were maintained during the long period of the Dark Age and the Early Medieval Period. During this period the smelting and smithing techniques did not change significantly. When the use of waterpower was introduced at the beginning of the 15 th century, the scale of the smelting operations did change.
4
There are three main types of iron ore: carbonate ores, haematite ores and limonite ores. Most common are the carbonate ores which occur in two forms, as nodules and as sedimentary deposits (sometimes called siderite of spathic iron ore). In both forms, the iron is present as ferrous carbonate (FeCO3). The haematite ores (Fe2O3) contain very little of the phosphorus which gave rise to difficulties in the early days of steel making, and were therefore much sought after in the Industrial Revolution[1]. The limonites are hydrated iron oxides that consist mainly of the crystalline oxide goethite (Fe 2O3.H2O or HFeO2) with various amounts of absorbed water. When the ores are roasted before smelting the carbon dioxide can be expelled in the case of carbonate ores, and the water content will be expelled in the case of haematite ores. The roasting of the ores is beneficial because it opens up the ore and makes it more porous, and therefore it can be reduced more easily. A distinct group of limonites, the bog iron ore, also contains a good deal of water, because it is formed by precipitation of groundwater flowing into wetlands. Because bog iron ore is already very porous, it is not necessary to roast it before smelting. The water will evolve in the higher levels of the hearth. A second benefit of roasting is that it makes extremely hard ores easier to break down to the desired size. The size of the ore is important, because if the lumps are too big, the carbon monoxide takes too long to penetrate and reduce the lumps but if, on the other hand, the lumps are too small, they will need a greater draught, as they increase the resistance of the furnace. The size of the lumps of ore also depends on the size of the furnace. The optimum size lumps can be the size of a walnut or something in the order of 6x3x3cm. In comparison to the non-ferrous metals, iron is difficult to obtain from the ore. The main reason for this is, unlike what would logically seem to be the case, not the fact that pure iron has a much higher melting point than any of the common nonferrous metals. Iron occurs in ores as an oxide, just as tin does, but in contrast to tin, iron does not separate easily from the oxygen. This reaction requires much stronger reducing conditions. It is possible to reduce iron from pure iron oxide at about 800ºC, well below its melting point of 1540ºC. Because the iron ore also contains, aside from the oxides, some unwanted minerals and it is not possible to remove these by washing, they will have to be separated by ‘slagging’ in the smelting process. The slag thus obtained consists mostly of a compound formed from iron oxide and silica. In order to separate the slag from the iron, liquation is required. To liquefy the slag, a furnace operating temperature of about 1150ºC is needed. This temperature is well above the minimum reducing temperature, but is still below the melting point of iron. In this way the iron is produced in a solid state as a bloom from which the slag partly drains away. The remainder of the slag left in the finished bloom will be hammered out when the slag is still in a fluid state. A bit of slag will always stay in the bloom. (A more detailed description of the process of making iron in a bloomery can be found in the next chapter.) To keep the reducing conditions strong at more elevated temperatures in a primitive furnace, a larger ratio of charcoal to ore is needed, and the total amount of iron ore and charcoal must be kept hot without consuming too much charcoal for this purpose. In other words, the furnace must be well lagged so the heat will not escape too easily. If this is not done, the reducing conditions will become more oxidising[1].
5
A bloomery is a furnace and forge in which wrought iron in the form of blooms is made directly from the ore, or (more rarely) from cast iron. In the furnace, or bloomhearth, iron ore is reduced to iron by heating it with charcoal. The air required for combustion is provided either by natural draught or, more commonly, by some form of bellows (manually operated or water powered). Iron ore, which is essentially an iron oxide, will be reduced according to a series of reactions when the burning of the charcoal in the air blast produces a sufficiently high temperature and at the same time supplies carbon monoxide. The following reactions occur at various temperatures (and thus at the various locations in the furnace): - at 300 – 500 °C (top of the furnace) o 3 Fe2O3 + CO => 2 Fe3O4 + CO2 o Fe3O4 + 3 CO => FeO + 3 CO 2 - at 500 – 700 °C (middle of the furnace) o FeO + CO => Fe + CO 2 - at 700 – 1100 °C (bottom of furnace) o FeO + C => Fe + CO In the region where air is blown into the furnace, through an opening called ‘tuyère’, the oxygen in it will react with the carbon from the charcoal to form carbon monoxide (CO) and carbon dioxide (CO 2). The gas flow indicated in Figure 1 should not be taken literally; there is a general passage of the products of combustion upwards through innumerable small channels in the upper charge materials. The iron (or particles of metallic iron) thus obtained will sink to the bottom of the furnace and form a lump there, the future ‘bloom’. But not all the iron oxide will be reduced; part of the oxide combines with the minerals present in the ore and will form a liquid slag, which covers the bloom [2]. The bloom will be an iron or steel with a maximum carbon content of 0.8%. During the whole process the temperature in the bloomhearth will not exceed 1100 – 1200 °C. The iron won’t be in a liquid state. In later processes, like the blast furnace, the temperature of the process will exceed the melting point of iron (1540°C), and iron will be produced in a liquid state, thus making the whole process of making iron more continuous. However, the steel produced in the blast furnace is saturated with carbon (4.5%, a cast iron) and cannot be used for making armour without removing (some of) the carbon first, because it is too brittle and will shatter upon forging. The removing of the carbon was usually done in a finery, a process developed later than the blast furnace, which also explains why the transition from the use of the bloomery to the use of the blast furnace was a gradual one. In the finery a quarter to a third of the cast iron would be lost, and some charcoal was also needed for fining, so per amount of produced malleable iron more charcoal was used in the blast furnace/finery combination than in the bloomery process. In fifteenth century Austria the Stückofen was developed, a high shaft bloomery (up to 3 m, which is common for a blast furnace, instead of 1 m for a normal ‘low’ bloomery), where ‘Stücke’ or blooms of 400 kg per day could be produced. This would be a wrought iron because they had reduced the fuel/ore ratio. It was more cost effective to produce a tonne of malleable iron in this way as there was no additional cost of fining [3]. The yield effectiveness of the bloomery was experimentally found by Tylecote [4] et al. and Sauder and Williams[5]. In the experiments of Tylecote et al. the yield was calculated by the amount of available iron in the ore divided by the amount of iron in 6
the bloom. The yields from various smelts differ between 29% and 80%. Sauder and Williams came to a yield of 57% and of 60% for two smelts (the weight of the bloom divided by the amount of available iron in kg). So overall it can be said that the yield effectiveness of the medieval bloomery was not exquisite, with an average of 55% yield, but sometimes higher yields could be obtained. On the other hand, in a blast furnace, where all ore is melted, only a little of iron will go to the slag, so the yield effectiveness is almost 100%.
Fig. 1:
1 2 3 4 5
Schematic drawing of a high-shaft medieval bloomery [2]
Drying and roasting of the ore. Indirect reduction of the ore takes place by reaction with the carbon monoxide produced below. Oxidation zone, where charcoal is burned to produce heat and also carbon monoxide for reduction. Hearth, where direct reduction of iron oxide by the carbon in the charcoal can occur. Slag bath, where any adventitious carburisation of the iron tends to be removed, and where the slag flows off, leaving a slag-coated bloom of metallic iron.
7
In his book De la Pirotechnia Vanoccio Biringuccio describes how a blast furnace can be built. Although the subject of this thesis is not the blast furnace, but the bloomery, the purposes of the blast furnace and the bloomhearth are similar, namely smelting iron ore. The construction of both furnaces is therefore also similar. In general, blast furnaces are slightly bigger and/or higher than bloomhearths. When building a furnace that is suitable for smelting iron ore, the first thing to do is to find a location that is conveniently near to a water supply, so that the water power can be used to blow the bellows that are attached to the tuyère. But you also don’t want to have to walk too far for ore and for wood for the charcoal. When such a location has been found, the building can start. Stone that resists fire is needed for building the furnace. The following stones are suitable: black flintstone flecked in white, Peperino and a certain dead stratified stone that is almost half talc. [6] Before constructing the furnace a decision on the shape of the furnace has to be made. Bear in mind that each master chooses his own shape: long and narrow, bent a little at the bottom, wide or less wide where the blast of the bellows enters. Furnace shapes can be varied by doubling the walls and/or by the addition of several pairs of bellows. All blast furnaces are built against the side of a wall that is next to the water so a wheel machine can be made to drive the bellows. Ordinary blast furnaces have the shape of a mill hopper, wide at the mouth and narrow at the bottom.
Fig. 2: Various kinds of blast furnaces[6]
The furnace “ should be made between two pilasters separated from each other 1 by about two and a half braccia and four braccia or more high, for, even if they surpass the height of the blast furnace, this does no harm. ”[6] The above-mentioned fire resistant stones are used to build the furnace between the pilasters, using a little mortar. When a straight walled furnace has been made, the next step it to give the inside of the furnace its specific shape. “ In order to give the inside space its shape, a bed that inclines somewhat forward is first made as a foundation for this blast furnace, half a braccio above the ground; above this one commences to build up and make a square hollow space one and a half palmi wide. At each corner of the 1
The units of length mentioned by Biringuccio are the Florentine ones, which where also used in Siena and elsewhere in Toscany: 1 braccio (meaning ‘arm’) equals 2 palmi and is 58.36 cm long. A palmo, sometimes said to be equal to four dita, equals 29.18 cm. To specify a certain length, these units of length are frequently used as fractions. The equality of the different units of length is not always certain, so 2 palmi are not always exactly one braccio.
8
foundation two strings are stretched, extending upwards and outwards like a reversed pyramid…”[6] The final height of the furnace will be between 1.75 and 2 braccia (so
between 1m and 1.5m high), although the exact sizes do not matter much, for every master has his own ways of making things. The tuyère is placed in the wall where the bellows are. The tuyère is made out of copper. “This tuyère should be arranged so that the wind strikes in a straight line almost against the middle of the blast furnace and is directed toward the bottom .”[6] This last remark of Biringuccio is also stated by Rehder, albeit more specific: “ For the smelting hearth, the angle of the tuyère should be 14 degrees, ‘too flat’ will give ‘a greater yield of lower quality iron’ too steep will be ‘economical in fuel but giving poor yield’ of iron from the ore. ”[7]
For a blast furnace it will suffice to make a ‘drain’ for the molten iron, to escape from the furnace. The slag will float on top of the molten iron. But because in a bloomery the iron does not melt, a larger opening in the front is needed to take the bloom out. It is best to make this larger opening closable, so no heat will be lost unnecessarily. However, a small drain for molten slag can be useful, made in the door for the opening, or at any other place desired at the bottom of the bloomery.
9
To find the right ore, or to be sure the ore that is about to be smelted has a sufficient yield of iron, it is necessary, and beneficial, to assay the ore.
5.1. Assaying the ore
In Lazarus Ercker’s Treatise on Ores and Assaying [8] Ercker mentions a way of assaying the ore by means of a lodestone or magnet. This only works with magnetite iron ore so it is not a very sound way of assaying. To find out whether the ore is good or not, it is necessary to roast it, crush it small and then pull the magnet through it. “ All the good iron ore will attach itself to the magnet. Wipe it off with a hare’s foot, and again lift up with the magnet as much of the iron ore as possible. If at the end something is left behind that cannot be lifted, it is barren matter – not good ore. This will show you not only whether a mineral contains iron but also whether a given iron ore is rich or poor in iron content, since, as you know, the magnet lifts no metal except iron and steel. ”
Some ores yield on smelting a better start for creating steel. Most old authors recognized this difference in ores. Biringuccio states in his description of iron ores and the procedure of roasting and reroasting so a good malleable iron will be obtained upon smelting: “ But if, after using all possible care, it happens that the ore does not produce soft iron because of its particular nature, but is instead still hard, then it is good for making steel; indeed it is much better than soft iron. For this reason some call this the ore of steel and not of iron, but in my opinion this is an error because there is not enough difference between steel and iron to distinguish an additional ore, nor as far as I know has this been attributed to it by these speculators. I think that it can well be called ore of an iron which is more disposed to make steel than is soft iron, as I shall tell you in detail in its own place.” [6]
Ercker, though he published his book 34 years after The Pirotechnia was published, does make the distinction between iron ore and ‘steelstone’. His description of this steelstone applies to siderite (FeCO 3). “Steelstone, however, is colored very differently from iron ore; some of it looks like a yellow spar and (like some iron ores) is not lifted by the magnet in its untreated state. When you roast steelstone, though, it changes color so that it resembles the rich kind of iron ore and is then lifted by the magnet very readily – as a matter of fact, even more willingly and eagerly than iron ore.” [8]
Agricola mentions the assaying of iron ore almost as an afterthought, but he also uses a magnet in the process. He does not make the distinction between ‘steel’ ores and iron ores: “ Lastly, we assay iron ore in the forge of a blacksmith. Such ore is burned, crushed, washed, and dried; a magnet is laid over the concentrates, and the particles of iron are attracted to it; these are wiped off with a brush, and are caught in a crucible, the magnet being continually passed over the concentrates and the particles wiped off, so long as there remain any particles which the magnet can attract to it. These particles are heated in the crucible with saltpetre until they melt, and an iron button is melted out of them. If the magnet easily and quickly attract the particles to it, we infer that the ores is rich in iron; if slowly, that it is poor; if it appears actually to repel the ore, then it contains little or no iron. ”[9]
However, he does elaborate on the assaying as to actually say to melt the selected pieces of iron ore in a crucible with the addition of saltpeter. Regrettably, he does not tell what is done with the obtained iron button, if any tests are done with it or not. 10
Before smelting the ore, it is necessary to prepare the ore before putting it in the furnace. This is briefly mentioned in the first chapter, but I will come back to that later in more detail. Equally important is the fuel you use in the furnace. This is commonly charcoal. But there are many different kinds of charcoal, and there are also two ways of converting wood into charcoal. Biringuccio describes the differences in his book Pirotechnia [6]. I will summarize in §5.2 what he has to say.
5.2. Making charcoal When making charcoal, a very important material for smelting, it is beneficial to pay attention to the differences between the woods and also to the way of making it. Charcoal should be made of vigorous and powerful wood, it will not suffice to use soft wood when you need a long-lasting powerful fire. On the other hand, if flames are needed, as in reverberatories, it is useless to use charcoal, instead seasoned and dried wood is best used. “ For strong charcoal, certain kinds of wood of an earthy 2
3
nature are chosen like that of the oak, cerris , holm oak , elm, beech and other similar large and hard trees. The soft charcoal is that which is made from all wood that seems to be most domesticated and to contain more of the airy nature, like that of the fir, willow, elm, alder, walnut and others that are of a softer and weaker quality. ” It is
noteworthy that Biringuccio mentions elm to be suitable for both the strong charcoal and the soft charcoal. It might be that different parts of the elm are used for either strong charcoal or soft charcoal. Wood burns differently than charcoal does. Burning wood yields much more lively flames, because it contains more humidity. But charcoal burns more vigorous, “ since it is freer form humidity and more united with more live forces, the air penetrates less, so that the fire, as actually happens, is more easily introduced there and comes nearer to the thing in a manner more united both in power and action .”
For making charcoal the desired goal that is to be achieved has to be kept in mind. It will be no use to take broom charcoal to smelt gold, silver or copper with, it would be labouring in vain. Likewise, if it is desired to boil iron, it is no use to take charcoal of willow, fir, poplar, white poplar or similar trees. Charcoal from the chestnut, broom or other trees of a similar nature is of no use to the smith. The right preparation of charcoal is important, for not every tree that is good by nature will make good charcoal without good preparation. Differences in quality of charcoal made from the same kind of tree can occur due to different baking times and to different kinds of earth used for covering while baking. You will get different results when you use similar woods but one batch contains young wood, and the other batch contains wood from an old tree, or whether the wood is clear or knotty. Also the freshness of the wood is important for the result, because fresh cut wood will yield another type of charcoal than charcoal made from dry and decayed boards or dry and well seasoned wood. The location where the trees grow has also influence on the quality of the charcoal obtained from the tree. There are two methods of making charcoal.
2 3
cherry a southern European evergreen oak
11
“The first and the best of all is that called appagliaro charcoal.4 To make this, a level place is chosen convenient to the wood that is cut up for the process. If the bed is not level, it is made so, and is given a round form. Four large poles are set up in the middle in a square, or three in a triangle, which make an empty space of little less than half a braccio. Then proceed around these, covering them upwards, circle by circle, with all your cut wood and with chumps split into smaller pieces, building it in
Fig. 3: The construction of charcoal piles[6]
the form of a round pyramid or a haystack, from which it takes its name. If you wish to make good charcoal, the wood should have been dried at least six months or a year. Proceed thus to stack piece on piece with certain intervals until you have arranged the width and height that you wish the charcoal pile to be. The middle space between the poles is left empty up to the top. When this pile has been made, it is well covered all over the outside with fern leaves and broom and then, on top of these, it is plastered well from the top with good, firm earth, as dry as it is dug up. This plastering is made a palmo or a little less in thickness, very well prepared and sealed so that it may not breathe except at the top where ten or twelve vents are left as outlets for the smoke and humidity contained in the wood and earth. This done, the fire is thrown to the bottom of the little hole that you left in the middle between the poles , and certain little dry parts of small branches and dry leaves are put on it so that they fill it up to the top or as far as necessary for the fire to spread throughout. Then this opening is also closed over with earth and only the vents are left open. In six or eight days the whole charcoal pile gradually takes fire and bakes. When it is seen that the vents cease to give off strong exhalations it is to be believed that it is baked. Then, with the same kind of earth, seal the top and all around and in every place so that none of the vents can breathe at all, and the fire inside immediately suffocates and dies out when it finds itself without respiration. Thus your wood is converted into wholly spent charcoal without ashes or any humidity .”
“There is still another method for making charcoal. The usual way for this is the one that smiths follow in making it of broom or of chestnut, and that makes it harder but smaller. In order to do this a hole is made in the ground about one and a half braccia in diameter and of about the same depth. It is filled, indeed it is heaped up full, with broom roots or small pieces of chestnut or some other wood. A space is left in the middle from the top to the bottom for lighting the fire. The rest of the space that is still uncovered is filled with fern or broom and then with earth just as I told 4
i.e. charcoal made in piles like a haystack.
12
you above is done for the large charcoal piles. Proceed in the same way in applying the fire and also in extinguishing it, but because it is a small quantity, it is well baked in eight or ten hours after the fire has been lighted. The charcoal thus made is for the smith’s forge. It is not good for smelting even though it be made of good wood, especially if the blast from powerful bellows is not used, because it does not burn like that made in the pile on account of its hardness. But when the fire is once introduced there, it continues very well .” “ In conclusion: charcoal which is to be called good must be made of good dry wood
Fig. 4: Charcoal-burning in pits[6]
that is well seasoned, and baked but not burned, because it then becomes small and weak. If it is rightly baked it is thick and strong and, when struck with another piece, is as resonant as glass. Therefore, whoever has to work with it must take care that it is good and attend to the processes of making it as I have said .”[6]
5.3. Preparing the ore After proper charcoal is made, or probably during making charcoal, the iron ore has to be prepared. As mentioned in the first chapter, there are different kinds of ore, and some ores are very easy to reduce, so you won’t need a big furnace for smelting it. For example, Elban ore is excellent and easy to acquire. This ore can be reduced and purified by just putting it in a forge fire in front of the tuyère. A very soft and malleable iron is obtained. No powerful fires are needed. The process is described like this: “… a sim ple forge fire is sufficient, together with a pair of bellows not much larger than ordinary ones. After the ore has been broken in little nutlike pieces it is arranged on the said place in a heap and around this a circular enclosure is made of the larger pieces of ore or of other dead rocks, put there only in order to hold the charcoal and fire close together. The amount that is wished to reduce is well covered with charcoal and then an arrangement of bellows connected with a water wheel which moves them is caused to blow. With a fire of only eight or ten hours duration the ore is smelted and cleansed of the earthiness which it contains; thus the iron remains all pure in a waxlike mass. The said rocks are removed from around it and it is removed from the forge while still hot. It is broken in pieces with mallets by hand, and then each piece is reheated, carried to the forge-hammer machine, and made into blooms ( masselli ). This process is done in every works twice a week and is called il far dela cola (i.e. making a heat, or a “running down”). When it is finished, the said blooms of the iron thus extracted are taken again and placed on the same forge, and well reheated. When hot they are cut up by the said forge hammer and extended into the form of a bar, squares or any other desired form. …”[6]
13
In this description of reducing an easily reducable ore, the following things are noted: -
the size of the lumps of ore, ‘nutlike pieces’, assumable the size of a walnut (a good amount of) charcoal is put on top of the ore (‘well covered’) position of the bellows is unclear time factor ‘a fire of only eight or ten hours duration’ rocks removed from furnace when hot and broken into pieces (while still hot?) and made into blooms frequency of the process in a workshop; ‘in every works (workshop?) twice a week’ blooms are well reheated and made into shapes The occurrence of these ores is rare, most ores “need to pass through large
furnaces with strong and powerful fires, consuming great quantities of charcoal and much l abor, for their wildness cannot otherwise be tamed since it is caused by bad elemental mixture or traces of other metals of which they are often so full that they can scarcely be liberated .”[6]
How good ore should look like: “Good ore must be clear, he avy, and firm of grain, free from earth, stone and all traces of any other kind of metal. The color of good ores is black and those that have the color of lodestone are worth very little, because almost all of them contain a trace of copper .”
The erroneous idea that small amounts of copper are harmful to iron is mentioned by Pliny[10] ( Natural History, Book XXXIV, para. 143) and persisted for nearly two thousand years [6]. Pliny does not exactly state that small amounts of copper are harmful, but one could get the implication from the following: ”But to come unto the nature of iron, herein appeareth still the same goodness of Nature, that this metal working such mischief as it doth, should be revenged of itself, and receive condign punishment by the own rust. See also the wonderful providence of Nature, who maketh nothing in the world more subject to death and corruption, than that which is most hurtful and deadly to mankind. But wonderful it is above all, that man’s blood should have such a virtue in it, as to be revenged of the iron blade that shed it; for being once imbrued therein, it is given ever after eftsoons to rust and canker.” From
reading this, I do not immediately get the implication that copper may be harmful to iron. What Pliny says is that man’s blood contained in the iron is harmful. Now I can imagine that if a smith sees small reddish dots in his iron (being copper) that he could compare it with small blood spatters. This way, it can be that when Pliny mentions the blood, he actually means copper. But this is only speculation. When the good ore is obtained, the amount of ore to sm elt should be weighed or measured. Break the ore in to small, bean-sized pieces. If evaporation by fire (roasting the ore) or cleaning by washing was needed, this should be done by now. “ A layer of this is arranged on a bed of planks, bricks, or flat stones in front of the blast 5 furnace and its companion is placed on top, then a quarter part of galena or a third if you can obtain it. Then an equal part of crushed iron slag or slag from the same or different ore is added. These materials are spread in layers one above the other .”[6]
5
Galena (origin in Latin; lead ore): a bluish gray mineral with metallic luster consisting of lead sulfide, showing highly perfect cubic cleavage, and constituting the principal ore of lead
14
It is not clear what is meant by ‘companion’ in the quoted first sentence. It could refer to a companion layer of pieces of ore, which makes sense, but it could also refer to another bed of planks, bricks or flat stones, or a second blast furnace. The latter two options are not logical to me and I will stick to the first option, a second (companion) layer of ore. ‘A quarter part of galena or a third’. This is also open to more than one interpretation. It could mean that an amount of galena is put on top of the two layers of ore equivalent to a quarter (or a third) of the volume of the layers of ore, or an amount equivalent to a quarter (or a third) of the weight of the layers of ore. The mentioned part of crushed iron slag can be equal to the amount of ore and galena on the bed of planks or can be equal to the amount of galena only. After having sorted the ore mixture and having a well constructed blast furnace that is full of well-kindled burning charcoal, the smelt ing process can start. First added is some more charcoal and at the same time water is admitted to the bellows machine. “ When you see that it is well kindled again from this wind and that the flames begin to come out forcefully on top, work it down with a little rake and fill it up again…”
I assume that another load of charcoal is meant. “…Then the pannier is filled with new charcoal, the blast furnace heaped full, and another basketful of the said composition of ore i s also put on top .” This would lead to the idea that previously not another load of charcoal is added, but a basketful of the ore mixture. “ Proceed this way, continually adding charcoal and ore as long as you have any, or as long as you wish to keep u p the work, maintaining the blast furnace always full, or continuing in this way until the bottom of the blast furnace is filled with melted materials. Decide this by judgment, or by seeing through the mouth of the tuyère where the wind of the bellows enters that they are level with this. Then with an iron tool open the hole that was left as an outlet in front of the blast furnace and allow all the metal with its slag to run out .”
In a bloomery the iron will of course not run out, but a part of the slag will. At the end of the process in the bloom ery, the fires within the furnace will die and the bloom is taken out while still hot and will immediately be hammered into the desired shape. While doing this, most of the slag still present in the bloom will be hammered out (occasionally the bloom will be reheated in a stringhearth before continuing hammering the slag out). For the next use of the furnace, the mixture of (iron oxiderich) slag and unreduced iron ore will be used again. This way most of the iron will be taken out of the ore. The roasting of iron ore is a special form of ore-dressing that was widely used in antiquity. Roasting would make the water go out and all the iron in the ore would be oxidised. This is favourable, because the reduction of the ore goes best if the iron is present as an ironoxide in the ore. Also, roasting would make the ore more porous, and thus more amendable to the reduction process. There is considerable archaeological evidence for roasting. Roasting sites have often left shallow circular depressions. Sometimes the base of these depressions are fire reddened, and pieces of slag and magnetic (roasted) ore can also be found there [11].
15
In 1408 the Bishop of Durham ordered that a water powered bloomery be built at Kyrkeknott (or Byrkeknott), Weardale. The account roll of the ironmaster of this bloomery has survived through time. This account roll gives valuable information about the costs of building and operating a bloomery in early fifteenth century England. The scale of the production of iron in the bloomery in Kyrkeknott was six times larger than in earlier times. This is ascribed to the fact that this bloomery was water powered (i.e. the bellows are operated by means of water power). The average daily production of Kyrkeknott was a bloom of 195 lb (~88 kg), whereas in about 1340 in Tudeley, Kent, the daily production averaged 30 lb. [12] The surviving records of the Bishop of Durham’s bloomery in Weardale give a very detailed and complete story of the weekly operation of a bloomery. Lapsley made a translation of the original Latin texts at the end of the 19 th century. He claimed that the text was written in very faulty Latin, but his interpretations of it were as faulty as the Latin itself. Nevertheless the transcription is a very valuable contribution. A summary of total costs is given below [12]. Receipts
From William Chauncellor, Constable of Durham From the same, per Ralph Eure, the proceeds of the coal mines of Rayley Total
£ (=20s.) 32
s. (=12d.) 0
d.
55
15
11
0 £87 15s. 11d.
Expenditure
(1) Preparing the site (2) Mason’s work, making the water course (3) Carpenter’s work making the water wheel, dam and other wooden equipment 6 weeks @ 3s. (4) Felling the trees, 12 weeks @ 2s. 6d. less 6d. (5) Making four buildings and the hearths; sawing and carriage of timber (6) Tools for the forge and ironstone mines and other metal work
2
15 10 18
4 4 0
1
9
6
1
5
7
3
17
9 £10 16s. 6d.
(7) 53 dozens of ironstone from Sir Ralph Eure royalty @ 2s. carriage @ 2s. mining @ 5s. carriage of cinders Total initial costs (8) Expenditure for 52 weeks of operation Total expenditure Credit balance
5 5 13
6 6 5 8
0 0 0 0 £24 5s. 0d. £35 1s. 6d.
47
18
11 £83 0s. 5d. £4 15s. 6d.
At the site where the bloomery was built, four houses or huts were also built, for the bloomer as well as for the foreman who lived on the spot. Their wives helped in breaking and sifting the ore, and sometimes they worked the bellows. The bloomer, assisted by another bloomer from a neighbouring bloomery, made the bloomhearth and the stringhearth[13]. In the stringhearth the blooms made in the bloomhearth are refined. This basically consists of reheating the (half)bloom between hammerings. The people working at the bloomery were paid a certain amount of money per bloom 16
made. The bloomer of this bloomery, John Gylle, received 6d. per bloom. In 47 weeks time he made 278 blooms, and so he earned an average wage of nearly 3s. per week, compared with 2s. for six days for a labourer. The total amount of production in one year, 278 blooms, weighed about 24,360 kg. The average weight of one bloom was about 88 kg. Of the 278 blooms, 217 were refined and cut into pieces, the remaining 61 were not. The total costs for labour during the 52 weeks in which the bloomery was operating and made those 278 blooms were £47 18 s. 11d . These costs include: coling (costs of charcoal), carriage (of charcoal), paying of the bloomer, the smith, the foreman and the wives of the bloomer (Mrs. Gylle) and the foreman (Mrs. Whenfell), cost of ale and cost of repairs. This would amount to operating costs of 41.39 d . per bloom. Added to this total should be the costs of the ore (with royalties), a total of £24 5 s. 0d . or 20.94d per bloom. In total, without taking into account the costs of actually building the bloomery, one bloom of 88 kg would cost 62.32 d. or 5 s.2d. (or, expressed in days of work for a labourer, fifteen and a half days of work). If you want to include building costs, then for a year of making blooms, 9 d . per bloom should be added to the total costs per bloom. Of course, the more blooms made in the bloomery before any major repairs are done, the less this amount will become. The information on the payroll does not express how long the bloomery was in operation after the credit balance was made up. To conclude this calculation, 100 kg (0.1 ton) of bloom iron costs 70.8 d . (almost 6 s.; which was what three labourers would earn in 6 days of work). With machinery costs (building costs) included 100 kg iron costs 81 d. (6.75 s.) Nowadays hundred kilograms of steel cost about 14 euro, and if you take the investments expenses into account, it costs €16. So in comparison, a hundredweight of medieval iron costs as much as a labourer would earn in twenty days of work, while the same amount of modern steel costs as much as a man can earn today in one hour of work. This can mean two things. Either the wages for a labourer have gone up a lot over the course of time, or the production of steel has become more costefficient. It is most likely that the latter effect played an important role in this. But, on the other hand, wages have indeed gone up during time, if only to keep up with inflation. Inflation, however, has not been included in this comparison, because that would unnecessarily complicate matters. The payroll can be used to find out how many people were needed to build the bloomery and how many to operate the bloomery. At the bloomery of Kirkeknott the following people had a fulltime job: - the bloomer of the bloomhearth (John Gylle, paid 6 d. per bloom) - the smith of the stringhearth (John del Logge (John Logge, John Loge), paid 7d . per bloom) - the foreman (Thomas Whenfell, paid per bloom, made on the string hearth 2 d. for helping John Gylle and 3 d. for helping John del Logge) - the colier (John Mody, he supplied the coles, a dozen at 2 s. (he is mentioned in the account roll frequently, his helpers and he cut down wood from a nearby park and burn it to make charcoal) The part-time jobs were for the wives of John Gylle and Thomas Whenfell, for operating the bellows and/or breaking the ore. Note that even though the bloomhearth and stringhearth were water powered, the water supply from the nearby stream was not always sufficient and it was possible to revert to manual operation to keep the process going[1]. The wives were always present when the bloomery was operating, 17
because they lived on the site. The last important person was the ironmaster, or the overseer of the forge, John Dalton. He wrote the account roll and was responsible for paying the people. His wages are not on the account roll though. For building the bloomery auxiliary help was needed in the form of: - a bloomer from a neighbouring bloomery to help John Gylle build the bloomhearth and stringhearth - a mason and his 4 helpers (labourers) to build a water course and to make a stone gutter. The wife of one of the labourers also helped in the work. These workers were employed for 30 days. - A carpenter (Thomas Gylle, maybe brother of John?) to build the forge house, the water wheel, the spouts, bellows and other necessaries for the forge. - Two men were hired to fell the trees needed by the carpenter. - Another carpenter (Roger Coley) and his helper built the houses where the bloomer and foreman would come to live with their families. - Someone to carry loads of timber and an axletree for the waterwheel to their destinations. The tools which were bought were made by various people, but they did not actually work on the site themselves, so I won’t include them. Sir Ralph Eure was responsible for delivering the iron stone, but other workers mined the ore and carried it to the final destination. All the ore necessary for a year was supplied at the bloomery in the beginning of the financial year which ran from Michaelmas to Michaelmas (November 1408 – November 1409). In total, some 15 people were hired for building the bloomery. Only six people (4 men, 2 women) were needed to operate the bloomery. For small tasks outside the routine of running the bloomery, some men were hired off and on to do such things. In a modern steelmaking factory, up to 5000 workers can be employed. It has to be said that most of these workers won’t busy themselves with operating the blast furnace. Most work goes to the secondary steelmaking process, where steel with a certain desired composition is made. In the process of preparing the cokes and ore and in operating the blast furnace, about 600 people are occupied.
18
The modern blast furnaces have gigantic proportions, they can be up to 110 meters high, but the principle of reducing iron ore to iron is still the same. Instead of using charcoal as a fuel, cokes are used. Coke is coal that has been heated to get rid of the oils and other impurities in it. Pieces of about 3 to 10 cm of coke in diameter, with a 90% to 93% carbon content, are fed into the furnace. Although cheap and plentiful, coal, without the conversion to cokes, isn't a feasible fuel because it contains sulphur that makes the iron too brittle to be of any use. The transition from charcoal as a fuel to cokes was an important one. At the beginning of the eighteenth century wood had become very expensive, because the forests were being cleared for farmland and timber, and in the centuries before that, a lot of wood was used to make charcoal. If the discovery of using coal (in the form of cokes) as a fuel was not made, it would have been very likely that Europe would not have any forests today. Also this invention made the Industrial Revolution possible, because it allowed a major expansion of the iron trade. [i-1] The iron ores used today are haematite or magnetite with an iron content varying between 50% and 70%. These ores can be used in the blast furnace without any preparation other than being made into small pieces of 0.5 to 1.5 inch. If the ores contain less iron, it is required to process it so its iron content will increase. This ore is crushed and ground into a powder, the waste material can be removed. The remaining powder that is rich in iron is rolled into balls and fired in a furnace to produce strong, marble sized pellets that contain 60% to 65% iron. When operating the blast furnace, aside from cokes and ore, limestone is also added in pieces of a size between 0.5 and 1.5 inch to become blast furnace flux. This flux can be pure high calcium limestone, dolomitic limestone containing magnesia or a blend of the two types of limestone. The calcium (or magnesium) will bind with the sulphur that is present in the furnace, and the bonded sulphur will end up in the liquid slag, and not in the iron. In short, iron ore, coke and limestone are dumped into the top of the blast furnace, and preheated air is blown into the bottom. This air is preheated by the burning of the gases that come out of the top of the working blast furnace. The raw materials require 6 to 8 hours to descend to the bottom of the furnace where they become the final product of liquid slag and liquid iron. These liquid products are drained from the furnace at regular intervals. The hot air that was blown into the bottom of the furnace ascends to the top in 6 to 8 seconds after going through numerous chemical reactions. Once a blast furnace is started it will continuously run for four to ten years with only short stops to perform planned maintenance. [i-2] The final product of liquid iron contains little bits of silicon, sulphur, phosphor, manganese and titanium (less or much less than 1%). The carbon content is 4.1% to 4.4%, which makes the product a cast iron. Before continuing to the secondary steelmaking process, this carbon is removed so the iron will only contain about 0.05% of carbon. In the secondary steelmaking process the desired amount of carbon will be added again. In those centuries of development since the first blast furnace was used in the fourteenth century, the basic principle of the chemical reactions inside has not changed much. It is only a bit more refined today because people know exactly what is going on in the furnace. Also the secondary steelmaking processes are much more advanced, and the result is a very clean steel, with no inclusions, made in a controllable process. As the early blast furnaces had a production of one ton of steel per day, the modern ‘giant’ blast furnaces can produce up to 13,000 tons per day. The production of one ton of steel per day in a medieval 19
blast furnace is already more than can be produced in a bloomery (maximum of about 150 kg iron per day, or about 400 kg per day for the Stückofen). So the big difference between modern and medieval processes is the scale on which the iron is produced. Another detail that should be mentioned here is that the modern steelmaking process is a 24h process, it does not stop unless something goes wrong or when it is time for periodical cleaning/reparation. The medieval bloomery was only operated during daytime, between sunrise and sunset. Because there are on the payroll of the bloomery of Kirkeknott no expenses for candles and such, and there is no increased production in summer when the daylight lasts longer, it can be assumed that during the year, be it summer or winter, similar operating hours were maintained. In Figure 5 you can see the weights of the blooms produced in various times. There is a steady growth in the size of bloom produced from 250 grams of the pre-Roman period up to the 15 kgs or so in the middle of the 14 th century. It can be clearly seen that the invention of the water powered bellows around that time gives a sudden increase in bloom weight. [1]
Fig. 5: Weights of blooms produced by the bloomhearth during its period of use[1].
An interesting comparison between modern and medieval methods of iron production can be made with the following calculation where the amount of produced (raw) iron per fulltime worker per year is calculated. In the medieval bloomery of Byrkeknott the production was about 24 tons of iron per year. Because the workers did not work on Sundays and feast days, my estimate is that they worked 275 days per year. For convenience sake, we’ll say that the people worked 10 hours per day on average. Another estimate is that in the production of the iron 4 people would be occupied fulltime (the 10 hours per day). In total about 115 days non-stop work is put into the 24 tons of bloomiron by 4 workers (1 for making charcoal, 1 for operating the bloomhearth, 1 for operating the stringhearth and 1 for all other auxiliary help). In a modern blast furnace (Corus IJmuiden, BF7) 6 in 115 days of continuous work about 1,150,000 tons of raw iron is produced by 100 employees working in 5 shifts of 20 workers working at the blast furnace and another 66 people working at technical service of the blast furnace. Additional working forces are employed at the 6
Information about BF7 (Hoogoven 7) obtained from email and conversation with R.Boom
20
cokesfactory (235), sinterfactory (76) and pelletfactory (76). In total about 550 people are busy making raw iron. These men work in shifts because no one is working 24 hours. An estimate is that an equivalent of 150 ‘working forces’ of 24 hours do all the necessary work. Per medieval worker 6 tons of bloomiron is produced in 115 days, whereas in a modern blast furnace about 7,600 tons of raw iron is produced per worker in the same time. The difference in production can be explained not just by the difference in technology, but also by looking at the demands for steel in medieval and modern times. In medieval times the demand for steel was mainly in the armour and weapon industry, but also a lot of farming and household implements had to be made of iron or steel. In modern times, particularly since the advent of the industrial revolution, you find steel being used almost everywhere, and one of the main applications is the building industry. But the packaging industry and car industry should also not be forgotten. So a rise in demand and the change of applications for steel meant that the production had to increase and be more economic according to the workforces, because it had to become cheaper to use.
21
In general the carbon content of a bloom does not exceed 0.8%. This percentage can be observed at some regions of the surface of the bloom, but the carbon content inside the bloom is generally much lower, typically a few points of a percent. After being removed from the furnace, the bloom was worked (wrought) with a hammer to remove the slag and to make bars or other usable shapes. A bloom made into these shapes is therefore called wrought iron, even though it was in fact low carbon steel. Most blooms were formed in a heterogeneous way, part of which having higher carbon content than others. This is confirmed by the fact that frequently medieval artefacts, including many examples of armour, show a banded microstructure, suggesting that they were forged from a heterogeneous bloom. [14] Early smiths would know the difference between harder and softer pieces of iron, but they would not have realised that they were handling steel mixed with iron. They did realise that the harder pieces of iron were better for making armour. An easy way of obtaining the benefits of steel is to convert a part of the iron artefact to steel. This can be achieved by forge-welding a steel edge to the iron back, or by ‘casecarburising’ the edge (heating the iron in contact with carbon for hours). This last technique would suggest that the smiths realised that iron could be converted into steel. This understanding was developed long before the Middle Ages, throughout which it was practised regularly. Case-carburising is very suitable to harden the edges of files, knives and swords, but it is less suitable to be used when making armour, and is therefore seldom found therein. When a piece of heterogeneous (banded) wrought iron was brought to an armourer, he could, in order to homogenise the iron, hammer it out and fold it, and repeat this a couple of times. The thickness of the bands in the banding observed will become much smaller and the distribution of carbon will be more even. When making pieces of armour, most things can be worked cold, but to make the more difficult curves and bends the piece needs to be worked when hot. Aside from the degree of homogeneity a measure of the quality of the armour is its hardness. The scale of hardness commonly used is the Vickers Pyramid Hardness scale (VPH) with units of kg/mm2. Iron (pure ferrite) has a hardness of about 80 VPH, but because medieval iron will contain slag and elements may be dissolved in the ferrite, its hardness may vary between 100 and 180 VPH. The presence of carbon in the iron will form iron carbides and this will make the thus formed pearlite harder, a carbon content of 0.2% will yield a harness of about 180 VPH whereas a carbon content of 0.6% will give 260 VPH. So, the hardness of wrought iron will be anywhere between these values. This is of course without taking into account a quenching and a possible tempering of the steel. If the steel is cooled by plunging it red-hot into cold water (quenching) a structure called martensite is formed which will cause a greater hardness. When at quenching an all-martensite structure is formed, the quenching is called ‘ fullquenching’. Quenching in oil, molten lead or some other less drastic coolant than water, or an interrupted or delayed quench will result in a combined pearlitic, bainitic and martensitic microstructure. This is called ‘slack-quenching’ and was regularly practised in the Middle Ages. Nowadays this technique is avoided. The only drawback of full-quenching is that the steel will become brittle, which is of course not desired. Therefore it was needed to carefully reheat it to ‘temper’ it. Tempering will reduce the hardness, but it will give an increase in ductility. These techniques are very advanced though, and it is very difficult to obtain the right result. Most 22
armourers would therefore abstain from full-quenching and tempering. Also few armourers would try slack-quenching, so most pieces of armour would be unhardened and have a ferritic/pearlitic structure. In his book ‘ The Knight and the Blast Furnace’[14] Alan Williams describes tests he performed on medieval steels. He did experiments on the slack-quenching and on the tempering of the steels and he did some mechanical testing on some samples of the armour. Results were compared with the behaviour of a modern steel (EN 42). He has used two large specimens from 16 th century armour. Specimen A: this was part of a vambrace (upper arm defence) made in North Italy ca. 1570. As received, this was a pearlitic-ferritic steel, of average hardness 183 VPH. Energy-Dispersive Analysis during the course of SEM gave these bulk percentages of elements (the carbon content was estimated by metallography). C 0.5 wt% Si 0.24 wt% Mn 0.02 wt% Ni, P and S neglible
Specimen B: This was a plate from a pauldron (shoulder defence) made in South Germany, probably in Innsbruck ca.1550. As received, this was a uniform tempered martensite, with an average hardness of 514 VPH. A spectrographic analysis gave these bulk percentages of elements: C Mn S P Si Al Ni
0.49 wt% 0.07 wt% 0.007 wt% 0.049 wt% 0.02 wt% 0.007 wt% 0.03 wt%
At Reading University Williams carried out experiments on delayed quenches of these two pieces of armour. A: the Italian vambrace c1570 Delay before quenching
(sec) 0 3 5 6
Hardness (Rc) 56 31 25 30
VPH 620 310 266 300
(sec) 5
Hardness (Rc) 28
VPH 287
B: the Innsbruck pauldron c1550 Delay before quenching
Modern steel of composition; C 0.70-0.85% Si 0.10-0.40% S max 0.05% P max 0.05%
23
Mn 0.55-0.75%
Rockwell C hardness (as received) Delay before quenching
(sec) 10 15 16 17 20
Hardness (Rc) 60 49-52 39 27 23
VPH 720 530-560 400 270 230
From these experiments it can be observed that the medieval steel is much more sensitive to delays in quenching, the hardness drops fast even after a few seconds of delay. The modern steel is, due to the alloying of Mn, less susceptible to a drop in hardness if the quenching is delayed. Other alloying elements like Ni, Cr and Mo will also influence the rate at which martensite is formed and will alter the behaviour of quenched steels on tempering. Generally, they do not soften as quickly because the carbide particles do not coarsen as rapidly. This is also confirmed by experiments on tempering, both specimens of armour plate show a faster rate of softening on tempering than would be expected from modern steel. Presumably this is because of the absence of alloying elements such as chromium, manganese, nickel and silicon. In medieval armour manganese is only found in slag inclusions, and chromium is also absent. The mechanical testing of the two specimens gave the following results [14]: A: the Italian vambrace of c1570, a pearlitic steel, hardness 183 VPH B: the Innsbruck pauldron plate of c1550, which was also made of a 0.5%C steel, but a tempered martensite of 514 VPH Tensile testing of armour Specimen
Yield Stress MN.m-2
UTS MN.m-2
Elongation %
Young’s Modulus
A B
107 132
426 513
40 19
105,000 130,000
These results may be compared with the mechanical properties quoted for a modern plain carbon (0.5%C) steel quenched and tempered at different temperatures. Temperature
VPH
UTS MN.m-2
Elongation %
300 ºC 500 ºC
470 300
1600 1000
7 20
8.1. Protection by armour Armour was made in the first place to protect the wearer from blows from swords, maces, spears, arrows, bolts, etc. When the quality of the weapons improved, also the protection against them should improve. But how effective was the protection? 24
A blow from a sword or any other moving object will have a certain amount of kinetic energy, which is transferred to the object it hits on impact. 1 2 mv 2 Where m = mass (kg), v = velocity (m/s) and E kin = kinetic energy (J). A 100 gram missile travelling at a velocity of 40 m/s will have a kinetic energy of 80 J. In hand to hand combat in the Medieval period, any blow from sword, spear or axe might deliver an energy of anything between 60 an 130 J. Depending on the hitting surface of the instrument delivering the blow, a lesser (big surface) or greater (small surface (point or sharp edge)) effect can be achieved. For example, to pierce a sheet of steel of 2 mm thickness striking at the normal, an arrow needs an energy of at least 175 J, whereas a bullet, being bigger and thus having a greater impactsurface than the arrow, needs an energy of at least 750 J. [14] In comparison to blades and clubs, arrows and crossbowbolts are another matter. Medieval archers used longbows of up to 120 lb draw weight. An arrow shot at 10 ft range with a 50 lb bow had a maximum striking force of 170 J, a 75 lb bow of 212 J. The initial energy of the larger longbow is comparable to the energy of a crossbow. After travelling a distance through the air, the arrows and bolts lose velocity and thus kinetic energy. The crossbowbolt will lose energy more rapidly, because of its weight and shape (it generates proportionally more drag). The early handguns, with a 50 g lead ball and a muzzle velocity of at least 100 m/s, deliver an initial energy of 250 J. At a distance of 30 m, this bullet can penetrate a ‘suit of light armour’. Later early handguns could have a muzzle energy of up to 500 J. The more extended use of guns and cannons in the sixteenth century announced the end of the glory of the armour, which had its peak in the late fifteenth century. When handguns were more commonly used as a weapon in battle, there is an increase of the maximum thickness of armour, from around 2 mm in the 15 th century to around 6 mm which is regularly found by the late 16 th century, some plates even as thick as 8 mm. At some point Montaigne (c1580) observed that as many men were lost by the weight of their armour as were saved by its protection. [14] At some stage it did not help to make the armour thicker to offer enough protection. Also little more could be done with the geometry of the different plates. Fluting and curves can do only so much. A suit of armour was expected to be proof against the crossbow. The medieval armourers of Germany and Italy did test their suits by ‘proving every piece with strong crossbows’. If the test did not end satisfactory, the master armourers would not award the full mark (the armourer’s mark imprinted on the pieces), and those suits were sold for a lot less than a suit of armour which did receive full mark. In general, the performance of a suit of armour depends on four factors, of which the first two could be manipulated by the armourer: thickness, shape, hardness, slag content. The last two factors depended on the raw material supplied, although the armourer could also influence the hardness to a degree. The influence of thickness is quite clear, the thicker the armour, the more protection it offers (up to a certain point where overly thick armour becomes impractical). With regard to the shape-factor, it should be noted that the impact of an arrow will not be maximal when it hits the armour at an angle. To a first approximation the proportion of the impact energy delivered depends upon the cosine of the angle of attack with respect to the vertical. The armourer can take advantage of this effect by E kin
=
25
designing his armours with a globose breastplate and cylindrical protection for limbs. A test done on a curved plate yields the following results [14]: 60J impacts on 1.2mm mild steel Angle of impact 0° 10° Penetration (mm) 31 30
20° 26
30° 20
40° 14
45° 11
50° 9
55° 5
The hardness influences the performance in such way that with higher carbon content, and thus a higher hardness, the steel will be penetrated at larger energies. But care must be taken here, for if the steel becomes too hard, it will also become brittle and thus will be unsuitable to make armour out of. Slag inclusions account for failures because they are a weakness in the steel and at these points cracks can initiate very easily. The fracture toughness of the steel and iron decreases with increasing slag content. The steelmaker, or the bloomer and the smith at the bloomery can influence these last two aspects. The men at the bloomery would have had such experience to be able to tell ‘good’ iron from ‘bad’. I can imagine that good iron was generally sold for high performance ends, such as making armour, and that bad iron was sold to make more common things out of, like farming implements, although their performance should also be good enough.
8.2. Microstructure of some pieces of armour In microstructural pictures of medieval iron and steel it can often be seen what process of making has been used. Also it gives information about the carbon content and the way the piece has been heat-treated (if at all). Additional hardness tests to the specimens will tell something about the quality of armour. Alan Williams has researched many specimens during the past 30 years. What is remarkable from all the tests is the great variety of quality of the iron and steel of the pieces of armour. The following pictures are taken from his book where he has summarised his research [14].
Fig. 6:
Microstructure (X 50) of a gauntlet of an embossed halfarmour made in Italy around 1580. Only ferrite and slag inclusions are seen.
26
a
c
b
Fig. 7:
Italian sallet c. 1460 (a). Cross-section (X 40) made from the rim of the sallet behind the visor (b). The microstructure consists of ferrite and bainite/low-carbon martensite with few slag inclusions (c, X 200). This is a low-carbon steel which has been hardened by some form of quenching.
a
b
Fig. 8: Lower plate from a right
knee defence. Italian, c. 1520 (a). The plate was examined in section, the microstructure consists of ferrite and pearlite with some slag inclusions(b, X 40). The plate was decorated with fluting, etching and gilding. This is a medium carbon steel (< 0.5% C) which has been air-cooled after fabrication.
27
a
b
Fig. 9:
Breastplate fabricated in Augsburg c. 1485(a). Specimen from within the breastplate was examined. (b) X 320, martensite, pearlite and widmanstättenferrite. The microhardness averages 405 VPH. This has not been quenched fast enough to form an all-martensite structure.
Fig. 10:
Microstructure (X 120) of an infantry breastplate with an Augsburg mark, c. 1495. It consists mostly of ferrite wit some areas of martensite, and slag inclusions. The microhardness averages 302 VPH. It is a low-carbon steel that has been hardened by quenching.
a
b
c
Fig. 11:
Form a fluted armour made about 1520 in Augsburg. The breastplate (a) shows a martensitic and ferritic structure (X 40) with very few slag inclusions. The average microhardness is 470 VPH. The microstructure of the backplate (b) consists of ferrite and granular carbides with few slag inclusions (X 160). The average microhardness is 290 VPH. The skull of the close helmet (c) has a microstructure that consists of ferrite and martensite with few slag inclusions (X 200). The microhardness averages 217 VPH.
28
a
b
Fig. 12:
Microstructures from the top plate of a left cuisse (upper leg defence) made in Innsbruck c. 1500. The microstructure consists of a band consisting mostly of tempered martensite and a wider band consisting mostly of ferrite with a few slag inclusions. (a) section X 40, (b) section X 160. It is a medium carbon steel (~0.5 % C overall) which has been hardened by quenching and tempering.
In Figure 11 it can be seen that even within the same suit of armour the microstructure and composition of the material used can vary a lot. Also between the various microstructures shown there is a lot of difference. Figure 12 shows the banding often observed in medieval steels. For comparison I will add some microstructure pictures of Japanese bloomiron. I have taken the pictures from a small piece of Japanese bloom, given to me by Alan Williams. The microstructures as shown are from the untreated bloom, so no heat-treating has been applied to the piece. The bloom was recently made in a reproduction process of the bloomery in Japan. The process is similar to the medieval bloomery process I’ve discussed in this thesis.
Fig. 13: Picture of the
unetched surface (X 200). Visible are the slag inclusions (black areas).
29
Fig. 14:
Pearlitic (gray areas) and ferritic (white areas) structure, along with slag inclusions (black areas) at 200x magnification. Notice the line of slag, along where the slag has been hammered out. In the upper right of the picture there is also a line of ferrite in the pearlite, with some slag inclusions.
Fig. 15: At 500x magnification, the pearlitic lamellas are well visible. Also the line of ferrite
through the pearlite can be seen clearly.
30
8.3. The cost of armour In the long period of use of armour, the cost of it varied considerably with quality as well as with other prices in general. [14] Armour made for princes, dukes and other important people could indeed cost a fortune, up to and over what a common labourer would earn in 10 years of hard work. Most of these costs would go to the gold for the gilding and to the decorative elements. On the other hand, the cheapest armour falls to a price equivalent to 1 or 2 days of labour. In the overview below prices of foot-soldier’s armour and horseman’s armour are listed, according to date and location. The equivalent in days’ wages is also given, when possible. Date
Place
9th c 1304 1388
Frankish mail Bruges mail Westphalia mail
1437 1441
Westphalia England (Milanese armour) England (Milanese) Innsbruck
1468 1527 1539 1540 1547
England (from Westphalia?) England (Greenwich) Innsbruck
1542 1551 1567 1568 1569 1571
England Landshut Milan Landshut Milan Landshut
1578 1583 1596 1599 15791630 1584 1590
Graz Landshut Graz Landshut Graz Milan England (ABC)
Cost of a Cost of a foot-soldier’s horseman’s armour armour 12 solidi £ Flem 10-20 rg 4.6 (£1.06) rg 4.3 £ 8.33 £5-£6 £7 fl 50 fl 25 (light horse) £ 0.38 £ 0.23 £8 fl 1258 (Adlergarnitur) £ 2.25 s 200 s 36 fl 220 s 6.5 fl 2550 (Rosenblattgarnitur) fl 35 fl 5 fl 8 (light horse) fl 5 fl 7 s 11 £ 3.30 £ 1.33
1600 Austria [ox = 15 fl] 1627 Landshut fl 4.5 £F = Flemish pound; rg = rhenish guilders; fl = florins; s = scudi ABC = Armourers’ Company of London
Equivalent days’ wages [6 oxen] 60-130 25 100-166 300 150 10 6 160 7500 45 1500 270 1000 48 15 000 175 25 50 25 40 80 66 26 20
You can see from the list above that armour came in all prices. You could compare it with buying a car today. A cheap car could be a Volkswagen Polo, but an expensive (sports) car, a Lotus Elise or a Rolls Royce, costs indeed a fortune. And like owning a 31
sports car says as much as ‘look at me, look how impressive my car is! And I can afford one!!’, in medieval times a fancy amour would say something similar. At some stage in the sixteenth century the battles were fought on such large scales that it became too expensive for the Lords to equip all their soldiers with armour. Also the quality of the armours had to be good enough to give at least some protection to the wearer when he was shot at with a gun, that the individual armours were more expensive. After this armour became much more a fashion thing, rather than a useful protective item. This is demonstrated by the very fancy looking armours of the sixteenth century, decorated skilfully with the strangest designs. Other evidence for this trend is that the microstructure of pieces of armour of the (late) sixteenth century show only (soft) ferrite and pearlite, and often just ferrite, whereas pieces of armour from the late fifteenth century consist most of the time of some form of hardened steel, being martensite, in combination with ferrite and/or pearlite.
a
b
Fig. 16:
Some masterpieces in the work of iron and steel. (a) open burgonet with the skull in the form of a dolphin. Made in Italy c. 1540. ferritic with a little pearlite. Less than 0.1 % C. (b) Italian burgonet made c. 1543. Microstructure of ferrite and pearlite. Carbon content varies between 0.2 % and 0.8 %. (c) Augsburg armour made in the form of a ‘puffed & slashed’ costume in about 1524. Microstructure consists of fine pearlite and a little ferrite with few slag inclusions. It has to be said that with this suit of armour came also exchange arms so it could be worn in combat[14]. c
32
To start with reviewing the process of making iron, I will begin with the charcoal. Making charcoal is a time consuming process, but necessary nonetheless. When doing so, it is important that great care should be taken in the selection of the wood that is to be converted to charcoal. To obtain good quality charcoal over and over again, the process must be repeatable so that the quality of the charcoal between loads will not vary much. This is of course in an ideal situation. I can imagine that the medieval people would not or could not be so accurate in the selection of wood, but on the other hand, they were free to cut down whole forests without getting the WWF on their backs, and an oak forest could supply lots of wood for the purpose of making charcoal, without creating the need for the people to be selective in their felling of the trees. There were some regulations about what forests can be used for chopping wood and what forests for hunting. The medieval forester was able to recognise the right kind of tree for making either hard or soft charcoal, and he or other woodsmen maintained the forests by means of coppicing, so a regular supply of good wood can be obtained from the forest. So I think that supplying charcoal for the bloomery wasn’t a big problem, but when a bloomery site would run out of nearby forests, the bloomery could go out of business because the charcoal could not be supplied anymore or at least not at a reasonable price. On assaying the ore, I would like to say that in my opinion it would be a tough job to assay every piece of ore before it is put into the furnace. In the three sixteenth century sources I have used in this thesis it is stated that the assaying should be done with a magnet or lodestone. This does not work on every type of ore, it is an useless technique for non-magnetic ores. The three authors all do not state how these nonmagnetic ores should be assayed. It is a fact that the actual iron-content of the ore can vary a lot, even within one small piece of ore. Therefore I can imagine that it is useful to assay every piece of ore before putting it into the furnace, but on the whole it can be very time consuming and it wouldn’t surprise me if medieval people put unassayed ore in the furnace to save time and work. Of course, this could mean that they were letting the furnace run to obtain iron from very poor quality iron ore, but they might have taken that risk, especially when they have found the supply of ore to be of a good consistent overall quality. Also, when breaking and sifting the ore before it is fit to put in the furnace, it is possible to take out pieces of dirt and ore that on appearance do not look like to be containing much iron. This ability would come from the everyday experience of the people handling the ore. Between the medieval blast furnace and the bloomery there is not much difference in construction. Since the purposes are similar, the materials used for building a furnace are the same. The blast furnace is higher than a bloomery, generally about 3 meters to 1 meter for a bloomery. This is one of the reasons why it is possible to get a high enough temperature to melt iron in a blast furnace. Another way of influencing the heat that is obtained in the furnace is by varying the fuel/ore ratio, and to use more or less power for blowing the bellows. Experiments with these variables are done by Tylecote, Austin and Wraith [4] and Sauder and Williams[5]. These experiments give a good insight on the actual process inside the bloomery. In medieval times the ironmasters and bloomers would have found the best way of smelting by trial and error and by experience and passed on knowledge. In their experiments Sauder and Williams found it difficult to keep the carbon content of the iron low, whereas in medieval times the carbon content of the iron produced in a bloomery seldom exceeded 0.8%. In a blast furnace only cast iron, saturated with 33
carbon, could be produced. Cast iron, being useless to make armour out of, had to go to a finery to have (almost) all carbon removed. This made for a time the malleable steels produced in a blast furnace more expensive than those produced in a bloomery. Therefore, with the invention of the blast furnace, there still was a (financial) need for bloomeries. In fact, bloomeries have been in use until the beginning of the 20th century, although in the end on a lesser scale than the blast furnaces. To make a distinction between pieces of armour made from bloomiron and pieces made from iron from the blast furnace is not always easy. In general, the iron from the bloomery contains more slag, but there are pieces of bloomiron with a remarkably low slag content, and on the other hand, pieces made of iron from the blast furnace may have some slag inclusions with the origin not necessarily in the process of making iron, but rather in the smithing process. Also the carbon content does not tell us much, other than that high-carbon steel (over 1% C and thus less suitable for making armour) most likely has a blast furnace origin. This is the reason why I have not specified the origin of the steel in the pictures in §8.2. Most likely is that the latter pieces of armour (after 1550) are made from blast furnace steel. In comparison with modern steel, both bloomiron and iron made in the early blast furnace are qualitatively inferior. This does not mean that the products made from the medieval iron are inferior to a similar product made from modern steel. It might just be that some impurities are necessary for the purpose of the steel. In a modern steelfactory the composition of the final product can be monitored. Concentrations of nonferrous elements in the iron can be regulated up to 0.01% or better. Also the influences of various elements in the steel are better understood, in fact, in medieval times they did not even know what elements were, let alone the influences of them in the steel. Of course medieval people had more practical knowledge and knew from experience how to make better quality iron, but most of the time they did not know exactly what caused which effect. Agricola [9] mentions that unslaked lime should be added to the crucible or furnace with the first load of iron ore, but he doesn’t mention why, so it’s not certain whether he or his contemporaries knew what the purpose of it was, other than that doing so would yield better iron. The skills of the medieval bloomer, smith, armourer and ironworker should never be underestimated. The armours they have made are sometimes of a very good quality, metallurgical as well as aesthetical. With their relatively limited means and ‘crude’ implements they could produce the most ingenuously designed pieces of art, just look at Figure 16b for a simple example. A lot of modern viewers, myself included, will think the actual products look too fancy and are not nice-looking at all, even ugly, but they must recognise the skill of the manufacturer. But not only the aesthetical part was of astounding workmanship. The mechanical quality was often good. Which stands to reason, because if it wasn’t, the wearer would not live through an attack, and nobody would see the purpose of wearing a heavy, uncomfortable suit of armour if it did not even protect you against physical attacks from the enemy. And good proof that the armour did its job can be seen in the fact that the armour business has thrived for over 300 years, with its peak in the late fifteenth century. In that period you can find that the suits of armour with best qualities are made. This is because the experience from ages past has given its fruits to the armourers of that period, and the weaponry was not yet of such good quality that it could sever any armour, no matter its quality. If you want to reproduce the process of making iron in a bloomery and do that in an authentic way, you are bound to run into some problems. These problems arise 34
from the fact that exact and concise data is hard to obtain. Archaeological remnants of bloomeries can tell us only so much. I have thought about some parameters that play a role in the bloomery process. Some of these parameters you can influence but sometimes there are parameters that you cannot influence, and those will lead to an uncertainty in your results. Also, you have to take into account that in a process there may be a few unknown parameters that can influence the results to a degree. Below the parameters of the bloomery process that influence the quality of the obtained bloom are listed, with a distinction between the ‘known parameters’, i.e. those that can be consciously influenced, and the ‘unknown parameters’, i.e. those that you know they exist but that cannot reliably or realistically be influenced. Obviously, the real unknown parameters cannot be listed here. Known parameters concerning the building of the bloomery o Shape and size of bloomery Material used for building bloomery o o Angle and position of the tuyère o Shape of the tuyère o Tuyère material •
•
•
when operating the bloomery o Size of the pieces of ore to be reduced o Temperature and temperature distribution in furnace during the reducing of the ore o Airspeed through the tuyère o Temperature of the air blown into furnace Ratio of charcoal to ore o o Operating time o Number of loads o Composition of loads (amount of charcoal, ore, lime and recycled material) after having taken the (hot) bloom out of the furnace o Temperature of forging o Amount of forging (time) o Number of times the bloom is reheated in the stringhearth
Unknown parameters o Quality/composition of ore o Quality/composition of charcoal o Influence of material used for building the furnace on process. When a number of blooms are obtained from the reproduced bloomery process by changing some parameters, it can be useful to make from each bloom a number of plates of certain sizes and give them different heat-treatments and then test the mechanical quality the plates. This way something can be said about the quality of the bloomiron and about the influences of the process on the quality of the final product. Some other thoughts that may be taken into account if one desires to continue along the line of this research are the following. It has been found by Bob Smith of the Royal Armouries in Leeds that the thickness of armourplates is extremely regular. 35
This led him to ask if how far back rolling mills went. Some people are convinced that the rolling of iron was rather later, but Smiths conclusion might result from there being water-powered plating forges. The earliest variety of rolling mill known as yet was the slitting mill (producing rods for nailmakers). This was introduced to England at Dartford in Kent in 1590 when a certain Godfrey Box was bought over from Liège7. It would be interesting to find out if indeed the medieval ironmakers used a roll mill or how else they accomplished such a regular thickness in the plates they supplied to the armourers (plating forges). Also, the concentration of armour production in a few places can have a reason. Is this reason linked with the existence of plating forges? Was the competition between armourers so large that only the best could survive? What is the correlation between the location of iron production sites (bloomeries and blast furnaces) and the location of armour production sites?
7
From email conversation with Peter King
36
The developments in the iron production industry have been of great importance. Iron and steel have played important roles in the history of mankind, and especially in the history of warfare. There was always a need to improve the quality of the steel used for protection against (steel) weapons and to improve the quality of the steel used for the weapons themselves. The end of the fifteenth century was a period of a gradual change from using the bloomery to using the blast furnace for producing iron and steel. The effectiveness of production increased a lot when the bellows of the bloomeries and blast furnaces became water powered. The cost effectiveness has increased a lot over time, the production costs of 100 kg of iron made in a medieval bloomery equivalent 18 days of wages for a labourer, nowadays a labourer has to work only two hours to be able to afford 100 kg of iron. The yield effectiveness is also different in a medieval bloomery than in a (modern) blast furnace. In a blast furnace the yield almost always approaches 100%, in a bloomery the yield can vary between 25% and 95%. The mechanical quality of pieces of armour made from bloomiron or iron from the blast furnace is in general less than the mechanical quality of a modern steel equivalent. The ultimate tensile strength is for the modern steel 2 to 4 times larger than for tested medieval specimens. The hardness is for the medieval steels more susceptible to drop with delayed quenching. However, the mechanical quality of medieval steel was good enough for the purpose of making armour, otherwise people would have abandoned the use of it for making armour a lot sooner than what history tells us. The microstructure pictures of medieval steel and iron show a varied quality of the various pieces, from very pure and mostly martensitic structures to mainly ferritic structures with lots of slag inclusions, and everything in between. To conclude, it can be said that in medieval times some people were able to make very good quality steel, and in general the armourers and smiths could distinguish between good and bad quality steel. Overall the quality of medieval iron and steel is too varied to make a general conclusion about the mechanical quality and composition. The skills of the craftsmen in the armour business were of a very good (practical) quality, which is hard to find in a modern craftsman.
37
[1]
R.F. TYLECOTE Metallurgy in Archaeology – A Prehistory of Metallurgy in the British Isles Edward Arnold (Publishers) Ltd 1962
[2]
K.C. BARRACLOUGH Steelmaking Before Bessemer, Vol. 1; Blister Steel – The Birth of an Industry The Metals Society 1984 ISBN 0-904357-53-8
[3]
R.F. TYLECOTE A History of Metallurgy The Metals Society 1976 ISBN 0-904357-06-6
[4]
R.F. TYLECOTE, J.N. AUSTIN AND A.E. WRAITH The Mechanism of the Bloomery Process in Shaft Furnaces Journal of the Iron and Steel Institute, May 1971, 342-363
[5]
LEE SAUDER AND SKIP WILLIAMS A practical treatise on the smelting and smithing of bloomery iron Historical Metallurgy, Volume 36 Part 2 for 2002, 122-131
[6]
The Pirotechnia of Vannoccio Biringuccio Trans. Cyril Stanley Smith & Martha Teach Gnudi Originally published in Italian in 1540 English translation first published in 1943 Dover Publications Inc. New York ISBN 0-486-26134-4
[7]
J.E. R EHDER The Mastery and Uses of Fire in Antiquity McGill-Queen’s University Press 2000 ISBN 0-7735-2167-8
[8]
Lazarus Ercker’s Treatise on Ores and Assaying Trans. Anneliese Grünhaldt Sisco and Cyril Stanley Smith (from the German Edition of 1580) Originally published in German in 1574 English translation first published in 1951 (1683) The University of Chicago Press, Chicago
[9]
De Re Metallica by Georgius Agricola Trans. Herbert Clark Hoover & Lou Henry Hoover Originally published in Latin in 1556 English translation first published in 1912 Dover Publications Inc. New York ISBN 0-486-60006-8
38
[10]
The History of the World commonly called The Natural History of C. Plinius Secundus or Pliny Trans. Philemon Holland (born 1552) Southern Illinois University Press, McGraw Hill Paperback 1964
[11]
R ADOMÍR PLEINER Iron in Archaeology: The European Bloomery Smelters Archeologický ústav AV ČR, Prague, 2000 ISBN 80-86124-26-6
[12]
R.A. MOTT English bloomeries (1329 – 1589) Journal of the Iron and Steel Institute, June 1961, 149-161
[13]
R.F.TYLECOTE The location of Byrkeknott, a 15 th century iron smelting site Journal of the Iron and Steel Institute, April 1960, 451-458
[14]
ALAN WILLIAMS The Knight and the Blast Furnace, A history of the Metallurgy of Armour in the Middle Ages & the Early Modern Period Koninklijke Brill NV, Leiden, 2003 ISBN 90-04-12498-5
[15]
WALTER PREVENIER Uit goede bron, Introductie tot de historische kritiek Garant, Leuven/Apeldoorn, 1992 ISBN 90-5350-116-9
World wide web resources [i-1] http://www.bbc.co.uk/history/games/blast/blast.shtml visited on 12-08-2003. [i-2] http://www.steel.org/learning/howmade/blast_furnace.htm visited on 12-082003.
39
A lot has been written on the subject of medieval iron production and everything that can be associated with that. Aside from the books and articles in my references you can find in this bibliography a selection of books and articles on the subject of medieval iron production, as they are written in the bibliography of Iron in Archaeology by R. Pleiner. 1963: Notes sur la commerce du fer en Europe occidentale du XIIIe au XVIe siècle. Revue d’Histoire de la sidérurgie 1, 7 – 35, 35 – 61. Bautier, R.-H.
1987: Bloom Smithing and Early Historic Smelting Sites in the Holy Cross Mountains. In: The Craft of the Blacksmith, 35 – 46. Bielenin, K.
and Voss, O. 1992: Iron Production in Viking and Medieval Times. In: Bloomery Ironmaking During 2000 years II, 31 – 43. Buchwald, V.F.
and Salter, Chr. 1991: Comparative data from iron smelting and smithing experiments. In: From Bloom to Knife, 15 – 22. Crew, P.
1995: Formation and properties of bog iron ore. Some relations between bloomery sites and topography. In: Paléometallurgie du fer et Cultures, 41 – 47. Espelund, A.
1995: A review of recent work on the archaeology of early ironworking sites in Europe (P. and S. Crew eds). Plas Tan y Bwlch, Maentwrog. Iron for Archaeologists
Karlsson, L.
1985: Cistercian iron production. In: Medieval Iron in Society I, 341 –
355. 1986: Der Übergang vom Stuckofen zum Flossofen aus metallurgisher Sicht. Ferrum 57, 28 – 31. Köstler, J.
1985: Methods of ore roasting and the furnaces used. In: Furnaces and Smelting Technology in Antiquity (P.T. Craddock and M.J. Hughes eds). British Museum Occasional Paper No 48. London, 29 – 41. Maréchal, J.R.
and Windgrove, J. 1972: Constitution of bloomery slags, Part II: Medieval. JISI 210: 477 – 478. Morton, R.
Pleiner, R. 1969b: Experimental smelting of steel in early medieval bloomeries. Pam.
Arch. 60, 459 – 487. 1993d: The Technology of Iron Making in the Bloomery Period. A Brief Survey of the Archaeological Evidence. In: Archeologia delle attività estrattiva e metllurgiche. Quaderni del Dipartemento di Archeologia. etc., sez. arch. Università di Siena, Firenze, 533 – 560. Pleiner, R.
40
