Smelting your own iron by digging up ore, clay and sand, and turning them into a bar iron using heat and hard work alone, is a transformative process which carries much mystery and appeal. The whole act of taking what is essentially dirt and transforming it into iron and steel, which can then be turned into objects both useful and beautiful, still captivates the imagination of artisans, archaeologists and chance observers alike.
I have already touched upon the topic iron smelting in the context of experimental archaeology in my article ‘The Smelters Tale’. But over the last year I was approached by many a young and enthusiastic smith, seeking advice on their early attempts at smelting iron.
Therefore it is high time that I put down some of the realisations gained over my short career as a historical iron smelter. By doing this I hope to help a few people to get started in this mad pursuit, as well as to dispel some of the unnecessary mystique and technical terminology which sometimes plagues this field.
I would like to state that I acquired this knowledge through a combination of having the honour of working with a few people far more experienced than me, as well as devoting a considerable amount of time to my own experiments.
I have been actively working with ancient iron production over the last two years. During this time I have taken in part in several iron-smelting symposia, such as the annual meeting of iron-smelters in Adamov, Czech Republic, and Fête du Fer in Neuves Maisons, France. I have also held or helped teach iron smelting workshops at living history events. Currently I am also working on a postgraduate thesis involves both iron smelting and blacksmithing.
Since I come from a background in archaeology and living history, this guide will revolve largely around how this was done in the past. The methods described in the forthcoming parts of this guide will urge you to, at least to a certain degree, emulate the practices of the ancients. The principles outlined here can be adapted for use with modern materials, such as fancy refractory mortars, thermocouples and whatnot. Nevertheless, these are not at all necessary. You may find that just like for most (pre)history, using materials which can be dug up in your vicinity might be the cheapest and quickest option.
Staying within the limits of what was historically possible in a certain time and region gives you a new understanding and respect for the work of the ancients. It should keeps this in a format which will be of equal interest to artisans, experimental archaeologists and those dealing with living history. In the end, you should of course adapt these basic principles to your given situation and aims.
I will not bore you to tears with details about every furnace subtype and variation on the practice interpreted by archaeology. Instead I will outline a general model, since I presume that you are primarily interested in just making some iron. You can then adapt this to a localised context, should you need to.
If you are interested in the details of historical iron production and the archaeology connected with it, then a good reference library is provided in Radomir Pleiner’s (2000) Iron in Archaeology: Early European Smelters, which is his collected life work on iron smelting across Europe published in English.
Lastly, before we dig into the details of iron smelting, let me clarify one last thing…
A guide like can be a valuable resource, but especially in your early steps, it is no replacement for first hand experience under the guidance of someone who knows what they are doing.
You can certainly learn this on your own by using knowledge gleaned from books, blogs and forums, but you will also need to allow for a lot trial and error in the beginning. Witnessing and taking part in a successful smelt, learning about common problems – how to recognise and solve them on the go – is a large part of this craft. If you are just starting out, then I would urge you to come in contact with someone who already has some decent experience with this. There is no need for you reinvent the wheel and repeat every mistake. This might involve taking a few days off for a trip to a gathering of smelters, where you should offer your help in building a furnace and running a smelt. Observe, ask questions and adapt to your situation. Thus you will be rewarded.
Since I believe that an understanding of the process and reasoning behind it is helpful in trying to learn a technique, the first part of this guide will give you a general theory of mechanics of bloomery iron smelting. In the following episodes I will delve into the more concrete practicalities of of building a furnace and running a smelt.
The Bloomery Process Explained
The Basic Mechanics
The smelting, or reduction, of iron ore is a thermochemical reaction, wherein the iron oxides are reduced to metallic iron. Impurities in the ore are melted out of the ore in the form of slags – most commonly iron silicates – while the iron itself is never truly molten. Instead the product of the smelt is a spongy mass of iron and slag known as a ‘bloom’. This then needs to be consolidated by hammering while the bloom is at a (white) welding heat. In doing so, the slags trapped in the bloom are squeezed out and the gaps are welded shut, resulting in a (hopefully) solid piece of metal.
In other words, the process of iron production is the transformation of rusty rock (iron oxides and silicates) into solid metal. This is made possible in an environment where there is sufficient heat and an excess of carbon. Such an environment is offered by the smelting furnace, which is full of burning charcoal.
That at least is a very basic and cursory glance at the process. Now let me break this down into its components and explain each in turn…
The raw source of iron is iron ore. Iron ore comes in the form of rocks or other sediments. These are composed of iron oxides, which we try to reduce, and the unwanted ‘gangue materials’ which take the form of silicates and alumina, in other words the ‘rock’ part of the ore.
Besides oxides, which are oxidised iron, iron ore also occurs in the form of sulphites and carbonates. I will focus on iron oxides, since this is what I have the most experience with and what you end up smelting in any case.
All types of iron turn into an iron oxide once they are roasted as part of the preparation for smelting. Ore is easily roasted by laying it on pyre such as a decent bonfire, where it heats up together with the embers. Roasting drives off excess water and makes hard and dense ores brittle enough to be crushed. Also it turns all ores which are not a simple iron oxide into one. Therefore sulphides such as iron pyrite (FeS₂) and carbonates such as siderites (FeCO₃) may be used, provided that they are thoroughly roasted first.
A note on Phosphorous – Besides iron, there are other mineral oxides which may occur in the ore and may impact the smelting process and the resulting bloom. The most important of these is phosphorous, which can be found in high concentrations in some bog ores. About a quarter of the phosphorous in the ore will end up in the resulting iron. Small amounts of phosphorous will harden iron, but also make it brittle and hard to work at higher concentrations. Small amounts of phosphorous are useful for pattern-welding, since phosphoric iron shows up brighter than plain iron or steel when etched. This is actual how the original pattern-welded knives, swords and spears of the migration, Viking and medieval period were made. But beware, too much phosphorous will make the iron too brittle and prone to cracking when forged, thus making it useless to the blacksmith.
Here are some of the most common ores:
Limonite (Fe₂O₃.2H₂O) – A form of hydrated iron oxide, also sometimes called brown ironstone or brown haematite. Quite common in many parts of the world and varying in iron content from very modest to quite rich (up to 60% FeO). Widely used in certain regions. Usually reduces easily. Benefits from some roasting to make crushing easier.
Bog Ore (Fe₂O₃.2H₂O) – Essentially another form of limonite. Very common in wetter parts of Europe, since as the name implies, it is formed in bogs. The go to ore for most of prehistory and early history in the areas where it was available. It usually quite crumbly once dry, and therefore full roasting is not necessary, but it does not hurt. Reduces easily due to the porous nature of the ore. It has a variable iron content, which generally ranges from low to moderate. In some cases it contains a lot of phosphorous, in which case it is usually best to look for a better source of ore.
Haematite (Fe₂O₃) – Haematites are rich ores, containing up to 60-70% iron. Prised by those who like to focus on steel production. Being hard and dense, they need a good roasting and crushing. The high proportion of iron to the amount gangue material means that less slag is produced compared to the ‘poorer’ ores. Sometimes you might actually need to add flux in the form of sand or slag from previous smelts to aid in bloom formation.
Magnetite (Fe₃O₄) – The richest and densest ores, containing above 70% iron. It got its name because it adheres to a magnet in its natural state. Rarely encountered in rock form on the surface, instead the eroded rock gets deposited on shores in the form of black ‘iron-sand’, which can be easily identified if you happen to have a magnet with you. In this form it was traditionally used by Japanese smelters to produce tamahagane steel, which is famous for its continued use by Japanese swordsmiths.
Siderite (FeCO₃) – A common iron carbonate, which occurs in clay layers of coal measures, shale layers, limestones and so on, where it forms ore lenses. Typically contains no sulphur or phosphorous and relatively rich in iron. Works well in a bloomery, provided that it is roasted well first, thus turning it into an iron oxide and making crushing possible.
Forge Scale – The flakes of oxidation which fall off hot iron while forging are not exactly iron ore, but technically speaking, this is a form of magnetite, and can be smelted. When charged into the bloomery furnace together with sand it can produce some good, usually steely, blooms.
Do not make the grave mistake on judging the quality of the ore on iron content alone. People often get deceived by the notion that a higher iron content is necessarily better. What you are looking for is an ore that reduces well, forms a slag which flows well, separating from the bloom with ease, and produces a bloom that consolidates nicely, without unnecessary cracking or other notions of misbehaviour. Some of the haematites and magnetites can be quite annoying to work with, often requiring fluxing in the form of sand to aid bloom formation. On the other hand, some bog ores might contain alarming amounts phosphorous, resulting in iron with a tendency to crack when forging.
In other words: Some very rich ores can be sometimes quite cheeky, while some lower yielding ores may give you beautiful iron, but you will need to smelt more ore to get a decent amount of iron from them.
In the end, you will learn to adapt and make best use of whatever you end up using. Be it bog ore, haematite or siderite. For example, I spent a large part of spring 2016 figuring out how to use the some very low-yielding limonite from southern Slovenia, because I happened to be given a large quantity of it. After several tries and some heartache, I managed to produce some very nice blooms out of it.
The smelting of ore requires a source of carbon in the form of a fuel which burns hot, clean and ideally leaves little residue. With very few exceptions, this has always been wood charcoal.
There are some regional references to the possibility of peat charcoal having been used, and also the use of pine wood instead of charcoal, but let us concern ourselves with charcoal for now since it is best tested and most easily available.
The type of wood used for charcoal production depended on the region, with various hardwoods, such as beech, hazel and oak, being used. Softer woods such as pine were also used in regions where this was the predominant wood species. Both can be used. My general recommendation would to get a hardwood charcoal of good quality which does not crumble too much.
The bloomery furnace is a purpose built feature which makes the process of reduction possible. There are many variations on the theme used depending on the region and time frame in question, but essentially they all serve the same purpose. They provide an enclosed space which contains the fire, have an air intake near the bottom, some space bellow the air intake and an opening to feed the furnace at the top. Let me again break this down…
The enclosed nature of the furnace has two principal functions. Firstly, by containing the fire it allows for a larger hot-spot to be formed. Secondly, it by containing the fuel it creates an environment with an excess of fuel. This leads to incomplete combustion of the charcoal, which in turn produces a lot of carbon monoxide. These gases then travel up the furnace, reacting with the ore, thus reducing it.
Furnaces therefore have some kind of shaft which provides this kind of enclosed environment, and this shaft is open at the top to allow the gases to escape and the furnace to be charged with ore and charcoal. This shaft is typically built of a mixture of clay and sand, often with some kind of organic fibre (e. g. straw) added. The shaft can be thick or thin, tall or short. It can have straight sides, forming a chimney, or it can taper towards the top, or in some cases it can start approaching an open dome in shape. The shaft can be free-standing, or it can be built into a hill bank.
The fire needs to be fed not only fuel, but also with a blast of air to keep it burning hot. Controlling the air intake is the principal way of controlling the temperature and therefore the nature of the smelt. It can take the form of a blow-hole, which is a narrow opening in the wall of the furnace, or a tuyere (blowpipe) may be used, which extends slightly into the interior of the furnace.
The position of the air intake will determine the location of the zone where the fire burns the hottest. From there on the heat and reducing gasses will travel upwards. Since it is impossible to force heat downwards, then naturally the air intake will be positioned somewhere near the bottom of the furnace.
The air is commonly pumped into the furnace with some form bellows and all of my instructions will relate to this method. Alternatively, there is ethnographic evidence for the use of ‘natural draft’ furnaces. These have very tall shafts, which cause sufficient amounts of air to be sucked into the furnace due to the ‘chimney effect’.
Space Beneath Air Intake
Although the air intake is positioned in the lower portion of the furnace, it is not found right at the bottom. Smelting produces waste material in the form of slag, which flows down out of the bloom and below the hotspot, where it starts to solidify. This means that there needs to be some space between the bottom of the furnace and the air intake, which allows the slag to collect without clogging the air intake and subsequently stopping the smelt.
This depth will vary depending on ore used, tradition employed, and the technological decision whether the slag is to be removed during the smelt tapping a hole in the shaft.
On iron-rich ores produce less slag, while tapping slag means that it can be removed once it starts rising to high. This means that much more than a shallow bowl won’t be needed. Smelting a dirty bog ore will result in a lot of slag being produced.
Reduction is a process wherein oxides are reverted to metal. It starts soon after the ore starts descending down the furnace shaft, as the ore is exposed to hot carbon monoxide.
The iron oxides react with the hot carbon monoxide, where the monoxide strips the oxygen off the iron to form carbon monoxide, while leaving us with metallic iron. This can be expressed in the form of the following formula:
3Fe₂O₃ + CO = 2Fe₃O₄ + CO₂
Fe₃O₄ + CO = 3FeO + CO₂
FeO + CO = Fe + 3CO₂
This is a very simplified, unidirectional, way of looking at this, but it suffices to illustrate the basic principle. The reality is that we are burning charcoal and ore in a shaft made of muck and straw, while utilising what is probably at least a somewhat inconsistent air-blast. Therefore the chemical processes are more complex, with a tug of war between reduction, re-oxidation, carburisation and decarburisation, while the walls of the furnace are partially melting, adding all kinds of elements to the ones already present in the ore. This is one of the reasons why the iron emerges as the heterogeneous mess that it is. In other words, it is one dirty deed. But let us keep moving on.
At this point, the ore has turned into a mixture of metallic particles trapped in between the unwanted material. Once the reduced chunks of ore start moving down towards the hot zone at the lower part of the furnace, fun things start happening. The temperatures around the blow-hole/tuyere exceed 1250°C, with the hot spot which can hover in the 1400’s. This has two effects:
- Separation of Iron & Slag – Above 1100°C (ideally) the gangue materials start melting and thus separating from the metal in the form of a glassy substance known as ‘slag’. This is for the most part an iron silicate, but other factors are also present, such the ash, the melting walls of the furnace and other chemical elements present in the ore.
- Consolidation – In the hottest part, the grains and bits of iron are at very good welding heat. Therefore they start joining on contact consolidating consolidating into one lump (ideally).
Generally speaking, a higher temperature will lead to better separation of slag from the bloom and a denser, more consolidated bloom.
Therefore, if you are having trouble with a smelt where you do not get one large bloom, but instead small bits scattered around the furnace, then this can often be solved by increasing the temperature of the smelt. This is done by pumping air into the furnace with a greater fury. But there is also a cut-off point where one can not simply keep increasing the temperature. If you push this too far, you run into the danger of actually destroying the furnace. Often the ideal temperature is just a bit before everything starts melting (provided that your clay and sand are good). The other danger is that at these very high temperatures, the iron might start absorbing more and more carbon from the surroundings, eventually turning the bloom into cast-iron.
This brings us to the mechanics behind the interaction of temperature, slag and carbon content of the final bloom.
Iron & Steel
Iron is a base chemical element, while steel is an alloy of iron and carbon. There are several standards by which these are named and distinguished. Nowadays all ‘iron’ which sold is termed as steel (e. g. mild, medium and high-carbon steel), because chemically pure iron is actually quite a rare thing.
Nevertheless, historically the terms and iron and steel were used to describe two types of metal used by blacksmiths. As blacksmiths we often distinguish iron and steel based on the fact that steel can be hardened, while iron cannot.
An iron bloom typically has a varying carbon content, with some bits higher and others lower in carbon. Several factors will determine whether the final result of a smelt is mostly low carbon (iron) or high carbon (steel).
First of all, once the iron ore has been reduced with iron, the process of iron reacting with carbon continues into the process of carburisation. Iron heated above 900°C again starts binding with the carbon monoxide in the hot furnace. Therefore the same carbon rich and hot environment which produced the iron in the fist place, now turns it into steel.
In case you are fond of chemical formulae, then this process can again be expressed in the following manner:
3Fe + 2CO = Fe₃C + CO
Again the variable environment of the furnace leads to parts of the nascent bloom also de-carburising – the steel oxidising back into iron.
The absorption of carbon is more rapid at higher temperatures. Therefore hotter smelts, with a larger hot-spot are more likely to produce steel, than slightly cooler (but still hot) runs. Very hot runs, using a very strong air-blast, can start producing cast iron as a part of the smelt. Such was sometimes the case with the late medieval water-powered bloomeries.
A richer ore tends to produce steel. Rich ores, such as magnetite, produce less slag in the smelt. Slag play a role in the smelt by coating the bloom, thus shielding it from the environment and aiding the consolidation. Because there is less slag, more of the iron is exposed and therefore absorbs carbon more easily. If a both very iron-rich and fine magnetite sand is used, then this means that the small particles have even more surface area, which is free to react with the environment.
Carbon content can be controlled with with the ore to charcoal ratio. Ore and charcoal are typically charged at a weight ratio of 1:1. By increasing the amount of charcoal charged relative to the amount of ore, then this not only increases the amount of free carbon available. It also raises the temperature somewhat and gives each charge of ore more time to burn down. This means that the ore has more time to reduce and pick up carbon.
You should also keep in mind that high-carbon blooms can sometimes be more difficult to process. They can be more prone to cracking, have a narrower range of temperauretes at which they can be worked, and other forms of misbehaviour. You can always carburise already refined bloomery iron, thus turning it into steel. Therefore if you want to make blades, that does not mean that you absolutely need to smelt steely blooms.
After The Smelt – Processing The Bloom
You should also keep in mind that smelting the bloom is only half the process. Afterwards the spongy mass of iron still needs to be hammered into a bar and possibly folded a few times to further compact it. This requires a considerable amount of familiarity with the hammer, forge and the techniques of forge-welding.
Therefore, just as with the process of smelting, there is a craft to bloom-refining. An inexperienced smith may quickly fragment a bloom instead of consolidating it. These skills are improved only through practice and learning. Having a mentor in the early stages is again beneficial, since they will point out the pitfalls in front of you. Nevertheless, I will address some of this in a future post as well.
A Conclusion to Part I
“Then the blacksmith, Ilmarinen,
Thus addressed the sleeping iron:
Thou most useful of the metals,
Thou art sleeping in the marshes,
Thou art hid in low conditions,
Where the wolf treads in the swamp-lands,
Where the bear sleeps in the thickets.
Hast thou thought and well considered,
What would be thy future station,
Should I place thee in the furnace,
Thus to make thee free and useful?”
(The Kalevala, Rune IX. John Martin Crawford (1888) translation)
Having read this, you should now understand the basic processes behind the act of smelting iron. If some of the detail about chemical formulae bores you, then you may forget it. The ancients did not have them either.
Nevertheless, this introduction was written from the perspective of somebody who engages with the topic practically. May it serve you as a useful vademecum in the pursuit of turning muck into iron.
Understanding the process will enable you to not just blindly follow someone’s instructions, but to reflect upon them and adapt your method according to the situation and materials at hand. Thus you will make us of models instead of following recipes.
Now that we got the necessary formalities of theory out of the way, let us move onto the practical steps, where I shall explain everything from preparing the raw materials, to building the furnace and running the smelt.
Any questions? Do post in the comments bellow?