Over the past few months the EXARN team has been working alongside three masters students from The Culture Lab at Newcastle University, Kypros Kyprianou, Katie Oswell and Jasmine Padgett.
We have written, filmed and created a documentary about experimental archaeology with the aim to bring archaeology and experimental archaeology to all audiences. We want to engage the public and academics alike in a subject we adore.
Last post I talked about the background to the project, making the glass, and building the furnace. Over the past week we began the first round of the experiment proper, with the first batch of ‘raw’ glass entering the furnace on Sunday evening, to then be repeatedly recycled every day until Friday!
The furnace was first lit on Saturday evening, and reached 1000°C at the top of the fire-pit ( thermocouple A), and 900°C just above the top of the crucible (thermocouple C) within 5 hours. It was then held at roughly these temperatures throughout Saturday night and into Sunday, being stoked with 2-3 logs every 7 minutes.
The wood used was a mixture of native hardwoods including oak, birch, ash, beech, and alder, split into logs 8-15 cm wide 20-25 cm long.
On Sunday evening the furnace was charged with glass for the first time. So that each round of recycling was as consistent and comparable to one another as possible; the glass was always held in the furnace for 12 hours, between charging and working, to ensure equal exposure to the furnace environment each time.
The temperature of the furnace was then raised to 1100°C at thermocouple A, and 950°C at thermocouple C. Again a stoking regime of 2-3 logs every 7 minutes was found to be ideal to maintain these temperatures, with additional regulation through opening and closing the top hole of the furnace (using a firebrick) in order to control the degree of draw.
This thing absolutely eats wood, with roughly 1.5 cubic meters of logs being used per day! Not only that, but due to the extremely high temperatures, particularly at the level of the fire pit (max recorded 1200°C!), the volatiles, mostly in the bark of the wood, were driven off almost instantaneously upon contact with the fire, releasing billowing clouds of soot (seriously it was like Mount Doom was erupting!) and I spent most of my time caked head to toe in black soot, although as I have learnt from previous experiments I am somewhat of a soot magnet and I will admit that no-one seemed to end up quite as covered as I did!
Whilst the glass was in the furnace temperatures were logged at all four thermocouples at 15 minute intervals and an effort was made to ensure, as far as possible, consistency in the temperature regime during each round of recycling.
As I’m sure you can imagine, maintaining a hungry and demanding furnace day and night for 5 days at a time is a LOT of work, even with working in shifts through the night it was difficult to get enough sleep!
In order to work the glass the temperature at thermocouples B and C needed to be within the range of 1000-1100°C, but ideally greater than 1050°C . On the first day the glass was worked (Monday) I unfortunately had not quite mastered control of the furnace yet and the glass had to be worked at just above 1000°C, additionally it had not been held at this temperature for long enough and as a result it was rather more bubbly than would be ideal. Because of this it was difficult to make an assessment of the workability of the glass on this day or to directly compare it to the subsequent two tests on Wednesday and Friday.
Nevertheless, actually seeing the glass come out of the furnace for the very first time and be worked was incredibly exciting, and this gave us to opportunity to refine the tests that would be used to assess the working properties of the glass on subsequent days. When I repeat the experiments next week with a fresh batch of ‘raw’ glass we should now be able to get consistent results on all three days.
Once all the glass had been gathered out of the crucible I took a couple of small fragments for chemical analysis. Whilst the rest was collected together and broken up, ready to go back into the furnace
Thankfully, from the Tuesday onward I gained better control over the furnace and was able to ensure, not only that the glass was worked at higher temperatures but that these temperatures were as consistent as possible across the next four days, with the glass being brought up to working temperatures 2 hours before it was due to be worked. Once I understood better how to regulate the temperature within the furnace, it became surprisingly easy to achieve consistent temperatures each day, to a timed schedule and to hold the furnace at a particular temperature when required, at least with the aid of thermocouples!
On Tuesday and Thursday no formal testing of the working properties of the glass was carried out (although I was able to make some of my own, much less expert, observations), and I simply gathered all of the glass out of the crucible myself, pulling it into canes in order to allow it to most easily be broken up for recycling.
With consistent glassworking temperatures now reachable, Colin was able to properly assess the working properties of the glass on day 3 (Wednesday) and day 5 (Friday). In order to do this two tests were carried out 1) blowing a small vessel (a palm cup), and 2) pulling a cane. These tests allow a feel for a) the workability b) the shortness and c) the stiffness (all somewhat interconnected properties) of the glass.
Between working the glass on day 3 (Wednesday) and day 5 (Friday) Colin observed that it was both noticeably shorter (the length of time the glass was workable after gathering from the furnace had decreased), and stiffer and that there had been an overall drop in the workability of the glass. This matched up with my own observations when pulling canes between days 2 and 4. However, he also pointed out that had glass been recycled that many times in a modern furnace, it would almost certainly be completely unworkable by that point, whilst this glass, whilst less workable than before was still evidently capable of being formed into vessels!
This definitely suggests that although it is likely that soda is being driven off by successive heating events, there is at least some degree of ‘alkali replacement’ in the form of potash from the wood fuel. This would mean that although Anglo-Saxon glassworkers would likely be experiencing a gradual decline in the workability of the glass they were using with increased reliance on recycled material, this decline would be significantly slower than might be predicted by glass recycled within a modern furnace environment, and that the wood vapour rich environment of the Saxon furnace itself essentially extended the life of the glass recycled within it!
However, with only two consistent tests so far it is difficult to draw firm conclusions, further working property tests with a fresh batch of glass next week will be needed and of course the results of the chemical analysis before then, but the results so far are very promising!
The final task for the first round, once the furnace had been left to go cold after glassworking had finished on Friday morning, was to retrieve the first week’s crucible, which despite the extreme conditions within the furnace remained remarkably intact! The crucible will be sectioned and the small amount glass left in the bottom will be analysed with a transect across it from crucible wall to centre, to help determine the degree (if any) of contamination leaching from the crucible walls into the batch.
So that’s round 1 done! I’ve just ordered a second enormous batch of wood in preparation to do it all again next week. I absolutely could not have managed 5 days on my own and so thanks once again to all the help from Marco, Ele, John, Robin, and Chloe, particually those who gave up their night so that I could actually get some sleep, I’m eternally grateful!
This blog post comes from EXARN member Victoria, you can follow her progress on our Twitter, Facebook and Instagram over the next month, as well as weekly blog posts here.
My PhD research uses glass chemistry to investigate recycling practices and access to glass, as a proxy for wider issues of trade and contact in the mid-late Anglo-Saxon period (c. 700 – 1000 AD).
Because recycling of glass involves remelting and the total loss of the ‘original’ objects from which a new batch of glass is formed, it is often an archaeologically invisible process, rarely leaving any physical traces.
For this reason chemical analysis is the primary tool used by archaeological scientists who study glass to identify and understand glass recycling in the past.
In order to do this we rely on a suite of ‘recycling indicators’; changes in the chemical composition of a glass that are believed to be the result of the alteration and deterioration of glass over successive recycling events.
Many of these changes likely take place within the furnace environmental itself, as a result of remelting and reworking of glass, these might include: contamination from the wood fuel used to heat the furnace, introducing potash, phosphates and magnesia into the glass; or contamination from iron from the tools used to work the glass. The image below shows how some of these changes are thought to take place.
However, little work has been done to test these underlying assumptions about how glass composition changes with recycling. This project is the first to be explicitly designed to experimentally test the effect of recycling on glass composition and working properties using period appropriate fuel (wood) and furnace structure.
The basic structure of the experiment is outlined below: in this blog post I’ll talk about the first two steps – production of ‘raw’ glass and construction of the furnace.
In order to test how the chemical composition of glass changes with repeated recycling, the first thing that is needed is glass that has never undergone remelting , this will be used as a baseline against which any changes can be measured . Glass that hasn’t undergone remelting is referred to archaeologically as ‘raw’ glass. Chunks of archaeological raw glass have been analysed and chemically characterised and we can then use this information to essentially backwards engineer a glass, produced in the laboratory, with the same composition.
For the purpose of these experiments the glass I made was based upon the composition of a type of ‘raw’ glass known as Levantine 1. Levantine 1 glass is composed of approximately 70% silica, 15 % soda, and 9% lime, along with smaller amounts of alumina, magnesia, potash, iron oxide, manganese dioxide and titania.
In order to replicate this glass in the lab I used pure laboratory reagents to ensure that the final composition of the glass is as close to the intended one as possible and that it contains no unknown contamination.
Once the powdered glass mix had been made up it was measured into crucibles and heated in a modern electric furnace up to 1200°C, in order to allow the mixture to fully vitrify into glass!
The pictures above show the finished glass, ready to be broken up and used in the wood-fired experimental furnace! Although some of it looks purple and some of it green/turquoise (and some a rather gorgeous looking gradient between the two) they are all compositionally the same and the differences in colour are just caused by how oxidising the furnace environment was during firing (purple = more oxidising).
Building the furnace:
And now to the (even more) fun bit! Building the furnace!
The design for the furnace was largely based upon the fantastic work by Taylor and Hill (2008), and their experiments in the reconstruction and use of Roman wood-fired glassworking furnaces. We don’t have a huge amount of information with regards to what Anglo-Saxon glassworking furnaces might have looked like, however what we do have suggests that they were probably not dissimilar to the designs proposed by Taylor and Hill for Roman furnaces. The other key thing about the designs are that they are tried and tested; we know they work!
Since my experiments are focused on recycling and not furnace design there seemed little need to reinvent the wheel. The ‘small’ furnace was chosen to replicate based primarily on it requiring fewer resources to construct and run (i.e. required less clay and less wood, the two main expenses of the experimental phase of the project!), than the ‘pot’ furnace design.
In a stroke of good luck a previous experimental project to reconstruct a Roman pottery kiln had been carried out at Jarrow Hall, where my furnace has been built and the brick-lined firing chamber was still intact. This enabled us to save a considerable amount of time in the construction as we were able to begin to build the superstructure of the furnace directly on top of the preexisting structure (after giving the firing chamber walls a sacrificial skin of clay).
The furnace superstructure was constructed using a daub mixture consisting of a heavily grogged fireclay (it is in fact sold as a ‘pizza oven’ clay, a fact about which my Italian EXARN colleagues were delighted!), mixed with approximately 30% by volume of chopped straw (one of the many perks of building at a working ‘Anglo-Saxon’ farm is that there is no shortage of straw!), the straw helps to give the clay green strength during construction as well as aiding with insulation of the structure once its in use.
After building up the walls of the firing chamber about 15 cm above ground level, it was time to construct the siege, this is where the crucible will eventually sit once the furnace is in operation.
The siege was constructed using left over fire bricks from the remains of the pottery kiln, half built into the walls of the furnace, these will hold a tile on which the crucible will sit.
The next two stages of the main construction were the creation of an archway over the stoke hole and building the furnace up above the height of the siege to a dome on top.
The weather was against us on Wednesday, with pouring rain all day, however somehow despite this the stoke hole was completed.
On Thursday, with the weather thankfully on the turn myself and fellow EXARN member Alicia were able to complete the dome.
The furnace is designed to have a removable top, since the gathering hole must be kept relatively small to prevent excessive heat loss, this is the only way to get a crucible in an out of the furnace! And in order to create this we laid down a layer of paper between the main body of the furnace and the start of the construction of the dome to separate them from eachother.
The very last thing left to do before introducing fire to the furnace was cutting in the gathering hole, through which molten glass will be removed from the furnace on the end of a blowpipe, and the heat from which will be used to reheat the glass during working. As well as a ‘warming’ hole into which blowing and punty irons can be placed to keep them hot.
And finally: on Sunday evening the furnace was fired for the very first time! Although it wasn’t brought up to the temperatures that will be needed for glassworking (approx 1000 – 1100 °C), this firing was primarily designed to ensure that the furnace was completely dry before being ramped up to glassworking temperatures to ensure that the structure doesn’t crack or fail due to thermal shock.
Thermocouples have been built in to the walls of the furnace and these will monitor hot face temperatures throughout the course of the experiment, and help ensure that the working temperature of the glass is kept as consistent as possible. During the pre-firing they were able to show that the furnace was functioning as intended with most of the heat concentrated at thermocouples B and C, just above and below the height of the siege.
Well that’s it for Week 1! At the end of this week the furnace will be lit in earnest with the first batch of glass going in Sunday evening, to start the cycle of being worked, tested, sampled and of course recycled next Monday evening!
A great deal of thanks go out for all the help I’ve received so far from fellow EXARN members (Marco, Ele, and Alicia), as well as Newcastle Masters student John Person and his partner Robin (who has also made the amazing crucibles that will go inside the furnace), my husband, and my supervisor Dr Chloe Duckworth who has been invaluable at every stage, but particually in running in and out of the lab firing glass and crucibles whilst I was a Jarrow Hall building the furnace!
If you are interested in helping to run the furnace over the course of the next few weeks please send me an email to firstname.lastname@example.org.
Last week three of us travelled to Trento, Italy, for this year’s Experimental Archaeology conferfence (EAC11) by EXARC, hosted by Trento University.
The two day conference was packed full of fascinating archaeology from archaeologists across the globe.
Here is an image of all the attendees, some came from as far afield as Australia and America!
We listened to presentations on a variety of experiments from sword shining, to bread ovens and projectile tests. Regions and time periods also varied, providing a range of interesting and thought provoking talks.
The three of us also presented our research at EAC11. In the first panel Amber presented part of her PhD research, using experimental to test dated interpretations of Early Bronze Age stone battle-axes.
In the same panel Alicia presented her ethnographic experimental research which aims at understanding fuel use in Viking Age Iceland.
Marco also presented his poster on the copper smelting tests he carried out as part of his PhD. You can see information about how these went in previous blog posts.
We love presenting our research as part of EXARN. It was brilliant to do this making fellow experimental archaeologists.
Last week EXARN members Eleonora Montanari and Victoria Lucas (chair) presented papers about their research for Newcastle Antiquaries Archaeological Glass night.
Ele spoke about her PhD research: Gendered Stories: Constructing identities through glass beads in Iron Age Italy and Iberia
She is currently carrying out ongoing experimental tests to understand the formation of wear on glass bead during use – this means that a number us get to wear pretty glass necklaces!
Our chair, Victoria Lucas, also presented a paper about glass for her PhD project: Looking Through the Glass: glass chemistry as a window on Early Medieval innovation, recycling, trade and contact, AD 700-1000
Victoria is about to start a long series of glass production and recycling experimental tests at Jarrow Hall. We are so excited to see how they turn out. Stay tuned for updates on their progress!
Early Medieval Archaeology Student Symposium 2019 (EMASS 2019) – Durham University and Newcastle University
This weekend just gone a student symposium held between Durham and Newcastle Universities happened. The organising team included EXARN member and Chair Victoria Lucas who also presented a paper. Victoria and EXARN member Alica Hart Sawyer brought Experimental archaeology to Early Medievalists with papers about Anglo Saxon glass production (Victoria) and the use of fuel in Iceland.
The 3rd and final speaker of the morning’s session was Alicia, discussing “Recycling dung: manure and fuel in resource-limited medieval Icelandic settlements”.
Victoria gave the final paper of the material culture session and of EMASS 2019! Victoria’s paper is entitled: ‘looking through the glass: recycling, experimentation and innovation in the mid-late Saxon glass industry’
Take a look at our events page to see what else we are doing, you might even be able to join in!
Hi everyone, welcome to our first blog post of 2019!
Today’s post comes from Eleonora Montanaria new EXARN member for this year. Ele’s research aims to shed new light on the role of glass beads in the construction of personal and group identity including gender, age, status, and ethnicity in Iron Age Spain and Italy.
As part of this research she is carrying out a series of use wear experiments on glass beads!
Here’s what she has to say:
The bead wearing experiment has finally started! A group of lovely volunteers will be wearing necklaces, each with two glass beads and an amber or copper alloy bead, for three years (gasp!). The aim of the experiment is to see if any recognisable traces of wear by use will develop on the glass beads, and how these traces might differ from what are observed on beads from archaeological contexts.
This experiment is an integral part of my research, and for this reason, each element of the necklaces reflects the most common types of materials present in ornaments in Spanish and Italian Iron Age burials. The glass beads have the same broad composition of the archaeological specimens of the first millennium BC, which means they are softer than modern ones. I should add as well that I made the glass beads myself, and I feel quite proud!
You can follow Ele and her experiments @EleVenerabl on Twitter, and we will keep you updated with more on our blog too!
I hope that’s whetted your appetite for experimental archaeology, we’ve got some big projects coming up in 2019 (including more glass and more beads!) so be sure to look out for more on those in the near future!
Last weekend EXARN hosted the first Experimental Archaeology Student Symposium (EASTs) at Newcastle University.
What a success! All speakers were very impressive and of a high quality. Their papers ranged from glass knapping to charcoal production and iron smelting experiments. Theory, practice and reflection provided interesting thought on experimental outcomes, objects biographies and interpretations. We would like to thank our speakers for helping to create such an interesting and engaging event.
We couldn’t have asked for a better conference and we a pleased for it to be continued next year at Sheffield University. We are so excited!
We also received some brilliant feedback from Dr Chloe Duckworth; we are confident the success of EASTs will continue for years to come.
The conference excursion continued the successful event. It was so much fun! We spent the day at Jarrow Hall Anglo-Saxon Farm and Bede Museum where EXARN have carried out a number of out smelting experiments. One of our speakers, John Pripani, held a glass knapping workshop for us. He taught us to use the base of glass bottles to fashion a barbed and tanged arrowhead using traditional Bronze Age flint knapping techniques, such as pressure flaking.
The whole group successful learnt the knapping and pressure flaking techniques resulting in a knapped glass arrowhead for each one of us. Many of us had previously found flint knapping very difficult. But, with John’s guidance, we were able to understand how the material flaked and the types of pressure and motions needed to remove flakes. We learnt that we could practice on so many different materials which reacted in the same way as flint, including glass and toilet ceramic piping. Many of us are now sourcing materials to continue to build on the skills that John helped us develop during this workshop.