Hypothetical Norse glass bead making furnace
Over the weekend of June 29-30 I undertook a participant demo of one of many possible furnace designs constructed primarily by Neil Peterson, who is the project lead for DARC's research into Norse glass bead making.
see Neil's research documentation on the DARC web site : https://www.darkcompany.ca/beads/index.php
There is considerable evidence of the use and form of Viking Age glass beads themselves, common in most Norse archaeology. There are a number of well documented bead production sites, the one at Ribe, south Denmark quite extensive. Debris at Ribe includes starting raw glass, various production elements, flawed and broken beads.
 |
Recovered raw glass, partial products, broken beads - Ribe Museum (2008)
|
The bulk of our research work has focused not as much on the creation of duplicate patterns from specific artifact beads, but on the equipment actually used to work the glass in the first place. As it turns out, there are only a very few 'hearths', potential furnace base plates, and no surviving actual furnaces. (1)
One of the furnace designs we have repeatedly tested is what is called a 'tea pot', basically a dome shape with a top spout. Filled with charcoal and provided with bellows air, this creates what in effect is a charcoal fired 'torch', where you work the glass just above the spout.
 |
Original hypothetical layout for a 'tea pot' furnace
|
Builds using this layout have proved successful in terms of proving adequate heat, a continuing problem has been with ash from the charcoal marring the surface of the beads. Also this ash scarring is seen on some artifact beads, our attempts have most commonly have considerably more of this flaw.
Details of the specific build used over June 28-29
This furnace had been constructed earlier by Neil. Composition was a mix of powdered clay, course sand and dry shredded horse manure (a variation on our standard mix, but exact proportions unknown).
The overall size was determined by the refractory plate used as a base, cut to 35 cm diameter and about 1 cm thick. Part of the experimental test was determining the resulting burn pattern on the base after repeated firing cycles.
(images below taken after use cycle)
outside diameter = 32 cm
top port diameter = 4 cm
overall height = 35 cm (irregular)
base section height = 22 cm (rear) by 18 cm (tuyere)
chimney top height = 14 cm
base to centre tuyere = 13 cm
charcoal fill above centre tuyere = 7 cm (within base section)
 |
furnace sections inverted to show interiors
|
interior diameter at base = 25 cm
average wall thickness = 3 cm
interior diameter at base top = oval 13 wide by 14 deep (from tuyere side)
 |
A similar sized 'Aristotle' furnace, mounted much the same as the bead furnace under discussion was, showing bellows position on the sand table forge. Rey Cogswell seen operating the bellows
|
|
Generally this furnace is somewhat on the small size, at least in terms of overall height. A particular problem proved to be the location of the tuyere mounting hole. This was cut fairly high up on the furnace wall, with proportionally more room for charcoal bellow the tuyere centre point (13 cm) than above this (only 7 cm) when the base section was filled with fuel. In past small furnaces it has been found that with the (also largely hypothetical) Norse twin chamber 'blacksmith's' sized bellows used, (2) the effective 'ball of heat is about 10 - 15 cm diameter. Because the bellows / sand table forge combination that was used to mount the furnaces is intended for a much lower tuyere point (flat angle into a bellows stone at about 8 cm), it proved necessary to set the bellows up on a pair of 14 x 14 square wooden blocks. This gave the tuyere a starting down angle of roughly 15 ° down angle (rough, measured by eye). Because the bellows could not be tied down as normally done, coupled with the relative inexperience of the various participant users, the air blast was actually constantly shifting. All combined, this meant that the volume of ignited charcoal above the air blast was significantly less than with other previous furnace builds.
The charcoal fuel used was an on hand but previously untested brand : Extra Charcoal, a company based in Laval Quebec. This material has source and type only given cryptically as "Made in South America". (3) The charcoal was significantly denser and physically harder (more difficult to break) than the oak (Royal Oak) or maple (Maple Leaf) normally used in the past. This fuel had not been previously broken for size down to 0.5 - 2.5 cm pieces as done for other tests. Instead it was used 'right out of the bag', although some attempt was made to take only the smaller pieces. A smaller amount of previously sized oak charcoal was on hand, and mixed in with the Extra brand.
Results
Over the two days, new workers made/attempted roughly 8 beads, with two previously skilled workers (Rey Cogswell and myself) made another four. There was only about a 50% 'finished success' rate (whole bead removed from mandrel). This primarily due to the lack of any significant annealing process on forming (simple air cooling).
 |
Two finished beads made by myself - 1/4 inch grid
|
In use this combination furnace - and charcoal fuel was found to just barely produced the temperatures required for working the soft glass on hand. Although it did prove possible to wind up the starting core spirals, getting the glass hot enough to contract into spheres and controlling centering via gravity slumping proved tedious. Attempts to add decoration (lines and dots were applied) proved 'just possible' and difficult to control. All of this because the temperatures produced were found to be marginal for the glass working.
The surfaces of all the beads made were severely marred with ash scarring, the worst level seen for any of the (considerable) previous work using various charcoal fired bead furnaces. When the top spout/lid section was removed after a firing sequence, the charcoal was found to be covered with a very heavy ash layer. Although this was a light powdery ash, the effect visually was similar to what is seen with the use of charcoal briquettes.
The combination of the reduced charcoal burn load (volume above air input), combined with the extremely high ash content of this specific charcoal brand, resulted in less than ideal operating conditions for effective glass bead making. When combined with the unacceptable level of ash scarring, the aspect of actual production is not considered successful.
 |
The furnace base, scarring after use. Tuyere location was to middle, left hand side.
|
The most important result of this test was the effect on the refractory plate base. The total firing time during bead making attempts was very roughly 6 hours (figuring roughly 30 minutes per making cycle x 12 beads) This would have been the burning using bellows forced air. The furnace also underwent two burning down sets, one each day, when any remaining charcoal was allowed to combust naturally to 'cold' overnight.
Although there are most certainly major problems (charcoal effects, overall furnace layout) in assessing the resulting burn pattern on the base plate, hopefully this experiment may be of some value in assessing the burn scarring observed on the very few artifact bead furnace bases recovered?
Notes:
1) for Neil's collected list of described archaeological remains : https://www.darkcompany.ca/beads/arch.php?submenu=B
2) For a description of this bellows unit - and the logic behind the re-construction, see : https://warehamforgeblog.blogspot.com/2008/01/bellows-reconstruction-2.html
3) The supplier's web site (in French, via translator) does not provide any information as to the actual wood species or country of origin. An attempt was made (ongoing ? 3 days) to contact them directly via e-mail for this information. A major concern, given the vague 'made in South America' is that this charcoal is actually the result of Rain Forest clearing - a major environmental problem.
Posts
