This related to the descriptions of a wood fired bloomery furnace and a re-melting hearth, by Evenstad in the late 1700's:
Anyway, that looks like a very cool experiment to try.
I still have my doubts about the wood converting, as well as reducing the ore, with one load. I would think, it would take some continuous feeding of wood and ore. I don't like that there seems to be no slag tap.
I was working within eyeshot of Ole Nielsen at the Heltborg Symposium (2008). Unfortunately, I had shot video of his process, via a camera which destructed and destroyed the footage. So I have only a few still photographs of the build and more importantly of the furnace in action.
Those who have run a smelting furnace have certainly noticed that the hot spot over the tuyere will consume the fuel on that side faster. This can result in a slope to the charcoal at the top. (Why you end up charging in a horseshoe around the far side of the furnace in the later stages of a smelt.)
The version of the Evenstad that Ole constructed was something like 2/3 scale. It was roughly a meter wide and a meter deep. It did in fact appear a bit wider than deep, so not quite the same proportions as the manuscript illustration shows. There was also not as pronounced difference between the top and bottom diameters - so in turn not that more trumpet shape seen in the illustration.
I suspect both of those aspects are important.
What I observed was that the surface of the raw wood packing was making the same slumping pattern over the tuyere side. So in effect, the raw wood from the back was travelling along a diagonal line - which certainly was giving it time to transform into charcoal. This in effect utilizing the waste heat from the smelting process.
I agree with what Mark comments on the lack of slag taping. There may be something going on with the size of the bellows system - related to force of blast and volume of air. Most of us in North America, if working with European types furnaces, are using the 'Sauder System' of high volume air. This puts the bloom below the tuyere, towards that side. The Europeans are often working with lower air volumes. This appears to move the bloom higher in the furnace (and produces a more lacy, smaller bloom as well.)
As you suggest, more direct experimentation with the Evenstad system is needed to really understand it.
We also need to remember that those illustrations are certainly 'ideal' - and may not actually match the functioning furnaces in fine detail!
My understanding from the manuscript is that this system makes a series of smaller blooms, in a continuous process running over several days (!). It describes pulling away excess slag (the bowl) after each extraction. Remember that this process is linked to use of a primary bog iron ore as well, which depending on source, might contain a lot less silica than the rock based ores that those of you from the eastern USA are used to using.
Under 'Conclusions' Evenstad says : "Smelting 24 pars of ore gives 36 lbs of raw bloom. Output should be 150 lbs per day. ... Three men can do five smelts a day."
Espelund notes that his own experimental trials suggest 4 hours per cycle, so that is a 24 hour day. Running the math gives us roughly 30 lbs for each bloom / 15 kg. So large, but not immense. Espelund describes a test he ran in 1991 had 15 kg of ore produce 6 kg of bloom. (Thats a very respectable 40% yield, but of course the ore quality is critical.)
Mark continued - Referring to 'Iron Production in Norway' - Arne Espelund
Not overly surprising. I think the Espelund book was almost 'vanity press', and I doubt all that many were printed. The publisher is Arketype in Trondheim Norway. The frontpiece states "The printing has been financed jointly by the Norwegian Metallurgical Society, INFACON, and the Historical Metallurgy Society in England" . Maybe some leads there?Is there a full description of the operation of it in that book? Which is no where to be had, at any price that I have found in the world.
I managed to score a photocopy directly from Arne when we first worked together at a workshop for Parks Canada in 2001. (Which got me started in this insanity.)
Iron in Norway devotes about 20 pages to the manuscript and Eslplund's comments on it. There is another 10 pages of background material.
What might surprise you (??) and is most specific to our starting point here :
The actual reference to a re-melting hearth in the Evenstad is very short and not overly descriptive. Its only a single paragraph!
I've scanned that specific section and attached it to the bottom of this.
The whole bladesmithing community owes a huge debt to Skip Williams, for researching and prototyping the first 'Aristotle' furnace. Then to Lee Sauder, for hosting the Early Iron group at Smeltfest 08 where we were introduced to this prototype. Jesus Herandez, Shelton Browder and Steve Mankowski, (and myself) for the week we spent figuring out how to make the system work and control the variables.