Vinland 4 Smelt - Draft Report

On June 12, the Norse had come to Wareham and undertook an iron smelt.

The framework for the smelt was inside DARC's full dress rehearsal for our major presentation at L'Anse aux Meadows NHSC coming up in August. As most of you are sure to know (?) this is the site of the first iron production in North America, by the Norse circa 1000 AD. This smelt was undertaken using all Viking Age equipment and methods and in historic clothing (other than modern safety equipment).

Duplicating the physical layout of the 'Furnace Hut' uncovered at L'Anse aux Meadows.
Pierre on the bellows, Darrell checking the tuyere, Ken adding charcoal and ore.

Keeping to the archaeology of L'Anse aux Meadows, the furnace was built with 5 cm walls, made of 50 / 50 clay and sand mix. I had made up the first course of clay a bit too wet, so the whole furnace had slumped. The major effect was to produce a 'pot belly' shape. Although the furnace started at the 20 - 22 cm ID suggested by the archaeology, the sagging expanded the diameter right at tuyere level to closer to 28 cm. This was sure to effect air requirements. The air was provided by that special 'smelt sized' Norse style twin bellows I had made up. I think in fact that we had exceeded the amount of volume it could dependably produce. Certainly we have all seen that any reduction in air volume below the Sauder and Williams Magic Numbers results in reduced bloom yields.

We had big problems with cracking. I *did* expect this from not using any kind of organic material in the wall material. Generally we were able to reduce massive voiding of working gases by placing stone slabs around the outside of the furnace, and packing the gaps with 50/50 sand and ash mix. The problem with that stuff is that if too much of it runs into the cracks and into the furnace itself, it makes a lot of extra slag. We managed to get through the bulk of the smelt with only a single tapping off however.
The 'disaster' happened right near the end of the smelt. The last heavy charge (about 2 kg worth) had been added, and we had just started the burn down. The top of the charcoal had dropped down maybe 10 cm from the top, so that large charge might have gotten to just above tuyere level. There was a huge self tap out of one of those large cracks, I'd estimate at least a half the available liquid slag. Must have been at least a litre, maybe more (think a milk carton's worth).

This is where being tired did not help. I tried picking up the pieces of slag and re-cycling them back into the furnace. with the addition of a bit more charcoal to cover. The loss of slag had also significantly dropped the internal temperatures (of course!). I certainly did not help that by putting the 'just barely solid' - but colder, slag back into the furnace. We kept going with the burn down, till we were maybe 10 cm above tuyere and slag bowl.

Initial working of the bloom.

Of course when we pulled the bloom that we had, it really was not hot enough to forge effectively. Really it was closer to a bright orange than at welding temperature like it should have been. What we ended up with was a rough brick shape, but more mechanically compressed than really solid. The resulting piece was about 10 x 4 x 6 cm (2/3 the size of a pound of butter), with a lot of cracks. When I spark tested it, it appears to be nice low carbon material.
The good news on this is that this small bloom will prove relatively easy to handle for me when I go to try to forge it down into a bar.

What I really *should* have done was to have immediately cleared the furnace and pulled the bloom when the slag poured out - so at least the mass would have been at welding temperature.
My guess is that there was considerable iron contained in that last flow of slag. We had run virtually the same smelt last time, the only main difference was that the earlier furnace was cylindrical at 22 cm ID. With the same air set up, the same ore, the yield was closer to 25%.

The purpose of this experimental series is to duplicate (?) the physical processes that were carried out by the Norse at Vinland. This puts our team in a bit of an awkward position. All the archaeological evidence suggests that although the Norse did smelt local bog iron ore into a small bloom, they certainly did not 'do it very well'. We know from our own experience that merely adding chopped straw to the clay mix would virtually eliminate the cracking problems. Our quandary : Do we do what our experiences have taught us - or do we 'screw up' like they did?

In truth, the iron smelt presentation at LAM on August 21 is really about demonstrating the *process* to the general public, not about the quality or size of the bloom produced at all.

But its hard to report poor results openly ...

PS - I expect to have my full report on this smelt, with images, posted up to the web site in a couple of days. Will send the link when I've got the materials available.