Those who have been following the recent work have seen a number of commentaries related to the ongoing series of bloomery iron smelts based on elements from the excavations at Hals, Iceland, by Kevin P. Smith.
The basic undertaking for the phase three experimental series was build a potential Hals type turf / sod furnace, containing a relatively thin clay liner, using a mixture simulating (as best possible) the material found close by that location. The full build was done in mid June, with the first complete firing cycle on June 20. The intent was to subject the furnace to multiple firings, including one after a winter exposure to freeze and thaw.
After the first use of the furnace, there was pretty much the normal expected damage to the front section of the furnace, with erosion of the walls around the tuyere, and breakage at the extraction arch.
 |
|
Furnace - just at end of the smelt
| 20 cm turtle - and damage
|
What was not expected was the invasion of the warm furnace by a large snapping turtle, which resulted in a large section of the front wall section breaking clear. Weather at Wareham has been unusually wet over later June, July and now into August. (This normally a period where there is little or no rainfall = Climate Change!). Although the top of the furnace was fitted with a metal cover, there has been considerable damage to the 'baked mud' outer sections of the furnace walls, plus considerable slumping to the original grass sod surrounding structure.
 |
Field drawing of the furnace - August 7, 2021
Note combination of metric and US Imperial units (single marked tape!) |
|
|
|
|
Typically, the next scheduled smelt attempt would take place over Canadian Thanksgiving, this being October 10. I have some concern over additional damage to the existing furnace structure, so have decided to undertake the required repairs to the lower front of the furnace, including breaking clear the internal slag bowl which was left in place after the first smelt.
 | Interior - just after smelt
(tuyere still in place - at top)
|
|
|  | Roughly same view, slag bowl broken clear
|
|
The slag bowl from the June smelt did nicely resemble those exposed at Hals, a distinctive C shape with a cupped cross section, the front edge broken away where the bloom had been pulled free during the front extraction. The remaining slag broke free of the interior wall surface quite easily, using a chisel tipped bar from above. There was little actual damage to the wall structure from this process.
 |
Condition of furnace - August 7
|
As seen in the images and drawings above, a considerable part of the front wall surface had been broken away.
The original line of the sods can be distinguished as the place where the upper exterior wall changes from smooth to a wrinkled texture. That top area, about 10 cm wide, despite the cover, had been flaking away in the rains. The slumping of the sod structure is obvious here as well. The heat from the smelt had largely destroyed the binding root structure, leaving basically baked earth. Again the rains were slowly washing this earth down and away.
The original build had constructed a full cylinder of clay, with the stones supporting the arch and tuyere placed to the outside. As a result, there was no exposure of the stones to heat, other than the short time of extraction. The archaeology at Hals indicates fire marked and slag adhering stones. For this reason, in the repair, a set of small flat stones were used to block in the eventual extraction arch, which will expose the inner surfaces to the heat of the lower furnace.
 | Interior after repair. Dark grey is the new clay added
|
|  | Lower section, replaced lintel stone with small flat stones filling extraction arch
|
|
In the initial build, it proved quite easy to keep the wall thickness to 4.5 cm overall. This simply was not possible when adding new clay to repair sections considered badly eroded, and those locations where the walls needed to be replaced entirely. new material had to be added by reaching up inside through the extraction arch, most often working from touch alone. There are certain to be places where the fresh clay will be much thicker. Additional clay had to be added around the edges of the small flat stones used to seal the extraction arch, as these where chosen from a random pile of available (gniess) pieces. The same granite lintel stone was used as in the first build, only this time it's inner surface will be exposed to the full heat at tuyere point. Once again, this slab serves to support the upper layers of grass sod (total of three).
 |
Finished repair build, tuyere yet to be installed
|
An extra row of grass sod was placed around the furnace liner, building the supporting structure back to approximately the same level as at the start of the first smelt ( about 65 cm above hard base). Extra dirt was placed in a ring around the top of the sods, with the hope of reducing any fire damage during the next firing. One additional aspect seen in this image is that the grass composing the sods laid in mid June has continued to grow, and least around the edges of the sod cone.
For this repair, a total of five prepared balls of the clay / sand / manure mix were required, each about the size of a large grapefruit. (This material left over from the initial construction of the clay liner.)
Depending on team availability and other ongoing work here at the Wareham Forge, the hope is to undertake the next smelt in this series some point over the next two weeks...
