Vinland 3 Smelt - Draft Report

Members of the Dark Ages Re-creation Company undertook their third smelt in the 'Vinland' experimental series over the weekend.

Smelter and Bellows System	Leather Y at Tuyere, Clay Furnace supported by stone slabs.	Later in the smelt, tap arch open showing slag bowl inside furnace.

Smelter:
22 cm interior diameter
55 cm tall
5 - 7 cm wall thickness
construction of clay mixed 50 / 50 with course sand
'Boxed Short Shaft' - a clay cylinder supported by stone slabs, wood ash packing
tap arch at 45 degrees to tuyere position

Tuyere:
2.5 cm ID steel pipe
started smelt with end taper to about 1.5 cm ID
inserted 5 cm past interior wall
set at 25 degrees down angle
interior tip set 15 cm above hard floor of smelter

Air:
Human powered
Norse style double bag bellows
estimated air delivery (@ average one stroke per minute) 700 LPM
Attached to tuyere via a leather Y joint, allowing both observation and ability to clear blockages.

Ore:
Dark Dirt 2 analog (about 65 % Fe)
18 kg added
Basic charge was 1 kg, added by 'standard scoop' at approximately 250 gm

Charcoal
graded to normal ' golf ball to pea' size (through 2.5 and .5 cm grids)
Basic charge was 2 kg, added by 'standard bucket'

Production
Bloom - 2.9 kg (after initial compaction after extraction)
further 1 kg of lacy metal was recovered (mostly too fine to compact)

High temperatures above Tuyere.	During Extration - bright mass to top of the smelter is the bloom.	Second sequence with the 'Thumper' to loosen the bloom.

General Notes:

The use of a steel pipe (standard 1/8 wall thickness) always results in erosion of the tuyere. Typically, this works back to the furnace wall, then causes significant erosion of the furnace wall itself. This was extreme in this case, the wall immediately around the tuyere was egg shell thin (about 2 mm) by the end of the smelt. At one point the wall burned through and had to be quickly patched. Continued use of our standard ceramic tuyere would certainly have avoided this problem.

The specially built 'smelting bellows' proved both easy to operate and able to produce volume required. There appeared to be some problem maintaining the correct pressure. Lower force levels used by some bellows operators resulted in an obvious shifting of the heat zone both towards the tuyere and also further up from the bottom. This in turn caused the slag bowl to sit higher and closer to the front than is ideal. Some problem with a shallow slag bowl attempting to drown the tuyere resulted, requiring tapping / modifying bowl position, especially early in the smelt.
The most obvious solution is better training for the bellows operators to allow for more consistent air delivery. (This *was* our first use of human powered air since our very first attempts, some years ago. Many of the operators 'taking turns' had no experience with any type of bellows at all!)
Increasing the angle of the tuyere slightly may also help with lowering the slag bowl formation.

This smelt also marked an absence of one modern aspect - time. In past experiments, elapsed time has been the main control point for our actions. For this smelt, a consistent air flow was established, and a set of standard measures for charcoal and ore used. The pattern of ore to charcoal additions was based on past experiences (rather than being modified to suit elapsed time). This did not appear to present a major problem, as the team has accumulated enough experience at this point to utilize other cues (primarily sound).
One thing that did become obvious is that some counting method is required. Over the long course of a smelt, things like 'how many buckets since...' and 'how many ore scoops so far this bucket...' become difficult to remember. A simple set of stone counters is the suggested method for keeping track.

It was clear that the end segment of the smelt sequence was rushed slightly. It was decided to merely cover over the last ore charge added with a partial bucket. The reason for this was the extreme fragility of the eroded sections of the lower furnace walls. In the end it was found a significant amount of reduced metal did not have time to adequately attach itself to the developing bloom. This resulted in the loss of a good 25 percent of possible bloom mass.

Hammering on the bloom to remove adhering slag, and compressing the mass.	Close up of bloom after initial compaction.

Despite these small problems, overall the experiment was a success. The bloom produced appears to be soft workable iron, and certainly closely matches the actual production by the Norse at Vinland which we are working to replicate.