Wind & Weathering : air delivery & long term erosion Darrell Markewitz, with contributions by Neil Peterson

Wind & Weathering :

air delivery & long term erosion

Darrell Markewitz, with contributions by Neil Peterson

 Introduction :

For the last bloomery iron smelt of the 2021 season, a standard pattern furnace was constructed, then fired using a typical sequence. There were two primary goals to this experiment :
Wind - The recent acquisition of a high quality air flow meter allowed for precise and frequent measurements of the actual 'in line' delivery of air into the working furnace. A by-pass system allowed for the shifting to human powered bellows at a number of points, also with accurate recording of volumes produced.
Weathering - A new furnace was constructed on a clean sand pad, set to one side of the main smelting area. Photographic and video recordings were made of the extraction sequence. The remains of the furnace and the resulting debris field will be exposed to weather, and the aging documented as the features erode. The intent is to continue these observations over the next ten years.

This version, prepared for the web site documentation, has additional images

Part One - The Smelt :

The furnace constructed for these experiments was the long proven ’Short Shaft’ type. (1) In this case the furnace structure was placed on a rectangular plinth formed from fairly uniform thickness limestone blocks. The clay walls were constructed of EPK powdered clay, course sand, and dry shredded horse manure, rough thirds by volume, again a standard mixture. Individual fist sized ‘bricks’ were placed against an internal metal form. To reduce the possibility of cracks generated from clay shrinking against these stones, a ‘Beardsley Break’ was incorporated at the top of the first line of clay. (2)


Figure 1 : Limestone blocks laid out for the plinth	Figure 2 : Setting the metal form (note cover of newsprint!)	Figure 3 : Early in the build, showing the ‘Beardsley Break’.

The initial construction would result in walls built to 65 cm height. On later consideration, It was decided to add a further 10 plus cm (to a total of 77 cm). That initial build would have allowed for roughly 50 cm of working stack, considered effective. It became obvious in the days leading up to the smelt however, that participants were going to be limited (only Markewitz and Peterson had confirmed). Adding the extra height would provide a bit of additional working time, so ensuring no interference with the reduction of ore.

One factor that may have had significant impact on the later durability of the construction was that the builder (Markewitz, working alone) had experienced significant nerve damage to the left hand and was still recovering, and so did not have full strength to apply.

Figure 4 : Furnace interior after the first layer completed, showing minor breaks on inner surface

As was done in past builds, the completed clay structure was wound with rope to support the exterior, then filled with ash / sand mix to both stabilize and assist in initial drying.


Figure 5 : Completed furnace after initial build, with rope spiralled over exterior, interior packing	Figure 6 : After addition the upper ring

The day before the smelt, the tuyere was installed, the arches cut, and the furnace underwent several hours of a slow drying fire


Figure 7 : With tuyere installed, arches cut	Figure 8 : Moisture venting off the clay surface during drying fire

Details : 

Figure 9 : Front Elevation of completed furnace

Wall Thickness : 8 cm at base - tapers to 5 cm 
Interior Diameter : 28 cm 
Total Shaft Height : 65 cm at first build - extended to 77 cm
Plinth : natural stone about 10 cm tall, filled with charcoal fines 
Extraction Arch : cut through liner, 22 - cm (inner) / 26 cm (outer) wide by 25 cm tall 
Tapping Port : 6 cm wide by 10 cm tall 
Tuyere : forged copper, set 4 cm beyond interior wall, 20 degrees down angle, placed 90 ° to arch 
Stack Height (above tuyere) : 59 cm  
Base Depth : 28 cm above hard base, charcoal fines layer adjusts depth below tuyere to 15 cm
Thermometer ports : start 20 cm above furnace base, set every 10 cm / one set above arch , one set at tuyere

Figure 10 : Overall view of the working area during the pre-heat, with the furnace on the sand pad, the air supply systems to the rear.

Darrell Markewitz was smelt master, with Neil Peterson as lead hand and record keeping, Richard Schweitzer assisting on bellows.

Preheat Phase : 100 minutes 
Air Volume : varies over sequence - see details in Part 2 
Total Slag Added : 1.5 kg
Ore Type : DD1 Analog - 51% Fe (3) 
Total Ore Added : 24.3 kg 
Total Charcoal Consumed : 60 kg (‘bucket’ = 1.9 kg) 
Primary Sequence (over ore additions) : 220 minutes 
Primary Charcoal Burn Rate : (average over ore additions) 8 minutes / kg 
Ore Rate : (average over additions) 9 minutes / kg

Over the smelting sequence a number of measurements of internal temperatures were made, using an OMEGA HH12B digital meter. Measurements were taken by using a number of different thermocouple types, inserted through the holes cut into the body of the furnace. Each was pushed through to about 2 cm beyond the interior wall. A number of physical problems hampered the use of these probes. There were a good number of fine wire thermocouples available, but these were rated to only 1100 C, and did fail into the higher interior temperatures that were reached. (4) There was also one rigid metal shielded probe, rated to 1330 C. Unfortunately, the holes prepared were just barely larger that the shaft, and so once clay expanded due to heating and slag accumulated, it proved impossible to insert that probe through the furnace walls.

TIME	1352	1402	1430	1439	1521	1530

ARCH 20	1170		1135		1210
30	930		1040		1190
40	780		840		1190
50	550		800		1000
60	320		650		905
70	250		460		715

TUYERE 20		1180		1180		1315
30		1030		1030		1350
40		1190		1190		O/L
50		800		800		-
60		780		780		-
70		610		610		-

Table A : Charting the interior temperatures
Distances above furnace base in cm
O/L indicates problems with the meter, no further measurements possible

As was fully expected, there would be a number of slag taps over the course of the full sequence. The first two of these were based on observations (sound and visual) of rising slag levels potentially interfering with air flow. As the first of these were very early into the addition of ore, the first at 6.5 kg (potentially only 2 - 3 kg fallen to bowl level), it was felt the most likely origin of that stag was from interior erosion of the furnace walls around the tuyere. Both of these taps were controlled to small volumes, attempting to drop the interior slag bowl and level of liquid slag, but to retain as much of a ‘working’ system as possible.
In the later stages of the smelt, there would be two much larger volume ‘self taps’. For both of those events, the hot slag was allowed to flow undirected, and initially remain where it solidified


Figure 11 : The results of self tap 1	Figure 12 : The results of self tap 2

Over the course of the smelt sequence there would be three intervals where a human powered ’smelter bellows’ replaced the air blast from an electric blower. (Details on this seen in Part 2 : Wind)

During the course of the smelt, it became clear that the fusing of the individual clay additions during construction was less effective than normal. At least two major cracks, running top to bottom, had developed. This could be attributed to the building process, as the cracks clearly jogged back and forth along joints of the individual ‘bricks’. There was also a clear break between the original shaft and the second additional ring, although this was fully expected. During the actual smelt sequence, these cracks, although clearly visible, were not considered a major structural problem, as there were no actual gaps large enough to vent furnace gasses visible. In retrospect, wrapping the furnace with several loops of iron wire should have been undertaken to strengthen the furnace.


Figure 13 : Cracking above the front arch	Figure 14 : At the start of the burn down phase, accumulated tap slag at the front.

To record the actual extraction phase, two different video systems were fixed in place, with two different viewing angles. Peterson undertook the extraction, with Richard Schweitzer utilizing the ‘thumper’ and hammer compacting, Darrell Markewitz provided general assistance and hammer compacting.

(Peterson seen in blue, Schweitzer in silver, Markewitz in grey) (5)

To clear a working area, the accumulated tap slag is shovelled to one side.
The slag bowl was obviously large, and securely adhered to the inner furnace surface. Even breaking the intended extraction arch free of the slag is seen to be difficult.
The remaining charcoal fines are cleared from underneath the slag mass, to create some working space. - The normal attempt to pry upwards to loosen the slag bowl is seen to lift the entire furnace structure.
To loosen the slag block, the mass is hammered downwards into the lower space.
This results in the breaking up of the majority of the furnace structure (6)

The mass pulled clear is clearly huge, including the majority of the interior slag, still attached to the bloom at it’s core. Initial striking by Schweitzer (employing large wooden ‘troll hammer’) can be seen to quickly knocking off the slag. Once the bloom is exposed, this was worked down with metal sledges. The way the metal compacts into a more solid block is an indication of a relatively ’spongy’ consistency.

Figure 15 & 16 : The roughly compacted bloom - top and bottom surfaces (as inside the furnace)

Final Weight : 7.6 kg
Total Yield : 31 %

On to Part Two : WIND

Notes :

1) see ‘If you don’t get any IRON…’

2) This feature named after George Beardsley, who had demonstrated it’s use during a cast iron session at the Scottish Sculpture Workshop in 2014.

3) Further details on the reasoning behind, and preparation of this analog mix can be found : Ore Analog Composition (blog post)

4) The various wire probes on hand have been purchased individually over many years, so detailed specifications are uncertain. The method was to insert the thermocouples for just long enough to get a reading, then pull them clear and allow them to cool before moving to the next point. This kind of short duration use will often allow measurements to temperatures beyond the rated melting point of the wires.

5) The first sequence (start to extraction) was shot on a cell phone camera (© Neil Peterson)
The second sequence (duplicates extraction to initial compaction) on a compact video camera (© Darrell Markewitz)

6) These two sequences combined do give a more complete depiction of events. On viewing the second (side view) footage, it is clear that the action of the log ‘thumper’ was in a diagonal direction, pushing against the front edge of the slag bowl at least, and perhaps actually against the interior furnace wall? To be fair, this was the first time Schweitzer had ever undertaken this specific task, and despite the fire suit and face shield worn, was still somewhat restrained working down inside the furnace.