Iron production weekend - 2004

Date: June 12 & 13, 2004

Location: Vinderheima

Premise: If they made iron -- then so can we using the same techniques

Conclusion: We can't do it this way either but this is much closer.

The Team:

Staff	Dave Kevin Gus Mark Michael Ken
Recorder	Neil Peterson
Leader	Darrell Markewitz

Reports of all of our iron smelting efforts along with more articles and information are available on the "Iron Smelting in the Viking Age" CD from the Wareham Forge.  Copies of the CD can be purchased here.

This year marked our third attempt to produce iron from ore. Once again we made some changes from the previous year. Once again we did not get any iron.

Again this year we had some visitors. Mark from the L'Anse aux Meadows, NL interpretative staff came out to act as an observer (so of course we put him to work). We also had a visit from Michael from the Farmers' Museum in Cooperstown. The museum website is here. He was on his way home after spending several weeks discussing and doing Japanese smelting. Michael organized and hosted the 'Pre Industrial Iron Symposium' in October.

In this year's experiment we reaffirmed our intention of targeting the reproduction of a NORSE method of iron production rather than making the SUCCESS of the experiment our priority. This choice controlled many of our design decisions this year. Our notes from these discussions can be found here. Some of the questions facing us can be found here. Of course this means that you may have to visit us for a few more years before you see an iron bloom from us.

Once again we would like to express our thanks to Royal Oak Charcoal who donated the charcoal used in this year's experiment. We have used Royal Oak Charcoal in all our experiments as they are a hardwood lump charcoal (pretty much what the Norse would have produced) and their quality is consistently high. From all of us -- Thank You!

The smelter itself was constructed differently this time. A circular smelter of clay/cobb was built. Around this base a stone box was added. The spaces between the two elements was filled with sand. Three weeks before the actual smelt we gathered to dig a local clay and build the smelter. Our target was a smelter 65 cm high with an average wall diameter of 10 cm. The internal dimension was to be 25cm. The bottom of the smelter was to be a shallow bowl with the air inlet (2.5 cm diameter) 10 cm above the bottom of the bowl and coming in on a flat angle. Our first mistake occurred here. The clay we dug was chunky and quite wet. Although we used a couple of techniques to break it up, the actual cobb we used in the build was neither as dry or as uniform as it should have been. This led to the smelter having a noticeable settling as it dried. The final dimensions of the smelter are listed in the table below. As these things always seem to work just as we finished building we found an article documenting a series of similar experiments back in the 1980s. Crow, Peter: Experimental Production of Historic Bar Iron, Historical Metallurgy v25, No 1, p27. We've included the numbers from his smelters in the table below for comparison.

Dimension	Planned	Actual	Crow
Wall Thickness	10 cm	10+ cm	20 cm
Interior Diameter	25 cm	25 cm	30 cm
Interior Height	65 cm	55 cm	60 cm
Diameter of air inlet	2.5 cm	2.5 cm	2.5 cm
Tube Angle	0	~15 deg. down	30 deg. down
Inlet Height	10 cm	2 cm	20 cm

Photos of the Smelter build weekend

Breaking clay with hand tools	Crushing the fresh clay	Breaking clay with power tools	Mixing the cobb
Digging out a flat base	Setting the inlet and front wall	Setting inner form on the base	Inner and outer forms in place
Packing cobb between the forms

The day of the Smelt

June 12th. We began this process with a few overall area shots.

Overview	Final check	Smelter top	Bellows

During our morning setup an ore-roast was set up as we did not have a full 7 Kg of ore ready to go. The ore roasted was more of the same ore we used last year. The leftover roasted ore from last year was screened and weighed into 1Kg charges. When it had cooled sufficiently the ore roasted today was also broken up, screened, and set into 1Kg charges. A total of 7 charges were prepared. The ore was screened twice. Once with a 25mm screen to remove large chunks. It is worth noting that no ore pieces failed this test. We estimate that the actual largest piece was probably around 20mm. The second screening was done to remove fines. A 4mm screen was used for this purpose. This allows us to note that ore added for this smelt was between 4 and 25 mm in size.

In a similar fashion the charcoal was broken up and screened for size. Charcoal used was between 7mm an 25mm in size. The charcoal was also weighed into 1 Kg charges to allow us to track the usage.

Setting up the burn	Weighing ore	Anne, Gus, Peter	Sarah, Anne

Actual Time	Elapsed Time	Notes
1036h		We began with the warm up burn. We began the warm up with some softwood.
1051h	T+0:15	The vacuum we were using for air was set to 1/4 and the opening into the furnace was noted as 10cm.
1053h	T+0:17	While the warm-up continued we had our final planning meeting.
1104h	T+0:28	The warm-up fire was switched to hardwood.
1153h	T+1:17	Finished the first milk crate of hardwood in the warm up fire.
1214h	T+1:39	First optical pyrometer measurements. Temperature inside was recorded as 2350°F (1288 C).
1234h	T+1:59	Second milk crate of hardwood finished.
1243h	T+2:08	Our old vacuum died. Oh well it happens.
1248h	T+2:13	New vacuum brought on-line and air flow resumed.
1314h	T+2:39	Time to begin the charcoal burn. The hardwood and ash was scooped out. New Air system was made ready.
1316h	T+2:41	The Smelter was filled with charcoal (approx. 3.5 Kg), bellows were hooked up and pumping began.
1324h	T+2:49	Live coals from the ore roasting were added to the top of the smelter
1330h	T+2:54	Remains of fourth kilogram of charcoal added (approx. 0.5 Kg) From this point the smelter was stirred/poked from time to time and small amounts of charcoal were added as needed. The experiment records only the time each Kg of charcoal was finished.
1341h	T+3:05	Fifth kilogram of charcoal finished. Optical pyrometer measurement gives the internal temperature as 2380°F (1304 C)
1350h	T+3:14	Sixth kilogram of charcoal.
1358h	T+3:22	Bellows stopped.
1400h	T+3:24	Bellows restarted.
1403h	T+3:27	Seventh kilogram of charcoal finished.
1406h	T+3:30	Optical pyrometer measurement: 2720°F (1493 C) At this point the ore was added. Each charge of ore was added in two steps (each of approximately 0.5 Kg).
1410h	T+3:34	Ore charge 1 - part one added
1414h	T+3:38	More charcoal and ore charge 1 part 2 added.
1416h	T+3:40	Eighth kilogram of charcoal finished.
1417h	T+3:41	Ore charge 2 - part one added.
1422h	T+3:46	Ore charge 2 - part two added.
1426h	T+3:50	Tuyure rodded out.
1429h	T+3:53	Ore charge 3 - part one added.
1431h	T+3:55	Optical pyrometer measurement 2640°F (1449 C).
1431h	T+3:55	Ore charge 3 - part two added.
1432h	T+3:56	Ninth kilogram of charcoal finished
1440h	T+4:04	Ore charge 4 - part two added
1441h	T+4:05	Tuyure rodded out
1443h	T+4:07	Tuyure rodded out
1445h	T+4:09	Ore charge 5 - part one added. Tenth kilogram of charcoal finished.
1450h	T+4:14	Tuyure rodded out
1452h	T+4:16	Eyeball estimate of airflow volume was made. It appears that we are pumping 9L/sec (540 L/min) into the smelter.
1454h	T+4:18	Ore charge 5 - part two added. It was around here that we realized that the shallow air inlet was in fact giving us the problems we had been hoping wouldn't occur. An attempt was made to "save" the situation.
1505h	T+4:29	Attempted to rod out the tuyure again. This time with a hammer and hitting it hard. We attempted to remove the pipe to clear the jam but it wouldn't move. At this time we decided to add a new air inlet on the back of the smelter 45 cm below the top of the smelter (hence 10cm above the base of the bowl.
1511h	T+4:35	Back door was holed through and a slag pull was done.
1515h	T+4:39	Eleventh kilogram of charcoal was finished.
1518h	T+4:42	Another attempt to rod the tuyure.
1531h	T+4:55	Another attempt to rod the tuyure. All of these attempts have produced only partial successes.
1540h	T+5:04	Tap the back hole again.
1545h	T+5:09	Move the bellows to the back hole. Front hole was plugged with clay.
1548h	T+5:12	Twelfth kilogram of charcoal finished.
1552h	T+5:16	airflow in the water measurement device attached is 4.5->6.5 cm of water.
1600h	T+5:24	Thirteenth kilogram of charcoal finished.
1609h	T+5:33	Fourteenth kilogram of charcoal finished.
1620h	T+5:44	Fifteenth kilogram of charcoal finished.
1630h	T+5:54	Rod tuyure
1641h	T+6:05	Stop bellows, connect new pipe to back and change bellows
1644h	T+6:08	Begin airflow with new bellows
1647h	T+6:11	Rod tuyure
1656h	T+6:20	Sixteenth kilogram of charcoal finished.
1704h	T+6:28	Change to rotary air pump
1709h	T+6:33	Seventeenth kilogram of charcoal finished.
1719h	T+6:43	Eighteenth kilogram of charcoal finished.
1725h	T+6:49	Nineteenth kilogram of charcoal finished.
1739h	T+7:03	Rod tuyure
1749h	T+7:13	Stop airflow and allow smelter to begin cooling.

Adding wood	Front view - air supply	Burn with air	Wood burning down
Removing wood	Preparing bellows	Charcoal added	Mark on bellows
Airways	Checking the airpath	Looking down the pipe	Preparing new airway
Opening the back	Rogering the tuyere	Measuring for hole	Plugging the new hole
Preparing to shift	Tapping the back	Tuyere plugged	Salg on radner
Switching air source	Wire testing the smelter

Day 2 - The Excavation

Sunday began with a photographic record of the site. It is worth noting that the smelter itself was still hot to the touch this morning (and the outside temp had been down below 10C the night before). After the record run we broke for breakfast and some discussion and then came out to disassemble the smelter.

Measurements of empty smelter. Above the original tuyure there was an area of was that had been burned away to a depth of 2.7 cm. This burned in area extended 19.5 cm up from the tuyure opening and had a maximum width of 15.5 cm. On the left side there was a burn in 1.6 cm in depth, on the right side a burn in of 1.5 cm depth existed. The distance across the top of the smelter from left to right was 23 cm, the base of the smelter was 27cm from front to back. The original tuyure opening began 46 cm from the top of the smelter.

More photos and begin some weights. We removed a large piece of slag earlier. The top of the slag was 53cm below the top of the smelter. The bottom of the smelter was intact and the slag was resting on the clay NOT on a layer of charcoal. There was no dead-zone and no place for the slag to go. The air path through the slag mass in these photos is obvious. Total bowl depth was 3 cm. The original tuyure was plugged with 5 cm of slag. For the accuracy of the following weights please note that they were recorded in pounds, ounces, and tenths of ounces and then converted to kilos.

Side view	Top view	Tuyere	Real view
Left side	Rear air port	Removing rock	Cobb smelter
Extending crack	Planning break	Removing first piece	Planning next piece
Removing charcoal	Brushing out pieces	Removing more charcoal	Removing more wall
Closeup of wall	Wall pieces	Closer view	Close up
First slag	Plugged tuyere	Level of rear tuyere	Level of slag bowl
Pieces of wall	Blowing out ash	Clear of ash	Piece of slag
Location of slag pieces	Tuyere	Rear tuyere	Recovering pieces
Team at work	Rear airway	Slag mass	Rear wall pieces
Line of airway	First airway exposed	First airway	Slag mass - bottom
Slag centre removed	Salg mass	Overview	Overview without slag mass
Scaled photo	Taking measurements	Plugged tuyere	Closeup of wall

Item	Weight
slag removed from bowl	1.06 Kg
Solid blocks of slag with Iron Blebs	2.96 Kg
Deepest slags (near base)	0.69 Kg
Slag from near tuyure	0.23 Kg
Magnetically separated slag	0.31 Kg
Fill from the tuyure	0.09 Kg
Total Mass of Slag Recovered	5.34 Kg
Total Mass of Ore used	5.0 Kg

Conclusions

Time to burn kilos of charcoal
Once the first four kilos of charcoal had been added it was possible to see how long each new kilo took to burn. Those times are listed in this table:

Charge	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19
Time	NA	NA	NA	NA	11	9	13	13	16	13	30	33	12	9	11	36	13	10	6

The smelter itself was a SOLID creation this year. If it had not been our intent to break it down for recording purposes we could have run a second smelt in the same body.

The biggest single issue this year was the lack of depth in the main bowl not allowing space for the slag and thus building up to where the air flow was significantly impeded.

In a nut shell - we pushed too much air. This created overly hot interior, which in turn created too large a volume of melted wall as fluid glass. This clogged the air intake. (That whole effect enlarged by slumping of inlet towards base during construction). Excess super heated oxygen literally vaporized the iron particles. The fluid glass slag drowned the air intake. Constant fiddling with bellows (plus break down of one) dropped interior temperatures and halted the reactions. Although an attempt was made to recover - the reaction had been halted and high temperature zone shifted to make restarting the reactions impossible.

Concepts for next year

This section is a little "free form" right now but will get some more work.

• Plan on four days next year: One to dig and prep the clay, one for building smelter, roasting ore, prepping charcoal, one for the burn, one for records work and disassembly.
• Shade near the smelter to make visuals of the flames easier to interpret.
• Plan the smelter for 75 cm.
• Drill the tuyure once the smelter is dry - don't use a pipe as we build it.
• Have a prearranged tool kit including: weigh scales, visual scales (non-meltable), whisks, drill bit & drill, scoops, protractor
• A workshop to make bellows first? Or remove the bellows and just use the rotary air source?
• Marks on the tuyure to record burn off
• Wall clock
• Pre-published schedule
• Number all charge containers and have more ready in advance (100 oz coffee cans for ore)
• Make a materials list
• Ensure internal walls are smooth to reduce problems with air flow patterns
• Build the smelter in two steps, fire the bottom before adding more wall height.
• Run a full theoretical model first with numbers
• Record additional measurements including: pump rate, air flow indicators, consumption rate, rate of ore drop, temperature
• Reduce ore size (aim for 4-7mm)
• Keep ore charges to the center of the smelter and spiral it out as the burn progresses and the area grows (as the walls burn)
• Perhaps we should do two smelters next year with slightly different designs
• Use a Y fitting instead of a T to add air plastic proved to be durable enough. Cut notch that seals with wood plug downstream of this - which will allow simple clearing of nozzle without breaking air flow. A simple tool was suggested for this purpose.
  
• More observation holes with simple wooden plugs
• The design of the smelter requires modifications - suggested by Michael. We will create the lower section (below air intake) into a conical shape, smallest at the bottom and tapering out to air inlet level. The volume of this area should roughly equal the volume of the glassy slag melted out of the walls at the base of the smelter as observed. (Roughly 10 cm at base and maybe 15 - 20 cm tall). This will give that fluid slag someplace to go. Exact location and angle of air inlet to be fixed by using sticks inside clay body to prevent slumping.
  
• Creation of durable forms for tamping clay/cob into to allow for standardized smelters.
• Significant work needs to be done to build a better bellows (Darrell's fault, he got rushed). Kevin's bellows very durable, but * may * be a bit small in terms of air volume. Larger bellows considered ideal from volume and pressure stand point.
• Construction of a moveable overhead shelter (tin roof) was discussed, this will allow easier observations of flame structure.

Reports of all of our iron smelting efforts along with more articles and information are available on the "Iron Smelting in the Viking Age" CD from the Wareham Forge.  Copies of the CD can be purchased here.