sml logo Dark Ages Re-Creation Company sml logo
Oct 29, 2016
Oct 12, 2015
Jun 20, 2015
Apr 25, 2015
May 11, 2013
May 27, 2012
Nov 5, 2011
Oct 9, 2011
July 28, 2011
June 11, 2011
Nov 6, 2010
Aug 21, 2010
June 12, 2010
Nov 7, 2009
Oct 11, 2009
May 30, 2009
Nov 8, 2008
Oct 12, 2008
Jun 14, 2008
Apr 13, 2008
Oct 27, 2007
Oct 8, 2007
Jun 10, 2007 - Excav.
Jun 9, 2007 - Second
Jun 9, 2007 - First
Nov 5, 2006
Sep 1-2, 2006
Jun 10, 2006
Jun 10, 2006 - Tools
Nov 4, 2005
Sep 1, 2005 - CanIron V
Jun 11, 2005
May 14, 2005 - OABA
Jan 14, 2005 - Consolid.
Oct 24, 2004
Jun 3, 2004
Jun 3, 2004 - Design
Jun 3, 2004 - Questions
May, 2003 - Analysis
May 2003
May 2002

Iron Production Experiment Nov 2005

Date: 12 Nov, 2005

Location: Vinderheima


We decided on two variables only, one major one minor :

1) Reuse a smelter structure. This entailed repairing the inner wall of  the smelter around the tuyere. A depth of about 1 1/2" of new clay  was applied to replace the material eroded from the earlier smelt. We had high confidence that this would not create a major problem, as archaeology shows this process, even several times over. The clay body used was the same as used to create the initial cob structure.

2) Use of one of the pottery kiln shelf support tubes as the tuyere. These were 2" OD and that magic 1" ID. The material was rated to 1250 C - which puts it into our operation range for the smelter. Again our confidence was high that this would perform well.

We had re-lined the smelter about a week before the smelt with pretty wet clay, and patched some cracks from the outside. Afterwords wrapped the structure up in its protective plastic bag. The repair area was still soft enough to take finger dints when we started.

In addition to these variables we decided to confirm our air flow numbers so that we can work on getting the bellows producing the correct flow.

Experiment Notes - Norse Short Shaft Smelter

The Team:

Leader Darrell Markewitz
Staff Kevin Jarbeau
Dave Cox
Ken Cook
Recorder Neil & Karen Peterson

Smelt_cd 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.

Setting up the smelter:
This was easy this time.  Patching the inside and outside of the wall near the tuyure.  A few other small patches were put in place.

A forty-five minute pre-warm was done with wood splits. Since we were waiting - what's a fire without some jiffy-pop! We've got to get around to asking them to help sponsor us...

It should be noted that due to a minor miscalculation we ran out of prepared taconite ore.  As a result we switched over to virginia rock ore part way through (the ore charges are identified below).  In addition some gromp was added towards the end of the burn.
thumbnail thumbnail thumbnail thumbnail
Patches around tuyere Adding charcoal First tapping Auto-tapping
thumbnail thumbnail thumbnail thumbnail
Cooled autotap Burning well Burn down Pulling the bloom
thumbnail thumbnail   
Consolodating Measurements

The morning after
As usual we deconstructed the site the morning after the smelt taking a variety of measurements to assist in our experiments.

thumbnail thumbnail thumbnail thumbnail
Light slag Wall top Above tuyere Tuyere comparison
thumbnail thumbnail   
Double wall Double wall

Time Event Air Flow Measures
Relative Absolute Total
Relative Elapsed Charcoal Ore
0:00 00 addition rough charcoal, occasional gentle air blast (begins at 1135h) B        
0:25 00 Increase air flow C        
0:38 32 first graded charcoal added          
0:44 19 Increase air flow C+        
0:48 10 charcoal   1 5 lbs* 5 lbs  
0:58 10 charcoal   2 5 lbs 10 lbs  
1:04   seed charge   1 .75 lbs*   .75lbs
1:06 8 charcoal   3 5 lbs 15 lbs  
1:14   seed charge   2 .75 lbs   1.5 lbs
1:15 9 charcoal   4 5 lbs 20 lbs  
1:21   seed charge   3 .75 lbs   2.25 lbs
1:25 10 charcoal   5 5 lbs 25 lbs  
1:31   seed charge   4 .75 lbs   3 lbs
1:33 8 charcoal   6 5 lbs 30 lbs  
1:37   seed charge   5 .75 lbs   3.75 lbs
1:41 8 charcoal   7 5 lbs 35 lbs  
1:50 9 charcoal   8 5 lbs 40 lbs  
ore charge (+ one and a half scoops added in this time)   6 1.1 lbs+   4.85 lbs
1:58 8 charcoal   9 5 lbs 45 lbs  
ore charge (++ two scoops plus the left overs from above)   7 1.9 lbs++   6.75 lbs
2:06 8 charcoal   10 5 lbs 50 lbs  
ore charge    8 1.5 lbs   8.25 lbs
2:13 7 charcoal   11 5 lbs 55 lbs  
ore charge   9 1.5 lbs   9.75 lbs
2:22 9 charcoal   12 5 lbs 60 lbs  
ore charge   10 1.5 lbs   11.25 lbs
2:30 8 charcoal   13 5 lbs 65 lbs  
ore charge   11 1.5 lbs   12.75 lbs
2:38 8 charcoal   14 5 lbs 70 lbs  
ore charge   12 2.25 lbs   15 lbs
2:45 7 charcoal   15 5 lbs 75 lbs  
ore charge   13 1.5 lbs   16.5 lbs
2:53 8 charcoal   16 5 lbs 80 lbs  
ore charge   14 2.25 lbs   18.75 lbs
3:02 9 charcoal   17 5 lbs 85 lbs  
ore charge   15 2.25 lbs   21 lbs
3:10 8 charcoal   18 5 lbs 90 lbs  
ore charge   16 2.25 lbs   23.25 lbs
3:18 8 charcoal   19 5 lbs 95 lbs  
ore charge (1 scoop taconite, 2 scoops virginia rock ore)   17 2.25 lbs   25.5 lbs
3:26 8 charcoal   20 5 lbs 100 lbs  
ore charge (virginia rock ore)   18 2.25 lbs   27.75 lbs
3:36 10 charcoal   21 5 lbs 105 lbs  
ore charge (virginia rock ore)    19 2.25 lbs   30 lbs
3:38   Scoop of gromp added          
3:41   Rogering the tuyure          
3:43 7 charcoal   22 5 lbs 110 lbs  
ore charge (virginia rock ore)   20 0.75 lbs   30.75 lbs
3:50 7 charcoal   23 5 lbs 115 lbs  
ore charge (virginia rock ore)   21 1.5 lbs   32.25 lbs
3:58 charcoal   24 5 lbs 120 lbs  
ore charge (virginia rock ore)   22 2.25 lbs   34.5 lbs
4:06 charcoal   25 5 lbs 125 lbs  
ore charge (virginia rock ore)   23 2.25 lbs   36.75 lbs
4:16 charcoal   26 5 lbs 130 lbs  
ore charge (virginia rock ore)   24 2.25 lbs   39 lbs
4:24 charcoal   27 5 lbs 135 lbs  
ore charge (virginia rock ore)   25 1.5 lbs   40.5 lbs
4:25   ore charge (virginia rock ore)   26 0.75 lbs   41.25 lbs
4:28   ore charge (virginia rock ore)   27 0.75 lbs   42 lbs
4:36   Begin adding gromp pieces          
4:37   charcoal   28 5 lbs 140 lbs  
4:42   charcoal   29 5 lbs 145 lbs  
4:45   Add remaining gromp (total 1.1 kg)          
4:47   reduction charge   28 1.5 lbs   43.5 lbs
4:48   charcoal   30 5 lbs 150 lbs  
4:57   charcoal   31 5 lbs 155 lbs  
5:07   charcoal   32 5 lbs 160 lbs  
5:16   charcoal   33 5 lbs 165 lbs  
5:25   charcoal   34 5 lbs 170 lbs  
5:35   charcoal   35 5 lbs 175 lbs  
    air flow reduced to minimal - remaining charcoal removed          
    consolidate bloom with wooden thumper          
    remove bloom and consolidate          
    charcoal   36 5 lbs 180 lbs  
    return bloom to smelter          
    remove bloom and consolidate          
    charcoal   37 5 lbs 185 lbs  
    return bloom to smelter          
    remove bloom and split          
6:25   smelt complete          

Air rates

Control Mark At mouth 1" from mouth Fully Blocked
m/s flow (L/s) m/s flow (L/s) m/s flow (L/s)
A 5.3 2.7 2.9 1.5    
A+ 10.1 5.1 6.1 3.1    
B 16.7 8.5 9.5 4.8    
B+ 21.2 10.7 12.0 6.1    
C 22 11.2 13.3 6.7    
C+ 23.7 12.0 14.4 7.3    
D 24 12.2 14.9 7.6    
D+ 30.5 15.5 16.5 8.4    

The Anemometer has exactly a 1" diameter fan.  The pipe has an internal diameter of 1.375".  The "At mouth" measurement is done by placing the device against the pipe.  This leaves 4 small holes where the air can escape with the majority going through the device.  The second sets of measurements are taken with the device one inch from the mouth of the pipe.  This provides a feeling for the directness/pressure of the air.  The final column is measurements made with all air being forced through the device (same as the first column but with the small holes blocked).

Charge times recorded at END of unit addition

*  All charcoal charges are based on a 'standard pail' - weight of 5 lbs
All ore charges are based on a 'standard scoop' - weight of 3/4 lbs
  Estimated from known marks

YIELD:  10 lbs (estimated) - looks like it may be good workable iron (spark test on interior surface)

An excellent smelt.

This was as easy as it ever gets. We re-used the cob smelter from June, so no construction work was required.

We started hauling out the piles of prepared gear at about 8 am or so. Set the canvas and prepared the work area - about 9:30 Mounted the ceramic tuyere, made new fittings to adapt the viewing T, patched a few remaining cracks.

The air system preformed very well. We only had to 'roger' the tuyere once, and might have been able to skip that one.

We did get a LOT of glassy gooey silica slag produced. Most of this was extremely porous, so the volume was high. We should compare WEIGHT  against clay erosion some time. We had to tap this at least three times,  as the level rose to start flooding the tuyere. A big crowd favorite!

We were concerned that we were not getting any of the iron rich 'gromp'  slag. When we sectioned the smelter the next day, it turned out that  there was an amount of this created, but it had run through the lighter glass slag of the lower bowl to collect under the slag bowl. We were able to pull this out as a single large piece.

Again, the slowest part of the preparation was the ore - with Ken and Kevin doing the lion's share. We started with the richer taconite ore, with 24 lbs prepared as the preheat was underway. One potential error / problem from the post-smelt review. For the past couple of smelts, wehave been roasting the pellets before crushing them. The material from STELCO is not magnetic as it was delivered. This time we used the ore pellets straight from source. This made the crushing more difficult, and may have had an effect on our yield. In the end we added a second lot of pre roasted Virginia rock ore (18 lbs).

There was also 1.1 Kg (2.4 lbs) of gromps (magnetic slag) left over from the OABA smelt in May that was added near the end of the process. This small amount, combined with the inability to access the newly created iron rich slag, likely combined to produce the slightly porous structure of the bloom.

Our process was halted partially by time, but also by the charcoal we had prepared. In the end we had used 42 lbs of ore.

A solid 'slug' consisting of 1 1/2 lbs ore (two scoops) was then added and allowed to settle to the bloom area. This judged by burn rate of the charcoal. (5 more pail loads, 48 minutes of straight charcoal. At this point the fire was allowed to burn down to about half the smelter volume before extraction. The air flow was kept at the same rate throughout the smelt (approx 12 L/s).

As with past smelts, the air was reduced and the top layer of burning charcoal was scooped out down to roughly the top of the liquid slag / tuyere level. (Since the sun had set at this point, the twilight made for spectacular photos!) The thumper was used to loosen the bloom from the slag bowl. The position was just as expected. This method has a couple of advantages. It first removes the fragments of 'mother' that cling to the bloom when  it is extracted from the tap arch. This makes the primary consoldation phase a bit easier. The second advantage is that its easier to utilize the heated smelter as a giant forge loading from the top.

The resulting bloom was positioned from the tip of the tuyere to about the mid line of the smelter, but extending roughly from one side to the other. This gave a rough lens shape roughly 10 x 5 x 5 cm.

It was roughly compacted using the wet stump method that Mike showed at Early Iron (with good control). The hammer team was Ken, Dave and Kevin. On the first compaction heat, the thickness was reduced by about 1/3. Three compaction heats were taken, reducing the size down to 6 x 5 x 2 1/2. There was virtually no fragmentation off the edges during the process.We took a last heat to section for the classic 'pleiner knotch'. Kevin on the hot set. The bloom cut relatively well. At about 1/2 way through, however, one of the lobes split off. (This same thing happened at Early iron) The bloom most likely needed a bit more compaction before cutting was attempted.

As mentioned, this smaller piece was later spark tested. The inside of  the cut surface was used. The sparks show large and red with little feather, which suggests a good low carbon and workable iron was produced.

The ceramic tuyere was exceptional. There was some, but actually very  little erosion. Over the entire course of the smelt, only about one inch of the tubes lenght was lost. The crisp cyliner did errode back to the classic conical shape encased with hard slag (as seen in artifacts), We figure we can get at least two - maybe three uses from each ceramic tube - at a cost of about $8 each.

The relining of the smelter proved sucessful as well. On Sunday the smelter was sectioned along the tuyere line. This allowed us to determine the relative erosion for the two uses of the structure. An imporant note was that the higher erosion rate always experienced with the metal pipe tuyeres had come close to completely burning through on the last smelt. The 3" wall had been reduced to a mere 1/2 just above the tuyere. On Saturdays smelt, less erosion was found, about 2" of thickness remained (starting from roughly 3" thick). This is primarily due to the durability of the ceramic tuyere. At the start of the smelt, it projected about 2" from the interior wall. At the end of the smelt it extended about 1 1/2 inches past wall (bearing in mind that both the tuyere and wall had changed shape during the smelt).

The position of the final slag bowl was about what was predicted for a well managed smelt. The overall structure of the smelter, with the cob cylinder resting on a circle of fire brick, created a problem. The lower edge of the slag bowl was slightly below the upper rim of the fire brick. This made first attempts to lower the slag bowl / decant slag difficult. In the end the front brick had to be split and pulled out in sections to allow access. If a large bloom had been desired, this difficulty in lowering the slag bowl would have made it hard to leave room for a larger bloom to form. This same restriction kept us from gathering the iron rich slag that had formed and penetrated the glassy slag at the last phases of the smelt. (Or frankly, even knowing it was there.)

Taken altogther, and extremely successful smelt!

One new experimental process proven - that of relining the smelter. There was some discussion of attempting a third use of this smelter. There was extensive cracking of the structure, with pieces missing from the top edge. More importantly, the area around the tap arch was in pretty poor shape. There was a large block of solidified slag that would have had to have been chipped out, with probable damage to the base area. It was decided that it would be better to build a new smelter for next years series.  Clearly though with additional forethought we can set a tap arch we won't have to damage as much to get the slag out.

One new piece of equipment tested and proven - the ceramic tube tuyere. The durability of this was excellent, and is combined with the low unit cost and ease of replacing. (almost as good as the heavy forged copper tuyere - and at 1/10 the cost). A new mounting pipe will be made up to match the ceramic to the existing T port system.

Smelt_cd 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.
      Updated: 4 Dec, 2007
Text © Neil Peterson, Darrell Markewitz, 2005
Photographs © Individual artists
Copyright details
Contact us