EXPERIMENTAL ARCHAEOLOGY
An overview of the smelting days
by Leo Moonen (Leo Ironsmelter)
PREFACE:
For several years the members of the “iron smelting group” at the Hunebed Centre in Borger have been extracting iron from ore, following a method which dates back to the Iron Age.
Iron can be found as iron ore in rocks but also near to the subsurface as bog iron, carried as a solution by water and deposited. This process of deposition happens quite quickly over decades, rather than centuries. Sometimes just a few decades are enough for farmers to decide to remove the iron-bearing layers a few feet beneath the surface of their land.
Here in Holland, as in several other parts of Europe, the basic method of extracting iron from bog ore was very simple, using small smelting furnaces, and this process is called “iron smelting”. (1)
This practice – and the necessary skill – are regularly kept alive at the Hunebed Centre (2) and at various other places in Europe and elsewhere! There are only a few smelters in each country and it is always a pleasure to meet and share our experiences, as we did this year. The meetings took place at:
Ulft Netherlands, April 18-21 (3)
Woodford Galway in Ireland, August 19 – 25 (4)
Eindhoven Pre-Historic Village /Prehistorisch Dorp, October 16 – 20. (5)
This report covers our iron smelting days in Eindhoven, organized by
“International Iron Smelting Days” and “Prehistorisch Dorp Eindhoven”
The team members:
Andre Hazewinkel Hunebedcentrum
Leo Moonen Hunebedcentrum (Leo Ironsmelter on facebook)
Paul Klaasen Blacksmith (Smederij Klein Projekt, Ulft) (6)
Jasper Swaak Blacksmith
Forty to fifty “iron smelters” worked in 11 teams, spread over the park.
Thanks to:
Alun Harvey, English translator at the Hunebedcentrum, for checking this document for grammatical and spelling errors.
CONTENTS:
Four days “iron smelting”:
page 4 day 1 Wednesday afternoon, roasting bog ore and start building the furnace
page 7 day 2 Thursday, finishing the furnace and preparing the roasted ore.
page 8 day 3 Friday, firing the first furnace: iron smelting
page 12 day 4 Saturday, building and firing the second furnace: 2nd irons smelt.
page 17 day 5 Sunday (morning), cleaning the area.
INTRODUCTION:
Andre and I arrived on Wednesday afternoon and were assigned a place in the
middle of a few very nice looking historic farmhouses near a lake. We started
collecting iron ore (140 kg) and made a good fire for roasting the ore. We also
started building (my) preferred type furnace: a German type
“rennofen”. The plan was to build the furnace strong enough so it
could last two smelts, on Friday and on Saturday. This would be the second time
for me to use a furnace twice – but it collapsed the first day after removing
the bloom…. The second furnace was built from scratch on Friday, very
rapidly, it hardly had time to dry before we started smelting on Saturday
afternoon.
Together with the changeable weather we also saw a varying number of visitors – ranging from “gezellig” during sunny periods to hardly any visitors during the frequent rain.
The bog ore was supplied by “IISD” and was dug up near Dalfsen on the River IJssel.
For many smelters, eating and sleeping was done inside the historic farmhouses – an adventure worthwhile in itself.
Several blacksmiths were present to work on the new blooms.
Although the bog ore contained a lot of iron, there was also an amount of phosphorus present, which prevented the iron sticking together in a nice solid bloom. As a result, all four blooms were rather “crumbly”, no matter which working methods we used, such as:
– using hand operated bellows, often blowing near horizontal into the furnace
– electric blower driven airflow, often blowing under an angle on top of a bloom.
Our first bloom (estimated total weight c. 6 kg) fell apart during compacting; we did not compact our second bloom so far but stopped and split the bloom in two halves. Total weight was c. 3 kg.
Preparation:
First the ore was roasted on a “good” fire. In this process all organic material is broken down and the ore gets brittle. The iron compounds in the ore (as iron oxyhydroxides) and is mostly converted into magnetite: Fe3O4. (7) As magnetite adheres to a magnet, it can be sorted out of the cooled roasted ore. In this way you can ensure that you have high-percentage iron in the ore but you are leaving some non-magnetic iron oxides behind in the roast… The remaining pieces must be small enough to allow the CO to enter and do their work (taking the oxygen away) and large enough not to be blown out of the furnace during charging.
We did not sort the roasted ore using a magnet, we used all the ore.
After the furnace had reached the desired temperature, we loaded equal amounts (weights) of ore and charcoal into the top opening of furnace. The desired ratio was 2 kg of ore and 2 kg of charcoal every 12 minutes. This 2 kg was divided in three portions of c. 650 grams.
DAY ONE – Wednesday October 16
After an introduction by the overall co-ordinator, Yvonne Lammers we started to roast 140 kg of bog ore.
Total height was 120 cm. The internal width was close to 35 cm at the bottom and c. 28 cm at the top. As it was difficult to look from the top inside the furnace during operation – to check the level of the ore + charcoal – we later removed a part of the top layer at the back (20 cm).
We inserted two tuyeres, one at 20 cm from the bottom and a second at a height of c. 28 cm. During the smelting process the level of molten slag rises. Slag tapping can be done a few times, but that becomes more and more difficult – then it is time to move the blower up to the second tuyere …. and continue until the system “clogs” the airflow.
DAY TWO – Thursday October 17
This was a sunny day with plenty of visitors.
We finished the furnace outer layer and gave it a special small platform near the top. Some children showed great interest in playing with our construction material (loam or clay) so we asked them to make something (like a cup) that could stand on top of the furnace to dry for a short time and then they could take it home.
The roasted ore was cool by now and broken in pieces of 1 to 3 cm.
We expected Paul and Jasper to join us the next day, Friday, late in the afternoon, so we decided to do some preparation work: we (mostly Andre…) weighed the crushed ore in portions of 650 grams and put it in thin plastic sandwich bags together with a spoon of the fine remaining “powder” ore.
We did not refine the ore by selecting all the magnetite by magnets. And nothing – such as soda, chalk or sand – was added to the ore.
DAY THREE – Friday October 18
The first smelting day.
Furnace: first tuyere 20 cm above bottom, under c. 20 degrees. Second tuyere at c. 28 cm. The thermocouple for measuring the inside wall temperature is stuck in a metal tube. The other end of this tube is flattened, bent 90 degrees and stuck to the inside of the furnace wall. The temperature of the wall is NOT the same as the “inside” temperature. When you look through the glass on top of the tuyere you “see” the actual temperature of the molten slag. It is more comparable to the colour-temperature relation of black body radiation. As this colour turns from bright yellow to almost white, and the intensity is too much for the naked eye (you need sun or welding glasses), we know by experience that “it is working fine”.
The inside temperature of the fluid (slag) will be around 1250 – 1350(°C), the actual flame temperature of the burning charcoal will be even higher…..
Higher up in the furnace the temperature drops. Various chemical processes take place. A T-shirt from Adrien M. (designed by Maria A.K.) shows what is happening inside the furnace:
Dirty burning of charcoal produces a lot of CO gas which takes the oxygen part of the iron3-oxyde4 to the upper part of the furnace (at T ~ 700 °C or higher)
If everything goes fine, iron atoms should find each other and connect to form pieces of iron. A lot of iron stays connected to oxygen as iron1-oxyde1 (wüstite) and can be found in the tapped slag – this is not wanted.
After about 2 hours we tapped the liquid slag from the furnace for the first time – this was quite fluid (low viscosity)
Looking through the tuyere you can see pieces of charcoal slowly dropping into the “moving” hot white surface of the molten slag. During the smelting process this surface level rises almost to the exit of the tuyere. As it touched the tuyere we did our first tap. If all went fine, some iron should be formed at the bottom of the bowl, iron is heavier than slag. Not all the slag is removed – the hot bath remains so that the iron mass is allowed to grow.
Sometimes the airflow is partly blocked by a piece sticking just in front of the tuyere and then you need to punch that away through the tuyere hole. This time a piece of molten “iron+ slag + magnetite” got stuck at the punch stick and froze in a brittle needle shape:
The whole smelting process went fine, but only lasted for 3.5 hours. The log file of charge loads is shown at the end of this report.
We expected a nice bloom to emerge just after dinner – it was getting dark already.
I have no pictures or video of opening the furnace.
I asked Matthias (of the German team Sachsenhof Greven) and Adrien Morat to help us compacting the bloom. (Matthias has a huge wooden hammer which is far too heavy for me to lift…)
A nice bloom came out, but it did not survive the compacting process. After some time it broke in pieces – one piece of 1500 grams below.
Some time later it turned out that all teams (except the first smelt of Jan J.) suffered from blooms that did not remain in one piece and could not be compacted and “cleaned”. Too much phosphorus in the ore prevented the iron sticking together… There are a lot of small bits of iron – magnet sticking – in the bloom. Refining these bloom parts is another process we did not do. (9) A similar piece of bloom from our neighboring German team:
DAY FOUR – Saturday October 19
The furnace we used yesterday did NOT survive. We started to build a new furnace, smaller, with prefab bricks made of loam. Height 80 cm, internal width at the bottom 28 cm, tuyere at 20 cm.
We finished the furnace early in the afternoon and started with the first charge at 15.46 .
It would (again) be dark at the end of this smelting process…..
Despite an increased distance between the blower exit tube and the tuyere inlet, the furnace became hotter and hotter after an hour! For a time we also had some “self tapping” at the end.
In the evening, when it was already pretty dark, we removed the bloom by flipping the whole furnace.
By now we knew (also from the other teams) that too much compaction would break up the bloom into smaller and smaller pieces, we stopped before that was expected to happen and split the bloom in two. Total weight was around 3 kg.
Further extraction of the iron from this bloom will require extra attention, for example in a small hearth furnace, and careful work in a smithy. (9)
Log files:
A total of 140 kg ore was collected (wet). After roasting, 105 kg remained (75 %) of which 88 kg from near Dalfsen, IJssel (Jan J.) and 17 kg was of other, “lesser” quality.
First iron smelt on Friday, 18 October 2019
This furnace, with internal diameter of c. 32 cm at tuyere: Desired burn rate: (2 kg ore + 2 kg charcoal) / every 12 minutes (+/- 1 minute)
Each time the height of the charcoal in the furnace dropped around 20 cm, another charge was given to the furnace. The amount of 2 kg was split in 3 parts of ~650 grams. So, noting the time, ore entered the furnace ( one bag of 650 grams ore, directly followed by ~650 grams charcoal) , 3 times directly after each other. Followed by a wait of 12 minutes before the next charge was given. If the rate increased, the airflow was lowered; if the rate decreased, the airflow was lowered. First we put the ore in the furnace , directly followed by the charcoal to cover the ore.
11.45 Start warming up, first wood, then charcoal, until the burn rate of charcoal was 2 kg/12 minutes
1 14h39 Start: 1 charge T=1040°C
2 14h51 T=1160°C
3 15h05
4 15h16 tuyere blocked, punched
5 15h26 T > 1160°C
6 15h38
7 15h51
8 16h04
9 16h22
10 16h48 first slag tap,”a lot”
11 17h03
12 17h15 second slag tap,”a lot”
13 17h48
14 18h10 last charge
Stopped charging the furnace (getting dark, dinner time….) .In total 14 * 2 = 28 kg ore was put into the furnace. Remaining charcoal burned for 1 hour
19h15
The bloom came out Too many too heavy blows on the bloom: in smaller bits…
Stopped
Average charge rate:
28 kg ore in (18h10 – 14h39) = 211 minutes
-> 1 kg ore 7.5 minutes, or 2 kg ore every 15 minutes ( and 2kg
charcoal every 15 minutes)
Second iron smelt on Saturday 19 October 2019
This furnace, with internal diameter of ~28 cm at tuyere, has a surface of ~76% (D1/D1)^2 of the first furnace. We reduced the burn rate to the burn rate of yesterday, accordingly to { 65% ore + 76% charcoal + } / every 12 minutes. The ore was already sorted in sandwich bags, 650 grams, so we decided on a multiple of 650 grams:
Desired burn rate of { (1.3 kg ore + 1.45 kg charcoal) / every 12 minutes }
Charging the furnace was split in 2: (0.65 kg ore + 0.72 kg charcoal) / every 6 minutes.
14h00 Start warming up, first wood, then charcoal,
until the burn rate of charcoal was 1.2 kg/10 minutes (= 1.45kg/12 minutes)
1 15h46 Start: 1 charge T=1040°C
2 15h52 T=1160°C
3 15h57
4 16h02 tuyere airflow de-blocked
5 16h10 T > 1160°C
6 16h11 burn rate too high, reduced airflow (increased distance pipe-tuyere)
7 16h17
8 16h22
9 16h27
10 16h36 first slag tap,”a lot”
11 16h50 tuyere airflow de-blocked
12 16h51
13 16h56
14 16h57 burn rate too high -still
15 17h05 second slag tap
16 17h06
17 17h14
18 17h15
19 17h17
20 17h18
21 17h24
22 17h25
23 17h31
24 17h32
25 17h36
26 17h
27 17h40
28 17h44
29 17h49
30 17h
31 17h57 last charge
Stopped charging the furnace (getting dark again, dinner time….). In total 20.1 g ore entered the furnace. Remaining charcoal burned for 1 hour
19h15 – 30 The bloom removed from the furnace. Stopped compacting the bloom after some time and split in two halves. Total weight was ~3 kg.
Average charge rate:
20.1 kg ore in (17.57 – 15.46) = 131 minutes
-> 1 kg ore 6.46 minutes, or 2 kg ore every 12.9 (~13) minutes ( + 2.2kg charcoal every ~13 minutes)
– The End –
Internet links:
(1) https://pediaa.com/difference-between-melting-and-smelting/
(2) https://www.hunebedcentrum.eu/en/
(3) https://www.regio8.nl/pionieren-met-ijzer-op-dru-terrein-in-ulft/content/item?1129709
(4) https://www.furnacefestival.org/
(5) https://www.prehistorischdorp.nl/nl/agenda/vuur-vonk-en-ijzer
(6) https://www.kleinprojekt.nl/
(7) https://pediaa.com/difference-between-fe2o3-and-fe3o4/
(8) Forging iron bloom to iron bar ……… the event, Vuur, Vonk & Ijzer, (fire, spark and iron) took place in the prehistoric village in Eindhoven,………
(9) Hurstwic: Converting Bloomery Iron to Steel in a Hearth Furnace
======================================================
