A Northern Smelt

I was up to northern Michigan, Traverse City, over the weekend. Dan
Nickels of Black Rock Forge has hosted a great workshop weekend at his
shop each mid February for the last dozen (?) years. About 40 people,
had braved the bad roads and serious cold to attend.

It was roughly -15 F at night and maybe hit about 5 F mid day Saturday.
We smelted outside in the wind - something like -25 with the wind chill
when we started. At least the snow had mostly stopped, there was maybe
another inch fallen over the day.

David Robertson and I (crazier than most) drove around closed roads and
through total white outs in Ontario to get there on Friday. A couple of
the local fellows had expressed an interest in seeing an iron smelt. One
of the smiths had generously paid for the charcoal (total of 240 lbs)
They also had a good quantity of taconite pellets from the mines in the
Upper Peninsula. I brought all the gear to set up and run an Econo Norse.

The smelt team was made up of Brent Cole, Kent Chambers and myself.

We used 67 lbs of the pellets, crushed to pea / rice plus the dust.
Since I forgot my measured ore scoop, I made an estimate of a standard
measure (which turned out after to be wrong). Instead of the more normal
1/2 to 3/4 lbs per charge - we really were adding in multiples of 1 lb.

One modification was that the packing material supporting the firebrick
core was made up of a mixture of wood ash (from a stove), construction
sand, and charcoal fines - roughly equal amounts. This proved relatively
stable, instead of flowing through the cracks in the bricks the ash and
sand combined in the heat to kind of a course particle glass and fused
the gaps closed. (See the files section for plans for the Econo Norse).

This smelt also used one of the ceramic tube kiln supports I had spoken
about in the fall for the tuyere.

The smelter was based on 8 standard fire bricks per layer, three layers
tall. This put the chamber at about 10 inches dia, total about 27 inches
tall. Tuyere was set about 8 inches from the base, set to the magic 25
degrees down angle.

Smelt progressed pretty much as expected. Roughly 200 lbs of charcoal
was used in the main series. We started pre-heat with split wood at
about 8:30, first course charcoal about an hour latter. Air was via my
calibrated vacuum blower system. We had the air set a bit higher than
what I use normally, closer to 800 l/m (an estimate).

I did my best not to mess with the slag bowl (yes, Uncle Skip!). A
couple of times, small amounts were tapped off, mainly to check the
reaction, and to demonstrate the process and explain slag development.
At roughly 3 1/2 hours in, the arch was opened and about a three inch
gap was cleared out of the fines below the slag bowl. Some liquid slag
was drawn off into this space to keep the tuyere clear, but mainly the
system was then left to its own activity.

The burn down was started at 5 1/2 hours into the sequence. A total of
64 lbs of ore had been added to that point. Once that charge was
estimated to have settled to the tuyere level, a slug of ore (for carbon
control) was added - the balance of 3 lbs. This was then allowed to
settle down to tuyere level.

Normally I would have then stopped the charcoal and let the level just
start to burn down. The other events of the workshop put us into a
holding pattern however. The air was reduced by maybe a third. The tap
arch was opened, and some slag was pulled off and recycled. This was
found to be dark, fluid and glassy - but almost totally non magnetic. I
was getting exited - as obviously the iron had gone some place!

At 7 hours 15 minutes I started to extract the bloom. First the
remaining 1/3 burning charcoal was scooped out to clear down to the top
of the mass - roughly at tuyere level. An attempt was made to use the
'thumper' to slightly compact and loosen the bloom in place. This did
not prove very successful - the mass seemed very dense. When probed with
the bloom hook it was discovered that the mass extended from side to
side of the chamber over the entire front 2/3 of the space available
(towards the tuyere). I just could not get it loosened.
So the brick structure was disassembled, the top two rows (of three)
being individually pried off and pulled away. The bloom when exposed was
quite large, with the classic Knob on the top just below the tuyere tip
and a less dense sponge filling from side to side. A number of whacks on
the Knob were enough to loosen the mass from the remaining brick chamber.

See data02-06.html for the experiment details

The bloom was swung over to a waiting stump, and hammered over its
surface to remove as much clinging slag as possible. Some consolidation
was done via hand hammering at this point as well.

Add much cheering from the assembled crowd. (easy to impress - none of
them had seen this before, save a couple who had experienced Lee and
Skip at ABANA a couple of years back)

A lot a shifting gears and team at this point. I was wasted, and told
the younger - and considerably fresher - participants it was up to them
to consolidate and cut.

The base of the brick smelter was quickly re-arranged to form a rough
pit forge. Steel pipe replaced the now damaged ceramic tuyere. The
remaining charcoal (about 30 lbs were used) kept the bloom at at least a
dull orange while more modern equipment was arrange for the next step.

A new team converted a three burner propane forge to hold the bloom with
fire brick. The mass was transferred inside  and worked over with an air
hammer to consolidate and re-shape the bloom so it could be cut. This
was proving to be fairly heavy work - but perhaps an hour latter two
pieces were separated. At this point my fishing scale could handle the
weight - one section was 12 lbs, the other 18 for a combined weight of
30 lbs.

My estimate of the bloom at extraction is closer to 35 lbs total. A
number of walnut sized pieces came off to the point of the cutting. We
also found at least one fist sized piece in the smelter base the next
morning. This gives us a rough yield of about 50% - with our source ore
thought to be something in the range of 65 % iron content.

David Robertson spark tested across on of the pieces produced the next
morning. His estimate was that the iron ranged from roughly .4 through
perhaps as much as 1.2 plus carbon in that fist sized piece. Given the
relative amount of work required to draw the bloom down into two first
stage bars (not welded) roughly 2 x 2 in diameter, the general
impression was a material about mid carbon content.


Extremely well done. I figure a new centre of iron insanity may have
been kindled...

(Edited from a posting to EARLY IRON Wed, 29 Mar 2006)