Friday, April 29, 2011

A 'Learning Experience'...

... Or 'Just pulled it off by the skin of our teeth, from the jaws of total failure ?'

Not just once, but TWICE.

Theoretical design of the smelter.

I had the golden opportunity to finally run a version of the 'Experimental Iron Smelting' workshop program I had designed for university students. At Brown University in Providence no less. In large part thanks to my college, sometimes adviser (and friend) Kevin Smith from the Haffenreffer Museum of Anthropology.

The planned session consisted of background lectures, then a full day with students building a furnace and preparing materials, then the next day actually conducting an iron smelt. The final part of the program would be for them to survey the remains as if it was an archaeological site.

I had intentionally designed the furnace so it stood on a brick plinth. This allows for easy correction of potential slag bowl problems, and also permits either top or bottom extraction methods. Flexibility the key here.

Good thing...

I had also provided a blend of differing ore types. I had just enough of the DARC Dirt 2 analog (9.4 kg) and a bit of granular haematite (1.5 kg) to produce a dependable bloom. Normally the yield on that blend would be expected to be something like 25 - 30 %. In addition I had a good amount of not so great Virginia rock ore, previously roasted (about 20+ kg). This was stuff Vandy and I had picked from one of the mine sites Skip and Lee use. Not the best mine source, and very unskilled hands gathering. In the past this stuff had proved to produce a lot of slag, but not much iron.
My plan was to start off with about 6 - kg of the poor ore, establishing the slag bowl. Then switch over to the high grade analog mix, forming a nice bloom. Then switch back to the low grade material, which should add some extra mass to that bloom. Along the way, there should be greater than normal slag production, especially during the later stages of the smelt. More taping events than normal, so good experience for the students.

Well, things got a bit more *exciting* than that.

Into the second hour of the smelt, there really was a fair amount of slag, perhaps sitting a bit higher than desired. Remember that I was making the students do most of everything - at least as much as possible. During one of the rogering sequences (clearing slag drops out of the front end of the tuyere) something I've never seen before happened. The ceramic tuyere got shoved into the centre of the furnace, actually dropping it inside the shaft. It fell down below its mounting hole in the furnace wall. I tried (frantically) to hook it back up and clear, but without being able to.
Remember this cuts the air blast entirely, so the whole time the furnace is quickly freezing up. As I had told the students repeatedly 'If you cut the air blast any more than about three minutes, the whole furnace temperature crashes, and you can't get it back.' Working like a mad man, I cut a double fist sized hole into the side of the furnace just around the tuyere mount, grabbed out the blocked tuyere tube, then shoved in the steel pipe mount used to attach the ceramic tube. Kevin and one of the students quickly (an working with amazing intuition) slapped on wet clay and back filled the area with sand. Turn the air on full blast.

Amazingly, we got the heat back. Talk about acting in the nick of time!

The damaged tuyere. The right end was originally (correctly) set inside the furnace.
The left end was broken off through (failed) attempts to grab it with tongs and pull it clear.

With clear head, the student managing ore and charcoal additions had just plugged along, following our previously established sequence. Pretty much like none of this other insanity had been unfolding around him. Great work! As the furnace moved back up to correct working temperature, the shift was made to the high iron content ore analog.

It was clear however, the there was most definitely too much slag, and sitting too high into the furnace. So working with Kevin, I pulled clear one of the bottom plinth blocks, then scrapped out the charcoal fines forming the lower base. This exposed the bottom of the slag bowl. Using a technique that Skip and Lee taught me, I tried to lower the bowl, by taping hot slag to flow below the bowl. The principle here is that the hot liquid slag warms the bottom of the bowl, which then sags down lower. Well, than did not work so well. After the first demonstration, I let mainly Kevin punch through the slag bowl and drain excess amounts. We did reach some kind of working equilibrium, at least keeping the tuyere clear for the rest of the smelt.

Even with aggressive tapping, it was getting harder to keep the air blast constant. The damage to the furnace, plus the switch to a steel pipe tuyere (over sized as well), was making it difficult to get the air flow into the centre body of the furnace. The last 'burn down' phase of the primary smelt sequence was rushed. (We were also running out of charcoal).

I decided to blend the two extraction methods, as much for teaching purposes as anything. The last 1/3 of the smelter stack was cleared with the charcoal scoop, working from the top. Then there was a short sequence with the thumper, one of the students attempting to compact the bloom (?) in place. I had hoped this would also drive the slag bowl down into the cleared base area below the bowl. This more or less did happen, but also resulted in blowing open a big piece of the lower area of the furnace above the tap arch.

Smelter area, day after the experiment

The good news was that this smashing from above, plus the large piece of wall tearing off, *did* expose the bloom mass. Another student was standing by, ready to snatch the hot bloom with tongs and transfer it to a waiting wood stubb. Two others were ready as hammer men. Working with a hand sledge, I guided these two in knocking free any remaining 'mother' and at least partially compressing the mass. With the delays instructing students with their unfamiliar tasks, the temperature of the bloom had really dropped to the point there was not much effective working time.

The end result was a rather lacy bloom, about 2.2 kg in weight. I made an attempt to slice it with a cut off disk on my angle grinder. The sparks generated suggested a bit more carbon content than mild steel, maybe a 1030C equivalent.

On the day, exhausted from frantic activity and guiding a 4 - 8 person student work team, I was a bit disappointed. The ore analog additions should have resulted on something closer to a 3 kg plus bloom (based on past experience). It was clear that there was a considerable amount of reduced and partially sintered iron material that had not yet made it down to bloom level, wasted when we scooped the charcoal.

But like Kevin said at diner later - it was an excellent *teaching* experience. The intent of the whole program was to introduce students to the reality of working experimental archaeology. They had just been given a dramatic example of 'anything can go wrong', 'you have to think fast on your feet' and 'practical experience matters'.

I have to congratulate the participating students (who will be credited in the more formal report). They did an excellent job in all counts. They pitched in full heatedly, followed directions well, were innovative and resourceful as required by the challenges they faced. I would be happy to work with you all any time.
It also should be mentioned that it poured rain pretty much the entire smelt day!

The participants:

Note that is is from Krysta's original schedule. Some of the Friday build people turned out on Smelt Day (and worked there too). Many of those who came on Smelt Day stayed for the whole sequence.

Instructors : Kevin - Krysta

FRIDAY (Build & Prep)

Morning Shift: Seekay - Slash - Kevin - Peter
Furnace building

Afternoon Shift: Max - Nick - Morgan - Anya
Furnace building, ore crushing, pre-heat


Morning Shift: Ian - Alicia - Julieta - Matthew

Afternoon Shift: Zoe - Jeremy - Hiu

Saturday, April 16, 2011

a MEDIEVAL Double Bag Bellows

- This just a short piece, cut and modified from an ongoing conversation about iron smelting bellows over on Don Fogg's Bladesmith Forum.

Now, something you see on Early Medieval illustrations of blacksmith's set ups is stuff like this:

(sorry, I can't cite the sources. I have been preparing lectures and just did some scavenging via Google)

The Early Medieval blacksmith forge is using an extension of the twin chamber bellows from the earlier Viking Age. The chambers have been increased in size. The forge is now at table height. Most commonly the smith is shown working standing at the anvil.

The historic illustrations are never clear on this system, though the cross shaped bars and ropes are often illustrated.

These images stuck me as most useful, as you can see more easily how this lever system might work. A long bar set in a diagonal, likely mounted to the building, can rotate down its axis. There is a cross bar set on to this pivot. From the cross bar two ropes or chains attach to the top plates of the pair of bellows. One side of the cross bar extends towards the operator. There is a handle hanging down from this free end.
I suspect you would place some large weights on top of the two bellows top plates.
Pulling down on the handle would raise the opposite side bellows chamber. This also would free the close side chamber to drop and exhaust under the action of the weight. Releasing the handle would allow the far side chamber to drop under its top weight, the ropes would raise the close side chamber at the same time.

Monday, April 11, 2011

'Off Cut Bowl' - Waterlife / Shadow Box show

I was contacted a while back by Dyan Jones of the South Grey Bruce Literacy Council.

She was helping to put together a special fund raiser:

The Film & Invitational Shadow Box Show & Sale
For Youth Literacy - May 10 - 14, 2011
Screening and Arist Reception - Saturday May 14, 6 PM
Victoria Jubilee Hall - Walkerton Ontario

A couple of things about this intrigued me:
- 'shadow box' concept
- low object cost limit
- fund raising via object donation

- The individual artists where supplied with a simple wooden box, roughly 1 inch thick by 5 x 6 inches, open on one flat side. The work had to fit in or on this box.
- All of the objects would have a flat cost of $125

My challenge was two fold:
- Working in such a small scale
- Working inside the low price limit
Normally I would consider either a small knife, or a piece of forged jewellery.

This is what I came up with:

'Off Cut Bowl' - forged mild steel - 2011

This piece *is* a bit larger than the shadow box, which will serve as a plinth. (I had checked with Dyan, and this was one of the possible applications of the restriction). The object is about 10 x 10 in size, standing about 3 inches deep. The individual segments are in fact off cuts left over when I was cutting up some 1/8 inch thick plate I acquired as commercial scrap. Taking a clue from a method used by Japanese artisan Takayoshi Komine, the individual segments were MIG welded together on the back, then the weld beads ground smooth. The resulting flat form was then worked hot to dish it.

At a workshop session here in Wareham back in Feburary, I had messed around with David Robertson and Kelly Probyn-Smith with some other potential methods. One was attempting to lay in bronze wire into hot punched lines and melt the bronze into the grove. It turned out this was not going to work as simply as I had hoped. (The concept will work, but will require considerably more steps than simple punching and melting!) Another thing we did some tests of was laying down MIG beads on to otherwise flat heavy plate surfaces. Then hot hammering the raise lines flat. This shows some potential as a decorative effect, but needs some more testing to fully develop.

All of this comes from some pondering what kind of objects I might make in the future. Bowls have the advantage of being relatively small (at least compared to gates!) This makes them easy to transport and show, then store and protect. They cross the line to potentially practical objects, of a size and cost that makes them more accessible to the general public. There are also a very large number of blacksmithing methods that can be applied to create unique objects. The fluid shaping possible with hot forging will dramatically contrast the solid rigidity of the finished forms.

More to come!

Saturday, April 09, 2011

'Steel' strikers - some considerations (?)

This came in from my friend (and mentor) Kevin Smith of the Haffenreffer Museum of Anthropology (Brown University, Rode Island). Kevin has done considerable research into a number of pieces of jasper uncovered at L'Anse aux Meadows in the Viking Age occupation layer. These show signs of being used as the stone element for 'flint & steel' fire starter sets. Most importantly, these stones are not native to the region, and are in fact imports from Iceland. This discovery tied another of the Saga story elements into fact - that some of the ships that travelled to Vinland had Icelandic origins.

" I've been asked by Parks Canada to look at another jasper fragment they uncovered during recent excavations at L'Anse aux Meadows that were focused on the prehistoric native components. ...

In thinking about what to do with it next, I've been putting together a mental list of possible projects to do with students ... One thing that I know should be done is some more experimental work on how much debris is produced by striking jasper fire-starters with strike-a-lights, what kinds of debris are produced, and how to distinguish them from other kinds of flakes. "

This may be a much larger kettle of fish that you may realize. Your primary focus is on the stone - but the variation within the actual steel striker is a HUGE 'random' factor.

All our modern metals (as you know) are much different structurally than historic bloomery irons. Varying carbon contents then also gets back into the mix.

Image is taken from the Faganarms online catalogue:
" C866-1067. Found in York England. Identical examples have been recorded in numerous Viking settlements. 2 3/4” overall with scroll tip recurved arms and high peaked striker bar, demonstrating the skill of the forger. "

This is what (I think) I know about strikers and how they work:

The hard flint tears off a small piece of the metal.
The force of the strike actually massively heats this fragment.
The resulting temperature is enough to actually 'burn' this piece. (energetic reaction with the oxygen in the air)
The carbon content alloyed with the iron is what causes the visible spark. The more carbon, the larger and brighter the visible spark.

Now comes the tricky bit.
The exact alloy chosen needs to have just the right resistance to the shearing action of the flint that this tearing process results in 'sparks' (actually the pieces) that are hot enough to effect the tinder. If you use too soft a metal (modern, a 0.2 % mild steel), the hard flint will certainly tear the surface, but the resulting fragments are not heated enough (show no spark).
If you use a high carbon alloy, but harden (quench) the material too much, the flint can't actually tear any pieces off. Few fragments (sparks) produced, those that do occur are too small to retain effective heat as they pass through the air.
The balancing act is to use a carbon content *plus* correct cooling/hardening cycle to produce a surface that falls into the correct 'hard enough to tear with hot fragments' balance.

I suspect there is also an aspect of the physical structure within the stone itself. 'That side doesn't work' is an often heard comment. There are a set of variables that have to do with the actual shape of the stone edge. I don't have enough experience to know what exactly is required here, save to comment that the stone edge needs to be 'sharp' enough to slice off that sliver of metal, but at the same time not so thin as to have the stone shatter on contact.

To make this all even worse.
Individual pieces of stone are hardly consistent. Even within a walnut sized piece of stone, the effective 'hardness' can vary noticeably. Even the same 'type' of stone (read flint) does vary in hardness considerably between physical locations. (This is why 'English' flints are prized for use as gun flints over generally softer 'American' ones.) This effect is going to be even wider with differing types of stone (flint, jasper, ...).

I know from my own (limited) use of flint and steel that there is an effective pairing of a specific steel striker with a specific chunk of flint. Because the flint is consumed more significantly than the steel, you often see the situation where a once effective steel is suddenly considered 'used up' - because the flint itself has been replaced. Unfortunately, I have had the *steel* blamed more often than not for being 'poor quality' (which always makes the metalworker responsible!). "But I always got great sparks with my old steel" is often heard.

Back to the metal.

That theoretical description of how the metal is effected by the striking flint now has to be considered in terms of the actual iron metals used in the Viking Age. These are all bloomery irons, not our modern day Bessemer steels. As such, the historic metals have an entirely different creation process, resulting in a quite different physical structure.
Bloomery iron will always have microscopic layers of glassy slag as threads through the metal. These layers may create what are in effect shear planes through the block. As glass, those layers are hard, but extremely brittle. The number and size of such layers will vary depending on the quality of the individual starting bloom from the furnace. The smelt master would desire an extremely dense and compact bloom as a finished product, but a huge number of individual factors conspire together to determine just what qualities of the product of each individual smelt event might be. So any given bloom might contain quite differing amounts of glassy slag - and deposited physically quite different through the parent metal.

The way a given bloom is consolidated into bar might also have an effect. When starting with a dense bloom (with less contained slag in the first place) it may be possible to compress to working bar with very few compress / fold / weld steps. (I had one bloom piece that only took a single folding over and weld step to produce a quite nice working bar.)
Is there a factor related to just how this 'grain' is worked up during bloom to bar?
The simplest way, if the quality of the starting bloom allows, is to make long draws and folds to re-weld (long rectangle, single fold back on itself). This is going to create slag lines all running in the same orientation, down the long axis of the bar.
Most often, the fragmentary nature of the starting bloom does not allow for this. With flaws running in all directions, you end up dealing with diagonal cracks. The way to deal with this is to use a series of compressions and folds, each set at 90 degrees to each other. The easiest first step is to just flatten the bloom into a irregular plate of fairly uniform thickness. The first fold is then across this width (called a 'book' fold). This is welded, then the piece is drawn out to a long rectangle, at 90 degrees to the initial fold and weld. This rectangle is then folded in half and welded. Last the resulting 'brick' shape is compressed downwards to make another rectangle, again force at 90 degrees to the previous step. Fold and weld to a block. The end result is more or less a 'brick' shape. You can see how this would create a series of shorter slag lines, running in differing directions, but individually quite short.

Just to throw yet another element dealing with potential grain. Normally, the pronounced grain in a piece of bloomery iron (or historic wrought iron) is in long lines running down the length of a bar. (Think of the grain in a piece of clear pine lumber.) The normal process of making an object from bar is to form the shapes by stretching as required long the length. Strikers are more or less a C shape, the centre of the C creating the striking surface. This puts the grain running in more or less the same direction as action of the individual strikes. The net effect should (at least potentially) be long thin shavings.
Or is this what you really want (??). If you forged your source material quite deliberately, you could work the grain at 90 degrees to the direction of strikes. Admittedly, this would be a pain to do, and might also effect the quality of shapes on the two terminal ends of the C. (I don't actually think this would be likely, but thought I'd mention the theoretical possibility.)

Before we leave historic bloomery metal entirely, remember our variation in carbon content.
Its easy with our modern metals to specify alloy content, and get completely even distribution of carbon through the mass. Not so in either case with bloomery irons (despite what some modern knifemakers claim). Each smelt even will produce a bloom that will vary in potential carbon content. Each bloom will also vary in carbon within itself. The easiest way to roughly select for relative carbon (thus potential hardness) is via the 'quench hard, smash, then eye ball select' method. (This best documented in Japanese traditional methods.)
Given what was just discussed concerning slag content, it *may* be that the most effective carbon content would hardly be a standard, but might have to be considered in light overall quality of a given bloom.

My own modern experience with making steel strikers is hardly consistent. At first I followed the 'traditional' advise - which is to use an old file. Honestly, some of those worked well, others not very effective. Thinking back now, I can see the huge range in possible alloy types existing by using such unknown materials might easily have been a bit part of the problem. Better quality modern files are made of more complex alloys (Chrome, Vanadium, who knows?) beyond a simple approximate 1% carbon content.
I did spend an entire afternoon forging strikers, using differing metal alloys and variations in quench method with outdoors instructor Peter Ferri. After a number of not so great pieces, we settled on using commercial O1 (drill rod) stock, then oil hardening it from critical (measured with a magnet). This alloy and method produced the most consistent and effective results. At least for the batch of flints he brought (which yes, were all from the same point source.)

David Cox of DARC starting a fire using a flint & steel - L'Anse aux Meadows NHSC, August 2011

Placed back into our Viking Age context, I suspect what might REALLY happen would be this:
A given smith most often was working from a fixed geographical location. Although he did not (usually) smelt his own iron, I suspect access to iron bars would also be relatively local and regular. He would thus gain a direct experience with how best to work up his available source iron materials to a product most effective to the locally / regular source of stones used in combination. I can imagine him tinkering with his process against a given customer's personal stone.

Although there is a lot of speculation here, it may cast a new light on just what the Icelandic Jaspers found at L'Anse aux Meadows mean in the larger context of Norse material culture?

PS - Any readers who have better experience (or more effective!) producing steel strikers, PLEASE comment. Those with experience working with gun flints on actual antique firearms may actually have some important insights.

Wednesday, April 06, 2011

Final Report - African passive draw furnace

'Africa' - Smeltfest '11
Germinal Ironworks, Lexington VA
March 18 - 25, - 2011

A passive draw furnace from Burkina Faso.

Now available on the main Wareham Forge Iron Smelting documentation

This is an expansion from the earlier progress reports from several weeks back (seen here). The full report includes the smelt data sheet, drawings of the furnace, eventually some video segments.

Monday, April 04, 2011

Aristotle Furnace - Photo Essay

The Aristotle Furnace is a small re-melting furnace originally researched and prototyped by Skip Williams. It was further developed by the Early Iron Underground at Smeltfest in 2009.

I've been waiting for Skip to publish up a full working description of his work. (There are a few references on this blog in earlier posts.)

A number of the EIU were (lucky!) enough to attend the annual conference of the Historical Metallurgy Society in West Dean (UK), 2–3 September 2010.
From that event, Donald Wagner has put together a rather good photo essay showing the construction and operation of an Aristotle furnace by Lee Sauder, Steve Mankowski and Sheldon Browder.

The essay can be found HERE.

Friday, April 01, 2011

OUT of 'Africa'

- This is the third of my draft reports on the activities at this year's Smeltfest in Lexington Virginia with other members of the Early Iron Underground...

In the last report, we had stumbled into bed at roughly 5:30 in the morning. The last ore charges were added and the furnace was left to burn down untended. I was in considerably better shape than some, as I had crashed at midnight and then gotten back up at 4 AM.
The next morning, individuals arrived back up to the work site at random times - largely based on how much sleep they had gotten the night before. Jake and Lee were first on the scene, but it was about 10:45 AM before serious work on opening up the furnace started.

This is an 'down the kilt' shot, taken from inside the loading gate, down the shaft into the hot base of the furnace. You can see a mound of charcoal towards the centre still remaining. The hot spots to the left and right corners mark the location of the east and north sets of tuyeres (respectively).

The first slag block to be pulled free was the one to the larger north facing arch. The block is seen here rotated 90 degrees CCW from its initial position. The bases of the clay tuyeres are seen embedded into the slag, and broken off at the point they would have been surrounded by the initial loose clay packing to seal the arch. The glow seen in the interior of the left most tuyere reminds you that this block is still orange hot in the centre!

This is a view down inside the furnace from that same north side arch. The larger masses seen on top of the charcoal are actually pieces of the interior wall of the furnace, which fell down on top of the remaining charcoal and slag blocks. This primarily due to the whacking and prodding required to pull the slag block free. You can see that the base of the furnace has slumped down, the slag blocks burning down into the supporting sticks to about even with the ground level.

Here, Jake is working on prying up one of the slag blocks from the west arch. Some better indication of the heat can be seen. In fact, in the couple of minutes required for this operation, the wooden block used as a support here would catch on fire due to radiant heat. (Although it was a good 18 inches back from the furnace.)

Here you see that with Jake on tongs and Lee assisting with a bloom hook, this second slag mass is extracted. Again the bases of the broken off tuyeres can be seen in the lower edge of the block.

Although there was some expectations of a large bloom mass (for Jake at least) no large bloom was found. A number of smaller, roughly tennis ball sized, fragments were broken out from within the various slag blocks.

This process of recording then breaking up slag to recover potential iron metal was still underway when I left Smeltfest on Saturday morning. When some final measurements are available I will finish this series of reports.

February 15 - May 15, 2012 : Supported by a Crafts Projects - Creation and Development Grant

COPYRIGHT NOTICE - All posted text and images @ Darrell Markewitz.
No duplication, in whole or in part, is permitted without the author's expressed written permission.
For a detailed copyright statement : go HERE