What happens after hearth melting

Last week I got together with Prof. Justin Fermann of the University of Massachusetts Amherst chemistry department, a great enthusiast for craft and obviously a learned chemist.  For all his technical knowledge, none of the magic is lost on him, and he actively pursues blacksmithing, glassblowing, pottery, and brewing, fully content with neither the science nor the practice. 

What we met to do was to take the hearth-melted material that I made with Jack McAuliffe, in many ways indistinguishable from ancient, home-made bloomery (iron smelted directly from iron ore), and compact its porous, spongy mass into a bar without seams or cracks that would compromise its usability in a blade.  

In order to squish all the holes shut and seal up the cracks, we had to first flatten all the little pieces we had at a yellow-white welding heat, in order to squeeze out the slag leftover from the charcoal furnace and fuse the voids shut.  This necessitated a very hot fire, so we burned charcoal in a coal forge and used a hand-crank blower.  Charcoal burns extremely hot and fast, and we hoped it would also lend some of its carbon to the steel's molecular matrix to help ease of welding and to give us a more hardenable product.

Slowly, we flattened all the pieces and then stacked them on top of each other.  At first, we had a very frustrating time welding each piece to the other. Justin would crank the blower and when we deemed it hot enough to weld, he would grab the pieces with two pairs of tongs and lay the hot ends on top of each other.  I would tap that end lightly with a smaller hammer in order to tack the weld.  After the end was tacked, on the next heat I would fully set the weld with a sledgehammer.

We worked from one end to the other, welding, scraping, fluxing, heating, welding.  We first fused two pieces, then began to add.  We had four small pucks of the hearth material Jack and I had made.  I had several similarly flattened cakes of Jeff Pringle's bloomery material about equal in mass, so we added those too.  We layered them all together, and once all of our flat pancakes were welded into a solid stack, I turned them into a bar shape on the power hammer.  The consolidated but unrefined bar cooled down and weighed in at 657 grams.  Completely covered with charcoal dust and little burns but elated with our learning curve and general level of success, Justin and I slurped down a bunch of water and called it a day. 


The next day I threw our bar in the propane forge and began to refine it.  Refining is essentially heating it up to welding heat again, and further solidifying welds and expelling slag by forging the bar out, folding it on itself multiple times, welding the fold seams shut, and drawing the bar out again, kind of like stretching and folding layers of play-doh. They all seemed to take, though I had a few minor bubble issues, but eventually I had a flat bar with no weld flaws.  I can't truly count the number of folds because many of them were half-folds.  It still happened though, and once my flaws were gone I let it cool again and put it on the scale.  454 grams meant that I had about 69% of the bar that I initially weighed.  I wish I'd weighed the material I had before welding any of it together, but I totally didn't.  So there.

Where did the weight go?  I'm guessing there were two main causes of material loss.  The first is slag.  The slag is mostly composed of various silicates from the solid fuel and the material the furnace is made off.  These are rock-like bits that melt while the iron is still solid at high temperatures, and as liquid, they fill up the cracks and get squeezed out when you hammer and fold.  The other cause of material loss was probably just oxidation.  In such a hot fire, iron converts incredibly quickly into iron oxide (in the form of black scale, a kind of rust), and the amount of welding heats I had to take probably reduced a lot of the material into that.

It just so happened that at MassArt last year I consolidated a bar of bloomery iron from when I was in England with Owen Bush.  It had almost the same proportions as my new bar (on the right in the above photo), so I decided to polish the two, and etch them to compare their pattern and what little I could tell about refinement/slag content, carbon content, and anything else.  The etch was much more revealing that I thought it would be. 

The left one is Owen's bloom.  The right one is the hearth material I made with Jack mixed with Jeff's bloom. Ours has a more complex pattern probably due to more slag and somehow less homogenous material.  Ours is also shinier.  What that means is that Owen's bloom, with the cloudy, dark parts, has much more carbon in it.  Carbon-rich steel etches dark (unless it has nickel or chromium in it, which this does not), and that's just something I know from experience.  If I were to pick one of these for a knife or a tool edge, Owen's would clearly be the better choice. 

 But I have to say that from a purely aesthetic perspective, the less homogenous steel has an incredible appeal.  Also, though I had a hand in making Owen's bloomery steel, the intimate involvement in the creation of the other one was so fresh in my hands and mind that I couldn't stop looking at it, couldn't stop looking deeper into its random matrix.

My favorite thing about this material is that it takes so long to make and is so imperfect that along with the slag and weld lines, it packs memories, conversations, frustration, and cooperation in between the layers.  It's a trophy of mutual discovery and a sedimentary painting of process.  It's also a little bar-shaped window into the random pattern-storms that only nature can make, full of beautiful tree-diseases and river-paths seen from on high.  It exposes the illusion of imposed order, as well as the cooperation of the chaos that is the true nature of order. 

Birch-moon

This knife belongs to a close friend, who spends a lot of time on the border between nature and dream.  This knife is cloudy contemplation and rainy dronescapes and layers on layers of sky, dirt, and the mist in between.  It is birch and moon, silver liminal aethereal gateways.  Check out its owner's dream-wave creations in this land. 

The blade is forged of Aldo Bruno's 1095, clay-hardened under the supervision of Matt Venier.  The handle is carved moose antler, cow horn, desert ironwood Jeff Pringle found, and stacked birch bark that my uncle collected.  The ferrule is silver and the pommel cap is salvaged wrought iron from Warner, New Hampshire.

The leather is tooled using only a knife and punches that I forged and filed out of 1084 (except my lantern touchmark).  All organic materials are fully sealed with a beeswax-linseed oil mix, for misty morning contemplations and solitary transcendence of time. 

More Hearth-Melting!

A few weeks ago, I did another round of hearth-melting at my friend Jack McAuliffe's shop in Worcester, MA.  He's definitely got this process down by now, having produced a number of daring and incredibly advanced Roman blades of various periods out of this character-rich iron of unparalleled complexity and historicity. 

Scraps for melting on the right, consolidated cakes on the left

For those who don't know, hearth-melting is basically iron-age steel recycling.  The idea is you can build a fire in a furnace, chuck in all of your iron scraps that are too small to forge-weld together and semi-melt them into a porous cake.  Think like a scone, kind of.  There are many small furnaces that have been excavated across the Indo-European sphere of iron-working which were too small or temporary to accommodate full-scale reduction of raw ores into iron.  Other archaeological evidence and some primary accounts indicate that nails and other scraps were compiled and slowly fed to the furnace, consolidating into a cake that can be forged into a bar. 


The other purpose of hearth-melting is to change the chemical makeup of a batch of steel.  There in the furnace at temperatures between welding and melting, where fundamental bonds dissolve and form in a whirlwind dance of elemental cataclysm, iron, carbon, oxygen and other ingredients  strap together and tear apart.  The product of this storm is determined by overall mixture of fuel and air, temperature and time. The fuel is charcoal and the air is forced into the side by bellows or blower, though a slanted pipe called a tuyere.

We built our furnace out of easily-disassembled firebricks under Jack's coal-forge hood.  There was a bed of ash inside it and a hole in one of the bricks for the tuyere to poke though.  We built a fire with wood and then started piling on the charcoal once it was going.  After that, we started adding our scraps piece by piece.  Mostly we had broken up old saw blades, pitchfork tines, and other pieces of formerly high-carbon steel.  It was sort of an experiment to see how high-carbon our product would be if we used high-carbon source material, but our conclusion was that the environment of the fire mattered more than the material. 

So, once we figured we had loaded enough source material into the furnace that we'd accumulated a good amount of mushy iron batter at the bottom, we stopped adding charcoal and let the fire burn down a bit.  We removed bricks from the side and dug out the little cake.  Jack handled it with tongs and laid it on the anvil, and I slammed on it with the sledge.  We had to be a little bit careful because what comes out can be brittle and portions of it can crumble off.  The most troublesome part is getting the extremities to weld on; usually they cool too fast to do that.  But we've learned that there's a significant amount of material loss, and the best thing to do is just re-melt whatever falls off. 

After we'd manageably flattened our little steelcakes, Jack cut them in half on the power hammer.  The idea was to expose the interior of the puck so we could spark test the carbon content on both the exterior and interior of the puck.  We also split it up so we'd have more pieces to layer and fold, in order to better homogenize the material.  Good thing we did that, because the steel we made ended up being incredibly non-homogenous.  The spark test showed very, very high carbon content on the outside of the puck and surprisingly rather low on the inside. 

That probably meant that fire was hot and oxidizing enough in places that the carbon burned out of what we put in, especially in the material that remained in the furnace for a time and became the center of the cake.  However, the extremely high carbon casing indicated that there was a significant amount of carbon migration into the material from the charcoal fuel.  The fact that it was not very deep illustrated that carbon can saturate iron to make a very high-carbon jacket, but penetration depth takes a significant amount of time, perhaps at lower temperatures so it is not also burning off at the same time.  Either way, that is how we concluded that regardless of the carbon content of the starting material, most of it burns out due to the extreme heat and oxidation, and any carburization that happens is the result of cementation from the charcoal.  Either way, it might follow to use higher carbon stock anyway, simply knowing that it will melt faster and more fully (more carbon means a lower melting temperature; castable iron is around 4% carbon and utterly unforgeable; malleable wrought iron has almost no carbon and takes very extreme temperatures to melt). 

The other variable we did not experiment with or research much was the shape of oxidation and reduction zones in the furnace as determined by tuyere height and angle.  That we'll have to save for the future.  As it stands now, I have a few small cakes of homemade steel to play with, and there's challenges ahead but lots of promising and exciting work to do!