Findings with Steel

It all began several years ago when I was first branching out of smelting. I was enamored by the process the Japanese used to create smaller chunks of steel through the remelt process known as Oroshigane.
This process is employed to this day by Swordsmiths in Japan in order to mix in with their Tamahagane smelted steel used for swords. This melted steel product can either increase or decrease the overall carbon content in the steel while not diminishing its quality or appearance. Oroshigane is also called hearth steel in other areas of the world and the naming is used interchangeably among practitioners.
After dozens of separate melts in, I was starting to play with the parameters of a common furnace found in antiquity modeled after a man named Ole Evenstad. I felt his methods of melting steel to be superior to that of Aristotle’s methodology, and so I mostly chose his furnace design to play with.
Changing certain parameters of the furnace yielded different results quite often. Sometimes its a tweak in floor depth, or air pressure, or charcoal size, or so many things that didnt seem to matter before, you find they actually do. It was finding a balance of things that finally allowed me to yield mostly what I wanted, and eventually lead me to a chase.
After a few years of melting in a hearth I ran across a scenario that created the most beautiful chunk of steel I have ever produced. It is the colorful chunk you see now and again on my page. That piece of steel caused me to spend the next 2 years trying to reproduce. The colors found in the grain are simply oxides on the surface of clean, newly formed steel. when the steel solidifies in the presence of O2 at a certain temperature, surface oxides are produced. At the time I thought it was a visual marvel, a nod to the term “jewel steel”.
For all of its beauty, that chunk of steel was hiding something that wasnt yet realized. Hiding in that chunk of steel and countless others was the presence of dendrites. Beautiful shards of cementite (Iron Carbide) lay scattered all across the surface. I thought this was quite a sight later on, but it wasnt so quickly revealed to me.
In my race to recreate the most beautiful piece of steel, I had ran into problems. Chief among them was the formation of pure iron on the tops of my chunks of ultra high carbon steel. My goal was to make steel worthy of a Japanese sword, and I knew a mix of iron would only bring down my overall carbon content. For the next 6 melts this problem persisted. I couldnt wrestle it out by just running again. I had to figure out my problem.
It was one evening while I was skimming through a book written by a Japanese sword smith describing the way the Tatara in the Shimane prefecture of Japan was ran. The Tatara being the large smelter than produces the country’s supply of precious Tamahagane steel for the country’s swordsmiths. It described how the furnace entered a phase of carburization. It was in this reading that I had an idea.
Once I was able to set up my furnace for another melt, I employed what I had learned to my small furnace. In my mind, this furnace is acting in a similar manner. While this isnt a slow, day long carburization process, it was a fast hour long carburization process. The principles remain the same. My process takes an hour because carbon is absorbed into the iron or steel much faster when said iron or steel is a liquid. In this furnace I am melting iron into a full liquid in the presence of a carburizing atmosphere (burning charcoal). Obviously the time it is liquid and the more carburizing my atmosphere is, the more carbon uptake.
The resulting work that followed started to really break the steel/white cast iron barrier, and the further I perfected my process, the easier it was to make a fully ultra high carbon product.
Eventually I started cutting these chunks cold and polish/etch the surfaces to inspect what was going on. Coupled with a slow cool, I started to produce iron carbide steel phases the reminded me of the dendritic character of crucible steel. I was producing coarse crystals of cementite among pearlite. Cementite being the pure iron carbide in the form of coarse crystals separate from grain boundaries in these examples. Pearlite being the room temperature crystal from containing ferrite and cementite up to around .78% Carbon. With reheating and cooling, these coarse crystals of cementite go into solution and precipitate back in the grain boundaries.
After using this material in a tekogane stack of flattened steel, and working it in the forge, witnessing its carbon decarb, I felt that I had comfortably made ultra high carbon steel throughout the chunks each and every melt. I felt I was regularly making steel in the 1.3-2% C range. These were perfect ranges for crucible steel I thought. I postulated for awhile that if I could without any technical equipment, consistently made a steel product in the 1.3-2% Carbon range, then I feel like the ancients could have as well. The ability to perfect parameters to be repeatable seems to me quite a lot eaiser than knowing how much charcoal to add to a crucible to up the C range of mixed quality bloomery iron/steel, as an ancient, with no knowledge of Carbon in steel.
It just made more sense to me to take a clean steel at generally the carbon level I wanted, but always created in the hearth because the end product of a hearth is more easily high in carbon than not. It could have on its own as a folded product make a high quality sword already. Why not with these queues would it be a giant leap to use it in a crucible?
Of course I took a hearth melted chunk and stuffed it in a crucible. My very first and next 10 melts in a crucible to produce water patterned crucible steel used my hearth melts as the feed material. No additives were used, and even using standard modern day clean steel as the original feed material in the hearth, I still had highly beneficial alloying elements in high enough concentration to act as carbide forming nodes in the steel to assist in pattern creation in the process of forging crucible steel.
I was able to secure elemental analysis of my crucible steel over time as I had worked out a deal with a university student in Canada. In one hand I was also extremely curious if the hearth melting process also strips Manganese. It was highly speculative if melting Mn rich steel in the hearth actually strips it in the final product. The presence of Mn actually assists in hardening. Without it, the steel would have been similarly reactive to steels used in antiquity, as Mn wasnt introduced to steel until the industrial revolution. As far as I know, there had not been a conclusive test to show Mn stripping, until I had it done.
It was after a few months that I had received my answer to the Mn idea, and if there were alloying elements present in the steel, including the correct concentration of carbon. I was pleased to find out that every single sample sent in, Mn was either undedetectable or of a concentration too low to notice. This meant hearth steel was suitable for shallow heardening work in the form of Japanese style swords. I also found out I had many elements present. Mo, Cr, Ni, W and especially V. The hearth strips some things and barely effect others. Sulphur seems to lower, but depending on how it is ran Phosphorous is either lowered or unaffected.
Of all information, Carbon level was the most satisfying. The most successful bar I had forged at the time had a carbon level right at 1.64%. A perfect concentration for crucible steel. I managed to make other crucible pucks at 2% and even higher, which resulted in white cast iron and would not properly work to create a watered pattern. These runs were re-melts of pucks that I had made with incorrect forging. I was experimenting on how much higher I could get with the carbon, not understanding that sparks and grain of lower end white cast are almost identical to UHC steel without the aid of a proper scope, i kept making overly high carbon material.
Pretty much in my experience now, and setting up my furnace parameters the same as a previous run, I can get pretty consistent 1.3-2% material. I have found a knack with hitting 1.5-1.8% on average now however. I feel like discerning between that and guessing the same way with bloomery product, or figuring out how much charcoal powder to add to bloomery in a crucible to nail a specific carbon content is far easier. No additives. Just insert steel that is already more or less at the target carbon concentration just in an undesirable form if you want watered crucible steel and melt it. The melting point is rather low and the melting run is relatively fast. Almost twice as fast in my experience than if you were to use a combination of cast iron and pure iron in the mix to hit target carbon percentage.
Im not trying to sell everyone on the use of hearth steel being the historic way of making crucible steel. I just feel out of the many methods I have read about and seen done, I feel like this method is rather simple and effective. I still have a long way to go with working crucible steel, but I thought chemical analysis and getting out my realizations is worth writing about. Its technically another way to produce crucible steel. With a few easy methodology changes in running the hearth or by simply forging the raw hearth steel down and with a few welding passes, I have found Phosphorous levels to greatly decrease all the way down to .001% levels like that have been found in folded hearth steel samples I had tested. Results shows to have greatly decreased P levels between raw hearth steel and product that has been folded 4 times.
My research also proved the stripping of Manganese from the parent steel. Some melts I was using feed material that had .8% Mn and by the time the melt was over, the mass spec could not find anything beyond trace, or <.001%!
I have also been able to grow comfortable with the material and where my carbon estimations have been landing. Looking at averages of my tests and some averages found in a historic site that I do not want to mention, 1.5-1.8% Carbon is rather common in these furnaces, even in antiquity.

Dendritic Oroshigane. Coarse cementite. -Ferric Chloride etch

More Cementite structures found in hearth steel (In this case white cast) -Ferric Chloride etch


Crucible steel at 2.06% C made using hearth steel as feed material. This is a cross section of puck showing grain boundary cementite and widmanstatten cementite. -Ferric Chloride etch



  • 2.06% C
  • .212% Si
  • .025% Mn
  • .045% P
  • .014% S
  • .026% Cr
  • .111% Ni
  • .010% Mo
  • .015% Co
  • .0042% V
  • .0073% W







Successful forging of crucible steel ingot made purely from hearth steel/Oroshigane









Chemical Analysis of above watered example

  • 1.64% C
  • .051% Si
  • <.0010% Mn
  • .042% P
  • .014% S
  • .019% Cr
  • .142% Ni
  • .0079% Mo
  • .021% Co
  • .003% V