This will describe the building and running of a tatara-like smelter. Operating the tatara alone proved to be exhausting and very gratifying in the end when steel was recovered from the bottom of the smelter.
Charcoal is the fuel, the heat source and the source of carbon for the smelter.Charcoal has to be chopped to a specific size for the operation to run smoothly.The bore of the furnace is 11 inches so that seems to be the right size for the charcoal: 8-10% of the inner diameter of the furnace.
The ore is from a ceramic supply place. The name of this particular compound is Spanish Red. It is a light dust that stains everything almost permanently red and contains 80% hematite.
The furnace design is based on the shell of a water heater. It has been cut in three sections for easier assembly and dis assembly. Here is a diagram.
The parts were cast with refractory made mostly of sand and cement with an inner lining of high temperature furnace cement. It held up pretty well to the heat but crumbled at the time of taking it apart.
The tuyeres were made of simple black iron pipe fittings. They stayed cool by the constant air flow through them and since they don’t protrude in the inner chamber they don’t become melted away.
Here is an additional picture of the bottom part of the assembly after pouring the refractory and letting it dry. It is shaped like a funnel as in the Japanese tatara
View of the assembled parts half way through the construction
Here is the setup with the smelter put together and starting the pre-heat.
Covered the smelter in refractory blanket to increase the thermal efficiency. That actually worked really well. The temperature measured at the bottom of the top third of the smelter reads 1000 degrees F.
Home made manometer to gauge the air flow pressure. The manometer was indirectly reading the amount of air flow very accurately. The air flow was kept at about 15%.
Here it is running on charcoal now and with the first charge in.The temperature reads 1800 degrees F and it remained there or above for most part of the run.Charges started at 1/2 kilogram of ore and half the way through the run went up to 1.25 kilograms. Using about 2 kilograms of charcoal per charge.
The manometer is indicating the change in pressure during the run.
Bubbling at the tuyeres. Now taping the arches and keeping the tuyeres open and trying to keep up with the schedule of charges.
Even at this point 2 hours into the run there are lots of sparks coming out of the tap arches.
Now started to take it apart. The top third of the smelter lift off easily but the refractory crumble to pieces on the rims. It held very well in the inner chamber. So much from trying to have a system that was reusable! The parts will need to be repaired each time.
The middle third has now been removed as well as the tuyeres.
After cleaning up the charcoal a little bloom is peaking inside. It looked like Mount Fuji covered in snow.
After cleaning up all the debris and removing the bloom it turned out that the walls held up pretty well.
Here is the bloom. A total of 5 pounds of steel were consolidated and lots of smaller chunks were recuperated with tongs or with a magnet
The spark test looked pretty good.
This is the largest chunk of the bloom divided in two with the band saw. Resembles a Rorschach Ink Blot Test. What do you see?
After polishing the surface and etching a bit you can see all these beautiful dendrites. This is an indication that the core of the bloom had reached liquidus.
PART TWO: WORKING WITH TAMAHAGANE
The second part will deal with the processing of the KERA or BLOOM, including compacting and forming into a bar of steel.
At this point about 15# of steel combined with slag and other debris. The next step is to consolidate the different chunks of steel into solid bars.Don't plan to use all of the steel. Save one of the halves of the bloom as a memento. Also set aside some pieces for chemical analysis.
Sorted the steel using the spark test. High carbon, medium carbon and lower carbon sections separated and ready for the forge.
Put a few chunks in the forge and waited to reach temp to start consolidating. A few remarks about working with this material: WOW! It welds at lower temperatures than steel from the mill. It uses less PSI from Venturi burner. It moves like butter and is self-fluxing. It sticks to itself so easily.
Flywheel press to flatten some of the chunks. Different strategies for the consolidation. For the smaller chunks flatten first then stack them on top of each other and weld them together.
For the smallest pieces simply stack them on top of a larger chunk that had already flatten. The properties of this steel allowed to weld those onto the larger chunk with such ease, keep using this method to grow the billet. At this point not done any folding, Simply growing the billet by adding little chunks and consolidating into a larger piece.
When the billet was big enough add more and more of the little pieces on top and work the processes of welding and consolidating at the same time. The amount of time spent in doing this was less than the amount of time normally spend in forge-welding regular steel. Since it sticks together so well and requires a lower temp to weld it also needed less time to soak in the forge to reach the right temperature. That will hopefully result in less carbon loss.
Here you can see a little chunk that added on top of the bar getting quickly hot enough to weld it to the bar.
The bar was growing in size quite rapidly.
To work the biggest chunks from the smelt use a variation of the previous strategy. Another interesting thing is that these larger pieces require a lot less work and seem to want to stick together and consolidate even better than the little chunks.
Here the large piece has reached enough temperature to start working with it.
Square the whole thing in one pass but now using the hydraulic press for the larger pieces.
Then once nice bar formed , add more little chunks to it grow it larger. Keeping the bars separated into 2 categories according to the carbon content based on the spark test mixing high and medium carbon chunks into what would call higher carbon bars and mixing medium and lower carbon chunks into lower carbon bars. Also separated the bars made directly from the main bloom.
Adding some more pieces.
The big bar is at the bottom and on top there are multiple irregular size chunks that will be welded on to the bar at the same time that the consolidation of the bar takes place.
A better view of the little chunks on top of the growing bar that is already consolidated.
After all that hot work here is the result.
Lower carbon bars, higher carbon and the bars that came from the larger portions of the bloom itself.
PART THREE: WELDING AND FORGING OROSHIGANE
Manipulating the carbon content of the steel to create different quality bars of steel.
Traditionally oroshigane is the name of two different processes that take place at the Japanese smithy in the preparation of the steel for welding and it also has extended its meaning to the steel made in this way. Since not all the steel that comes from the tatara has the ideal carbon content (TAMAHAGANE) the Japanese smith will either by reduction or oxidation alter the carbon content of those pieces of steel that have either too little or too much carbon and adjust it to the desired level to make either shingane (core steel) or kawagane (jacket steel). Chemical analysis of the steel:
Made several bars of steel , have ranked according to estimated carbon content or whether they came from the larger bloom (kera) or from other chunks of steel from the smelter. The bars will be stacked up together alternating according to carbon content to hopefully create some contrast in the final billet.
By weight have about 7# of steel after processing all the chunks.
Divide the bars in half and make two separate billets. The first billet will used primarily to make two tantos, use to experiment with the heat treating process. Here is a picture of the bars bound with wire. Did not tack-weld with the MIG so as to no introduce any "foreign" steel. The wires will be removed .They are there just to hold the bars in place temporarily and get the welds started.
In to the forge it goes.
Flipped on the side to even out the heat.
By this picture already made the first welding pass and removed the wire and did a second pass to consolidate the bar and draw it out a bit.
Used the press to speed up the drawing out. Thenhammer it straight.
And it is back in the forge to get hot enough to split it in the middle.
Put the first fold into it.
Add flux at this point to ensure that the welds go well.
Back in the forge to get hot for welding.
At this point it has reached the right temperature and taking the bar out of the forge and to the press for welding. Sparks are flying off the steel. Maybe that was a little too hot but better safe than sorry.
After that you do it again and again, as many times as you want. It was 5 folds.By the second fold the steel was behaving like regular factory steel and needed higher psi to weld and definitively flux to complete the welds.
The last step was to draw out the bar.
The bar is now split in two halves to make the two tanto.
Polished and etched one section of the bar to see the grain. The face of the bar will have itame grain (random pattern). The pattern is very subtle. Very little contrast. The color of the steel is different from refinery steel. It looks grey-blue. Similar to blades from Japan made during the koto period. Quite nice actually.
The side of the bar will have a straight laminate pattern (masame grain). Again very subtle and difficult to photograph.
PART FOUR: FORGING THE SUNOBE INTO A BLADE
Forge-welding and folding of the different steel bars to create a billet and forging it into a SUNOBE, then into a blade.
The next step is to make those two steel bars into blades.Start by manipulating the pattern in the steel. This is not a foreign concept to the Japanese smith. They use it commonly to make mokume out of soft metals and some schools use it to enhance the visual appearance of the steel grain in their blades. Most commonly the Gassan school in creating their ayasugi hada. Here is the plan in drawing. It is a modified ladder pattern that generates the wavy look of the ayasugi hada.
First lay out marks on the steel bar.Doing this to one bar only.
Used a 3/8 round file to cut the grooves.
In to the forge it goes.
Reversed the tip of the sunobe to make the kissaki of the blade. What the Japanese call sunobe is not other than a steel bar that has been modified to have the distal taper and width taper in it thus making it easier to forge the flat faces and bevels from it.
Here is how it looks.
A little bit ahead in the shaping.
And further ahead. Now working the bevels in. Meaning the distal and width tapers are already forged.
Forging the bevels at the tip.
Finish on the nakago (tang). At this point choose to not forge the blade any further. It is easier to set the machi (notches) for both the ha (edge side) and mune (spine side) on the grinder.
Go ahead and thermal cycle the blade 3 times at this point. Start above austenizing temperature for the first cycle and just at austenizing temperature for the last cycle.
Here you can see the blade pass the recalescence point in a sequence of pictures.
Forged the other blade and thermal cycled just the same way. So now two identical blades ready for the grinder.
No pictures of the grinding process. There is nothing of interest there. These are the two blades at the end of grinding. They both have 220 grit grinding marks along the length of the blade. The mune machi and ha machi areas have been cut in and the mune has been beveled as well. They are ready for the clay.
In case you wonder how much they have "grown" from forging.
PART FIVE: THERMAL PROCESSING - YAKI IRE
Heat treat: Clay-coating, hardening and tempering of the blade.
This is the critical step in developing the beauty and functionality of the blade. Most will refer to this process as hardening and it started already when thermal cycled the blade in what has become known as normalizing. At this point clay-coat the blade in preparation for quenching and hardening the blade. That is called differential hardening (hard edge - soft back). And to finish the process the blade will be tempered or drawn back a little to reduce the brittleness of the edge.
In order to gain as much knowledge as possible from this steel planning on using three different clays applied to the blades in different sections and with different thicknesses. Also vary the soak time at austenizing temperature in the heat treating forge to see how that affects the steel.
Start by mixing own clay formula: 50% fireclay + 50% charcoal.
Developed a set of simple tools for laying down the clay: a brush and a spatula with some shaping of the tips to form the clay. A little plastic box to mix the clay to the consistence needed. A hard plastic board to serve as an easel.
Here mixing clay to a thick paste consistency first.
The second formula using is well-known by everyone: satanite.
And the third formula is based on a recommendation from Don Fogg: an anti-scale compound that is quite plastic and easy to work with.
Applying the satanite to the back of the blades. One blade will have a thick coat and the other a thinner coat.
Then Don's anti-scale compound. Thick coat and thin coat again
Now formula for the kissaki. A tricolor blade.
After letting the clay dry over night quench on the next evening. There will be no pictures of the actual quench. But here is a picture of the setup. Long heat treating oven made out of a water tank, a pyrometer to guide temperature. The final decision as to when the blade is ready to be quenched is a visual one based on the colors of the blade and not the reading on the pyrometer. And a fish tank for tanto-size blades. The fish tank contains water saturated with salt and some soap at about 110 degree F. One blade was heated to 1450 degrees F and the temperature was held there for 5 minutes then quenched. The other blade was quickly brought up to 1480 degrees and then quenched.
Another useful tool is a blade holder. A piece of rebar to which welded two attachments on the ends to hold the blade. One small sized for tantos and the other for larger blades.
Here is a detail view of the holding attachment:
And at this point yaki ire (quenching) has been completed and the blades are in the oven cooking at 300 degrees F for one hour and a half to temper the edge and reduce the brittleness of the martensitic edge. Prior to putting them in the oven checked the edge with a file to see that they hardened properly.
Next morning this is what they look like.
After a quick clean-up noticed that the satanite and the anti-scale compound did not leave any residue but clay mix interacted with the steel a little deeper and left some residue (carburization?).
Now two twin blades ready for polish.
PART SIX: POLISHING
In polishing a blade there is not one process that fits all. Over years of practice you develop a technique that works for you and for the kind of steel that you use. You change your polishing techniques according to what you want the final effect to be. In this particular case venturing into new territory. Using a kind of steel that have not polished before. In general When polishing Japanese style blades or any blade for which you intend to display a hamon, there are two general ways to proceed: one is to use stones (whether synthetic or expensive natural ones) and go through the different stones according to a set sequence and the other is to use wet/dry sandpaper and progress along the different grits. What really changes everything is the final step in polishing, what the traditional Japanese polishers accomplish with the hazuya, jizuya stones and nugui. For the Western, non-traditional or hybrid polish or whatever you want to call it, this is usually a combination of a mild etchant (ferric, vinegar, lemon juice) followed by a cleaning of the oxides left by the acid with a paste polish (Pikal, Simi chrome, etc) and finally a bit more polishing with pumice or other compound mixed in oil or water and backed with cotton or leather. It is a process of trial and error that tends to be different for each blade.
Show you first simplistic setup. You can polish by moving the blade on a fixed stone or polishing board to which you would attach the sanding paper or you can keep the blade fixed and move the stone or the paper along the blade. Choose to use sandpaper and to keep the blade fixed. Use a piece of corian to back paper and hold it by pinching the paper against the corian between my index and thumb fingers. That is what works. Get plenty of light coming from two different sources aimed at different angles on the blade and clamp the blade to a table near the edge with quick-grip vises.
Finished the blade on the grinder with a 220 grit along the length of the blade.
Depending on how well the machined 220 grit looks start with either 180 or 220 grit paper. In this case an inspired day at the grinder so start with 220 grit paper. Work at a 45 degree angle so that can clearly see the scratches from the previous round. The goal is to evenly remove all of those scratches completely. This tanto is hira zukuri so don’t need to worry about a shinogi line. The hand-worked 220 grit looks quite different from the machined 220 as you can see.
Then move to 400 grit paper and work at 90 degree angle from the previous marks. Again, there are many ways to do this and some people work first at 90 degree across the blade and progressively tilt the angle until at the very last steps they are only working along the blade. Prefer to alternate each pass at 90 degrees from each other and switch to working along the blade at the end.
Once finished with the 400 grit stop there and start working on the fittings (koshira). Some slight damage may occur to the blade during the fabrication of the koshira and it is easier to erase those by redoing the 400 grit than having to redo all the steps if took the blade to a final polish before doing the koshira. For these two blades make two copper habaki and a two shirasaya.
Apply a patina to these two habaki to take them up a notch. The patina is called niage and it is based on the Japanese rokusho patination with the addition of copper sulfate. Niage means “to boil” and immerse the habaki in a heated coffee mug with the solution for two hours depending on the desired color. Copper put to this process goes from yellow to brown to orange to red to dark brown. Stop the patination at orange-red.
Here is the result.
Made two shirasaya out of poplar and alder.
They will be coated with a few layers of clear lacquer.
And a sayagaki (inscription) will be printed on them
Now continue the polish with the 600 grit paper working at 90 degree angle from the previous marks.
Then 800 grit paper working at 90 degree angle from the previous marks.
Then move to 1000 grit paper and switch to working along the length of the blade from now on. The surface is becoming more mirror-like but no signs of a hamon yet. By the way at this point adding some water to make a mud that acts as a lubricant and polishing paste of sorts.
Then move to 1500 grit paper and continue to work along the length of the blade but stacked a piece of leather between the paper and the corian and continue to use water as lubricant. As you can see in the pictures the hada has now become visible to the eye of the camera.
Here is a picture of the polishing “tool”. Corian, leather and paper with the water slurry on the blade.
Continue to use the same backing and move up to 2500 grit paper. The hada is visible but no hamon. Something had gone wrong.
Etch the blade at this point rather than using the hazuya and jizuya stones. Need to etch anyhow since worried that something did not work well during heat treatment. Here is setup for this process. Have a swivel vise that allows to flip the blade for one side to the other quickly and prepare a solution of lemon juice, vinegar and soap to be heated up in coffee mug.
Use a cotton ball to first clean up the blade with Fantastic and then apply the hot etching solution.
And the bad news: the hamon shows up as a very thin line along the edge. About 3/16 of an inch. The file test after the quench was fine and the blades survived my usual cutting tests (wood chopping, hair shaving) before it had harden. Just wasn’t expecting this shallow hardening.
Continue with polishing as wanted to figure out the steps to bring the hada out with this steel..Here is bag of tricks for this. First use a ceramic polishing compound that contains citric acid backed with a 3M polishing pad. Follow that with an automotive industry polishing paste backed on the same pad. Then use a clean sanding pad to clean up all the residues. Follow that with nugui (iron oxide) diluted in cutting fluid. Then pumice backed with leather to clean up the residues. follow with 2500 grit paper with water first and then dry and then go back to the 3M soft sanding pad. This sequence is specific to these blades and worked this one out through trial and error. Each blade depending on the steel they are made of and the kind of heat treatment tends to behave differently at this stage
And these are the results: First the tanto with the itame hada. 3 pictures follow.
Second the tanto with the ayasugi hada. 3 pictures follow.
A final picture.
Sunday, February 20, 2011
60 responses to "Tatara Project : Guy makes blade out of homemade steel (125 Pics)"
I visited a swordmaker in Ryujin, Wakayama, in 1995-6 and made a Hassun Tanto with him. We made steel in a very similar way to you, using a method developed by his father. This was unusual, as opposed to buying from the central smelter.
Thanks again. The tanto is my most cherished possesion and that time is one of my most cherished memories. As a matter of fact, we once ate a rice bowl with fresh-caught eel grilled over coals from the forge. The first time I ate eel on rice after returning to Canada, I cried.
My good man, you are some kind of lunatic... taken in a good sense of the word! :-) That means you definitely don't lack dedication and handskill. Had I had some more time on my hands, I would gladly replicate these above steps to get my own tamahagane steel... until then, I just make knives out of factory steel. You are a genius, keep up the good work!
This is very interesting. However I don't see any recognition of the person who did this. Maybe I am missing it somewhere but I don't find it. His name is Jesus Hernandez and I think it would be appropriate to prominently give him credit for his work.
I enjoyed this sir and added your work to my knowledge. Congradulations on a job well done. One question, is there a better way (more time efficient) to reclaim ore into steel or iron in a home base or small shop enviroment?
Very well done sir. I am an amateur knife maker. You have given me inspiration to be a better bladesmith. I will definitely be studying up on making my own steel now. Japanese blades are what got me interested in forging in the first place. Time to take it to the next level.