Damascus and Pattern Welding

True damascus steel is made by a casting process utilizing very high carbon steel. The material is held at or near the temperature of fusion for an extended period of time. This causes the formation of large, high carbon dendrites in a matrix of approximately eutectoid steel.

This is a process of growing crystals in the molten metal, and no welding is involved. This material, called "wootz," is used in the traditional Persian blades of Indian steel, to which the trade in Damascus gave its name. The term Damascus is incorrectly applied to the welded, patterned steels that have become common in North American cutlery during the last decade. To try and encourage the distinction between the two processes, I term the welded metal "pattern welded," and I would like to encourage the use of this term. It may seem a small point, but smiths have been "discovering" the "lost secrets of Damascus" for nearly 200 years. This distinction has caused confusion in every generation of metalsmith since.

Pattern welding is the process of welding together plates of high, low and sometimes intermediate carbon content steels, drawing them out, cutting and rewelding repeatedly until the desired fineness of material distribution is achieved. The number of layers increases nearly exponentially with the number of welds. However, some material is lost from the surface due to firescale and surface cleaning processes.

I was originally attracted to this process partly by the mythology that  has grown up around it and partly for esthetic reasons. On the practical side, much has been claimed for the material with regards to its edge-holding ability ("the sword that never dulls"), cutting ability, toughness and strength. While it is true that combining billets of different compositions can produce a blade that is excellent in each of these categories, I have come to believe that the performance of pattern welded material can be equaled or surpassed by the best of the modern homogenous tool steels, even without taking into consideration their functional advantage of stain resistance.

Pattern welding for me addresses the issue of the static and formal effect produced by mirror and brushed finished, which I feel sacrifices the vitality of the objects to perfectionism and concerns of detail. At the same time, I reject deliberate roughness of technique, feeling that that approach can produce only a hollow mockery of the vital workmanship of the medieval craftsman, whose rough work reflected his rough working conditions and demands for production. Pattern welding offers a middle way out of the dichotomy between rigorous development of detail and warmth and vitality of the object.

Forge Welding

The bonds in patter welded steel are solid state welds made in the forge; that is, welds made below the temperature of fusion. The plates are heated to increase deformability of the material, allowing them to conform to each other very closely under the pressure of the hammer. The heat also has the effect of increasing the rate and amplitude of molecular vibration (this is what heat is). The combination of these two effects bring the plates to within intermolecular distances. As soon as two molecules have been brought sufficiently close together, interatomic bonds (shared electrons) form, making the two molecules effectively one piece of material. The percentage of molecular bonds across the weld interface compared to the number of bonds in the parent metal determines the strength of the weld relative to the strength of the material. It is possible with some types of solid state welding, particularly explosive welding, to make a welded joint stronger than the parent metals due to work hardening of themetal at the interface.

A weld automatically forms when you achieve interatomic proximity between two pieces of metal. The chief barrier is the presence of contaminants between the two plates. In addition to loose dirt, which may be physically removed, oxides and silicates are present, which form a thin, tenacious layer bonded to the surface of the metal. These metals are coated with a flux, often anhydrous borax, but possibly borax and sand, or borax and clay, mixtures.

The purpose of the flux is to reduce the melting point of the contaminants so that they become liquid well below the temperature of fusion of the metal. This allow the oxide flux mixture to be driven out of the joint by the force of the hammer blows, making possible the intimate contact required for the success of the joint. The flux also acts to protect the metal surface from loss of carbon. The clay/straw fluxes used by the Japanese smiths seem to serve this purpose better than the straight anhydrous borax, and for this reason are recommended during the drawing-out heats between welds.

I do not feel that any advantage is attained by the introduction of iron filings in the flux. The theory outlined above would lead one to believe that there is no critical temperature below which it is impossible to get a weld and above which a weld is certain. The success of the weld depends on the condition of the metal surface and the amount of pressure. I have succeeded in obtaining a high-quality weld by hammering two abrasively cleaned steel plates (one mild, one 1095) in an airtight wrap of "ticronic" foil, at a temperature of 2000°F. Ticronic foil is a commercial heat-treating foil. Check your Yellow Pages under "Heat Treating Supplies" for sources. This is about 250-300°F less than the temperature necessary for a conventional forge weld. It is common industrial practice to produce high-quality welds between thoroughly cleaned parts in a high-vacuum furnace. The procedure is known as diffusion welding. In this case, the metallic atoms actually diffuse across the weld interface.

A Procedure for Pattern Welding

A number of plates of mild and high carbon steel are cleaned by abrasive cleaning (this cleaning while not strictly necessary is desirable). The plates are beveled slightly on the ends, then tack welded together on one end with an oxyacetylene torch. Welding on one end allows any slag or impurities to be driven our the opposite end of the stack. A handle of ½" mild steel is welded to the plates. This involves welding to weld metal. This produces a weak bond which must often be redone later in the process. The plates are heated in the forge to about 1300°F (dull red) and then sprinkled with anhydrous borax. The plates are then replaced in the forge and heated slowly with a reducing flame (or with very light blast in a solid fuel forge). Normally, the outside plates heat more quickly than those on the inside. This seldom presents a problem with a gas forge. The temperature will not usually come up to the "burning" temperature of the metal. In solid fuel forges, this risk should be dealt with by heating as slowly as possible, bringing the forge to welding heat at the same time as the inner layers of metal.

Observing the Process

The outside plates glow a dull yellow (observed through oxyacetylene welding goggles to avoid eye damage) fairly quickly. The color progresses inwards towards the center plates. As the lamination reaches dull yellow, the flux begins to bubble violently and flows with about the same viscosity as light oil. Gradually you will see the billet brighten, but the color change is small. Normally, the furnace wall heats up much more quickly than the billet does, so a comparison of colors will give an indication of the heat in the billet. During the first heat, it is difficult to know exactly when welding temperature has been achieved because there is no firescale present. When the bright yellow billet increases in incandescence, it is ready to weld. If you wait too long, a very tenacious oxide will gradually form, which is difficult to deal with in subsequent steps.

Hammering proceeds from the handle to the center and then to the edges. You are forcing the liquid impurities outwards towards the edges of the billet. It is important not to trap them inside the billet. You will often produce a sound weld everywhere but on the bottom plate. This is caused by rapid head transfer to the anvil. It is a good idea during your initial forging of a billet to take a welding heat on  one side and then the next one on the opposite side. During the second heat, there is firescale on the billet, which helps in gauging temperature (although not in getting the weld). Disregard the flux bubbling; pick a bit of firescale to watch. When the firescale bubbles, you know that the impurities have gone liquid and it is safe to weld.

The soundness of the weld can be estimated by looking at the edge of the billet after heating. If the color is even through the billet, the welds are sound. Sharp lines of color contrast represent weak welds because the heat will not travel as easily across a tightly shut void as it will across a weld. When you have sound welds throughout the billet (three or four heats), it is time to draw the billet out. The drawing-out process consists of hammering the orange-hot billet until it is about half its original thickness. Be careful not to heat the metal above dull yellow heat, as this can cause surface cracking due to the relatively high carbon content in 1095. The metal may be either lengthened or widened in drawing. The Japanese practice is to alternate the two. In either case, during the last heat, flux and clean off the firescale from one flat face and forge with that face against the anvil. This face will form the inside in the next weld. This makes grinding the weld face clean much easier. In drawing out, avoid unnecessary hammering on the side faces of the billet.

After the billet is drawn out, one face is cleaned and it is cut part way through from the rough face, leaving about ¹/₁₆" of metal. The ¹/₁₆" metal left after the cut acts as a hinge in the folding. The next weld may then take place after fluxing, folding and heating the billet. The above process is repeated as necessary to obtain the desired number of layers. The theoretical rate of increase is as follows: 1^st weld, 4; 2^nd, 7; 3^rd, 13; 4^th, 25; 5^th, 49; 6^th, 97; 7^th, 193; etc. These figures overestimate the actual number of layers accumulated in practice, since there are losses with each weld from scaling and cleaning.

Creating the Pattern

With fine material, the billet may simply be forged, ground and polished without any attempt of artificial pattern generation. To my mind, this method creates the most beautiful patterns in the object. It normally produces a gently varying wood grain known as itame.

An artificial wood-grain pattern known as mokume may be created by dimpling the article with a ball peen hammer in final forging before grinding and polishing. This pattern can be made more random and lively by doing so on the weld face prior to cleaning during the preparations for each weld. Mesame or line grain is formed either by grinding the billet on a bias or by forging the billet on the bias, then grinding flat the plane thus formed. Ayasugi-hada, or curly grain, is produced by incising the billet in straight, perpendicular lines, then forging the billet flat. The Japanese disapprove of this grain for a working sword, feeling that it weakens the metal.

Finishing

After the desired grain is formed and the billet has been forged, filed and ground to size, the piece must be treated to render the pattern visible. In the Western tradition, this is normally done by polishing, cleaning and etching. I have used 10% sulphuric, concentrated muriatic and 5-10% ferric chloride solutions. These may be used alone or in sequence. I usually finish with ferric chloride. It creates differential coloring where the others merely produce a difference in level. It is tempting to stop polishing prematurely, telling yourself that the scratches will etch out. They won't, so produce as clean a surface as you can before etching unless you want the look of a scratchy/etched surface.

In etching the work, one is taking advantage of the faster and darker etch experienced in cutting martensite crystals (the structure of hardened, tempered high carbon steel) compared with the effect of the mordant on the ferritic/pearlitic structure found in mild steel. In the Japanese techniques, it is simply differences in hardness that are being shown up. In both cases, it is necessary to harden and temper the article before polishing and surface treatment in order to get a clearly visible grain.

The Japanese look down on "chemical polishing" and reveal the grain in their swords entirely by mechanical polishing. They might begin with a soft, large-grained (approximately 1000) abrasive on a hard back, followed by a final polish with a fine, hard abrasive on a soft back. For this final step, the wet powder grit may be used on the pads of the thumbs. The process produces a glossy finish that shows the grain quite clearly. The finish is beautiful but tends to be a bit subtle for Western-style work, which is seldom the object of deliberate contemplation.

Adam Smith has been a working knifemaker since 1975. He is currently president of the Ontario Crafts Council.

By Adam Smith

Metalsmith Magazine – 1987 Fall

In association with SNAG‘s

Metalsmith magazine, founded in 1980, is an award winning publication and the only magazine in America devoted to the metal arts.