Tempering (metallurgy)

Tempering is most often performed on steel that has been heated above its upper critical (A₃) temperature and then quickly cooled, in a process called quenching, using methods such as immersing the hot steel in water, oil, or forced-air. The quenched steel, being placed in or very near its hardest possible state, is then tempered to incrementally decrease the hardness to a point more suitable for the desired application. The hardness of the quenched steel depends on both cooling speed and on the composition of the alloy. Steel with a high carbon content will reach a much harder state than steel with a low carbon content. Likewise, tempering high-carbon steel to a certain temperature will produce steel that is considerably harder than low-carbon steel that is tempered at the same temperature. The amount of time held at the tempering temperature also has an effect. Tempering at a slightly elevated temperature for a shorter time may produce the same effect as tempering at a lower temperature for a longer time. Tempering times vary, depending on the carbon content, size, and desired application of the steel, but typically range from a few minutes to a few hours.

Tempering quenched steel at very low temperatures, between 66 and 148 °C (151 and 298 °F), will usually not have much effect other than a slight relief of some of the internal stresses and a decrease in brittleness. Tempering at higher temperatures, from 148 to 205 °C (298 to 401 °F), will produce a slight reduction in hardness, but will primarily relieve much of the internal stresses. In some steels with low alloy content, tempering in the range of 260 and 340 °C (500 and 644 °F) causes a decrease in ductility and an increase in brittleness, and is referred to as the "tempered martensite embrittlement" (TME) range. Except in the case of blacksmithing, this range is usually avoided. Steel requiring more strength than toughness, such as tools, are usually not tempered above 205 °C (401 °F). Instead, a variation in hardness is usually produced by varying only the tempering time. When increased toughness is desired at the expense of strength, higher tempering temperatures, from 370 to 540 °C (698 to 1,004 °F), are used. Tempering at even higher temperatures, between 540 and 600 °C (1,004 and 1,112 °F), will produce excellent toughness, but at a serious reduction in the strength and hardness. At 600 °C (1,112 °F), the steel may experience another stage of embrittlement, called "temper embrittlement" (TE), which occurs if the steel is held within the temperature range of temper embrittlement for too long. When heating above this temperature, the steel will usually not be held for any amount of time, and quickly cooled to avoid temper embrittlement.[3]

Steel that has been heated above its upper critical temperature and then cooled in standing air is called normalized steel. Normalized steel consists of pearlite, martensite and sometimes bainite grains, mixed together within the microstructure. This produces steel that is much stronger than full-annealed steel, and much tougher than tempered quenched-steel. However, added toughness is sometimes needed at a reduction in strength. Tempering provides a way to carefully decrease the hardness of the steel, thereby increasing the toughness to a more desirable point. Cast-steel is often normalized rather than annealed, to decrease the amount of distortion that can occur. Tempering can further decrease the hardness, increasing the ductility to a point more like annealed steel.[8] Tempering is often used on carbon steels, producing much the same results. The process, called "normalize and temper", is used frequently on steels such as 1045 carbon steel, or most other steels containing 0.35 to 0.55% carbon. These steels are usually tempered after normalizing, to increase the toughness and relieve internal stresses. This can make the metal more suitable for its intended use and easier to machine.[9]

Steel that has been arc welded, gas welded, or welded in any other manner besides forge welded, is affected in a localized area by the heat from the welding process. This localized area, called the heat-affected zone (HAZ), consists of steel that varies considerably in hardness, from normalized steel to steel nearly as hard as quenched steel near the edge of this heat-affected zone. Thermal contraction from the uneven heating, solidification and cooling creates internal stresses in the metal, both within and surrounding the weld. Tempering is sometimes used in place of stress relieving (even heating and cooling of the entire object to just below the A₁ temperature) to both reduce the internal stresses and to decrease the brittleness around the weld. Localized tempering is often used on welds when the construction is too large, intricate, or otherwise too inconvenient to heat the entire object evenly. Tempering temperatures for this purpose are generally around 205 °C (401 °F) and 343 °C (649 °F).[10]

Modern reinforcing bar of 500 MPa strength can be made from expensive microalloyed steel or by a quench and self-temper (QST) process. After the bar exits the final rolling pass, where the final shape of the bar is applied, the bar is then sprayed with water which quenches the outer surface of the bar. The bar speed and the amount of water are carefully controlled in order to leave the core of the bar unquenched. The hot core then tempers the already quenched outer part, leaving a bar with high strength but with a certain degree of ductility too.

Tempering was originally a process used and developed by blacksmiths (forgers of iron). The process was most likely developed by the Hittites of Anatolia (modern-day Turkey), in the twelfth or eleventh century BC. Without knowledge of metallurgy, tempering was originally devised through a trial-and-error method.

Because few methods of precisely measuring temperature existed until modern times, temperature was usually judged by watching the tempering colors of the metal. Tempering often consisted of heating above a charcoal or coal forge, or by fire, so holding the work at exactly the right temperature for the correct amount of time was usually not possible. Tempering was usually performed by slowly, evenly overheating the metal, as judged by the color, and then immediately cooling, either in open air or by immersing in water. This produced much the same effect as heating at the proper temperature for the right amount of time, and avoided embrittlement by tempering within a short time period. However, although tempering-color guides exist, this method of tempering usually requires a good amount of practice to perfect, because the final outcome depends on many factors, including the composition of the steel, the speed at which it was heated, the type of heat source (oxidizing or carburizing), the cooling rate, oil films or impurities on the surface, and many other circumstances which vary from smith to smith or even from job to job. The thickness of the steel also plays a role. With thicker items, it becomes easier to heat only the surface to the right temperature, before the heat can penetrate through. However, very thick items may not be able to harden all the way through during quenching.[11]

Pieces of through-tempered steel flatbar. The first one, on the left, is normalized steel. The second is quenched, untempered martensite. The remaining pieces have been tempered in an oven to their corresponding temperature, for an hour each. "Tempering standards" like these are sometimes used by blacksmiths for comparison, ensuring that the work is tempered to the proper color.

If steel has been freshly ground, sanded, or polished, it will form an oxide layer on its surface when heated. As the temperature of the steel is increased, the thickness of the iron oxide will also increase. Although iron oxide is not normally transparent, such thin layers do allow light to pass through, reflecting off both the upper and lower surfaces of the layer. This causes a phenomenon called thin-film interference, which produces colors on the surface. As the thickness of this layer increases with temperature, it causes the colors to change from a very light yellow, to brown, then purple, then blue. These colors appear at very precise temperatures, and provide the blacksmith with a very accurate gauge for measuring the temperature. The various colors, their corresponding temperatures, and some of their uses are:

• Faint-yellow – 176 °C (349 °F) – gravers, razors, scrapers
• Light-straw – 205 °C (401 °F) – rock drills, reamers, metal-cutting saws
• Dark-straw – 226 °C (439 °F) – scribers, planer blades
• Brown – 260 °C (500 °F) – taps, dies, drill bits, hammers, cold chisels
• Purple – 282 °C (540 °F) – surgical tools, punches, stone carving tools
• Dark blue – 310 °C (590 °F) – screwdrivers, wrenches
• Light blue – 337 °C (639 °F) – springs, wood-cutting saws
• Grey-blue – 371 °C (700 °F) and higher – structural steel

Beyond the grey-blue color, the iron oxide loses its transparency, and the temperature can no longer be judged in this way. The layer will also increase in thickness as time passes, which is another reason overheating and immediate cooling is used. Steel in a tempering oven, held at 205 °C (401 °F) for a long time, will begin to turn brown, purple or blue, even though the temperature did not exceed that needed to produce a light-straw color. Oxidizing or carburizing heat sources may also affect the final result. The iron oxide layer, unlike rust, also protects the steel from corrosion through passivation.[12]

A differentially tempered sword. The center is tempered to a springy hardness while the edges are tempered slightly harder than a hammer.

Differential tempering is a method of providing different amounts of temper to different parts of the steel. The method is often used in bladesmithing, for making knives and swords, to provide a very hard edge while softening the spine or center of the blade. This increased the toughness while maintaining a very hard, sharp, impact-resistant edge, helping to prevent breakage. This technique was more often found in Europe, as opposed to the differential hardening techniques more common in Asia, such as in Japanese swordsmithing.

Differential tempering consists of applying heat to only a portion of the blade, usually the spine, or the center of double-edged blades. For single-edged blades, the heat, often in the form of a flame or a red-hot bar, is applied to the spine of the blade only. The blade is then carefully watched as the tempering colors form and slowly creep toward the edge. The heat is then removed before the light-straw color reaches the edge. The colors will continue to move toward the edge for a short time after the heat is removed, so the smith typically removes the heat a little early, so that the pale-yellow just reaches the edge, and travels no farther. A similar method is used for double-edged blades, but the heat source is applied to the center of the blade, allowing the colors to creep out toward each edge.[13]

Interrupted quenching methods are often referred to as tempering, although the processes are very different from traditional tempering. These methods consist of quenching to a specific temperature that is above the martensite start (M_s) temperature, and then holding at that temperature for extended amounts of time. Depending on the temperature and the amount of time, this allows either pure bainite to form, or holds-off forming the martensite until much of the internal stresses relax. These methods are known as austempering and martempering.[14]

Time-temperature transformation (TTT) diagram. The red line shows the cooling curve for austempering.

Austempering is a technique used to form pure bainite, a transitional microstructure found between pearlite and martensite. In normalizing, both upper and lower bainite are usually found mixed with pearlite. To avoid the formation of pearlite or martensite, the steel is quenched in a bath of molten metals or salts. This quickly cools the steel past the point where pearlite can form, and into the bainite-forming range. The steel is then held at the bainite-forming temperature, beyond the point where the temperature reaches an equilibrium, until the bainite fully forms. The steel is then removed from the bath and allowed to air-cool, without the formation of either pearlite or martensite.

Depending on the holding-temperature, austempering can produce either upper or lower bainite. Upper bainite is a laminate structure formed at temperatures typically above 350 °C (662 °F) and is a much tougher microstructure. Lower bainite is a needle-like structure, produced at temperatures below 350 °C, and is stronger but much more brittle.[15] In either case, austempering produces greater strength and toughness for a given hardness, which is determined mostly by composition rather than cooling speed, and reduced internal stresses which could lead to breakage. This produces steel with superior impact resistance. Modern punches and chisels are often austempered. Because austempering does not produce martensite, the steel does not require further tempering.[14]

Martempering is similar to austempering, in that the steel is quenched in a bath of molten metal or salts to quickly cool it past the pearlite-forming range. However, in martempering, the goal is to create martensite rather than bainite. The steel is quenched to a much lower temperature than is used for austempering; to just above the martensite start temperature. The metal is then held at this temperature until the temperature of the steel reaches an equilibrium. The steel is then removed from the bath before any bainite can form, and then is allowed to air-cool, turning it into martensite. The interruption in cooling allows much of the internal stresses to relax before the martensite forms, decreasing the brittleness of the steel. However, the martempered steel will usually need to undergo further tempering to adjust the hardness and toughness, except in rare cases where maximum hardness is needed but the accompanying brittleness is not. Modern files are often martempered.[14]

Tempering involves a three-step process in which unstable martensite decomposes into ferrite and unstable carbides, and finally into stable cementite, forming various stages of a microstructure called tempered martensite. The martensite typically consists of laths (strips) or plates, sometimes appearing acicular (needle-like) or lenticular (lens-shaped). Depending on the carbon content, it also contains a certain amount of "retained austenite." Retained austenite are crystals which are unable to transform into martensite, even after quenching below the martensite finish (M_f) temperature. An increase in alloying agents or carbon content causes an increase in retained austenite. Austenite has much higher stacking-fault energy than martensite or pearlite, lowering the wear resistance and increasing the chances of galling, although some or most of the retained austenite can be transformed into martensite by cold and cryogenic treatments prior to tempering.

The martensite forms during a diffusionless transformation, in which the transformation occurs due to shear-stresses created in the crystal lattices rather than by chemical changes that occur during precipitation. The shear-stresses create many defects, or "dislocations," between the crystals, providing less-stressful areas for the carbon atoms to relocate. Upon heating, the carbon atoms first migrate to these defects, and then begin forming unstable carbides. This reduces the amount of total martensite by changing some of it to ferrite. Further heating reduces the martensite even more, transforming the unstable carbides into stable cementite.

The first stage of tempering occurs between room-temperature and 200 °C (392 °F). In the first stage, carbon precipitates into ε-carbon (Fe_2,4C). In the second stage, occurring between 150 °C (302 °F) and 300 °C (572 °F), the retained austenite transforms into a form of lower-bainite containing ε-carbon rather than cementite (archaically referred to as "troostite").[16][17] The third stage occurs at 200 °C (392 °F) and higher. In the third stage, ε-carbon precipitates into cementite, and the carbon content in the martensite decreases. If tempered at higher temperatures, between 650 °C (1,202 °F) and 700 °C (1,292 °F), or for longer amounts of time, the martensite may become fully ferritic and the cementite may become coarser or spheroidize. In spheroidized steel, the cementite network breaks apart and recedes into rods or spherical shaped globules, and the steel becomes softer than annealed steel; nearly as soft as pure iron, making it very easy to form or machine.[18]

Embrittlement occurs during tempering when, through a specific temperature range, the steel experiences an increase in hardness and a reduction in ductility, as opposed to the normal decrease in hardness that occurs to either side of this range. The first type is called tempered martensite embrittlement (TME) or one-step embrittlement. The second is referred to as temper embrittlement (TE) or two-step embrittlement.

One-step embrittlement usually occurs in carbon steel at temperatures between 230 °C (446 °F) and 290 °C (554 °F), and was historically referred to as "500 degree [Fahrenheit] embrittlement." This embrittlement occurs due to the precipitation of Widmanstatten needles or plates, made of cementite, in the interlath boundaries of the martensite. Impurities such as phosphorus, or alloying agents like manganese, may increase the embrittlement, or alter the temperature at which it occurs. This type of embrittlement is permanent, and can only be relieved by heating above the upper critical temperature and then quenching again. However, these microstructures usually require an hour or more to form, so are usually not a problem in the blacksmith-method of tempering.

Two-step embrittlement typically occurs by aging the metal within a critical temperature range, or by slowly cooling it through that range, For carbon steel, this is typically between 370 °C (698 °F) and 560 °C (1,040 °F), although impurities like phosphorus and sulfur increase the effect dramatically. This generally occurs because the impurities are able to migrate to the grain boundaries, creating weak spots in the structure. The embrittlement can often be avoided by quickly cooling the metal after tempering. Two-step embrittlement, however, is reversible. The embrittlement can be eliminated by heating the steel above 600 °C (1,112 °F) and then quickly cooling.[19]

Malleable (porous) cast iron is manufactured by white tempering. White tempering is used to burn off excess carbon, by heating it for extended amounts of time in an oxidizing environment. The cast iron will usually be held at temperatures as high as 1,000 °C (1,830 °F) for as long as 60 hours. The heating is followed by a slow cooling rate of around 10 °C (18 °F) per hour. The entire process may last 160 hours or more. This causes the cementite to decompose from the ledeburite, and then the carbon burns out through the surface of the metal, increasing the malleability of the cast iron.[23]

Ductile (non-porous) cast iron (often called "black iron") is produced by black tempering. Unlike white tempering, black tempering is done in an inert gas environment, so that the decomposing carbon does not burn off. Instead, the decomposing carbon turns into a type of graphite called "temper graphite" or "flaky graphite," increasing the malleability of the metal. Tempering is usually performed at temperatures as high as 950 °C (1,740 °F) for up to 20 hours. The tempering is followed by slow-cooling through the lower critical temperature, over a period that may last from 50 to over 100 hours.[23]