BP0078 The Metallurgy of Heat Treating for Blacksmiths

It is not the purpose of this article to present the most highly technical aspects of metallurgy. However, some of the concepts will be new to you and may look rather intimidating. Don't worry about what it looks like, read it several times if necessary. Other metallurgists will note that some topics have been greatly simplified in an effort to illustrate certain basic principles. The purpose is to present some of the basics of ferrous metallurgy that will give the Blacksmith some insight into what is happening on a microstructural level during normal smithing processes. It is hoped that this will aid him or her in utilizing enough theory to avoid some trial-and-error disasters.

It is not the intent of the author to impress you with big words, complex theory, or difficult concepts. However, metallurgy is a difficult and inexact science. The terms used are universally accepted and if you learn them, you will understand what others are trying to explain when discussing heat treating and metallurgy.

Before we discuss the theory, we need to define some terms that are used to describe the characteristics of the metals we work with.

Iron: Atomic No. 26 on the periodic chart. Comprises much of the earth�s core and is present almost everywhere on the earth as an oxide (iron combined with oxygen, as in rust).

Alloy: a mixture of 2 or more elements, one of which is a metal.

Steel: An alloy (mixture) of iron and carbon. The carbon level in steel can be from .02% to about 1.8%. The carbon is usually dissolved into the iron and is invisible even under high magnification. Higher carbon levels are associated with cast iron. Other elements may be added for specific purposes. It is the carbon content that controls the ultimate hardness that can be achieve when the part is quenched.

Cast Iron: Iron with over 1.8% carbon. The carbon exists as flakes or spheres and gives the iron a sooty gray appearance (hence, gray iron). Not suitable for forging.

Mild Steel: Low carbon steel, .10% to about .20% carbon.

Hot Rolled Steel: Steel that has been rolled to it�s final shape while hot, usually in the range of 1650F to 1950F.

Cold Rolled Steel: Hot rolled steel that has been put through an extra process while cold, ie, at room temperature. This is done to improve the shape and dimensional tolerances. Cold rolling will increase the strength of hot rolled steel but this extra strength disappears when it is heated to forging temperature and cooled.

Tensile strength: The load, applied in tension in pounds, divided by the cross sectional area of a sample, that causes it to break. This is expressed as pounds per square inch (PSI) in the US.

Yield Strength: The load, applied in tension in pounds, divided by the cross sectional area of a sample, that causes it to exceed the elastic limit by a specified amount of stretch. More on this later.

Hardness: The resistance of a material to deformation. This is a relative property and must always be determined by comparing the hardness of a sample with the hardness of a known standard.

Toughness: The ability of a material to resist crack propagation.

Ferrite: Essentially pure iron and the name of the crystal structure of iron at room temperature.

Pearlite: Ferrite with fingers of iron carbide arranged in parallel rows.

Austenite: Iron at a temperature over about 1340F. Austenite usually dissolves all of the carbon in the steel as well as most alloys. It is the name of the crystal shape of iron in this temperature range.

Martensite: Iron that has been first transformed to austenite, then quenched very rapidly to room temperature. It also has all the carbon dissolved in it and the name of the crystal shape of rapidly cooled iron.

Cementite: Iron Carbide, a chemical mixture of iron and carbon.

Curie point: The temperature at which iron becomes non-magnetic, about 1400F.

What is a Metal?

Metals account for more than half of the known elements on Earth. They are usually solid at room temperature (Mercury is not), they have relatively high melting points, they conduct electricity and heat, they are malleable (formable), and mixing them with other elements may alter their properties.

Metallic Atoms

Metallic atoms bond together in a very unique way. When metallic atoms bond, they form crystals. Each atom shares their electrons but the electrons are free to move to any other atom in the crystal or any other crystal. The flow of electrons is what we call electricity and that is why metals conduct electricity so well.

Iron Crystals

If the austenite is very rapidly cooled (quenched) it changes back to a cube with one atom at each corner and one in the center. However, it is slightly distorted, with one dimension of the cube longer than the other two. This crystal is called �martensite� (named after another British metallurgist). It is this ability to change crystal shapes that allows iron to be heat treated to so many different properties.

When the iron changes crystal shapes, the spacing between the iron atoms in the crystal changes. For example, when ferrite (room temperature iron) changes to austenite (iron at an orange heat), there is more room between the iron atoms. Atoms like carbon, which are much smaller than the iron atom, can now fill in the spaces between the iron atoms. Austenite is able to dissolve much more carbon than ferrite. This is an extremely important point, as we will see in the section on heat-treating.

When more carbon, or other alloy, is added to iron, the distortion of the crystal is greater. This is especially true when the steel has been quenched and tempered. The distortion of the crystal makes it more difficult for rows of atoms to move across one another, making the steel stronger.

Metal Grains

When the metal is totally solidified, the metal has formed �grains� consisting of rows of atoms all oriented in one direction within the grain. Each grain has rows that are oriented differently from the grain beside it. Think of the piece of metal as a jar filled with sand with each grain of sand touching its neighbors at different angles (Fig.4). As soon as a metal solidifies, it is fully crystalline. It does not �crystallize� at some distant time, years after solidification.

A newly solidified ingot has a crystalline structure very different from a piece of steel that has been hot or cold rolled. The ingot has long finger-like grains growing in from the edges of the ingot. The center is often highly segregated, containing a lot of impurities. When steel is hot rolled, the internal structure is broken up and the grains become more uniform in size. Hot working is extremely important to obtaining a homogeneous (uniform crystal structure) piece of steel.

Deforming of Grains

When grains of steel are hot worked, that is, worked at a temperature well above where the steel is fully non-magnetic, and forging is deforming the grains, the grain boundaries immediately reform, creating new normally shaped grains (Fig. 6). When steel is worked cold, the grain boundaries do not reform, leaving massive distortion of the grains. The atoms in the deformed grains cannot easily move because the slip planes are totally disrupted. This is called work hardening and it is why cold working causes the steel to increase in hardness and yield strength.

When cold-worked steel is heated up to a good red heat, the grains re-form back into regular shapes. The atoms are again in nice neat rows and slip planes can facilitate movement with very little resistance. That is why normalizing and stress relieving are usually applied to heavily cold-worked steel.

When steel is heated, the iron atoms vibrate. The higher the temperature, the greater the oscillation above and below the normal position for the room temperature atom. When the atoms are vibrating significantly out of position, it is easier for them to jump over other atoms that might block their path when being deformed. That is why steel is easier to forge at higher temperatures. Alloy elements, like chromium, block the path of the atoms being moved and alloy steels require even higher temperatures to get the iron atoms around the alloy atoms.

The Iron-Carbon Phase Diagram

The phase diagram (Fig. 7) is nothing but a map of what kind of iron crystals (ferrite, austenite, pearlite, etc) exist at different temperatures and different carbon contents. It is simplified in that it only shows the phases in plain carbon steel, ignoring the effects of all of the other alloy elements. The diagram shown here is only part of the whole diagram but it is the part that we are concerned with and it will get us close enough to reality. Every alloy has its own specific phase diagram. However, only two or three elements can be graphed with a simple two dimensional diagram. The effects of more than three elements are simply too complex and cannot be graphically represented.

Heating a .5% carbon steel just to non-magnetic will make it almost fully austenitic. Following the dotted line at Curie toward the right it intersects another diagonal line. It intersects the line at about .95% C. We can assume that any steel with a carbon content between about .5% and .95% carbon will be austenitic if heated to non-magnetic. Steel with less than .5% or more than .95% carbon will require a higher temperature to achieve a fully austenitic structure.

So why do we care if the steel is fully austenitic before we quench it? Martensite forms from austenite. Any ferrite or cementite left in the steel when it is quenched will still be ferrite or cementite after it is quenched. That means the steel did not fully harden. Furthermore, it takes a while for the carbon that is combined with iron and other alloys to dissolve back into the austenite. Any carbon that is tied up as a carbide will not contribute to the hardness. It is like quenching a lower-carbon steel.

Normalizing

Many good heat-treaters prefer to normalize a piece before they quench and temper it. Normalizing consists of heating to 100 degrees hotter than is required for hardening. Then, the part is air-cooled. Heating it hotter dissolves more of the carbides. Air cooling causes the carbides to re-form quickly, leaving them very small and widely dispersed. This puts them into an ideal condition for heat-treatment. Being small, they dissolve quickly and being widely dispersed, they give the steel a uniform distribution of carbon.

Normalizing has the additional benefit of stress relieving the forging prior to hardening. To properly normalize a forging, first allow it to cool to room temperature and then re-heat it. Since forging at high temperatures can cause grain growth, normalizing to a temperature lower than forging temperature forms smaller austenite grains and this improves toughness. Normalizing several times, at a lower temperature each time, will make the grains smaller each time. Normalizing at the same or higher temperature can actually make the grains grow larger.

Hardenability

Hardness is a physical property that is defined as the resistance to deformation. Hardenability is a measure of how DEEP into the steel a given hardness can be achieved. This concept is easily explained with a virtual lab experiment. Assume we have two pieces of steel. One is a plain carbon steel, say .6% C, and the other is an alloy steel with the same carbon content but also has about .7% Chromium. The pieces are conveniently 1� in diameter and about 6� long.

Both pieces are heated to the appropriate temperature and allowed to �soak� at temperature to assure that all of the carbides are dissolved. The plain carbon steel is removed from the furnace and a stream of water is directed onto the bottom of the piece as it is suspended vertically from a fixture. All of the heat is being removed from one surface only, the bottom end. When it is cooled to room temperature, the other piece is heat treated in the same manner.

Both bars are ground longitudinally to remove about 1/8� so there is a flat running full length of the bar. Using our virtual Rockwell C tester, we take hardness readings every 1/16� on the flat starting on the end and moving up the bar. When we plot the hardness against the distance from the quenched end, we get a graph that looks like Fig. 8.

Curve A is the chromium alloy steel and curve B is the carbon steel. As you can see, both achieve the same hardness at the surface. However, as the hardness is plotted further and further away from the surface, Steel, A, is harder than the other. Note that, for example, at a distance of 2/16� from the surface, A is a 52Rc, while B is 26. Steel A is a deeper-hardening steel; it has more hardenability than Steel B. This test is one that is routinely performed to determine hardenability and is called the Jominy Test. There is so much data available that the Jominy hardness curves can be calculated from the chemistry of the steel.

The Jominy test illustrates the concept of hardenability. The practical application for the smith is that the type of steel chosen for an application may achieve the hardness on the surface but leave a very soft core inside. Sometimes, this is satisfactory for the application, such as a hammerhead. For a power hammer die, it may result in early failure. The next question we will address is �What controls hardenability?�

Isothermal Transformation Diagrams

Most smiths learn early that steel quenched in water gets harder than steel quenched in oil. This assumes that the part does not crack in the water quench. The difference in results is due to the difference in cooling rate. To demonstrate the effect of cooling rate we will conduct another virtual experiment.

Using the same piece of .60% carbon steel from the first experiment, and all the real and imaginary lab equipment necessary, we heat the steel to a fully austenitic state (about 1500F). Under a microscope built especially to view hot samples, we see that the microstructure is, indeed, entirely austenite. Using molten salt quench bath heated to a temperature of 1200F, we instantly cool the piece to 1200F.

Again under the microscope, we notice that nothing happens for 2 seconds. Then, it begins to transform back into ferrite. A second later some pearlite (ferrite with carbide fingers) begins to show up. A second later, all of the austenite is gone. The structure is entirely made up of ferrite and pearlite. Once again we make a graph (Fig. 9). The first dot is where the transformation started, and the second dot is where the pearlite began to form and the third is where it stopped.

The next sample is heated as before, quenched to 1000F, and held there while we look at it under the microscope. This time, the ferrite begins to show up at 1 second, the pearlite begins to form 1 second later and is finished in about 3.5 seconds. We continue to do this for each of the temperatures down to 500F, each of them being plotted as before. At 500F, we find a new structure beginning to form: martensite. However, no matter how long we hold it at 500F, only a small portion changes to martensite. The rest becomes another structure, bainite, which is of limited interest here. We do the experiment again for 400F and so on down to 100F. We discover is that as we cool the steel below 500F, more and more martensite forms. At 100F, about 90% of the structure is martensite.

To summarize the chart, we can see that for this steel, austenite will transform back to ferrite and pearlite at various times and temperatures down to about 500F. At 500F, it begins to transform to martensite. To form more martensite, the steel must continue to cool. Every different steel alloy has a different Isothermal Transformation diagram. High alloy steels have curves that are shifted far to the right. Low carbon, non-alloy steels have curves shifted to the left. What this curve suggests is that if we can cool a piece of steel from 1350F down to 500F in less than one second (line A on graph) it will not have a chance to start forming ferrite and carbide (pearlite). It will immediately begin to form martensite. Once it begins forming martensite, it will continue to transform to martensite as the piece is cooled and no ferrite or carbide will form.

When alloys are added to steel, the effect is to shift the curves further to the right, allowing more time to reach the martensite transformation temperature without forming ferrite and carbide. If we are quenching a thin section, we can cool the piece very rapidly and get martensite all the way through. However, if the section is thick, the surface will cool quickly but the interior of the piece cools more slowly. By adding alloys to steel, we allow the slow cooling interior more time to reach the martensite transformation temperature without forming ferrite and carbide. When a steel is called �deep hardening� it means it has sufficient alloy to allow very thick sections to transform to martensite.

Now, consider what happens when we cool the steel more slowly. Look at the graph and notice that Line B cuts through the curves. It crosses the first line, the ferrite start line at about 1100F. This means that for this cooling rate, ferrite will begin to form at 1100F. Line B crosses the second line, the pearlite start line, at about 1000F. Pearlite will begin to form at about 1000F. The line intersects the martensite line before it intersects the transformation end line. This means that there is still some austenite remaining that has not transformed to anything. When the untransformed austenite reaches the martensite start temperature, it will begin to transform to martensite.

For the steel cooled by Line B, the microstructure will be mostly ferrite and pearlite, with some martensite mixed in. For the steel cooled by Line C, the structure will be entirely ferrite and carbide (pearlite). All of the austenite was transformed to ferrite and pearlite before it reached the martensite start temperature. Now, think back to the first experiment on Jominy hardenability. The reason that the hardnesses got lower and lower as we tested further and further into the bar was due to the fact that the cooling rate slowed down inside the bar. Less martensite was forming while more ferrite and carbide were forming and martensite is harder than ferrite and carbide.

There is a very important lesson in all this. First, steel will usually not harden all the way through a heavy section unless it is an alloy designed to do so. Second, and most important, martensite does not begin to form until the steel reaches a temperature of 500F to 700F and it must continue to cool, at any cooling rate, to get more martensite to form. If you bring the part out of the quench at a temperature above the martensite start temperature, you probably did not get a martensitic structure and the part will be soft because it will contain some pearlite and ferrite.

Let's conduct another quick experiment. We take another �� diameter piece of our test steel, austenitize it, quench it down to 350F, and pull it out of the quench, and allow it to cool in air. Two things are happening. The steel was quenched below the martensite start temperature so it should continue to form martensite as it air cools, and it does. However, as the outer section cools and transforms, the heat from inside the piece is tempering the martensite, reducing its hardness and improving its toughness. As it cools, there is less and less tempering taking place so the core of the steel is not tempered. Untempered martensite is hard and brittle and the core may be a place where cracks could form. To address this, the part must be re-tempered at a temperature ABOVE the temperature at which the part was removed from the quench. But what was that temperature? It was always changing as it cooled. You can see that for critical parts, auto-tempering is not a good method.

So what really happens when we temper? Remember that when we heated the part to form austenite, the carbon dissolved into the spaces between the iron atoms in the face-centered cubic structure. When the part is quenched, the face center cubic structure distorts and forms martensite. However, the carbon is still trapped in between the iron atoms, causing a tremendous strain on the bonds between the iron atoms. When the iron is re-heated for tempering, the atoms begin to vibrate. As they move up and down, it allows room for the carbon atoms to escape, recombine with other iron atoms and form carbides. The hotter the piece is tempered, the more carbon will escape, and the softer the part will get. You can do the same thing using lower temperatures for a longer time.

If you temper martensite hot enough or long enough, virtually all of the carbon escapes, forms carbides, and the martensite will revert back to being basically just ferrite with a lot of carbides in it. Because the carbides are smaller and more evenly dispersed through the steel, if we re-harden it, the hardness will be greater than the first time. The carbides are smaller and will dissolve faster. More carbon in solution will give a higher hardness after quenching. Heating and quenching several times will improve the hardness and toughness of the steel but it also exposes the part to more stresses due to thermal shock. This may cause it to crack if the part has sharp corners or a rough surface.

Heat Treating Processes

The most important factor in good heat treating, other than getting the right temperature, is uniform heating and cooling. Failure to heat or cool uniformly will almost always lead to distortion or cracking.

Stress Relieving

Stress relieving is really designed to relieve stress created by cold working, welding, bending, etc. Hot forged pieces are usually not stress relieved unless they hardened while cooling from forging.

Stress relieving is normally done at temperatures in the 700F to 1000F range, independent of the alloy type. For stress relieving to be effective, time at temperature is important and normally, one hour per inch of thickness is the rule. Allow to slow cool from stress relieving.

Annealing

Annealing is very similar to tempering. The metal is heated and held at a temperature for a period of time and then allowed to cool very slowly. The temperature allows the carbon in the steel to form carbides and then allows the carbides to grow larger. By allowing the carbides to grow large, the steel becomes softer. A full anneal involves heating the steel until it is fully austenitic. As we learned in the discussion about phase diagrams, this temperature will be determined by the carbon content. However, if the carbon content of the steel is not known, heating to a bright red heat will have to do. Holding is difficult to do in a coal forge but any time spent at high heat will achieve more than just heating and slow cooling. By slow cooling, we mean cooling in ashes or vermiculite for 12-24 hours. This gives more time for the carbides to grow and results in a softer steel.

Sub-critical annealing, or heating to a lower temperature, will soften steel but not as much as a full anneal. Sub-critical annealing involves heating to a dull red and slow cooling. This may be sufficient for low carbon steels.

Normalizing

Normalizing is also similar to annealing. It requires the steel to be fully austenitic and this means heating to an orange to yellow heat, depending on the carbon content. Cooling is done in still air rather than slow cooling in ashes or vermiculite. Normalized steel will be harder than annealed steel. Some steels should not be normalized, such as the air-hardening grades of tool steel. They will get almost as hard from normalizing as they would if you intended to fully harden them.

Tempering

In many old books on blacksmithing, the word "tempering" actually referred to the hardening process. "Drawing" was the reheating of the piece to achieve the final hardness. Today, the process of heating and quenching is called "hardening" and the reheating afterwards is "Tempering".

The alloy content of the steel affects how quickly the carbon can form these carbides. Molybdenum is a very potent alloy used to create a deep hardening steel and it slows down the movement of the carbon during tempering. This has the effect of maintaining high hardness at high heat. That is why hot work die steels like H12 or H13 have a lot of molybdenum in them. Molybdenum is also used in high speed steels (the "M" series) so that they do not become soft when they heat up due to friction.

The surface of the steel will change colors during tempering if you remove the scale and brighten the surface before tempering. The color indicates the temperature of the steel, NOT the hardness. For this method to be effective, you must have some idea of the alloy content of the steel so you can choose an appropriate temperature for tempering.

A simple way to determine the appropriate temperature (and color) is to harden a sample piece of the steel you are working on, preferably something about 6" long. After hardening the entire piece, grind the scale off, sand the surface smooth, and heat it from one end only. Stop when the opposite end is just a light straw. Quench immediately to preserve the colors. Take a file and strike a line through each color. When you hit steel that is harder than the file, it will leave no mark. This will give you an idea what color is appropriate for the steel. This assumes that you are trying to achieve a high degree of hardness. Tempering to lower hardnesses will improve toughness.

Tempering of tools usually involves heating to the range of 400-600F. This can actually be done in your kitchen oven with a high degree of accuracy and success. Get the oven up to temperature before you put the steel in it and allow at least one hour per inch of thickness for a holding time. Minimum tempering time should be one half hour regardless of thickness. Double and triple tempering often improves hardness and toughness but every time you temper, you should increase the temperature by about 25-50 degrees.

Quench and Tempering

This is usually the most confusing part of Blacksmith heat-treating and there is a lot of semi-scientific information on how to do it. The part about aligning your part with true north before you quench is mostly just showmanship. However let's discuss the steps and point out what is really important.

Surface Finish. Time spent filing, grinding, and sanding the rough surface is good insurance. Cracks can start in a deep scratch, a hammer scale pit, or even a sharp corner. When possible, provide generous radii, make section changes gradual, round the corners, and smooth the surface. This becomes more important as carbon content increases. Although air-hardening steels are usually less susceptible to cracking due to the slow cooling, they can also crack if the part is not well finished.

Holding at temperature is difficult but will assure that the heat is uniform and most or all of the carbon has gone into solution. The phase diagram suggests that only at a carbon content from about .50 to .80% does the iron become fully austenitic at the same time it is non-magnetic. For higher and lower carbon contents, you must heat a little hotter. Judging how much hotter can be difficult when you are doing it by eye for the first time. An error to the low side will result in less cracking but also lower hardness. Error to the high side will get the hardness but may cause distortion or cracking. This is the art of blacksmith heat-treating and it is an acquired skill.

If you consistently use steel with about the same carbon range, determine the appropriate tempering temperature and buy a Tempil Stick from a local welding supply that can be used at that specific temperature. To use the Tempil Stick, swipe it across the hot steel. When the steel reaches the temperature indicated on the Tempil Stick, the stick will melt. If it smokes and burns off, you are hotter than the indicated temperature. If it leaves a powdery mark, you are colder than the indicated temperature.

Quenching

The objective of quenching is to cool the piece rapidly enough to get it to the martensite start temperature before it has a chance to start forming other things like pearlite and ferrite. However, not so rapidly that it cracks due to thermal shock. Generally, any steel with a carbon content less than .30% can be quenched in water. Steel with a carbon content of .30-.50% might be water quenchable, depending on the other alloys in it. W1 tool steel may have up to 1.4% carbon but is still a water-quenching grade because is has almost nothing else in it. Over .50% and oil quenching is usually necessary.

Quenching in water may allow you to quench only part of the piece, keeping some of the piece above water. This is often done when quenching knife blades or chisel bits. However, if the steel is high carbon you run the risk of cracking the blade right at the water line due to the difference in expansion of the steel as it transforms. DO NOT partially immerse a hot piece in oil. The part above the oil can ignite the vapors and start a fire.

Immerse the piece into the quench and move it in a figure 8 pattern and leave it submerged until it has reached the quenchant temperature. Adding salt to the water can harden very low carbon steels a little deeper. Sometimes even ice is added to speed up the cooling rate. A cup of table salt in 5 gallons of water is usually sufficient. There are �special� quenchants that add a variety of salts and surfactants that help the water �wet in� to surface of the steel more quickly. One such is called Super Quench and formulas for it can be found on the Internet.

Failure to move the part in the quench allows a vapor barrier to be created between the surface of the hot part and the quenchant. This slows down the heat transfer and may cause spotty hardening.

When quenching high carbon steels, the carbon content sometimes causes the transformation to martensite to finish at temperatures well below zero. When quenching high carbon, high alloy steels, you may improve the hardness and toughness by refrigerating the part. After quenching, give the tool a light temper, about 300F, put in the ice box, or better yet, in dry ice, and allow it to sit over-night. Allow it to warm to room temperature, and temper again at 400-500F.

Tempering of knives and cutting-edged tools

If you are using an oil quench, clean the part before tempering to prevent a smoky burn-off. If you are tempering at the forge, you should brighten the edge of the tool to bare metal. Heat behind the edge and let the heat conduct slowly to the cutting edge. Quench the part as soon as the edge achieves the desired color. Tempering to color is not an exact science but it can often get you close enough. Finding an exact color chart is difficult but one can be had from Tempil Corp, the same people who make the temperature indicating crayons.

It is usually preferred to harden the entire part and then selectively temper the piece if necessary. Placing the spine of a hardened knife onto a hot piece of steel will soften the spine more than the edge, leaving the edge hard and the spine softer and tougher.

Machinery�s Handbook has the following list of colors versus temperature in Degrees F.

430*F - - very pale yellow

440*F - - light yellow

450*F - - pale straw yellow

460*F - - straw yellow

470*F - - deep straw yellow

480*F - - dark yellow

490*F - - yellow brown

500*F - - brown-yellow

510*F - - spotted red-brown

520*F - - brown-purple

530*F - - light purple

540*F - - full purple

550*F - - dark purple

560*F - - full blue

570*F - - dark blue

580*F - - light blue

Choose a tempering temperature appropriate for the alloy and application. Remember that the temper color indicates the temperature of the piece, NOT the hardness! A pale straw color on a piece of 5160 will be MUCH harder than the same color on a piece of A36. Be aware that if you over temper the tool, it must be re-hardened before you temper it again.

To keep heat away from areas you do not want heated, wrap a dripping wet towel tightly around it. Knives can be selectively tempered by brightening the cutting edges, placing the spine of the blade on a heated block and allow the heat to conduct to the cutting edge. The tip will heat faster than the body, so try to keep the tip out of contact with the block.

When steel is heated to hardening or forging temperatures, the oxygen in the air combines with the carbon on the surface of the steel. It forms CO2 and is carried away, leaving the surface of the steel without carbon. This process is call decarburizing (not de-carbonizing). It leaves the surface of a forged part with a very soft skin of pure iron (ferrite). If a blade has not been ground to shape prior to hardening, this skin will remain on the finished part. The ferrite skin is very soft and will scratch easily, ruining a high polish. The skin is usually only a few thousandths deep and can be easily removed with light grinding with a fine grit abrasive.

Summary of important points.

* Steel is an alloy of iron, carbon and alloy elements.

* Carbon content controls the hardness and strength of the steel.

* Higher carbon and alloys make the steel stiffer to forge.

* Alloys are added to allow heavy sections to form martensite deeper into the steel.

* It is the cooling rate when quenched that determines what microstructure will be created. Steel must cool fast to form martensite.

* Some steels can be cryogenically treated (frozen) to improve hardness.

* Uniform heating and cooling are important to prevent distortion and cracking.

* Avoid rough surfaces and rapid section changes to prevent cracks when quenching.

* Multiple normalizing prior to hardening can improve hardness and toughness.

* Multiple tempering can improve toughness.

Cryogenic Treating Steel

Some high carbon steels, like tool steels, do not complete the transformation from austenite to martensite until it gets well below zero. This leaves some retained austenite scattered throughout the martensite. When you temper the part, the retained austenite will transform to things other than martensite, like ferrite, pearlite or bainite, none of which is as hard and abrasion resistant as martensite.

Freezing the part is a continuation of the quench and it allows all the austenite to transform to martensite. This improves the hardness, abrasion resistance, and toughness because there are no secondary phases mixed in with the martensite to degrade its performance.

It really does work, especially on high carbon tool steels. Some people even cryo treat razor blades, musical instruments, and golf balls.