Heat-Treating Tool Steels
by Dave Smucker
In the article "Steels Useful for Tools", we have detailed a number of tool steels that could be useful to the blacksmith. In this article, we will look at ways to heat-treat those steels and try to understand a little about what is happening with the steel as we go through the heat-treating process.
Do you need to heat-treat all of your tools? The simple answer is No. Depending on what steel is used and how hot you get the tool in use there may be only a small benefit from heat-treating the tool. For example, Dan Tull likes to use 5160 spring steel, either from coil springs or leaf springs for almost all of his tools. This includes hot cuts, punches, drifts etc. For all of the applications where you are going to get the tool very hot – Dan's experience is that there is no reason to heat-treat these tools. This is because in the process of using the tool and thereby heating the tool you are most likely to fully draw the temper of the tool anyway. His suggestion is to let the tool air cool after making and, thereby, normalizing it. I agree with Dan on this point for tools made of 5160 that you are going to get very hot such as a hot cut or punch. I also have found in my experience that 5160 is a good material for such tools. To quote Dan "why would you use anything else – it works, doesn't it?" Well, yes, but here I think that for some of the very toughest applications in blacksmith work S7 and H13 work even better, really hold up and are worth the effort. Tom Clark, for example, makes his hammer drifts out of H13. I love hot cuts out of S7 or H13. I always heat treat these steels. By heat-treating these tools, you will obtain a tool with a very high life. For items that you are not going to get above the tempering temperature such as cold cuts, eye punches, center punches, drifts, hammers, etc. then a full heat treat is very worthwhile. Remember all of these steels, (4140, 4340, 5160, 1080, W1, S7, A2, H13) are designed to be heat-treated. I am getting a little ahead of myself here, talking about tempering temperatures when we haven't even talked about what is going on in the heat treat process itself yet, so bear with me for a bit if the above doesn't really seem to make complete sense to you.
Do I need to be a Metallurgist to do Blacksmithing? I think that it can be helpful to the blacksmith to understand some of what is going on inside the metal as it is heat treated – but you don't have to. Many blacksmiths have produced outstanding work, including making all of their tools, without a deep understanding of what is happening inside the metal. Later in this article, I will give you some specific steps for heat-treating S7, H13, W1, O1, 4140, 5160, and 1080 but for now I would like to give a shot at trying to explain some aspects of the metallurgy of heat treating. If you want to learn more about this a basic book, that I would highly recommend is Metallurgy Fundamentals by Daniel A. Brandt ISBN 0-87006-922-5. Norm Larson has this book. It is written by a mechanical engineer and provides information on the basic principles of metallurgy without trying to turn you into a metallurgist. Its text is clear, simple, and very well illustrated. Again, it is just a great book for the average blacksmith. By the way, I am not a metallurgist either, but rather a mechanical engineer, so for the metallurgists out there that is my excuse for anything I get wrong in this article.
Carbon steel is a remarkable material. Depending on how much carbon it contains, its process and thermal history, steel can be "soft and ductile" or "hard and strong" and many points in between. We can also change its performance by adding other metals to the basic carbon and iron and further enhance its properties. The reason that we can control so many of the characteristics is that steel is one of a limited number of materials that changes their crystalline structure as we change the temperature – even while it remains in a solid state. It is not just the temperature of the steel that is important – it's the temperature change vs. time that is also critical. It makes a difference how fast we heat the steel and especially how fast we cool (quench) the steel.
Three crystal structures of steel are important to us. The first of these is the body centered cubic space lattice, a crystal that has an iron atom in each corner and one in the center of the cube. We call this ferrite and it is the normal structure of annealed (soft) carbon steel at room temperature. A single cube contains 9 iron atoms. It is also magnetic in nature. The second crystal structure is a face centered cubic space lattice. This is called austenitic iron. It is the normal structure above the "transformation temperature", and as you might have guessed, is non-magnetic at this temperature and in this structure. A single cube contains 14 atoms.
Body Centered Cubic
Face Centered Cubic
Now here is where it gets interesting, at least from my viewpoint. If we take those 14 iron atoms and rearrange them from face centered cubic to body center cubic – we get two body centered crystals. What? How do you get from 14 atoms to two cubic with 9 atoms each – that would be 18, right? But the two cubes are next to each other, and share the same 4 atoms – so it only takes 14 to make 2 body centered cubic crystals. What good is all of this? I had trouble understanding this until I made a sketch and counted all of the atoms in the two forms. Well here is one of the things that create problems when we heat treat – because the volume of the body centered cubic is greater than the face center. This means that as we cool through the transformation temperature we actually expand the steel. More on this later.
Body Centered Tetragonal
If we reheat this highly stressed martensitic structure, body centered tetragonal, we release a great deal of the stress, without removing all of the hardness and strength. If we heat it high enough, back above the transformation temperature, and slow cool it we will get a fully annealed structure.
When Heated Carbon Steel becomes Non-Magnetic at 1420F. For many blacksmiths, one of the key indicators that they use in heat-treating is that when heating plain carbon steels they become non-magnetic at just below the transformation temperature. This works very well to tell when we are at the transformation temperature (critical point) for plain carbon steels. For these steels, it can tell us we are ready to quench the steel. One major caution here, this does NOT WORK for the high alloy tool steels. For them, the transformation temperature is quite a bit above the non-magnetic point.
High temperature means larger grain size. Why not just heat it up good and hot, and know for sure we are above the transformation temperature before we do our quench? Well, the reason is that if we do this we will get a very large grain structure, something we generally don't want to have in our finished product. I also think quenching from too high a temperature leads to "quench cracks" in steels such as W1. OK, that makes sense, but why don't we have the grain size problem when we are heating our material for forging? We do, but one of the major things that happen in the forging process is that we generate a fine-grained structure by the work of forging. So if we have made a tool by forging it, and now want to heat-treat that tool, we don't want to overheat the tool and drive it to a larger grain structure. Grain size change is not just a function of over heating, but also how long we are at those elevated temperatures.
What is the Carbon doing? In the above discussion of ferrite, austenite and martensite, we have left out an important part of the final structure of our steel – the carbon and what it is doing. What we have left out of this discussion is that we can form a compound of iron and carbon that is called cementite – or iron carbide (chemical formula Fe3C). Like other carbides, iron carbide is a very hard material, and at room temperature is magnetic. In steel such as 1080, with 0.8 of 1 percent carbon (80 points carbon) we get a fixture of the cementite and the ferrite that forms a structure that is called pearlite. Pearlite gets it name from the way the structure looks under the microscope – looks like mother of pearl, or others describe it as looking like an aerial view of many newly plowed fields. Pearlite looks this way because it is made up of layers of ferrite and layers of cementite adjacent to one another. Below this 80 points of carbon, we get a mixture of ferrite areas and pearlite areas and above this 80 points of carbon; we get a mixture of pearlite and cementite. These structures all depend on the thermal processing history of our steel. We can obtain other structures, such as martensite, from a very rapid quench, bainite and mixtures of these structures. They vary depending on our rates of cooling, and depth within the steel. (We can't cool the inside of the steel as quickly as we can the material near the surface.) All of these structures depend on the alloying, carbon content and thermal processing (heat-treating) used to make the finished product. Ferrite, martensite, austenite, cementite, pearlite, bainite, do I need all of that? No. Just remember that the information on this is out there and you can get into it in any depth you want or need. For example, the knife makers among us often go into this in great detail because they are working very hard to get very high and consistent performance out of their heat-treating.
Volume change in Heat-Treating. There is one area that I want to again talk about and that is the volume change as steel goes through structure changes as it is quenched or cooled. As we quench steel, we have two volume change things happening. One that we are all very familiar with is that as we heat a metal it expands, as we cool it the metal contracts. This by itself can create major stress in a part that is being quenched. The outside of the tool is shrinking while the inside, since it is still "hot" is still larger than when cool. This means we want to stretch the outer section of the tool. Now to this, we also add the effect of the transformation, – where we get a volume increase as we cool – the outside of the part has gone through this expansion, and now starts to shrink as it cools more, and then the inside reaches the transformation point and expands, before it then also starts to shrink. If the stress in the material becomes greater that what the material can hold it must do one of two things or both – it must yield and / or crack. If it cracks, we usually call that an "oh shit". If we step back and think about this a bit we will also see that the rate at which we quench an item will also greatly affect the stress in the part. This means that the slower we can quench an item the less likely we are to have cracks or failure. Now when we talk about quenching we can think of a number of different quenches – from very fast to slow – Caustic solution, brine, water, oil, air. Yep air can be a quench; it is just a rather slow one. Ok, then why not just quench everything with air, and have little or no problems with cracks? We can't do that because the Good Lord said, "For some steel you must have a very very fast quench, and for others a medium speed quench and for still others you can have a very slow quench and still get a hard, strong, and brittle material". In general, the plain carbon steels, 1045, 1080, W1, 1095 all require a rapid quench. Hence, we think of them as water hardening steels. In general, the medium alloy steels such as 4140, 4340, 5160, O1 require a medium speed quench and we can use oil or they become oil-hardening steels. We can also quench these steels in water, (they may crack) and this is done especially where a much deeper depth of hardening is wanted. The third class of steels here are generally higher alloy steels and we can use air as the quenching fluid. Examples would be H13, A2, and S7. The size of the part also has a lot to do with what quench we use, because it is not the quench material that counts but rather the rate of cooling. Very large industrial parts, for example, made of H13 are water spray quench – because they hold so much heat that air would not be fast enough. There is a great deal more to this area of quench speed, alloys, depth of hardness and the structures formed in the steel than we could cover in this article. For more information, I suggest you start with the Daniel A. Brandt book Metallurgy Fundaments.
How to Heat-Treat Your Favorite Tool Steels. Ok, the time has come in this article to talk about just how to heat-treat some of our favorite tool steels. To start with, I am going to recommend another book, if you are looking for complete, straightforward detailed heat-treating directions for tool steels. It is Heat Treatment, Selection and Application of Tool Steels by Bill Bryson, ISBN 1-56990-238-0. I believe Norm Larson also carries this book. It is expensive at $ 40 for a 200-page paperback but a very good book. It is written for the "tool maker" in a machine shop or manufacturing operation – but the blacksmith can use it even if he doesn't have a heat treat furnace, and other "professional" heat-treating equipment.
In our basic heat-treating procedure, we are going to do the following.
1.) Normalize After you have finished your forging and before you start the heat-treating normalize, you piece. When done forging take you work back up to just above the critical temperature and let it slowly air cool. This helps reduce stress in the piece and gives smaller grain size. 2.) Slowly heat your work piece or at least the working end, to above the transformation temperature (critical temperature), depending on the tool steel you are using. Do NOT over heat. 3.) Quench the tool in water, oil, or air depending on the type of tool steel. 4.) Quickly temper the tool by reheating to the tempering temperature depending on the desired hardness vs. toughness and the intended end use. A second temper cycle is often used for critical industrial parts. Quickly is important. I didn't used to pay much attention to the "quickly" part of this last point – going directly to the tempering operation – but it is important. After we have quenched the part, it is in a very high state of retained stresses and wants to relieve those stress by, you guessed it, cracking. If we move quickly to lower those stresses by tempering, we decrease greatly the level of residual stress in the part and the likelihood it will develop stress relieving cracks. These cracks of course lead to failure of the tool. More on tempering temperatures and methods at the end of this article. Slowly is important too. The suggestion to slowly heat our work piece in the 2nd. step is the blacksmith's answer to a stress relieving hold in the industrial practice. In industrial practice, the piece is heated to 1200 F and held for 10 to 20 minutes, before heating to the critical transformation temperature. With our small parts, we can accomplish the same thing by heating slowly. It doesn't have to be dead slow, just don't rush it. Because we are used to being able to quickly heat low carbon steel such as A36 for lots of items we make we may tend to heat tool steel too fast. It is just not as forgiving as mild steel. Do Not Over Heat. I have seen students with cracking problems especially in small sizes of W1 because they over heat. Start on top of your fire and work slowly and carefully.
Knowing the temperature at which you have completed transformation is not easy with the higher alloy steels. The reason for this is that these critical temperatures are above the magnetic / non-magnetic point especially for the A2, S7 and H13. You will have to use metal color as your guide.
Heat-Treating Steps for Various Steels
|Steel||W1 - 1095||O1||4140||A2||H13|
|1080 (railroad rail)||4340||S7|
|Slowly heat to the critical temperature. Slowly means 10 to 15 minutes.||1420 F magnet is good test cherry red||1500 F must be a little above non-magnetic||1575 F must be above non magnetic bright cherry / salmon||1750 F magnet can't be used beginning orange heat from bright salmon||1850 F magnet can't be use lemon yellow|
|Hold at critical temperature for X minutes per inch of thickness||3 to 5 minutes soak time||3 to 5 minutes soak time||3 to 5 minutes soak time||10 to 30 minutes soak time||10 to 30 minutes soak time|
|Quench in the following.||Water; keep the part moving to prevent formation of steam layer.||Quench Oil, keep part moving||Quench Oil, keep part moving||Still Air -- set part on a fire brick and let it cool, yes this is a quench||Still Air -- set part on a fire brick and let it cool, yes this is a quench|
|Temper by heating to draw temperature in oven for 2 hours and let part slow cool. See table 2 for hardness vs. tempering temperature||Tempering Temperatures 300 F to 600 F||Tempering Temperatures 300 F to 600 F||Tempering Temperatures 300 F to 800 F||Tempering Temperatures 300 F to 1000 F||Tempering Temperatures 1000 F to 1100 F|
What do you mean by soak time at critical temperature? After we have brought our tool to the critical temperature, we should hold it there for the time noted per inch of thickness. That means if your tool is 1/2 inch thick you only need 1/2 the time. For the high alloy complex steels like A2, S7 and H13 the handbook values are actually 1 hour per inch of thickness and that is a very long time at a forge. No one is going to do that. Being a little hotter helps a lot, the heat transfer in the forge is much high than in a heat treat furnace so we get by with a shorter time. In addition, we may just not get the maximum results in the steel.
Hardness vs. Tempering Temperature for Various Steels
|Tempering Temperature||Hardness Rc||Hardness Rc||Hardness Rc||Hardness Rc||Hardness Rc||Hardness Rc||Hardness Rc||Hardness Rc|
Hardness vs. Tempering Temperature. The chart on the previous page shows the approx. hardness you will obtain using various tempering temperatures after quenching. Use the highest tempering temperature that makes sense for your tooling application. Remember that high hardness values will mean you have a strong but brittle material. The lower hardness material will be softer and somewhat weaker but much tougher. Also, if the tool will get hot – then your tempering temperature needs to be at or above the temperature that you will get the working end of your tool to in use. Let me use an industrial example from my old day job. In the aluminum industry hot mill work rolls (H11) are tempered at 1000 F because we would get them to that temperature in operation. Foil mill work rolls (52100 bearing steel) were tempered at 212 F because they don't reach that temperature in normal service. For tools made from W1, 5160, 4140, 4340 you can cool them in use to keep them from getting too hot. Now this doesn't work if you get them to a red heat, but as long as you stay below that, it helps retain their hardness and edge. The great advantage of the S7and H13 is that you can get them into the low end of the red heat without much damage.
Oven Tempering vs. Color Tempering. The chart shown assumes you use an oven to do your tempering. I think this is the best way to temper because it is very repeatable and almost no work. You can use your kitchen oven for many of these temperatures or pick up a used toaster oven for your shop. In either case, buy an oven thermometer at Wal-Mart to check your oven temperature. Go by the oven thermometer, not by the dial settings especially on a toaster oven. You can pick up toaster ovens for about $ 5 at the local thrift shops. The oven works well to temperatures of 550F. For temperatures above this, you will need to use either the forge or a torch unless you build or find a special heat treat furnace. We again have to turn to the heated color of the steel to judge the tempering temperature for the high alloy tool steel such as H13. In reduced light you can judge 900 to 1000 F as a very faint red, 1000 to 1100 F as dark red. You can also temper tools at the forge or by using a torch and watching the temper colors. Blacksmiths did this for centuries. If you use this method, you will need to pick a temperature 50 to 75 degree hotter than those shown in the chart above. The reason is that these values are based on time at temperature and with the color method; you stop the process by quenching when reaching the desired temper color. To judge tempering temperatures you can use the color on a bright surface, (polish with emery cloth) and watch the color run from the heat source to the critical working surface of your tool. There are a number of good charts showing those colors and descriptions. See the article on "Break-Testing" in this issue for a listing of colors vs. temperature.
I never Temper, it works for me. (Sometimes)
I hear some blacksmiths brag that they "Never temper" or they may mean I never temper air hardening tool steels. "Just get them real hot after forging and let them air cool." Well as the saying goes, there are two types of Blacksmiths in this world, those who have had tools break and those that are going to have tools break. Not tempering just increases the number of tools that we will have fail in use. Remember, if you want fewer tools to crack, heat slowly, get to above critical temperature, don't over heat and always temper. In fact, always temper right away when finished with your quench.
Copyright 2001, 2005 by David E. Smucker Note to other editors of blacksmith newsletters. You are free to use this article in your publication provide you used it in its entirety and credit the Appalachian Area Chapter of Blacksmiths and author. I can provide you with an electronic copy by contacting me at firstname.lastname@example.org It may not be reproduced in any form for commercial use.