Annealing, Austenitizing, Forging, Heat Treating and Processing, Quenching, Tempering

How to Use a Steel Datasheet to Develop a Heat Treatment

Thanks to David Reem, Colton, camilo, Brendan Gildea, Vince Koacz, Monery Custom Cutlery, Curt E, Shannon Sanders, Brazilian Blades, Rory Kelly, Brunhard, Zachary Chumley, Noel, Adam Nolte, Gundam lupus, and Jan Huch for becoming Knife Steel Nerds Patreon supporters! I still don’t know when it will be time for the announcement of my new steel so until then you can get all of the data on its properties exclusively on Patreon.

Sources for Datasheets

The first place to look for a steel datasheet from the steel manufacturer. Places to look include Crucible, Carpenter, Bohler, Uddeholm, Bestar, Zapp, or sij. The ASM Heat Treater’s Guide is also available as an app on Android and iOS though the charts and micrographs are missing. Datasheets will sometimes be missing pieces of information, such as recommended forging or normalizing temperatures. I would recommend looking at several datasheets to see if more information is available from one manufacturer or another.

Forging, Normalizing, and Annealing

These steps are only required for bladesmiths that are forging their blades. Steel comes forged and annealed from the manufacturer so if the blade is being produced for stock removal the steel is ready for drilling, grinding, and heat treating.

Forging, normalizing, annealing data for 52100 from the Heat Treater’s Guide app

Looking at different datasheets, forging and normalizing temperatures are often missing. The ASM Heat Treater’s guide is a good place to look for many low alloy steels. In the absence of recommended forging temperatures I would use no higher than 2100°F. High carbon steels melt at lower temperatures, and even if the entire blade hasn’t melted, the grain boundaries can melt leading to poor “hot ductility” leading to cracking even when the steel is very hot. And of course lower temperature means that grain growth is minimized. Forging at temperatures that are too low also means poor hot ductility and increased risk of cracking, of course. I wrote more about forging temperature and hot ductility in this article.

The 52100 datasheet from Crucible has no forging or normalizing recommendations

The goal of normalizing is to dissolve all of the carbide and form a new grain structure followed by air cooling to form pearlite. This step is necessary because the structure can be very uneven after forging. The goal is not necessarily grain refinement though the grain size will also be reduced. For normalizing it is again difficult to find recommendations, with the Heat Treater’s Guide being the best source, usually. My book Knife Engineering also has recommendations for a range of steels that are not found in the ASM guide (and also those found in the guide). I don’t have a specific article on normalizing on this website yet so those recommendations are only found in the book right now. A generic normalizing temperature would be about 1600°F. That is somewhat higher than necessary for some steels but won’t hurt it. Only steels with very high carbon or chromium need more temperature, like White #1, 26C3, 52100, 1.2519, Blue Super, 1.2562, etc. High alloy steels (A2, D2, stainless steels, etc.) do not need normalizing and in fact can’t be normalized, at least not in the way we typically think about normalizing. When you normalize, heat the steel through for 10-30 minutes and cool in air. Once the steel is fully magnetic again that means the transformation is complete. Some high hardenability steels like O1 or L6 will partially harden when normalizing and will need to be annealed to be workable.

The Latrobe 52100 datasheet has no forging temperature recommendation, but does have a recommended normalizing temperature. The range is higher than in the Heat Treater’s Guide. I think the Latrobe recommendation is better as the 1.5% Cr in 52100 increases the temps required for dissolving carbides.

Many knifemakers then perform “grain refining” steps at lower temperatures than then initial anneal. It is common to see recommendations of  descending temperatures, such as 1600, 1550, 1500°F. The descending method isn’t a perfect way of doing it. The normalize needs to be sufficiently high in temperature to dissolve all of the carbide which is why it is at higher temperature. The grain refining step only needs to be high enough in temperature to transform to austenite, so heating to nonmagnetic is sufficient. So the in between step doesn’t accomplish either of those goals, as the temperature is higher than necessary for austenite transformation, potentially leading to slightly larger grains. And it isn’t high enough in temperature to accomplish the goals of normalizing. Note: nonmagnetic is usually not the “Curie” temperature in high carbon steels as I wrote about in this article. A generic recommended temperature for grain refinement of low alloy steels would be 1400-1450°F (or nonmagnetic) followed by air cooling. This can be done once or twice prior to annealing.

S35VN datasheet information on forging, annealing, and stress relieving. Many datasheets also make recommendations on forging and annealing of high alloy steels.

Annealing makes the steel soft for machining and also sets it up for the final heat treatment. The final austenitize can also be performed from the normalized condition but spheroidized carbides resulting from annealing results in better machining behavior than pearlite from normalizing. Annealing cycles are usually found in datasheets though sometimes they can be very generic, such as simply a temperature with the instruction to “slow cool.” For most low alloy steels they can be cooled somewhat faster than is recommended in the datasheet, which results in a bit higher hardness (more difficult to machine), but a finer microstructure and better response to heat treatment. A good cooling rate for 52100, for example, is 670°F/hr, much faster than the 25-50°F/hr found in the datasheets. I wrote about the mechanisms behind annealing in these articles: Part 1 and Part 2. My book also has specific recommended annealing temperatures for most of the available low alloy knife steels used in forging.

Stress Relieving

May datasheets have a recommended temperature for “stress relieving” which are intended to reduce stresses induced from machining prior to heat treating. Those stresses can exacerbate warping during austenitizing and quenching. So if you are commonly overheating steel during grinding, prior to heat treating, you may consider a stress relief treatment. The common recommendation is 1200°F for 2 hours.

S35VN hardening data from datasheet

Preheating

Datasheets usually have a recommendation for a “preheat” prior to the final austenitizing temperature. I wrote about preheating in this article on austenitizing. The preheat is intended to ensure the piece is heated consistently through before the final high temperature, and to minimize the time at the peak temperature. The recommended temperature is also often high enough to transform the steel to austenite at a lower temperature, so that the size changes related to the phase change have completed before the final temperature. If you have two furnaces a preheat won’t hurt anything and may help. However, because blades are thin the preheat isn’t as necessary. You can go straight into the furnace at the final austenitizing temperature.

Hardening data for 52100 from Crucible

Austenitizing Temperature

I wrote about austenitizing of steel in these articles: Part 1, Part 2, and Part 3. The steel must be transformed to austenite (hence the name), and the appropriate amount of carbide dissolved prior to quenching. Different steels have different ranges of austenitizing temperature possible. Optimal properties may be within a relatively narrow range, especially with low alloy steels, such as CruForgeV which has a very sharp drop in toughness when austenitizing at 1550 instead of 1450-1500°F. High alloy steels have more stable carbides that dissolve over a relatively wide range of temperature so they are typically less sensitive to temperature. Usually a datasheet will have a recommended temperature (1950°F) or a recommended range (1900-2000°F). Generally using the middle of the range is safe. For many steels I have specific recommendations in articles on this website or in the book. For some of those we did a series of experiments to find optimal combinations of hardness and toughness, such as AEB-L, CPM-154, and many more.

Austenitizing Time

Some knifemakers are afraid of austenitizing time because they are concerned about grain growth. However, if the selected temperature is appropriate, it can be held at that temperature for a while and not have significant grain growth. Heating the piece through is more important. So I tend to recommend a 10 minute hold with many low alloy steels rather than the 5 minutes recommended by some. High alloy steels have more stable carbides so even though they are austenitized at higher temperature they typically need longer soak times. High alloy steels with a large amount of carbides will “pin” grain boundaries even better than low alloy steels, and so again you don’t need to be afraid of the 15-30 minute soak times recommended by the datasheets. Sometimes reduced times are recommended for higher temperatures, such as in the CPM-M4 datasheet below. At higher temperature the time for carbides to dissolve is reduced so shorter times are possible, and the shorter time helps to minimize grain growth.

CPM M4 austenitizing time recommendations change based on temperature

The exact amount of time can be tricky to judge because generally you don’t know exactly when the knife reached the temperature of the furnace. Somewhat counterintuitively, it takes a knife less time to reach temperature at higher temperatures rather than lower. One way to know that the knife has reached the temperature is simply to see when the knife has reached the same color as the furnace. Steel is very conductive and knives are thin so once the surface has reached the furnace temperature so has the knife. Opening the door of the furnace isn’t optimal but after some experience you will know about how long it takes and then can leave the door closed. However, one simplification recommended by some metallurgists (and in some datasheets) is to start the time when the furnace has rebounded back to the target temperature (opening the door and placing in the cold knife leads to a drop in temperature). Adding on extra time may be necessary with large knives, especially with a thick cross-section.

Quenching

O1 hardening data from Crucible

Different steels require different quenching rates, though datasheets can be somewhat loose guides on this. For low alloy steels it may just say “quench in oil” without guidance on what speed of oil may be required. This is necessary on their end because different sizes, applications, etc. will require different quench rates. Steels come in a range of “hardenability” that dictates how slowly they can be quenched and still achieve full hardness. I wrote about hardenability in this article and ranked steels according to quench speed there and also in Knife Engineering. Some steels that will recommend a water quench can still be quenched in oil as long as it is a very fast oil and the cross-section of the knife isn’t too thick. Steels with higher Mo, Mn, and Cr can be quenched more slowly to help prevent warping and cracking. High alloy steels and stainless steels are often “plate quenched” by knifemakers to accelerate cooling over an air quench and to help maintain flatness. Any of the common steels with 4% or more chromium can be quenched in this way.

Cryo and Cold Treatments

I wrote about cryo in these articles: Part 1, Part 2, and Part 3. Datasheets can vary in whether they recommend a cold treatment and the specifics of their recommendation. For some datasheets the cold treatment isn’t mentioned at all. For low alloy steels I have seen many knifemakers say that they don’t “need” a cryo treatment or even that nothing changes in the steel with cryo. For high alloy and stainless steels I see many knifemakers state that cryo is “required” for good performance. Both of these points of view are incorrect and demonstrate a misunderstanding of cold treatments. High alloy steels can be heat treated without cryo as long as the austenitizing temperature isn’t too high. See the chart below on AEB-L, for example. However, higher hardness is possible when using cryo. The selected austenitizing temperature must be at the peak hardness or lower temperature. At temperatures above peak hardness there will be excessive retained austenite and poor properties, even if the resulting hardness is the same.

And low alloy steels still see a reduction in retained austenite and an increase in hardness, such as shown for 52100 below. So contrary to popular belief these steels are affected by cryo treatments. With low alloy steels you should avoid austenitizing above the peak hardness of the “non cryo” condition, even when using cryo. The excess retained austenite comes primarily from carbon, so if it is forming enough retained austenite to drop hardness that means you are getting brittle “plate martensite.”

Datasheets will sometimes say to perform the cryo treatment in between tempering steps. This helps to reduce the chance of cracking or warping. However, it also reduces the effectiveness of the cold treatment. The most effective way to do a cold treatment is go directly in after quenching. It is essentially an extension of your quench. For some reason many knifemakers will check hardness after quenching but before a cold treatment. Delaying the cold treatment also limits the effectiveness of the cold treatment. If you were quenching in 300°F oil you wouldn’t care what the hardness is before cooling the rest of the way to room temperature, and the same is true when you are at room temperature prior to going into liquid nitrogen. When using a freezer for a small bump in hardness, this is especially true. With T1 high speed steel it was found that a freezer would no longer lead to any transformation if the steel was left at room temperature for only a few minutes prior to the freezer. Liquid nitrogen was more forgiving as up to an hour before putting in LN2 led to little difference, and dry ice was in between. Either way I recommend going directly into the freezer, dry ice, liquid nitrogen, etc. after the quench.

As far as time spent in the freezer or liquid nitrogen, only enough time to cool down to the cold temperature is required. So generally an hour is sufficient. However, there are some (in my opinion, dubious) studies that claim that long hold times in cryo are beneficial, and I have left steel in liquid nitrogen overnight many times without problems, and the knifemakers I have spoken to have reported the same. If it is more convenient to leave the steel in liquid nitrogen longer, such as for waiting for a furnace to cool down, that is generally fine.

Tempering

I wrote about tempering in this article. The goal of tempering is improve toughness, because the as-quenched martensite is brittle. Some datasheets stress the importance of tempering immediately after quenching. This is to avoid cracking. However, as I mentioned in the above section on cold treatments, knives are pretty simple and cracking is usually not an issue. Doing a cold treatment after quenching is fine without worrying about cracking. Tempering is generally done twice for two hours each time. One exception is high alloy steels tempered in the high temperature range (>750°F) where sometimes 3 tempers are necessary. In the first tempering step some retained austenite transforms to brittle martensite which then needs to be tempered. Transformation also occurs in low alloy steels being tempered at low temperatures (~400°F). The time required for that austenite transformation is part of why at least an hour is recommended. Steel could be tempered for a short time at a higher temperature, but that leads to high sensitivity to small changes in tempering time. Longer tempering at lower temperatures leads to more consistency. Things level out after some amount of tempering and then longer times lead to very small changes, for consistent behavior between different size pieces.

S35VN tempering data

The tempering temperature is also important to hardness and toughness. Many datasheets include hardness vs tempering temperature charts, sometimes for multiple austenitizing temperatures, and occasionally along with toughness data. A lower temper typically leads to higher hardness but lower toughness. 400°F is generally a good starting point for balanced properties. I usually don’t recommend any lower than 300°F even when seeking maximum hardness. The best case scenario is to use toughness data to select a tempering temperature rather than hardness. In other words, select the tempering temperature that offers a good balance of hardness and toughness, and don’t focus only on having some target hardness. There are a few steels that will see a steep drop in toughness below a certain tempering temperature. 5160 had excellent toughness at 375°F and then a big drop when tempering at 350°F, for example. The L6 data from Crucible below shows only 15 ft-lbs with a 300°F from a Charpy C-Notch toughness test, but a jump to 43 ft-lbs with 400°F. Knives usually favor high hardness, so if the datasheet recommends 300-600°F tempering you can generally ignore those higher temperatures. Also, tempering between about 450-750°F can lead to “tempered martensite embrittlement” where the hardness is lower from higher tempering temperature, but the toughness isn’t improved, leading to worse properties. In the L6 data below, tempering at 500°F led to slightly less toughness than with 400°F even though the hardness was reduced by about 2 Rc. So those factors combine together for my general recommendation of 400°F as a starting point for tempering.

L6 data from Crucible

Many high alloy and stainless steels can be given that high temperature temper (>750°F) I mentioned. For stainless steels I almost never recommend this because it leads to a reduction in corrosion resistance. For high alloy steels I still don’t recommend it because we have found somewhat better toughness with a 400°F temper with equivalent hardness. However, the high temperature temper gives better “tempering resistance” for knives that are being overheated in grinding, or for knives given high temperature coatings. Many datasheets for high alloy steels only show tempering curves in the high temperature range, so it is safer to stick with that range without experimentation or heat treating recommendations from sources like Knife Engineering that looked at the low temperature range. Often a lower austenitizing temperature is required when tempering at 400°F, for example.

Summary and Conclusions

Datasheets can vary in the amount of information they provide, but are usually the best starting point when developing a heat treatment. Cross-checking multiple datasheets, and understanding why the datasheet recommends what it does, helps in selecting the processing the steel will undergo. Some missing information can be found in other sources such as articles on specific steels on this site or the book Knife Engineering. Before ignoring recommended temperatures in a datasheet you must be very sure about why you are ignoring that advice. Even experienced knifemakers can provide poor heat treating advice so I would generally recommend following the datasheet unless you have reliable hardness and toughness data that says otherwise.

2 thoughts on “How to Use a Steel Datasheet to Develop a Heat Treatment”

  1. thank you for these great informations .
    I have some question :
    – high speed steels are known for relatively low hardenability of large sized parts due to precipitation of vanadium carbide , I want to make large punches “up to ~8 inches” with highest possible compressive strength . can heavy brine “i.e. 25% NaCl” solution either hot or not used for quenching instead of oil to ensure “full hardening” ?

Leave a Reply

Your email address will not be published. Required fields are marked *