Cryogenic Processing of Steel Part 3 – Wear Resistance and Edge Retention

Thanks to Ian Rogers for becoming a Knife Steel Nerds Patreon supporter!

Intro to Cryo and Wear Resistance

In Cryogenic Processing Part 1 I covered the effects of cryo on retained austenite and hardness. In Cryogenic Processing Part 2 I looked at the studies on cryo and toughness. Wear resistance is the most controversial aspect of cryogenic processing of steel. In particular there are claims that the use of cryogenic processing (liquid nitrogen) leads to an improvement in wear resistance that is not found with subzero processing (dry ice). Sometimes it is claimed that cryo can lead to massive increases in wear resistance [1]:

These claims lead to skepticism. How can cryo provide 6.5x the wear resistance of conventionally heat treated steel? I combed through the journal articles and figured it out if it does. 

Wear Resistance from Hardness and Carbides

Higher hardness, in general, means better wear resistance. In the chart above lower is better because it means that less material was lost during the wear test. Harder materials are more difficult to scratch or abrade. Knife steels usually also have carbides which are hard particles that contribute to wear resistance. Those carbides are the reason that 10V has much better wear resistance than L6 even when both are heat treated to 60 Rc. You can read about the contributions of hardness, carbide amount, and carbide type in these articles on edge retention of knives:

Which Steel Has the Best Edge Retention? Part 1

Which Steel Has the Best Edge Retention? Part 2

Detour – Carbides

In the chart at the top of the article there was a very large improvement in wear resistance from cryo processing shown relative to subzero processing. Both subzero and cryo can be effective for transforming retained austenite to martensite and increasing hardness. There are many studies claiming that the cryo process leads to microstructure changes that improve wear resistance apart from hardness and retained austenite transformation. The majority of the proposed mechanisms for improvement in wear resistance from cryo involve changes in carbides. Therefore some background is necessary to understand the changes they are talking about. There are many different types of carbides, and I am going to focus on a few major categories:

Primary Carbides

Primary carbides are formed during casting (solidification) of steel, often forming in the liquid iron. These carbides are highly stable, typically large, and are generally not completely eliminated through forging and heat treating. I wrote about why these carbides form and how powder metallurgy helps reduce their size in this article on powder metallurgy. 

Secondary Carbides

Secondary carbides are those that dissolve during high temperature processing like forging and re-form, or re-precipitate, at lower temperatures. A high alloy steel like D2 or 154CM contains a mixture of both primary and secondary carbides. Here is the same micrograph of D2 from above but instead with the secondary carbides pointed out:

A size cutoff between secondary and primary carbides is difficult to pinpoint, there can even be some overlap. Primary carbides are larger than 1 micron, and typically larger than 3 microns. Some steels do not have primary carbides after typical forging and heat treating and only have secondary carbides. They are designed so that the carbides fully dissolve at forging temperatures and precipitate at lower temperatures where they are smaller than primary carbides that form during casting. Examples of these steels include AEB-L/13C26 and low alloy carbon steels like 1095. Here is a micrograph of 13C26 showing its small secondary carbides:

Tempering Carbides

After quenching steel to form hard martensite, the carbon is supersaturated in the martensite of the steel. During tempering, very tiny carbides are formed as the carbon leaves the martensite, even smaller than secondary carbides. You can read more about the formation of tempering carbides in this aticle on tempering of steel. At low temperatures, “transition” carbides are formed, referred to as ε-carbides or η-carbides. Transition carbides are very tiny, typically only viewable with transmission electron microscopy (TEM), which is an expensive and difficult method of microscopy, less common than scanning electron microscopy (SEM) in the micrographs shown above for D2 and 13C26. At typical tempering temperatures, those transition carbides are replaced by cementite, Fe₃C. Here is a TEM image of a steel tempered at 150°C (300°F) with a combination of transition carbides and cementite. Tempering carbides are in the nanometer size range rather than the micron size range. In the image below the cementite is the linear carbides around 50-100 nm (0.05-0.1 microns), and the transition carbides are the rows of very tiny round features (2-4 nm) [2]:

At yet higher temperatures, “alloy” tempering carbides are formed. These are Cr, Mo, W, or V carbides that lead to secondary hardening, which is a bump in hardness that is seen with tempering temperatures greater than about 400°C (750°F) in steels with significant amounts of those alloys. The larger cementite or alloy carbides are often visible with SEM images, where you can see the size difference with secondary carbides. Below are images of 440M, a steel similar to 13C26, where you can see the “large” secondary carbides on the order of a micron without any tempering carbides after quenching (a) tempering carbides from a 300°C (572°F) temper leading to cementite (b) and alloy carbides after a 500°C (932°F) temper (c) [3]:

Wear Resistance from Cryo

OK, we are back on track now. To look at the effect of cryo on steel wear resistance I gathered 13 articles [4-16] and have summarized the results below. I did not include the study from the chart at the top of the article [1], because its data is presented in that chart. Otherwise, I included every study that I found. 

The % improvement above is for wear resistance at a consistent hardness. If hardness data is not available in the study then I do not have an improvement listed, which includes the study in the chart at the top of the article. Each improvement is for the greatest difference measured in each study.

The tempering temperature is listed because many of the claims about wear resistance improvement involve carbides so tempering is important. Also cryo time is listed because there are a couple studies that claim that the time in cryo controls the carbide formation behavior. In general, the studies reported an improvement in wear resistance of approximately 20%, but there are a few which reported no change. When averaged the result is about 13%. Because of the variety of different steels, heat treatments, and wear resistance results, it is difficult to come to conclusions based only on the reported numbers. Therefore I am going in deeper and have analyzed a few of the specific studies. 

Wear Resistance Tests

The majority of the experiments above were for “sliding” tests, specifically a pin-on-disc test where a hardened steel ball is held with a fixed force against the test “disc” while the disc spins at a constant rate. The wear of the disc material is typically what is measured. Why these tests are so common for cryo wear studies I’m not sure. Probably because the early cryo studies used it and it became standard. How well this type of test correlates with knife edge retention is debatable but perhaps we can hope it is an indication of wear resistance changes at least. This type of test favors powder metallurgy steels and steels with vanadium carbides. 

The scratch test listed is somewhat similar except instead of a hardened steel ball is uses a diamond stylus which scratches into the steel.

The dry sand/rubber wheel is an abrasive wear test that uses sand as the abrasive. The test above listed as “abrasive” was a test with aluminum oxide (alumina). These types of tests are probably more applicable to slicing edge retention. However, the large abrasive size favors steels with large carbides, and carbide size does not have a strong effect on slicing edge retention. 

No Difference in Wear Resistance with Cryo

Three of the tests in the table showed no improvement in wear resistance when compensated for hardness [4, 11, 12]. Steel can usually be given a lower tempering temperature to achieve a higher hardness so if the wear resistance is the same when at the same hardness that means that there are no other mechanisms in play from cryo for increasing wear resistance. Here are plots of wear resistance vs hardness showing that there is no clear difference between the cryo-treated and conventionally heat treated steel:

All of these studies looked at several different variables, including a range of hold times at cryogenic temperatures, low vs high temperature tempering, and even powder metallurgy vs conventionally cast steel. The only similarity between them is they used sliding wear tests, but so did many other studies that did show an improvement from cryo. Therefore it is difficult to pinpoint any reason why these studies would find no improvement in wear resistance from cryo apart from hardness, even though other studies did. 

Wear Resistance from η-carbides

The first study generally cited claiming that carbide formation leads to an improvement from cryogenic processing is from 1994 on D2 steel [10]. They found that the use of subzero processing with dry ice led to little difference in sliding wear resistance but that liquid nitrogen processing led to a significant improvement in wear resistance (lower is better):

There was little difference in retained austenite between subzero and cryo so the researchers thought there must be some other mechanism to explain the improvement in wear resistance:

They analyzed the tempering carbides of the D2 steel with transmission electron microscopy (TEM) and determined that the cryo treated steel contained η-carbides rather than the ε-carbides in the subzero or conventionally heat treated samples. 

Since the publication of that study there have been many more cryo studies where they proposed that η-carbides led to an improvement in wear resistance. However, almost none of them directly observed differences in transition carbides but simply cited the above study. This is in part because of the difficulty and expense involved in TEM studies. However, even with TEM studies, differentiating between η- and ε-carbides is difficult as they have a very similar structure. In more recent studies on tempering of high carbon steels it is found that transition carbides are η-carbides rather than earlier studies which thought they were ε-carbides [2]. These are not cryo studies but general tempering studies. Therefore not only do we not expect cryo to lead to a change from ε- to η-carbides, but we expect any steel, with cryo or without, to form η-carbides during low temperature tempering.

Furthermore, as described in the carbides section, ε- and η-carbides are called “transition carbides” because they transform to the more stable cementite with typical tempering temperatures, even with “low” temperature tempering around 200°C. Five of the studies in the table cited above use high temperature tempering where transition carbides would certainly not be present. 

More recent studies on cryogenic processing seem to be distancing themselves from the η-carbides explanation. From a 2017 review of cryogenic literature [18]:

“Several later investigations have not been able to detect the presence of 
η-carbide in cryotreated tool steels; this is expected because η is a transition carbide and hence, either dissolves or gets converted into stable [cementite] or alloy carbides…It is established now beyond doubt that [cryo] does not alter the nature of carbides for a particular material.” 

Wear Resistance from Secondary Carbides

Cryo researchers have instead shifted to a different proposed mechanism for wear resistance improvement: an increase in density of “secondary” carbides. They claim [18] that the cryogenic processing leads to “conditioning of martensite” which means that defects like dislocations are created which leads to clustering of carbon atoms at these defects. During tempering more carbides are then formed because of the dislocations and carbon clustering. You can read more about dislocations in this article on grain refinement. Below are example micrographs showing the difference in secondary carbide density, where they differentiate between large secondary carbides (LSC) and small secondary carbides (SSC) and show that there are more SSCs in the cryo treated D2 [18]:

The increase in secondary carbide density is the claim in most of the articles cited in the summary table. One example is this study on D2 [9] where they looked at conventionally heat treated, subzero treated, and cryo treated steel. They found that the liquid nitrogen treated steel had somewhat better wear resistance than the subzero treated steels even though the hardness was essentially the same. However, the improvement was small; there was a larger effect of austenitization/hardening temperature. 

There is reason to be skeptical of this secondary carbide explanation. Few studies have compared subzero with cryo so very few of them are showing that liquid nitrogen shows an improvement beyond dry ice. But more fundamentally, the carbides shown in these micrographs are much too large to have formed during tempering, relatively great diffusion distances of carbon would be necessary to form such large carbides. It seems highly unlikely that such a huge change in carbon diffusion has occurred based on the use of liquid nitrogen. The micrographs above were from D2 tempered at 210°C, a temperature where very tiny tempering carbides are formed. Therefore the researchers are proposing that an increase of several orders of magnitude in carbon diffusion is the result of a liquid nitrogen treatment after quenching.

All of this despite the fact that tempering cryo-treated steel looks only marginally different than conventionally quenched steel. Tempering occurs through the precipitation of carbides and therefore any change in carbide precipitation should show up with a hardness change in tempering. The “cryogenically treated” curve is shifted up because it started out at higher hardness because the retained austenite was transformed to martensite prior to tempering. However, other than the increase in hardness from transformation of retained austenite above 450°C, the tempering curves look similar. The behavior up to the 210°C tempering used to generate the images above is virtually identical [9]:

The reason for the differences observed in secondary carbides is difficult to explain. Perhaps simply because it involves the selection of images and statistical comparisons there is some bias that is creeping into the analysis. In a study of M2 steel they show micrographs of cryo treated and conventionally heat treated steel, showing a vast difference in carbide structure. They even show that primary carbides are no longer present in the cryo-treated steel. This is an extreme example of cherry picking micrographs to show carbide differences [19]:

Some researchers have been critical of other studies claiming an increase in carbide density using optical microscopy because they say that the carbides that form are too small to view with optical microscopy. They have not applied the same scrutiny against their own studies that claim that they see a difference with scanning electron microscopy [18].

Time in Liquid Nitrogen

Some studies have also claimed a difference in wear resistance based on the amount of time at cryogenic temperature. These claims rely on a mechanism involving an optimum time for generating the highest density of carbides during tempering. In a study on D2, they found an improvement in wear resistance up to 36 hours, and somewhat lower wear resistance with longer cryo times [8]:

When plotted vs hardness the trend is clear; hardness is the greatest controlling factor for wear resistance:

Only one test breaks from this pattern, which was the 36h cryo time. This makes me suspect the 36h wear test is an outlier. To explain why I think so, here is a summary of the reported data including hardness (in Vickers rather than Rockwell C), secondary carbide fraction (%), and the wear rate reported in the chart above:

The hardness increased up to 36 hours and remained flat with increased time beyond that, while the secondary carbide amount increased up to 36 hours and decreased slightly. There was a minimum amount of wear at 36 hours (best wear resistance), with somewhat more wear with other cryo times. The authors explain the peak in wear resistance at 36 hours because of the peak in secondary carbide density. There is a relatively large reported increase in carbide fraction from 0 hours to 36 hours, but the carbide fraction only decreases slightly between 36 and 132 hours. Despite that, all of the values on the chart are on the same trend line vs hardness except for the 36 hour cryo condition. In fact, the 0h cryo condition is actually slightly below the trend line (better wear resistance) despite the reported lower secondary carbide fraction. This makes it appear that differences in hardness are the controlling factor for wear resistance and the 36h sample performed better because of experimental variability. 

If it is not an outlier, it is still concerning, as the authors propose that different steels may have different behavior and a different optimal cryo time. If the D2 has a somewhat different composition, or a different austenitization temperature, is 36 hours still optimal? Extensive studies would be required for every steel and heat treatment to ensure that the optimal cryo time is used, and otherwise only hardness changes would lead to differences in wear resistance. This is outside the capabilities and time of the average knifemaker. I think the 36 hour test being an outlier is more likely due to my reasoning explained above along with the reasons for being skeptical of the reported carbide fractions. 

Real Carbide Differences from Cryo

There are changes to carbides that do occur from cryo. This is because where retained austenite is present tempering carbides do not form. Where there is no martensite there are no tempering carbides. However, this would not explain why liquid nitrogen processing (cryo) would lead to superior wear resistance relative to dry ice processing (subzero). Here is a micrograph of M2 heat treated to have a large amount of retained austenite showing the smooth nature of it free of tempering carbides [20]:

Other Studies Showing Improvement from Cryo

These two studies on 52100 [5] and D6 [16] show relatively simple behavior, with cryo leading to an increase in hardness, but also an improvement in wear resistance even after accounting for hardness. 

In this study on Vanadis 6 [13] they found that subzero led to a decrease in hardness and that liquid nitrogen led to a further decrease. Wear resistance, however, improved with subzero and liquid nitrogen despite the reduction in hardness. The reduction in hardness is somewhat puzzling and was explained in Part 2 as being from the shift in the tempering curve with secondary hardening. The improvement in wear resistance despite the reduction in hardness seems to indicate that the cryogenic processing was effective in improving wear resistance by a mechanism unrelated to hardness. 

Cryo vs Subzero Wear Resistance Summary

Either cryo or subzero treatments can lead to an increase in hardness from the transformation of retained austenite which we would expect to lead to an improvement in wear resistance. As shown in the table above, there are studies which demonstrate that cryogenic processing can lead to an improvement in wear resistance above and beyond hardness alone. However, there are a few studies which showed no improvement, and I don’t see any commonalities among them which would explain the difference. The proposed mechanism for the difference in wear resistance from cryo, the increased density of secondary carbides, does not make sense to me and I don’t believe it to be real. There is the possibility that more studies are not finding a difference, or even a small negative difference in wear resistance from cryo, and not publishing it because positive studies are more likely to be published. However, when summarizing the studies cited above, an average improvement in wear resistance of ~13% was found relative to conventional heat treating. Now I will summarize the limited studies which have been presented with knife edge retention testing to see if we find a similar 13% improvement. 

Edge Retention Tests with Knives

CATRA Testing

In a 154CM CATRA study that I previously wrote about, many factors were compared including the effect of cryo. The steel was given a 1 hour snap temper at 300°F followed by 4 hours at liquid nitrogen temperatures, and finally tempered at 960-1000°F. The tempering had a range to achieve the same hardness regardless of cryo which was successful, as the cryo-treated steel was 61.7 Rc on average and the non-cryo was 61.5 Rc. There was no difference between cryo and no-cryo, in fact if anything the non-cryo steel was slightly better (higher is better):

Rope Cutting Tests

Wayne Goddard performed rope cutting tests with and without cryogenic processing with 440C, 154CM, and ATS-34 [21]. Cryogenic processing led to an improvement in edge retention due to an increase in hardness, but when compensated for hardness there was no difference. For example, the cryo treated ATS-34 was 60 Rc while the non-cryo 154CM was also 60 Rc, with a cryo step the 154CM reached 61 Rc. These steels are identical so the 1 Rc difference in hardness is likely due to small differences in composition or heat treatment between the two knives. However, both of the 60 Rc knives achieved the same number of cuts with 38. The 61 Rc cryo-treated 154CM reached 44 cuts. Therefore, cryogenic processing did not lead to an improvement in edge retention apart from the change in hardness from the transformation of retained austenite:

Proposed CATRA Study

Perhaps a good CATRA study to look at the proposed improvement in wear resistance from cryogenic processing apart from simply converting retained austenite would be the following, using three knives from D2 steel:

1. Subzero (dry ice) for 1 hour after quenching, followed by tempering
2. Cryo (liquid nitrogen) for 1 hour after quenching, followed by tempering
3. Cryo for 36 hours after quenching, followed by tempering

This would allow us to test subzero vs cryo and the effect of cryo time on wear resistance. Using D2 would allow a direct test of the optimization study, without worrying about the possibility that different steels may have a different optimal time. 

Summary and Recommendations

Cryogenic processing can improve wear resistance due to an increase in hardness and transformation of soft austenite to martensite. Some studies have claimed that there is a further improvement in wear resistance apart from those two factors, usually attributing the improvement to some change in carbides from cryogenic processing that is not seen with subzero. These carbide explanations are not very convincing, but there are studies that found an improvement in wear resistance beyond that explained by hardness when adding a cryogenic treatment step. One influential study showed that the time in cryo is an important factor but their results appear to be based on an outlier and not a real effect. More time in cryo can also increase the chances of cracking or splitting because tempering is being delayed. Edge retention tests with CATRA and rope cutting have not revealed an improvement from cryo outside of the hardness difference. Edge geometry and selected steel have a much stronger role in controlling edge retention than cryo. For these reasons, and more given in Parts 1 and 2, I recommend that knifemakers use a short (<60 min) liquid nitrogen treatment directly after quenching to maximize the retained austenite transformation and temper immediately after. 

---

[1] Barron, R. F. “How cryogenic treatment controls wear.” In 21st InterPlant Tool and Gage Conference. Shreveport, LA, USA, p. 205. 1982.

[2] Krauss, George. Steels: processing, structure, and performance. ASM International, 2015.

[3] Lin, Yuli, Chih-Chung Lin, Tsung-Hsien Tsai, and Hong-Jen Lai. “Microstructure and mechanical properties of 0.63 C-12.7 Cr martensitic stainless steel during various tempering treatments.” Materials and Manufacturing Processes 25, no. 4 (2010): 246-248.

[4] Zurecki, Zbigniew. “Cryogenic quenching of steel revisited.” In Heat Treating: Proceedings of the 23rd ASM Heat Treating Society Conference. 2005.

[5] Paydar, H., K. Amini, and A. Akhbarizadeh. “Investigating the effect of deep cryogenic heat treatment on the wear behavior of 100Cr6 alloy steel.” Kovove Materialy-Metallic Materials 52 (2014): 163-169.

[6] Amini, K., A. Akhbarizadeh, and S. Javadpour. “Investigating the Effect of the Deep Cryogenic Heat Treatment on the Mechanical Properties and Corrosion Behavior of 1.2080 Tool Steel.” International Journal of Iron & Steel Society of Iran 12, no. 2 (2015): 24-29.

[7] Sehri, Masoud, Hamid Ghayour, Kamran Amini, Masaab Naseri, Habib Rastegari, and Vahid Javaeri. “Effects of cryogenic treatment on microstructure and wear resistance of Fe-0.35 C-6.3 Cr martensitic steel.” Acta Metallurgica Slovaca24, no. 2 (2018): 134-146.

[8] Das, D., A. K. Dutta, and K. K. Ray. “Optimization of the duration of cryogenic processing to maximize wear resistance of AISI D2 steel.” Cryogenics 49, no. 5 (2009): 176-184.

[9] Collins, D. N., and J. Dormer. “Deep Cryogenic Treatment of a D2 Cold-Work Tool Steel.” Heat Treatment of Metals 3, (1997): 71-74.

[10] Meng, Fanju, Kohsuke Tagashira, Ryo Azuma, and Hideaki Sohma. “Role of eta-carbide precipitations in the wear resistance improvements of Fe-12Cr-Mo-V-1.4 C tool steel by cryogenic treatment.” ISIJ international 34, no. 2 (1994): 205-210.

[11] Molinari, A., Mater Pellizzari, S. Gialanella, G. Straffelini, and K. H. Stiasny. “Effect of deep cryogenic treatment on the mechanical properties of tool steels.” Journal of materials processing technology 118, no. 1-3 (2001): 350-355.

[12] Pellizzari, M. “Influence of deep cryogenic treatment on the properties of conventional and PM high speed steels.” Metall Ital 9 (2008): 7-22.

[13] Sobotova, Jana, Petr Jurci, and Ivo Dlouhy. “The effect of subzero treatment on microstructure, fracture toughness, and wear resistance of Vanadis 6 tool steel.” Materials Science and Engineering: A 652 (2016): 192-204.

[14] Çiçek, Adem, Turgay Kıvak, Ilyas Uygur, Ergün Ekici, and Yakup Turgut. “Performance of cryogenically treated M35 HSS drills in drilling of austenitic stainless steels.” The International Journal of Advanced Manufacturing Technology 60, no. 1-4 (2012): 65-73.

[15] Candane, D., N. Alagumurthi, and K. Palaniradja. “Effect of cryogenic treatment on microstructure and wear characteristics of AISI M35 HSS.” International journal of materials science and applications 2, no. 2 (2013): 56-65.

[16] Naravade, R. H., and A. P. Aher. “ANALYSIS OF WEAR BEHAVIOR OF D6 TOOL STEEL BY INFLUENCE OF CRYOGENIC TREATMENT.” IRIJET 4, no. 5  (2017): 1146-1150.

[17] Laino, S., J. A. Sikora, and R. C. Dommarco. “Development of wear resistant carbidic austempered ductile iron (CADI).” Wear 265, no. 1-2 (2008): 1-7.

[18] Ray, K. K., and D. Das. “Improved wear resistance of steels by cryotreatment: the current state of understanding.” Materials Science and Technology 33, no. 3 (2017): 340-354.

[19] Gill, Simranpreet Singh, Jagdev Singh, Rupinder Singh, and Harpreet Singh. “Effect of cryogenic treatment on AISI M2 high speed steel: metallurgical and mechanical characterization.” Journal of Materials Engineering and Performance 21, no. 7 (2012): 1320-1326.

[20] LOMNO, VPLIV MIKROSTRUKTURE NA, and ILAVOST VAKUUMSKO TOPLOTNO OBDELANEGA HITROREZNEGA. “THE INFLUENCE OF MICROSTRUTURE ON FRACTURE TOUGHNESS OF VACUUM HEAT TREATED HSS AISI M2.” MATERIALI IN TEHNOLOGIJE 35, no. 5 (2001): 211.

[21] https://sharpeningmadeeasy.com/edge.htm