Knife Steel Metallurgist Juha Perttula

Thanks to the new Knife Steel Nerds Patreon supporters! Steve, Bruce Blake, Michael Glaser, James F Khoury, Paul Horner, and Shane Sibert.

Intro

Juha Perttula has a doctorate in metallurgy and he has published several studies in scientific journals on heat treating knife steels. So for this article I want to feature not just an individual study but a person’s work. His studies have focused on traditional bladesmithing, low alloy and carbon steels, and wootz Damascus. He also makes and sells knives which you can read about at his website, juhaperttula.com. Juha’s website also has links to the articles I will be discussing below, and I recommend you read them in addition to my commentary.

Video

Here is a video version of the following content:

Wootz Damascus

Two of Juha’s earliest publications are on Wootz Damascus, published in 2001 [1] and 2004 [2]. Wootz is an ancient steel where unlike pattern-welded Damascus, Wootz is made of a single steel. The pattern instead comes from how the steel is processed where layers of carbide bands are created that are visible macroscopically after etching. This material has been the subject of many legends and much study, including many attempts to recreate the methods by which it was produced. Wootz had very high carbon (~1.5%) which we would expect to lead to relatively high wear resistance (for a simple carbon steel) though toughness would also be reduced by the high carbon content.

The 2001 study by Perttula [1] reported a replication of Wootz by using a small chromium addition, whereas John Verhoeven and Al Pendray used a small vanadium addition based on analysis of historical blades. Historical Wootz blades were not quenched and tempered for high hardness as modern knives are, but had a “fine pearlite” structure which would result in a hardness in the mid-40s Rc, as opposed to modern blades in the range of 52-65 Rc. Cutting tests performed showed that Wootz greatly outperformed mild steel showing that for a time, anciently, Wootz would have shown much higher strength and cutting ability. And in certain cutting tests Juha said that the “cutting capacity” of a 46 Rc Wootz blade was similar to a 1075 steel knife hardened to 60 Rc.

In the 2004 study [2] he expanded the study to include Wootz with different microstructures, and cutting tests with 63 Rc martensitic Wootz or 1075 were found to hold an edge longer when cutting leather than steel at lower hardness. He also reported that if the coarse carbide bands were found within the edge the performance was higher, but in many cases these bands were not found directly in the edge leading to lower performance.

Another interesting test he did was measuring the toughness of Wootz Damascus, which largely has not been studied in the past. He even heat treated the Wootz steel to have a uniform carbide structure rather than the banded structure to see what the effect of this is on the toughness of Wootz. Perhaps unsurprisingly, a uniform carbide structure had superior toughness. A band of brittle particles provides a path for crack formation which then reduces toughness. He also compared with lower carbon steels and predictably a reduction in carbon meant an improvement in toughness. Lower carbon generally means tougher martensite for a constant hardness, and also means less carbide. As noted, carbides are brittle particles so they also reduce toughness. I referenced this Wootz toughness study in my book Knife Engineering in the chapter on Damascus steel.

Toughness of Low Alloy Knife Steels

In 2015 [3] and 2022 [4] Juha published articles on the toughness of a range of low alloy steels. The first study used 1075 and unique compositions with a base carbon content of ~0.75% along with alloy additions of nickel, manganese, vanadium, or aluminum. The more recent study focused on 80CrV2. Both of these studies used heat treating in a forge using traditional methods, which is not my preferred method for heat treating but that may make the information more relevant to the people typically using these steels in knives. Rather than using impact testing he used bending specimens of fixed dimensions and measured the angle of bending at fracture. Materials with greater ductility can bend further prior to fracture.

As a good example of why I am not the biggest fan of hardening with a torch or a forge (as opposed to a furnace), look no further than the studies that Juha performed. In his 2015 study, his 1070-type steel (labeled 69C) had lower toughness than his other tested steels because the grain size had coarsened somewhat, “despite an exceptional careful hardening.” When aluminum or vanadium were alloyed with the steel the grain size was fine and the toughness was higher. The vanadium leads to a fine distribution of vanadium carbides which pins grain boundaries to keep the grain size small, while aluminum leads to the formation of aluminum nitrides. He also tested two nickel alloyed steels which were lower in hardness than the other steels, which may explain the slight improvement in toughness that was measured in some of the specimens, though with the scatter inherent in testing it is difficult to tell how much ductility may have actually been improved, if at all. Juha did not provide a hypothesis on why the nickel steels were lower in hardness, but it might be explained by either the slightly lower carbon content (perhaps not), or the high nickel leading to retained austenite, which may also explain the potential increase in toughness. Nickel is a strong “austenite stabilizer” and reduces the temperature at which martensite forms, thus it can leave some austenite untransformed, or “retained.” Nickel can improve toughness though other mechanisms, however, and I have an old article on how nickel improves toughness here. The fine grained 80MnV and 82MnAl are not shown in the same chart as the 67Ni and 65NiMn steels but if we look only at 60 Rc, the 80MnV is around 20 degrees while the Ni-alloyed steels each have one value at 20 but the rest at 30 degrees, which is why I say the nickel steels may have shown better ductility but there was some scatter. But the study has perhaps definitively showed that steels with grain-pinning elements like vanadium may be inherently better for uncontrolled heat treating in a forge because they can be overheated without grain growth (within reason). Another thing that should perhaps be pointed out is that no amount of cycling prior to the final austenitize would have fixed this issue; overheating prior to quenching will wipe out a fine grain size previously developed.

I found his recent study on 80CrV2 interesting because I recently completed my own extensive study on heat treating of 80CrV2, though our tests were performed independently. One of the things I looked at in my study of 80CrV2 was the effect of prior microstructure, and one of the comparisons I made was between the coarse spheroidized material from the manufacturer, fine spheroidized from my own recommended treatments, and a pearlite starting microstructure. I had previously found a pearlite microstructure to be easier to austenitize when using a forge and explained the reasons why, and the process, in two articles/videos: How to Thermal Cycle Knife Steel and How to Heat Treat Knife Steel in a Forge. Juha found something similar with his experiments on 80CrV2: “80CrV2 steel plate was received from the factory in a soft annealed condition. The soft annealed initial microstructure is not good for forge hardening because large spheroidized carbides need too long soaking time. But anyway, it was tested, and as expected, the reaction to heating was slow and hardening was difficult to execute. Further hardening tests were made with normalized initial microstructure (pearlite).” I also found in my study that the relatively small chromium addition (0.5%) in 80CrV2 makes the spheoridized structure much slower to dissolve, and simple carbon steels like 1084 can be forge heat treated even from the manufacturer condition. Below I have simple schematics showing that with spheroidized carbides (black circles), there is a much greater distance that carbon has to move than pearlite (black lines). Because eventually this carbon needs to be evenly distributed throughout the steel before quenching.

Spheroidized Carbides

Pearlite

Juha found the vanadium addition to help prevent grain growth in 80CrV2 just like his previous study. To test this he let the steel heat up in the forge past nonmagnetic for different lengths of time. The grain size remained “superfine” when heating for 4-6 seconds past nonmagnetic, with an increase in grain size visible after 8 seconds, and a significant increase after 16 seconds. Juha says that 4-6 seconds was “a small but noticeable increase in the brightness of heat color…the disappearance of magnetism plus a shade brighter heat color.” This is typical advice for forge hardening of knife steels, a “shade brighter,” though that can be difficult to achieve without experience. So as long as the steel wasn’t overheated, Juha observed an increase in ductility (bending angle) with decreasing hardness from higher tempering. He also found that whether the steel was hardened from “fine” pearlite, “very fine” pearlite, or martensite the final properties were relatively similar (as opposed to starting with the manufacturer-supplied coarse spheroidized microstructure).

With the simple carbon 1075-type steel, however, an increase in grain size was observed with only a one second hold past nonmagnetic. Again this is why I don’t generally recommend forge heat treating as controlling the proper temperature is very difficult. He did a comparison by heating the C75 for four seconds past nonmagnetic to “one shade brighter” and compared the toughness with fine grained specimens quenched from nonmagnetic. The increase in grain size also led to a significant drop in toughness. At 58 Rc the samples bent 30-40 degrees with the fine grain size (ASTM 10), while the coarser specimens were only 5-10 degrees. This great decrease in toughness occurred after heat treating only one shade brighter than nonmagnetic, which again shows how difficult it can be to perform heat treatments in a forge with simple carbon steels that do not have grain pinning elements added.

Juha wrote in his conclusion: “Some bladesmiths have started to use 80CrV2 instead of unalloyed steels. The current work compares 80CrV2 with C75 (1075) and reveals forge hardening advice for 80CrV2. The studied C75 had no Al deoxidation. Despite that, careful temperature control prevented grain growth. However, it was difficult and not necessarily repeatable in real-life bladesmithing. A smith who uses unalloyed carbon steel without Al has a high risk for grain growth and unsuccessful hardening results. Forge hardening of 80CrV2 is easy because V prevents grain growth, so bladesmiths can repeatedly get good hardening results with it.” I agree with Juha that the vanadium makes it much easier to heat treat without good temperature control, though I might emphasize that the chromium can make it more challenging without the right prior microstructure, and perhaps the best steel would be a 1075/1084 with vanadium, though such steel is not regularly available as far as I know. W2 has vanadium but it also has a bit higher carbon which makes brittle “plate martensite” more difficult to avoid, and low manganese which makes hardening in oil more difficult.

Summary

This was a fun article since I don’t often get to write about the studies of a single individual. I think there is a lot of interesting stuff in what Juha has published, and more in what he hasn’t published. Sometimes people ask me if there are rivalries between different metallurgists or knife testers, but obviously if we are all mature and supportive it is much better to appreciate the contributions of everyone and enjoy learning together. There is some particularly interesting stuff from Juha on Wootz and forge hardening of blades and there is some very instructive stuff in there.

[1] Perttula, Juha. “Reproduced wootz Damascus steel.” Scandinavian journal of metallurgy 30, no. 2 (2001): 65-68. [2] Perttula, Juha. “Wootz Damascus steel of ancient orient.” Scandinavian journal of metallurgy 33, no. 2 (2004): 92-97. [3] Perttula, Juha. “Effect of Ni, Mn, V, and Al on Toughness of Blade Steels.” ISIJ International 55, no. 10 (2015): 2225-2228. [4] Perttula, Juha. “Hardening of 80CrV2 in Bladesmith Forge.” ISIJ International 62, no. 11 (2022): 2397-2401.