Toughness Improvement of High Carbon Tungsten Steel 1.2562

Thanks to all the new Knife Steel Nerds Patreon supporters that have joined since our big article on CATRA testing. The response has been phenomenal. We only need 4 more supporters to reach 200! Brandon Percival, Jason Stone, Fixall, John Schmidt, Adam Ringgenberg, Adam Webb, Shah Paul, Andy, Chad Shoopman, Dylan Giagni, Adam Nolte, John Scarborough, Datafox, Ben Secrist, Noah, Josh – REK, Robert Towe, NJO, Dd, Andrew Bounds, Nathan Gee, and hsc3.

1.2562 Steel

1.2562 along with its American counterpart F2 are probably the highest wear resistance low alloy steels. The steel gains its wear resistance due to the higher carbon for higher content of cementite (iron carbides) but especially the tungsten addition which leads to tungsten carbides (WC) which are high hardness carbides that contribute to wear resistance. I wrote about tungsten-alloyed low alloy steels in this article which covered other steels like Blue Super, 1.2519, and O7. The low overall alloy content of 1.2562 makes it well-suited for forging, you can read more about what makes a steel easier or more difficult to forge in this article.

One important thing to note is that high speed steels with ~4% Cr do not form the same WC carbide as low alloy steels even when they have as much as 18% W. Instead the high speed steels form a W₆C carbide which is lower in hardness and does not contribute as much to wear resistance. You can read about these alloying element interactions for different carbide types in this article.

1.2562 Toughness

I previously reported a toughness value for the low alloy tungsten steel 1.2562 which scored relatively low. I wrote about the toughness of this steel in the context of other low alloy steels typically used by forging bladesmiths. You can read that article here. I concluded that 1.2562 had low toughness due to “plate martensite” which is difficult to avoid with high carbon low alloy steels, as well as poor carbide structure which included both carbides along grain boundaries and some larger tungsten carbides.

1.2562 area 1 – many carbides visible along grain boundaries

1.2562 area 2 – a large tungsten carbide

My toughness measurements were consistent with earlier reported values for F2 steel, which was found to have relatively poor toughness:

Data from [1]

New 1.2562 Toughness Measurements

Knifemaker Marco Guldimann contacted me after my toughness measurements of 1.2562 and said he wanted to show that the toughness could be improved. I said I would be interested in seeing the results of whatever experiments he wanted to perform and recommended that he use the same toughness specimen dimensions so that his results could be directly compared to ours. We use a subsize unnotched charpy specimen which is 2.5 x 10 x 55 mm, the exact specifications can be seen on this page. All of the experiments, toughness testing, microscopy, hardness testing, etc. were performed by Marco and a metallurgy technician. I did not perform any experiments or evaluate any specimens. These are the first toughness results I am reporting on that were not tested by myself.

Processing of Steel

Marco wrote about the processing of his steel and how it was tested in a report; you can read the full report here. Marco wanted to have as much reduction in forging as possible, and this steel does not have particularly great availability and so thickness options are limited. Instead, he forge welded several pieces together, forged them out, cut and restacked, and forge welded and drew out to a final thickness of 3.3 mm. All forge welding and forging was performed between 900 and 750°C (1650-1380°F). These are relatively low temperatures for forging; Marco chose these temperatures because his primary goal was achieving a fine grain size.

Marco next performed a “temper anneal” by austenitizing at 770°C (1420°F), quenching, then tempering at 690°C (1275°F). The specimens were then given one of 5 heat treatments and tested for toughness and hardness. He heated the samples in a furnace and quenched in molten salt at the tempering temperature of each condition, followed by air cooling. Then each steel was tempered at the indicated temperature.

And in Fahrenheit instead:

Carbide Structure

No grain boundary carbide was observed in the micrographs shared by Marco, showing that the “thermomechanical processing” (temperature plus forging) led to the elimination of those carbide networks. There were still regions of larger carbides, especially when found in bands in the steel. Perhaps with a higher initial forging temperature those larger carbides could be eliminated, as Thermo-Calc predicts that the tungsten carbides would be dissolved at about 2075°F.

Area with larger carbides in a sample of heat treatment 2

Area with more even carbide structure in a sample of heat treatment 3

Grain Structure

Revealing grain boundaries in metallographic specimens is very difficult, often requiring specialized etchants and even then revealing them completely is challenging. To demonstrate grain refinement, Marco and the technician helping him pointed out a few visible grains. This is not a good way to measure an average grain size, and no comparison was made to an unprocessed specimen but the visible grains are relatively small, on the order of several microns, and Marco claims an ASTM grain size between 11.5 and 13.

1.2562 steel with a few visible grains

Hardness 

The hardness of each specimen was measured with Vickers and with Ultrasonic Contact Impedence (UCI). I will show both the Vickers value and a conversion to Rockwell C. Conversions between hardness scales is never perfect but gets into the right ballpark. I used two different conversion scales for hardness which are shown as HRC-1 [2] and HRC-2 [3]. The HRC-1 value comes from a hardness conversion from Professor Paul Beiss who created it by using with a large database of values taken with high hardness high speed steels. The HRC-2 value comes from the standard ASTM conversion which agrees better with Marco’s UCI hardness measurements. Three toughness specimens were tested and averaged for each heat treatment which is also shown.

The hardening response was surprisingly lower than in the specimens that I heat treated, such as heat treatment 4 which used 1475°F and a 300°F temper which resulted in about 60.6-62.3 Rc where instead I got over 66 Rc. Higher hardness values were instead seen when austenitizing at about 1500°F. 1500°F was where I was seeing a drop in hardness, presumably due to excess retained austenite. I used hold times of 10 minutes and Marco used 12 minutes, and neither of us used cryo in heat treating.

There are several reasons why the optimal austenitizing temperatures could be shifted higher, such as furnaces that are reading different temperatures, decarburization, different starting structure, or grain refinement. I don’t think that our furnaces would be so far off from each other though that can’t necessarily be ruled out. Decarburization would presumably show up in the micrographs and they did not report any. Different starting structures resulting from different processing and annealing can show up in the final heat treatment. For example, a martensite structure leads to rapid response in heat treatment so lower temperature or shorter hold times are possible. Pearlite structures are a bit slower, and spheroidized carbides can be similar to pearlite or slower depending on how coarse the carbides are. You can read more about the effect of these starting structures here. However, the temper anneal performed should lead to a relatively fine starting structure so I wouldn’t expect a major difference in diffusion time required. Grain refinement can lead to poorer hardening response, however, sometimes requiring higher austenitizing temperatures. Sometimes this is to get enough grain growth to have sufficient hardenability, as 1.2562 is already a low hardenability steel and with some grain refinement it would be even more difficult to harden. It could be that the molten salts Marco quenched in were somewhat slower than the Parks 50 which would be even more significant after grain refinement. Molten salts are typically faster than oil at higher temperatures and slower as the temperature decreases.

Toughness

The best heat treatment, #2, in particular looks much better than my toughness measurement, resulting in values that look close to what was achieved with CruForgeV, 1.2519, 1.2442, and V-Toku 2. I have noted both HRC-1 and HRC-2 values because of the uncertainty with hardness conversion.

To separate out the different effects that resulted from the forging process (grain refinement and elimination of boundary carbides), I performed some toughness experiments after cycling steel in a way designed to eliminate the grain boundary carbides. I heated the steel to 2100°F for 1 hr and air cooled (to dissolve all carbides and reform them), then 1700°F for 15 minutes followed by air cooling (to normalize), then annealed by heating to 1435°F for 30 minutes followed by furnace cooling at 665°F/hr, and austenitizing at 1450°F for 10 minutes, quenched in Parks 50 oil, and tempered at 400°F. The measured toughness was still low at 2.8 ft-lbs. I have not taken any micrographs to ensure the carbide structure was improved, but this likely shows that the carbide structure is not the limiting factor for toughness, but rather plate martensite.

Grain refinement does not necessarily lead to large improvements in toughness, as can be seen in toughness studies done on CruForgeV and AEB-L. The CruForgeV study, for example, included comparisons of forging temperatures between 1550 and 2000°F and also multiple quenches in an attempt to refine the grain size and differences between different processing were small. However, with sufficient grain refinement plate martensite can be avoided, which is likely the reason for such poor toughness with 1.2562. The very high carbon in 1.2562 and low chromium makes avoiding plate martensite challenging.  With CruForgeV we apparently avoided plate martensite with the major exception being the conditions which were austenitized at 1550°F which had very low toughness, similar to the values we obtained with the original specimens of 1.2562. Some steels are more or less prone to plate martensite which I wrote about in the previously referenced article about low alloy steel toughness. It could also be that the marquenching that Marco performed affected the formation of plate martensite for improved toughness.

Summary and Conclusions

Marco achieved a very excellent improvement in toughness of 1.2562, despite the very high carbon content of the steel and the difficulty in avoiding brittle plate martensite. He also successfully eliminated the undesirable grain boundary carbide networks through the forging process. This improvement was probably due to grain refinement which led to the reduction in plate martensite, but marquenching may have also been a factor. This shows the desirability of grain refinement in very high carbon, low alloy steels to avoid plate martensite and greatly improve toughness. This allows forging bladesmiths to achieve good toughness in combination with improved wear resistance with these types of steels.

---

[1] Roberts, G.A., et al. Tool steels. American Society for Metals, 1962.

[2] Samal, P., and J. Newkirk. “Properties of Powder Metallurgy Tool Steels.” (2015).

[3] Standard, A. S. T. M. “E140-12be1, Standard Hardness Conversion Tables for Metals Relationship among Brinell Hardness, Vickers Hardness, Rockwell Hardness, Superficial Hardness, Knoop Hardness, Scleroscope Hardness, and Leeb Hardness.” West Conshohecken, PA: ASTM International (2012).