Thanks to Larry Elletson, Tim Sigwarth, Steve Matthes, Curtis Seizert, Kristian Gasparius, and Will Collins for becoming Knife Steel Nerds Patreon supporters! We now have over 150 supporters!
History of VG10
Both VG10 and SG2 were developed by Takefu Special Steel, a company started in 1954 and known for their laminated steel produced for knives [1]. They have a range of “V-series” steels like V1, V2, V-Toku1, V-Toku2, V-Gin1, and the “V Gold” series which includes VG10 [2]. The website states that VG10 was developed for knives “over 60 years ago” [3] and with the current website having been updated in April 2019 that would put the development of VG10 at 1959 or perhaps a couple years earlier. That means the steel would have been developed in a similar time to 154CM. The design of VG10 includes 1% Mo and 0.2% V which are common additions to older steels like A2 and D2. The vanadium helps to refine the grain size and the molybdenum addition helps to improve hardenability so that it can be air cooled and reach its maximum hardness. The oddest part of VG10 is the cobalt addition; I explored the reasons for its addition in this article. The cobalt was likely added to increase tempering resistance so that VG10 can be used with coatings that must be applied at high temperatures. Cobalt additions to steels for this purpose were not unheard of but VG10 is the first stainless steel I know of with a cobalt addition for tempering resistance.
(Note: Sometime during the 1940’s it became common to add 0.4-0.5% Mo to 440C)
History of Super Gold 2
Super Gold 2 (SG2) is another grade designed for knives by Takefu, the patent was filed in 1991 [4]. They don’t appear to make Super Gold 1 anymore though it could be one of the other grades in the patent. Super Gold 2 is produced with powder metallurgy to have a fine microstructure and good toughness. 1991 is relatively early for powder metallurgy stainless steel, in general. S60V and M390 already existed though M390 had not yet been used in knives at that point. And SG2 predated steels like S90V and S30V. S30V, in particular, got a lot of press for being designed specifically for knives, but SG2 predated it by approximately a decade and according to the patent was developed for cutlery. With its combination of relatively low Cr with high Mo and V it looks somewhat similar to S30V as well, at least superficially. Here is a comparison with a few other stainless steels and the approximate date each was released.
Super Gold 2 and S30V both have lower chromium (14-15%) in combination with high Mo (2-3%). Molybdenum helps to improve corrosion resistance and therefore can make up for somewhat reduced chromium. The first knife steel I know of that combined high Mo with reduced Cr was 154CM, you can read the history of that steel here. The reduced Cr content in SG2, S90V, and S30V helps to reduce the amount of chromium carbides and therefore improve toughness. The vanadium leads to the formation of vanadium carbides and increases the hardness of the chromium carbides, and both of those improve wear resistance and edge retention. You can read more about those carbides in this article.
In the patent, Takefu lists several steels that were analyzed to develop SG2 (1-7), and those were compared to the standard grades D2 and 440C:
The patent does not describe much where these different alloy amounts came from. The 14-15% Cr probably came from their previous VG steels. VG1, VG2, and VG5 all have 14% Cr and VG10 15% Cr. They state in the patent that Mo and W were added to improve hardness, toughness, and corrosion resistance. And the vanadium addition for an improvement in wear resistance without an excessive addition. Cobalt additions were also tried “to strengthen the metal matrix.” The developers then tested the steels for toughness, wear resistance, and corrosion resistance.
Toughness
Toughness was measured with the 3-point bend test. Higher values mean better toughness. The best toughness was found with grade 1; grades 2, 4, and 5 were a small step down. Grade 3 (SG2) was a step below any of those, followed by the relatively poor values of D2 and 440C (6 and 7).
These toughness values correlate with carbon content of the steel, as can be seen in the plot below comparing the developmental steels 1, 2, 3, and 4. This makes sense as higher carbon means both more carbide and higher carbon in solution, both of which negatively affect toughness.
Wear Resistance
The inventors also performed a wear resistance test using the Ohgoshi abrasion tester. The test piece hardness was 60 Rc of SCM435 with a friction distance of 200 meters and a load of 6.3 kg. I’m not familiar with this wear test so I can only give the test parameters.
The wear resistance also correlated with carbon content, as would be expected from the increasing amount of carbide:
Corrosion Resistance
The inventors performed a series of corrosion tests to relatively rank the corrosion resistance of each steel. They tested the mass loss in 10% hydrochloric acid, 15% sulfuric acid, and 3% hydroflouric acid (lower mass loss means less corrosion). They also performed a “natural weather exposure” test in the far right column which was given a rating by circles, triangles, or x. Presumably the double circle is best, followed by circle, triangle, x, and xx. The patent doesn’t explain the rating system. They performed the corrosion test with two tempering temperatures: 200°C and 500°C. The low tempering temperature almost always leads to better corrosion resistance, because the high tempering temperature leads to the precipitation of tiny chromium and molybdenum carbides which reduces corrosion resistance. They are interested in the high tempering range in case the steel is used for coatings or other applications requiring high tempering resistance. Similar to the reasoning behind the cobalt addition to VG10. I don’t recommend the high temper with stainless steels so I won’t be analyzing that data closely.
Corrosion also correlated pretty well with carbon content. More carbon leads to more carbide which reduces chromium “in solution.” The chromium in solution is what contributes to corrosion resistance. Read more about corrosion resistance here.
The correlation isn’t perfect because there are different chromium and molybdenum contents in each steel, and both elements contribute to corrosion resistance. We can use Thermo-Calc software to estimate the amount of chromium in solution and rank corrosion resistance that way across all 7 of the steels:
That gives a pretty decent correlation (R2 of about 0.7) apart from the HCl data which seems pretty scattered. However, this also does not utilize the contribution of Mo. In my previous corrosion resistance experiments, I devised an equation that can predict behavior based on Cr, Mo, W, and N: Cr + 1.6*Mo + 0.8*W + 6*N. The contribution of Mo, W, and N max out at 2.6 as they can support chromium but not wholly replace it. This gives us a good chance to test that equation using an independent dataset:
The R2 was increased above 0.9 for the sulfuric and hydroflouric acids, it even improved the hydrochloric but there is one potential outlier throwing everything off. I am very happy with that result, it is a good indication that my corrosion resistance predictions are working.
Balancing Properties
So as you can see there is a balance in properties based on the carbon contents that the inventors explored. Lower carbon content means better corrosion resistance and toughness but worse wear resistance. It also means lower hardness though maximum hardness was not a discussion point within the patent. The patent specifies a carbon range of 0.9-1.5% carbon so it is somewhat interesting that the only grade that remains in production is SG2 which is at the top end of that range. Perhaps Super Gold 1 was lower in carbon but I haven’t been able to find solid information on the other Super Gold grade. Presumably they ended up with SG2 to favor higher hardness and wear resistance.
Knife Steel Nerds Experiments
I was able to get some VG10 and SG2 from Richard Airey of Barmond Special Steels. I was excited to get some because these steels are often difficult to obtain outside of Japan. But many knife enthusiasts are interested in the steels because of how common they are in Japan-produced knives. So I ran the steels through some of my common experiments.
Microstructure
Super Gold 2 looks a lot like other PM stainless steels with a low vanadium carbide content. In other words, it looks pretty close to Elmax. Perhaps Elmax is a bit finer in its structure. I calculated SG2’s chromium carbide at 12.8% and its vanadium carbide at 0.5%. The vanadium carbide is relatively low because of the interactions with chromium which limit vanadium carbide formation. You can read more here. Using point counting with the micrograph itself I found 16.5% carbide volume, which is a bit higher than what Thermo-Calc estimated but that makes sense as the calculation is for an infinite hold time at temperature where the carbide volume would be lower. You can compare with other steels in the mega micrograph article.
Super Gold 2
Elmax
I expected VG10 to have somewhat less carbide than it has. It has the same ~1% carbon as 19C27 and 440C, but intermediate chromium content of 15%, whereas 19C27 has 13% and 440C has 17%. The more chromium the more carbide is formed. However, VG10 looks closer to 440C than it does 19C27. All three of those steels are contain primarily chromium carbide. VG10 looks to be somewhere in the 12-16% carbide volume range.
VG10
440C
19C27
Toughness
I heat treated the two steels using the approximate standard temperatures recommended on the Takefu website. I used a 1975°F austenitize for VG10 and a 2000°F austenitize for SG2, each for 20 minutes, plate quenched, liquid nitrogen, and then tempered twice for 2 hours at 400°F. The hardness for both ended up very similar, about 60.7 Rc. I did three toughness tests using our standard subsize unnotched charpy specimen. The VG10 ended up averaging 5.8 ft-lbs and the SG2 6.5 ft-lbs. This puts it in line with other high carbide stainless steels. SG2 is a bit higher in toughness than S30V but below S35VN, putting it about in line with S45VN. It also appears to be a bit better than Elmax when compensated for hardness. M390 measured better for toughness but I am still not sure why M390 did as well as it did in this toughness test, more experiments are coming. It is a bit surprising that VG10 and SG2 are so similar for toughness considering the large difference in carbide structure. It may be that the carbide volume itself is more the limiting factor for these steels.
Corrosion Resistance
I did my set of corrosion resistance experiments that I have been performing with a range of stainless steels. First I heat treated the specimens with the same heat treatment used for toughness testing. Then I ground the samples and finished them to 400 grit, which is intended to simulate the finish of a factory-produced knife. I then did my “is it stainless?” test where I sprayed each sample with distilled water every 8 hours for 4 days. Neither sample showed any corrosion. So I consider both steels to be stainless.
I next did the same test with 1% salt water after re-finishing them to 400 grit. And I rated them according to the method laid out in the testing article. Thermo-Calc estimated that VG10 has 11.7% Cr and 0.9% Mo in solution, and SG2 10.4% Cr and 2.4% Mo in solution. VG10 ended up roughly where it was expected to be though SG2 ended up a bit below the trendline of other high Mo grades. Perhaps this was simply the result of re-spraying the samples at a slightly different rate or some other source of variation. Or a somewhat inaccurate estimate of alloy in solution from Thermo-Calc. Regardless, the difference from prediction is relatively minor. This also shows that the cobalt addition to VG10 is not strongly affecting its corrosion resistance, either positively or negatively.
I have a more general 10-point rating scale for each knife steel that is different than the experimental rating given above. The rating for VG10 is 7.9 and SG2 is 7.8. They performed pretty similarly in the 1% saltwater test so I will not be adjusting them. You can see how they rank vs other steels in the chart below:
Edge Retention
I have an experimental result for CATRA slicing edge retention testing of VG10. It is a small notch below 440C which is about in line with where it would be expected based on its volume of chromium carbide:
I do not have an experimental value for edge retention of SG2, but it can be estimated using its carbide content and composition. With its 2% vanadium we would expect it to be a bit below Elmax, S30V, and S35VN, which is also where the estimate puts it. The edge retention was estimated using the method described in this article. SG2 is still higher than VG10, 440C, and CPM-154 in terms of wear resistance, however. Therefore in slicing tasks that lose sharpness due to wear we would expect SG2 to be a bit below S35VN.
Finishing and Sharpening
VG10 has no vanadium carbide and SG2 only a small amount (~0.5%) according to Thermo-Calc. Vanadium carbides are harder than the standard abrasive aluminum oxide which means that polishing steel with significant amounts of vanadium carbide is more difficult, whether for finishing a blade or sharpening to high grit. SG2 has less vanadium carbide than steels like Elmax, S35VN, S30V, and M390, which likely makes finishing it easier than those alternatives. It still is likely not as easy as a steel like CPM-154 which is free of vanadium carbide. VG10 is relatively easy to polish, similar to 440C or 19C27, the only added challenge with VG10 is the occasional large carbides relative to powder metallurgy steels.
Summary andConclusions
Looking at these steels was fun because they have been around for a while but I haven’t learned much about them until the research I did for this article. VG10 has been around much longer than I would have expected, being from 1959 or perhaps even a couple years earlier. And it surprised me that SG2 is a powder metallurgy stainless developed for knives that predated S30V by almost 10 years. In terms of properties neither really stands out as being an amazing performer, though they are in line with expectations. VG10 is a decent Japan-developed alternative to 440C with some added cobalt for tempering resistance. And SG2 is in the same approximate property range as other PM stainless steels like Elmax and S35VN.
[1] http://www.e-tokko.com/profile.php?lang=en
[2] http://www.e-tokko.com/original.php?lang=en
[3] http://www.e-tokko.com/v_gold_10.php?lang=en
[4] “Uniform organization of stainless steel cutlery.” Japanese Patent 2764659, issued June 11, 1998.
