Which Knife Steels Have the Best Corrosion Resistance?

Thanks to John Scarborough and Robert Angerer for becoming Knife Steel Nerds Patreon supporters!

Update 1/6/2020: I have since performed my own corrosion resistance testing and provided updated analysis.

Corrosion Resistance of Steel

I have written about steel corrosion resistance in the past, like the article on adding chromium to D2 to make it a stainless which also talked about the effect of heat treatment on corrosion resistance. In the most basic terms, if we start with iron or a very low carbon steel and add chromium to it, the resistance to corrosion improves with increasing chromium [1]:

Corrosion resistance comes from the chromium forming a “passive film” of chromium oxide that prevents further corrosion. Without the passive film, rust forms which tends to flake off and corrosion of the steel continues. More chromium means that the passive film is more complete on the steel surface to better prevent corrosion. There is no fixed cutoff where a certain amount of chromium is enough to prevent corrosion, it depends on the environment. Sometimes 11% or 12% is given as the point where a steel is called “stainless” though there are many stainless steels with more chromium than that.

This image from [2]

Carbon Content and Heat Treatment

Another factor that affects corrosion resistance is the carbon content of the steel. Carbon and chromium in steel tends to lead to the formation of chromium carbides, and when chromium is tied up in carbides it is not available to form the passive film to prevent corrosion. As covered in the D2 article I linked to above, with enough carbon, even a 12% chromium steel is no longer stainless because insufficient chromium is available “in solution.” Even with stainless steels, in the non-heat treated condition where as much carbon is tied up in carbides as possible, the steels are not very corrosion resistant. During the high temperature austenitizing/hardening process, chromium carbides are dissolved allowing carbon and chromium to go into solution to give the steel higher hardness and corrosion resistance. Higher temperatures means more carbide is dissolved and corrosion resistance can be further increased, read the D2 article to learn more. In the plot below, I have calculated chromium in solution for different austenitizing temperatures with JMatPro, a thermodynamic software package, where you can see that even if the steel is heated to just under the melting temperature the steel does not reach the arbitrary 11% chromium for being stainless:

Pitting Resistance and Molybdenum

The above discussion relates to general corrosion of steel. Another type of corrosion is called pitting, where localized corrosion occurs leading to a small cross-section “pit” that can be deep and difficult to remove with surface-level re-finishing. It has been found that elements other than chromium can help to prevent pitting corrosion, notably molybdenum and nitrogen. Studies of resistance to pitting has led to the development of the “Pitting Resistance Equivalent Number” or PREN:

PREN = Cr + 3.3Mo + 16N

As you can tell, molybdenum is shown to have a stronger affect than chromium, with 1% Mo being equivalent to 3.3% Cr. This contribution of molybdenum to corrosion resistance has been used in the design of several knife steels. I wrote about this in the article on the development of 154CM, a steel with 14% Cr and 4% Mo. The Mo addition was not made to 154CM for corrosion resistance, but the recognition of good corrosion resistance in 154CM with relatively low Cr in solution was used to develop other steels designed to take advantage of Mo, such as S30V and S110V. However, PREN was designed for one area only, pitting resistance, which is usually tested in salt water solutions, and it is primarily in chloride that Mo has the greatest contribution. Also, the PREN equation was developed for austenitic stainless steels that already have a high chromium content, so it is not clear if the contribution of Mo is still so strong when the chromium in solution is lower. Does the Mo only improve corrosion resistance if the chromium oxide passive film is already very complete? Or can it replace chromium to some extent? I have not been able to find any studies that looked specifically at high carbon stainless steels and how much Mo contributes to corrosion resistance from a quantitative standpoint. Therefore, I have taken what data is available and performed an analysis myself below.

Thermodynamic Software

Just like in the D2 graph above, chromium and molybdenum in solution can be calculated with thermodynamic software. These estimates are not perfect. First of all, thermodynamic software estimates alloy in solution after an infinite hold time, so the estimates are generally going to be a little higher than is determined experimentally. Also, it does not predict composition fluctuations or segregation of elements, and corrosion can be localized to areas of steel that may be low in chromium. One example is when chromium carbides form on grain boundaries during slow cooling, leading to a depletion in chromium. This is known as sensitization. And more fundamentally, the software is performing a simulation, and any simulation has the possibility of being slightly (or greatly) different than would be experimentally measured for a variety of reasons. However, in general, thermodynamic software is quite good. Companies like Crucible and Uddeholm have used it extensively for designing their steels which would not be possible if it wasn’t at least approximately correct.

Corrosion Experiments

So it is possible to calculate the chromium and molybdenum in solution, but I still needed an experimental dataset of corrosion tests to correlate the alloy in solution with corrosion resistance. I found two good datasets, both in Crucible steel patents. In the 420V/S90V patent [3], they looked at many compositions to decide what the final composition of S90V would be. This included high Mo versions and high nitrogen (N) versions, which provides a good test of the PREN equation. They used two different austenitizing temperatures for all of them which also provides an indication of the effect of heat treatment. They also compared with many other commercial steels including K190, Elmax, S60V, a modified S60V, M390, MPL-1, and 440B. They tested each with two different corrosion tests, one in boiling Aqua-Regia, and one in 10% Acetic acid. Another good dataset was in the S110V patent [4], where they measured the pitting potential in a 1% salt solution. We would expect that test to compare the most directly to the PREN equation. They tested S110V, 440C, S90V, Elmax, S30V, X235, M390, and MPL-1. So for all three corrosion tests, I calculated the Cr, Mo, and N in solution for every steel using the austenitizing temperature indicated in the patent, and then compared that against the measure corrosion resistance. I then calculated the contribution of each element relative to Cr:

The contribution of Mo was calculated to be much less than the PREN equation (3.3) with the aqua-regia and acetic acid tests, though the pitting test was not far off at 2.6. This makes me wonder if the pitting test is a good test of general corrosion or just one specific test. After all, a steel like 154CM has good corrosion resistance, but not excellent corrosion resistance. Also interestingly, nitrogen was found to have a negative contribution to corrosion resistance in two of the datasets. However, even with the high nitrogen versions of S90V that were tested, with as high as 0.46% nitrogen, the thermodynamic software estimated only 0.011% nitrogen in solution. Those high nitrogen versions had excellent corrosion resistance, but it was because the nitrogen addition allowed more chromium to be in solution, not because the nitrogen itself improved corrosion resistance. You can read more about the contribution of nitrogen in the article on nitrogen-alloyed knife steels. Therefore I’m not sure if the nitrogen actually affected anything, or if the contribution to corrosion resistance was so close to zero that statistically it just happened to end up negative.

Rating the Steels

Using the same JMatPro thermodynamic software, I have calculated the Cr, Mo, and tungsten (W) contents in several different knife steels. Because the nitrogen in solution was generally very small, and the nitrogen content in different steels is often not available, I have not included it in the calculations below. Experimental values for alloy in solution (as opposed to Thermodynamic calculations) would be better, of course, but such values are not widely available, and only for a few knife steels.

Some versions of the PREN equation indicate that W has half the contribution of Mo. So even though I did not directly calculate a value for W, I used a coefficient for W that is half of Mo for each equation.

Each steel has recommended austenitizing temperatures in its datasheet; usually a range is given. I used the mid-point of the austenitizing temperature range for each steel for the calculation. Higher austenitizing temperatures could potentially be used to somewhat increase Cr and Mo in solution and therefore improve the corrosion resistance. The exception is H1 which does not have available heat treatment information but it has all of its alloy in solution at relatively low temperatures.

Another assumption I made is that none of the steels were given a high temperature tempering treatment (>750°F). These tempering treatments are very common with tool steels and high speed steels, but tempering that way reduces the Cr and Mo in solution. This assumption is somewhat unrealistic with high speed steels, but I don’t have a good way of estimating the effect of a high temperature tempering treatment. However, if the knife maker/manufacturer wants to optimize corrosion resistance they should use a low tempering temperature instead. Here is a chart showing how corrosion is increased with high tempering temperatures [5]:

Ranking the Steels

To estimate corrosion resistance I used the following equations which were derived as described above, giving a “chromium equivalent” based on chromium, molybdenum, and tungsten in solution:

Acetic acid = Cr + 0.45(Mo + 0.5W)

Aqua-Regia = Cr + 0.79(Mo + 0.5W)

Pitting potential = Cr + 2.6(Mo+0.5W)

Many steels are known under multiple names, here are a few you might be looking for:

Cronidur 30 is also sold as LC200N and Z-finit

M390 = 20CV, 204P

AEB-L = 13C26

154CM is the same composition as ATS-34 and CPM-154

10V = A11

CruWear = Z-Wear, PD1

Vanadis 4E = 4V

Pitting Potential vs Acid Corrosion

The rankings for acetic acid and aqua-regia are similar, as their Mo coefficients are close (0.45 and 0.79), but the pitting potential equation has a much stronger contribution of molybdenum (2.6). Which equation is more realistic is hard to say, perhaps in salt water the pitting potential would be a better predictor and with general corrosion the two acids would be better. One thing that makes me think the equations derived from the acid experiments are more realistic is the fact that M2 and M4 have a higher Cr-equivalent than several stainless steels with the pitting equation because of their high Mo and W. This does not line up with experience. Using the Cr-equivalents from the two acids, however, gives an estimate of corrosion resistance which seems more realistic.

I scoured the internet for pitting potential tests on non-stainless steels and came up nearly empty. In general, if corrosion is a concern then stainless steel is used, so corrosion tests are rarely performed on non-stainless steel. Then with the rare pitting potential tests I did find they generally used something other than 1% NaCl as was used in the Crucible patent. The closest I found for anything was a test on M2 where they used a 0.58% NaCl solution. They measured a pitting potential of -320 mV, which is similar to the lowest measured by Crucible, 440C, with -140 mV with a low temper and also -320 mV with a high temper (lower voltage is worse). The M2 measured in the study was also heat treated with a high temper, being a high speed steel. Because of the difference in NaCl concentration I’m not sure how closely the numbers can be compared. However, the lower concentration of NaCl with M2 leading to the same voltage tells me the M2 has lower corrosion resistance, despite its higher Pitting Cr-equivalent number (14.7 vs 12.8). This makes me think that the Molybdenum contribution is not as effective with low chromium in solution. Perhaps because the Mo supports the chromium oxide film but cannot outright replace it with its own passive film [6].

Because of the analysis above I think the Cr-equivalent equations derived from the two acid tests are a better indicator of resistance to general corrosion. It may be that when a minimum chromium in solution is used (such as 10% as suggested in the S110V patent), that the pitting Cr-equivalent is valid, particularly in salt water. However, more tests with non-stainless steels would be necessary to confirm if this is the case. The tests in the two acids, however, did include a non-stainless steel so the equations derived included conditions with low Cr in solution. For these reasons I will primarily be discussing the Cr-equivalent predictions from the two acid tests below.

Corrosion Resistance Predictions of Individual Steels

The highest steel on the chart is Vanax, a nitrogen-alloyed steel with high bulk chromium that forms very little in chromium carbides or nitrides. Interestingly, the Cr in solution was calculated somewhat lower in the article on nitrogen-alloyed steels. In that article I used ThermoCalc rather than JMatPro, and apparently the two software packages do not calculate nitrogen-alloyed steels identically. In fact, the top five steels were all alloyed with nitrogen. However, nitrogen does not automatically lead to high corrosion resistance, as Nitro-V is somewhat middling in the chart.

Another steel which does not get its due is S110V, which despite being designed as a corrosion resistance upgrade over S90V, is often not recognized for superior corrosion resistance. At the bottom of the chart is carbon steels, alloy steels, and tool steels, which is expected due to their low bulk chromium content.

The best tool steels for corrosion resistance include A8 Mod, CruWear, and 3V, which are ~8% Cr steels. They have a higher Cr-equivalent than D2 and K190 despite those steels having ~12% Cr, due to carbide formation.

The lowest stainless steels on the list are XHP and ZDP-189, both of which do not have a Cr-equivalent of even 11% with the two acid tests, and XHP having just 11.1% Cr-equivalent with the pitting equation. In fact, there are four tool steels with a higher predicted Cr-equivalent than ZDP-189. This at first glance may make one think that the thermodynamic software is simply not calculating the Cr in solution correctly. However, I have a couple reasons to believe that these values are at least qualitatively correct. First of all, ZDP-189 achieves very high hardness levels, which usually means that less chromium is in solution, which I have written about in the 154CM article and the nitrogen-alloyed steel article. Secondly, there have also been reports about the poor corrosion resistance of ZDP-189 [7]. It has always been somewhat puzzling to me how ZDP-189 can achieve such high hardness levels while still being stainless. These calculations may provide a clue as to how it does so; it isn’t actually a stainless steel. I purchased a small ZDP-189 knife and hope to experimentally measure the Cr in solution in the future. I will have to destroy the knife but it will be worth it to check.

Conclusions

Chromium improves corrosion resistance of steel by forming a passive film that prevents further corrosion. Molybdenum and tungsten can improve corrosion resistance, to a lesser degree than chromium in tests with acids, but to a great extent in pitting corrosion of stainless steels.

The bulk chromium content of the steel cannot necessarily be correlated with corrosion resistance because the chromium may not be available in solution to contribute to the passive film. Heat treatment can, to some extent, put more chromium in solution to contribute to corrosion resistance. Thermodynamic software or experimental measurements can be used to determine the Mo, W, and Cr in solution, though software is a much faster way of doing so. In general, experimental values are not available.

Equations for a “Cr-equivalent” in solution were developed from corrosion tests in acetic acid, aqua-regia, and pitting potential in salt water. The relative effect of Mo estimated from the tests with acetic acid and aqua-regia make more sense intuitively than the 2.6 times Mo of the pitting equation, particularly with non-stainless steels. I probably wouldn’t look at the Cr-equivalent of the pitting equation unless the steel has at least 9 or 10% Cr in solution in the chart.

Nitrogen-alloyed stainless steels tend to have the highest corrosion resistance, though some steels not alloyed with nitrogen like S110V and 440A are also high on the chart. Some stainless steels have a significantly higher Cr-equivalent than others, with ZDP-189 and XHP being the worst according to these calculations. Some tool steels rank relatively high like 3V and CruWear despite not being stainless steels. This information can be used to select steels based on corrosion resistance for applications where that is important. The Cr-equivalent equations can also be used in knife steel design to target different levels of corrosion resistance.

---

[1] http://emrtk.uni-miskolc.hu/projektek/adveng/home/kurzus/korsz_anyagtech/1_konzultacio_elemei/stainless_steel_case_study.htm

[2] https://sassda.co.za/about-stainless/introduction-to-stainless-steel/

[3] Pinnow, Kenneth E., William Stasko, and John Hauser. “Corrosion resistant, high vanadium, powder metallurgy tool steel articles with improved metal to metal wear resistance and a method for producing the same.” U.S. Patent 5,936,169, issued August 10, 1999.

[4] Kajinic, Alojz, Andrzej Wojcieszynski, and Maria Sawford. “Corrosion and wear resistant alloy.” U.S. Patent Application 11/124,350, filed November 9, 2006.

[5] http://www.crucible.com/PDFs/DataSheets2010/Datasheet%20CPM%20S110Vv12010.pdf

[6] Hashimoto, K., K. Asami, and K. Teramoto. “An X-ray photo-electron spectroscopic study on the role of molybdenum in increasing the corrosion resistance of ferritic stainless steels in HC1.” Corrosion Science 19, no. 1 (1979): 3-14.

[7] https://www.bladeforums.com/threads/zdp-189-corrosion-resistance-compared.992801/#post-11297843