Carbide Volume
As covered in Part 1, carbide volume decreases with increasing temperature. As an example, here are micrographs showing carbides in a spray-form version of the original Vanadis 4 (non-Extra) [1]:
This steel is a good representation of what can occur with more complicated steels versus our simple low-alloy steels illustrated in Part 1. In this case we have two major carbide types: vanadium carbides and chromium carbides . The MC carbides are larger in part because they form at a higher temperature during solidification. Thermodynamically the vanadium carbides are more stable, they form at higher temperatures and they also dissolve at higher temperatures when austenitizing. So at the lowest austenitizing temperature of 850°C, many small chromium carbides are present after quenching, with some of the somewhat larger vanadium carbides as well. With an austenitizing temperature of 1150°C, those vanadium carbides are mostly unchanged but there very few chromium carbides remaining. This is confirmed by thermodynamic calculations for this composition:
| Austenitizing | Cr carbides | V carbides |
| 850°C | 8.5% | 8.3% |
| 950°C | 5.2% | 8.3% |
| 1050°C | 2.7% | 8.2% |
| 1150°C | 0.0% | 7.8% |
All of this is important because we know that carbide size and volume greatly affect toughness, as I summarized in a previous forum post [2].
Grain Size
Another important factor we covered in Part 1 was grain size. Here are images of the prior austenite grain size in Uddeholm Caldie with increasing austenitizing temperature [3]:
Here there is a pretty dramatic difference in grain size with austenitizing temperature, though the 1075°C austenitizing temperature is above what is recommended by the Caldie datasheet [4].
That change in grain size had an obvious impact on toughness. After tempering at 200°C, the hardness between all three was similar but the toughness was quite different:
| Austenitizing | Retained austenite (%) | Grain size (microns) | Hardness (Rc) | Toughness (ft-lbs) |
| 1020°C | 19 | 11 | 60.3 | 156 |
| 1050°C | 23 | 16 | 60 | 131 |
| 1075°C | 28 | 45 | 59.7 | 37 |
Or I put together this chart to illustrate the effect of grain size a little more clearly:
Carbon Content in Martensite
As the carbon content increases with higher austenitizing temperature, that also increases the carbon content of the final martensite, which is why higher austenitizing temperatures lead to higher hardness [5]:
The amount of carbon in martensite also changes the type of martensite that forms; there is a transition from “lath” martensite to “plate” martensite [6]:
This transition from lath to plate martensite is important because plate martensite has poorer ductility, in part due to the formation of microcracks during the transformation of martensite [7]:
Higher carbon martensite has poorer toughness than lower carbon martensite, even when tempered down to the same hardness. Here is an example from a study on 52100, where an increase in austenitizing temperature led to a reduction in toughness even when the grain size and hardness were found to be the same [8]:
| Austenitizing | Matrix C | Carbide volume | Grain size | RA @ 150°C | RA @ 230°C |
| 1500°F | 0.49% | 8% | 16 microns | 8 | 0 |
| 1550°F | 0.63% | 6% | 16 microns | 15 | 0 |
| 1600°F | 0.67% | 5% | 16 microns | 17 | 0 |
| Austenitizing | As-quenched | 135°C | 150°C | 170°C | 195°C | 230°C |
| 1500°F | 63 | 63 | 62 | 61 | 60.5 | 60 |
| 1550°F | 64.5 | 64 | 63 | 62 | 61 | 60 |
| 1600°F | 65 | 64 | 64 | 62.5 | 61.5 | 61 |
Most of the factors are constant between the three selected austenitizing temperatures except for carbon in solution, which is what the authors proposed was the cause for the differences in toughness. When I plotted carbon vs toughness for the 170°C tempering temperature, I got a very convincing trend:
Retained Austenite
As could be seen in the 52100 and Caldie studies above, retained austenite is higher with higher austenitizing temperatures. Stable retained austenite increases toughness [3]. However, there are reasons why high retained austenite may not be good for a knife blade, but that is a topic by itself. If retained austenite needs to be minimized, lower austenitizing temperatures should be used.
Toughness Summary
Grain size is refined and carbon in solution is reduced by using the lower austenitizing temperature range; however, carbide volume is reduced by using higher austenitizing temperatures. How do those factors interact? In general, the effect of carbide volume is less than the effect of grain size and carbon in solution. Perhaps this is due to the largest carbides still being present with high austenitizing temperatures as was shown in the Vanadis 4 images. Therefore, a lower austenitizing temperature usually leads to superior toughness, as can be seen with K390 that was all tempered to the same hardness [9]:
This is assuming the tempering temperature selected is sufficient, of course, but tempering is not the focus of this article. But in general it is better to set the final hardness with the austenitizing temperature than with tempering.
Edge Retention
How does the selection of austenitizing temperature affect edge retention? Does using a lower austenitizing temperature for greater carbide volume lead to superior wear resistance? There was a study conducted using CPM-M4 with CATRA edge retention testing where they compared different heat treatment parameters [10]:
| Austenitizing | Tempering | Hardness | Carbide volume | CATRA |
| 2000 | 925 | 63.5 | 17.4 | 399 |
| 2000 | 1125 | 57.7 | 17.3 | 371 |
| 2100 | 1025 | 63.9 | 19.6 | 401 |
| 2200 | 925 | 66.6 | 14.7 | 440 |
| 2200 | 1125 | 60.8 | 14.2 | 365 |
If we plot hardness vs CATRA result we get a very good correlation:
However austenitizing temperature or carbide volume does not appear to correlate with CATRA. Therefore, for a given steel hardness is the most important factor for edge retention.
Conclusion
In general, lower austenitizing temperature is better for a given hardness in terms of toughness. If heat treating to higher hardness an increase in austenitizing temperature may be required, of course. As summarized in Part 1, higher austenitizing does lead to an improvement in corrosion resistance with stainless steels, however. The selected austenitizing temperature has little impact on edge retention except when it comes to achieving the desired hardness. In part 3 I will cover the various multiple quench treatments and their effect on steel properties.
[1] Yan, Fei, et al. “Microstructure evolution during hot rolling and heat treatment of the spray formed Vanadis 4 cold work steel.” Materials Characterization 59.8 (2008): 1007-1014.
[2] https://www.bladeforums.com/threads/predicting-toughness-with-steel-composition.1534942/
[3] Rehan, Muhammad Arbab, et al. “Effect of Austenitization and Tempering on the Microstructure and Mechanical Properties of a 5 wt% Cr Cold Work Tool Steel.” steel research international87.12 (2016): 1609-1618.
[4] https://www.uddeholm.com/files/PB_caldie_english.pdf
[5] Krauss, George. “Heat treated martensitic steels: microstructural systems for advanced manufacture.” ISIJ international 35.4 (1995): 349-359.
[6] Krauss, G., and A. R. Marder. “The morphology of martensite in iron alloys.” Metallurgical Transactions 2.9 (1971): 2343.
[7] Krauss, G., and W. Pitsch. “Deformation twins in martensite.” Acta Metallurgica 12.2 (1964): 278-279.
[8] Santiago, Rescalvo, and Jose Antonio. Fracture and fatigue crack growth in 52100, M-50 and 18-4-1 bearing steels. Diss. Massachusetts Institute of Technology, 1979.
[9] Schemmel, I., et al. “Böhler K390 Microclean–A new powder metallurgy cold work tool steel for highly demanding applications.” Proc PM2004, Oct: 17-21.
[10] Lian, Sidi. Heat treatment effects on CPM-M4 tool steel performance as edged blade material. Diss. 2014.
Hello.
Nice and comprehensive article. You certainly went though all the variables I can think of. Quite a bit of information to chew and digest. Thank you!
But something caught my eye and left me with a lot of questions. It’s a bit off topic, but still: In the 52100 study you quoted [8], there is a sharp drop off in toughness at 230°C tempering, after a relatively stable plateau for 135°C-195°C. I can’t tell if this is because of the retained austenite vanishing (there are only 2 data points for RA in your article – are there more in the quoted study?). Or is it already tempered martensite embrittlement? Or are these 2 phenomena inextricably linked anyway?
And the practical question is: how do you temper for getting a sweet spot of hardness and “usable” toughness? I guess I’m keeping my eye out for the future tempering article(s) discussing toughness.
Thanks again for the incredibly informative article.
You can read the thesis here: https://dspace.mit.edu/bitstream/handle/1721.1/70610/07354787-MIT.pdf?sequence=2
You are correct that tempered martensite embrittlement occurs from both retain austenite transformation and also cementite precipitation and it can be difficult to separate them without cryo testing. In the thesis they attribute the loss of toughness to loss of retained austenite but say that it is puzzling.
Hopefully I can write more tempering articles soon. There is a long list.
Thanks for the link. 200 delicious pages of PhD thesis…
Found the data I was looking for. And yes, at first glance the retained austenite is relatively stable up until 195°C, and vanishes at 230°C. Good correlation.
But then I would expect the samples with the most RA to also loose the most toughness. and… it is not the case! The 815°C austenitized sample lost the most toughness, while having by far the least RA.
So the phenomenon seems to be a bit more complicated then just some extra toughness linearly dependent to the amount of RA.
Anyway, thanks again!
Br, Dorin
Hi Larrin,
thank you for sharing your article!
In the first picture, about the carbide volume, i noticed a lot of voids in the matrix. It might be related to the spray technology, but made me wonder if in general when big carbides dissolve they might leave voids in the matrix (if we don’t forge).
Maybe the polishing technique might not always show this in the micrography of the samples?
Do you know if the matrix is able to “fill” the spaces formerly occupied by the carbides? We usually don’t bother too much because we assume that omogeneization would take care of carbide size and distribution, but might we be missing something? We are concerned about microcracks in plate martensite, but if we start with big carbides we might end with micro voids acting as crack initiation sites, or am i out of track?
Don’t worry Stefano, with “normal” steel dissolving carbides does not leave voids behind.
🙂 it makes sense!
thank you for your kind reply
Hi Larrin, great article. It did leave me curious about the consequences/ effects of austenizing a steel for too long. Some data sheets recommend preheating and equalizing before austenizing. If a steel is held at a certain temperature for so long to equalize the entire peice, than what more would an extended hold time do to the steel especially at austenizing temperature. A few times I have missed the end of the hold time by ten to twenty minutes on my evenheat kiln when hardening a blade and find the kiln below temperature by a few hundred degrees. At that point I assume the blade itself had cooled a bit and was nolonger hot enough for full hardening; so I would retire the kiln to the same temperature and hold it there for another ten minutes. Are there any negative effects of prolonged austenizing before quenching?
If holding for long periods of time there is a danger of grain growth. With normal holding times this is rarely an issue at the proper temperature. Don’t forget the steel in the furnace. If the blade has cooled by several hundred degrees it has been trasforming to ferrite/pearlite/carbide which is obviously not what we are looking for.