Carbon vs Hardness

[kel_aa]

This carbon vs hardness graph is from the fourth edition (I think the latest was the fifth in 1995) of Contemporary Manufacturing Processes, by J Barry DuVall. It showes that the hardness attainable by steel as function of the carbon content, up to about 0.55 weight percent, where the maximum hardness levels out at 66/67, which is the hardness of marsenite.

The author cites research evidence to say that the hardness is only dependent on carbon content, no on other alloys. The points on the graph certainly leads one to that conclusion. Is this correct and complete information?

Aside from marsenite hardness, I understand preciptates can also increase the hardness. Is this of minor importance compared to marsenite hardness?

CRKT cites their 420J2 to be 0.32% carbon, with a final hardness of 54-56. From the graph, it looks like 56 is about on the line. Does this mean there was relatively little tempering?

http://www.crkt.com/steelfct.html

 

[Cliff Stamp]

Allen has a similar graph but notes the increase is present above 0.6% as a horizontal asymptote which looks to be about 68/69, it is a very slow rise above 0.6%. Generally secondary precipitates increase martensite hardness by essentially reversing the tempering process. During tempering the grains and carbide coarsens, secondary hardening induces the formation of very fine alloy carbides when act to basically "pin" the martensite in place making it harder. In extreme high alloy steels this can be so strong that the steels reach 72 HRC. There is also work hardening, see the HRC tests that Spyderco had done on their H1 blades which showed a very high hardness.

In regards to the CRK&T, in general there is an increase in hardness and strength initially during tempering as transition carbide will precipitate from the martensite. In general many of the low alloy steels tend to actually have toughness and strength maximums at the lower tempering range because there is an embrittlement range at 500 F. If you can believe that hardness then yes, the temper was low, the soak was high, the quench fast and there was a cold treatment. For perspective, Phil Wilson, custom heat treats his 420HC gets about 55 HRC and he does a full liquid nitrogen treatment.

The numbers often posted are the maximal results which mean optimal soak temperatures and oil quench + cold treatment. Production blades usually are limited in several of these aspects and as well can be influenced by large batches. I have used specifically CRK&T's 420J2 and Wilson's 420HC which as noted are spec'ed at the same hardness but Wilson's 420HC is *MUCH* harder. Note in many alloy steels the alloy elements will lock up the carbon and prevent it entering the austenite, this is the case in most vanadium alloys for example because it takes extreme temperatures to dissolve vanadium carbide. Most stainless steels also have large fractions of chromium carbide also present after austenizing which reduces the effective carbon content. 440C for example has only 0.56% of carbon in the austenite when soaked at 1100 F.

-Cliff

 

[kel_aa]

So increases hardness by preciptation and toughness by releving quenching stress. Would you be an idiot to want an unquenched blade?

Does plain carbon steel form any carbides, or does it revert to alpha iron and cemetite upon tempering?

 

[Cliff Stamp]

There are lots of steels which are not quenched but just left cool, some will still transform to martensite because of elements like chromium which stronly retard the isothermal reactions. Bainite has started to be used/discussed now as a alternative to martensite in large blades because it is much tougher and more ductile at similar hardness. Bainite is an isothermal reaction so you don't quench to form bainite (well you do in some cases, but it is to avoid pearlite not form the bainite).

If you meant untempered, there is little reason to want untempered martensite and thus leave a quench blade untempered because martensite is highly stressed and can stress crack on its own. Plus it is both weaker and more brittle than optimally tempered martensite anyway. Some large edge quenched knives are often don't have a specific tempering step because the heat in the spine will auto-temper the edge after the quench.

Plain carbon steel will form iron carbide which is cementite. After tempering there are a number of distinct reactions, first the martensite loses carbon which forms transition carbide (Fe2.4C), in really high carbon steels cementite (Fe3C) can precipitate. As the temperature increases the retained austenite transforms to bainite (ferrite + cementite), as you go higher the cementite coarsens and the martensite loses it tetragonality.

The alloying elements can retard these reactions significantly to the extent that they won't happen in some steels in usual tempering times (1-2 hours), D2 for example strongly resists forming bainite both during the quench and tempering for the same reasons.

-Cliff

 

[Doug C]

Note in many alloy steels the alloy elements will lock up the carbon and prevent it entering the austenite, this is the case in most vanadium alloys for example because it takes extreme temperatures to dissolve vanadium carbide.-Cliff

Does not forging break up the Vanadium,and uniformly(more or less) distribute it.

Doug

 

[Cliff Stamp]

Most steels with the exception of Boye's tend to be used by the maker after forging to break up the dendritic nature of the as cast ingot steels and forging temperatures are high enough to dissolve some carbides. But vanadium is present in many alloys which would not be forged hot enough to get it to dissolve and there will always be recrystallization after forging which is how the grain is refined.

Vanadium will strongly resist dissolving during austenizing and tends to stay as primary carbide. This is why for example if you look at steels which raise the vanadium significntly ( S30V vs 154CM ) the carbon content needs to ramp up massively or they will have a much retarded hardening responce.

CPM 9V for example has lots of carbon 1.8%, but has a soft heat treat responce because so much of the carbon is lock in primary carbides, especially the 9% of the vanadium. In short, while carbon controlls the hardness of martensite, the amount of alloy carbide formers will control how much carbon will dissolve in the austenite at a given austenization temperature.

For example the carbon amounts in 440A, 440B and 440C are 0.68, 0.85, 1.13, but dissolved in the austenite at 1100F they are 0.48, 0.52, 0.56. Even though 440B has about 0.2% more carbon than 440A the hardening responce is increased by effectively only 0.04% carbon. These numbers will be effected by the austenization temperature, you can force more or less to dissove by raising/lowering it as well as increasing/decreasing the time.

You then have to real with other issues like level of retained austenite after quench and grain expansion during the soaking if left too hot or for too long.

-Cliff

 

[kel_aa]

see the HRC tests that Spyderco had done on their H1 blades which showed a very high hardness.

Do you have any idea of how this is accomplished? I understand that this is more of an effect in copper alloys, but of course steel exhibit this too when cold worked. Imagine telling someone their brand-new H1 knife has been "pre-worked and conditioned."

Thanks Cliff for the discussion. I don't know why more people wouldn't join in. Maybe it's not that interesting, and people would rather be told whether their steel X is "good" or "bad" instead of trying to understand why.

 

[HoB]

Well, yes the Salt knifes have been preworked....they are rolled to a certain hardness and another hardness increase appears apparently during the grinding. The lock bar btw. which is also H-1 is rolled to a different hardness.

 

[kel_aa]

So all the hardness comes from cold working and the nitrides? I must say that is pretty amazing. Is there any annealing in the whole process at all?

 

[Cliff Stamp]

So all the hardness comes from cold working and the nitrides?

The precipitation steels come to the manufacturer generally in solution annealed state, this is achieved by heating them to around 1900 F and quenching to put them in a supersaturated solid solution. After they are shaped they are age hardened by heating them to a fairly low temperature (around 1000 F) and held as precipitates form. You can raise the hardness responce of the age hardening by cold working.

Some have put forth the idea that sharpening a H1 knife cold works it and thus the edge retention will increase as it is used and honed. However while there is some cold working in any honing, there is also a lot of direct abrasion. I have also tended to see the opposite where the edge retention will decrease if you do the runs right after each other but if you start them again several weeks later the edge retention will have rebounded.

HOWEVER I just looked at it in a rough way to this point and the numbers are not statistically conclusive and recently I have improved by sharpening which has eliminated some problems of that type so I am not sure it wasn't just a random deviation. It is one of the things I intend to check out in some detail. Given the work hardening issues, H1 could also have a different responce to steeling for example than AUS-6A, which is what it is often compared to edge retention wise on abrasive materials. It is also one of the more durable edges I have seen on a stainless steel.

I reground the primary hollow on a pacific salt to flat and removed the secondary edge bevel so the edge is formed at 6.5 degrees per side. I did this mainly to make it easier to apply bevels for comparisons however my brother took a liking to it and I have yet to see the edge chip or get visibly distorted. It gets significantly blunted, it cuts materials so abrasive and hard that they scratch the primary grind deeply, but the damage has never penetrated the micro-bevel which is sharpened less than 0.005" thick. I reset the primary full every sharpening which is generally less than bi-monthly.

The good thing about such feedback is the very wide range of materials cut in a natural and unconstrained work manner as when you constrain something to make the comparisons more focused there is always the possibility that one of the elements you removed actually has a critical effect on performance. However the problem is that the random work load means it it thus hard to gauge relative performance from one knife to the next (a large statistical sample would make this easier). When I get the Endura back from the pass around I am going to adjust the profile to a similar one and see if I can't convince him to carry them together and see how they respond if he alternates cutting.

-Cliff