The average ranges of impact testing between the numbers that I provide generally do not overlap ...
Larrin, again, they will overlap if you look at the comparison in detail as I showed in the above and do a simple error propogation calculation. You have to consider not only the variance introduced by the test but all factors which would influence it if it was repeated by someone else because that is in fact what is done when you compare results on different steels.
This is basic numerical analysis and it should have been made very clear to you when you were giving the results that they should not be used in such an exact manner. Again, take a steel like 440B and look at its ranges and see how close it gets to both 440A and 440C and what this implies about what you would expect if you were testing samples of these steels. Now add on to this the extra variance from the machine measurements and the heat treating.
Do I have to add a disclaimer about what testing means and the estimation that goes on behind it whenever I provide numbers?
Generally if you are mixing in estimates with actual measurements you would specify them and if you are using test data which has inherent problems you would note that as well. As I said, check Carpenters steel book and you will see that they constantly warn that impact testing on high strength steels is an estimate and not something which you would use to compare in the manner you did in the above.
In general tool steel books such as Robert's and Cary will warn significantly not to make the types of comparisons that you have done in the above, especially when you restrict it to one such statistic. This is why if you look up the reference data in actual materials texts you see full curves on every possible statistic and you will note that the superiority often switches depending on which one you use. Now obviously someone selling you something picks the one where theirs does the best and if you ask the guy competiting with them you get a different statistic which is more favorable to them.
You also have to be aware that there is a huge different between impact testing on cm blocks of steel vs sub mm edges. For example, the shallow hardening steels do very well in impact testing because the core is actually nor hardened. Similar, steels with a high amount of retained austenite do well because that actually changes to martensite during the impact and soaks up the impact energy. However in an edge there is no soft core and retained austenite is a negative because you are not concerned with the results of one impact but long term and you will end up with a more brittle blade due to it eventually containing untempered martensite.
In other words, it takes more strength to torque the steel at a higher hardness but if too brittle it breaks too easily ...
Torsional strain graphs will show both the ultimate torque as well as the ductility, you would pick the point where both of them are high. For example A2 has an ultimate torque maximum at 350F but if you draw back to 400F then you gain 50% in ductility for about a 10% loss in strength, so it you could argue there is benefit for that tradeoff. The impact value is also a maximum at 400F.
In general torsional data tends to be more sensitive and shows embrittlement regions that charpy/izod do not. These are real physical problems which were known before they were measured, such as "blue embrittlement". This obviously shows the torsional data is superior because it can discriminate.
I have also used knives so hardened and they have been discuss in rec.knives for about 15 years now so it isn't just a issue of theory.
If S30V does have 4 times greater toughness in side impacts, that does not mean it has four times greater toughness overall, so I don't see why it would have to have the same toughness relationship as M2 vs. L6.
Ok, fine, restrict it to side impacts only then, you still would claim that it matches L6 vs M2 in that regard. Just stop and consider just how large of a difference in toughness that actually implies. Note that these types of lopsided comparisons hold for all P/M vs ingot steels, even the high carbide ones will still have very high relative toughness vs ingot steels in that regard. However look around in the cutlery industry, have you really seen evidence they are that much tougher which would support the claim that is a critical statistic? Is there any evidence for this at all?
-Cliff
