HT Techniques and Procedures

quint said:
Bluntcut are you referring to hardening the entire blade or the cutting edge. I think (if I am not mistaken) that you are still referring to a thickness of under a quarter inch right. When we are hardening blades the cross section is very small. That is the reason that you can get away with using a fast oil and still get reasonable hardenability vs using water or brine which may be the recommended quenchent for items that are thicker of the same steel. This is for carbon steel and associated alloys which is the only thing I know anything about.
Click to expand...
I was thinking that regardless of fast or slow quenchant, the thin part (edge) will cool down faster than the thicker area (spine). So aren't we settle for a compromise quenchant that give best result per scenario/target? Thus my 'shade of grey' view rather than definitive right & wrong.
 
Ahh ok I see what you were saying. I think it may have to do with how forgiving the steel is also. Say like 1084 which will get decently hard in various oils even canola oil although the hardest I think is with brine. However the higher scale of hardness usually is not used with that steel anyways so its kind of overkill.
 
bluntcut said:
I was thinking that regardless of fast or slow quenchant, the thin part (edge) will cool down faster than the thicker area (spine). So aren't we settle for a compromise quenchant that give best result per scenario/target? Thus my 'shade of grey' view rather than definitive right & wrong.
Click to expand...

Best I understand it, and maybe someone further along in their walk will correct me, but yes the edge will quench faster cross sectionally speaking. Thus, as I have experienced I think, a fast quenchant will harden the whole blade where as a slower quenchant may only harden the edge and a certain depth of the body of the blade... I believe, and have been told at least, it is simply a function of the rapidity of the temperature change in the steel, no matter the depth at which that steel lies. Thus, quenching with a fast quenchant allows the heat to transfer out of the entire blade quickly and a slower quenchant only transfers the heat from an area of uniform depth throughout the sample. In other words if quenchant x can transfer heat from a sectional depth of steel 1/16" thick, then anywhere the blade cross section is less than 1/16" it will be fully hardened. Anywhere it is thicker than that (total from all sides etc... ) it will only be hardened 1/16" deep.
Yes? No?
 
Stacy sory I wasn't trying to make it personnel. I know what the jominy test does and is designed to do, After the official tests were done to satisfy his curiousity the tester put the samples in the lab freezer overnight then rerun the tests right next to the the spots he had tested beforeat every spot the readings were ~one point harder than the original tests. I asked the oil question because you said something like it dosen't matter how the steel gets hard as long as it gets hard. this is a paraphase not a quote and I could have taken what you said out of context pardon me if that is the case. In tests that I have done in my shop where I have taken two blades and done all the same thermal cycles etc with the exception of triple quenching one and single quenching the other the triple quenched blade cut better and held it's edge longer than the single quenched blade. these were stock removeal blades from the same bar of steel using the recommeded heat treating times and temps from crucibles orange bible. I am not saying that everyone should be doing a triple quench or their blade are junk. I have found that it works for me. If a single quench works for a person and gives them everything out of the knife they made then that is great for them. I do think that it is just as ridiculous for guys to jump all over others who have found that they get the results that they want using a triple quench.
 
I agree with most everything written (very concise analogies BTW). However, I don't think it's always smart to immediately write off as 'false' claims that do not have full known science backing them. I won't take their claims verbatim on face value, but I will say "show me your methods and results". As an R&D and process engineer I have stumbled upon some interesting things - many times the reasons I thought they were happening were later proved false. However, they were still truly happening. I no longer make knives or I would make and heat treat two identical knives, one with industry developed HT recipe, and one using the witch doctor juju recipe, and then test them. All that might be lost is a bit of materials and time, and or some pride if the one you're betting on doesn't win. Ultimately either one's beliefs will be strongly reaffirmed, or strongly trounced.

The best historical knives and swords were not made using sound modern metallurgical theory, but they were made using sound metal working techniques.
 
Without knowing more about the specimen in the Jominy test referenced by Bill, there isn't really a way to answer that question. Also, I would ask if the same results could be repeated and how many more times was it tried? Why was the specimen put in the freezer? At a certain point, much of the microstructure of the bar would be the same as if the steel were air cooled, ie much of it wouldn't be martensite and retained austenite. Those 2 things are all that count if you want the freezer to change anything. If each of the readings was increased on all the hardness test locations, more was going on than just a temperature dip to home freezer temperatures. While interesting, this is going to get well beyond the concise statements of the intial post very quickly.
 
me2 said:
Without knowing more about the specimen in the Jominy test referenced by Bill, there isn't really a way to answer that question. Also, I would ask if the same results could be repeated and how many more times was it tried? Why was the specimen put in the freezer? At a certain point, much of the microstructure of the bar would be the same as if the steel were air cooled, ie much of it wouldn't be martensite and retained austenite. Those 2 things are all that count if you want the freezer to change anything. If each of the readings was increased on all the hardness test locations, more was going on than just a temperature dip to home freezer temperatures. While interesting, this is going to get well beyond the concise statements of the intial post very quickly.
Click to expand...

it was 52100 steel, as quenched. they where put in the freezer to check the claims that we where seeing a change caused by the -15 degree temp. I believe that the test was repeated on several different samples of 52100 steel.
 
LucyCustomKnives said:
Best I understand it, and maybe someone further along in their walk will correct me, but yes the edge will quench faster cross sectionally speaking. Thus, as I have experienced I think, a fast quenchant will harden the whole blade where as a slower quenchant may only harden the edge and a certain depth of the body of the blade... I believe, and have been told at least, it is simply a function of the rapidity of the temperature change in the steel, no matter the depth at which that steel lies. Thus, quenching with a fast quenchant allows the heat to transfer out of the entire blade quickly and a slower quenchant only transfers the heat from an area of uniform depth throughout the sample. In other words if quenchant x can transfer heat from a sectional depth of steel 1/16" thick, then anywhere the blade cross section is less than 1/16" it will be fully hardened. Anywhere it is thicker than that (total from all sides etc... ) it will only be hardened 1/16" deep.
Yes? No?
Click to expand...

Sure, I think it will harden to the steel hardening depth limit per quenchant; however the thinner cross-sectional molecular will reach lattice configuration ready for martensite or being transform into martesnite before thicker section. So perhaps grain will growth in direction from thin to thick and diverging at the same time. I am new in this area (aka newb), for these over my head stuff, I envision steel lattice being form similar to ice/salt crystal form except at rapid rate. So excess cooling will lead to vertical crack or wavy edge or blade face warp because of dimensional stress. And for those steel at greater depth where elements are free to move and form prevalent molecular bond (pseudo anneal).

If my thinking (heheh based on almost zero experiences) is off base, please correct & educate - much appreciated.
 
Bill Burke said:
it was 52100 steel, as quenched. they where put in the freezer to check the claims that we where seeing a change caused by the -15 degree temp. I believe that the test was repeated on several different samples of 52100 steel.
Click to expand...

Any idea of the austenizing temperature or the hardness measurements?
 
Bill Burke said:
it was 52100 steel, as quenched. they where put in the freezer to check the claims that we where seeing a change caused by the -15 degree temp. I believe that the test was repeated on several different samples of 52100 steel.
Click to expand...

I assume, you meant a -85F delta right (70F room to -15F freezer), right?

OK experts, the majority think freezer treatment is silly. I am curious and want to learn - please pardon if I mis-use terminologies. We know thermal affect steel dimensional state contract/expansion. So for a freshly ht steel, can a -85F temp drop contract/squeeze some mis-aligned lattices and break up weak/mal-formed grains (grain refinement)? Yeah a tiny freezer hammer blow vs big sub-zero or mondo cryo :)
 
Good info, but a caveat. Hardening isn't the only factory in knives. From what I have seen and ben told, some steels you "take what the chemistry will give you." Others can be heat treated slightly differently to make them better suited for knives at least when compared to the metals intended purpose. 52100 comes to mind and I am talking about lowering the austenizing temp to below the "saturation line" to around 1475F like Mr. Cashen says. One thing that he also suggested was that triple quenching may actually do some good by allowing you to have less retained austenite on the subsequent quenches IF you are using a less than ideal method. But the implication there is that because of your chosen method, you are not able to get the most from your steel on the first try for sure.
 
bluntcut said:
I assume, you meant a -85F delta right (70F room to -15F freezer), right?

OK experts, the majority think freezer treatment is silly. I am curious and want to learn - please pardon if I mis-use terminologies. We know thermal affect steel dimensional state contract/expansion. So for a freshly ht steel, can a -85F temp drop contract/squeeze some mis-aligned lattices and break up weak/mal-formed grains (grain refinement)? Yeah a tiny freezer hammer blow vs big sub-zero or mondo cryo :)
Click to expand...

Typically, cold treatment is about getting a "stubborn" high alloy steel down to the Mf point and eliminating retained austenite.
 
bluntcut said:
I assume, you meant a -85F delta right (70F room to -15F freezer), right?

OK experts, the majority think freezer treatment is silly. I am curious and want to learn - please pardon if I mis-use terminologies. We know thermal affect steel dimensional state contract/expansion. So for a freshly ht steel, can a -85F temp drop contract/squeeze some mis-aligned lattices and break up weak/mal-formed grains (grain refinement)? Yeah a tiny freezer hammer blow vs big sub-zero or mondo cryo :)
Click to expand...

In a word, no. Even liquid nitrogen won't do that.
 
Bill, Sorry if I took your post wrong.

I do admit that 52100 can be pushed this way and that way more than most steels. What I was trying to state in general is that if the first quench was right in all parameters, it should produce the same results as the third quench of a triple quench.

Exceeding of being insufficient in any HT parameter, like temperature, quench speed, etc. will always affect the blade quality.

I always have problems with "test" results on knife performance, as it is very difficult to do in any controlled way. Sometimes the impression of one knife being better is enough to make the results seem better.

I will be specific about a claim of improvement in 52100 steel. There is a regularly statement that the ASTM theoretical limit for grain size in 52100 is "G10". The statement continues that the triple quench and freezer treatment have reduced this to "G15". The grain size number is just a reference to how many grains there are per sq.in. The formula is n=2(to the power of) G-1 . This is an exponential increase, not linear. An increase of five times would be huge, but going from 10 to 15 is an increase of 32 times in the number of grains. For those who don't do head math, ASTM 10 has 512 grains in a sq.in. ASTM 15 has 16,382 . This is an amazing increase. Even if the steel got a point and a half harder, I can't understand anything in the triple quench/freezer treatment that would account for a thirty-two fold decrease in grain size. Such claims need better proof than just a claim that they happened. The ASTM numbers were calculated in laboratories by repeated tests done hundreds of times in very well controlled circumstances, and have been replicated in industry processes that have been checked in the labs for decades with the same results, so anyone setting this on its ear by thirty-two times is making an extraordinary claim.

One problem that will surely occur in the knife test pass-around being discussed is that unless the two blades are extremely carefully duplicated from the same bar of steel and done by the same maker, any comparison is going to be skewed. The HT can be done separately, but a single HT on a single blade won't quantitatively tell you anything. As we have all experienced, a blade will not always come out as we expected it would. If they made 20 or 40 identical blades, and one HT overwhelmingly showed a better result, then that would indicate something.
 
Some good Points Stacy. the steel that I used in my comparison test was 15n20, that I was told was l6 when I did the testing. So I did use the L6 heat treat info and soak times instead of the proper times and temps for 15n20. since both were stock removed and heated/soaked in a paragon kiln then tempered in that same kiln after cooling in the freezer long enough for the kiln to cool down for tempering. the blade were removed from heat and quenched at the same time(one in each hand). the differences between the two blade where different enough that peception didn't really come into play. Both blades where tested by two other people on butchering animals. neither a maker and neither new which was which. both chose the triple quenched knife as the better cutter.

As far as grain size claims all I know is what the lab told me. Grain size measurement is an approximation based on the number of grains observed under a microscope in a specifc amount of area which as I remember is a square micrometer but don't remember for sure.

Bluntcut,

you are right the samples were room temp when going into the freezer.
 
Ilmarinen - I'm an older guy getting into this 'cuz I love it - not an authority by any stretch.

I agree with many of your comments but in the spirit of dialog I've got a few thoughts:

Statement One (what is a "good" knife) - well said!

Statement Two (Occam's razor/chicken fat) - I'm totally on board with Occam's razor - do the simplest thing that gets the job done right. But I'm not so sure homebrew quench oils will /necessarily/ degrade as you suggest - given the empirical evidence of old timers still using their quench and getting good Rc.

Statement Three (0f is not -100f) - 0f will not get to Mf for most steels used in bladesmithing. But it isn't 70f any more than -100f. Since Ms and Mf are two separate temps, I don't think it's a binary choice (did you get to Washington DC or didn't you). If it was a binary choice then Ms and Mf would be the same temp. I gather from Verhoeven's document that % retained austenite varies as you go from Ms to Mf - so while 0f isn't -100f, throwing the blade in the kitchen freezer is better than quenching to room temp. And yah - as I understand it martensite forms at nearly the speed of sound - so leaving it in the freezer overnight would be a convenience, not a necessity - and ensure that the core of the blade gets chilled.

Statement Four (science is static/the slate is cleaned at phase change) - I expect that metallurgy theory will remain /mostly/ unchanged for the rest of my days, and I study what I can absorb to help me be a better bladesmith - but science does change over time. I would not say that steel is governed by metallurgical science: I would say that metallurgical science is governed by steel. I agree that the slate is /mostly/ cleaned at phase change, but new grains form on old boundaries. There may be other "memories" in the steel that survive. Carbides or the distribution of alloying elements etc. Maybe Multi-quench is just refining the grain size as you suggest - and doing thermal cycling followed by a single quench might produce the same result. But the only way to know is to conduct independently verifiable testing. Otherwise both sides are just speculating. As for Occam's razor - I don't see one as much simpler than the other - I've been tempted to do some tests to see what looks right to me. To really settle this someone in a truly controlled environment would have to compare results from both methods. Otherwise we are arguing beliefs rather than results.

Statement Five (cult of personality/0f is not -100f) - Ah, teachers and mentors - where would we be without them? But yes - they are fallible. Yep - claims are just that unless they can be independently verified. But if I'm understanding the metallurgy properly, while 0f won't get you -100f, it will get you more than 70f.

Final Comments (a good knife/learn the science) - well said! HT, geometry, and right steel for the task.

That's my 2 cents.
 
Think in these terms: when heating to quench, will you get a better HT at 1300f than 1200f? Neither made it to the critical, so neither is better than the other.
 
You must log in or register to reply here.
Back
Top