Backyard heat treat of Hitachi shirogami and aogami (white and blue) steels

M

milkbaby

Joined
Aug 1, 2016
Messages
644
For a Christmas or Hanukkah present to myself, I'm thinking of buying a little piece of laminated shirogami or aogami from the eBay. I only have stock removal tools and am thinking of doing a santoku out of the 0.130" thick san mai steel. I only have a two brick forge and canola oil for heat treating.

My questions about how to work this steel: I would profile and grind to about 0.020 to 0.030" edge? Then it looks like I would need to normalize 3x at just under or just at nonmagnetic about 1415F with air cool to black? Then for hardening go a shade past nonmagnetic 1475F and soak ideally 10 to 15 minutes then quench in canola oil? Probably realistically can only soak a minute or two by pumping the blade until I get too antsy about way overheating it. Should canola oil still be heated to 130F for quenching? Tempering, I don't know... figured 2x at 375F will hit around HRc 60?

I'm a newbie only have used 1084. I fully understand that I won't be getting optimal performance out of this steel, but I wanted to try this steel for kicks. If I can normalize properly and not overheat during the hardening, then what is the main metallurgical issue with not soaking long enough? Won't the vast majority of austenite be formed after a minute or two soak at the austenitizing temp? What are the various problems? Carbides formed, distribution, grain boundary issues, etc?
 
Last edited:
The san mai steel will be received in a condition that is ready to harden. That is to say, no normalizing or thermal cycling is needed. So simply harden. Canola oil is OK, but not ideal, as the Hitachi steels are extremely shallow hardening. If you don't want to do a brine/water quench (I don't blame you if you didn't), and you don't want to spend the $$ on a fast oil (understandable...but if you will be working these steels or W2/1095 in the future....get it...it's more than worth it), then 130F canola will work. 1475F with a 10 minute soak will work with Blue and White. 375F with fast oil or water quench is still going to be around 64HRC. For 60HRC probably need to go hotter, maybe 450F. But because you are using canola...these numbers may be off. 375F may put you around 62-63 with canola...not too sure.
 
Given the lack of real heat control or ability to soak, I wouldn't worry about the quenchant overly. Canola is fairly fast.

Since soaking doesn't really work with your forge, something that might work instead is to do several "thermal cycles", where you heat the blade just above quench temps (1500-1550°F) 2 or 3 times and letting it cool a bit between each one, then heat to your best approximation of quench temp and dunk in the canola. You could also quench between cycles, which would help get your oil up to temp for the final quench.

This isn't really a "technique" as much as an possible method that might get more carbon into solution than a single heat would. It isn't ideal, but nothing is ideal about trying to heat treat 1% carbon steels in a forge. If someone thinks a cycles with or without multiple quenches is a bad idea, please speak up.
 
RX-79G said:
Given the lack of real heat control or ability to soak, I wouldn't worry about the quenchant overly. Canola is fairly fast.

Since soaking doesn't really work with your forge, something that might work instead is to do several "thermal cycles", where you heat the blade just above quench temps (1500-1550°F) 2 or 3 times and letting it cool a bit between each one, then heat to your best approximation of quench temp and dunk in the canola. You could also quench between cycles, which would help get your oil up to temp for the final quench.

This isn't really a "technique" as much as an possible method that might get more carbon into solution than a single heat would. It isn't ideal, but nothing is ideal about trying to heat treat 1% carbon steels in a forge. If someone thinks a cycles with or without multiple quenches is a bad idea, please speak up.
Click to expand...

This won't work. You only want to get about 0.80% carbon into solution or you end up dealing with retained austentite, the bane of hypereuctoid steels when working without temperature control. You aren't going to accidentally get better structure than the way hitachi steels come delivered. It's already ideal. The rest of the 0.3% carbon will stay in fine carbides if you don't mess it up. Backyard heat treat poses no advantage to steel, just multiple ways to screw it up. Do the least you can to mess it up.

I would worry about the quenchant. DT-48 and Parks 50 are on the slow side for these steels. I would recommend 3 seconds in brine, then the rest in heated canola oil, followed immediately into a 375-400f 2h temper. Not ideal, but probably the best chance of getting a blade at least as good as 1084.


Not to be too critical, but you've posted metallurgical inaccurate information a few times now. You seem to know the jargon. Maybe review the verhoeven text and clarify some of the processes and outcomes.
 
Willie71 said:
This won't work. You only want to get about 0.80% carbon into solution or you end up dealing with retained austentite, the bane of hypereuctoid steels when working without temperature control. You aren't going to accidentally get better structure than the way hitachi steels come delivered. It's already ideal. The rest of the 0.3% carbon will stay in fine carbides if you don't mess it up. Backyard heat treat poses no advantage to steel, just multiple ways to screw it up. Do the least you can to mess it up.

I would worry about the quenchant. DT-48 and Parks 50 are on the slow side for these steels. I would recommend 3 seconds in brine, then the rest in heated canola oil, followed immediately into a 375-400f 2h temper. Not ideal, but probably the best chance of getting a blade at least as good as 1084.


Not to be too critical, but you've posted metallurgical inaccurate information a few times now. You seem to know the jargon. Maybe review the verhoeven text and clarify some of the processes and outcomes.
Click to expand...

Specific to the OP, it doesn't sound like he is going to be buying professional quenchant, and as Stuart indicated, you might not want to risk a brine quench when the raw material is so expensive. Are you are saying that canola oil specifically is not fast enough?


As far as thermal cycling below normalizing temperatures, I was not saying it would definitely improve the final product - just that it possibly could. But I don't understand how it would hurt the final product, as you are implying. I do not know why you think I was implying some advantage to backyard heat treats. It would be nice for you to address the issue directly rather than making a more general statement about my knowledge.

If you are talking about our discussion about 52100 in the O1 thread - I apologize: I missed you mentioning both normalizing AND thermal cycling.



In particular to this thread, the OP is specifically asking how to do something the wrong way - harden hypereutectoid steel without any temperature or time controls. And that's pretty speculative territory. As far as I know, no professional has spent any time analyzing backyard techniques to see if there are any processes that can make up for the lack of precision or soak.
 
I wasn't referring to you advocating for backyard heat treat. When judging temp by eye, even mastersmiths are out by 200f on average underestimating temp. So, your guess of 1500-1550 becomes much more likely 1700-1750f, probably at best. This will not help the steel at all, and will mess up the fine grain structure the steel already has. In regards to the other thread, you said you thought the normalizing and cycling the steel if it's already fine spheroidized is still necessary. That is not true. We do the cyclibg and subcritical anneal to get to that state. No repeat needed if it's already in that state. Same with the OP, the steel is ideal. Don't do anything to mess it up.

Watching for decalescence is the best the OP can do, and that will not get the most out of the steel, but mess it up the least.

As for canola oil, it's too slow. If the OP got an as quenched low Rc60's, I'd be surprised. 3s in brine if the steel isn't messed up through unneeded steps that could harm the steel is relatively safe. I just did a hitachi blue #2 in brine for 3s, into DT-48, and it survived fine. I broke 2/10 or so I did this way this year. The cladding makes it even safer, as it shields the steel from the shock.
 
I'm with Willie on this.

If the OP isn't able to properly austenitize the steel, has no Parks #50, and doesn't want to send it out, he should do a minimal conversion quench.

I would heat to non-magnetic and then barely a shade redder (about 1475F). As soon as the blade edge is evenly at that shade, quench in 8% brine for a three count, pull out and quench in a fast oil ( canola will work) for a five count, and then pull out and hang in still air until able to hand hold.. Temper at 400-450F for an hour twice.
 
Willie71 said:
I wasn't referring to you advocating for backyard heat treat. When judging temp by eye, even mastersmiths are out by 200f on average underestimating temp. So, your guess of 1500-1550 becomes much more likely 1700-1750f, probably at best. This will not help the steel at all, and will mess up the fine grain structure the steel already has.
Click to expand...
So it isn't that the cycles would hurt it, it is that a backyard smith can't be trusted to get them right.

In regards to the other thread, you said you thought the normalizing and cycling the steel if it's already fine spheroidized is still necessary. That is not true. We do the cyclibg and subcritical anneal to get to that state. No repeat needed if it's already in that state.
Click to expand...
I don't believe I said anything like that at all, since I was talking about forged 52100 and didn't mention spherodizing at all:

O1 is a decent choice if you are forging AND have decent heat control. 52100 is more difficult to heat treat,
Click to expand...
I don't know your process - whether you were forging or grinding and what state the steel started in. But if you went up to normalizing temps but never did any thermal cycling around quench temps, then it is unlikely 52100 got as hard as it can or should.
Click to expand...

However, I do not think it is 100% universally agreed that spherodized 52100 from the supplier will prepare and refine the carbides as well as thermal cycling stepping down to final austenizing temp. Especially if you don't know that it is fine vs heavy spherodized. I'm getting that impression from Cashen's site and other things I've read about the difficulty of dealing with that much carbon and producing fine carbides, not from some mumbo-jumbo triple quench blowtorch stuff.

I do not pretend to be a metallurgist, but 52100 used to make extremely fine grain/fine carbide knives still gets a considerable amount of attention from the pros for the subtlety of methods useful for getting it right compared to most of the .95% C steels. 52100 is one of those steels that takes more futzing to live up to its ideal, but will perform reasonably when heat treated in a more common fashion.
 
Last edited:
RX-79G said:
So it isn't that the cycles would hurt it, it is that a backyard smith can't be trusted to get them right.

I don't believe I said anything like that at all, since I was talking about forged 52100 and didn't mention spherodizing at all:




However, I do not think it is 100% universally agreed that spherodized 52100 from the supplier will prepare and refine the carbides as well as thermal cycling stepping down to final austenizing temp. Especially if you don't know that it is fine vs heavy spherodized. I'm getting that impression from Cashen's site and other things I've read about the difficulty of dealing with that much carbon and producing fine carbides, not from some mumbo-jumbo triple quench blowtorch stuff.

I do not pretend to be a metallurgist, but 52100 used to make extremely fine grain/fine carbide knives still gets a considerable amount of attention from the pros for the subtlety of methods useful for getting it right compared to most of the .95% C steels. 52100 is one of those steels that takes more futzing to live up to its ideal, but will perform reasonably when heat treated in a more common fashion.
Click to expand...


Please explain what you think you are accomplishing metallurgically with your cycling after a steel us properly set up.
 
I have worked with White steel and warmed veg oil is definitely not fast enough. Applied very thin clay coat on the blade and water quench if you don't have fast oil.
 
There should be no need for clay. He is talking about using laminated ao/shiro and this will also increase the need for a fast quench San-mai and suminagashi have a very thin core that needs a really rapid quench. Water/brine is the best for these. Parks #50 at room temp will barely get it done if all the temps and such are perfect.
 
The very thin wash of clay* is used to speed up/even out the cooling provided by a plain water quench (it apparently does this by creating nucleate boiling points). I don't know why that became the standard way to water quench in Japan, whereas brine became accepted in the West, but both methods are supposed to prevent the vapor jacket from forming.

*Edit: I'm calling it clay, but it's usually a mixture of clay, powdered sharpening stone, and charcoal (or just powdered stone and charcoal, depending on the region in Japan). They tend to use a coarser grit powder for the thin wash and a finer grit of stone for the clay meant to insulate when creating hamon.
 
Last edited:
RX-79G said:
To reduce the xize of the spheroids in the marstenite and form smaller carbides. Read Cashen's comments on this page:

http://www.hypefreeblades.com/forum...le&sid=caa5420c709cc0c0304979bf6a07d8d6#p4795

What is "properly set up?" and who set it up? Are you talking about trusting that the mill's spheroidization?
Click to expand...
Properly set up means that it is fine spheroidized (as opposed to coarse), and yes, the mill knows what they are doing. This is Hitachi Yasuki, after all. The steel is received in the best condition you're gonna get it in. You can thermal cycle, but all you will do is reduce hardenability on a steel that is so shallow hardening already. You're not going to reduce aus grain by any appreciable amount by thermal cycling, you'll just make things harder on yourself with these steels. They are ready to harden as received. The only time you'll need to normalize/cycle these steels is if you've forged them.

To the OP and the quench, the interrupted quench was mentioned and is a good route to take with these steels. 1 or 2 seconds into water, finish of in 130F canola. Warren is right....P50 is just fast enough, almost not fast enough. VERY fast quench needed for Shiro and Ao steels, especially with a jacket of iron around them.
 
samuraistuart said:
Properly set up means that it is fine spheroidized (as opposed to coarse), and yes, the mill knows what they are doing. This is Hitachi Yasuki, after all. The steel is received in the best condition you're gonna get it in. You can thermal cycle, but all you will do is reduce hardenability on a steel that is so shallow hardening already. You're not going to reduce aus grain by any appreciable amount by thermal cycling, you'll just make things harder on yourself with these steels. They are ready to harden as received. The only time you'll need to normalize/cycle these steels is if you've forged them.

To the OP and the quench, the interrupted quench was mentioned and is a good route to take with these steels. 1 or 2 seconds into water, finish of in 130F canola. Warren is right....P50 is just fast enough, almost not fast enough. VERY fast quench needed for Shiro and Ao steels, especially with a jacket of iron around them.
Click to expand...
Since the thread had drifted, I was speaking generally and not just about Hitachi. It is pretty well known that Aldo put out a bunch of 52100 that was not properly prepared and had large spheroids, and 52100 is one of the highest carbon low alloy steels there is, which causes other issues.

I apologize for giving the OP poor advice on Hitachi heat treat methods - I presumed that thermal cycle temps would be held in the correct temp range to not damage anything. The problem of overheating is perfectly valid, even though it applies as much to the final quench as any cycles.
 
Even IF you thermal cycle these Hitachi steels "in the correct temperature range", no you will not "damage" anything, but you will reduce hardenability on steels that are extremely shallow hardening to begin with. In which case fast oil will not be fast enough to harden them, and you will have to use water/brine.
 
samuraistuart said:
Even IF you thermal cycle these Hitachi steels "in the correct temperature range", no you will not "damage" anything, but you will reduce hardenability on steels that are extremely shallow hardening to begin with. In which case fast oil will not be fast enough to harden them, and you will have to use water/brine.
Click to expand...

I'm not sure I understand. Thermal cycling can reduce grain size and reduce the maximum as quenched hardness. But that is not the same as saying that the steel is "less hardenable", is it?

Lower hardness due to cycling or lower austenizing temps is not the same as lower hardness due to incomplete marstenite formation and having a lot of pearlite in the matrix due to incomplete hardening, as far as I understood.


Wouldn't it be actually better to get a slightly lower quench hardness and then temper at lower temps than get such a high hardness that you have to temper at 450° F, which is in the TME range that starts at 446°?
 
The finer the grain, the less hardenable the steel will be, requiring faster quenches.

Regarding the 52100, as Stuart said, Aldo's stock is course spheroudized. That steel needs normalizing and thermal cycling. If you get 52100 from AKS, it's already fine spheroidized. You do not need to do anything to this steel. It's a great choice for stock removal. You won't improve it by cycling. It's already in optimum structure.

You want maximum hardness out of quench. Then temper to the desired hardness. What do you think the advantage of not hardening properly? TME is a very minor problem in our application.
 
Thermal cycling, strictly speaking, has nothing to do with the hardness levels.....it is to reduce aus grain size. However, reducing aus grain size reduces hardenability. This means, at the end, you need a faster quench to get that max hardness. In other words, take Shiro and quench as received, in P50, you'll get, what? 66HRC+. Take Shiro and thermal cycle it a few times, lets be excessive here and thermal cycle 5 times, you will reduce the hardenability of the steel to such an extent that P50 will no longer give max as quenched hardness. So in that sense, yes, thermal cycling does reduce the as quenched hardness...but it is because the "standard" (I hate to use that term here) quench is now too slow to get the job done. To get that back, you need to to quench in water/brine, or....."start over" and normalize it.

I'm not sure about that last statement. What do mean by that? You always want to shoot for max as quenched hardness and then temper for desired HRC. TME should not be an issue with these steels, as there is no reason to temper them that high. These steels are made for high hardness, thin edges. If you are tempering Shiro and Ao in TME range, then you're lowering hardness levels to such an extent you should be using a different steel to begin with. I'll let others chime in....got work to get done.
 
I just did a hitachi blue #2 kitchen knife with a brine interrupted quench, into DT-48. It came out at Rc67. I had to temper at 400f to get a final hardness of Rc63/64. If you want these steels softer than Rc62/63, you are using the wrong steel.
 
You must log in or register to reply here.
Back
Top