Jump to content

Inducing positive curvature (upward sori) in a shinogi-zukuri katana using OIL

Recommended Posts

I know the first thing everyone is thinking is that I fell from a tree and hit my head really hard. Everybody knows that when you quench a katana-like sword in water the tip goes up and when you do it in oil the tip goes down. But water causes blades to crack and oil doesn’t, so some people will pre-curve the blade and quench in oil for that reason alone.

Up until recently, I have used water or water-to-peanut oil as my main quenchant. Why? The hamon are more vivid and I can grind a blade straight easier than I can a curved one and when it goes in the water the curvature is very natural. Since I moved to the new shop, I had a number of blades crack with the new setup. More than usual and I had Parks#50 on hand, so I said why not use it for the next batch of blades I am making.


The chance observation:

I prepared four blades related to upcoming commissions. I pre-curved the blades knowing that the curve will relax some in the oil.

(All measurements are approximate. I am not good outside the metric system).

First blade went in. W2 steel. The curve relaxed from 1 ¼” to 1”.

Second blade in. W2 steel. From 1 ½” to 1 ¼”.

Third blade in. W2 steel. From 2” to 1 ¾”

Fourth blade in. Sanmai 1050 core and folded cable on the sides. The curve went from 2 ¼” to almost 3”.

What the heck?

Later that night, I took one of the W2 blades from above but did not want to spend the time applying the clay and drying it, so I thought I would try to heat up the edge more than the spine to get the same effect. Warmed up the oil to where I thought it was during the 4th quench. Around 180 degrees. The blade curved from 1 ½” to 7/8” in a dramatic quench. Boiling oil is a bit scary.

Again, what the heck?

I went to bed but I couldn’t sleep. I exchanged emails with Howard Clark and he confirmed that no one to his knowledge had reported this before. The next morning I headed to the shop.


A series of experiments:

The first I needed to do, was to prove to myself that this wasn’t an accident and that it could be repeated so I planned to do exactly the same thing I did the day before using the same blades and at the suggestion of Howard, I kept track of the oil temperature. But being me, I couldn’t help it but to introduce a couple of changes in the sequence.

The day before I started from room temperature oil. That temp was around 50 degrees F and for the repeat series I warmed up the oil before the first sword went in. I did that by quenching a sword-like object (approximately the same mass and same austenizing temp). The temp of the oil climbed up to 110.

First sword goes in. W2 steel. Sori went from 7/8” to 5/8”. Oil temp climbs up to 150 afterwards.

Second sword goes in. W2 steel. Sori from 1” to 5/8”. Oil temp up to 155.

Third sword goes in. W2 steel. Sori from 1 ¼ to 1 ½. Oil temp about 200.

Here we go again.

Next was the sanmai sword but I decided not to put it in. I figured it will become a scimitar after the quench. Instead I took the third sword of these series, scraped off the clay brought it up to austenizing temp and quenched it. The curved relaxed quite a bit in a dramatic boiling quench.

At that point I have concluded that:

1) The phenomenon was repeatable.

2) It happened to both a hypereutectoid steel (W2) and a hypoeutectoid one (1050).

3) It required the presence of clay.

4) It seemed to be mainly related to a certain oil temperature above 155 degrees.

5) No cracks.

I had some idea of what was happening and now I needed to put it to practice. Not to bore you with the details of the additional experiments over the next couple days, I will tell you that I set to determine how much the curve was moving and to get a better approximation to the oil temperature responsible for the effect.

As for the curve. It is a very consistent 1/4” movement up every time for a given amount of clay and oil temp.

The oil temperature seems to be more like 180 degrees. At this temperature the oil easily boils and/or ignites. So be cautious if you want to try this.

There is a larger effect to the presence of clay than what I initially thought. To the point, that the amount of clay on the blade determines the degree of up-curve that occurs.

It does not seem to be affected by the austenizing temperature. At least within the range of 1400 and 1500 degrees.

Still no cracks. :D


To wrap up this long diatribe, I want to point to the fact that I would have never find out, if it weren’t for the fact that I decided to harden four blades instead of the usual one.


From now on, the axiom of “the curve goes up in water and down in oil” is no longer.

  • Like 2
Link to post
Share on other sites
  • Replies 61
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Popular Posts

I know the first thing everyone is thinking is that I fell from a tree and hit my head really hard. Everybody knows that when you quench a katana-like sword in water the tip goes up and when you do it

Preliminary pictures of the second sword. This is a sanmai construction 1050 for the core and folded cable steel for the sides. The hamon is a simple suguha (straight). It worked on one side. The oth

Posted Images

wow...i hope you realize that you have just revolutionized a new method of heat treating...seriously, i would consider this ground breaking...great job and thank you very much for sharing

Link to post
Share on other sites

thanks for doing all of the systematic work before you broke this to us.


So - is the hamon as vivid with warm Parks as it is with warm water? After all, you said you primarily use water due to hamon effects. I, in my neophyte way, have been using water rather than oil for the same reason for clay heat treatments.


However, if the hamon is equally active, and the sori can be positive or negative depending upon need, then you have really opened the door to something wonderful for all of us.


In fact, you still have opened the door to something wonderful, and thank you.(but - what do the hamons look like?).



Edited by Kevin (The Professor)
Link to post
Share on other sites

Very interesting, Jesus. Good to see that this craft is still full of dicovey after so many time practising. From my point of view, the cracking risk is part of the game, making yaki ire a rich human adventure... For a production/commission work, I understand your concern though.

Can't wait to admire the result anyway

Link to post
Share on other sites

The jury is still out on the hamon, Kevin. I think it could be very beautiful too but I am having to change many of my previous ideas about lay out, thickness, etc. The grain on these blades was very fine after using them for so many cycles of quenching and I suspect that could also have an effect on how high the hardenability and therefore the hamon.

Link to post
Share on other sites

Interesting, pretty fascinating actually. I've been quenching clayed kitchen knives in hot canola all year, and typically get zero movement either way...unclayed I do get a bit of negative sori, so I guess that may mean I'm seeing a little of this effect myself...? Much more noticeable in a 25"+-long blade, I'm sure.

Link to post
Share on other sites

I remember reading something about this(oil temperature/aggressiveness) in a steel manufacturer or quenchant manufacter's PDF about heat treating(no, by no means do I remember which one). I always kinda wondered about how it effected sword like objects with clay, but have always been pretty happy, or at least, set in my ways with water.

Thanks for sharing the experiment and please post pics of how the hamons turn out.

My assumption is that the positive curved blades will be more defined. Im curious to see if thats *really the case though. What are your preliminary thoughts?


Is there a noticeable shift in viscosity around 155F for the oil?

Link to post
Share on other sites

Is there a noticeable shift in viscosity around 155F for the oil?


Yes. It's almost like water. There is also a significant volumen expansion.


Too early to call the hamon. It will take several weeks to polish these blades to see what I really get but I will certainly share the pictures.

Link to post
Share on other sites

Ive noticed these things when making "fry bread". Ive used olive oil and lard.

I could almost say both seem "thinner" than water at elevated temps. Lard seeming to be the "thinner" of the 2. They are both noticebly more volumous when at a good heat.


Youve inspired me to actually check the temperatures next time I make frybread and maybe start to do my own experiments with oil as a quenching medium. Ive used it very little in the past, but Im certain it could come in handy for certain things.

Link to post
Share on other sites

I'm confused by the term "relaxed" refering to the last blade on the second night. Did this blade experience positive sori, or just the san-mai?

Link to post
Share on other sites

HMM Very cool Jesus. For a while I was doing 4 or 5 in a row into a long 6" diameter tube of Park's #50. The effect of the oil temp changing over time did cause me some trouble! I'd start with the oil at near room temp. Usually on about the 3rd quench the cloud of Park's ignites and needs to be covered. The temp of the oil is around 150f at this time. Parks says between 50f and 120f is the desired range, since the viscosity is already very high.


If we could watch the quench in slow motion to see what the vapor jacket looks like at different temperatures, that might tell us something.

I suspect that since the oil is more prone to vaporizing when it's lightened by higher temps, the gas jacket collapses later in the quench, like it does with water.

In a water quench, the bubbling doesn't really stop until the sori has come back up.


I do the grinding while the blade is still straight, then pre-curve it over a piece of wet wood or copper bar with a soft hammer to avoid making any marks or dings that would require more grinding.

Thanks for sharing your findings with us as always.

Link to post
Share on other sites

That's how I pre-curve too: Rubber mallet and the blade supported on a 2x4.

Your thoughts on the vapor jacket are very interesting.

I must say that on two of the blades (foundation polish only) the hamons are looking very Japanese. :D

Edited by Jesus Hernandez
Link to post
Share on other sites



If we could watch the quench in slow motion to see what the vapor jacket looks like at different temperatures, that might tell us something.

I suspect that since the oil is more prone to vaporizing when it's lightened by higher temps, the gas jacket collapses later in the quench, like it does with water.

In a water quench, the bubbling doesn't really stop until the sori has come back up.




i don't think that this is quite what's going on - i have no experience with parks, and my physics is a little hazy, so take this with a grain of salt, but i'm pretty sure that most oils have little to no vapour phase; instead, nucleate boiling starts pretty much immediately.


when you quench in water, the real business end of the quench doesn't start until the vapour jacket collapses, which is one of the reasons a thin clay wash speeds up the quench, as it disrupts the vapour jacket.


nucleate boiling is where the real work of the quench gets done, and you beat the pearlite nose. i believe that this stage of the quench is both much faster in water than in oil, and goes on longer in water.


the final stage of the quench is the convection phase, which is comparatively fast in water and slow in oil.


i think that this fast convection phase in water is the reason for most cracks (though the violent collapse of the vapour jacket and the longer, more aggressive nucleate boiling stage must also cause issues), as the blade is still cooling very fast after the martensite has started to set up.


the upshot of this is that it is the rate of cooling in the latter phases of the quench which determines if the blade goes up or down, as the slower oil quench allows the martensite to organise itself into a less stressful configuration.


so basically what i think is going on is that at such high temperatures the oil has lost enough viscosity that the rate of heat extraction through convection has crossed a tipping point and is now more analogous to water than a standard oil, and possibly that the life expectancy (probably) and rate (almost certainly) of the nucleate boiling stage has also been raised.


this would seem to raise a couple of issues: firstly, as quenching is a physical process rather than chemical, i'm not sure that i'd expect blades quenched this way to have a much greater survival rate than those quenched in water (though thinking about it some more it could be that the convection phase is just fast enough to produce sori, while still being considerably slower than water, so that could be bunk... but i suspect that the fact that the edge is under tension rather than compression has a lot to do with the failure rate in water, and this would still seem to be an issue here); and secondly that i'm fairly convinced that the collapsing vapour jacket is what causes a lot of the cool traditional looking effects from a water quench, and i'm not sure that an oil quench can ever reproduce them exactly.


no idea if this will make any sense to anyone, and of course, i could be totally wrong about any and all of this - i've just got back from Paris and my brain is pretty much mince...

Edited by jake cleland
Link to post
Share on other sites

Well, Jake, this is exactly the kind of discussion that I wanted to see in this post and the main reason I provided the information.

I think what you say makes perfect sense even in your Paris state-of-mind but as you also point out why neither of these blades cracked when I basically tortured then through multiple cycles and quenches. That will hopefully open a talk about what people think actually causes the blades to crack. I will venture to give my opinion: a different rate of formation of martensite from the surface to the core of the blade. And secondly, the hamon. So far, it appears to me that the hamon (following the same laws of physics) on these blades behaves a lot more like it does when doing it à la Japanese. And just to be a bit more precise about what I mean: it seams to follow the rules of applying clay that work better for tamahagane blades.

Link to post
Share on other sites
  • 3 weeks later...
  • 4 weeks later...

The J Hernandez phenomenon. I think that would be a good name for it. I can't wait to see these blades.

Link to post
Share on other sites
  • 2 weeks later...

As promised, these are the pictures of the polish of one of the blades. I have two more to polish and I will post those when I complete them.

I must say that I started this thread in the spirit of hoping that someone else would try to see about the reproducibility of this phenomenon.

I am still hopeful...


Now for the pictures.













Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

  • Create New...