Can anyone explain san mai warping to me?

L

Leon Husock

Joined
Jul 17, 2019
Messages
313
Every so often I try to make a san mai chef's knife, it drives me nuts warping, I give up on it. But most recently I noticed that when it would warp from uneven grinding, the only thing that would somewhat correct the warp would be grinding on the side with less cladding, i.e. making the grind *more* uneven. Does anyone have any idea why this would be? Logically you would think that to correct a warp caused by uneven grinding you'd want to try to even out how much cladding is on each side. This was done on a knife with 52100 core and medium-high layer random pattern Damascus cladding, in the end I think the spine was about .1" thick.
 
was the Damascus cladding hardenable? or was it a low carbon steel?

thin laminated steel (with soft cladding) requires constant checking and straightening throughout the process. Better to embrace it.
use a straightening stick and a brass mallet on a wooden stump.

keep the core centered while grinding
i don't believe it's warping from uneven grinding...
often I'm straightening 6x during working on a kitchen blade.
the twists and propeller warp is the challenge, you have to find the high side ridgeline to fix those.
 
Last edited:
It was hardenable Damascus, 1084 and 15n20. Honestly I'm not a huge san mai fan (though that might partly be because it gives me so much grief), I'll probably just stick to edge bars when I want a monosteel edge on something unless someone in my area who's really good at it does a workshop. This was going to be a wedding present for a friend who likes Japanese stuff, but he'll get a normal Damascus blade and he'll like it lol.
 
Leon Husock said:
It was hardenable Damascus, 1084 and 15n20. Honestly I'm not a huge san mai fan (though that might partly be because it gives me so much grief), I'll probably just stick to edge bars when I want a monosteel edge on something unless someone in my area who's really good at it does a workshop. This was going to be a wedding present for a friend who likes Japanese stuff, but he'll get a normal Damascus blade and he'll like it lol.
Click to expand...
so it's effectively like a mono steel blade, using a hardenable cladding you get no benefit of san mai construction.
so it's going to warp more like a monosteel blade,

san mai is meant to be used with a soft cladding...otherwise you don't get the benefits of that construction.
but I understand many like the visual aspect of san mai.
 
HSC /// said:
so it's effectively like a mono steel blade, using a hardenable cladding you get no benefit of san mai construction.
so it's going to warp more like a monosteel blade,
Click to expand...
That's what I assumed, which is why I went for fully hardenable steel after having warping difficulties with wrought iron cladding. But both of these knives warped way more than my usual monosteel knives, and were very resistant to fixing the warp while tempering which isn't something I usually encounter. Maybe if I'd used a core which was also present in my cladding, as opposed to a third steel. *shrug*
 
The issue is that even though they are both hardenable, the mono-steel core and the layered cladding harden at different speeds. This makes the blade pull to one side or the other. In some cases, it shears the weld between the core and the cladding. along the spine.

It is best to profile the blade and then harden, Grind the bevels post-HT.
 
Ahh that makes sense. The only time I've ever really successfully made san mai it was 1084 core with 15n20 jacket, and of course those quench at the same rate. The 52100 and the 1084/15n20 Damascus must have thrown things off on this batch.
 
A carbide hammer is a quick and easy way to deal with warps, with traditional san mai you can often just use a brass mallet and hit the soft cladding to straighten.
 
Steels grow when hardened and shrink when tempered but have a net growth.

Carbon and low alloy steels grow more than high alloy steels.

Soft cladding does not grow as much as the core.

Over austenitizing causes retained austenite which results in less growth.

52100 has a higher austenitizing temperature than 1084 or 15n20.

Cryo causes more growth in hardening because it eliminates retained austenite.

The Japanese cold forge before hardening, I believe this helps eliminate some stress because the cladding is stretched beforehand.

Incorrect quench temperature, usually too high, is a cause of stress in sanmai.

Soft nickel in between the core and the cladding helps.

Tempering right after the quench helps. (because the core will shrink some)

There’s probably something I’m leaving out, that’s most of it.

Hoss
 
Thanks Devon.

When I try to explain the metallurgical reasons for why a high carbon mono-steel sword (1095, W2, and similar) will curve down upon quench and then suddenly curve upward some folks' eyes glaze over. Words like super-saturated austenite, TTT curve, and such aren't common knifemaking words.
The reason a fair percentage of water quenches break themselves in half or form severe cracks is due to the principles Devon listed.

When the blade is clay coated for a hamon, it gets worse, as you have a blade with a pearlite spine and a martensite edge. These do not expand and contract at the same time or speed.
The blade is all austenite at austenitization temp ... even under the clay. Upon quench the edge cools quicker than the clayed spine.
Because the edge cools rapidly enough to not transform into pearlite it misses the pearlite nose and stays as super-cooled austenite (which is very soft). As the clayed spine drops below 1000°F if stays within the pearlite nose and converts to pearlite. Thus, the spine expands causing the downward curve. When the steel at the edge then transforms into martensite at 400°F it becomes less dense than the pearlite spine, so it expands greatly, causing the backwards curvature. As the structure changes from soft super-saturated austenite to martensite it becomes very brittle and can tear itself apart as the curvature changes.


Clay thickness and placement, quench timing and rates, blade thickness and geometry, auto-tempering or delaying the martensite conversion, and controlling the temperatures of the steel and the quench rate are all things that have to be exact to get it right. I wish there was a precise recipe for doing yaki-ire, but it is a trial-and-error system specific to you and your forge setup and materials. The best you can get is a guideline of where to start. It may take years or even decades to become proficient at it.
 
Last edited:
No, it was "volume per mass" = lower density. Density and volume are inverts. Density decreases when volume increases. My example is always feathers vs lead, "A pound of feathers occupies more volume than a pound of lead. It has higher volume per mass and lower density".

I did have one part slightly wrong and adjusted the wording. I edited it. Here is the better explanation:
The blade is all austenite at austenitization temp ... even under the clay. Upon quench the edge cools quicker than the clayed spine.
Because the edge cools rapidly enough to not transform into pearlite, it misses the pearlite nose and stays as super-cooled austenite (which is very soft). As the clayed spine drops below 1000°F if stays within the pearlite nose and converts to pearlite. Thus, the spine expands causing the downward curve. When the steel at the edge then transforms into martensite at 400°F it becomes less dense than the pearlite spine, so it expands greatly, causing the backwards curvature. As the structure changes from soft super-saturated austenite to martensite it becomes very brittle and can tear itself apart as the curvature changes.

Here is an in-depth thread about it. Kevin Cashen and I both posted longer explanations of the reasons for curvature.

Questions about sori...

I just forged a Japanese style short sword (wakizashi?) and I'm wondering how straight it should be prior to heat treat. From what I've read, it sounds like water makes it curve one way, and oil the other. So I'm wondering if a fast oil like Parks #50 will make it curve the same as water...
www.bladeforums.com www.bladeforums.com
They are the same basic reasons that warp happens in a billet of two different steels. One thing that is a common misunderstanding is the confusion with thermal expansion/shrinking caused by heating and cooling vs structural expansion/shrinking caused by a change in the atomic arrangement of the structure. FCC is denser than BCC. Austenite is FCC and martensite is BCC (technically BCT in high carbon steel). While thermal changes are there, the largest factor is structural.


Just to finish up this tech talk:
In the conversion from austenite to pearlite, the spine gets harder. The pearlite is not as hard as martensite, but it is harder than the austenite it used to be. While harder, the spine is also less tough. The edge is soft, rubbery, super-cooled austenite so it doesn't care about the change in curvature. So, when the edge expands as it becomes brittle martensite the spine does not want to change its shape. It will lose the battle and curve, but this causes extreme stress between the two areas, and something has to give. If all goes right the spine curves upward and the edge hardens. A quick temper removes the stresses, and all is well. Sometimes they fight it out until one of the two strictures can't take the stress anymore. It is usually the brittle edge. I have seen a blade split lengthwise along the transition area following the hamon line due to this phenomenon. This is what shears a san-mai billet down the spine.
 
Last edited:
Sorry to anyone asking for details, I forgot if you don't check back in you don't get emails about new posts! I'm heat treating in a propane barrel furnace with a thermocouple and pyrometer, so sort of splitting the difference between an electric furnace and a forge. My guess is, based on everything that's been said, the issue was the 52100 and the 1084/15n20 behaving differently when heat treated. In the future if I want to take another crack at it I'll do a core steel with a more similar heat treat to the cladding (or maybe try soft cladding and straightening with a brass hammer or something like that). Thanks everyone for the wealth of information on this stuff! I know a bit about metallurgy but the stuff I don't know could fill a warehouse.
 
You must log in or register to reply here.
Back
Top