Log in

No account? Create an account
Tanto heat treatment, round one - Doug Ayen's Blacksmithing Blog [entries|archive|friends|userinfo]
Doug Ayen

[ userinfo | livejournal userinfo ]
[ archive | journal archive ]

Tanto heat treatment, round one [Jul. 16th, 2003|12:12 am]
Doug Ayen
[mood |productive]

Yesterday, monday evening, I fired up the heat-treat oven and attempted to heat-treat the tantos. If you remember, I had clayed up the tantos beforehand, using furnace cement and baking it in my oven to set it. With five tantos , and a fairly good-sized heat treat oven, I thought heat treatment would be no problem.

I was a trifle optimistic, I think.

Now, according to the heat treat data, the opitmal temperature for the heat treatment of a 10xx steel such as I'm using should be between 1485 and 1550 deg. F. (yes, I double checked to make sure I'm using the right scale.) So, I heated the oven until it was above that range, let it soak for a good half hour, then cooled it down to 1550 and put in two tantos. I set a timer for 15 minutes and let the temperature fall down to around 1480 during that time.

Some notes about the oven: this is an ancient beast, basically a metal box lined with heavy duty furnace cement, with a blower-driven two-burner heating system. There's a raised platform above the burners (opposed design, one burner on the left and one on the right, one in front one in back) on which the work rests. The thermocouple is at the top of the box, enclosed in an iron pipe. The thermocouple element is a type K, moden design, hooked up to a digital readout. The oven originally had a much larger type K thermocouple, hooked up to an analog gage, but I prefer the digital readout as it's far more accurate and easy to calibrate. The regulator for the forge has two outputs -- one for a minimum output that results in the oven losing about one degree every two seconds, good for annealing and other slow cooldowns, and an auxiliary valve, that lets you really dump in the gas, which will heat up the box at the rate of about 300 deg/minute. This valve is a bit tricky to operate, as it's definitely non-linear in operation. You also have to fiddle with the air intake on the blower to adjust it to the gas flow you select.

Anyhow, after letting the tantos soak in what should have been the right temperature for about 20 minutes, I quenched one in a fast oil. It should have come out at about RC 65, but a quick file test showed that the metal hadn't hardened. I increased the temperature a bit, to 1600, and gave the 2nd tanto an additional half hour, but it didn't harden fully either.

Having failed on two, I had three more to go. I put them in the furnace, and changed the fast quenching oil to water. After a half-hour soak at 1600, I pulled one out and quenched it. Still too soft, though the very tip got hard. I increased the temperature to 1800, and pulled the fourth blade. More of the blade was hardened, but there were still spots on the blade, particularly near the tang, that were not hardened enough. I increased the temperature to 2000, and let the last blade sit for another half hour at this temperature. It was quenched, and fully hardened.

Now, obviously something is wrong here. On the one hand, this is the first time I've tried heat treating a japanese style blade in what I hope is a truly controlled environment -- it may be that with the extra insulation and thermal mass of the clay coating it just takes much longer, requires a higher temperature, or possibly my thermocouple setup is mis-reading the temperature somehow. Maybe with all that extra-thick, high density cement I need to pre-heat a lot longer (the outside of the oven wasn't even boiling hot after over 2 hours of operation!) in order to get the true temperature of the interior to match that of the relatively exposed, lightweight thermocouple. Still, if I have to heat each blade to the 2000 deg. mark and hold for a while, that works, though technically I'm probably getting a lot of grain growth and decarb. I'll clean off the four failed blades, re-clay, and heattreat them again.

This time maybe they'll all work.


[User Picture]From: perspicuity
2003-07-15 11:06 pm (UTC)
would one of those point and shoot remote heat sensors be useful?

i wonder too, what with geeks we are and know, if you can't figure out a way to measure the temp of a given blade inside it's little snausage encasing of clay. hmmm. yeah, i bet that the clay is acting to keep it cooler initially.

what would happen if a piece say 8" had 4" of clay on one side but none on the other in terms of when its quenched? assuming even and proper heating of the clayed side?

i was wondering if having bare metal on the handle end, that might be beyond the desired final length and removed, might conduct interior heat? that's probably not quite the point of the claying though.
(Reply) (Thread)
[User Picture]From: blackanvil
2003-07-16 05:05 pm (UTC)

clay hardening

Nope, those point and shoot things crap out at around 500 deg. F, at least those in the omega catalog that are in my pricerange do. They sense IR -- at these temperatures, there's visible light. If I had better color vision, I could even learn to gauge temperature by color like the traditional nihonto makers do.

I really don't care what the temperature of the metal inside the clay is at -- it's not supposed to get hard, that the's whole point. The exposed edge bit is the part that is supposed to get hard, and isn't at the temperatures that I think I'm getting.I'm guessing that the clay may just be taking much longer to heat up by holding it at temperature than I'm used to -- in the full blast of a 3000 deg. F forge, even the furnace cement heats up fast. I'm wondering if I need to leave the steel in the furnace for half an hour or more at temperature to get the whole works up to temp. We'll see. I'll sacrifice a blade to see what the grain looks like afterwards.

If you had a piece 8" long with 4" of clay on only one side, and quenched it at temperature, the blade would warp drastically, the clay would fall off, and you'd end up with a warped blade. trust me, I've had enough clay fall off before discovering the furnace cement trick.

the very end of the tang is, indeed, left exposed, as it is going to be drawn back fully soft anyway after initial heat treatment.

The point of the claying is to retain the heat of the heat-treatment long enough that the blade under the clay does not harden, and that there's a clear transition between the hardened and the unhardened steel -- the hamon. Hope that helps.

(Reply) (Parent) (Thread)
[User Picture]From: perspicuity
2003-07-16 05:15 pm (UTC)

Re: clay hardening

*smacks head* - i somehow misread what you were saying as totally encasing the blade! yeah, thinking about it, the point is to harden the edge, so that would be exposed.

oh, anywhere i said clay, i meant whatever material you were using: furnace cement in this case.

i ran into someone that did remote temperature sensing hardware at MIT and for various probes that people like NASA pick up. she was doing VERY high temp stuff. this was 5 years ago or so, and i was hoping the prices had come down a bunch. pity they haven't, just a thought.

well, at least one thing, if it takes longer with your setup than initially thought, and it's consistent: then win! figuring out why will be the fun challenge for later.

hamon on :)
(Reply) (Parent) (Thread)
[User Picture]From: madbodger
2003-07-16 03:14 pm (UTC)
Thermocouples are cheap -- maybe encase one in clay and watch the heating profile that way? I bodged together a logging system out of a multimeter with a serial output and a computer -- if your temperature gauge has an analog output, we could use that to graph the heat/cool behaviour.
(Reply) (Thread)
[User Picture]From: blackanvil
2003-07-16 05:07 pm (UTC)

clay hardening

Hmm, I got the cheapest omega handheld thermocouple readout, so I don't think it has an output. It's worth a try, I've got a couple extra thermocouples kicking around. It might even survive the process.

(Reply) (Parent) (Thread)