Log in

No account? Create an account
Pizza cutter heat treatment - Doug Ayen's Blacksmithing Blog [entries|archive|friends|userinfo]
Doug Ayen

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

Pizza cutter heat treatment [Dec. 4th, 2011|07:13 pm]
Doug Ayen
My back was doing better this weekend, thank you magnesium salicylate. I took the plunge and heat treated the pizza wheels that have been sitting around waiting for me to get my courage up. Three look like they succeeded, but one is definitely a casualty unless I can fix it up. More, including some pics, behind the cut

Friday night, I pulled the turned and ground discs from storage: one W2 disc nearly 4" across, one of case-hardened wrought iron (CHWI), one sound disc of pattern-welded O1 and L6, and one unsound of the same admixture. And I mean sound and unsound literally -- when tapped, the sound one rang like a bell, the other quickly dampened from some hidden flaw inside the blade.

After a thorough cleaning and degreasing, I got out my tub of furnace cement, and discovered it had set to a stiff heavy-clay like lump, but was still recoverable with the addition of water and agitation. After trying by hand a bit, I switched over to the Kitchenaid, which did the job then got a thorough cleaning. I rigged up some bamboo holders for the disks, used some of the scraps to make the tools, and made some art.

With the W2, I tried for a pinwheel effect, swirling out from the center with grooves in the cement accompanied by ashi in the gaps; the first of the damascus was a clover pattern; the other I'm not sure if I should call it "pizza" or "8-lobed jasmine," but you get the idea.

Pinwheel clay

Clover and the other one

The CHWI (pronounced "chewy") was a fairly unique situation. Cut and ground out of some 1/4" thick 300 year old wrought iron plate, claimed to be part of the armor of a colonial Rhode Island bank-vault, and pitted enough that I believe it, it was then put in a crucible with a carbon donor and heated until a surface skin of steel formed. In theory, at least. This is called blister steel, and except that in this case my crucible melted halfway through the process (oops), and apparently some of the stainless steel from the crucible welded to the disc, so I'm not sure how much if any penetration I have. On the other hand, no guts, no glory, so I just clayed a rough circle.

CHWI and pinwheel

The B sides

I'd meant to get up earlier and actually do more on Saturday, but ended up sleeping in until about 10. By the time I was really awake, fed, and caffeinated, I only had a few hours until I wanted to leave for a holiday party. And, as it turned out, it's still humid enough after the rains earlier this week that the discs hadn't dried out completely out yet. Into the oven while getting the rolls ready to rise for the party.

Oven off, wheels out, down to the shop, where I waste another hour getting the heat treat oven, the thermocouples, and the quench ready. Park50 is some of the best stuff you can use for a fast oil or slow water type quench, and I just got a 5 gallon pail in.

Heat treat info for the steels involved:

w2:1400-1450F for 15, quench in water (the "W" part of W2 stands for "water quenching") or fast oil

L6/O1: O1 1450F for 15, quench in oil; L6 1500 and air quenched. The best compromise I could think of is to quench from 1500F after 30 minutes of soak time. In retrospect, I probably should have been more conservative.

CHWI:Temperature 145F0 - 1460F is good, hold for 5 min, quench in Park 50, hold fo 4-5 seconds, flash temper at 350F, temper @400F. Ok, I'm pretty much guessing here, this is the best formula I've seen for straight iron/carbon low alloy steels, but with unknown penetration of unknown levels of carbon into lightly-forged slag-riddled

So, my plan A was to heat to 1450 and hold 15 minutes. quench the W2 and the CHWI, then raise the heat to 1500F and hold 30 minutes. quench damascus and bring in to temper at 400. I'd skip the "flash temper" as that really requires a molten salt pot furnace, and mine's out of commission.

As it turns out, the new, larger, fancier thermocouples I'd bought, clearly marked "Type K" , were registering temperatures in the 90s even when glowing slightly, though the cheap one worked fine, so I went with that, and will troubleshoot the problem later. Those were expensive, dammit.

Still, I was able to achieve, by dancing back and forth between the gas pressure valve and the air intake, a reducing atmosphere at 1450F for 15 minutes. The W2 and CHWI came out and into the oil. The clay came right off, except for a few small patches. Up to 1550, and again with the dance.

The dance was necessary because, like a lot of my equipment, my heat treat oven is ancient. The previous owner didn't have a firm date, but said he thought it was 1930s, the current blower being a 1950's era replacement. It's big, heavy lump of cast iron and asbestos, and I love it. I might modernize it at some point, and already have by replacing the inaccurate analog thermocouple system with an inaccurate digital thermocouple system.

I say inaccurate because when the 15 minutes were almost up, I was moving the thermocouple DRO and the temperature suddenly jumped from 1550 to 1680 and rising. I quickly killed the burner before it could rise any further -- this was getting way too hot too fast. The first damascus disc, the one with the clover leaf pattern, went into the oil. After a couple of seconds I heard the noise most hateful to the knifemaker's ear: TINK!

The second one survived the quench without issue, but the first needed some further examination.After tempering (lesson learned the hard way: always always always temper immediately after hardening), I took a good hard look. The hidden flaw I'd suspected was there earlier had caused the blade to split, was clearly present as a jagged split on the bevel of the blade, next to which the edge was bent into a shallow U.

The Crack (polished with some sandpaper to bring out the contrast)

The Bend

Some careful hammering, first with a wooden mallet on a hardwood board, then with a ball-pein hammer on the anvil directly, showed mixed news. On the bad side, yes this blade was ruined, unusable as it was -- not only was the bend just popping back in place when stuck, but the delamination was clearly quite deep, both showing considerable movement when stuck and having caused a considerable bulge and cracks on the other side of the blade. On the other hand, this split and bend proved that the other blades were almost certainly properly hardened.

Why did the blade delaminate? By the look of it, nearly 1mm worth of steel was torn along a 4cm length. Even at the temperatures involved, it takes an almost absurd amount of tension energy to tear steel jagged like that. Looking at the other side of the blade, there's a small bulge right under the tear, as well as what may be a hairline crack right at the base of the U-bend.

During a differential quench like this, grossly oversimplifying things, the unclayed steel is rapidly cooled and hardened. It's set hard long before the clay-covered portion could even start cooling -- that's the point of the clay, to delay and prolong the cooling so the covered metal doesn't harden. It doesn't harden, and as it cools, it shrinks, by as much as 10% in volume compared to the hardened material. On the other wheels, the steel held it's shape because, being a disc, the edges were supported evenly all the way to the center holes on both sides, so it couldn't move out of position, even when the core tried to shrink.

On the failed piece, it was another story. The edge hardened, and a skin of hardened steel formed on the exposed surfaces, merging with and connected to the central core. Then the time has passed: the core could no longer harden. The edge, cooled, tries to shrink; the still malleable core obliges,then cools and shrinks itself, putting tension on the edge. On a katana, the curve of the blade is traditionally entirely due to this process, as the edge freezes and the back shrinks, but this is a circle. The center pulls, the edge resists, and something gives -- the flaw focuses all that force into a stress riser until the steel tears, the crack propagating out to the surface. The edge suddenly isn't balanced, so bends in the direction of the uncracked side. The bulge is from where the fault displaced on the other side of the blade. The smaller cracks are probably from the flexing and bending during the quench, when the steel was fully hard and brittle, and not yet tempered, though my wacking it later with a ball-pein might have done some damage as well.

Crack, 150X

Crack, edge on

The best idea I have for fixing it it to anneal it, straighten it, TIG weld the cold shut, and just put it through a gentler, non-clayed heat treatment. Another option would be to anneal, then grind down past the flaw, but I think it's too deep for that to work. It will never be a salable object, but even having a ruined piece to practice and experiment on would be useful.

Next step: grind off the clay residue and any oxides, and polish.

Other projects:

--The replacement vacuum pump arrived, just need to assemble the new chamber and make some bacon-micarta.

--I'm waiting on a shipment of a specialized titanium etchant before I anodize the next set of blades. If that doesn't help, I may have to build a much bigger anodizer. Not happy with the colors. Sorry, Mark, it's going to take a bit longer.

Edit 11:36pm: Almost forgot to include the post-heat-treatment photos:


[User Picture]From: blackanvil
2011-12-05 03:24 am (UTC)
I know this is a long post, but insomnia stuck last night.

If you have any questions, please ask!
(Reply) (Thread)