Armour Forge
A Sheet-Metal Heating Forge

For Armour and Repousse’
Copyright Eric G. Thing, 2007

When I started making armor as a hobby some 20 years ago, I was faced with a problem.

I wanted to make deep-drawn shapes, particularly helmets, without modern welding techniques. For forming small or shallow pieces like breastplates, knee-cops, and so forth, working with cold sheet steel was not much of a burden. Starting material was usually .080″ at thickest, and often much thinner. Cold-working techniques used by sculptors of non-ferrous stock (such as silver or copper) worked all right – although scaling the tools up in weight by a factor of two or three really helped!

The noise was really irritating, though, especially when working against metal forming tools. Steel gives bright, high-pitched bangs when hit cold, especially after it work hardens a bit. I didn’t like the repeated shocks to my hands and arms, either, when forming the heavier gauges.

It was helmet-making that finally defeated me. I made many helmets by forming heavy-gauge (14 to 12) halves in a metal die, then welding the halves together. I hated it. The trimming, fitting, clamping, welding, and cleanup were processes that (I thought) weren’t teaching me anything I wanted to know, and were boring to boot. I wanted to make helmet skulls out of one-piece blanks, with no weld seams – the grand old way. But a couple of attempts at sinking and raising large pieces of 12 gauge plate utterly vanquished me. I simply did not have the strength or endurance!

The solution had to be the use of heat. But was this “period”? Back in the 1980’s, when I started out, most literature on the subject of actual armor fabrication did not deal much with the use of heat during forming operations. Brian Flax, in his SCA periodical “The Hammer”, published in the early 80’s, dealt with raising as a helmet-making technique, but his experiments were done with cold metal, although frequently annealed in a forge. Mainstream books published on arms and armor usually described manufacturing operations only in the broadest, briefest terms (e.g., “metal was cut from plates, then hammered over stakes”, etc).

Then one day I talked to Robert MacPherson, one of the most accomplished armorers on the planet (and an excellent fellow, by the way); he heard out my troubles, and explained to me that cold-raising heavy steel plate was “an act of heroic madness”. He used heat. That was good enough for me!

So, I tried some sinking and raising of 14 gauge plate using a coal forge that I had built some years previously for making tools. I had some success; I actually managed to roughly raise a small sallet using the forge and an improvised stake made from a torch-cut piece of railroad rail. The awkwardness of constantly tending the fire, dealing with bursts of smoke, and the fact that I was working in my carport and annoying my neighbors, convinced me to find another way.

What about a propane forge? Mr. MacPherson (as I recall) used an oxy-fuel torch to heat plate for raising. An oxy-acetylene rosebud heating tip can indeed bring a good area of steel plate up to workable temperature fast, but I wanted to try a different approach: make a propane-air rig for the purpose. The fuel would be relatively cheap, and if properly designed, such a forge could hold the work while heating it.

So, around 1995 (I can’t remember the exact date), I set about it.

To start, some obvious parameters. First: Helmets start out as disks. Well, to be more precise, their starting blanks are based on disks; they fit egg-shaped human heads, so they have to finish approximately as domes or cylinders. Simple skull-type helmets (Norman, say) start as disks of 15″ in diameter, or so. The biggest helmet blank I ever could conceive of using was about 22″ in diameter. Therefore, I needed a forge with a “throat” of about 12″; if it could heat the center of a 24″ diameter blank, I figured I would be covered. Even large breastplates could be accommodated.

Next parameter: size of the heated zone. In theory, you can build a huge furnace and heat the entire blank for every “stanza”, as the silversmiths call it. Not desirable. Enormous waste of fuel, heating areas that don’t get worked. Great loss to firescale. Handling a big, totally orange-hot sheet. If I were feeding the blank into a giant toggle press, a la 1930’s auto industry, heating the whole thing might make sense; but alas, I can only hammer a small area of the blank before it needs reheating. So, I wanted to heat an area of about 5″ square. I estimated that was about the limit I could work during a useable heat.

Next: The forge has to be able to keep heating the piece as it changes shape! Most helmets start as a flat disk, teardrop, oval, etc; then they become deeper and narrower bowls, and (sometimes) can become nearly a foot deep, and less than 7″ wide in some internal dimensions.

From 1995 to about 1999 I messed around with various configurations. I had some embarrassing failures, but around the year 2000 I came up with a design that I have now been using for seven years with very minor modifications.

I’ll skip the trauma of the formative years; here are photos of my current forge.
Nasty-looking thing, really, but I am fairly happy with it. There are about a dozen refinements that I would like to install, but I feel no great urge for another “Mark” in design; this basic configuration does what I need. Most of the time.

Forge Design

The forge consists of the following units: Frame, chamber, deck (or stage), and burner. The gas hoses, valves, regulator, tank and so on are all easily available commercial products, and can be found in just about any large town in the USA.

Frame:

Most of the frame is 1″ square tubing, 14 gauge wall. It is cheap, easy to work, and fairly strong. If I ever build another forge, I will go up one size, to 1.25″ tubing, just for the extra strength and mass the bigger stuff affords.

The base of the frame is a welded-up table. I just wanted four legs for the forge to rest on. The one really essential criterion: the base must be long enough to keep the forge from tipping over when a heavy work piece is put on the deck! My base is about as long as the chamber support arm. Note the locking casters. I wanted to be able to roll the forge around the shop, and lock it in place. It turned out that the casters were basically useless; I never move the forge more than a foot or so in the shop. No wheels, next time. But by happy accident, the base gives me shelf space to put forge bits on: the coiled main gas hose, wrenches, and so forth.

I bolted the main upright of the forge to the back of the base, rather than welding it, to allow disassembly. Naturally, those bolts have been undisturbed for seven years!
Closer look at the upper works, including the chamber support arm, and the deck support arms. The chamber arm is welded to the upright, and a diagonal brace is welded in, for a sort of gallows-look. All 1″ tubing, but very strong. The arm is pierced in the middle to take the deck support adjustment screw. The deck support is the old classic parallelogram structure. This configuration has the huge advantage of keeping the deck horizontal to the floor, at whatever height you set it. (I pay for this, of course, by having the deck move slightly horizontally, retracting and extending, as it is lowered and raised. The distance is small, and doesn’t affect forge operation.) The top arm of the support is pierced vertically in the middle for the adjustment screw, and is pierced horizontally for the pin that takes the screw and holds up the deck support.
At the end of the chamber arm is a 1.25″ socket to receive the chamber’s mounting fixture. The socket is 1/8″ wall, so it receives 1″ tubing very snugly.
At the end of the deck support is another 1.25″ socket, to receive a 1″ shank on the deck. The deck support socket is a bit funky in shape: it is displaced downward to adjust for the thickness of the deck. Note how it is pinned to the two arms. I use 1/4″ clevis pins on both ends of the support.
Closer look at the installation of the pivot pin, which is technically a nut. It’s a piece of 1/2″ rod, drilled and tapped for the 3/8″-13 threaded rod that makes up the adjustment screw. (Yes, that’s right — a 3/8″ hole through 1/2″ rod. This leaves darn little meat on each side of the hole, but the pivot has held up well for years. I would like a stronger one, though.) It pivots in two 1/2″ holes drilled in the sides of the top deck support arm. Slots had to be cut in the top and bottom wall of the arm to allow the screw some play, as the support changes angle when it is raised and lowered.

Side note: The mechanically adept among you may shiver a bit when you see how often I pierced these frame members with holes that decrease their wall strength. Mea culpa. This is the main reason why I will eventually rebuild this forge with bigger, thicker tubing. The frame has held up fine for years, but those holes do bother me. This is especially true around the pivot pin; this is the weakest area of the frame.

Naturally, this design limits the travel of the deck support. It can only go up until the deck hits the bottom of the chamber (no great reason for it to go much higher than that), or the screw hits the bottom arm (unless you slotted the bottom arm, of course). It can only go down until you run out of screw length.

If I rebuild, I will probably keep most of the dimensions, or maybe enlarge them 10-15%; this setup has worked well for me. But I would almost certainly beef up the pivot pin section. I might go to a 3/4″ pin, and maybe to a 7/16″ or 1/2″ screw. I might also mount the pin between ears outside the top arm, thus getting rid of the side holes.

The top-crank design for raising and lowering the deck support seems awkward; you have to reach around the (hot!) chamber to work it. Curiously, though, I got used to this fairly quickly, and I don’t really mind it. It was very simple to build.

One smith of my acquaintance built a version of this forge and installed a foot-operated lever arrangement to raise/lower the deck support from below. Very ingenious, but he is very mechanically skilled, and I am definitely not. I will probably stick to the hand-crank.

Now, why the sockets to hold the chamber and the deck? Well, that suggestion was actually made by John Segura, a member of the Arizona Artist Blacksmith Assocation (AABA), our chapter of ABANA. In an earler model, I bolted them onto the arms. He pointed out that with sockets, you could very quickly change chambers and decks, if you had different sizes of them. That turned out to be a great idea, as will be shown later.

Deck:

More properly, decks; plural. I have two. The one on the left is 8″ wide, the other is 6″ wide. I start blanks on the big one; when the helmets (or other pieces) get narrow, I switch to the little one.

Side view of the smaller deck. A shank of 1″ tubing is welded to the back, to fit into the support socket. It has a spacer to make the deck line up better with the chamber base.

These decks were built up and somewhat modified over time. I’ve settled on a basic design. I bend 1/8″ metal strap into a U-shape of width I want. I weld another straight piece of same metal across open end of U, giving a D-shape. I then weld on a sheet metal bottom, 12 or 14 gauge; doesn’t much matter. The depth of the deck is determined by the refractories I have on hand. In the bottom of the deck, I put insulating refractory, like soft kiln brick. On top of that, I put a slab of hard alumina kiln shelving. Both the insulating layer and top layer must be sawn to fit the deck, fitting very loosely, since they will expand when highly heated. So, ideally, the height of the deck walls is the sum of the two refractory layer thicknesses.

As you can see in the photos above, neither of these decks are ideal height; the 8″ one had to have pieces spliced on to hold the kiln shelf, and the 6″ one was a tad too high. Oh, well. They work fine.

It’s easy to saw soft insulating brick; I just ruin a cheap hacksaw blade. I cut kiln shelf with a wet diamond tile saw, which I rent from a big box store.

The hard deck floor absorbs a lot of heat when the forge operates, especially when I am working on the piece and the forge is empty. When the piece is put back in, the hot floor helps reheat it. Of course, this means the floor goes through big temperature cycles many hundreds of times; eventually, it develops at least one big hairline crack. But that doesn’t really affect performance.

Chamber

This was the most trying part of the design. The chamber I use today is the fifth or sixth one I’ve tried. I’m fairly happy with it, although I may try a slightly larger one in the future.

Basic design: A truncated cone (or a frustrum — one of my favorite words!) I lay out and roll up a truncated cone of 14 gauge steel: 8″ diameter at base, 5″ diameter at top, 8.5″ high. I weld up the seam. I cut out a disk of 12 gauge to fit top, then punch out the hole in the top, just a teeny bit (1-2 mm) larger in diameter than the nozzle of the burner I plan to use. I weld the disk to the top, making sure hole is well centered. I drill and tap three holes around the base, to fit small machine screws (yes, tapping 14-gauge metal is a bit weird, but it works, if threads are fine enough – I think mine are 10-24). I cut out a ring of fairly heavy gauge stainless steel to fit the bottom of the chamber; 8″ OD, 6″ ID. This ring holds in and protects the insulation I later install.

Cone Layout
I weld on a T-shaped fixture to the side of the chamber, made of 1″ square tubing and 1″x2″ tubing. It’s angled, as can be seen, so that the top of the “T” is vertical. Why a T-shaped fixture? Well, this means I can mount the chamber in the deck support socket upside down, if I want a bottom-fired work session, for some reason. I’ve used this feature several times; I need it rarely, but sometimes I really need it.

A burner holder is welded to the top of the chamber. I consists of a right angle (cut from large angle iron), to which is bolted a short piece of 1″ wide, 1/2″ deep channel iron. The channel iron cradles the burner nozzle firmly; I make it fast with a small stainless steel band clamp.

To insulate the inside: I use 1″ thick ceramic wool (Kaowool). I lay out and cut a truncated cone of wool, slightly smaller in radius than the chamber, and 1″ shorter. I cut a disk of wool to fit the roof of the chamber, and cut a hole slightly larger than my burner nozzle in this disk. I keep the wool misted with water to cut down particle shedding while cutting. I put in the wool ceiling, then the wool cone (the ceiling is why the wool cone is cut 1″ short, of course). I spread the hole in the roof so it flares very slightly going into the chamber.

I coat the whole inside with ITC-100 ceramic paint. I love the stuff. I just dab it on with a cheap brush. The hole in the roof gets heavily coated; it will be the flame entrance.

The stainless steel ring goes on loosely, held by three stainless clips, which in turn are held to the chamber outer wall by three very short stainless machine screws. They’re short, so they don’t go into the insulation wall any great distance.

I mount the burner so that only 2-3 millimeters of the nozzle pokes into the chamber; barely passing through the metal, into the hole in the ceramic wool ceiling. The slightly flared hole in the wool functions as a sort of nozzle extension; the flame seems to propagate well. This is why the ceiling hole must be heavily coated with ceramic paint. One benefit of this is that the burner nozzle functions for hundreds of hours with almost no erosion.

As noted above, my current chamber is 8″ in diameter at the base. With 1″ wool insulation, that means the ID is 6″ at the base. I figured this would heat a round spot of 4″ to 5″ in diameter fairly quickly, and so it does. For some operations, like massive sinking, I would like a bigger zone, so I may build a 10″ OD chamber someday, which will probably heat a 7″ zone.

The chamber being 8.5″ high, it seems a rather skinny shape. I’ve found this half-angle to work well; the flame seems to propagate and “wash” the inside of the chamber, and get the inside up to incandescence in a few minutes. Shorter, squatter chambers did not seem to work as well. If I build that 10″ OD chamber, I will stretch the height accordingly to keep the same proportions as my current 8″ chamber.

Burner:

I use a 3/4″ T-Rex Burner, purchased from hybridburners.com. I’ve made a few of the Ron Reil design burners, and they functioned fairly well; but getting a professionally-made burner upped the performance of the forge markedly. In the future, I will probably buy more burners of the T-Rex line, rather than making my own.

Operation

I adjust the deck so that about a 3/4″ gap shows all around the base of the chamber. This slight restriction helps ignition and keep combustion steady during warmup. I open the choke on the burner, then feed it gas while holding a gooseneck spark igniter in the chamber exit, sparking away. The forge catches immediately. I let it run for 2-3 minutes at least, to heat the walls and floor up a bit.
A couple of views, after the forge reaches full heat. Input gas pressure is about 8 lbs. The deck and walls get to a good orange, and mild steel plate reaches working heat fairly quickly.
Cooking a piece of plate. It’s great to have my hands free while the work is heating! I can drink water, do stretching exercises, fuss with my tools, stand under my shop air blast (this is Arizona, so my shop has a 600 CFM evaporative cooler that I run full tilt while the forge is on). The importance of the stainless ring at the base is apparent; it protects the insulation from abrasion by the work.
As the work progresses, the deck still holds the bowl fairly well. The bowl is actually lightly clamped by the deck and chamber bottom ring. There is an awkward phase, during the making of the average helmet skull, when the bowl is 70-80% formed; it has a tendency to rool and slide right out of the deck. This could be fixed with an extendable rack on the deck; I’ll make it someday, but for now I just hold the work in with tongs when it gets contrary. A bit hot and tiring, but the awkward phase does not last long.

This was my solution to the sheet metal heating problem. One thing I must emphasize: This forge is a specialized tool. It works well for heating plate; but for almost any other purpose, it is very inefficient. It is definitely not a general purpose design for blacksmithing! The 360 degree access, and its very “openness”, means that the chamber never gets super hot. A horizontal enclosed gas forge of good design will get much hotter, even reaching welding heat with sufficient gas pressure.