Kiko Denzer  @  ChelseaGreen

View All of Kiko Denzer's Posts

Jumping bricks, or: inside out oven building

Posted on Wednesday, June 3rd, 2009 at 11:53 am by Kiko Denzer

I built this oven for a local CSA farmstand restaurant (gathering together farm). We held a public workshop; folks came to make mud and learn and we built the basic oven in a weekend. BUT! (and this was my fault for not watching more closely), the dome came out a little flat. Usually, when it's not quite right, I tell folks, "OK, time to tear down and rebuild." This is a great way to conquer the fear of doing it wrong—

and it's the only way to prove to folks the truth of my favorite oven-building adage: "the second time is easier and faster." But I let myself be convinced that the dome was adequately curved. It wasn't. A year later, it was bulging downward, at the rate of about one eighth of an inch per month. Collapse was inevitable.

"Well," you must be thinking, "didn't you say the second time is easier and faster!"

Yes, well, OK, but…even had I been able to halve the time for building the oven and the large, sculpted shell containing the loose perlite insulation, I would have needed at least a week, not including the time required to dry out the damp oven itself. The restaurant uses the oven 5 days a week, and I didn't want to be responsible for two weeks of down time—

nor did I want to put in that much (unpaid) time myself.

So the need for speed provided a perfect opportunity to try something I'd been wanting to try: making a brick dome using thin mud bricks laid on an angled bed of mortar, with successive courses in a series of shrinking circles to close in the dome—

without formwork! It's traditional for ovens and vaulted or domed ceilings. I'd seen ceilings done this way in Mexico and wondered, "how did they do that!?" as well as photos of German mud-brick ovens made the same way.

Now I can say from experience that it works! And it makes me appreciate bricks. They are pre-dried, pre-shrunk, easy to work with, and quick. Since they are relatively small, they shrink without cracking, which means you can use mud with a high percentage of clay. Clay holds up better to the thermal demands of an oven than a typically sandy cob mix, which relies on lots of sand to limit shrink. And if you're working in a situation where time = money—that is, if you're building an oven for someone else—you can make bricks at your leisure, and store them for when you need them. Mud has higher value when it's made into a brick, so you can charge a unit price for each nicely squared blob of mud, but since bricks make for a quick build, you don't have to spend so much time on the site building, drying out the oven, etc. (Here starts the slippery slope of professionalism, which is, as Collette said of writing, much like sex: first you do it for love; then you do it as a favor just for friends; then you do it for money. And then you set up guilds and unions, a licensing board with bonding and contract requirements with related insurance and legal industries, then lobbyists and trade agreements and complete control—

which spawns renegade activities like how-to books for home-bakers and oven builders, backyard mud ovens, internet groups, and here we are!)

note the cracking bulge at the top center of the oven ceiling. I cut a story stick that just fit under the bulge without scraping, to keep track of movement. After about a month, I could no longer slide the stick through without scraping. I figured the bulge had dropped another 1/8-3/16 of an inch. Fortunately, our raw clay-sand bricks had dried by then.

Lisa, one of the cooks, helps make bricks—

here she wets the brick form (1.5" deep: scrap 2×2 — thin bricks are best for domework, as they make better curves, and their lighter weight means less risk of slippage when you're setting them).

Next, she sprinkles sawdust (we didn't have any sand), so the brick will slip off the board.

Throw two handfuls of mud into the form; if you throw it right, the corners come out clean and sharp.

smooth the top surface…

we learned a lot from watching Caramelo, from Oaxaca, an expert adobe-maker… I didn't get his picture, unfortunately…

pull the form: the brick is ready to slide off onto a flat surface for drying.

here's our stack of about 220 bricks, with form.

in mexico, they say you can tell the top of the adobe by the dog prints…

The outer surface of the over is a thin shell containing perlite insulation. I cut a hole in the bottom side to drain out the perlite. I used a piece of old roofing tin to make a chute to direct the flow of perlite. It poured out like water. Nasty dust. A good mask is well worth the money.

here you can see the layers, from outside to inside: colored final clay-sand plaster, insulating sawdust-clay mix, and gypsum-impregnated burlap

a Sawzall came in very handy to cut a hole in the top, and eventually, to cut out a large, wedge-shaped piece of the entire shell. There was virtually no cracking as a result of any cutting. The shell was super-strong and rigid. Note the bamboo armature underneath.

perlite inside the shell, avalanching down to the drain hole…

Lisa peering into the emptied out shell…

good insulation! The bamboo, which was just temporary formwork for the fireproof plaster shell while it dried, is still completely intact: not even charred!

The bamboo supports a layer of burlap impregnated with clay slip and gypsum plaster for a quick-setting, stiff, and fireproof surface that could be plastered immediately.

even the jute twine that I'd used to tie the bamboo hadn't charred—

indeed, it was barely toasted! I would guess that the temperature at that level barely got over 200 degrees F.

In cross-section, you can see the layers of the fire-proof shell that holds the insulation. From top to bottom: white is gypsum and clay impregnated burlap with some sandy-clay plaster, then a layer of insulating and sculptable plaster made of clay and sawdust, then a layer of fine clay-sand finish plaster. 

The surface at the bottom of the photo is the top of the oven dome, made of pure sand and clay. Above that, looking like the edge of a cliff, is a layer of sawdust mixed with clay slip. This close to the heat, the sawdust gets hot enough to completely burn out, leaving an insulative clay foam. It works well, but crumbles at a touch. The white above is perlite, which was pretty well contaminated with the crumbs of crumbled clay-foam, so we had to buy new when we refilled the insulation cavity.

The thermal layer comes down. Note thickness of shell.

The dense thermal layer was not hugely massive because the oven is used primarily for pizza and just a bit of bread. However, with 6-10 inches of perlite all around, it held heat extremely well, and performed beautifully.

Thickness ranged from a bit more than three inches to almost five inches.

Thermal layer completely cleaned out. Note the bottles in perlite exposed around the edges of the floor bricks. The bricks on edge around the perimeter are a "bumper course" to protect the softer mud dome from peels and firewood.

Front arch

the hole in the shell was just big enough for me to step into the oven and sit on the floor. I'm very glad I didn't have to wiggle through the door in order to work!

first courses of bricks, showing the mortar wedge that sets the angle of the dome

a wedge of bricks at the corner nearest the door helps define the line from the (low) front end of the oven to the higher rear end…

story stick shows the 16" target height for the rear of the dome: I angled the bricks to meet the top of the story stick about 6-8" off the back of the oven…

the mortar is just clay and sand, like the bricks. Building an oven with bricks this way means that you can use more clay in the mix; because the bricks are relatively thin and small, they shrink without cracking. The clay is generally more durable under constant use than a typical "cob" mix, which uses a higher percentage of sand to limit cracking.

at work: I was very glad when I realized I could just cut out a big doorway right through the shell. Originally I'd thought I would have to wiggle through the door and work lying down!

the walls start to close in; each course describes a slightly smaller circle and tilts at a steeper angle

I'd only ever seen pictures of this, so was both pleased and amazed at how well the mortar held the bricks against the force of gravity

getting tighter!

I had to clean the bottoms of the mortar joints by feel because I couldn't get low enough to see 'em!

I tried to work as symmetrically as possible, but it's definitely very different than laying up a straight wall.

toward the end, I had to enlarge the opening of the shell just to be able to maneuver.

I used the spray bottle to wet out dry surfaces to take the mortar better.

the dome closing in to the final keystone courses…

these last bricks were almost vertical, but were also held in place by being wedge-shaped. Since the bricks were raw (unfired) they were easy to cut with an old Sawzall blade.

final courses close the dome.

note the angled courses shaping the "throat" leading to the door opening

hard to clean the insides of the final mortar joints; I made a long handled tool that reached through the doorway to scrape the joints clean.


patching the shell was a matter of splitting some new bamboo and wedging it into place

not much structure is needed to support the burlap and gypsum plaster, which is self-supporting once it sets. However, I did "tie" the main bamboo struts with "straps" of gypsum-soaked burlap that wrapped around the bamboo and over the outside of the shell. These got trimmed after the gypsum set.

from burlap to finish plaster was a matter of an hour or so…

I started a drying fire immediately (using the old bits of bamboo for kindling—

nice and dry!)

almost ready for pizza!


Leave a Reply