Crazy Cold Climate Cob Vault Idea: Shred This!

In my youth, I had a passion for cob, but then I moved to a cold climate with unsuitable soils and so I built my house as a timber frame instead. Recently I have discovered a local source of clay, and also heard of a foamglass factory opening nearby (man-made pumice from recycled bottles, basically; we are too far from volcanic areas to reasonably ship natural pumice). These have given me the itch to try something muddy after all.

What’s really prodding me is a longstanding fascination with intersectional vaults: groin vaults in their simplest form or, for the true madmen, star vaults. And I fell down a bit of a rabbit hole with the flying concrete guy in Mexico, or Gustavino’s timbrel vaulting, or … But of course there are a thousand and one reasons NOT to think that’s a good idea with cob; in my experience, it will only corbel just so far under its own slumping weight, and of course, a soggy cob vault would be catastrophic, and then you’re planning to waterproof it … how? But still, I played around with the form, making one out of some cardboard, and sat there and stared at it for a while, and maybe just thought up a whole new building technique. So tell me if I’m insane:

You’re going for a hexagonal star vault, a la Constable’s Chapel, only less churchy and ornate.
Stripped to its simple geometry, it would look more like the following pics.

So say you build 6 arch-shaped solid cob walls. So far, simple enough. (I’m leaning towards catenary both for aesthetics and to avoid excessive corner buttressing, but if you like buttressing, I suppose you could do Roman. You’d have to find a way to keep those buttresses out of the rain.) As they go up, you bed lots and lots and lots of tie-downs all along the upper edge. Possibly you’d tie them to a small buried form jigsawed from heavy plywood. (I’m trying to avoid rebar here, but that would work as well.) I know more than enough farmers to have an unlimited supply of short lengths of used plastic baling twine. Stuff is wicked strong and lasts … well, until the end times and beyond, probably. Might as well stick it in the wall as in the landfill. You use these to tie sticks of wood horizontally, perpendicular to said arch-shaped wall. So you’ve got two pieces of wood coming inward from two different arches, and just before they meet you add a third cross-wise between them to make the beginning of the outer star. Is that making any sense? The thing would look like a vault-shaped log cabin, or perhaps like a hogan: a lot of intersecting horizontal bits of wood stacked on top each other.  Only you probably wouldn’t fool with notching a tight fit on each because you would want to leave wide gaps between. While you might tie all those intersecting ends together with said plastic twine, you’d be counting on a sticky cob mix mashed onto it from both sides, thumbed both down into the layer below and horizontally into itself from either side through the gaps, to cement it in place. It’d look a bit like the old mud-and-log chimneys they used to do in the South, only a complex semi-dome of intersecting curved planes. The wood would bear most of the initial weight and keep the cantilever rigid; the mud, once dry, would turn that pile of sticks from a tottery stack of dominos into a unified matrix.

The central star itself would be wood framed as something like a compression ring, also with rigid spokes from the star points bearing onto the arches of the walls, just to help out. It would wind up sitting completely flat. Given the complexities of a star vault, we’d be talking very short spans here: a 12’ interior diameter room wouldn’t have any members longer than 3’. Also, they’d be bearing in universal compression rather than shear, I think. So they really could be sticks. Like, they could be the leftover limbs that are too small to efficiently buck up for firewood. Or they could be chunks of weathered old two-by salvaged from the collapsed barn, because no single stick would be bearing significant weight, but the whole ought to be quite strong. Say a 12’ diameter room would need about 600 little roundwood billets 3” in diameter. Like I said, I don’t think I’d fool with shaping them to fit at the corners like a proper log cabin. I think I’d just tie them together with baling twine and bury them in cob. But they’d be rigid enough to let the cob go flying out in cantilever more than it naturally likes to do.

At 5’ high or so you’d bed stubby eave extensions at each of the 6 corners and then above that the horizontal sticks would start also extending 2’ to the outside of the wall. So that would create an undulating eave. Once you had the primary frame mudded into place with a fairly sticky mix, you’d add a straw-rich layer over the whole, plus maybe a thick layer of perlite plaster for insulation, creating a self-supporting dome-like structure that, as a fail-safe, would still have the buried interior frame. (This would be handy if, god forbid, it ever got wet.) Then you’d slap a sheet of EPDM on top and maybe plexiglass over the star for a skylight, voila!

At this point a problem becomes obvious: That EPDM ain’t gonna breathe. Which means about half the mass of your cob building can’t breathe on its outer side. Now it’s still open on the inside, and in a dry climate with the mass all very well cured before you covered it and no interior sources of moisture (shower, cooking, etc.), maaaybe you’d get away with it. However, a dry climate we have not got. In fact, winter here is brutal, and if this is a heated space, there would be quite a temperature gradient from inside → out. Insulated or not, you know there’s going to be a dewpoint in there somewhere!

So here’s the crazy part: You bed over the whole vault in 4” of foamglass or pumice, with a network of perforated pipes running through and possibly even little pancake fans on the ends wired to a switch you can flip during our rare occasions of dry weather, putting positive pressure on the system. Even without the fan, there should be some passive air exchange. THEN you bury the whole array in waterproof rubber, with your pipe ends sticking out under the eaves. Sprinkle 4 more inches of foamglass gravel on top and plant your living roof.
*
Weigh in, folks! What am I missing here?
Obviously you’d have to do the initial build under a tent, because it couldn’t get wet until it were fully roofed over. I don’t know how long a thick cob would take to cure in our cool damp summers even so, and it would need to have the membrane on and the heat cranked up before the freeze-thaw cycles of winter hit. I can’t imagine damp cob stands those well at all.

I expect you’d get shrinkage around every obnoxious little stick, the way you do with “cordwood masonry,” but that would be buried entirely inside the wall, so maybe it wouldn’t be a problem. Or maybe it would. It might be beastly to keep the inner layer of cob/plaster stuck where it belongs as it starts to cantilever out overhead, especially if it’s pulling away from the sticks. You’d really be counting on the couple-inch gap between the “logs” to join the inner cob and the outer cob into each other as one piece. You’d probably need a fairly particular maximum fiber / minimum shrinkage recipe to make it adhere around all those interruptions. Anybody got one? (Did I mention I also have an unlimited supply of cow poo? But that breaks down over time, no?) This is the sort of challenge someone with good wattle-and-daub experience has probably solved, only I’m trying to get a bearing mix, not just a plaster.

You’d live in dread of roof leaks, and ambient moisture could still be trouble. At present I’m thinking of a bedroom add-on only, well separated from kitchen and bath, and I heat with wood, so my interior air in winter is nose-prickling dry. I’m wondering if that pipe-punctured air exchange layer, insulative though foamglass/pumice may be, would also let in all the cold?

Maybe not if a thick perlite layer was beneath it, but most perlite recipes I see call for waterglass to harden it, which is itself waterproofing? Anybody have any good experience with super-thick exterior perlite-slip or perlite-waterglass (or perlite-waterglass-slip) plasters? Breathability? Workability? Adhesion?

And then there’s whatever I haven’t considered, but which I’m sure you’ll tell me about….

You may want to look into AIRCRETE.

Thank you, Lorinne.
Yes, Aircrete and other foamcretes are interesting chemistry! They do make a nice balance of lightness and strength.
For this project, however, I'm trying to avoid cement entirely, except perhaps as a bond beam at the base.

Foamglass? Ah, FoamglasTM. Fascinating product. This is the first I have heard of it.

That is a beautiful design, April, even just built out of cardboard! In my experience, though, setting out to build a masterpiece on your first try rarely works out the way you planned. Now that you have access to some clay, (and hopefully a lot of time) maybe try something much simpler to get more experience with the material? The extent of my cob experience is tiny little ovens and the like, but man oh man, it is a lot of work to mix all that mud!

Building a cob house would be major undertaking - building a cob cathedral - even a small one, is on a whole other level. Maybe you could try building an ornate pizza oven with some interesting arches as a place to start?

I would like to try and help but I just cannot read your entry, too many underlines, sentences strrating in the middle of a line and generally very hard to comprehend.
Would it be possible to clear it up please?

Mr. Nystrom,
Thank you, but in fact I do have quite a bit of cob experience and 20+ years of general construction experience. I have worked on three cob houses and a greenhouse, and did not give the technique up until I moved to this area, where good clay deposits are rare. I believe my design respects the limitations of the material; however, if you see a specific problem with the design itself, I'd be happy to hear about that.

Mr. Daley,
Huh. I did not add any of those underlines; I presume they were added by the host as links to other threads. Nor did I add broken paragraphs, and in fact, I don't see any when I look at the post. I can well believe that you do, because I have seen the exact same website appear completely different from one computer to another, perhaps because of differences in browser programs or screen layout. If you want to private message me your email I can send you the post that way, and we could continue the discussion on this thread.

Alternately, if handing over your personal email to a random stranger on the internet seems unwise (understandable!) you might try pasting the bulk of the post into a different text program, see if that clears it up. Appreciate your thought.

Hey April, it seems you were shy about sharing your extensive experience; without knowing your familiarity with the medium I can understand how some comments could inadvertently seemed challenging.

I guess where I erred was assuming you were a novice; it seems the challenge is adapting your expertise to a different climate and seeking others with experience in the specific wet clime?

I'm on the west (wet) coast of BC Canada, and we do see some cob building in the Pacific Northwest. There is a renowned fellow in Washington state, whose name currently escapes me (Sunshine, Sunlight???) who has done some fantastical, fairy tale cob dwellings with living roofs, that are generally round. I would bet there might be some adaptations there - but it rarely gets below freezing here for long periods of time.

For one build he made bricks from used mushroom growing medium, mycilium bricks I think he called them (in an attempt to add insulation, or perhaps it was to speed up the building process?) he seemed thrilled with the outcome (is it just me? I always wish they would do follow up broadcasts on unique ideas like that!).

I am assuming the clay/sand components naturally available would differ in composition from place to place also - or would that be an issue?

Hello Lorinne,
Well, I had kind of hoped that the specificity of the post and questions would speak for themselves. Nevertheless, I thought your comment was entirely appropriate: If my interest were solely in the star vault form and not also in trying to create that form without the use of concrete, then aircrete or some other lightweight ferrocement would be an excellent solution.

The thing about cob is, although it is very cool and we all love it dearly, it is not, in fact, the strongest material humankind has ever discovered. It’s reasonably good in compression, when fully dry. It pretty much sucks in cantilever, in shear, in tension, or under any other sort of lateral loading. IN THEORY, a properly designed vault should transfer all forces through catenary lines of thrust in perfect compression. In Actual Reality, I am not nearly suicidal enough to put that many tons of relatively weak material over my head without some serious reinforcement!

So in your example of aircrete, the method, as I understand it, would look something like this: You build a formwork from rebar and chicken wire. You trowel a 2” coat of hard cement over it. This cement solidifies into the true bearing shell, which then receives a 6” pour of lightweight concrete, probably with a few more layers of wire mesh for good measure.

There’s no way that would work with cob, of course. 2” of cob wouldn’t bear jack. Rebar and chicken wire could not possibly support the slumping weight of the 8” of wet mud that *might* be able to bear, eventually, after 6-10 months of drying. Any bendable form of reinforcement is obviously going to deform under the weight → then pass outside of the line of thrust → down she goes. So this is why I was trying to think of some kind of rigid embedded frame.

I have not, however, tried anything exactly like what I am proposing: a relatively lightweight but extensive wooden frame embedded in a curved plane of cob. No one I know has. So I thought I’d throw it out there to people I don’t know.

So that’s one specific question I have for the forum. Also, I have no idea whether or not the proposed layer of loose gravel, with or without embedded ventilation pipes or pancake fans, would work to ensure breathability between the cob and the rubber membrane. Also, has anyone observed what happens to green-set but not *fully* dry cob when hard freeze/thaw cycles hit?

Looks like I should take my perlite-waterglass question over to the Paints and Finishes forum.

Yes, sand and clay have a fair amount of regional variation. Now that cob has an appendix in the IBC, it includes methods for testing things like moment of rupture. Unfortunately, as I read it, that test seems to be that one takes a 12” block, props up either end, and pushes down on the middle with a hydraulic press. I haven’t got a hydraulic press. Anyone know of a good home test for that sort of thing? (No, I’m not trying to meet code; it just seems like something I ought to know.)

Appreciate your thoughts.

If you have access to lumber or small trees, you can make a press that would give a rough approximation of load capacity.

make a strong base frame with support rails at the correct distance apart, and an upward extension a precise small distance away which can hold the thick end of a "hinged" pole which will put a load on the top center of the brick. Put a weight on the far end of the pole hung from a rope, and increase the weight until failure. Simple ratios (hinge to brick vs. hinge to weight) will tell you the effect of bare pole plus weight at the center of the brick.

Thank you, Glenn.
I knew I needed to get around to making that cheese press one of these days....

April, thanks for the changes.
If I read it correctly you are after a hexagonal shaped structure which is leaning in to a 6 pointed star light well at the top?

More or less. A barrel vault: half a cylinder. A groin vault: two intersecting barrel vaults, absorbing each other's lateral thrust. A hexagonal vault: three intersecting barrel vaults. A hexagonal star vault: three barrel vaults intersecting with each other and with the surface of a dome, plus the compression ring at the center spoked into a second star, which is framed by sturdy ribs. Traditionally those are masonry but I was thinking wood and perhaps slap an oculus on top, so long as you have a compression ring anyway. I would think the design was completely insane if it weren't for all the centuries old examples still standing. In practice it behaves kinda sorta like a dome.

Hey April, I read "had a passion for cob" to mean "really always wanted to try cob." Sorry for making a condescending assumption.

I sounds like you are possibly talking about testing the outer limits of what this material is capable of; and as someone who has the other kind of passion for cob (Really, one of these days I am going to build that bigger pizza oven :) I would be quite interested to hear about what you come up with in your research and experiments. Are there domed cob structures out there that are surviving in a similar climate? I live not far from the cascadia fault, so unless a cob house out here could stand up to 4 or 5 minutes of intense shaking, it would probably be a bad idea.

Some definitions;
What is an oculus in a building?
oculus, (Latin: “eye”), in architecture, any of several structural elements resembling an eye.
A small window that is circular or oval in shape, such as an oeil-de-boeuf window (q.v.), is an oculus.
HOGAN built by Navajo
A hogan can be round, cone-shaped, multi-sided, or square; with or without internal posts; timber or stone walls and packed with earth in varying amounts or a bark roof for a summer house,
with the door facing east to welcome the rising sun for wealth and good fortune.
Traditional structured hogans are also considered pioneers of energy efficient homes. Using packed mud against the entire wood structure,
the home was kept cool by natural air ventilation and water sprinkled on the dirt ground inside.
During the winter the fireplace kept the inside warm well into the night. This concept is called thermal mass.

I have studied the construction of many of that style of building.
From memory some were built with layers of different materials or even different sized bricks and the loads were carried differently in such a manner the forces equaled each other out.
Its tricky stuff to remember, I always go back to my books, the issue is finishing with vertical forces, not angular which would require buttresses.

The centre may require a strong 'star' to hold / carry the lateral load from the vaulted arch, if it was designed to lean into the centre a little bit.
That way it would remain stable.

Infill double skinned vault shaped panels can then be inserted between the ' Vaults' maybe using scaffolding to hold the shape until all the blocks are in place.
Filippo Brunelleschi's red-tiled cupola on Florence's Cathedral is an example

Carl, thank you for your graciousness. You are completely right; this is a terrible idea in a seismic area!! Anytime an arch or vault slumps or gets knocked outside of its line of thrust, even for a few seconds, it fails. We’re seismically quiet here.

No, I don’t know of any examples of what I’m trying to do. Pure cob vaults and domes are generally considered impracticable, because the material is too soft. Hence my idea of the buried frame.

The nearest existing example I can think of is perhaps the Navajo hogan, as John has posted above. As I understand it, this is what the interior roof framing of a hogan looks like:

You see each member is spanning a relatively long distance (6’, then 5.5’, then 5’, etc.). You see each is bearing the weight of everything above it as a point load in the center of the span. For this reason, the Navajo use chunky timbers, usually whole logs. Once the outer mud is dry, it probably does begin to take up its own weight and act as a shallow dome, but the frame has to support it until then, and support it again throughout the life of the building whenever it becomes wet (which, in Navajo country, happens sometimes but not often). The hogan is a good, simple, sensible, well-proven design for a dry climate. It might even take some seismic shocks.

I can’t make this (very basic) computer sketch program show you the whole interior view of the structure I propose, so I have instead included six horizontal cross-sections of the proposed frame at increasing heights. Thick lines represent the narrowing sections of the solid, plumb vertical arch-shaped walls. Thin lines represent the pattern of sticks at that height, each projecting a few inches out from the layer beneath. Apologies for the horrible sketch quality. Not to any kind of accurate scale!

As you can see, there are no long spans and no point-load bearing until you reach the very top, which, as John points out, would need to be framed quite sturdily, especially the interior hexagon which ought to act in compression. (The spokes off from it are just insurance, bearing on the peaks of the solid vertical walls. They also represent the top line of the barrel vault arches.) So this whole “dome” of intersecting arches is formed by steadily deepening corrugation. At each indented corner, not shown here, there is vertical cob infill up to the point of the lower eaves, which aids in buttressing.

SEEMS LIKE, as I said above, a frame like this could be made of fairly small sticks, so long as the cob acted as a monolithic mass. Any structural engineers out there? Can anyone see a reason why that’s wrong? And can cob be trusted to do so?

Suppose for clarity those low-point eave extensions should be moved back to sketch #4. You get the idea.

like so