Cold Cob

[M J Epko]

        As usual, I've jumped through many more mental hoops than should
have been necessary to figure this one out. With any luck, there won't be
anything left for me to think about when buildin' time finally rolls around...

        Eric's hesitations are supported by the evidence. R-Values are
quoted as (sometimes substantially less than) 1/" in the very few specific
references I've found in the last couple days. (The thermal-mass qualities
are then quickly pointed out in defense.) It's interesting (perhaps to me
alone) to note that cob's thermal resistance per inch is about the same as
wood. (Could I possibly have used any more parenthetical asides in this
paragraph?)

Ann Edminster wrote on the SB list:
>... cob mixes vary substantially, and those with greater proportions
>of straw in them are more insulative than those with less; the other
>ingredients will also affect the cob's insulating capacity.

        I also got my rammed earth book back this afternoon (having lent it
out before I'd read it [hey, it's in queue]), and Mr Easton states that
"dense materials such as adobe, concrete, stone, brick, and rammed earth
have R-values roughly equivalent to .25 per inch." And later, "Extremely
cold or oppressively hot outside temperatures will render uninsulated
thermal mass ineffective."

        He suggests the possibility of installing a couple inches of rigid
foam in the center of the wall as it's being made, just like Shannon's
doghouse. Except that he's talking about rigid foam and Shannon's talking
about straw.

        Andrew Alcorn had a good post today to the SB list in response to my
(in hindsight kinda stupid) question about thermal mass. (It's safe for me
to assume that everyone here is also there, right? If not, email me & I'll
send you the post I'm talking about.)

*

        And so, cob having been deemed unwise as sole-component for this
beastly clime (however excellent it is in more moderate areas), but being
still-enamored with it's beauty and beneficial thermal possibilities, we
move into design considerations...

Similar to Andrew's above-referenced post to the SB list, John Swearingen
wrote there on this topic:

>... let the sun in through windows (not walls) and trap it with mass. Put 
>your mass on the walls and places where the thermal transfer will be 
>efficient (like the floor in front of a window). The bale walls, being well
>insulated, stay close to the ambient air temperature; and cob benches, 
>interior walls, etc. provide a thermal flywheel. 

powerhouse Ann Edminster echoed that:

>There are homes in the southwest U.S. that combine adobe and strawbale, with
>the adobe as thermal mass (TM) on the building interior, and strawbale for
>the building envelope... [snip] -- store some of the building's heat in the
>interior TM for release at night, and protect the TM from the exterior cold.

and Rob Tom, for whose opinions I have enormous respect for some reason,
wrote in that thread:

>... Cob/SB Hybrid offers interesting possibilities.
>
>It would be wise to take advantage of SB's high thermal
>resistivity (R = 2.1 to 2.9 per inch according to JoE) to help
>reduce the heat-load requirement of the building volume.
>
>Roof snow loads will be a big concern. Rather than resort to a
>timber frame/SB infill, it seems reasonable that an interior cob
>wall, utilized for its thermal mass capability, could do double
>duty in a roof-load-sharing arrangement with the SB thermal wall.

which prompted Jorg Ostrowski to respond with:

>Having designed/built with rammed earth interior walls and SB exterior walls
>(with shallow footings & rubble trenches) in 11,000 DD/year climate, it is a
>challenge from both a design and construction point of view to accommodate
>solid structural systems that don't move on the interior and flexible wall
>systems (initially) on the exterior that do move.

        Jorg's concern seemed initially valid; but I think that Rob might
have been suggesting something Eric touched on a few days ago - build the
cob structure, then add a bale cladding. Jorg seemed to be talking about
interior inflexible (read unsquooshable) interior divider walls and
loadbearing SB shell walls, which would squoosh.

        I think I might be more comfortable building the two simultaneously,
rather than stacking bales against an already-dry cob wall. My thoughts
aren't so much with adhesion (wouldn't the cob mixture be too thick to
adhere well to anything but itself? hence the porcupine method of attaching
door & window frames?) but rather with conforming the cob to the shape of
the bales. Probably not important, but I don't want to inadvertantly make a
chase for critters. Anyway, that would result in loadbearing cob with SB
cladding that wouldn't squoosh much if at all.

        At 6 or so inches a day, the walls alone would be a 16-day *minimum*
building process.

        Another possibility would be a post & beam / pole framework with the
posts set flush with (or immediately interior to) the interior of the cob.
(I think it was Ann E who proposed this.) That way the posts can be got at
in case of a post emergency of some sort, and it also allows a handy place
to attach shelves and stuff. A person could router in electric chases too.
And the roof could be weathered in prior to building the walls. Win / win. I
like it. Anybody want to take a shot? (Yeah, the p & b structure would be
redundant, I know... consider it a means to an end. A construction-process
convenience. A safety net.)

So it all comes back to what I'd said initially, & I've wasted a lotta time
in the interim:

>We here have been considering using standard sb construction, using regular
>stucco on the exterior, then using cob on the interior for its functional
>mass and artistic sculptability. That would give us the outstanding
>insulation from the bitter cold *and* a lot of interior mass for the sweet
>warmth.

*

Shannon wrote, and I requote it here for no real reason other than that I
really like it:

>I prefer round buildings in general for the simple reason that they are 
>more efficient.  A single story 900 square foot house built using straight 
>walls requires at least 120 linear feet of wall to build.  A round 
>building would require only 106 linear feet for the same square footage, 
>a 12% reduction in materials, labor, and most importantly thermal losses 
>through the completed walls.

*

M J