Cob-Fiber Composite

I was thinking of ways to build a green roof on wooden beams without covering the whole area with wood…
Clay alone isn't terribly strong especially when it gets wet. covering the whole area with wood could be expensive and a waste of trees.

I have seen that fermenting cob (clay) with chopped hay gets really solid even if not completely dry. (The sugars in the plant matter are broken up by the bacteria and form a "giant molecule".)
Now there is this thing called fiberglass which is fiber made of glass embedded in a glue (some resin). The fiber provides tensile strength and the glue holds it together and fills the space, thus creating compression strength.

Would it be possible to do this composite thing with cob and plant fiber? With out chopping up the grass? Will dried or fresh grass yield better results?

I have just rolled up some grass embedded in clay. Now it has to ferment for a few days and dry.

In the field of composites there are both chopped fiber and continuous fiber, which would be used depending on all the materials involved and the desired properties. For example carbon fiber stuff is often braided continuous fibers, but a lot of glass fiber composites used chopped fiber. There aren't any rules, just different properties for different applications.

Dried vs non dried? My guess is drying before would be better since it will eliminate extra moisture from inside the composite once it's put together.

Hello Sebastian,

There are a lot of misconceptions concerning fiber in plaster and cob. These fibers do provide some tensile strength, but mostly they help while the mud cures by providing silica for chemical reactions and moisture regulation to prevent cracking. Many adobes that I've seen from 100 to 200 year old homes have very little plant material.

It sounds like you should look into building a vaulted adobe or CEB roof. The Auroville Earth Institute is the world leader in these.

All Blessings,
Bill

This is a complex and interesting subject, for decades I've been a part of the design of many of the commercial and military aircraft in the world and there are some very close similarities to building constructions. Zach is correct, you can design any composite layup as desired by placing the continuous fibers (braid and/or unidirectional tape) in different directions along with the type and size of fiber and therefore have total control over the properties. There are continuous or chopped graphite and fiberglass usually in an embedded two part epoxy mostly today called pre-impregnated or pre-preg vs wet lay-ups that gets laid up in a clean room cured in a autoclave under vacuum for high strength, and high temp resigns, in a factory setting. There is automated "fiber placement machines" used on the better aircraft today like the Boeing 787 Dreamliner has the largest.

So the answer is yes definitely the right type of binders and fibers can make COB far more robust than the past not needing large thick walls, and heavy roofs, the same concept has proven itself in many other industry for decades now.

Site lay-ups as we know today are more of challenge. A few American products I know of in the construction industry have really raised the bar to bring the concepts into practice.

1. Hemp has one of the highest tensile of plants due to it's high silica content and the way it grows tall. It binds best will magnesium oxy chlorides (MOCS), limes, some clays. It has limited strength as Bill said since it has no nodes like rebar and is reliant solely on the binder bond.

2. Helix has created a steel torsional chopped fiber that is twisted to gain tensile or torque resistance that claims higher than steel rebar, I'm not a fan of corrosive metals, be interesting to test some core samples with clays. I'm working with a university hope to be doing this soon, more for rammed earth maybe some COB while we are at it.

3. Zodiac makes FRPs (Fiber Reinforced Polymers) rebars and chopped...They sell to alot of civil and marine, industrial, both chopped and continuous make for the best structures in the world for like bridges, etc...since the properties are muti-directional. Working with them on some test too.

4. Rapid Protoype, look on u-tube for fiber reinforced COB and as the technology grows more materials and fibers will work with the machines, no more hand lay or site manufacturing just like other industries, its just a matter of time. I designed a factory HVAC duct back in 2014 there was talk of machines that could handle fibers, not sure now.

Bill, thank you for pointing me at the Auroville Earth Institute!
I visited their website a year ago during research about arches, but forgot about it later.
The architecture is amazing… The thin arches are gracile; unlike most modern buildings.
The challenge is probably to find a compromise between a vault and the shielding it provides and "simple" architecture that doesn't draw your attention. I also don't want to build a church…

The idea was to build a mold for an arch, fill it with aligned grass (or other plant fiber) and clay, let it ferment and dry. When enough of these arch segments are ready, they are lifted up and assemble the first layer of the roof.
Then additional layers of grass and cob with different alignment are applied. This should give an extremely robust roof that can then carry the load of earth to make a living roof.

While I appreciate the tendency of the industry to replace concrete with friendlier materials, I prefer buildings to be created by the people that will be living there.
Carrying huge amounts of clay, plants and other materials and energy into the factory where they are processed to highly optimized structures and then transported to the building site for assembly is definitely faster, but I doubt it is a good solution.
Especially when almost everything is present around the building site.

The grass for the second test and the first test-"cylinder" (hand shaped object that is similar to a cylinder) are drying. It was wrapped in black rubber sheet and exposed to the sun for 1 1/2 days and smelled horrible afterwards (it still does).

@Bill: I have used only enough clay to bind the grass together and fill most of the space in between. It is certainly not usable as plaster in this form and it behaves more like a thick sticky rope.
Also the smell that develops might limit the amount of plant matter that would be used normally.


I am waiting for the mess to dry and stop smelling before making any possibly destructive measurements.

Ditto, thanks Bill I found a link for "CAL-Earth" http://calearth.org/contact.html on the AuroVille site I'll get in touch with. I wont be far from Hesperia, CA beats the Mojave desert I was in last time I was here. I'm traveling across the US from arriving in a cold east cost last winter now on my way to a VERY hot 115 F (47 C) west cost in Arizona heading to So CAL. One can really get a feel for earth construction or plaster that dominates our hot dry desert climates(NM, AZ, CA) vs rare in freeze thaw-cold-wet where COB roofs would be more of challenge and more than likely have all kinds of issues if not properly designed. It would not take as much engineering here in the desert the biggest issues may be UV protection and over heating from too much mass that is slow to react. I have read of some high mass over heats even in cold Chicago. I don't agree that the average person should structures composite design there own house, most don't or stick to code which has none or has limited like OPC concrete. Next time I am in the Pacific North West I want to visit SIREWALL also linked, and see how the do rammed earth in a wet cold climate. They keep trade secrets to themselves developed over decades unfortunately.

I've chatted with some in SO cal, codes are restrictive for those arche spans and most are in a R&D developments approved by and being monitored by local jurisdictions. I advise taking caution pulling websites and photos off the internet without knowing the entire life cycle story or being part of the design and maintenance records. Most of what your are looking at has or should be approved by a licensed PE (structures engineer) for safety reasons if there is no local or international code. We do have earth code in some US states but that again requires a PE and core sample testing by an "approved lab" such as compression, sieve, atteburg limits, plastic indexs, shear in some cases, in approved ASTM, or other test fixture, to develop the properties I mentioned above so the PE can do the math. Most of these structures are in high seismic and/or wind zones where the fiber has to be shown to take most of or all the loads.

Sebastian is basically describing what has been done for decades I mentioned above. The concept of fibers cured and embedded in a binder aka "pre-preg". The concept came from the aircraft industry more than 30 years ago since I been in it, or shall we say has been refined to produce some of the strongest structures in the world. CFRP "Carbon Fiber Reinforce Plastics" or fiberglass/bassalts. They have been molded into shape using injection molds, compression molds, nothing new here but very old outdated technology rarely used anymore. One big issue I see is the elasticity of earth is low along with other properties like compression, tensile, shear unless mechanically rammed, so low with COB using clay binders alone with no admixes will be a challenge to build molds then lift into place without cracking and lost rework and labor. The strength-to-weight ratio is too low. In most industries tooling cost pays for itself over some production run, not one-offs where there are less costly less time consuming methods.

The other issues with "better concretes" is they can yield entirely different results depending on the environment they are manufactured in. For example, Mag cements the Romans used long ago, I was just reading a 2013 lab patented report that slight temp differences on the order of 5F (15 C) caused far less leaching of MAG Chloride (corrosive salts) and a breakdown of the mechanical properties in the first 30 days of curing, and the finished product. Clay's or most binders (glues) will behave the same way, for example if drying is in high humidity vs low or different temps or drying speeds, lime another one depending on type. Some drying speeds are far too slow and many not conducive to many site environments, some mags types cure in 24 hours and reach max strength, some in 1 hour too fast. Much of this to get the most robust composite is done in a factory with controlled environments. Even then as in the case with MAG and other wall boards some that tried never did get it right a folded. Those Romans must have had a mix that worked for them without the leaching of corrosive salts or efflorescence clays and limes can also yield in their environment that probably cannot be duplicated anywhere, and we need to see all the maintenance records since there is ALOT to know here. The higher the kiln temp the better the mechanical properties that is hard to get around and why OPC is so cheap and strong but has a high carbon foot print. There are alot of admixes on can add to earth just to change the binder properties and what it binds to.

There is a myth out there that these glass fibers only slows down shrinkage on OPC concrete, read Helix or Zodiacs reports and you will see the increase in properties on the order of at least 2-3xs steel rebar. I don't see that happening with grasses as Bill said, that could be tested for properties based on weight or probably has. I have seen some that used the wrong fibers and foundations failed had to be removed in less than a year, same with bad COB designs, or earth.

The first test has failed.
It was very promising while it wasn't dried out completely. Then the bond between clay and grass broke. Now it's just dry grass and dry, crumbly clay.
Either it has to be used in the humid stage, needs a binder that works when dry, or something else that I am not aware of.

The idea behind the statement that buildings should be created by the people is that it makes them aware of the work required and creates an understanding of the principles that are used in the construction.
Still, an experienced builder is required to ensure that the building is save and to teach the involved people.

Sounds like you have a weak binder for soil that needs to be stabilized.

Compressed (pounded) straw-clay has been used to make domed roofs over storage silos in China, "China: Grain Storage and Structures: Report on a FAO/UNDP Workshop Study Tour in the Republic of China", description of dome construction on pg 36-38.

Hi Sebastian,
We have made arches of cob packed tight in formwork, at our school. We've done a couple of spanned vault cob roofs, and those have stayed good for 15 years. We used whatever plant material was available, or none.
In the battery house (the most finished little house in these photos), we laid seabuckthorn branches near the bottom of the forms. My impression is that it caused gaps in the earthen mix right around it, but the whole roof was thick enough that it is still fine. We filled the outside to make it look almost like a typical local house, because we were concerned about local aesthetics. It is thermally very comfortable under that thick earthen roof; the purpose was to keep the solar batteries from getting cold, and indeed it stays warm all winter.
The cow shed with an 8-foot span is intentionally left unplastered, because visitors were convinced that there was some structural secret other than earth in the roof of the battery house. We didn't use much plant material in this one, I think. Again, the thermal comfort is wonderful, and we attach a greenhouse on the south side in winter, and keep a small vent hole open at the top, and still two cows keep their own shed at a minimum of +16C (+60F) when winter nights go down to -23C (-10F).
Both the battery house and the cow shed are about 15 years old, and were built with a simple mix of the on-site sandy soil with some fine clay-like white silt which is available right nearby, which for years we thought was clay, as it makes very strong and good earthen walls.
The low-arch roof was built by students of a natural building course, and isn't a good example. We did use a good mix of lots of straw with a good sticky red clay that we had to truck from a half hour away, but the experiment with the low arch hasn't been great. We intended to remove the wood planks, but it doesn't seem that will possible. The low arch seems to exert too much outward pressure on the walls below it, and they seem to be leaning away. Maybe if we made it as thick as our old ones, it would have enough of its own integrity, I don't know.
I'm dubious about combining a living roof with an earthen roof. The reason our cob roofs survive is because our climate is so, soooo, very dry. Even though a living roof would have a waterproof membrane under the soil, you have to expect that sometime in the coming decades it will spring a leak. In a non-soluble roof, that's an annoyance that can be repaired. In an earthen roof, not so much.
That said, I'd love, love, love! to have a living roof, that would cool by evaporation in summer, and be left to dry out for winter. Ooh!

Hello Rebecca,

the cow shed looks nice!

The cow shed with an 8-foot span is intentionally left unplastered, because visitors were convinced that there was some structural secret other than earth in the roof of the battery house.

Did you show them suspended chain? Maybe build a small arch from bricks without mortar …

We intended to remove the wood planks, but it doesn't seem that will possible. The low arch seems to exert too much outward pressure on the walls below it, and they seem to be leaning away.

From the picture I would also guess it would not stay upright.
I made a design where a wooden arch was kept together with strong rope/fiber. (I was looking for aramid fiber, as it resists fire and could be thin).

I'm dubious about combining a living roof with an earthen roof. The reason our cob roofs survive is because our climate is so, soooo, very dry. Even though a living roof would have a waterproof membrane under the soil, you have to expect that sometime in the coming decades it will spring a leak.

I have build a small shed with a green roof (rubber membrane and a mix of dead soil with high clay content and perlite 5-8cm thick), that is doing reasonable well. Some of plants at the upper edge dried out, yet there are survivors. The strawberries at the lower edge are fine.
I would love to find a way to build a roof with cob and earth where plants can grow and that does not use any barrier other than clay. The part of me that somehow often predicts the right outcome (I have no idea how) says it is possible. The logic based and formally reasoning part has not found the solution yet.
Soil physics is incredible complicated and the existing simplifications fail when the soil has two air interfaces. The effects of evaporation, condensation, different levels of hygroscopy of clay, soil and sand and their changing insulative behavior with moisture content doesn't make it easy to get a model of it into the head.

Again, the thermal comfort is wonderful, and we attach a greenhouse on the south side in winter, and keep a small vent hole open at the top, and still two cows keep their own shed at a minimum of +16C (+60F) when winter nights go down to -23C (-10F).

That's great! I was wondering how much insulation would be required. How much heat does a cow produce? 500W or less? Oh and... do you have a lot of sun in the winter months?

Right, we did base that arch on the curve of a suspended chain.

We didn't use any insulation in the cow shed except the thick earthen roof, probably 50 cm or more throughout. And the north side is the hay bin. Yes, we get plenty of sun in the winter, so even though the outdoors night time temperature goes down to -23C or -25C, the days are usually sunny and only -5C or warmer.

I am impressed by your confidence in the living roof over an earthen roof. We intentionally put those vaulted earth roofs over rooms where people would not be sleeping or spending a long time. We have not had any seismic event over 4 something since they were built, and they are holding up well, but what if we get an earthquake? And that's eithout trying to grow a living roof.

Well, go ahead with your experiments. I'll be very happy to hear if you succeed with the living roof over earth structure!

Looks like I'm resurrecting a dead topic.  Sorry!

Regarding the idea of making mixtures of cob into ropes, I remember seeing that on YT and thinking what a nice alternative to wattles that would be.  In one video they were forming long ropes and weaving them horizontally just like one might weave wattles, and in another they were draping shorter ropes over periodic horizontals, like wattles but I can't remember whether or not the members were woven.

Regarding the fermenting concept, I remember reading something about that in ancient Eqyptian adobe construction.  I believe they used raw water from the Nile and let the mixture ferment for either one or two weeks.  I had the impression that water specifically from the Nile was an important point, leading me to believe that it had special chemical/biological properties, although it could just be that the Nile was the only significant source of water in the area.

Finally, the rice mortar used in China for the Great Wall I believe relied on some kind of chemical reaction between the sticky rice (just the juice, I think) and the soil's mineral content.  I'm not sure about that, though.

For my part, I have also been trying to think of a way to combine cob with a living roof, although my climate (central Missouri) is definitely very wet.  I have been considering something like an animal den under tree roots, where a living root structure provides soil and moisture retention as well as the mechanical strength to handle narrow vaults.

Dampness is an obvious issue with that concept, but I think additionally the roots would not appreciate being exposed to air and would begin to die back or would be susceptible to disease, yielding either a short life expectancy or requiring constant replanting.  All of that said, protruding roots could be woven together to enhance the matting along the underside of the arch, and it is possible that experimentation with species and selective breeding could evolve a happy, healthy root structure.

Such a construction would probably only be suitable for something like a root cellar, and I imagine it would be a very bad idea to try and heat the space because it might dry out the protruding roots, or worse, confuse their biological clocks and get them all killed by trying to wake up mid-winter.

I am strongly opposed to the use of membranes in a roof because, as another poster stated, leaks are inevitable.  It sounds to me like an issue where a future owner would say 'that's too much work to replace!  Let's just tear the house down.'

But what other options are there?  Membranes seem to be the universal solution in modern attempts.

Regarding the use of too much wood in an earthen roof, I completely agree.  But what if the life expectancy of a 100% natural house, sourced all on-site, could be matched to the time it would take to regrow the lost timber, after which point the house could be gutted for glass, fixtures, wiring, etc and allowed to rot away?  That seems acceptable to me if it could be accomplished.

The roofing solution I have decided to go with is wood shakes over an attic whose floor is an earthen ceiling, formed by vaulting small members between beams and then plastering a cob mix until level on top (or very slightly domed to shed drips).  The vaulting and beams would support the cob if it got wet and the dry, breezy attic would help to dry the cob.  The wood shakes won't last forever, but they can eventually be replaced by clay tiles made on-site, which will be re-usable when/if the house reaches the end of its life.  I'm a little nervous about the use of earth overhead, but if the ceiling framing is designed for wet weight it should be sound even if a tornado blows half the shakes off.

As a last point, I am looking into the use of long wood fibers for my cob mix because I don't happen to have acres of straw on hand.  I don't have any knowledge of the chemical aspect, but I have read that one of the main benefits of straw in cob is not many things want to eat it, which is definitely a point against the use of wood fiber.  Still...wood framing is made of wood fiber.

I would probably be trying the horizontally woven 'rope' idea as mentioned above in order to take advantage of my opportunity to cut fibers as long as the source poles will allow.  Right now I am thinking the best approach is to strip off long sections with my drawknife and then further split them down with my pocket knife.  Soaking and then flexing and/or crushing splits could aid the process.  It also seems like a good time to experiment with varying the thickness of the fibers to include something in between a wattle and a straw fiber.

A sample I made of pure (undoctored) sifted earth, soaked for several days and allowed to dry to a gooey paste, yielded a small brick that cracked a little upon drying but eventually required a heavy hammer blow to break.  This in combination with wood fiber should be more than adequate for a ceiling comprised of miniature vaults between beams, layered to...4"?  I will need to keep spans fairly short, I am sure, but it's a promising solution in combination with the shakes overhead.  I only wish I could be assured of several decades of service life for the shakes, which will be white oak heartwood from diseased trees, but nonetheless quite old.

Well...those are just some thoughts on a semi-earthen roof technique.  Certainly not a living roof, but it has at least a bit of mass.  An earthen vault under a shake roof could yield far greater mass but with the disadvantage of greater thrust in the bearing walls, and it would be difficult to navigate the attic for inspections of the roof framing, not to mention the space would be less usable.