34[hreg] Papercrete/Fidobe Strength Tests
- Sep 13, 1999The following is from Earth Quarterly issue No. 5. In this article they
refer to fidobe as paper adobe. I have inserted the appropriate fomulae
in square brackets.
Papercrete Strength Tests
In May we had the good fortune to meet Kenneth Leitch, a graduate
sstudent in Civil
Engineering at New Mexico State University, who agreed to do some
compression and tensile
strength tests for us. We made samples of 5 different formulations, and
met with Kenneth on
August 19 at NMSU's Civil Engineering materials testing lab, where we
tested them with a couple
of Tinius Olsen multitesters, which are essentially huge hydraulic
presses with calibrated
dials that tell you how much pressure is being applied.
For the compression tests, we had made papercrete cylinders, 6" in
diameter, and 12" tall.
The cylinders of pure paper pulp were only 10" tall, because of excess
Here are the formulas we used:
#1. Pure paper pulp (newsprint).
#2. 1/2 bag of cement in a 200gallon batch; no sand. This is the "roof
panel mix" described on
page 16 of this issue. [160 gallons water; 60 lbs. paper; 47 lbs
#3. 1/2 bag of cement + sand This is the "formula for large tow mixer"
on page 16 of this issue.
[160 gallons water; 60 lbs. paper; 47 lbs cement; 66lbs sand.]
#4. Identical to #2, but with a full bag of cement rather than 1/2 bag.
#5. Paper adobe, using the formula on pages 16-17.[160 gallons water; 60
lbs paper; 240 lbs dirt.]
We made three cylinders of each formula. One of the paper adobe
cylinders broke when wet,
so we only tested two cylinders of this formula.
When testing these cylinders, we noticed how elastic papercrete
and paper adobe are. Under
a compressive load, they behave more like wood than like concrete. Wood,
when subjected to
modorate compressive loads, will compress down without breaking.
Concrete, on the other hand,
will retain its original shape as pressure is applied, until it
Papercrete/paper adobe behaved like an accordion--we could compress
the cylinders from their
original 12" down to about 9' when we got to the 85 psi range, but they
would regain about half
of this when the pressure was released.
We decided to measure how much pressure was required to compress
each sample by 2". Here
are our averages:
#1 59 psi
As one would expect, the higher the non-elastic content (cement,
sand, or dirt), the more
pressure is required to deform the sample.
We then took one cylinder of each formula to a larger multitester
to see if we could
destroy them. This unit was set up for concrete testing and had a
limited range of motion, and
couldn't smash samples 1-3 small enough to cause them to fail. (For
example, we applied 12,000
pounds of force (424 psi) to the pure paper sample, and compressed it
from its original 10"
down to 3 1/4". When the pressure was released, it rebounded to 4 3/4".)
Samples 4-5, more brittle with a higher non-paper content, did
fail. #4 failed at 248 psi,
and #5 failed at 212 psi. (After the pressure was released, samples 4
and 5 were both 8 1/2"
tall 3 1/2" less than their original height.)
The formula for #4 is similar to that used by Mike McCain; the
"260 psi" figure we've
been quoting in EQ is based on a compression test he had done in
Alamosa, CO (EQ #1, p. 13).
Considering the range of experimental error (a single test done on two
the 248 psi figure we obtained at NMSU is equivalent to the 260 psi
we've been quoting all along.
But-and this is an important issue--long before papercrete and
paper adobe lose structural
integrity, they will compress down as more pressure is applied. So
rather than asking, "At what
pressure will papercrete/paper adobe lose structural integrity," a more
immediate question would
be, "What level of compressional shrinkage is acceptable?" If you put X
amount of weight on top
of a wall and it squeezes down by-Y inches, is this acceptable?
The papercrete pioneers would answer: Yes of course it is
acceptable. In the real world,
the weight of the wall itself is insufficient to compress the bottom
layers at all; even adding
the heaviest possible roof (vigas, heavy planks, etc.) will not compress
the wall much, if at
all; any compressional shrinkage can be easily compensated for, and the
structural integrity of
the wall will not be compromised.
From the point of view of structural engineers and building codes
people, who are used
to dealing with inelastic wall systems, there might be some issues here.
I think that straw bale
walls, which surely compress a little when heavy loads are applied to
the top, could provide a
precedent. It might well be that more conservative building codes will
insist that papercrete
be used only as infill with post-and-beam walls. However, I remain
convinced (and I think that
most seatof-the-pants papercrete experimenters would agree) that
load-bearing papercrete and
paper adobe walls are perfectly safe, particularly if they are
reinforced with rebar.
For the tensile strength tests, we made little beams, 30" long
with a 3x3" cross section.
We supported these beams on their ends and applied a load in the middle.
As expected, papercrete
had a low tensile strength, much like unreinforced concrete. Samples 1
and 2 could support
approximately 20pounds; the other three samples could support
approximately 40pounds. Kenneth
told us that a piece of wood this size could support 1000pounds. So
clearly, tensile strength
is not papercrete's strong point.
The conclusion to be drawn from all this is that papercrete/paper
adobe are unique
materials, with much more compressive strength than tensile strength. I
wish we had tested to
see how much pressure was necessary to cause the samples to deform even
slightly, since I know
that people will be asking this question. My sense is that any kind of
real-world loads will cause, at most, only a slight deformation, not
enough to be concerned
about. But I think that further testing is called for, because I know
that the elasticity of
papercrete/paper adobe will be of concern to people who are used to
working with perfectly