FOREST PRODUCTS LAB

MADISON, WISCONSIN
MARCH 12, 2004

E NG IN EE R I NG R E PO RT
O F LI GH T C L AY
S PE C I ME NS
THERMAL CONDUCTIVITIES FOR DESIGN
COALITIONS STRAW/CLAY FORMULATIONS
EXTEND VOLHARDS K-VALUE VS. DENSITY
CURVE IN LOW CONDUCTIVITY END

TES T RE SULTS FOR THE RMA L CO NDUCTI VITY ME ASU REMENT S, TES T
CON DITION S, AND FORMULATIO NS OF SPECIM ENS P ROVIDE D BY
DES IGN CO ALITIO N

Four formulations of light weight straw loam provided by Design Coalition where
tested for thermal conductivities with the results as shown in table 1 below. The
specimens where first removed from their lay up forms and band sawed to 4 or 6 inch
nominal uniform thickness. The specimens were then equilibrated to 75 deg. F and 50%
relative humidity in the room in which they were thermally tested in a Lasercomp Fox
600 series thermal tester.

Table 1
Specimen

Density
(#/ ft^3)

Density
(kg/m^3)

Conductivity
(W/m*K)

delta temp.
(deg. C)

temp_median
R/inch
(deg. C)
(hr*F*ft^2/BTU/inch)

Low Dens. I
Low Dens. II

10.2
13.0

164
209

0.08
0.09

4.54
approx 4.5

23.21
approx 23

1.80
1.69

So. Dakota I
So. Dakota II

15.8
13.3

254
213

0.09
0.09

4.51
4.54

23.40
23.39

1.55
1.67

Reg I
Reg II

13.3
13.7

213
220

0.08
0.09

4.55
4.54

21.30
23.59

1.72
1.66

NM I
NM II

38.1
43.9

612
705

0.13
0.16

4.58
4.19

23.21
23.40

1.11
0.90

After thermal testing the specimens were oven dried at 105 deg. Celsius and reweighed to estimate their moisture content at the time of testing with results shown in
table 2 below.

Table 2

Specimen

Test weight
(grams)

Oven dry
(grams)

Moisture content
(% dry wt. bases)

Low Dens. I
Low Dens. II

6878
8652

6426
8121

7.0
6.5

So. Dakota I
So. Dakota II

10953
9330

10383
8839

5.5
5.5

Reg I
Reg II

8867
9158

8382
8644

5.8
5.9

NM I
NM II

17660
20379

17059
19774

3.5
3.1

The formulation as provided by Design Coalition is as follows:

S T R AW-C L AY I N S U L AT IO N T E S T S A M P L E P R E PA R AT IO N

Three components were mixed to make the straw-clay insulation infill for the thermal tests. These
were Wisconsin wheat straw, brown clay from surface excavation near Madison Wisconsin (here
after called Madison clay) and tap water. The straw was in 25 pound bales that tested 13%
moisture with a range of 11-15%. Test depths were 4 and 8 inches using a wood moisture
reading from a Delmhorst BD10 moisture meter. The locally mined clay when analyzed was 41%
clay, 40.6% silt and 18.3% sand & 0.1% gravel. Straw, clay and water were combined in 1:1:7
weight ratios respectively. Clay slip was made first by combining the clay and water. This was
done by mixing the clay into the water with a Silverson HydromaxTM high shear mixer or by
soaking the clay overnight in the water then mixing with a paint stirrer on a drill. Straw was then
put on a plastic tarp and the clay slip used to coat the straw by pouring it onto the straw while
tossing the straw with forks. When homogeneous, the mixture was stuffed in a five sided frame
made from 2x6s with OSB sides measuring exactly 2 feet by 2 feet on the interior and the width
of the 2x6s. Stuffing was done by lightly stomping on the mixture as it was added. When full,
the top 2x6 was added and the OSB sides removed after the sample had set 2 hours. Narrow
wooden diagonal corner pieces were added after the OSB was removed to maintain frame
geometry. The samples were allowed to dry in ambient conditions protected from any direct
weather exposure as rain or snow.
___________________________________________________________________

Conclusions:
The light weight specimens provided by Design Coalition appear to extend the measured
properties of denser formulations of Straw and clay/loam materials sited in the building
design literature (Minke, 2000: Earth Construction Handbook,pp35). The curve fit to
thermal conductivity of the Volhard data (Volhard,1983) places materials with densities
of 300 kg/m^3 at 0.1 W/mK. As the plot below shows the FPL measurements fall on
the trend line of Volhard data.
Thermal performance at mean temperatures other than the room temperature averages
chosen for our tests were not measured and may be presumed to change.
Thermal performance at moisture levels other than those shown in table 2 were not
measured and may be presumed to change.
The weights before and after thermal testing were taken with only nominal losses of
around 10 grams measured- near the accuracy of accounting for straw fiber being chaffed
off the specimen. Moisture movement at higher temperature differentials than we used
may be significant.