The geosynthetic reinforcement used in all model configurations
was a woven polyester geotextile with tensile strength Tult
5 40 kN=m and ultimate strain at failure ɛult 5 11%. The same
geotextile was used in all eight different setups. This was a rather stiff
reinforcement considering that the stiffness of the geotextile in the
prototype would be 4 and 16 times the stiffness in the model for 1=2
and 1=4 scale models, respectively. The soil used as backfill was
a well-graded silty sand (SW-SM) with coefficient of uniformity,
Cu 5 9:13 and coefficient of curvature,Cc 5 1:17. The percentage of
fines (mainly silt-size particles) was 6%. The soil was used at its
natural water content, which was approximately 4–5%. Although the
fine content did not exhibit plasticity, when a molded sample of the
backfill was oven dried, it still kept its shape even when it was
stroked slightly. When the fine content was removed, the molded soil
kept its shape at its natural water content, but completely segregated
in the oven. This observation suggested the presence of a small
amount of apparent cohesion