Factors of safety against overturning and sliding of tilted model walls were evaluated
based on a pseudo-static approach. The backfill earth pressures were estimated using
the two-wedge method for the reinforced soil-type wall, as described by Horii et al.
(1994) (Figure 9), and the Mononobe-Okabe method (using a single wedge) for the other
types of model walls. In both methods, earth pressures due to the self-weight of the
backfill were assumed to be hydrostatically distributed, and those due to the surcharge
at the top of the backfill were assumed to be uniformly distributed.
It should be noted that in previous studies (e.g. Bathurst and Cai 1995), a dynamic
increment of the earth pressures was assumed to have a higher point of application than
the hydrostatic distribution. In this study, however, the assumption of a hydrostatic distribution
was employed because it was broadly used in the current, aseismic design
practice for retaining walls in Japan.
Based on the plane strain compression test results for the Toyoura sand which was used
in the model tests, at a low confining pressure (9.8 kPa) the shear resistance angle, φ, of
the backfill and subsoil layers was evaluated to be 51_ when the major principal stress,
σ1 , direction was vertical. For the cantilever-type wall, having a wall base underlain by
the backfill, a virtually vertical back face was assumed within the backfill as shown in
Figure 10. The portion of the backfill located above the wall base and between the back
of facing and the virtually vertical back face was regarded to be a part of the wall.