However, GRS RWs with geogrids may encounter the following potential problems: 1) for ordinary
structures, the local soil materials which may be poorly graded or include larger particles would be
inevitably used as backfill. This would result in a decrease in its deformability. On the other hand, to
obtain a high friction resistance with geogrids, the backfill soil is restricted to sandy soil and heavy compaction work should be conducted which lead to an increase in construction cost; 2) for some
important structures such as high-speed railways, the requirement of seismic stability and allowable
deformation of GRS RWs are higher than ordinary structures, which requires the reinforcement to
have a higher resistance. In order to alleviate these problems and improve the seismic performance of
GRS RWs, a new type of geocell (Fig. 1a) was developed by Han et al. (2013), which has a different
cell shape compared with traditional type geocell (diamond-shaped geocell, Fig. 1b), called
square-shaped geocell having straight longitudinal members with transversal walls at separated
locations. The pullout test results indicated that square-shaped geocell showed only slightly
progressive deformation and therefore showing higher pullout resistance and initial stiffness than
diamond-shaped geocell. What’s more, Han et al. (2013) conducted a series of pullout tests using
square-shaped geocell models and a prototype geogrid (Tensar SS-35) embedded in gravelly soils
having different particle sizes indicating the important benefit of square-shaped geocell which can
confine large soil particles in their three dimensional cells and respective cells provide large anchorage
capacity when pull laterally compared with geogrids. In this paper, the seismic performance of GRS
RWs reinforced by square-shaped geocell models and geogrid models was evaluated by two shaking
table model tests. In addition, a gravity-type retaining wall model was also tested for comparison.