Acceleration-amplified responses of geosynthetic-reinforced soil structures with a wide range of input ground accelerations

A series of dynamic centrifuge tests is performed to investigate the accelerationamplified and de-amplified responses within geosynthetic-reinforced soil (GRS) structures. Further, a database from various dynamic centrifuge and shaking-table tests is compiled from literature to cover a wide range of input ground accelerations in the range of 0.01.1.0g. This study demonstrates that among all factors in GRS structures (i.e., structural configuration, backfill and reinforcement material, and seismic characteristics), input ground acceleration, ag, location, z, and input motion frequency, f, have the most significant effects on acceleration-amplified responses of GRS structures. The magnitude and variation of acceleration amplification factor, Am, which is the ratio of horizontal acceleration inside GRS structures, ah, to input ground acceleration, ag, decrease as ag increases. Am is larger than 1.0 and non-uniformly distributed with height at approximately ag<0.40g; while Am is less than 1.0 and generally uniformly distributed with height at ag≥0.40g. Experimental results show that acceleration-amplified responses are highly dependent on input frequency, f. Acceleration inside GRS structures increases markedly when the predominant and fundamental frequencies are close. Further, this study examines the Am and amax relationships (i.e., Am=1.45-amax/g) adopted in current GRS structure design guidelines. Comparison results indicate that the Am and amax relationship adopted in current design guidelines follows well the trend line (Am=-0.69lnag+0.43) regressed from complied physical data at ag≥0.40g, but underestimates Am at ag<0.4g. The influence of location and frequency on Am, as observed from physical data, is not considered in the current design guidelines.