Abstract: Seismic response testing of dry sand sites serves as an effective method for evaluating boundary effects of flexible laminated shear boxes. For this purpose, a centrifuge shaking table model test on a dry sand site was designed and conducted under 50g centrifugal acceleration. Representative near-fault pulse-like and non-pulse ground motions were used as input at the shaking table platform. Acceleration records from sensors installed at various depths within both the central zone and boundary regions of the site were comparatively analyzed to assess the shear box’s boundary effects. The model test results demonstrate that boundary effects diminish with increasing soil burial depth and greater distance from the box walls; furthermore, boundary effects are more pronounced under the pulse-like ground motion. When only ground-motion amplitude is considered, the free-field zone lies at least 3/20 L from the short-side wall and at least 1/5 W from the long-side wall (L = box length, W = box width). When ground-motion spectral characteristics and phase lag are also considered, the extent of the free-field zone requires further determination through experiments. The derived conclusions can provide critical reference for enhancing the accuracy of centrifuge shaking table model tests.

Key words: flexible laminated shear box, boundary effect, dry sand free field, centrifuge shaking table model test