Applicability of rigid block based approaches in predicting sandy slope displacements by 1 g shaking table tests

Various approaches have been developed by researchers to predict earthquake induced landslide displacements. This study evaluates the applicability of these approaches on sandy slopes. For this purpose, nine physical models are constructed inside a rigid box and thirty-six shaking table tests are conducted. Dynamic responses are then converted to the full scale model by applying similitude laws. Five statistical criteria are applied to compare the measured and predicted displacements and evaluate the precision of the approaches. By combining the outcomes and using a scoring procedure, the approaches are scored. Consequently, the approaches of Fotopoulou and Pitilakis [23] and Hsieh and Lee [22] score the highest. Additionally, a correcting coefficient is assigned to each approach to convert the predicted and observed displacements. An exponential equation is then developed to correlate the correcting factor to K-y/PGA ratio for each approach.