Prediction of Liquefaction Potential of Sandy Soil around a Submarine Pipeline under Earthquake Loading

Abstract

Liquefaction of sandy porous soil under earthquake waves is the most important feature governing the serviceability of underground pipelines. In this study, an artificial neural network (ANN) is combined with the mesh-free local radial basis function differential quadrature (LRBF-DQ) method to estimate the effect of soil properties such as the hydraulic conductivity, unit weight, Poisson’s ratio, and deformation module on the excess pore fluid pressure and the liquefaction potential surrounding a submarine pipeline under earthquake loading. The LRBF-DQ method was used to solve the governing equations. The results obtained by the LRBF-DQ codes and ANN show that with an increase of Poisson’s ratio, the deformation module, and hydraulic conductivity of the porous seabed, the pore fluid pressure and the liquefaction potential are reduced. Moreover, the sensitivity analysis of the ANN model showed that hydraulic conductivity has a significant impact on the excess pore pressure compared with other parameters.