In recent years, natural disasters combining rainfall and earthquakes, or seepage flow and earthquakes have been occurring frequently. To deal with such problems, it is necessary to evaluate the response by considering the infiltration of pore water into the unsaturated zone. Therefore, we modified a conventional effective stress analysis program to reproduce the behavior of unsaturated ground by using a quasi-three-phase method that treats pore water and pore air as mixed fluids. The improved program was then used to reproduce centrifuge experiments that considered seepage flow into the small scale earth dam, and it was confirmed that the program could accurately reproduce the excess pore water pressure ratio, displacement of the top, and deformation modes. Furthermore, a parametric study was conducted for different soil water characteristic curves, and the effect of the soil water characteristic curve settings on the seismic behavior was discussed.

This paper presents details of a third-order accurate and unconditionally stable time-discontinuous Galerkin Space-Time Finite Element Method (ST/FEM) for solving solid-dynamics and earthquake engineering problems. This method is derived by employing the newly developed concept of displacement functions for satisfying the displacement-velocity relationships. The proposed method, which is designated as LC v-ST/FEM, is derived from the linear combination of displacement functions of single-field and two-field ST/FEMs. Consequently, LC v-ST/FEM has a user-defined parameter 0 ≤ α ≤1, which is introduced for controlling the high-frequency dissipation characteristics. From numerical analysis and solutions of benchmark problems, it is demonstrated that the proposed method is the third-order accurate in time, unconditionally stable, and contains negligible numerical dispersion error for all 0 ≤ α≤1. Moreover, for α ≠ 0, the method can attenuate the spurious high-frequency components from the velocity and displacement fields.

This paper presents a numerical investigation of the impacts of post-peak strength loss on ground motion amplification and reduction for a soil column with a layer of strain-softening clay. Stress attenuation and amplification were modeled with a one-dimensional soil column using the finite difference program FLAC 8.1 with the PM4Silt constitutive model. Three calibrations of an idealized soil were developed consisting of different rates and magnitudes of post-peak strength loss. Alternative column geometries were devised, each with different clay layers thicknesses to establish how ground surface motions were affected. Impacts were quantified using both a cumulative ground motion intensity (Arias Intensity) as well as different spectral components (base to surface transfer function). The results illustrate that the thickness of the clay layer surface and strength loss in the layer have an impact on the magnitude and frequency content of the earthquake motion measured at the ground surface. The impact of these results on practice and future research needs are presented.

Liquefaction has been primarily interpreted as an undrained phenomenon and liquefaction susceptibility assessment procedures are largely based on the undrained hypothesis. However, analytical, experimental, and field evidence have revealed the importance of co-seismic water flow. If water flow causes localized volumetric expansion during an earthquake, for instance at the interface of a liquefiable layer with an overlying layer of lower permeability, it can facilitate the manifestation of liquefaction. To understand the implications of drainage on the co-seismic cyclic behaviour of sand, the element level response needs to be evaluated and quantified using appropriate constitutive models. In this paper, experimental results from cyclic triaxial tests on Hostun sand under undrained and partially drained conditions, imposed through constant ratios of volumetric strain rate, are studied. The two-surface plasticity model of Dafalias and Manzari (2004) is calibrated in triaxial space using the undrained monotonic and cyclic test results and then assessed for its ability to simulate the partially drained cyclic experiments. The results of this assessment provide insight on the performance of two-surface plasticity models that are calibrated using cyclic undrained tests when predicting soil behaviour under partially drained conditions. Implications for constitutive model calibration and the reliability of numerical simulations of earthquake-induced liquefaction are discussed.

Earthquake-induced landslides are natural disasters that can pose significant threats to human lives and infrastructure. To assess the risk and the extent of damages, it is essential to develop a robust computing method capable of predicting the large deformation and failure process of geomaterials caused by seismic loads. Over the years, several numerical methods have been developed for this purpose. Among these methods, the smoothed particle hydrodynamics (SPH) method has emerged as a powerful approach for simulating the large deformation and post-failure behaviour of geotechnical problems. However, one of its limitations is the lack of suitable boundary conditions for analysing seismic responses in unbounded boundary domains or infinite computational domains. This study addresses this knowledge gap by presenting a general SPH framework for the large deformation seismic response analysis of unbounded slopes. It will be shown in this paper that the proposed SPH framework can accurately reproduce wave propagation and dissipation, in particular for a long duration. The proposed SPH framework is then employed to simulate an earthquake-induced retrogressive landslide in sensitive clay. The simulation results reveal several aspects of the slope failure under seismic loading conditions, and demonstrate that SPH is a suitable tool for the risk assessments of landslides and other geohazards triggered by earthquakes.

In this study, the effects of sustained initial shear stress and cyclic stress amplitude on loose sand liquefaction potential in undrained cyclic simple shear tests are evaluated using an energy-based approach. The testing results demonstrate that various stress conditions lead to two deformation patterns, namely cyclic mobility and plastic strain accumulation. The accumulated strain energy is calculated by the stress-strain hysteresis loop during loading cycle. A different pattern of dissipated energy during cyclic loading is observed for the two failure modes, with a clear tendency of dissipated energy to increase less rapidly as lower cyclic stress amplitudes are applied for all samples. The dissipated energy at failure is found to be practically independent of the cyclic stress amplitude, and it is little influenced by initial static shear stress. For a given static shear stress level, a power law-based model is established for the pore pressure build-up prediction using dissipated energy. A decreasing trend of the ultimate pore water pressure ratio values with the increase of sustained shear stress is apparent. When the residual pore water pressure ratio normalised by its ultimate value is plotted against the dissipated energy normalised by the corresponding value at failure, data points fall in a narrow band regardless of various static and cyclic shear stresses.

This study proposes improving a method for designing a water pipeline against fault displacement by incorporating long collar into hazard resilient ductile iron pipe (HRDIP) pipeline. A hazard resilient joint pipeline is capable of absorbing the large ground displacements by movement of its joint (expansion, contraction and deflection) and the use of the joint locking system. Existing hazard resilient joint pipelines have been exposed to several severe earthquakes such as the 1995 Kobe Earthquake and the 2011 Great East Japan Earthquake, and there has been no documentation of their failure in the last 40 years. These performance records of hazard resilient joints are also highly evaluated in west coast of U.S, therefore hazard resilient joint has been widely adapted as a measure for fault, earthquake and landslide etc. On the other hand, there are many faults which are estimated with even larger movements in west coast of U.S. In those locations, additional countermeasures are required when the fault displacement exceeds 1.8 m because this displacement could stress the hazard resilient joint pipeline beyond the elastic limit. Therefore, special countermeasures are adapted to increase the performance of the pipeline such as using long collar, which has an expansion/contraction performance more than 10 times as much as regular hazard resilient joint pipe joint. In this study, researchers conducted long collar joint deflection tests with fully expanded and contracted situations to make the relation between expansion/contraction and deflection performance clear. Based on the test results, researchers set the long collar’s joint deflection feature considering joint expansion/contraction accurately for finite element modeling (FEM) analysis and optimized the design of hazard resilient joint for fault movement.

From an engineering viewpoint, pumice sand particles are problematic because of their crushability and compressibility. While laboratory and in-situ tests can be implemented to characterise their behaviour, these tests are generally time-consuming and expensive. In this paper, the liquefaction behaviour of crushable pumice sand specimens is examined using the Discrete Element Method (DEM). Each pumice particle is modelled as a sphere, which, when the maximum contact force reaches the limit condition, will break and split into 14-ball inscribed tangent spheres. Initially, the results of laboratory single-particle crushing tests are used to represent the breakage characteristics as a function of particle size. Next, using the open-source code YADE, 3D model specimens in a loose state are prepared and isotopically consolidated at prescribed levels of confining pressure. The numerical specimens are then subjected to cyclic loading under the undrained conditions. The results showed that the DEM model could replicate the laboratory-obtained experimental results, offering explanations on the effect of particle crushing on the cyclic deviator strain. The microscale observations also provide better insights into the evolution of force chains within the specimens, resulting in a clearer understanding of pumice sand behaviour during cyclic loading.

This paper presents the results of a study to assess the effect of ground motion directionality on the stability of a slope under seismic loading. Stability conditions were determined through two-dimensional (plane strain) nonlinear dynamic analyses using the finite element method and the Newmark rigid block method. In the numerical calculations, the soil was characterized by a nonlinear elastoplastic constitutive model, including features such as small strain stiffness degradation, plastic straining due to primary compression, and a Mohr-Coulomb-type limit envelope. The adopted motions correspond to acceleration time histories recorded at rock outcrops. One of the selected records presents a horizontal particle motion that is significantly polarized and, therefore, directionality effects are expected to be more relevant. Analyses were performed following the so-called complete rotational approach, which implies the execution of calculations where the input motion is derived from the linear combination of the horizontal as-recorded components, accounting for different, non-redundant, incident angles. Obtained results provided relevant insights regarding the effects of ground motion directionality on the stability of slopes and showed that ignoring this aspect can lead to non-conservative evaluations. In addition, the comparison between the numerical nonlinear dynamic analyses and the Newmark method allowed us to broaden the interpretation of the results of the traditional rigid block method as a general stability index.