This paper analyses the seismic response of natural slopes within a probabilistic framework with reference to the Italian seismicity. Once combined with a proper semi-empirical relationship linking the permanent earthquake-induced displacements of slopes to one or two synthetic ground motion parameters, the probabilistic approach allows to evaluate the permanent displacements for a given return period and a given yield seismic coefficient, the latter accounting synthetically for the seismic resistance of the slope. The results are illustrated in terms of hazard maps showing the distribution of the displacements at a regional scale and are therefore useful to identify slopes in areas that are prone to earthquake instability. In detail, scalar and vector probabilistic approaches based on the ground motion parameter PGA and the couple PGA, PGV, respectively, are employed to produce hazard maps for two Italian regions characterised by severe seismicity. Finally, the results obtained through the fully probabilistic approach are critically analysed and compared with those determined through a simplified decoupled method that combines a standard probabilistic seismic hazard analysis (PSHA) with the displacement semi-empirical relationships. The maps presented in this work represent a reliable and simple tool for preliminary, screening level analyses of natural slopes in terms of seismic-induced displacements.

Rock slopes with a dominant, persistent joint set are naturally occurring and have two common failure mechanisms: flexural toppling and cross-joint shear. Although the static stability of this slope category has been well studied, the stability of these slopes during earthquakes has remained opaque. In this study, the dynamic stability of persistently-jointed rock slopes is investigated using a dynamic implementation of the bonded particle model. In response to the dynamic motions, the slopes experience dynamic forces that break interparticle bonds and begin to fail. The study shows that the behavior of the slope is sensitive to the frequency content of the input ground motion. Further, the nature of the frequency sensitivity is dependent on the initial failure mechanism and the accumulation of damage over time. Persistently-jointed rock slopes that fail in flexural toppling are sensitive to low-frequency dynamic loading. Slopes failing in cross-joint shear are sensitive to tuning ratio. The nature of the frequency sensitivity has implications for the applicability of simplified (e.g., pseudostatic) analysis methods. Increased damage accumulation in rock slopes results in softening, which decreases the slope’s natural frequency and, therefore, the critical input frequency.

Excess pore water pressure has a significant effect on the occurrence of landslides under earthquake shaking. For seismic displacement calculation using the sliding block method, the prevailing idea about the generation of excess pore water pressure is mostly in fully saturated soils. However, the existence of shallow landslides with partially saturated soils has been demonstrated. In the present study, an empirical partially saturated excess pore water pressure model is applied to the nonlinear analysis of flexible sliding block method for sandy slope, and the degree of saturation ranges from 10% to 100%. The dynamic shear modulus and damping ratio are considered, and the effect of degree of saturation variation on yield acceleration and slope displacement is investigated in this study. Detailed explanations are made to the effects of suction stress, soil mass, and excess pore water pressure on the seismic response of slopes at different degrees of saturation. The results clearly demonstrate that the variation of degree of saturation is of significance, the excess pore pressure ratio increases with saturation, showing an exponential function relationship, and the change of yield acceleration and displacement with saturation is complex, which can be divided into two cases: suction stress dominant and excess pore water pressure ratio dominant.

Many existing earth dams were designed before the establishment of a seismic code or in a period when the seismicity of the dam sites was approximately assessed. Accordingly, the evaluation of their seismic performance is a crucial issue and, for both new and existing dams, the design or retrofitting procedures should aim to the improvement of their seismic resilience. In this framework the dynamic analyses devoted to the prediction of the dam seismic performance are usually carried out accounting only for the horizontal component of the expected ground motion. However, the vertical seismic motion can play an important role since it could affect the earthquake induced plastic mechanisms and the magnitude and distribution of the corresponding permanent strains. In this vein, the paper presents the results of a large set of 2D finite-elements dynamic analysis carried out with reference to a zoned earth dam located in a high seismic area of Southern Italy, using an isotropic hardening elasto-plastic hysteretic model. The analyses point out the influence of the amplitude and energy content of the vertical component of the input motion on the dam performance. Specifically, the numerical results are presented and discussed focusing on the on the magnitude of the crest settlements, considered a good proxy of the seismic-induced level of damage, and on their relationship with the seismic parameters describing the energy and frequency content of the vertical component of the input motion.

In Japan, there has been an increase in the failure of dry masonry retaining walls due to strong earthquakes especially in urban residential areas. Currently, there are no established optimal seismic reinforcement methods for such retaining walls due to issues related to workability and construction cost. This study proposes a novel reinforcement method where tilting reinforcement bars were installed and integrated from the top of the retaining wall. The reinforcement mechanism was attempted to be investigated by conducting DEM analysis of previous centrifuge tilting experiments that were conducted to examine the seismic resistance of such reinforcement method. Results indicate that DEM can be used to successfully reproduce the centrifuge experiment in terms of the horizontal displacement of the retaining wall and the behavior of the mobilized earth pressure. It was also found that the stiffness of the reinforced retaining wall increased due to the downward vertical compressive force on the bottom retaining wall block in response to the axial force in the tensile direction of the reinforcement.

The seismic slope performance is usually evaluated by estimating the permanent downslope displacement of potential sliding mass based on Newmark-type sliding block procedures. As a key input of these procedures, the critical acceleration (a_c) is typically derived as the horizontal seismic acceleration that produces the unity factor of safety in iterative pseudo-static slope stability analyses. This process is time-consuming especially when considering a number of slope configurations (e.g., in regional landslide hazard mapping). This preliminary study aims to present a single prediction equation for a_c using a neural network. The input parameters of the prediction equation include the cohesion, friction angle, and unit weight of soil, as well as the slope height and slope angle, which are common in practice. The dataset for model development is obtained from iterative pseudo-static stability analyses for thousands of generic slope configurations. The results indicate that the proposed model can reasonably predict a_c under various scenarios. A slope example is used to illustrate the application of the prediction equation in seismically-induced slope displacement analysis. Further efforts are needed to enhance the generalization capability of the a_c prediction.

Dynamic properties of Red river dyke sand such as vertical Young’s modulus and damping ratios were experimentally investigated by using cylindrical specimens (diameter Di = 5cm, and height H = 10 cm). The specimens were taken from the boreholes at the foundation of the Red river’s right dyke with different depths (K73+500-K74+100). At some stable stress states, vertical Young’s modulus was measured with the non- contact transducer, by applying a series of unload-reload cycles with vertical strain single amplitudes varying in the range of 0.001% to 0.1%, at different confining stresses. Test results show that the degradation of the vertical Young’s modulus and the increase of damping ratio on the single vertical strain amplitude were observed and could be well simulated by empirical equations.

Unlike the conventional Newmark slope model (CNM), a potentially sliding soil block may undergo small displacement immediately before sliding in a thin layer underneath. Hence, a Spring-Supported Newmark model (SSNM) has been developed by adding a linear spring for the pre-sliding yield displacement u₀ to a slider of the CNM, which has demonstrated a good reproduction of slope sliding in a model test. Dynamic responses of typical slopes to harmonic and earthquake motions have been calculated incorporating the yield displacement u₀ of only a few millimeters to find out that the acceleration for slide initiation tends to overshoot the yield acceleration of the CNM by a larger margin. Accordingly, the slope displacement tends to decrease by more than a half, demonstrating a great impact of the SSNM considering a small yield deformation u₀ as a specific input parameter.

Most of the slopes that are found in nature, occurred naturally without any intervention by humans. Failures of these slopes can be triggered due to earthquakes and heavy rainfall. The failure characteristics such as the slip surface shape and the collapsed magnitude vary greatly due to the dominant direction of seismic motion, heterogeneity of the ground, and unsteady infiltration of rainfall. Although an extensive number of experimental studies have been accomplished to understand slope failure, it is not enough to grasp the probabilistic nature of slope failure phenomena.

In this study, experimental and statistical analysis was carried out using the sand pile model to understand the slope failures and their magnitude. The sandpile model is a model that gradually distributes sand to a pile of sand and repeatedly observes the collapse of the sandpile. The experiment using of silica sand no.8 (fine sand) was carried out with different base plates of 5cm to 12cm at a constant rate of adding sand. Shape variations of the sandpile were captured with a depth camera.

Results show that the number of slope failures decreased when the diameter of the sandpile increased. However, slope failure mass and time interval between consecutive slope failures increased. The statistical distributions of slope failure mass and time intervals exhibit deviations from a normal distribution pattern. Notable geometric changes were predominant on the apex of the sandpile. Results conclude that the magnitude of the slope failures and the time taken to occur a failure depends on the metastable slope angles.