Seismic behaviour of caisson foundations supporting bridge piers has been attracting the interest of researchers over the last few years, mainly due to their complex interaction phenomena with the soil and the superstructure, usually referred to as Kinematic and Inertial Interaction effects. Such embedded, rigid, and very large foundations are typically resorted to when dynamic horizontal forces generated by strong earthquakes are to be resisted. One of the main parameters influencing the Dynamic Soil-Structure Interaction of such structures is the fundamental, compliant-base period, T_eq. In this paper, the fundamental period of caisson foundations supporting bridge piers is computed from the results of dynamic centrifuge tests performed on reduced-scale models at University of Cambridge, UK. A rigid and a flexible bridge pier founded on a cylindrical caisson foundation, embedded in a typical alluvial deposit, were reproduced. The two systems were subjected to a series of both sine waves and acceleration time histories from ground motion records. The analysis of the experimental data allowed to estimate the fundamental period of the system, which was then compared with those resulting from relationships available in the literature. It is shown that the empirical equations may be profitably used at a preliminary design stage, provided that nonlinear soil behaviour, bringing the soil stiffness to decay during the seismic event, is considered.

The past large earthquakes have caused masonry train platforms to collapse due to large deformation. The collapse of the platform can jeopardize the safety of train operations and pose a threat to human life. Therefore, the purpose of this study is to confirm the seismic performance of the embankment-type platform with masonry during an earthquake and to propose appropriate seismic reinforcement. In order to understand the behavior of the masonry train platform during an earthquake, the shaking table tests were carried out on 1/3 model in the gravitational field. As a result of the experiment, it was confirmed that embankment-types platform with masonry may collapse in a large earthquake in Japan. Therefore, we developed an optimal seismic reinforcement method based on the behavior confirmed in the experiment. The shaking table tests were carried out again with a model simulating the proposed seismic reinforcement method. As a result of the experiment, it was confirmed that the seismic resistance performance was greatly improved compared to before reinforcement.

During large earthquakes, pile heads are subjected to repeated axial load variations owing to the overturning moments caused by the inertial forces of the superstructure. Such axial load variations may cause the residual and unequal settlement and damage to superstructures and lifelines. However, the characteristics of pile settlement during large earthquakes remain unknown. Therefore, this study investigated the settlement characteristics of pile foundations during large earthquakes using a dynamic centrifuge test on a soil–pile–superstructure system. The following results were obtained. (1) The residual settlement of piles caused by a large earthquake could be estimated using the maximum axial load variation, skeletal curve, and correction factor. (2) The unequal settlement between the front and rear piles could be estimated based on the relationship between the dynamic vertical displacement and the axial load variation instead of the skeletal curve. Moreover, it was found that the unequal settlement based on the skeletal curve may have been significantly overestimated

The effect of liquefaction on the piled raft foundation with grid form DMWs (Deep Mixing Walls) is studied numerically. The building simulated is a 12-story RC seismically isolated residence on soft ground in Tokyo. The 3DFEM SSI (Soil Structure Interaction) model that was calibrated using seismic observation data of the 2011 off the Pacific coast of Tohoku earthquake is used. The constitutive model that can consider post-failure softening is used for the DMWs. The input wave is an extremely rare earthquake under the Japanese Building Standard Law. The analysis results shows that the area of tensile stress failure is extended in the lower part of the DMWs, mainly at both ends in the shaking direction. Some parts of the failing area lost strength due to softening. And outside of the grid form DMWs, the layers are liquified almost completely. However, the overall grid-form DWMs dose not lose the function to prevent the occurrence of liquefaction inside the grid and liquefaction-induced deformation. And the cross-sectional forces of the piles are kept within the damage limit state.

In this study, dynamic centrifuge model tests were conducted on various foundation types with identical embedments to investigate the characteristics of their horizontal and rotational resistance. The results indicated that (1) the nonlinearity of the ground and between the soil and foundation affected the building response. (2) The front and rear side earth pressure was generally consistent with the Coulomb evaluation equations. (3) The difference in the historical characteristics of the front and rear side acted as the horizontal and rotational resistance, which was minimal for each foundation type. (4) The front and rear side earth pressure in the horizontal resistance and the front and rear side vertical friction in the rotational resistance are important factors.

The earthquake probabilities worldwide prove that the structure’s quality, especially in the foundation, is needed. Performance Based Design (PBD) is also required tremendously in current society to make more efficient and effective construction. This research aims to identify and study the emergence of the overturning moment in the foundation under earthquake loads. The foundation was performed by pushover analysis until 70 mm displacement. The foundation was modelled as a constitutive model of a pile group with a 3x3 configuration and embedded in one layer of clay soil with different values of undrained shear strengths (Su), which were 40 kPa – 100 kPa with 20 kPa increment as the representation of the soil stiffness parameters. A whole model was modelled as Beam-on-Nonlinear-Winkler-Foundation (BNWF), and the soil was presented as a series of Winkler springs using the nonlinear p-y method. OpenSees application, which uses the finite element method (FEM), was used for this research. The pushover analysis has shown that stiffer soil makes more significant lateral responses in the same displacement. The sectional and reinforcement-yielding phases of stiffer soil occur in a smaller displacement than the softer soil, except for Su 40 kPa. The overturning moment was identified as the axial responses increased in the lead pile and decreased in the middle and rear piles. Combining the bending moment responses, axial responses, and yielding phase data, it can be concluded that overturning occurs when sectional yielding happens and ends when the bending moment reaches its maximum value. A stiffer soil results in a shorter period of overturning moment. The overturning moment is caused by the change of the relative stiffness in every pile row due to the development of the plastic hinge and the difference of P-Multiplier values in every pile row, where the change impacts the transformation of the axial response. The PBD concept should be more provided as the overturning moment is related to stiffness variation, resulting in different responses.