This paper offers dynamic centrifuge model tests on a piled raft foundation supporting a plate-shape building on soft clayey ground. The dynamic tests were conducted in 50 g field in order to investigate the lateral load sharing ratio of piles and sectional forces on the piles during large earthquakes. The model ground was made of wet tamping loam to reproduce the soft clayey ground with a shear wave velocity from 100 m/s to 200 m/s. The vertical load sharing ratio of piles was about 20 % before and after the shaking and also during the shaking. On the other hand, the lateral load sharing ratio of piles was considerably small about 10 % even in a large earthquake. In addition, the shear force and bending moment acting on the piles were evaluated by simplified equations based on Mindlin’s solution which is derived from making some approximations and assumptions. The experimental results were compared to the calculated values and the lateral load sharing between piled and raft was discussed

Research on foundation response to seismic shaking has shown that full mobilisation of bearing capacity and the promotion of rocking foundation response may be beneficial for structural integrity – particularly in the case of seismic motions that exceed design limits. Full mobilisation of foundation bearing capacity acts as a safety valve, limiting the inertia loading on the structure. While most research has focused on rocking shallow or embedded foundations, a rocking pile group foundation has attracted much less attention. Today, the growing need for retrofit of existing structures, which were typically designed and constructed long before the adoption of modern seismic design provisions, makes such a rocking pile group design an appealing solution, as it allows optimisation (or even complete avoidance) of foundation retrofit, which can be a major operation, especially in dense urban environments. Compared to the bearing capacity of surface shallow or embedded foundations, the failure modes of a rocking pile group are more complex. They may involve structural damage below the ground level, which is perceived as something to avoid, as it is difficult to repair or even detect. To shed more light on the problem, the present study develops a finite element (FE) model of an idealised yet realistic single bridge pier supported by a rocking 2x2 pile group. The FE model employs a carefully calibrated and thoroughly validated kinematic hardening model for the soil, combined with the concrete damaged plasticity (CDP) model for the reinforced (RC) piles and pier. The FE model is subjected to pushover and preliminary dynamic time history analysis. It is shown that the rocking pile group exhibits a ductile response.

Shear walls are earthquake-resistant elements in buildings. Superstructures and foundations are generally designed separately. The knowledge of the effects of shear walls on pile stress is limited. In this study, to investigate the effects of shear walls on the stress of piles, centrifuge tests and numerical analyses of the soil-pile superstructure for frame and shear wall structures were performed. The following results were obtained: 1) For the frame structure, the bending moments at the pile heads in the middle rows were larger than those in the front and tail rows. This was because the rotational restraint of the pile heads in the middle rows was higher than those in the front and tail rows. 2) For the shear wall structure, the bending moments at the pile heads in the middle rows were as large as those in the front and tail rows. The bending moment in the middle rows caused by the vertical forces was opposite to that caused by the horizontal forces because of the rotation of the footing beam beneath the shear wall.

In recent years, there has been growing momentum to build structures on the sea. Examples include bridges, offshore wind turbines and SEP ships (Self-Elevating Platforms). When constructing such structures, it is necessary to assess their safety during and after construction. The soil-structure evaluation is in this regard of significant importance. In particular, there is a possibility of collapse due to liquefaction, seismic and wind loads. It is therefore important to consider the behaviour of foundations and the ground of structures built on the sea when they are subjected to cyclic horizontal seismic and wind loads. Simulation of the behaviour of a pile installed in subsoil needs three-dimensional analysis because two-dimensional plane strain assumption is inapplicable. In addition, it is difficult to reproduce the actual loading conditions in pile loading tests in terms of time and cost. In this study, a three-dimensional analysis of the behaviour of a single pile and the ground under monotonic and cyclic horizontal loading was carried out using a finite element analysis method based on Biot's two-phase mixture theory with a cyclic elasto-plastic constitutive equation for sand. After verifying the validity of the analysis method and parameters through simulations of monotonic loading experiments conducted on actual ground, numerical experiments were conducted under cyclic loading.

This paper presents a novel elastodynamic model for the kinematic response of single, end-bearing piles embedded in vertically inhomogeneous soils, excited by vertically propagating P-waves. The response in terms of displacements is expressed in the form of a generalized Fourier series, and extends a previous work of some of the authors to account for soil inhomogeneity. Contrary to formulations for homogeneous soils, the associated Fourier coefficients are now coupled, and can be obtained as a solution to a system of algebraic equations of rank equal to the number of soil modes considered in the analyses. The pile is modelled as a rod, using the strength-of-materials solution, and the soil as an approximate continuum of the Tajimi type. For the axisymmetric problem at hand, the Tajimi approximation lies in adopting the physical motivating assumption that the vertical normal and vertical shear stresses in the soil are controlled exclusively by the vertical component of the soil displacement. This approximation results in reducing the number of governing elastodynamic equations to one, which satisfies the equilibrium in the vertical direction. The proposed model can predict the steady-state and transient response of piles in inhomogeneous soil strata over a rigid rock, subjected to vertically-propagating harmonic compressional waves and actual earthquake recordings, respectively. The predictive power of the model is verified through comparisons with finite element analyses. Results are presented in terms of a kinematic response factor which relates the motion of the pile head to the free-field surface motion. It is shown that a pile foundation may significantly alter the motion transmitted to the base of a structure.

In recent years, the development of macroelement approaches to include the macroscopic nonlinear response of soil-foundation systems in the assessment of structures is receiving an increasing interest by virtue of the minimal computational effort required. However, existing formulations commonly neglect any undrained or partly drained soil behaviour, that may be crucially important for simulating the response under dynamic loading. The present study provides an insight into the effects of the hydro-mechanical coupling of the soil on the macroscopic multiaxial cyclic response of shallow foundations. This is accomplished through a series of nonlinear transient analyses on a fully coupled soil-foundation numerical model implemented in OpenSees, providing an explicit description of the pore water pressure build-up induced by the nonlinear soil behaviour. The numerical study explores different assumptions for the hydraulic regime, from drained to undrained conditions. The effect of the volumetric-deviatoric coupling on the cyclic response of the reference foundation is examined, highlighting the key role played by the drainage conditions on the stiffness and dissipative features of the foundation system. The effect of non-linearity on the above effects is discussed and interpreted in terms of degradation of the system response at the macro scale.

Thanks to their resilience, self-centering frame structures are increasingly being utilized in practice. Releasing the fixity at some of the frame joints allows for dissipation of seismic energy through rocking, while un-bonded pre-stressed rebars can be used to enhance the self-centering mechanism. So far, most studies focused on such frames resting on a rigid foundation, not accounting for soil-structure interaction (SSI). This paper presents preliminary results of shaking table tests, recently conducted on the 4 m × 4 m shaking table of Tongji University (China). Two six-floor self-centering frame structures (SCF) made of reinforced concrete (RC) were tested at a 1:10 scale on: (a) fixed-base, with the structure directly connected to the shaking table; and (b) compliant-base, with the structure founded on a pile-group on a layer of sand-sawdust-mixture, filled in a 3 m diameter cylindrical container, thus accounting for SSI. The comparison of the test results allows for quantification of the role of SSI. A tensile anchor was specially designed to apply the prestressing and anchoring of high-strength rebars to the RC columns. The uplifting of the self-centering frame was captured by an NDI Optotrak Certus dynamic measurement system. The experimental results indicate that both self-centering frames can successfully self-center after the earthquake, experiencing only limited uplifting and sustaining limited structural damage of the RC columns. Subjected to successive seismic excitations, the natural frequency of the SSI system was found to decrease, indicating some accumulation of damage, but to a lesser extent than that of the fixed base system. The presented experimental results offer useful preliminary insights for the design of such self-centering structures.