With the soil constitutive model developed for the dynamic large deformation of soft soil, a finite element model was constructed to simulate the soil–underground structure static and dynamic coupling interaction system, in which the dynamic contact between soil and underground structure and the dynamic damage of reinforced concrete were considered. The seismic performance and damage mechanism of the large underground subway station in different soil foundations were investigated and evaluated in detail. It was proven that the interaction mechanism between the soil and underground structure would change with the type of soil foundation and the peak acceleration of input ground motion, which can be evaluated by the interaction coefficient between the soils and underground structure. At the same time, some regulations on the seismic performance of underground structure in China code are already unsuitable for evaluating the seismic performance of a large underground subway station in the soft soil foundation or subjected to a strong earthquake, which should be investigate by the specialized time–history analysis method instead. In this study, five seismic performance levels were firstly defined by the inter-storey drift angle of underground structure and corresponding earthquake damage states.

The near-surface part of the crust, also called the skin of the earth, is the arena of human activity of which the stiffness is of great concern to engineers in infrastructure construction. The stiffness reduction of near-surface geomaterials also plays a vital role in geohazards triggering. However, the physical mechanism behind the material softening is still not fully understood. Here, we report a coseismic shear-wave velocity reduction in the near surface by up to a few tens of percent during the strongest shaking from the 11 March 2011 Tohoku-Oki Earthquake and a subsequent two-stage healing process including a rapid recovery within a few minutes and a slow recovery over many years. We also present a theoretical contact model between mineral grains in geomaterials containing multiple metastable contacts at small separations due to the oscillatory hydration interaction, which can explain the emergence of different stages in the healing process.

Fundamental period has been recently proposed to be an alternative to time‐averaged shear‐wave velocity to 30 m (⁠$VS30$⁠) and used as a proxy for site amplification in recent ground‐motion prediction equations. In this study, we propose a simple approach to construct continuous‐variation velocity models and present analytical‐numerical solutions to the fundamental period for the following three cases: (a) shear‐wave velocity increasing or decreasing linearly with depth, (b) shear‐wave velocity increasing as a power of depth, and (c) shear modulus increasing or decreasing linearly with depth. Eight practicable methods of estimating the period from shear‐wave velocity profile are also developed and discussed. The accuracy of the eight methods is evaluated by comparing the estimated values with the analytical solutions and empirical results. The results show that the linear velocity model method has high correlation coefficient and low standard deviation among the eight approximate methods for both the comparison between estimated and analytical periods and that between the estimated and empirical periods. In addition, we find that there is a significant improvement in the accuracy after modifying four times the total travel time (previous solution) as $π$ times the total travel time (proposed solution) to give a preliminary estimation of fundamental period for profiles with at least two layers.

The small strain shear modulus G_max is an important parameter in geodynamic calculation and earthquake response analysis of site. To have a better fundamental understanding of the deformation behaviour of sand-fines mixtures, a total of 105 bender element tests have been performed. It is found that the G_max is significantly dependent on fines content FC, initial effective confining pressure σ′_3c and void ratio e. Furthermore, FC has an obvious influence on the trend of Gmax – e curve, but the impact decreases as σ′_3c increases. According to Hardin’s equation, a modified model and an evaluation method of G_max of sand-fines mixture are proposed and verified. Tests results and example of application indicate that the modified equation accounting for the influence of FC can predict the G_max of sand-fines mixture satisfactorily.