Insights into the seismic response of pile supported wharves subjected to liquefaction-induced ground deformations: implications for design.

Current seismic design provisions and guidelines for the seismic design of pile-supported wharves (PSWs) explicitly require considering the combined effect of kinematic and inertial loads. However, there is a lack of consensus regarding the adequate simplified methodologies for assessing and combining inertial and kinematic loads for cases in which the wharf structure is subjected to liquefaction-induced lateral spreading ground deformations. Current seismic design provisions and guidelines are based on the distinction between cases of no-liquefaction and cases of liquefaction. The present study showed that the seismic demands on PSWs can be represented following more general framework, that makes distinction between cyclic and lateral spreading phases (or components) of the response. Meaning that, for a given soil deposit, the relative importance of the cyclic and lateral spreading phases is determined by the earthquake intensity. This was achieved by making extensive use of time-history effective stress dynamic-soil-structure interaction analyses to estimate the seismic response and seismic demands on pile-supported wharves founded on liquefiable ground. Two case studies were considered, one large diameter PSW located in southern Italy, and one small diameter PSW located in Wellington, New Zealand. The former is found on a heterogenous medium-to-dense sand deposit, while the latter is placed atop of an uncompacted gravelly land reclamation. Numerical simulations were conducted with FLAC. In both cases, advanced soil constitutive models were used to capture the response of the liquefiable ground, namely PM4Sand and SDm (Stress-Density model). These models were successfully calibrated to well-known empirical liquefaction triggering relationships. Likewise, the analyses also considered the non-linear response of the wharf structure by employing a distributed plasticity model. This strategy (1) ensured the rigorous modelling of modes of deformation and interaction between soil and wharf, and (2) maintained consistency with performance-based earthquake engineering assessments. The large diameter PSW, wharf BAF, was subjected to a more extensive probabilistic seismic demand analysis. In this case, the epistemic uncertainty, represented by using PM4Sand and SDm in parallel, was systematically addressed during the different stages of the study. In essence, for this typology of wharves, inertial demands estimated for the cyclic phase, are well captured by simplified displacement-based methodologies, while kinematic loads were well correlated to 1D ground response estimates. Kinematic loads for the lateral spreading were insensitive to the inertial loads and proved more difficult to be predicted by 1D site response parameters. Lateral spreading displacements were highly dependent on (1) the post-liquefaction strain rate reproduced by the constitutive models, (2) ground motion characteristics. In terms of optimal intensity measures, the seismic demand analysis reveals that the modified acceleration spectrum intensity, MASI, is the most suitable candidate for an optimal intensity measure as in resulted in large correlation with all the response parameters considered, while ranking high in proficiency.