Hydraulic conductivity plays a significant role in the evolution of liquefaction phenomena induced by seismic loading, influencing the pore water pressure buildup and dissipation, as well as the associated settlement during and after liquefaction. Experimental evidence indicates that hydraulic conductivity varies significantly during and after seismic excitation. However, most previous studies have focused on experimentally capturing soil hydraulic conductivity variations during the post-shaking phase, primarily based on the results at the stage of excess pore water pressure dissipation and consolidation of sand particles after liquefaction. This paper aims to quantify the variation of hydraulic conductivity during liquefaction, covering both the co-seismic and postshaking phases. Adopting a fully coupled solid-fluid formulation (u-p), a new back-analysis methodology is introduced which allows the direct estimation of the hydraulic conductivity of a soil deposit during liquefaction based on centrifuge data or field measurements. Data from eight well-documented free-field dynamic centrifuge tests are then analysed, revealing key characteristics of the variation of hydraulic conductivity during liquefaction. The results show that hydraulic conductivity increases rapidly at the onset of seismic shaking but gradually decreases despite high pore pressures persisting. The depicted trends are explained using the KozenyCarman equation, which highlights the combined effects of seismic shaking-induced agitation, liquefaction, and solidification on soil hydraulic conductivity during the co-seismic and post-shaking phases.

Retaining walls and other waterfront structures were seen to suffer severe damage due to soil liquefaction in previous earthquakes. As part of the LEAP project, cantilever retaining walls with loose, saturated backfill were tested at various centrifuge centres participating in this endeavour. The toe of the retaining wall penetrated about 0.5 m into the dense sand layer underlying the loose sand layer. Retaining walls with different ratios of the retained height h over the penetration depth d were tested. As part of the LEAP project, additional testing was carried out at Cambridge to consider the effect of the wall size on its deformation following liquefaction. It will be shown that a larger wall will suffer more rotation and wall top displacement than a smaller wall with the same h/d ratio. This can have implications for numerical modelling in terms of how well the constitutive models capture the suppressed soil dilatancy at higher confining pressures.