The K & uuml;& ccedil;& uuml;k & ccedil;ekmece-Avc & imath;lar corridor of the D100 highway constitutes a critical component of Istanbul's transportation infrastructure. Given its strategic importance, ensuring its operational continuity following the anticipated major Istanbul earthquake is imperative. The aim of this study was to investigate the liquefaction-induced geotechnical risks threatening the K & uuml;& ccedil;& uuml;k & ccedil;ekmece-Avc & imath;lar segment of the D100 highway. Initially, the study area's liquefaction susceptibility was assessed through Liquefaction Potential Index mapping. Subsequently, post-liquefaction ground displacements were quantified using semi-empirical methodologies and advanced numerical analyses focused on representative critical sections. Numerical simulations incorporated various constitutive models for liquefiable soils, enabling a comparative assessment against semi-empirical estimations. The results revealed that semi-empirical approaches systematically overestimated the lateral displacements relative to numerical predictions. Moreover, the analyses highlighted the sensitivity of model outcomes to the selection of constitutive parameters, underscoring the necessity for careful calibration in modeling liquefiable layers. Despite considering the most conservative displacement values from numerical analyses, findings indicated that the D100 highway is likely to experience substantial damage, potentially leading to extended service disruptions following the projected seismic event.

Seismic liquefaction is one of the most devastating natural hazards that can cause significant damage to structures and infrastructure. The liquefaction behaviour is simulated in the finite element code PLAXIS by the UBC3D-PLM constitutive model that is 3-D generalized formulation of the 2-D UBCSAND model developed at the University of British Colombia. The UBC3D-PLM model used in this work was successfully employed in many recent studies, e.g. to evaluate the liquefaction effects on the seismic soil-structure interaction, to assess the dynamic behaviour of earthen embankments built on liquefiable soil and to investigate the seismic performance of offshore foundations. Moreover, UBC3D-PLM model involves many input parameters to model the onset of the liquefaction phenomenon. Therefore, their determination becomes a crucial concern. Previous studies elaborated a specific formulation that requires the corrected Standard Penetration Test (SPT) blow counts as input. However, the Dilatometer Marchetti Test (DMT), compared to the SPT, is more sensitive to several factors that affect the liquefaction resistance such as aging, stress history, overconsolidation and horizontal earth pressure. For this reason, a new parameter selection procedure, which uses the horizontal stress index derived from DMT, was developed in this study. The new relationships were applied for determining the initial parameters of the UBC3D-PLM model to describe the behavior of several liquefiable deposits located in eastern Sicily (Italy) that experienced destructive earthquakes in the past. For each site, the model was calibrated to the DMT-based liquefaction triggering curve, developed by combining DMT correlations with the current method based on SPT test, by the simulation of cyclic direct simple shear tests (CDSS). Finally, CDSS tests were performed by means of the CDSS device at the Soil Dynamics and Geotechnical Engineering Laboratory of the University Kore of Enna (Italy). This allowed to validate the applicability of the proposed procedure in simulating the liquefaction behavior of sandy soils.