Particle-Morphology-Aware Prediction of Cyclic Liquefaction Resistance in Calcareous Sands

Abstract

The prediction of cyclic liquefaction resistance in highly irregular geomaterials, such as calcareous sands, remains a significant challenge in modern geotechnical and earthquake engineering. Unlike terrigenous silica sands, which possess relatively uniform particle morphologies and solid mineral structures, calcareous sands are biologically derived, resulting in extreme angularity, high intra-particle porosity, and distinct susceptibility to particle crushing. Traditional empirical models for evaluating the cyclic stress ratio required to trigger liquefaction often fail when applied to these marine deposits, primarily because such models omit the profound influence of microscopic particle shape on macroscopic volumetric strain and pore water pressure generation. This paper presents a comprehensive, particle-morphology-aware predictive framework designed specifically for calcareous sands. By utilizing advanced dynamic image analysis to quantify particle shape parameters including sphericity, convexity, and aspect ratio, we establish a robust empirical relationship between these microscopic morphological descriptors and macroscopic cyclic liquefaction resistance. Extensive undrained cyclic triaxial testing was conducted on samples with varying morphological characteristics under multiple confining pressures. The experimental results demonstrate that decreased particle convexity and sphericity significantly enhance interlocking, thereby increasing resistance to initial pore pressure generation but promoting sudden catastrophic failure upon the onset of particle breakage. A novel mathematical model is proposed, incorporating a composite morphological index alongside traditional state parameters. This research provides a critical step forward in the reliable design of marine infrastructure, offshore platforms, and coastal reclamations subjected to cyclic loading.