MPM Modeling of Internal Collapse Under Levees Caused by Underground Openings

Abstract

Evaluating the mechanisms of internal collapse in earth-retaining structures, such as levees, is essential for ensuring their stability and safety. This study investigates the mechanics of underground trapdoor collapse in frictional and cohesive soils, which is relevant to the behavior of water-retaining structures like levees, e.g., subjected to pipe leakage. The Material Point Method (MPM), an advanced numerical tool capable of simulating large deformations, is used to model an underground trapdoor collapse process in order to validate its accuracy in addressing such problems. Numerical simulations based on experimental benchmarks are performed to explore ground surface settlement and failure mechanisms. The findings show that the MPM effectively captures the collapse process, aligning with experimental data, empirical relations, and previous literature. The research highlights the importance of determining strength parameters within the correct stress range to avoid overestimating the soil strength. Additionally, the research highlights the role of cohesion in predicting collapse outcomes and surface settlement. Following the validation process, this thesis examines the formation of underground cavities and the collapse behavior of levee systems subjected to seepage. Using MPM, the failure initiation and progression of underground openings are explored with different geometries and under various soil conditions, including one-phase total stress formulation, one-phase submerged formulation, and two-phase fully coupled hydro-mechanical formulation. A sensitivity analysis is conducted to assess the impact of mesh size and material point density on simulation accuracy. The results highlight the complex interactions between opening location, opening size, and seepage in cavity formation and levee stability. Failure progression and collapse results are compared with experimental centrifuge test data. The study highlights the challenges of numerically simulating this complex problem using advanced numerical tools and provides insights into the factors influencing internal collapse in levees.