Simulation of cohesive-frictional artificial soil-to-blade interactions using an elasto-plastic discrete element model with stress-dependent cohesion

Abstract

The discrete element method (DEM) has become a valuable computational technique for simulating soil dynamic loading during bulldozer cutting processes. It allows for the virtual design of Ground Engaging Tools (GETs) and predicting energy expenditure during earthmoving operations. Few studies exist for modeling dynamic soil-cutting processes of soils exhibiting elasto-plastic behavior with stress-history-dependent cohesive soil behavior. The study aimed to calibrate an elasto-plastic DEM soil model, with cohesion, for a cohesive-frictional artificial soil and predict soil reaction forces from soil-to-blade interaction. Plackett-Burman screening design of experiment (DOE) and inverse profiling techniques were applied to calibrate the elasto-plastic DEM soil model, with cohesion, predicting soil compaction energy with a percent relative error (PRE) of 3 % and maximum normal stress (PRE of 1 %) using cohesive-frictional artificial soil in a uniaxial confined compression test. Validation of the calibrated DEM soil model resulted in good prediction of soil reaction forces versus blade displacement for a narrow planar blade, a wide planer blade, and a geometrically scaled curved bulldozer blade, with RMSE values of 2.04 N, 14.89 N, and 7.42 N, respectively. The findings showed that elasto-plastic soil behavior with stress-dependent cohesion can be modeled using DEM for simulating the cutting and moving of earthen materials, offering valuable insights for optimizing GET design and development of digital twins of earthmoving operations.