Calibration of contact model parameters for aggregates: considering particle shape and moisture content

Abstract

Aggregates, as typical granular materials, are widely used in pavement and construction engineering, where inter-particle contact significantly influences their mechanical properties. The discrete element method (DEM) effectively simulates nonlinear interactions in aggregate systems, but its accuracy largely depends on contact model parameters. However, most existing calibration approaches simplify particle shape and neglect moisture content effects, leading to inaccurate parameter estimation and limited generalizability. To address this, a systematic and efficient calibration framework is developed that explicitly incorporates aggregate morphology and moisture. Distinct from previous approaches that often focus on either dry or wet conditions and rely on a single modeling representation, the proposed framework integrates particle morphology reconstruction, sequential design-of-experiments (combining Plackett–Burman screening, steepest ascent, and Box–Behnken optimization), and dual contact models (Hertz–Mindlin and Johnson-Kendall-Roberts (JKR)) into an integrated workflow, thereby enhancing both physical fidelity and practical applicability. SMA-11 gradation aggregates made of basalt were used as the study material, and their real geometries were captured via 3D scanning and modeled as balls, clumps, and rigid blocks to evaluate shape influence. Simulations for dry aggregates yield average relative errors of 1.37 %, 1.98 %, and 1.72 % for the three shaped aggregates, while the AOR error for wet aggregates remains below 2 %. These results confirm the accuracy and robustness of the proposed calibration approach, offering reliable support for high-fidelity DEM modeling of aggregate behavior across various conditions.