An integrated framework for mesoscale modeling of concrete elements based on spherical harmonic expansion

This study presents a novel integrated mesoscale modeling framework for concrete that couples high-resolution laser scanning, spherical harmonics-based shape representation, stochastic aggregate synthesis, and density-controlled packing by means of discrete element simulations. Laser scanning is first employed to capture aggregate morphologies with sub-millimeter precision, providing empirical input for the spherical harmonics decomposition. A novel phase-scaling spectral perturbation method is proposed to regulate stochastic shape variability in the reconstructions while preserving fine-scale features. This method addresses unrealistic small-scale surface characteristics of unconstrained conventional randomization, such as wavy surface textures on the sharp edges and angular facets. A discrete element-based modeling approach is proposed to incorporate a velocity-driven compaction algorithm that efficiently achieves uniform aggregate packing. The full modeling pipeline is validated by means of data collected from experimental tests, with simulation outputs benchmarked against consolidation indicators, including aggregate spacing, mortar distribution, and centroidal shifts. This framework advances mesoscale concrete modeling towards a validated, mechanistically grounded diagnostic tool capable of quantitatively assessing consolidation performance in practical construction scenarios. • Laser scanning enables rapid and accurate aggregate geometry acquisition. • Spherical harmonics compactly represent irregular aggregate shapes across scales. • Phase-scaling controls morphological variability while preserving angular features. • The porosity-driven DEM approach achieves uniform packing with fast computational convergence. • Simulation validated by comparison with experimentally obtained consolidation metrics.