Dam-break granular flows on various slopes and substrates: Large-scale experiments and numerical modeling

This study presents a series of large-scale dam-break granular flow experiments along a solid surface, systematically varying grain size, basal roughness, and slope inclination. This dataset is compared with two continuum numerical models: a Navier–Stokes solver (Basilisk) and a depth-averaged model (IMEX_SfloW2D), both implementing μ ( I ) -rheology with parameters independently calibrated from small-scale steady avalanche experiments. The main findings of this study are fourfold: (i) both models accurately reproduce large-scale experiments from horizontal to steep rough slopes (up to 36°), although the Navier–Stokes model shows limited sensitivity to the μ ( I ) -rheology; (ii) beyond the dominant role of inclination, basal conditions critically influence flow dynamics and deposition within a narrow parameter range; (iii) both models qualitatively reproduce the observed variations induced by basal conditions using rheological parameters derived from empirical relationships accounting for both internal and basal dissipation; and (iv) at steep slopes, experiments exhibit higher mobility than predicted by the depth-averaged model, reminiscent of the high mobility commonly reported for natural flows. Overall, this study advances landslide modeling across diverse slope and substrate conditions, highlights the strengths and limitations of current models, and provides a valuable large-scale benchmark dataset, which would be of interest for hazard assessment.