Rockfall, a typical kinematic process near slope surfaces, poses challenges due to its high energy, unpredictability, and dependence on both analytical methods and topographic accuracy. In this study, the classical three-dimensional discontinuous deformation analysis (3D-DDA) is reformulated using cone complementary theory to better capture nonlinear contact interactions. Unmanned aerial vehicle (UAV) photogrammetry is employed to construct an accurate numerical model of the complex slope terrain at the No. 2 transverse tunnel of the Layue Tunnel along the Sichuan–Tibet Railway. Simulations reveal that dangerous rock masses primarily fail by sliding, with both isolated boulders and massive rock formations exhibiting diverse 3D motion patterns. The rockfalls traverse the tunnel construction zone, threatening traffic and river safety before deposition. UAV-assisted 3D-DDA effectively characterizes trajectories, displacements, and kinetic energy, enhancing prediction of impact zones and deposition sites. These results provide insights into rockfall mechanisms and support hazard mitigation strategies.
Liu et al. (2026) studied this question.