Evaluation of the anti-collision performance of reinforced concrete protective wall for railway bridges under high-speed train impact

With the expansion of railway bridge networks, the exposure of train operations to elevated environments has grown; in such settings, any derailment carries higher consequence severity, underscoring the need to predict post-derailment collision dynamics. This study focuses on a typical Chinese dual-track ballastless railway bridge system and develops a multi-scale dynamic collision model for high-speed trains and reinforced concrete protective wall (RCPW). The model employs a mass-spring-damper system to represent the vehicle’s non-impact dynamics and applies a zoned hybrid approach that integrates fiber-beam and three-dimensional solid elements, thereby enabling accurate RCPW damage simulation while reducing computational time by approximately 70 %. Furthermore, a novel bogie motion plane method is introduced, which analyzes the train’s dynamic behavior from wheel-track board contact failure to collision by computing spatial angles and displacement time histories. An augmented uniform design sampling approach, combined with a Gaussian process surrogate model and Sobol sensitivity analysis, is employed to establish for evaluate the train-RCPW coupling effect. Key findings reveal that (1) lateral train speed (v_x) governs collision patterns—v_x ≤ 2.5 km/h results in an initial collision with the base slab followed by a secondary collision with the RCPW, while v_x ≥ 7.7 km/h leads to a direct collision with the RCPW—and (2) optimizing the yield strength of longitudinal reinforcement is more effective for energy dissipation than increasing concrete compressive strength.