Development and evaluation of an optimized wheel design for tram-train wear and derailment performance

Abstract

The rising demand for dual-operation vehicles capable of seamless transitions between urban tramways and heavy rail networks has intensified global interest in tram-train systems. These systems offer improved operational efficiency, reduced infrastructure costs, and enhanced passenger convenience. However, geometric disparities across rail types present a critical technical challenge. This study addresses the issue through an optimized wheel design compatible with six distinct rail configurations. A baseline wheel profile is proposed, incorporating a specialized geometry and unique rim structure to support smooth transitions at crossings and turnouts. A validated multi-physics simulation model is developed to assess dynamic performance across ten representative rail combinations. Derailment risk and wear number are defined as key quantities of interest (QoIs) and analyzed across three rail categories. Six geometric parameters, selected from a practical design perspective, are evaluated, with three screened out via the Morris sensitivity method. Based on algorithmic benchmarking, the Elliptical Basis Function (EBF) is selected to construct a surrogate model. The resulting meta-model captures multidimensional sensitivities and supports iterative optimization to yield optimal wheel profiles for each QoI. Beyond simulation—including dynamic response and cumulative wear prediction—full-scale experiments are performed on a newly constructed tram-train testbed featuring a dedicated transition lane. The optimized designs demonstrate superior dynamic stability, reduced wear, and decreased derailment risk, with good agreement between numerical and experimental results. This study contributes to the advancement of hybrid rail transit by integrating surrogate-based optimization with comprehensive computational and experimental validation, laying a foundation for safer, more efficient tram-train operations.