For rail infrastructure, its service life and operational safety are significantly affected by the rail and wheel wear on sharp curves. To figure out the inherent mechanisms behind the phenomenon and develop corresponding strategies, this study establishes a vehicle-track dynamic simulation framework for heavy-haul freight railways coupled with a 25-ton axle load using UM multibody dynamics software. Based on the Archard wear model for simulation, this paper investigates the wear on both inner and outer rails across different cant conditions to determine the correlations among the side friction coefficient, dynamic wheel-rail interactions, and wear evolution. The results illustrate that necessary maintenance interventions should be prioritized to slow down the wear process when rail wear intensifies markedly at a cumulative traffic tonnage of approximately 90 million tonnes. Moreover, an under-superelevation ranging from 10% to 20% alleviates side wear on curved tracks. Rail-specific wear patterns are observed, where side wear dominates on the outer rail and vertical wear prevails on the inner rail, with peak wear zones located in the horizontal coordinate range of [20, 36] mm and [10, 16] mm, respectively. The side friction coefficient significantly affects side wear on the outer rail in sharp curves. When the value of this coefficient rises from 0.15 to 0.35, the side wear depth increases by 179% accordingly. In contrast, the vertical wear change of the inner rail is negligible. Therefore, it is recommended to keep the side friction coefficient rather low by using friction management techniques, thereby reducing outer rail wear and extending rail service life.