Rail profile optimization of a subway's sharp curve considering multiple curve sections and metal removal.

Autor: Wang, Shuguo, Xu, Jingmang, Liu, Caiwei, Qian, Yao, Wang, Ping, Wang, Pu, Ma, Qiantao, Liu, Boyi
Zdroj: Proceedings of the Institution of Mechanical Engineers -- Part F -- Journal of Rail & Rapid Transit (Sage Publications, Ltd.); Oct2024, Vol. 238 Issue 9, p1045-1056, 12p
Abstrakt: In order to effectively address the engineering problem of optimizing the worn rail profile on subway sharp-radius curves, this paper focuses on the inconsistent wear characteristics of rails on such curves. The contact stress, wheel-rail lateral force, and wear index are taken as objective functions, and key rail profiles at the points of straight-transition, transition-curve, curve-midpoint, circular-transition, and transition-straight are selected. A multi-objective and multi-section optimization scheme based on the optimized limit curve is proposed to design the grinding profile for rails on subway sharp-radius curves. In order to compare its optimization effectiveness, two additional optimization schemes are presented: the second scheme involves multi-section profile optimization with no optimized limit curve as constraints, and the third scheme deals with optimizing a single profile for grinding with equal cross-sections. Through a comparison of the three optimization schemes, the results indicate that all three schemes lead to a reduction in the contact stress, wheel-rail lateral force, and wear index of the outer and inner stock rails after optimization. Scheme 1 shows a more extensive and uniform distribution of wheel-rail contact points, with larger reductions in contact stress and increases in contact area. Scheme 3 only improves contact in severely worn sections, while in less worn sections, the contact distribution becomes concentrated, accelerating wear evolution. Furthermore, in Scheme 1, the cumulative wear after 100,000 and 200,000 cycles is reduced by a maximum of 8.44% and 22.33% compared to before optimization, making it the most effective among the three schemes. Therefore, the optimized profile output by the optimization method proposed in this paper is considered the best solution, which is represented by Scheme 1. [ABSTRACT FROM AUTHOR]
Databáze: Complementary Index