Abstract: In order to compare the differences between Hertzian and semi-Hertzian contact algorithms in wheel-rail wear simulation processes, a dynamic model of CR400BF electric motor train unit(EMU) was established. Long-distance simulations were conducted based on Hertzian and semi-Hertzian contact algorithms(STRIPES), and the wear was calculated using Archard wear model. The calculated results of the two algorithms were compared respectively with the measured tread wear of 250 000 km. Results show that the wear amount and distribution of different rail profiles are obviously different. The shapes of the semi-Hertzian contact patch are obviously different under different wear mileages, the contact patch areas increase and the maximum contact stresses decrease gradually. At early stage of wear, the abrasion and distribution obtained by the two contact algorithms have little difference and are close to the measured results. But at later period, the abrasions of two kinds of algorithm are larger than the measured results， and the semi-Hertzian contact is more in line with the measured results than that of Hertzian contact. The reason is that with the increasing abrasion mileage, wheel/rail contact no longer meets the Hertzian contact hypothesis.

Key words: wheel-rail contact, wheel wear, semi-Hertzian contact, Archard model

摘要： 为了对比赫兹与半赫兹接触算法在轮轨磨耗仿真过程中的差异，建立CR400BF动车组动力学模型，分别基于赫兹接触与半赫兹接触算法（STRIPES）进行长距离仿真，用Archard磨耗模型计算磨耗量，以实测250 000km踏面磨耗量为参照分别比较两种算法的计算结果。结果表明：不同钢轨廓形对应磨耗量及磨耗分布差异明显；不同磨耗里程下半赫兹接触的接触斑形状差异明显，接触斑面积逐渐增大且最大接触应力逐渐减小；磨耗前期，两种接触算法得到的磨耗量及磨耗分布差异不大，且与实测结果较为接近，但磨耗中后期，两种算法得到的磨耗量较实测结果都偏大且半赫兹接触较赫兹接触更加符合实测结果，原因在于随着磨耗里程增大，轮轨接触不再满足赫兹接触假设。

关键词: 轮轨接触, 车轮磨耗, 半赫兹接触, Archard模型