Nihon Kikai Gakkai ronbunshu (Feb 2021)

Estimation of coefficient of friction between flange and rail utilizing vehicle dynamics simulation (Investigation based on roller-rig test with different lubrication condition)

  • Yohei MICHITSUJI,
  • Yosuke ICHIYANAGI,
  • Akira MATSUMOTO,
  • Yasuhiro SATO,
  • Hiroyuki OHNO,
  • Seigo OGATA,
  • Masuhisa TANIMOTO,
  • Tomoki FUKUSHIMA,
  • Takuya MATSUDA,
  • Takanori MATSUMI

DOI
https://doi.org/10.1299/transjsme.20-00357
Journal volume & issue
Vol. 87, no. 895
pp. 20-00357 – 20-00357

Abstract

Read online

The coefficient of friction (COF) between the outer rail and the leading-outside wheel flange of a railway bogie running on sharp curved track is an important parameter for the progress of wheel/rail wear and running safety. Therefore, it is desired to grasp the actual state of COF of wheels especially at the leading-outside wheel flange. In the author’s previous studies, a roller-rig experiment was conducted with changing lubrication condition of each wheel of the bogie to figure out relationship between lubrication conditions and wheel/rail contact forces such as tangential, lateral and vertical directions. In this paper, a method to estimate the value of COF of leading-outside wheel with look-up tables built by repetitive numerical simulations is explained. In the construction of the look-up tables, friction coefficients for each wheel in the numerical simulation model are changed while collecting calculated wheel/rail contact forces. A railway bogie running on roller-rig test is considered for both numerical simulations and experiment with different lubrication condition, therefore look-up tables are built with a half-vehicle model running on the roller-rig. By inputting the wheel/rail contact forces measured in roller-rig test into the look-up tables built by simulations, and the COF of leading-outside wheel flange is estimated. The result shows that the lubrication condition of leading-outside wheel flange can be estimated by the proposed method especially when the value of COF is large and is useful for the detection of high COF condition.

Keywords