Railway Air Brake Model and Parallel Computing Scheme

Qing Wu, Colin Cole, Maksym Spiryagin, Yucang Wang, Weihua Ma, Chongfeng Wei

Research output: Contribution to journalArticlepeer-review

29 Citations (Scopus)

Abstract

This paper developed a detailed fluid dynamics model and a parallel computing scheme for air brake systems on long freight trains. The model consists of subsystem models for pipes, locomotive brake valves, and wagon brake valves. A new efficient hose connection boundary condition that considers pressure loss across the connection was developed. Simulations with 150 sets of wagon brake systems were conducted and validated against experimental data; the simulated results and measured results reached an agreement with the maximum difference of 15%; all important air brake system features were well simulated. Computing time was compared for simulations with and without parallel computing. The computing time for the conventional sequential computing scheme was about 6.7 times slower than real-time. Parallel computing using four computing cores decreased the computing time by 70%. Real-time simulations were achieved by parallel computing using eight computer cores.
Original languageEnglish
Article number051017
Number of pages11
JournalJournal of Computational and Nonlinear Dynamics
Volume12
Issue number5
Early online date15 May 2017
DOIs
Publication statusPublished - 1 Sept 2017
Externally publishedYes

Keywords

  • air brake
  • gas dynamics
  • method of characteristics
  • railway train dynamics
  • parallel computing

Fingerprint

Dive into the research topics of 'Railway Air Brake Model and Parallel Computing Scheme'. Together they form a unique fingerprint.

Cite this