Parameters of a discrete element ballasted bed model based on a response surface method

• Published in: Journal of Zhejiang University. A. Science 2019-09, Vol.20 (9), p.685-700
• Main Authors: Xiao, Jie-ling, Liu, Gan-zhong, Liu, Jian-xing, Dai, Jia-cheng, Liu, Hao, Wang, Ping
• Format: Article
• Language: English
• Subjects: Angle of repose; Civil Engineering; Classical and Continuum Physics; Coefficient of friction; Computer simulation; Discrete element method; Engineering; Friction; Friction resistance; Industrial Chemistry/Chemical Engineering; Laboratories; Laboratory tests; Mechanical Engineering; Optimization; Parameters; Railroad ballast; Railway tracks; Regression analysis; Response surface methodology; Rolling resistance; Static friction
• ISSN: 1673-565X; 1862-1775
• DOI: 10.1631/jzus.A1900133

Discrete element simulation on ballasted beds is an important method to study the service characteristics of ballasted tracks; an effective simulation should be based on proper ballast parameters. Ballast contact parameter, which exhibits a high discreteness affected by factors such as material, shape, and gradation, can effectively be calibrated by an angle of repose test. Based on the testing principles of a multi-parameter response surface method, the Box-Behnken method is adopted to design the angle of repose test under the influence of restitution, static friction, and rolling friction coefficients; laboratory-measured results are combined with the simulation; regression analyzed angle of repose is considered as the goal; parameters optimization and ballasted bed resistance simulations are verified for multiple parameters. The results demonstrate that Chinese special-grade ballasts exhibit an average laboratory-measured angle of repose of (39.78±1.27)°, and the optimal combination of parameters in this discrete element simulation based on the response surface method are as follows: the restitution coefficient is 0.72, the static friction coefficient is 0.56, and the rolling friction coefficient is 0.27. The results of the lateral resistance simulation are in accordance with the laboratory test, indicating that the optimal parameters are usable. The multi-parameter response surface method effectively helps calibrate the parameters of the discrete element simulation on ballasted beds.