Quantification of the Effect of Train Type on Concrete Sleeper Ballast Pressure Using a Support Condition Back-Calculator

Abstract

Monitoring ballast support condition and improving current sub-structure and ballast maintenance strategies is critical to ensuring safe and efficient railroad operations. Researchers at the University of Illinois at Urbana-Champaign (Illinois) have developed a ballast support condition back-calculator, a non-destructive instrumentation method and corresponding analysis tool that quantifies ballast pressure distributions under concrete sleepers without interrupting revenue service train operations. This laboratory-validated non-intrusive method uses concrete sleeper bending moment profile and rail seat loads as inputs to back-calculate the reaction distribution using a Simulated Annealing optimization algorithm that incorporates Pareto Distribution as the random variable generator. In order to further understand in-service ballast support conditions, concrete surface strain gauges were installed on concrete sleepers at a revenue service field site to measure strains that could subsequently be converted into bending moments. This site is on a shared use rail corridor with traffic ranging from high speed passenger to heavy axle load (HAL) freight trains. Rail-mounted strain gauges were used to measure strains that were used to calculate the vertical wheel-rail loads to approximate rail seat loads. This paper quantifies the ballast pressure distributions beneath concrete sleepers under different types of rolling stock and evaluates how ballast support condition changes as a function of accumulated tonnage. A wide range of loads were observed at the field site, ranging from 4 to 35 kips (18–156 kN). Corresponding ballast pressures ranged from 14 to 175 psi (97–1,207 kPa), with sleeper-ballast contact area corresponding to 60% of the bottom of the sleeper area. The accumulation of 12.24 million gross tons (MGT) (12.44 million tons) did not generate a quantifiable change in ballast pressure values nor did it generate a change in the ballast support condition. The research results presented in this paper demonstrate the potential of the back-calculator to provide a stand-alone non-invasive method to quantify ballast support conditions, sleeper health, and sleeper bearing stress. Back calculator data will aid the rail industry in optimizing tamping cycles, enhancing safety, and developing more representative concrete sleeper flexural designs based on actual support conditions.

Research on the Realization Path of Railway Intelligent Construction Based on System Engineering

“…The effective integration of new technology and traditional railway construction technology can not only give new capabilities to various production factors in the railway engineering construction system but also bring a new revolution. At present, these emerging technologies are widely used in the construction of smart construction [13,[19][20][21][22][23]. Ning, X. et al [13] established a construction simulation model based on a virtual geographic environment to solve the spatial conflict problem arising in the planning and design phase of high-speed railroad construction, and the obtained parameters can provide a scientific method for preferring the construction plan and assessing the full-cycle performance, and improve the project design efficiency to a certain extent.…”

“…Shin, M. [21] studied the prospects of BIM technology in the Korean railroad industry, verifying the advantages of this technology over traditional construction through field research. Silva, C. P. et al [22] introduced a simulated annealing optimization algorithm to calculate the monitoring of the forces on railroad concrete rail sleepers, based on which the safety performance of the sleepers was evaluated, and an optimized vibrating scheme was given accordingly. Yuan et al [23] studied the design of a railroad line selection scheme based on machine learning, which introduced the Dueling Double-Deep Q Learning Network (D3QN) to train the model, and the method can realize the automatic finding of the line scheme with the optimal objective function, and it has been well verified in the engineering examples.…”

Prediction of Thermal-Induced Buckling Failures of Ballasted Railway Tracks Using Artificial Neural Network (ANN)

“…recycled concrete [23], rubberised concrete [24] etc. More examples on railway track applications includes track response quantification [25], train weight prediction [26], fault detection [27][28][29][30], railway safety and accident identification [31,32]. Hence, this paper is the first to use machine learning in predicting buckling phenomena of ballasted railway tracks.…”