Research Of Railway Ballast Geometric Characteristics Based On3D Model Reconstruction

Due to China’s high-speed railway development strategy to the central and western regions, ballast track will be more widely used in high-speed railway construction to reduce the cost. The research of performance of railway ballast under high speed cyclic load is important. For granular materials such as coarse aggregate and ballast gravel, the relationship between macroscopic mechanical properties and the geometrical characteristics of the particle shape is still insufficient in research. To find out the corresponding relationship between them, the mesoscopic geometry characteristics such as shape, angularity and surface texture, must be measured quantitatively and accurately. So the measurement is a problem that must be solved to analyze geometry characteristics.In this paper, an image acquisition system was established to measure the granular material by using two planar mirrors and a CCD camera. Reconstruct the3D models of the ballast by computer graphics algorithm. Based on this method, the image acquisition hardware device and software system were invented. The device is used to analyze the granular materials used in civil engineering, especially the ballast gravel used in railway construction. Based on the theory of3D model reconstruction, the optimization analysis of outer epipolar tangency error, outline resolution, grid precision was taken. Some quantitative characteristics were recommended and measured to evaluate particle shape and angularity. A comparison between image method and manual caliper (mechanical screening test) measurement was taken, the result shows high consistency in the measured value. Angularity index, roundness, sphericity, convexity and other geometric parameters was obtained by analyze the3D models. Based on these result, classification and comparison was taken. The probability distribution of the geometric parameters can distinguish between different types of granular materials (granite, limestone, and dolomite).3different types of gravel were divided into3form types:rounded, sub rounded sub angular, angular. Through the analysis the data distribution under different parameters combination, two kinds of parameters combination (F&E ratio-sphericity, sphericity-convexity) were recommended to aid in quantitative classification of granular materials...