Topological characterization and Gaussian projection reconstruction of ballast 3D contour

Geometric description and reconstruction of ballast is the basis of ballasted track bed refinement research. In this paper, Laser scanning is used to establish the ballast contour sample library. The characteristic parameters of ballast contours, such as flattening, rectangularity are analyzed. A reconstruction method is proposed through deconstructing the ballast into the optimal ellipsoid and surface undulation feature by Gaussian projection; the ballast reconstruction function libraries and digital elevation-like model (DEM) are established to realize the regeneration of ballast. The results show that the ballast samples have various parameters of favorable morphological characteristics with no special particles, and the samples are of high quality. The optimal ellipsoid of ballast regenerated by the function libraries conforms to the probability distribution of the sample, and the correlation coefficients of middle-minor axis ratio (Em/Es) and middle-major axis ratio (Em/El) are 0.9514 and 0.9628, respectively; the optimal ellipsoid generated by the method of equivalent ellipsoid is well fitted, the average overlap rate between the ellipsoid and the ballast reaches 90.62 %; the ballast reconstruction is excellent, all parameters are consistent with the sample distribution. This method can provide an effective way of particle modeling for a more realistic discrete element simulation of ballasted track bed.