Investigation and evaluation of railway ballast properties variation during technological processes

Railway ballast (RB) layer has to limit tie movement by resisting vertical, lateral and longitudinal forces from the train and the track, to reduce the stresses from train loads applied to the subgrade, thus limiting permanent settlement. It also has to provide immediate water drainage from the track structure, to aid in alleviating frost problems, to facilitate maintenance surfacing and lining operations, to provide support for ties with the necessary resilience to absorb shock from dynamic loads. Secondary functions include retarding vegetation and resisting effects of fouling from surface deposited materials, absorbing airborne noise, providing adequate electrical resistance between rails, facilitating the redesign or reconstruction of the track. These properties are obtained through the use of proper gradation crushed granite, by determining its proper laying and compacting technological parameters, by ensuring sufficient thickness and profile of the layer. This paper presents statistical investigations of the four samples of crushed granite taken from transporter belt (TB), plant stockpile (PS), wagon (W) and railway construction (RC) uncompacted layer used to RB course. The gradation (particle size distribution), density of crushed granite particles (DENP), water absorption (WA24), Deval index (MDE), Los Angeles coefficient (LARB coefficient) and resistance to crushing (SZRB) of four samples were measured. Statistical parameters of crushed granite qualitative indicators, histograms, theoretical curves of normal distribution are presented. Their compliance with normal distribution was verified by employing the criteria of skewness, kurtosis, Pearson, Shapiro-Wilk and Kolmogorov-Smirnov. Dependance of standard deviations of percent passing through sieves on the means of percent passing through these sieves was obtained from regression analysis. The use of Kruskal-Wallis test by ranks showed that means obtained in different sample-taking locations did not differ statistically. The maximum value of standard deviation of this dependance, equal to mean of 50% percent passing, was used to evaluate the homogeneity of crushed granite used to construct the ballast layer according to the variation of its gradation. Absolute allowable error of minimum sample size n, equal to 5%, 10%, 15% and 20%, was calculated. Investigation results indicated that due to segregation the homogeneity of the gradation of crushed granite used to construct the ballast layer from its production site to its exploitation site decreased by 78%. During the technological processes of loading, transporting and spreading the gradation has hardly changed.