TTCI R&D Predicting lateral buckling in rails TTCI investigates thermal buckling of rail to develop an algorithm to help predict and prevent potential occurances. by David H. Allen Ph.D., P.E., senior research engineer, director, Center for Railway Research, Texas A&M Transportation Institute and Gary T. Fry Ph.D., P.E., senior scientist II, Transportation Technology Center, Inc.
T
hermal buckling of rail is a complex phenomenon influenced by many factors, including temperature distribution within the rails, relative stiffness of the rail-crosstie interface, structural configuration of the underlying track structure and the highly nonlinear interactions among the rail-tie-ballast interfaces. Previous research shows that rail temperature, rail-ballast interface friction and rail-crosstie structural configuration must be included within any rail buckling model in order to accurately predict rail buckling.1,2,3 Furthermore, friction between the crossties and ballast has been shown to be highly nonlinear; see Figure 2. Accounting for all of these contributing factors result in a mathematical model that cannot be solved analytically. Accordingly, the authors are developing a self-contained computational algorithm for predicting lateral buckling in rails4 that is based on the large deformation Euler-Bernoulli beam theory5,6 cast within the finite element method 7; see Figure 3. Nonlinearity in the algorithm results from geometric nonlinearities in the rail in the deformed configuration, finite strains and nonlinearities in the friction between the crossties and ballast. This nonlinearity is accounted for using Newton iteration. Typical results obtained with the model are shown in Figure 4, wherein the critical temperature change necessary to induce buckling is plotted against buckle length for a variety of (constant) coefficients of friction, kz, between crossties and ballast for a typical track structure.8 A significant feature of the model is the ability to account for the degradation of material properties within the track structure that arise from cyclic loading and environmental conditions. Toward this end, the authors have developed a model for predicting the coefficient of friction between the crossties and ballast as a function of loading history (see Figure 5). Experiments are being designed, but have not been performed in the field, to determine how this property Figure 1, top: Photograph showing a local railway buckle. Figure 2, bottom: Typical lateral load versus displacement from single yie push tests (STPT).11
10 Railway Track & Structures
April 2017
www.rtands.com