Experimental Evaluation of Crack Evolution in Rails Using a Phased Array
Crack growth in rail heads can lead to rail breaks that cause delays in train operation and, in rare cases, train derailments. As a result, it is important to detect cracks and develop models that are able to accurately predict crack growth rates in order to avoid such events. Accordingly, it is necessary to develop complex rail testing protocols that are capable of validating such models, thereby increasing the safety, reliability, and efficiency of railway operations.
For the last five years, Texas A&M University along with MxV Rail has been running fatigue tests on rails with pre-existing cracks under simulated loading conditions of moving freight cars. The loading includes the following: 1) vertical bending to simulate wheel load cycles and 2) axial tension to simulate thermal stresses. Non-destructive evaluation (NDE) of internal rail cracks using phased array ultrasonic testing (PAUT) has been used in this series of tests. This research focuses on our efforts to utilize PAUT methodology to nondestructively measure internal crack geometry at various intervals during the long term multiaxial cyclic loading. These efforts are furthered by an exploration of residual stress maps of specimens post failure, via the contour method. It was found that the residual tensile and compressive zones within rails correlate with the stable growth direction of each crack.
Our efforts have in some cases resulted in anomalous measurements of crack geometry, thereby impacting the accuracy of the results of these complex fatigue tests. In this paper we present an overview of our experimental results to date as well as our efforts to improve this critical testing methodology for assessing crack growth rates in rails subject to cyclic loading.