Insulated Railroad Joint Design Evaluation by Coordinated Test and Finite Element Analysis

The railroad industry faces challenges with bonded insulated joint designs in the present practice. A program initiated by Virginia Tech and the Transportation Technology Center Incorporated (TTCI) has been in progress to analyze and test a class of insulated joint designs featuring non-adhesive bolted connections. A hierarchical approach to finite element modeling with a parametric model maintaining essential mechanics of the joints has been applied to develop a bolted insulated joint design. The current paper reports on the recent phase of the program including development of experimental tests along with finite element analyses on scaled simplified insulated rail joint models. Two baseline rail joint configurations with simplified sections were considered for studying dominant mechanics under the AREMA (American Railway Engineering and Maintenance-of-Way Association) rail joint acceptance standard test loading and boundary conditions. The finite element models developed based on three-dimensional continuum elements incorporated bolt preloads and full-contact analysis. In the experimental tests, the strain analyses on 1/4 scaled polycarbonate rail joint specimens were performed by means of an array of strain gauge transducers mounted on the joint bars and a photoelasticity technique. The results of the experimental stress analyses were employed to validate the finite element models quantitatively and qualitatively in terms of load transfer mechanics and stress distribution. The validated models serve as baseline insulated joint configurations for developing fracture-mechanics-based fatigue-failure analysis. To investigate the role of cracks on the performance and reliability of joint bars, a damage tolerant analysis is performed on the rail joints utilizing linear elastic fracture mechanics. The locations of most critical type defects are estimated based on high stress/strain regions from stress analyses along with past experiences on failure of rail joints. To characterize the severity of theses defects under alternating loading conditions, stress intensity factors are computed as a function of crack length. Cracks of different lengths are introduced in the vicinity of the most fatigue-prone locations of the joint bar in a parametric modeling fashion. The fatigue-crack-growth-rate properties in terms of Paris Law scaling constants are selected from a survey of available material data. The number of loading cycles to failure is obtained by employing the computed stress-intensity factors as well as initial and final crack sizes. Predicted lifetimes as a function of pre-existing crack sizes and geometry of joint configuration can be used as a fracture-mechanics-based function for more accurate design of the rail joints.