Train carbody and truck structures are designed to exhibit primary natural frequency modes great enough to avoid unwanted resonant oscillations with normal track interactions. Critical bounce modes can be excited by typical track in the 2–4 Hz range. Trains are designed with first modes above this threshold. Historically, simplified approaches are employed to predict natural frequencies of the main truck and carbody train structures independently. Since the advent of high powered computing, more detailed finite element analysis (FEA) eigenvalue approaches have been used to more accurately predict natural frequency of structures. Still, the typical FEA approach uses simplified boundary conditions and partial models to determine natural frequencies of individual components, neglecting the interaction with other connected structures. In this paper, a detailed, holistic approach is presented for an entire Light Rail Vehicle (LRV). The analysis is performed on a fully detailed FEA model of the LRV, including trucks and suspension, carbody structures, non-structural mass, articulation, as well as intercar and truck-carbody connections. The model was developed for detailed crashworthiness investigations, which requires a high level of fidelity compared to what is typically required for static and modal analysis. Using the same model for multiple purposes speeds up development while also improving the accuracy of the analyses. In this paper, the modal analysis methodology developed is described. A case study is presented investigating the often neglected contribution of windows, cladding, and flooring on the overall carbody natural frequency.

This study develops a detailed multi-body dynamic model of the Virginia Tech Roller Rig (VTRR) using multi body simulation software package SIMPACK. The Virginia Tech Roller Rig, a single-wheel roller rig with vertical plane roller configuration, is a state of the art testing fixture for experimental investigation of wheel-rail contact mechanics and dynamics. In order to have a better understanding of the dynamics at the contact, dynamic behavior and interaction of various components and subsystems of the rig need to be understood. In addition, it is essential to make sure that the measurements are only due to particular subject of study and not any intermittent source of disturbance. Any unwanted vibration at the contact needs to be compensated in the data measurements. To this end, a fully detailed model of the rig including all the components is developed in SIMPACK. The coupled multibody dynamic model represents all the major components of the rig and their interactions. The multibody dynamic model is employed for conducting noise, vibration, harshness (NVH) analysis of the rig. An Eigenvalue analysis provides the modal frequencies and mode shapes of the system. The modal analysis predicts the first natural frequency of the rig to be approximately 70 Hz, providing a relatively high bandwidth for evaluating the dynamics at the wheel-rail interface. Only dynamic that could have higher frequencies than the rig’s bandwidth is wheel-rail squeal. The model is also used to evaluate the performance of the contact force measurement system designed for the rig. The results show that the contact forces can be estimated precisely using the force measurement system.

This study uses a nonlinear multibody dynamics model of a railway vehicle with three-piece trucks to perform a parametric study on a wide range of parameters influencing the performance of the truck on tangent track. One of the major disadvantages of the three-piece truck is that its performance can be greatly influenced due to worn parts or contaminants on friction surfaces, such as the friction wedge. The influence of worn parts is modeled as increased gap between bodies, using dead-band springs. The influence of environmental contamination is modeled as change in coefficients of friction. Lateral accelerations, measured at the center of gravity of the leading axle of the leading truck (Axle 1) are used for the hunting performance of the vehicle. For the baseline case, an eigenvalue analysis of a linearized model is used to evaluate the effect of speed on damping ratio and lateral motion frequency. A comparison of the linear model and non-linear time series analysis show that the linear model is less conservative, in that it predicts that the onset of hunting occurs at higher speeds. The results indicate a direct relationship between various coefficients of friction and hunting velocity of the rail vehicle. Increasing the gap between the pedestal legs and bearing housing does not influence hunting velocity, but reduces lateral acceleration magnitudes during hunting. The weights of the carbody and cargo also has a direct relationships with hunting velocity, meaning that hunting occurs at lower speeds for unloaded car and at higher speeds for a loaded car.