Dynamics of high speed vehicles traversing structures with random surfaces

In recent years, interest in high speed ground transportation has encouraged the study of vehicle-guideway systems that can operate at 240 to 480 kph (150 to 300 mph). The system considered here consists of a multiple axle, passive suspension vehicle model, random surface roughness model and a linear elastic structure model. This research provides insights on system modeling, system analysis and system design.; The vehicle models are one-car and three-car systems with a constant velocity and a fixed wheelbase or distance between axles. The vehicle is coupled with the structure at the spatial positions of the axles. The interface between the vehicle and the structure determines the appearance of parametric terms in system matrices. If a mass-spring-dashpot interface is used, then all the system matrices are parametric. If a spring interface is used, then only the stiffness matrix is parametric.; Surface roughness is modeled as a stationary spatial random process. A Markov vector model that includes multiple interface points is formulated.; A series of two-span beams is used as a structure model. It is defined in the modal domain by natural frequencies, mode shapes and damping values. All vehicle, roughness and structure parameters are nondimensionalized and the equations of motion of the vehicle-guideway coupled system are written in state form. Then, by taking expectations, stochastic state equation is decoupled into two matrix equations. One is for the evolutionary mean vector and the other is for the evolutionary covariance matrix. The matrix equation for the evolutionary mean vector has parametric and deterministic additive excitation. The equation for the evolutionary covariance matrix has parametric and random additive excitation. Structural responses are normalized by maximum static responses, so the evolutionary means and variances of dynamic amplification factors are computed.; Convergence rate of different structural responses are investigated. Then effects of parameters on mean and variance responses are studied. The nondimensional span passage rate and the nondimensional axle arrival rate are two parameters that affect responses significantly. At very high velocities, the axle arrival rate can be equal to fundamental frequencies of the structure, causing large mean values of responses. Effects of the initial conditions of the covariance matrix of the vehicle degrees of freedom on the covariances of the structure degrees of freedom are also studied. Effects of different surface roughnesses, including those corresponding to current surface roughness specifications, on the responses of the high speed vehicle-guideway coupled system are presented. Probabilities of responses exceeding high dynamic amplification levels are estimated using Rice's assumption. Dynamic amplification factors corresponding to probabilities of exceedance ranging from 10−2 to 10−6 are presented. If an acceptable probability of exceedance is decided on, then the corresponding dynamic amplification factors may be used for design.