Efficient non-stationary random vibration analysis of vehicle-bridge system based on an improved explicit time-domain method. (15th March 2021)

• Record Type: Journal Article
• Authors: Yu, Helu; Wang, Bin; Xia, Cuipeng; Gao, Zongyu; Li, Yongle
• Is Part Of: Engineering structures. Volume 231(2021)
• Publication Date: 2021-03-15
• Subjects: Vehicle–bridge interaction -- Random vibration analysis -- Track irregularity -- Improved explicit time-domain method -- Response statistics -- Probability density; Structural engineering -- Periodicals; Structural analysis (Engineering) -- Periodicals; Construction, Technique de la -- Périodiques; Génie parasismique -- Périodiques; Pression du vent -- Périodiques; Earthquake engineering; Structural engineering; Wind-pressure; Periodicals; 624.105
• DOI: 10.1016/j.engstruct.2020.111786
• Languages: English
• ISSNs: 0141-0296

Abstract

Highlights: Improved explicit time-domain method for random vibration analysis of vehicle-bridge system. Spectral representation-based formulation for calculating excitation covariance matrix. Explicit solutions for statistical moment and probability density of stochastic responses. High accuracy and efficiency of proposed method validated by numerical examples. Abstract: Based on the recently developed explicit time-domain method (ETDM), this paper presents a new efficient computational framework for non-stationary random vibration analysis of the vehicle-bridge system (VBS) subjected to random track irregularity excitation. In this framework, the time-varying VBS equation is discretized using the time-stepping integration method, leading to an explicit expression for the stochastic system response in terms of the track irregularity excitation. Based on this expression, the response moments of the VBS can be easily evaluated in a direct way; the response probability density can also be obtained by using the Gaussian characteristic of the irregularity excitation. To further improve the computational efficiency, an efficient spectral representation-based formulation is proposed to calculate the excitation covariance matrix required in the analysis, which avoids tedious Fourier integrals required in the traditional ETDM. In the numerical application, the accuracy and efficiency of the proposed method are validated by the solutions obtained using the Monte Carlo and pseudo-excitation methods; the effects of several parameters on the stochastic responses of the VBS are investigated.