Structural flexibility identification via moving-vehicle-induced time-varying modal parameters

To overcome the limitation of existing methods only identifying basic dynamic properties of bridge structures, a moving vehicle strategy was proposed to identify more meaningful parameters including mass-normalized mode shapes and structural flexibility matrix, which can be further adopted for static deflection prediction and condition assessment of bridge structures. The contribution of this article is to investigate the intrinsic relationship between mass-normalized scaling factor and time-varying modal parameters identified from moving vehicle-induced responses. In the proposed method, the variational mode decomposition (VMD) algorithm was firstly utilized to decompose vehicle-induced responses into several intrinsic mode function (IMFs), then time-varying frequencies of the vehicle-bridge coupled system are identified by performing Hilbert transform of each IMF; Second, the relationship between mass-normalized scaling factor and time-varying dynamic properties of the coupled system was theoretically derived, and structural flexibility matrix was computed by combining calculated scaling factor and dynamic properties of the bridge without moving vehicle; Finally, a numerical example of simply supported bridge was employed to verify the correctness of proposed method, in which influences of predefined parameters (i.e. penalty parameter and number of decomposed IMFs) in VMD algorithm, road roughness, measurement noise, vehicle weight, vehicle spring and driving speed on identification results were investigated. To further verify the correctness of this method, an experimental model of steel beam was designed in the laboratory for structural flexibility identification and static deflection prediction. A good agreement between predicted deflections using the proposed method and directly measured values by using displacement transducers has successfully verified the proposed method. (C) 2020 Elsevier Ltd. All rights reserved.