Identifying Dynamic Characteristics of a Short-Span Viaduct from Vehicle-Induced Vibrations Considering Different Pavement and Parapet Conditions

Abstract:

In analyzing the vehicle-inducedvibrations of short- to medium-span bridge, this study adopts the conventionalmultivariate autoregressive (MAR) model along with a stabilization diagram (SD) that is introduced to resolve the difficulty indeciding the optimal model order and to reduce the variation of identifiedcharacteristics. Such a combined identification technique is applied to identify dynamic characteristics of a simple steel highway viaduct in Osaka, Japan, with three pavement andparapet conditions. Two issues are investigated through the field experiment: firstly, the accuracy and precision of the present technique is verified, especially in the frequencyidentification; and secondly, the vehicle-bridge interactions (VBI) issues areinvestigated. The bridge frequencies vary due to differentpavement and parapet conditions, but no obvious variation in the mode shapes is observed. Observations also demonstrate that the correlations betweenvehicle and bridge responses are not strong enough to guarantee the successfulidentification of bridge parameters from the raw vehicle responses. Somefurther data processing techniques should be applied to realize the indirect identification of bridge’s dynamic characteristics utilizingvehicle vibrations.n analyzing the vehicle-induced vibrations of short-to medium-span bridge, this study adopts the conventional multivariate autoregressive (MAR) model along with a stabilization diagram (SD) that is introduced to resolve the difficulty in deciding the optimal model order and to reduce the variation of identified characteristics. Such a combined identification technique is applied to identify dynamic characteristics of a simple steel highway viaduct in Osaka, Japan, with three pavement and parapet conditions. Two issues are investigated through the field experiment: firstly, the accuracy and precision of the present technique is verified, especially in the frequency identification; and secondly, the vehicle-bridge interactions (VBI) issues are investigated. The bridge frequencies vary due to different pavement and parapet conditions, but no obvious variation in the mode shapes is observed. Observations also demonstrate that the correlations between vehicle and bridge responses are not strong enough to guarantee the successful identification of bridge parameters from the raw vehicle responses. Some further data processing techniques should be applied to realize the indirect identification of bridges dynamic characteristics utilizing vehicle vibrations.