TY - GEN
T1 - Application of wavelet synchro-squeezed transform (WSST) method to railway bridge health monitoring
AU - Mostafa, Neda
AU - Loendersloot, Richard
AU - Di Maio, Dario
AU - Tinga, Tiedo
PY - 2020
Y1 - 2020
N2 - Typically, the identification of resonant frequencies in railway bridges is carried out from free-decay stationary signals as a train leaves the bridge. The same identification proves very challenging when nonstationary vibrations are measured as a train traverses the bridge. Despite the numerous attempts, nonstationary signals with low modulating frequencies are still difficult to be processed. This paper attempts to evaluate the bridge-vehicle first bending resonance by a method known as Wavelet Synchro-Squeezed Transform (WSST). The significant advantage of this signal processing method is to deal with low-frequency modulations, which are typical of long bridges. This research focusses on a Finite Element Model (FEM) of a bridge simulating the nonstationary vibration responses exerted by a spring-mass model traversing the bridge. The paper sets two objectives, and the first one is to investigate how the WSST analyses nonstationary signals generated by the FE model. The instantaneous frequency trace of the bridge-vehicle system will be compared to a similar frequency trace, that is created by performing several modal analyses at different locations of the bridge. The second objective of the paper is to investigate if the instantaneous frequency obtained from WSST is suitable for damage detection, as the FE model is fitted with damages. Both objectives are met, and the results will be presented. The trace of the first natural frequency matches well the one calculated by the WSST, and the instantaneous frequency shows to be capable of detecting damages included in the model.
AB - Typically, the identification of resonant frequencies in railway bridges is carried out from free-decay stationary signals as a train leaves the bridge. The same identification proves very challenging when nonstationary vibrations are measured as a train traverses the bridge. Despite the numerous attempts, nonstationary signals with low modulating frequencies are still difficult to be processed. This paper attempts to evaluate the bridge-vehicle first bending resonance by a method known as Wavelet Synchro-Squeezed Transform (WSST). The significant advantage of this signal processing method is to deal with low-frequency modulations, which are typical of long bridges. This research focusses on a Finite Element Model (FEM) of a bridge simulating the nonstationary vibration responses exerted by a spring-mass model traversing the bridge. The paper sets two objectives, and the first one is to investigate how the WSST analyses nonstationary signals generated by the FE model. The instantaneous frequency trace of the bridge-vehicle system will be compared to a similar frequency trace, that is created by performing several modal analyses at different locations of the bridge. The second objective of the paper is to investigate if the instantaneous frequency obtained from WSST is suitable for damage detection, as the FE model is fitted with damages. Both objectives are met, and the results will be presented. The trace of the first natural frequency matches well the one calculated by the WSST, and the instantaneous frequency shows to be capable of detecting damages included in the model.
KW - Health monitoring
KW - Vehicle-bridge interaction
KW - WSST
UR - http://www.scopus.com/inward/record.url?scp=85099728717&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:85099728717
SN - 978-618-85072-2-7 (Set)
SN - 978-618-85072-0-3 (I)
T3 - Proceedings of the International Conference on Structural Dynamic , EURODYN
SP - 1388
EP - 1396
BT - EURODYN 2020 - XI International Conference on Structural Dynamics
A2 - Papadrakakis, Manolis
A2 - Fragiadakis, Michalis
A2 - Papadimitriou, Costas
PB - National Technical University of Athens
CY - Athens
T2 - 11th International Conference on Structural Dynamics, EURODYN 2020
Y2 - 23 November 2020 through 26 November 2020
ER -