Abstract:
Over time, bridges degrade due to various processes like creep, corrosion, and cyclic loading, raising concerns about their structural health. Detecting potential damage has become crucial to prevent sudden failures. Vibration-based damage identification, a part of Structural Health Monitoring (SHM), relies on changes in a structure's dynamic properties as it deteriorates. By analyzing the vibrations caused by passing vehicles, the Vehicle-Bridge Interaction (VBI) method can identify bridge damage without needing knowledge of the force applied. This thesis introduces a VBI-based approach for assessing bridge condition through damage identification.
Initially, the bridge's structure is simulated through the Finite Element Method (FEM), while a half-car dynamic model represents the vehicle along with its suspension system. Equations of motion (EOM) for both the bridge and vehicle are developed using FEM, the mode superposition method, and D'Alembert's principle. The interconnected dynamics are resolved using the Newmark-beta method, considering the road surface roughness. Artificial damage is introduced by reducing the stiffness of a particular element in the bridge.
Secondly, the vehicle induced acceleration response, obtained from sensors in both the damaged and undamaged conditions which follows Gaussian distribution.An early damage detection index, based on the Mahalanobis Distance (MD) of matrices incorporating different statistical parameters i.e., mean, standard deviation, skewness and kurtosis of the acceleration responses, isproposedthrough both numerical simulations and experimental. Moreover,the calculation of the Statistical Moment (SM) is performed using the Power Spectral Density (PSD) at bridge sensor locations.A damage localization index, named as the Normalized Exponential of Percent Change in Statistical Moment (〖EPCSM〗_n),is then developed to identify the probable location of various damage scenarios for numerical as well as experimental study.The study's findings showcase the method's effectiveness in identifying damages across various positions along the bridge span for different damage levels.
