Damage detection for civil structural health monitoring application - A case study of the steel grid bridge structural model

Abstract

The procedure for detecting the location and severity of damage of complex structural systems using their modal properties is an important tool of Structural Health Monitoring (SHM) of civil infrastructure. The herein presented research proposes procedures for damage detection based on two heuristic optimization methods: Simulated Annealing (SA) and Tabu Search (TS). In order to test the proposed procedures in different frequency ranges, experimental and numerical analyses were conducted on a steel grid bridge model in two configurations, according to the total mass of the structure, as well as for two simulated damage cases. The calibration of model parameters, according to experimentally extracted modal properties, is carried out using the proposed procedures. Numerical computations were conducted using ANSYS package and developed routines under MATLAB environment for model calibration and damage detection procedures. Experimental modal properties were extracted from ambient vibration measurements, as state-of- the art in SHM of complex structures, by the Frequency Domain Decomposition (FDD) technique, using ARTeMIS software. Both of the proposed procedures for model calibration and damage detection, with adopted objective functions including frequency and mode shape differences, exhibit accuracy, efficiency and robustness.