An algorithm for damage localization in steel truss structures: Numerical simulation and experimental validation

It is imperative to study the damage detection methods of steel truss structures that are always employed in extreme environment. Accurate structural damage localization is still a challenge due to high noise and low accuracy of the structural finite element model. To develop a dependable damage localization technique for truss structural health monitoring, a novel idea of damage localization is proposed: the curvature difference method of strain waveform fractal dimension, based on fractal theory and curvature method. To validate the approach, a simply supported bailey steel truss benchmark model has been designed and constructed in the laboratory. Based on the model, both experimental and numerical simulation results using the procedure under pulse excitation indicate that it is feasible and effective to detect the change of boundary conditions and the stiffness reduction of a truss member. In addition, the proposed technique exhibits high-noise insusceptibility (e.g. it works for noise levels up to 20% for a 10% truss member stiffness reduction). Moreover, the proposed technology is robust against the accuracy of the finite element model of measured structures, which decrease the workload of model updating dramatically. All these lay a good foundation for its engineering application.