Milad Jahangiri

Numerous uncertainty sources can influence the intrinsic characteristics of the structures. Among these, the seismic noises and ambient vibrations are seamlessly subsisted in reality. Moreover, the finite element modeling errors are unavoidably occurred in numerical simulations. The aforementioned uncertainty sources can lead to an erroneous analysis of damage detection in structures. Hereupon, to improve the accuracy of structural damage detection, the current research attempts to foster a novel probabilistic approach called “probability of damage existence (PDE).” For this reason, two robust techniques including the Monte Carlo simulation (MCS) in interaction with the HOF-IAS model updating technique are implemented. The MCS technique is employed to build a baseline of the random variables. The model updating technique minimizes the discrepancy of the inherent vibrational characteristics for the damage detection of structure. PDE makes use of the advantages of the model updating technique and compensates its limitations. To assess the efficiency of the proposed approach, four different un-damped benchmark truss and frame structures are evaluated. The obtained results confirm the superiority and reliability of the probabilistic approach over the deterministic ones. Furthermore, in spite of the presence of uncertainty sources, the proposed probabilistic strategy by a high level of PDE not only predicts the damage location accurately but also estimates the damage intensity precisely.