Structural damage analysis using finite element model updating enhanced by response surface methodology

Over time, structural damage can manifest due to external loads, underscoring the importance of early identification to mitigate risks and minimize financial and safety concerns. Maintaining an up-to-date finite element model supports informed decision-making throughout a structure's lifecycle, enabling effective risk assessment. Additionally, finite element model updating ensures structural analyses accurately reflect real-world conditions. However, the process of model updating, which involves adjusting FEM parameters to accurately reflect the behavior of the structure, can be time-consuming and challenging, particularly for complex structures. To address this issue, this research investigated the effect of different parameters on Finite Element model updating using statistical methods. In all the controlling factors, two specific ones, the Convergence Index (COI) and the Closeness Index (CI), were employed in 81 numerical experiments. Following the presented model updating technique, these experiments were structured using the central composite design method within a 2D truss framework. ANOVA was then employed to analyze the effect of different variables, and predictive models were developed and verified. To achieve appropriate convergence time and accuracy based on project sensitivity, the models were used to optimize the FEMU performance. Drawing from the findings regarding the CI parameter, implementing a multi-objective optimization strategy has notably elevated its value. Originally hovering between 75 and 85%, it has surged to 97.5%. Through optimizing the FEMU process with the suggested methodologies, pinpointing structural damage becomes more precise, rapid, and cost-effective, consequently diminishing maintenance expenses.