Comparative Analysis of Fractals-Homogeneity-Entropy Algorithms Applied on a FEM Bridge Model to Identify Damage

Abstract

Vehicular bridges accumulate damage over time due to overloads and material degradation. Non-visible structural damage in such large structures poses a serious risk, as small defects in critical elements can rapidly grow, potentially leading to catastrophic failure. Therefore, implementing simple yet effective methods for damage identification within a structural health monitoring (SHM) system is crucial for ensuring bridge reliability. This study presents a systematic comparative analysis of multiple damage detection algorithms, including six different fractal dimensions (FDs), the homogeneity index (HI), and the Shannon entropy index (SEI). These methods are applied to a high-fidelity finite element method (FEM) model of the Rio Papaloapan Bridge (RPB), a cable-stayed structure, to detect and localize two different types of damage (deck and cable failures) with varying severities and positions. To enhance practical applicability, realistic conditions are simulated by introducing noise to the vibration signals collected from both the undamaged and damaged bridge scenarios while a moving load, simulating a vehicle, is crossing. The results indicate that the HI and SEI not only detected and localized all damage scenarios but also effectively distinguished between different levels of severity, making them highly promising for SHM applications. Additionally, two of the six FD algorithms successfully identified all damage cases with minimal variation from the healthy condition, demonstrating their potential utility. The findings presented in this study are consistent with previous experimental and real-world bridge assessments, reinforcing their validity for real-life applications.

Keywords

• bridges
• damage identification
• entropy
• FEM model
• fractals
• homogeneity