Methods for Improving the Damage Recognition Ability of Curvature Damage Index

The curvature damage index (CDI) has gained broad application in the fields of structural damage detection and localization due to its high damage sensitivity and accurate localization. However, this index exhibits significant limitations in weak-damage detection: When structural damage severity is low, damage-induced curvature perturbations are prone to be obscured by the global fluctuations of modal shape curvature (MSC), leading to diminished localization accuracy and weakened identification capability. To address these limitations, this study first proposes an improved curvature damage index (ICDI) by integrating abrupt geometric change features of modal shape difference curves into the CDI. Subsequently, a cantilever beam model was used to compare the weak-damage focusing capability, inter-class discriminability, and localization performance between the original and improved indices under varying damage severities. Finally, a finite element (FE) model of a wind turbine blade is established to validate the applicability of the ICDI in blade damage identification and localization. The effects of damage severity and location on the performance enhancement magnitude and early-stage damage identification capability were systematically analyzed. The results demonstrate that the ICDI exhibits superior weak-damage focusing capability, inter-class discriminability, localization accuracy, and early-stage damage identification capability. The performance enhancements of the ICDI index over the CDI index show significant spatial dependency: The ICDI index achieves significantly greater improvements in weak-damage focusing, inter-class discriminability, localization accuracy, and early-stage damage identification for weak-damage in the fore-mid blade segments, while exhibiting relatively smaller improvements for damage near the blade root and mid-span regions.