Robust Baseline-Free Damage Localization by Using Locally Perturbed Dynamic Equilibrium and Data Fusion Technique.

Area of Science:

• Structural Engineering
• Mechanical Engineering
• Vibration Analysis

Background:

• Mode shape-based methods are effective for damage localization but sensitive to measurement noise.
• Conventional methods struggle with multi-damage scenarios and require specific modes.

Purpose of the Study:

• To develop a robust and accurate structural damage identification method.
• To address limitations of existing mode shape-based techniques regarding noise and multi-damage localization.

Main Methods:

• A joint singular value decomposition (JSVD) technique for robust mode shape estimation from noisy data.
• An improved pseudo-excitation method for baseline-free damage localization index calculation.
• A data fusion scheme to integrate information from different modes for enhanced localization.

Main Results:

• The proposed JSVD method shows superior performance in handling measurement noise compared to traditional transmissibility methods.
• The novel damage localization index effectively quantifies local dynamic equilibrium changes without prior structural knowledge.
• The data fusion scheme outperforms Bayesian fusion, providing more accurate and robust multi-damage localization.

Conclusions:

• The developed method offers significant improvements in damage localization accuracy and robustness.
• The approach is validated through numerical and experimental studies on cracked cantilever beams.
• This technique enhances structural health monitoring by overcoming noise and multi-damage challenges.