Baseline-Free Adaptive Crack Localization for Operating Stepped Rotors Based on Multiscale Data Fusion

Area of Science:

• Mechanical Engineering
• Structural Health Monitoring
• Vibration Analysis

Background:

• Crack localization in running rotors is crucial for machinery operation and maintenance.
• Existing characteristic deflection shape methods face challenges with steps and noise without baseline data.
• Accurate crack detection is vital for preventing catastrophic failures and ensuring system reliability.

Purpose of the Study:

• To develop a baseline-free adaptive crack localization method for running rotors.
• To address challenges of interference from rotor steps and improve performance in noisy environments.
• To enhance the efficiency and practical applicability of crack localization techniques.

Main Methods:

• Proposed a novel baseline-free adaptive crack localization method using data fusion of multiscale super-harmonic characteristic deflection shapes (SCDSs).
• Utilized crack-induced asymmetry and crack breathing nonlinearity to eliminate step interference without a reference model.
• Employed Gaussian multiscale space and the Teager energy operator to create multiscale Teager super-harmonic characteristic deflection shapes (TSCDSs) for noise robustness.
• Used fractal dimension for adaptive selection of TSCDSs and Dempster-Shafer's (D-S) evidence fusion to derive a new damage index for localization.

Main Results:

• Successfully localized single or multiple cracks accurately, overcoming interference from rotor steps.
• Demonstrated improved crack localization performance in noisy environments.
• Verified the feasibility and effectiveness of the proposed method through numerical simulations and experimental investigations.

Conclusions:

• The developed baseline-free adaptive method effectively localizes cracks in running rotors.
• The data fusion of multiscale TSCDSs offers enhanced robustness against noise and interference.
• This approach provides a promising solution for practical machinery diagnostics and maintenance.