Time reversal imaging for sensor networks with optimal compensation in time

Area of Science:

• Structural Health Monitoring
• Non-Destructive Testing
• Computational Mechanics

Background:

• Assessing structural integrity is crucial for safety and maintenance.
• Traditional methods like travel time migration struggle with complex environments and signal scattering.
• Accurate damage localization requires advanced imaging algorithms for ultrasonic sensor data.

Purpose of the Study:

• To analyze distributed sensor imaging algorithms for detecting localized structural damage.
• To develop and evaluate a novel imaging method overcoming limitations of conventional techniques.
• To image point-like defects in complex structures using ultrasonic transducer arrays.

Main Methods:

• Extensive numerical simulations using the two-dimensional wave equation.
• Numerical back-propagation of recorded sensor traces with background knowledge.
• Minimization of Shannon entropy or bounded variation norm to determine optimal refocusing time.
• Singular value decomposition of the response matrix for multi-defect scenarios.

Main Results:

• The developed imaging method successfully produces tight images of defects at the correct location and time.
• Numerical simulations validate the effectiveness of back-propagation and entropy/variation minimization.
• The approach is robust even in complex structural environments with significant delay spread.
• Singular value decomposition aids in analyzing damage when multiple defects are present.

Conclusions:

• The proposed numerical back-propagation imaging method is effective for localized damage detection in structures.
• Minimizing image entropy or bounded variation provides a reliable criterion for defect imaging.
• This technique offers a significant improvement over traditional methods for complex structural analysis.
• The study demonstrates the potential for advanced algorithms in structural health monitoring.