Automatic Fracture Characterization Using Tactile and Proximity Optical Sensing.

Area of Science:

• Materials Science
• Mechanical Engineering
• Sensor Technology

Background:

• Traditional methods for mechanical fracture detection often rely on visual inspection, which is limited in challenging environments.
• Electronic sensing devices commonly used in fracture detection can be susceptible to damage from radiation and extreme conditions.

Purpose of the Study:

• To demonstrate the efficacy of integrated tactile and proximity sensing for automatic mechanical fracture detection.
• To develop a robust sensing method suitable for extreme environments where conventional sensors fail.

Main Methods:

• Implementation of a custom-designed integrated tactile and proximity sensor utilizing fiber optics.
• Analysis of sensor data (deformation and proximity) using machine learning techniques for fracture detection and classification.
• Development of a real-time classification system for online surface analysis.

Main Results:

• Achieved an average crack detection accuracy of approximately 94% and width classification accuracy of 80%.
• Statistically significant improvements (p < 0.001) were observed when combining deformation and proximity measurements.
• The fiber optic-based approach shows promise for applications in radiation-heavy environments like nuclear facilities.

Conclusions:

• Integrated tactile and proximity sensing offers a viable alternative for automatic mechanical fracture detection.
• The proposed fiber optic sensor technology is more resilient to extreme conditions compared to traditional electronic sensors.
• This method enhances safety and reliability in critical infrastructure monitoring and maintenance.