An ultrasonic inspection method for diffusion-welded components with variable curvature and thickness.

Area of Science:

• Materials Science
• Robotics
• Non-Destructive Testing

Background:

• Automated ultrasonic inspection of diffusion welds with variable curvature and thickness is difficult.
• Challenges include focusing on curved interfaces, detecting small defects, and the absence of accurate CAD models.

Purpose of the Study:

• To develop a reliable methodology for automated ultrasonic inspection of complex curved diffusion-welded components.
• To address challenges posed by variable geometry and lack of precise surface data.

Main Methods:

• Developed an integrated system combining laser profile scanning and multi-axis robotic ultrasonic testing.
• Resolved synchronization issues between surface data, ultrasonic signals, and robot motion.
• Generated adaptive scanning trajectories based on laser profiles for unknown curved surfaces.
• Utilized ultrasonic field simulations to optimize transducer parameters and analyze geometric effects on defect detection.

Main Results:

• Successfully conducted multi-axis robotic C-scan experiments on titanium alloy diffusion-welded curved blade specimens.
• Demonstrated effective detection of internal defects in complex curved diffusion welds.
• Achieved improved inspection stability and reliability through the proposed methodology.

Conclusions:

• The integrated laser-ultrasonic approach provides a reliable method for automated inspection of complex curved diffusion-welded components.
• The system effectively synchronizes multi-source data and enables adaptive scanning on surfaces with unknown geometry.
• This methodology enhances the feasibility of non-destructive evaluation for challenging industrial applications.