Real-time imaging and geometric characterization of laser ultrasound based on array scanning optimization and delay-multiply-and-sum

• Published in: Mechanical systems and signal processing 2025-02, Vol.224, Article 112206
• Main Authors: Chen, Long, Liu, Zenghua, Zhang, Zongjian, Zhu, Yanping, Liu, Xiaoyu, Hu, Jingdong, He, Cunfu
• Format: Article
• Language: English
• Subjects: Array scanning; Delay-multiply-and-sum; Laser ultrasound; Real-time imaging; SAFT
• ISSN: 0888-3270
• DOI: 10.1016/j.ymssp.2024.112206

•1. A bilateral excitation method is proposed based on the characteristics of detection sensitivity.•2. Real-time laser ultrasound delay-multiply-and-sum SAFT imaging.•3. Geometric characterization of transverse hole and crack defects. Laser ultrasound array scanning imaging has extremely high flexibility in space arrangement and scanning parameter adjustment. However, a major challenge is how to achieve high-speed inspection and high-resolution real-time defect imaging using a limited number of scanning configurations. In this study, the detection sensitivity of laser ultrasonic synthetic aperture focusing technique (SAFT) is analyzed based on ray theory. According to the theoretical calculation results, a bilateral excitation method is proposed. An optimized delay-multiply-and-sum (DMAS) beamforming method was employed for defect imaging. The optimization efforts encompassed refining the mathematical expression, integrating matrix operations with GPU acceleration to enhance imaging speed, implementing coherent noise suppression, and utilizing image fusion techniques to improve imaging quality. The experiments involve two types of defects: transverse holes at different depth positions, and cracks at different angles of additive manufacturing. The results demonstrate that optimizing the scanning configuration improves image resolution, particularly for additive manufacturing of highly scattering materials. Furthermore, the morphological characteristics of defects are revealed, and better imaging and geometric characteristics can be achieved through image fusion. The proposed method outperforms the time-domain delay-and-sum (DAS) and frequency-domain phase shift migration (PSM) algorithms, achieving a real-time imaging frame rate.