Research on Laser Ultrasonic Total Focusing Imaging Method Based on ZYNQ Acceleration
Chanjuan Wu,Kaining Ying,Lunan Dai,Junfu Tan,Zhonghua Shen,Chenyin Ni
DOI: https://doi.org/10.3788/CJL221223
2023-01-01
Abstract:Objective With the continuous emergence of various materials and the improvement of the quality requirements of modern industry for mechanical parts, non-destructive testing technology has been rapidly developed. Ultrasonic non-destructive testing has been widely used because of its high sensitivity, penetrability and other advantages; it has become an indispensable tool in aerospace, railway, oil pipeline, and other industrial testing. As a hot spot in the field of industrial non-destructive testing in recent years, phased array testing technology exhibits the characteristic of flexibility in focusing. However, all ultrasonic sub beams are focused at a constant depth, and reflectors outside the focus area cannot be detected sensitively. Synthetic aperture focusing technology (SAFT) can synthesize multiple small numerical aperture transducers into large numerical aperture transducers, which can obtain higher resolution reconstructed images and provide a more reliable basis for the qualitative analysis of defects. The total focusing method (TFM) is developed by combining phased array detection technology and synthetic aperture focusing technology. The total focusing method uses all the transceiver combinations in the phased array elements to detect defects and performs image post -processing on the obtained full matrix data. It uses all the information of the data to recover the detection signal to the maximum extent, and obtains imaging results with a stronger defect characterization ability and higher resolution. However, it has many limitations such as a large data demand, complex calculations, and the need for a substantial amount of time. To address these issues, the laser ultrasonic total focusing imaging method based on ZYNQ acceleration is studied in conjunction with laser ultrasonic testing technology. Methods First, the total focusing imaging method based on laser ultrasound is established, and the laser ultrasound directivity coefficient is then introduced to optimize the algorithm by analyzing the imaging principle. Furthermore, a laser ultrasonic scanning detection device is built for experimental verification. The ultrasonic signal is excited by a linear laser source and the echo signal is detected by a Doppler vibrometer. The laser ultrasonic total focusing imaging method is then used to detect and locate the internal defects, and the results are compared with those obtained by synthetic aperture focusing technology. Finally, the personal computer (PC) is connected to the ZYNQ-7000 development board to test the data received by the PC. The total focusing imaging method is accelerated based on the loop expansion and pipeline principles of the ZYNQ programmable logic (PL) part and the dual core design principle of the processing system (PS) part, and the results are analyzed. Results and Discussions Before and after the introduction of the laser ultrasonic directivity coefficient, the imaging results reflect the defect location (Fig. 10). In contrast, in the results obtained by the original total focusing method, the high amplitude signals are scattered around the defects. In the results obtained by introducing the laser ultrasonic directivity coefficient, the high amplitude signals are mainly concentrated at the defects, and the image signal-to-noise ratio is higher. Therefore, the introduction of the excitation directivity coefficient can effectively suppress noise and improve the image signal-to-noise ratio. Moreover, compared with the SAFT image (Fig. 11), the laser ultrasonic total focusing imaging method has a smaller defect position error, a higher signal-to- noise ratio of the defect image, and a stronger defect characterization ability. The algorithm is then transplanted to the ZYNQ platform for acceleration. Compared with the PC based imaging results, the imaging effects of the two are essentially the same (Fig. 12); however, ZYNQ has the advantages of time and cost (Table 2). The time consumption is reduced to 317/1000 that of the PC, and the cost is reduced to 3/10 that of the PC. This shows that ZYNQ has a higher cost performance ratio. Conclusions An algorithm of the laser ultrasonic total focusing imaging method based on ZYNQ is investigated in this study. The results indicate that the signal-to-noise ratio of the full focus image is improved significantly after the introduction of the laser ultrasonic directivity coefficient. Upon comparing the imaging results of synthetic aperture focusing technology and the total focusing imaging method, it is noted that the total focusing imaging method image has a smaller error and higher signal-to-noise ratio. The laser ultrasonic total focusing imaging method has a stronger defect characterization ability and higher signal-to-noise ratio, which further verifies the feasibility of the laser ultrasonic total focusing imaging method. After using ZYNQ to optimize the laser ultrasonic total focusing imaging method, it is deduced that the final imaging effect based on the ZYNQ calculation is essentially the same as that based on the PC calculation. However, the calculation time of ZYNQ is reduced by 86%, and the cost is reduced by 68.5%; this further confirms the feasibility of the laser ultrasonic total focusing imaging method based on ZYNQ.