GPU accelerated 3-D ultrasonic flaw imaging using full waveforms sampling and processing technique

Abstract

We develop a three dimensional (3-D) flaw imaging system using the full waveforms sampling and processing (FSAP) technique and an ultrasonic matrix array transducer. In FSAP, each array element is sequentially used as an emitter and all other array elements are used as receivers. By changing the emitting element, we obtain a set of signals of every two-element combination. The combination of all signals enables the generation of focal beams at any point in the region of interest. Using the focal beams, we can reconstruct high-resolution 3-D images of flaws. In this study, we introduce massively parallel calculations with graphics processing unit (GPU) to accelerate the beam-forming in FSAP. Here, we implement the GPU accelerated FSAP on a signal acquisition system with a multiplexer, which can rapidly switch the ultrasonic transmission and reception. We validate the performance of the FSAP imaging system by 3-D numerical simulations with the elastodynamic integration technique and experimental measurements with an acrylic specimen. The results show high-speed calculation of the beam-forming, and accurate and clear flaw images.

Keywords

• ultrasonic testing
• matrix array transducer
• 3-d flaw imaging
• full waveforms sampling and processing (fsap)
• full matrix capture (fmc)
• gpu computation
• efit simulation