Automated Flaw Detection in Multi-channel Phased Array Ultrasonic Data Using Machine Learning

Abstract

Modern ultrasonic inspections utilize ever-richer data-sets made possible by phased array equipment. A typical inspection may include tens of channels with different refraction angle, that are acquired at high speed. These rich data sets allow highly reliable and efficient inspection in complex cases, such as dissimilar metal or austenitic stainless steel welds. The rich data sets allow human inspectors to detect cracks with low signal-to-noise ratio from the wider signal patterns. There’s a clear trend in the industry to even richer data sets with full matrix capture (FMC) and related techniques. Convolutional neural networks have recently shown capability to detect flaws with human level accuracy in ultrasonic signals at the B-scan level. To enable automated flaw detection at human-level accuracy for critical applications, these neural networks need be developed to take advantage of today’s rich phased array data-sets. In the present paper, we extend previous work and develop convolutional neural networks that perform highly reliable flaw detection on typical multi-channel phased array data on austenitic welds. The results show, that the modern neural networks can accommodate the rich ultrasonic data and display high flaw detection performance.

A Systematic Review of Advanced Sensor Technologies for Non-Destructive Testing and Structural Health Monitoring

“…In contrast with conventional PAUT, the recorded data files are huge, resulting in slower acquisition speeds. To improve and automate the data analysis, the capabilities of using machine learning models machine on FMC data were investigated by Siljama et al [ 182 ]. The researchers compared the model performance to the results of a human inspector using real thermal fatigue cracks with similar weld geometries used in training.…”

Ultrasonic and phased-array inspection in titanium-based alloys: A review

“…For instance, machine learning with developed neural networks was employed to optimally adjust and accommodate the rich ultrasonic data and enhance high flaw detection performance. [115][116][117] One of the serious issues in composite materials is delamination which should be thoroughly addressed since composites have a tendency to delaminate, resulting in the reduction of strength, stiffness, and piece shelf life, 118,119 therefore, delamination defect detection in composite components is of high significance. Tian et al 120 use a rapid-guided ultrasonic wave inspection method by PA beamforming approach to detect delamination location, size, and shape.…”

“…For instance, machine learning with developed neural networks was employed to optimally adjust and accommodate the rich ultrasonic data and enhance high flaw detection performance. 115–117…”

Adaptive Enhancement of High Greyscale Weld Images Based on the New Pixel Self-transformation Method

“…With the development of computer technology, the traditional method of X-ray film inspection using searchlights is gradually eliminated, while the industrial X-ray films captured by existing line array cameras are 12-bit high dynamic range images, which cannot be displayed properly on existing monitors. Secondly, to extract high-quality industrial radiographic films for weld port number, weld date identification, and defect detection [1] [2], corresponding enhancement is required to improve image clarity for discriminating image information [3] [4].…”