Artificial Intelligence for the Output Processing of Phased-Array Ultrasonic Test Applied to Materials Defects Detection in the ITER Vacuum Vessel Welding Operations

Abstract

Artificial Intelligence (AI) has been applied to accelerate the analysis of outputs from the phased-array ultrasonic (PAUT) non-destructive testing (NDT). The aim of this work is the development and analysis of AI tools for the output processing of PAUT applied to welding defects detection in the ITER Vacuum Vessel manufacturing. The development of the AI model, parameters of the data and data processing are described herein. This development shows that AI is an appropriate tool to process PAUT data, resulting in an accuracy of prediction of over 83%. This allows for prompter data availability and gives an additional information set in order for projects to take informed decisions. The human error factors are decreased through this automation, as is the large time required to process each PAUT output, which can be decreased from an average of a week to a matter of minutes. A successful AI application for UT has a potential to save time and cost.