Porosity evolution under increasing tension in wire-arc additively manufactured aluminum using in-situ micro-computed tomography and convolutional neural network

Abstract

Internal defects such as porosities are often formed in additively manufactured metal components. The pores nucleate, grow, and coalesce to form cracks under loads, leading to eventual catastrophic failure. In this paper, the full-field porosity evolution, including pore growth and coalescence in a wire-arc additively manufactured (WAAM) aluminum alloy cylinder under tension is observed with in-situ X-ray micro-computed tomography (μCT). The pore size distribution, density, and tensile stress are calculated from the volumetric images analyzed by a convolutional neural network (CNN) algorithm, which provides rapid analysis of 12,950 slice images from μCT volumetric images at the reference state and 13 tensile strains. The results show the quantitative evolution of the growth and coalescence of macropores under tension. A strong correlation is found between the local pore volume fraction and the true tensile stress when the tensile strain is larger than 5%.