Assessment of Segmentation-Induced Deviations of Porosity Metrics in Powder Bed Fusion Additively Manufactured Components
NONDESTRUCTIVE CHARACTERIZATION AND MONITORING OF ADVANCED MATERIALS, AEROSPACE, CIVIL INFRASTRUCTURE, AND TRANSPORTATION XVIII(2024)
摘要
Post processing X-ray computational tomography (CT) inspection data for additively manufactured (AM) components can induce deviations in defect quantification, affecting subsequent fatigue and failure predictions. To assess the influence and potential impact of segmentation-induced measurement deviations, this paper applies several segmentation techniques to X-ray CT data for powder bed fusion Ti-6Al-4V specimens exhibiting porosity conditions. X-ray CT reconstructions were segmented with varying techniques including Otsu's thresholding, random forest, k-nearest neighbors, and the multilayer perceptron. Metrics such as pore size and global porosity were compared for internal validity. Then, top-down X-ray CT measurements of surface-breaking porosity were compared to optical profilometry for cross-validation.
更多查看译文
关键词
Data fusion,X-ray computed tomography,image segmentation,supervised classification surface,profilometry,additive manufacturing,image registration
AI 理解论文
溯源树
样例
![](https://originalfileserver.aminer.cn/sys/aminer/pubs/mrt_preview.jpeg)
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要