A Machine Learning Approach for Segmentation and Characterization of Microtextured Regions in a Near-α Titanium Alloy

• DOI: 10.3390/cryst13101422
• Publication Date: 2023-09-26T07:53:00Z
• Authors: Haodong Rao; Dong Liu; Feng Jin; Nan Lv; Jungang Nan; Haiping Wang; Yanhui Yang; Jianguo Wang
• Keywords: Characterization

The development of automated segmentation and quantitative characterization microtextured regions (MTRs) from the complex heterogeneous microstructures is urgently needed, since MTRs have been proven to be critical issue that dominates dwell-fatigue performance aerospace components. In addition, in Ti alloys similarities encountered other materials, including minerals biomaterials. Meanwhile, machine learning (ML) offers new opportunities. This paper addresses MTRs, where an ML approach, Gaussian mixture models (GMMs) coupled with density-based spatial clustering applications noise (DBSCAN) algorithms, was employed order process orientation data acquired via EBSD Matlab environment. Pixels information through electron backscatter diffraction (EBSD) are divided colored into several “classes” within defined c-axis misorientations (i.e., 25°, 20°, 15°, 10°, 5°), precision efficacy which verified by morphology pole figure segmented MTR. An appropriate range for MTR derived, i.e., 15~20°. contribution this innovative technique compared previous studies. At same time, were statistically characterized global region.