|About this Abstract
||Materials Science & Technology 2017
||Fifty Years of Metallography and Materials Characterization
||Next Step in Complex Microstructure Classification – How to Replace Subjective Expert Bias by 3D Information and Autonomous Procedures?
||Frank Muecklich, Dominik Britz, Jessica Pauly
|On-Site Speaker (Planned)
The precise quantification and classification of different phases and microstructural constituents, e.g. of steel, is essential for a reliable quality control, since the microstructure tolerances are getting increasingly tighter in high performing materials. Thus, the subjective bias of individual experts poses an increasing problem in reliable microstructure classification. Autonomous classification would offer a smart solution. In complex steel microstructures, the geometrical details of morphologies but also the substructure of the phases has to be recorded with sufficient accuracy with correlative microscopy methods to overlay corresponding light- and electron microscopy images. The study gives evidence that smart data mining combined with an additional pixel-based analysis of the substructure as well as 3D information of the microstructural constituents from nonrecurring tomography is a powerful toolbox to reliably discriminate broad variations of steel microstructures in an autonomous classification methodology with striking performance.
||Planned: Publication outside of MS&T