Novel texture analysis method for optimising material property in extruded 6xxx alloys using artificial neural networks
Year: 2025
Type: article
Abstract: This study investigates the extruded texture of a 6xxx series high-strength aluminium alloy as a function of profile geometry using Electron Backscatter Diffraction (EBSD) and X-Ray diffraction pattern (XRD). A novel texture analysis method was designed to acquire and prepare reliable texture data for machine learning applications. The method categorizes textures into five distinct groups, with volume fractions calculated for each group. Furthermore, finite element analysis of the extrusion process revealed that axial tensile strain promotes a combination of 〈100〉 and 〈111〉 //ED texture components, while shear deformation induces 〈211〉 //ED texture components. The results were subsequently fed into an artificial neural network (ANN) model developed to link the texture to profile geometry, which governs the deformation modes experienced during the material flow. This approach represents a significant advancement towards real-time control of material properties during extrusion. • A novel texture analysis method to acquire reliable data for machine learning applications. • FE analysis shows axial tension promotes <100> and <111>//ED texture components. • Shear deformation induces <211>//ED texture components. • Using digitised texture data, an ANN model was developed to predict deformation modes.
Source: Materials Characterization
Authors: Mian Zhou, Chrysoula Tzileroglou, Carla Barbatti, Hamid Assadi
Institutions: Brunel University of London, Constellium (France)
Language: English
Subfield: Materials Chemistry
Field: Materials Science
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $3,320
OpenAlex ID: W4407800484