|About this Abstract
||2018 TMS Annual Meeting & Exhibition
||Phase Transformations and Microstructural Evolution
||Using Multiparadigmatic Approach in Microstructure Evolution Prediction of Two-Phase Titanium Alloys: Linking Artificial Neural Networks, 2-point Statistics, Multiphase-field Methods and Self-consisting Analytical Models. Building Integrated Computational Materials Engineering (ICME) and Materials Data Infrastructure (MDI)
||Anton Ektov, Surya R. Kalidindi, Yuksel C. Yabansu, Xinyi Gong, Jeoung-Han Kim
|On-Site Speaker (Planned)
To predict microstructure evolution in Ti-alloys under wide range of temperature-deformation conditions - multiparadigmatic approach was developed on the base of coupling different simulation software into one integrated feed-forward workflow. A representable Ti-database of properties was collected from the laboratory electronic protocols of mechanical properties. Using 2-point statistics allow to transform origin microstructure images from SEM into unique correlation fields. Coupling correlation images, chemical composition and HT routes on the input, makes possible to predict mechanical properties of the alpha/betaTi-alloys as a function of the input parameters using ANN. Developing analytical models is the significant layer in multiparadigmatic system. New rate formulation for the static and dynamic recrystallization was implemented on the base of FE DEFORM-3D software. Parameters such as temperature, strain/stress tensor components are passed into multiphase-field software MICRESS in order to predict the evolution of Ti-microstructure. Corrected Ti-database from Thermo-Calc software was used to calculate required thermo-kinetic parameters.
||Planned: Supplemental Proceedings volume