Just another WordPress site - Ruhr-Universität Bochum

Abstract

Metastable austenitic steels may undergo a phase transformation from austenite to ɛ- and/or α-martensite due to stress or plastic deformation. The transformation is highly dependent on local phase stability and crystallographic orientation with respect to the load direction. This phase transformation is similar to the two-way effect observed in shape memory alloys, except that it is irreversible upon unloading, resulting in a stable martensitic phase, at least when a-martensite is the transformation product. The primary goal of this research is to develop a data-driven model for the phase transformation in metastable austenitic steels that incorporates microstructural, thermodynamic, and mechanical loading information. The project is founded on the idea that there exists a broad class of descriptors that encode the topological information of three-dimensional microstructures in a compact format suitable for use as input in supervised machine learning (ML). The primary criteria for these descriptors are that they can be used to characterize both experimental and simulated microstructures in order to generate time-dependent data sets of microstructure dynamical evolution under various loading circumstances. It will be demonstrated in this research that hybrid experimental and modeling data may be generated to serve as the basis for supervised training of a suitable machine learning algorithm. The learned machine learning approach will be used to robustly and numerically efficiently describe the phase transformation of metastable austenite caused by deformation.