|About this Abstract
||2016 TMS Annual Meeting & Exhibition
||Computational Methods for Spatio-temporal Scale-bridging: from Atomistics to Mesoscale
||Towards Real-time Multi Scale Modeling
||GŁnter Gottstein, Markus Kuehbach, Luis Barrales-Mora
|On-Site Speaker (Planned)
The physical properties of metallic materials are defined on a macroscopic scale but are controlled on a mesoscopic scale, i.e. by the microstructure, which changes during processing through the activity of atomistic mechanisms. Prediction the final properties of metallic products requires a computation of microstructural change during the entire processing chain. This involves a variety of computer simulations with typically extensive demand on computer power and computation time, typically orders of magnitude above the real processing time.
To substantially speed up computation we propose a new approach that combines mesoscopic modeling with statistics in order to make fast predictions. For this purpose we maximize the benefits of massive parallel computer architectures and utilize independent microstructural operators in combination with appropriate ensemble statistics to anticipate a response on a macro scale.
We will present examples for the modeling of deformation texture as well as recrystallization kinetics.
||Planned: A print-only volume