Probing Defect-Controlled Small Scale Plasticity via Machine Learning Approach and Defect Dynamics Modeling

Mechanical behaviors of metals at small scale have attracted much attention due to their widespread applications in modern MEMS/NEMS devices due to their enhanced properties. To understand the underlying microscopic deformation mechanisms that control the mechanical properties of nanostructured metals, an insight into the intricate interaction of dislocations and grain boundaries is vital. Computational simulations of materials have advanced significantly in recent years, shifting from analyzing experimental observation to providing capability to predict mechanical behaviors for use in advanced material development. Machine learning has been a topic of great interest in recent years within the engineering community.<br/>In this study, we employed a multiscale modeling approach to investigate defect interaction. For the prediction of dislocation interactions with grain boundaries, we will present an efficient physically informed machine learning framework that has potential to significantly improve predictive capability of computational modeling, which in turn would reduce the number of prototypical experimental validation. With the atomistically informed mesoscale defect dynamics model, we explore the effect of varying misorientation angle for pure twist and pure tilt grain boundaries on plastic deformation of nanostructures, which could provide a better understanding of dislocation driven plasticity for polycrystalline metals at small scale. For complex loading and environmental condition, the unified defect dynamics model which coupled dislocation dynamics (DD) and finite element model (FEM) has used, which could play an important role in obtaining a fundamental understanding of deformation mechanism at small scale. The developed multiscale framework will shed light on fundamental investigation of “defect-controlled” mechanical behaviors in metallic materials.