Characterizing the Local Yield Surface in Simulated Glasses

Amorphous materials, unlike their crystalline counterparts, have no obvious defects that help facilitate plastic instability, and so methods that predict the locations of plastic events are highly sought after. Two-dimensional model glasses have been used extensively to study amorphous materials and to develop a theoretical basis for determining the location of plastic events. One method that has resulted from such investigations is the Local Yield Stress (LYS) method in which nanoscale regions of the material are probed to quantify the incremental stress necessary to induce plastic rearrangement. The resultant local yield stress depends on the sense of the mechanical load applied on a local region. Higher accuracy predictions can be obtained by sampling many disparate mechanical loads. Doing so allows one to develop a more accurate picture of the “local yield surface,” the limiting stresses that the material can withstand before yielding along all possible mechanical loadings. In three dimensions, the number of distinct mechanical tests needed to accurately sample the local yield surface increases. The resulting dataset is computationally expensive to produce. We seek to use this data to compactly characterize this high-dimensional “surface” and to enable extrapolation along deformations that were not explicitly tested in the original dataset. This can, in principle, be accomplished by describing the local yield surface in terms of a discrete number of Shear Transformation Zones (STZs). By characterizing the number of STZs in a local region we will be able to predict areas of high plasticity more accurately in amorphous materials, eventually leading to the accurate modeling of plastic response and the design of materials with mechanical properties catered to specific needs.<br/><br/>SNL is managed and operated by NTESS under DOE NNSA contract DE-NA0003525.