Understanding the ductile-to-brittle transition in disordered solids using structuro-elastoplaisc models

The deformation of disordered solids is governed by a complicated interplay between elastic deformation, plastic particle rearrangements, and their underlying disordered packing structure. Pinpointing the exact influence of microscopic factors such as particle interactions and preparation history on the ductility of a solid is challenging. In this study, we use a recently developed structro-elastoplasic (StEP) framework to model the ductile-to-brittle transition of three distinctive disordered solids: (i) simulated two-dimensional repulsive disks annealed at different temperatures, (ii) experimental two-dimensional rafts of granular materials at an air-oil interface with different particle interactions and (iii) simulated three-dimensional oligomer nanopillars deformed under different temperatures. In all three systems, we used a machine learning method to develop a structural descriptor, softness, which predicts the propensity of a particle to rearrange. Microscopic interplay between softness, elasticity, and rearrangements was extracted and used to inform the StEP models, which in turn captures the shear band formation and the brittle-to-ductile transition in all three systems. Through analysis of the models, we demonstrate how the StEP model can be used to establish the pathway from the respective microscopic factors of the three systems, i.e., preparation history, particle interaction, and temperature, to their ductility upon deformation.