Multiscale plasticity of geomaterials predicted via constrained optimization‐based granular micromechanics

Abstract

A general framework to derive nonlinear elastic and elastoplastic material models from granular micromechanics is proposed, where a constraint‐based variational structure is introduced to classical grain contact‐based homogenization methods of hyperelasticity. Like the classical hyperelastic methods, reference solutions for closed‐form hyperelastic material models are analytically derived from the grain‐scale contact mechanics. However, unlike prior methods, the proposed homogenization framework defines closed‐form hyperelastoplastic material models that extend multiscale variational methods to granular plasticity. The proposed framework is used to develop novel granular micromechanics‐based macroscopic models for a Mises type solid, Drucker–Prager type plasticity, and grain‐contact cohesive‐debonding with a deviatorically and volumetrically coupled nonlinearly elastic response. Macroscopic plastic parameters and yield criteria are explicitly related to their microscale counterparts, for example, the friction coefficient governing intergranular slip. Numerical examples and comparison to measurements from the literature, including triaxial compaction of concrete, are provided to investigate model predictions and demonstrate calibration to experimental data.

Multiscale numerical investigation of ratchet growth damage effects in PBX 9502

“…Because our Automated segmentation method is affected to a lesser extent by the constraints of geometry and size, this method can easily be expanded for changes in geometry: larger fields of view, and progressing to 3D microstructural models, which is part of our current and ongoing work. Finally, we note that the methods and results provided here are important to inform our present granular homogenization and data-driven modeling research [33][34].…”

An integrated EOS, pore‐crush, strength and damage model framework for near‐field ground‐shock

“…Geomechanics models for elastoplasticty and damage have of course also been a topic of much development. Pressure dependency of strength has long been well-understood, 21,22 but continues to be developed, for example, for multiscale modeling, 23 coupling to damage, [24][25][26] anisotropy [27][28][29] and for various other needs. 30 Similarly, computational models inclusive of compactive yielding have been well elucidated, 31 but there remain various topics of active development, such as coupling to damage, [32][33][34] thermodynamic-consistency at large deformation, 35 and dynamic applications.…”