CCES Unicamp

Multiscale Modeling of Dynamic Failure in 3D Polycrystalline Materials using BEM and MD

Date: 01/25/2019, 14:00

Candidate: Andrés Felipe Galvis Rodriguez

Advisor: Prof. Paulo Sollero

Sala JD2/FEM/Unicamp

Abstract:

This thesis presents a multiscale approach to analyze the dynamic intergranular failure in 3D polycrystalline materials. The model comprises the meso and atomistic scales using the boundary element method (BEM) and molecular dynamics (MD), respectively. A detailed polycrystalline structure is considered in the mesoscale, where stochastic grain morphologies, random crystalline orientations and initial defects are included in the physical model. Owing to its heterogeneous character, polycrystal aggregates tend to be macroscopically isotropic when the number of crys- tal grains is large. This fact facilitates the evaluation of the influence of the dynamic effects on the mechanical behavior. The absence of analytical dynamic models for these stochastic materials has been a challenge in validating the numerical results. Therefore, a computational framework is proposed to show the validation of the elastodynamic BEM formulation for these materials. Stress and strain waves propagate through the polycrystal, inducing the material to be more susceptible to fail. The intergranular failure is governed by the critical energy density, taking into account the energy density dependency on the interface lattice structures of a set of nano-grain boundaries. In order to connect the scales, and due to the high variation of the mechanical properties with respect to the scale size, the asymptotic scaling methodology applied to the yield strength is adopted as an approximation. Finally, numerical results of the dynamic intergranular failure are presented for various dynamic loads.

Related posts

Contact-based models in the study of folding and thermal diffusion in proteins

cces cces

Topology Optimization of fluid-actuated cellular mechanisms.

cces cces

Hyper-reduction and calibration techniques for projection-based reduced order models of unsteady flows

cces cces