Multiscale modeling of plasticity in a copper single crystal deformed at high strain rates

Abstract

A hierarchical multiscale modeling approach is presented to predict the mechanical response of dynamically deformed (1100 s −1 −4500 s −1 ) copper single crystal in two different crystallographic orientations. An attempt has been made to bridge the gap between nano-, micro-and meso-scales. In view of this, Molecular Dynamics (MD) simulations at nanoscale are performed to quantify the drag coefficient for dislocations which has been exploited in Dislocation Dynamics (DD) regime at the microscale. Discrete dislocation dynamics simulations are then performed to calculate the hardening parameters required by the physics based Crystal Plasticity (CP) model at the mesoscale. The crystal plasticity model employed is based on thermally activated theory for plastic flow. Crystal plasticity simulations are performed to quantify the mechanical response of the copper single crystal in terms of stressstrain curves and shape changes under dynamic loading. The deformation response obtained from CP simulations is in good agreement with the experimental data.

Comparison of Dislocation Density Tensor Fields Derived from Discrete Dislocation Dynamics and Crystal Plasticity Simulations of Torsion

“…7b, given the Peach-Kohler forces in Eq. (15). For the two cases, the orientations of the slip planes relative to the wire are different, but elastic stress field is same, up until yield, due to the isotropic moduli assumption.…”

“…Since that time, a number of researchers have made connections between dislocation mechanisms and crystal plasticity, see e.g. [13][14][15].…”

Dislocation density-based plasticity model from massive discrete dislocation dynamics database

“…( 1) and other constitutive relations [8,9,10,11] commonly used in engineering calculations [12] are still phenomenological, i.e., they are not derived from fundamental dislocations physics [13]. These models are typically parametrized by fitting to experimental data under uniaxial loading [14,15,16]. As a result, their extensibility to conditions beyond those used for fitting as well as their predictive capabilities under more complex scenarios remains unclear.…”