Search

The ductile versus brittle fracture in crystalline materials depends on the relative values of K Ic and K Ie as defined by well-known Rice theory, where K Ic and K Ie are the critical values of stress intensity factor corresponding to cleavage and dislocation emission, respectively. For K Ic < K Ie , the brittle fracture (or cleavage) takes place in atomically sharp pre-cracked crystal subjected to Mode I loading. For K Ie < K IC , the dislocations are emitted from the crack front resulting in ductile fracture. To this end, molecular static simulations are used to explain the crystal orientation dependent fracture behaviour of FCC single crystal and its contradiction with respect to Rice theory based on stress triaxiality at the crack front. The stress triaxiality at crack front changes with crystal orientation due to transformation of stiffness tensor C ijkl . It is shown that high stress triaxiality suppressed the dislocation initiation leading to cleavage failure even for the case when K Ie < K Ic .

Many engineering challenges occur on multiple interacting length scales, e.g. during fracture atoms separate on the atomic scale while plasticity develops on the micrometer scale. To investigate the details of these events, a concurrent multiscale model is required which studies the problem at appropriate length- and time-scales: the atomistic scale and the dislocation dynamics scale. The AtoDis multiscale model is introduced, which combines atomistics and dislocation dynamicsinto a fully dynamic model that is able to simulate deformation mechanisms at finite temperature. The model uses point forces to ensure mechanical equilibrium and kinematic continuity at the interface. By resolving each interface atom analytically, and not numerically, the framework uses a coarse FEM mesh and intrinsically filters out atomistic vibrations. This multiscale model allows bi-directional dislocation transition at the interface of both models with no remnant atomic disorder. Thereby, the model is able to simulate a larger plastic zone than conventional molecular dynamics while reducing the need for constitutive dislocation dynamics equations. This contribution studies dislocation nucleation at finite temperature and investigates the absorption of dislocations into the crack wake.

In this paper we attempt to understand the origins of strain hardening response in glassy polymers at large strains through well-designed molecular dynamics (MD) simulations on an atomistic model of a glassy polymer. In the existing constitutive theories of glassy polymers, strain hardening is assumed to be the result of affine orientation of an underlying entanglement network with stretch. This model is inspired by theories of rubber elasticity. However, in the glassy state the length scale of the network is uncertain although satisfactory fits to experimental stress–strain curves can be obtained using rubber elasticity-inspired constitutive theories. In this work we probe whether the network of entangled macromolecules in the glassy state is capable of deforming affinely obeying Langevin or Gaussian statistics. We also investigate the thermodynamics of the deformation process and try to ascertain if the free energy description of deformation in rubbery materials can be used to model irreversible plastic deformation of glassy polymers. We observe in particular that the 'locked in energy', which is the part of the energy dissipated during deformation that is not converted to heat, shows remarkably different behavior in MD simulations as compared with rubber elasticity based continuum models.