Nanoscale twin boundaries with initial defects were found to be easy dislocation sources in experiments. Here, deformation of nano-twinned single crystals was performed by using atomistic simulations. It was revealed that perfect twin boundaries were potentially not dislocation sources. Another observation was that plastic deformation under uniaxial tensile loading could be dominated by twin boundary migration in samples with an inclined twin boundary orientation. Twin planes with disconnections induced by dislocation/twin-boundary interactions were obtained during nano-indentation. Under further plastic deformation, it was observed that nucleation of dislocations occurred at the disconnections where the coherency of the twin boundaries was destroyed. In general, the stress–strain curve first underwent an elastic linear process, and then dropped abruptly due to dislocation nucleation in the sample. Twin boundary migration was here confined by the vertical orientation of the twin boundaries. Plastic deformation was strongly suppressed by the absence of easy dislocation sources; which was clearly indicated by the stress–strain curve. Dislocations were nucleated from the interior of the sample between two adjacent twin boundaries at very high stresses (about 12GPa). It was demonstrated that ideal twin boundaries were not dislocation sources in homogenous deformation when the monocrystalline samples were defect-free. In nano-indentation force versus displacement curves the initial abrupt load drop was caused by dislocation nucleation under the indenter. Confined by the geometry of the simulated sample, two slip systems were active for dislocation motion. One was parallel to the twin boundaries, while the other was inclined to the twin boundaries at an angle of 70.5°. Attention was focused here upon whether or not pre-existing perfect twin boundaries were dislocation sources under inhomogeneous deformation. Dislocation gliding along the inclined slip system was hindered by the twin plane. It was found that a twin boundary step was left on the twin plane after the dislocation/twin-boundary interaction. Meanwhile, an extrinsic partial dislocation was initiated and glided into the neighbouring twin grain; leading to stacking-fault pinning on the twin plane. As the simulation continued, these residual twin-boundary steps, and pinning stacking faults due to dislocation/twin-boundary interactions, became potential dislocation sources. It had been speculated that pre-existing defects on twin boundaries could be easy dislocation sources. In molecular dynamics simulations, on the contrary, twin planes were initially perfect. They gradually lost coherency due to dislocation/ twin-boundary interaction, and acted as dislocation sources with further deformation.

Can Nanoscale Twin Boundaries Serve as Dislocation Sources in Single Crystals? S.Qu, G.Wang, H.Zhou, Z.Huang: Computational Materials Science, 2011, 50[4], 1567-70