LAUSR

Abstract A model for the formation and growth of type II twins is described using the topological theory of interfacial defects and interface structures. A type I twin forms and grows to macroscopic dimensions by the generation and expansion of disconnection loops, provided these defects are sufficiently mobile. However, if their mobility is limited, they accumulate into a tilt wall, which, after accommodational relaxation, forms the type II conjugate twin. Thus, whether the type I or type II conjugate twin forms is the outcome of competitive mechanisms, depending primarily on disconnection mobility. The plausibility of this model is discussed with reference to experimental observations of twinning in α − U . Disconnection mobility is shown to be limited by atomic shuffling in the cases of " { 17 6 ¯ } " and " { 1 7 ¯ 2 } " type II twins, as compared with the higher mobility expected for the active disconnections in { 1 3 ¯ 0 } compound twins. In the topological model, the twinning shears are identical to those predicted by the classical model of deformation twinning. However, while type I twins are formed directly by shear due to the motion of disconnections on their glide planes, the mechanism of type II twinning is different, involving not only shear by disconnection motion but also accommodational relaxation. This understanding prompts a reassessment of the physical significance of the twinning elements of the classical geometrical approach.