Identification of grain boundary dislocations

Abstract

The simultaneous two-beam method for identifying defects in grain boundaries developed in Part I is demonstrated by using image matching to determine the Burgers vectors of GBDs introduced into coherent twin boundaries in copper by deformation at room temperature. The choice of this special twin system involves no loss of generality in the identification procedure, but provides an idealized test case where the diffraction parameters, the twin crystallography and the expected Burgers vectors of the GBDs are well defined. Furthermore, this system also allows the ‘same-g’ diffraction condition to be set up and tested. It is found that for the general simultaneous two-beam case the Burgers vectors of GBDs can be easily and positively identified and the high degree of agreement between matching sets of theoretical and experimental micrographs indicates that the assumptions used in computing the theoretical images are a good approximation to the experimental conditions. However, the matches between theoretical and experimental images in the same-g case, whilst adequate for identification purposes, were not quite so good as those obtained for the general case. This is thought to be due to the fact that, in the particular experimental cases encountered here, the diffracting vectors in each grain were not exactly the same so that the perfect coupling between the diffracted beams at the boundary, assumed in the theoretical images, was not being achieved in practice. The Burgers vectors which are identified in the coherent twin boundaries are of the type 1/6〈112〉 lying in the boundary, 1/3 〈111〉 normal to the boundary and 1/2〈110〉 oriented relative to the dislocation line direction so that they are sessile (Lomer) dislocations. It is shown how the positive identification of these dislocations leads to a description of the detailed structure of the boundary and the way in which the GBDs have been formed. The question as to whether the dislocations lie in the boundary or merely close to it is discussed.