It was noted that the proposed explanation for the origin of the long-range internal stresses in dislocation wall structures was closely related to the composite model, and the basic idea of compatible deformation of the channels and walls of the dislocation structure remained the same. The averaged local stresses again had to equal the applied stress. However, the shape of the glide dislocations in the channel was not assumed  a priori, but was calculated by using the balance of forces which acted on the glide dislocations. This was made possible by assuming a typical configuration for the glide dislocations which passed across the walls. In this way, long-range glide of the dislocations was constrained by the longitudinal walls, which was the reason for bowing of the glide dislocations. Thus, instead of edge dislocation segments at the interfaces, the bowed glide dislocations in the channels were considered to be the primary source of internal stresses. The role which was played by the interface edge segments was not underestimated. They were expected to permit the penetration of new dislocations into, and through, the walls. However, the corresponding temporary, and strictly local, high stresses were not involved in the simple model that was proposed. The original composite model explained the internal stresses in dislocation wall and cell structures with thick boundaries. Internal stresses of the same type were also measured in sub-grain structures, but could not be explained within the original framework. The proposed model was essentially independent of the width of the walls or boundaries, and its extension to the latter case was possible. The original model could explain the internal stress of edge segments at the interfaces. However, this length was not amenable to direct experimental verification. On the other hand, the proposed model predicted definitely measurable results on the basis of experimentally based input data. Thus, experimental disproof of the computed results was feasible.

R.Sedlácek: Scripta Metallurgica et Materialia, 1995, 33[2], 283-8