The effect of dislocation density on the creep of dispersion-strengthened copper crystals

Abstract

Creep tests were conducted on copper crystals containing dispersions of SiO 2 or Al 2O 3 produced by internal oxidation. A comparison was made between well-annealed crystals with a dislocation density ϱ of 4 × 10 10 m −2 and swaged-and-annealed crystals with a dislocation density ϱ of about 10 14 m −2; the changes in dislocation distribution were followed by transmission electron microscopy. The well-annealed crystals exhibited a creep limit which corresponded to the tensile yield stress, although no such limit was observed in the crystals containing a dislocation substructure. The steady state creep rate data are discussed in terms of the empirical equations dot ϵ = Aσ A n exp(− Q/kT) and dot ϵ = K exp( Bσ A) exp(− Q/kT) . The dislocation microstructure of the crept specimens consisted of a high proportion of aligned dipoles, mainly of edge orientation. It is concluded that the main effect of the dispersed phase is to retard the rate of recovery in the glide-recovery sequence. From the comparison of the well-annealed and swaged-and-annealed material the major contribution to creep strength arises from the presence of a dislocation substructure. The application of particle-induced internal stress concepts does not provide a satisfactory interpretation of the creep behaviour, and an approach is used which is based on the role of the mobile dislocation density. At low dislocation densities the dislocation link length distribution is determined by the effective particle spacing. When the dislocation density exceeds D s −2 (where D s is the particle spacing), dislocation nodes will influence the dislocation segment length, thus decreasing the number of mobile dislocations.