The dependence of cross-slip on stacking-fault energy in face-centred cubic metals and alloys

Abstract

The results of an experimental study of the temperature and strain-rate dependence of τ_III for Cu–Zn alloys are described and interpreted in terms of Seeger's theoretical analysis of τ_III. The values of the stacking-fault energy, γ, derived in this way are compared with the estimates of γ for the same alloys obtained directly from electron microscope observations of dislocation nodes. The two sets of values are found to disagree, and the nature of the discrepancy is such as to throw serious doubts on the applicability of the Seeger analysis to Cu-based alloys with e/a >1.10. The lower limits of γ for pure Cu and Ag, from electron microscope data, are ∼60 ergs/cm² and ∼20 ergs/cm²; the values of γ deduced from Seeger's τ_III analysis are ∼170 ergs/cm² and ∼30 ergs/cm² respectively. The lower limit of γ for Cu is inconsistent with the previously accepted figure based on the assumption that γ is twice the twin boundary energy, and this assumption is now held to be invalid. Seeger's model of cross-slip at Lomer–Cottrell barriers is examined critically, and found to be incompatible with the observations in Cu and Al that screws are held up preferentially. It is proposed that screws are stopped by becoming heavily jogged in the dense tangles observed by transmission microscopy, and that cross-slip occurs at these tangles by processes controlled by jogs.