Abstract
Calculations of stacking-fault energies in aluminum, magnesium, beryllium, copper, silver, and gold are reported. For the polyvalent metals it is shown, by means of comparisons of numerical results based on several different energy-wave-number characteristics, that exchange and correlation corrections are not as significant in the present context as they are in most other defect calculations. Nonlocal effects are quite important but can be approximately accounted for by empirical adjustments of the form factors based on simpler local models. Although stabilities against faulting and the relative magnitudes of stacking-fault energies are correctly predicted, quantitative agreement with experiment is not obtained. Possible reasons for this discrepancy are discussed. Results for the noble metals indicate a severe sensitivity to the detailed curvature of the energy-wave-number characteristic in the vicinity of the smallest reciprocal-lattice vector. Failure to obtain agreement with experiment in these cases may therefore be due to minor inaccuracies in the energy-wave-number characteristics for noble metals.