Low stress creep of copper and some aluminium and magnesium alloys at high and intermediate homologous temperatures.

Creep behavior of high purity copper under low stress has been investigated in tension for bamboo grain-structured wires 25 -500 pm diameter and foils 0.4 and 0.6 mm thick and, in bending, for foils 100 and 250 pm thick. Additionally, creep behavior of polycrystalline 7075 aluminium and AZ61 magnesium alloy has been investigated under low applied tensile stress. The conditions explored feature diffusional creep and related mechanisms expected to be operative at the high temperatures and low stresses involved. The creep and surface profile of copper (>99.99%) wires has been investigated close to their melting temperature (0.93 Tm) under stresses up to 0.35 MPa for which strain rate varied linearly with stress. For the thinnest wires (diameter 25- 125 pm), the strain rate was about twice that expected from Nabarro-Herring diffusional creep theory and between one and two orders of magnitude larger than expected from Harper-Dorn creep. For 500 pm diameter wire, the measured rate was initially near to Harper-Dorn prediction but became constant only at longer durations at a level about five times lower than this. The lower rates were about 1.5 times that expected from diffusional creep. The surface profile observations indicated a small contribution of grain boundary sliding to the creep process when grain boundaries were not closely perpendicular to the stress. The observed effect of grain aspect ratio on the creep rate is shown to provide better correlation with theory. Tensile creep tests were carried out on OFHC copper foils at 850°C and 990°C in the stress range 0.1-0.6 MPa. The stress exponent for creep was found to be close to 2 and measured rates were about two orders of magnitude faster than expected from diffusional creep. Slip lines, approximately 30 pm apart, were observed on the surface after creep. The creep process in these foils under tensile loading is ascribed to glide of dislocations controlled by the rate of generation of dislocations at Bardeen-Herring sources about 30 pm apart. The creep tests in bending (which are novel) were carried out at 950°C in cantilever configuration loaded under self weight. The measured profile of the crept foils confirmed the linear dependence of strain rate on stress with final curvature 7-13 times lower than predicted from diffusional creep theory. A hundred nanometer thick alumina coating was applied to some copper foils prior to creep exposure. The associated localization of strain at grain boundaries was found to result in fracture of a 100 nm thick alumina coatings there at extremely low applied stress and overall strain. Tensile creep test of thermomechanically treated 7075 aluminium alloy of initial grain size 48 (am at <5MPa and 350 to 410°C showed a stress exponent close to 1. After correcting for grain growth to 79 pm during the test, the creep rates were within a factor of two of those expected for Nabarro-Herring creep. The creep rates were found to be lower for longer test durations evidently due to grain growth at test temperature and thus indirect evidence for dependence of N-H creep rate on grain size was obtained. True activation energy for creep was found to be close to 165 kJ/mol comparable to the aluminium self diffusivity. For AZ61 magnesium alloy at 250 to 346°C, and stresses upto 6 MPa, Bingham type behaviour was observed with threshold stress decreasing with increasing temperature. The corresponding activation energy for creep was 106 ± 9 kJ/mol comparable with that expected for grain boundary self diffusion in magnesium with the resulting values of grain boundary diffusivity closely matching those obtained previously for Coble creep in pure magnesium. Grain elongation in the direction of the application of tensile stress was observed also to be consistent with operation of Coble creep. Strain rate versus stress for both these materials are shown to be continuous with published results for superplastic flow under comparable conditions.