Primary creep at low stresses in copper

Abstract

Creep tests at very low stresses are often not run long enough to reach the stationary stage. This means that the test results must be analysed with the help of primary creep models. A previously published such model that does not involve adjustable parameters is reformulated to simplify its integration. It is applied to three sets of test data for copper at 95-150 °C, 600 °C and 820 °C. Primary creep data including that at a very early stage follow the φ model, i.e. the creep rate decreases exponentially as a function of time. The primary creep model can accurately reproduce this behaviour. Previously unpublished creep data at 600 MPa and 1–2 MPa are presented. These data are controlled by dislocation creep. This is demonstrated by a stress exponent of 3, a minimum creep rate that is fully consistent with a stationary creep model and primary creep that can be modelled. In spite of this, the predicted Coble diffusion creep rate lies about an order of magnitude above the observations. The third data set is believed to be controlled by diffusion creep. It can also be represented with primary dislocation creep that gives a stress exponent of 1 at low stresses and a transition to stationary creep at higher stress. This shows that dislocation and diffusion creep are competing processes at high temperatures and low stresses and that careful testing is needed to distinguish between the mechanisms.