Creep tests at very low stresses are often terminated before the stationary stage is reached, which makes it essential to use a model for primary creep to assess the results. Such a model was developed in a recent paper by the author. The model is applied to creep at very low stresses where dislocation creep is the controlling mechanism. Two cases are investigated. i) Aluminium at very high temperatures. This is usually referred to as Harper-Dorn creep. ii) The austenitic stainless steel 316H at 700ºC. Most aluminium data for the considered case can be modelled assuming stationary conditions, but some data has to be represented with the primary creep model giving stress exponents down to 1. Contrary to the case for aluminium, the data for 316H at low stresses are very far from stationary conditions. The model results give in agreement with observations a stress exponent of 1 and stress strain curves consistent with the ϕ (phi) model, i.e. an exponentially decreasing strain rate with increasing time. The consistency with the ϕ model verifies that dislocation creep is the controlling mechanism in spite of the fact that the Coble model gives a higher strain rate for diffusion creep.
Read full abstract