Modelling and prediction impression creep behaviour of Al–Cu cast alloy

Abstract

In this study, the effect of cooling rate on microstructure and impression creep behaviour of the as-cast Al–3.7 wt.% Cu alloy was investigated. Impression creep was performed in the temperature range of 423–513 K in stresses of 380–470 MPa for about 5000 s and cooled with different rates. It is deduced that for all cooling rates, lattice, pipe diffusion and climb controlled dislocation creep control the creep rate. The activation energies vary between 82 and 124 kJ/mol which is due to changing the diffusion mechanism from pipe to lattice self diffusion. Moreover, secondary dendritic arm spacing decreases with increasing the cooling rate because of higher constitutional undercooling and formation of secondary dendrite arms. In this research, an artificial neural networks (ANN) structure was proposed to estimate creep of a cast aluminium alloy. Results show that the ANN is a robust approach to predict the creep behaviour of this alloy.