Modification of the Wilshire strain equation: An application to RR1000

Abstract

Being able to accurately predict creep strain as a function of time is important both in preventing aeroengine blades rubbing against their outer casings, but also in being able to convert small punch test data into equivalent uniaxial test results using finite element models. Modern studies have found success in applying the Wilshire equations to uniaxial creep test results. The capability of the Wilshire equation to interpolate creep curves was assessed using uniaxial creep tests carried out on RR1000. In this paper, an artificial neural network (ANN) was used to modify the Wilshire equation for the time taken to reach various strains so that the parameters of the Wilshire equation could be interpolated as a function of strain. The model was then evaluated using statistics on predictive accuracy, which showed that the model was capable of predicting the shape and scale of the creep curves with high accuracy. These modifications also revealed that the activation energy is dependent upon the average internal stress and thus strain.