Optimising the multiplicative AF model parameters for AA7075 cyclic plasticity and fatigue simulation

Abstract

Purpose This study presents the improvements of the multicomponent Armstrong–Frederick model with multiplier (MAFM) performance through a numerical optimisation methodology available in a commercial software. Moreover, this study explores the application of a multiobjective optimisation technique for the determination of the parameters of the constitutive models using uniaxial experimental data gathered from aluminium alloy 7075-T6 specimens. This approach aims to improve the overall accuracy of stress–strain response, for not only symmetric strain-controlled loading but also asymmetrically strain- and stress-controlled loading. Design/methodology/approach Experimental data from stress- and strain-controlled symmetric and asymmetric cyclic loadings have been used for this purpose. The analysis of the influence of the parameters on simulation accuracy has led to an adjustment scheme that can be used for focused optimisation of the MAFM model performance. The method was successfully used to provide a better understanding of the influence of each model parameter on the overall simulation accuracy. Findings The optimisation identified an important issue associated with competing ratcheting and mean stress relaxation objectives, highlighting the issues with arriving at a parameter set that can simulate ratcheting and mean stress relaxation for load cases not reaching at complete relaxation. Practical implications The study uses a strain-life fatigue application to demonstrate the importance of incorporating a technique such as the presented multiobjective optimisation method to arrive at robust parameters capable of accurately simulating a variety of transient cyclic phenomena. Originality/value The proposed methodology improves the accuracy of cyclic plasticity phenomena and strain-life fatigue simulations for engineering applications. This study is considered a valuable contribution for the engineering community, as it can act as starting point for further exploration of the benefits that can be obtained through material parameter optimisation methodologies for models of the MAFM class.