Prediction of ratcheting and LCF damage using a memory surface in cyclically hardening additively manufactured AlSi10Mg with enhanced uniaxial ratcheting resistance

The fatigue life prediction of additively manufactured AlSi10Mg alloys under complex loading, particularly under conditions involving both low-cycle fatigue (LCF) and progressive strain accumulation (ratcheting), remains a significant challenge. This study presents a comprehensive experimental and theoretical investigation to address this issue. The experimental campaign included LCF and ratcheting tests performed on AlSi10Mg specimens fabricated via Selective Laser Melting using two different laser powers (300 W and 175 W). Post-mortem fractographic and tomographic analyses were conducted to identify the underlying failure mechanisms. A novel, path-dependent fatigue damage model is proposed to capture the observed behaviour. The core of the model is a dynamic memory surface in plastic strain space that governs damage accumulation. This surface evolves through both expansions, to record overloads, and contraction, to model the “fading memory” of prior load history. The framework integrates the Smith-Watson-Topper parameter for LCF damage with a new ratcheting damage formulation directly linked to the evolution of the memory surface size. A modified summation rule is introduced to account for the effect of compressive ratcheting. A robust, non-iterative calibration procedure, which decouples the initial damage from standard LCF tests, is also presented. The proposed model demonstrates significantly improved predictive accuracy for complex; non-stationary loading histories compared to both a simple linear summation rule and the original competitive model. The memory contraction mechanism is shown to be decisive for accurately predicting fatigue life under variable-amplitude loading. Furthermore, the 175 W material variant exhibits superior resistance to ratcheting due to its pronounced cyclic hardening behaviour, which represents a key finding for process optimization.