Modification of the Anand Constitutive Model for Enhancing Predictive Accuracy of 96.5sn-3.0ag-0.5cu Inelastic Deformation Behavior

This paper presents the modification and improvement of the Anand constitutive model for accurately predicting the mechanical behavior of 96.5Sn-3.0Ag-0.5Cu (SAC305) lead-free solder. The original Anand model had limitations in capturing strain hardening significantly at low-strain applications. The modified Anand model addressed this limitation by correlating the initial value of deformation resistance (s0) and hardening coefficient (h0) as a temperature and strain rate function. The Anand and modified Anand model parameters were determined through two-stage optimization-type curve fitting using Genetic Algorithm (GA) and Nelder-Mead (NM) optimization methods. Experimental data indicated good agreement with both models. However, the modified Anand model showed a significantly lower MSE (0.2216) than the Anand model (0.3396), suggesting improved agreement with experimental data. The modified model's yield and saturation stress prediction were also enhanced, with maximum errors of 4.36% and 4.39%, respectively. The modified Anand model also exhibited higher coefficient of determination (R2) value (0.9996), surpassing the original model's R2 value of 0.9981. These results demonstrated the enhanced predictive capability of the modified model at various conditions. Thus, the modified Anand model could represent the SAC305 solder joint deformation behavior.