Single crystal nickel-based superalloys are extensively used in turbine blade applications due to their superior creep resistance compared to their polycrystalline counterparts. With the high creep resistance, high cycle fatigue (HCF) and low cycle fatigue (LCF) become primary failure mechanisms for such applications. This study investigates the fatigue life prediction of CMSX-4 using a combination of crystal plasticity and lifetime assessment models. The constitutive crystal plasticity model simulates the anisotropic, rate-dependent deformation behavior of CMSX-4, while the modified Chaboche damage model is used for lifetime assessment, focusing on cleavage stresses on active slip planes to include anisotropy. Both qualitative and quantitative data obtained from HCF experiments on single crystal superalloys with notched geometry were used for validation of the model. Furthermore, artificial neural networks (ANNs) were employed to enhance the accuracy of lifetime predictions across varying temperatures by analyzing the fatigue curves obtained from the damage model. The integration of crystal plasticity, damage mechanics, and ANNs resulted in an accurate prediction of fatigue life and crack initiation points under complex loading conditions of single crystals superalloys.

Single crystal nickel-based superalloys are extensively used in turbine blade applications 1 due to their superior creep resistance compared to their polycrystalline counterparts. With the high 2 creep resistance, High Cycle Fatigue (HCF) and Low Cycle Fatigue (Single crystal nickel-based 3 superalloys are extensively used in turbine blade applications due to their superior creep resistance 4 compared to polycrystalline counterparts. As a result, fatigue becomes the primary failure mechanism 5 in such applications. Specifically, High Cycle Fatigue (HCF) and Low Cycle Fatigue (LCF) are of 6 significant concern under these conditions. This study investigates the fatigue life prediction of 7 CMSX-4 using a combined crystal plasticity and lifetime assessment model. The constitutive crystal 8 plasticity model simulates the anisotropic, rate-dependent deformation behavior of CMSX-4, while 9 a modified Chaboche damage model is used for lifetime assessment, incorporating anisotropy by 10 focusing on cleavage stresses on active slip planes. Both qualitative and quantitative data obtained 11 from HCF experiments on single crystal superalloys with notched geometry were used to validate 12 the model. Furthermore, artificial neural networks (ANNs) were employed to enhance the accuracy 13 of lifetime predictions across varying temperatures by analyzing the stress-strain data obtained from 14 simulations. The integration of crystal plasticity, damage mechanics, and ANNs resulted in accurate 15 predictions of fatigue life and crack initiation points under complex loading conditions of single 16 crystal superalloys.