Abstract
The hot deformation behavior of nickel-based superalloy 925 under different strain rates and elevated temperatures is inherently complex because of its strong dependence on strain, strain rate, and temperature. Constitutive modeling becomes essential to capture and predict this behavior accurately. In this study, two modified versions of the Kobayashi-Dodd (KD) and the Khan-Huang-Liang (KHL) models were introduced based on a detailed analysis of the alloy's hot deformation characteristics, aiming to increase their predictive capabilities. The accuracy of the modified models, along with their original versions, was rigorously evaluated via the key statistical parameters correlation coefficient (R), average absolute relative error (AARE), and root mean square error (RMSE). The findings revealed that the modified KD and KHL models agreed well with the experimental data, indicating a remarkable fit. Both modified versions demonstrated high predictive accuracy, each achieving an R-value of 0.997. The modified KHL model reported an AARE of 3.07% and an RMSE of 7.49 MPa, whereas the modified KD model achieved an AARE of 3.19% and an RMSE of 6.64 MPa. These results confirm that the improved models offer superior performance in capturing the hot flow behavior of nickel-based superalloy 925, making them valuable tools for high-temperature forming process simulations and improving the overall performance of components. Furthermore, by enabling more efficient and optimized forming processes, these models contribute to sustainable manufacturing by minimizing material waste and optimizing energy usage, supporting global sustainability initiatives in line with SDG 9, SDG 12, and SDG 13.