Abstract: The computational efficiency and real-time performance of creep-fatigue life prediction for aerospace engine turbine disks are low, making it difficult to meet the requirements of life management in practical engineering applications. Therefore, a digital twin model for predicting the creep-fatigue life of turbine disks is proposed. The model employs reduced-order model (ROM) methods to rapidly derive structural stress fields under varying operating conditions, followed by life prediction based on stress-driven fatigue-creep damage models.First, a parametric ROM is constructed to replace high-fidelity finite element analysis (FEA) for stress field computation. By projecting the full-order governing equations onto a low-dimensional subspace using techniques such as singular value decomposition (SVD), the ROM achieves a 99.74% reduction in computational time compared to conventional FEA—enabling stress field predictions within seconds under varying operational conditions. Leveraging these real-time stress results, the model realizes instantaneous prediction of creep-fatigue life across diverse operational scenarios, effectively resolving the latency issues of conventional methods.To validate the reliability of the model, six groups of creep-fatigue tests were conducted on specimens corresponding to the turbine disk bottom under varying conditions. Results demonstrate that the relative errors between predicted and experimental lifespans consistently fall within a 1.5-fold error band, confirming the model’s high prediction accuracy.The proposed framework significantly enhances computational efficiency while maintaining precision, offering a practical solution for real-time lifespan monitoring and management of turbine disks in engineering applications. This advancement bridges the gap between theoretical models and industrial requirements, providing a robust foundation for proactive maintenance strategies in aviation systems.