Reduced order models for small-scale turbopumps

In this study, the developments of reduced order models using Proper Orthogonal Decomposition (POD) applied to CFD data from small-scale turbopumps are investingated. The capabilities of the reduced order model reconstruction are tested, demonstrating high accuracy and efficient computational performance in predicting the working conditions of new geometrical configurations for turbopumps rotating at a constant velocity. Subsequently, the developed reduced order model is utilized to establish a rapid optimization process based on low-order models, showcasing its potential for expedited design iterations and performance enhancements in turbopump applications.