Abstract Computational fluid dynamics (CFD) simulations based on the Reynolds-Averaged Navier-Stokes (RANS) equations are essential for aerodynamic analysis and compressor design. However, structural uncertainties arising from the assumptions embedded in the constitutive model limit the precision of these simulations. This study proposes an improved eigenspace perturbation framework (EPF) to quantify structural uncertainties in RANS simulations of the transonic compressor NASA Rotor 67. By combining a non-uniform eigenvalue perturbation technique with relaxation regulation and a partial eigenvector perturbation method, the improved EPF enhances robustness while maintaining physical realizability. The study systematically assesses the impact of this strategy on overall performance metrics, including total pressure ratio and isentropic efficiency, as well as localized parameters like total pressure and temperature. Analysis of turbulent kinetic energy distribution, shock structure, and loss mechanisms reveals that eigenspace perturbation-induced changes in Reynolds stress significantly influence shock position, intensity, and flow structure near the blade trailing edge and channel. Moreover, the spatial variations in relative total pressure loss indicate that flow modulation patterns due to structural uncertainty differ across regions.
Wang et al. (2025) studied this question.