Data-Driven Modal Decomposition Analysis of Unsteady Flow in Multistage Turbine

Two data-driven modal analysis approaches, proper orthogonal decomposition (POD) and dynamic mode decomposition (DMD), are applied to analyze the unsteady flow obtained by solving the Reynolds-averaged Navier–Stokes (RANS) equations in a 1.5-stage axial turbine. The reduced-order reconstructed pressure, dominant mode shapes, and dynamic features in the downstream stator of the turbine are compared between POD and four DMD variants. It is found that the DMD methods based on the amplitude criterion, the Tissot criterion, and sparsity-promoting DMD (SP-DMD) achieve reconstruction accuracies comparable to those of POD, while the frequency criterion yields larger reconstruction errors. The second and third POD and DMD modes capture the dominant pressure fluctuation structures within the stator, and there is similarity between the corresponding POD and DMD spatial modes. The unsteady flow is primarily dominated by neutrally stable DMD modes, which are characterized by relatively high modal amplitudes and low frequencies corresponding to harmonics of the rotor-passing frequency. While POD provides accurate reconstruction for the original snapshots, it is not capable of identifying the fundamental dynamic components of the system.