ISBN 978-3-8439-5717-5

978-3-8439-5717-5, Reihe Strömungsmechanik

Nima Fard Afshar
Analysis of Turbulence and Loss Mechanisms in Turbomachinery Cascades Using High Fidelity Simulations

188 Seiten, Dissertation Rheinisch-Westfälische Technische Hochschule Aachen (2025), Softcover, A5

Zusammenfassung / Abstract

This dissertation investigates turbulence processes and flow loss mechanisms in turbomachinery cascades using high-fidelity numerical simulations. Three-dimensional simulations of a low-pressure turbine (LPT) cascade and a high-pressure compressor (HPC) cascade are performed to identify and quantify key mechanisms driving turbulence and losses. Particular attention is given to interactions between free-stream turbulence (FST), transition, and loss generation under off-design conditions. Although full 3D domains with sidewalls are considered, the analysis focuses on mid-span behavior. Numerical results are validated against experimental blade pressure distributions and wake losses. Advanced statistical, spectral, and modal analyses characterize dominant turbulence structures. In the LPT cascade, low-frequency structures are linked to transitional and separated flow, while high-frequency components relate to turbulence during reattachment. In the HPC cascade, dominant high-frequency structures arise within the boundary layer due to strong FST–pressure-gradient interactions. Quadrant analysis shows that ejection and sweep events dominate turbulent exchange in both cascades. For the LPT cascade, higher turbulence intensity increases mixing, reduces separation length by 20%, and lowers total losses by 13%, with major losses originating in the separated shear layer and reattachment region. RANS underestimates wake losses by 32% compared to LES. In the HPC cascade, higher FST triggers earlier transition but yields similar total losses due to comparable wake mixing; main losses occur near the suction-side boundary layer, where RANS overpredicts losses by 55%. The results support validation and improvement of RANS models. By clarifying turbulence structure formation, this work advances turbomachinery flow modeling and shows that high-fidelity simulations can serve as virtual experiments to complement measurements and improve turbine and compressor performance.