Data-driven estimation of entropy production by large scale motions in an intermediate turbine duct

Chenxing Hu, Mingqiu Zheng, Ziming Yang*, Runnan Zou

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

In modern gas turbines, the physics of three-dimensional inflow for aggressive intermediate turbine ducts are strongly associated with large-scale coherent motions, which contribute to the production of irreversible aerodynamic and thermodynamic energy losses. To address this problem, an extended data-driven approach for estimating the irreversible losses through viscosity dissipation and heat transfer in turbulent flows is proposed. An aggressive intermediate turbine duct connected to a high-pressure turbine is numerically investigated using the detached-eddy simulation framework. The influence of tip leakage vortices and rotor wakes on the aerodynamics in the duct is assessed through quadruple proper orthogonal decomposition analysis. The entropy production rate corresponding to the coherent structures is then identified, and the irreversible losses in the duct subject to unsteady flow motions are quantitatively evaluated. The results indicate that the coherent structures in the intermediate turbine duct are strongly associated with the upstream conditions, including the tip leakage flow and rotor wakes. Direct viscous dissipation accounts for the majority of the irreversible losses, with coherent perturbed dissipation contributing to 7.25% and 5.54% of the total for tip gap sizes of 1.5% and 3.5% span, respectively. The streamwise development of coherent perturbations is restricted compared with that in the mean structure, and the stochastic part contributes little to the viscous dissipation. The proposed method provides a systematic procedure for estimating the irreversible energy losses in unsteady turbulent flow, and highlights the significance of unsteady aero-thermodynamics in the design of turbomachinery.

Original languageEnglish
Article number108577
JournalAerospace Science and Technology
Volume142
DOIs
Publication statusPublished - Nov 2023

Fingerprint

Dive into the research topics of 'Data-driven estimation of entropy production by large scale motions in an intermediate turbine duct'. Together they form a unique fingerprint.

Cite this