Flux-splitting finite volume method for turbine flow and heat transfer analysis

• Published in: Computational mechanics Vol. 27; no. 2; pp. 119 - 127
• Main Authors: XU, C, AMANO, R. S
• Format: Journal Article
• Language: English
• Published: Heidelberg Springer 01.02.2001; Berlin Springer Nature B.V
• Subjects: Algorithms; Applied sciences; Computational fluid dynamics; Computational methods in fluid dynamics; Computational techniques; Continuous cycle engines: steam and gas turbines, jet engines; Convergence; Discretization; Energy conservation; Engines and turbines; Exact sciences and technology; Finite volume method; Finite-difference methods; Fluid dynamics; Fluid flow; Flux; Flux difference splitting; Fundamental areas of phenomenology (including applications); Heat transfer; Mathematical analysis; Mathematical methods in physics; Mechanical engineering. Machine design; Navier-Stokes equations; Numerical analysis; Numerical methods; Other topics in fluid dynamics; Physics; Turbines; Conservation law; Three dimensional model; Numerical integration; Time domain method; Aerofoil cascade; Navier Stokes equation; Turbine; Heat flow; Finite volume method; Numerical method; Time integration; Heat transfer
• DOI: 10.1007/s004660000219
• ISSN: 0178-7675

A novel numerical method was developed to deal with the flow and heat transfer in a turbine cascade at both design and off-design conditions. The Navier–Stokes equations are discretized and integrated in a coupled manner. In the present method a time-marching scheme was employed along with the time-integration approach. The flux terms are discretized based on a cell finite volume formulation as well as a flux-difference splitting. The flux-difference splitting makes the scheme rapid convergence and the finite volume technique ensure the governing equations for the conservation of mass, momentum and energy. A hybrid difference scheme for quasi-three-dimensional procedure based on the discretized and integrated Navier–Stokes equations was incorporated in the code. The numerical method possesses the positive features of the explicit and implicit algorithms which provide a rapid convergence process and have a less stability constraint. The computed results were compared with other numerical studies and experimental data. The comparisons showed fairly good agreement with experiments.