Trends in turbomachinery turbulence treatments

• Published in: Progress in aerospace sciences Vol. 63; pp. 1 - 32
• Main Author: Tucker, P.G.
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.11.2013
• Subjects: CFD; LES; RANS; Turbomachinery; URANS; CFD; Turbomachinery; RANS; LES; URANS
• ISSN: 0376-0421; 1873-1724
• DOI: 10.1016/j.paerosci.2013.06.001

General forms of turbulence models are outlined along with their defects and palliatives for these in relation to turbomachinery. The turbulence modelling hierarchy available in turbomachinery is set out, moving from RANS (Reynolds Averaged Navier–Stokes) to the eddy resolving DNS (Direct Numerical Simulation) approach. New vistas for techniques are discussed. A modular RANS turbulence modelling strategy is outlined. Simple scaling arguments for Unsteady RANS (URANS) spectral gaps in turbomachinery are presented and the presence of such gaps shown not always to be guaranteed. The power of computers continues to steadily rise. Hence, the use of eddy resolving simulations in their various forms is expected to increase and also their use for the refinement of lower order models. Current examples for the latter are given. The use of eddy resolving simulations in the coupled and sometimes multi-physics turbomachinery environment is considered. The need for improved measurements with well defined boundary conditions that have Reynolds stress and even spectral information, at Reynolds and Mach numbers that connect with typically powerful turbomachinery systems is identified. This is necessary to refine both RANS and eddy resolving strategies. Most available ‘Best Practices’ are centred on RANS. Hence, new guidance needs to be developed for eddy resolving methods. Expert systems, based around flow taxonomies, that can assist with for example making initial grid estimates and guiding aerodynamicists through the eddy resolving simulation process are discussed. The need for more turbomachinery relevant strategies for generating turbulence inflow is identified.