Numerical simulations of diffusing S-duct and vortex-generator-jet flows

This study evaluates several turbulence models and a particular CFD algorithm with emphasis on the prediction of flows related to high-performance aft-craft engine intakes. The intake in question is a diffusing duct with an S-shaped centerline wherein vortex-generator jets are to be used for active flow control. Although the ultimate case would be to consider the instance where the vortex generator jets are installed in the S-duct, for the purposes of a controlled study, the present effort examines the plain S-duct and vortex-generator-jet flows separately. A couple of simpler flows precede the simulations of these more complex flows to establish or verify the basic parameter settings for the solution algorithm and the turbulence models.; The objective of this study is to establish whether or not the resultant combination of solution algorithm and turbulence model yields sufficiently accurate and fast calculation of diffusing-S-duct and vortex-generator-jet flows so as to allow its use in iterative design-cycle analysis of high-performance intakes. The turbulence models considered are a low-Re k-epsilon model utilising damping functions, a two layer k-epsilon model, a shear-stress-transport model, a k-o model and a one-equation eddy viscosity model. Overall, the models are noted to provide fairly similar accuracy with the exception of the low- Re k-epsilon model. The minimum requirements for spatial resolution are found to differ, and these are highlighted in the context of grid optimisation studies. The solution algorithm based on explicit pseudo-time marching with implicit residual smoothing, multigrid acceleration, and preconditioning based on the artificial compressibility concept is found to provide favourable convergence and minimal numerical dissipation in all instances, and the required parameter setting to realise such results are presented.