This work investigates the applicability of three different elliptic-blending RANS-based turbulence models to predict the turbulent flow separation over a 6:1 prolate-spheroid at a length-based Reynolds number of 4, 200, 000 for 20 degrees of angle of incidence. The simulations are performed using the turbulence models with and without wall-functions within the commercial finite-volume CFD code STAR-CCM+. A comparison is drawn against the available experimental data regarding local and global variables. Furthermore, to better assess the capability of the turbulence models to reproduce correctly the physics of the flow, contours of turbulence kinetic energy and turbulence anisotropy around the body are evaluated for various models. Overall, the elliptic-blending Reynolds stress turbulence model, especially with wall-functions, by giving the most accurate predictions among all the turbulence models, is found to be successful to some extent in predicting the key flow characteristics.
展开▼