Effect of Mesh Shape and Turbulence Model on Aerodynamic Performance at NACA 4415

Abstract

This study uses three turbulence model variations, i.e., S - A, k - ε, and k – ω turbulence models. In addition, there are two variations of cell shape and three variations of cell number. The number of cells is 500, 5000, 50000, and 100000. Verification is carried out in the mesh refinement study and validated by aerodynamic performances. Based on the mesh refinement study, quadrilateral cells with the k - ε are in the asymptotic convergence range. Based on the Cl, it can be concluded that the quadrilateral mesh with 50000 and 100000 cells simulated using the k-ε turbulence model shows very low errors, namely 4.1151% and 3.8643%, respectively. It shows consistency based on the quadrilaterals Cd mesh data with the k-ε and k-ω turbulence models. However, k-ε shows the lowest error with the number of cells 50000 and 100000, i.e., 127.7682% and 110.4175%, respectively. However, choosing mesh 50000 cells are advisable because it only takes 23 minutes 48 seconds in computation, while mesh 100000 cells take 1 hour 17 minutes 21 seconds. Only Cm from quadrilateral mesh with the turbulence model k-ω shows consistency. An error of mesh 50000 cells is 22.0717%, and the error value for 100000 cells is 18.1630%. By considering computation time, mesh 50000 cells are preferable because it only takes 27 minutes 16 seconds, which is faster 43 minutes 14 seconds than 100000 cells.