Validation of computations around high-lift configurations by structured- and unstructured-mesh

In this study, flow computations are performed on two-dimensional and three-dimensional high-lift configurations on multiblock structured and unstructured meshes. The objectives are to assess and improve the reliability in simulating the flow around such devices. First, two-dimensional computations on two-element and three-element airfoils are performed to investigate mesh dependency in detail. For unstructured meshes, this is carried out by using a mesh-refinement approach. The results show that sufficient mesh density away from the wing surface is required for accurate drag prediction because the flow has large circulation and wake. A mesh-refinement approach using entropy increment as the refinement indicator is also applied, and its effectiveness is validated. The method is effective in increasing the mesh points in the region of physical importance, such as wake, and the region where the numerical error is high. Three-dimensional computations on three-element trapezoidal wings with fuselage are also performed, and some modest grid studies are performed with structured meshes. Two configurations with full-span flap and part-span flap are used. Ways to improve the reliability of structured and unstructured meshes are discussed by comparing the results. The prediction of aerodynamic forces is quite reasonable even on the unstructured mesh at moderate settings of slat and flap, despite minor differences in local flow physics. However, it is also shown that the unstructured meshes should be further refined to resolve the wake near trailing edge, wing-tip vortices, and the flow separation at the juncture of the wing and body for more accurate prediction.