An automated, viscous, adaptive-Cartesian grid generation method for complex geometries is presented. The fully automated nature of the method is demonstrated by its requirement for only a single user input, the Reynolds number. Given a surface geometry, an Octree based adaptive Cartesian grid is produced automatically. A viscous-layer grid is then "inserted" between the Cartesian grid and the body surface through a projection technique. The thickness of the viscous layer grid is determined from the user-defined Reynolds number. The method completely avoids cell-cutting, and guarantees the convexity of all computational cells. The new technique is demonstrated on several test geometries, including an F-16 aircraft.
展开▼