A new fast hybrid adaptive grid generation technique for arbitrary two-dimensional domains

This paper describes a new fast hybrid adaptive grid generation technique for arbitrary twodimensional domains. This technique is based on a Cartesian background grid with square elements and quadtree decomposition. A new algorithm is introduced for the distribution of boundary points based on the curvature of the domain boundaries. The quadtree decomposition is governed either by the distribution of the boundary points or by a size function when a solution-based adaptive grid is desired. The resulting grid is quad-dominant and ready for the application of finite-element, multigrid, or line-relaxation methods. All of the internal angles in the final grid have a lower bound of 45 and an upper bound of 135. Although our main interest is in grid generation for unsteady flow simulations, the technique presented in this paper can be employed in many other fields. Several application examples are provided to illustrate the main features of this new approach.