An Adaptive Grid-Based All Hexahedral Meshing Algorithm Based on 2-Refinement

Adaptive all-hexahedral meshing algorithms have many desirable features. These algorithms provide a mesh that is efficient for analysis by providing a high element density in specific locations, such as areas of high stress gradient or high curvature and reduced mesh density in other areas of less importance. In addition, inside-out hexahedral grid based schemes, using Cartesian structured grids for the base mesh, have shown great promise in accommodating automatic all-hexahedral algorithms. In these algorithms mesh refinement is generally used to capture geometric features. Unfortunately, most adaptive mesh generation algorithms employ a 3-refinement method. This method, although easy to employ, provides a mesh that is coarse in most areas and highly refined in other areas. Because a single refined hex is subdivided into 27 new hexes, regardless of the desired refinement, there is little control on mesh density. This paper will present an adaptive all-hexahedral grid-based meshing algorithm that employs a 2-refinement insertion method. 2-refinement is based on dividing a hex to be refined into eight new hexes. This allows greater control on mesh density which in turn increases computational efficiency.