Parallel implementation and performance analysis of a linear octree finite element mesh generation scheme

摘要

We show a parallel implementation and performance analysis of a linear octree-based mesh generation scheme designed to create reasonable-quality, geometry-adapted unstructured hexahedral meshes automatically from triangulated surface models. We present algorithms for the construction, 2∶1 balancing and meshing large linear octrees on supercomputers. Our scheme uses efficient computer graphics algorithms for surface detection, allowing us to represent complex geometries. An isogranular analysis demonstrates good scalability. Our implementation is able to execute the 2∶1 balancing operations over 3.4 billion octants in less than 10 s per 1.6 million octants per CPU core.