Efficient cloud refinement for kinetic meshless methods

摘要

In the present work efficient cloud refinement to carry out adaptation in meshless framework is proposed. Locally refining point density in meshless adaptation modifies the connectivity of points in this region. Previously [1] the initial connectivity was ignored and quadtree based algorithms [2] generated new connectivity for every node from scratch. This is particularly inefficient and an efficient connectivity update must exploit the fact that the node distribution would be largely unaffected except in the region of adaptation. Hence connectivity modification and generation needs to be done locally, only in these regions. We present a fast, efficient algorithm which uses the existing connectivity from the initial cloud to modify or generate the connectivity over the refined cloud.