Improvement of Wall Distance Pre-Classification Overset Grid Assembly and Its Applications

Recently, we proposed an efficient pre-classification overset grid assembly (OGA) using wall distance. In contrast to conventional OGA, which first performs a donor search and then performs classification based on a specific criterion, this methodology first performs a classification based on wall distance and subsequently conducts a donor search on a reduced set of cells. Since the donor search is limited to this reduced set, the process becomes highly efficient, but the trade-off is that the exact wall distance from all surfaces must be calculated for all query points, introducing a non-negligible overhead. In our previous work, we minimized this overhead by optimizing the methodology and achieve a speed-up of one to two orders of magnitude over conventional OGA in general grid configurations. Nonetheless, we were unable to achieve satisfactory speed-up when the proximity between objects was high due to limitations in the implemented algorithm. In this study, we improved the algorithm to enhance the efficiency of pre-classification in high-proximity scenarios and also present some applications of the pre-classification OGA.