2D automatic body-fitted structured mesh generation using advancing extraction method

Stretched Meshing in a 2D Computational Domain with Elbow Edges for CFD Applications

“…To tackle this complexity, we begin with a simplified domain featuring a single elbow and its boundary, resulting in six initial corners (Figure 1(a)). This geometric shape mirrors the bending domain model used by Zhang and Jia [10]. We define the computational domain by mapping physical coordinates (x, y) within the 2D plane, with the main computational domain (level 0) serving as the foundation for subsequent decomposition.…”

“…While effective, this method requires careful decomposition, which can be time-consuming and complex. Another option is the advancing extraction method (AEM) by Zhang and Jia [10], which can directly generate structured meshes for complex 2D geometries. However, this method still produces meshes with uniform element sizes within each subdomain, limiting its ability to adapt to varying flow characteristics.…”

Non-iterative generation of an optimal mesh for a blade passage using deep reinforcement learning

“…However, generating such a mesh is a challenging and time-consuming endeavor that often necessitates extensive manual intervention. Despite geometric complexities, meshes with high orthogonality and smooth resolution variation are typically required to ensure accurate simulation results [1,2]. This challenge is particularly pronounced in the CFD analysis of a blade passage in turbomachinery, where a structured mesh is preferred due to the wall-bounded nature of the flow, which exhibits specific directions [3][4][5].…”

“…To address the time-consuming and labor-intensive task of mesh generation, various mesh generators have been developed [2,[6][7][8][9][10][11][12][13]. These tools automate the process, eliminating the need for manually specifying the location of individual nodes.…”

The Coupled Volume of Fluid and Brinkman Penalization Methods for Simulation of Incompressible Multiphase Flows

“…Moreover, during the frequent reconstructions, one must avoid the appearance of geometrical singularities and small angles in order to obtain high-quality meshes. Although the automatic generation of body-fitted grids has been developed [58,59], the task remains challenging and labor-intensive [60], especially when bodies move. Such methods are also difficult to generalize to high-order schemes, so most body-fitted methods are low-order.…”