|Abstract:|| A discretization of a computational domain is the starting point in many
different numerical applications. Anisotropic meshes allow to obtain simulation results of similar accuracy with significantly less effort.
TetGen offers an efficient triangular and tetrahedral data structure for robust and efficient local mesh operations. So far, all these features are employed in an isotropic framework. While they can be exploited to get anisotropic triangular or tetrahedral mesh. The key concept is the following: an anisotropic mesh corresponds to an isotropic one in a proper distorted space. Indeed, in this distorted space we may re-use all the routine developed for the isotropic case,but now, the resulting mesh will be anisotropic.
In this talk, we present how we have developed this idea to make TetGen an
anisotropic triangular mesh generator. More in detail, we provide two differentways used to distort the space and get anisotropic triangular meshes.