Gittersorten

Gittersorten is a German term that translates roughly to “grid types” or “types of lattices.” In technical literature it is used to describe categories of spatial subdivisions or grids that are employed to partition a domain for analysis, computation, or representation. The concept applies across disciplines such as numerical simulation, computer graphics, and geographic information systems, and the exact meaning can vary by context.

A central distinction is between structured (regular) grids and unstructured (irregular) grids. Structured grids use a regular connectivity pattern and simple indexing, which simplifies numerical methods. They can be further divided into Cartesian grids, where cells are axis-aligned and of uniform shape; and curvilinear grids, where cell lines follow a curved geometry and spacing may be nonuniform. Within structured grids, there are also variations like staggered grids, where variables are defined at different nodal positions to improve stability or accuracy, and grids with nonuniform spacing that concentrate resolution where needed.

Unstructured grids do not follow a single regular pattern, allowing cells of various shapes (such as triangles or quadrilaterals in 2D, tetrahedra or hexahedra in 3D). They are well suited to complex geometries and adaptive refinement, where mesh density increases in regions of interest. Hybrid approaches combine structured and unstructured regions to balance efficiency and geometric flexibility.

Dimensionality is another axis of classification: 1D grids for line discretizations, 2D grids for surfaces, and 3D grids for volumes. Grids may also be static or adaptive, with local refinement driven by error estimates or feature complexity.

Applications include finite difference and finite element methods in engineering simulations, computational fluid dynamics, graphics rendering, and spatial data modelling. The choice of gittersorten influences numerical stability, accuracy, and computational performance.