Unstructured Grid

From XMS Wiki
Jump to navigationJump to search

What is an Unstructured Grid?

An unstructured grid (UGrid) is spatially referenced information made up of points (x, y, z locations). An unstructured grid may also have information about connections between the points. The connections between the points may define lines and faces—the faces may also be connected to define volumes. The connecting area between points are referred to as cells. Thus a UGrid is always made up of points and may also have cells.

Vocabulary

How the words "grid" and "mesh" should be used and what type of geometry each should refer to differs among professions. Some prefer to reserve the word grid for structured spatial data and the word mesh for unstructured spatial data. Traditionally the XMS programs (GMS, SMS) have followed this same approach.

Therefore, UGrids were introduced into XMS as a new type of geometry that could represent both grids and meshes (as defined by XMS).

Confusion can also exist by the words used to describe components of a grid or mesh: points vs. nodes and cells vs. elements. Traditionally in XMS, different names have been used for different modules. For example, the x, y, z locations for scatter points are referred to as points in contrast to the locations of a mesh being referred to as nodes while the locations of a grid have either been ignored or been referred to as nodes. Also, the word element has been used to describe the connections between points in a mesh and the word cell has been used to describe the connections between points in a grid.

UGrids refer to the x, y, z locations as points and connections between points are referred to as cells.

Types of UGrids

  • UGrids without cells – Spatial point data can be represented using a UGrid. This is similar functionality to the 2D Scatter Points and 3D Scatter Points in XMS. The difference between 2D and 3D only becomes important if points exist in the grid at the same x, y location and different z locations.
  • 2D UGrids – A UGrid is considered to be 2D if there are no points with same x, y coordinates and if all defined cells are 2D.
  • 3D UGrids – A UGrid is considered to be 3D if there are points with the same x, y coordinates and different z coordinates, and if all defined cells in the grid are 3D.

Constrained UGrids

XMS supports UGrids that have certain constraints. These UGrids have specific applications and tools that work with these grids. Examples of constrained UGrids are: 2D grids, 3D grids, quadtree (octtree) grids, voronoi grids (2D or 3D), all triangle cell grids, etc.

Why UGrids?

The vision for UGrids is to encompass all of the functionality present in 2D Grids, 2D Meshes, 2D Scatter Points, 3D Grids, 3D Meshes, 3D Scatter Points, and TINs without having separate modules and tools. Additionally, UGrids can also be used to represent other types of "grids" that were previously not supported in XMS such as quadtree and voronoi grids. Future users of XMS will not have to learn what differences exist between TINs, meshes, grids, and scatter or wonder why some tools or commands only operate on scatter sets but not meshes. Display options, tools, and models (simulations) will work with UGrids.

In summary, XMS is moving toward using UGrids to represent grids, meshes, and scatter. Supporting one type of geometry has the following advantages.

  • One set of tools for working with all types of grids
  • One set of display parameters for visualizing the grids
  • One type on grid to be used with the toolbox
  • Smaller code base with increased test coverage to make sure that tools and visualization function properly
  • Support for point datasets and cell datasets on one type of grid

Related Topics