In STAR-CCM+, space is divided into regions separated by boundaries.
A region is a volume in three-dimensional space, or a surface in two-dimensional space, which need not be contiguous. It is discretized by a conformal mesh consisting of connected faces, cells and vertices.
Boundaries are what surround and define a region. A boundary is a surface in three-dimensional space, or a line in two-dimensional space. The following figure shows a simple region. The line around the region can represent one or multiple boundaries. The number of boundaries around a region will depend on how the mesh was set up and on the particular problem being solved. At a minimum, a separate boundary will be required for each boundary type in the problem, but one boundary type (for example, an inlet) could consist of multiple boundaries in the simulation. Conversely, multiple physical boundaries, like the two slip walls in this problem, can be modeled as a single boundary in the simulation.
The discretization of a boundary corresponds to the surface mesh bounding the volume mesh for the region. Regions never share the same boundary; each boundary belongs to only one region, and can coincide with a boundary belonging to a neighboring region. The figure below shows an example of three separate regions with boundaries coinciding.
When the mesh is imported, each boundary gets a unique default name, though you can rename them.