<< Click to Display Table of Contents >> Navigation:  Problem Descriptor Reference > The Sections of a Descriptor > Boundaries > Regions

A REGION is a portion of a two-dimensional problem domain (or of the projection of a 3D problem domain), bounded by boundary paths, that encloses an area and contains a single material (but see Regions in One Dimension for exceptions).  

Each material property in the REGION has a single definition, although this definition may be arbitrarily elaborate.

A REGION may consist of many disjoint areas.

 

Example:

 

REGION 1        { an outer box }

START(0,0)

LINE TO (10,0) TO (10,10) TO (0,10) TO CLOSE

 

REGION 2        { two embedded boxes }

START(1,1)

LINE TO (2,1) TO (2,2) TO (1,2) TO CLOSE

START(5,5)

LINE TO (6,5) TO (6,6) TO (5,6) TO CLOSE

 

Overlaying regions:

 

RULE:  

REGIONS DEFINED LATER OVERLAY AND OBSCURE REGIONS DEFINED EARLIER.  

AREAS COMMON TO TWO REGIONS BECOME PART OF THE LATER DEFINED REGION.

 

So, in the example above, the two smaller boxes overlay the large box.  The material parameters assigned to the large box pertain only to the part of the large box not overlaid by the small boxes.  

 

It is customary to make the first region define the entire outer boundary of the problem domain, and then to overlay the parts of the domain which differ in parameters from this default region.  If you overlay all parts of the outer domain with subregions, then the outer region definition becomes invisible.  It may be useful to do this in some cases, since it allows a localization of boundary condition specifications.  Nevertheless, one of the subregions is superfluous, because it could be the default.