ANSYS ICEM CFD™ Topology Library Description

Last revised on August 6 2003.

 Contents:

• Introduction
• Open, Read, Close
• Number of entities of each topological type
• Domain Information
• Face Information
• Subface Information
• Edge Entity Information
• Vertice Information
• Node Information
• Free Subface Information
• Grid Connectivity Information
• Superdomain Information
• Linear Domain Information
• Quadratic Mesh
• Error Handling
• Non-Matching Grids
• Periodic Data

The topological information characterizing structured meshes of one or more blocks is written in a file named topo_mulcad_out. This file contains information such as the number of blocks in the model, the block size, the block-to-block connectivity, etc.

The topology library topolib.a contains routines to access the information present in the topology file. To use these routines, application codes must be linked to the topology library located in the directory /$ICEM_ACN/icemcfd/output-interfaces/lib..

This chapter describes the functions of the topology library. Unless otherwise specified, the return code ier is equal to zero when the function is executed successfully. Most functions are available for both the FORTRAN and C programming interface.
 
 

Open, Read and Close
 
 

Number of entities of each topological type

Use the functions listed in the following table to request the number of entities of each topological type.
 

Domain Information
 

This function returns the type of degenerated domain, type_deg.

The types of domain degeneration are:

• 0: not degenerated
• 1: wedge with three quadrangular faces and two triangular faces
• 2: wedge with one quadrangular face and four triangular faces
• 3: pyramid with one quadrangular face and four triangular faces
• 4: wedge with two quadrangular faces and two triangular faces
• 5: pyramid with four triangular faces
• 6: two-dimensional domain

Range of a structured domain, imin(3), imax(3):
 

Edge entities of a structured domain, edgees:
 

Vertices of a structured domain, vertices:
 

        (1, 1,  1), (NI, 1,  1), (1, NJ,  1), (NI, NJ,  1),
     (1, 1, NK), (NI, 1, NK), (1, NJ, NK), (NI, NJ, NK)

Face Information

Global face number face_no of a domain face:
 

Coordinate type for a face of a domain,tcor:
 

Coordinate type for a face of a domain,tcor. This function is similar to tp_face_side except that it does not return the error code.
 

Range of a face of a structured domain,imin(3), imax(3):
 

List of subfaces composing the face of coordinate type face_type, within a domain:
 

input:  integer domain          domain number
        integer face_type       1<=face_type<=6, for face IMIN to KMAX 
output: integer nsub            number of subfaces on face
        integer subfaces(*)     array of subfaces
        integer ier             return code

Subface Information

Dimension of a subface, dimension(2):
 

Range of a subface within a structured domain, imin(3), imax(3):
 

Region of a subface within a structured domain, first(3), last(3):
 

According to the orientation definition, it is not necessary that ifirst(n)<=ilast(n).

Three points of a subface within a structured domain, p1(3), p2(3), p3(3):

The three points are transformed from the local subface coordinates (1,1), (m,1), (1,n), where m*n is the dimension of the subface.
 

The three points are transformed from the local subface coordinates (1,1), (m,1), (1,n), where m*n is the dimension of the subface. The first and last points of the subface may be found from these 3 points:

                            first = p1
                last  = p1 + (p2-p1) + (p3-p1)

Node dimension vector ndim of a subface within a domain:
 

Cell dimension vector cdim of a subface within a domain:
 

List of edge entities composing an edge of a subface:
 

The sign of the edge entity numbers in the list determine the flow direction of the edge entities within the edges. The convention for edge numbering and directions are shown below.

(... Figure missing... )

Coordinate type of a subface within a domain:
 

Coordinate type name for a subface of a domain:
 

List of domain(s) sharing a subface:
 

Boundary condition code of a subface, bc_code:
 

Edge Entity Information
 

Name of an edge entity, entname:
 

Number of an edge entity, edgee:
 

________________________

Range of an edge entity within a structured domain, imin(3), imax(3):
 

Region of an edge entity within a structured domain, ifirst(3), ilast(3):
 

Dimension vector ndim of an edge entity within a domain:
 

In C, the indices must be reversed, so the coordinate types are in ndim[2][dir], the indices of the 1^st node in ndim[0][dir] and the number of points in ndim[1][dir]. Note that only the first direction is meaningful here since edge entities are one-dimensional.

Dimension vector ndim of an edge entity within a subface:
 

( figure missing ... )

Vertices of an edge entity, vertex:
 

Start or end vertex of an edge entity, vertex:
 

List of subfaces sharing an edge entity, subfaces:
 

List of domains sharing an edge entity, domains:
 

Vertice Information

Coordinates i,j,k, of a vertex in a domain,ijk(3):
 

Description of a vertex within and edge entity, ndim(2):
 

Node Information

Global (not merged) node number of a node in a certain domain, nod_posit:
 

Free Subface Information

Number of loops, nloop:
 

Number of prescribed curves, ncurve:
 

Number of prescribed points, npoint:
 

Number of edge entities for the outer loop of a free subface, nedgee:
 

Edge entity number for an edge of the outer loop of a free subface, edgee:
 

Array of loop numbers for a free subface, loops:
 

Free subfaces data:
 

        info_no                 data 
           1                                number of loops in the free subface 
                    2                                number of prescribed curves in the free subface
                    3                                number of prescribed points in the free subface

Array of loop status for a free subface, status:
 

Array of prescribed curve numbers for a free subface, curve:
 

Array of prescribed point coordinates for a free subface, coord:
 

Number of shared subfaces (block interfaces) in the entire mesh:
 

List of shared subfaces (block interfaces) in the entire mesh:
 

Number of subface(s) of a domain shared with other domain(s), nadsub:
 

Number of edge entity(ies) of a domain shared with other domain(s), nedge:
 

Sum of subface(s) and edge entity(ies) of a domain shared with other domain(s), ndomad:
 
 

Face and subface numbers of block interfaces of a given domain, domface, subface: The arrays are allocated to nadsub, the number of interfaces for this domain. The value of nadsub is found using the function tp_ndomad.
 

One-to-one face connectivity of a domain: All arrays are allocated to nadsub, the number of interfaces for this domain. The value of nadsub is found using function tp_ndomad.
 

Connectivity information, domad_record(5): For each domain the connectivity information is stored in the array domad(5,ndomad), where ndomad is the return value of the function tp_ndom_nsub_ad(domain). The following function returns the connectivity information for one interface of a domain:
 

The meaning of domad_record(5) is different for face connectivity and edge connectivity:
 

One-to-one edge entity connectivity of a domain: All arrays are allocated to nedgee, the number of shared edge entities for this domain. The value of nedgee is found using function tp_nsubad.
 

Superdomain Information

Global coordinates of a 3D domain within a 3D superdomain:
 

Global coordinates of a 2D domain within a 2D superdomain:
 

Number of edge entities in the linear domain, nedgee:
 

List of edge entities in the linear domain, edgees:
 

List of edge entities in the linear domain, vertices:
 

Update the topology information (in memory) to be consistent with a quadratic mesh:
 

C: (void)tp_quad_elements();

FORTRAN: call topquad()

The topology file is always written for a linear mesh even if the mesh is saved with the option "quadratic". For a linear structured mesh of NI*NJ*NK nodes, the corresponding quadratic mesh has (2*NI -1) * (2*NJ -1) * (2*NK -1) nodes. However, the distinction between linear and quadratic is not included in the topology file. As a result, functions such as tp_domain_rangereturn the size of the linear grid, even if the mesh is quadratic.

This problem is easily corrected by callingtp_quad_elements, immediately after reading the topology file ( tp_read). The function tp_quad_elementsupdates the topology information (the in-memory copy) to reflect the higher order elements. If the function tp_domain_range is called after updating the topology, the correct range of (2*NI-1), (2*NJ-1), (2*NK-1) will be found.
 

When the topology library detects an error, it records the error message and returns an error code in the form of an integer smaller than zero. Only the last detected error message is recorded. To obtain the error message, the C function tp_error_print()may be called. It prints the error message to the standard output (stdout). Similarly, the functiontp_error_exit() prints the error message to the standard output and exit the program.

The error handling in the fortran interface is a bit different than for the C interface. In the Fortran interface the topology library returns and error code integer smaller than zero, and automatically prints the error message to the standard output.
 

Non-Matching Grids

Group number of a topological entity, group_number:
 

List of twin edge entities of an edge,edge_twins: The twin edge entities are defined as edge entities having the same entity name and the same geometry, but different edge entity number and possibly different number of  nodes. Twin edges belong to different entity groups.
 

List of twin subfaces of a subface, sf_twins: The twin subfaces are defined as subfaces with the same geometrical edges but belonging to different entity groups. Twin subfaces have the same intrinsic orientation.
 

Periodic Data

Number of periodic pairs of a topological entity type, npair:
 

Lists of periodic data of a topological entity type, list1, list2: