Mapper

SIMULIA Co-Simulation Services provide mapping services to conservatively map fields between points, surfaces, and volumetric domains.

This page discusses:

Mapping consists of two operations:

  • a search where an association is made between members (nodes, elements, or integration points) between source and target meshes; and
  • a calculation where field values are transferred from the source mesh to the target mesh using interpolation and/or integration algorithms.

Domain Coupling

When coupling numerical domains in multiphysics and multiscale simulations, you must consider whether the physics domains are connected through discrete points, abut one another, or overlap.

Domain coupling includes the following:

Point coupling
Point coupling occurs at discrete locations associated with a body.

Coupling a physics application with a reduced-order model is an example of point coupling. In this case, the points of the reduced-order model must be co-located with the physics model nodes.

Point coupling is different from logical coupling (for example, sensors and actuators), in which case there are no coordinates associated with the sensor or actuator.

Surface coupling
Surface coupling involves distinct domains that abut one another, where the common interface is a surface.

Fluid-structure interaction is an example of surface coupling, where the fluid and structure domains are distinct, and field data is exchanged across the common interface surface.

Volume coupling
Volume coupling involves domains that partially or fully overlap.

A sequential thermal-stress simulation, where temperatures are exchanged over the volume of a body, is an example of volume coupling.

The mapper supports mapping between similar domain types (for example, point-to-point, surface-to-surface, and volume-to-volume).

Neighborhood Search

The search algorithm establishes a one-to-one association between members of the source mesh and target mesh. Members can be nodes, elements, or integration points depending on their field position. The mapping algorithm determines how this association is formed and whether members of interest are searched on the source mesh or the target mesh.

When coupling surface domains, the surface normal is used to distinguish the bottom and top faces of membrane and shell elements. The normal direction is defined by the right-hand rule of the surface facet connectivity. The search is performed such that the surface normal of the source and target regions are facing each other.

Mapping Algorithm

A mapping category defines a set of similar mapping operations performed when mapping a field from the source mesh to the target mesh. Table 1 shows the four possible mapping categories. The following factors determine the mapping category:

  • the field's physics properties (for example, intensive quantity if the property does not depend on the amount of matter, or extensive if the property depends on the amount of matter), and
  • the position where the field is evaluated.
Within a mapping category, there might be one or more algorithms.

Table 1. Mapping Categories
Category Source Field Target Field
Nodal Integration Point Nodal Integration Point
Intensive Extensive Intensive Intensive Extensive Intensive
1 √ √ √
2 √ √
3 √ √ √
4 √ √

The four mapping categories are described below. To simplify the discussion, the term "element" is used to represent an element for a volume domain and a surface facet for a surface domain.

Mapping Category 1: Interpolation based on the element's shape function

Applicable when mapping intensive nodal fields from the source mesh to either intensive nodal or integration-point fields on the target mesh. Examples include mapping nodal displacements between meshes or mapping nodal velocity on a structural mesh to the cell-center velocity of a fluid mesh.

If the source mesh is element-based, the target node or integration point is projected onto the source mesh. The search algorithm finds the closest source element and the location within the source element where the target point is projected. The target field is then interpolated using the source element's shape function. If the source mesh is discrete, a similar algorithm is employed using a point-cloud search algorithm.

Mapping Category 2: Interpolation based on the target element's shape function

Applicable when mapping extensive nodal fields to extensive fields; for example, mapping nodal forces between source and target meshes.

If the source mesh is element based, the source node is projected onto the target mesh. The search algorithm finds the closest target element and the location within the target element where the source node is projected. The source field is then distributed using the target element's shape function.

If the target mesh is discrete, a similar algorithm is employed using a point-cloud search algorithm. The point-cloud search algorithm finds the closest target nodes and forms a virtual element to distribute the field values. In this case, the total quantity is conserved across the interface; however, the distribution might be less accurate.

Mapping Category 3: Extrapolation followed by interpolation on the source element

Applicable when mapping intensive integration-point fields to either intensive nodal or integration-point fields on the target mesh. An example is mapping stresses between source and target meshes.

The source mesh must be element based. The field is first extrapolated to each element's nodes on the source mesh. The search algorithm projects the target node or integration point onto the source mesh and finds the closest source element. The target field is then interpolated using the source element's shape function.

Mapping Category 4: Integration of intensive to extensive fields

This category is applicable when mapping intensive fields to extensive fields (for example, mapping surface traction vector to nodal forces). This mapping category uses one of two available integration algorithms: the subelement or the global conservative algorithms.

The subelement algorithm provides conservative mapping with a smooth distribution. It requires that the target mesh is element based. Each target element is divided into subelements. For each subelement, this algorithm locates the source field values according to the procedure described in Mapping Category 1 and sums the values over all the subelements. The number of subelements is determined automatically to provide accurate results.

The global conservative algorithm provides conservative results over the entire region. It requires that the target mesh is element based. This algorithm is computationally less expensive than the subelements algorithm; however, the mapped field might show an uneven distribution in some cases. The global conservative algorithm makes the source field conserved by integrating the field over the source elements and then, for every source node, computes field values according to the procedure described in Mapping Category 2.

The mapping algorithm is determined automatically by the SIMULIA Co-Simulation Services based on the source and target fields. You can overwrite the default setting by providing a mapping specification in the Co-Simulation Configuration file (see Mapping Specification).