Understanding Data Mapping

Mapping is the process of interpolating a set of data between two separate models. For example, pressure data from a fluid domain can be mapped onto the corresponding surface of a solid domain and used as a traction load.

Mapping data is an important part of the file-based coupling method, as it allows you to use data computed in Simcenter STAR-CCM+ to specify a boundary or field condition in a third-party CAE code. Similarly, you can use data computed in a third-party code to set boundary conditions in the Simcenter STAR-CCM+ model. Typically data is mapped at an interface between two models, and this may involve interpolating data between two surfaces, a surface and a set of vertices, or between a region and a volume. Mapping presents several challenges to the user, such as how to overcome differences in mesh density or topology at the interface between the two models, or how to preserve the accuracy of the solution. These challenges, along with the type of mapping algorithms used in Simcenter STAR-CCM+ and the typical workflow, are discussed below. For details on how to carry out mapping operations, refer to the following chapters:

All of the formats in Simcenter STAR-CCM+ file-based coupling use the same core interpolation algorithms to map data from one surface to another. In the current version of Simcenter STAR-CCM+ the accuracy of the mapping techniques has been improved from zeroth-order to at least first-order. Mapping in Simcenter STAR-CCM+ is carried out by performing operations via the object tree as described below.

Mapping Operations

A key operation in coupling with imported models is to map data between Simcenter STAR-CCM+ boundaries and/or regions and imported surfaces and/or volumes. It is also possible to map data between Simcenter STAR-CCM+ boundaries and imported beam curves. The pop-up menus for imported surfaces, beams and volumes present several map actions. Each of these map actions creates a Data Mapper object that contains the mapping functionality and is listed under the Tools > Data Mappers node for you to modify or reuse. Depending on whether the map action is associated with surfaces, beams, or volumes, a Surface Data Mapper, Beam Data Mapper or Volume Data Mapper is created. For more information, see Using Data Mappers.

The map action dialog (Map Imported Model Data) allows you to choose a Simcenter STAR-CCM+ boundary/region, an imported surface/volume, and a field function to map between the source and target boundaries/regions. In addition, you can specify certain parameters to control the underlying data mapper. These options include proximity tolerance and normalcy tolerance (only for the surface data mapper). For more information, see Setting Data Mapper Properties. Your selections from the map action dialog are automatically populated in the created data mapper object, thereby allowing you to achieve the same mapping either by creating mappers as described here or by manually creating them. Moreover, in case you perform subsequent mappings between the same surface or volume pairings, the interpolation stencil will be reused, thereby optimizing the computational time.

Interpolation Techniques

The interpolation techniques are based on either the least squares method or shape function method. The choice is based on whether the source data is face-centric or node-centric. For instance, consider a conjugate heat transfer (CHT) application in which the thermal analysis of the solid is being done using ABAQUS and the flow/thermal analysis of the fluid is being performed using Simcenter STAR-CCM+. The coupling of these analyses takes place in two steps:

  1. Mapping the heat transfer coefficient and reference temperature from the Simcenter STAR-CCM+ fluid boundary to the ABAQUS solid surface.
  2. Mapping the surface temperature from the ABAQUS solid surface to the Simcenter STAR-CCM+ fluid boundary.

In Step 1, the Simcenter STAR-CCM+ boundary is the source surface and the source data is face-centric. The target points are the ABAQUS face centers. In this case, we use least squares interpolation to map boundary face-centered data to ABAQUS face centers. In Step 2, the solid surface temperatures are imported from ABAQUS. In the above-mentioned CHT case, the ABAQUS surface is the source surface and the source data resides on the nodes. The shape functions for the ABAQUS elements are used to interpolate functions to map data from the ABAQUS nodes to Simcenter STAR-CCM+ boundary face centers. The shape functions are implemented for linear/quadratic tri/quad elements only.

Meeting Challenges

One of the challenges in mapping data is to identify the matching faces or surfaces on the source and target sides. In cases of complex applications like the under-hood of an automobile, heat transfer between solid and fluid takes place across many common surfaces attached to many different parts. These surfaces and parts vary in size, shape and thickness. Depending on the analytic requirement, the parts are grouped together differently. For instance, the fluid region can group many of the exposed surfaces together in doing the CFD analysis. However, the solid region needs to treat each of the parts separately to account for different material properties and to analyze the thermal stresses in each of them. Moreover, the discrete representation of the common surfaces can be drastically different in the solid and fluid regions. The mesh topology and mesh density on the solid and fluid surfaces can also be very different. For instance, Simcenter STAR-CCM+’s mesh engine generally creates polygonal faces and cells, while most of the finite element packages operate on tri/quad elements of various orders. In addition, the CFD analysis generally requires higher mesh densities (than solids) to capture boundary layers. Depending on the quality of mesh generated, the geometric detail of common surfaces can be compromised. These challenges can be addressed by using the additional mapping options. For more information, see Nearest Neighbor Search Constraints.