For 1D discrete profiles, the topology of the data can be determined by ordering the raw data based on the given single spatial coordinate. Linear interpolation is performed between the ordered raw data points. The data is sorted so that the order of specification is not important.
For 2D and 3D discrete profiles, a cloud of points algorithm is used to perform the interpolation. The process involves a fast lookup of the three nearest raw data points to the evaluation point, and then application of an inverse distance weighted averaging procedure. If a raw data point lies precisely at the evaluation point location, the raw data value at that point will be used.
Note: NOTE: For 2D and 3D discrete profiles that have fewer than three data points defined, only those data points will be used in the nearest point search. For profiles with two data points, the inverse distance weighted averaging procedure is still applied, although the results could appear misleading in regions away from those points. For profiles with just one data point, the exact data value from that point will be returned with no spatial variation.
During the solution process, the solver requires values at various locations. For example, at integration points, nodes and face center locations, as required by the specifics of the discretization and numerical integration process. In all instances, the required location is determined and the raw data is interpolated to that location.