In order to get started with Parametric Information Modelling (PIM), it is first necessary to understand some of the theory.
In order to allow for greater control, the PIM approach requires models to be defined using a topological modelling approach. Topology is a mathematical approach that defines components in terms of adjacency and connectivity. For example, a mesh may be described as being an object with a set of faces, each bounded by a set of straight edges, each defined by a pair of vertices, each of which is associated with a point in space.
Topology is useful in that it allows models to be described and analysed in more advanced ways. For example, in the case of the mesh, the faces adjacent to any particular face can be easily discovered by navigating through the topological data structure. However, the downside of topology is that it makes models heavier, due to the fact that additional data needs to be stored.
In order to ensure that models remain lightweight and easy to work with, the PIM approach defines topology implicitly. This means that geometric objects can stored using simple lightweight representations, and topology is generated on demand only when needed, in response to certain user operations.
Elements in the model consist of three types: geometric objects, geometric points, and implicit topological components.
Objects are defined through points. For example, a polyline is defined by connecting a series of points. Topological components then form a data structure that creates a well defined relationship between the objects and the points.
Semantics can be added to the model in two ways: by defining attributes and by defining groups.
Attributes are data values that can be attached to elements in the model. Attributes can be attached to all element types, including geometric objects, geometric points, and topological components. Attributes allow elements in the model to have different values. For example, creating an attributed called “colour” for faces, means that all faces in the model can be assigned a different colour.
Groups are collections of elements in the model. All elements types can be grouped, including geometric objects, geometric points, and topological components. Each group can be assigned a set of properties, as key value pairs. Groups can also be nested, thereby allowing various more complex types of data structures to be created, including hierarchies.
The use of attributes and groups allows rich 3D data sets to be created by users, suited and customised to their specific purposes.