Reconsideration of T-spline data models and their exchanges using STEP

• The drawbacks of the existing T-spline definitions are discussed and remedied using the redesigned T-spline data models.
• Some characteristics of the new T-spline data models are investigated and evaluated.
• The STEP-compliant data models based on the new T-spline definitions are proposed for data exchanges between different CAx systems.
• A prototype system is developed to test the T-spline data models and exchanges.

T-spline is a new approach to define freeform surfaces with relatively less control points than NURBS and is able to represent a model using a single surface without joining errors. Whereas, the complexity of T-spline data models leads numerous difficulties in its programming, which hinders the research and development of T-spline technologies. In addition, the data exchange of T-spline models still remains on a primitive level, and no standardized data format has been published so far. This article gives a reconsideration to the existing T-spline definitions, and proposes a set of redesigned data models which have much more understanding conveniences to both human and computer. Moreover, STEP-compliant data models are designed using the proposed T-spline models to standardize their data exchange between different CAx systems. The combination of T-spline with other product models in ISO 10303 makes it convenient to exchange the versatile resource data in a hybrid neutral file. A prototype system is developed for the validation purpose, which consists of a TSM-to-STEP convertor, a STEP parser and a T-spline kernel. Using the developed prototype system, one can automatically convert a Rhino system exported TSM file to a STEP file in the P21 format, which can be then parsed using the STEP reader and processed by the T-spline kernel. Some testing examples show that the proposed data models are much more efficient in processing and exchanging the T-spline data.