The presence of fillets, rounds, chamfers and grooves is typical of many real-world mechanical components. Such features introduce in the model various topological and geometrical modifications, which may alter the correct understanding of its main features. Another drawback which hinders the correct engineering evaluation of a part is that its geometrical and topological representation is not unique, since it depends either on the different procedures used to generate CAD models or on the different internal representation of the geometric kernels used by CAD packages. The overall aim of this work is to define an intermediate representation scheme, which incorporates information about minor features such as chamfers, fillets, rounds and grooves into a graph based representation and overcomes the problem of non-uniqueness in the geometric representation. The representation scheme discussed here takes as input a solid model in B-rep form, and provides a description of a part at a higher level of abstraction in comparison with the raw B-rep description. A procedure for secondary features recognition is also described in this paper. The proposed algorithm consists of two steps where both topological/geometrical properties and dimensional attributes are investigated in order to fulfill a correct recognition. The recognized secondary features are then inserted into the model as labels, which qualify the relationships between the faces of the graph representation. The resulting modeling scheme can be conveniently used for primary feature recognition as well as for other engineering analysis/simulation purposes.

A geometric primitive for CAD implementation is presented (Bezier neural network, BNN). It is specifically designed to reproduce geometric shapes with functional requirements such as aerodynamic and hydrodynamic profiles. This primitive can be useful when a known and well defined map between functional requirements and geometric data does not exist, and it have to be deduced by a physical or numerical experimental analysis. BNN gives rise to a typical CAD representation, a Bezier curve, of a functional profile, once the functional parameters are supplied. In BNN the capability of neural network to approximate very complex and non-linear function has been combined with the capability of Bezier functions to describe geometry, in a unique neural network. In this work BNN is used in the representation of aerodynamic profiles starting to their typical functional parameters: lift and drag coefficients, Reynolds number and angle of attack. BNN is tested in reproducing the wing profile of the 4-digit NACA series. The output of BNN is compared with the results of a fluid-dynamic analysis performed by commercial software.