Towards declarative CAD modeler built on top of a CAD modeler

Abstract

Today, industrial CAD software rely on an incremental B-Rep (Boundary Representation) modeling paradigm where volume modeling is performed iteratively using planar sketched contours subjected to mainly extrusion or revolution operations. Even if CAD modelers provide operators (e.g. pad, pocket, shaft, groove, hole, fillet) to get rid of the direct use and manipulation of canonical surfaces and NURBS, working with a CAD modeler is almost procedural with a lot of intermediate operations required to obtain the desired shape of an object. Actually, all those intermediate operations are time-consuming and generate complex construction trees that are not particularly needed to describe the final shape. Moreover, using such a procedural approach, the designers have to make a mental gymnastic to break down the object body into several basic shapes linked to the different operators of the CAD software. Clearly, an approach closer to the designers’ way of thinking is missing and there is still a gap between the ideas designers have in mind and the available tools and operators used to model them. Ideally, it would be more convenient to enter a semantic description of the shape, the CAD modeler being in charge of generating it.