Theoretical And Application Study On CAD Reconstruction Of Free Surface Based On Shape Characteristic

As one of the important tools of design and manufacturing for complicated industrial products, it is very wide to attach great importance for the reconstruction of free-form surface on the engineering field of CAD/CAM. With the development of computer technology, automatic control technology and CAD/CAM technology etc., coordinate measuring machine(CMM) is also applied to reconstruction of surface because of measuring precision and intellectualized level. But CMM measuring efficiency usually becomes constraint to its wide application in the reconstruction of free-form surface. Starting with improving the measuring efficiency of CMM, the reconstruction theory of free-form surface and engineering application are profoundly studied on the basis of the shape characteristics of free-form surface in this dissertation. In addition, curves and surfaces constructed by the dissertation can be represented in the form of B SPLINE or NURBS, so it is easy to blend into current CAD/CAM systems.The dissertation puts forward and sets up the adaptive sampling model based on the concept of gravity center. On the basis of the sampling model, the model of curvedness measure is analyzed deeply, and the gauss model of curvedness measure, which represents the curved magnitude on surface, is proposed. Meantime, the accelerating generation of sampling mesh is realized by the mesh-initialized method. The research result will establish a theoretical foundation for sampling planning of free-form surface.According to the feature of contacting measurement, the dissertation first puts forward the twice-sampling planning method for free-form surface in order to improve measuring efficiency. This method makes a unknown surface become a known surface successfully by introducing a base surface. Then the automatic sampling of surface is realized, which provides reasonable information support for smooth fitting of sampling points.Surface reconstruction approaches are studied at the same time from two aspects of precision and visual effect. First, in order to obtain the fitted surface with higher precision, the least-squares technology, the randomly points parameterization technology and the constraint non-linear least-squares optimization technology are combined together originally, and precise smooth approximation is realized for sampling points. Second, by applying the base surface parameterization, the surface reconstruction model based on shape constraint is constructed from physical deformation optimization model and geometric constraints.In terms of the feature of radius compensation for coordinate measurement, the natures of offset curve and surface are studied in the dissertation. One three-dimensional compensation method on probe radium is put forward, and the three-dimensional compensation precision is probed by simulating computation. The simulating results show the compensation method is stable and reliable.According to research results above, reconstruction experiments of object have been done. By blending reconstruction surface models into current CAD/CAM system, some applying experiments are carried into execution on CAD models of object, NC tool path generation, rapid prototyping manufacturing and machining inspection. The experiments indicate the proposed reconstruction method in the dissertation is characteristics of high sampling efficiency, fitting precision and integration, so it has engineering application value.