High-order implicit blending surfaces of low degree (Q1183542): Difference between revisions

The potential method for blending surfaces [cf. \textit{C. Hoffmann} and \textit{J. Hopcroft}, The potential method for blending surfaces and corners, in: G. Farin (ed.), Geometric modeling: Algorithms and new trends. SIAM, Philadelphia (1987; Zbl 0636.53002)] is based on the possibility of introducing new parameters that locally map edges and corners of intersecting surfaces onto edges and corners defined by intersecting hyperplanes in parameter space. Therefore, the blending problem is reduced to the problem of smoothing intersections of coordinate planes in Cartesian \(n\)-space. The author explicitly constructs blending surfaces for standard cases using a very restrictive notion of \(k\)-contact; the very restrictive notion used makes the explicit formulation possible. His examples blend \(k\)-smooth convex corners by surfaces of degree \(k+1\) and nonconvex corners by surfaces of degree \(2k\). The blendings constructed in this way are automatically geometrically \(k\)-smooth \{or, in mathematical language, at points of contact generate identical \(k\)-jets\}.