Reverse engineering of sculptured surfaces by four-axis non-contacting scanning

Abstract

Reverse engineering can be divided into Soul steps: surface data extraction; extracted data processing; surface reconstruction; and CNC part program generation. The goal of this research is to deal with data extraction fr-om complex sculptured surfaces, redundant data elimination, and surface reconstruction problems of reverse engineering. To deal with the complex-sculptured-surface data extraction problem, an efficient four-axis non-contacting surface data scanning system, which integrates a PC-based XYZ table, a laser displacement meter (LDM), and a personal computer is developed. In this fonr-axis system, an LDM an XYZ table and a PC are used as a displacement sensor, sensor carrier and central controller respectively. In extracted data processing, we apply the medium filter technique to smooth data points, and propose a simple algorithm to discard efficiently those redundant measured points according to the required degree of accuracy. While in the surface reconstruction, we first try To convert all eliminated column or row data into spline curves. Methods for 2D spline (u, v directions) curve construction are described Non-meshed 2D spline curves are then blended to a spline surface in terms of a sparse matrix data structure. Experimental results show that the proposed four-axis surface data scanning machine can be programmed to measure most of the complex sculptured surfaces which is nor possible using a three-axis machine. The proposed extracted dataprocessing method can reduce the surface reconstruction time substantially for only the cost of a little extracting lime and minor modelling errors.