یک رویکرد یکپارچه از فرآیند بازسازی مهندسی معکوس برای اجزای پوشیده شده

The worn mechanical components/parts arrived in the remanufacturing system exhibit highly uncontrolled variabilities in failure conditions as well as structures and shape complexities. With the aid of reverse engineering (RE) technologies, a quick and accurate acquisition of the damaged areas of the worn part is attainable and thereby facilitates remanufacturing operations necessary to bring the parts back to like-new conditions. In this paper, a reverse engineering based approach is proposed to aid the remanufacturing processes of worn parts. The proposed approach integrates 3D surface data collection, nominal model reconstruction, fine registration, extraction of additive/subtractive repair, tool path generation and actual machining process, seeking to improve the reliability and efficiency of manual repair process. For nominal model reconstruction, a Prominent Cross-Section algorithm embedded with curvature constraint is proposed to automatically identify the boundary of the part's damaged area and thereby eliminate the defective point clouds from the reconstruction process. With the nominal reconstruction model and the 3D model of the worn part, a modified ICP algorithm integrating curvature and distance constraints is proposed to achieve a best-fit position of the two models by automatically identifying and eliminating the unreliable corresponding pairs through iterations. The proposed approach is demonstrated through remanufacturing of two different mechanical components and is approved to be efficient and effective.