Abstract
Reverse engineering (RE) is the process of creating engineering design model from existing components. It is widely applied in design of sculptured surfaces such as molds, airplane wings, and turbine blades. The first step of RE process is collecting data lying on surface. When using a coordinate measuring machine (CMM) with a touch trigger probe for digitizing, probe radius must be compensated according to normalvector such that accurate coordinates of the measurement data points can be obtained. However, when the measured target is a sculptured surface, the inaccurate compensation for the probe may cause errors in the generated model owing to the unknown normals of the part surface. The common solution to this problem is to use an iterative measurement method to refine the probing directions. The drawback of this approach is time consuming. This research proposes a software iterativeapproach incorporating a surface skinning method to get the accurate digitized points and generate a NURBS surface model of the part. Since only one digitization is performed, the processing time can be dramatically saved. Advantages of using CMMs include its measuring flexibility, good accuracy, and fast processing speed. When applying CMMs for automated inspection,sampling problem, probe accessibility analysis, and path planning are three important tasks. Since the regions with sharp curvature change need to be paid more attention, this research develops a quadtree average curvature method to distribute the sampling points. More points will be allocated in high curvaturedregions such that errors can be minimized. To determine the probe accessible angles of concave features, a mathematical model based on the analytical equations of the probe is developed in this research. For the path planning problem, this research develops a simulated annealing algorithm to search a best path to minimize the measurement time. Finally, the reverse engineering and automated inspection technologies are integrated together to expedite the production cycle.