Development Of On-line Three-dimensional Geometric Measurement System Based On Automated Laser Scanning Technologies

In many industrial fields(for example, aerospace, ship, railways and automotive) 3D dimensional measurements of medium/large-scale objects should be easily and rapidly taken. Nowadays, the broad diversity of large-scale metrology applications(optical, mechanical, electromagnetic, etc.), involving numerous manufacturing areas, bring out the interests in designing innovative and efficient measuring systems to achieve this target. However, each of these systems is more or less adequate, depending on measuring conditions, users’ experience and skills, or other factors like time consuming, cost, dimensions, accurateness, portability, etc. Meanwhile, this ever-increasing demand for faster and more accurate 3D geometric measurements has already initiated the search for new solutions in large-scale coordinate metrology, coupled with an incessant development of the existing ones, but the need for a highly automated high-speed high-precision low-cost flexible-assembly and less-working environment constraints that would answer the above problems still persists.In this thesis large-spatial mobile coordinate measuring method, characterized by an integrated use of liberally distributed portable laser scanners, is proposed to rapidly and automatically capture the high-quality point clouds of critical medium/large features synchronously, whose 3D geometrical relationships can be real-time assessed and described too. Moreover, through the design of the developed measuring system, the idea of virtual datum is adopted to flexibly enlarge measuring range and assist on-line measurement of medium/large-scale workpiece having multiple regular geometric features.The main contributions of the dissertation are as follows:(1) The proposed measuring system uses virtual environment adopted by large-spatial mobile coordinated measuring method to provide a controllable framework for geometric assessment, thus enabling measuring range stretch actively and flexibly. Established techniques for virtual measurement datum planes have been introduced to on-line step-by-step follow with the scanning trajectory, as well as reference to primary datum. Then, according to virtual measurement datum planes, 3D geometrical relationships between different medium/large regular features are automatically assessed and accurately described using a series of intelligent routines.(2) A mathematical model of general linear scanning path planning is proposed to automatically guide and continuously adjust the movements of laser scanner in order to strategically control them for high-quality sampling of critical regular surface area. With the application of volumetric 3D decompositions method, geometric properties of previous neighboring critical point can be divided with reference to virtual measurement datum plane, and thus the relative displacements along three orthogonal directions can be determined, which direct the scanner to move to the next position. Consequently, an appropriate 3D scanning path is on-line automatically yielded step by step by an orderly collection of local connected critical points. Finally, in term of specified view angles of laser scanner, our path planning enables the position of laser scanner approximate to the stand-off distance with respect to measured area, and hence high-quality point clouds of large regular surface can be effectively captured. On the other hand, globally optimized scanning strategy is additionally introduced to fill out local missing areas.(3) In this work, in order to plan measurement strategies for measured large-scale feature in the early inspection stage, a mathematic model for describing the relationship between overall inspection costs and sample size has been proposed. According to this model, the inspection cost is then minimized by means of a suitable optimization algorithm, thus choosing an optimal sample size, which corresponds to finding an agreement between the measurement cost and the inspection error cost. The model is illustrated and validated using case studies on 3D geometric measurements of designed artifact.(4) This work aims to experimentally investigate the location accuracy of knee point or corner point of edge features using a commercial laser stripe scanner, which is common in mechanical parts. The location accuracy can be affected by the relative posture between the scanned surface and the laser scanner, described by the scan depth, scanning orientation, the incident angle and projected angle. Therefore, a series of experiments considering those factors have been implemented by the use of designed artifact. Besides, other experiments on relative moving speed and temperature have also been conducted and analyzed.In this thesis, by integrating liberally distributed portable laser scanning technologies with the state-of-the-art equipment, a rapid length measuring system for mobile and large-scale regular workpiece and an automated 3D geometrical measuring system for large-scale workpiece having multiple regular geometric features, are designed and constructed, respectively. Moreover, the novel integrally assistive software of measuring system is developed and laid out in order to build large-scale three-dimensional virtual measurement fields, therefore promoting free exchange and keeping interoperability between measurement information and design information. The experiments on several typical medium/large scale workpieces are performed in the manufacturing field. According to the actual operation results, those proposed measuring systems have the capability of simultaneously statisfying great-pace, high-precision and increasing-reliability requirements, thus effectively establishing a unified environment for the exchange of information between digital design, machining process planning and inspection.