| Popis: |
The development of parallel manipulators involves new challenges related to the design of the mechanical, actuating and information-processing subsystems. In this chapter, we limit ourselves to the design of the mechanical subsystem. It typically includes a structural and a dimensional synthesis. Whereas the first one consists in finding the a priori most appropriate mechanical architecture, i.e. the types and the arrangements of the joints and the links that make up the robot, the latter deals with the determination of its dimensions in order to match the requirements of the task at hand as closely as possible. Structural synthesis may be achieved either by combining in a systematic way the different types of joints and links allowed by the task in order to obtain all possible arrangements, or by considering preexisting solutions and customizing them. Clearly, this step relies on engineers’ intuition, whereas dimensional synthesis can more easily be automated. Still, it remains a very delicate task, especially for parallel manipulators. Indeed, the performances of these manipulators heavily depend on the chosen geometry. As underlined by many authors (Gosselin, 1988; Merlet, 2006), they also possess kinematic features that vary in opposite directions when their dimensions are modified. In this chapter, we propose an approach to the optimal design of parallel manipulators that helps the designer to find the appropriate dimensions of the mechanical structure he has opted for. For the sake of clarity, we illustrate our approach by a practical example: the design of a guidance mechanism to be used in a stitching unit. This challenging task results from the continuous demand for speeding up the assembly process of reinforcement textiles needed for the manufacture of fibre composites This demand has led to an increased automation over the last decade in the textile industry. In order to reduce the process duration and to improve both the productivity and the quality of the assembly seam, robot stitching units have been introduced. Recently, we have proposed a new sewing technology in (Kordi et al., 2006). In contrast to conventional ones, all mechanical parts of the proposed sewing head are arranged only on one side of the work pieces. This enhances chances for the automation of the assembly process, as the free side can be more easily attached to manipulators. The next step is to design an appropriate manipulator that takes into consideration the peculiarities of this technology. |
