Design and analysis of hybrid mechanical systems

点击翻译按钮获取中文摘要

In many manufacturing situations, the handling of parts, or materials, is often performed by reprogrammable mechanical devices such as robotic manipulators, i.e., serial chains of links coupled by actuated joints, and thus, called serial manipulators (SMs). Alternatively, parallel manipulators (PMs) are made of two main bodies coupled by several serial chains on which only some joints are actuated, while others are left passive. Although the presence of the latter, the whole mechanism is still drivable by the actuated joints alone under some geometric conditions. The mechanical constraints coming from the passive joints provide very interesting behaviours relative to its serial counterpart, and hence, could be used in situation where the characteristic of serial ones do not satisfy the application requirements. Consequently, PMs could have a higher load-carrying capacity, a higher positioning accuracy or higher velocities, but also a smaller workspace, more problems of singularities, relative to serial ones of similar dimensions. These theoretical expectations coming from the differential calculus are all conditionals to a well understanding of the motion transmission right from the beginning of the design process and a careful control of joint fittings. Whenever a mechanism does not fall into any of these particular two categories, they are all classified by default as hybrid manipulators (HMs), which thus encompass all other possible kinematic architectures. Recently, the journal of Mechanism and Machine Theory published a review paper on the design of parallel manipulators and mentioned that “The mechanism design contains three important parts: evaluating characteristics, type synthesis and dimensional synthesis, among which type synthesis is the most original and the most inventive one”. My research program will focus on this innovative area. The main objective of this research program is to develop modelling methodologies of both topology and geometry of hybrid mechanical systems that could be used in optimization procedures to produce new designs that would have enhanced properties. The first objective is to develop the concept of topological maps that we have recently proposed. This concept has been specifically developed to solve design problems, where the mobility of the end-effector is known as a requirement, while the geometry and partially the topology are yet unknown. The second objective is to extend the kinematic analysis concepts using the topological maps, and third, to design novel architectures of parallel and hybrid manipulators. Finally, we will use our analysis tools and synthesis methodology to design novel demonstration prototypes for specific applications.