An approach of topology optimization of multi-rigid-body mechanism

Topology synthesis of multi-rigid-body mechanisms has always been a very important stage in the mechanism design process. In most cases, the topology of the multi-rigid-body mechanism for particular task is obtained by designers' experience and ingenuity, rather than automatic approach. In this work, an approach of topology optimization of multi-rigid-body mechanisms is investigated. The core process of the approach is an automatic optimization design process. In this approach, we construct kinematics mapping from truss structures to the joint-linked mechanisms, which transforms the topology optimization problem of multi-body system into the truss structure optimization problem. We also develop a new strategy for topology optimization of statically determinate truss, the advantage of which lies in the ability dealing with statically determinate truss topology optimization problem compared to the existing methods. By automatically optimizing the topology of the truss structure, the topology of the multi-rigid body mechanism is optimized automatically, accordingly. Here, we utilize the investigated approach to design suitable layout for multi-rigid-body micro-displacement amplifying mechanisms (MMAMs) with a large amplification ratio (>50). The layout consists of not only the topology information of the mechanism, but also the dimension parameters of the mechanism. The procedure of the approach is carried out in steps, and a human-computer interaction program has been developed for it. Using the developed program, different MMAMs are achieved. Meanwhile, the direct kinematics analysis of the MMAMs is achieved automatically, the existence of dead point position in the mechanism within movement range is checked and the micro-displacement amplification ratio is calculated out. The computing results are validated by the ADAMS (R) motion simulation, which proves that the achieved MMAMs fully fulfill the functional requirement. Along with two of the achieved MMAMs, the approach is explained, its functionality is shown, its advantages, limitations, some open problems and future works are discussed. (C) 2016 Elsevier Ltd. All rights reserved.