Abstract
Abstract. The aim of this article is to introduce a new topology optimisation formulation for optimal robust design of Micro Electro Mechanical Systems. Mesh independence in topology optimisation is most often ensured by using filtering techniques, which result in transition grey regions difficult to interpret in practical realisations. This problem has been alleviated recently by projection techniques, but these destroy the mesh independence introduced by the filters and result in single node connected hinges. Such features in the design are undesirable as they are not robust with respect to geometric manufacturing errors (such as under/over etching). They can be avoided by optimising for several design realisations which take into account the possible geometry errors. The design variations are modelled with the help of random variables. The proposed stochastic formulation for the design variations results in nearly black and white mechanism designs, robust with respect to uncertainties in the production process, i.e. without any hinges or small details which can create manufacturing difficulties.
Highlights
The focus in this article is on the design of compliant mechanisms by topology optimisation
Mesh independence in topology optimisation is most often ensured by using filtering techniques, which result in transition grey regions difficult to interpret in practical realisations
Compliant mechanisms gain their mobility from the flexibility of the building components and they have found wide applicability in the production of Micro Electro Mechanical Systems (MEMS) – small mechanical devices coupled with electronic circuits
Summary
The focus in this article is on the design of compliant mechanisms by topology optimisation. Topology optimisation (Bendsøe and Sigmund, 2004) has been utilised widely in the industry for optimising machine elements and assemblies. It is an iterative process where the aim is to minimise predefined objective, such as weight, cost or compliance, by distributing material in the design domain and fulfilling prescribed constraints. In order to utilise gradient based optimisation methods, the 0/1 design problem is relaxed and the design variables are allowed to take values continuously between zero and one. The regularised topology optimisation problem results in designs with grey transition regions between the void and solid.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
