Simulation, Fabrication, and Control of Biomimetic Actuator Arrays

Abstract

Ctenophores are small sea creatures that achieve propulsion using thousands of locally controlled ciliated paddles. The human heart is a flexible network of actuators coordinated with electro-chemo-mechanical waves of contraction. Stomata are self-controlling variable apertures that work together to solve the gas exchange problem for plants. It can be seen that flexible networks of actuators coordinated using distributed intelligence and sensing provide robust and adaptable performance in the face of complex and varied challenges. This thesis aims to create networks of flexible actuators with distributed sensitivity and intelligence that realise the robust and adaptable performance seen in nature. Dielectric Elastomer Actuator(s) (DEA) were identified as ideal for achieving this goal, however, their implementation into biomimetic arrays was limited by a lack of modelling tools for engineering design; the fact that DEA self-sensing has never been used to underpin a control strategy; and rigid and bulky sensor, logic, and driver circuitry. The objectives of this thesis were to overcome these three problems: Firstly, a novel low degrees-of-freedom modelling approach was developed and experimentally validated using dielectric elastomer minimum energy structures. The approach was rapid and robust and applied as a virtual experimentation tool where it reduced prototype development times by an order of magnitude. Secondly, biomimetic distributed control using capacitive self-sensing was explored and applied to a variety of DEA arrays including a ctenophore inspired mechano-sensitive artificial cilia array. Actuator coupling, multilevel control, travelling waves of actuation, and performance under partial failure were explored. Finally, Dielectric Elastomer Switch(es) (DES) were invented. DES enable distributed sensor, driver, and logic circuitry to be integrated directly into a biomimetic array and substantially reduce the need for rigid, heavy, and expensive external circuitry. DES were used to create NAND gates, smart oscillator circuits, and memory elements. These contributions have led to two journal articles, three conference papers, two provisional patent applications, and should substantially accelerate progress towards the development and control of large scale biomimetic actuator arrays. Future work involves applying modelling and control techniques to micro-scale systems and developing DES applications, materials, and fabrication techniques further.