Reconfiguring smart structures using approximate heteroclinic connections

Rights statement
Awarding institution
  • University of Strathclyde
Date of award
  • 2017
Thesis identifier
  • T14870
Person Identifier (Local)
  • 201467937
Qualification Level
Qualification Name
Department, School or Faculty
  • The term smart structures is commonly used to describe structures which have the ability to actively change their geometry or mechanical properties. Potential applications can be found in the aerospace, energy and marine sectors, e.g. use of MEMS-type devices which require frequent switching of compliant components and morphing of advanced aerofoils to generate additional lift. Traditional reconfigurable smart structures are designed with multi-stable characteristics.;In particular, such structures can use stored strain energy to enable motion from one stable position to another stable position. However, the means of reconfiguring smart structures between stable con-figurations requires the input of, and then dissipation of energy to cross the potential barrier separating the stable configurations. Therefore, the accumulated work done for frequently actuated devices in reconfiguring between stable states can be significant.;Considering reconfigurable smart structures for power and energy constrained applications, this thesis investigates a novel concept of reconfiguring smart structures between unstable states. The vision is to take advantage of modern dynamical system theory to develop entirely new devices that use the instability of mechanical systems to deliver energy-efficient shape-changing structures.;This thesis indicates that theoretically in a simple model, transitioning between unstable states (so-called heteroclinic connections) can be more energy-efficient than traditional structures which transition between stable states and so need to cross a potential barrier. However, further experimental work will be required to verify this initial finding for real engineering systems. Clearly, energy is required to stabilize the unstable configurations, but if the energy required for active control of the instability is sufficiently small, or devices need to be frequently switched between different states, this concept is likely to be of benefit.;The concept of using instability for reconfiguration is demonstrated first by controlling a mass-spring chain model through a simple cubic nonlinearity, which is sufficient to provide the required qualitative behaviour of the system. A sufficiently smooth set of functions is then used to generate a path to approximate the heteroclinic connection, which is then used as reference trajectory for reconfiguring between different unstable configurations.;Moreover, the model is extended to a smart surface as a two-dimensional spring-mass array without dissipation. It is shown that the activere configuration scheme can be used to connect equal-energy unstable (but actively controlled) configurations for the purpose of energy-efficient morphing of the smart surface. However, in consideration of the difference between the cubic and real spring model, a spring-mass model with fully geometric non-linearity is also developed to verify the possibility of using heteroclinic connections to reconfigure future real smart structures.;Furthermore, by considering a compliant mechanism, the concept of reconfiguration of a four-bar mechanism using heteroclinic connections is also investigated. Different models varying from fully rigid to purely elastic are employed to be controlled for reconfiguring between different unstable configurations. In addition, a continuous buckled beam model has been investigated with its characteristics based on the Euler-Bernoulli beam theory. An experimental beam was fabricated with shape memory alloy actuators for active control. Although the shape memory alloy was a slow response to time, it illustrates the possibility of reconfiguration of smart structures by using heteroclinic connections.;In summary, this thesis demonstrates the potential of using heteroclinic connection to reconfigure smart structures with both numerical investigation and experimental validation.This entirely new approach to smart structures offers potentially significant benefits for power and energy constrained applications which require frequent reconfiguration.
Advisor / supervisor
  • Macdonald, Malcolm
  • McInnes, Colin R.
Resource Type
Date Created
  • 2017
Former identifier
  • 9912599693102996