| Περίληψη: | The work presented in this dissertation comprises a second phase of research on the topic of morphing structures using shape memory alloy (SMA) actuators. In this context, a novel methodology for the design of SMA actuated morphing structures is developed and presented addressing issues associated with the actuator material and the host structure response, from concept level till prototype testing. To this end, specialty numerical tools accounting for the simulation of SMA actuators are used. In particular, design aspects related with the SMA material and geometric non-linearity, the thermomechanical coupling of SMA actuators and their thermal response are accounted for. Comparison between numerical and experimental results is conducted, considering two representative morphing components adopting different actuator configurations. Thus, the prediction accuracy of the numerical tools and the respective FE models is evaluated and the morphing capability introduced by each adopted actuator configuration is assessed. Subsequently an actual morphing structure using the most efficient actuator configuration is developed. The challenging task of the design of a morphing structure able to achieve a predefined response is handled in simulation level by taking into account the critical phenomena impacting significantly the prediction accuracy. In order to achieve an efficient design, a novel optimization method is implemented seeking for the values of the design variables which produce the desired structural response with reasonable use of materials and limited power supply, averting any failure of the structure’s components. Once the optimal design of the morphing structure has been defined, its controlled response in time domain is sought. To this end a three-term controller scheme is implemented and integrated in morphing structure FE model and lab scale prototype. Its capability to define the response of morphing structures with SMA actuators is demonstrated both through simulation and experimental characterization of a representative morphing component. Correlation of the results validate the prediction accuracy of the numerical tool implementing the three-term controller. Then a novel numerical tuning process is developed and implemented, able to successfully handle the non-linear response of the morphing structure and seeking the controller parameters values which achieve a predefined response.
Initially, a literature review of the research works and industrial achievements relative to the design of morphing structures is presented to identify the current state of the art and state the progress beyond, of this work. A comprehensive description of the SMA material behavior is then presented and the main effects associated with its response under thermal or mechanical stimuli are provided. A concise description of the SMA constitutive model used in this work follows and its predictive capabilities are summarized. Additionally, the available beam finite element used to model the SMA actuators is outlined and the main aspects considered in its formulation regarding thermomechanical coupling and inclusion of large displacements and rotations framework are presented. Subsequently, the material characterization test campaign and the extracted SMA properties required to calibrate the constitutive model are given. Furthermore the material thermal response is characterized and modeled considering two different heating mechanisms while also the electrical properties are extracted and presented.
Next, the effect of two important phenomena associated with the SMA actuator response, i.e. the geometric non-linearity and the coupled thermal and mechanical behavior of the actuators on the response of morphing structures, is quantified and evaluated. To this end two representative morphing components are considered and their response is investigated both numerically and experimentally. Their predicted response is correlated with respective experimentally acquired results and the effect of the above-mentioned phenomena on the prediction accuracy is quantified. Moreover the capability of each actuator configuration to morph the respective component is assessed. Subsequently, the development of a multi-objective optimization method for SMA-actuated morphing structures is presented. The optimal values of the design variables producing a predefined response of the morphing structure satisfying conflicting requirements and imposed constraints are sought through an iterative process. The definition of the optimization problem includes the formulation of a multi-objective function aiming to compromise the conflicting requirements from the performance of morphing structures as well as the imposition of constraints related with both the active and the passive components aiming to ensure the structural integrity. Based on the validated simulation results, a structure adopting the most efficient actuator configuration able to morph it towards a single direction is considered in order to demonstrate the capabilities and evaluate the results of the optimization process. A scheme aiming to concurrently optimize the design of the passive components of the structure and the SMA actuators is firstly developed. Then a multi-level scheme is formulated splitting the optimization into two sub-levels: one tuning the structural stiffness of the host structure and one addressing the optimization of the actuators. The required interactions between the two levels are identified and implemented while the proper modifications of the objective functions, design space, performance constraints and adopted finite element models are presented. Comparison of the outcome and the required time to converge to the optimal solution of each scheme is performed to evaluate their efficiency and quantify the time savings of the latter.
Next, a controller scheme adopting the Proportional-Integral-Derivative (PID) algorithm is developed and integrated in the morphing structure to regulate its response in the time domain. Initially the capability of the controller to prescribe the non-linear response of the morphing structure is demonstrated and the prediction accuracy of the finite element model including the controller scheme is validated against experimental results of a lab scale prototype. Afterwards, the optimized design of an actual morphing structure is considered and the respective finite element model is adopted to develop and demonstrate a numerical tuning method aiming to specify the proper controller parameters producing the desired response in the time domain. The capability of the PID control to prescribe the response of the SMA-actuated morphing structure is demonstrated and the efficiency of the developed tuning method to define the gains of the controller to obtain the desired response is proved.
Summarizing, in this dissertation the development of a procedure for the design of a morphing structure has been presented. Design aspects related to the effect of the SMA actuator’s material and geometric non-linearity as well as the thermomechanical coupling on the predicted response of a morphing structure have been successfully addressed. A novel optimization method including design variables of both the passive components and the actuators of the morphing structure has been successfully implemented to address the design of an SMA-actuated shape adaptive structure. Additionally, a multi-level optimization scheme has been proposed to produce time savings in the design process of morphing structures. Finally the design of the morphing structure is completed by integrating a controller scheme in order to prescribe its response in the time domain.
|
|---|
