University of Patras
Εφαρμογές και έλεγχος έξυπνων υλικών με έμφαση στην χρησιμοποίησή τους ως επενεργητές
Abstract
dc:descriptionSmart materials are the target of an ongoing research effort due to their increasing use as parts of commercial devices from the wider area of mechanics (medical devices, robotics, aeronautics, etc.). Standard classification metrics are their ability to produce work with respect to their geometric dimensions (external volume), the range of actuation frequencies and the energy source used for their actuation. Shape Memory Alloys (SMAs), as part of this family of materials, are selected for the actuation of devices found in the area of robotics applications. This thesis focuses on this type of materials and examines them from the scope of robotics applications.The macroscopic SMA behavior is characterized by memory effects linked to the transformation mechanism, between two phases, and to the material loading history. These phenomena are described, in micromechanics models, by the energy equilibrium that is formed by the values of the constitutive quantities (force, deformation, temperature) and the loading history. When operating under the conditions present in robotics applications, these quantities may not be directly accessible and the recorded values may contain variations (measurement noise). In these conditions the existing models are modified so that the role of the structural components, that are sensitive to noise, is suppressed.This direction has been used for two types of SMAs, distinguished by their geometric properties and hence their dynamic characteristics. The difficulty to access the material temperature is initially shown by making the comparison between the recorded values and the theoretically expected responses. The fast actuator dynamics allow for the infiltration of exogenous dynamic characteristics, in a theoretically expected static hysteretic relationship. The presence of this additional dynamics, combined with the system sensitivity to variations of the environmental conditions prohibits the use of static SMA models for the construction the control law. In this environment, the construction of local approximations to the hysteresis trajectories is proposed, as a means to produce updated system descriptions from the input - output pairs. The validity of this method is verified by incorporating a low output sampling scheme in the feedback loop, under which the local trajectory approximate is used in place of the recorded output.At a second phase the investigation focuses on SMA actuators of a helical spring shape, in an antagonistic configuration. For this type of actuators, the effect of the heat transfer dynamics is less drastic and allows the use of a complete model. An SMA model is constructed by formulating a generalized elastoplastic model through the approximation of a phenomenological model.This reformulation enables the actuators' characterization directly on the force - deformation plane and reduces the hysteresis memory length to provide the model with robustness characteristics against input variations (noise). The examination of the antagonistic operation focuses on the magnitude of the internal forces, for different relationships of state between active and passive springs. The examination of this relationship leads to the proposition of a systematic approach in the design of antagonistic devices and the selection of the appropriate actuation strategies.
Degree
thesis:*- Grantor dc:publisher
- University of Patras
- Year dc:date
- 2018
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Kolyvas, Efthymios-Iason
- Κολυβάς, Ευθύμιος-Ιάσων
Subjects
dc:subject × 10- Έξυπνα υλικά με μνήμη σχήματος
- Επενεργητές
- Αυτόματος έλεγχος
- Smart materials
- Actuators
- Automatic control
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Engineering and Technology
- Electrical Engineering, Electronic Engineering, Information Engineering
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/42713