Abstract
Rare-earth-iron magnetostrictive alloys discovered by A.E. Clark are featured by "giant" strains when excited by a magnetic field. Among these materials, terbium-dysprosium-iron alloy, often called Terfenol-D, is the most attractive one because of its high magnetostrain level (typically 1600 ppm). One of the main keys of success in the design of many applications is the control of the largest possible dynamic strain amplitude. For example with appropriate bias and prestress conditions, a dynamic peak-to-peak strain of 2440 ppm has been measured. In order to use this exciting "giant" strain in applications, it is necessary to solve one particular engineering problem: a high magnetic bias has to be applied into Terfenol-D. The chosen method to deal with such a problem consists in studying three actuators, the difference of which concerns their magnetic bias system. The first actuator is biased using a coil. The second actuator is biased using permanent magnets in a parallel configuration. The third actuator is biased using permanent magnets in a series configuration. The aim of this paper is to present the experimental results on the three actuators and to explain the particular behavior of the third actuator. Thanks to their high bias, these Terfenol-D actuators have much more output power capability (a maximum of 2.4 kW of mechanical power) than PZT actuators. These actuators could be used as components in sonar transducers, industrial actuators for machining or sealing, systems for active control of vibration or friction motors. The method used for the analysis of the 3D behavior of the third actuator is modeling taking into account piezomagnetic coupling with the ATILA finite element program. This analysis has permitted to explain why its coupling coefficient is lower than expected. A first reason is that a part of the exciting magnetic energy is stored unusefully in the magnets. A second reason comes from magneto-mechanical conversion. Due to its structure, the actuator works on a combination of the longitudinal mode and the thickness mode. The coupling coefficient of this last mode is smaller than that of the longitudinal mode. As a result of this analysis, one can list a new validation of the ATILA code, numerical values for thickness-modes coefficients and a new method of characterization from device analysis.
Get full access to this article
View all access options for this article.
