A linear inchworm motor was developed for structural shape control applications. One motivation for this development was the desire for higher speed alternatives to shape memory alloy based devices. Features of the subject device include compactness (60 x 40 x 20 mm), large displacement range (6 mm), and large holding force capability (200 N). There are three active piezoelectric elements within the inchworm: two "clamps" and one "pusher." Large displacements are achieved by repetitively advancing and clamping the pushing element. Although each pusher step is small, on the order of 10 microns, if the step rate is high enough, substantial speeds may be obtained (8 mm/s). In the past, inchworm devices have been used primarily for precision positioning. The development of a robust clamping mechanism is essential to the attainment of high force capability, and considerable design effort focused on improving this mechanism. To guide the design, a lumped parameter model of the inchworm was developed. This model included the dynamics of the moving shaft and the frictional clamping devices, and used a variable friction coefficient. It enabled the simulation of the time response of the actuator under typical loading conditions. The effects of the step drive frequency, the pre-load applied on the clamps, and the phase shifts of the clamp signals to the main pusher signal were investigated. Using this tool, the frequency bandwidth, the optimal pre-load and phase shifts which result in maximum speed were explored. Measured rates of motion agreed well with predictions, but the measured dynamic force was lower than expected.

No abstract

Electroacoustic transducers can be divided into an active part, the driver, and a passive part, the structure. The driver ensures electromechanical transduction, while the structure performs various mechanical and acoustical functions, such as support, shock and pressure protection, and impedance transformation. For design purposes, one often needs an equivalent circuit model which gives a relationship between the acoustic characteristics of the overall device and that of its components. Transducer equivalent circuits are usually either physical or modal. Physical equivalent circuits lend themselves to the treatment of a transducer as an assembly, but in general yield frequency dependant parameters. Modal equivalent circuits are more adequate for resonant transducers, but describe transducers as a whole. This works shows how these two types of equivalent circuits can be obtained from a full finite element model by using common structural dynamics techniques: substructuring and modal expansion. It also shows that a third, hybrid type of equivalent circuit can be obtained by using a component mode synthesis technique derived from the Craig-Bampton method. This hybrid equivalent circuit combines the advantages of physical and modal equivalent circuits, enabling to express transducer modal parameters in terms of driver and structure modal parameters.

Underwater transducers with variable and controllable resonance frequency can efficiently transmit stepped or linear chirp signals of greater bandwidth by continuously matching their resonance frequency to the dominant frequency of the transmitted signal. This paper presents a technique to vary the first resonance frequency of some widely used underwater transducers: flexural piezoceramic bars and disks. Positive-bias electric fields are added to the driving ac field to generate either a tensile in-plane load in k31 coupling, or a compressive in-plane load in k33 coupling. This load modifies the flexural rigidity of the transducer, which in turn affects its resonance frequencies. Ideally, supported transducers are first analyzed, from the point of view of the electrochemical coupling coefficient at the first resonance, and the frequency shift per applied bias field. Transducers using elastic hinges to approximate simply supported boundary conditions are then examined. The response of the transducers to the combined dynamic bias field and driving ac field is simulated to look for possible dynamic effects, such as the excitation of high-frequency extension and bending modes. Finally, experimental measurements of the coupling coefficient and frequency shift per bias field of prototype transducers are presented. [Work supported by ONR.]

No abstract