A smart structure involves distributed actuators and sensors, and one or more microprocessors that analyze the responses from the sensors and use distributed-parameter control theory to command the actuators to apply localized strains to minimize system response. A smart structure has the capability to respond to a changing external environment (such as loads or shape change) as well as to a changing internal environment (such as damage or failure). It incorporates smart actuators that allow the alteration of system characteristics (such as stiffness or damping) as well as of system response (such as strain or shape) in a controlled manner. Many types of actuators and sensors are being considered, such as piezoelectric materials, shape memory alloys, electrostrictive materials, magnetostrictive materials, electro- rheological fluids and fiber optics. These can be integrated with main load-carrying structures by surface bonding or embedding without causing any significant changes in the mass or structural stiffness of the system. Numerous applications of smart structures technology to various physical systems are evolving to actively control vibration, noise, aeroelastic stability, damping, shape and stress distribution. Applications range from space systems, fixed-wing and rotary-wing aircraft, automotive, civil structures and machine tools. Much of the early development of smart structures methodology was driven by space applications such as vibration and shape control of large flexible space structures, but now wider applications are envisaged for aeronautical and other systems. Embedded or surface-bonded smart actuators on an airplane wing or helicopter blade will induce alteration of twist/camber of airfoil (shape change), that in turn will cause variation of lift distribution and may help to control static and dynamic aeroelastic problems. Applications of smart structures technology to aerospace and other systems are expanding rapidly. Major barriers are: actuator stroke, reliable data base of smart material characteristics, non-availability of robust distributed parameter control strategies, and non-existent mathematical modeling of smart systems. The objective of this paper is to review the state-of-the-art of smart actuators and sensors and integrated systems and point out the needs for future research.

This paper presents an active moving blade tip for a smart rotor with torsional actuation via a piezo-induced bending-torsion coupled composite beam. A novel spanwise variation in the beam layup and piezoceramic element phasing is used to maximize the twist response, while minimizing the bending response. The composite beam is located lengthwise within the blade with surface bonded piezoceramic elements excited to induce spanwise and chordwise bending of the beam. The bending results in an induced twist via the structural coupling and this twist deflects the moving tip. A proof-of-concept actuator beam was developed and tested. Tip twist with essentially zero tip bending deformation was demonstrated. Two operational 1/8^th scale model rotor blades were manufactured for hover testing. In non-rotating calibration tests, at a 60 Hz actuator excitation, a peak to peak moving tip deflection of 4 deg was achieved.

The unique configuration of the rotorcraft generates problems unknown to fixed wing aircraft. These problems include high vibration and noise levels. This paper presents the development and test results of a Terfenol-D based actuator designed to operate in an individual blade control system in order to reduce vibration and noise and increase performance on Army UH- 60A helicopter. The full-scale, magnetostrictive, Terfenol-D based actuator was tested on a specially designed testbed that simulated operational conditions of a helicopter blade in the laboratory. Tests of actuator performance (strike, force moment, bandwidth, fatigue life under operational loading) were performed.

A shape memory alloy rod, if it is pre-twisted to create stress induced martensitic transformation, will exhibit a torsional shape memory effect, namely, some of the pre-twist will be recovered once the SMA rod is heated above its austenitic start temperature. If the pre- twisted SMA rod is constrained, a large recovery torque will be generated. Torsional shape memory alloy actuators have broad applications. This paper focuses on their applications of active helicopter rotor blade control. In this paper, we first review the effort of active rotor blade control, introduce the basic properties of shape memory alloy actuators, present the important characteristics of torsional SMA actuators governing their applications, and address issues involved in the use of torsional SMA actuators for automatic rotor blade tracking systems.

This paper present an experimental-theoretical study on the active tuning of a supercritical composite shaft with shape memory alloy (SMA) wires. To predict the constrained recovery stress in SMA wires, free SMA wires were tested for their thermo-mechanical characteristics, such as the stress-strain curves at different temperatures, the recovery strain at different temperatures and the change in the phase transformation temperatures at different applied stresses. These thermo-mechanical properties were used to determine the coefficients needed in the description of a typical constitutive model for SMA. The predicted value of recovery stress was experimentally correlated. A finite element analysis of a supercritical composite shaft with SMA wires inserted in embedded sleeves was conducted to predict the rotating natural frequencies at a specified recovery force. A 23% increase in the fundamental frequency of a non-rotating composite shaft with 7.4% volume fraction of SMA wires embedded is shown.

This paper presents the results of research in using piezoelectric stacks to power a trailing edge flap for use as an individual blade control device to suppress vibrations in helicopter rotor blades. A flap actuator was designed and constructed using two piezoelectric stacks arranged in series. Mechanical amplification was provided through a hinged L-arm. An analytical model was formulated to calculate the predicted flap deflection using quasi-steady aerodynamics. The model showed that the high force piezoelectric stacks were capable of maintaining flap deflection over a wide range of freestream velocities. An experimental model consisting of an 8 inch chord NACA 0012 airfoil was constructed incorporating a single 4-inch span, 1.6 inch chord trailing edge flap. The flap performance of the model was tested in no load conditions and in the wind tunnel under freestream velocities of up to 118 ft/sec. The flap deflections were lower for higher velocities and below analytical predictions.

The concept of a rotor blade with a smart flap has received considerable attention lately due to the potential for vibration suppression using individual blade control (IBC). In this paper curved polymeric piezoelectric actuators, called C-block actuators, which exhibit significant advantages over other types of actuators are proposed to drive a smart flap for IBC. The efficient implementation involves the design of both the actuators and the flap. Therefore, it is appropriate to use a formal optimization technique to address this problem. The optimization problem is complex since it includes both continuous (flap size) and discrete (number of actuators) design variables. Therefore, a newly developed hybrid optimization procedure, which can include both types of design variables, is used to maximize flap performance using the C-block actuators. Optimization results indicate that the C-block actuators provide comparable control authority without the drawbacks, such as brittleness, of conventional bimorph actuators.

This paper examines the static and dynamic behavior of initially twisted composite spars for tilt-rotor applications, as well as documenting the fabrication of five pairs of initially twisted advanced composite single-cell spars. Spars having five different material stacking sequences were fabricated for a given geometric and pretwist definition so as to determine the maximum and minimum spar untwist for a given unit load (stiffness approach) and a given maximum load (strength approach). Analytical static stiffness and strength results are presented using a detailed shell-type finite element model, along with a two-dimensional elasticity approach to show how the maximum elastic beam untwist varies as a function of the composite spar geometric parameters and material ply orientation. Experimentally measured free-free and cantilevered natural frequencies and damping levels of the spars are presented along with a correlation to a highly refined shell-type finite element model. These vibration results reveal that both ply angle orientation and initial twist significantly alter the spar natural frequencies, where the effects are largest in the higher bending and torsion modes. Damping level changes were clearly observed with ply orientation changes (i.e. increases in extension-torsion coupling produced bending mode damping increases and torsion mode damping decreases).

On high performance military aircraft, small changes in both wing twist and wing camber have the potential to provide substantial payoffs in terms of additional lift and enhanced maneuverability. To achieve the required wing shape, actuators made of smart materials are currently being studied under an ARPA/WL contract for a subscale model of a fighter aircraft. The use of the shape memory alloy TiNi for wing twist actuation was investigated using shape memory effect (SME) torque tube actuator configurations. The actuator configurations were sized to fit inside a 16% scale model of an aircraft wing and the torque's supplied to the wing were similarly calculated from full-scale requirements. The actuator systems were tested in a conventional laboratory setting. Design and calibration of the actuators for wing twist are discussed.

This paper presents an overview of several activities of the Aeroelasticity Branch at the NASA Langley Research Center in the area of applying adaptive materials and integrated systems for controlling both aircraft aeroelastic response and airfoil shape. The experimental results of four programs are discussed: the Piezoelectric Aeroelastic Response Tailoring Investigation (PARTI); the adaptive neural control of aeroelastic response (ANCAR) program; the actively controlled response of buffet affected tails (ACROBAT) program; and the Airfoil THUNDER Testing to ascertain charcteristics (ATTACH) project. The PARTI program demonstrated active flutter control and significant reductions in aeroelastic response at dynamic pressures below flutter using piezoelectric actuators. The ANCAR program seeks to demonstrate the effectiveness of using neural networks to schedule flutter suppression control laws. The ACROBAT program studied the effectiveness of a number of candidate actuators, including a rudder and piezoelectric actuators, to alleviate vertical tail buffeting. In the ATTACH project, the feasibility of using thin-layer composite-unimorph piezoelectric driver and sensor (THUNDER) wafers to control airfoil aerodynamic characteristics was investigated. Plans for future applications are also discussed.

This study outlines active flight control materials, structural arrangements, and several new active flight control methods for rotorcraft, airplanes and missiles. A system-level comparison shows that flight control actuator systems using materials like piezoceramics have approximately double the mass-specific energy and 4 to 6 times the volume specific energy of conventional actuators. New fabrication techniques centered on the principal of directional attachment allow wings and rotor blades to become twist active. Using these new methods, directionally attached piezoelectric (DAP) actuator elements were built into graphite-epoxy sandwich structures. When compared to conventionally attached piezoelectric (CAP) elements, twist deflections (important for flight control) of DAP plates were an order of magnitude greater. By using such twist-active elements in a torque-plate configuration, an active helicopter rotor was built. This Froude-scaled solid state rotor was whirl-stand tested and showed steady blade pitch deflections in excess of plus or minus 8 degrees with good correlation between theory and experiment rates up to 42 Hz (which corresponded to 2.5/rev) and no degradation in deflection as RPM was increased. DAP elements were also used in high aspect ratio subsonic and supersonic wings, demonstrating static twist deflections of plus or minus 2 degrees and plus or minus 6 degrees respectively, with good correlation between experiment and finite element results. The final section compares all-moving active stabilator structural arrangements and pitch deflections, which range up to plus or minus 12 degrees, generating lift coefficient changes in excess of plus or minus 0.8.

The Adaptive Neural Control of Aeroelastic Response (ANCAR) program is a joint research and development effort conducted by McDonnell Douglas Aerospace (MDA) and the National Aeronautics and Space Administration, Langley Research Center (NASA LaRC) under a Memorandum of Agreement (MOA). The purpose of the MOA is to cooperatively develop the smart structure technologies necessary for alleviating undesirable vibration and aeroelastic response associated with highly flexible structures. Adaptive control can reduce aeroelastic response associated with buffet and atmospheric turbulence, it can increase flutter margins, and it may be able to reduce response associated with nonlinear phenomenon like limit cycle oscillations. By reducing vibration levels and loads, aircraft structures can have lower acquisition cost, reduced maintenance, and extended lifetimes. Phase I of the ANCAR program involved development and demonstration of a neural network-based semi-adaptive flutter suppression system which used a neural network for scheduling control laws as a function of Mach number and dynamic pressure. This controller was tested along with a robust fixed-gain control law in NASA's Transonic Dynamics Tunnel (TDT) utilizing the Benchmark Active Controls Testing (BACT) wing. During Phase II, a fully adaptive on-line learning neural network control system has been developed for flutter suppression which will be tested in 1996. This paper presents the results of Phase I testing as well as the development progress of Phase II.

The Rotor Acoustic Monitoring System (RAMS) is an embedded structural health monitoring system to demonstrate the ability to detect rotor head fatigue cracks and provide early warning of propagating fatigue cracks in rotor components of Navy helicopters. The concept definition effort was performed to assess the feasibility of detecting rotor head fatigue cracks using bulk- wave wide-bandwidth acoustic emission technology. A wireless piezo-based transducer system is being designed to capture rotor fatigue data in real time and perform acoustic emission (AE) event detection, feature extraction, and classification. A flight test effort will be performed to characterize rotor acoustic background noise and flight environment characteristics. The long- term payoff of the RAMS technology includes structural integrity verification and leak detection for large industrial tanks, and nuclear plant cooling towers could be performed using the RAMS AE technology. A summary of the RAMS concept, bench-level AE fatigue testing, and results are presented.

Numerical simulation results are presented for a novel damage detection algorithm for detecting damage in smart structures. The method finds damage using a dynamically measured static flexibility matrix. Damage is found through a parametric correction of a large order static finite element model using a much smaller order measurement set. This is accomplished without any model reduction or eigenvector expansion step. The parametric correction is formulated to minimize a residual which is formed using the pseudoinverse relationship between structural flexibility and stiffness matrices. By posing the update problem in terms of measured flexibility matrices instead of experimentally determined modal data, the problems of modal correspondence, mode selection, and modal truncation are avoided. In this paper, numerical simulation results demonstrate that the technique is capable of locating and quantifying damage manifested as a change in local stiffness in the full order finite element model using a measured flexibility matrix obtained for a much smaller measurement degree- of-freedom set. This is accomplished using a non-iterative linear solution and is possible even when the damage is contained entirely in an element not touching the measurement degree-of- freedom set.

This paper presents a computer code, IDIMPACT, based on a proposed impact load identification method for identifying both the amount of the impact load and its location in composite plates from distributed build-in piezoelectric sensor measurements. The proposed method consists of a system model characterizing the dynamic response of composite plates and of a response comparator estimating the impact location and force history by comparing the sensor measurements with the simulated ones. To verify the computer code and the proposed method, instrumented hammer impact tests were conducted on a 36-inch-by-30-inch composite plate with distributed surface-mounted piezo-ceramic sensors. The computer code was implemented in a PC with data acquisition boards to digitize the sensor measurements. The code would automatically run and graphically display the predicted location and impact force-time history from a monitor. The code has reliably with high accuracy predicted the force and location of every impact event that was tested.

A high-frequency, impedance-based structural health monitoring technique developed at the Center for Intelligent Material Systems and Structures (CIMSS) has been extended to the NDE of complex precision parts. Like many other non-destructive evaluation (NDE) techniques, this method relies on the detection of change in the dynamic properties of the structure when damage occurs. This NDE technique relies on the measurement of the electrical impedance at high frequencies (100 kHz - 275 kHz) with bonded non-intrusive piezoelectric actuator/sensors. Since the electrical and mechanical impedances of the bonded actuator/sensor are directly coupled, the proposed NDE method is able to detect incipient damage in the structure. Gears were chosen as complex precision parts for the experimental procedure because of their tight tolerances, high quality, and broad use. The goal is to show that incipient damage in the gear teeth, which are an extension of the cylindrical base structure, can be monitored. The most common types of damage in gears, i.e., abrasive wear and bending fatigue, were successfully detected. The impedance measurements before and after damage were converted into a scalar damage metric, which was used to detect the presence of damage when a threshold value was exceeded. Also, quality inspection was successfully demonstrated using the impedance-based technique.

Piezoelectric actuator technology is reviewed, and the performance of current state-of-the-art piezoelectric actuator devices estimated. Comparisons between actuator configurations show a wide range of available force and displacement performance. General design guidelines for the selection and use of piezoelectric actuators for active vibration and control applications are given. Examples of state-of-the-art actuator devices are given for each actuator case.

Voltage control is by far the most common strategy for driving piezoelectric actuators, but it is, at times, handicapped by large, inefficient amplifiers and nonlinear piezoelectric actuator behavior. Optimizing voltage-feedback amplifiers for driving capacitive-type loads with amplifier schemes such as pulse-width-modulation promises to drastically reduce amplifier weight and improve efficiency. This development work is currently in progress and is discussed in this paper. Also presented are some methods that take advantage of direct charge control of piezoelectric actuators. Direct charge control removes much of the hysteresis that is inherent in voltage control so it is a likely option if a high level of positioning accuracy is needed in a given application. Unlike voltage-feedback control, charge amplifiers can accurately control a piezoactuator open loop. The main drawback to charge control is that periodic removal of charge bias is necessary to prevent actuator drift. Discussed are strategies that attempt to accomplish charge control while simultaneously avoiding actuator drift and conserving power.

The thermal response of a specific piezoceramic induced-strain actuator (the Penn State SPICES 'frame' actuator) was investigated under two conditions: (1) as a free device; and (2) embedded in a woven glass/epoxy composite panel. Actuators were driven at various combinations of electric field strengths and frequencies. Field strengths ranged from 75 kV/m to 1.5 MV/m (10 to 200 Vrms over a 135 micron thickness), while frequencies ranged from 100 Hz to 2000 Hz. The 50-ply composite panel was instrumented with thermocouples at 4 locations through the panel thickness. Temperature measurements were recorded continuously from an initial ambient isothermal state until a steady state temperature distribution was reached. Temperatures increased with frequency and field level, with heat generation roughly proportional to the frequency and to the square of the field level, consistent with a dielectric loss mechanism. The temperature rise at the actuator-composite interface, when driven at 100 Vrms and 500 Hz, was 50 degrees Celsius. The data indicate that self-heating in applications involving a combination of high field levels, high frequencies, and thick composites can result in high internal temperatures, and possibly lead to reduced performance and reliability.

As the field of smart materials and structures emerges, there is an increasing need for high- force, high-displacement actuators. An actuator architecture that shows potential in meeting this need is the C-block. C-blocks are curved piezoelectric laminated beams, poled in the radial direction, that flex when voltage is applied. One of the main advantages of C-blocks is they can be combined into distributed arrays to generate more force and/or deflection. The original C-block design was a polymeric bimorph and required the piezoelectric layers to experience both tension and compression. Unfortunately, ceramic versions of the actuator would fail as a result of the tensile stress. Therefore, the C-block design has been modified to an asymmetrically loaded unimorph. This paper introduces a ceramic C-block actuator design along with a simple model for deflection-voltage performance. To verify the model, piezoelectric polymeric and ceramic C-block prototypes were fabricated and deflection-voltage experiments were performed. These experiments demonstrate that the model predicts well the performance of asymmetric C-blocks. The model may be used to design a variety of C-block actuators, potentially including a high-force, high displacement actuator for smart structures applications.

A study of published literature and vendor information about solid-state induced-strain actuators has been undertaken to establish their dynamic-operation mechanical and electrical characteristics, and to compare their output power and energy density criteria. A theory was developed from first principles to predict the input and output power and energy amplitudes under dynamic operation including the piezo-electric counter electro-motive force and the bias- voltage effects. The effective electro-mechanical coupling coefficient, k, for an assembled active-material stack was calculated from vendor data using a remarkably simple formula. Values between 0.45 and 1.038 were found, somehow in disagreement with the generally accepted value of 0.7 for the basic material. Dynamic energy output at the optimal stiffness match condition was found to have values of up to 0.260 J for some off-the-shelf actuators. Dynamic energy density per unit volume and unit mass varied in the range 0.5 - 3.7 J/dm³ and 0.058 - 0.482 J/kg, respectively. Conversion efficiency between input electrical power and output mechanical power was found to vary widely in the range 3.74% - 32.39%. This paper presents, with priority, two remarkably simple formulae: one for calculating the effective electro-mechanical coupling coefficient, K, using standard vendor data; the other for estimating the peak reactive power for bias-voltage operation of the actuator.

Piezoelectric materials have been known to have temperature dependency regarding their basic properties, such as the dielectric constant and the piezoelectric coefficient. In this paper, this temperature dependency is investigated. The motivation of this work is linked to the impedance-based nondestructive evaluation (NDE) technique which employs piezoceramic (PZT) sensors for tracking changes in the structural impedance, by measuring the electrical impedance, to qualitatively identify damage. However, for this NDE technique to be successful in all types of environments, it must be insensitive to temperature variations. As mentioned earlier, piezoelectric materials have strong temperature dependency and a temperature compensation procedure is necessary. The approach used in this paper is empirical due to the complexity of the thermoelectromechanical constitutive modeling. Through experimental investigation, it was found that temperature will have the effect of shifting the electrical impedance magnitude of the piezoelectric sensor, while leaving the impedance phase unaffected. For a PZT PSI-5A, the variation was found to be linear in the 80 degree F to 160 degree F temperature range. To characterize the temperature effects in piezoelectric materials, a temperature coefficient which is independent of frequency has been defined. Finally, based on the defined temperature coefficient, a simple temperature compensation technique has been implemented successfully, eliminating the effects of temperature on PZT sensors while not eliminating the effects of temperature on the structure.

A smart system for monitoring the structural integrity of pressure vessels throughout service life might consist of a limited number of piezoelectric transducers permanently bonded to the vessel, together with a system for intelligent real-time data processing. In the present work it is shown that the growth of subcritical cracks can be detected using the modal properties of a vibrating cylindrical tube. Finite element analysis and experimental measurements are in agreement in showing that the growth of cracks in a variety of locations can be detected using various strategies based on modal analysis. One strategy corresponds closely with a whole- field method of detection in that measurements made at a single point are capable of revealing crack growth in a number of different locations in the cylinder.

Fiber reinforced thermoplastics offer some substantial advantages over fiber reinforced thermosetts. Besides high specific stiffness and strength and excellent resistance to impact damage, especially with respect to the manufacturing process, considerable cost reductions can be realized due to the reduction of the processing times. However high process temperatures (250 degrees Celsius to 400 degrees Celsius) can be problematic with respect to the embedding of shape memory wires as actuators. Moreover the adhesion between the sma wires and the matrix must be carefully taken into consideration. Therefore different thermoplastic matrix systems such as polyetherimid and polyamid were chosen. To get some comparative results a classical duroplastic matrix was included into the investigations. Before embedding the sma- wires were characterized with respect to their thermo-mechanical properties. Among other things differential scanning calorimetric tests were performed to determine the phase transition temperatures. Moreover the necessary current for the phase transition as a function of time was measured. To manufacture specimens by different manufacturing processes such as autoclave forming, power press, or diaphragm forming, a universal mold was designed which allows a variable pre-straining of the sma-wires. Moreover the number of embedded wires can be adapted to control the fiber volume fraction of active wires. With the manufactured specimen numerous verification tests were performed as static and dynamic deflection measurements as function of time and current. The adhesion of the embedded wires was measured by pull-out tests. As a global result it turns out that the polyamid matrix is the best host material among the compared matrix systems.

A structural health monitoring technique developed at the Center for Intelligent Material Systems and Structures (CIMSS) relies on the high frequency impedance characteristics of the structure to qualitatively detect incipient damage. An important feature of this technique is the limited PZT actuator-sensor sensing area, which has been attributed to energy dissipation through material damping and nonconservative joints. In this paper, a study of the structural damping effect at high frequency on the localized sensing region is presented. A theoretical model of the energy dissipation, using a wave propagation approach and the correspondence principle is derived to obtain the specific damping capacity. For this purpose, an infinitely long bolted beam has been simultaneously analyzed using numerical and experimental procedures to obtain a quantification of the energy dissipated through the structure. The results of this study are presented in part II of this paper.

A procedure for mapping strain measurements to nodal displacements for a 3.74 meter laboratory truss is presented and validated using experimental data. Assumptions of small displacements and a linear displacement-strain relationship were used to develop the strain- displacement mapping. Due to the assumed discrete nature of the space truss structure, the transformation depends only on kinematics and hence only geometric data is required for the mapping. The procedure is therefore valid for quasi-static deformations as well as dynamic deformations. Estimated displacements for several nodes are compared with truth measurements obtained from both a laser interferometer system and accelerometers. It is shown that the accuracy of the predicted displacement for a limited number of sensor depends not only upon the deformation state but also which degree of freedom is being estimated.

A major difficulty in designing active controllers for structural systems is the lack of mathematical models with high enough fidelity for effective control design. In order to generate models appropriate for active control design, it is often necessary to compute models directly from test data via system identification methods. Active Control eXperts (ACX), Inc. has developed two solutions to this problem. The first consists of an interactive, menu-driven program run from the Matlab computing environment. This software package, Smart ID^TM, generates state space models from transfer function data using a combination of model synthesis, non-linear curve fitting, model reduction and model augmentation techniques. The second solution, and the focus of this paper, takes the Smart ID program a step farther. For large multivariable systems, it can be tedious to construct models in the interactive fashion dictated by the structure of the Smart ID program. Furthermore, in automated system ID/control design strategies, it is vital to have a system ID program that requires no user input. ACX has developed a fully automated version of the Smart ID program. This work discusses the steering logic and functionality of the program and presents system ID results using this program on several representative systems.

Conventional non-destructive evaluation (NDE) methods are not very effective in monitoring the material conditions of advanced composite and adhesive joints. A technology that has been proposed to enhance the inspectability of advanced composites is the particle tagging technique. Two theoretical models were recently proposed to characterize the dynamic behavior of ferromagnetic and magnetostrictive tagging particles. These theoretical models concerning the development of an active tagging technique with embedded ferromagnetic and magnetostrictive particles and magnetic excitation are now experimentally verified. The experimental results of the active particle tagging shows a variation in the dynamic response of the specimens when defects and/or damage are present. The sensory signature from a tagged polymer is extracted as a result of the interaction between the embedded particles and their host matrix. A study of various types of composites and tagging particles for passive and active tagging was performed. Experimental validation of concepts for tagging of structural materials for on-site inspection prior to installation have also been explored. The on-site particle tagging inspection has been verified on laboratory specimens obtained from industry and was shown to be very efficient.

Noise transmission through a flexible kevlar-epoxy composite trim panel into an acoustic cavity or box is studied with the intent of controlling the interior sound fields. A hybrid noise attenuation technique is proposed which uses viscoelastic damping layers in the composite plate for passive attenuation of high frequency noise transmission, and uses piezo-electric patch actuators for active control in the low frequency range. An adaptive feedforward noise control strategy is applied. The passive structural damping augmentation incorporated in the composite plates is also intended to increase stability robustness of the active noise control strategy. A condenser microphone in the interior of the enclosure functions as the error sensor. Three composite plates were experimentally evaluated: one with no damping layer, the second with a 10 mil damping layer, and the third with a 15 mil damping layer. The damping layer was cocured in the kevlar-epoxy trim panels. Damping in the plates was increased from 1.6% for the plate with no damping layer, to 5.9% for the plate with a 15 mil damping layer. In experimental studies, the improved stability robustness of the controller was demonstrated by improved adaptive feedforward control algorithm convergence. A preliminary analytical model is presented that describes the dynamic behavior of a composite panel actuated by piezoelectric actuators bonded to its surface.

A new type of piezoelectric air transducer has been developed for active noise control and other air acoustics applications. The transducer is based on the composite panel structure of a bimorph-based double amplifier, that is, two parallel bimorphs or bimorph arrays with a curved cover plate as an active face attached to the top of the bimorphs. The electro- mechanical and electro-acoustic properties of the double amplifier structure and the transducer are investigated in this paper. The displacement of the cover plate of the double amplifier structure can reach millimeter scale with a relatively low driving voltage, which is more than ten times larger than the tip displacement of bimorphs. The sound pressure level (SPL) of the transducer can be larger than 90 dB (near field) in the frequency range from 50 to 1000 Hz and be larger than 80 dB (far field) from 200 Hz to 1000 Hz, with the largest value more than 130 dB (near field). Because of its light weight and panel structure, it has the potential to be used in active noise control.

Numerical studies undertaken to determine the structural intensity for active vibration of plates bonded with piezoceramic patches are presented. The formulation, which is based on finite- difference schemes, allows for thin isotropic plates with general boundary conditions. Active and reactive intensities are computed for the different cases, in which the considered plate is excited by a planar sound field and/or one or more piezoceramic patches. The numerical results indicate the importance of considering reactive intensities for active vibration control applications.

Active noise control has shown much promise in recent years as a viable and practical technique for attenuating low frequency sound levels in complex enclosures such as motor vehicles and propeller aircraft fuselages. However, the optimal placement of actuators and sensors to reduce the global interior noise levels still remains a significant design problem. In the absence of a direct analytical method for determining the optimal actuator location configurations, alternative methods have been sought, all too often defaulting to a partial or complete search of possible configurations until adequate system performance is achieved. As the search space becomes larger, difficulty arises due to an increasing number of potential location configurations, often far exceeding practical search limitations. The possibility of searching only a small portion of the total search space to find an overall optimum solution therefore becomes increasingly desirable. One method displaying significant potential in achieving this task is the genetic algorithm. Here, it is specifically developed for application to the potentially large problem of optimizing the locations of up to four vibration control actuators on a stiffened cylinder with floor structure (representative of an aircraft fuselage) to actively control harmonic interior sound levels. For the purposes of demonstration, the primary excitation was simulated by up to ten point forces acting on the cylinder boundary. The development of the genetic algorithm for optimal search performance was undertaken using a generalized multi-modal technique. The performance characteristics of various genetic algorithm operators were tested, and the optimal configuration was then applied to the stiffened cylinder with integral floor model with good results.

Experimental investigations into the control of noise in the interior of a three-dimensional enclosure with a flexible wall are presented for harmonic and multiple frequency disturbances. Lead zirconate titanate (PZT) patches mounted on the flexible wall are used as actuators and condenser microphones are used as acoustic sensors in a digital feedforward control scheme. The control system is effective in identifying the dominant resonances and achieving significant noise reduction for the dominant modes. Based on the studies on the effect of actuator redundancy on noise reduction, it is found that less noise reduction is achieved when there are more number of actuators than sensors. Modal spillover is also observed in some of the controlled cases.

This paper describes an active/passive system designed to provide a stable, isolated platform for vibration-sensitive equipment. The UltraQuiet Platform includes three subsystems: a six- strut, six-axis passive-active vibration isolation mount, a damped support bench, and specialized vibration isolation mounts which reduce transmission of narrowband vibration from individual noisy components on the quiet platform. This paper emphasizes the six-axis Stewart platform active isolation system, which uses a novel series approach to active-passive isolation within each strut. The stiff electromagnetic actuator and geophone velocity sensors in the struts are described. Control design and active vibration isolation performance are summarized. Vibration reduction of up to 20 dB was demonstrated over a 100 Hz bandwidth. Reduction in isolation performance which resulted when the system was mated with a highly flexible base structure is noted.

This paper describes the design and implementation of local feedback controllers for active vibration suppression of a laboratory truss referred to as the Naval Research Laboratory (NRL) space truss. The NRL space truss is a 3.7 meter, 12-bay aluminum laboratory truss used as a testbed to explore smart structures technologies for future Navy spacecraft missions. To conduct real-time control and data acquisition for the implementation of controllers, a digital signal processor based system is used. Two piezoceramic active struts are employed in this experimental study. Each strut is instrumented with a force transducer and a displacement sensor. Modal strain energy computed using a refined finite element model was used to select the optimum locations of the two actuators to ensure controllability of the first two structural modes. Two local feedback controllers were designed and implemented, an integral force feedback and an integral plus double-integral force feedback. The controllers were designed independently for each active strut using classical control design techniques applied to an identified model of the system dynamics. System identification results and controller design procedure are described along with closed loop test results. The test results show up to a factor of 1/110 attenuation of the truss tip motion due to sinusoidal resonant input disturbances and up to 100 times increase in damping of the lower frequency modes of the truss.

Hybrid modal/wave distributed sensors and their application to active structural control are studied for one-dimensional structures. The unique spatial filtering properties of these convolving sensors are developed illustrating their direct benefit to solving the structural control problem. This work is motivated by problems encountered with structural control design methodologies using point sensors/actuators, including spillover, model uncertainty, non-causal compensators and sensor noise. Typical noncausal sensor dynamics such as gain rolloff without phase lag, are possible in real-time because of the spatially distributed nature of the sensor. In this paper a hybrid distributed sensor scheme is employed to exploit the spatial frequency properties of distributed piezofilm (PVDF) sensors to eliminate problems with control spillover and model uncertainty. At low frequency, a modal model is assumed to be valid while at high frequency, a wave model is assumed to be valid. This modeling is incorporated into a hybrid distributed sensor design that attempts to capture the relevant dynamics for structural control at low and high frequencies. Simulated control is attempted on a simple pinned-pinned beam.

The large millimeter-wave telescope (LMT), a bi-national project between the U.S. and Mexico, is a 50 m diameter radio telescope which will operate at 1 mm wavelengths. To achieve its surface accuracy requirements, the telescope relies on an actively controlled primary surface. A telescope reflecting surface is usually subjected to slowly varying loads. However, environmental conditions, as well as those observing techniques which require high accelerations, excite vibration in the structure. In a conventional telescope, the observer must wait for the vibration to decay naturally before continuing an observation. However, for a telescope such as the LMT, the mass of the surface and the large number of actuators can be used to add damping to the support structure. This paper investigates the use of active surfaces in vibrational damping of their supporting structures. This is accomplished by using the surface sensors to determine the modes present in the vibration of backstructure and using the surface control system to damp the dominant modes. The surface figure is controlled simultaneously.

With the availability of the magnetostrictive material, Terfenol-D, in particle form a new type of composite is emerging. One or more layers of Terfenol-D layers may be embedded in conventional fiber reinforced plastic composites. Utilizing the magnetostrictive property, the Terfenol-D layers can be used to introduce actuation forces. In this paper, we present investigations on vibration suppression of simply supported and cantilevered magnetostrictive composite beams utilizing a thin embedded magnetostrictive layer to introduce actuation stresses. A velocity feedback, constant gain control law is used considering the beam as a distributed parameter system. The magnetostrictive composites can be tailored to meet a variety of requirements with regards to their vibration suppression capability, as there are a number of variables the designer can choose, such as, lay up sequence, location and thickness of the magnetostrictive layer, span length over which the coils enclose the magnetostrictive layers, coil radius and number of turns in a coil. Numerical results are presented to bring out the effect of such parameters on vibration suppression performance.

The objective of this paper is to give a detailed insight into the behavior of an open-section beam actuated by skewed piezoelements. It is shown that significant changes in torsional and bending displacements occur even for free, not directly in contact with the actuator, portions of the beam. The effect of those phenomena on the correlation between the results obtained using currently available theoretical models and experimental results is demonstrated. Methods to improve the currently used models are outlined.

This paper presents an extension of the authors' previous analyses on the one-dimensional pseudoelastic theory of shape memory alloys (SMA), in which an interaction energy has been introduced to represent the energy dissipation during the phase transformation. From the equilibrium condition of the two-phase state, we show that the partial derivative of the interaction energy with respect to the phase fraction represents the thermodynamic driving force for the phase transformation. Two functions for the interaction energy are derived from the assumptions about the driving force of the equilibrium transformation. Using this interaction energy, we also examine the stability of equilibrium state and the formation of subloops due to incomplete transformations. Analytical predictions by the proposed theory agree well qualitatively with available experimental observations.

This paper deals with the modeling of the macroscopic behavior of magnetostrictive particulate actuators (MPA). MPA is made of magnetostrictive particles dispersed in a polymeric matrix. A load line equation is developed that relates the free strain with the mechanical stress experienced by the magnetostrictive particles. The load line equation combined with the material property gives the macroscopic behavior model of MPA. Numerical results to show the effect of volume fraction of particles and Young's modulus of matrix on the overall actuating performance are presented.

Electrorheological (ER) fluid behavior has two predominant rheological characteristics. ER fluids behave like viscoelastic materials in the pre-yield region and like viscous fluids in the post-yield region. The transition between the two regions is a strong function of the electric field, strain amplitude and the frequency of excitation. A phenomenological model is introduced that combines these two dominant characteristics to describe the behavior of ER fluid dampers. Linear shear flow mechanisms in the form of mechanical analogs are chosen for each characteristic region and are combined using nonlinear weighting functions. The process of design and fabrication of an ER fluid moving electrode damper is outlined and the testing methodology discussed. Experimental data obtained from the damper is used to determine the model parameters via a system identification technique. The parameters obtained from the system identification procedure show an increasing trend as the electric field is increased. The hysteresis loops are reconstructed using the identified parameters and the results show that the model is capable of accurately capturing the linear and nonlinear behavior of ER fluid dampers.

This paper presents a proof-of-concept investigation on an active vibration control of a hybrid smart structure (HSS) consisting of a piezoelectric film actuator (PFA) and an electro- rheological fluid actuator (ERFA). Firstly, an HSS beam is constructed by inserting a starch- based electro-rheological fluid into a hollow composite beam and perfectly bonding two piezoelectric films on the upper and lower surfaces of the structure as an actuator and as a sensor, respectively. As for the PFA, a neuro-sliding mode controller (NSC) incorporating neural networks with the concept of sliding mode control is formulated. On the other hand, the control scheme for the ERFA is developed as a function of excitation frequencies on the basis of field-dependent frequency responses. An experimental implementation for the PFA and ERFA is then established to perform an active vibration control of the HSS in the transient and forced vibrations. Both the increment of damping ratios and the suppression of tip deflections are evaluated in order to demonstrate control effectiveness of the PFA, the ERFA, and the hybrid actuation. The experimental results exhibit a superior ability of the hybrid actuation system to tailor elastodynamic responses of the HSS rather than a single class of actuation system alone.

Based on previous investigations an ultrasonic traveling wave motor for a linear motion is developed. The stator consists of straight and curved beams. In order to match the impedance of the piezoceramic elements the straight beams are notched. The notches have to be designed carefully to get two orthogonal bending modes with equal eigenfrequencies. A prototype stator was manufactured. The experimental results are encouraging.

The purpose of the present paper is to develop methods allowing the design and optimization of traveling wave type ultrasonic micrometers. These methods are based on both mechanical models. The first one concerns the contact mechanism and allows us to perform the analytical calculation of the contact rate and the pressure distribution at the stator-rotor interface for a given external preload. Then, these results are included in a second model describing the driving mechanism from kinematic considerations. Both models have been validated and applied to design and manufacture micrometers fabricated from watch-making technologies. The mechanical potentialities of such piezomotors have been confirmed. Finally, a concept for batch-fabricated piezoelectric micrometers is exposed and different induced problems are discussed as perspectives.

A high force to volume ratio magnetostrictive reaction mass actuator has been designed and developed. The actuator operates as a resonant device allowing the stored strain energy to be utilized. A discussion of the design issues associated with this actuator are presented. In addition, preliminary data is presented for this actuator. This data includes a linear analysis, evidence of parameter variation, and preliminary small signal tests intended to explore this behavior.

An adaptive algorithm is proposed for the control of a large space truss structure which uses modal filters for independent modal space control and a simple neural network that provides an on-line system identification capability. The modal filters are computed off-line using measured frequency response functions and estimated pole values for the modes of interest, and provide a coordinate transformation that yields modal coordinates from physical response measurements. The time histories for the modal coordinates are then processed in real time by the neural network, which models a single degree of freedom system transfer function and provides estimates of modal parameters, namely, frequency, damping ratio and modal gain. The modal filters are used to implement independent modal space control on a 3.74 meter space truss using a single reaction-mass actuator and 32 accelerometers. The performance of the modal filter based controller is compared to that of a local rate feedback controller using the same actuator. The applicability of the adaptive filter to adaptive control is demonstrated by real time estimation of the modal parameters of the truss with and without control. Because the modal filter control gain can be adjusted to maintain a desired closed loop damping ratio, which is tracked by the adaptive filter, adaptive control of individual modes in a time-varying system is possible. The goal of this work is to field a control system which can maintain desired closed loop damping ratios for mode frequency variations as high as 10%.

This paper presents a parameter based form of a wave absorbing controller (WAC) in the time domain. The controller is applied to a flexible structure with the piezoceramic transducers bonded onto it. A distributed piezosensor gives the moment rate of the wave absorbing region as an output current. This current is then fed back to a piezoceramic actuator, resulting in a wave absorbing controller. The controller is originally designed in the frequency domain by minimization of the ratio of total output to total input wave amplitudes. Although the controller performance appears satisfactory in the frequency domain, its time domain behavior is of interest for overall performance evaluation. However, the wave absorbing method usually produces controllers which are difficult to realize in practice. In order to study the behavior of the controller and closed loop system in the time domain, and also to address these realization issues, we introduce a method of approximation of the controller enabling it to be represented in a simple form. Frequency domain comparison of system performance under the original controller and this approximation demonstrates the accuracy of the latter. The system is then studied in the time domain, showing that this novel wave absorbing controller is effective and can be easily implemented in practice. The main contribution of this paper is the introduction of a new controller for flexible structures, based on the wave absorbing method (WAM), which depends on only system parameters, i.e., E, I, (rho) , and natural frequencies. Therefore, rather than the usual controller design based on dynamic model (such as LQR) for obtaining control gains, these gains can be easily calculated from a simple formula based only on the mechanical parameters of the flexible structure.

This paper shows the design of command input shaping and smart structure actuation for vibration suppression of a single flexible link to produce a rest-to-rest, residual vibration-free maneuver. This included the development of equations of motion, numerical implementation and simulation, optimized trajectory design, system identification, and control law implementation using the University of New Mexico (UNM) single flexible link testbed. The testbed was designed, fabricated, and used to perform control law development, dynamic model identification, and sensor/actuator validation.

In recent years the authors have engaged in the development of a special technology to optimally modify a structure's physical parameters (mass, damping and stiffness) in time- dependent loading conditions. Specific efforts have been devoted to reduce structural responses under seismic loading. In this regard, there are many other research activities dealing with different devices and control algorithms generally classified in the following categories: passive control, active and hybrid control, adaptive control, semi-active control, etc. All together, this newly emerging field of structural control has generated significant momentum of development. Real-time structural parameter modification (RSPM) is a different approach from the categories mentioned above. A detailed explanation of RSPM is given in the references. In this paper, we discuss a number of fundamental issues facing the development of various structural control technologies. These issues are summarized from our development of RSPM, and are examined based on the view that structural control devices may be considered from only two distinct types: material type and mechanical type. Such a view helps to classify some advantages and disadvantages of different structural protective systems including RSPM.

An improved polarization-effect tuning technique, which suppresses the vibration-induced phase noise of quartz crystal oscillator, is presented. Unlike the earlier approach, a smart signal conditioner is used to provide varying magnitude and varying phase to the tuning signal in order to correct the nonlinear distortion of accelerometer and the phase delay of tuning mechanism. Experiment result demonstrates that the suppression of vibration noise is significantly improved.

The primary objective of this work is to develop a standard methodology for characterizing structural actuation systems intended for operation in high electrical and mechanical loading environments. The designed set of tests evaluates the performance of the active materials system under realistic operating conditions. The tests are also used to characterize piezoelectric fiber composites which have been developed as an alternative to monolithic piezoceramic wafers for structural actuation applications. The performance of this actuator system has been improved using an interdigitated electrode pattern, which orients the primary component of the electric field into the plane of the structure, enabling the use of the primary piezoelectric effect along the active fibers. One possible application of this technology is in the integral twist actuation of helicopter rotor blades for higher harmonic control. This application requires actuators which can withstand the harsh rotor blade operating environment. This includes large numbers of electrical and mechanical cycles with considerable centripetal and bending loads. The characterization tests include standard active material tests as well as application-driven tests which evaluate the performance of the actuators during simulated operation. Test results for several actuator configurations are provided, including S2 glass- reinforced and E-glass laminated actuators. The study concludes that the interdigitated electrode piezoelectric fiber composite actuator has great potential for high loading applications.

An approach based on approximate concepts for reliability-based optimal design is presented. Limit-state functions are approximated explicitly in terms of system random variables. These approximations are then used in conjunction with a shooting Monte Carlo simulation to estimate system reliability. The overall optimization process is numerically efficient since it drastically reduces the number of real analyses required to obtain a near optimal design. An example problem is presented to illustrate the ideas set forth.

When a shape memory alloy rod or tube is pre-twisted then heated, part of the initial rotation is recovered as a result of shape memory effect. If the pre-twisted SMA rod or tube is constrained from rotating, an enormous blocking torque is generated. As a shape memory alloy rod is twisted the central region remains elastic while the outer layer undergoes martensitic transformation. This central elastic region can be thought of as an internal bias spring. Because of this self-biasing mechanism, a torsional SMA actuator can also exhibit an 'intrinsic' two-way effect. Since the maximum recoverable shear strain is 70% more than the maximum recoverable tensile strain for shape memory alloys, a torsional SMA actuator can provide large recovery torque at high rotation output while maintaining excellent repeatability. This paper discusses both theoretical modeling and experimental testing of torsion SMA actuators. The behavior of the torsional SMA actuators to be addressed includes stress/strain relations, controlled recovery behavior, blocked recovery, free-recovery, and actuator repeatability.

The data of previous researches on the structure formation of electrosensitive suspensions under the action of an external electric field and on the resulting changes in the mechanical parameters of the medium (viscosity, plasticity, resilience) were used to develop a method of pendulum damping in a damping element of a seismoreceiver. The design of a seismoreceiver is described in which the medium serves both as a sensor and an output regulator. An automatic correction system ofpendulum damping in the seismoreceiver is developed on the basis of the algorithm which relates variations of the viscosity due to temperature changes and input power conditions with the electrical control signal. Keywords: seismoreceiver, electrorheological fluid, vibration, damping, viscosity

The responses of structures under seismic conditions present very high nonlinearity due to both the parameter variations of the structures and the nonlinearities of the forces induced by ground motions. On the other hand, an adaptive controller (stabilizer) can adjust itself continuously to the changes in the system. Therefore, this paper examines the applicability of the adaptive controller to the nonlinear seismic structures so as to enable us to design economic and attractive structures which can successfully withstand the earthquake loading. The approach to the desired active control employs model reference adaptive control (MRAC) technique. Initial parameters of the system, which are mass matrix, damping matrix and stiffness matrix, are identified using off-line algorithms. An adaptive controller is, then, introduced to change the system characteristics of the target structure so that the responses of the controlled structure, under seismic conditions, will follow the expected ideal responses which are the output of a reference model. The reference model is constructed according to the specifications on the performance requirements of structures under earthquake excitation. It is shown that the MRAC technique can be applied to solve the active control problem of nonlinear seismic responses of structures in an effective and efficient fashion. The controller proposed demonstrates rational self-tuning capability to the random changes of earthquake excitation as well as to the time-varying property of nonlinear seismic structures under strong- motion earthquake. The theoretical results obtained are illustrated and verified with numerical examples.

An experimental approach is used to identify Terfenol-D material properties under magnetic bias and mechanical prestress conditions typical of transducer applications for the magnetostrictive material Terfenol-D. The approach is based on low signal transduction models, a mechanical model of the test transducer, and on the theory of vector impedance and admittance analysis. The material properties being investigated, measured directly or calculated from measured quantities, are: two Young's moduli, magnetomechanical coupling factor (or 'figure of merit' of the material), linear coupling coefficient or axial strain coefficient (d₃₃) and two magnetic permeabilities. Electrical impedance and admittance and acceleration per unit current complex functions are measured at frequencies from 100 Hz to 5 kHz using a swept sine excitation at five applied magnetic field levels. Ten Terfenol-D samples are used in randomized performance tests to obtain the preliminary material property information that is presented. Of these ten rods, two were used in randomized performance repeatability tests for which data is also presented. These data sets demonstrate the dependence of material properties on applied magnetic field levels, and provide a preliminary assessment of the variability in material properties given known operating conditions.

Quantification of Terfenol-D material properties is not readily achieved through standard stress-strain test procedures due to the sensitivity of material properties to changes in magnetic and mechanical loads. A Terfenol-D transducer can be used to measure material properties under varied operating conditions. The development of the measurement and calculation procedures used to obtain material properties is based on the model of low signal, linear magnetostriction, linear transduction equations for a transducer, and a one degree of freedom mechanical model of the transducer. Vector impedance and admittance analysis is used to determine the transducer resonant, anti-resonant, and half power point frequencies. Using these frequencies, acceleration from an accelerometer mounted on the transducer, readily measurable transducer and Terfenol-D rod parameters (mass, length, etc.), and the calculation procedures described, pertinent material properties such as the Young's moduli, permeabilities, magnetomechanical coupling factor, and the linear coupling coefficient (d₃₃) can be determined. The subtleties of applying this theory to real world non-linear Terfenol-D operated in a transducer are addressed, including proper design and operation of a transducer to achieve circular Nyquist mobility loops. Experimental results from this method are compared with other research and the material property trends are found to be consistent.

The authors have developed new multidegrees of freedom (dof) piezoelectric micromotors using both stationary bending modes and traveling wave technology. These devices are respectively based on the conversion, through frictional contact, of bending vibrations sustained in one or several vibrators into rigid body displacements of a moving element. They have allowed the design and the development of a micromanipulator with five degrees of freedom composed of stationary wave type linear translators (2 dof) and traveling wave type spherical micromotors (3 dof). These two motors can be driven reversibly in both directions with linear speeds between one mm/s^-1 and few 10^-1 ms^-1. The paper shows however that micrometer steps can be activated for micropositioning applications without displacement limitations. A prototype of an autonomous mobile minirobot integrating a behavior-based intelligence is realized in order to study the behavior designing problem of a multiple robot system incorporating a behavior-based dynamic cooperation strategy for multiple behavior-based robots.

The topic of smart structure and materials is extremely broad and so encompasses an immense amount of activity. This paper presents a very restricted snapshot of this activity and highlights ongoing work in materials, components technology, software support and applications. The highlights are selected primarily on the basis of the author's personal interests and seek to convey at least some of the exciting potential offered by the engineering disciplines encompassed within the smart structures concept. Perhaps the theme which emerges is that the applications which have already made their mark have yet to fully exploit the potential offered by the contributing technologies. There is much more to come.