The structural and aerodynamic characteristics of a new class of active flight control surface are presented. This new type of surface uses a symmetric, subsonic aerodynamic shell which is supported at the quarter-chord by a main spar and actively pitched by an adaptive torque- plate. The structural mechanics of the torque-plate and several actuator elements are detailed, including newly invented interdigitated electrode and constrained directionally attached piezoelectric elements. Laminated plate models demonstrate that both generate similar deflections with comparable torsional stiffnesses. An experimental torque-plate specimen constructed from PSI-5A-S2 piezoceramic shows high torsional deflections and stiffness as well as excellent correlation with theory. The constrained torque-plate was integrated into a 12.5 cm plan X 5 cm chord adaptive missile fan which was designed for Mach 0.6 flight under standard conditions. The specimen showed static pitch deflections up to +/- 8.1 degree(s) and dynamic deflections of +/- 19 degree(s) at resonance. The active surface was also wind tunnel tested up to 40 m/s and demonstrated invariant pitch deflections as a function of airspeed, a steady break frequency of 50 Hz, no flutter, buffet or divergence tendencies and steady lift coefficient changes up to +/- 0.51.

A servoflap that uses a piezoelectric bender to deflect a trailing edge flap for use on a helicopter rotor blade was designed, built, and tested. This servoflap design is an improvement over a design developed previously at MIT. The design utilizes a new flexure mechanism to connect the piezoelectric bender to the control surface. The efficiency of the bender was improved by tapering its thickness with length. Also, the authority of the actuator was increased by implementing a nonlinear circuit to control the applied electric field, allowing a greater range of actuator voltages. Experiments were carried out on a bench test article to determine the frequency response of the actuator, as well as hinge moment and displacement capabilities. Flap detections of 11.5 deg or more were demonstrated while operating under no load conditions at frequencies up to 100 Hz. The data indicate that if properly scaled, the actuator will produce flap deflections greater than 5 deg at the 90% span location on a full- scale helicopter. In addition, the first mode of the actuator was at 7/rev frequency of the target model rotor. Proper inertial scaling of this actuator could raise this modal frequency to greater than 10/rev on an operational helicopter, which is adequate for most rotor control purposes. A linear state space model of the actuator was derived. Comparisons of this model with the experimental data highlighted a number of mild nonlinearities in the actuator's response. However, the agreement between the experiment and analysis indicate that the model is a valid tool for predicting actuator performance.

The lateral buckling characteristics of flexible composite beams are actively controlled by activating sets of shape memory alloy (NITINOL) fibers which are embedded along the mid- plane of these beams. With such active control capabilities, the beams can be manufactured from light weight sections without compromising their elastic stability. This feature will be invaluable in building light weight structures that have high resistance to failure due to buckling. Finite element models are developed to analyze the mechanical and thermal characteristics of the composite beams. The models account for the individual contributions of the matrix, the NITINOL fibers and the shape memory effect to the overall performance of the composite beams. A closed-loop computer-controlled system is built to validate the finite element model. The system is used to control the lateral buckling of a composite beam which is 40.6 cm long, 5.1 cm wide and .25 cm thick which is reinforced with two NITINOL-55 fibers that are 0.55 mm in diameter. The results obtained confirm the validity of the developed theoretical model and indicate that the buckled beams can be brought back completely to their unbuckled configuration by proper activation of the shape memory effect. Also, the results obtained suggest that reinforcing the beams with NITINOL fibers is effective in enhancing their lateral buckling characteristics.

Many of the Smart materials being investigated (e.g., Shape Memory Alloys (SMAs), piezoceramics, and magnetostrictives) exhibit significant hysteresis effects, especially when driven with large control signals. In this paper the similarity between the microscopic domain kinematics that generate static hysteresis effects, or ferromagnetics, piezoceramics and SMAs is noted. The Preisach independent domain hysteresis model, and its derivatives, have been shown to be a comprehensive class of hysteresis operator that captures the major features of ferromagnetic hysteresis, and hence it is proposed here as a suitable model for piezoceramic and SMA hysteresis also. This basic Preisach model is used to model piezoceramic sheet actuators bonded to a flexible aluminum beam, and a Nitinol SMA wire muscle that applies a bending force to the end of the beam. A numerical inverse Preisach hysteresis series compensator is also proposed and applied in a real time experiment thereby reducing the apparent nonlinear hysteresis effects for the piezoceramic actuator quasi-static case.

This paper contains theoretical and experimental results of controlling cantilever beams using distributed sensors. Shape functions for custom-made sensors are derived which allow for the identification and measurement of the states of a distributed system. This proposed technique allows for the direct identification of single-input single-output and multi-input multi-output system models as well as the implementation of full state feedback controllers without state estimation. Simple analog hardware circuits suffice to implement full state feedback controllers. System models can be identified quickly from time domain data for on-line controller adaptation. The suggested system identification method has the capacity to model only the modes of interest without relying on model reduction techniques.

The performance of various passive and active vibration control devices for equipment aboard spacecraft is analyzed. Several vibration control mechanisms, including pure friction, spring- damper, resilient friction and active friction control systems, are studied. Peak responses of equipment and their light subsystems under spacecraft g-jitter excitation are evaluated, and the results are compared to those for a payload with a fixed-base (rigidly attached) condition. For the frictional vibration control mechanisms, the nonlinear Coulomb model is used in the numerical simulation and the stick/slip condition at each time step is carefully evaluated. A tentative design for a new active friction control mechanism, by varying the normal force, is described. The device uses a velocity feedback to control the normal interface force with the aid of a magnet. By eliminating and/or minimizing the stick duration, the overall payload response reduces. Sensitivity of the performance of various vibration isolation mechanisms, to variations in the isolator properties, spacecraft excitation amplitude and equipment and subsystem frequency is also analyzed. The results show that the passive and active vibration control devices considered are generally highly effective in controlling the g-jitter vibration. In addition, these vibration control systems are insensitive to small deviations from their design conditions.

This paper presents a formulation of the transient response of elastomeric rods with embedded shape memory alloy actuators that incorporate the inherent coupling between the dynamics of deformation at the structural level, the thermal response and the constitutive law describing the shape memory alloy. Previous work by the authors has shown that the incorporation of shape memory alloy actuation in distributed parameter systems can induce a large class of nonlinearities including hysteresis effects in the SMA constitutive law, nonlinear kinematics of large deformation, and, in some cases, local plasticity effects. To derive a control methodology for the class of SMA embedded elastomeric rods considered in this paper, it is essential that the system characteristics of the nonlinear distributed parameter system be identified accurately. This paper presents an identification theory applicable to the coupled system of partial differential equations. The results are validated using both numerical simulation and experimental results.

We use the first-order shear deformation theory to study quasistatic deformations of a plate with piezoceramic elements (PZTs) mounted symmetrically on its top and bottom surfaces. The Galerkin formulation of the problem is derived. After having validated the computer code, the voltage to be applied to the PZTs in order to nullify the deflection of preassigned points of a plate deformed quasistatically is determined. It is shown that the deflections of the centerline of a simply supported plate and the tip deflection of a cantilever plate can be controlled by applying suitable voltages to the PZTs. The voltage to be applied to the actuators as a function of the surface area covered by them in the former case, and as a function of their distance from the fixed end for the latter case is depicted graphically.

Constrained layer damping has been used for many years to increase the damping in engineering structures. Active damping using smart materials has been investigated a great deal recently. Several researchers have suggested the concept of using viscoelastic materials with piezoceramics as the constraining layer. Since the piezoceramics are active materials, this concept can be referred to as active constrained layer damping. The paper presents a finite element model for a sandwich beam consisting of a host layer, a viscoelastic layer, and a piezoelectric layer. Lesieutre's method (Augmenting Thermodynamic Fields) for modeling damping was modified and applied to active constrained layer damping. The effects of the thickness and stiffness of the viscoelastic layer are studied. Experimental transfer functions for various configurations are presented. Active damping and passive damping are individually of interest, however, here we propose to combine these two types of damping to produce an active constrained layer system with the best of both technologies.

In this investigation, a theoretical model for adaptive beam structures with various boundary conditions was developed, and its potential applicability to electrorheological (ER) material based adaptive beams was introduced. The cross-sectional configuration of the beam selected was based on sandwiching a damping material between elastic face plates. The transverse continuous vibration model developed is based on thin plate theory. In the model shear and rotary inertia effects are neglected, and it is assumed that no slipping between the elastic layers and the damping layer occurs. All layers are considered to have the same transverse displacement. The resulting analytical model is able to predict the structural vibration response at any location on the beam surface, and is applicable to generalized sets of boundary conditions. In this study, four different boundary and external excitation conditions were considered, namely; clamped-clamped, clamped-free, clamped-simply supported and simply supported. The transverse vibration response was studied for each boundary condition set. Variations in the mode shapes, natural frequencies, and loss factors as functions of mode number and excitation frequency were analyzed. ER material based adaptive beam semi-active vibration control capabilities were analyzed from the observed model results.

There is widespread agreement among engineering scientists that the next major step in the design, development and manufacture of structural components will most likely be a focus on building smart structures. Therefore the question posed above will be examined in the context of design of engineering structures and structural components. We begin by defining the key words here: smart and structures. By structure we mean an assembly or its components which serves an engineering function. Therefore, in addition to familiar examples such as airplanes, helicopters, bridges, buildings, roads, railways, ships, etc., a wide variety of objects such as microchips, semiconductors, accelerometers, strain gages, tennis rackets, musical instruments, cooking utensils, etc. quaIif' to be described as structures or structural components. These components may be subjected to wind, ice, heat, loads, noise, radiation, etc. Extending the context a little further, we are mainly concerned with these structures or structural components maintaining their structural integrity in performing their design function throughout the period of their estimated life. The dictionary deñnition of the word '1smatt" is not very complimentary. However, the engineering community has adapted the word and in its usage over nearly a decade now, the word has come to define a certain extraordinary ability of a structural component in performing its function. Thus a structure or structural component may be defined as "smart" if it can sense a variable such as strain, pressure, temperature, etc. and process the same to initiate any action needed, in order to continue to perform the functions for which it was designed. Some prefer the word intelligence to describe the above characteristics but that word implies features fr and above those intended here, such as, for example, cognitive abilities. For the present we interpret smartness, in realistic terms, to mean the inherent ability of structural members to sense, diagnose and actuate in order to retain their integrity. Within the scope ofthese definitions, we study some specific examples and examine the extent oftheir smartness and relevance to engineering design

In this paper we are presenting the experimental dynamic response of an adaptive beam containing embedded mini-devices (sensors and actuators). The specimen used is a cantilever beam made of ALPLEX plastic as host material and PZT-5H as active devices for sensing and actuation. The experimental setup, and experimental results are presented and discussed. The capability of mini-actuators to change the dynamical behavior of the adaptive beam is demonstrated through adaptive stiffening and adaptive damping examples.

Pressure-shear plate impact experiments have been carried out on Cu-14.44Al-4.19Ni (wt. %) single crystals in order to study the kinetics of the stress-induced phase transformation occurring in this shape memory alloy. In these plane wave experiments, the crystal is oriented in such a way as to activate only a single variant of martensite. The habit plane of the variant to be induced is parallel to the impact plane and shearing is parallel to the shearing direction of the transformation. Wave profiles are monitored at the rear surface of the target plate. Arrival of step-like quasi-longitudinal and quasi-shear waves are evidence of the nearly elastic response of the material. Velocity profiles provide several indications of the existence of a propagating phase boundary. The measured transverse free-surface velocity after arrival of the three elastic waves, is less than the imposed velocity at the impact plane. This velocity difference, in the absence of any evidence of plastic deformation, is indicative of a stress induced phase transformation that propagates into the crystals from the impact face. The resolved shear stress driving the transformation is approximately the same in experiments conducted at different impact angles and velocities. Velocity profiles at the rear surface of the specimen also exhibit features consistent with the interaction of the quasi-longitudinal wave with a phase boundary. Reflection of the quasi-longitudinal wave from the phase boundary results in an increase in free surface particle velocity in both the normal and transverse velocity records. The velocity profiles appear to provide an effective means for probing the kinetics of the phase transformation.

Shape Memory Alloys (SMA) are characterized by a unique solid-solid phase transition during which significant strains are observable. On reversal of this phase transition, the strains also are almost completely recoverable. When properly constrained, a shape memory alloy is capable of producing large actuation forces. The frequency of actuation, however, depends on the heat transfer mechanism employed. Here, we undertake a theoretical study of a heat removal approach using semiconductors. These solid state elements employ the thermoelectric Peltier Effect for cooling or heating of the SMA. Compared to free or forced convection cooling mechanisms, cooling by the Peltier Effect is significantly faster. The phase transition is accompanied by a significant exchange of latent heat, and it is seen that the time period for a complete thermal cycle with the phase transition almost doubles as opposed to one without a phase transition. Partial phase transformation results in a lower period of cycling: 25% of transformation translates to the time period being 30% of the one with full transformation. AS a first step in the design of the actuator, a thermoelectric module is assembled in the laboratory for cooling/heating the SMA. Transient temperature profiles are recorded for the heating and cooling runs for two different materials--copper and SMA (with or without the phase transformation). These recorded profiles are then compared with the predictions from the model; the agreement is reasonable, particularly during the cooling process.

Experimental data for 55/45 Ni-Ti alloy (Nitinol) is presented in this work for correlation to phenomenological and thermodynamic/micromechanical constitutive and transformation models for shape memory alloys. Hysteresis data is used to obtain the transformation temperatures M_s-martensitic start, M_f-martensitic finish, A_s-austenitic start, and A_f-austenitic finish. Nitinol wires are mechanically tested in tension at five different temperatures: below M_f, at (M_s + M_f)/2, at (M₂ + A_s)/2, at (A_s + A_f)/2, and above A_f. For the phenomenological models, several characterization parameters are required. These are obtained experimentally and the models are correlated with the data.

In this paper, a macro-scale, phenomenological constitutive model for shape memory alloys is used in conjunction with energy balance equations to study the temperature profile and hence deformation profile seen in SMA wires under specific boundary conditions. The general, fully coupled thermo-mechanical problem is formulated, but analytical solutions are determined only for the decoupled case. Results for two specific boundary value problems are presented here: (1) resistive heating of an SMA wire--initial detwinned martensite leads to strain recovery (contraction) on heating; (2) deformation migration in semi-infinite SMA wire cooled at the boundary--deformation zone propagates and expands as initially austenitic wire transforms to martensite (expansion). In both cases, the region and extent of transformation is identified, indicating the magnitude of actuation obtained. Implications of the modeling for active control of structures are discussed.

Shape memory alloys have been used almost exclusively in components subjected to simple loadings such as uniaxial tension, pure bending or torsion. As part of the development of smart structures and adaptive components, shape memory alloys must be designed to function in complex, multi-dimensional stress states. In order for these materials to participate as active elements in smart structural designs, the alloys must undergo a phase transformation, during which their properties change. For 1D stress states, data generated in simple stress-strain- temperature tests is sufficient to define the onset and completion of a phase transformation. When multi-dimensional stress states are encountered, data from more complicated tests will be needed to define the transformation. In this paper, a model for multi-dimensional phase transformation criteria is introduced and discussed, and tests for defining the boundaries of the initiation and completion of phase transformation surfaces are proposed.

Presented here are some modifications of a previously developed hysteretic model for shape memory alloys (SMAs) and other metallic hysteretic deforming materials, for use in vibrational analysis. This work begins by modifying an existing rate-independent SMA constitutive model to eliminate some undesirable features. The SMA model is also simplified by reducing the number of parameters without sacrificing physically significant SMA characteristics. To enhance the ease of evaluating the model parameters, a systematic method for determining these parameters is developed and illustrated by example. The results of uniaxial tests on a Nickel-Titanium SMA are presented in conjunction with the previously mentioned parameter determining examples. The model response to small amplitude alternating strain inputs is also explored in detail. The modified SMA constitutive model is then extended to a 3D form.

A unified thermodynamic constitutive model for Shape Memory Alloy (SMA) materials, derived by a proper extension of the thermodynamic model proposed by Boyd and Lagoudas is presented in this paper. The specific selections for the form of the mixing energy associated with the martensitic volume fraction are identified for several earlier models. The heat energy released or absorbed during the forward or reverse transformation predicted by the different models is compared with the experimental data from calorimetric measurements. The constitutive model is implemented in a finite element analysis scheme using a return mapping integration technique for the incremental formation of the model. Finally, the constitutive model is utilized to analyze the thermomechanical response of an active metal matrix composite with embedded SMA fibers. Both tetragonal and hexagonal periodic arrangements of SMA fibers are considered in the calculation and the results from the different arrangements of SMA fibers are compared.

Thin film layers of shape memory alloys and ferroelectric ceramics can produce a family of `smart' heterostructures capable of performing both sensing and actuating functions. Important issues in these synthesis of these active structures are the ability to generate the appropriate crystalline phases of each material, while producing defect-free homogeneous high quality films. The compatibility of sol-gel processed Pb(Zr,Ti)O₃ (PZT) thin films with thin film shape memory effect TiNi substrates were investigated. Thin film TiNi was deposited on quartz substrates by physical sputter deposition utilizing a TiNi target in a ultra-high vacuum chamber, and followed by in-situ vacuum annealing. PZT was deposited on TiNi by sol-gel and spin coating processes. The ferroelectric tetragonal phase of PZT was obtained by a 600 degree(s)C anneal for 5 m in air. The heterostructures were nominally defect-free, unlike those obtained through deposition onto bulk TiNi substrates.

Ferroelectric and ferroelastic switching cause ferroelectric ceramics to depolarize and deform when subjected to excessive electric field or stress. Switching is the source of the classic butterfly shaped strain vs. electric field hysteresis loops and the corresponding electric displacement vs. electric field loops. It is also the source of a stress-strain curve with linear elastic behavior at low stress, non-linear switching strain at intermediate stress, and linear elastic behavior at high stress. In this work, a series of experiments on lead lanthanum zirconate titanate are modeled with a computer simulation of the ceramic microstructure. The polarization and strain for an individual grain are predicted from the imposed electric field and stress through a Preisach hysteresis model. The response of the bulk ceramic to applied loads is predicted by averaging the response of individual grains that are considered to be statistically random in orientation. The random orientation yields essential non-linear behavior of the observed strain and electric displacement hysteresis loops and the non-linear stress- strain curve for the polycrystalline ceramic. The linear piezoelectric effect opens up a butterfly shape to the strain vs. electric field hysteresis loop but the model fails to predict the observed effect of 90 degree(s) switching. The grain to grain residual stress and residual polarization are estimated from inclusion calculations. These are both a function of the remanent strain and remanent polarization of the ceramic. This constraint opposed switching but has little effect on the butterfly shape in the strain vs. electric field hysteresis loop.

Experimental investigations suggest that microcracking may be the major cause for the commonly-encountered degradation of ferroelectric actuators, namely electric fatigue. The present analysis, based upon the lattice distortions occurring during the formation of electric twins, indicates that intersections of domain walls with grain boundaries are likely to be origins of microcracks because of the concentrations of stress and electric fields arising due to the incompatibility of the lattice distortions with the grain boundary constraints. This further motivated an asymptotic analysis on the stress and electric field concentrations near intersections of domain walls with grain boundaries. This analysis is carried out within a frame work of electrostatics for deformable continua. Assuming that the electromechanical state in each of the ferroelectric domains is slightly distorted from one of the natural states of the crystal, the authors developed a piecewise linearized model for which concentrations of stress and electric fields correspond to singularities of these fields. The asymptotic analysis concludes that both the stress and electric fields exhibit a power-law singularity. Several examples indicate that the order of singularity depends upon the crystal orientation and the orientations of the domain walls with respect to the grain boundary which they intersect.

A system of 2D equations for the flexural vibrations of symmetric composite plates with both unelectroded and electroded portions of piezoelectric actuators attached to the top and bottom surfaces is obtained. The resulting system of equations is applied in the analysis of the cylindrical bending of a composite plate caused by voltages applied to partially electroded piezoelectric actuators. It is shown that no shearing stress is transferred across the interfaces between the electroded portions of the actuators and the composite plate and that an exponentially decaying shearing stress is transferred across the interfaces in the unelectroded regions of the actuators. The resulting finite shearing stress shows that when the electrodes end a short distance from the edge of the actuator, the undesirable singular shearing stress distribution that exists when the electrodes extend to the edge of the actuator is eliminated.

Piezoelectric actuators are useful for smart structure applications because they are light weight, compact and fast compared to other types of actuators. The two most common configurations, bimorph and stack, have significant limitations on the available force and deflection, respectively, which can be produced. This paper describes a new class of piezoelectric bimorph actuators that supply larger force than the traditional straight cantilever beam bimorph design. These actuators, called C-blocks because of their half circle shape, can produce approximately five times the force of the traditional design with only a slight decrease in the deflection. In addition, they can be combined in series or in parallel like building blocks to form larger actuators that provide improved deflection and/or force capabilities. This paper presents a mathematical model that provides an approach to designing and predicting the force performance of a C-block architecture. This paper also summarizes a set of manufacturing techniques developed to fabricate C-block architectures from polymeric piezoelectric material Polyvinylidene Fluoride (PVdF). These techniques were used to manufacture prototypes used in a Force-Voltage experiment that was performed to measure force performance. The results from these experiments confirm the mathematical model is accurate and the force performance of a C-block architecture is substantially increased in comparison to the force generated from a traditional straight beam bimorph.

Shape memory alloy (SMA) materials such as nitinol have unique properties associated with the shape recovery effect and the material's phase changes that have been used in a variety of actuator and sensing applications. By embedding SMA elements into host composite materials, control or modification of the SMA hybrid composite's structural properties can be accomplished in-service, thereby increasing the hybrid composite's structural functionality. Utilizing the SMA's substantial capacity to dissipate strain energy to increase the hybrid composite's static functionality is addressed herein. Specific applications of SMA hybrid composites for improving polymer matrix composite material's impact damage resistance is presented. Hybridizing composites with nitinol improves their impact resistance because of nitinol's tremendous capacity to dissipate impact strain energy through the stress-induced martensitic phase transformation. The amount of impact damage is reduced and the material's resistance to impact perforation at various velocities is improved. The experimental response of nitinol hybrid composites to low and high velocity impact perforation is presented. The unique toughness and resistance to permanent deformation that is a result of the stress-induced martensitic phase transformation enables the nitinol to dissipate on the order of 4 times the strain energy of high alloy steel and 16 times that of many graphite/epoxy composites.

A mathematical model is presented for the dynamic response of contractile fiber bundles embedded in an elastic medium. The fiber bundle is assumed to consist of a parallel array of contractile fibers made from contractile shape memory alloy wires of a few mills in diameter. The proposed model considers the electrically induced (ohmic heating) contraction of the fibers. A model is then presented for the electro-thermo-mechanics of such SMA assemblies including the heat transfer effects. For the proposed model, we consider the fiber bundle of SMA to be electrically insulated from the host elastic medium by means of a dielectric liquid to enhance the transfer of heat from the SMA wires to the host medium. The host elastic medium acts as a heat sink to quench the SMA wires during the power-off cycle, mainly by conduction heat transfer. Thus, the fibers can be electrically heated and subsequently contracted to compress the elastic medium. Design details are first described. In essence the dynamic behavior of the assembly depends on the interaction between the current supplied to the wires and the heat transfer from the wires to the elastic medium. Further, a mathematical model is presented to simulate the electro-thermo-mechanics of expansion and contraction of such assemblies. The proposed model takes into account all pertinent variables such as the strain (epsilon) , the temperature of the fibers T(t) as a function of time t, the elastic medium initial temperature T₀, the martensite fraction (xi) , the elasticity of the host medium and the overall heat transfer coefficient h. Numerical simulations are then carried out and the results are displayed in the form of graphs. Computer simulation have been performed for a 1 meter aluminum cylinder of 50 mm in diameter with an elastic modulus of 70 GPa, with 2400 embedded SMA wires of diameter 0.254 mm under a duty cycle of 10 seconds on followed by 20 second off, to allow the temperature of wires to drop back to the martensitic phase. The dynamic response of such assemblies appear to be quite robust for a number of engineering applications.