Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 47
Expanding the Vision of Sensor Materials 4 SELECTED SENSOR APPLICATIONS FOR STRUCTURAL MONITORING AND CONTROL Advanced sensors and actuators, together with exponential improvements in computer technology, are causing a surge of interest in the development of "intelligent" structures and equipment. Scientists and engineers are investigating materials systems that are capable of monitoring condition, changing shape, controlling vibrations, accommodating changes in the environment. Numerous applications are possible that range from the mundane to the exotic (Ramamurthy, 1992). They include automotive springs, smart skins on aircraft, smart bridges, improved biomedical devices, and advanced military systems. The shorter-range interests address incorporation of sensors into systems that monitor structural performance throughout the life cycle. The long-range goal is the deployment of active systems that are able to autonomously adapt in response to a change in the environment. These technologies offer the potential of substantive advancements for better, more reliable structures. Within a product's total life cycle, sensors can play a significant role during manufacture and customer use. Sensors are able to monitor manufacturing operations and ensure that the final product meets its design specifications, as discussed in Chapter 3. Throughout use, sensors could perform many important functions such as monitoring the condition and performance of the product, controlling use, and aiding in maintenance and repair decisions. LIFE-CYCLE MANAGEMENT Life-cycle management (LCM) has the ultimate goal of integrating the fabrication and customer-use-periods, in order to create a birth-to-death concept of product design. Three major considerations are pushing the implementation of LCM: reduction in operating costs through optimizing maintenance procedures and avoidance of premature component replacement; enhanced product safety in the face of product liability requirements; and product life extension.1 Four disparate applications (condition monitoring of equipment, control and condition monitoring of aeronautical jet engines, structural monitoring of test prototypes, and structural monitoring during operation) are discussed below to illustrate the broad range of current LCM applications and associated sensor technology requirements. Condition Monitoring of Equipment Preventive maintenance is commonly employed in industry to reduce the probability of equipment failure. In line with this approach, maintenance actions are typically scheduled on the basis of operating time without regard to whether the equipment actually needs specific maintenance or repair. Condition monitoring is a more "intelligent" approach, which detects impending problems early
OCR for page 48
Expanding the Vision of Sensor Materials FIGURE 4-1 Factory equipment condition monitoring system. enough to allow corrective maintenance actions to be prioritized and scheduled during noncritical time periods. Properly designed and implemented, such systems can reduce unnecessary work, allowing concentration on the most important maintenance and repair actions. In the design of a typical installation, components to be monitored are identified and measurement points designated. Sensors are specified in terms of functional requirements, so that a variety of sensor options are available. This "open architecture" strategy is based on the premise that suitable sensors are commercially available.2 These systems typically employ sensors that target one or more equipment degradation modes. They then provide early warning of an impending problem when signs of degradation begin to appear. The sensor framework described in Chapter 2 can be used very effectively in developing a core set of sensor specifications and in comparing the attributes of candidate sensors. Condition monitoring requires periodic polling of the sensors and analysis of the data. Increasingly, the analysis is facilitated by computerized expert systems. A basic assessment compares the value of a parameter with an established alarm level.3 A higher level of analysis examines the trendline of the data to estimate when future maintenance actions may be necessary. In addition, an automated diagnostic tool can aid in determining the root cause of equipment problems; accuracy can be enhanced by fusing the readings from multiple sensors. The architecture of a typical automated equipment condition monitoring system is depicted in Figure 4-1. The core of this architecture is an expert system that analyzes the inputs from a distributed network of sensors. Such a condition-monitoring system is being used to monitor turbomachinery and other expensive rotating equipment (e.g., that used for electrical power generation). This equipment has well-characterized warning signs of impending failure. Failures typically result from misalignment of the rotating elements, imbalance, antifriction bearing defects, and mechanical looseness. A vibration sensor monitors each machine, since a change in a machine's vibration level is usually indicative of a change in the machine's performance, perhaps as a result of deterioration that will lead to failure (Voegtle and Bever, 1992).
OCR for page 49
Expanding the Vision of Sensor Materials Control and Condition Monitoring of Aeronautical Jet Engines Modern jet (gas turbine) engines that power today's aircraft are very complex, highly engineered systems. The centerpiece of these engines is rotating turbomachinery that is built and maintained to very exacting standards. Engine operation is well understood at the macro level. Hence, even the most sophisticated jet engines developed for front-line military fighter aircraft can be controlled by monitoring and adjusting relatively few critical engine parameters. In order to accomplish this, jet-engine technologists have been among the leaders in implementing sensor-based control technology. Advanced jet engines use an electronic digital control to manage engine operation and performance. These controls are more capable and reliable than the hydromechanical control they replaced. For instance, through March 1993, approximately 788,000 flight hours had been accumulated by digital controls installed on F100-PW-220 engines; during this period, 19 controls were removed for reported problems. This unscheduled removal rate of 1 per 41,500 engine flight hours is a factor-of-five improvement over the previous generation of hydromechanical controls (Khalid, 1994). The electronic digital control receives engine sensor and actuator position feedbacks, processes these signals with its control logic, and sends commands to drive engine effectors to provide optimum engine and aircraft performance. The basic control parameters are engine pressures and total airflow, which are derived from the engine sensor inputs. An overview of a typical control system, along with sensed parameters, is indicated in Figure 4-2. Also shown are typical engine-control interfaces with the engine's hydromechanical components that control main gas-generator fuel flow, augmentor4 sequencing, exhaust nozzle position, augmentor ignition, and bleed air. A more advanced control is being used by the Air Force's F-22 Advanced Tactical Fighter engine (F119-PW-100 engine). Two full-authority digital electronic controls (FADECs) are employed, each with its own dedicated set of sensors. They operate "dual active," with both channels continuously active and performing identical functions. Each FADEC also communicates with the aircraft's flight-control system. Each channel provides half the control output; if one channel fails, the other channel assumes full authority.5 Cross-channel communication allows sensor values to be compared between FADEC channels. These electronic digital controls employ a number of different sensors to detect engine speeds, temperatures, pressures, and actuator/valve positions. The control's ability to accurately sense these critical parameters is important so that the control can recognize when engine limits are being approached, and appropriate accommodations can be made in fuel flow scheduling. The sensors must FIGURE 4-2 Jet engine digital control system.
OCR for page 50
Expanding the Vision of Sensor Materials have high reliability and repeatability in a severe environment. Some examples of sensors that are currently used include: a magnetic pickup sensor to measure engine fan speed; thermocouples to measure engine gas temperature, as well as fuel, oil, and hydraulic fluid temperature; an optical pyrometer to measure temperature of a rotating set of turbine blades; pressure transducers to measure gas pressures, such as inlet and burner pressures; strain-gauged diaphragm sensors to monitor oil, fuel, and hydraulic pressure; linear and rotary variable differential transducers to measure actuator and valve positions; and an ultraviolet light photodetector to monitor the thrust augmentor flame. A significant advantage of a digital control is that it can be readily integrated with an engine diagnostic system to continuously assess the health of the engine. Table 4-1 summarizes the key storage functions of an onboard engine diagnostic system. Figure 4-3 is a simplified block diagram depicting how the diagnostic system interfaces with the engine control for the F100-PW-220 engine mentioned above (Khalid, 1994). When a control system fault or an anomalous engine event is detected, the pertinent aircraft and engine conditions are stored in memory for later use by maintenance technicians. Also, a fault indicator is set on an aircraft status panel to alert maintenance personnel that a problem occurred in flight. A significant payoff has resulted from being able to quickly isolate faults, at a high confidence level, that are quite difficult to duplicate on the ground. The reliability and maintainability of the digital electronic jet-engine control systems are sufficiently high to have allowed their certification by the Federal Aviation Administration and successful transition to commercial aircraft engines. Advanced architectures for future control systems are being developed that have potential benefits in terms of lower cost and weight. Distributed control architectures that utilize engine-mounted smart sensors and actuators that communicate to the propulsion system controller through high-speed optical data bus are being studied (Skira and Agnello, 1991). The sensors and actuators can be TABLE 4-1 Functions of a Jet Engine Diagnostic Unit Function Typical Stored Parameters Engine event detection Low-pressure turbine rotor speed High-pressure turbine rotor speed Turbine airfoil over-temperature Hot start Engine stall Augmentor blowout or mis-light Engine performance determination Fuel flow Engine pressures Mach number Thrust-level request Life-usage data collection Engine operating time Engine flight time Hot section time Control-system fault detection Electronic control faults System faults
OCR for page 51
Expanding the Vision of Sensor Materials FIGURE 4-3 Jet engine engine diagnostic system. copackaged with electronic modules to provide signal conditioning directly at the source. These "smart" devices will have the ability to provide their own environment/temperature compensation, use built-in test features to assess their health, compute necessary engineering units conversion, resolve actuator position requests from the engine controller, and perform actuator loop closures. Use of these smart devices would eliminate the need for point-to-point wiring for sensors and actuators at extended distances from the engine controller and would, in turn, greatly reduce engine harness weight (Tillman and Ikeler, 1991). The use of smart sensors and actuators for jet-engine control is currently limited by the availability of mature high-temperature electronic components that can withstand the engine operating environment. 6 As this technology advances, smart devices will increasingly appear in engine applications. One of the most challenging turbine-engine sensor requirements is measuring the gas temperature as it exits the combustor and enters the turbine. As engine temperatures have increased, the durability and performance limit of engine temperature sensors are an issue. Thermocouples are commonly used for engine temperature sensing, but their lifetime above 1100 ¹C (2000 ¹F) decreases rapidly. Since the first-stage turbine in advanced engines currently operates above this temperature, the temperature sensor has been moved downstream to a cooler environment. The turbine inlet temperature is then estimated using an empirically derived relationship. However, this approach results in inaccuracies. For instance, it does not take engine-to-engine variations into account, nor does it compensate for engine operational changes due to normal "wear and tear." In addition, future-generation turbines are expected to have turbine inlet temperature well above 1650 ¹C (3000 ¹F), which renders the thermocouple approach unusable (Bird et al., 1990). On the latest advanced military engines, an optical pyrometer that measures turbine blade temperature is used to improve the accuracy of the turbine inlet temperature. An improved temperature sensor that can provide accurate temperature measurement over a broad temperature range is under development. It uses an opaque coated sapphire rod to act as a near-blackbody cavity. A fiberoptic guide collects the radiation emitted from the sapphire rod and transmits it to a photodetector as shown in Figure 4-4. The resultant voltage is proportional to the engine temperature (Bird et al., 1990). This sensing technology will be an advancement, but it suffers from materials temperature limitations of the sapphire rod and silicon carbide housing that restrict the upper end of temperature measurement to about 1650 ¹C (3000 ¹F). Thus, although this technology
OCR for page 52
Expanding the Vision of Sensor Materials FIGURE 4-4 Fiberoptic high temperature sensor. offers an advancement beyond thermocouples and will probably be adequate for commercial jet engines for the foreseeable future, it does not represent the ''ultimate" solution. Another temperature-sensing approach, which is much less mature, involves exciting the molecules in the gas path with a laser; the resulting spectra can be calibrated against temperature over a broad range. Using this method, it may be possible, in the long run, to monitor the temperature at the exit of the combustion chamber directly, providing an even greater opportunity for exact engine control. This sensing method has the added advantage of being nonintrusive (Eckbreth and Stufflebeam, 1985; Hall and Eckbreth, 1984). Structural Monitoring During Validation Testing Oftentimes particularly difficult sensing needs arise during the development of a testing program that is necessary to validate an innovative engineering design of equipment. When the overall program involves public safety concerns as well as the expenditure of considerable resources, there is a heightened interest in ensuring that the design is correct. Design models offer a list of parameters that should be monitored to validate the design assumptions and calculations, and these in turn establish the sensor requirements. Frequently, there are a multitude of sensor candidates, none of which exactly "fits the bill." As an example of sensor needs during a development program, the National Aerospace Plane (NASP) program, a joint effort of the U.S. Air Force and the National Aeronautics and Space Administration had advanced far enough that planning for the flight-safety validation testing of critical airframe components was initiated (Lockheed, 1988). NASP's flight envelope is substantially different from conventional experience; for instance, it will cruise in the range of Mach 6 to Mach 12.7 (Conventional passenger jets cruise well below the speed of sound at Mach 0.7. Even supersonic military fighters usually do not exceed Mach 2, and then only for a limited time.) NASP's mission profile will result in aerodynamic heating of some sections of the airframe to peak temperatures on the order of 1650 ¹C (3000 ¹F) for a large fraction of flight duration, in excess of 1 hour. In addition, aerodynamic and aerothermal loads will be significantly larger than those of previous aircraft experience and will experience significant transients. This hypersonic flight regime demands that the testing program be able to rapidly simulate changing aerodynamic and thermal loading on structural components and establishes instrumentation needs for monitoring the test conditions and the resultant structural response. Validation testing methods include wind-tunnel, shock-tunnel, and structural simulations. Both static (e.g., normal, side, axial forces, and related moments) and dynamic (e.g., pitch, yaw, and roll damping) measurements are required. Measurement of static surface pressures are important for structural panel proof testing, engine inlet evaluation for propulsion system performance, and assessment of aerodynamic interactions. Dynamic pressure measurements are required for analysis of acoustic problems, panel flutter, and unsteady flow phenomena. Knowledge of actual surface heat transfer rates and temperature distributions is critically important to structural survivability and integrity. Structural engineers determined that strain measurement at elevated temperature was the prime requirement, followed by heat flux and structural
OCR for page 53
Expanding the Vision of Sensor Materials temperature measurement. Instantaneous nonintrusive flow diagnostics measured at a point, across the plane, and in the entire three-dimensional field were also judged very important. Two additional daunting sensor requirements were identified: rapid thermal mapping of entire components or structural elements and measurement of pressure fluctuations in an aerodynamic flow close to a surface in a high-temperature environment. As a first step in developing an appropriate sensor suite, all possible sensor technologies were identified. Table 4-2 lists a generalized set of sensing concepts for four of the key parameters. Evaluation of each of these sensor options was a significant undertaking. The framework presented in Chapter 2 could be used to develop a structured approach that examines each option against the requirements and determines the gaps in capability. These gaps could then serve as the basis for sensor development efforts. TABLE 4-2 Initial Sensor Options for NASP Ground Simulation Testing Parameter Sensor Options Pressure absorption/ scattering/ fluorescence bourdon tube capacitance fluidic fiber optic laser piezoelectric semiconductor strain gauge Temperature acoustic antiStokes Raman spectroscopy fluid expansion infrared radiation isotope radiation laser Raman/ fluorescence phase change paint resistive element semiconductor thermoelectric thermal phosphor Velocity drag force hot film hot wire ionization laser fluorescence/ Doppler/fiberoptic laser particle pitot pressure ultrasonic velocimetry/ Doppler/time Strain capacitance foil carbon filament fiberoptic laser speckle resistive wire thin film ultrasonic Structural Health Monitoring Structural health monitoring involves the evaluation of a structure's capability to carry useful load. Structural health monitoring is being given greater emphasis as a result of the interest in extending the life of equipment and facilities as long as possible (Gerardi and Hickman, 1992; Rotherford and Westerman, 1992). This evaluation can be done passively (e.g,, through deflection or distortion measurement), actively (e.g., analyzing ultrasonic response), or by inference (e.g., examining time history of load cycles). Of these three techniques, the active approach is the most desirable, since the signal-to-noise ratios of the measurement are typically high, and sensors can target the critical areas where precursors to serious damage or failure are expected to originate. On the other hand, this approach typically adds cost and complexity to the structure's design, and benefits would not accrue until later in the life cycle. Monitoring the integrity of structures is a high-interest application of this technology. Early detection of fatigue cracks in metallic structure can allow relatively inexpensive repairs to be made, rather than replacing or extensively repairing structural components. In the case of aircraft (or bridges, etc.), the cost of replacement (in terms of capital cost and loss of use) is extremely high, and patch-type repairs are universally made. Once the repairs are made, continued monitoring can provide a high degree of confidence in continued structural integrity. How to effectively monitor the degradation of structures is currently an active area of research, with many different sensor approaches under exploration. Several of the approaches are discussed below. The Air Force has a need to repair fatigue cracks in aluminum structural components of C-130 transport aircraft. A current repair process uses composite boron/epoxy patch. After repair, the component strength greatly exceeds that of the original aluminum structure. However, if the epoxy interface weakens or delaminates over time, the patch effectiveness can eventually degrade to the point of uselessness. The Air Force must therefore assess the integrity of this patch over an expected 20-year lifetime, as schematically depicted in Figure 4-5. The example sensor technologies described below
OCR for page 54
Expanding the Vision of Sensor Materials FIGURE 4-5 Embedded sensors in a composite repair section. are not a comprehensive list of the many different sensor methods that could be employed to accomplish this task but are representative of sensor approaches that have been or are being applied to satisfy this particular requirement. Two different approaches have been used: embedded fiberoptic sensors8 (passive) and ultrasonic transmission (active). If the strains exceed preset values, the sensor could issue an alert (Pollack, 1992). Embedding fiberoptic probes for strain measurement
OCR for page 55
Expanding the Vision of Sensor Materials in epoxy was found to be difficult. First, the interface between the fiber and the epoxy was unreliable, and intermittent separation occurred. Second, moisture was absorbed over time in the epoxy and ultimately permeated to the silica fiber; this caused surface microcracks that resulted in light loss and eventually fiber breakage. A materials development effort resulted in improved reliability and survivability for the optical fibers due in part to the development of hermetic coatings for the fibers. System reliability was also improved through better connectors and electronically simpler processing systems. The ultrasonic approach involved attaching transducers to the structure in order to generate ultrasound waves around and through the patched area.9 If the bond degrades or the patch deteriorates (e.g., through cracking, delamination, modulus changes) the ultrasonic pulse spectrum will also change. Problems of implementation included finding a way to couple sufficient pulse energy into the structure and to securely attach the transducers for long-term environmental survivability. Energy coupling was achieved by two methods: (1) using wires as acoustic waveguides to direct ultrasonic energy into specific portions of the patch and surrounding structure and (2) developing piezoelectric and electrostrictive wafers capable of generating ultrasonic energy that could be adhesively bonded to the structure. Most significant in this latter case was the application of the tape casting process that enabled routine production of these ceramic wafers with thicknesses down to 5 mils. In addition to fatigue cracks, a significant cause of gradual structural debilitation in aircraft is metallic corrosion. Visual inspection by trained technicians has traditionally been the detection method of choice. However, corrosion normally occurs in areas of the aircraft that have restricted access and thus are difficult and expensive to inspect visually. Several solutions have been proposed, although none has yet been demonstrated to be fully cost effective. One method involves the continuous monitoring of the environment (e.g., detecting the presence and concentration of moisture, corrosive chemicals [NaCl] in the air, corrosion by-products) and then using this data in a computerized analysis program to estimate the likelihood of corrosion. The computer would then cue the inspection process, identifying when and where to look for indications of incipient corrosion. Such monitoring could be performed by chemical sensors, which have considerable promise (see Chapter 6). However, these sensors must be able to survive for many years in the operational environment, detecting the critical corrosion precursors. Further maturation of chemical sensor technology is necessary before this approach becomes feasible. Other sensor solutions include a variety of nondestructive evaluation techniques, such as conductivity measurement, capacitance thickness measurement, neutron adsorption, and high-resolution radiography. A most promising near-term inspection system is the High Resolution Real Time Radiographic system that is being developed by the U.S. Air Force. The unique part of this system is the development and use of scintillating doped optical-fiber matrix screens coupled to solid-state charge-coupled device video camera chips. This replaces conventional photographic-film recording media and enables images of much higher resolution at much lower dosage of x-ray illumination. Video spectroscopy is emerging as an important sensing method for this application. It involves remote inspection of a part using a vision system that can image the corroding surface in any part of the optical spectrum, from ultraviolet to infrared. Corrosion products can be observed and the extent of damage characterized by analysis of the image's optical spectrum. For this method to be effective, the optical system must be able to view a remote area over a wide bandwidth. Moreover, it may be necessary to illuminate the area with a pulsed laser. These requirements have led to the development of an image translation system that employs a coherent bundle of flexible optical fibers with a very broad passband. Both materials development and innovative equipment design are crucial to the advancement of this technique. SMART MATERIALS AND STRUCTURES Smart materials and structures are defined in the literature in many different ways. Smart materials are primarily distinguished by being able to perform both sensing and actuating functions.
OCR for page 56
Expanding the Vision of Sensor Materials Newnham (1993) provides definitions of the degrees of "intelligence" of smart materials. Passively smart materials respond to external change without external control. Actively smart materials utilize a feedback loop enabling them to function like a cognitive response through an actuator circuit. Very smart materials sense a change in the environment and respond by altering one or more of their property coefficients, turning their sensing and actuating capabilities, i.e., materials with built-in learning functions are smarter than those without learning. "Intelligent" materials integrate the sensor and actuator functions with the control system. The success of engineering is ultimately measured by the ability of a structure or component to perform its intended function without failure; oftentimes these structures must operate in an environment that was not fully known or defined at the beginning. Traditional design practices use a philosophy of "defense in depth." This dictates that a system incorporate numerous reinforcements, redundant subunits, and backup systems that add mass and cost to ensure safety in an uncertain environment. This typically results in conservative designs that require considerable human effort to postulate and analyze all possible contingencies, the application of more resources than required to implement the structure, and the consumption of additional energy to produce and maintain the structure. The ultimate goal of "intelligent" materials systems is to be able to adapt to an unpredictable operational environment. In the long term, the conventional "defense in depth" design philosophy could become outmoded. The significant effort in developing the theories, simulations, and hardware implementations for the control of machines, as discussed earlier in this chapter, has spurred the development of smart structures. Modern control approaches, including adaptive control and neutral networks, are finding widespread use. However, the tremendous number of sensors, actuators, and associated power sources that are required for smart structures do not lend themselves to conventional, centrally processed computer architectures. A distributed architecture appears to have significant advantages. Further discussion of actuators and signal processing, communication, and controls is beyond the scope of the present report. The bibliography contains references for further information on these topics (Wada et al., 1990; Crawley and de Luis, 1987; Rogers and Robertshaw, 1988). Sensing is a critical function within smart material systems and structures. Damage control, vibration damping, acoustic attenuation, and intelligent processing all require accurate information provided by sensors to describe the state of a structure. Sensing capabilities can be added by externally attaching sensors or by incorporating them within the structure's material during manufacture. Sensing materials used for this purpose include optical fibers and piezoelectric materials, possibly in conjunction with "tagging" particles. Optical fibers can be used either extrinsically or intrinsically in sensing, as explained in Appendix D. When used extrinsically, the optical fiber is not itself a sensor; it merely transmits light. Intrinsic sensing relies on changes in the light transmission characteristics of the optical fiber. The use of optical fibers to perform intrinsic sensing in smart structures was first investigated in 1979 at the National Aeronautics and Space Administration's Langley Research Center. In this early research, optical fibers were used to measure strain in low-temperature composite materials. (Measures, 1989a,b; Hickman et al., 1990). The work in "smart skins" (as it was then called) provided a catalyst for the development of a variety of fiberoptic sensors. Interferometric, refractometric, blackbody, evanescent, modal domain, and time-domain sensors were investigated for use in nondestructive materials evaluation, in-service structural health monitoring, damage detection and evaluation, and composite cure monitoring. Researchers examined the use of fiberoptic sensors as magnetic field sensors, deformation and vibration sensors, accelerometers, and sensors in propulsion systems. Resistance to adverse environments and immunity to noise from electrical or magnetic disturbances are among the many advantages of fiberoptic sensors (Claus, 1991).
OCR for page 57
Expanding the Vision of Sensor Materials Piezoelectric materials have found widespread use as sensors in intelligent material systems. Piezoelectric ceramics and polymers produce measurable electrical charges in response to mechanical stress. Recently piezoelectric materials have been used to simultaneously actuate and sense. For example, when bonded as a patch to a structure and powered by an alternating voltage, these materials induce structural vibrations that in turn modulate the current flowing through the patch. The current flowing through the piezoelectric transducer can be divided into two parts, one due to the actuation function and the other due to the sensing function that is representative of the mechanical strain. The primary advantage of controlling structures with colocated velocity feedback is that such systems are inherently stable at all frequencies. Piezoelectric polymers, such as polyvinyldene fluoride (PVDF), are used in many sensing applications. PVDF can be formed as thin films and bonded to many surfaces. Uniaxial films, which are electrically poled in one direction, can measure strain along one axis, while biaxial films can measure strain in a plane. The g-constants, which represent the voltage generated in response to a mechanical strain, are typically 10 to 20 times those of piezoceramics. PVDF film also produces an electric voltage in response to infrared light; the pyroelectric coefficient defines this relationship. As a sensor, it behaves like a strain gauge, but does not require a conditioned power supply. The output signal is also typically greater than an amplified strain gauge signal. This high sensitivity is due to the thinness of the typical PVDF film (˜ mil). Even a very small extensional force creates a large strain because of the small cross-sectional area. Specially shaped PVDF sensors have also been used as modal sensors for monitoring particular modes of vibration of structural elements such as beams, plates, and cylinders. The sensitivity of PVDF films to pressure changes has been utilized in tactile sensors that can read the Braille alphabet and distinguish different grades of sandpaper. Tactile sensors with ultrathin (200–300 µm) PVDF films have been proposed for use in robotics.10 The pyroelectric effect, which allows piezoelectric polymers to sense temperature, limits their use of these types of sensors to lower temperature ranges (Lovinger, 1983; Tanaka, 1981; Amato, 1989; Carlisle, 1986). One of the key factors limiting the application of ceramic sensors, such as a piezoelectric, is their brittle nature and low tensile strength. These limitations can be overcome to an extent by combining ceramics with polymer materials. The development of electroceramic composites offers two distinct advantages: overcoming the limitations of brittle ceramics and tailoring properties to suit specific needs. The piezocomposite consists of long, thin piezoceramic rods, poled along their axis, held parallel to each other and perpendicular to the faces of the plate by a piezoelectrically passive polymer, as shown in Figure 4-6. In this configuration, the thin rods can expand or contract laterally at the expense of the softer polymer even though the plate as a whole is restrained. This results in an increased bandwidth and higher sensitivity, which allows the external high-frequency uniaxial strain of the ultrasonic waves to be effectively coupled electrically within the rod composite. Piezocomposite materials are widely used in hydrophones and medical ultrasonic transducers and have improved sensitivity and mechanical performance compared with the traditional piezoelectric ceramics (Vest, 1991). A good example of this is the 1–3 piezoelectric transducer composite pulse-echo transducer used in medical diagnostic imaging (Newnham, 1986; Newnham and Ruschau, 1991; Newnham et al., 1978). Polymers containing piezoelectric powders have also been investigated FIGURE 4-6 Piezocomposite structure.
OCR for page 58
Expanding the Vision of Sensor Materials for use as sensing materials. Piezoelectric paint and coatings are being developed that can be applied to complex shapes to provide information about the state of stress and health of the underlying structure. Passive and active tagging are sensing techniques that involve adding taggant particles to materials. Embedded particles that are piezoelectric, magnetostrictive, electrostrictive, or magnetic can be used to provide inherent information about the in-process or in-service state of adhesives and polymers. Tagging offers the advantage of distributed in-situ sensing, which is not possible with many types of sensors. If the material system is properly designed, the tagging particles will have a minimum adverse effect on the properties of the host material (Clark, 1991, 1992; Zhou et al., 1993). Passive techniques involve sensing the distribution of the particles. The earliest literature regarding tagging for nondestructive evaluation is a patent in which ferromagnetic particles were added to concrete structures containing nuclear waste to monitor the integrity of the structures. Another example of a passive technique is adding magnetic particles to an adhesive and then employing an eddy current probe to detect voids in a tagged adhesive bond. Active tagging involves exciting the sensing particles and measuring the response of the host material. Applying an alternating magnetic field to a polymer tagged with magnetic particles and measuring the resulting force is an example of active tagging. Applications of passive and active tagging techniques include characterization of adhesive bonds, cure monitoring, intelligent processing, nondestructive materials evaluation, damage detection, and in-service health monitoring. The tagging technique is now being investigated for nondestructive evaluation applications in composites, in adhesive bonding layers, and even in intelligent manufacturing processes. Some of the preliminary research involves using active ferromagnetic tagging to determine the complex modulus of polymer specimens using magnetostrictive tagging particles to detect delamination in thermoplastic composites, using active ferromagnetic tagging particles to monitor the curing process of composite specimens and using ferroelectric particles to monitor adhesive bonds. Using the tagging method for nondestructive evaluation usually requires the selection of the appropriate sensor particles, determination of the quantity of particles, preparation of the proper transducers (to pick up signatures resulting from the tagging particles), and the proper interpretation of the measured signal. Current knowledge of the above factors and issues is limited, and the investigations are in the preliminary stages. Although the investigations so far have demonstrated great potential for both the active and passive tagging techniques for nondestructive evaluation, much further development and research is required to make them competitive with the well-developed existing nondestructive evaluations methods. SENSOR MATERIALS DEVELOPMENT OPPORTUNITIES IN STRUCTURAL MONITORING AND CONTROL The technology required for structural monitoring and control is still evolving. Sensors constitute an enabling technology, as do the other elements of the system, such as computer processing, actuators, and controls. Sensor technology can be viewed as falling into one of four categories, listed in decreasing order of maturity (Lockheed, 1988): proven, off the shelf, and commercially available; proven, but requiring validation in the service environment over the extended operational envelope; new concepts currently undergoing laboratory development; and new concepts proposed with undetermined applicability and difficulty of implementation. For the near term, the critical issues are engineering-related. For any given measurement requirement that typically arises for condition monitoring, there usually are a great many sensing options from categories (1) and (2) above. A critical challenge is to assess these different options against the requirements, so that the most appropriate selections can be made. The framework described in
OCR for page 59
Expanding the Vision of Sensor Materials Chapter 2 can be used very effectively to specify the requirements for a sensor technology and then serve as a framework to compare different sensors. If no candidate sensor meets all of the requirements, then a development effort can provide incremental improvements to the technology. Such developments often require innovative equipment design or enhanced computer analysis programs, as well as materials improvements. An example is video spectroscopy to examine hard-to-inspect structures for signs of corrosion. Flexible optical fibers with a broad passband were required (conventional optical fibers do not transmit infrared radiation), as well as a very capable image translation system. LCM has somewhat different requirements for sensors than those associated with the manufacturing process discussed in Chapter 3. The most challenging additional sensor needs for in-service monitoring relate to the following three areas: Requirement for very long sensor lifetimes with high reliability; Sensor information displayed in an user-friendly format; and Sensor-product integration. The lifetimes of sensors used in manufacturing operations can be relatively short, and frequent opportunities normally exist for sensor calibration, maintenance, and reliability checks. In contrast, in-service performance-monitoring sensors must have a lifetime that exceeds that of the product; this can translate into tens of years of reliable, maintenance-free use in uncontrolled environments. Despite the use of accelerated aging tests, there is a lack of confidence regarding sensor performance over a lifespan of twenty years or longer. As a result, the development focus of sensor suppliers is beginning to change from one dominated by sensitivity and resolution enhancement to issues of long-term reliability and calibration. However, this change in strategy is not easy to make, particularly considering the fact that the typical major developers and suppliers of sensors are small businesses supporting niche market applications. About half of all commercially available sensors have been fielded within the last five years; thus the long-term reliability track record is incomplete. A major problem encountered in LCM is the lack of field experience in using many of the relevant sensors. Sensors within a manufacturing environment are typically used by skilled personnel in a relatively controlled setting. However, once the product is sold, the sensor readout must be readily interpretable by maintenance technicians who may be relatively unfamiliar with that particular sensor. There is a general concern about avoiding interference of the sensor with the performance of the product. For instance, if the sensor is embedded in the structure, degradation of the structural strength could result (Davidson and Roberts, 1992). The sensor may also become detached during service, reducing its effectiveness or possibly causing total failure of the sensor. Design approaches to integrate the sensor with the product must be developed by detailed analysis, test results, and experience in the service environment. LCM is expected to continue to provide a technology pull for incremental materials development efforts that are needed to solve the myriad of problems that arise as sensors are incorporated into structures that must satisfactorily endure their environment. The need for long sensor lifetimes with high reliability in the operational environment requires sensors that are inherently stable and have very high operational reliability, which may result in sensing solutions that do not directly depend on materials for transduction. An example is turbine temperature sensing. Thermocouple technology is being eclipsed by a completely different sensor technology: one that acts as a blackbody cavity. But this technology also has materials-related limitations that will eventually restrict its application. The high-temperature sensing technology of the future will measure molecular spectra. LCM and smart materials have many similar sensor requirements. Fiber optic sensors, both intrinsic and extrinsic, are very important, and their potential has yet to be fully exploited. Chemical sensors, possibly used in conjunction with fiber optics, have significant potential for the future (see Chapter 6). Piezoelectric polymers, piezoelectric ceramics, and piezoelectric composites offer advantages for strain measurements. Smart materials have the added requirement for actuation, and
OCR for page 60
Expanding the Vision of Sensor Materials these piezoelectric materials can do "double duty." Active and passive taggants placed in structures offer the possibility for low-cost in situ sensing; their potential is only beginning to be explored. In the long term, the full potential of intelligent material systems cannot be achieved unless appropriate sensors are available (Amato, 1992). Existing sensor technology is often less advanced than required for many of the envisioned smart structures applications. However, major advances in control technology are also needed; a discussion of these is outside the scope of this report, but Appendix A contains appropriate references. REFERENCES Life-Cycle Management Davidson, R., and S.S.J. Roberts. 1992. Do embedded sensor systems degrade mechanical performance of the host? Pp. 109–134 in Active Materials and Adaptive Structures. Knowles, G.J., ed. Philadelphia: Institute of Physics Publishing. Gerardi, J., and G. Hickman. 1992. Health monitoring system for aircraft. Pp. 403–406 in Active Materials and Adaptive Structures. Knowles, G.J., ed. Philadelphia: Institute of Physics Publishing. Lockheed Aeronautical Systems Co. 1988. Instrumental Development for the National Aerospace Plane. Executive Summary, Report No. NASP-445, prepared under Contract No. F33657-86-2125. Palo Alto, California: Lockheed. Ramamurthy, S. 1992. Business opportunities in smart materials systems. Pp. 507–512 in Active Materials and Adaptive Structures. Knowles, G.J., ed. Philadelphia: Institute of Physics Publishing. Rotherford, P.S., and E.A. Westerman. 1992. Aircraft structural integrity and "smart" structural health monitoring. Pp. 267–270 in Active Materials and Adaptive Structures. Knowles, G.J., ed. Philadelphia: Institute of Physics Publishing. Turbine Engine Control Bird, V.J., P.C. Sweeny, and P.J. Kirby. 1990. Fiber-optic turbine inlet temperature measurement system. American Institute of Aeronautics and Astronautics (AIAA) Paper 90-2033. Washington, D.C.: AIAA Press. Eckbreth, A.C., and J.H. Stufflebeam. 1985. Considerations for application of CARS to turbulent reacting flows. Experiments in Fluids 3(6):301–314. Hall, R.J., and A.C. Eckbreth. 1984. Coherent Antistokes Raman Spectroscopy (CARS) application to combustion diagnostics. Pp. 213–309 in Laser Applications, Vol. 5. New York: Academic Press. Khalid, S.I. 1994. personal communication to Robert Schafrik, National Materials Advisory Board, February 1994. Skira, C.A., and M. Agnello. 1991. Control systems for the next century's fighter engines. Journal of Engineering for Gas Turbines and Power, Transactions of the American Society of Mechanical Engineers (ASME) 114(4):749–754. Tillman, K., and T. Ikeler. 1991. Integrated flight/propulsion control for flight critical applications. Journal of Engineering for Gas Turbines and Power, Transactions of the American Society of Mechanical Engineers (ASME) 114(4):734–739. Voegtle, W.B., and K.D. Bever. 1992. Helping to reduce turbomachinery losses through advanced technology and on-line expertise. Pp. 76–81 in Turbomachinery International Handbook. Norwalk, Connecticut: Business Journals, Inc. Smart Structures Amato, I. 1992. Animating the material world. Science 255(17 Jan):284–286. Crawley, E.F., and J. de Luis. 1987. Use of piezoelectric actuators as elements of intelligent structures. American Institute of Aeronautics and Astronautics (AIAA) Journal 25(10):1373–1385. Newnham, R.E.. 1993. Smart, very smart and intelligent materials. MRS Bulletin 18(4):24–25. Pollack, A. 1992. Someday Bridges May Have Feelings Too. New York Times, February 16, 1992, p. 6E. Rogers, C.A., and H.H. Robertshaw. 1988. Shape Memory Alloy Reinforced Composites. Engineering Science Reprints 25, ESP25.88027. 25th Annual Technical Meeting, June 20–22, 1988, University of California. Houston, Texas: Society of Engineering Sciences. Wada, B.K., J.I. Fanson, and E.F. Crawley. 1990. Adaptive structures. Mechanical Engineering 112(11):41–46. Optical Fibers for Sensing Claus, R.O. ed. 1991. Proceedings of the Conference on Optical Fiber Sensor-Based Smart Materials and Structures, Fourth Annual Smart Materials and Structures Workshop, Virginia Tech, Blacksburg, Virginia, April 3–4. Lancaster, Pennsylvania: Technomic Publishing Company, Inc. Hickman, G.A., J.J. Gerardi, and Y. Feng. 1990. Application of smart structures to aircraft health monitoring. Pp. 966–984 in Proceedings, 1st Joint U.S./Japan Conference on Adaptive Structures held November 13–15, 1990, in Maui, Hawaii. Lancaster, Pennsylvania: Technomic Publishing Co., Inc. Measures, R.M. 1989a. Fiber Optic Sensors: The Key to Smart Structures. Vol. 1120 in SPIE Conference on Fiber Optics held April 1989 in London. Bellingham, Washington: SPIE Press.
OCR for page 61
Expanding the Vision of Sensor Materials Measures, R.M. 1989b. Smart structures with nerves of glass. Progress in Aerospace Science. 26:289–351. Piezoelectric Materials/Sensors Amato, I. 1989. Fantastic plastic. Science News 136(Nov):328–329. Carlisle, B.H. 1986. Piezoelectric plastics promise new sensors. Machine Design 58(25):105–110. Lovinger, A.J. 1983. Ferroelectric polymers. Science 220(4602): 1115–1121. Tanaka, T. 1981. Piezoelectric devices in Japan. Ferroelectrics 40(3–4):167–187. Piezoelectric Composites Newnham, R.E., D.P. Skinner, and L.E. Cross. 1978. Connectivity and piezoelectric-pyroelectric composites. Materials Research Bulletin 13(5):525–536. Newnham, R.E., 1986. Composite electroceramics. Journal of Materials Science 16:47–68. Newnham, R.E., and G.R. Ruschau. 1991. Smart electroceramics. Journal of the American Ceramics Society 74(3):463–480. Vest, R.W. 1991. Recent developments in piezoelectric composites. Ceramics Information Center Newsletter, Center for Information and Numerical Data Analysis and Synthesis, Purdue University 1(4):1–7. Tagging Clark, W.G. 1991. Material tagging for improved inspection and control. Pp. 233–246 in Proceedings, 23rd International SAMPE Technical Conference held October 21–24, 1991, in Anaheim, California. Covina, California: SAMPE Press. Clark, W.G. 1992. Magnetic particle tagging for improved material diagnostics. Pp. 274–284 in Proceedings, Conference on Recent Advances in Adaptive and Sensory Materials and Their Applications, Virginia Polytechnic Institute and State University held April 27–29, 1992, in Blacksburg, Virginia. Lancaster, Pennsylvania: Technomic Publishing Co. Zhou, S., C. Liang, C., C.A. Rogers, F.P. Sun, and L. Vick. 1993. An in-situ technique for in-service quality monitoring—Measurement of the complex young's modulus of polymers. Pp. 14–23 in Smart Structures and Materials, 1993: Smart Sensing, Processing, and Instrumentation. Claus, R.O. ed. Albuquerque, New Mexico: 1–4 February, 1993. SPIE Proceedings Vol. 1918. Bellingham, Washington: SPIE Press. NOTES 1. In addition to the obvious benefit of being able to use a product longer, there is a secondary benefit of reducing environmental impact by postponing the disposal of the product and the production of its replacement. 2. This approach also facilitates testing of candidate sensors and allows ready insertion of new sensors. 3. Alarm levels are usually first set on the basis of industry standards or manufacturer recommendations. However, these often are adjusted to take the peculiarities of the service environment into account. 4. Fighter aircraft use an augmentor, also known as an afterburner, to provide a very quick increase in thrust. Usually this boost is on the order of 50 percent. The augmentor burns fuel in a separate stage after the turbine, and is not very fuel-efficient. 5. In the event of dual sensor failure, the control's software uses an on-board model of the engine's operation to synthesize the value of the parameter. 6. A National Materials Advisory Board study is under way that addresses materials for high temperature semiconductor devices. Publication is expected by mid-1995. 7. Mach numbers are multiples of the speed of sound. Thus, Mach 1 is the speed of sound, Mach 6 is a velocity that is six times the speed of sound. 8. Optical fiber sensor technology is discussed in Appendix D of this report. 9. A piezoelectric driver supplies an impulse to the structure, and the transducers pick up the transmission through the patch and surrounding structure. 10. For example, a skin-like sensor with temperature-and pressure sensing capabilities can be used in different modes to detect edges, corners, and geometric features or to distinguish between different grades of fabric.
Representative terms from entire chapter: