Fiber optic sensors have the small size, low weight and environmental ruggedness to enable their being embedded directly into composite materials. The resulting smart structures have the potential to greatly enhance the usage of practical composite materials in aerospace applications by reducing manufacturing costs, enabling structural integrity monitoring through nondestructive evaluation and providing sensing capability to augment control systems.

The success of the concept of smart skins depends in large part upon the use of new composite materials. Widespread incorporation of composites into new aerospace vehicles has been limited due to the incomplete predictability of the failure modes and lifetimes of structures made up of these materials. Fiber optic technology offers a possible method whereby a greater understanding and use of these types of materials can be obtained through the use of embedded fiber optic sensors. The types of sensors which could be used will be discussed with a focus upon the benefits that they would provide for smart skins applications.

This paper reviews the smart structures and avionics research and teaching program at Virginia Tech. Started in 1979 with support from the NASA Langley Research Center, this program has grown to include interdisciplinary participation between more than twenty faculty from seven departments in the Colleges of Engineering and Arts and Sciences, graduate and undergraduate engineering classes which specifically address smart structures issues, and directed research programs conducted in cooperation with more than a dozen aerospace companies and government organizations. Current smart structures research includes major efforts in 1) the development of embedded and attached optical fiber and acoustic fiber sensors for a wide range of applications including cure monitoring, in-service lifetime structural monitoring, nondestructive evaluation, and impact and damage detection and analysis, 2) sensor signal multiplexing, processing and data handling to achieve near real time distributed structural analysis, and 3) the combined use of embedded or attached sensors and actuators to achieve controlled structural response. Special campus facilities which have been used for this work include an optical fiber fabrication facility, an autoclave for composite structure fabrication and curing, and laboratories for optical fiber sensor development, materials response and non-destructive evaluation, structural control testing, and computer engineering. The work in all of these areas during the past few years by different Virginia Tech faculty, staff, graduate student and visiting scientist groups is reviewed.

Future aerospace vehicles will incorporate embedded sensor and computer networks to monitor flight loads/environment, structural integrity, and hostile threats and to initiate corrective actions. The corrective actions will consist of real-time reconfiguration of flight controls to ensure mission performance and flight safety. Moreover, post-flight repair and maintenance will be automatically scheduled and guided by the sensor/computer system. This concept, called smart structures/skins, is one of the Project Forecast II technology areas. Smart structures/skins will incorporate a variety of sensor technologies (including fiberoptic sensors), advanced computers and communication/power transmission links. An attractive feature of fiberoptic sensors is that they can be embedded into the structure without affecting the external aerodynamic shape of the vehicle. The present paper defines our smart structures/skins concepts and addresses the role that fiberoptics may play in their success. Emphasis is placed on the capabilities and deficiencies of current fiberoptic sensors. Areas requiring research and development are discussed. Finally, Air Force activities and plans in smart structures/skins for aircraft and space vehicles are summarized.

A fiber optic analogue of the electrical strain rosette is proposed and shown to be necessary if two-dimensional strain fields are to be mapped using fiber optic sensors. The first fiber optic strain rosette has been constructed and preliminary testing demonstrates its ability to determine principal strain magnitudes and axis orientation. Polarimetric transduction was used for this device; the strain sensing region along each arm was localized using 45° fusion splices. This ensured that the strain rosette was lead-in and lead-out insensitive. Possible non-polarimetric designs for optical strain rosettes are also suggested.

An optical fiber embedded between layers of structural fibers in a composite deforms together with a composite structure and modulates light passing through the optical fiber. This effect has been implemented in the form of so-called "smart skins". By analyzing the changes of light transmitted by an embedded fiber dangerous strain levels in the structure as well as failure of the material can be detected. A system of such optical sensors embedded in a structure could act as the nervous system of this structure. Several important questions must be answered before the concept of embedding optical fibers in a composite structure can be commercialized. How does the presence of optical fibers in a composite affect its mechanical performance? How does an embedded fiber affect the strain distribution in its vicinity? What is the relationship between the strain distribution and the layup of the surrounding composite? What can be done to prevent the negative effects of the introduction of an extra fiber into a structure? This work addresses some of these questions. Preliminary results of experimental tests performed on representative composite specimens are presented. A discussion of the results and plans for the future are provided.

Current and projected advances in aerospace structures technology for such programs as the Advanced Tactical Fighter, Orient Express, and NASA space platforms are used as the starting point for a top-level structural monitoring systems analysis. Performance requirements are considered for such structures, and used to identify needs for integral sensors and systems. Implications of recent work are discussed, which have demonstrated the potential for fiber optic sensor systems which can assess and report on structural condition in real time. The essential system elements are defined, and constraints imposed by vehicle environments are considered. The resulting development needs, for fiber optic sensors and systems are stated in the conclusion.

Power by light has been previously demonstrated with photodetectors whose open circuit voltage. V was approximately 1 volt and where the remote system voltage requirements were 1 to 2 volts. This paper explores the power by light cases where Voc is about 0.5 volt (e.g., silicon) and where the voltage requirement is about 5 volts (e.g.. TTL or ECL compatibility). A hybrid power by light/DC-DC converter power supply is proposed for these cases and its performance is compared with the all optical power by light power supply.

In this paper, many of the critical issues which affect the application of embedded fiber-optic sensors for smart structures applications are identified, discussed, and - to a limited extent - experimentally verified. In particular, fabrication, durability, and characterization issues are addressed. The fabrication procedures for embedding these sensors into advanced composite components have been refined. This has been demonstrated with the successful fabrication of a 1.5 inch diameter by 58 inch long composite tube with an embedded fiber-optic pressure sensor. Preliminary characterization results using an optical time-domain reflectometer are presented.

Dynamic loading conditions have been used to characterize a polarimetric fiber optic strain gauge developed for integration into structural control systems. The gauge can be used to provide sensor feedback for dyanmic structural control needs in large composite space structures. The optical sensing technique measures changes in polarization state due to strain induced birefringence in an optical fiber. Dynamic strain measurements were made on a vibrating contilevered beam with the gauge and compared to measurements made on the same system using resistive strain gauges that were colocated. Results over the frequency range of 40-500 Hz show excellent agreement between the two gauges. The experimental data also compared favorably with standard vibrating beam theory.

The use of a photodetector array to process the full far-field output of a modal domain fiber optic sensor is reported. The array is linked to a laboratory computer, which reduces the data in real time and measures the transfer of power between the two lobes of the far-field output of a dual-moded fiber. The power transfer is generated by imparting a longitudinal strain to the fiber. The ratio of powers in the two lobes can be used to determine the magnitude of strain.

A novel class of 'Smart' materials utilizing embedded actuators and sensors has been developed and demonstrated at Virginia Polytechnic Institute & State University. This class of adaptive material with sensing capabilities utilize a shape memory alloy (Nitinol) in a laminated, fiber reinforced composite as the embedded distributed actuators and embedded optical fiber sensors for evaluating the dynamic response of the structure. Shape Memory Alloy (SMA) reinforced composites have the capabilities to: change their material properties, induce large internal forces in the material, modify the stress and strain state of the structure and alter its configuration, all in a controlled manner. The SMA reinforced composite adaptive abilities has resulted in two new concepts for active vibration control; i) Active Strain Energy Tuning, and ii) Active Modal Modification, both of which will be briefly explained in this paper. Results from the experimental demonstration of 'Active Strain Energy Tuning' using the first known composite material containing embedded distributed actuators and sensors will be presented. The variation and controllability of the dynamic response of a beam made of shape memory alloy reinforced composites was determined from a single optical fiber sensor embedded in the material with the distributed actuators. The optical fiber sensing method used was modal domain sensing and the experimental results are compared to accelerometer data. Fast Fourier Transforms (FFT) of both the optical fiber and the accelerometer outputs were performed to compare the two methods for deducing the natural frequency of the structure.

The success of the concept of smart skins depends upon the development of sensors that can be integrated into the vehicle skin and can provide accurate measurement of some parameter of interest. In particular, measurements of distributed parameter fields over the vehicle surface are of prime importance. Work to date has focused upon using a combination of optical fiber sensor and optical radar (OTDR and OFDR) techniques to provide distributed measurements of stress and strain. In this paper, the results of an investigation into the use of integrated optical radar sources/detectors are presented. The work focused upon incoherent OFDR using a laser diode - PIN photodetector combination. The laser diode was driven with an RF "chirp" with its output coupled into an optical fiber. Delayed optical signals returned from the fiber were fed back into the laser cavity and mixed. The integral packaged photodetector was used to provide system output. Such topics as laser coherence length and wavelength stability all influence this technique. Experimental results are presented and compared with measurements based on external photodetection and mixing. Finally, the applicability of the technique for smart skins applications is discussed.

Progress toward the development of a fiber optic damage assessment system for composite materials is reported. This system, which is based on the fracture of embedded optical fibers, has been characterized with respect to the orientation and location of the optical Fibers in the composite. Together with a special treatment, these parameters have been tailored to yield a system capable of detecting the threshold of damage for various impacted Kevlar/epoxy panels. The technique has been extended to measure the growth of a damage region which could arise either from impact, manufacturing flaws, or static overloading. The mechanism of optical fiber fracture has also been investigated. In addition, the influence of imbedded optical fibers on the tensile and compressive strength of the composite material has been studied. Image enhanced backlighting has been shown to be a powerful and convenient method of assessing internal damage to translucent composite materials.

The development of infrared-transmitting optical fibers as sensors for monitoring the cure of advanced composite materials is reported. Fourier transform infrared (FfIR) spectra are presented which were remotely sensed during the cure of a graphite/ polyimide matrix prepreg using a 0.2 mm OD x 3m chalcogenide optical fiber and a graphite/epoxy matrix prepreg using ~0.1 mm OD x 2m fluoride optical fiber. This marked the first time infrared-transmitting optical fibers have been used to sense and transmit infrared spectra of polymeric materials. A discussion of the fiber and sensor element, absorption mechanism and potential applications is presented.

We present recent experimental verification of a new in-situ fiber optic sensors for detecting the curing of thermoset composites. This device is sensitive to the difference between the refractive index of the curing polymer matrix within the composite and that of a fully-cured reference fiber made of the same thermoset material. The method is indifferent to temperature variations imposed during the cure process. The process of polymerization usually results in an increase in the refractive index of a thermoset resin. Increases on the order of one percent are common. In our method, a fiber optic core, comprising the fully-cured neat resin, is placed within, and in direct contact with, the same curing thermoset composite matrix, which functions as a cladding. When the matrix is not fully cured, its lower index of refraction permits waveguiding in the fully-cured core. As the composite is subjected to curing conditions, the matrix cures and the index of refraction difference diminishes. At final cure, the waveguiding characteristics of the fiber optic cure sensor are lost, resulting in a vanishingly small transmitted intensity, at any cure temperature, and without calibration. Our sensor does not suffer from the limitations exhibited by other methods, in large part because it is self-referencing, and relies on a differential rather than an absolute measurement. Our technique is sensitive to the extent of cure even at the latest stage and could be used within the body of large composite structures undergoing a curing process in an autoclave. It is further expected that because of its small size, the sensor may be left in the final product without detrimental effect to the eventual application.

This paper describes the implementation of a distributed fiber optic sensing system for monitoring strain, temperature and absorbed moisture at multiple discrete locations throughout the volume of an organic matrix composite structure. The system employs a network of passive fiber optic sensors fabricated along a continuous single or dual channel waveguide that is embedded within the composite part during manufacture. The single channel sensor networks are interrogated by a conventional Optical Time Domain Reflectometer (OTDR) while the dual channel networks are operated as coupled mode systems. In both cases, physical information is encoded as relative changes in the amplitude of the signal returned from individual sensors. Operational theory, design and experimentally observed characteristics of each of the sensor types as well as the overall system are presented. Distributed sensors capable of measuring strain in the range of 1.8 microstrain to tens of millistrain, temperature from zero to 170°C with one tenth of a degree C resolution and absorbed moisture from zero to two weight percent at up to ten discrete locations have been achieved. Theoretical extrapolation of existing data indicate that as many as 140 sensors can be fabricated along a single sensor fiber.

We review the performance of a microbend fiber optical time domain reflectometry (OTDR) based distributed strain sensor. A testbed has been developed to evaluate and compare this system to conventional discrete strain gauge sensors. The testbed consists of an aluminum cantilever beam to which have been attached several pairs of strain gauges (for reference) and several microbend transducers. A single standard multimode optical fiber is threaded through the microbend transducers. The fiber is pulse illuminated by a commercial OTDR instrument. The microbend losses are detected as step changes in the returned OTDR signal wave form. The transducers consist of small pairs of opposing sets of teeth, one set mounted on the flexible beam testbed, the other machined into a small, rigid cantilever bar mounted to produce scissor action as the beam flexes. The transducers are biased to produce a step loss with the beam in its neutral position. The direction of beam deformation is then determined by the increase or decrease in the step size. Several transducers along the beam, incorporating the same fiber, permit the acquisition of strain measurements from multiple locations on the structure by processing a single OTDR wave form. The waveform is obtained as the logarithm of the Rayleigh backscattered intensity. The change in the step response of the logarithmic signal caused by each transducer is linear in the in-plane strain produced by displacing the tip of the beam. Repeatability, hysteresis and sensitivity are dependent on the nature and thickness of the fiber coating, signal processing, transducer design, and other less significant factors. We have demonstrated a resolution of 2-3 microstrains over a range of 120 microstrains using transducers with 2 mm pitch in tooth spacing and a 2.5 cm interaction length, mounted on cantilever bars 5 cm in length, and averaging multiple wave forms to improve signal-to-noise ratio.

Understanding of critical variables in coating properties and fiber design will allow optimization of sensing and communication performance in structurally embedded applications. The severe environmental requirements for optical fibers in smart-skin applications requires advanced research and development in fiber/coating systems. Several approaches to high temperature, thin coating systems are discussed. Optical fiber sensing systems further require a wide range of ability to couple or de-couple external perturbations from propagation characteristics. The effects of coating physical properties, such as modulus and cross-link density on this coupling are investigated.

Single mode couplers have been utilized as the basis for thermally switched sensors. Fabrication of the switches is described, and the response of the sensors to temperature changes. Wavelength-dependent operation of the devices is also discussed.

Fiber optic sensors embedded in composite aircraft skins offer the capability of detecting physical parameters with extremely high sensitivity and bandwidth in unique "intralaminar" locations. The ultimate extent of usage of these sensors will be determined by the cost savings that they will provide through the availability of the additional sensed information versus life cycle costs in terms of producibility and maintainability. In the manufacturing process, fiber optic sensor "lay-in" must be compatible and integrable with the overall automatic or robotic manufacture of composite aircraft parts and should include automatic calibration and checkout. In addition, it would be highly desirable for these sensors to be used in cure monitoring and optimized control of processing. Over the lifetime of the aircraft, maximum sensor array utility should be achieved by multiple uses such as structural integrity monitoring, multi-color, low and high speed data buses, and new types of air data sensors. These sensors should have a reliable interconnection system and should be designed with adequate redundancy so that graceful degradation of the sensor array will be achieved. Maintenance policies and procedures for repair must be developed for dealing with failures in the fiber optic sensing system due to malfunctions and/or damage to the composite skins, e.g. bird strikes and combat damage. This paper discusses these aspects and others needed to yield a cost effective smart skin system.

When a single-mode fiber is operated at a wavelength shorter than its nominal wave-lenght, few lower-order modes are excited and an interference pattern develops which is sensitive to differential phase variations of the modes. Intensity patterns resulting from the interference of the pairs of modes LPoi, LP11 and LP11, LP21 are examined. By proper choices of phase and amplitude of the interfering modes, one,two and three-lobe patterns can be synthesized which exhibit oscillatory or rotational behavior in response to an external disturbance. Launching conditions for achieving such patterns are discussed. Sensor applications of few-mode fibers, particularlyfor strain and temperature sensing, are addressed.

Interference between the modes of an optical fiber results in specific intensity patterns which can be modulated as a function of disturbances in the optical fiber system. These modulation effects are a direct result of the difference in propagation constants of the constituent modes. In a companion paper, a detailed theoretical analysis of the modal phenomena responsible for the far field intensity pattern modulation of various few-mode (V 5.0) modal domain sensor systems is presented [1]. This paper will present experimental results which verify the expected system behavior and which serve as proof of the significance of the modal domain sensing technique in general. The results of both static and dynamic strain measurements are presented, as are data which demonstrate the strain sensitivities achievable with modal domain sensor systems. System linearity and dynamic range are also considered.

Several planned United States Air Force (USAF) and National Aeronautics and Space Administration (NASA) space systems such as Space Based Radar (SBR), Space Based Laser (SBL), and Space Station, pose serious vibration and control issues. Their low system mass combined with their large size, precision pointing/shape control and rapid retargetting requirements, will result in an unprecedented degree of interaction between the system controller and the modes of vibration of the structure. The resulting structural vibrations and/or those caused by foreign objects impacting the space structure could seriously degrade system performance, making it virtually impossible for passive structural systems to perform their missions. Therefore an active vibration control system which will sense these natural and spurious vibrations, evaluate them and dampen them out is required. This active vibration control system must be impervious to the space environment and electromagnetic interference, have very low weight, and in essence become part of the structure itself. The concept of smart structures meets these criteria. Smart structures is defined as the embedment of sensors, actuators, and possibly microprocessors in the material which forms the structure, a concept that is particularly applicable to advanced composites. These sensors, actuators, and microprocessors will work interactively to sense, evaluate, and dampen those vibrations which pose a threat to large flexible space systems (LSS). The sensors will also be capable of sensing any degradation to the structure. The Air Force Astronautics Laboratory (AFAL) has been working in the area of dynamics and control of LSS for the past five years. Several programs involving both contractual and in-house efforts to develop sensors and actuators for controlling LSS have been initiated. Presently the AFAL is developing a large scale laboratory which will have the capacity of performing large angle retargetting manuevers and vibration analysis on LSS. Advanced composite materials have been fabricated for the last seven years, consisting mostly of rocket components such as: nozzles, payload shrouds, exit cones, and nose cones. Recently, however, AFAL has been fabricating composite components such as trusses, tubes and flat panels for space applications. Research on fiber optic sensors at NASA Langley Research Center (NASA LaRC) dates back to 1979. Recently an optical phase locked loop (OPLL) has been developed that can be used to make strain and temperature measurements. Static and dynamic strain measurements have been demonstrated using this device.' To address future space requirements, AFAL and NASA have initiated a program to design, fabricate, and experimentally test composite struts and panels with embedded sensors, actuators, and microprocessors that can be used to control vibration and motion in space structures.

A method for producing rugged, continuous porous glass fiber optic sensors was developed. pH and temperature sensors based on this technology have been successfully produced. The sensor portion of the fiber is made porous by selective leaching of a specially formulated borosilicate glass fiber. This results in a strong, monolithic structure where the sensor portion of the fiber remains integrally attached to the rest of the fiber (which acts as a light pipe), essentially eliminating losses at the sensor-light pipe interface. Pore size in the sensor can be controllably varied by modifying heat treatment conditions, making these sensors suitable for chemical concentration measurements in liquids and gases. Appropriate dyes were chemically bonded by silanization to the large interior surface area of the porous sensors to produce the pH and temperature sensors. Cresol red and phenol red were used for pH and pinacyanol chloride was used for temperature sensing. The sensitivity of these devices can be controlled by varying the concentration of the chemically bonded dye and the length of the porous region. Optical absorbance measurements were made in the visible range. The tip of the sensors was coated with a thin, porous layer of gold to reflect the incident light, resulting in a double pass across the porous sensor. Experimental measurements were made over a pH range of 3 to 8 and a temperature range of 28-70 C. These porous glass fiber optic sensors were found to be rugged and reliable due to their monolithic structure and large interior surface area for attachment of active species. A broad range of sensors based on this technology could be developed by using different active species, such as enzymes and other biochemicals, which could be bonded to the interior surface of the porous glass sensor.