Our previous work has shown that elastic textiles covered with an epitaxial layer of conducting polymer show piezoresistive properties. They can be used to fabricate sensorized garments such as gloves, leotards, socks and seat covers as man-machine interfaces. A purposely designed screen printing process has been implemented to realize sensors/tracks patterns. Polypyrrole/lycra fabrics were prepared using the method developed by Milliken Co. (Spartanburg, USA). The epitaxial deposition is obtained controlling the concentrations of monomer and the temperature of the reaction medium. Investigation on mechanoelectric transduction properties (static and dynamic) of the fabrics, calibration of wearable sensing devices and ongoing R&D efforts in multimedia, sport and rehabilitation fields are reported.

Many of the sensor technologies considered for integration into 'smart' materials systems have been adopted from a host of other applications. Consider, for example, fiber optics; these materials were originally designed for telecommunication and have since been modified to serve as effective sensor elements. The present research, by contrast, has focused on the development of a new generation of sensor specifically designed to meet a range of critical operational constraints. The sensor consists of conductive particles selectively distributed in a non-conductive matrix. The matrix could be an elastometer or the host matrix used in a polymer composite. Selective distribution of the particles allows for unprecedented amplification of the sensor signal, thus augmenting its ability to detect relatively weak disturbances in the host structure. The signals exhibit both analog and digital-like response characteristics, and hence they have been called 'fuzzy' sensors. The advantages of this approach include a practical sensor element that exhibits very low cost (less than 10 centers per sensor), structural compatibility with the host structure, and high sensitivity. The fuzzy sensors have been successfully implemented in a structural composite component and a real time data retrieval system has been developed.

A linear array of embedded optical fibers was utilized to produce an ultrasonic beam in carbon composite plates as a potential tool for structural assessment. The embedded fibers were used to deliver laser pulses from low power (1.5 W) modulated laser diode sources for the generation of the ultrasound. These were modulated with a pseudorandom binary sequence, which was used in conjunction with a correlation based detection technique using interferometric fiber optic sensors. Furthermore time delays could be used to adjust the excitation of the elements in the array, thus enabling the direction of the generated ultrasonic beam to be controlled.

Ultrasonic Lamb waves have been extensively investigated for non-destructive testing of materials. Embedded or surface bonded optical fiber, acting as the signal arm of a Mach- Zehnder interferometer, is one method previously utilized to detect the Lamb waves. Optical fibers therefore have potential as permanent sensors for structural monitoring of damage and defects in materials. A greater understanding of the ultrasound interaction with the optical fiber sensor will bring application closer. In order to probe this interaction we built a two channel interferometer allowing ultrasound traveling through a material to be monitored simultaneously by a Mach-Zehnder interferometer and also a Michelson interferometer. The Michelson interferometer allows a non- constat measurement to be made of the absolute surface displacement associated with an ultrasonic Lamb wave. Comparison of the ultrasound signals detected by the two different interferometers provides a greater insight into the detection mechanism and sensitivity of the Mach-Zehnder interferometer. The work is then extended to look at embedded fibers in composite materials and damage detection.

The electrostatic self-assembly (ESA) process was utilized to fabricate a novel optical fiber humidity sensor based on a nano interferometric cavity. A wide operation range, from 11.3% to 100% relative humidity with a maximum variation of 4.77 dB, was experimentally demonstrated. Due to the short length of the interferometric cavity, less than 400 nm, an LED was used as the light source instead of a laser. The fast response time of this humidity sensor, less than 1.5 seconds, makes it possible to monitor human breathing.

White light interferometric sensing systems are very attractive for application in smart structures due to the fact that they read out the absolute path differences from the transducers so that calibration schemes and device history do not have to be used and that they can be serially multiplexed. The basic output from such systems consists of arrays of data containing complex interference patterns. In this paper, a new technique is presented in which the signal from one or more white light interferometers can be demodulated with high resolution. In the technique, the conventional demodulation approach of a Fizeau wedge and linear detector array is replaced by a rotated wedge and a 2-D array such as a CCD. This results in a sampling resolution equal to the linear sampling number of points squared. A theoretical description of the technique is given and planned experimental validation is described.

We present a novel class of optical fiber sensors based on fiber Bragg gratings written in germanium-doped silica multimode fibers that can be implemented with inexpensive components. The gratings are inscribed by the phase mask technique in 50/125 μm step-index profile fibers and their temperature and strain responses are investigated.

A powerful, innovative diagnostic technique for assessing railcar wheel health characteristics is to listen to the operating sound of a bearing. The emission of acoustic energy (sound) from a wheel bearing is a frequency-dependent and load-related phenomenon. A complementary set of bearing health data to audible acoustic emission (AE) data, referred to as 'stress-wave' data, exists at an order of magnitude above traditional vibration-based (i.e., accelerometer) data and contains information about friction and shock conditions in a bearing under highly loaded conditions. These health-related symptoms of shock and friction will provide an early warning capability to prevent flat wheel failures and train derailments through condition-based maintenance (CBM). This paper describes Honeywell's recent work with applying stress- wave AE to perform CBM for today's railroad industry. A detailed description of the railcar wheel bearing problem, failure modes and effects, and technical approach are presented. The summary of a field-test effort capturing AE signatures from a 100-ton railcar truck are provided. A preliminary set of data analysis is presented.

Reductions in Navy maintenance budgets and available personnel have dictated the need to transition from time-based to 'condition-based' maintenance. Achieving this will require new enabling diagnostic technologies. One such technology, the use of acoustic emission for the early detection of helicopter rotor head dynamic component faults, has been investigated by Honeywell Technology Center for its rotor acoustic monitoring system (RAMS). This ambitious, 38-month, proof-of-concept effort, which was a part of the Naval Surface Warfare Center Air Vehicle Diagnostics System program, culminated in a successful three-week flight test of the RAMS system at Patuxent River Flight Test Center in September 1997. The flight test results demonstrated that stress-wave acoustic emission technology can detect signals equivalent to small fatigue cracks in rotor head components and can do so across the rotating articulated rotor head joints and in the presence of other background acoustic noise generated during flight operation. This paper presents the results of stress wave data analysis of the flight-test dataset using wavelet-based techniques to assess background operational noise vs. machinery failure detection results.

Acoustic Emission signals (AE), generated by the formation and growth of micro-cracks in metal components, have the potential for use in mechanical fault detection in monitoring complex- shaped components in machinery including helicopters and aircraft. A major challenge for an AE-based fault detection algorithm is to distinguish crack-related AE signals from other interfering transient signals, such as fretting-related AE signals and electromagnetic transients. Although under a controlled laboratory environment we have fewer interference sources, there are other undesired sources which have to be considered. In this paper, we present some methods, which make their decision based on the features extracted from time-delay and joint time-frequency components by means of a Self- Organizing Map (SOM) neural network using experimental data collected in a laboratory by colleagues at the Georgia Institute of Technology.

The feasibility of an integrated system for permanent detection of damaging impacts on composite structures has been evaluated. This system is based on the existence, during the damaging impact, of an intense acoustic emission in the high frequency range. This acoustic emission is registered by a network of piezoelectric sensors. Experiments are carried out with carbon-epoxy plates equipped with four miniaturized disc- shaped piezoelectric sensors. Both fully embedded and surface positioned sensors are used. Each plate is impacted using a weight drop machine equipped with a Boeing window. The impact energy varies from 2 J to 6 J for 16 ply coupons and from 4 J to 8 J for 32 ply coupons. The procedure used to identify the damaged area as a noise source from the signals received by the sensors allows localizing them with an accuracy of a few centimeters. It also allows to correlate the <<amplitude>> of the identified noise source to the area of the damage. It is then possible to estimate the damaged area from the amplitude of the noise source.

Ultrasonic Lamb waves have been investigated extensively for damage detection in advanced composite materials. They are particularly suitable for proving thin plate structures of large area, where the Lamb wave approach offers a considerable saving in time over through-the-thickness inspection. However the potential complexity of the propagation can introduce significant difficulties to the technique. We present a review of work conducted at The University of Strathclyde in collaboration with several European partners into the feasibility of Lamb wave inspection. Specifically we will address issues of Lamb wave generation, propagation, defect interaction and detection.

Optical fiber sensor systems have been recently used for ,health monitoring of fiber reinforced plastic (FRP) composite laminate structures. However, the diameter of optical fibers for optical communication is normally 125 μm, which is similar to the thickness of one ply in FRP composite laminates. In order to embed optical fibers within one ply of FRP without introducing any defects in laminates, small- diameter optical fibers are necessary. The present paper describes development of small-diameter optical fiber sensors for damage detection in composite laminates. The coating diameter of the prototype small-diameter optical fiber is 52 μm, while the diameters of the cladding and the core are 40 μm and 8.5 μm, respectively. FBG (Fiber Bragg Grating) sensors were also fabricated with these fibers. Gratings were inscribed into the optical fiber by ultra-violet irradiation method. The wavelength shift and the response by strain of FBG itself were measured. The strain response, when small-diameter optical fiber and FBG sensors were embedded in composite laminates, were evaluated for practical application to detect the damage evolution.

It is well known that the compression after impact (CAI) strength of carbon fiber reinforced plastic (CFRP) laminates decreases by impact damage, especially delamination. The impact damage has a close relation to impact energy, which can be derived from the time history of impact load. Thus, it is important to detect the impact load applied to the composites. In this study, single-mode or multi-mode small-diameter optical fibers embedded in CFRP laminates were used as a sensor for detecting the impact load. Diameters of the cladding and the polyimide coating are 40 μm and 52 μm, respectively. Such optical fibers embedded inside laminas cause no serious effect on the mechanical properties of composites. The optical fiber sensors were able to detect the impact by bending loss in the vicinity of impact point. The optical fibers were embedded parallel to reinforcing fibers in CFRP composites. Charpy impact tests were performed for the CFRP specimens. The strain on the surface of the specimens, the optical loss and the impact load were measured as a function of time. Then, the relationship between the optical loss and the impact load was discussed experimentally and theoretically.

We developed a surface mounting technique where fiber-optic Bragg grating (FBG) sensors are glued to the surface of structures and tested the technique on the surface of a CFRP- wing at the DASA Airbus test center Hamburg for over one year. The FBG sensors were interrogated with a measurement system capable of determining the Bragg wavelength in a few seconds over a spectral range of 60 nm (around 1.53 μm) with an absolute accuracy better than 1 pm. A polarization scrambler was used to account for polarization effects. Excellent consistence between the values of electrical strain gauges and the FBG sensors were found during all measurements. However because this method shows drawbacks in a harsher environment such as a flight test, we are currently investigating the possibilities of integrating FBG sensors into the varnish of the structures. For reasons of their better mechanical performance we use FBG sensors produced on the fiber draw-tower with a special UV-curable coating. The sensors are integrated into an original Airbus varnish build- up. We observed linear strain sensitivities in a temperature range between -50 and +100 °C. Furthermore, at negative temperatures we found a vanish- induced polarization dependence which could be used to differentiate between strain and temperature effects.

In this paper embedding of surface mount pressure and temperature sensors in the Carbon fiber composites are described. A commercially available surface mount pressure and temperature sensor are used for embedding in a composite lay- up of IM6/HST-7, IM6/3501 and AS4/E7T1-2 prepregs. The fabrication techniques developed here are the focus of this paper and provide for a successful embedding procedure of pressure sensors in fibrous composites. The techniques for positioning and insulating, the sensor and the lead wires, from the conductive carbon prepregs are described and illustrated. Procedural techniques are developed and discussed for isolating the sensor's flow-opening, from the exposure to the prepreg epoxy flow and exposure to the fibrous particles, during the autoclave curing of the composite laminate. The effects of the autoclave cycle (if any) on the operation of the embedded pressure sensor are discussed.

In this paper we present the results of our experiment for the characterization of mechanical vibrations, using a multichannel interferometer with sensing optical fibers embedded into a carbon fiber board. We investigate the application of this measurement system for dynamic gradients of vibrations at medium frequency (50 Hz to 1 kHz). The optical phase is biased to operate at quadrature conditions (optical phase = π/2) in order to get the maximum sensitivity. We measure the differential optical phase of four spatially separated embedded fiber-optic channels relative to a common reference path. The tests are the calibration of the sensor response to vibrations and the spatial propagation of vibrations. This scheme provides an improvement of previous characterizations with strain gages at the surface because of the high sensitivity of the sensor, better spatial resolution (separation of the sensing regions up to 5 mm), and measurements performed inside the board. The signal processing required for phase measurements is also analyzed, as well as it is presented a comparison with the electronic measurements at the surface.

The need to obtain information on the performance and lifetime of a tool in service is of prime importance to many industries. It calls for on-line acquisition of information such as temperature and strain values from tools and structures. With embedded sensors, structures are capable of monitoring parameters at critical locations not accessible to ordinary sensors. To embed sensors in the functional structures, especially structures, Shape Deposition Manufacturing (SDM) is a methodology capable of integrating sensors during the production of tooling or structural components. Thin film sensors and fiber optic sensors have been identified as two promising candidates to be integrated in metallic structures. Embedded thin film strain gages have been characterized in a four-point bending test and the results, showing linearity and no hysteresis, match with those from the theoretical model and commercially available strain gages. Fiber optic sensors have been successfully embedded in nickel and stainless steel structures. The embedded fiber optic sensors have been used to measure temperatures and strains. They provide higher sensitivity, good accuracy, and high temperature capacity. Based on fiber optic sensor embedding techniques, a remote temperature/strain sensing system suitable rotating objects, such as turbine blades, has been developed. The developed techniques can be harnessed for rapid prototyping of smart metallic structures.

An optical waveguide sensing method is proposed for monitoring the corrosion of steel in reinforced concrete structures. The method depends on the absorption of light in the waveguide by a metallic cladding which is applied at a particular location. When this cladding is exposed to a corrosive environment, it degrades, and the optical waveguide output increases. This sensing principle is analyzed, as is an electrochemical method proposed to lay the metal film onto the waveguide. In order to establish a reliable procedure for electroplating an Fe-C alloy film onto fused silica, we first investigated the use of a planar waveguide substrate, and electroless plating of silver onto SiO₂ was introduced as the electroplating cathode. After analysis, an optical fiber version of the corrosion sensor was then fabricated using the electrochemical method. The sensor was intensively corroded with HNO₃, NSS, and simulated concrete pore solutions. The experimental results are presented and to some degree coincide with theory.

It is essential to study the dynamic behavior of the soil to make clear the characteristics of ground behavior during earthquake. However, the relationship between the dynamic characteristics and the strain of the soil is not completely studied, because there is no device to measure the strain of the real ground directly. Therefore, it is necessary to develop the ground strain measuring system which can be applied to the real ground. This study presents a system to measure the ground strain, using the fiber Bragg grating (FBG) sensors. Using optical fiber sensor makes the devices simple in mechanism and highly durable. We improve the strain measuring device which was proposed by Sato et al. and also develop a new strain measuring device based on a different mechanism. Their applicability is studied in the experiments. The results in the experiment indicate that it is possible to measure the ground strain by the presented systems with the same level of accuracy as that of the systems by Sato et al. It is also important to recognize the necessity to improve the accuracy.

The electrical characteristics of fiber reinforced plastics (FRP) composites have been investigated in order to develop the self-diagnosis function suitable for health monitoring of structural materials. The electrical conductivity was achieved by adding carbon particles or fiber as a conductive phase into FRP. The self-diagnosis function of the composites was evaluated by the measurement of change in electrical resistance as a function of stress or strain in tensile tests. The resistance of carbon fiber in the composite slightly changed at a small strain level and increased nonlinearly with the applied stress due to the fracture of carbon fiber. The conductive FRP composite containing carbon particles had high sensitivity and linear response of the resistance in a wide strain range. In the cyclic loading tests, the phenomenon of residual resistance was observed at an unloading state in the composites with carbon particles. The residual resistance increased with an applied maximum strain, showing that the composite with carbon particles possesses the function to memorize the applied maximum strain or stress. These results indicate that the FRP composite containing carbon particles has a promising possibility for simple diagnosis of dynamic damage and for damage hysteresis with high sensitivity.

Fiber optic sensors have already demonstrated to be suitable for the monitoring of full-scale structures. Usually, most effort has been concentrated in improving the different sensing techniques and it remains often the problem of adapting the sensor to the harsh and dusty environment of the building site. In this context, the SOFO deformation sensors have successfully been tested in different types of structures. The current measurement system, based on low- coherence interferometry, is specially conceived for the long- term assessment of structures, where high precision, resolution and stability is required. Since each measurement takes a few seconds, the set-up is not adequate to monitor the dynamic behavior of structures, where measurement frequencies of up to 1 KHz are necessary. In order to take advantage of the good performance of the current SOFO sensors, a new technique to demodulate at a high frequency the signal from these sensors has been investigated. Based in the radio- frequency amplitude modulation of a low-coherence source, this intensity-based method offers the required dynamic range of several mm to measure the deformations of the structure under dynamic loads. In this paper we present the theoretical background together with the experimental verification of the principle.

This work reports two modeling and control attempts performed on a positioning system comprising of linking SMA wires and an overlooking video system for on-line measurements. The first attempt takes the model by Ikuta and identifies experimentally the parameters of the SMA wire. The identified single wire model is then extended to a system of two SMA wires joining together at their tips, based upon which open loop position control of the linkage is then conducted. The approach, however, becomes too complicated when more SMA wires are involved. The second attempt utilizes a neuro-fuzzy based approach for positioning control of a linkage point joining together four SMA wires. The second approach involves four ANFIS neuro-networks with hybrid learning algorithm trained to model the currents to the SMA wires as functions of present and target positions of the linkage point. Experimentation for both the two-wires and four-wires system yield quite satisfactory performance.

The modal transfer matrix method (MTMM) is a new general- purpose technique developed by the authors as an alternative to coupled mode theory (CMT) to study complex optical waveguide systems. It is applied here to overcome the numerical accuracy limitation of the beam propagation method (BPM) when simulating long-period gratings (LPG). Simulation of a LPG optical fiber sensor with a broad-wavelength source using the MTMM is compared with analytical results based on CMT. A simple grating parameter study shows that both methods are in agreement, indicating that the MTMM is a promising method of analysis. In addition, the results reveal the potential effect of combined variation in grating amplitude, spatial wavelength, and length on sensor performance. By discretizing the waveguide into small elements, the MTMM reduces the problem domain size and makes use of proven electromagnetic numerical simulation methods. An appropriate numerical method, in this case the BPM, is selected to compute each element's modal transfer matrix. The elements are then combined to predict the output from the overall waveguide system.

This paper studies the correlation between mechanical deformation, damage behavior and the change of electric resistance for unidirectional CFRP and CFGFRP (carbon fiber/glass fiber reinforced plastics) composites. This study is performed using experimental and numerical methods, and the results of two methods are compared. The change of electrical resistance is measured under the condition of simple tension and repeated loading-unloading, and the failure process of two types of CFRP is investigated using an optical microscope. A Monte Carlo simulation is performed to predict the change in electrical resistance due to strain and damage (fracture) of carbon fibers. Through these studies, it is revealed that the change of electrical resistance had a close relation with the damage process and there is a good agreement between experimental and predicted results. It is also suggested that this detecting technique is applicable to the health- monitoring of composite structures.

Embedded optical fibers and sensors must survive and remain functional for the lifetime of the structure being monitored, as repairs are generally impossible. Thus, the feasibility of an embedded optical fiber monitoring concept depends heavily on the durability of the optical fiber. Processes that degrade the mechanical properties of these fibers are therefore of great concern. During the process of writing a Bragg grating sensor in a fiber, the polyimide coating is damaged locally by ablation, making the fiber vulnerable to moisture degradation. To rectify this situation, the coating in the area around the grating is commonly removed and the fiber is recoated. However, this procedure itself makes the fiber susceptible to degradation by moisture and handling. Tensile experiments were conducted on both virgin fiber and on fibers that had been recoated to study deterioration related to the recoating process. Weibull theory was used to model the strength distributions and a fracture mechanics approach was used in conjunction with microscopy to study failure initiation and to evaluate the relative significance of coating defects. The results indicated that two independent flaw populations existed in the fibers, one associated with manufacturing defects and the other with inherent flaws on the surface of the glass fiber. The failure was always initiated on the glass surface, not in the coating, and the condition of the coating did not effect the failure location. The recoated fibers always failed in the recoated section at a significantly reduced load, due to degradation after exposure of the glass to the environment. This suggested that the recoating process might actually worsen the situation.

Fiber optic grating sensors written into polarization preserving optical fiber may be used to monitor multidimensional strain fields in composite materials. This paper provides an overview of the characterization and test of multiaxis fiber grating sensors formed by writing 1300 and 1550 nm fiber gratings into polarization preserving optical fiber. A discussion of the usage of these multiaxis fiber grating sensors to measure two and three dimensional strain fields will be made. A brief review of practical applications of the technology to measure shear strain, transverse strain gradients as well as axial and traverse strain will be made with emphasis on aerospace and civil structure applications.

We report on application tests of novel sensor elements for long term surveillance of tunnels. The sensors are made of glass fiber reinforced polymers (GFRP) with embedded optical fiber Bragg gratings. The tests were made in a tunnel near Sargans in Switzerland and we will present strain and temperature data of more than one year of operation of the sensor elements. Two sensor types were tested. First, GFRP rockbolts with a diameter of 22 mm were produced. They have a load-bearing function as anchors for tunnel or mine roofs and in addition measure distributed strain fields and temperature with embedded optical fiber Bragg grating arrays. Rockbolts are key elements during construction and operation of tunnels. Data about strain inside the rockbolts can support decision about precautions to be taken and reveal information about the long term movement of the rock. Second, thin and flexible GFRP wires of 3 mm in diameter were found to be robust and versatile sensors not only for tunnel surveillance but for many civil engineering applications where they can be attached or embedded (e.g., in concrete). The fabrication of both sensor types and solutions for the connection of the embedded fiber sensors to a fiber cable will be presented. Moreover, laboratory and tunnel data of functionality and long term stability tests will be discussed and compared.

The single peak of the spectrum of a fiber Bragg grating written in a standard low birefringence optical fiber splits in two peaks when a transverse strain field is applied due to the promotion of a strain induced birefringence. In a composite laminate, the theory predicts a plane stress state, which allows using a single embedded Bragg grating to obtain simultaneously both components of the strain field. After a brief review of the photoelastic theory, and the experimental verification of the suitability of the plane stress state hypothesis, some experiments are done with thick graphite fiber composite laminates with different lay-up sequences. Strong residual stresses are detected inside the laminate, in accordance with the theory of composite laminates. The response of embedded Bragg gratings must take into account these internal stresses in order to adequately interpret the experimental results.

In the present research, fiber Bragg grating (FBG) sensors were applied for the detection of transverse cracks, which cause strain distribution within the gage length, in carbon fiber reinforced plastic (CFRP) cross-ply laminates. An uncoated FBG sensor was embedded in 0° ply on the border of 90° ply in a CFRP cross-ply laminate. The reflection spectra from the FBG sensor were measured at various tensile stresses. As a result, the reflection spectrum became broad and had some peaks with increase of the transverse crack density in the 90° ply. After the crack density was saturated, the spectrum became narrow and had one large peak again. For confirming that the change in the spectrum form was caused by transverse cracks, the spectra were calculated theoretically. The calculated result reproduced the change in the measured spectrum form very well. These results show that the occurrence of transverse cracks can be detected from the change in the form of the reflection spectrum, and the spectrum width at the half-maximum is a good indicator for the quantitative evaluation of the transverse crack density on real-time.

In this paper, we present a new fiber optic Bragg grating sensor for petroleum hydrocarbon leak detection. The developed sensor includes fiber section with imprinted in the fiber core Bragg grating covered by special polymer material. This polymer reversibly swells under hydrocarbon influence and strains the fiber section with the Bragg grating inside. As a result of the fiber elongation, the Bragg wavelength shifts to aside longer wavelength. Experimentally demonstrated shift of the grating resonant wavelength was more than 2 nm for 20-min gasoline influence, which significantly exceeds a shift due to possible environmental temperature variation and the width of the grating reflection spectrum, which was about 0.5 nm. The paper presents also the results of the swelling-behavior test of the loaded and unloaded sensitive polymer material under liquid and vapor gasoline influence and the results of the theoretical and experimental investigation of the hydrocarbon sensor performance.

Fully-distributed optical-fiber sensing (FDOFS) systems are developing rapidly and are offering significant advantages for measurement functions in a variety of structural applications, especially in the oil industry, the power supply industry and the aerospace industries. Polarization techniques are well established in FDOFS, and in PMD analysis for optical-fiber telecommunications. However, a major problem has been that of determining, in backscatter, the full polarization properties of a monomode optical fiber, as a function of position along the fiber, with some specified spatial resolution. This paper will present a new technique for providing this full information, and thus for measuring the distribution of any parameter, external to the fiber, which can modify its polarization behavior. As a result, for example, it becomes possible to measure simultaneously the distributions of temperature and of the full strain field, comprising, in the latter case, the longitudinal and the two transverse components of strain, plus the shear (or differential transverse) strain. Magnetic field and electric field measurements also become more readily accessible. The technique comprises a conceptual extension of POTDR, and necessitates on-line processing. Details of the physical principles, the algorithms and the polarimetry will be presented, together with results illustrating the measurement accuracies which can be achieved. The potential for further development, and application to both distributed sensing in smart structures and to PMD diagnostics in optical telecommunications, will be reviewed.

In this study, we developed a health monitoring system using a fiber optic distributed strain sensor for International America's Cup Class (IACC) yachts. Most structural components of an IACC yacht consist of an aluminum honeycomb core sandwiched between carbon fiber reinforced plastic (CFRP) laminates. In such structures, delamination, skin/core debonding and debonding between adhered members will be result in serious fracture of the structure. We equipped two IACC yachts with fiber optic strain sensors designed to measured the distributed strain using a Brillouin optical time domain reflectometer (BOTDR) and to detect any deterioration or damage to the yacht's structures caused by such failures. And based on laboratory test results, we proposed a structural health monitoring technique for IACC yachts that involves analyzing their strain distribution. Some important information about structural conditions of the IACC yachts could be obtained from this system through the periodical strain measurements in the field.

We constructed a new health monitoring system to detect damage using a fiber optic distributed sensor, namely a Brillouin optical time domain reflectometer (BOTDR), and installed it in International America's Cup Class (IACC) yachts, the Japanese entry in America's Cup 2000. IACC yachts are designed to be as fast as possible, so it is essential that they are lightweight and encounter minimum water resistance. Advanced composite sandwich structures, made with carbon fiber reinforced plastic (CFRP) skins and a honeycomb core, are used to achieve the lightweight structure. Yacht structure designs push the strength of the materials to their limit and so it is important to detect highly stressed or damaged regions that might cause a catastrophic fracture. The BOTDR measures changes in the Brillouin frequency shift caused by distributed strain along one optical fiber. We undertook two experiments: a pulling test and a four point bending test on a composite beam. The former showed that no slippage occurred between the optical fiber glass and its coating. The latter confirmed that a debonding between the skin and the core of 300 mm length could be found with the BOTDR. Next we examined the effectiveness with which this system can assess the structural integrity of IACC yachts. The results show that our system has the potential for use as a damage detection system for smart structures.

Novel optical fiber sensor architecture has been developed. The actual element of the sensor is highly curved multimode fiber. However, the feed to the multimode fiber is through a single mode fiber to ensure that only the lowest order spatial mode is launched. Similarly the receiver is also coupled to the sensing element through a single mode fiber. The fundamental mode within graded index multimode fiber proves to be very insensitive to macrobends, if bend radius is larger than certain critical value. If bend radius is reduced below critical value the loss increases very rapidly and this allows for construction of relatively sensitive macrobend fiber optic sensor. In this paper we describe a quantitative theoretical model and a corresponding experimental investigation of the proposed structure. A proposal for simple and practical sensor design based on the proposed structure is presented. It is consisted of a miniature fiber optic coil that is deformed proportionally to the measured environmental parameter. We practically demonstrated sensitivities in the range of ΔI/Δx=130%/N and ΔI/ΔF= 1.1%/μm. Even higher sensitivities are possible by proper mechanical construction of the sensor element. The proposed structure can configured in variety of different distributed and quasi-distributed architectures and is suitable for embedding into the composite materials.

A novel distributed fiber optic sensor that incorporates liquid swellable polymers to transduce the swelling into a microbend loss is presented. Interrogation of the sensor using standard optical time domain reflectometry (OTDR) instruments provides the possibility of detecting target chemicals and fluids at any location along the sensor length. The location of multiple events along a sensor, which may extend to 4 km is readily achievable. In this paper we present an overview of the work conducted on the characterization of a distributed optical fiber water sensor. Following a discussion of the basic principles of the water sensor and the underlying technology we present a review of the significant developments achieved. Tests incorporating the sensor in civil engineering applications, which range from monitoring of concrete curing to leak detection in highways, are described. In addition to this, more recent developments to utilize the sensor technology to detect other fluids are discussed, in particular for the monitoring of pH changes and liquid hydrocarbons. We discuss some of the significant advantages in using this type of sensor construction and areas in which it can be practically used.

The use of low frequency (sub 20 MHz) incoherent optical frequency domain reflectometry (IOFDR) provides a low cost alternative to the conventional OTDR techniques often used in the interrogation of optical fiber microbend sensors. We have modeled the operation of the IOFDR and experimentally characterized the operation of this technique for monitoring distributed water sensors based on water-swellable polymers (hydrogels). We demonstrate that the IOFDR is capable of detecting and locating sections of increased loss in a graded index multimode optical fiber and discuss the fundamental limits on spatial resolution and range.

We report on the performance parameters and a specific system application for a quasi-distributed absorption based sensor system. Coherence addressing of a series of open-path micro- optic sensing cells using FMCW is achieved by the interferometric mixing of signal pairs originating from glass/air interfaces in each cell. The interference patterns of all sensor cells add up (incoherently) at the detector, and by arranging the cells to be of differing lengths, each individual cell may be identified by its unique power spectrum in the frequency domain. We show theoretically and experimentally how to avoid 'cross terms' arising from the undesired interference between signals from optical paths associated with more that one cell. Results on system parameters including dynamic range, output stability, resolution, temperature and fiber bend influences are presented. For the specific application to distributed gas detection, where absorption linewidths of approximately 5 GHz are involved, we present a comparison of system performance for broad- and narrow-band absorbers.

We report on the instrumentation of a high-speed air-cushion catamaran (Surface Effect Ship) with more than 50 fiber optic Bragg grating strain gauges, as well as conventional resistive strain gauges, accelerometers, a Motion Reference Unit and Global Positioning System. A bow mounted wave radar was used to characterize the sea-state in order to estimate the wave loads on the hull. The relatively large number of strain gauges enabled us to determine the global deformation modes of the hull as well as local stress concentrations. This instrumentation was installed on a new Norwegian naval vessel and employed during sea-keeping tests in smooth and rough seas off the Norwegian coast. The measurements enable a detailed characterization of the vessel's dynamic response to wave loading and comparison with Finite Element Analysis modeling of the ship. The experimental results provide invaluable information for the subsequent development of a system for health monitoring of the structure. We present the instrumentation layout and selected results.

Closed meshed instrumentation or sensor networks with conventional sensors for temperature and strain measurements may result in excessive penalties in terms of weight constraints, sensitivity to environmental conditions and complex interfaces. The FOS is a multiplexed sensor system for up to 50 single strain and temperature measurement points comprising of a fiber network and an optoelectronic unit. The FOS sensor was designed and developed by Kayser-Threde, Munich, for demanding space environment, but can also be emphasized as a promising sensor technology with high potential for non-space applications. A Fiber Optic Sensor (FOS) measures strain and temperature by means of wavelength shifts due to tensile stress on a Bragg grating. Slightly shifts in the reflected wavelength are proportional to temperature or strain acting on the fiber at the corresponding grating location. Dependent on the fixation of the fiber to the structure, either floating or attached to the surface, local thermal or mechanical loads can be determined. The fibers can be mounted at the monitored structure or embedded (e.g. into composite materials). The FOS sensor is very suitable for structural health monitoring of large structures, i.e. to determine thermal and mechanical load profiles during operation, assessment of residual strength of structural elements or to detect irregular conditions. In comparison to conventional sensors like thermocouples and strain gauges, a FOS network significantly reduces the amount of required Front End Electronics (FEE) and harness.

We present three types of intensity-based fiber-optic accelerometer with two axes of sensitivity. Transmission and reflection, or single-ended sensor configurations are compared and bare fibers or a machined brass element are used a seismic masses. The devices are shown to be capable of measuring accelerations at relatively low frequencies (below 1 kHz). Linear responses to acceleration were observed up to the limit of the testing apparatus and the crosstalk between vibration directions was always -12 dB or less.

Embedding fiber optic sensors in composites has achieved sensing of strain or vibration at micron-levels. Wavelet signal and image processing are potentially a means to detect actual cracks (and future fractures). Successes and obstacles of composite-embedded fiber optic sensors will be examined and new capabilities emulating from wavelet signal and image processing will be projected.

Magnetoelastic materials are obtaining an increasing interest in the last few years. Magnetoelastic wave resonant frequency is the sensory parameter generally used to reliably obtain critical information, nevertheless, it was shown that wave amplitude also is very sensitive to measure displacements or magnetic fields. Both parameters, amplitude and resonance frequency could be used as the 'sensitive parameter' in no- contact vibration sensors or stress, flux or magnetic field sensors. To make them advantageous to be used, some problems should be solved: the excitation problem (to induce resonance) and the conversion problem (to transduce frequency or amplitude information into a convenient electric signal); in other words, the signal conditioning problem. The paper describes a signal conditioning technique for magnetoelastic sensors developed by using metallic glass materials. Analysis and synthesis phases of design for the signal-conditioning prototype are presented. From the frequency response analysis of a magnetoelastic metallic glass ribbon, a very low damping factor is pointed out. This means that a stationary oscillation could be easily induced into the ribbon and it was seen that the oscillation frequency changed when the boundary conditions (mechanical stress or magnetic field) changed. So, the ribbon can be used as a sensor. The proposed device (PLG 3) for the sensor signal conditioning is described in detail. Finally, PLG 3 plus a magnetoelastic sensor is used for a no contact vibration measurement and compared to the results from traditional tools.

This paper reports on a new generation of Fabry-Perot fiber optic sensors to be used in parallel or in replacement of conventional instruments for monitoring of structures. The new generation of sensors is based on a unique fiber optic strain sensor that represents a breakthrough in fiber optic sensing. The novel technique is based on extrinsic Fabry-Perot white- light interferometry which offers outstanding accuracy and repeatability. Furthermore, all sensors are completely immune to lightning surcharges which opens new possibilities in the field of reliable long term structure monitoring. Instruments such as piezometers, embedment strain gages, surface strain gages, temperature sensors and displacement transducers are all available in Fabry-Perot fiber optic version. Furthermore, all these instruments use a common multimode optical fiber to carry the signal to the readout units and multi-channels dataloggers. Both static and dynamic measurements are possible with this technology. The paper presents results of laboratory and field studies on fiber optic sensors mentioned above including integrated fiber optic sensors in carbon and glass fiber reinforced polymer. A case study of bridge strain monitoring with Fabry-Perot sensors is also reported.

The paper describes the necessity and feasibility of digitizing output signal from the fiber-optic sensors (FOS), and how to realize the digital optics for FOS with intensity modulation. A method of obtaining information by using special optical fibers is described. The auto-adaptive optics technology is utilized to derive fiducial information. In order to increase computational flexibility, optical technology is used in place of several electronic technologies.

White light interferometric sensing systems are very attractive for application in smart structures due to the fact that they read out the absolute path differences from the transducers so that calibration schemes and device history do not have to be used. The basic output from such systems consists of arrays of data containing complex interference patterns. In this paper, the output of such sensing systems is simulated and used as a common basis for the comparison of a number of different signal processing algorithms that could be used to reduce the raw output to a more usable form.

Optical fiber sensors, because of their small size, low weight, extremely high information carrying capability, immunity to electromagnetic interference, and large operational temperature range, provide numerous advantages over conventional electrical based sensors. Current and future aircraft designs require reduced sensor size and weight while maintaining resolution and accuracy in the extreme flight environment. Unmanned air vehicles also require more accurate sensor information to improve aircraft control systems. This paper presents preliminary results from optical fiber sensor designs for monitoring acceleration, pressure, and skin friction in-flight.

The prevention of fracture in a structure is one of the most important problems of mechanical engineering. The fracture of a structure occurs at mechanical failure locations. It would be possible to prevent fracture if it were possible to discern failure locations. One possible method of detection might be to make use of the visible properties of the material. Yield point elongation is accompanied by the formation of visible bands known as Lueders' lines or stretcher strains. These Lueders' lines result not only from monotonic loading, but also from cyclic loading. Utilization of the appearance conditions of Lueders' lines enables the naked eye to detect yielding failure locations of ductile machine members. The appearance of those striped patterns indicates where a fracture is about to occur. It is possible to monitor the position where fracture may occur, over time, if changing striped patterns can be detected by digital signals. It is possible to detect the failure locations of a structure from a remote location. Change of striped patterns indicating maximum stress during cyclic loading is examined. Direct observation was made of change in surface conditions at the vicinity of a hole and round corners where cyclic loading was applied.

Electrical time domain reflectometry (ETDR) distributed strain sensing technique has been successfully used in the health monitoring application of civil concrete structures to detect crack damages and in geotechnical application to monitor rock deformation and longwall movement. Although promising results of using commercial ETDR coaxial sensing cables have been shown in recent studies, the low signal-to-noise ratio of the sensors is a research issue needs to be addressed. Since all the commercial coaxial cables are specifically designed for the transmission of electrical signals, the cable configuration is to sustain a virtually constant electrical property under environmental loading effects. For structural strain sensing application, on the other hand, the electrical impedance of the sensor is required to proportionally vary with respect to externally applied loading. Thus, the commercial coaxial cables are indeed not in an optimal configuration for strain sensing application. In this paper, a newly developed high-sensitivity ETDR coaxial strain sensor prototype is presented. The construction of the prototype sensing cable as well as its electrical properties will be described in details. Experimental characterization of the high-sensitivity prototype coaxial sensor was also conducted. Test results of the sensitivity and tension responses of the ETDR signal of the prototype sensor are presented and compared with those of commercial coaxial cables. It is shown that the prototype sensor has a much superior sensitivity and properties for distributed strain sensing application.

In many situations, it is desirable to measure the load acting in a specific direction by measuring the strain induced by Poisson effects in a direction perpendicular to the load direction. For this to be possible, a fixed relationship between the strains in both directions must be known. This can be useful, for example, when the geometry is such that there is not sufficient room to locate a strain gauge parallel to the load direction but a gauge can be placed in a transverse plane. In this paper, we investigate the use of a fiber Bragg grating in such an arrangement with the fiber embedded within the host material. The investigation is done by theoretical, numerical and experimental approaches and we concentrate on two aspects: (1) the non-uniform strain transfer, particular in axial strains, due to shear-lag effects, and (2) the effect of induced birefringence in the optical fiber due to a load cross to its axis. The results of these approaches indicate that the strains of an embedded fiber sensor subjected to transverse loads are dependent on the location of the embedded sensor and the material properties of the host material. The results also show that when the Young's modulus of the host material is much less than the modulus of the embedded sensor, the Bragg spectrum broadening due to induced birefringence is not significant. However, a lower host Young's modulus also results in longer sections on non-uniform axial strain near the ingress and egress sections of the optical fiber. These two factors must be balanced if we desire to use conventional methods of Bragg grating interrogation that measure only the central wavelength of the Bragg grating's spectrum. In the case investigated (Host Young's modulus of 4.83 GPa) full strain build-up requires approximately 4 mm of fiber length at each end. Likewise, the transverse stress coupling into the fiber modifies its wavelength-shift-to-axial-strain- coefficient by about 6%.

A soil pressure transducer by using fiber Bragg grating (FBG) sensors associated with a circular diaphragm is developed. The FBG based transducers can be used for pavement performance study and weigh-in-motion measurement. We consider three methods of bonding the FBG to the diaphragm: (1) radially, (2) radially, inside a glass capillary, and (3) circumferentially. The investigation of strain-gradient induced spectral broadening in FBG-based transducers is conducted since spectral broadening can have adverse effects on the sensor interrogations. We derive analytical closed form results for describing measurand-induced strain gradients in circular geometry transducers, which allow us to experimentally demonstrate novel FBG bonding approaches that eliminate spectral broadening. In addition, Bragg spectral broadening analysis using T-matrix calculation is also conducted to validate some of the experimental results. Two prototypes of soil pressure transducers are field tested at the Cold Region Research Engineering Laboratory (CRREL). The buried pressure transducers are impact-tested by use of a Falling-Weight- Deflectometer (FWD), and detected by NRL-developed FBG interrogation device. Lastly, we use the Boussinesq equation to verify the soil stress measured by the buried transducers.

The newly developed TEFPI (transmission-type extrinsic Fabry- Perot interferometric) optical fiber sensor can distinguish the direction of measurement more simply and effectively than the conventional reflection-type EFPI optical fiber sensors. The output signal of the TEFPI optical fiber sensor has the characteristics that the signal level of fringes shows a negative slope for a tensile direction and a positive slope for a compressive direction. Based on these characteristics, the direction of measurement of the TEFPI optical fiber sensor can be distinguished with ease. In this paper, the signal processing algorithm adequate to the TEFPI optical fiber sensor was developed. This algorithm can process signal with recognition of the positions of peaks, valleys and signal levels of fringes. Thus this can determine a measurement direction and the positions of direction changes by using the change trend of signal levels. The developed algorithm makes the post-process and real-time process of the signal of the TEFPI optical fiber sensor possible.

Fiber cavity etalon (FCE) sensors have demonstrated ultrahigh static strain sensitivity (~1 nε) when they are either surface-mounted to, or embedded, in graphite reinforced resin composites. Although a significant amount of data has been acquired at very low strain, little is known about their performance and durability in typical installations. Graphite/epoxy composite test specimens were fabricated to address practical concerns and to evaluate the reliability of embedded FCE sensors. Two different specimen configurations using two different composite fabrication methods were selected for sample installations: thin flat laminates and cylindrical struts. After fabrication, the FCE sensors were interrogated to ensure that they were still intact, to record a baseline response, and to determine any changes in response that might have occurred during manufacturing. Next, to determine the survivable strain limits of the embedded sensor, the specimens were loaded in tension to a predetermined strain level, unloaded, and then interrogated. Once these limits were found, the specimens were subjected to cyclic loading and periodically interrogated until sensor failure. The results from these tests provide practical strain limits for the embedded FCE sensor and show that the response does not change as a result of tensile cyclic loading.

We have studied stimulated Brillouin scattering in single-mode optical fibers as a sensor for both temperature and strain. Shape Memory alloy Nitinol is also studied for enhancement of active sensing and control in structures. We have introduced the hybrid sBs amplifier/oscillator scheme, in which the short fiber amplifier performs sensing and the long oscillator fiber provides the required signal. The sensing fiber can be coated with SMA thin film for ruggedness, increased sBs sensitivity to temperature and strain, and memory or trainability. The oscillator/amplifier scheme also serves as a building block in the design of optical threshold logic circuits, optical computation, and more sophisticated sensing schemes. Such sensing schemes can be highly competitive with those based on fiber Bragg grating. The incorporation of SMA thin films provides memory capability to all these applications.