Multiplexed Fibre Bragg Grating Monitoring System for Dynamic and Static Strain Measurements:
Sensor System Design and Evaluation Report D5.4
PRIORITY 6
SUSTAINABLE DEVELOPMENT GLOBAL CHANGE & ECOSYSTEMS
INTEGRATED PROJECT
permissible speed of passenger trains. This should be possible without causing unnecessary disruption to the carriage of goods and passengers, and without compromising the safety and economy of the working railway.
The Project has developed improved methods for computing the safe carrying capacity of bridges and better en- gineering solutions that can be used in upgrading bridges that are found to be in need of attention. Other results will help to increase the remaining life of existing bridges by recommending strengthening, monitoring and repair systems.
A consortium, consisting of 32 partners drawn from railway bridge owners, consultants, contractors, research in- stitutes and universities, has carried out the Project, which has a gross budget of more than 10 million Euros. The European Commission’s 6th Framework Programme has provided substantial funding, with the balancing funding coming from the Project partners. Skanska Sverige AB has provided the overall co-ordination of the Project, whilst Luleå Technical University has undertaken the scientific leadership.
The authors of this report have used their best endeavours to ensure that the information presented here is of the highest quality. However, no liability can be accepted by the authors for any loss caused by its use.
Copyright © Authors 2007.
Figure on the front page: Photo of a optical fibre integrated into a CFRP bar.
Project acronym: Sustainable Bridges
Project full title: Sustainable Bridges – Assessment for Future Traffic Demands and Longer Lives Contract number: TIP3-CT-2003-001653
Project start and end date: 2003-12-01 -- 2007-11-30 Duration 48 months
Document number: Deliverable D5.4 Abbreviation SB-5.4
Author/s: M. Gebremichael & W.J.O. Boyle, City University, London, UK Date of original release: 2007-11-30
Revision date:
Project co-funded by the European Commission within the Sixth Framework Programme (2002-2006)
Dissemination Level
PU Public X
PP Restricted to other programme participants (including the Commission Services) RE Restricted to a group specified by the consortium (including the Commission Services) CO Confidential, only for members of the consortium (including the Commission Services)
3
Summary
The method of WDM monitoring system methodology with scanning Fabry-Perot interferome- ter for multiplexing has been numerically analysed to highlight the design considerations for individual component parts of the system. This has been achieved with the aid of a computer simulation, modelling the elements of the complete system.
It has been shown that wavelength measurement resolution of 3pm, acceptable for most monitoring applications, can be achieved using an electronic detection system having a modest SNR of 20dB, although a higher a higher SNR (30dB) has been achieved with the detection electronic circuitry designed. This result shows that the electronic noise in the de- tection circuitry is not critical to warrant stringent circuit design specifications, implying that cheaper and low power light sources of the form of LED’s can be used, replacing the more expensive SLED or ASE sources, thus making the system cost effective.
The results also show that although a higher Finesse Fabry-Perot interferometer is required for accurate peak position location, the narrow transmission peak means that the system SNR is reduced.
In practice, the measurement system designed suffered from other sources of errors such as FPF scan non-linearity and ambiguity in the mirror separation and also in the inability of the centroid peak position detection algorithm to deal with dynamically changing numbers of sample points per peak. In this regard, least square quadratic fit based peak location is pre- ferred.
Following the numerical design analysis, an advanced high bandwidth, multiplexed FBG in- terrogation system has been developed and evaluated as a field trial prototype. The multi- plexed optical system is much more versatile and vastly superior to a system based upon the same number of discrete and individually connected conventional sensors, an advantage that is of increasing value, as the number of sensor points is also increased. It has been ex- tensively evaluated in the laboratory environment to analyse its performance. The capability to measure multipoint data using the multiplexed system successfully developed for potential use in strain mapping across a large structure has been demonstrated with strain uncertain- ties of <±5μS. Comparative measurements with resistive strain gauges showed close corre- lation. High measurement repeatability to within <5μS was achieved, highlighting the per- formance of the system and that there has been no degradation observed of the sensor or the bonding mechanisms that have been employed for attaching the FBG sensors to the structure during repeated cyclic tests in the lab.
Tensile test results on post embedded array of Bragg grating sensors into carbon fibre struc- tures in an attempt to create a smart sensing strips or strengthening rods have shown good results correlating with surface attached electrical strain gauges. Such robust sensor packag- ing will result in preassembled sensors that could routinely be used in harsh environments, large concrete or masonry structures with ease. Following this exercise, 96 FBG sensors have been embedded into glass fibre composite rods which in turn where embedded into a number of 12m long concrete beams to monitor the effects on corrosion on the steel rein- forcement bars by introducing accelerated corrosion electrically.
Further fibre Bragg gratings have been embedded in composite rods and tubes to make repair and re-enforcement components that can be monitored throughout the structure life- times is seen as a major goal of the work This work was done to repair and re-enforce a working concrete rail bridge in order to increase its lifetime and its load capacity in coopera- tion with Luleå university and Sto Ltd. Initial results form this work are reported, showing the response of the sensors to dynamic strain due to loading by rail traffic. Analysis of the results of this work is ongoing and will be reported in conferences and journal papers in the near fu- ture.
stalled for long term monitoring on two 50m deep reinforced concrete foundations during the construction stages. Initial tests showed that the attachment and protection methods em- ployed gave predictable strain transfer efficiencies. The purpose of this field trial exercise was to demonstrate the sensor survivability during installation and the performance of the sensors in harsh environments similar to those expected in monitoring rail bridge structures.
5
Acknowledgments
This guideline has been drafted on the basis of Contract No. TIP3-CT-2003-001653 between the European Community represented by the Commission of the European Communities and the Skanska Teknic AB contractor acting as Coordinator of the Consortium. The authors ac- knowledge the Commission of the European Communities and the City University for its fi- nancial support.
1 Introduction...2
2 Background: Fibre Bragg grating for sensing applications ...5
3 Monitoring system requirements for practical use ...7
4 Fibre Bragg grating multiplexing techniques...9
5 Theoretical Considerations and Experimental Estimation of Wavelength Division Multiplexed Fibre Bragg Gratin Sensors...10
6 Development of FBG interrogation systems ...12
6.1 Prototype multi-channel interrogation unit...12
6.2 Bragg grating sensors and transducers development...13
6.3 Real-time data acquisition and processing software ...13
6.4 Modular and Handheld systems ...14
7 Performance evaluation and pilot tests in controlled environments...17
7.1 Measurement accuracy and wavelength shift calibration test ...17
7.2 Strain sensor calibration ...18
7.2 Dynamic strain measurements: cantilever beam...18
8 FBG Sensors Incorporated and Embedded in Rebars and re-enforcement composite components for repair and enhanced specification. ...20
8.1 Carbon fibre rod tests...20
8.2 Steel rebars ...22
8.3 Carbon Fibre Panel ...23
8.4 Class fibre rods ...23
8.5 Composite carbon fibre reinforcement rods ...24
Örnsköldsvik Bridge ...24
Frovi Bridge ...24
9 Conclusions ...26
References ...27
Appendix 1: Wavelength division multiplexed monitoring system: design considerations ...30
9.1 Scanning Fabry-Perot filter multiplexing principle ...30
9.2 Light source and detection electronics ...31
9.3 Data acquisition, signal processing and data reduction requirements ...32
Appendix 2: Experimental and simulation results...34
9.4 Electronic noise and intensity fluctuation ...34
9.5 Fabry-Perot filter characteristics...35
2
1 Introduction
The upkeep and enhancement of existing civil infrastructure is a major concern in the Euro- pean Union. This report discusses and presents work done in the Sustainable Bridges (SB) project to develop and apply Bragg grating based optical strain monitoring technology, Fibre Bragg Grating (FBG) sensors, for monitoring strain in rail bridge structures and the compo- nents under development within SB and other programmes to enhance lifetime, load capabil- ity and rail traffic speed. The incorporation of resident, cost effective damage detection and monitoring systems is an important consideration which can be used to monitor long term behaviour and/or determine the effectiveness of any subsequent repairs that may be carried out.
Current monitoring such civil structures is limited by the accepted available technology. This consists on the most part of a combination of electrical strain sensing and load cells. Electri- cally based sensors are time consuming to install and require a large number of electrical connections and thus complex cabling. Electrical systems are also difficult to embed during either the construction or during any subsequent repair processes that may be carried out.
Consequently, much monitoring is usually undertaken by periodic visual inspection or through the use of conventional strain gauges, which are not well suited either for long-term use or in harsh environments (e.g. during exposure to chlorides). Fibre Bragg Grating sensor (FBG) sensors based monitoring technology on the other hand enables online in-situ moni- toring of temperature and strain with unobtrusive, chemically inert (relatively) and electrically immune sensing operation.
FBG strain monitoring technology has developed significantly over the last five years [Maier- hofer, Cremona, Alampalli, Pullin, and Shigeishi]. It has been applied worldwide to enable research programmes in structural engineering and material science. Many of these pro- grammes have been concerned with assessing the methodology and applicability of FBG sensors in civil structures and related components including; rail and road bridges, founda- tion piles and material studies on composite materials for construction and avionics, and smart materials (composite panels) for repair. Studies by our group and others have estab- lished FBG technology in structural monitoring and assessment in a number of important ap- plications.
In particular, studies on rail and road bridge structures show that FBG offer considerable ad- vantages over electrical strain sensors: Namely higher sensitivity, sensor and cabling com- pactness, higher sampling rate, and with careful design robustness. The technology is now at the stage where it is comparable in cost than electrical strain gauge technology. Dissemina- tion of the skills required to implement it is however now the main rate limiting factor in take up of the technology but there are signs in the number of studies and in the exposure in the trade press that take up, at least as an enabling technology in R&D studies is quickening as structural engineers and material scientist become more familiar with FBG technology.
The relevance of the technology has been raised by several partners in the Sustainable Bridge programme and there have been discussions on its applicability at programme meet- ings to several key areas of interest to the project partners. Amongst these issues are, its applicability to:
Steel, Concrete and Steel-Concrete composite, bridges
1. Many bridges in Europe are in this category and their construction dates back some 60 years or more. Many are in a condition where useful life can be ex- tended and enhanced by remedial repair. Regular condition monitoring then be- comes more important as bridges are taken beyond original design limits.
tation and suitable transducer technology required to address the issues raised here. In particular City University has worked on developing more compact re- mote controlled instruments for interrogation of FGB transducer-sensors. Also in collaboration with Luleå University, City has developed suitable transducers con- sisting of FBG sensors incorporated in carbon-fibre composite panels, rods and
‘coupons’ and long (10m) glass-fibre rods for incorporation in reinforced concrete and of possible use in masonry structures.
Foundations
1. With the desired reuse of bridge structures, the question of the integrity of founda- tions required addressing.
1. City have through collaboration with a FP5 programme “reuse of foundation piles” installed 15 sensors in two 50m deep construction piles for later moni- toring of third party effects and eventual assessment for reuse. The sensors were located in steel rebars and successfully deployed within the tight con- struction schedule. Strain and temperature changes induced by the concrete curing, and the building schedule of a ten storey building were followed. The work has direct relevance to this programme because it demonstrates the ability to incorporate FBG sensors in re-enforced concrete as the construction stage.
Rail Traffic axial loads
1. The FBG technology steps required for real-time measurement of strain and tem- perature on rails has been developed. The sensors can be addressed at very high sampling rates allowing detailed analysis of the dynamic changes in rail strains.
Also as many sensors can be deployed at intervals down the track from a single optical fibre, the technology offers the possibly of studying the complex dynamics of wagon and bogey movement.
1. This area of work was not considered as a main priority within this pro- gramme. It is logistically difficult to implement due to heath and safety consid- erations controlling access to track for sensor installation.
And Masonry bridges
1. Masonry bridges represent a major part of the stock in many European countries including Italy, the UK, Spain and France. Monitoring and assessment of masonry bridge condition and repair has been identified as a goal of the programme by WP2 and partners concerned with condition monitoring in WP4-WP8 have been required to identify and address this issue. Masonry Bridges are recognised as the most complex of structures to monitor, because the stress field within the structure can be spatially very complex due to the interaction of masonry blocks and cement. A major issue is the effect of fast rail traffic on the structure, as the forces can dislodge blocks and produce undetected voids with unpredictable ef- fects of structure integrity.
1. City have had discussions with WP4 partners at Bradford University about the applicability of FBG sensors to masonry structures, but have been unable to apply the considerable effort required within this programme to investigate the application of FBG sensors directly. However our work, with Luleå University, developing of composite glass fibre rods incorporating FBG sensors can be employed in masonry structures.
In particular City have through collaboration with Luleå University developed technologies for embedding FBG sensors into rebars and onto composite panels, rod and tubes for monitor- ing of repaired and improved structures. It is considered that a considerable effort will be af-
4
forded to maintain and improve the existing stock of bridges in the EU and that the renovated structures will require more monitoring to ensure the correct loading is taken up by re- enforcement and that this loading stays within specification.
The main body of this report discusses two main aspects of our work undertaken in the pro- ject: The design and development of FBG sensor systems specifically for use in civil engi- neering applications and the application of FBG monitoring in several key studies that ex- plore and elucidate its potential in monitoring railway bridges and related structures (Founda- tions, track) and particularly explore and develop the potential for FBG sensors embedded in composite components for repair and enhancing the performance of bridges.
For the studies a ruggedly engineered high bandwidth optical fibre Bragg grating based pro- totype sensor interrogation units has been designed and constructed. This has the capability for monitoring, lifetime prediction and maintenance control of large strategically important civil engineering structures, such as rail bridges. This design is underpinned by a full and de- tailed numerical analysis of the dynamic response of the electronics and optics components of the system. This analysis has allowed us to determine the factors with the technology that affect the overall sensitivity and reliability of strain measurements.
The performance of the new prototype interrogation system design has also been demon- strated in pilot tests in the laboratory environment for and later in large-scale laboratory tests where the system was assessed during a study of concrete beam corrosion. In this study, in collaboration with Luleå University, 96 FBG sensors were attached to 12m long concrete beams and embedded inside 12m concrete, Sweden. These long term tests assessed accel- erated corrosion and induced damage to assist the study of various concrete repair tech- niques. These experiments could not have been undertaken using conventional foil gauges.
The particular advantage of FBG sensors embedded in the glass fibre rods is the resulting 12m long, 5mm dia. linear transducer structure with up to 24 strain and temperature sensors.
The sensors and attaching optical fibre cables take up a cross-sectional area of less than 0.5mm2. The glass fibre rods developed and used in the study have potential application in studies of masonry structures and it is envisaged they will be further assessed in later trials on bridges. The results of this work on concrete beams will be presented when the tests are completed.
Further to this, through direct experience in field use of sensor systems a conclusion that smaller systems will be more acceptable to structural engineers is made and a smaller com- pact system that can be more easily transposed and installed in construction sites is desir- able. Consequently two systems are designed, developing and contorted for instrument evaluation and field work.
The first is a redesign of the compact interrogation system discussed above into several small modular units with small external power supplies. Several advantages result. The overall weight of the interrogation system is less than 8kg compared with 19kg with no mod- ule more than 4kg; the modules can be more easily transported and set up by one person and the modules can be maintained separately.
The second small -almost handheld system is near completion. This is a two channel sys- tem. Currently the system is constructed in two (7 x 20 x 10 cm3) modules with small exter- nal power adaptors. This system is capable of addressing some 30 sensors with strain ranges up to +- 1000uS. The sample rate of this system is less than the compact system, however the construction cost is much less, ~<€5000, compared with >€10,000 for the com- pact system. This system will be integrated into one (7 x 20 x 10 cm) unit within the time scale of the project but not in time for this report.
The remainder of this report discusses the technology developed and its application within the programme. Much of this work has a level of technical complexity outside the scope of the programme and is therefore included as appendices to maximise accessibility to this document.
2 Background: Fibre Bragg grating for sensing applica- tions
Condition monitoring is an ongoing requirement in the civil engineering industry. It is used to assess the response of structures to test or service loading as well as monitor changes re- sulting from environmental conditions. The management of structures relies partly on such techniques to discover and monitor known defects and allow economic scheduling of main- tenance. Established techniques such as ultrasonic scanning [Bastianini], transient pulse and infrared thermography [Clark, Wiggenhauser] and ground radar [Scott, Maierhofer] are used for condition monitoring applications. Although these techniques are adequate and reliable, they are not suitable for real-time monitoring which can be desirable when dynamic loading is an issue, for example to study the effects of high speed and high load traffic, or to study the long term stress in a structure from a known starting point. This limitation can be overcome by using electrical resistance strain gauges [Cremona, Alampalli] although these require electrical cabling which can deteriorate over the long term. Continuous monitoring with acoustic emission [Pullin, Shigeishi] measurements can detect the presence of release of fatigue where stress released suddenly can result in acoustic noise. The use of Fibre Bragg Grating Sensors (FGBS) is growing in importance both for intermittent and for continuous monitoring of strain on new and existing structures. [Zhang, Casas, Singh, Spillman, Udd].
Passive sensing devices optical fibre sensors are superior in many applications over conven- tional monitoring techniques due to immunity to electromagnetic (electrical) interference, their small size, lightweight construction [Grattan]. As a result, there is an increased interest been in the widening application of optical fibre sensors for a range of engineering use. In particu- lar, there is considerable interest in developing fibre Bragg grating (FBG) sensing techniques for a range of applications in structural integrity monitoring for physical, chemical and biologi- cal applications commonly measuring strain, temperature, pressure, vibration, acceleration, humidity, corrosion and other monitoring parameters for industrial and civil structures such as bridges, foundations, high-rise buildings, landfills, nuclear waste repositories, reservoirs, oil wells and tankers, mining, aerospace and other structures [Fernando, Maaskant, Aleixandre].
Written directly into the core of the optical fibre, FBG sensors encapsulate all the advan- tages of fibre optic sensors: the sensors are small in size, unobtrusive and lightweight and hence suitable for embedding into new or retrofitting into existing steel, concrete or compos- ite structures [Gebremichael, Lau, Gebremichael, Measures]. Being passive sensing ele- ments that are dielectric in nature, there is no electric current flowing at the sensor and they are therefore insensitive to interference from electrical interference and are thus suitable for use in potentially explosive and harsh environments (e.g. high voltage environments, oil re- fineries, chemical processing plants etc.) [Willsch, Yamate]. Being made from glass they are relatively resistive to chemical attack. They can be operated and can measure temperatures over an extended range, typically from -100oC to +120oC.
In comparison with conventional electrical sensors, the versatility of these sensors and their diverse application mainly stems from their ability to be connected in series, with up to 20 or more sensors along a single optical fibre of typically 125μm dia. and the potentially very fast response. For a few sensors, typically 5, this can be as fast as 500 samples per sensor per second; however for many sensors the sampling rate is lower. FBG sensors are more sensi- tive than electrical sensors, typically better than 1 μS at one sample per second, and poten- tially better that 0.001 μS at this sample rate. This capability allows distributed sensing and parameter mapping across large structures with very little requirement for cable space. This is a major advantage of the use of this approach. One drawback of optical fibre sensors is that they are relatively fragile and thus for most real life applications a robust sensor protec- tion techniques [Glisic] must be developed whilst maintaining efficient measurand transfer
6
between the sensor and the structure. A great deal of research has gone into FBG based measuring techniques and applications, and thus the knowledge and experience gained in the subject is now significant and available from the literature [Grattan, Kersey, Rao].
The principle of operation of fibre Bragg grating sensors is based on the Bragg grating sen- sor element. This is typically a small region of photo sensitive optical fibre (1cm length, 125μm dia. that has been subject to a spatially modulated Ultra Violet (UV) laser beam that produces disruption of the glass that results in changes in periodic modulation of the refrac- tive index down the length of the sensor element with a period pitch of about 1.0 μm. The grating sensor element thus created essentially acts as a wavelength filter for incoming light in which a narrow spectral band is reflected while the remaining light is transmitted and may be used to interrogate gratings at different centre wavelengths further along down the fibre.
This results in a multiplexed sensor system.
The centre wavelength of the reflected spectral band from a grating, λB, is defined by the Bragg condition, λB=2neΛ, where ne is the average effective refractive index of the material and Λ is the pitch length of the grating [Kashyap, Othonos]. The sensing principle arises from the fact that strain in the fibre induced by changes in temperature, longitudinal stress, or other physical or chemical influences will alter the pitch period or refractive index of the grat- ing and in consequence the wavelength of the reflected light. This change is determined by measuring the shift from some initial value of λB using a wavelength spectrometer.
The result of this is a sensor that transduces strain to an optical wavelength measurement with a sensitivity of: 1.2pm/µS and 10pm/°C for telecommunications grade single mode fibre around 1550nm wavelength and works over the whole Hook’s law linear region, approxi- mately >±2,500μS. This provides a spectrally encoded sensor signal, which allows an abso- lute measurement of the wavelength of the sensor which is a direct measurement of the length of the sensor. The resulting signal is independent of the signal power fluctuation (which may occur due to factors such as the light source variation, the fibre bending loss or the connector attenuation). This is a major advantage of Bragg grating sensors for long term monitoring in large engineering structures where the service lifetime of the structure is con- siderable, and may be as long as several decades in suitable conditions.
3 Monitoring system requirements for practical use
A FBG sensing system architecture in its simplest form consists of one or more channels (linear arrays) of FBG sensor networks illuminated with a broadband light source though a multiplexing device coupled with detection electronics to receive the reflected spectra from each of the channels of FBG sensors. The spectrum of the reflected light is analysed by feeding the output to a digital signal processing schemes to give the centre wavelength of each sensor in the channel.
There are several design requirements that must be met for implementing reliable FBG sen- sor systems for monitoring applications. These are:
9 The intensity of the reflected signal must be strong enough for detection
9 The detection electronics must avoid adding noise, it must have high signal to noise ratio (SNR)
9 Very accurate timing between scanning the spectrometer and detecting the reflected light is required
9 Calibrated against accurate wavelength standard so that variations in signal are accu- rately and reliably determined as being due to the changes in the parameter of interest rather than other effects from environmental or instrument noise.
The criterion is that the induced changes are detected to a sufficient degree of accuracy with large dynamic range and high bandwidth for transient monitoring.
In general a practical monitoring system should exhibit many of the following qualities de- pending on application:
• Capable of multiplexing a large number of sensors sufficient to give multi-point and multi- parameter measurement.
• Flexible, re-configurable, expandable and upgradeable open system in both its software and hardware design with no specialised expertise required for its routine use.
• Expandable data acquisition for additional analogue and digital interfaces to enable inte- gration of other sensors and transducers coupled with intelligent data reduction and in- terpretation methodologies with database recording of historical data of interest, e.g. ex- treme loading conditions, mean values, and natural frequency of a structure.
• Low power or battery operation, rugged, compact and easily transportable for use in harsh and limited access environments e.g. construction sites for both periodic monitor- ing and permanent installation.
• Accessible through the Internet (wired PSTN network or wireless GSM) for online moni- toring, remote data access and instrument control. It must also be capable of self- monitoring for its performance, i.e. signalling faults or self resetting
• Cost effective and within market requirements.
Often these considerations need to be traded. For example; the higher the number of sensor points on each optical fibre channel the lower the maximum range per sensor and the higher the cost of getting enough power to illuminate all the gratings and accurately detecting the changes. Further more, large number of sensors inherently leads to large volume of data for processing and communication, which also adds to the cost of the system. Further with more and more reported applications, FBG sensor elements are steadily and interrogation
8
instrumentation is steadily coming down in price and the cost of implementation will soon be comparable to the cost of electrical strain gauges. Uptake of the technology is however lim- ited by lack of user experience and this is a bottle neck in bringing this new technology to the market for wider industry.
4 Fibre Bragg grating multiplexing techniques
There are several major techniques for multiplexing Bragg grating sensors in multi-point sen- sor applications and thus decoding the information provided by individual sensors in the form of identifiable wavelength shifts. These have been discussed in detail elsewhere in the litera- ture and include wavelength dependent filters [Melle], interferometric demodulation [Kersey], tuneable laser configurations, tuneable Fabry-Perot filters [Kersey, Ning] and time domain reflectometry systems [Berkoff, Ashoori]. Following an analysis of these various approaches, in this work a spatial wavelength division multiplexing (WDM) system was employed for ap- plication to structural health monitoring. The system has been configured to address a cluster of gratings in several arrays, based on serial multiplexing, using a scanning Fabry-Perot filter and a multi-way passive splitter. An array of single-element photodetectors (PIN diodes) is used to measure multiple sensor arrays in parallel. The measurement of each wavelength shift corresponds to measurand value over that particular grating. The centre wavelength of each sensor in any of the arrays is separated spectrally by a few nanometers which is wide enough to accommodate the maximum strain levels experienced by large structures such as rail bridges in service, allowing for the creation of a series of wavelength division multiplexed transducers.
10
5 Theoretical Considerations and Experimental Estimation of Wavelength Division Multiplexed Fibre Bragg Gratin Sensors
A considerable amount of effort has been given under WP5 to improving the design and to estimating the accuracy and sensitivity of FBG interrogation systems. As these more scien- tific considerations and effort are considered to be outside the direct scope of this pro- gramme, namely the interests of the rail operators, detail of this work is taken out of the flow of the report and included instead in appendices. There they will provide for the attention of the other technology developers and for a more complete account of the work.
The aim of this work is to define the design specifications for a high-resolution multiplexed fibre Bragg grating monitoring system and analyse the expected performance of a compact 8-channel system developed. To summarise the work: a theoretical based computer simula- tion model of the complete sensing system has been developed and experimentally verified in lab based measurement studies. This is used to optimise the sensitivity of WDM wave- length estimation and hence that of strain measurement:
The basic sensing principle of the system is measuring wavelength shifts of the spectral out- put of the system, which is a convolution of the spectral peak of the interferometer in trans- mission and the Gaussian reflection profile from the Bragg grating sensor elements in re- sponse to changes in a measurand of interest. The main factors that affect the measurement of wavelength are:
9 The available Signal to Noise Ratio (SNR) of the spectral peak heights, as depicted in figure 1a above the level of background noise
9 The uncertainty in the accuracy in the wavelength due to jitter in the scanning of the Fabre-Perot spectrometer mechanism
9 and the half height spectral peak width of peaks in the spectra
-3 0 3 5 8 10 13 15 18 20
1543.70 1543.72 1543.74 1543.76 1543.78 1543.80 1543.82
Wavelength (nm)
Time (s)
Figure 1a Transmission peak of a QI microfilter with bandwidth of 0.4nm
Figure 1b Wavelength resolution measured with a QI microfilter of bandwidth of 0.4nm
wavelength measurement resolution of 3pm is achievable. However the analysis shows the measurement is quite sensitive to an increase in the half-height spectral width of peaks with the minimum resolution increasing from 3pm to over 20pm. The results of this computer simulation have been born out in laboratory based measurements as discussed in the ap- pendix.
12
6 Development of FBG interrogation systems
6.1 Prototype multi-channel interrogation unit
A portable and compact multiplexed FBG sensor interrogation unit utilising a 10mw broad- band light source centred at 1550nm has been designed and developed specifically for long- term field application use, Figure 2. The broad band LED diode source is driven by a custom designed and built stabilised current source and feedback controlled thermo-electric control- ler. This ensures good stability of its output power and spectral profile and prevents thermal run-away and device damage.
The instrument comprises x8 optical fibre channels and is capable of interrogating 8 sensors per channel for high level strain measurements (each capable of measuring >±2,500μS).
With suitable programming the instrument can interrogate many more sensors per channel at smaller measurement ranges depending on the installed measurement configuration.
The return signal from each of the optical fibre is detected via one of 8 low noise (30dB SNR) PIN diode detectors. The detector circuits developed consists of a two stage electronic circuit consisting of a transimpedance amplifier that converts the current produced by the optical signal on the detectors into a voltage signal. This voltage signal is further amplified and the output filtered to a limiting frequency of 10 kHz to reduce noise. Signals from the gratings are typically of a Gaussian shape as shown in Figure 2.
The signals from the x8 channels are then sent to a plug-in Digital Signal Processor (DSP) (ADwin-light-16) board which converts the analogue input signals into digital signals suitable for further processing in a PC to determine the Bragg wavelengths. The DSP board is de- signed around a SHARC DSP processor with its own local memory for fast data acquisition (conversion time of 10μs), on-line processing (processing of each measurement can occur immediately after acquisition). The digital signal processor also produces a tightly controlled synchronous analogue signal to control scanning of the Fabry-Perot filter in real time.
The board uses proprietary real-time programming language (ADbasic) which allows accu- rate control of the Fabry-Perot filter and synchronous measurements of the sensor signals.
The DSP temporally resolves the optical spectra received from each of the 8 channels used, for every scan of the Fabry-Perot cavity and thus the position of each peak is identified.
In addition, post processing and data visualisation in the PC can be handled with many popu- lar programming environments including Visual Basic and LabVIEW.
strain and temperature readings, the measurements are made relative to two thermally stabi- lised (to 0.15pm) reference Bragg gratings which are incorporated in the interrogation sys- tem. These are assembled in a thermally controlled package for the rejection of common mode noise and multiplexing non-linearity, as well as wavelength scaling.
The reference gratings are also used to eliminate any long term or short-term thermal or shock vibration-induced drifts arising from the scanning filter. The accurate detection of the peak position in the wavelength domain is the key to accurate strain measurement.
Peak location involves least square curve fitting or threshold level detection around the peak from which the centroid of the reflected grating spectra is calculated. Long term test results show that the strain resolution of the system at full bandwidth (>10ms) is less than ±5μS.
6.2 Bragg grating sensors and transducers development
In most applications, Bragg grating sensors are either surface bonded to existing structures or embedded within the structures during manufacture. During the various laboratory evalua- tion tests, the sensors were either directly attached to the surface of the structure (e.g. steel) or embedded into concrete test beams by attaching the fibre sensors to steel re-enforcement bars during the casting phase. Several bonding adhesives and techniques of attachment have been investigated to optimise the strain transfer and to achieve high repeatability and sensor integrity in long-term measurement use. It is found that the sensor gratings must be attached to the structure directly with suitable glue. When tested, cynoacrylate type adhe- sives achieved near to 100% strain transfer in a manner that was consistent over many tests.
Prior to attachment, the acrylate coating on the Bragg gratings was stripped off to enhance the bonding to the surface, which was cleaned by using an appropriate solvent.
Further to this, as fibre optic sensors can be fragile, where fibres are not embedded, a robust sensor protection mechanisms must be employed which are compatible with the operating environment, while maintaining efficient measurand transfer between the sensor and the structure.
Fibre Bragg gratings long-term stability is achieved by using an annealing procedure in which the gratings are heated at ~100°C for a period of few hours in an argon atmosphere. Argon prevents degradation of the acrylate coating. Equally important to their effective use in in- strumentation is the thermal response of the sensors. Bragg gratings respond to temperature in the same manner as an applied strain (by expansion) and thus by giving a change in the centre wavelength. As a result, for long-term structural fatigue analysis, temperature com- pensation is necessary to enable strain to be uniquely determined. In this work, thermally compensated transducers have been developed by inclusion in the sensor “package” of strain-decoupled sensor to measure temperature, from which the thermo-optic effects can be corrected and strain determined, as required.
6.3 Real-time data acquisition and processing software
To facilitate real time monitoring an advanced software framework is being developed for in- telligent, remote-monitoring applications in a high level programming language environment.
The aim is to create more flexible applications for data acquisition, data processing and data reduction/fusion algorithms. Furthermore, the software design includes either event driven or on demand sensor system remote control options to enable a range of sensing configura- tions to suit an event.
The data management infrastructure will enable networking and archiving of sensor data in a database. These data can be called up for statistical analysis, interpretation and for use in accurate numerical modelling of complex structural systems as well as visualised access of real time data or display of key information requested by a client user. Such system can be used to support knowledge based decision-making process on risk assessment and struc-
14
tural maintenance. So far, significant progress has been made on developing web interfaced software for remote system control, data logging and communication of data sets over the Internet for storage on databases and display on web clients using PHP and MySQL soft- ware.
6.4 Modular and Handheld systems
Figure 4a shows a schematic for the compact system. Here the =various system compo- nents: Led light source, spectrometer, Optical fibre assembly, photodiode detectors and amplifiers, digital sampling processor, PC and power supplies are all integrated into one case with size and weight as depicted in figure 4a.
In the modular system, the component parts in figure 4a have been repackaged into the scheme shown in figure 4b below. This system so configured consists of six small units and is less than half the weight of the compact system, with similar performance.
Figure 4a: Schematic for the compact system.
A Design evaluation has been done to further develop the compact system into a battery op- erated handheld system which is robust, affordable, field deployable measuring tool suitable for routine static/quasi-static or dynamic strain measurements in structural integrity monitor- ing applications. The interrogation unit is designed for low-power consumption. The scheme for the hand held system is shown in figure 4c. This system utilises similar LED source and photodetectors and amplifiers as the compact system. However, here a much more compact miniature optical spectrometer and A/D sampling Field programmable gate array are used instead of the spectrometer, Digital Sampling Processor and PC arrangement. The handheld system has a USB interface and can be controlled directly via the Internet. This system has two channels for arrays of up to 16 sensors with a range of +- 1250uS per sensor. More sen- sors can be accommodated with smaller range.
The system is currently configured in two small modules with external power supplies. This system has been used to monitor the installation of strain in sensors embedded in carbon fibre rods and tubes for the re-enforced concrete railway bridge at Frovi; it has also been
Figure 4c Very small system
Figure 4b: Component scheme for the modular system
16
used on that bridge in preliminary field trials to measure strains induced by rail traffic. Cur- rently the instrument requires adjustment to follow the small strains of <20uS measured on the bridge with an alternative optical interrogation system and electrical strain gauges. Power consumption for his design is less than 5 Watts.
Figure5a shows this prototype system in use during the installation of Bragg grating sensors into the carbon-fibre rods. The system consists of the small black box and the circuit in the pink lined cardboard box. Figure 5b shows the optical spectra from the instrument.
Optical Spectra from array of Bragg grating sensors
0 500 1000 1500 2000 2500 3000 3500
1500 1520 1540 1560 1580 1600
approximate grating wavelength (nm)
Signal Voltage (mV)
Figure 5b Grating Spectra obtained with small system
7 Performance evaluation and pilot tests in controlled en- vironments
The hardware and software systems designed are tested at various stages of the develop- ment process to evaluate the functional performance of the compact prototype system in the laboratory environment. These tests were followed with extensive laboratory evaluation tests to characterise the performance of the system built and the transducers developed for moni- toring applications.
Detailed test plans were devised for laboratory tests prior to field trials, which were designed to serve as a means of verifying the performance of the optical fibre instrumentation and sensor attachment techniques in relation to strain transfer from the structure to the grating and sensor calibration issues. These tests were conducted using various structural elements to provide experimental test conditions mimicking real applications. Structural evaluation of the pilot structures using conventional laboratory based test equipment also provides infor- mation on the validity of the results. Thus a series of static and dynamic loading tests of vari- ous structural elements was performed in the laboratory and comparisons of the resulting strain were made with corresponding resistive strain gauges. These tests were conducted mainly on steel structures such as cantilever beams, steel plate coupons, steel reinforced concrete beams and carbon fibre rods. The outcomes of some of these tests are discussed below.
7.1 Measurement accuracy and wavelength shift calibration test
Prior to experimental tests, calibration tests were conducted to discover the performance of the monitoring system. Initially the measurement sensitivity is analysed and as shown in Fig- ure 6, the system is accurate to within ±2μS as demonstrated over repeated long-term tests.
The linearity of the measurement is demonstrated by a simple experiment where a fibre grat- ing sensor was anchored at one end and the other attached to a (high precision) micrometer driven stage. Incremental strain loading of up to 800μS was applied by adjusting the mi- crometer bezel. The results are shown in Figure 7 which shows that FBG sensors have ex- cellent linearity with sensitivity of 1.2pm/μS. The level of strains, at this stage were selected to ensure that the performance was well matched to the levels of strains that may be experi- enced during the bridge trials, but testing at higher strains also produced a similar linear re- sult.
Figure 6 Data showing measurement
-100 0 100 200 300 400 500 600 700 800
-12.00 -8.00 -4.00 0.00 4.00 8.00 12.00
Strain (με)
Sample points
Figure 7 Linearity of strain and wavelength calibration tests results
18 7.2 Strain sensor calibration
In earlier tests FBG sensors were also calibrated against measurements from conventional
resistive strain gauge. Figure 8a and figure 8b show wavelength shift measurements with an FBG sensor and corresponding strain measurements from a foil gauge. Both sensors were attached on the same strain axis on a cantilever beam for wavelength/strain calibration. From this experiment, a calibration factor of 1.2pm/μS (100% strain transfer) was achieved. This figure was consistently repeated with further tests on a number of similarly attached FBG sensors on a steel structure, indicating a successful sensor preparation.
7.2 Dynamic strain measurements: cantilever beam
Time-varying strain measurements are of interest in large structures such as the rail bridges in order to capture the transient behaviour of the structure under dynamic loading. A 1m-steel cantilever beam was chosen to demonstrate both the performance of the system and the ef- fectiveness of the sensor attachment method under dynamic loading, noting the response in
comparison to that from a foil strain gauge. Here, two fibre gratings were bonded on the top and the bottom surfaces of the cantilever beam. A foil gauge was also fixed adjacent to the top surface fibre sensor. The beam was deflected to a set displacement point and released into its natural (damped) oscillation at around 10Hz. The responses of the two fibre gauges
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
-800 -400 0 400 800 1200 1600 2000
Strain (με)
Time (s)
Figure 9a. Cantilever test, damped strain with FBG sensor
Figure 9b. Cantilever test, damped strain with resistive gauge
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
-1.5 -1.0 -0.5 0.0 0.5
Strain (AU)
Time (s)
0 20 40 60 80
-250.0p -200.0p -150.0p -100.0p -50.0p 0.0
Wavelength (m)
Time (s)
0 50 100 150 200 250
-150 -100 -50 0 50
Strain (με)
Time (arbitrary units)
Figure 8a wavelength shift, FBG Figure 8b Strain, foil gauge
the foil gauge was recorded using a separate high bandwidth data acquisition system. Fig- ures 9a and 9b show the comparison of the responses from the fibre and foil gauges just be- fore and after the release of the beam. It is seen that they have identical dynamic responses and using the known calibration scale factor, the figures can be made to show an identical strain response on the time axis. Similarly, when two FBG sensors were attached to the front and back faces of the beam so that approximately equal but opposite strain magnitudes and peak strain deflections measurements were obtained, further demonstrating the highly re- peatable nature of the measurement system and the optimisation of the sensor attachment techniques.
20
8 FBG Sensors Incorporated and Embedded in Rebars and re-enforcement composite components for repair and enhanced specification.
This section of the report discusses work done in developing embedded FBG sensors.
Throughout the course of the ‘Sustainable Bridge’ programme, considerable effort was given to developing equipment and software for interrogating arrays of FBG sensors, City has also collaborated with Luleå university industrial partners to developing methods for embedding FBG sensors into steel rebars (with Skanska UK and Cranfield University, EU REFUS FP6) programme, and glass and carbon composite components for repair and improvement with Skanska SA, STO SA and Luleå University with WP6 partners. This work on embedded sensors contributes significantly to the viability of using composite reinforcement panels to repair and upgrade existing bridges. There may well be a requirement to continually or peri- odically monitor the take-up of strain by the reinforcing components and this will best be achieved by embedding FBG sensors in reinforcing rods, tubes and panels. Currently only preliminary results are available for this report. This part of the report will present four case examples in the development of embedded FBG technology to date. These are:
- Carbon rod tests in the lab
- Attachment and protection of FBG sensors on steel rebars in 50 meter concrete piles - Embedding of FBG sensors in glass fibre rods for incorporation in concrete sections - Embedding of FBG sensors in carbon fibre rods for reinforcement of railway bridges - And embedding of FBG sensors in carbon fibre rods and tubes for concrete railway
bridges
8.1 Carbon fibre rod tests
In most applications, FBG sensors are surface bonded to the structure with some level of protection against mechanical as well as chemical hazards during usage in as much the same manner as conventional foil gauges. Although surface bonding allows optimised linear strain transfer, long-term sensor integrity can be significantly improved if the sensors are embedded within the structure; this is more practical for use in carbon or glass fibre compos- ite structures. Further to this, pre-assembled or embedded sensors minimise the risk of sen- sor dropouts both during filed installation and long-term usage.
In this work, results from experiments carried out with surface bonded and embedded FBG strain sensors are presented with comparative data obtained from surface bonded resistive strain gauges. A test set-up was constructed to characterise the response of embedded FBG sensors to applied strain loading. Both embedded (Figure 10a) and surface attached (Figure 10) FBG sensors and also electrical resistive strain gauges (ESG) were tested. In Figure 10a, two FBG sensors are embedded within 1mm deep and 1mm wide pre-cut grove inside the carbon fibre road. Embedding the sensors is of great importance for the use of Bragg gratings in mechanical and civil structures with ease of installation and for improved long- term survivability in harsh environments e.g. for concrete embedding. The integration of fibre FBG strain sensor system into carbon fibre composite materials is an essential step towards a continuous, remote monitoring of composite material in the design of "smart" structures
The carbon fibre test elements were subjected to an incrementally changing static-loading as well as to slowly varying dynamic-loads applied pneumatically by employment of a hydraulic test rig. Two such tests were conducted on (i) a 350mm long carbon fibre strip with surface bonded FBG strain sensors and resistive gauges (ESG) and, (ii) a 450mm long carbon rod with FBG strain sensors embedded in a 1mm deep grove and surface bonded resistive strain gauges (ESG).
For each type of test conducted, one set of recorded test data is shown on Figures 11, and 12. Figure 11(a) and (b) show data from the carbon fibre rod with embedded FBG sensors and corresponding surface bonded ESG gauges. Figure 12 (a) and (b) show the measured strain against time from both the surface bonded FBG sensors and co-located ESG gauges.
In each of the tests conducted the shift in the reflected wavelength and thus the strain meas- ured by the FBG sensors, whether surface bonded or embedded within the carbon fibre composites, shows similar temporal strain variations as that from the co-located ESG strain gauges with a good overall linear agreement.
Figure 10b: x3 FBG strain sensors and a resistive strain gauge (RSG) surface bonded to a carbon fibre strip
Figure 10a x2 FBG strain sensors em- bedded in a 1mm wide grove in the car- bon fibre rod and x3 RSG surface bonded sensors for comparison
0 500 1000 1500 2000
0 500 1000 1500 2000 2500 3000
Strain (με)
Time (AU)
ESG
0 100 200 300 400 500 600 700
-500 0 500 1000 1500 2000 2500 3000
Strain (με)
Time (AU)
FBG
Figure 11: Showing tensile strain readings from, (a) surface bonded FBG sensor and (b) co-located ESG sensor on a carbon fibre test strip
(a) (b)
22
From the results it can be seen that use of embedded fibre Bragg grating sensors for strain monitoring has been demonstrated to be as good as surface bonding in relation to strain transfer and sensitivity with added benefits of sensor protection for long term harsh environ- ment use. Embedded sensors also offer sensor packages, which are easy and quick to in- stall even in the most remote/limited access and harsh environments.
This extends the use of Bragg grating sensors as smart packaged sensors for routine applications either in all- composite structures or as part of smart composite repair elements using com- posite materials.
8.2 Steel rebars
In collaboration with the EU 5th Frame- work project: “Re-use of Foundations for
Urban Sites” (RuFUS) (http://www.reuseoffoundations.com/),
FBG sensors were attached to rebars and strain and temperature changes monitored during the curing cycle of concrete piles and the subsequent ef- fects of building construction. Here the smooth edge on the rebar is polished and the sensors glued and then pro- tected by a clip-on carbon fibre compos- ite (Cranfield University) against com- pressive strain across the fibre radius and finally protected by a layer of silicone resin. Such rebar sensor transducers can be used in monitoring foundation piles of load bearing concrete beams and panels.
0 100 200 300 400 500
-500 0 500 1000 1500 2000 2500 3000 3500
Strain(με)
Time(AU)
ESG2
0 1000 2000 3000 4000 5000 6000
0 800 1600 2400 3200
Strain(με)
Time (AU)
FBG2
Figure 12: Tensile strain readings from, (a) embedded FBG sensor and (b) co-located sur- face bonded ESG
(a) (b)
Figure 13 Scheme for using FBGS on rebars
Figure 14 Carbon composite protection sheathing
8.3 Carbon Fibre Panel
FBG sensors can be glued, using cyano-acrylate or epoxy, to the surface or in channels milled in reinforcement composite panels or rods incorporated at the build stage or later to repair or enhance the load bearing capacity of structures.
Figure 15 shows an example of gratings attached to carbon fibre composite installed on one of the concrete box section railway bridges for
the light railway transport system in Stockholm, Sweden. The figure also shows gratings at- tached to the surface of buffed smooth concrete.
This is acceptable for short term measurements, however concrete is relatively hydroscopic and uptake for water in the concrete will affect the peal strength of the bond between the grating and the concrete that may result in creep or de- tachment of the sensor.
8.4 Class fibre rods
In collaboration with Luleå University FBG sen- sors were embedded in Glass fibre rods and used as embedded transducers in concrete structure monitoring. Figure 16 shows the as- sembled concrete beams used for accelerated testing of concrete corrosion. This is an example where foil gauge sensors are less suitable. The figure shows 12 meter long concrete beams in- corporating glass fibre rods with arrays of 8 FBG sensors each. The beams were constructed and used in accelerated corrosion test in Luleå Uni- versity.
Figure 15 Fibre Bragg sensors on polished concrete and re-enforcement carbon com- posite panels
Figure16 12-meter long concrete beams incorporating glass fibre rods with arrays of 8 FBGS each.
24
8.5 Composite carbon fibre reinforcement rods
Örnsköldsvik Bridge
Composite fibre re-enforcement rods have application in repair and performance enhancement of existing strucure as well as application in monitoriing new structures. Figure 17 and 18 shows FBG sensors Carbon Rods embedded in 18 metre long carbon fire rods used to re-enforce concrete structures in the Sustinable Bridge test bridge at Örnsköldsvik. This structure was re-enforced and then load tested to descruction
For the assembly of the FBG sensors a groove,
~2mm wide and ~1mm deep, was cut into the centre of a square cross section of 10mm of carbon fiber rod in order to accommodate fiber optic as shown in fig (18). The fibres were glued with low viscosity Cyano-Acrylate and then covered with epoxy.
Frovi Bridge
This is an ‘in service’ concrete railway bridge in Sweden. In September 2007, FBG sensors were embedded into Carbon fibre rod and tubes as part of a repair and strengthening pro- gramme to ensure the future performance of the structure. The method of installation is simi- lar to that used on the Örnsköldsvik Bridge. Figure 19 shows views of preparing the carbon fibre tube, testing the sensors on the tube in the lab, and installation on the bridge.
Figure 20 shows the response of the embedded FBG sensors to rail traffic on the bridge taken over a two hour period. The figures show, in figure 20a, the response of a temperature sensor and five strain sensors in the rod and, in figure 20b, from five strain sensors in the rod to rail traffic.
Figure 17 application of epoxy resin over fibre sensors incorporated in car- bon fibre rod
(a) (b) (c)
Figure 19 (a) Grinding a 1mm grove in the tube, (b) Carbon fibre Rod with embedded FBGS and in the foreground the prototype miniature FBGS interrogation unit consist of the two small black boxes in the near view. (c) Installing the tube in the bridge at Frovi
approximately 50μS. Comparing the response of the rod and tube sensors you can see that the strains in the rod are significantly higher than those in the tube, where there is little ob- served response. This is consistent with modelling and the stiffness of the bridge. The re- sponse of the sensors in the rod fall into two groups corresponding to the traffic on the two railway lines on the bridge. The first set consisting of rod sensors 2, 3 and 4, respond to one set of rail traffic and the other consisting of rod sensors 4, 5 and 6 responding to a second set of rail traffic. Sensors 5 and 6 respond in opposition with sensor 5 showing tension and sensor 6 compressive
strains.
The FBG sensors are attached and protected within groves in the rods and tubes using adhe- sives and epoxy previ- ously developed for com- posite reinforcement.
Therefore it is probable that the sensors will be available for measure- ment throughout the life- time of the reinforcement.
It is expected that any significant changes from the initial wavelength measurements of the sensors when they were installed will correspond to real changes in the static strain in the struc- ture. It is therefore likely that they can be used in long term monitoring of the continuing effective- ness of the reinforcing rods and tubes.
Figure 20 the response of the embedded FBG sensors to rail traffic
26
9 Conclusions
This report discusses the development and construction of two of fibre Bragg strain sensing instruments and the methods used in applying the technology as an enabling technology in the monitoring of strain in rail structures. The instrumentation and methodology are evaluated in experiments undertaken in a range of laboratory and site tests of relevance to the Sustain- able Bridges project. The work shows that FBG technology can be routinely applied in the lab and on construction sites albeit by trained personnel. Further, the application of this technol- ogy in combination with the technology of composite re-enforcement for structural strength- ening and reuse, is demonstrated and shows promise for contributing significantly to the up- take of composite re-enforcement.
References
♦ Alampalli S, Fu G, “Remote Bridge monitoring systems for bridge condition”, J Low Freq Noise Vibration, 16:43-58, 1997.
♦ Aleixandre, M., Corredera, P., Hernanz, M.L. and Gutierrez-Monreal, j., “Development of fibre optic hydrogen sensors for testing nuclear waste repositories”, Sensors and Ac- tuators B, 107, pp. 113-120, 2005.
♦ Ashoori R, Gebremichael YM, Xiao S, Kemp J, Grattan KTV and Palmer AW, “Time domain multiplexing for Bragg grating strain measurement sensor network,” Proceed- ings of Optical fibre Sensors, (OFS 13), S. Korea, 1998, pp. 308- 311.
♦ Bastianini F, Di Tommaso A and Pascale G, “Ultrasonic non-destructive assessment of bonding defects in composite structural strengthening”, Compos Struct., 563, 463-7, 2001.
♦ Berkoff TA, Davis MA, Bellemore DG and Kersey AD, “Hybrid time and wavelength di- vision multiplexed fibre Bragg grating sensor array,” SPIE vol. 2444, 1995, pp. 288- 294.
♦ Casas JR, Cruz PJS. “Fibre optic sensors for bridge monitoring”, J Bridge Eng., 8, 362- 73, 2003.
♦ Clark M R, McCann DM and Forde MC, “Application of infrared thermography to the non-destructive testing of concrete and masonry bridges”, NDT&E Int 36, 265-75, 2003.
♦ Cremona C and Carracilli J, “Development and use of smart acquisition systems for bridge monitoring”, Key Engineering Materials, 204-205, 47-56, 2001.
♦ Ferdinand P, Ferragu O, Lechien JL, Lescop B, Marty V, Rougeault VS, Pierre G, Re- nouf C, Jarret B, Kotrotsios G, Neuman V, Depeursings Y, Michel JB, Uffelen MV, Ver- bandt Y, Voet MRH and Toscano D, “Mine operating accurate stability control with opti- cal fibre sensing and Bragg grating technology,” The BRITE EURAM STABILOS Pro- ject Prc. SPIE 2360, 1994, pp 162-166.
♦ Gebremichael, Y.M., Li, W., Boyle, W.J.O., Meggitt, B.T., Grattan, K.T.V., McKinley B., Fernando, G.F., Kister, G., Winter, D., Canning, L. and Luke, S., “Integration and as- sessment of fibre Bragg grating sensors in an all-fibre reinforced polymer composite road bridge”, Sensor Actuat. A- 118, 78-85, 2005.
♦ Gebremichael, Y.M., Li, W., Meggitt, B.T., Boyle, W.J.O., Grattan, K.T.V., Mckinley, B., Boswell, F., Arnes, K.A., Aasen, S.E., Tynes, B., Fonjallaz, Y. and Triantafillou, T., “A Field deployable, multiplexed Bragg grating sensor system used in an extensive high- way bridge monitoring evaluation tests”, IEEE Sensors Journal, Vol.5,No.3, 2005.
♦ Glisic, B.and Inaudi, D., “Sensing tape for easy integration of optical fibre sensors in composite structures” 16th International Conference on Optical Fibre Sensors, OFS-16, pp. 482-485, 13-17 October 2003. Japan.
♦ Grattan KTV and Meggitt BT (Eds), “Optical Fibre Sensor Technology: Advanced Ap- plications,” Kluwer Academic Publishers. Dordrecht, The Netherlands 2000, pp. 79 – 187.
♦ Grattan KTV and Meggitt BT, “Optical Fibre Sensor Technology: Fundamentals,” Klu- wer Academic Press, 2000.
♦ Hecht, E. “Optics” Addison Wesley, 3rd Edition, 1998.
