About this Journal Submit a Manuscript Table of Contents
International Journal of Distributed Sensor Networks
Volume 2012 (2012), Article ID 804394, 18 pages
Research Article

Integrated Optical Fiber Sensing System by Combing Large-Scale Distributed BOTDA/R and Localized FBGs

School of Civil Engineering, Dalian University of Technology, Dalian 116024, China

Received 5 July 2012; Accepted 3 December 2012

Academic Editor: Gangbing Song

Copyright © 2012 Zhi Zhou et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.


Structural health monitoring (SHM) has been regarded as a significant tool for the safety of civil infrastructures. Local fiber Bragg grating (FBG) sensor and distributed Brillouin optical fiber sensor have been successfully applied in civil engineering fields. Unfortunately, neither single FBG nor single Brillouin sensing technique can satisfy the requirements of simultaneously positioning full-scale structural damages and accurate local damage details. So it still matters to establish balance between localized high-precision measurement and large-scale distributed measurement with lower accuracy. This paper introduces an integration system by combining distributed Brillouin and local FBG sensors on one single-mode optical fiber. The temperature self-compensation and the data processing method of the integration system have been investigated. The integrated FBG/Brillouin sensor has been further packaged by glass fiber-reinforced polymer to be used as a robust and smart structural component. Finally the proposed integrated sensing system has been applied in three case studies, including investigation of reinforced concrete beam, smart cable, and smart steel strand to verify the feasibility. The test results illustrate that the integrated system can simultaneously provide accurate strain measurement at critical points by FBGs and the rough distributed strain measurement by Brillouin sensor, which can be further applied to various applications.

1. Introduction

In the last two decades, a large amount of critical infrastructures, characterized as high cost, large scale and required long service life, have been built all over the world. These critical structures are expected to survive after a long-term service and during or even the rare events such as earthquakes and fires. To secure the survivability of the critical structures in a long run, their health conditions need to be monitored and assessed in real time. Thus, structural health monitoring (SHM) technology has attracted considerable interests worldwide. SHM systems have been successfully deployed to understand the loads, environmental changes, and structural behaviors, which have also been applied to a number of case studies, for example, structural behavior monitoring of Guangzhou TV Tower, Water Cube, and Tsing Ma Bridge [16].

In general, damages are randomly distributed through the structure, and their locations are usually hidden. Also, the structural damages are time dependent and interrelated with various factors. Thus, a cost-effective SHM system is desired to provide damage distribution and current local damage assessment at the same time. Optical fiber sensor, with its unique advantages of immunity to electromagnetic interference, absolute measurement, high durability, distributed sensing capability, and locally accurate measurement, potentially provides a promising solution towards these issues. Among various optical fiber sensors, fiber Bragg grating (FBG) sensor shows excellent sensitivity to the strain and temperature [716], leading broad applications in the civil, marine, and aerospace engineering. However, the FBG sensors can only provide the strain measurement at localized locations, missing the damages away from the FBG sensors.

On the other hand, Brillouin optic time domain analysis/reflectometry (BOTDA/R) is a fully distributed optical sensing technique, which attracted worldwide attentions in recent years. BOTDA/R can provide strain or temperature information along the whole length of a single-mode fiber. The BOTDA/R sensing technique has been employed to a large number of applications, including bridges, dams, pipeline, and tunnels [1723]. Unfortunately, BOTDA/R also possesses its own disadvantages such as low sampling rate, low accuracy, and poor spatial resolution, which violates the measurement requirements of high accuracy, high frequency real-time monitoring capability for the structures.

Therefore, taking the advantages of both FBGs and BOTDA/R technique, a sensing system by combining FBGs and BOTDA/R technique together may yield a possible solution to simultanously provide distributed damage distribution and localized damage details. The integrated sensing technique has been investigated for strain and temperature measurements [24, 25]. However, the signal interference between the FBGs and BOTDA/R technique has not been fully studied. In addition, its temperature compensation for strain measurement still opens. In the past few years, a number of temperature compensation have been proposed, including multiple parametersensing, employing inverse expansion materials, applying polymer package, combining with other sensing techniques, and absolute temperature compensation [26, 27]. However, the compensation temperature sensors used in these methods do not measure the same temperature as that of the strain sensor. Besides, extra spaces are needed to install these additional temperature sensors, which may not be satisfied in some cases.

In an effort of providing an SHM system with both accurate localized and distributed measurements, in this paper, an integrated optical fiber sensing system is proposed by combining distributed BOTDA/R sensor and localized FBGs in one single optical fiber. The proposed integrated optical sensing system is then called BOTDA/R-FBGs. The system setup of BOTDA/R-FBGs has been investigated. With measurement signal from these two different sensing principles, the proposed integrated optical sensing system owns the capability of temperature self-compensation. In addition, the data processing method to distinguish the influences from the FBGs to the Brillouin sensing signal has also been investigated. The proposed integrated optical sensor has been packaged by GFRP and further applied into several structural components, which are not only robust but also smart for SHM purpose. At last, the smart components have been employed to several case studies to verify their feasibilities, including reinforced concrete beams and smart cables. The test results show that the proposed integrated BOTDA/R-FBG system works well and can be further applied to other associated applications.

2. System Integration of the BOTDA/R-FBG Sensing System

2.1. Configuration of the BOTDA/R-FBG Sensing System

Figure 1 shows the operational principle of the proposed optical sensing system. In this BOTDR-FBG system, the FBG sensors and the BOTDR sensor share the same single-mode optical fiber, which function as both sensing and transmitting unit. For the FBG system, the FBGs in the optical fiber are used as local measuring points, while the optical fiber itself is used as distributed sensor for the BOTDR system.

Figure 1: Operational principles of the BOTDR-FBG system.

Figure 2 shows the schematic configuration of the BOTDR-FBG system. Here, we can use a fiber loop to design testing circuit at the end of OF sensor, so the BOTDA-FBG system can also be constructed. In the BOTDA-FBG system, light switch is used to separate the Brillouin and the FBG signal to avoid the damage of the FBG sampling system. In the following, this two hybrid sensing systems (BOTDA-FBG and the BOTDR-FBG systems) are named as BOTDA/R-FBG system. Light comes out from the light source, comes up to the light switcher, and travels through the single-mode optical fiber. When light encounters the FBGs, they get reflected back and the reflected signals are recorded on the FBG and BOTDA/R interrogators. With the light switcher, the light signals from BOTDA/R and FBGs can be recorded in time division independently. In this paper, an SI-720 instrument from Micron Optics Inc. has been used as the FBG interrogator and a DiTeSt STA 202 from Omnisens has been adopted as the BOTDA/R interrogator. On the single-mode optical fiber, several FBGs have been written to form an OF-FBGs sensor. To protect the OF-FBG sensor from external damage, it has been packaged by the highly durable material, such as polypropylene (PP) and fiber reinforced polymer (FRP). In field applications, the OF-FBGs sensor is installed along the structure member with the FBGs located at the potential damage locations. The actual damage locations and their spatial distribution will be identified from the BOTDA/R measurements while the detailed damage information can be acquired by the FBGs, respectively.

Figure 2: Configuration of the integrated BOTDA/R-FBG system.
2.2. Temperature Self-Compensation for the BOTDA/R-FBG Sensing System

Without considering the cross-sensitivity of strain and temperature, both the Bragg wavelength shift of FBG and the Brillouin frequency shift of BOTDA/R show excellent linearity to the strain and temperature as expressed below in binary matrix: where and are the changes of the Bragg wavelength and the Brillouin frequency; and are the changes of strain and temperature; and are the strain and temperature sensitivity of FBGs, which equal to 1.2 pm/με and 10.8 pm/°C, respectively [28]; and are the strain and temperature sensitivity of BOTDA/R, which have values of 0.05 MHz/με and 1.0 MHz/°C, respectively. So (1) can also be expressed as follows:

With measured changes of Bragg wavelength () and Brillouin frequency shift (), the changes of strain () and temperature () then can be predicted accordingly. Thus, the temperature effect to the BOTDA/R-FBG system can be canceled out and self-compensated.

Test has been set up to verify the self-temperature compensation as shown in Figure 3(a). One unpackaged OF sensor, one OF-FBGs sensor, and one unpackaged FBG have been placed into a water tank with a gauge length of 1 m as can be seen in Figure 3(b). The unpackaged OF and the unpackaged FBG are set without mechanical loads and used as temperature sensors for absolute temperature compensation. On the other hand, both mechanical and thermal loads have been applied on the OF-FBGs sensor by the tensile instrument and the water tank. All the measurements from the three sensors have been recorded by the SI720 instrument for FBGs and DiTeSt STA200 instrument for BOTDA/R. Figure 4 shows the test results of self-temperature compensation and Table 1 lists the measured strain and temperature from the OF-FBGs sensor by using (2). The strain measured by the OF-FBGs sensor with self-temperature compensation agrees well with the reference strain calculated from absolute temperature compensation by OF or FBG, with a maximum relative error of 9%. Some of the measured strains or temperatures have large deviation with the references because of the temperature instability of water tank. However, taking the advantage of the convenience of the simultaneous measurement of strain and temperature, the OF-FBGs sensor can be a promising and cost-efficient sensing technique for field practices.

Table 1: Simultaneous temperature and strain measurement by the OF-FBGs sensor.
Figure 3: Test setup of self-temperature compensation verification.
Figure 4: Test results for self-temperature compensation verification of OF-FBG.

The self-temperature compensation test has been demonstrated for the GFRP packaged OF-FBG sensors as well. A static load is applied to the GFRP packaged OF-FBGs sensor for a short period. Figure 5 compares the variance of the strain measured by GFRP packaged OF-FBGs sensor with and without self-temperature compensation for two temperature states of the test. It can be seen that compared to the strain measured by the OF-FBGs sensor without the self-temperature compensation, the corresponding strain with the self-temperature compensation is much smaller than the actual strain. The self-temperature compensation property of the OF-FBGs sensor needs no additional temperature sensors, providing a promising temperature compensation solution for the sensing system for various applications.

Figure 5: Field application of temperature intercompensation.

3. Signal Interactions and Processing

FBGs use the broadband light source, but BOTDA/R uses a light source with a certain wavelength of 1550 nm. Thus the potential signal interactions exist between the BOTDA/R and FBGs signal, including interaction of the reflected light from different optical light sources. In addition, FBGs are somehow regarded as damages on the single-mode optical fiber. Their structures, typically the Bragg period, change with various mechanical and thermal loading, resulting in a complex Brillouin gain spectrum at the locations of FBGs, and the Brillouin gain curve may turn complex under complex external force within the spatial resolution.

3.1. Interactions between BOTDA/R and FBG Signals

Two GFRP packaged OF-FBG sensors, named as GFRP-OF-FBG1 and GFRP-OF-FBG2, had been tested for strain sensing under tension loads, and the two GFRP-OF-FBG sensors are connected to each other by using a free optical fiber. Each GFRP-OF-FBG sensor has two FBGs. GFRP-OF-FBG1 has a gauge length of 40 m, and the two FBGs, FBG1 and FBG2, have initial center wavelengths of 1525 nm and 1527 nm, respectively, while GFRP-OF-FBG2 has a gauge length of 70 m, and the initial center wavelengths of its two FBGs, FBG3 and FBG4, are 1540 nm and 1545 nm, respectively. Figure 6(a) shows the strains measured by GFRP-OF-FBG1 under various tension loads. With the initial wavelengths of the FBG1 and FBG2 far away from 1550 nm, the strains measured by the BOTDA/R agree well with the strain measured by FBG1 and FBG2. From Figure 6(a), it has been seen that FBGs do not affect the Brillouin signal much if the center wavelength of FBGs is away from the wavelength of BOTDA/R instrument light source, 1550 nm. Figure 6(b) shows the strains measured by GFRP-OF-FBG2 under various tension loads. The obtained distributed strain of the BOTDA/R has a huge intensity loss at the locations closed to FBG3 and FBG4, which have initial center wavelengths close to 1550 nm. The more the wavelength of FBG is close to 1550 nm, the more the intensity loss of the BOTDA/R signal will be. In the two figures, there is a wave trough between the two FBGs at each load, which is the measuring value of the free optical fiber.

Figure 6: Strain measured by GFRP packaged OF-FBGs.
3.2. Signal Processing of the OF-FBGs Sensor

In an effort to eliminate the interactions of the FBGs to the Brillouin signal in the condition when the FBGs have initial center wavelengths around 1550 nm, in this paper, several signal processing techniques have been investigated and employed to denoise Brillouin signal, including wavelet analysis and curve fitting algorism.

3.2.1. Wavelet Denoising for the OF-FBGs Sensor

Denoising is a filter process that can get rid of the high-frequency noise and reconstruct the low-frequency signals. It uses the signal property that the Lipschitz index of noise is negative. The maximum modulus of noise decreases as the scale increases, but the opposite applies for the actual signal. Various denoising processes exist, including fast Fourier transfer method and wavelet analysis. Wavelet analysis is an effective signal processing method to denoise signal in field application [29, 30]. The capability of multiscale enables the wavelet analysis to comprehensively show various abnormalities and changes of signal at different scales. In this paper, the nonsymmetric Daubechies wavelet (dbN) is used to denoise the measured Brillouin signal. Experiments had been employed to verify the signal processing based on wavelet analysis by installing one GFRP packaged OF-FBGs sensor on a structure under various tensile loads. The GFRP-OF-FBGs sensor has a gauge length of 20 m and the initial wavelengths of the two FBGs are 1525 nm and 1527 nm, respectively. Figure 7(a) shows the strain distribution of the GFRP-OF-FBGs at various loads. The measured strain spectrum shows a number of burrs or noise, caused by the interactions between FBGs and BOTDA/R and the instability of the measuring equipments such as the FBG and Brillouin instruments themselves and the loading devices. To eliminate the effect from these interactions, a db5 wavelet was selected to denoise the measured signal. Figure 8 shows the decomposition of the measured signal by using the db5 wavelet. The maximum modulus of the high-frequency components, denoted as D1, D2, and D3, decreases as the scale factor increases, as shown in Figure 8(a). On the other hand, the maximum modulus of other high-frequency components, named as D4 and D5, indicates good correlation with low-frequency component, named as A5, as shown in Figure 8(b). So the noise signal can be reconstructed from D3, D2, and D1, while the actual signal can be reconstructed by combining the low-frequency component of the measured signal, A5, and the high-frequency components, D5 and D4. Figure 7(b) shows the obtained strain after denoising by wavelet dB5. The strains flow smoother than the initial measured signals.

Figure 7: Measured Brillouin signal from OF-FBG sensor.
Figure 8: Signal reconstruction for 5-scale one-dimensional wavelet decomposition.

One strain gauge was installed at the end of GFRP packaged OF-FBGs to validate the strain measurement of the GFRP-OF-FBGs sensor. Figure 9(a) illustrates the strain measurement comparison of the GFRP-OF-FBGs sensor before and after db5 wavelet denoising to that from the strain gauge. Figure 9(b) shows the corresponding relative error. With the wavelet denoising, both the strain measured by the BOTDA/R and that by the FBGs agree well with that from the strain gauge. The maximum relative error of the GFRP-OF-FBGs sensor is 18% before denoising and 4% after denoising. The results indicate the feasibility of the wavelet denoising method for the data processing of OF-FBGs sensors.

Figure 9: Strain measurement comparison before and after denoising.
3.2.2. Brillouin Gain Spectrum Curve Fitting Algorithm for Complex Brillouin Gain Curve

Brillouin distributed optical fiber sensing technology is based on the assumption that Brillouin frequency shift (BFS) obtained from Brillouin gain spectrum linearly depends on the environmental temperature or applied strain on optical fiber. With an axial tensile load on the optical fiber, the Brillouin gain spectrum has a Lorentz profile, while it becomes complex for complex stress condition or damage within the spatial resolution. Figure 10 shows the comparison of the 3-D mapping of one GFRP packaged OF-FBGs’ Brillouin gain spectrum to that of an unpackaged OF at certain temperature. Brillouin gain spectrum for the GFRP packaged OF-FBGs sensor turns to be more complex than that of the unpackaged OF, especially at the locations of FBGs. This complexity is caused by the initial stress induced by the packaging process and the interaction between the FBGs and the Brillouin spectrum. Typically, the algorithm to calculate the Brillouin frequency shift is not flexible, which cannot meet the required measurement accuracy of temperature or strain. Thus, in this paper, local Lorentz single-peak and localized least square fitting algorithms are proposed to fit the Brillouin gain spectrum. By applying optimization, a selected bandwidth of 50 MHz has been employed to fit the measured Brillouin gain spectrum with the proposed algorithms.

Figure 10: 3-D mapping of Brillouin gain spectrum.

Experiments had been set up to verify the two proposed spectrum fitting algorithms, as shown in Figure 11. One GFRP packaged OF-FBGs sensor had been placed into a water tank with various temperature measurements. The test condition includes heating process and cooling process. The temperature of the water tank is ranged from 23°C to 64°C. Figure 12 shows the Brillouin gain spectrum and the corresponding fitted curve by using the local Lorentz single-peak fitting algorithm at FBG location at various temperatures. Figure 13 illustrates the obtained Brillouin gain spectrum by using the localized least square fitting algorithm. Figure 14(a) shows the comparison of measured temperatures by BOTDA/R after using these two fitting algorithms. Figure 14(b) shows the corresponding relative error. The maximum errors from local single-peak Lorentz, localized least square fitting algorithms, and BOTDA/R measurement are 3.7%, 6.9%, and 7.5%, respectively. The test results indicated that, with the proposed curve fitting algorithms, the interaction of the FBGs to the Brillouin gain spectrum of OF-FBGs sensors can be eliminated.

Figure 11: Setup of temperature test.
Figure 12: Fitted BFS spectrum by Lorentz local single-peak fitting.
Figure 13: Fitted BFS spectrum by localized least square fitting.
Figure 14: Comparison of sensing temperature from various fitting methods.

4. Experimental Validation of GFRP-OF-FBGS Sensors

Considering the fragility of the optical fiber, proper packaging method is required for various practical applications. GFRP has the advantages of high strength, corrosion and fatigue resistance, and linear elasticity. More importantly, GFRP is naturally compatible with the optical fiber. Thus, GFRP is a potential packaging candidate for OF-FBGs sensors. In this paper, the fabrication of the GFRP packaged OF-FBGs (GFRP-OF-FBGs) sensors, their feasibility, and applications had been investigated as follows.

4.1. Fabrication of GFRP-OF-FBGs Sensor

In an effort to meet various layout requirements of the installation in different applications, two types of GFRP-OF-FBGs sensors are proposed, including the GFRP-OF-FBG-based rod and the GFRP-OF-FBG-based belt as shown in Figures 15(a) and 15(b), respectively. The GFRP-OF-FBG-based rod is able to be embedded into various structures for internal strain measurements. On the other hand, the GFRP-OF-FBG-based belt can be easily bonded to the surface of structures for surface strain measurements. Figures 16(a) and 16(b) show the fabricated prototypes of these two types of sensors. The anchors on both ends of the sensor are designed to improve the bonding between the host matrix and the sensors. The length of these anchors can be designed based on the requirements of the dynamic range of the strain measurements. In addition, with the GFRP packaging layer, several OF-FBGs sensors can be packaged into one GFRP layer for multiple parameters sensing.

Figure 15: Types of GFRP-OF-FBG-based sensors.
Figure 16: Sensor prototypes of the GFRP-OF-FBG-based sensors.
4.2. Experimental Verification of the OF-FBG-Based Sensors for Strain Sensing

The feasibility of the single-line OF-FBG sensors had been verified by series of experiments. One unpackaged OF-FBG sensor, one unpackaged OF sensor, and one GFRP-OF-FBG sensor were tested under static tension loads. One FBG had been placed on each of the OF-FBGs and the GFRP-OF-FBG sensor. The initial wavelengths of FBGs of the unpackaged OF-FBG and the GFRP-OF-FBG sensor were 1555 nm and 1566 nm, respectively. The gauge length of the unpackaged OF-FBG and OF sensors was 4.5 m. Considering the small diameter of the unpackaged OF and OF-FBG sensors, their tensile test had been set up as shown in Figure 17(a). The sensors were attached to two stages by using epoxy resin. One of the two stages was automatically controlled for axial movement. Thus, the reference strain applied to the attached sensors can be calculated by the relative movement of the stages. Figure 17(b) compares the measured strains by the unpackaged OF and OF-FBG sensors with the reference strain. The maximum relative error of the unpackaged OF-FBG is 6.5%, indicating that the measured strains by the unpackaged OF-FBG sensors agree well with the reference strain. The GFRP-OF-FBG sensor has a gauge length of 0.7 m. It had been tested by a tension machine as shown in Figure 18(a). The reference strain of the GFRP-OF-FBG sensor can be obtained by the attached extensometer as also shown in Figure 18(a). Figure 18(b) shows the strains measured by the GFRP-OF-FBG sensor compared to the reference strains. Comparing to the reference strains, the maximum relative measurement error of the GFRP-OF-FBG sensor is 11% application. The large error is mainly caused by the spatial resolution of the BOTDA/R measurement for the GFRP-OF-FBG sensor.

Figure 17: Strain measurement by unpackaged OF-FBG sensor.
Figure 18: Strain measurement by the GFRP packaged OF-FBG sensor.
4.3. Strain Measurement of Concrete Beam Based on the GFRP-OF-FBG Sensors

The proposed GFRP-OF-FBG sensors had been applied to measure the strain distribution of a referenced concrete (RC) beam to prove their feasibility of field applications. The RC beam had a dimension of 2700 mm × 200 mm × 400 mm and it was loaded by a two-point loading. Loads were supplied by a hydrostatic machine with a load step of 6 kN. In each load step, a static load had been maintained for 5 min for the strain measurement purpose. Figure 19 shows the schematic test setup and Figure 20 shows the test scene of the sensor installation. Various sensors had been placed on the bottom surface of the RC beam, including commercial electrical resistance strain (ERS) gauges, two unpackaged OF sensors, one unpackaged FBG sensor, one GFRP-OF sensor, and one GFRP-OF-FBG sensor. The GFRP-OF and GFRP-OF-FBG sensors had the same gauge length of 120 cm and they were fabricated in the shape of the rod. Both ends of the GFRP-OF and GFRP-OF-FBG sensors were welded onto the preembedded steel block on the RC beam to ensure perfect bonding conditions. The GFRP-OF-FBG, GFRP-OF, and unpackaged OF sensors were parallel to each other, and the ERSs were located at middle of the RC beam. The initial wavelength of FBG of the GFRP-OF-FBG-based rod is 1565 nm, which is far away from 1550 nm.

Figure 19: Schematic of the test setup and sensor layout.
Figure 20: Test scene of sensor installation.

Figures 21(a) and 21(b) show the experimental results of strain measurements of the RC beam from various sensors mentioned above, for loading and unloading cycles. Strains measured by the GFRP-OF-FBG sensor agree with that from the GFRP-OF, unpackaged OF, and reference ERS sensors for the loading-unloading cycles. The maximum relative error between strain measured by the BOTDA/R and that by the FBGs is of 6%. Figure 22 shows strain distribution of the GFRP-OF-FBG sensor measured by the BOTDA/R system. The sensed strain from BOTDA/R system at the middle of the RC beam agrees well with that measured from the FBG technique. Figure 23 shows the comparison of dynamic range of the strain-to-load measurement for various sensors at the middle of the RC beam. The measured strain under load of 45 kN was away from the initial strain-load curve, indicating the occurrence of potential damages. Visual inspection found out that cracks with a length of 0.3 mm were viewed at the middle bottom of the RC beam. As the load applied on the RC beam increases, cracks expand and the measured strain-load curves from various sensors began to separate from each other. When load went beyond 160 kN, the measured strain from the GFRP-OF-FBG sensor dropped quickly with the increase of the load, while that from other sensors increased. Visual inspection found that slippage occurred on the anchor of the GFRP-OF-FBG sensor, resulting in the sudden increase of the measurement error. The slippage indicated that the dynamic range of the GFRP-OF-FBG sensor for strain measurement was highly dependent on the bonding between the sensor and the host matrix. The comparison of the GFRP-OF-FBG sensor to other types of sensors indicates that with proper sensor installation the proposed GFRP-OF-FBG sensor is able to measure strain accurately. Thus, the proposed GFRP-OF-FBG sensor is a promising solution for simultaneous localized and distributed strain measurement in practical applications.

Figure 21: Comparison of the strain of the RC beam by various sensors in the loading-unloading cycle.
Figure 22: Measured strain distribution by the GFRP-OF-FBG sensor.
Figure 23: Strain-load measurement of failure of the RC beam from various sensors.

5. Smart Structural Components Based on the GFRP-OF-FBG Sensors and Their Applications

5.1. Smart Steel Strand Based on the GFRP-OF-FBG Sensors

A smart steel strand has been proposed based on the GFRP-OF-FBG sensors to monitor its health condition during its service life. Figure 24 shows the schematic of the configuration of the proposed smart steel brand. It consists of a GFRP-OF-FBG-based rod and six steel wires. So the internal force of the smart steel brand consists of the force applied on the steel wire and the FRP rod. The GFRP-OF-FBG-based rod is located in the middle of the smart strand and it has a diameter of 5 mm. As the strand deforms, the GFRP-OF-FBG-based rod will deform together with the six steel wires. With the real-time measurement of the strain change of the GFRP-OF-FBG-based rod, the deformation of the steel strand can be directly obtained. The health condition assessment of the steel strand can then be employed based on these measured strain information. In the practical application, the installation of the proposed smart steel strands is the same as the traditional strands.

Figure 24: Configuration of a smart steel strand.

With the measured strain from the GFRP-OF-FBG-based rod, the internal force of smart steel strand can be obtained as follows: where , and are the strain of steel wires, the strain of the GFRP-OF-FBG-based rod, internal force of steel strand, cross-section of the six steel wires, and the cross-section of the GFRP-OF-FBG-based rod, respectively.

One proposed smart steel strand with a length of 3 m was fabricated and tested. The initial wavelength of the FBG on the GFRP-OF-FBG sensor was 1527 nm. The smart steel strand was tested under tension as shown in Figure 25. Figure 26 shows the measured strain distribution of the smart steel strand under various loads. The GFRP-OF-FBG-based rod not only provided accurate strain in local FBG location, but also gave out the strain distribution along the whole GFRP-OF-FBG sensor. The strain distribution measured by the GFRP-OF-FBG sensor agrees well with that measured by the localized FBG. Thus, the internal force distribution along the smart steel brand can be obtained based on (3). Table 2 illustrates the comparison of the calculated internal force compared to the applied force. The maximum error between the BOTDA/R measurement and the FBG is 8.7%, indicating that the proposed smart steel strand is able to assess the health condition of the steel strand.

Table 2: Comparison of the calculated internal force compared to the applied force.
Figure 25: Tension test of the proposed smart steel strand.
Figure 26: Measured strain distribution of smart steel strand by the GFRP-OF-FBG sensor.
5.2. Smart Stay Cable Based on the GFRP-OF-FBG Sensor

A smart stay cable is proposed based on the combination of the GFRP-OF-FBG sensors with the stay cable wires. Two GFRP-OF-FBG-based rods were symmetrically placed in the stay cable together with other steel cable wires as shown in Figure 27. Two ends of the GFRP-OF-FBG-based rods and the steel cable wires were anchored together by the mixed epoxy and iron sand in the anchor cup. With torsion-induced frictions between the steel wires and the GFRP-OF-FBG-based rods, the GFRP-OF-FBG-based rods are supposed to deform cooperatively with the steel wires. Thus, the behavior of the stay cable can be monitored by the GFRP-OF-FBG sensors. Figure 28 shows the fabrication procedures for the proposed smart stay cable. Firstly, two GFRP-OF-FBG-based rods with the same length of the cable wires were symmetrically laid along with the cable wires as shown in Figures 28(a) and 28(b). The smart stay cables were then protected by polyethylene tube, as shown in Figure 28(c). Secondly, the GFRP-OF-FBG-based rods were placed in series and were fusion-spliced together. The splicing point was protected by a brass tube as shown in Figure 28(d). Finally, a mixture of epoxy and iron sands was infused in the anchor and solidified by heating, as shown Figures 28(e) and 28(f).

Figure 27: Layout of the smart stay cable.
Figure 28: Fabricating process of the proposed smart stay cable.

Figure 29 shows the schematic diagram of the real-time monitoring system for the proposed smart stay cable. The FBGs are located at critical locations, including the middle and the anchorage zone of the smart cable. With the strain information measured from the GFRP-OF-FBG sensors embedded in the smart cable, the cable fore () can be calculated as follows: where and and and and are the forces, elastic modulus, area of the cross-section, and strain of the steel wires and the GFRP-OF-FBG-based rods, respectively; is the number of wires in the steel cable.

Figure 29: Schematic diagram of the real-time monitoring system for the proposed smart stay cable.

Based on the calibrated strain sensitivities of FBGs and Brillouin technique, the obtained force sensitivities of four types of smart stay cables were calculated and listed in Table 3. It can be seen that when the number of the steel wires increases, the force sensitivity of the GFRP-OF-FBG sensor decreases significantly.

Table 3: Cable force sensitivity of the proposed smart cables with various numbers of steel wires.

To verify the feasibility of the proposed smart stay cable for strain measurements, a series of experiments had been employed. A smart steel cable with two GFRP-OF-FBG-based rods, (named GFRP-OF-FBG1 and GFRP-OF-FBG2, resp.) had been fabricated. The initial wavelengths of the two FBGs in the two FRP-OF-FBG sensors were the same, with values of 1525 nm and 1527 nm, which were far away from 1550 nm. The smart cable had a length of 20.3 m. It had 61 steel wires. Each wire had a diameter of 7 mm and an elastic modulus of 210 Gpa. The fracture force of the smart cable, , was expected to be of 3990 kN. An ultratension test was conducted on a horizontal tensile testing machine at the JuLi Group Ltd. in China, as shown in Figure 30. The tension machine has a loading capacity of 1000 T. The load sequences of test were 878 kN, 1000 kN, 1317 kN, 1756 kN, 2195 kN, 1756 kN, 1317 kN, 1000 kN, and 878 kN. Two pressure sensors were applied on both ends of the test machine to control the applied force. The applied force was kept for 3 min at each load step for strain measurements.

Figure 30: Static test of the smart stay cable.

Figure 31(a) shows the measured cable force distribution at various loads by the two installed GFRP-OF-FBG sensors and Figure 34(b) shows the average measured cable force by the two sensors. The variance of frictions along the cable induced the fluctuation of measured cable forces along the cable length under axial tension. Here, the measuring signals were denoised by the wavelet above mentioned.

Figure 31: Distribution of measured cable force.

Figures 32(a) and 32(b) show the comparison of the measured cable force and the relative error by the distributed GFRP-OF-FBG sensor and the localized FBGs on the sensor to that from the commercial presser sensor. The cable forces measured by the proposed smart stay cable agree well with the applied load measured by commercial pressure sensors with a relative error of less than 5%. Therefore, the proposed smart stay cable is applicable for the practices.

Figure 32: Cable force comparison by various sensing techniques.
5.3. Prestress Monitoring of a Steel Frame Using the Proposed Smart Steel Strand

The prestress loss of a steel frame structure was monitored by the proposed smart steel strand. Figure 33 shows the dimension of the monitored steel frame and the setup of the prestress. The steel frame was made up by three 1.0-meter-long sections and was prestressed by a hydraulic jack. The jacking force was measured by a pressure sensor. A seven-wire smart steel strand was placed along the steel frame to monitor the prestress loss. After the jacking force reached a predetermined level, the smart steel strand was anchored. Other than the GFRP-OF-FBG sensor, a GFRP-OF sensor was also installed in the smart steel strand. The prestress loss in the smart steel strand mainly resulted from the elastic deformation of the steel frame and the initial slippage of the anchorage after the jacking force had been transferred to the frame structure.

Figure 33: Setup of a steel frame for prestress loss monitoring by the proposed smart steel strand.
Figure 34: Monitoring of prestress loss by the proposed smart steel strand.

Figure 34 illustrates the changes of prestress and the corresponding prestress loss at the middle of the steel strand as a function of time over a period of 100 hours. It can be seen that the prestress loss occurs mainly at the beginning, induced by the transfer of a jacking force to the steel frame. The measured prestress loss from the proposed smart steel strand agrees well with that from the pressure sensor and the FBG sensor, indicating that the proposed smart steel strand is able to apply in prestress loss for structural practices. Figure 35 shows a temporal-spatial prestress force distribution of the steel strand with various time periods. Figures 34 and 35 indicate that the stress in the steel strand remains almost constant after the initial prestress loss.

Figure 35: Stress distribution along the steel strand within 100 hours.

6. Conclusions

This paper presents a novel optical integrated sensing technology by combing distributed fiber-optic Brillouin sensing technology (BOTDA/R) and local fiber Bragg grating (FBG) sensing technique, named the BOTDA/R-FBG sensing technique. Based on the discussion of principles, signal processing methods, self-temperature compensation, packaging, and applications, the following conclusions are obtained.(a)In the BOTDA/R-FBG system, FBGs do not affect the Brillouin signal unless the FBGs’ wavelengths are close to 1550 nm. Wavelet denoising method and local single-peak Lorentz fitting algorithm can effectively eliminate the signal interaction between FBGs and the BOTDA/R technique, when the FBGs’ wavelength is close to 1550 nm. Thus, the strain and temperature measurement accuracy can be improved. (b)The measured strain by self-temperature compensation agrees well with that by the absolute temperature compensation. Considering that the self-temperature compensation can reduce the cost and complexity of sensor installation significantly, the proposed integrated sensing technique is cost efficient. (c)To protect the OF-FBG sensor from potential frugality, GFRP material is proposed to be used as packaging materials. With comparisons to various sensing techniques, the GFRP-OF-FBG sensor had been validated for strain measurement by experiments.(d)Smart steel strand and smart stay cable based on the GFRP-OF-FBG sensors had been investigated. The feasibility of these smart components for strain and force monitoring had been verified by experiments and practical application case studies.

Based on all the above investigation, it is confident to draw the conclusion that the proposed BOTDA/R-FBG collinear technique introduced in this paper is a promising method to meet the need of the large-scale strain measurements for practical application, such as airports, large-scale dams, and highways.


The authors are grateful for the financial support from the National Natural Science Foundation of China (NSFC) under Grant nos. 50978079, 61040031, and 51108065, China Postdoctoral Science Foundation under Grant no. 20100481233, the National Science-Technology Support Plan Projects (2011BAK02B02), the National Program on Key Basic Research Project (973 Program) (2011CB013700), and the Ministry of Education Program for New Century Excellent Talents.


  1. H. Li, Y. Q. Bao, and J. P. Ou, “Structural damage identification based on integration of information fusion and shannon entropy,” Mechanical Systems and Signal Processing, vol. 22, no. 6, pp. 1427–1440, 2008. View at Google Scholar
  2. J. Ou and H. Li, “Structural health monitoring in mainland china: review and future trends,” Structural Health Monitoring, vol. 9, no. 3, pp. 219–231, 2010. View at Publisher · View at Google Scholar · View at Scopus
  3. F. Marazzi, P. Tagliabue, and F. M. Corbani, “Traditional versus innovative structural health monitoring of monumental structures: a case study,” Structural Control and Health Monitoring, vol. 18, no. 4, pp. 430–449, 2011. View at Publisher · View at Google Scholar · View at Scopus
  4. H. Li, J. Ou, X. Zhao et al., “Structural health monitoring system for the Shandong Binzhou Yellow River Highway Bridge,” Computer-Aided Civil and Infrastructure Engineering, vol. 21, no. 4, pp. 306–317, 2006. View at Publisher · View at Google Scholar · View at Scopus
  5. Y. Q. Ni, Y. Xia, W. Y. Liao, and J. M. Ko, “Technology innovation in developing the structural health monitoring system for Guangzhou New TV Tower,” Structural Control and Health Monitoring, vol. 16, no. 1, pp. 73–98, 2009. View at Publisher · View at Google Scholar · View at Scopus
  6. T. H. T. Chan, L. Yu, H. Y. Tam et al., “Fiber Bragg grating sensors for structural health monitoring of Tsing Ma bridge: background and experimental observation,” Engineering Structures, vol. 28, no. 5, pp. 648–659, 2006. View at Publisher · View at Google Scholar · View at Scopus
  7. I. Talebinejad, C. Fischer, and F. Ansari, “Low frequency fiber optic accelerometer for civil structural health monitoring,” in Nondestructive Characterization for Composite Materials, Aerospace Engineering, Civil Infrastructure, and Homeland Security, vol. 7294 of Proceedings of SPIE, pp. 1–13, March 2009. View at Publisher · View at Google Scholar · View at Scopus
  8. A. Hongo, S. Kojima, and S. Komatsuzaki, “Applications of fiber Bragg grating sensors and high-speed interrogation techniques,” Structural Control and Health Monitoring, vol. 12, no. 3-4, pp. 269–282, 2005. View at Publisher · View at Google Scholar · View at Scopus
  9. H. N. Li, D. S. Li, and G. B. Song, “Recent applications of fiber optic sensors to health monitoring in civil engineering,” Engineering Structures, vol. 26, no. 11, pp. 1647–1657, 2004. View at Google Scholar
  10. Z. Zhang, Z. Zhou, C. Wang, and J. Ou, “Long-term monitoring FBG-based cable load sensor,” in Smart Structures and Materials: Smart Sensor Monitoring Systems and Applications, vol. 6167 of Proceedings of SPIE, pp. 1–7, March 2006. View at Publisher · View at Google Scholar · View at Scopus
  11. S. W. Lu and H. Q. Xie, “Strengthen and real-time monitoring of RC beam using “intelligent” CFRP with embedded FBG sensors,” Construction and Building Materials, vol. 21, no. 9, pp. 1839–1845, 2007. View at Publisher · View at Google Scholar · View at Scopus
  12. T. A. Berkoff and A. D. Kersey, “Experimental demonstration of a fiber bragg grating accelerometer,” IEEE Photonics Technology Letters, vol. 8, no. 12, pp. 1677–1679, 1996. View at Google Scholar · View at Scopus
  13. H. Li, J. Ou, and Z. Zhou, “Applications of optical fibre Bragg gratings sensing technology-based smart stay cables,” Optics and Lasers in Engineering, vol. 47, no. 10, pp. 1077–1084, 2009. View at Publisher · View at Google Scholar · View at Scopus
  14. S. C. M. Ho, H. Razavi, A. Nazeri, and G. B. Song, “FBG sensor for contact level monitoring and prediction of perforation in cardiac ablation,” Sensors, vol. 12, no. 1, pp. 1002–1013, 2012. View at Google Scholar
  15. W. Chung and D. Kang, “Full-scale test of a concrete box girder using FBG sensing system,” Engineering Structures, vol. 30, no. 3, pp. 643–652, 2008. View at Publisher · View at Google Scholar · View at Scopus
  16. L. Li, X. L. Tong, C. M. Zhou, et al., “WenIntegration of miniature Fabry-Perot fiber optic sensor with FBG for the measurement of temperature and strain,” Optics Communications, vol. 284, no. 16, pp. 1612–1615, 2011. View at Publisher · View at Google Scholar
  17. T. Horiguchi, T. Kurashima, and M. Tateda, “Tensile strain dependence of Brillouin frequency shift in silica optical fibers,” IEEE Photonics Technology Letters, vol. 1, no. 5, pp. 107–108, 1989. View at Publisher · View at Google Scholar · View at Scopus
  18. D. Culverhouse, F. Farahi, C. N. Pannell, and D. A. Jackson, “Potential of stimulated Brillouin scattering as sensing mechanism for distributed temperature sensors,” Electronics Letters, vol. 25, no. 14, pp. 913–915, 1989. View at Google Scholar · View at Scopus
  19. X. Bao, C. Huang, X. Zeng, A. Arcand, and P. Sullivan, “Simultaneous strain and temperature monitoring of the composite cure with a Brillouin-scattering-based distributed sensor,” Optical Engineering, vol. 41, no. 7, pp. 1496–1501, 2002. View at Publisher · View at Google Scholar · View at Scopus
  20. Z. Wu, B. Xu, K. Hayashi, and A. Machida, “Distributed optic fiber sensing for a full-scale PC girder strengthened with prestressed PBO sheets,” Engineering Structures, vol. 28, no. 7, pp. 1049–1059, 2006. View at Publisher · View at Google Scholar · View at Scopus
  21. Z. Zhou, J. P. He, and K. Yan, “Fiber-reinforced polymer-packaged optical fiber sensors based on Brillouin optical time-domain analysis,” Optical Engineering, vol. 47, no. 1, pp. 1–10, 2008. View at Publisher · View at Google Scholar
  22. W. Zhang, J. Q. Gao, B. Shi, H. Cui, and H. Zhu, “Health monitoring of rehabilitated concrete bridges using distributed optical fiber sensing,” Computer-Aided Civil and Infrastructure Engineering, vol. 21, no. 6, pp. 411–424, 2006. View at Publisher · View at Google Scholar · View at Scopus
  23. F. Bastianini, A. Rizzo, N. Galati, U. Deza, and A. Nanni, “Discontinuous Brillouin strain monitoring of small concrete bridges: comparison between near-to-surface and “smart” FRP fiber installation techniques,” in Smart Structures and Materials. Sensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems, Proceedings of SPIE, pp. 612–623, March 2005. View at Publisher · View at Google Scholar · View at Scopus
  24. M. A. Davis and A. D. Kersey, “Separating the temperature and strain effects on fiber Bragg grating sensors using stimulated Brillouin scattering,” IEE Proceedings. Optoelectronics, vol. 144, no. 3, pp. 151–155, 1997. View at Publisher · View at Google Scholar
  25. D. H. Liu, Study on super-long distance distributed optical fiber sensing technology and its application [Doctoral dissertation], College of Civil Engineering and Architecture, Zhejiang University, 2005.
  26. M. G. Xu, L. Dong, and L. Reekie, “Temperature-independent strain sensor using a chirped Bragg grating in a tapered optical fiber,” Electronics Letters, vol. 31, no. 10, pp. 823–825, 1995. View at Publisher · View at Google Scholar
  27. B. O. Guan, H. Y. Tam, S. L. Ho, W. H. Chung, and X. Y. Dong, “Simultaneous strain and temperature measurement using a single fibre Bragg grating,” Electronics Letters, vol. 36, no. 12, pp. 1018–1019, 2000. View at Publisher · View at Google Scholar · View at Scopus
  28. J. Gao, B. Shi, W. Zhang, and H. Zhu, “Monitoring the stress of the post-tensioning cable using fiber optic distributed strain sensor,” Measurement, vol. 39, no. 5, pp. 420–428, 2006. View at Publisher · View at Google Scholar · View at Scopus
  29. H. Jianping, Z. Zhi, C. Genda, and O. Jinping, “Measurement accuracy improvement of brillouin signal using wavelet denoising method,” in Smart Sensor Phenomena, Technology, Networks, and Systems, vol. 7293 of Proceedings of SPIE, pp. 1–7, March 2009. View at Publisher · View at Google Scholar · View at Scopus
  30. Y. Y. Wang, Y. H. Yang, M. W. Yang, et al., “Wavelet transform de-noising technology for distributed optical fiber sensor,” in Proceedings of the, vol. 7853 of Proceedings of SPIE, pp. 1–6, 2010. View at Publisher · View at Google Scholar