Gangopadhyay2009.pdf

Sensors and Actuators A 150 (2009) 78–86

Contents lists available at ScienceDirect

Sensors and Actuators A: Physical journal homepage: www.elsevier.com/locate/sna

Fibre Bragg grating strain sensor and study of its packaging material for use in critical analysis on steel structure Tarun Kumar Gangopadhyay ∗ , Mousumi Majumder, Ashim Kumar Chakraborty, Asok Kumar Dikshit, Dipak Kumar Bhattacharya Central Glass & Ceramic Research Institute (CGCRI), Council of Scientific and Industrial Research (CSIR), 196 Raja S.C. Mullick Road, Kolkata 700032, India

a r t i c l e

i n f o

Article history: Received 21 August 2008 Received in revised form 20 November 2008 Accepted 21 December 2008 Available online 31 December 2008 Keywords: Strain measurement Fibre-optic sensor Packaging of FBG sensor Composite materials Polymer materials

a b s t r a c t Strain studies in civil structures, aircrafts, oil pipelines, etc. are pivotal in avoiding unexpected failures. Long-term strain study of structures also helps in setting the design limits of similar structures. Conventionally, most structures rely on maintenance schedules, visual inspection and a few conventional sensors. But the high cost of maintenance, lack of precision in visual inspection and susceptibility of sensors to harsh environmental conditions have made structural health monitoring (SHM) a necessity. Over the past few decades, fibre Bragg grating (FBG) sensors have emerged as a suitable, accurate and cost-effective tool in SHM. Fibre Bragg gratings are obtained by creating periodic variations in the refractive index of the core of an optical fibre. These periodic variations are created by using powerful ultraviolet radiation from a laser source. Periodic structure acts as a Bragg reflector of particular wavelength. Minute change in the periodic structure due to external perturbation will cause appreciable wavelength shift. This shift in turn can be translated to measurand related to perturbation. The main advantages of FBGs over other optical sensor schemes are its low cost, good linearity, wavelength multiplexing capacity, resistance in harsh environments and absolute measurement. FBG sensor technology is now on the verge of maturity after almost two decades of active research and development in this field. Efforts are now concentrating on delivering complete FBG sensor systems including front-end electronics. This paper demonstrates with the aim to provide different design and experimental packaging procedures of indigenously developed FBG sensors for strain measurement. Various model of loading on FBG have been tried to explore with particular attention on the primary packaging of the sensor for application on steel cantilever structure and cement concrete. Preliminary packaging has been done with composite materials such as epoxy resin casting and fibre reinforced plastic (FRP) composites. Encouraging results are obtained and presented in this paper. The results are compared with the standard FBG sensors and with mechanical strain gauge. © 2009 Elsevier B.V. All rights reserved.

1. Introduction Strain studies in civil structures, aircrafts, oil pipelines, etc. are pivotal in avoiding unexpected failures. Long-term strain study of structures also helps in setting the design limits of similar structures. Conventionally, most structures rely on maintenance schedules, visual inspection and a few conventional sensors. But the high cost of maintenance, lack of precision in visual inspection and susceptibility of sensors to harsh environmental conditions has made structural health monitoring (SHM) a necessity. Over the past few decades fibre Bragg grating (FBG) sensors have emerged as a suitable, accurate and cost-effective tool in SHM.

∗ Corresponding author. Tel.: +91 33 24649329; fax: +91 33 24730957. E-mail addresses: [email protected], [email protected] (T.K. Gangopadhyay). 0924-4247/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.sna.2008.12.017

The main advantages of FBGs over other optic sensor schemes are its low cost, good linearity, wavelength multiplexing capacity, resistance in harsh environments and transduction mechanism which eliminates the need for referencing as in interferometric sensors. FBG sensor technology is now on the verge of maturity after almost two decades of active research and development in this field. Efforts are now concentrating on delivering complete FBG sensor systems including front-end electronics. Several review papers on fibre Bragg grating applications have been published [1–4]. Strain and temperature have so far been the dominating measurand of interest [5–7]. Author is also made a critical review on structural vibration using FBG and Fabry–Perot sensors [8]. Recently the authors’ reviewed [9] the previous work of FBG as strain sensors in structural health monitoring including the present status and applications along with various encapsulation techniques. FBG’s are basically strain and temperature sensitive devices. Bragg gratings can be inscribed directly in a standard optical fiber at

T.K. Gangopadhyay et al. / Sensors and Actuators A 150 (2009) 78–86

79

Fig. 1. Transmission and reflection spectra from an FBG [14].

any position, and several of them can be configured in series or parallel on different fibers and interrogated from the same light source enabling flexible sensor configurations. This paper demonstrated with the aims to provide different design and experimental packaging procedures of FBG sensors for strain measurement. Various model of loading on FBG have been tried to explore with particular attention on the primary packaging of the sensor for application on steel cantilever structure and cement concrete.

Fibre Bragg gratings are obtained by creating periodic variations in the refractive index of the core of an optical fibre. These periodic variations are created by using powerful ultraviolet radiation (holographic method). Fig. 1 shows the internal structure of an optical fibre with an FBG written in it. In a single mode optical fibre, light travels in the fundamental mode along the axis of the core of the fibre. When light passes through an FBG, Fresnel reflections take place due to the variations in refractive index of the fibre. This is called coherent reflection. If the criterion for constructive interference is met, then the incident light satisfies the Bragg condition is given by [1]. (1)

where B is the Bragg wavelength, n is the effective refractive index of the FBG and  is the grating period. When the Bragg condition is satisfied, reflections from each successive period will be in phase. Light that does not satisfy the Bragg condition passes through the FBG as if it were of uniform refractive index n (Fig. 1). 3. Strain measurement using FBG sensors When an FBG is strained, the Bragg wavelength, B changes due to both the change in grating pitch,  (due to the simple elastic elongation) and due to the photoelasticity-induced change of the refractive index. The relative change in Bragg wavelength is given by [2] B = (1 − e )ε B

The Bragg wavelength B is also susceptible to temperature changes. The change in wavelength is due to the combined effect of the thermal expansion of the core material and the thermo-optic behaviour that induces a change in the refractive index of the fibre. The relative change in the Bragg wavelength due to temperature change is given by B = (˛ + )T B

2. Principle of operation of FBG sensors

B = 2n

Fig. 2. A stainless steel cantilever structure has been fabricated in-house and FBGs are placed on it for strain calibration.

where T is the change in temperature experienced at the FBG location, ˛ is the thermal expansion and  is the thermo-optical coefficient. For an FBG of central wavelength of 1550 nm, typical temperature sensitivity B /T = 13 pm/◦ C [5]. However, the strain and temperature sensitivities of FBG sensors depend on the type of fibres as well [6]. 3.1. Experimental setup for FBG based strain measurement Although the application of bare FBG on real structural application is not advisable due to fragility of the silica fibre, a few tests have been performed in the laboratory with bare FBG on a stainless steel (Fig. 2) structure fabricated in-house. However these tests are essential to ascertain the strain-opto coefficient of the bare fibre. FBGs are placed on the axis of the cantilever for strain calibration as shown in Fig. 2. The optical-setup is presented in Fig. 3. A tuneable fibre laser source emitting (1520–1570 nm) with a maximum power of 2 mW is used for interrogation of 80–90% reflectivity FBG elements centered at 1551 nm. The laser, illuminated with a single-mode fibre pig-tail, is directly connected to the FBG via a 50:50 (3 dB) fused coupler (developed at CGCRI) which is also used to collect Bragg wavelength reflected by the grating. The reflection coming from

(2)

where ε is the longitudinal strain experienced by the optical fibre at the FBG location and e is the effective photo-elastic constant of the fibre core material e =

n2 [p12 − v(p11 + p12 )] 2

(3)

where pij are the silica photo-elastic tensor components and v is the Poisson’s ratio. For an FBG of central wavelength of 1550 nm, typical strain sensitivity B /ε = 1.2 pm/microstrain [5].

(4)

Fig. 3. Schematic diagram of experimental set-up and steel cantilever.

80

T.K. Gangopadhyay et al. / Sensors and Actuators A 150 (2009) 78–86

Fig. 4. Fibre Bragg grating interrogation instrument has shown the wavelength peak of a FBG made in-house by CGCRI (micron optics, wavelength range 1520–1570 nm, resolution 0.25 pm).

FBG is detected by the PIN photodiode with wavelength accuracy 1 pm and resolution 0.25 pm (si720, Micron Optics). Bare FBG is placed on the surface of the steel cantilever; hence, static and dynamic strain can be applied to the FBG via cantilever. It can be placed for the measurement either on upper surface for strains due to elongation or on the bottom surface for strain due to contraction. Theoretical strain has been calculated at the point of location on which the FBG is placed. Before applying load on structure the wavelength reflection is recorded as shown in FBG interrogation system (Fig. 4). Two mechanical strain gauges are also placed side by side on the cantilever to compare the data with FBG sensor at the same location and to get direct strain through data logger (Fig. 5). While applying load on the cantilever structure it starts to bend and strain is translated to FBG. The complete set-up is shown in Fig. 6. 4. Tensile tests and results with bare FBG Theoretical values of strain on the axis of the cantilever (Fig. 2) has been calculated using conventional strain equation [10]. ı=

 M = Y Yz

(5)

Fig. 5. Strain gauge data logger for strain measurement and associated display of strain on screen. (Make: TML, Japan, measurement range 1–20,000 microstrain, resolution 0.1 microstrain).

Fig. 6. FBG interrogation system for strain sensing and data logger for strain gauge measurement (inset: stainless steel cantilever structure).

Table 1 Comparison of strain responses of CGCRI-make FBG sensors and recorded value by conventional strain gauge with theoretical strain. Force (N)

Strain (calculated) ␮␧

Strain (measured) ␮␧

Wavelength shift of FBG in pm

4.9 9.8 14.7 19.6 24.5 29.4 34.3

61.04 122.09 183.13 244.18 305.22 366.27 427.31

61 117 184 244 315 363 421

80 140 220 300 380 440 500

where ı is the strain developed at the axis of the cantilever,  is stress due to load applied, Y is Young’s modulus of SS steel bar, M is the total moment and z is the section modulus. Strain due to different load has been calculated and plotted in Table 1. While starting measurement initial wavelength spectra of the bare FBG is recorded. A test for static strain measurements has been performed on FBG. Actually the response of the laser system is investigated for subsequent static loads on the FBG. A load is gradually increased on one side of the cantilever and corresponding wavelength shift is measured as shown in Fig. 7. Comparison of strain responses of CGCRI-make FBG sensors, recorded value by conventional strain gauge and theoretical strain is shown in Table 1. Wavelength shift versus microstrain is depicted in Fig. 8. From the data it is observed that the strain sensitivity of the developed sensor is 1.25 pm/␮␧ which is tallied with theoretical value. The system is tested to perform strain measurements up to about 450 ␮␧ with

Fig. 7. FBG responses of CGCRI-make sensors before strain and after strain (used bare FBG).

T.K. Gangopadhyay et al. / Sensors and Actuators A 150 (2009) 78–86

81

Fig. 8. Response of CGCRI-make FBG sensor force applied to steel structure (used bare FBG before packaging).

linear range and manual loading on cantilever. To get reproducible responses of CGCRI-make FBG sensors the strain has been measured on consecutive days (Fig. 9) using bare FBG before packaging. To avoid the error on damping of steel structure during manual loading, the linearity of the bare FBG has been tested on a calibration set-up using load cell. This FBG calibration set-up has been developed at CGCRI as shown in Fig. 10 in which load cells are controlled by personal computer and simultaneously wavelength shift is captured on specially designed software (Lab view). Response of CGCRI-make bare FBG against load applied through load cell apparatus is shown in Fig. 11. The system is tested to perform strain measurements on FBG with automatic loading up to about 200 g with linear wavelength shift and strain sensitivity was tailed 1.25 pm/␮␧.

Fig. 10. FBG-strain calibration setup developed at CGCRI.

alone. The final goal in manufacturing a composite is to combine similar or dissimilar materials in order to develop specific properties that are related to the desired characteristics. The concentration of components is generally considered as the single most important parameter influencing the composite properties [12]. Also, it is easily controllable all processable variables for altering the composite properties.

5. FBG strain sensor packaging with composite materials 5.1. Fibre reinforced plastic (FRP) composites By definition a composite is a composed material developed by the synthetic layer assembly of two or more constituents, selected filler or reinforcing agent and a compatible binder (i.e., resin) in order to obtain specific characteristics and properties [11]. Each of these constituent materials plays an important role in the processing and the final mechanical performance of the end product. The composite matrix improves the physical properties through specific molecular rearrangement in the end product. The reinforcement filler provides to improve the mechanical strength. The fillers and the additives are used as process or performance enhancement and to impart special properties to the final product. The properties of a composite cannot be achieved by any of these components acting

In recent years, many experiments and researches have been carried out on laminated composites to obtain optimal mechanical properties. Composite properties are highest in the orientation direction of fibers. In practical application, most of the structures are not loaded in single direction, it is necessary to orient fibers in multiple directions. This demands evaluation of mechanical properties for different fiber orientations. The fibre weight fraction is an important parameter influencing the mechanical properties of composites. Bar-Yoseph and Pian [7] developed a new method for calculating interlaminar stress (ILSS) concentration in angle-ply

Fig. 9. Reproducible responses of CGCRI-make FBG sensors to monitor strain on consecutive days (used bare FBG before packaging) [14].

Fig. 11. Response of CGCRI-make FBG sensor against load applied through load cell (used bare FBG on load cell apparatus).

82

T.K. Gangopadhyay et al. / Sensors and Actuators A 150 (2009) 78–86

laminates. The aim of the work was to study the effects of fibre orientations and fibre weight fraction on tensile strength, heat deflection temperature (HDT), impact strength and hardness of glass/epoxy laminated composites. 5.2. FRP composite for structure To develop a composite structure the following points are very important. (i) In order to understand how composite materials perform in structures, it is necessary to understand some basics about their nature. Composite materials contain a mixture of two or more types of fundamentally different components. They have properties that are some combination of the properties of their components. (ii) All materials that contain more than one component are not necessarily composite materials. For example, pearlitic steel is not considered a composite, although it contains more than one component, since its various parts are of the same nature. (iii) Some materials that are considered composites are concrete, steel-reinforced concrete, fiber-reinforced polymers (like graphite/epoxy or glass/epoxy), laminated wood, and rubber tires. Concrete- and steel-reinforced concrete contain more than one type of component (aggregate, cement paste, and steel). Graphite- and glass-based FRPs contain high-strength fibers surrounded by a more ductile resin. Rubber tires contain the polymeric rubber-type material, carbon particles, and, frequently, steel or other types of reinforcement. (iv) The driving force behind the development of modem composite materials has been their high strength and stiffness when determined on a weight basis. Most of the original work on modem composites was in the aerospace industry. These industries are very weight sensitive, and a decrease in weight is a very important issue. This is the case even if the FRP parts are more expensive than the parts they replaced. Composites are now being used in the surface transportation industry. They are frequently used on automobiles and lightweight boats. Composites have also penetrated a number of consumer sports areas, such as graphite/epoxy golf clubs and skis. (v) One way to better understand composite materials is to examine some current applications of composites. There are a number of aerospace applications. They are used as structural parts on many modem jet airplanes, such as the Boeing 767. The Voyager was the first airplane to fly around the world nonstop without refueling. Its superstructure was mostly made of composite materials. Similarly, the Gossamer Albatross became the first human-powered vehicle to fly across the English Channel. Such a vehicle could not be built with traditional metallic materials, for it would have been too heavy. (vi) Composites have also been used on land-based vehicles. For example, the auto industry has formed a consortium to do research on composite materials. They have successfully built a Taurus whose superstructure is composed of five composite panels that have been glued together. Glass-fiber-based composites have been used to form the hopper in railroad cars. The U.S. Army has recently designed and built an armoured vehicle that has a composite material hull. The U.S. Navy has used composite materials to make mine sweeper ships. (vii) Composite materials are being used extensively in the sporting world. Glass-based FRP poles are commonly used in pole vaulting. Graphite/epoxy golf club shafts are highly desirable because of weight. Graphite/epoxy skis were also popular as they are light and because of their light-weight glass-based

Fig. 12. Comparison of properties of fibre, resin and composite.

FRPs have been used in small consumer-oriented sporting boats for many years. (viii) Composite materials are being used in civil engineering structures. The tent-like roof on the new Denver International Airport is made from a glass-based FRP that has been coated with Teflon. The same basic material has been used as the roof for a number of sports stadiums, such as the Metrodome in Minneapolis. Glass-based composites have been used in nearly 100,000 underground fuel storage tanks; this use is growing rapidly. Uses also include sandwich shell roofs for exhibition structures, large-diameter pipe, and numerous gratings and structural shapes. 5.3. Properties of FRP composites Properties of composites are strongly influenced by the properties of their constituent materials, their distribution and the interaction between them. The composite properties may be the volume fraction, sum of the properties of the constituents, or the constituents may interact in a synergetic way as to provide properties in the composites that are not accounted for by simple volume fraction of the constituents. Concentration is usually measured in terms of volume or weight fractions. The contribution of a single constituent to the overall properties of the composite is determined by this parameter. The concentration is generally considered as the single most important parameter influencing the composite properties. Also, it is easily controllable all manufacturing variables for altering the composite properties. The orientation of the reinforcement affects the isotropic property of the system. In continuous fiber reinforced composites, such as unidirectional or cross-ply composites anisotropy is desirable. The primary advantage of these composites is the ability to control the anisotropy by design and fabrication. The potential advantage for considering composites for any application is manifold, but for majority of cases the interest will be on: (i) Light weight manifested in the form of high specific properties. (ii) Corrosive resistance. Comparison of properties of fibre, resin and composite is shown in Fig. 12. 5.4. Primary packaging of FBG with polymer materials In the present experiment FBG has been packaged with Araldite LY556, epoxy-A and Araldite HY951, epoxy B as hardener which is obtained from Huntsman & Co. Ltd., Basel 4057, Swizerland. Araldite LY556 is the epoxy resin and used with a liquid diglycidyl ether of bisphenol A and N,N -bis (2-aminoethyl) ethane-1,2-diamine. Both resins were used as received with no additional purifica-

T.K. Gangopadhyay et al. / Sensors and Actuators A 150 (2009) 78–86

83

Fig. 13. FBG casting in epoxy-resin (78–22%) as primary packaging for strain measurement.

tion. The chemical structure of the resin components are as follows [11–13]:

Fig. 15. Response of CGCRI-make FBG and polymer packaged sensor to monitor strain (when FBG is placed on the top surface of the cantilever) [14].

(a) DGEBA-diglycidyl ether of bisphenol A. (b) AEEDA-N,N -bis (2-aminoethyl) ethane-1,2-diamine.

5.4.1. Tensile tests and results with FBG after packaging This packaged FBG (sensor pellet) has been used on the steel cantilever structure for testing with load in two manners. Firstly, it is rigidly fixed on the top of the cantilever. Wavelength shift due to loading on cantilever has been recorded in the similar way as it is done in the case of bare fibre (discussed in Section 4). Wavelength shift vs. load has been plotted in Fig. 15. As observed in the result 420 pm of wavelength shift (strain ≡ 336 ␮␧) has been observed for 29N of applied load and strain sensitivity was tailed 1.25 pm/␮␧. Secondly, the sensor has been fixed rigidly at the bottom of the cantilever and wavelength shift due to loading on cantilever has been recorded (Fig. 16). Due to downward loading, the bending of the cantilever has been turned downward direction and the FBG sensor exerted a compressive load. As observed in the result 0.22 nm (220 pm) of wavelength shift (strain ≡ 176 ␮␧) has been observed for 1.4 kg of gradually applied load and strain sensitivity was confirmed 1.25 pm/␮␧.

The two resins are mixed in the molar ratio of 4:1 at room temperature. The curing mechanism is exothermic due to chemical reaction between hydroxyl groups of bisphenol epoxy and diamine of AEEDA. The molecular structure of epoxy network is due to cross-linking density, enhances mechanical property at room temperature. The extension of network formation in the chemical reaction depends on how much diamine is involved with hydroxyl group of epoxy. The details curing kinetics has been reported by Musto et al. [15]. Primary packaging of FBG with epoxy-resin (78–22%) has been done as shown in Fig. 13. Here FBG is embedded inside the casting. The size of the cured sheet is 50 mm × 15 mm having 3 mm thickness. PVC sleeves of 250 ␮m diameter have been used to protect the fibre at both edges of the sheet. The chemical bonding in between epoxy-diamine and epoxy-dianhydride is an exothermic reaction. The exothermic heat from chemical bonding in between epoxy A and epoxy B at initial curing period affect on wavelength curve. This phenomena occurs during curing through chemical bonding due to the cross linking of polymer chains, as a result a network structure is found. The process has been made on-line i.e., FBG was connected with interrogation system so as to observe the wavelength shift from the grating while curing of the polymer packaging materials. The on-line data shows in Fig. 14 in which a few airy peaks are significant in the starting region (around 1-min duration) of the reaction due to thermal stress on the grating. The reflected wavelength is gradually shifted nearly 0.14 nm which is equivalent to 112 ␮␧ to settle down the packaging material. After the curing process this small pellet can be used as primary packaging.

Fig. 14. Wavelength shifting during polymer packaging on bare FBG.

5.5. Packaging material testing Properties of composites and packaging materials on the sensor are very much important as applied load has been exerted on FBG in reduced manner. Load distribution will be uniform on FBG if the composites materials have high storage modulus and sheer strength and homogeneously network structure formation occurs. Different tests are performed to evaluate the characteristics of the sensor pellet. 5.5.1. X-ray diffraction The amorphous phases of epoxy resin coated silica fibre was recorded with an X’ Pert Philips MPD system (PW1710) Cu LFF X-ray

Fig. 16. Wavelength response of compressive load on the FBG inside polymer packaging using steel structure (PS294-1, dated 13-04-06).

84

T.K. Gangopadhyay et al. / Sensors and Actuators A 150 (2009) 78–86

Fig. 17. XRD profile of epoxy-resin packaging materials.

tube operating at 40 kV and 20 mA using Cu K␣ ( = 0.154056 nm) radiation. A scanning step size of 0.020 and scanning speed of 0.5 s per step were adopted for this measurement. Fig. 17 represents wide angle X-ray diffraction (XRD) profile for epoxy polymer resin packaged silica fibre. There are two amorphous halos observed at 18.8◦ and 42.6◦ which corresponds to amorphous phase of epoxy resin. At 2 of 18.8◦ corresponds 1 1 0 and 0 2 0 lattice plane [11]. The peak halo intensity is too much reflects packing of molecular chains with high density. Amorphous silica peak at 25.7◦ was also observed when scale resolute in between 24◦ and 30◦ . 5.5.2. Thermo gravimetric analysis Schimadzu model TGA-50 instrument was employed to measure thermal history from room temperature 30 ◦ C to 600 ◦ C using few milligrams of samples in the pan at 10◦ /min heating rate. Fig. 18 shows thermal history of resin up to 600 ◦ C, which represents that there was no significant weight loss up to 300 ◦ C. The loss of weight is 4.71%, between 300 ◦ C and 340 ◦ C, further weight loss of 48% between 340 ◦ C and 425 ◦ C and finally total loss of 85% at 600 ◦ C. The weight loss taken place at the different stages of thermal treatment due to either structural change of the materials or thermal decomposition reactions of the components in presence of

Fig. 18. TGA figure of epoxy-resin packaging materials.

Fig. 19. DSC profile of epoxy-resin packaging materials (a. endotherm, and b. exotherm).

oxygen with evaporation of gas. From the graph it has been observed that the sample is thermally stable up to 300 ◦ C temperature. The molecular epoxy interpenetrating network structure of the cured samples are unaffected at low temperature, the crosslink density at room temperature not bring significant change of the mechanical property [11]. 5.5.3. Differential scanning calorimetry The thermal study of the epoxy resin was done in a PerkinElmer DSC-diamond instrument using pyris software under dry nitrogen atmosphere. The samples of (2–8)mg were made in aluminium pan. They were then heated at the scan rate of 10◦ /min. The peak temperature and the enthalpy of epoxy resin melting were measured from the endotherm using a computer attached to the instrument. The enthalpy of melting and the crystallization temperature were measured using a dynamic heating and cooling method. The DSC was calibrated with indium before the experiment. The crystallization temperature (Tc ) and the melting temperatures (Tm ) were taken as the temperatures of the minimum and the maximum of both endothermic and exothermic peaks, respectively. Fig. 19a and b represents DSC thermogram of chain packing of epoxy crystal melting (Tm ) and crystallization (Tc ) endotherm and exotherm, respectively. The crystals are melted at 76.86 ◦ C and crystallize at 66.32 ◦ C, respectively. 5.5.4. Scanning electron microscope In order to observe the phase structure of epoxy blends, the samples were fractured under cryogenic condition using liquid nitrogen. The etched specimens were dried to remove the solvents. The epoxy resin was carbon coated under vacuum sputtering, and SEM micrographs were taken through a SEM (LEO S-430 i), UK instrument under active voltage of 15 kV at different magnifications.

T.K. Gangopadhyay et al. / Sensors and Actuators A 150 (2009) 78–86

85

Fig. 21. FBG casting in FRP composites as primary packaging for strain measurement.

Fig. 22. Wavelength response of compressive load on the FBG inside FRP packaging using steel structure (PS294-1, dated 12-04-06).

Fig. 20. (a and b) SEM images of epoxy-resin packaging materials.

The following two figures (Fig. 20a and b) are represented scanning electron micrographs (SEM) of epoxy resin at different magnification which was derived from polymeric resin A (diamine) and resin B (dianhydride). The fractured surface of epoxy resin shows that fibril structure network structure. The fibrils are thin and widely dispersed in the whole structure of the sample. There was no porosity observed, which reveals that it was continuous homogeneous interpenetrating network (IPN) molecular polymeric domain structural matrix. The degree of cross linking that occurs during curing inside the molecular chain networks, cross linking of polymer chains surrounded by uncured materials as a result of flexibility increased [12]. The epoxy networks structure show a single-phase morphology. It can be explained that the effects of solubility parameter and chemical interaction in-between epoxy and hardener. The combine equation of the Flory–Huggins equation and the Hildebrand equation [16], the free energy of mixing can be expressed as Gm 2 = ϕe ϕr (ıe − ır ) + RT V

ϕ

e

Ve

ln ϕe +

ϕr ln ϕr Vr

 (6)

where ϕe , ϕr are the volume fractions, and ıe , ır are the solubility parameters and Ve , Vr are the molar volumes of epoxy and hardener, respectively. Since ϕe , ϕr are fractions (<1), the second term (change in entropy) is always negative. During curing of epoxy resin, the value of Ve and Vr increases which results in a decrease in second term in the above equation. At a critical point of overlapping concentration during chemical reaction, the change of molar free energy (Gm ) becomes positive and as a result phase separation occurs. However, if (ıe − ır ), i.e., the difference between the solubility parameters of the epoxy and the hardener is very low, then the change in entropy due to the curing reaction, cannot make

Fig. 23. Schematic model shown glass fiber and epoxy composite.

free energy change of mixing (Gm ) positive prior to overlapping concentration [17]. 5.6. Primary packaging of FBG with fibre reinforced plastic (FRP) composites Primary packaging of FBG with FRP composite has been successfully done (Fig. 21). Glass fiber and epoxy resins are composed with one by one layer with high impact hot press. Here FBG is embedded inside the FRP casting. This process is also continued on-line connected with interrogation unit. The dimension of the sheet casting is (47 × 12) mm having 2 mm thickness. The details studies have been done and will be published in next publication. In similar test (as Section 5.4.1) has been performed using FRP composite packaging material and FBG is bonded in side the sheet. Wavelength vs. load is depicted in Fig. 22. It has been observed 0.35 nm (350 pm) of wavelength shift (strain ≡280 ␮␧) for 1.6 kg of gradually applied load and strain sensitivity was tailed 1.25 pm/␮␧. Fig. 23 has shown schematic model of glass fiber and epoxy composite where glass fiber are aligned directionally and epoxy network structure bonded glass fibers through layer by layer structure gives a high mechanical impact strength. 6. Conclusion This paper presented design and experimental packaging of indigenously developed FBG sensors for the strain measurement. A few models of loading on FBG have been tried to explore on bare FBG and placing the FBG on steel cantilever structure. Preliminary

86

T.K. Gangopadhyay et al. / Sensors and Actuators A 150 (2009) 78–86

packaging with epoxy resin casting has been used on steel structure. Encouraging results are demonstrated. The results are compared with the standard FBG sensors and with mechanical strain gauge. Practical modelling has been tried for the bonding composite at the interface of bare fibre and coating, coating and adhesive layer, adhesive layer and host material and validated the results using load application. Finally the study of packaging material has been performed to evaluate the characteristics of the sensor pellet. X-ray diffraction profile, thermo gravimetric analysis (TGA), differential scanning calorimetry (DSC) and scanning electron microscope (SEM) for epoxy polymer resin casting with FBG have been performed to confirm the packaging performance. These are the new tests to observe the packaging characteristics of fibre-optic sensors. Encouraging observations obtained and presented in this paper. The study conclude that for better strain transfer from the host material to the FBG sensors, a thin layer of adhesive, a high modulus coating material and a sufficient embedding length of the sensor is appreciable. Further scope of study is to optimise the thickness of the epoxy-layer on FBG surface considering strength of the packaging embodiment as well as load transfer to sensor. Acknowledgement The authors would like to acknowledge the support, guidance and encouragement provided by the Director, Central Glass & Ceramic Research Institute (CGCRI), Kolkata, India. The work has been carried out under the project Technology for Engineering Critical Assessment (TECA), CORR 0022. The authors would also like to thank Dr. Dipten Bhattacharya, Scientist and Dr. N.R. Bose, Ex. Scientist of CGCRI for their valuable suggestions on characterization of polymer materials. They thank the staff members of Fibre Optics Laboratory, CGCRI for their help and co-operation. References [1] W.W. Moorey, G.A. Ball, H. Singh, Applications of fibre grating sensors, Proc. SPIE 2839 (1996) 2–7. [2] P. Ferdinand, S. Magne, V. Dewynter-Marty, C. Martinez, S. Rougeault, M. Bugaud, Applications of Bragg grating sensors in Europe, in: Proceedings of the 12th International Conference on Optical Fibre Sensors, Williamsburg, USA, 1997, pp. 9–14. [3] Y.J. Rao, Recent progress in applications of in-fibre Bragg grating sensors, Optics Lasers Eng. 31 (1999) 297–324. [4] B. Lee, Review of the present status of optical fibre sensors, Optical Fibre Technol. 9 (2003) 57–79. [5] Y.J. Rao, Fiber Bragg grating sensors: principles and applications, in: K.T.V. Grattan, B.T. Meggitt (Eds.), Optical Fiber Sensor Technology, vol. 2, Chapman & Hall, London, 1998, pp. 355–389. [6] X. Shu, Y. Liu, D. Zhao, B. Gwandu, F. Floreani, L. Zhang, I. Bennion, Dependence of temperature and strain coefficients on fiber grating type and its application to simultaneous temperature and strain measurement, Opt. Lett. 27 (9) (2002) 701–703. [7] P. Bar-Yoseph, T.H.H. Pian, Calculation of interlaminar stress concentration in composite laminates, J. Composite Mater. 15 (1981) 225–239. [8] T.K. Gangopadhyay, Prospects for fibre Bragg gratings and Fabry–Perot interferometers in fibre-optic vibration sensing, Sensors Actuat. A: Phys. 113 (2004) 20–38. [9] M. Majumder, T. Kumar Gangopadhyay, A.K. Chakraborty, K. Dasgupta, D.K. Bhattacharya, Fibre Bragg gratings as strain sensors—present status and applications, Sensors Actuat. A: Phys. 147 (2008) 150–164. [10] S. Timoshenko, D.H. Young, Engineering Mechanics, 4th ed., vol. 1, McGraw-Hill, New York, 1956. [11] X. Kornmann, M. Rees, Y. Thomann, A. Necola, M. Barbezat, R. Thomann, Epoxylayered silicate nanocomposites as matrix in glass fibre-reinforced composites, Compos. Sci. Technol. 65 (2005) 2259–2268. [12] Neeraj Gupta, I.K. Varma, Effect of structure of aromatic diamines on curing characteristics, thermal stability, and mechanical properties of epoxy resins, J. Appl. Polym. Sci. 68 (11) (1998) 1759–1766.

[13] I. Stewart, A. Chambers, T. Gordon, The cohesive mechanical properties of a toughened epoxy adhesive as a function of cure level, Int. J. Adhesion Adhesives 27 (2007) 277–287. [14] T.K. Gangopadhyay, M. Majumder, A.K. Chakraborty, N.R. Bose, A.K. Dikshit, K. Dasgupta, S.K. Bhadra, D.K. Bhattacharya, Development of on-line strain sensor using FBG for engineering critical analysis on structure, in: International Conference on Sensors and Related Networks (SENNET’07), VIT-Vellore, India, 12–14 December 2007, pp 42–46. [15] P. Musto, M. Abbate, G. Ragosta, G. Scarinzi, A study by Raman, near-infrared and dynamic-mechanical spectroscopies on the curing behaviour, molecular structure and viscoelastic properties of epoxy/anhydride networks, Polymer 48 (2007) 3703–3716. [16] Y.-D. Lee, S.K. Wang, W.K. Chin, Liquid-rubber-modified epoxy adhesives cured with dicyandiamide. I. Preparation and characterization, J. Appl. Polym. Sci. 32 (8) (1986) 6317–6327. [17] M. Ochi, J.P. Bell, Rubber-modified epoxy resins containing high functionality acrylic elastomers, J. Appl. Polym. Sci. 29 (1984) 1381–1391.

Biographies Tarun Kumar Gangopadhyay graduated bachelor of Electrical Engineering in 1989 and Master of Electrical Engineering in 1991, both from the Jadavpur University, Calcutta, India. He graduated Ph.D. in December 2005 in the field of fiber-optic sensor from the University of Sydney, Australia. His Ph.D. thesis title is: ‘Measurement of vibration using optical fibre sensors’. During 1995–1999, he went to Australia in AusAID scholarship for commonwealth countries and there he was involved in the research and development of optical fibre vibration sensors for electrical power industry with High Power Testing and Optical Fibre Sensors Group, School of Electrical and Information Engineering, The University of Sydney, Australia. He worked there in the field of intrinsic and extrinsic single-mode fibre-based sensors and fibreoptic interferometry. He has got many awards and prizes for his research work from The University of Sydney and Australian Photonics. Presently he is employed as a Senior Scientist in the Optical communication Fibre Division at Central Glass and Ceramic Research Institute (CSIR), Kolkata, India. Recently he worked to develop Fibre Bragg Grating (FBG) temperature sensor for high voltage (400 kV) power transmission line application. He is currently involved in R&D work of optical fibre sensors, FBG sensors for smart structures, FBG sensors for power line application, chemical sensor using fibre ring resonator and development of fibre-optic components such as bi-directional coupler and WDM couplers. His current research interests are development of FBG sensors, chemical and gas sensors, bio-medical sensors and PM fibre coupler for Gyro application. He has authored several journal papers and international conference papers. He is a member of Optical Society of America (OSA), USA. Mousumi Majumder obtained her B.Tech in Electronics and Telecommunication Engineering from the North-Eastern Hill University, India in 1997. She had joined the Council for Scientific and Industrial Research, India in 1998. Presently she is employed as a Scientist in the Instrumentation Division at Central Glass & Ceramic Research Institute, Kolkata, India. Her professional interests lie in Material characterization, and usage of various strain sensing platforms in Structural Health Monitoring and Refurbishment. Ashim Kumar Chakraborty has obtained his M.E. (Hons.) degrees in Instrumentation and Electronics from Jadavpur University, Kolkata, India in 1990. He served Simon Carves Limited, India before joining as a scientist in the Instrumentation Section of Central Glass and Ceramic Research Institute (CSIR), Kolkata, India in 1987 where he is presently a senior scientist and the head of the section. He has active interest in technology development and has participated in a number of sponsored research projects. He has served as the research project leader for development of FBG strain sensors under TECA (2004–2007), a network project among CSIR laboratories, India. His current research interests are in the area of technology development for specialty glasses and optic fiber sensors. Dipak Kumar Bhattacharya, a graduate in Metallurgy obtained his doctoral degree from the Indian Institute of Science, Bangalore, India in 1995 working on the subject of correlation of Barkhausen Signal and magnetic hysteresis loop parameters with microstructures in steels. His professional interests lie in materials processing & characterization, and R&D project management. He has worked in various NDT techniques for the evaluation of remaining life assessment of engineering materials and structures in the power and petrochemical plants. He is presently heading the Analytical Facility Division and Programme Management Division in Central Glass & Ceramic Research Institute, Kolkata, India. He has more than 70 publications in International Journals. He is the Chief Editor of the Journal of Nondestructive Evaluation published by the Indian Society for NDT.