Sensing tape for easy integration of optical fiber sensors in composite structures

Size: px
Start display at page:

Download "Sensing tape for easy integration of optical fiber sensors in composite structures"

Transcription

1 Sensing tape for easy integration of optical fiber sensors in composite structures Branko Glisic, Daniele Inaudi SMARTEC SA, Via Pobiette 11, 6928 Manno, Switzerland Phone: , fax: , ABSTRACT The integration of optical fiber sensors in composite structures is often limited by the difficult handling of the bare fibers. In this contribution we present a novel packaging consisting in a composite tape that can be easily manipulated, embedded and surface mounted. INTRODUCTION New application domains for thermoplastic and thermoset composite materials, notably in gas and oil industry 1 and civil engineering 2 created a need for appropriate monitoring systems, able to record important parameters relating to design and service performance. The collected data supports cost effective management 3 and increase structural overall safety 4. While standard measurement systems and software can be directly used for composite materials monitoring, sensors must be mo dified and adapted to each host material. Fiber optic sensors are proven to be ideal transducers for composite material monitoring. Their material compatibility with composites, and small cross-section make them less intrusive compared with traditional, electrical sensors. Composite materials are often manufactured in form of tapes or sheets, while sensors are to be embedded within the structure, depending on the structural layers that have to be monitored. Improper embedding of the sensors may be a source of delaminating that causes a significant decrease of mechanical properties. Sensor can also be installed on the surface of the structure, and in this case the optical fiber has to be protected against environmental influences. On the other hand, if the sensor is designed to monitor strain or deformation, it is necessary to guarantee a good bonding between the optical fiber and the composite. Finally, for an industrial deployment of fiber optic sensors in this domain, it is necessary to package the sensors in a way that makes them as easy to handle as other components used for composite production. In this contribution, we propose to pre-package the sensing optical fiber in a thin composite tape, that can than be embedded or surface mounted on the composite structure. The tape gives to the optical fiber necessary protection against an accidental damaging during handling and installation. The fiber-reinforced composite tape with integrated optical fiber is called a sensing tape. This paper presents the design, production, testing, sample applications and performance of such a sensing tape. The selected example are based on the use of SOFO 5, 6 deformation monitoring systems, however, this design has also been used in conjunction with distributed Brillouin scattering systems and can be extended to fiber Bragg grating sensors. SENSING TAPE DESIGN CRITERIA A strain measuring fiber is to be in permanent mechanical coupling with the host structure. The transfer of strain is to be complete, with no losses due to sliding. Therefore an excellent bonding between the strain optical fiber and the host structure is to be guaranteed. To allow such a good bonding it has been recommended to integrate the optical fiber within the tape in the same way as the reinforcing glass or carbon fibers. In the presented case, a glass fiber thermoplastic composite was selected. Since its production involves heating all components at high temperature (in order to melt the matrix) it is necessary for the sensing fiber to withstand such temperatures temperature without damage. Furthermore, the bonding between the optical fiber coating and the matrix has to be excellent. For these reasons, polyimide coated optical fiber where selected for this application. The typical cross-section of thermoplastic composite tapes used for tape winding manufacturing of composite structures presents a width of about 13mm and a thickness as low as 0.2 mm. This later dimension is critical being close to the external diameter of polyimide-coated optical fiber: approximately 0.145mm. A tape cross-section, with typical dimensions, is presented in Figure 1. Thermoplastic tape NOT TO SCALE, Rh=10:1, R v=100: Polyimide fibre, φ0.145 Figure 1: Typical tape cross-section

2 TAPE MANUFACTURING A 100m batch of the sensing tape prototype was produced in order to demonstrate feasibility, test the quality of bonding between optical fiber and tape, and to determine mechanical and optical performance. The tape was produced by Gurit Suprem using a sandwich technique: two half-thickness (~0.10 mm) glass fibers reinforced thermoplastic tapes were first produced, then the optical fiber was placed between them, and finally tapes and optical fiber were assembled under high pressure and temperature. A unidirectional glass fiber (type S2) reinforced thermoplastic tape with PPS matrix was used. The production setup is presented in Figure Microbending due to crossing with reinforcing glass fibers; 2. Microbending due to shrinkage of matrix. 3. Birefringence introduced by lateral pressure Optical fiber Black box Figure 3: Sensing tape with visible light (encircled) Sensing tape, full thickness However, the attenuation was not critical for sensors with lengths of a few meters and could be used for sensor manufacturing. Thermoplastic tape, half of thickness Figure 2: Production of the sensing tapes using the sandwich technique This procedure was selected in order to keep the optical fiber in the center of the sensing tape cross-section, i.e. to prevent eccentricity of the optical fiber with respect to both horizontal and vertical axis (see Figure 1). This first batch was tested optically, microscopically and mechanically. Microscopic tests Microscopic tests were performed in order to verify the position of the optical fiber in the cross-section, to detect possible delamination caused by integration of optical fiber and to find out if the source of the losses in geometry of the optical fiber. The microscopic analysis was performed on 14 samples of the crosssection taken along a 15 meters long sample. The most important results are presented in Figure 4. TESTING OF THE SENSING TAPE Optical tests The aim of the optical test was to evaluate for the sensing tape the suitability to be used as a sensor. Small optical losses are not critical for SOFO sensor manufacturing, but it is obviously important that the fiber is not damaged or broken during embedding. In order to perform optical tests, the optical fiber was extracted from the tape and connectors were spliced. Visible light was inserted at one and observed at the other end of the 100 m long tape, as shown in Figure 3. When the test with visible light confirmed that the optical fiber was not broken, the other end was also pig-tailed, both ends were connected to an insertion loss meter and optical losses at 1300nm are evaluated. This test demonstrated high attenuation of light, in excess of 0.3 db/m. The following reasons could be in origin of attenuation: Figure 4: Microscopic observation of the tape cross-section. The position of optical fiber was not in the center of the cross-section and up-and-down variations were observed. The optical fiber was surrounded with the composite matrix, as shown in Figure 4, and this arrangement characterized 13 out of 14 samples. No delamination was observed but a varying amount of glass-reinforcing fibers surrounds the optical fiber and

3 in 3 cases only very thin layer of composites covers the optical fiber (~0.01 mm). The surfacing of the optical fiber, as well as delamination of composite, was observed in only one sample. This anomaly is considered as local, and can be improved by an improved control of manufacturing process. Therefore the possible source of attenuation is probably micro bending in vertical direction due to shrinkage of the matrix after the cooling. This effect can be avoided by pre-straining the optical fiber during the manufacturing process. Mechanical tests Three 17 cm long samples of the sensing tape were spliced to optical pigtails and provided with mirrors at the other end, in order to used as SOFO sensors. The samples were put in traction using a testing device and their deformations were registered using the SOFO system and traditional micrometer measurements. The aims of these mechanical tests were to compare the measurements performed with the SOFO system with those performed with micrometer in order to detect any sliding of the optical fiber, to determine the maximum measurable strain and to verify the mechanical properties of the tape. The samples were tested for linearity, repeatability, relaxation (creep) and maximal elongation. The samples presented in Figure 5. Sensing tape [ ] Micrometer [me] Figure 6: Correlation between micrometer and sensing tape readings The tests show that the sensing tape can be used as a deformation (average strain) sensor. However, improvements are required in manufacturing procedure in order to decrease the optical losses, avoid the local delamination and surfacing of the optical fiber. A second batch was later produced with a modified setup and processing parameters. It displayed much lower losses, but similar mechanical properties. ON-SITE TESTING OF SENSING TAPES An on-site test was performed on rails with the aim to compare the results obtained using sensing tapes and standard SOFO sensors. For this purpose, two onemeter long sensing tapes were installed on a rail in parallel with a SOFO standard sensor, as shown in Figure 7. Figure 5: Sensing tape samples Standard SOFO Sensor (behind, not Visible) CMP (in Car) Extension Cables The gage lengths of the sensing tape and the micrometer were 186mm and 130mm, respectively. Therefore the average strain given by each instrument is calculated as measured deformation divided with gage length and transformed in micro-strain [µε]. The average strain is calculated as a measured force divided by the area of the cross-section of the sensing tape. The full dynamic range was better than 3%. At approximately 3% the sensing tape was broken, which corresponds to specifications of the tape without optical fiber. No hysteretic behavior of the sensing tape was observed. The linear correlation coefficient between the tape and the micrometer during the force rising was and the degree of correlation was R 2 = The correlation between the micrometer and sensing tape measurements is presented in Figure 6. Sensing tapes (behind scotch) Thermocouples (behind PVC) Connection Box Figure 7: Sensing tapes and SOFO sensors installed on rails during testing The measurements were performed before the arrival of the train, and then repeated when the wagon was parked with the wheel exactly on the sensor location. One more measurement was performed after the wagon was moved. Since the location of the tapes and standard sensor with respect to the height of the cross-section was not the same, and the rail was subject to bending, the comparison was performed by testing for co-linearity of the strain measurements, as shown in Figure 8. The results of test have shown a correlation of the order of

4 the measurement resolution indicating an excellent behavior of the tape sensors. S. tape 1 37 mm Loading -11µε Unloading 1µε sensing tape produced and tested. Optical, microscopic and mechanical tests have assessed the good performance of the sensing tape and indicated necessary improvements in manufacturing procedures in order to better control of the position of the optical fiber within the tape s cross-section and decrease the optical losses. On-site tests on rails (surface installation) and high-pressure vessels (embedding) confirmed the good sensing performance of the tape and its applicability in real conditions. SOFO S. tape 2 6 mm Before Loading Under Load +20µε +24µε -1µε -4µε After Unloading Figure 7: Comparison of the strain reading as a function of load and vertical position This test continues in order to evaluate the long-term performance of the sensing tape exposed to different environmental influences such as wind, rain, snow, temperature variations and mechanical actions (shocks, vibrations, straining etc.). In order to prove the use of the sensing tape for composite material monitoring, several such sensors were embedded in a composite high-pressure vessel produced by filament winding. Figure 8 shows tape sensors embedded between carbon and glass fiber reinforcement layers in both longitudinal and hoop directions. Static and dynamic tests were performed on the tanks and confirmed a perfect bonding between sensing tape and host composite. Figure 8: Tape sensors embedded in a composite highpressure vessel CONCLUSIONS The integration of sensing polyimide-coated optical fiber into a fiber reinforced thermoplastic composite tape was presented in this contribution. The integration procedure was developed and a prototype of the ACKNOWLEDGEMENTS The Authors are indebted to Dr. Jannerfeldt, Dr. Maier and Dr. Vodermayer from GURIT SUPREM, Flurlingen, Switzerland, for collaboration and fabrication of the sensing tape, Mr. Rickenbach, Mr. Luthy and Prof. Dr. Ermanni from Laboratory for Mechanical Systems (IMES) of Swiss Federal Institute of Technolgy, Zurich (ETHZ) for their advises, availability and testing of the sensing tapes, the Post Center of Manno, Switzerland for rail testing, Mrs. Claire Nan from RouteAero Tech.&Eng., Taipei, Taiwan for making the rail data available, the ZEM EU project consortium for the composite vessel test and Mr. Cerulli and Mr. Rossi for their precious help during the whole project. REFERENCES 1. D.W. McClatchie, H.A. Reynolds, T.J Walsh, C. Lundberg, Applications Engineering For Composite Coiled Tubing, 1999 SPE/IcoTA Coiled Tubing Roundtable, March 25-26, Houston, TX, USA 2. Europe s 1 st composite Highway Bridge Opens in Oxfordshire, Journal of Composites Technology, pp 11, December D. M. Frangopol, A. C. Estes, G. Augusti, M. Ciampoli, Optimal bridge management based on lifetime reliability and life-cycle cost, Short course on the Safety of Existing Bridges, ICOM&MCS, pp , EPFL, Lausanne, Switzerland (1998) 4. B. Glisic, D. Inaudi, S. Vurpillot, Whole lifespan monitoring of concrete bridges, IABMAS'02, First International Conference on Bridge Maintenance, Safety and Management, Pages , on conference CD, July 14-16, 2002, Barcelona, Spain 5. "Field testing and application of fiber optic displacement sensors in civil structures", D. Inaudi, 12th International conference on OFS '97- Optical Fiber Sensors, Williamsburg, invited paper, OSA'(1997), Vol 16, p

5 Sensing tape for easy integration of optical fiber sensors in composite structures Branko Glisic, Daniele Inaudi SMARTEC SA, Via Pobiette 11, 6928 Manno, Switzerland Phone: , fax: , 35-WORD ABSTRACT The integration of optical fiber sensors in composite structures is often limited by the difficult handling of the bare fibers. In this contribution we present a novel packaging consisting in a composite tape that can be easily manipulated, embedded and surface mounted.