Warp Knitted Textile-Based Sensors for In-Situ Structural Health Monitoring of Wind Turbine BladesEric Haentzsche*, Ralf Mueller, Georg Bardl, Andreas Nocke and Chokri Cherif
AbstractThe structural health monitoring of large-scaled fiber-reinforced composite components plays a crucial role for the further advancement of lightweight design approaches for a large-range application spectrum. Using textile-based and technological integrated stress sensors within the composite’s textile reinforcement, the detection of serious structural damages on early stages as well as an in-situ monitoring of mechanical loading conditions in inaccessible areas within immediate distance of the load-bearing layers of the subsequent composite component can be realized by those in situ condition monitoring systems, enabling the possibility of just in time maintenance or even local repairs before full structural failures occur.
IntroductionThe continuous structural health monitoring of large-scaled
fiber-reinforced composite components, e. g. wind turbine blades, plays a crucial role under safety engineering aspects and for reasons of economy as well as for the further advancement of lightweight design approaches. The detection of serious structural damage on early stages (crack initiation and crack propagation processes) as well as a continuous in-situ monitoring of local and global mechanical loading conditions in inaccessible areas within immediate distance of the load-bearing layers can be realized during system operation by structurally integrated and textile-based strain sensors. Therewith, idle periods of wind turbines due to blade inspections will be significantly decreased or even avoided because local repairs and even overhauls can be initiated in time before full structural failures occur.
The suitability of carbon filament yarns (CFY) acting as basic- material capable of being integrated as n-dimensional sensors in thermoplastic composites for in-situ load monitoring and condition monitoring task has been shown in previous investigations [1-3] with special regard to the reachable characteristic sensor values under
*Corresponding author: Eric Haentzsche, Institute of Textile Machinery and High Performance Material Technology (ITM), Faculty of Mechanical Science and Engineering, Technische Universitaet Dresden, Hohe Straße 6 01069, Dresden, Saxony, Germany, Tel: +49 351 463-31635; Fax: +49 351 463-34026; E-mail: [email protected]
Received: August 11, 2016 Accepted: September 28, 2016 Published: October 02, 2016
quasi-static mechanical loads. Using the CFYs’ piezoresistive effect, mechanical strains can be calculated and visualized due to a correlative change of the carbon filaments resistance. Even structural changes, e g. crack initiation and propagation processes can be measured and visualized based on this measurement principle.
For the integration of the textile- based sensors into textile reinforcement structures for composite applications, integrative manufacturing processes have been successfully applied [4,5]. The force-fit linkage with non-crimp fabrics (NCF) during the multi-axial warp knitting process is done through a warp yarn path manipulation (WPM) device, enabling the translational displacement of individual sensor yarn bands over the fabric’s width in a range of 0° to 90°. Using this technology, customized two-dimensionally integrated sensor layouts with adjustable integration lengths within the reinforcement fabric have been manufactured successfully.
This contribution presents current research activities aimed at the technological integration of CFY sensors into glass fiber NCFs acting as reinforcement of large-scaled composite components. Furthermore, the electro-mechanical behaviour and long-term stability of CFY sensors will be shown under quasi- static and dynamic mechanical loadings of a functional model of a wind turbine blade, built up in thermoset composite design. With this approach, in-situ sensor systems within the load-transferring areas of composites have been realized successfully, allowing a continuous and non-destructive global and even local structure- and load-monitoring with excellent gauge protection.
ExperimentalIntegrative sensor manufacturing
A novel innovative approach aims at the 2-dimensional integration into and force-fit linkage of textile-based strain sensors with non-crimp fabrics (NCF), respectively. Therefore, a special warp yarn path manipulation (WPM) device is used during the multi-axial warp knitting process, enabling the translational positioning of 0° yarns in the production direction and of individual sensor yarn sheets at exact positions over the whole fabric’s width in a stacking angle range of 0° to approximately 90°. Due to the synchronization of Karl Mayer Malimo’s mutltiaxial warp knitting machine MALIMO® 14024 with the WPM device developed at ITM of TUD, a damage-free and non-displaceable setting of the CFY sensor yarns into the NCF basic structure can be guaranteed. Stacking sequences of the reinforcement yarns, e. g. glass or carbon fiber, between ± 45° up to approximately 90° become realizable, whereat the sensor layer is principally the second to last layer from the top ply (0° warp yarn ply).
The design of the new WPM device allows an offset of two additional warp yarn sheets in user -definable stacking angles relating to the production direction at defined positions and over defined numbers of weft yarns to generate additional functions. Figure 1 shows a conceptual integration of the WPM device in a multi -axial warp knitting machine. The laying device of the WPM device traverses immediately in front of the stitch-bonding area allowing an offset adjustment according to the gauge of the grid element (Figure 2a). Therewith, the translational offset yarns can be accurately
Citation: Haentzsche E, Mueller R, Bardl G, Nocke A, Cherif C (2016) Warp Knitted Textile-Based Sensors for In-Situ Structural Health Monitoring of Wind Turbine Blades. J Fashion Technol Textile Eng S2:004.
• Page 2 of 5 •
doi: 10.4172/2329-9568.S2-004
Proceedings of AUTEX-2015 Conference
bonded over the grid intersection. Crossings of warp yarn patters due to the manipulation can be realized by separate aligned laying devices extending across the whole working width of the multi- axial warp knitting machine. The number of lateral offset warp yarns per laying device is extendable according to the requirements on the reinforcement structure. Each warp yarn is fed from individual spools into the tubular guide (Figure 2b).
The spare warp yarn lengths resulting from the pattern-related motion of the laying devices is compensated by special tensioners with yarn accumulators. A single yarn feeding system for the bonding yarn is also realized as varying stitching lengths are required for each yarn depending upon the manipulation path. Therewith, the required yarn length for the loop formation can be provided for each knitting needle. The translational offset of the laying device is numerically controlled and synchronized with the drives of the multi-axial warp knitting machine. The movement for the weft alignment process, path and speed of the transport chain are recorded by an external absolute encoder connected to the transport chain. Width and distance between two subsequent weft yarns is measured by a photoelectric barrier. The drive control and translational offset are synchronized with Siemens’ modular SIMOTION® system. The offset data for the laying device is derived from an offline yarn course calculation plugin module designed for commercially available CAD programs. In order to ensure a secure fixation of the manipulated warp yarns, the movements of the manipulated yarns are synchronized with the knitting needle with a photoelectric barrier that detects the needle position.
Textile-based in situ sensors for structural health monitoring
Using this integrative manufacturing technology, customized two-dimensionally integrated sensor layouts with adjustable integration lengths within the reinforcement fabric have been manufactured successfully (Figure 2c). After drapery and stacking of suchlike functionalized semi-finished products in conventional molding tools, the textile-based sensor’s electrical interconnection and contacting to special signal interfaces within the area of the subsequent composite components periphery, the stack of plies can be infiltrated and consolidated with conventional thermoset or thermoplastic matrices to obtain the final composite component (Figures 3a and 3b). Due to the integrative sensor manufacturing and electrical contacting before the consolidation, no further time- and cost -consuming steps are necessary, as they have to be spent comparatively for the application of conventional metal-foil strain gauge(s) chains.
Based on the described translational WPM technology for the MALIMO® 14024 stich-bonding machine, the technological integration of multiple CFY strain sensors into glass fiber NCFs acting as reinforcement for large-scaled composite components has been successfully realized. The electro-mechanical behaviour and long- term stability of the CFY bending sensors have been investigated under quasi-static and dynamic mechanical loadings on a functional model of a wind turbine blade (Figure 3c) in thermoset composite design. With this approach, in-situ sensor systems within immediate distance to the load-transferring areas of composite components can be successfully manufactured during the textile reinforcement’s fabrication, allowing a continuous in-situ and non-destructive global and even local structure-and load-monitoring with excellent gauge protection. The bending strain is measured with integrated and calibrated CFY strain sensors within the warp knitted reinforcement of the blade semi shells. There is equal number of sensor slopes integrated in each semi shell, located below and above the blade’s neutral axis, respectively. Thereby, geometrical as well as electrical similar sensors in each shell are boarded pairwise in Wheatstone half-bridges. This allows in addition to an excellent compensation of thermal drifts, a gain of the change of resistance (bridge detuning) indicated on the superposed mechanical bending strain of such kind resistive sensors.
ResultsMeasurement setup
The wind turbine blade is mounted horizontal in a special test rig. For applying quasi- static load profiles, hourly increasing weights are attached on the blade at two different load transmission points (1,150 mm and 2,300 mm behind the blade’s restraint), generating negative bending moments of MB=-(9.0…85) Nm. The nominal bending of the blade is captured by inductive displacement transducers (W1T3 and WTA series, HBM Hottinger Baldwin Messtechnik HBM GmbH) at three measurement s01-s03 points behind the blade’s restraint. The bending strain is measured with integrated and piezo-resistive CFY strain sensors within the NCF reinforcement of the blade’s semi shells. Performing quasi-static load profiles, three CFY sensor pairs (Ch01 - Ch03) graded by integration lengths located below and above the neutral axis of the blade are boarded pairwise in Wheatstone half-bridges. This allows in addition to an excellent compensation of thermal drifts, a gain of the change of resistance (bridge detuning) indicated on the superposed mechanical bending strain of such kind resistive sensors. Figure 4 depicts the used measurement setup schematically.
(A) (B)
Figure 1: Integration concept of the translational warp yarn path manipulation (WPM) device into a multi-axial warp knitting machine MALIMO® 14024: CAD model view (A) and cross-sectional view (B)
Citation: Haentzsche E, Mueller R, Bardl G, Nocke A, Cherif C (2016) Warp Knitted Textile-Based Sensors for In-Situ Structural Health Monitoring of Wind Turbine Blades. J Fashion Technol Textile Eng S2:004.
• Page 3 of 5 •
doi: 10.4172/2329-9568.S2-004
Proceedings of AUTEX-2015 Conference
Wind turbine blade monitoring under quasi-static loads
Performing the quasi-static bending tests, the accumulated bending strain can be measured as overlain tension and compression strain along the particular integration path length of the integrated CFY sensors above and below the blade’s neutral axis because of their circuitry as bending sensors in Wheatstone half bridge. Although the highest deflection is located at the blade’s tip and converges to zero toward the blade’s restraint, the nominal amount of accumulated strain measureable along the sensors length increases continuously toward the blade tip. This structure mechanical effect becomes apparent by comparison the measured bridge detuning of the CFY sensor pair Ch01 (integration length: 2 × 1,130 mm; basic resistance: each 993 Ω) with CFY sensor pair Ch03 (integration length: 2 × 2,285 mm, basic resistance: each 1,997 Ω), which exhibits a difference of +50% (load transmission point located at 1,150 mm; (Figure 5a)) or - 31 % (load transmission point located at 2,300 mm, (Figure 5b)) respectively, in each case at the maximum bending moment of MB=83 Nm.
Because of the complex NACA airfoil profiled shape of the blade, the mentioned correlation is not strictly consequent. A spatially resolved stress/strain measurement has been assured as possible.
The time elapsed of the blade’s deflection due to the quasi-static loads, its transient strain as well as its thermal induced reset over a period of approximately 13 h back into neutral position after the exposure as well as thermal caused relaxation due to an 1.9 K ambience heating of the test rig is summarized in Table 1. Figure 5 depicts the measureable FRP behaviour with the integrated CFY in situ bending stress sensors. Applying hourly increased quasi-static loads (0 < time < 30000 s), a good correlation between bending strain caused bridge detuning depending to the CFY sensor pairs particular integration lengths can be observed. As expected, the longer the CFY sensors integration length, the higher the measureable accumulated strain mapped to the bridge detuning of the measuring bridge. After complete unloading, also an excellent correlation between the blade’s transient strain and thermal caused reset due to an temperature difference of about -3.9 K into its neutral or horizontal position can be observed within 30000 s < time ≤ 76000 s. Due to an abrupt rise of the ambient temperature of about 1.9 K (76000 s < time ≤ 82500 s), the thermal induced relaxation of the wind turbine blade can also be measured by the CFY in situ sensors (Figure 6). The repeat accuracy of the integrated CFY sensors for a multiple repetition of these tests concerning their time response is therefore obvious. In each case, a good correlation between the particular load conditions mapped with the inductive displacement transducers and the integrated CFY sensors can be observed.
(A) (B) (C)
Figure 2: WPM device within multi-axial warp knitting machine MALIMO® 14024 with laying device in front of the stitch-bonding area (A), Its single yarn feeders for the translational manipulated warp yarns (B), and there with manufactured glass fiber NCF with integrated CFY strain sensors (C).
(A) (B) (C)
Figure 3: Manufacturing sequence of FRP wind turbine blade with integrated CFY sensors graded by integration lengths acting as bending sensors for spatially resolved SHM task: drapery of NCF reinforcement into molding tools (A), electrical contacting of CFY sensors and consolidation of blade shells due to thermoset matrix (B), and finalized in situ SHM system for wind turbines (C).
Citation: Haentzsche E, Mueller R, Bardl G, Nocke A, Cherif C (2016) Warp Knitted Textile-Based Sensors for In-Situ Structural Health Monitoring of Wind Turbine Blades. J Fashion Technol Textile Eng S2:004.
• Page 4 of 5 •
doi: 10.4172/2329-9568.S2-004
Proceedings of AUTEX-2015 Conference
Figure 4: Measurement setup for evaluating of the in situ CFY sensor’s behaviour at quasi-static bending stress applied to the FRP wind turbine blade.
Figure 5: Bending stress depending strain measurement with in situ CFY sensors for load transmission point at 1,150 mm (a) or 2,300 mm (b) respectively: change in resistance (mapped to bridge detuning) depending to the accumulated bending strain along the CFY sensor’s particular integration lengths (Ch01: 2 x 1130 mm Ch02: 2 x 1,900 mm Ch03: 2,285 mm).
ConclusionsThe translational WPM technology can as well be transferred
for the appropriate aligning of reinforcement yarns during the manufacturing of the sensor carrying NCF structure allowing tailored or bionic reinforcements suitable for flux of force that can be realized in a one- step procedure. The new technological development of the WPM has now opened up new and innovative application areas of integrative producing textile reinforced structures with textile-based sensors and sensor networks for in-situ SHM tasks according to the flux of forces and in consideration of user-specific restrictions for hybrid formations of the directed composite component. This offers a great flexibility to adapt the mechanical behaviour as well as the in-situ SHM sensor layouts to the expected load profiles over the FRP component’s destined lifespan. The realization of stress-related designed and even bionically inspirited reinforcement structures becomes possible, avoiding structural oversizing and therewith
inefficient usage of cost-intensive and raw fiber materials, e.g. carbon or aramid fiber. The consequent textile-technological in-line production allows the creation of a new generation of cost-efficient and adaptive reinforcement structures for the processing to multi-functional and innovative FRP components for a large-scaled application spectrum. The novel mindset requires the use of fundamental technologies to be made suitable for a resource -efficient mass-production of lightweight structures with high -performance and functional density. Combining discrete and material- specific processes to the technologies suited for large-scale manufacturing will enable a multi-material application specific design that makes the lightweight composite structures suited for complex demands.
The new WPM technology exhibits great potential for serial large-scale production of complex FRP components.
Table 1: FRP blade’s transient strain and thermal caused reset indirectly measured via deflection with inductive displacement transducers (s01 - s03) and integrated CFY sensor pairs (Ch01 - Ch03).
Citation: Haentzsche E, Mueller R, Bardl G, Nocke A, Cherif C (2016) Warp Knitted Textile-Based Sensors for In-Situ Structural Health Monitoring of Wind Turbine Blades. J Fashion Technol Textile Eng S2:004.
• Page 5 of 5 •
doi: 10.4172/2329-9568.S2-004
Proceedings of AUTEX-2015 Conference
Acknowledgement of Financial Support
The IGF research project 17529 BR/1 of the Forschungs-vereinigung “Forschungskuratorium Textil e. V.” is funded through the AiF within the program for supporting the „Industriellen Gemeinschaftsforschung (IGF)“ from funds of the Federal Ministry of Economics and Energy (BMWi) by a resolution of the German Bundestag. Financial support is gratefully acknowledged. We like to thank all the participating companies for their technical support and the supply of test material as well as all further partners supporting our research work within this application area. The final project report and continuative information are available at Institute for Textile Machinery and High Performance Material Technology (ITM) of TU Dresden.
Figure 6: FRP wind turbine blade’s stress measured with three CFY in situ sensors graded by integration lengths during hourly increased quasi-static loads at load transmission point 2,300 mm behind blade restraint.
References
1. Haentzsche E, Nocke A, Matthes A, Cherif C (2013) Sensory characteristics of carbon fiber based strain sensors and integration techniques into textile reinforced structures for in situ monitoring of thermoplastic composites. Sensor 2013-16th International Conference on Sensors and Measurement Technology, Proceedings, Nuremberg, Germany.
2. Haentzsche E, Matthes A, Nocke A, Cherif C (2013) Characteristics of carbon fiber based strain sensors for structural-health monitoring of textile-reinforced thermoplastic composites depending on the textile technological integration process. Sensors and Actuators A: Physical 203: 189-203.
3. Haentzsche E, Unger R, Nocke A, Hutloff D, Cherif C (2014) Carbon filament yarn-based sensor networks for spatially resolved monitoring of fiber-reinforced composites. Technical Textiles 57: 22-25.
4. Haentzsche E, Nocke A, Cherif C (2013) Textile-based sensor networks for the structure monitoring of lightweight structures and membrane constructions. 17th International Conference on Composite Structures (ICCS17), Proceedings, Porto, Portugal.
5. Haentzsche E, Kluge A, Nocke A, Cherif C (2014) Multi-functional fiber-reinforced plastics with integrated textile-based sensor and actuator networks. 16th European Conference on Composite Materials (ECCM16), Proceedings. Seville, Spain.
Author Affiliations Top
Technische Universitaet Dresden, Institute of Textile Machinery and High Performance Material Technology (ITM), Dresden, Germany
This article was originally published in a special issue, Proceedings of AUTEX-2015 Conference
Submit your next manuscript and get advantages of SciTechnol submissions
80 Journals 21 Day rapid review process 3000 Editorial team 5 Million readers More than 5000 Quality and quick review processing through Editorial Manager System
Submit your next manuscript at www.scitechnol.com/submission
