Skip to main content

Reliable optimal use of materials for wind turbine rotor blades (OPTIMAT BLADES)

Deliverables

The OPTIMAT research programme is an extensive material testing program that has conducted over 3000 individual tests over four years. The testing focus has been on the mechanical properties of the composite materials commonly used in modern wind turbine blades, specifically epoxy GFRP. This report aims to analyse the key research outcomes of the OPTIMAT research programme, and provide a summary of these key findings in the context of practical design guidance. The TG6 work captured in this report has reviewed all of the work conducted throughout OPTIMAT and summarised this in the context of practical applications in design for industry. As part of this review, the following key areas were reviewed: test coupon selection, stiffness assessment, S-N curve development, Constant Life Diagrams, rain flow counting, strain rate effects, variable amplitude loading, complex stress states, extreme conditions, thick laminates and repairs, residual strength, and the use of OptiDAT data in design. The current industry approach, the key OPTIMAT observations, and new perspectives have been identified for each of these areas. In all, 39 design recommendations have been developed in the following key areas: life time prediction, complex loading, residual strength, extreme conditions, and repair. These design recommendations will be considered for inclusion of the next version of DNV/GL design guidelines.
Within the scope of the OPTIMAT BLADES project was the development of repair methodologies suitable for application on the load carrying laminates of fiber reinforced wind turbine rotor blades, so as to avoid possible rejection of products both during production and service life. Currently there are no recommendations available for repairing structural parts of blades. Blades that are damaged or have production deficiencies in the thick structural parts are being destroyed, even if the damage or deficiencies are local. As the blades become larger more material is wasted due to such localized deficiency. However, material and costs savings can be made by adequate repair of locally damaged blades or blades with localized production deficiencies. To this end, as a first approach the location, type and importance of damaged zones was defined. Defects encountered during production, like dry spots and web to skin delaminations will need different repair techniques than those caused by lightning strikes or impact. Nevertheless, it was selected to concentrate on flaws that are found on parts of the primary structure of the wind turbine blades, e.g. the girder part of the blade and not parts that include structural foam. Repair techniques were surveyed and evaluated on aspects like complexity and suitability to large thickness and application on site. From the available repair methods for application to composite material structures in general, two were found suitable for the repair of the load carrying laminates of rotor blades, namely the scarf repair and the plug/patch. Semi-empirical stress calculation procedures are proposed, as selected from literature for a preliminary repair design. A minimum target value for the repair efficiency was stated with respect to both strength and stiffness. The most promising techniques were selected and applied to the repair of flat specimens. Within an extensive testing campaign these repaired specimens were tested in uniaxial tension and results with respect to both strength and exhibited stiffness were compared with that of flawless specimens, which were tested to form the necessary baseline. More than 100 static tension tests and 60 fatigue tests have been carried out on Glass/Epoxy coupons to investigate the effect of a number of parameters of the repair techniques, such as repair depth, slope of repair (in case of the scarf repair), form of the material of repair (liquid resin as opposed to use of prepregs), etc. Applied non-destructive inspection techniques for the quality assurance of the repaired area were also shortly assessed. Static test data show that the most promising from the selected repair techniques is the scarf repair with a slope of at least 1:50, which leads to a strength restoration of over 90% and scatter comparable to the reference coupons with adequate material use. Regarding the exhibited stiffness, results are within 10% of the reference elasticity. Verification of behaviour under tension fatigue loading was carried out for the selected repair method with similar results. Investigation on the effect of other conditions on the behaviour of the repair part, such as compression stresses both during static and fatigue loading will be continued in the near future. Other details of repair techniques applicable to composite material blades, such as curing cycles, environmental effects, etc. will also be investigated in order to arrive at a set of recommendations for the manufacturers leading to reliable repairs, thus prolonging the operating life of the blades and acceptance by accreditation bodies.
The laminate strength after (partial-lifetime) fatigue is measured at pre-determined life fractions. Apart from providing useful insights in the material strength behaviour, this forms the basis of strength degradation modelling, which can be used in lifetime prediction methods. Using strength degradation models, lifetime prediction can be improved relative to the "classical" Miner damage rule, by taking into account the effect of loading sequence. The OPTIMAT database contains results from more than 700 tests on two different laminate lay-ups, a unidirectional (UD) material and a multi-directional (MD) glass-epoxy laminate (see OPTIMAT specimen definition). The UD material was tested both longitudinally and transversely. Three R-ratios (R=10, R=0.1, and R=-1) were tested for each material, except where buckling instablity proved a problem. The database includes specimens, which failed prematurely during the fatigue phase of the test as well as the residual strength of surviving specimens at three life fractions (20%, 50%, and 80%). This data forms the basis for the model developments reported in the references.
The objectives of Task Group 2 in investigating blade material behaviour under complex stress states were to generate test results of basic plies and multidirectional laminates for the reference material to implement advanced FEM formulations, to define and validate experimentally multi-axial failure theories in static and fatigue loading and to quantify complex stress state effect on blade design by contributing with design recommendations. Concerning in-plane static and fatigue characterization, a comprehensive experimental program was performed for the UD material in the frame of TG2. To this end a number of at least 25 tests per category were performed, to have a statistical description as well, for measuring 19 engineering constants including elastic constants, strength properties and thermal expansion coefficients. All experiments were performed at the same lab (UP), same test rig, same procedures so as to minimize variations in material property. Test coupons were delivered directly by LM. Test methods and detailed results were presented in a number of OPTIMAT reports. For fatigue characterization of the UD material, definition of S-N curves at three R-ratios, 0.1, -1 and 10, both in the fibre and the transverse direction as well as in-plane shear fatigue strength under R=0.1 was performed. In all static and fatigue tests, except for in-plane shear characterization tests which were based on ISO 14129 geometry and test method, the standard OB coupon geometry was used, introducing the concept for just one geometry coupon valid for all types of tests, e.g. static or fatigue in tension and compression, residual strength etc. Only thickness was varying from coupons tested in the fibre direction to those loaded transversely to the fibre. Comparison of test results with ISO based test methods and coupon geometries revealed that strength and elastic moduli are in very good agreement except the compressive strength in the fibre direction where the OPTIMAT specimen performs not so well due to bending deformation. Concerning in-plane shear strength and modulus, a detailed comparison of several different methods and standards was presented showing large discrepancies in test results. It seems that the method selected in TG2, i.e. ISO 14129, tensile test of +/-45°laminated coupon yields fair results concerning both the shear modulus and strength. It is believed that the database created for the in-plane characterization of the UD material can be used for similar epoxy laminates if a single batch of confidence testing for a material combination gives similar results. With respect to the complex stress states in a rotor blade, it is believed that a rotor blade even if loaded in a complex mode, it is after all a thin-wall beam structure. As such, the tangential stress resultant in the shell thickness is negligible compared to the axial one. However, the shear stress resultant is of comparable magnitude to the axial one in areas as the shear webs or trailing and leading edges respectively. In conclusion, 1D stress analysis, i.e. beam theory formulations, could be acceptable but failure prediction should be done in a layer-by-layer basis, i.e. complex stress states at the ply level should be taken into account. As for failure prediction under complex stress states, typical plane stress states as those developed in the layers of the shells of a rotor blade can be simulated, besides sophisticated biaxial tests, by uniaxial testing in off-axis UD coupons. Failure prediction under static loading is performed satisfactorily using criteria such as Tsai-Wu, Puck and Tsai-Hill. Under cyclic loading, life prediction is also satisfactorily performed when using similar quadratic in stress functions compared to simplistic approaches, such as to consider separately damage from each stress component and add at the end. Experimental results from cyclic biaxial tests, using specimens either of cruciform or tubular geometry, were not conclusive. Static tests results from bi-axial tension of cruciform specimens made of MD lay-up suggest that failure prediction according to limit functions as those cited in the above is fair, in general, although many other parameters such as stiffness degradation scenarios or material non-linearity for example are of paramount importance.
Sixteen years back a load sequence for variable amplitude testing of materials in wind energy applications has been defined. The sequence has been synthesized from the measured flat wise blade root bending loads of 9 wind turbines varying from18 kW to 3 MW in power and from 12 m to 100 m in diameter. Very different operating philosophies have been covered. This load sequence called WISPER has found international acceptance and is widely used in variable amplitude testing of wind turbine rotor blade materials. In the context of the OPTIMAT BLADES project that aims at optimising materials and design recommendations for wind turbine rotor blade it has been proposed to set up a NEW WISPER standard load sequence that reflects today’s state-of-the-art in wind energy conversion technology. The idea is that material characteristics like fatigue life limits can be provided with better confidence for use in modern wind turbine rotor blade design when a test sequence reflecting today’s turbine technology is used to establish such characteristics. Following this line of thinking a work group within the OPTIMAT BLADES project has been formed to work out a NEW WISPER standard load sequence. The work group consisting of CRES, ECN, DEWI, DLR and WMC represents considerable experience in the field of wind turbine load determination and material testing. The reports present the major issues that have been discussed when creating NEW WISPER. The final resulting NEW WISPER sequence is presented and compared to the old WISPER standard sequence. The comparison is carried out on the basis of the rain flow range pair load spectra, 1-Hz equivalent load calculations and even more complex damage calculations using GFRP-material Goodman-diagrams and advanced damage accumulation models. Participants: Bernard Bulder, Johan M. Peeringa, ECN - NETHERLANDS ENERGY RESEARCH FOUNDATION, Wind Energy, P.O. Box 1, Westerduinweg 3, 1755 ZG Petten, The Netherlands Denja Lekou, Fragiskos Mouzakis, CRES - CENTRE FOR RENEWABLE ENERGY SOURCES, 19th km Marathonos Ave.,190 09 Pikermi, Greece Rogier P.L. Nijssen - WMC, P.O. Box 43, 1770 AA Wieringerwerf, The Netherlands Christoph Kensche, Olaf Krause, DLR - Deutsches Zentrum für Luft- und Raumfahrt, Pfaffenwaldring 38-40, D-70569 Stuttgart, Germany Holger Söker (Work Package Leader), Theo Kramkowski, DEWI - Deutsches Windenergie-Institut, Ebertstr. 96, D-26382 Wilhelmshaven, Germany
Introduction: OptiDAT is the database from the OPTIMAT BLADES project. The database is not directly meant for certification purposes, but the information and methodology is used for guidelines for testing of wind turbine blade materials. Contents: The OptiDAT database contains results of various tests on some 3000 coupons. In terms of testing time, this is 40.000 hours, the total number of fatigue cycles approximates 900 million. Tests were performed on two material systems; both are glass-fibre reinforced epoxy, but the resin systems are slightly different. The majority of the tests have been performed on the standard OPTIMAT geometries for Unidirectional and Multidirectional material. The focus is on fatigue results, and test results for various loading conditions can be found in the database, describing material behaviour in fatigue, static, bi-axial (tubular and cruciform), repairs, residual strength, variable amplitude, extreme conditions, thick laminates, shear loading, 4-point bending, loading in transverse direction, etc. Detailed information is also available on the thermal expansion coefficients, and for most plates, the fibre volume fraction and glass transition temperatures were measured and reported. Lay-out: The database is constructed in a spreadsheet format (MS Excel). This is for historical reasons as well as for compatibility. All the test results are reported in a single worksheet. Every row in the worksheet represents a test; every column contains data on various aspects of the test; coupon identification, laminate, dimensions, test type, test results. Although some 100 columns are reserved for detailed information on the various test types, further information may be required by the user. One column contains a reference to the various (ca. 150) technical reports with further information on the particular test. In these reports, information can be found which is, at the time, not feasible to include in the database (without making it difficult to handle), such as stress-strain diagrams, a thorough description of the test set-up, and images of the coupons’ failure modes. The database contains other sheets, with information on how to use the database, plate, geometry, and test type details, and contact information of the project participants. Also, the FACT database is included in a separate sheet. This set-up allows for quick comparison of test results. Storage: During the OPTIMAT project, the OptiDAT database was used for storage of the test results. The participating laboratories submitted the data as soon as possible after completion of the tests. Ensuring consistency, a data submission sheet was available for the contributing partners. Thus, all test results were available for review and further analysis during the test programme, rather than at the end of the programme. The database was uploaded to the OPTIMAT Web site after each addition of test data. Progress tracking The availability of the results facilitated tracking the test progress. The database included several progress tracking features. The Manufacturing and Delivery sheet, from the project’s main supplier, LM Glasfiber A.G., was included in the database. This showed for each batch of coupons the shipment date, recipient, coupon type, plate number, etc. The information was related to the data in to the database. Thus, for each delivered batch, the user could see directly how many of the coupons had already been tested hence, how many should be left. Also, the information from each Task Group’s Detailed Plan of Action was included, describing the test plan. This information was also linked to the test results already in the database, giving a detailed overview of the progress. Finally, the number of records, test hours, or number of cycles, could be visualised automatically using an interactive chart, giving a quick overview of the cumulative test results. Tools: The database contains some pre-programmed tools, to increase user-friendliness. The vast amount of columns and rows can be viewed in various modes using a specialised toolbar. This allows the user to limit the visible records to a particular test type, selection of columns, or both. A user-defined column view can also be set. To display fatigue data quickly, the database is equipped with a plot tool, which allows the construction of an S-N diagram using a selection of data. The result is a chart with the selected data, and a separate sheet with a copy of the selected records, for further processing. Prospects: OptiDAT is available from www.kc-wmc.nl from May 2006 for a small maintenance fee. It is freely available for OPTIMAT members and students). A subscriber set-up warrants regular updates, as well as inventorising the user-population, which helps tailoring the data to user requests. It will be expanded with the results from WP 3 of UPWIND.
Recommended testing methods Static tension: The following coupon geometries were experienced to work well in static tensile testing for the OPTIMAT glass epoxy materials: - Static tension 0° (UD). ISO 527-5, type A, with the exception that 25 mm (not 15 mm) wide and 2 mm thick (not 1 mm) specimens are suggested, OPTIMAT geometry I01. Alternatively OPTIMAT geometry R03 can be used. - Static tension 90°(UD). ISO 527-5, type B, with the exception that 2-6 mm thick (not only 2 mm) specimens are suggested, OPTIMAT geometry I01. Alternatively OPTIMAT geometry R03 can be used. - Static tension 0° (MD isotropic or orthotropic). ISO 527-4, type 3, 25 mm wide, OPTIMAT geometry I01. Alternatively OPTIMAT geometry R03 can be used. - Static tension 90° (MD isotropic or orthotropic). ISO 527-5, type B, 25 mm wide, OPTIMAT geometry I01. Alternatively OPTIMAT geometry R03 can be used. - Although not tested, ASTM D 3039 provides equivalent performance. Static compression: The following coupon geometries were experienced to work well in static compression testing for the OPTIMAT glass epoxy materials: - Static compression 0° and 90° (1 and 2 directions, both UD and MD). ISO 14126, type B1, with the exception that the gauge length can be 10-35 mm and the width can be 10 to 25 mm. - Although not tested, ASTM D 695 provides equivalent performance. Static shear: The following coupon geometries were experienced to work well in static shear testing for the OPTIMAT glass epoxy materials: - V-notch beam (Iosipescu) (UD and MD). ASTM D 5379, OPTIMAT I04-00 - Static shear 30° tension (UD). ISO 527-5, fibres at 30° relative to longitudinal axis of test specimen - Static shear ±45° tension (0°/90° special lay-up). ISO 14129, fibres at ±45° relative to longitudinal axis of test specimen. Fatigue The following coupon geometries were experienced to work well in fatigue testing for the OPTIMAT glass epoxy materials: - Tension fatigue (R=0.1):. OPTIMAT I01, OPTIMAT R03-00, OPTIMAT R04-00 - Compression fatigue (R=10): (UD and MD). OPTIMAT R03-00, OPTIMAT R04-00 - Tension-compression Fatigue (R=-1):° (UD and MD). OPTIMAT R03-00, OPTIMAT R04-00 The Frequency of fatigue testing should be chosen to limit the temperature rise of the test specimen to a maximum of 10 degrees centigrade above the ambient temperature. In case of the OPTIMAT material, frequencies were chosen to be linear with the inverse of the square of the strain in the specimen, in order to keep dissipated energy constant.
Extended damage mechanics modelling is used to describe the fatigue damage in polymer composite materials The measurements of the stiffness degradation are used to determine the corresponding internal state variable function experimentally. The relationships between particular ISV and corresponding stiffness degradation is formulated for considered damage mechanism. It was observed experimentally that the normalized stiffness degradation of conditioned and reference materials are rather close, and are approximated with the same polynomial. The progressive fiber fracture is considered as main contributing mechanism. The outlined micromechanics methods gives possibilities to study the interface strength and debond crack growth effects on formulated damage tensor. The continuum damage mechanics approach is used to characterize the damage evolution in long fiber-laminated composites. The internal state variable evolution law associated with damage in fiber direction is formulated in terms of used theory, and determined experimentally and theoretically. The theoretically and experimentally determined evolution law of internal state variable is used in damage dependent constitutive relationships for laminate in order to predict the static strain - stress behaviour of arbitrary laminates. A methodology is proposed to measure the values of the internal state experimentally. The FEM analysis and Monte Carlo simulations are used in order to predict the internal state variable according to the outlined theoretical formulation. The predicted values of internal state variables are compared with experimental results.