Back to the programme printer.gif Print



Thursday, 13 March 2014
11:15 - 12:45 Advanced operation & maintenance
Science & Research  


Room: Llevant
Session description

The session covers the entire area of wind farm and wind turbine operation and maintenance, e.g. how to access, repair and organise operation and maintenance logistics onshore and offshore. In order to keep in hand the current health of turbines and farms, failure detection, identification and prognosis methods are also presented. Maintenance operations are also addressed from the viewpoints of required activities and efficiency. In order to cover management aspects, operation and lifetime cost calculation methodologies are also introduced. Experts from various European countries share their results during the session.

Learning objectives

  • Advanced operation and maintenance
  • Fault detection methods
  • Reliability calculation techniques
  • Monitoring on the field
  • Statistical and artificial intelligence-based solutions for diagnostics data- and model-based solutions for fault detection
Lead Session Chair:
Zsolt Viharos, Hungarian Academy of Sciences, Hungary

Co-chair(s):
Christopher J. Crabtree , University of Durham, United Kingdom
Alexandros Antoniou Fraunhofer Institute for Wind Energy Systems, Germany
Co-authors:
Alexandros Antoniou (1) F P Matthias Saathof (1) Stefan Krause (1)
(1) Fraunhofer Institute for Wind Energy Systems, Bremerhaven, Germany

Printer friendly version: printer.gif Print

Presenter's biography

Biographies are supplied directly by presenters at EWEA 2014 and are published here unedited

Dr. Antoniou has been working in the WInd Energy sector since 2000. Currently he is a senior scientist in Fraunhofer Institute for Wind Energy Systems, responsible for the development and application of new testing and monitoring technologies for wind turbine rotor blades. He studied mechanical engineering in the University of Patras in Greece where he also finished his PhD in the department of applied mechanics. He has worked as researcher and project manager in many industrial and public funded projects. His research is focused on material and structural performance of composites in wind turbine rotor blades.

Abstract

Trailing edge monitoring with acoustic emission during a static full scale blade test

Introduction

Offshore wind turbines installation will grow the upcoming years, thus more wind turbine blades will come in operation. However, these parts of the wind turbine are inadequate monitored in terms of structural integrity. Unexpected blade failures are resulting in power production loss and high repairing costs. A 20% catastrophic failure of the rotor blades is addressed to the adhesive joints [1]. Hence, an Acoustic Emission system is implemented in a full scale large blade test towards investigating and developing sensor network and data analysis strategies for the structural health monitoring of the trailing edge adhesive joint.

Approach

Previous research has shown the potential gains of using the acoustic emission monitoring technique. Small rotor blades have been monitored with Acoustic Emission investigating the correlation of the damage accumulation after static and fatigue loads and the wave emission performance in different structural damage mode cases [4-5]. Acoustic Emission has been used in large full scale blade tests monitoring the overall blade performance [6] in a static test and predicting the spar cap buckling failure real time. Based on data of material acoustic emission tests under quasi- static loads, acoustic emission was implemented in order to monitor a fatigue full scale blade test and discriminate between separate damage events [7].
In this study the damage mechanisms developed in mechanically loaded thick adhesive joints of wind turbine blades is investigated, implementing acoustic emission real time monitoring. Therefore:
• Piezoelectric sensors are detecting the stress waves released from the damage occurring in the structure under externally imposed loads.
• Linear location algorithm is used for the location of the events. Provided a known wave propagation velocity, the damage source can be located real time by the arrival time differences of multiple sensors registering the same signal.
• An adequate sensor setup is developed for the monitoring of the trailing edge of a large rotor blade (over 55m).
• Data analysis is proposed for discriminating between different failure modes
A step wise health monitoring approach is followed in different structural test levels. Due to the complexity of the rotor blade geometry, the monitoring features of the acoustic emission technique are assessed under mechanical static and fatigue testing of a sub-component i.e. a 1.5m long composite I-beam structure. The sub-component is modeling the joint between spar caps and a shear web [2], as it is commonly found in rotor blade structures. Separate damage modes occurring in the aforementioned mechanical tests are discriminated implementing advanced data analysis. Intensity factor and Historic index [3] are highlighting small slope changes in the energy vs. time graphs indicating for occurring damage. The monitoring technique is further developed and applied on a full scale blade test.


Main body of abstract

Asymmetric three point bending static beam tests have been performed in order to evaluate the capability of the acoustic emission testing to identify and locate occurring damage in the adhesive bond line. The initial objective is to determine an experimental setup apt for monitoring the beam mechanical tests and the damage mechanisms, see Fig.1.

Fig. 1: Experimental set-up
During the static test and before the final failure of the sub-component there are transverse cracks occurring in the bond line between the shear web of the beam and the spars. The acoustic emission system can accurately detect the nucleation of the transverse tensile cracks (mode I) in the beam bond lines, both in terms of location and time of occurrence, Fig. 2. Signals with high amplitude are temporally indicating the emerging damage, while the implemented linear location algorithm is deriving the crack position along the beam. There is very good correlation with the visual inspection.

Fig. 2: Results of the source location and correspondence to actual damage (beam test)
In the fatigue tests the aforementioned geometry sub-components are subjected under alternating loading. The monitoring experimental setup derived during the static tests is applicable in the fatigue tests too. However, the fatigue-specific damage modes in the adhesive joint could not be detected in the recorded data as obvious as the static tensile cracks. Therefore, a combination of different data analysis methods is used to identify different stages of damage progression until the final failure of the beam.
Source linear location is still used in order to identify the occurrence and saturation of mode I cracks, implementing the same procedure developed for the static tests. The same mode I cracks as already recognized in the static tests are emerging in the fatigue loading too. They are also registered with relative high amplitudes giving a local peak in the recorded energy of the closest sensors recording the event. At the transverse crack tips, shear-driven cracks are starting to propagate along the adhesive joint. These interface cracks (mode II) which are propagating very slow, generate very low energy acoustic emission signals. Thus they had a very small contribution in the cumulative energy of the adjacent sensors recorded signal. In Energy vs. cycle fatigue number graph, see Fig.3 it is shown that after a phase I where most of the transverse cracks are developed, another phase II is following indicated with a reduced accumulation energy slope. As soon as new drastic slope occurs, the fatigue life enters the last phase III of its fatigue life which is relative small to the previous two. This also indicated the upcoming structural failure of the component.

Fig. 3: Typical cumulative energy vs. fatigue cycles AE signals performance during fatigue life

Intensity signal analysis and Historic index [3] are used to identify patterns in the energy released throughout the test, enhancing the comparability between separate beam tests.
The developed know-how in the experimental setup and the recorded acoustic emission data analysis is applied in full-scale blade tests. The sensor network is designed accounting for the larger attenuation of the blade composite structure in comparison with the beam structure i.e. 23m are covered using 16 sensors in total, with 1-2m distance from each other, see Fig.4.

Fig.4 Full-scale rotor blade test setup for monitoring trailing edge damage
The monitored area is the adhesive joint in the trailing edge of the blade, see Fig.5, during the ultimate tensile test.

Fig.5 Applied Acoustic Emission system on the blade trailing edge
Most of the developed cracks and delaminations in the bond line are identified and confirmed with a visual inspection, see Fig.6.

Fig.6 Emerged damage in the trailing edge. Acoustic emission vs. visual inspection


Conclusion

An extreme static full scale blade test is successfully monitored real time with the Acoustic Emission technique. During the tensile test, a sensor network is applied on the blade trailing edge laminate, which is the area of interest. Twenty three meters (23m) are covered with 16 sensors having 1-2m from each other. The emerging transverse cracks and small size delaminations in the bond line of the trailing edge were identified both temporally and spatially. A linear location algorithm is implemented resulting in very good comparison with the location of the identified damage during the visual inspection.
Before the implementation on the full scale blade structure, the Acoustic Emission experimental setup and performance are thoroughly investigated during small geometry scale structural tests. Asymmetric three point bending tests are performed on I-beam profile sub-components under static and fatigue loadings. These are composite adhesive joints, designed to simulate representative stress states of a wind turbine blade adhesive connection.
Different damage modes are distinguished through data analysis. Transverse tensile cracks (mode I) in the bond lines are registered with high amplitude events, both in static and fatigue loadings. These are visually compared very well with the mode I identified damage on the blade. Horizontal-shear cracks propagating in the interface between spars/shear webs and the bond lines are resulting in slow increase of the general emitted signal energy and therefore could be identified indirectly. Acoustic Emission intensity signal analysis could enhance the identification of the presence of different failure modes. Moreover, a strong peak in the emitted energy indicated the upcoming final failure. This performance could be used in order to predict the component catastrophic failure.



Learning objectives
This is a primary investigation for the applicability of the Acoustic Emission technique in the monitoring of large scale blade tests. The sensor network, the location algorithm and the acoustic emission data analysis had to be developed and applied. These will be used for the evaluation of fatigue tests and furthermore for operating rotor blades in the field.


References
[1] D. J. Lekou, "Scaling limits & costs regarding WT blades", Project Upwind report, 2010.
[2] F. Sayer, A. Antoniou and A. van Wingerde, "Investigation of structural bond lines in wind turbine blades", International Journal of Adhesion & Adhesives, pp. 129-135, 25 Januar 2012.
[3] T. J. Fowler, J. A. Blessing, P. J. Conlisk and T. L. Swanson, "The MONPAC System", Journal of acoustic emission, pp. 1-8, 1989.
[4] P. Joosse, M. Blanch, A. Dutton, D. Kouroussis, T. Philippidis and P. Vionis, "Acoustic Emission Monitoring of Small Wind Turbine Blades", American Institute of Aeronautics and Astronautics, American Society of Mechanical Engineers, 2001.
[5] A. Dutton, M. Blanch, P. Vionis, V. Kolovos, D. van Delft, P. Joosse, A. Anastassopoulos, D. Kouroussis, T. Kossivas, J. ter Laak, T. Philippidis, Y. Kolaxis, G. Fernando, G. Zheng, Z. Liu and A. Proust, "Acoustic Emission Monitoring from Wind Turbine Blades undergoing Static and Fatigue Testing", Proc. of 15th World Conf. on NDT, Rome, 2000.
[6] B. F. Sørensen, L. Lading, P. Sendrup, M. McGugan, C. P. Debel and O. J. Kristensen, "Fundamentals for Remote Structural Health Monitoring of Wind Turbine Blades - a Preproject" Risø National Laboratory, Roskilde, 2002. (http://orbit.dtu.dk/fedora/objects/orbit:90533/datastreams/file_7712416/content)
[7] E. Schulze , L. Schubert , B. Frankenstein, "Monitoring of a wind turbine rotor blade with Acousto Ultrasonics and acoustic emission techniques during a full scale fatigue test", European Workshop on Structural Health Monitoring, Sorrento, Naples, Italy, June 28 - July 4, 2010