Can fiber optics bring light into the darkness? – Grout monitoring on offshore wind turbines

Safe and reliable operation and a long service life of offshore wind turbines are key targets on the road to achieving climate neutrality, and the implementation of a predictive maintenance strategy is particularly relevant when it comes to reducing the associated maintenance costs. As such, monitoring of the integrity of the support structure and its components will continue to increase in importance for the kind of future support structures required in farms of current performance classes and greater water depths. The structural health monitoring (SHM) concept is an important element here. As the primary task of an SHM system is to detect damage to a support structure at an early stage, it must be capable of identifying structural changes early and differentiating them from natural variations occurring as a result of evolved environmental and operating conditions. For example, age-related changes in the behavior of the foundations (soil-structure interaction), the support structure, and its connecting components must be identified, differentiated from each other, and assessed taking the current overall dynamic operating behavior of the turbine into consideration. Doing so makes it possible to assign anomalies in the structural-dynamic behavior to potential local damage on individual components. Good interaction between the sensor technology, measuring technology, and evaluation systems is essential here, with self-learning monitoring systems or supervised machine learning for the identification and localization of specific damage processes being of great importance.

Infrastructure for physical model tests

The findings from large-scale model tests can be of key advantage in the design and successful implementation of SHM systems. In the physical model test, damage can be introduced into the structural components under investigation in defined load situations and analyzed using measuring technology. Fraunhofer IWES develops, monitors, and analyzes such tests at the Test Centre Support Structures (TTH) in Hannover as part of a cooperation with the Leibniz University Hannover launched a decade ago. The grouted connection of offshore wind turbines represented a special application in the Grout-WATCH joint project. This is a hybrid connection in which two steel tubes (pipe and sleeve) with different diameters are inserted into each other axially and the space between them (annulus) filled with high-strength grout in order to create a force-fit bond. In cooperation with the associated partners, Fraunhofer IWES was responsible for the planning and implementation of tests on grouted connections subjected to cyclic bending stress. The focus was on documenting the local, non-linear material damage in connection with the development of cracks in the grout material by means of fiber-optic strain gauging.

Grout sensor technology for use in SHM systems

The processes in the highly stressed grout were documented using an optical measuring system based on fiber Bragg gratings (FBG). FBG sensors allow determination of strain from measured shifts in the wavelength of the light signals reflected in the sensor. To this end, one of the companies involved in the joint project developed FBG grout sensors (see Nuber et al., 2023) for strain gauging in an 18.4-mm-wide grout annulus. The aim was to draw conclusions about possible crack initiation or development from the measured strains or from the shifts in wavelength occurring in the locally placed sensors. The usability of this approach was confirmed in the joint project in small-scale tests in the laboratory (see Nuber et al., 2023).

Test design and integration of the FBG sensors

In the scope of an experimental test on the large-scale model of a monopile support structure, the integrated grouted connection (see Fig. 1 (left, right)) was systemically damaged by means of gradually increased cyclic alternating loads, and it proved possible to document the damage process locally via novel FBG sensors in the grout. The FBG sensors positioned in the grout annulus (see Fig. 1 (center)) thus made it possible for the first time ever to derive correlations between local damage processes and the otherwise typical sensors of a monitoring system (see Nuber et al., 2023) and ultimately to evaluate the usability of the sensors.

Fig. 1: Technical drawings of the test setup in the clamping field with enlarged view of the grouted connection and sensor technology (left, center) (Kohlmeier et al., 2024). View from above of the test setup in the clamping field at the TTH (right) © Fraunhofer IWES

Time series of the FBG sensor technology positioned in the grout

An experimental modal analysis and a free vibration decay test were performed to verify the measuring technology and installed sensor technology prior to the damage tests. This was done using an electrodynamic shaker for dynamic excitation from another company involved in the project, which was operated by means of remote monitoring and using standard offshore SHM sensor technology for the monopile structure. To teach the SHM system, the company performed SHM reference measurements under stochastic excitation prior to the load tests and repeated them after each of the total of eight load levels to evaluate the damage progress at the respective stage. In cooperation with the TTH, 1,000 load cycles were then applied per load level and supplemented by defined load profiles (e.g., low-frequency sine wave, linear ramps, and constant loads) before and after every load level. This ensured there was a wealth of data available for successful verification of the measuring technology (Kohlmeier et al., 2024). The results showed that the FBG sensors worked exactly as intended. Due to their high sensitivity, the FBG sensors distributed throughout the grout annulus (see Fig. 2 (left)) reflected the load situation clearly even at low deflections of the tower. Accordingly, it proved possible to identify the initiation of cracks in load level 2 and track their subsequent development. Fig. 2 (top right) shows the damage progress in the form of increasing, non-linear behavior. It was also possible to derive valuable information about the local distribution of the damage process based on the positioning of the sensors between the upper, central, and lower shear key pairs.

Fig. 2: Positioning of the eight FBG sensors with color coding (left) and time series of the FBG strain gauging in the form of wavelength shifts in load level 4 (top right). Image of the 50 mm FBG sensor (bottom right) © Fraunhofer IWES

Numerical simulation environment as a virtual experiment

A 3D finite element model for numerical simulation was created with the help of the Support Structures department’s virtual test bench as an aid for use when designing the physical model and transferring experimental results to real scale. The data acquired in the experiment formed a sound basis for the subsequent validation of the material models employed as part of a continuous optimization process. The following points represent an established approach:

  • All decisive geometry variables and their dependencies are compiled as a parameter set with the help of a CAD-based test design.
  • The model setup is parameterized and implemented automatically by means of object-oriented programming in Python utilizing the program interface in the finite element program Abaqus (see Fig. 3 (left)).
  • The levels of detail required for the analysis of different issues are implemented step by step using the flexible model structure.
  • The validated simulation model can be transferred to existing or new types of structures in the offshore wind farm with manageable effort.

Once the characteristic damage behavior of a component has been identified in the physical model test, numerical models can be implemented to analyze this characteristic behavior (see Fig. 3 (right)). The data derived from the physical model test are then employed to validate the numerical material models used. Detailed explanations of the simulation results can be found in Kohlmeier et al. (2024) and Terbach (2024).

Fig. 3: Finite element model with representation of the boundary and contact conditions (left) and development of the pressure damage areas across the selected load levels (right) (Kohlmeier et al., 2024) © Fraunhofer IWES

When real geometric parameters are used, the numerical modeling system presented can be simply scaled to the typical size of the support structure in the wind farm and adapted for the respective type of turbine. Damage scenarios can be derived taking the findings gained from the physical experiment into consideration. The following applications are possible:

  • For monitoring of the support structure, the support structure characteristics significant for an SHM system can be established and used to quantify threshold values for the triggering of warning levels.
  • Numerical damage models can be utilized as the basis for generating data sets that can be used in the context of artificial intelligence, for example for the supervised teaching of a monitoring system. These could be developed as prototypes and adapted to suit different systems.

Findings from the Grout-WATCH project

The sensor technology developed is very well suited in the physical model test, see Fig. 4, to documenting the damage processes (which it was not possible to measure objectively in the past) including at a local level in the form of non-linear strain increases. Improved understanding of the temporal and local damage progression can be derived at the respective sensor positions from this. The following development approaches for future work thus present themselves:

  1. Correlation of the measurable variables with the current grout condition and integration of the findings into the monitoring process.
  2. Characteristic findings on the failure process of grouted connections can improve the interpretation of offshore measurement data.
  3. Application of numerical models to transfer the knowledge gained to the real support structure in the offshore wind farm.
  4. Development of an approach for the integration of the sensor technology developed into the installation process in the offshore wind farm.

In this way, the Grout-WATCH project has contributed to better understanding of grout failure in grouted connections subject to high loads. The results can be utilized to improve the validation of numerical simulation models for high-strength grouts, predict the service life of grouted connections in offshore wind turbines more accurately, and prevent possible damage.

Fig. 4: Video of the last step of the load test: loaded support structure by means of a hydraulic actuator (left), displacement measurement by laser distance sensor at the upper flange of the transition piece (top right) and view into the viewing window under the grouted joint: broken grout material in the main loading direction (bottom right) © Fraunhofer IWES

Detailed results of the project can also be found in the following publications:

Kohlmeier, M., Heinrich, D., Spill, S., Collmann, M., Dreger, D. & Schossig, T. (2023). Design of a Large-scale Model Test to Validate Monitoring Systems for Grouted Connections in Offshore Monopile Foundations. In J. S. Chung (Ed.), The proceedings of the Thirty-Third (2023) International Ocean and Polar Engineering Conference: ISOPE-2023 : Ottawa, Canada, June 19-23, 2023 (pp. 920–926). International Society of Offshore and Polar Engineers (ISOPE).

Nuber, A., Borgelt, J., Collmann, M., Dreger, D., Friedmann, H., Kohlmeier, M., Schossig, T., Römgens, N., Tsiapoki, S., Wernitz, S. (2023). Grout-WATCH – Untersuchung des Tragverhaltens von Offshore-Grout-Verbindungen unter Wasser an Tragstrukturen mit dynamischen Wechselwirkungen. In D. Janecek (Ed.), Schriftenreihe Projektträger Jülich: Vol. 17. Statustagung Maritime Technologien: Tagungsband der Statustagung 2023 (pp. 151–168). Jülich: Forschungszentrum Jülich GmbH Zentralbibliothek Verlag. Available at: https://edocs.tib.eu/files/e01fn24/1878626833.pdf

Kohlmeier, M., Heinrich, D., Spill, S., Collmann, M., Dreger, D., Schossig, T. (2024). Results of Initial Large-scale Model Tests to Validate Monitoring Systems for Grouted Connections of Offshore Monopiles. International Journal of Offshore and Polar Engineering 34(03): 322–331 (https://doi.org/10.17736/ijope.2024.cl26).

Terbach, J. (2024). Weiterentwicklung eines Finite-Elemente-Modells zur Abbildung des Tragverhaltens einer Grout-Verbindung unter Biegebeanspruchung in experimentellen Untersuchungen, master thesis, Institut für Baustoffe, Leibniz Universität Hannover, Hannover.

Joint project Grout-WATCH: https://www.iwes.fraunhofer.de/en/research-projects/finished-projects-2023/grout-watch.html

More information here:

Leave a Reply

Your email address will not be published. Required fields are marked *