Hybrid Simulation of Wind Turbine Blades

Wind turbines are progressively used as a substitute to fossil fuels enhancing the demand for larger and more energy efficient wind turbine blades. These wind turbine blades are typically made from composite materials among those glass and carbon fibre reinforced plastics along with lightweight cores.

Hence, the ambition to improve the structural and operational performance of the wind turbine blade has resulted in extensive research within large composite structures. In these efforts testing has primary been focusing on two length scales including laminate and structural scale testing. However, to reveal the structural response of the wind turbine blade during service, experimental testing covering length scales from micro through structural scale testing is required.

To establish a link between laminate and structural scale testing within the industry of wind energy a Hybrid Simulation (HS) technique is implemented which facilitates substructural scale testing. Structural assessment through HS is a substructural technique where the behaviour of emulated structure is revealed by combining the advantages of numerical modelling with those of experimental testing.

The coupling governed through the interface between the numerical and experimental substructure – referred to here as the shared boundary – is achieved by maintaining compatibility and equilibrium at the interface. During the test, a predefined external load is applied the numerical substructure and the corresponding response computed.

Through a communication loop, the displacement at the shared boundary is induced on the experimental substructure through an e.g. Proportional Integral Derivative (PID) regulated servo-hydraulic actuator – referred to here as the transfer system. The forces required to deform the experimental substructure – referred to here as the reaction force – are fed back to the numerical substructure to reveal the response of the emulated structure.

The experimental and numerical substructure, communication loop and transfer system combine to form the HS. The research within HS has to date expanded upon numerous branches including civil and mechanical engineering – referred to here as conventional HS. Common to conventional HS is that the shared boundary is defined by a discrete point operated within a few Degree-Of-Freedom (dof)s.

This configuration has become a mature and reliable approach however; it imposes some limitations in the effort of spreading the HS technique within new application areas including large composite structures. Therefore, a new generation of HS is presented capable of handling a shared boundary covering a continuous edge or plane – referred to here as single-component HS.

The implementation of single-component HS induces some distinctive challenges in the experimental substructure including compliance in the transfer system driven by slack and deformations in the load train and boundary introduction zone along with inertia effects induced by the mass of the load train and boundary introduction zone.

These errors governs a significant impact on the accuracy and stability within single-component HS, hence two compensators are introduced named high precision tracking compensator and inertia compensator. The high precision compensator is capable of reducing the discrepancy between the desired and achieved displacement by tracking the shared boundary through an external Data Acquisition (DAQ) system using i.a.

Digital Image Correlation (DIC). The compensator proved successful in both the Quasi-Static (QS) and Real-Time (RT) regime. The inertia compensator revealed sound performances in erasing the majority of the inertia effects induced by the mass of the load train and load introduction zone in the RT regime.

A communication loop capable of accommodating single-component HS in the QS and Pseudo-Dynamic (PsD) regime is designed and implemented in the Laboratory Engineering Workshop (LabVIEW). Here the numerical substructure, transfer system capable of operating the experimental substructure on an extended time scale along with relevant interface compensators are operated sequentially in a state-machine framework.

This configuration provides a simple and flexible multi-processing platform, which is easy to extend and modify throughout the design phase. The system architecture is successfully verified through a single-component and conventional HS application. To reveal the inherent dynamics of the experimental substructure a Real-Time Hybrid Simulation (RTHS) communication loop capable of accommodating single-component applications is designed and implemented in LabVIEW.

To attain a continuous time history of displacement, velocity or acceleration at the shared boundary an operation rate that is 10-25 times faster than the mode of interest is required. Given the enhanced complexity of the numerical model within single-component HS, an integration time equivalent to the required operation rate of the experimental substructure, can be difficult to attain.

Hence, a multi-rate Real-Time Hybrid Simulation (mrRTHS) approach is implemented capable of operating the numerical and experimental substructure at two different rates while including rate transitioning to link the substructure appropriately. Here the numerical substructure, transfer system capable of operating the experimental substructure with RT constraints along with relevant compensators is operated in parallel across multiple threads.

Implemented on a RT-target which provides reduced latency and tight jitter tolerances the system architecture is successfully verified through a single-component and conventional HS application. A representative experimental substructure of an SSP34m wind turbine blade is identified through a numerical analysis for evaluating the increase of stresses in the leading edge governed by the cross section of the blade being distorted in transverse shear.

Here an 8m root section of the wind turbine blade is identified as a representative substructure, capable of physically replicating the cross sectional shear distortion. Furthermore, a boundary introduction zone of 6m is added to erase the distortion induced by the load train, entailing that the entire experimental substructure covers the inner 14m root section of the wind turbine blade.

A fatigue rated multi-axial test setup is designed to accommodate the inner 14m inner root section of the wind turbine blade. Finally an initial HS architecture and strategy is presented to form the basis for an upcoming single-component HS on the SSP34m wind turbine blade. Altogether, this PhD thesis presents a single-component HS approach, which aims to form an important milestone in the effort of extending the application portfolio within HS for structural assessment of large composite structures.

Two compensation techniques were designed capable of enhancing the accuracy and stability within singlecomponent HS. A communication loop capable of accommodating single-component HS were designed and implemented in LabVIEW. The system proved successful within the QS and RT regime for the operation of a shared boundary including a discrete point with up to three dofs.

The presented work is based upon seven appended papers along with related research activities, which were not possible to convey through scientific publications.