When compared to open-loop configuration, full-scale wind turbine nacelle testing with Hardware-In-the-Loop (HIL) configuration allows for coupled electro-mechanical as well as full operational certification tests with the native nacelle controllers. This configuration requires a full turbine real-time simulation running in parallel to the nacelle under test. In this study, a baseline simulation model is used to investigate the nacelle fidelity necessary to capture dynamic characteristics of interest while meeting the real-time requirements. The same model is also utilized to understand the influence of different boundary conditions seen by the nacelle when mounted on a test bench without a rotor, tower, and platform. The results show that the torsional dynamics are mainly governed by the flexibility of the main shaft and the gearbox supports. It is also demonstrated that the abstraction of the nacelle leads to a torsional frequency shift and higher frequency content in component responses necessitating compensation techniques for proper implementation of HIL testing.

Ground testing of full-scale wind turbine nacelles has emerged as a highly favorable alternative to field testing of prototypes for design validation. Currently, there are several wind turbine nacelle test facilities with capabilities to perform repeated and accelerated testing of integrated turbine components under loads that the machine would experience during its nominal lifetime. To perform accurate and efficient testing, it is of significant interest to understand the interaction between coupled test rig/dynamometer and nacelle components, particularly when applying extreme loads. This paper presents a multi-body simulation model that is aimed at understanding the responses of a coupled test rig and nacelle system during specific tests. The validity of the model is demonstrated by comparing quasi-static and dynamic simulation responses of key components with experimental data obtained on an actual 7.5 MW test rig. A case study is conducted to analyze a transient grid-loss event; a Low Voltage Ride Through (LVRT) test on the dynamometer and drivetrain components. It is shown that the model provides an efficient way to predict responses of the coupled system during transient/dynamic tests before actual implementation. Recommendations for mitigating the impact of such tests on the test bench drive components are provided. Additionally, observations of differences between transient events in the field and ground based testing are made.

We investigate the potential of harvesting vibration energy via a bi-stable beam subjected to subharmonic parametric excitations. The vibrating structure is a buckled beam with two stable equilibria separated by a potential barrier. The beam is subjected to a superposition of a static axial load beyond its critical buckling load and a harmonic axial excitation which frequency is around twice the frequency of the first vibration mode. A micro-fiber composite (MFC) is attached to one side of the beam to convert the strain energy resulting from the beams oscillation into electricity. The study considers two regimes of excitations: an amplitude sweep and a frequency sweep. In the first regime, the amplitude of excitation is varied while the excitation frequency is tuned at twice the natural frequency of the first vibration mode. In the second regime, the excitation frequency is swept forward and backward around the subharmonic resonant frequency while the amplitude of excitation is kept constant. A theoretical model which governs the electromechanical coupling of the transverse vibrations of the beam and the output voltage is used to monitor the response as the excitation parameters are changed. An experimental setup is built and a series of tests is performed. The theoretical results are in good agreement with their experimental counterparts. The experiment also shows that this type of bi-stable energy harvesters exhibits a broadband frequency response as compared to the classical linear harvesters.

This paper aims to investigate the response and characterize the effective frequency bandwidth of tri-stable vibratory energy harvesters. To achieve this goal, the method of multiple scales is utilized to construct analytical solutions describing the amplitude and stability of the intra- and inter-well dynamics of the harvester. Using these solutions, critical bifurcations in the parameter’s space are identified and used to define an effective frequency bandwidth of the harvester. A piezoelectric tri-stable energy harvester consisting of a uni-morph cantilever beam is considered. Stiffness nonlinearities are introduced into the harvesters design by applying a static magnetic field near the tip of the beam. Experimental studies performed on the harvester are presented to validate some of the theoretical findings.

This article exploits nonlinear modal interactions, namely, a two-to-one internal resonance energy pump to improve the steady-state bandwidth of vibratory energy harvesters. To demonstrate the enhanced performance, an L-shaped cantilevered structure laminated with a piezoelectric patch and augmented with frequency tuning magnets is considered. The magnets serve to adjust the first two modal frequencies of the structure such that the second modal frequency is twice that of the first mode activating a nonlinear energy transfer mechanism. The structure is then subjected to a harmonic excitation with a frequency close to the first modal frequency of the structure. The voltage frequency-response curves are generated for different excitation levels, illustrating an improved bandwidth and output voltage over a case involving no nonlinearly-interacting modes.

Performance characteristics of the giant magnetostrictive alloy, Terfenol-D, have been studied by many researchers for actuation, sensing and energy harvesting applications. Mathematical models characterizing the magneto-elastic behavior and describing the effects of bias conditions — compressive prestress and magnetic bias — on the material performance, have been developed. For the most part, the models used to describe the material are linear models that can hide essential features of the dynamic performance. While nonlinear constitutive models of Terfenol-D exist, such models have not been utilized to study the dynamic frequency response characteristics that are essential towards a comprehensive understanding of its performance in actuation, sensing or energy harvesting. To address this problem, this effort investigates the role of empirically determined material nonlinearities in the dynamic performance of Terfenol-D. Towards that objective, a polynomial type stress-strain relation is used to construct a nonlinear distributed-parameters model for a Terfenol-D rod fixed at one end and mass loaded at the other while being subjected to a sinusoidal base excitation. Additionally, the model accounts for the rod being subjected to an axial prestress prior to excitation. Using the method of multiple scales, the nonlinear frequency response of the rod is investigated by obtaining analytical expressions for the steady-state response amplitude. It is demonstrated that the axial prestress results in a shift in the fundamental vibration frequencies of the rod and a change in the effective nonlinearity of the system. A qualitative analysis of the solution reveals that, the magnitude of axial load can be used to maximize the response amplitude over a larger bandwidth of frequencies.

Metrology systems take coordinate information directly from the surface of a manufactured part and generate millions of (X, Y, Z) data points. The inspection process often involves fitting analytic primitives such as sphere, cone, torus, cylinder and plane to these points which represent an object with the corresponding shape. Typically, a least squares fit of the parameters of the shape to the point set is performed. The least squares fit attempts to minimize the sum of the squares of the distances between the points and the primitive. The objective function however, cannot be solved in the closed form and numerical minimization techniques are required to obtain the solution. These techniques as applied to primitive fitting entail iteratively solving large systems of linear equations generally involving arithmetic intensive operations. The current problem in-process metrology faces is the large computational time for the analysis of these millions of streaming data points. This paper presents a framework to address the bottleneck using a Graphical Processing Unit (GPU), to optimize operations and obtain significant gain in computation time.