Wind Turbine Aeroelasticity introduction
Aeroelasticity is a complex field of expertise, in wind energy the field of wind turbine aeroelasticity it is a relatively new field of expertise with only a hand full of people that really understand this field. On this page of our website we will try to give a short introduction to this field. There is a short general introduction and there are some mini-lectures that can be downloaded which provide a more detailed explanation of some (interesting) phenomena.
If you would like to find out if you already know something about wind turbine aeroelasticity, you may enjoy looking at our quiz (see the links at the top or bottom of the page to open the quiz).
On this page we will first provide a short introduction to wind turbine aeroelasticity, followed by our mini-lectures. These mini-lectures address an aeroelastic phenomenon that is relevant for wind turbines and explains what, why and how. We hope you will enjoy these. Then there are some suggestions for further reading material for those that are interested in finding out more about wind turbine aeroelasticity. Finally some open source tools that can be useful are shortly discussed. If you feel we can further improve this site with other subjects, please feel free to let us know!
Short introduction to wind turbine aeroelasticity
A first distinction that has to be made is the distinction between resonance and instabilities. A resonance occurs when there is an excitation frequency that is close to some natural frequency of the system. Every wind turbine has infinitely many natural frequencies corresponding to infinitely many modes. Only the modes at the lowest frequencies for the different components are relevant as the higher modes have a lot of structural damping resolving any possible issues with resonance or instabilities.
On a complete turbine there will of course be interaction between the different components, so a blade mode will show up on a complete turbine as a combination of all blades deforming combined with some tower deformation. So if there is some external force at a frequency close to a natural frequency of the turbine, there can be resonance. From linear vibration theory it is known that this can lead to an infinite amplitude if the frequencies are equal and there is no damping (figure 1), to smaller amplification factors for cases where there is some damping and/or some difference between the excitation frequency and the natural frequency (see figure 2).
An excitation frequency can come from the rotational speed (tower shadow, rotational sampling of the turbulence,…) or perhaps from pitching actions or electronics. In calculations one has a lot more possibilities in creating an excitation frequency, for example one could have a sinusoidal wind speed. On a real turbine the excitation is usually related to the RPM and its multiples and possibly from the controls.
An aeroelastic instability is not caused by any excitation frequency. In case of an aeroelastic instability the shape of a natural mode at the corresponding natural frequency adds energy to the vibration from the aerodynamics. In one cycle the mode shape changes the aerodynamic forces in such a way that energy is extracted from the air to increase the vibration of the structure. So for instabilities the shape is the most important property, and usually changing the model such that this shape changes is the solution to the instability. This also illustrates why focusing purely on the frequencies and avoiding resonance is not enough to provide an aeroelastically sound design, one needs to ensure that all natural modes on the turbine are well damped. Such that some sudden change in for example the wind speed, results in a vibration that is quickly damped, as illustrated in figure 3.
Note that the amplitude of a vibration that is damped follows an exponential function, in case the vibration is linear. In the same way an instability will have increasing amplitudes that follow an exponential function, while a resonance for an undamped mode will be linearly increasing, as illustrated earlier.
Many website pages could be filled with all the knowledge available on wind turbine aeroelasticity, so this short introduction stops here for now. Explaining what whirling modes are, why the change in frequency occurs when looking at stand still frames of reference is explained in the mini lecture that can be downloaded below. Also a short description concerning classical flutter is provided in another mini-lecture.
Other important knowledge for wind turbine designers would include at least: which modes have the least damping, which instabilities we know of, how to perform a classical flutter speed calculation and how one can increase the classical flutter speed and so on and so on. However the experience is that it takes more than simply reading the material to get to grips with this complex subject. A more active approach which includes discussion and assignments, is much more effective. And time, to let it all sink in. One option is to follow our training in the field of wind turbine aeroelasticity.
For more information on wind turbine aeroelasticity, please refer the list at the bottom of the page.
Also we have some mini-lectures provided below.
MINI LECTURES
In this mini-lecture we explain the frequency shift from rotating frame to non-rotating frame: why, when and how.
Our second mini-lecture, this provides more information on classical flutter.
Why edgewise frequency at 6P could perhaps be a good idea…
Reading material
- Explanation of wind turbine modes
- Modal dynamics of structures with bladed isotropic rotors and its complexity for two-bladed rotors Morten Hansen Wind Energy Science 1-2016
- Modal properties and stability of bend–twist coupled wind turbine blades Alexander Stäblein et al. Wind Energy Science 2-2017
- PhD thesis Aeroelastic modal dynamics of wind turbines including anisotropic effects, Peter Fisker Skjoldan, 2011
- Multiblade Coordinate Transformation and Its Application to Wind Turbine Analysis G. Bir, 2008 – NREL
- General introduction to WT aeroelasticity
- PhD thesis Aeroelasticity of Large Wind Turbines Jessica Holierhoek, 2008
- A chapter Aeroelastic Stability Models in the book Handbook of Wind Energy Aerodynamics (not open access).
- Instabilities
- Aeroelastic Instabilities of Large Offshore and Onshore Wind Turbines, G. Bir and J. Jonkman 2007 – NREL
- Idling instabilities:
- Aeroelastic stability of idling wind turbines Kai Wang et al. Wind Energy Science 2-2017
- Stability analysis of parked wind turbine blades using a vortex model Vasilis Riziotis et al.
- Overspeed instabilities (classical flutter or edgewise):
- Field validation of the Stability Limit of a Multi MW turbine Bjarne S. Kallesøe, The science of making torque from wind 2016
- Flutter behavior of highly flexible blades for two- and three-bladed wind turbines Mayank Chetan, Shulong Yao, and D. Todd Griffith Wind Energy Science 7-2022
- Journals
- Wind Energy Science: a very good open access journal
- Wind Energy – Wiley: a very good journal with more and more open access articles being published
Open Source Tools – Free Tools
This part of the page is still under construction….
As a small company, we have to rely on free tools for most of our activities. Of course we also often use licensed tools, usually from the client, but for research activities and training purposes, we also require open source tools. Over the years we have found that there are many different open source tools that we use in the process of our analyses, therefore we would like to specify some of them here for those that are interested.
First there are tools that can be used for full aeroelastic simulations. OPEN FAST, from NREL, is a good tool to perform aeroelastic simulations of wind turbines. For current size and larger wind turbines, one should really use the beamdyn model for the structural blade modelling, as the nonlinearities become too relevant to use the original linearised approach.
QBlade has made great progress the last few years and is probably more user friendly than OPEN FAST. The aerodynamics in QBlade include more advanced model options. However, the license is not for free use in commercial settings.
For post-processing and graphing some of the data, most tools come with some options for this, but in general it can als be necessary to do some of your own post-processing. Note that our speciality is aeroelastic issues, so we usually do very different post-processing than that which is done during the IEC and/or DiBT load set calculations. For example, we want to be able to plot different loads or deformations against time or against each other and often after band-pass filtering. Therefore a tool such as Python becomes extremely useful, but also R and Octave are freely available tools that can help you with these kinds of processes.
When plotting Bladed data and also some other wind turbine aeroelastic simulation tools, there is a nice open source option called pyDatView. As the name already suggests, it is a Python code.