Analysis and modeling of unsteady aerodynamics with application to wind turbine blade vibration at standstill conditions
Since the introduction of computational models for the dynamic aeroelastic response of wind turbines, numerous dynamic issues have been investigated to assist in understanding and overcoming problems as well as help designing reliable turbines. In some cases specific problems have necessitated evolution of new or improved sub-models in the aeroelastic computational tools, such as for instance the so called dynamic stall models that models the dynamic aerodynamic response from the onset of separation until the flow over the airfoil is fully separated.
Applying the existing aeroelastic tools to investigate the dynamic response for wind turbines in standstill have indicated that under some conditions, the edgewise vibrational mode of the wind turbine blades are negatively damped, leading to very big fatigue loads or even failure. Since the unstable regions occur for angles of attack far from normal operation, in deep stall, where the underlying aerodynamic coefficients are uncertain, it was initially assumed that the problem existed only in computations. In recent years, however, turbine failures at standstill conditions reported from the industry spurred new analytic and numeric investigations.
The phenomenon is observed on wind turbines in standstill when both yaw and pitch regulation are offline. This situation where all forms of control are absent mostly prevail in connection with disconnection from the electrical grid, for instance during the erection phase before the wind turbines are connected to the electrical grid. In absence of control of the wind turbine, the angles of attack of the flow on the blades are determined by the direction of the free wind, and are therefore not only confined to the narrow normal operational range where the unsteady aerodynamic response is well understood and the aerodynamic coefficients that the tools use are well determined. The investigations in, concluded that the standard aerodynamics existing in aeroelastic codes for these cases is effectively quasi-steady, and that the aerodynamic damping can be either augmented or removed by adjusting the steady aerodynamic coefficients within the limits of uncertainty of the underlying aerodynamic data. On the other hand it is not known what mechanisms govern the mean response of a wing in deep stall, what influence the self-generated stochastic forcing in deep stall has or how far the quasi-steady approximation is from what occurs in real life.
PhD Project:
Even though the previous research has shed some light on the standstill issue, a more detailed investigation is required to determine the mean mechanisms in deep stall and to produce an engineering model applicable for aeroelastic codes and blade design methods. To do so both the basic mechanisms in deep stall unsteady aerodynamics as well as stationary force coefficients must be investigated.
In order to uncover the basic mean force coefficients and unsteady mechanisms prevailing in the deep stall case it is envisaged to start out investigating three different existing resources: results from existing 3D CFD computations [ ], analysis of detailed 3D measurements on turbines in wind tunnels and analysis of measurements on a heavily instrumented turbine (surface pressures, accelerometers and strain gauge measurements).
An investigation of whether the effect of the self induced stochastic forcing on the loading in the deep-stall region is significant, the stochastic model developed by Bertagnolio et. al. [ ] will be used in the standstill case by implementing it in the aeroelastic computational tool HAWC2.
After the initial investigations, the specific details pinpointed will be investigated using a combination of 3D CFD [v] and possibly also steady and unsteady test in wind tunnels. One problem with airfoil section measurements in the deep stall case is tunnel blockage issues which can greatly influence the results. However, since the problem is recognized by both the scientific society and the industry it is likely that it will be possible to apply for a research project for funding, establish cooperation with industry or other research institution with access to test facilities of appropriate size and properties. No matter what method of analysis of the phenomenon will be chosen, focus will be on the angles of attack associated with the negatively damped aerodynamic forces.
Possible ways of modeling the unsteady aerodynamics in deep stall could be the computationally very efficient indicial function method, as used in some dynamic stall models [ii] or for instance by modeling the aerodynamic response using a set of differential equations. This cannot be determined prior to pinpointing the mechanisms and physics responsible for the negative damping.
As mentioned earlier one of the main goals of the present work is the development of engineering models applicable to the standstill problem. Once a model for the unsteady mechanism exists, suggestions for avoiding the standstill problems can be formulated and numerically tested in the aeroelastic computational tool HAWC2. Further it will be possible to investigate to what extent the problem can be avoided by changing the aerodynamics or by other simple engineering means.
Further, it should be noted that there is currently an ongoing PhD project at Risø DTU aiming at solving the problem. This project treats the problem from a different perspective, aiming at damping the vibrations by means of a specific composite setting, thus treating to some extent the symptoms of the edgewise vibration problem. The present project aims at understanding and modeling the cause. Results from both projects combined will provide an improved basis for avoiding the standstill problems.