The trunk and upper arm motion might cause the flow structure difference of the wake. Two distinct vortexes generated by the arms produced a downwash flow in the wake of jumper A. It is considered that the positioning of the arms in a very low position strongly influences the flow structure. These results suggested that the vortexes generated by the arms seem to be very important for the aerodynamic lift generation.
9th Applied Aerodynamics Conference | Fluid Dynamics and Co-located Conferences
It is shown that detailed aerodynamic measurements are an absolute necessity for the validation and improvement of wind turbine design codes, where it is also concluded that the amount of measurement data which has been produced until now is still far too limited. Smoke visualization of tip vortices in the New Mexico experiment. The present article is devoted to detailed aerodynamic measurements on wind turbines. Such measurements are important for a better understanding of the aerodynamic wind turbine behavior and the subsequent improvement of aerodynamic models in wind turbine design codes.
As such every aerodynamic model inherently suffers from simplifications. For wind turbine aerodynamics, an additional difficulty arises due to the fact that the computational effort for design calculations is more extreme than it is for most other applications e. This necessitates the use of very efficient, but also very simplified aerodynamic models i.
Peer-reviewed journal papers
Obviously more advanced methods like computational fluid dynamic CFD codes are applied in the wind energy society, but their use is, due to the calculational demands, restricted to specific studies and load cases. From a practical point of view these simplifications inevitably go together with a large uncertainty band, see for example, Schepers et al. Alternatively the loads may be lower than expected which implies an over dimensioned and costly design.
From a more fundamental point of view, the simplifications in wind turbine aerodynamic models can obscure physical phenomena crucial to wind turbine operation, which may remain hidden and unnoticed in the results obtained with these models. Moreover, the load instabilities or extreme load excursions may occur due to the machine operating state, while others are produced by inflow turbulence or extreme events. Unfortunately, it is generally not possible to discern whether the origin lies with the machine itself or the inflow. The common denominator in these rotor aerodynamic experiments lies in the measurement of the pressure distributions at different locations along the rotor blades.
In these experiments, sometimes the underlying flow field around the wind turbine is measured too. Such dedicated aerodynamic experiments are carried out since the end of the s and many of them were brought into four subsequent IEA Tasks 14, 18, 20, and 29 see Boorsma et al. The present review relies largely on results and analyses from these IEA tasks. Thereto, Section 2 describes the content and the historical context of the IEA tasks. Section 3 elaborates on measurement uncertainties, including those due to inflow spatial and temporal availability.
It also explains the difficulty in the determination of angle of attack and dynamic pressure in wind turbine measurements, which to some extent can be seen as a measurement uncertainty too. Then, Section 4 selects a few analyses on aerodynamic measurements and identifies research areas on which the present measurements provided important information. It must be noted, however, that a huge amount of relevant results have been generated over the last decades from which a small selection can be presented only.
The chapter includes a long list of citations.
It is emphasized that the present article limits itself to the application of aerodynamic measurements for validation and model improvement purposes. Obviously, measurements could be used for the demonstration of aerodynamic innovations as well one can think of the demonstration of new airfoils or flow devices like vortex generators.
The aerodynamic modeling of these innovations does fall within the context of this article but the demonstration of them with, for example, practical recommendations on how to apply these flow devices is excluded. Obviously, there is a strong relationship between the 2D and 3D measurements. Among others, the 2D measurements are used to compare with the 3D measurements in order to extract the rotational effects.
Hence, in the validation of lifting line codes with 3D measurements, these 2D measurements play an important role in the interpretation of results. The article also limits itself to public experiments. Although some results of this experiment have been published in the works of Madsen et al. The present article also excludes wind farm aerodynamic measurements as described in the studies by Barthelmie et al. Such measurements are generally taken in the far wake several rotor diameters behind a wind turbine where one might expect the next wind turbine in a wind farm.
Still the distinction between wind farm and rotor aerodynamics is not always very pronounced in view of the fact that the near wake which is seen as part of rotor aerodynamics and considered in this article forms the starting point for the far wake. Hence, this experiment is important not only for understanding the aerodynamics of an actuator disc i.
Acoustics is also considered to be outside the scope of this article, even though it is well known that wind turbine noise is mainly aerodynamically driven Schepers et al. The first detailed aerodynamic measurement programs, as described in this article, were performed at the end of the s.
Until that time all experimental programs measured global loads only, that is, total blade or rotor loads. Differences between calculations and measurements were found but the validations could not explain the physical causes of these discrepancies. This was only possible from more detailed measurements of aerodynamic properties along the blade as carried out in the wind tunnel or in the field.
As a result several institutes initiated programs to measure the pressure distributions at different radial positions. Thereto Figure 1 shows how pressure taps around the airfoil, which are drilled into the surface, are connected with tubes to pressure sensors. The pressure scanners measure differential pressures relative to a reference pressure.
This reference port was generally connected to the hub of the rotor where the pressure is assumed to be close to the known atmospheric pressure, but Section 3. Note that in the later Mexico experiment, individual absolute pressure sensors were mounted on the blade surface eliminating the uncertainty from the unknown reference pressure. The resulting pressure distributions are then integrated to sectional forces, that is, the tangential force along the chord and the force normal to it from which the lift and drag forces can be derived, although this requires the angle of attack, which as explained in section 3.
The measurements of local aerodynamic loads are often complemented with probe measurements of inflow velocity and angles ahead of the blade section, measurements of loads e. Some wind tunnel experiments also map the detailed flow field upstream, in and downstream of the rotor plane which is important information on induction aerodynamics which drives the rotor aerodynamics.
The various facilities from the participants and their instrumentation are described in detail in Schepers et al.
- Aerodynamics of the Wing/Fuselage Junction at an Transport Aircraft in High-Lift Configuration?
- CCSP SNRS Exam Certification Guide.
- National Technical Reports Library?
- Viral Hepatitis.
- ABOUT FMFP.
- The Performance Consultants Fieldbook: Tools and Techniques for Improving Organizations and People (Essential Knowledge Resource)?
The difference between the phases lied in the blade design untapered and untwisted for Phase II, untapered and twisted for Phase II and IV, and in the instrumentation to measure the blade inflow angle either a wind vane or a five hole probe, see Figure 2. Angles of attack were set either by the wind speed, the pitch angle, or the rotor speed rpm dependent on the experiment. Apart from time series, measurements of airfoil coefficients 2D values as well as rotating values as functions of angle of attack were added to the database. This was due to the unsteady inhomogeneous and uncontrolled wind conditions.
The free stream conditions were usually measured with a mast some distance away from the turbine not necessarily upstream of the turbine where this mast was instrumented with only a few anemometers. In the best case, this yields a reasonable estimate for the mean wind conditions at the rotor plane but it anyhow does not provide the instantaneous inflow conditions at every position in the rotor plane which is needed to study high frequency, local, effects like dynamic stall.
In this experiment a heavily instrumented rotor with a diameter of 10 m was placed in the world's largest wind tunnel, that is, the NASA Ames As such measurements could be performed at stationary, known and homogeneous conditions which facilitated the interpretation of measurement results and detection of subtle aerodynamic phenomena extending down to small, briefly lived structures and events.
This results in a blade Reynolds number in the order of 1 million. Nine institutes from eight different countries participated.
Most of the analyses in IEA Task 20 were focused on measurements for the upwind rotor with fixed hub and without transition strips. It is noted that within IEA Task 29 Mexnext additional Phase VI measurements were analyzed at higher rpm, in order to investigate elevated rotor loading conditions. The Mexnext consortium differed slightly from phase to phase but generally some 20 parties from 10 or 11 countries were active in Mexnext. The rotational speed was either At This tip speed is relatively high to increase Reynolds number but sufficiently low to prevent noticeable compressible conditions.
Measurements were done at different tunnel speeds, tip speed ratios, and different yaw angles and pitch angles. The setup of the Mexico experiment is given in Figure 6 and described in Schepers and Snel and Boorsma and Schepers In addition to pressure and load measurements, the 3D flow field was mapped using stereo PIV where two cameras are mounted on a traversing tower which can move in the horizontal streamwise and radial direction this tower can be seen in Figure 6 at the right.
As such a comparison can be made between calculated and measured loads where the underlying flow field which drives these loads can be assessed too. The test matrix of New Mexico was designed to explain the nonunderstood phenomena from the Mexico experiment taking into account the lessons learned from the Mexico experiment. In the LST the aerodynamic characteristics of the blades at standstill have been measured including flow visualization where the blade instrumentation and data acquisition could also be tested and recalibrated.
The New Mexico experiment then partly repeated its predecessor experiment but an even wider variety of conditions in terms of wind speed, yaw, etc, were considered where advantage was taken of the improved measurement techniques which meanwhile became available e. Moreover, several blade and wake flow visualization techniques were applied. The improved quality of the New Mexico experiment compared to the Mexico experiment answered most of the nonunderstood questions from the first experiment and the new results obeyed the momentum balance between loads and velocities.
Although the focus of Mexnext lies on the New Mexico experiment, the project is overarching, that is, it aims to consider all public aerodynamic measurements. Moreover, some other experiments are considered of which the results are not always fully explored yet.
The role of tip leakage vortex breakdown in compressor rotor aerodynamics
Among other the following experiments were found of relevance. An important item in measurement programs is the identification of measurement uncertainties. These are usually assessed on visual inspection, consistency checks, checks on reproducibility, checks of measurement sensor specifications, etc.
In general the uncertainties introduced by the measurement equipment itself in particular the pressure sensors were found to be limited at least if frequent calibrations are carried out. An exception is formed by measurements at standstill or measurements at the inner part of the blade where pressure levels were sometimes within the lower part of the measurement range. One of the items investigated were the damping characteristics of the tubes between the pressure taps and scanners. Most of the results used in the IEA Tasks are averaged over a number of samples acquired during several consecutive rotor revolutions.
sesarilpre.tk In general, the standard deviation for the wind tunnel measurements was limited in the order of the plotting symbol size! In this respect a difference should be made between closed and open tunnels. Tunnel effects in a closed tunnels are relatively easy to determine but they are also known to be more severe than tunnel effects in an open tunnel Barlow, Rae Jr.
On the other hand, corrections for open tunnels are more difficult to determine due to the fact that the tunnel flow will interact with the outer flow through a turbulent process. Such turbulent process is inherently subject to a large uncertainty e.
Other uncertainties may be introduced from the processing of data, for example, the integration of pressure distributions into normal and tangential forces. Moreover, the use of processed data like angle of attack and dimensionless airfoil characteristics require careful interpretation as will be explained in Section 3. It is important to recall the difficult interpretation of field measurements due to the uncertainties from the limited spatial resolution of inflow measurements.
Some data from these model descriptions may suffer from uncertainties, other data may be lacking or ignored e. The data from the experiments described in this article are often given as aerodynamic coefficients, that is, they are nondimensionalized with the dynamic pressure and presented as function of angle of attack. It should be realized, however, that an angle of attack on a rotating blade as an equivalent to the 2D wind tunnel angle of attack cannot be measured straightforwardly nor modeled with a flow field calculation from CFD.
Now suppose that a device i. In order to derive the angle of attack from this local inflow angle, this upwash should be subtracted. Some methods to derive the angle of attack are explained in section 5. The differences in methods are diverse.