power one house from a single wind turbine. With growing concerns over use and .... Simulation of Maple Seeds and. Performance Analysis as a Wind Turbine.
Aerodynamic performance of swept blade wind turbine Adarsh H Guruprasad Department of Aerospace Engineering University of Cincinnati OH 45220 Abstract and Background
Literature study
The Wind is available throughout the year. Although, non-uniform, on an average we can harness enough electrical energy to power one house from a single wind turbine. With growing concerns over use and availability of fossil fuels, it is very important to develop methods to get the maximum benefit out of the freely available wind energy. Over the last decade, there have been a lot of research on wind turbines and the optimization of the wind turbine design. The blades, however, were not studied in detail due to their complicated shapes and negligible cost benefits of such cases. With the development in Manufacturing technologies, it is today possible to generate complicated shapes for the same cost.
Wind Turbine Design Theories
This paper aims at one such blade optimization study and presents the results of various sweep configurations of a wind turbine blade. The blade in question is a reengineered NREL phase IV blade with the s809 airfoil. The design point conditions are discussed in detail. Various 3D CFD runs are done for forward and backward swept blades and studied the efficiency parameters which are, the coefficient of power (Cp) and Axial induction factor (a). The main aim is to arrive at the most optimum sweep configuration to get the maximum efficiency out of the wind turbine blades.
Blade Momentum Theory (BMT) – This theory was developed by Betz with an aim to find the amount of power and thrust generated by an ideal rotor. He used a control volume approach with the turbine being an actuator disk in this case. It was found that the power generated by a wind turbine depends on the swept area of the rotor, wind speed, and axial induction factor. The power output is given by, P = ½ (ρAU34a(1-a)2 Where, P = Power A = Swept area of the rotor U = Wind speed a = Axial induction factor ρ = Density of air Further, Betz also proved that the maximum power that can be extracted from the wind is about 59% of the wind power available. This limits the Co-efficient of power (Cp) value to 0.59. Blade Element Theory (BET) – This theory uses the airfoil theory and calculates lift and drag over small elements of a blade and then they are integrated to obtain the lift and drag coefficient over the entire span of the blade. This provides information on the thrust distribution over the span of the blade.
Blade Element Momentum Theory (BEMT) – This is a combination of both the theories and makes use of the thrust values from both the theories.
optimum level, AoA must be changed to keep the Cl same. Effect of sweep on swept area of wind turbine
Swept blades
In swept blades, the meridional co-ordinates change along the span. If the sweep is along the wind direction, it's called the Backward sweep and if it's opposing the wind, it is the forward sweep. The below picture depicts a forward swept blade in the axis-symmetric plane. Figure 2: Swept area of wind turbine The swept area of a wind turbine is the area enclosed within a circle subtended by the blades of the rotor. This depends on the effective radius of the blade. The effective radius of the blade is given by, Reff = Rcos(A) where A = Coning angle or swept angle. Baseline Design Wind turbine blade- Re-engineered NREL phase IV blade Figure 1: Forward swept blade [6]
Number of blades on rotor- 2
Coefficient of lift (Cl)
Radius of the blade- 5.029 m
The coefficient of lift is a measure of the lift and hence the power generated in the blades. It is a function of the angle of attack (AoA) and coning angle (A)
Design point AoA – 7 degrees
Cl = 2πCos(A)α
Tip speed ratio- 7.5
A = Coning angle or sweep angle
Airfoil section- s809 [1]
Wind speed- 7 m/s Rotational speed- 72 rpm
This Co-efficient lift varies with the sweep angle. Therefore, for a swept blade, with constant AoA, Cl decreases. To keep Cl at an 2
Methodology •
•
•
• •
•
• •
•
The re-engineered NREL phase IV blade design points are the inputs for this study Using these inputs, the Reynolds number and the coefficient of lift for each blade section are calculated using Xfoil analysis tool [2]. By using the equation for Cl, the AoA is then adjusted for a swept blade to keep the Cl same. The value is Cl is checked by running the Xfoil again. The 3D Blade geometry builder is then executed to generate the blade shape for the given case. The Geomturbo script converts 3dbgb section files into a “geomturbo” file which can be read by the autogrid mesh generator. The geometry is then imported to autogrid and meshed. The boundary conditions are set as required and the 3d CFD run is initiated. The outputs of this run will be the power, torque, thrust developed by the turbine along with various other parameters obtained by post processing.
• • • • •
15 degrees forward and backward sweep 20 degrees forward and backward sweep 6 degrees backward sweep 8 degrees backward sweep 12 degrees backward sweep
The below figure represents a flowchart of the entire process.
Test cases The following test cases are planned to study the swept blade configuration • • •
Straight blade (Baseline design) 5 degrees Forward and backward sweep 10 degrees forward and backward sweep
Figure 3: Process flow chart
3
CFD Analysis Inputs and boundary condition
pressure gradient, Velocity in Z-direction and static pressure with streamlines in the axissymmetric plane.
Number of grid points- 4.8 Million Fluid- Perfect air Flow model- Steady turbulent Navier-Stokes S-A model (for blades) Euler solid wall for hub and far field Abu-Ghannam and Shaw (AGS) transition model Reynolds number- 12000000
Figure 6: Radial static pressure gradientStraight blade
Standard temperature and pressure Mesh configuration Meshing is through the autogrid grid generator tool. The geometry input is through the geomturbo script file. There are about 4.8 million grid points and no negative volumes are present.
Figure 7: Radial static pressure gradient- 5 Deg backward sweep
Figure 4: Meshed blade
Figure 8: Radial static pressure gradient- 5 Deg forward sweep Figure 5: Mesh near the leading edge (100% span) Post processing Each run is post-processed to obtain power, thrust, torque and other items of interest which include, Radial component of static
Figure 9: Radial static pressure gradient- 10 Deg backward 4
substantiated by the Axial induction factor plots discussed later.
Figure 10: Radial static pressure gradient- 10 Deg forward Figure 13: Mass averaged Vz – Straight blade
Figure 11- Radial static pressure gradient- 20 Deg backward
Figure 12- Radial static pressure gradient- 20 Deg forward To summarize from the Radial static pressure gradient contours, the backward swept blades have a positive gradient across the blade while the forward swept blades have a negative pressure gradient. This negative pressure gradient reduces the velocity of the wind that hits the blade thereby causing a reduced power generation. The effect of this is also
Figure 14: Mass averaged Vz- 5 deg backward
Figure 15: Mass averaged Vz- 5 deg forward
Figure 16: Mass averaged Vz- 10 deg backward 5
Figure 17: Mass averaged Vz- 10 deg forward
Figure 18: Mass averaged Vz- 20 deg backward
Figure 20: Static pressure contours with streamlines- Straight blade
Figure 21: Static pressure contours with streamlines- 10 Deg backward
Figure 19: Mass averaged Vz- 20 deg forward The mass averaged Vz in the axis-symmetric plane provides a further insight into the velocities seen by the blades. There is a subtle difference between the cases. The velocity seen by the blade seems to be the highest for 20 Deg backward sweep while lowest for 20 Deg forward sweep. This father substantiates the negative pressure gradient phenomenon discussed earlier.
Figure 22: Static pressure contours with streamlines- 10 Deg forward
Figure 23: Static pressure contours with streamlines- 20 Deg backward 6
Figure 24: Static pressure contours with streamlines- 20 Deg forward
Figure 28: Relative velocity vectors at 50% span20 Deg backward
Figure 25: Relative velocity vectors at 50% span- Straight blade
Figure 29: Relative velocity vectors at 50% span20 Deg forward The relative velocity directions as seen in the contours are almost the same for all the cases. There are minor changes in the angle of attack which cannot be seen in the contours.
Results Coefficient of Power (Cp)
Figure 26: Relative velocity vectors at 50% span- 5 Deg backward
Figure 27: Relative velocity vectors at 50% span- 5 Deg forward
Table 1: Coefficient of power values 7
Axial induction factor is a measure of the efficiency of the wind turbine. Higher than the value of a, lower the efficiency.
Table 2: Coefficient of power- Backward sweeps Figure 31: Span-wise Axial induction factor for backward sweeps
Figure 30: Effective radius Vs Cp for backward sweeps As seen from the comparison, the Coefficient of power increases for backward sweeps with the highest being that for 8 degrees backward swept case with a value of 0.259. Axial Induction Factor (a)
Axial induction factor is the ratio of the difference between wind velocity just upstream of the blade and free stream velocity to the free stream velocity. a = Vfree - Vup/Vfree
Figure 32: Span-wise axial induction factor for forward sweeps From the graphs, the backward sweeps have lower axial induction factors and hence higher efficiencies. This can be attributed to the wind speed at blade approach being higher than that of forward sweep due to the positive radial pressure gradient in backward sweeps. 8
Conclusion and Recommendations Based on the 3D CFD analysis results, the backward sweep, in particular, 8 degrees backward sweep seems to be the most optimum case as far the efficiency of the wind turbine blade is concerned. This increase in efficiency is likely because of the positive radial pressure gradient in backward sweeps which allows higher wind speeds at blade approach than those for forward sweeps. While the blade in study is not completely optimized to provide higher Cp values, it is sufficient to prove the case of backward sweep. The swept area effect increases with increase in coning angle resulting in a drop in Cp for higher sweep angles. Since area is exponentially related to effective radius, it decreases exponentially after 10 degrees of sweep. Therefore, although there is no practical limit for sweep angle, it is most effective within 10 degrees.
References 1. Airfoil tools: http://airfoiltools.com/airfoil/details?ai rfoil=s809-nr 2. Xfoil: raphael.mit.edu/xfoil/ 3. Kedharnath Sairam, 2013, The Influence of Radial Area Variation on Wind Turbines to the Axial Induction Factor, MS Thesis Defense. 4. Jacob Holden,2016 Experimental Testing and Computational Fluid Dynamics Simulation of Maple Seeds and Performance Analysis as a Wind Turbine. 5. Korn Saran – Yasoontorn, 2002, Mechanics of Wind Turbine, University of Texas- Austin 6. Adarsh, Athreya and Akshay, 2016, The effect of sweep and translation on wind turbine performance.
The Axial induction factor decreases with higher sweep angles resulting in higher efficiencies but has predominant effect only within the first 20 degrees after which the swept area effect overshadows it. While this study sufficiently concludes that the smaller degree backward sweeps are in general good for higher power and efficiency, there is still a lot of study to be made regarding the structural ability of the backward swept blade, optimizing the blade angles for maximum lift, manufacturing considerations and the overall bearing loads coming onto the shaft.
9
