Experiments on the Performance of Small Horizontal Axis Wind ... - MDPI

energies Article

Experiments on the Performance of Small Horizontal Axis Wind Turbine with Passive Pitch Control by Disk Pulley Yu-Jen Chen and Y. C. Shiah * Institute of Aeronautics and Astronautics, National Cheng Kung University, Tainan 70101, Taiwan; [email protected] * Correspondence: [email protected]; Tel.: +886-6275-7575 (ext. 63623) Academic Editor: Frede Blaabjerg Received: 23 March 2016; Accepted: 4 May 2016; Published: 7 May 2016

Abstract: The present work is to design a passive pitch-control mechanism for small horizontal axis wind turbine (HAWT) to generate stable power at high wind speeds. The mechanism uses a disk pulley as an actuator to passively adjust the pitch angle of blades by centrifugal force. For this design, aerodynamic braking is caused by the adjustment of pitch angles at high wind speeds. As a marked advantage, this does not require mechanical brakes that would incur electrical burn-out and structural failure under high speed rotation. This can ensure the survival of blades and generator in sever operation environments. In this paper, the analysis uses blade element momentum theory (BEMT) to develop graphical user interface software to facilitate the performance assessment of the small-scale HAWT using passive pitch control (PPC). For verification, the HAWT system was tested in a full-scale wind tunnel for its aerodynamic performance. At low wind speeds, this system performed the same as usual, yet at high wind speeds, the equipped PPC system can effectively reduce the rotational speed to generate stable power. Keywords: blade element momentum theory; passive pitch control; disk pulley; small horizontal axis wind turbine

1. Introduction Nowadays, it has been widely recognized that wind is one of the promising green energy sources for its non-regional availability. This has driven significant resource efforts devoted to the study of wind energy (e.g., [1–3]). In principal, there are two major classifications for wind turbines, namely vertical axis wind turbine (VAWT) and horizontal axis wind turbine (HAWT). Typically, a VAWT can generate more power at a relatively low wind speed in urban areas compared to the HAWT. Nevertheless, when they both operate at the same wind speed, a HAWT is expected to generate more power since the aerodynamic drag is less and the entire rotation of all blades receives more wind power. Currently, small-scale wind turbines are gaining more attention due to the fact that rising energy costs have driven the consumer to seek alternative solutions of independent power supply. Despite the high efficiency of the HAWT, its main drawback is the potential risk of circuit burnout or structural failure when the wind speed exceeds its critical value. The present research is to provide a solution to this problem of the small-scale HAWT by an innovative passive pitch-control mechanism. Over the years, there are various solutions proposed, including the short-circuit braking system, Choudhry et al. [4] has outlined the control requirements for dynamic stall [5] for wind turbines, where three passive control methodologies have been investigated. By applying this approach of dynamic stall control, Yen and Ahmed [6] showed synthetic jet actuation to be effective for a vertical axis wind turbine (VAWT). By the stall-controlled approach, the system may sustain its stable operation

Energies 2016, 9, 353; doi:10.3390/en9050353

www.mdpi.com/journal/energies

Energies 2016, 9, 353

2 of 13

generating power under a critical wind speed. However, when operating beyond the critical wind speed, the system will be at the risk of over-running due to a rapid rise of receiving excessive wind power. Another approach proposed by Xie et al. [7,8] was to design a folding blade of turbine rotor, controlled by a servo motor [7] or a passive mechanism regulation [8]. The main drawback of this design is its difficulty of retaining dynamic equilibrium at variable speeds of rotation. Another innovative design of a flywheel rotor system was proposed by Jauch and Hippel [9] to control the rotation of HAWT via inertia effect. Still, this flywheel mechanism has very limited capability of braking when the HAWT is subjected to very high wind speeds. Having borrowed the design of trailing edge flap of airplanes, Barlas et al. [10] used active flaps on a small-scale HAWT. An alternative design to protect the HAWT system is the so called “yaw control” mechanism, which applies the principal of misalignment to achieve the goal of reducing rotational speeds. Under operations at high wind speeds, the system shall keep yawing to seek a balance with the incoming wind. Among many innovations in this category, Kragh [11] proposed a rotor speed-dependent yaw control of wind turbines. Although applicable, this design has a potential problem of structural fatigue occurring at its support due to the dynamic loads arising from the frequent oscillations of the system. For this, Ekelund [12] presented mathematical models obtained from the equations of motion and proposed to use a yaw servo for attenuation of structural dynamic load oscillations. Additionally, Shariatpanah [13] developed another model for PMSG-based wind turbine with yaw control. Despite its effectiveness, the active control system requires a complicated design of an electro-mechanical sub-system and thus, the system reliability is still an issue when operated under long lasting sever conditions. Some other active pitch-angle control can be found in [14–17]. Still, these systems require relatively delicate electro-mechanical systems for fulfilling the function of active control and its reliability is always an issue of concerns. Moreover, the issue of consuming extra electric power is another drawback of the consumer’s concern for using the active control. To the authors' personal opinions, designs of passive pitch-angle control on these aspects are comparably more ideal in consideration of system reliability and saving of electricity. Li [18] presented a self-powered passive adaptive control to adjust pitch angle of blade. The principal of the self-powered passive adaptive control is to use aerodynamic force acting upon on the blade as the control power under balance with the spring force. Hertel et al. [19] developed a simple strip theory model for the analysis of a conceptual passive pitch control, providing the aerodynamics as a function of the relative wind vector’s magnitude and local angle of attack. Penghu, the largest island of the Pescadores Islands in the Taiwan Strait, has a wind turbine test unit. Having complete facilities, this platform has acquired the international certification of small wind certification council (SWCC). For collecting data from the original HAWT system without pitch-control, we set up HAWT systems of 200 W, 400 W, and 1 kW output at the Penghu test unit (Figure 1). Figure 2a displays all collected data of the 400 W system, as an example, for the monitoring period of one month (March 2012). From the collected data, it can be clearly seen that the data meets the designed power under normal operation as shown by the green region (before the rated wind speed). However, the blue region, the wind turbine system is operated over the rated wind speeds and should be braked for protection of over power outputs. Under the overloaded operation of up to 600 W, the long-term operation shall lead to generator corruption (Figure 2b). As the objective of this research, this has inspired us to make the best use of the good wind-farm to stably generate electric power even at over rated wind speeds. The present work aims to design a disk pulley mechanism for a prototype of a small-scale HAWT system to passively control the blade pitch angle such that the system is capable of outputting stable power at over rated wind speeds. SD8000-airfoil [20] is adopted in the blade design for its excellent lift-to-drag ratio at low Reynolds numbers to acquire a higher power coefficient. This article is divided into three parts: (1) BEMT design and software development; (2) design of the disk pulley mechanism; and (3) wind tunnel experiments. Through the complete processes of design, analysis, and experiments, the feasibility of the prototype is verified at the laboratory stage. This has under laid the foundation of possible extended applications to practical small-scale HAWT systems with various power outputs.

Energies 2016, 9, 353

3 of 13

Energies 2016, 9, 353 Energies 2016, 9, 353

3 of 13 3 of 13

Figure 1. Small-scale HAWT systems tested at Penghu. Figure Figure1.1.Small-scale Small-scaleHAWT HAWTsystems systemstested testedatatPenghu. Penghu.

(a) (a)

(b) (b)

Figure 2. (a) Monitoring data of the 400 W HAWT; and (b) damage of the generator of the 400 W Figure2.2.(a) (a)Monitoring Monitoring data of the W HAWT; and (b) damage of the generator of W theHAWT. 400 W HAWT. Figure data of the 400 400 W HAWT; and (b) damage of the generator of the 400 HAWT.

2. Blade BladeModeling Modelingby byBEMT BEMTand andDevelopment Developmentof ofGUI GUISoftware Software 2. Blade Modeling by BEMT and Development of GUI Software 2.1. Theoretical Modeling 2.1. Theoretical Modeling of of Blade Blade Design Design 2.1. Theoretical Modeling of Blade Design Our Our blade blade design design is is based based on on BEMT, BEMT, which which has has been been widely widely applied applied in in industry industry for for the the blade blade Our blade design is based on BEMT, which has been widely applied in industry for the blade design flowchart collating the blade design process according to [21,22] is shown design of ofwind windturbines. turbines.The The flowchart collating the blade design process according to [21,22] is design of wind turbines. The flowchart collating the blade design process according to [21,22] in Figure 3a. Table 1 lists all parameters for this blade design. Figure 3b shows the prototype and shown in Figure 3a. Table 1 lists all parameters for this blade design. Figure 3b shows the prototypeis shown indrawing Figure 3a. Table lists all parameters for this blade design. 3b shows the prototype the CAD of the design following processes depicted byFigure the flowchart. Table 2 lists all and the CAD drawing of blade the1 blade design following processes depicted by the flowchart. Table 2 lists and the CAD drawing oflengths the blade design following by the the SD8000 flowchart. 2 lists designed values of chord and pitch angles forprocesses ten sections of the of airfoil [20]Table as shown all designed values of chord lengths and pitch angles for ten depicted sections airfoil SD8000 [20] as 1 all designed values of chord lengths and pitch angles for ten sections of the airfoil SD8000 [20] in Figure 4a. All defined angles are indicated in Figure 4b, where a is the axial induction factor and aas shown in Figure 4a. All defined angles are indicated in Figure 4b, where a is the axial induction factor shown Figure 4a. All definedfactor. angles are indicated in Figure 4b, where a is the axial induction factor is thea'angular induction factor. and isinthe angular induction and a' is the angular induction factor.

Energies 2016, 9, 353 Energies 2016, 9, 353 Energies 2016, 9, 353

4 of 13 4 of 13 4 of 13

(a) (a)

(b) (b)

Figure 3. (a) Flowchart ofofthe design from BEMT; and (b)(b) CAD for for ten sections of the and Figure theblade blade design from BEMT; and CAD sections of blade the blade Figure3.3. (a) (a) Flowchart Flowchart of the blade design from BEMT; and (b) CAD for tenten sections of the blade and prototype. and prototype. prototype.

(a) (a)

(b) (b)

Figure 4. (a) Geometry of airfoil SD8000; and (b) BEMT angle definition for an airfoil. Figure4.4.(a) (a)Geometry Geometryof ofairfoil airfoilSD8000; SD8000;and and(b) (b)BEMT BEMTangle angledefinition definitionfor foran anairfoil. airfoil. Figure Table 1. The parameter of the blade. Table 1. The parameter of the blade. Table 1. The parameter of the blade. Item Parameter Item Parameter Item Parameter Item Parameter −3 ρ Item 1.225 (kg·m −3) R 0.41 (m) Parameter Item Parameter ρ 1.225 (kg·m ) R 0.41 (m) TSR 3.7 Root of blade 0.14 (m) ´3 TSR ρ 3.7 Root of blade 0.14 (m) R 0.41 (m) 1.225 (kg¨ α 5° m ) B 3 α TSR 5° B 3.7 Root of blade 0.14 (m) 3 Vs 12 (m/s) Rated Power 200 (W) ˝ B Power 3 200 (W) Vs α 12 5(m/s) Rated Vs 12 (m/s) Rated Power 200 (W) Table 2. The Geometry parameter of the blade. Table 2. The Geometry parameter of the blade. Table Geometry parameter of the blade. Section r/R 2. TheChord Length (m) Pitch Angle (deg) Section r/R Chord Length (m) Pitch Angle (deg) 1 0.341 0.06 0 1 0.341 0.06 0 Chord Length (m) Pitch Angle 2 Section 0.415 r/R 0.093 18.48(deg) 2 0.415 0.093 18.48 3 0.4880.341 0.107 15.27 0.06 015.27 3 1 0.488 0.107 4 0.561 0.095 12.81 2 0.415 0.093 18.48 4 0.561 0.095 12.81 5 0.6340.488 0.085 10.86 3 0.107 15.27 5 0.634 0.085 10.86 4 0.095 12.81 6 0.7070.561 0.076 9.29 6 0.707 0.076 9.29 5 0.085 10.86 7 0.7800.634 0.070 8.00 7 6 0.780 0.070 8.00 0.076 9.29 8 0.8540.707 0.064 6.92 8 7 0.854 0.064 6.92 0.780 0.070 8.00 9 0.927 0.059 6.00 9 8 0.927 0.059 6.00 0.854 0.064 6.92 10 1.000 0.055 5.21 10 9 1.000 0.055 5.21 0.927 0.059 6.00 10 1.000 0.055 5.21

Principally, the effective radius of rotation R [21–25] is determined by: Principally, the effective radius of rotation R [21–25] is determined by:

R  0.5Vs C 3  P R  0.5Vs Cp p P 3

(1) (1)

Energies 2016, 9, 353

5 of 13

Principally, the effective radius of rotation R [21–25] is determined by: b R “ 0.5ρVs3 C p πP

(1)

The tip speed ratio (TSR) of the system, denoted by λ [21–25] here, is defined by: λ“

Rω Vs

which is 3.7 for the present design. The speed ratio for each blade section [21–25] is: ´r¯ λr “ λ R

(2)

(3)

where r is the rotational radius of each blade section as defined in Table 2. Then, the tip speed ratio at each section given in Equation (3) can be substituted into the following equation to give the local inflow angle by [21–25]: ˙ ˆ 2 ´1 ϕ “ tan (4) 3λr For giving more detailed derivations, consider the Figure 4b. From the figure, one may readily have the following relationships from geometrical considerations [21–25]: tanϕ “

1´a Vs p1 ´ aq “ 1 ωrp1 ` a q p1 ` a1 qλr

(5)

For the case when a1 = 0 (no wake rotation) and a = 1/3 (the Betz optimum rotor) [21–25], one obtains: tanϕ “

2 3λr

(6)

Thus, the values of λr and ϕ obtained above can be further substituted into the following equation to give the corresponding cord length c [21–25]: c“

8π rsinϕ 3 B C L λr

(7)

For gaining optimal performance, the angle of attack α should be selected to achieve its maximum lift-to-drag ratio. With reference to Figure 4b, the pitch angle of blade [21–25] for each section of the rotor is: θp “ ϕ ´ α

(8)

A piece of GUI-software, called HAWT-Blade, was developed in MATLAB, where all pertinent formulas in Equations (1)–(8) are integrated altogether, as shown in Figure 5. As can be observed from this figure, the pitch angles and chord lengths near the root are modified for the prototype design. This is mainly due to the large pitch angles and chord lengths near that area, which would cause manufacturing difficulties and overweight of the blade; however, their contributions to the overall resulting torque are very trivial.

Energies 2016, 9, 353

6 of 13

Energies 2016, 9, 353

6 of 13

Energies 2016, 9, 353

6 of 13

Figure 5. The analysis by HAWT-Blade. Figure 5. The analysis by HAWT-Blade. Figure 5. The analysis by HAWT-Blade.

2.2.GUI GUIofofBlade BladePerformance PerformanceDevelopment Development 2.2. 2.2. GUI of Blade Performance Development Having provided all design parameters for the blade design, one may then calculate the resulting Having provided all design parameters for the blade design, one may then calculate the resulting Having provided torque Q [21–25] by: all design parameters for the blade design, one may then calculate the resulting torque Q [21–25] by: torque Q [21–25] by:  B 2 10 10´ (i ) (i ) Q ρ B 2U rel ÿ  10 CpiLq sin  i  CpiDq cos ¯i ci ri (9) B rel 2∆ i Q “ 2U C sinϕ ´ C ci ri (9) ( i ) ( icosϕ ) 1 L i i D Q 2 U rel   C L sin  i  C D cos  i ci ri (9) i “ 1 2 i 1 (i) (i) where Δ denotes the constant interval length between two consecutive sections, CpiLq /CDpiqrepresents (i) (i) where ∆Δdenotes the length sections, CCL /C represents where denotes theconstant constant interval length between twoconsecutive consecutive sections, represents the lift/drag coefficient of the interval i-th section, i isbetween the localtwo inflow angle of the i-th section, cDiDand ri stand L /C the lift/drag coefficient of the i-th section, φ is the local inflow angle of the i-th section, crii stand and i the lift/drag therotational i-th section, i isof thethe local angle of the i-thAnother section, cpiece i and of for the cord coefficient length andofthe radius i-thinflow section, respectively. GUI-ri stand for the cord length and the rotational radius of the i-th section, respectively. Another piece for the cord length and the rotational of the i-th section, respectively. Another piece of GUIsoftware, called HAWT-Perform, was radius developed using Equation (9) to calculate the output torque. ofsoftware, GUI-software, calledofHAWT-Perform, was developed using Equation (9) to calculate thetorque. output called was developed using Equation (9)oftothe calculate the torque, output Figure 6 is the HAWT-Perform, GUI HAWT-Perform, showing such analysis resulting output torque. Figure 6 is the GUI of HAWT-Perform, showing such analysis of the resulting torque, output Figure 6 is power, the GUI ofthe HAWT-Perform, showing analysisofof resulting mechanical and corresponding Cp (powersuch coefficient) thethe blade design.torque, output mechanical power, and the corresponding Cp (power coefficient) of the blade design. mechanical power, and the corresponding Cp (power coefficient) of the blade design.

 

 

Figure 6. The performance curve of blade predicted by HAWT-Perform. Figure 6. The The performance performance curve Figure 6. curve of of blade blade predicted predicted by by HAWT-Perform. HAWT-Perform.

As displayed in Figure 6, one may readily obtain corresponding performance curves with an in Figure may readily obtain with an inputAs ofdisplayed pitch control angle 6, forone a known geometry of corresponding a blade design.performance From this, itcurves may be easily input of pitch control angle for a known geometry of a blade design. From this, it may be easily

Energies 2016, 9, 353

7 of 13

Energies 9, 353 in Figure 6, one may readily obtain corresponding performance curves with an input 7 of 13 As2016, displayed

of pitch control angle for a known geometry of a blade design. From this, it may be easily observed observed how theof variation of pitch control affects the resulting performance curves. Just as how the variation pitch control angle affectsangle the resulting performance curves. Just as independent ˝ ˝ ˝ ˝ independent for demonstrations, the angles pitch 0control angles were input examples for examples demonstrations, the pitch control /10 /20 /30 0°/10°/20°/30° were input separately in separately in HAWT-Perform forequaling wind speed equaling m/s. set ofare outcomes areand recorded and HAWT-Perform for wind speed 12 m/s. Each12set of Each outcomes recorded plotted in plotted in Figure forcoefficients the powerand coefficients and torques,From respectively. From plots, is Figure 7a,b for the 7a,b power torques, respectively. these plots, it isthese evident thatitthe evident that the system work with and reduced torque and power The remains to system shall work with shall reduced torque power coefficients. Thecoefficients. issue remains toissue design a proper design a proper mechanism to adjust pitch control angles at high wind speeds in a passive way. mechanism to adjust pitch control angles at high wind speeds in a passive way. Before elaborating this Before this point of designing the mechanism, it is of important grasp the effect of point ofelaborating designing the mechanism, it is important to grasp the effect changingtopitch control angle on changing pitch control onblade, the resulting flow aroundnext. blade, which will be addressed next. the resulting flow field angle around which will befield addressed

(a)

(b)

Figure 7. Performance Performancecurves curves sample pitch control angles, simulated by HAWT-Perform at Figure 7. forfor sample pitch control angles, simulated by HAWT-Perform at 12 m/s: 12 m/s: (a) Cp vs. TSR curves; and (b) torque vs. RPM curves. (a) Cp vs. TSR curves; and (b) torque vs. RPM curves.

3. 3. Mechanical MechanicalDesign Designof ofPitch PitchControl Control In principal, the pitch-control mechanism can be broadly categorized categorized into two two kinds: kinds: active control and passive control. The former requires delicate electro-mechanical designs with relatively high costs, while the latter is less costly with a simpler simpler design. The most obvious obvious merit of using a passive pitch pitchcontrol controlis is its reliability when operating under sever conditions. There are designs several its reliability when operating under sever conditions. There are several designs of passive pitch control proposed over the years. The novel design of this present worka of passive pitch control proposed over the years. The novel design of this present work employs employs a disk pulley with combined with newly-designed components achieve goal of disk pulley combined newly-designed shift-driveshift-drive components to achieveto the goal ofthe passively passively controlling the pitch angle of turbine blades. In what follows, details of the motion principal controlling the pitch angle of turbine blades. In what follows, details of the motion principal along alongthe with the functions each component be elaborated with functions of eachofcomponent will bewill elaborated next. next. 3.1. Principle of Motion Verifying by the foregoing analyses by theories, one may proceed with the process to design the mechanism of of the thepassive passivepitch pitchcontrol controlsystem. system. functioning in motor scooters, pulley mechanism AsAs functioning in motor scooters, the the diskdisk pulley can can be applied for this purpose. To meet this working requirement, a disk pulley mechanism has be applied for this purpose. To meet this working requirement, a disk pulley mechanism has been been designed as shown in Figure 8a. When in operation, of the start bladetostart to rotate, as a designed as shown in Figure 8a. When in operation, rotors rotors of the blade rotate, and as and a result, result, centrifugal are produced to drive to move forward (Figure8a). 8a).With Withthe the forward forward centrifugal forces forces are produced to drive the the diskdisk to move forward (Figure moving of the disk pulley at high rotational speeds, pitch angles of the blades are adjusted by an assemblage of bearings. Figure 8b shows pictures of the HAWT blades when the disk pulley triggers 90˝ . From From Figure Figure 8b, one may clearly see how the disk pulley functions the rotation at the pitch angle 90°. to control the pitch angle.

Energies 2016, 9, 353

8 of 13

Energies 2016, 9, 353

8 of 13

(a)

(b)

Figure 8. Mechanism of the passive pitch control: (a) cross-section of the design; and (b) blade motion Figure 8. Mechanism of the passive pitch control: (a) cross-section of the design; and (b) blade motion of the prototype. of the prototype.

3.2.Components ComponentsFunction Function 3.2. When the the disk disk pulley pulley operates operates under under high-speed high-speed rotation, rotation, weight weight rollers rollers in in the the disk disk will will be be When pushed toward the outside wall by the centrifugal force and then drive the blades to rotate by both pushed toward the outside wall by the centrifugal force and then drive the blades to rotate by both aa cam-bearingand andaalinear linearbearing bearing(Figure (Figure displayed Figure 9, the complete system consists cam-bearing 9).9). AsAs displayed in in Figure 9, the complete system consists of of five main components, namely the disk pulley, bearings, spring, torque sensor, and the blades. In five main components, namely the disk pulley, bearings, spring, torque sensor, and the blades. In most most practice, active pitch control is generally applied with various types of gears powered by servopractice, active pitch control is generally applied with various types of gears powered by servo-motors motors topitch change pitch angle Thisofdesign power transmission by gears generally faces a to change angle (e.g., [7]).(e.g., This[7]). design powerof transmission by gears generally faces a problem problem of assemblage and backlash. To improve the conventional design of gears, this system of assemblage and backlash. To improve the conventional design of gears, this system adopts linear adopts linear and ball bearings to alter the direction of movement. The cam bearings, cam bearings, bearings,cam and bearings, ball bearings to alter the direction of movement. The cam bearing is bearing is designed to displace along the planned track to drive blades to rotate. Additionally, designed to displace along the planned track to drive blades to rotate. Additionally, a spring unit isa springat unit placed the center to resist the motion of the pulley and to position it back placed theiscenter toat resist the rigid motion ofrigid the disk pulley anddisk to position it back to the original to the original place when the rotation slows down at low rpms. place when the rotation slows down at low rpms. According to to each PPC system mainly consists of four subAccording each respective respective working workingfunction, function,thethe PPC system mainly consists of four systems, including (1) the pulley system; (2) the forward-moving system; (3) the shift-drive system; sub-systems, including (1) the pulley system; (2) the forward-moving system; (3) the shift-drive and (4) the system. For the sub-system (1), the main(1), partthe is the disk pulley applied system; andposition-recovering (4) the position-recovering system. For the sub-system main part is the disk in scooters. As shown in Figure 10, there are six weight rollers placed inside the disk. Figure 8a pulley applied in scooters. As shown in Figure 10, there are six weight rollers placed inside the disk. schematically depicts the forward motion of the disk pulley driven by the weight rollers when Figure 8a schematically depicts the forward motion of the disk pulley driven by the weight rollers subjected to high-speed rotation. BeingBeing connected to thetodisk the linear bearing of the when subjected to high-speed rotation. connected the pulley, disk pulley, the linear bearing of subthe system (2) is, thus, driven to move forward. The sub-system (3) consists of the ball bearing and the sub-system (2) is, thus, driven to move forward. The sub-system (3) consists of the ball bearing and the cam bearing, functioning to replace the conventional design of a rack and pinion. For this sub-system, cam bearing, functioning to replace the conventional design of a rack and pinion. For this sub-system, themoving movingforce forceconveyed conveyedfrom from is used to turn rotors adjust the pitch angles. By the (2)(2) is used to turn the the rotors andand thenthen adjust the pitch angles. By this, this, the shortcoming of requiring tolerance precision when applying the conventional mechanism of the shortcoming of requiring tolerance precision when applying the conventional mechanism of rack rackpinion and pinion be overcome. Sub-system (4) includes the components a spring and an and can be can overcome. Sub-system (4) includes the components of a springofand an adjustable adjustable nut. The spring functions to recover the sub-systems back to their original positions; nut. The spring functions to recover the sub-systems back to their original positions; the nut is usedthe to nut is the used to adjustforce, the resistance controlling the of triggering of rotation adjusting adjust resistance controllingforce, the triggering speed rotation speed for adjusting pitchfor angle. pitch angle.

Energies 2016, 9, 353

9 of 13

Energies 2016, 9, 353

9 of 13

Energies 2016, 9, 353

9 of 13

Figure 9. 9. Components Components of of the the pitch pitch control control mechanism and the prototype. Figure Figure 9. Components of the pitch control mechanism and the prototype.

Figure 10. Regional components of the PPC system. Figure 10. Regional components of the PPC system. Figure 10. Regional components of the PPC system.

4. Experiment of ABRI Wind Tunnel 4. Experiment of ABRI Wind Tunnel 4. Experiment of ABRIof Wind Tunnel After confirmation the performance curves, the wind turbine system was manufactured and After confirmation of the performanceand curves, the wind turbine system was manufactured and experimented in the ABRI (Architecture Building Research Institute: Tainan, Taiwan) wind After confirmation of the performance curves, the wind turbine system was manufactured and experimented in the ABRI (Architecture and Building Research Institute: Tainan, Taiwan) wind tunnel (Figurein11) measure the generator voltage,Research current,Institute: and torque. TheTaiwan) ABRI environment experimented thetoABRI (Architecture and Building Tainan, wind tunnel tunnel (Figure 11) to measure the generator voltage, current, and torque. The ABRI environment wind the National Chengvoltage, Kung University (Tainan, employedwind to test the (Figuretunnel 11) to at measure the generator current, and torque. Taiwan) The ABRIwas environment tunnel wind tunnel at the National Cheng Kung University (Tainan, Taiwan) was employed to test the aerodynamic performance of the full scale rotor models. Thiswas windemployed tunnel is atoclosed-circuit type with at the National Cheng Kung University (Tainan, Taiwan) test the aerodynamic aerodynamic performance of the full scale rotor models. This wind tunnel is a closed-circuit type with major components schematically shown in Figure 12. performance of the full scale rotor models. This wind tunnel is a closed-circuit type with major major components schematically shown in Figure 12. components schematically shown in Figure 12.

Energies 2016, 9, 353

10 of 13

Energies 2016, 9, 353 Energies 2016, 9, 353

10 of 13 10 of 13

(a) (a)

(b) (b)

Figure 11. ABRI wind tunnel tests: (a) the small-scale HAWT prototype and its PPC mechanism; Figure 11. 11. ABRI wind tunnel tests: (a) (a) thethe small-scale HAWT prototype and its its PPC mechanism; and Figure ABRI wind tunnel tests: small-scale HAWT prototype and PPC mechanism; and (b) experimental setups. (b) and experimental setups. (b) experimental setups.

Figure 12. Sketch of the ABRI wind tunnel. Figure12. 12.Sketch Sketchof ofthe the ABRI ABRI wind wind tunnel. Figure tunnel.

The test section is 36.5 m long with cross-section dimensions of 4.0 m × 2.6 m. The maximum The test section is 36.5 m long with cross-section dimensions of 4.0 m × 2.6 m. The maximum The speed test section 36.5section m longiswith cross-section 4.0 m ˆ 2.6 m. maximum wind wind in theistest 36 m/s. This winddimensions tunnel has of been verified for The many parameters, wind speed in the test section is 36 m/s. This wind tunnel has been verified for many parameters, including structure stability, flow verified uniformity, turbulence intensity, flow speed in the test sectionvibration, is 36 m/s.temperature, This wind tunnel has been for many parameters, including including structure vibration, temperature, stability, flow uniformity, turbulence intensity, flow angularity, and boundary layer thickness. mean flow uniformity is about and the structure vibration, temperature, stability, flowThe uniformity, turbulence intensity, flow0.37% angularity, and angularity, and boundary layer thickness. The mean flow uniformity is about 0.37% and the turbulence intensity is lessThe thanmean 0.25%flow as well as the boundary thickness is turbulence about 60 mmintensity at the boundary layer thickness. uniformity is aboutlayer 0.37% and the turbulence intensity is less than 0.25% as well as the boundary layer thickness is about 60 mm at the inletthan cross-section. The average flow angularity the pitchisand yaw60angles are, about is less 0.25% as well as the boundary layer of thickness about mm at therespectively, inlet cross-section. inlet cross-section. The average flow angularity of the pitch and yaw angles are, respectively, about ˝ 70.415° and 70.97° below the wind speed of 20 m/s. A pitot-static tube was installed to measure the ˝ The70.415° average flow angularity ofwind the pitch yaw angles are, respectively, 70.415 and 70.97 and 70.97° below the speedand of 20 m/s. A pitot-static tube was about installed to measure the free stream velocity in the test section because it is stable, accurate, and convenient to calibrate. The below wind speed in of the 20 m/s. A pitot-static was installed toand measure the free velocity free the stream velocity test section becausetube it is stable, accurate, convenient to stream calibrate. The variable reluctance pressure transducer is connected to the pitot-static tube and converts the pressure in the test section because it istransducer stable, accurate, and convenient to calibrate. variable variable reluctance pressure is connected to the pitot-static tube andThe converts the reluctance pressure difference between the total pressure and the static pressure into electric voltage. The small analog pressure transducer connected to the pitot-static tube and converts the pressure between difference betweenisthe total pressure and the static pressure into electric voltage. difference The small analog signal was then amplified and converted using an analog/digital converter. The pressure transducer thesignal total was pressure and the static pressure into electric voltage. The small analog signal was then then amplified and converted using an analog/digital converter. The pressure transducer was calibrated by a micro manometer and verified before the experiments. The flow speed in the test amplified and converted using an analog/digital pressure The transducer wasincalibrated was calibrated by a micro manometer and verifiedconverter. before theThe experiments. flow speed the test section was calculated based on the incompressible steady Bernoulli’s equation. by section a microwas manometer verified the experiments. The flowequation. speed in the test section was calculatedand based on thebefore incompressible steady Bernoulli’s For the prototype to function properly in the ABRI wind tunnel, the choice of 21.5 g per single Forbased the prototype to function properly the ABRI wind tunnel, the choice of 21.5 g per single calculated on the incompressible steadyinBernoulli’s equation. weight roller was made and there were six pieces in all. Additionally, the spring coefficient was weight roller was made and there were six pieces in all. Additionally, spring coefficient was For the to as function properly indesign the ABRI wind tunnel, thethe choice of 21.5 g per chosen to prototype be 260 (N/m) the companioned parameter. Under these test conditions, the single PPC chosen to bewas 260 made (N/m) and as the companioned design parameter. Under these test conditions, the PPC weight roller there were six pieces in all. Additionally, the spring coefficient was system started to work at the wind speed 8 m/s below its rated wind speed (12 m/s). For providing system started to work at the wind speed 8 m/s below its rated wind speed (12 m/s). For providing chosen to be 260data, (N/m) theexperiment companioned parameter. Undersystem, these test conditions, PPC comparative the as same wasdesign repeated for the HAWT yet with its pitchthe fixed comparative data, theat same experiment was repeated for the HAWT system, yet m/s). with its pitch fixed system started to work the wind speed 8 m/s below its rated wind speed (12 For providing by a screw (Figure 9), denoted here by PF-system. Since the both experiments of PF/PPC-system were by a screwdata, (Figure denoted here by was PF-system. Since both experiments of PF/PPC-system wereby comparative the9), same experiment repeated forthe the HAWT system, yet with itssuch pitchthat fixed conducted for the same prototype, no issue of different hub-weight needs to concern the conducted for9), thedenoted same prototype, no issue of different needs to concern such that the a screw (Figure here by the hub-weight both experiments PF/PPC-system performance difference between the PF-system. both could beSince experimentally identified for of various wind speeds.were performance difference between the both could be experimentally identified for various wind speeds.

Energies 2016, 9, 353

11 of 13

Energies Figure 2016, 9, 13 353shows the electric output powers and RPM of the both systems at various wind speeds. 11 of 13

As can be observed from this figure, both started to generate power at 4 m/s. Additionally, it is evident to see the difference of the both once the wind speed reached 8 m/s when the PPC-system started to for work. ourprototype, expectation, PPC-system functioned properly withthat thethe conducted the To same nothe issue of different hub-weight needsintoaccordance concern such original design whereas, in contrast, the PF-system output power inidentified a mannerfor of various cubic growth performance difference between the both could be experimentally windwith speeds. Energies 2016, 9, 353 of 13 the wind speed. At this now, it is powers very clear that operation of thesystems small HAWT system is11 very Figure 13 shows the point electric output and RPM of the both at various wind speeds. particularly when its electric controllergenerate malfunctions; at high wind speeds, theit system Asdangerous, can be observed fromthe this figure, both started atsystems 4 m/s. Additionally, is evident Figure 13 shows electric output powers to and RPM ofpower the both at various wind speeds. either break down with itsonce generator burned out reached for overloading or fallthe apart from the great tomay see the difference of the both the wind speed 8 m/s when PPC-system As can be observed from this figure, both started to generate power at 4 m/s. Additionally,started it is centrifugal force. As can be observed from Figure 13, the PPC-system started functioning to protect to work. To our expectation, the PPC-system functioned properly in accordance with the original evident to see the difference of the both once the wind speed reached 8 m/s when the PPC-system the system at wind speed reaching 6 m/s. Under the no-loading condition, the stable rotation speed design whereas, in contrast, the PF-system power in a manner of cubic withwith the wind started to work. To our expectation, the output PPC-system functioned properly in growth accordance the of the PPC-system is at 720 rpm, when operating with the maximum power, the PPC-system speed. At this point now, itin is contrast, very clear operation of the smallinHAWT system is very dangerous, original design whereas, thethat PF-system output power a manner of cubic growth with generates stable power at about 500–550 rpm. particularly when At its this electric at high wind speeds, the system may either the wind speed. pointcontroller now, it is malfunctions; very clear that operation of the small HAWT system is very break down with its generator burned out for overloading or fall apart fromwind the great centrifugal force. dangerous, particularly when its electric controller malfunctions; at high speeds, the system either break from downFigure with its burned started out for functioning overloading to or protect fall apart the at great As may can be observed 13,generator the PPC-system thefrom system wind centrifugal force. As can be observed from Figure 13, thethe PPC-system started functioning to protect speed reaching 6 m/s. Under the no-loading condition, stable rotation speed of the PPC-system the system windoperating speed reaching 6 m/s. Under the no-loading condition, the stable rotation speed at is at 720 rpm, at when with the maximum power, the PPC-system generates stable power of the PPC-system is at 720 rpm, when operating with the maximum power, the PPC-system about 500–550 rpm. generates stable power at about 500–550 rpm.

Figure 13. Experiment data of PF and PPC at 4 m/s to 16 m/s.

Figure 14a shows our experimental results for wind speeds at 8 m/s, 10 m/s, and 12 m/s; those for wind speeds at 14 m/s and 16 m/s are in Figure 14b. Obviously, accompanied with the increase of wind speed is the greater percentage of power decrease. At the wind speed 8 m/s, the maximum blade Cp would drop by about 10% while, at 14 m/s, the drop would increase up to 30%. From the experiments, the functioning of the PPC system is very clear that it may be applied to protect the Figure Experimentdata dataofofPF PFand andPPC PPC at at 44 m/s m/s to Figure 13.13. Experiment to16 16m/s. m/s. system from aerodynamic overloading. Figure 14a shows our experimental results for wind speeds at 8 m/s, 10 m/s, and 12 m/s; those for wind speeds at 14 m/s and 16 m/s are in Figure 14b. Obviously, accompanied with the increase of wind speed is the greater percentage of power decrease. At the wind speed 8 m/s, the maximum blade Cp would drop by about 10% while, at 14 m/s, the drop would increase up to 30%. From the experiments, the functioning of the PPC system is very clear that it may be applied to protect the system from aerodynamic overloading.

(a)

(b)

Figure 14. Experimental results: (a) blade Cp—TSR at 8 m/s to 12 m/s; and (b) blade Cp—TSR at Figure 14. Experimental results: (a) blade Cp—TSR at 8 m/s to 12 m/s; and (b) blade Cp—TSR at 14 m/s to 16 m/s. 14 m/s to 16 m/s.

Figure 14a shows our experimental results for wind speeds at 8 m/s, 10 m/s, and 12 m/s; those for wind speeds at 14 m/s and 16 m/s are in Figure 14b. Obviously, accompanied with the increase (a) (b) of wind speed is the greater percentage of power decrease. At the wind speed 8 m/s, the maximum Figure 14. Experimental results: (a) blade Cp—TSR at 8 m/s to 12 m/s; and (b) blade Cp—TSR at 14 m/s to 16 m/s.

Energies 2016, 9, 353

12 of 13

blade Cp would drop by about 10% while, at 14 m/s, the drop would increase up to 30%. From the experiments, the functioning of the PPC system is very clear that it may be applied to protect the system from aerodynamic overloading. 5. Conclusions Under operation at high wind speeds, the small-scale HAWT system will be subjected to aerodynamic overloading that may cause burning out of its generator and/or structural breakdown. There are several techniques to protect the system from the overloading condition. In this article, an innovative disk pulley mechanism is designed to protect the small-scale HAWT system by passive control of the blade pitch angle according to the rotation speed. The theoretical analysis was performed using our self-developed GUI software, called “HAWT-Perform”, which can be used to predict the blade performance. Through the theoretical analysis of HAWT-Perform, one may learn the significant performance change when blade pitch angle is adjusted by our disk pulley mechanism. For verification, experiments were carried out for a HAWT prototype with blades (using a SD8000 airfoil) designed by BEMT. From the experiments, one may learn the performance change from the use of our PPC system. This PPC system is featured with the advantages of great reliability, easy maintenance, and low cost. The present work simply determines a feasible approach to protect the small-scale HAWT; however, more extensive studies are required regarding the design parameters of the disk pulley mechanism to match different HAWT systems with other output powers. Supplementary Materials: The following are available online at www.mdpi.com/link, Video S1: Operation of passive pitch control system. Acknowledgments: The authors gratefully acknowledge the funding support by National Science Council (NSC 102-2221-E-006-290-MY3). This research is also partially supported by Research Center for Energy Technology and Strategy, National Cheng Kung University and Digisine Energytech Co., Ltd. Author Contributions: The author Y.C. Shiah planned the experiments and organized the research directions; Yu-Jen Chen performed the mechanism design, theoretical analyses. Conflicts of Interest: The authors declare no conflict of interest.

Abbreviations The following abbreviations are used in this manuscript: NCKU PPC PF

National Cheng Kung University Passive pitch control Pitch fixed

References 1. 2. 3. 4. 5. 6. 7.

Mc Garrigle, E.V.; Leahy, P.G. Quantifying the value of improved wind energy forecasts in a pool-based electricity market. Renew. Energy 2015, 80, 517–524. [CrossRef] Browne, O.; Poletti, S.; Young, D. How does market power affect the impact of large scale wind investment in “energy only” wholesale electricity markets? Energy Policy 2015, 87, 17–27. [CrossRef] Gillenwater, M.; Lu, X.; Fischlein, M. Additionality of wind energy investments in the U.S. voluntary green power market. Renew. Energy 2014, 63, 452–457. [CrossRef] Choudhry, A.; Arjomandi, M.; Kelso, R. Methods to control dynamic stall for wind turbine applications. Renew. Energy 2016, 86, 26–37. [CrossRef] Choudhry, A.; Leknys, R.; Arjomandi, M.; Kelso, R. An insight into the dynamic stall lift characteristics. Exp. Thermal Fluid Sci. 2014, 58, 188–208. [CrossRef] Yen, J.; Ahmed, N.A. Enhancing vertical axis wind turbine by dynamic stall control using synthetic jets. J. Wind Eng. Ind. Aerodyn. 2013, 114, 12–17. [CrossRef] Xie, W.; Zeng, P.; Lei, L.P. A novel folding blade of wind turbine rotor for effective power control. Energy Convers. Manag. 2015, 101, 52–65. [CrossRef]

Energies 2016, 9, 353

8. 9. 10.

11. 12. 13. 14. 15.

16. 17. 18. 19.

20. 21. 22. 23. 24. 25.

13 of 13

Xie, W.; Zeng, P.; Lei, L.P. Wind tunnel experiments for innovative pitch regulated blade of horizontal axis wind turbine. Energy 2015, 91, 1070–1080. [CrossRef] Jauch, C.; Hippel, S. Hydraulic-pneumatic flywheel system in a wind turbine rotor for inertia control. IET Renew. Power Generat. 2016, 10, 33–41. [CrossRef] Barlas, T.K.; van Wingerden, W.; Hulskamp, A.W.; van Kuik, G.A.M.; Bersee, H.E.N. Smart dynamic rotor control using active flaps on a small-scale wind turbine: Aeroelastic modeling and comparison with wind tunnel measurements. Wind Energy 2013, 16, 1287–1301. [CrossRef] Kragh, K.A.; Fleming, P.A.; Scholbrock, A.K. Increased Power Capture by Rotor Speed-Dependent Yaw Control of Wind Turbines. J. Sol. Energy Eng. 2013, 135, 031018. [CrossRef] Ekelund, T. Yaw control for reduction of structural dynamic loads in wind turbines. J. Wind Eng. Ind. Aerodyn. 2000, 85, 241–262. [CrossRef] Shariatpanah, H.; Fadaeinedjad, R.; Rashidinejad, M. A New Model for PMSG-Based Wind Turbine with Yaw Control. IEEE Trans. Energy Convers. 2013, 28, 929–937. [CrossRef] Lesic, V.; Vasak, M.; Peric, N.; Wolbank, T.M.; Joksimovic, G. Fault-tolerant Control of a Wind Turbine with a Squirrel-cage Induction Generator and Rotor Bar Defects. Automatika 2013, 54, 316–328. Lee, J.; Kim, J.; Kim, Y.H.; Chun, Y.H.; Lee, S.H.; Seok, J.K.; Kang, Y.C. Rotor Speed-based Droop of a Wind Generator in a Wind Power Plant for the Virtual Inertial Control. J. Electr. Eng. Technol. 2013, 8, 1021–1028. [CrossRef] Simley, E.; Pao, L.Y.; Frehlich, R.; Jonkman, B.; Kelley, N. Analysis of light detection and ranging wind speed measurements for wind turbine control. Wind Energy 2014, 17, 413–433. [CrossRef] Chiang, M.H.; Wang, C.S.; Chen, C.S. Intelligent Pitch Control for a 2MW Wind Turbine. Int. J. Fuzzy Syst. 2012, 14, 89–96. Li, S.C. Self-powered passive adaptive control of pitch angle and Betz-shaped wind tunnel. In Proceedings of the International Conference on Renewable Energies and Power Quality, Madrid, Spain, 4–6 May 2010. Hertel, J.; Nygaard, T.; Duque, E. Passive pitch control of small horizontal axis wind turbines. In Proceedings of the 42nd AIAA Aerospace Sciences Meeting and Exhibit, Aerospace Sciences Meetings, Reno, NV, USA, 5–8 January 2004; pp. 2004–1369. Barlow, J.B.; Rae, W.H.; Pope, A. Low-Speed Wind Tunnel Testing, 3rd ed.; Wiley-Interscience: New York, NY, USA, 1999. Burton, T.; Jenkins, N.; Sharpe, D.; Bossanyi, E. Wind Energy Handbook; John Wiley & Sons Ltd.: Hoboken, NJ, USA, 2001. Manwell, J.F.; McGowan, J.G.; Rogers, A.L. Wind Energy Explained—Theory, Design and Application; John Wiley & Sons Ltd.: Hoboken, NJ, USA, 2002. Lee, M.-H.; Shiah, Y.C.; Bai, C.-J. Experiments and numerical simulations of the rotor-blade performance for a small-scale horizontal axis wind turbine. J. Wind Eng. Ind. Aerodyn. 2016, 149, 17–29. [CrossRef] Hsiao, F.B.; Bai, C.J.; Chong, W.T. The Performance Test of Three Different Horizontal Axis Wind Turbine (HAWT) Blade Shapes Using Experimental and Numerical Methods. Energies 2013, 6, 2784–2803. [CrossRef] Bai, C.J.; Hsiao, F.B.; Li, M.H.; Huang, G.Y.; Chen, Y.J. Design of 10 kW Horizontal-Axis Wind Turbine (HAWT) Blade and Aerodynamic Investigation Using Numerical Simulation. Procedia Eng. 2013, 67, 279–287. [CrossRef] © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).