Download - Wind-tunnel Study on Aerodynamic Performance
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
1/31
1
WIND-TUNNEL STUDY ON AERODYNAMIC PERFORMANCE OF SMALL
VERTICAL-AXIS WIND TURBINES
J. J. Miau*1, S. W. Huang1, Y. D. Tsai1, S. Y. Liang1, C. H. Hsieh2, S. J. Chen3,
C. C. Hu4
, J. C. Cheng5
, and T. S. Leu1
1 Cheng Kung University, Taiwan2 Institute of Nuclear Energy Research, Taiwan
3Temple University, USA4Kao Yuang University, Taiwan
5Formosa University, Taiwan
ABSTRACT
Wind tunnel experiments were carried out to study the aerodynamic performance of
three vertical axis wind turbines (VAWTs) including a Darrieus VAWT, a Giromill
VAWT, and a helical VAWT. The performance curves regarding the power
coefficients against the tip speed ratios were reduced for the Darrieus and helical
VAWTs; whereas the reaction torques of the Giromill VAWT at different azimuthal
angles under static condition were measured and discussed. Moreover, the effect of
free-stream turbulence on the performance of the helical VAWT was studied by
having a turbulence generating grid installed at the inlet of the test section. As found,
the wind turbine actually performed better under the condition of high free-stream
turbulence intensity. On the other hand, the characteristics of unsteady flow around
the helical wind turbine were studied with a hot-wire probe positioned at the
peripheral of the wind turbine, under the condition of low free-stream turbulence
intensity, at two tip speed ratios.
Keywords: wind-tunnel experiment, VAWT, aerodynamic performance
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
2/31
2
1. INTRODUCTION
Strong growth of utilizing wind energy in the past decade [1] stimulates extensive
research efforts on the wind turbine technology nowadays. Among which, studies
on the small vertical-axis wind turbines (VAWT) have attracted a great deal of
attention, because of their potential applications in urban environment, for instance,
the idea of installing a small wind turbine on the roof of a building was explored [2].
Referred to such applications, a VAWT can be so small in physical size that its
full-scale can be fitted into a wind tunnel for testing. The advantage of using wind
tunnels for research and development is that one can validate the design specifications
in a relatively short turn-around time. A case of providing the detailed comparisons
between the wind-tunnel tests and the full-scale field measurements was reported by
Sheldahl [3].
Speaking of the performance test in wind tunnel, a wind turbine should be
examined under various operating conditions. Thus, the results obtained can serve
for validating the specifications. Motivated by this consideration, a series of research
efforts were set out by the present authors to study the aerodynamic performance of
small VAWTs using the experimental and numerical methods. This paper mainly
presents the results obtained by the wind-tunnel experiments with three wind turbines,
namely, a Darrieus VAWT, a Giromill VAWT, and a helical VAWT. The results
include the performance curves regarding the power coefficients reduced against the
tip speed ratios, the reaction torques measured with a wind turbine at different
azimuthal angles under the static condition, and the hot-wire velocity measurements
around the peripheral around a wind turbine. Additionally, discussion is made
concerning the effect of free-stream turbulence on the performance of the helical
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
3/31
3
VAWT and the unsteady flow around this wind turbine under the condition of low
free-stream turbulence intensity. The blockage effect due to the confinement of the
test-section walls of the wind tunnel also calls for attention, because the blockage
effect was identified as a critical factor from an uncertainty analysis of the power
coefficient and tip speed ratio.
2. The ABRI Wind Tunnel
Experiments were carried out in the ABRI (Architecture and Building Research
Institute) environmental wind tunnel situated in the Kuei-Ren Campus of National
Cheng Kung University. The wind tunnel is a closed-loop circuit, which has two test
sections in series. The first test section is 4 m (width) by 2.6 m (height) at the inlet,
and 36.5m in length. The second is 6 m (width) by 2.6 m (height) at the inlet and 21m
in length. [4] The first test section is long enough to have a thick boundary layer
developed at the downstream end, where the height of the test section is 3 m. In the
present study, the wind turbine was installed on a turntable immediately downstream
the inlet, centered at 2.9 m from the inlet, where the thickness of the boundary layer
was negligible compared to the dimensions of the test section. Therefore, the present
wind turbine was regarded as tested under the uniform flow condition.
According to the wind tunnel calibration results reported by Kao [5], the
maximum velocity achieved in the first test section flow was higher than the design
speed, 36 m/s [4]. The flow uniformity reduced at the inlet of the test section was
about 0.37%, and the turbulence intensity measured was less than 0.3%.
3. EXPERIMENTAL METHODS
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
4/31
4
In the present study, a wind turbine was situated on a turntable near the inlet of
the test section mentioned. The wind turbine was supported by a stainless steel strut of
circular cross section. See also Fig. 1 for a photo of a Darrieus wind turbine
installed in the test section. Note that the strut height should be appropriate, in order
that the wind turbine was located in the center region of the test section. The
coordinate system employed for the present study is given in Fig. 2, where x, y and z
denote the streamwise, spanwise and vertical directions, respectively. The origin, (x,
y, z) = (0, 0, 0) is denoted at the center of the wind turbine. In addition, is denoted
as the azimuthal angle associated with the wind turbine, where =0 is aligned in the
positive y direction.
In the present study, the power coefficient, Cp, of a wind turbine is defined below.
[6]
Pdenotes the power output of a wind turbine, measured by a DC electronic load
device; denotes the density of the free stream flow; V denotes the reference
velocity with respect to the wind turbine; and AS denotes the swept area of the wind
turbine.
Further, it was noted that the blockage effect caused by the presence of the wind
turbine in the test section could not be ignored in the determination of the power
coefficient, which can be seen below. Let the blockage effect be represented by a
parameter, t. Therefore, V in (1) can be corrected from the free stream velocity u,
measured by a Pitot tube at the inlet of the test section, as follows. [7]
V= u(1+t) (2)
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
5/31
5
in (1) can be expressed in terms of the pressure and temperature of the free
stream,P and T, respectively, by the ideal gas relation.
In (3), R is the universal gas constant. Combining the relations (1) to (3), Cp
can be expressed as follows.
q,udenotes the dynamic pressure based on the free stream velocity u. Therefore,
concerning the total uncertainty ofCp, it can be expressed as follows. [8]
As seen in (5), the term containing the blockage effect parameter is identified as an
influencing factor on the right-hand side. Similarly, with the tip speed ratio defined
below, where represents the angular velocity of the wind turbine,
the total uncertainty of the tip speed ratio can be expressed as follows.
In (7), it is also seen that the term containing the blockage effect parameter can be a
critical factor influencing the uncertainty of .
In the present study, the torque produced by a wind turbine was measured by a
torque meter, in-line connected to the shaft of the turbine. Literally speaking, the
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
6/31
6
torque generated by a wind turbine can be categorized into two kinds, namely, the
reaction torque and the rotary torque. The former is measured when the wind
turbine is in the static situation, whereas the rotary torque is measured with the wind
turbine in rotation. A non-zero rotary torque induces a change of the angular
velocity of a wind turbine, which can be seen in the following expression.
(8)
Qa, Qf , and Qem denote the aerodynamic torque, the loss due to the mechanical
friction, and the counter electromagnetic torque exerted by the electric generator,
respectively; Jrdenotes the inertia of the wind turbine, and is the rate of change
of the angular velocity mentioned. On the other hand, when is zero, the wind
turbine is rotating at a constant speed, which implies that no rotary torque is
generated.
In addition to the output power and torque measurements mentioned above, a
cross-type hot-wire probe was employed to gather the instantaneous velocity
information at the peripheral of a wind turbine. The velocity data obtained are of use
to examine the unsteady flow around the wind turbine.
In this study, a turbulence generating grid could be positioned at the inlet of the
test section to produce considerably high turbulence intensity downstream. The grid
was made of wood rods in a pattern of squared meshes, each of which was 0.3 m by
0.3 m in dimension. Each wood rod was of squared cross section, 0.09 m in width.
As calculated, the area blockage ratio of this grid was 49%. At 2.3 m downstream of
the grid, where was about the upstream edge of the wind turbine for test, the
turbulence intensity measured was 12 %. The measurements were conducted in the
absence of a wind turbine in the test section, for the free stream velocity up to 11 m/s.
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
7/31
7
The integral length scale of turbulent fluctuations at this streamwise location was
about 0.06m, equivalent to two times the mesh size.
4. EXPERIMENTAL RESULTS
4.1 The Results of a Darrieus Wind Turbine
The Darrieus wind turbine shown in Fig. 1, which was acquired from a local
manufacturer, was consisted of 3 turbine blades with the rated power at 400 Watts.
The maximum diameter of the turbine is 1.2 meter, called Dthe height, called h, is
1.2 meterthe chord length of a turbine blade, called C, is 90 mm. The solidity,
defined as NCh/As, is 0.225, whereN=3 represents the number of the turbine blades.
Figure 3 presents the power coefficient Cp against the tip speed ratio for the free
stream velocity, u, in a range of 6 to 13.1 m/s. Note that the reference velocity V
was corrected with t in (2), based on that the turbine blades were situated statically in
the test section; later, further discussion will be made concerning the validity of the t
value adopted. As seen in the figure, the maximum power output was normally
occurred at slightly larger than 2.5; the higher the flow speed, the higher the
maximum Cp value reduced. More specifically, the maximum Cp value reaches
about 0.2 at u =13.1 m/s, while it is only about 0.05 at u =6 m/s. Therefore, one
can see that the aerodynamic performance strongly depends on the incoming flow
speed. Moreover, as noted in the figure, few data points were obtained at the tip
speed ratios lower than 2.5. This is due to a fact that, for this wind turbine the wind
power was difficult to be drawn at low tip speed ratios.
4.2 The Results of a Giromill Wind Turbine
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
8/31
8
Figure 4 presents a photo of the self-made Giromill (H-type) wind turbine
installed for test. The wind turbine employed, 1.2 m in diameter and 1.2 m in height,
was consisted of three straight blades, each of which was a NACA0015
cross-sectional shape, 90mm in chord length. To study the starting characteristic of
this wind turbine, measurements of the reaction torque in the static situation were
carried out for u =3.7, 4.6 and 5.5 m/s, respectively. [9] The results obtained are
presented in Fig. 5, in terms of the distribution curves of torque values in Nm against
. Interestingly noted is that the turbine blades generated negative torques at in the
neighborhood of 50, 170 and 290, respectively. This observation explains why the
self-starting problem of a vertical axis wind turbine is intrinsically existed. By the
double multiple-streamtubes model [10], Hsieh [9] further estimated the starting
torque under the flow condition of u=5.5 m/s. A comparison of the experimental
data and the results obtained by the double multiple-streamtubes model is given Fig. 6.
The two curves shown in the figure indicates that not only the angles where the
negative torques take place, but also those of the maximum positive torques occurred,
are well coincided. This comparison gives a strong support to using the double
multiple-streamtubes model for the prediction of the starting torque.
To alleviate the self-starting problem, Gorlov [11] proposed a design of the
three-dimensional helical blade for hydraulic reaction turbines. Later, this idea was
extensively applied in wind turbines, one of which will be seen in the next section.
On the other hand, the ideas of pitch control have been proposed for the wind turbines
with straight blades [12-14]. The control laws were aimed to overcome the
self-starting problem under the static condition, as well as improve the aerodynamic
performance at different tip speed ratios.
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
9/31
9
4.3 The Results of a Helical Wind Turbine
Figure 7 presents two photos of a helical wind turbine situated in the test section
of the wind tunnel, one of which is with no turbulence generating screen situated at
the inlet of the test section, and the other is with the presence of the grid. The wind
turbine with the rated power at 400 Watts was acquired from a local manufacturer.
The wind turbine is featured with 1.25 m in diameter and 1.4 m in height, consisted of
three twisted blades, each of which is twisted 30 with respect to the vertical axis and
0.27 m in chord length. The solidity of this wind turbine estimated is 0.605, which is
noted significantly higher than the two wind turbines studied earlier. The
aerodynamic performances of the wind turbine were then studied with respect to the
two inlet conditions mentioned.
In calculating Cp and, the reference velocity, V, was taken to be the velocity
measured by a hot wire at (x, y, z)= (-0.7 m, 1.4 m, 0), indicated in Fig. 2 with an
open circle symbol. Further, to examine the flow uniformity under this flow
condition, additional hot-wire velocity measurements were made at y=0, 0.67 m and
-0.67 m, at the same streamwise location and z=0, without the presence of a wind
turbine. It was found that the mean velocities measured were well coincided.
Based on the wind-tunnel data obtained, the performance curves subjected to two
inlet conditions are shown in Fig. 8. Notable differences learned from a comparison
of the two plots in Fig. 8 are described below. Under the condition of high
free-stream turbulence intensity, the wind turbine appears to perform better, as far as
the maximum Cp values reduced are concerned. Moreover, it is noted that under this
condition, the tip speed ratios corresponding to the maximum Cp values reduced are
slightly lower than those found under the condition of low free-stream turbulence
intensity. Also noted is that under the condition of high free-stream turbulence
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
10/31
10
intensity, the Cp curves obtained at V = 5.5- 11.1 m/s appear to be almost collapsed
together, a strong indication that the performance of the wind turbine is insensitive to
the flow speeds tested. On the other hand, under the condition of low free-stream
turbulence intensity, the Cp curves obtained at different wind speeds show significant
scatter for lower than 2.
It is noted that the maximum Cp value obtained in the situation of high
free-stream turbulence intensity is higher than that obtained in the situation of low
free-stream turbulence intensity by about 0.01. This difference is noted rather
significant and deserved further discussion. Following Homicz [15], the
performance of a wind turbine in a turbulent stream can be analyzed as follows.
Theoretically, the output power of a wind turbine is proportional to the cubic of the
incoming velocity, say, the reference velocity V.
P(1/2)V3AS (9)
Meanwhile, considering a highly turbulent free stream, let the turbulent fluctuations in
the streamwise direction be u, and the time-mean reference velocity be
V .
Therefore, the instantaneous reference velocity can be expressed as
V= V +u (10)
Plug this relation in (9),
P
(1/2)[ V3
+3u V2
+3u2
V +u3
]AS (11)
By taking a time average of (11),
P(1/2)[ V3+3 2'u
V ] (12)
Here, P denotes the time-averaged output power. As a result, due to the presence
of the second term on the right-hand-side of (12), the time-averaged power coefficient
is increased by an amount of 3( 2'u /
V2).
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
11/31
11
For the present case of high free-stream turbulent intensity, ( 2'u /
V2)1/2 =0.12,
the amount of increase in Cp is estimated about 0.04, which is about four times the
increase of Cp noted in the experimental data. Hence, Homiczs argument [15]
gives a support to the trend seen that the efficiency of a wind turbine gets improved in
a turbulent stream. Nevertheless, the reasoning apparently is too simplified, without
taking into account the effect of free-stream turbulence on the unsteady flow
phenomenon around a wind turbine.
In Fig. 8b the Cp values at low tip-speed ratios are noted relatively insensitive to
Reynolds number, in comparison with those in Fig. 8a. Since at low tip-speed ratios,
flows around the rotor blades would experience large variations in angle of attack
with respect to the azimuthal angle, the dynamic stall phenomenon [16-18] is
anticipated to come into play. Based on the experimental observations above, one
can further argue that the presence of free-stream turbulence cause the dynamic stall
phenomenon less sensitive to the variations of Reynolds number.
To gather the instantaneous velocity information of the unsteady flow field,
hot-wire measurements were performed at 0.1 m away from the edge of the wind
turbine in the plane of z=0. The locations of hot-wire measurements are indicated in
Fig. 2 with open circles, i. e., =0, 30, 90, 150, and 180.
Figure 9 presents the streamwise (u) and vertical (v) velocity traces obtained at
=0, 90, and 180, subjected to = 1.73 and 2.58 while the free stream was at low
turbulence intensity. Also included in each figure are the signal traces obtained from
an optical sensor situated at =0 for phase reference. Note that in each plot the
horizontal axis is scaled with a normalized time, t/T, where Tdenotes the time period
of one revolution of the wind turbine. As seen from the output signal traces of the
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
12/31
12
optical sensor, the time length Tspans over three consecutive square waves due to the
three rotor blades passing over the optical sensor.
In Fig. 9a, the velocity signal traces obtained at =0 show that prior to a turbine
blade reaching this angular location, the streamwise velocity shows a strong
acceleration followed by deceleration. As the leading edge of the wind turbine blade
reaching this angular location, the vertical velocity is decelerated to the lowest value,
about zero. After the trailing edge of the turbine blade passing over, the streamwise
velocity is leveled off. Note that the vertical velocity measured at this location
appears positive always, which can be explained with the Conservation laws below.
Since the mass flux (momentum flux) of the incoming flow should be greater the
mass flux (momentum flux) downstream of the wind turbine, due to the extraction of
wind energy, a portion of the incoming fluid tends to be displaced outward in the
lateral direction. By the same argument, at =180, seen in Fig. 9c, the vertical
velocity measured appears to be negative most of the time, except for a period of time
between two turbine blades passing over this location, the flow is dominated by the
wake motions. In Figs. 9a, b and c, the time instants of a turbine blade reaching the
measured angular locations =0, 90 and 180, respectively, are marked by T* for
reference.
In Fig. 9b, the velocity signal traces obtained at = 90 basically show that the
streamwise velocity reaches the lowest value when a blade reaching this angular
position, corresponding to a situation of flow impinging on a solid surface, meanwhile
the vertical velocity measured is about zero, and gets increased as the blade passing
over the location. Moreover, it is interesting to point out that in Figs. 9a and b, the
velocity signal traces corresponding to the two tip speed ratios are almost coincided.
This observation can be reasoned that the unsteady flow experienced by a rotor blade
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
13/31
13
as it advancing from = 0 to 90 is of an attached-flow type basically. Alternately
speaking, the process of vortex shedding has not yet commenced, thus little variations
are seen with respect to the two tip speed ratios.
On the other hand, in Fig. 9c it is seen that either the u orv signal traces show
significant differences with respect to the two tip speed ratios. Moreover, in every
signal trace one can see intermittent fluctuations, which are attributed to turbulent
fluctuations associated with the vortices shed, as a result of the dynamic stall process.
It is anticipated that the phenomenon of dynamic stall and subsequently vortex
shedding take place when a turbine blade is travelling through the azimuthal region of
= 90 to 180. Recently, Cheng [19] conducted a numerical analysis on studying
the three-dimensional flow field around a wind turbine, whose geometric
configuration is similar to the present one. Cheng [19] presented and discussed the
numerical results obtained at = 2 and 3. In both cases, the results clearly indicate
that vortex shedding take place as a turbine blade advancing from = 90 to 180.
It is also worthwhile to mention that the phenomenon of vortex shedding due to the
dynamic stall process of flows around the VAWT blades was unveiled by the flow
visualization experiments performed in the water-channels [20, 21], whose Reynolds
numbers are considerably lower than those of the present wind tunnel experiments.
5. DISCUSSION
5.1 The blockage effect
A primary interest of this study is to obtain the performance curves of a wind
turbine, namely, the curves of Cp versus , under various operating conditions.
Along this consideration, the accuracy of the Cp and values reduced from the
measured data deserves further discussion here. According to (5) and (7), one can
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
14/31
14
identify that the major contribution to the uncertainties ofCp and, can be due to the
blockage effect, represented by t. Moreover, the error or uncertainty resulted in Cp
would be three times larger than t. Thus, the Cp values shown in Fig. 3 or 8 are
critically dependent upon the reference velocity, V, determined. On the other hand,
since the wind turbine was situated in the test section, the blockage effect could not be
ignored, nor could it be estimated by a simplified model of flow over fixed turbine
blades. [9]
Physically, it can be argued that the blockage effect be a function of the tip speed
ratio. To clarify this issue, recently a numerical simulation was carried out by the
present authors to examine the blockage effect of a wind turbine in a confined channel
at different tip speed ratios. A two-dimensional physical model based on the
schematics shown in Fig. 2 was employed, but the diameter of the wind turbine was
1.2 m. The NACA 0015 airfoil with chord 0.15m was chosen as the wind turbine
blade. Two types of boundary conditions were considered in the simulation, i. e. the
free and solid wall boundary conditions, with the free stream velocity ufixed at 6
m/s. Table 1 shows the time-averaged x- and y-component velocities, called u and v,
respectively, at the point indicated in Fig. 2 where V was measured, subjected to =1,
1.5 and 2. In the free boundary case u is decreased slightly, whereas v is increased
considerably as gets increased. On the other hand, in the solid wall boundary case,
both of u and v are increased as gets increased. At the measurement point, u is
always greater than the free stream velocity, irrespective of the tip speed ratios.
Table 2 lists the values of V, where V=22
vu , and the blockage effect
parametert with respect to the tip speed ratios. For the free boundary condition,
the velocity at the measurement point, V, is actually smaller than u. On the
contrary, for the solid wall boundary condition, V, is greater than u, irrespective of
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
15/31
15
the tip speed ratios. Notably, the blockage effect parameter, t is increased from
1.35 to 4.56 as the tip speed ratio is increased from 1 to 2.
5.2 The Reynolds number effect
In Figs 3 and 8, the Reynolds number effect to the characteristics of
aerodynamic flow around a wind turbine can be realized by comparing the Cp-
curves obtained at different free stream velocities. For instance, based on the
appearances of the Cp- curves in Figs. 3 and 8, one can say immediately that the
performance of the Darrieus VAWT is much more sensitive to the Reynolds number
than that of the helical VAWT.
By defining the Reynolds number based on the free stream velocity, u, and the
chord length of a turbine blade, C, the Reynolds numbers in the present study are
noted to fall within a range of 103 to 105. Within this Reynolds number range, the
phenomena of boundary layer transition and separation on a turbine blade are known
very sensitive to the influencing parameters, including Reynolds number, surface
roughness and free-stream turbulence intensity. Moreover, given the fact that a
wind turbine is in rotation, the turbine blades actually experiences various angles of
attack in time, which depend on the tip speed ratio, thus in most of the situations the
dynamic-stall phenomenon comes into play, and makes the prediction of aerodynamic
forces even more difficult. In summary, the results presented in this paper evidence
the importance of the effects of Reynolds number and free-stream turbulence.
6. CONCLUDING REMARKS
The experimental results reported in this paper show that the aerodynamic
performance of a small VAWT can be studied in a wind tunnel, by providing the Cp-
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
16/31
16
curves, the reaction torque of a wind turbine, and the characteristics of the unsteady
flow field using hot-wire velocity measurements. Meanwhile, there are key issues of
concern learned from this study, namely, (1) the accuracy of the Cp and values
reduced, which is critically dependent upon the correction of the blockage effect, (2)
the Reynolds number effect, to which the characteristics of aerodynamic flow over a
turbine blade is sensitive, and (3) the free-stream turbulence effect, which could better
the performance of a wind turbine and make it less sensitive to Reynolds number.
These issues are being under investigation by the authors. More results will be
reported in the future.
ACKNOWLEDGMENT
The authors would like acknowledge the funding support by National Science
Council, Taiwan, under the contract number 98-3114-E-006-007 for this research
work. Support of the ABRI Wind Tunnel Laboratory in conducting the experimental
work of this study is also greatly appreciated.
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
17/31
17
REFERENCES
[1] World wind energy report 2009, World Wind Energy Association, March (2010).
[2] Bussel, G. J. W., van Mertens, S., Polinder, H. and Sidler, H. F. A., TURBY:
concept and realization of a small VAWT for the built environment, Paper
presented at the EAWE/EWEA Special Topic conference The Science of making
Torque from Wind, Delft, The Netherlands ISBN 90-764768-10-9, 19-21 April,
pp. 509-516 (2004).
[3] Sheldahl, R. E., Comparison of field and Wind Tunnel Darrieus Wind Turbine Data,
SAND80-5469 (1981).
[4] Miau, J. J., Chou, J, H., Cheng, C. M,, Chu, C. R., Woo, K. C., Ren, S. K., Chen, E.
L,, Hu, C. C., and Chen, Z. L., Design aspects of the ABRI wind tunnel, The
International Wind Engineering Symposium, IWES 2003, November 17-18, 2003,
Tamsui, Taipei County, Taiwan (2003).
[5] Kao, Y. M. Calibration of the ABRI environment wind tunnel and experimental
study of 2-D bluff-body aerodynamic flows, M. S. Thesis, Department of
Aeronautics and Astronautics, National Cheng Kung University, Tainan, Taiwan,
ROC (2005).
[6] Manwell, J. F, McGowan, J. G., and Rogers, A. L., Wind energy explained theory,
design and application, 2nd ed., John Wiley & Sons (2009).
[7] Rae, Jr. W. H., and Pope, A.,Low-Speed Wind Tunnel Testing, 2nd ed., John Wiley
& Sons (1984).
[8]ISO guide to the expression of uncertainty in measurment. ISO, 2nd ed. (1995).
[9] Hsieh, C. H., Experimental and Numerical Studies of Torque and Power
Generation in a Vertical Axis Wind Turbine, Master thesis, Department of
Aeronautics and Astronautics, National Cheng Kung University (2009).
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
18/31
18
[10] Islam, M., Ting, D. S. K., and Fartaj, A., Aerodynamic Models for Darrieus-Type
Straight-Bladed Vertical Axis Wind Turbines, Renewable & Sustainable Energy
Reviews, Vol. 12, No. 4, pp. 1087-1109 (2008).
[11] Gorlov, A. M., Unidirectional Helical Reaction Turbine Operable under
Reversible Fluid Flow for Power System, US patent no. 5,451,137 (1995).
[12] Lazauskas, L., Three Pitch Control System for Vertical Axis Wind Turbines
Compared, Wind Engineering, Vol.16, No.5, pp. 269-282 (1992).
[13] Paraschivoiu I., Trifu O., and Saeed F., H-Darrieus Wind Turbine With Blade
Pitch Control, International Journal of Rotating Machinery, Vol. 2009, Article ID
505343, 7 pages (2009).
[14] Chen, S. J., Chen, Z., Biswas, S., Miau, J. J., and Hsieh, C. H., Torque and power
coefficients of a vertical axis wind turbine with optimal pitch control, ASME
2010-27224, ASME 2010 Power Conference, July 13-15, Chicago (2010).
[15] Homicz, G. F., Numerical Simulation of VAWT Stochastic Aerodynamic Loads
Produced by Atmospheric Turbulence VAWT-SAL Code, Sandia National
Laboratories, SAMD91-1124 (1991).
[16] McCroskey, W. J. Unsteady Airfoil, Annual Review of Fluid Mechanics, No. 14,
pp. 285-311 (1982).
[17] Carr, L. W., Progress in Analysis and Prediction of Dynamic Stall, Journal of
Aircraft, Vol. 25, No. 1, pp. 125 (1988).
[18] Paraschivoiu, I. and Dsy, P., Aerodynamics of Small-Scale Vertical Axis Wind
Turbine, Journal of Propulsion and Power, Vol. 2, May-June, pp. 282-288 (1986).
[19] Cheng, P. L. Three-Dimensional and Arm Effects on Aerodynamics Simulation
for the Slant-H-Rotor Vertical Axis Wind Turbine, Master Thesis, Department of
Mechanical Engineering, National Central University, Chung-Li, Taiwan (2011).
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
19/31
19
[20] Brochier, G., Fraunie, P., Beguier, C., and Paraschivoiu, I., Water Channel
Experiments of Dynamics Stall on Darrieus Wind Turbine Blades, Journal of
Propulsion, Vol. 2, pp. 445-449 (1986).
[21] Fujisawa, N., and Takeuchi, M., Flow Visualization and PIV measurements of
Flow Field around a Darrieus Rotor in Dynamic Stall, Journal of Visualization,
Vol. 1, No. 4, pp. 379-386 (1999).
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
20/31
20
Table 1 Comparison of the numerical results obtained at = 1, 1.5 and 2 subjected to the
free and solid wall boundary conditions, regarding the time-averaged u and v velocities
at the point where V was measured in the wind tunnel experiments. (unit: m/s)
Boundary
Condition Type
=1 =1.5 =2
u v U v u v
Free BC 5.845 0.536 5.822 0.814 5.811 1.178
Solid Wall BC 6.079 0.165 6.170 0.257 6.265 0.328
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
21/31
21
Table 2 Comparison of the numerical results obtained at = 1, 1.5 and 2 subjected to the
free and solid wall boundary conditions, regarding the velocities ofV, where
V=22
vu , and the blockage effect parametert. (unit: m/s)
Boundary
Condition Type
=1 =1.5 =2
V t() V t() V t()
Free BC 5.870 -2.18 5.879 -2.02 5.929 -1.18
Solid Wall BC 6.081 1.35 6.175 2.92 6.274 4.56
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
22/31
22
Fig. 1 The Darrieus wind turbine installed, a view taken from the inlet of the test
section.
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
23/31
23
Fig. 2 A schematic drawing of the wind turbine at the plane of z=0, and the coordinate
system employed.
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
24/31
24
Fig. 3 The Cp versus curves of the Darrieus wind turbine studied at u = 6 to 13.1
m/s.
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
25/31
25
Fig. 4 A photo of the self-made Giromill wind turbine.
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
26/31
26
Fig. 5 The wind-tunnel measurement results of the reaction torque versusat u =3.7,
4.6 and 5.5 m/s.
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
27/31
27
Fig. 6 Comparison of the measured and estimated reaction torque versus at u
=5.5 m/s. [9]
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
28/31
28
Fig. 7 The photos of the helical wind turbine studied in the situations: (a) without and
(b) with a turbulence generating grid situated at the inlet of the test section.
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
29/31
29
(a)
(b)
Fig. 8 The Cp versus curves of the helical wind turbine studied in the situations: (a)
without and (b) with a turbulence generating grid situated at the inlet of the test
section.
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
30/31
30
(a)
(b)
T*
T*
T*
T*
T*
T*
T*
T*
T*
T*
T*
T*
-
8/22/2019 Wind-tunnel Study on Aerodynamic Performance
31/31
(c)
Fig. 9 Streamwise and vertical velocity traces obtained by an X-type hot-wire around
the helical wind turbine at (a) =0, (b) 90, and (c) 180 subjected to the incoming
flow at low turbulence intensity and two tip speed ratios. T* indicates the time
instant when the leading edge of a turbine blade reaches the angular location
measured.
T*
T*
T*
T*
T*
T*