You are on page 1of 12

DESIGN AND SIMULATION OF HIGH EFFICIENCY AND LOW

ENVIRONMENTAL IMPACT OF VERTICAL AXIS WIND


TURBINE (VAWT) BLADE

Student: Muhamad Fahmi bin Mazlan


UiTM ID: 2012391515
Supervisor: Dr. Nor Afifah binti Yahaya

ABSTRACT

VAWT investigation was initially dropped out in the late 1970’s, with the HAWT system improvement and
modification. It was until recently researchers found that VAWT have highlighted advantages which HAWT do not
like they can rotate in any oncoming direction of wind and also requires no yaw mechanism to rotate [1]. However,
problems arise with the selection of VAWT blade design; which related to the high rotational speed that can endanger
wildlife, induce noise and at the same time also produce negative torque that can affect the efficiency of the VAWT
system. The project intended to solve the environmental issues problem by diagnosing characteristics of VAWT blade
type matching with Malaysia average wind speed. And as Malaysia average wind speed is low (below 4m/s)
throughout the year(Chong, Fazlizan, Poh, Pan & Ping, 2012), Savonius type was selectively made as it is always
able to self- rotate. New modification of geometrical shape of Savonius blade was introduced and compared to the
original Savonius blade by numerical analysis software, ANSYS. Coefficient of torque for each 10 o and 5o angle of
blade spacing interval were obtained in the simulation for both designs and it was proven that the proposed blade
induced more torque (reduced negative torque). The best no. of blade ranging from 4, 6, and 8 was found to be at 6,
which gave highest average torque coefficient and may achieve up to 0.5 compared to the proposed blade that only
ranging at 0.1. These values were then analyzed and discussed with the contours of velocity and pressure distribution
plotted to verify the results.

1.0 Introduction

Malaysia’s average wind speed throughout the year is low (average at 1.5 to 2 m/s) as it is placed in between the
equatorial zone. In addition, two major issues that was intended to solve are environmental impact and at the same
time efficiency problem. Therefore, to implement wind energy in Malaysia, study and consideration of various types
of VAWT system needs to be done in order for wind turbine to be workable in Malaysia’s condition. VAWT has 2
categories based on its driven types; namely the Savonius (drag driven) and Darrieus (lift driven).

Savonius type basically is a drag-driven type of wind turbine; where wind speed strikes the upfront
(hemisphere like cross section) facing of the blade, making it rotates. In short, it very much related to the torque
induced on the center of the wind turbine as it rotates. For the darrieus type, it is based on aeroplane wing concept;
that are lift driven type. The blade rotates opposing the direction of wind it pass through, where lifting forces are
induced when wind is cut by the aerodynamic shape of the wind turbine blade itself.
Comparing both types, lift types of VAWT (Darrieus) is workably to self-start in higher wind speed rather
than Savonius. This is because for Darrieus to be rotate itself, it needs high wind speed to create enough lift on the
back of the aerodynamic design of the wind turbine to rotate. But such high speed will induce high rotational rate
and produce noise as well as safety issues will be put aside. For drag driven type of Savonius, it is not a problem
since the design itself can rotates from drag force induced on the upfront surface facing the wind direction in a low
wind speed. In short, If the shape is designed as it is for optimization and efficiencies, safety considerations were to
be put aside in the design. So, such design, if installed, need to be installed in such a way that it is isolated from outer
surrounding. This in turns will affect cost or efficiencies or both in order for safe environment for wildlife to be met.

For Savonius type of rotor, it performs best on low wind speed area mainly because given low wind speed,
it manages to create enough drag force for the turbine to self-rotate. Unfortunately when rotating, it suffers from
problem of negative torque attacking the back-facing (back of the cup) of the blade profile, making the revolution
per minute limited. That’s why researches are done to eliminate the negative drag [2], This initially solve the problem
of negative torque of that and only that direction of wind speed only, making it unreliable when it comes to other
incoming direction of wind. For Darrieus type of rotor, they work at best on high power output per rotor weight [3],
making it can rotates faster given higher wind speed. The only thing this blade design lack off is the ability to self-
start. The blade design itself also will suffer from flow interference during rotation and at self-start position [4]. This
will also provide negative torque on the blade during rotates or self-start. Continuous blade adjustment is one of the
proposed designs done to mitigate the problem, but yet weight of the rotor blade will get affected and gearing system
may affect the rotational speed and induced negative torque.

In conclusion, when trying to implement wind turbine for generation of electricity in Malaysia Savonius
blade type is the best selection yet. This project intended to design new geometrical shape of wind turbine blade
using CAD software. Using ANSYS as the numerical analysis software to animate the flow; steady and time
dependent simulation were to be done. Coefficient of torque at each 5o and 10o angle of blade spacing were tabulated
for both design to distinctively compare the different values of torque induced and hence verify the proposed blade
was indeed better than the original one. Other manipulation variable was introduced like number of blade in 4, 6, and
8. Simulation also has been done for all sets of number of blade in order to determine which sets turns out to be the
best setting that could improve the wind turbine efficiency. Contours of pressure and velocity distribution done helps
to verify the values of torque coefficient obtained. In short, this paper would verify best set no. of blades with the
best design; whether the original or the proposed one.

2.0 Methodology

In doing design and simulation of newly blade design, it adopts the Computational Fluid Dynamics (CFD) approach
in which the research starts with Pre-Processor, Solver (Fluent) and Post Processing Stage, where all results was
obtained. As direction of air cannot be acknowledged and what is the orientation of the rotor blade at that particular
time, simulation were intended to be done at interval, within spacing of one blade from the other.

2.1 Pre-Processor Stage

In this stage, original and proposed blades were designed in CAD software in the same nominal dimension for both
design.(Figure 2.1)The blade then had been brought into the meshing in ANSYS: Meshing, where parameters of
minimum orthogonal quality at 0.2 and above and maximum skewness at 0.95 and below needs to be met.(Figure 2.3)
The last element that needs to be done is the boundary condition, characterized by the inlet and outlet far-field in
ANSYS : Design Modeller (Figure 2.4). All sets of blade number (4, 6, and 8)(Figure 2.2) with sets of angle interval
spacing (Table 2.1) were setup in this stage.
Table 2.1: Number of Simulation Intended to be done at each Number Of Blade

4 blade 6 blade 8 blade


Orientation angle 360/4 = 900 360/6 = 600 360/8 = 450
spacing
Different Torque 0 – 80 (0), interval at 0 – 55 (0), interval at 50 0 – 40 (0)interval at 50
Coefficient Obtained 100
Total angle to be 9, 12, 9,
simulated (0, 10, 20, 30, 40, 50, (0, 5, 10, 15, 20, 25, 30, (0, 5, 10, 15, 20, 25, 30,
60, 70, 80) 35, 40, 45, 50, 55) 35, 40)

Figure 2.1: Geometrical Construction using CAD Software for Original (left) and the Proposed blade(right)

Figure 2.2: Sketch of Regular Rotor Blade 4, 6 and 8 no. of Blades


Figure 2.3: Highlighted Meshing Parameters Met (left) with close up Meshing of Original 6 Savonius Blade

Figure 2.4: 6 Named Selection made for Boundary Condition


2.2 Solver (Fluent)

This was the stage where all the input was inserted into the Solver for calculation later on. The important inputs in
Solver (Fluent) were the Viscous Model and the Input of wind speed as well as the angular velocity of the blade. This
is to ensure that torque coefficient can be produced along with plotting contours of velocity and pressure distribution.

2.2.1 Turbulence, Viscous Model


The turbulence model selected is K-Epsilon, which includes two-additional transport equations model to
account for history effects made by the turbulent energy. In short k is the turbulent kinetic energy transported while
for the epsilon, it is the turbulent dissipative energy. This model also is widely used for industrial purposes as they
take shorter time in order for the solution to be converged and can be said to be accurate to compare between two
designs in the simulation. Futhermore, as said by S. Derakshan and A. Tavaziani, they concluded that K-Epsilon
model can be said to be similar in low speed with the other models like SST K-Omega and they perform better at high
wind speed [5]. In conclusion, K-Epsilon model was selected to be the turbulence model in this research.

2.2.2 Inputs and Others


There were 2 vitals input that needs to be set, that are the inlet velocity and angular velocity of the blade.
Input velocity was set to be constant at 2m/s and 4rad/s for the angular velocity of the wind turbine. Interface meshing
was linked between the two non-conformal meshes in order for the data to flow between the two domains.

2.2.3 Post-Processor
There were 3 types of data sets taken from the simulation; namely from the contour plots of pressure and
velocity with also from the torque coefficient value as shown in figure below.

Figure 2.5: Screenshot obtained from Fluent: Post-Processor Stage showing Contours of Pressure (left) and Velocity (right) as
well as torque coefficient value at the bottom
3.0 Results and Discussion
Analysis of the simulations was discussed starting from analyzing the torque coefficient value. Comparing the
values, it was found that not a single value from the proposed design at any orientation recorded

3.1 Torque Coefficient Value

Table 3.1: Torque Coefficient Values

No. of Blade : 4 No. of Blade : 6 No. of Blade : 8

Coefficient of Rotor Blade Coefficient of Rotor Blade Coefficient of


Torque Orientation Torque Orientation Torque
Blade Blade Rotor Blade Rotor
Rotor Design Design
Design Ori Proposed Ori Proposed Ori Proposed

Rotor Rotor Rotor


Orientation Orientation Orientation

0 0.2519 0.3563 0 0.1476 0.3461 0 0.2352 0.3775

10 0.2866 0.3594 5 0.2341 0.3462 5 0.1999 0.4086

20 0.3287 0.4040 10 0.3127 0.3146 10 0.2583 0.4143

30 0.2897 0.3032 15 0.3138 0.3441 15 0.2319 0.4054

40 0.2173 0.2355 20 0.3221 0.3646 20 0.2328 0.3655

50 0.1603 0.2297 25 0.2010 0.4073 25 0.2149 0.3625

60 0.1439 0.3323 30 0.1657 0.3885 30 0.2156 0.3621

70 0.2047 0.3889 35 0.1611 0.4337 35 0.2024 0.3222

80 0.2418 0.3785 40 0.1318 0.4901 40 0.1791 0.3058

- - - 45 0.1268 0.4650 - - -

- - - 50 0.1201 0.4069 - - -

- - - 55 0.1366 0.3559 - - -

Average 0.2361 0.3320 Average 0.1978 0.3886 Average 0.2189 0.3691


From Table 3.1, it was found that not a single value from proposed blade at any sets of blade number compute torque
coefficient lower than the original one, verifying that proposed blade produce more torque coefficient when compared
to the original blade design at all orientation of the rotor blade. 6 number of blade in a wind turbine shows the highest
value of Torque Coefficient among other sets of blade number, suggesting best number ofProposed Bladethe can
blade between three.
This can be further seen in contour plots of velocity (Figure 3.1) and pressure (Figure 3.2) continuously
below. receive
high velocity of
incoming air despite of
Original having build-up low
velocity around the
Proposed profile

Air
Direction
Air
Direction

Figure 3.1: Contours of Velocity for Original (top) and Proposed Blade (bottom)
Original

Air
Direction

Large negative pressure


induced, indicating
proposed blade was
having high positive
torque
Proposed

Air
Direction

Figure 3.2: Contours of Pressure for Original (top) and Proposed Blade (bottom)

From both contours of pressure and velocity, it can be said that newly proposed design definitely induced higher
torque with the indication of high velocity that manages to get through the front facing blade despite of low velocity
build-up and also negative pressure induced based on Figure 3.2, indicating that proposed blade experiencing more
positive torque compared to the original one. In the real situation, newly blade design will be experiencing high angular
velocity compared to the original blade design.
3.2 Number of Blades

As proposed blade had been proven to be better than the proposed one, investigation to the number of blades was
hereby proceed in order to find the best number of blades possible that could improve overall efficiency of the wind
turbine. Set of 4, 6, and 8 number of blades were taken into the simulation and from the Table 3.1 tabulated earlier,
graph summarizing the proposed blade torque coefficient values was tabulated as below in Figure 3.3.

0.6

0.5

0.4
Torque Coefficinet

0.3 blade 4
blade 6
0.2 blade 8

0.1

0
0 10 20 30 40 50 60 70 80 90
Blade Angle

Figure 3.3: Graph of Blade Angle vs Torque Coefficient for Proposed Blade VAWT with Different Number of Blade

From the graph shown above, 4 number of blade was fluctuating throughout its blade angle, with the 6 blade number
slightly decrease before increasing up to its highest torque coefficient. For the 8 blade number, it experienced slight
increase in the value before decrease to its lowest value of torque coefficient. 4 number of blade could affect the
efficiency of the newly proposed wind turbine blade as at the angle of 55 0, the torque coefficient value falls to the
lowest point ever recorded for the newly blade design. Therefore, as 6 blade number was experiencing the highest
torque coefficient with medium decrement of torque coefficient values between the three, it can be said that 6 number
of blades was the best sets of blade number in this research.
This can be explained by observing the plot contours of velocity and pressure distribution of these new blade setting
at different number. Random blade angle were selected and plots of contour of velocity and pressure was obtained as
in Figure 3.4 and Figure 3.5 below.

4 blades

Comparing between the three,


blade 6 has the larger area where Blade number 6 has the
high wind speed can still lesser back facing surface,
penetrate through the turbulence inducing the least
that dissipates in front of the 6 blades negative torque between
blade.
the three sets of blade
This will help the wind turbine number. This thus giving
to gain additional torque, highest torque coefficient
contributing to high angular
between the three.
velocity of the wind turbine

8 blades

Figure 3.4: Contours of Velocity for 4,6 and 8 number of blade


4 blades

6 blades
Blade number at 6 shows the most
value of negative pressure, giving
indication that the wind turbine
was experiencing higher positive
torque than the other 2.

8 blades

Figure 3.5: Contours of Pressure for 4,6 and 8 of blade number


4.0 Conclusion and Recommendation

The study carried out has been successfully met the goals of this project, that were to design a new blade, simulating
through ANSYS both time-dependent and steady type and also comparing the proposed blade with the original one.

Proposed blade was introduced with its irregular shape and more in length in order to receive more incoming air from
various directions and also separating air easily on the back facing in order to reduce negative torque. Simulation of
the proposed blade was compared to the simulation of the original one, based on 3 key features of the simulation; that
were torque coefficient, contour plots of velocity and pressure distribution.

From the simulation, proposed blade induced higher torque at each blade angle, indicating that the proposed bladed
does help improving the efficiency of the wind turbine. This can be explained through contour plots of velocity and
pressure distribution that shows proposed blade can still receive incoming air to the surface of the blade despite having
built up low velocity of air around it. Negative pressure also been noticed more around orientation of the blade
compared to the original blade, showing the blade was having more positive torque at that particular time.

Several initiative was also done like finding the best possible number of blade ranging from 4, 6 and 8. These 3
different sets of blade number was simulated through ANSYS as well and results had shown that increasing number
of blades do not necessarily increase the torque coefficient value obtained. In fact, result from the post-processor stage
shows that the best sets of blade number was found to be at 6,which having minimal decrement, but experiencing the
highest torque coefficient value between the 3.

The best design that could induce more torque coefficient and with the best set of blade number if implemented
together can significantly improve the wind turbine efficiency and can harness more energy from the wind flowing in
Malaysia.

For future research, the simulation can be done in 3-Dimensional, so that torque can directly be obtained in the
simulation which can compute the value of Coefficient of Performance of the Wind Turbine. Tip- Speed Ratio also
can be computed with the surface area of the wind turbine when the simulation was done in 3-Dimensional.

5.0 References

[1] Agostino De Marco, Domenico P. Coiro, Domenico Cucco, and Fabrizio Nicolosi (2014). A numerical study
on a Vertical Axis Wind Turbine with Inclined Arms,1-15
[2] M.H. Mohamed, G. Janiga, E. Pap, D. Thévenin (2011). Optimal blade shape of a modified Savonius turbine
using an obstacle shielding the returning blade, 52, 236-242
[3] R. Gupta, R.Das and K.K. Sharma (2006). Experimental Study of A Savonius-Darrieus Wind Machine
[4] Samuel Bruce Weiss (2010). Vertical Axis Wind Turbine with Continuous Blade Adjustment, 1-26
[5] S. Derakshan and A. Tavaziani (2015). Study of Wind Turbine Aerodynamic Performance Using Numerical
Methods

You might also like