Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

Ocean Engineering 144 (2017) 374–383

Contents lists available at ScienceDirect

Ocean Engineering
journal homepage: www.elsevier.com/locate/oceaneng

Design and hydrodynamic analysis of horizontal axis tidal stream turbines


with winglets
Yiru Ren a, b, *, Bingwen Liu a, b, Tiantian Zhang c, Qihong Fang a, b
a
State Key Laboratory of Advanced Design and Manufacturing for Vehicle Body, Hunan University, Changsha, Hunan, 410082, China
b
College of Mechanical and Vehicle Engineering, Hunan University, Changsha, Hunan, 410082, China
c
Strategic Development Department, China Three Gorges Corporations, Beijing, 100038, China

A R T I C L E I N F O A B S T R A C T

Keywords: Three horizontal axis tidal stream turbines (HATST) with winglets are proposed and investigated to enhance the
Tidal current energy energy conversion efficiency. Based on the conventional tidal turbine, three different winglets including trape-
Horizontal axis turbine zoid, triangle and blended types are equipped on the tip of the blade. The power coefficient, the thrust coefficient,
Hydrodynamics the pressure distribution and the tip vortex are analyzed and compared to investigate the effect of the winglets on
Winglet the hydrodynamic performance of HATST. Numerical simulation results show that the hydrodynamic model
Tip vortex
(Song et al., 2012) agrees well with the experimental test. At the optimal tip speed ratio (TSR ¼ 5), the proposed
tidal turbines could improve the energy conversion efficiency. The best hydrodynamic performance is demon-
strated for the HATST with the triangle winglet, which reduces the intensity of the tip vortex and improves the
energy conversion efficiency at all TSRs. The triangle tip turbine performs 4.34% increment in power coefficient
and 3.97% increment in thrust coefficient at the optimal TSR.

1. Introduction performance of HATST. To better predict the hydrodynamic perfor-


mance, Lee et al. (2012) developed two computational procedures based
Conventional electricity generation using fossil fuels alone is no on the blade element momentum theory and computational fluid dy-
longer sufficient and sustainable to meet the demands of the global namics (CFD) method. Three different numerical modeling techniques
population (Kai-Wern et al., 2013). Additionally, the burning of fossil including multiple reference frame (MRF), sliding mesh and actuator disk
fuels emits undesired greenhouse gases. Thus, developing the potential model were conducted and compared to study the downstream wake
renewable energy is an urgent affair, and much focus has been placed on field of a horizontal axis tidal current turbine by Liu et al. (2016). By
this field. Among the plenty of renewable energy resources, the regular consideration of turbine installation depth, a numerical study was con-
and predictable tidal current energy is of great potential which is envi- ducted to design, analyse and predict the hydrokinetic performance of
ronmentally friendly to reduce both visual and noise pollution (Charlier, horizontal axis marine current turbine with and without the gravity
2003; Zhang et al., 2015). waves by Noruzi et al. (2015). The hydrodynamic performance and the
Zhoushan Island in Zhejiang province of China has a plenty of tidal characteristics of the near and far wake profiles of a horizontal axis tidal
current energy resources. Based on the Marine Renewable Energy current turbine were predicted using an improved actuator surface model
Research Project of China, to exploit the tidal stream energy around by Bai et al. (2015). Seo et al. (2016) conducted an experimental study of
Zhoushan Island, the current researches focus on the design and hydro- a 100 kW horizontal axis tidal stream turbine to investigate the mo-
dynamic investigation of horizontal axis tidal stream turbines. mentum balance in terms of the energy conversion mechanism. Due to
To better utilize the tidal current energy, plenty of researchers focus the high density of seawater, the significant thrust and torsional loadings
on the analysis and design of tidal current turbine. The horizontal axis are applied on the tidal turbine blade. To design a cost-effective hydro-
tidal stream turbine (HATST), the most developed type, can be used to kinetic composite turbine system, a reliability-based fatigue life analysis
extract a large amount of tidal current energy from tidal streams methodology was developed for a medium-scale horizontal axis hydro-
(Goundar and Ahmed, 2013). Numerical simulations and experimental kinetic turbine blade by Li et al. (2014). The performance, efficiency and
tests are important methods to investigate the hydrodynamic stability of the tidal turbine are greatly influenced by the turbine

* Corresponding author. College of Mechanical and Vehicle Engineering, Hunan University, Changsha, Hunan, 410082, China.
E-mail address: renyiru@hnu.edu.cn (Y. Ren).

https://doi.org/10.1016/j.oceaneng.2017.09.038
Received 9 March 2017; Received in revised form 29 August 2017; Accepted 24 September 2017
Available online 29 September 2017
0029-8018/© 2017 Elsevier Ltd. All rights reserved.
Y. Ren et al. Ocean Engineering 144 (2017) 374–383

Nomenclature p estimated order of accuracy ()


Q rotor torque (Nm)
A swept area (m2) R radius of turbine (m)
CP power coefficient () S source term ()
Cpress pressure coefficient () T rotor thrust axis thrust (N)
CT thrust coefficient () TSR tip speed ratio ()
e relative error () u free stream velocity (m/s)
Fs safety factor () φ computational solution variable for uncertainty
GCI grid convergence index () assessment ()
n cell count () ρ fluid density (kg/m3)
P local hydrodynamic pressure (Pa) μ dynamic viscosity (N⋅s2 )
Pk turbulence production rate () Ω turbine's angular velocity (rad/s)
Pref reference hydrodynamic pressure (Pa)

configurations. Huang et al. (2016) designed and improved a coupling simple algorithm was employed to solve the Reynolds-Averaged
bi-directional counter-rotating type horizontal axis tidal turbine to Navier-Stokes (RANS) Equations. The incompressible continuity and
convert tidal energy in terms of ebb and flood tides. To gain advantages momentum equations are solved as Eqs. (1) and (2)
in the powertrain design, two shrouds were designed on a horizontal axis
hydrokinetic model turbine and experimentally tested by Shahsavarifard ∂ρ ∂
þ ðρui Þ ¼ 0 (1)
et al. (2015). Under the reliable numerical model, a design procedure for ∂t ∂xi
the tidal turbine blade is developed by Chul et al. (2012). The cavitation  
during the operation of the tidal turbine is an extremely important ∂ ∂   ∂ρ ∂ ∂ui
ðρui Þ þ ρui uj ¼  þ μ  ρu'i u'j þ Si (2)
problem because it would affect the structural life and reliability. To ∂t ∂xj ∂xi ∂xj ∂xj
predict the cavitation, an effective approach was proposed by Usar and
Where u is the fluid velocity, μ is the dynamic viscosity, ρ is the fluid
Bal (2015). Weichao et al. (2016) designed a HATST with leading-edge
tubercles, and its effects on cavitation and underwater radiated noise density and S is a source term. The ρu'i u'j is Reynolds stress term that is
were also investigated. Although considerable researches have been modeled by a turbulence model. The k-ω shear stress transport (SST)
conducted on HATST, there is still room to improve the energy efficiency. turbulence model (Menter, 1994) was selected in the present method for
Consequently, it is worthwhile to investigate the method to improve the accurate boundary layer detection due to its ability to capture the in-
energy conversion efficiency of HATST. fluence of different factors that affect transition such as the free-stream
For the finite wingspan blade of HATST, the flow near the suction side turbulence and pressure gradients. Additionally, it has been success-
moves inward toward the pressure side with the effect of the blade tip. fully used to simulate turbulent flow over airfoils.
Consequently, the tip vortex will be formed behind the blade tip. The tip The turbulence viscosity is related to turbulence kinetic energy and
vortex will cause power loss and affect the performance of the HATST. turbulence frequency as follows:
Therefore, reducing the size of tip vortex and minimizing the induced
a1 k
drag is of great significance for the performance improvement of the μt ¼ (3)
blade. The winglet is a component in the wing tip of aircraft to enhance maxða1 ω; SF2 Þ
the aerodynamic performance first proposed by Lancaster in the late The transport equations for turbulence kinetic energy (k equation)
1800s (Jupp, 2001). A winglet may carry an aerodynamic load that and specific dissipation rate (ω equation) are introduced as Eqs. (4)
produces a flow field which interacts with that of the main wing thereby and (5)
reduces the amount of spanwise flow (Blackwell, 1976). In essence, the
 
winglets diffuse the tip vortices such that the downwash is weakened, in ∂ ∂   ∂ ∂k
ðρkÞ þ ρuj k ¼ ðμl þ σ k μt Þ þ Pk  β* ρωk (4)
turn, the induced drag is reduced. To mitigate the cavitation problem, ∂t ∂xj ∂xj ∂xj
Song et al. (2012) designed a HATST with a raked tip. In their study, the
HATST with a raked tip also showed better energy conversion efficiency  
∂ ∂   ∂ ∂ω
than the base turbine. However, the reason of power increase was not ðρωÞ þ ρuj ω ¼ ðμl þ σ ω μt Þ þ αρS2  βρω2
∂t ∂xj ∂xj ∂xj
further investigated. In addition, only limited sorts of winglets on HATST
have been studied. To enhance the energy conversion efficiency of 1 ∂k ∂ω
þ 2ð1  F1 Þρσ ω2 (5)
HATST and investigate the effect of winglets on the hydrodynamic per- ω ∂xj ∂xj
formance, three types of winglets are proposed as the component of tidal The constant coefficient in the k equation and ω equation are solved
turbine here. by ϕ ¼ ϕ1 F1 þ ϕ2 ð1  F1 Þ, and ϕ1 is a constant coefficient in k  ε tur-
The present studies focus on the design and hydrodynamic investi- bulence model equation, ϕ2 is a constant coefficient in k  ω turbulence
gation of horizontal axis tidal stream turbines with different winglets. model equation.
Firstly, numerical simulations are carried out using CFD and compared In Eq. (5) closure coefficients and auxiliary relations are as
with the existing experimental data. Secondly, three HATSTs with Eqs. (6)–(9)
different winglets configurations are designed and investigated using
CFD. Finally, numerical results are presented and compared to give    pffiffiffi  
4

k 500μ 4ρσ ω2 k
design guidance of HATST. F1 ¼ tanh min max * ; 2 l ; (6)
β ωL L ω CDkω L2
2. Mathematical and numerical method  
1 ∂k ∂ω 10
CDkω ¼ max 2ρσ ω2 ⋅ ; 10 (7)
In the present study, the steady state CFD simulations were per- ω ∂xj ∂xj
formed. A finite volume method with a pressure-based pressure-velocity

375
Y. Ren et al. Ocean Engineering 144 (2017) 374–383

Table 1 3. Numerical model and validation


Geometry of the HATST model.

r/R Chord (mm) Twist Angle (deg) Foil type 3.1. Turbine blade geometry and experiment description
0.15 26.25 0 2:1 Ellipse
0.20 26.25 0 2:1 Ellipse The present CFD model was verified by comparison with the experi-
0.25 transition transition transition mental data reported by Song et al. (2012), who performed a number of
0.30 59.86 16.98 NACA63-418 experiments in a tow basin of Busan National University, Korea. The
0.35 57.32 14.59 NACA63-418
experiment gives an important reference for the design of HATST. It was
0.40 54.78 12.66 NACA63-418
0.45 52.25 11.07 NACA63-418 carried out in a 100 m long, 8 m wide and 3.5 m (with 5 m pit) deep
0.50 49.71 9.75 NACA63-418 towing tank. The turbine was equipped with a 3-blade rotor of 70 mm
0.55 47.17 8.64 NACA63-418 diameter, and 17 blade sections were defined to control the blade profile.
0.60 44.63 7.69 NACA63-418
The section details of the blade are demonstrated in Table 1.
0.65 42.10 6.87 NACA63-418
0.70 39.56 6.15 NACA63-418
The experiment was performed at a yaw angle of 0∘ , i.e., the velocity
0.75 37.02 5.50 NACA63-418 direction at the inlet was normal to the boundary. The power coefficient
0.80 34.48 4.91 NACA63-418 (CP) of the turbine was acquired for a wide range of TSRs in a number of
0.85 31.94 4.33 NACA63-418 experiments. The power coefficient was defined as CP ¼ QΩ=0:5ρu3 A,
0.90 29.41 3.74 NACA63-418
0.95 26.87 3.02 NACA63-418
and the tip speed ratio was defined as TSR ¼ ΩR=u. In order to obtain a
1.00 24.33 2.50 NACA63-418 target TSR, the towing speed was changed while the turbine's angular
velocity was kept constant (270 rpm) for both the experiments and the
  pffiffiffi 2  CFD simulations.
2 k 500μ
F2 ¼ tanh max * ; 2 l (8)
β ωL L ω
3.2. Computational domains with boundary conditions
 
∂ui The computational domain with boundary conditions is shown in
Pk ¼ min τij ; 10β* kω (9)
∂xj Fig. 1. The whole domain was one-third of a cylinder, whose length was
9D and radius was 3D, where D was the turbine diameter. For MRF was
In these equations, Pk is the turbulence production rate while
employed in this study, only one blade was modeled considering periodic
α1 ¼ 5=9, α2 ¼ 0:44, β1 ¼ 3=40, β2 ¼ 0:0828, σ k1 ¼ 0:85, σ k2 ¼ 1:0,
boundary condition. In the present CFD computation, the hub was not
σ ω1 ¼ 0:5, σ ω2 ¼ 0:856 are constant values.
considered. The whole computational domain was mainly divided into
To deal with the blade rotation in open water condition, the multiple
two zones, the rotor zone which contained the whole blade and the outer
reference frame (MRF) model was employed for economy of computation
zone that was outside the rotor zone. A velocity inlet boundary was
time. The MRF is a steady-state approximation method in which it sim-
applied at the inlet which was located at 3D upstream, where a uniform
ulates the rotation of the rotor without physically rotating the grid but to
and steady velocity was applied. A pressure outlet boundary was applied
form corresponding governing equations. The flow in the rotor domain is
at the outlet which was located at 6D downstream, where the pressure
solved using the rotating reference frame, while in the outer domain the
was kept constant. The symmetry boundary conditions were applied at
equations reduce to their stationary forms.
the outer boundary to ensure no stress at outer boundaries, and rotational
To discretize the convection terms, a second order accuracy upwind
periodic conditions were applied at the two slice sections. A non-slip
interpolation scheme was employed. The solution gradients at the cell
boundary condition was imposed at the blade surfaces, namely, the
centers were evaluated by the cell center-based Least Squares method.
turbine blade was static relative to the adjacent zone. Between the
The commercial CFD code, FLUENT, was used for the CFD computations.
adjacent subdomains, non-matching interfaces were imposed at the
overlap faces, and simple linear interpolation was used for the transport
of the flow properties through the interfaces.

Fig. 1. Computational domain with boundary conditions.

376
Y. Ren et al. Ocean Engineering 144 (2017) 374–383

coefficient of the three types of meshes is presented in Fig. 2. As shown in


the picture, the relative difference of Cp between the coarse meshes and
the medium meshes is 7.9%, meanwhile, the relative difference of Cp
between the coarse meshes and the fine meshes is 6.9%. The relative
difference of Cp between the medium meshes and the fine meshes was
quite small. Therefore, when the cell number of the rotor zone was more
than 1.32 million, the effect of the cell number on the power coefficient
was believed to be small. In consideration of the computation time, the
1.32 million cells were selected as the cells of the rotor zone. The yþ of
the blade surface was in the range of 30–300 which is acceptable for the
high-Reynolds-number k-ω SST model.

3.4. Numerical uncertainty assessment

The uncertainty in the computational simulations was assessed for the


power coefficient at TSR ¼ 5 with the three different meshes above. The
grid refinement ratios need to be greater than 1.3 for the error estimation
Fig. 2. Grid-dependence. results to be acceptable (Noruzi et al., 2015). The estimated order of
accuracy was calculated as

Table 2 jlnjφMC =φFM j þ qðpÞj


p¼ (10)
Numerical uncertainty assessment. ln rMC
Coarse Medium Fine
 p 
Number of cells/106 rMC  s
0.78 1.32 2.53 qðpÞ ¼ ln p (11)
Cp 0.378 0.408 0.404 rFM s
p 3.18
GCI 0.0029
s ¼ signðφMC =φFM Þ (12)

φij ¼ φi  φj (13)


rij ¼ ni nj (14)

where φC , φM , and φF are the Cp solutions at the coarse (0.78 million


cells), medium (1.32 million cells), and fine (2.53 million cells) levels,
respectively. p is the order rate of convergence, r is the grid refinement
ratio and n is the cell count.
The grid convergence index (GCI) can be calculated by equations (15)
and (16) as
eMF
GCI ¼ Fs p (15)
rMC 1

φM  φF
eMF
¼ (16)
φ
F

Where Fs is 1.25, which is a safety factor. eMF is the relative error.


Table 2 summarizes the numerical uncertainty assessment results. The
Fig. 3. Comparison between CFD results with the experiment data (Song et al., 2012). solutions show good mesh convergence with the GCI is less than 0.3%.

3.3. Grid generation 3.5. Validation result

The mesh of the whole computational domain was generated sepa- The comparison between the present CFD results with the experi-
rately in several subdomains and then merged together. The rotor zone mental data is presented in Fig. 3. It can be observed that the present CFD
with the blade was meshed with unstructured cells for its complex ge- results agree well with the experiment results, and there is still a little gap
ometries, which could avoid the difficulty of mesh generation and between those two results. Both highest Cp of those two results occurred
improve computational efficiency. The subdomain with the blade was at approximately TSR ¼ 5, where the CFD's Cp was about 0.408 and the
filled with finer grids to acquire the accurate hydrodynamic perfor- experiment's Cp was about 0.422, the relative error was about 3.3%. In
mance. In order to minimise the number of grid points and at the same addition, the mass and momentum balance of the CFD solutions were
time accelerate solution convergence and improve computational accu- checked and met the requirement.
racy, approximately 0.24 million structured cells were generated in the
outer zone for its simple geometries.
4. HATST design with winglet
The grid-dependence in the computational simulations was assessed
for the power coefficient with three different meshes at TSR ¼ 5. The
The improvement method of horizontal axis tidal stream turbines is
coarse (0.78 million cells), medium (1.32 million cells), and fine (2.53
focused. This section firstly introduces the conventional HATST design
million cells) meshes in the rotor zone were considered. The power
concept, and then the phenomenon about the tip vortex of finite

377
Y. Ren et al. Ocean Engineering 144 (2017) 374–383

Fig. 4. Conventional HATST blade.

The conventional finite wingspan blade would be affected by the tip


vortex effect. When the blade produces lift there must be a pressure
difference between the pressure and suction side. With the effect of the
blade tip, the flow near the pressure side moves outward to the suction
side, thus the spanwise flow is produced and the flow field shows the
three-dimensional effect. Once these two opposite flows meet at the
trailing edge, and due to the viscosity of the water, a vortex is formed
behind the blade tip. The energy conversion efficiency and hydrody-
namic performance would be affected by the tip vortex.

4.2. The origin of winglet

The application of winglet is an effective way to reduce the formation


of a tip vortex. The concept of winglets originated from the wingtip
endplates that proposed by British aerodynamicist Lancaster in the late
1800s (Jupp, 2001). He presented the idea that a vertical surface at the
wingtip (endplates) would reduce drag by controlling wingtip vortices.
However, due to the extra form drag or the profile drag of the endplates,
Fig. 5. Trapezoid tip turbine blade. the concept never demonstrated its effectiveness. It was not until the
1970's the concept of winglets was developed by Whitcomb, an engineer
at NASA Langley Research Center. Whitcomb designed a more complex
winglet installed on the wingtip of a first-generation jet transport. His
research reveals that properly designed winglet could reduce the in-
tensity of the tip vortices, thereby, reduce the lift-induced drag and
remarkably improve L/D ratios (Whitcomb, 1976). Whitcomb's research
set the ground upon the modern winglet design, after that, many re-
searchers carried out a great deal of investigation on winglets that
applied on planes and wind turbines (Brocklehurst and Barakos, 2013;
Maughmer, 2003; Monier et al., 2014). Among these studies, Monier
et al. (2014) reported around 9% increase in the power production of an
optimized winglet for a wind turbine. To improve the energy conversion
efficiency of HATST, the design concept and hydrodynamic analysis of
horizontal axis tidal stream turbines with different winglets are proposed
and investigated.

4.3. HATSTs with winglets

Fig. 6. Triangle tip turbine blade. The HATST with winglets were designed to reduce the formation of
the tip vortices. With the introduced winglet, the induced drag would be
wingspan blade is exhibited. To enhance the energy conversion effi- reduced, thus the energy conversion efficiency could be improved.
ciency, the HATSTs with different winglets are designed and compared. However, adding a winglet also results in a redundant wetted area, and
the profile drag would be increased (Maughmer, 2003). Therefore, to
decrease the total drag, the design concept should ensure that the profile
4.1. Conventional HATST drag of the winglet is smaller than the decrease in the induced drag.
Under the above consideration, three turbines with different types of
The concept of the water current turbine was investigated by different winglets including trapezoid, triangle and hybrid types were designed.
researchers since 1979. The rotor blade of HATST (Fig. 4) is the key To verify the design concept, just the blade tip was modified, while the
component that extracts energy from the tidal current. Except for the other parts were kept constant. In addition, all the blades were insured to
blade root, the blade is made up of airfoil sections, each consisting of keep the rotor radius unchanged.
different chord and twist parameters distributes along the blade, which The trapezoid tip turbine blade was mounted with a blended winglet
will lead to optimal hydrodynamic performance. is exhibited as Fig. 5. The winglet bended towards the suction side and

378
Y. Ren et al. Ocean Engineering 144 (2017) 374–383

was NACA63-418. As shown in Fig. 6, the triangle tip turbine blade is also
mounted with one blended winglet. The shape of the winglet resembled
triangle without the sharp corner. The winglet bended towards the suc-
tion side and the cant angle was kept at approximately 90 . The section
shape of the winglet used 18:1 ellipse type rather than an airfoil. The tip
chord of the winglet was 2 mm, and the center line was tangent with the
center of the original tip. The height of the triangle winglet was also
22 mm. Unlike the trapezoid and triangle tip, the hybrid tip turbine blade
mounted with a hybrid blended winglet is given in Fig. 7. The hybrid
winglet was made up of two winglets, one bent towards the suction side,
and the other bent towards the pressure side. Both of the two winglets
tilted tangentially towards the trailing edge. The shape of each winglet
resembled trapezoid and the section shape of each winglet took the
profile of NACA63-418. The tip chord, the cant angle and the height of
the winglet of the suction side was 10 mm, 75∘ and 22 mm, respectively.
The tip chord, the cant angle and the height of the winglet of the pressure
side was 8 mm, 85∘ and 16 mm, respectively.

Fig. 7. Hybrid tip turbine blade.


5. Results and discussion

The HATSTs with winglets were analyzed and compared with the
conventional turbine over a range of TSRs (from 4 to 10). The turbine's
angular speed was kept at 270 rpm, while the current speed was adjusted
to acquire the target TSR, which was the same for both the experiments
and the CFD computations. The computational models of the three
designed turbines were all the same except for the subdomain with the
blade. Accordingly, the mesh was the only part of the subdomain with the
blade that changed, and the rest of the mesh in the outer subdomain was
the same with the conventional turbine. The number of grid points of the
three designed turbines was kept approximately the same with that of the
conventional turbine.

5.1. Power coefficient

The comparisons of power coefficient between the conventional


turbine and the other three turbines with winglets are addressed as Fig. 8,
and the percentage of increase rate in power coefficient for the three
HATSTs compared with the conventional turbine is exhibited as Fig. 9. It
Fig. 8. Power coefficient comparison between the conventional turbine and the is evident that the winglets could improve the energy conversion effi-
three HATSTs. ciency of the turbines in most conditions, and all the blade configurations
acquire their best energy conversion efficiency at approximately TSR ¼ 5.
When the TSR was about less than 7, the CP of the trapezoid tip tur-
bine was greater than that of the conventional turbine. As shown in
Fig. 9, at TSR ¼ 4 the CP of the trapezoid tip turbine is 5.03% greater than
that of the conventional turbine. When the TSR was increasing, the
percentage of increase rate in CP was decreasing gradually. At the opti-
mum TSR (TSR ¼ 5), the trapezoid tip turbine achieved 3.65% increment
in power coefficient compared with that of the conventional turbine. At
TSR ¼ 10, the CP dramatically decreased by 45.59% compared with that
of the conventional turbine.
The CP variation of the hybrid tip turbine showed the similar trends as
that of the trapezoid tip turbine. From TSR ¼ 4 to TSR ¼ 10 the per-
centage of increase rate in CP was reduced. The CP of the hybrid tip
turbine was 3.19% greater than that of the conventional turbine at
TSR ¼ 4. At the optimum TSR, the hybrid tip turbine achieved 2.01%
increment in power coefficient. However, at TSR ¼ 10 the CP of hybrid
winglet tip decreased by 38.94% compared with that of the conven-
Fig. 9. Increase in power coefficient compared to the conventional turbine. tional turbine.
The triangle tip turbine performed power improvement at all TSRs. In
tilted tangentially towards the trailing edge, and the cant angle was kept Fig. 9, it is evident that the performance of the triangle tip turbine is the
at approximately 90 . The tip chord and the height of the winglet were best. From TSR ¼ 4 to TSR ¼ 7.4, the percentage of increase rate in CP
10 mm and 22 mm, respectively. The intersection between the blade tip varied from 4.37% to 0.89%. However, from TSR ¼ 7.4 to TSR ¼ 10, the
and winglet was a smooth curve instead of a sharp angle to reduce the increase rate in CP changed from 0.89% to 14.11%. At the optimum TSR,
intersection drag formed at the junction of blade and winglet. The shape the triangle tip turbine achieved 4.34% increment in power coefficient.
of the winglet resembled trapezoid while the section shape of the winglet

379
Y. Ren et al. Ocean Engineering 144 (2017) 374–383

Fig. 10. Thrust coefficient comparison between the conventional turbine and the
three HATSTs.

Fig. 11. Increase in thrust coefficient compared to the conventional turbine.

5.2. Thrust coefficient

The thrust is an indication of the axial forces acting on the blade, and
an increase in axis thrust means the increase of the load. The comparisons
of thrust coefficient (CT) between the conventional turbine and the other
three HATSTs with winglets are presented in Fig. 10, and the thrust co-
efficient was defined as CT ¼ T=0:5ρu2 A: Fig. 11 presents the increase
rate in thrust coefficient of the three HATSTs with winglets compared
with the conventional turbine.
As demonstrated in the figures the thrust coefficient of all the three
HATSTs configured with winglets is greater than that of the conventional
turbine. For the trapezoid tip turbine, the increase rate in CT ranged from
5.79% to 13.40%. At the optimum TSR, the trapezoid tip turbine pro-
duced 5.66% increment in thrust coefficient compared with that of the
conventional turbine. Overall, the percentage rise in CT increased with
the TSR increased. The CT variation of the hybrid tip turbine and the
triangle tip turbine all showed the similar trends. The increase rate in CT
of the hybrid tip turbine ranged from 2.53% to 7.08%. While the increase
rate in CT of the triangle tip turbine ranged from 3.72% to 6.69%. At the
optimum TSR, the hybrid tip turbine and the triangle tip turbine pro-
duced 2.74% and 3.97% increment in thrust coefficient, respectively.
The triangle tip turbine yielded the best power results of all the
HATSTs, meanwhile, the axial thrust was not the most. Although the
axial thrust of the triangle tip turbine was higher than that of the Fig. 12. Comparisons of pressure coefficient distribution on four span sections at TSR ¼ 5.

380
Y. Ren et al. Ocean Engineering 144 (2017) 374–383

conventional turbine, it was much less than that of the trapezoid tip
turbine and similar to that of the hybrid tip turbine. In the condition of a
large TSR, the increase rate in CT of the three HATSTs with winglet was
much greater than that of the conventional turbine. However, the current
speed was relatively low and the axis thrust was small. As a result, its
effect on the blade structure was expected to be small. Under the con-
dition of a low TSR, the current speed was relatively high, but the in-
crease rate in CT was relatively small especially for the hybrid tip turbine
and the triangle tip turbine.

5.3. Pressure coefficient distribution

To further investigate the effect of the winglets on the hydrodynamic


performance, the pressure coefficient distribution on four span sections
along the blade at TSR ¼ 5 (Fig. 12) and TSR ¼ 10 (Fig. 13) is exhibited.
The pressure coefficient was defined by:

P  Pref
Cpress ¼
0:5ρu2

where P and Pref were the local and the reference hydrostatic pressures,
respectively. The horizontal axis (x/c) was the cross-section of the span
normalized.
As illustrated in Figs. 12 and 13, the current close to the blade tip is
most affected by the winglets. On the 0.95 span section, the pressure
coefficient on the suction side varied remarkably due to the fact that the
winglets were designed to bend towards the suction side. The pressure
coefficient on the suction side of the hybrid tip turbine decreased the
most. However, the pressure coefficient on the pressure side of the
trapezoid tip turbine and the triangle tip turbine was nearly identical to
that of the conventional turbine. Because the hybrid tip turbine had two
winglets which bended towards both sides of the blade, respectively, the
pressure coefficient of the pressure side was affected as well. With the
pressure coefficient decreased remarkably on the suction side of the
HATSTs with winglets, the difference of the net pressure was larger. In
other words, the HATSTs with winglets provided better current pattern
and generated higher torque on the span near the blade tip. On 0.75 span
section and 0.5 span section, the winglets had a very small influence on
the pressure coefficient. The 0.25 span section was close to the root of the
blade, the section transited from ellipse to airfoil, where the separation
effect occurred; which affected the pressure coefficient and could do
some bad effect on the power coefficient. However, the blade root con-
tributes very little to the energy production, the effect is expected to
be small.

5.4. Tip vortex

In order to investigate the effect of the winglets on tip vortex, the


magnitude contours near the blade tip of all the configurations at TSR ¼ 5
(Figs. 14–17) and TSR ¼ 10 (Figs. 18–21) are presented. The first plane of
vorticity magnitude was plotted 0.066R away from the blade tip, and the
interval between the planes was 0.017R.
As shown in pictures, it is evident that the conventional turbine forms
the strongest vortices. As expected, the three HATSTs with winglets could
suppress the tip vortex. As demonstrated in Fig. 15, the tip vortex of the
trapezoid tip turbine is made up by two coherent patches of vorticity.
One is behind the tip of the winglet the other is near the root of the
winglet. The intensity of the tip vortex that near the tip of the winglet was
stronger than the one near the root of the winglet, and the intensity of
these vortices was less than the single vortex of the conventional turbine.
As they convected downstream, the tip vortex near the end of the winglet
gradually disappeared. That presumably was the reason why the vortices
of the trapezoid tip turbine dissipated faster than that of the conventional

Fig. 13. Comparisons of pressure coefficient distribution on four span sections


at TSR ¼ 10.

381
Y. Ren et al. Ocean Engineering 144 (2017) 374–383

Fig. 14. Vorticity magnitude contours of conventional turbine at TSR ¼ 5.


Fig. 18. Vorticity magnitude contours of conventional turbine at TSR ¼ 10.

Fig. 15. Vorticity magnitude contours of trapezoid tip turbine at TSR ¼ 5.

Fig. 19. Vorticity magnitude contours of trapezoid tip turbine at TSR ¼ 10.

Fig. 16. Vorticity magnitude contours of hybrid tip turbine at TSR ¼ 5.

Fig. 20. Vorticity magnitude contours of hybrid tip turbine at TSR ¼ 10.

that of the conventional turbine. At the fifth plane, the vortex can't be
observed in Fig. 16 which can be observed clearly in Fig. 14.
It was quite clear that the intensity of the vortex behind the tip of the
triangle tip turbine blade was the lowest of all the HATSTs. The intensity
of the vortex at each plane was similar to each other and decreased as the
vortex convected downstream, in other words, the vortices behind the
blade tip were relatively stable. As a result, the triangle tip turbine
showed the best energy conversion efficiency of all at TSR ¼ 5. It can be
concluded from the above that reducing the intensity of the tip vortex is
beneficial to the energy conversion efficiency of HATST.
The vorticity magnitude contours near the blade tip of all the turbine
Fig. 17. Vorticity magnitude contours of triangle tip turbine at TSR ¼ 5.
configurations at TSR ¼ 10 are given as Figs. 18–21. It is evident that the
intensity of the vortex at each plane at TSR ¼ 10 is much less than that at
turbine. A strong vortex at the first plane near the tip of the hybrid tip TSR ¼ 5, this is because at high TSR the velocity of the current was very
turbine blade could also be observed, but it dissipated much faster than

382
Y. Ren et al. Ocean Engineering 144 (2017) 374–383

The work could give design guidance of HATST. Under the support of
Marine Renewable Energy Research Project of China, future work will
focus on the uncertainty and optimization analysis and design of HATST
to increase energy conversion efficiency and reduce the thrust load. In
addition, the composite tidal current turbine and fluid-structure inter-
action would be investigated in the future.

Acknowledgments

This research is funded by the Marine Renewable Energy Research


Project of State Oceanic Administration of China (Grant No.
GHME2013GC03 and GHME2015GC01), and the Fundamental Research
Funds for the Central University (No. 201401390741).

References
Fig. 21. Vorticity magnitude contours of triangle tip turbine at TSR ¼ 10.
Bai, Guanghui, Li, Guojun, Ye, Yanghui, Gao, Tieyu, 2015. Numerical analysis of the
hydrodynamic performance and wake field of a horizontal axis tidal current turbine
low. The intensity of tip vortices of the trapezoid tip turbine and the using an actuator surface model. Ocean. Eng. 94, 1–9.
hybrid tip turbine was similar to that of the conventional turbine. In other Blackwell Jr., J.A., 1976. Numerical Method to Calculate the Induced Drag or Optimum
words, at TSR ¼ 10 the winglets of these two HATSTs had a much smaller Loading for Arbitrary Non-planar Aircraft. Vortex-lattice Utilization, NASA SP-405,
1976, p. 49.
effect on tip vortex, especially for the trapezoid tip turbine. Therefore, Brocklehurst, A., Barakos, G.N., 2013. A review of helicopter rotor blade tip shapes. Prog.
the decrease in intensity of the tip vortex for the trapezoid tip turbine and Aerosp. Sci. 56, 35–74.
the hybrid tip turbine was quite small, thus the reduced induced drag was Charlier, R.H., 2003. A ‘‘sleeper’’ awakes: tidal current power. Renew. Sustain Energy
Rev. 7 (3), 187–213.
even smaller than the profile drag brought by the winglet. That might the
Chul, hee Jo, Jin, young Yim, Kang, hee Lee, Yu, ho Rho, 2012. Performance of horizontal
reason why the energy conversion efficiency for the trapezoid tip turbine axis tidal current turbine by blade configuration. Renew. Energy 42, 195–206.
and the hybrid tip turbine was less than that of the conventional turbine Goundar, Jai N., Ahmed, M. Rafiuddin, 2013. Design of a horizontal axis tidal current
at high TSR. However, the intensity of the vortex at each plane behind the turbine. Appl. Energy 111, 161–174.
Huang, B., Zhu, G.J., Kanemoto, T., 2016. Design and performance enhancement of a bi-
blade tip of the triangle tip turbine was much less than that of other directional counter-rotating type horizontal axis tidal turbine. Ocean. Eng. 128,
HATSTs, especially for the vortex at the first plane. As a result, the tri- 116–123.
angle tip turbine still performed higher energy conversion efficiency at Jupp, J., 2001. Wing aerodynamics and the science of compromise. Aeronaut. J. 105
(1053), 633–641.
high TSR. Kai-Wern, Ng, Wei-Haur, Lam, Khai-Ching, Ng, 2013. 2002-2012: 10 years of research
progress in horizontal-axis marine current turbines. Energies 6 (3), 1497–1526.
6. Conclusions Lee, Ju Hyun, Park, Sunho, Kim, Dong Hwan, Rhee, Shin Hyung, Kim, Moon-Chan, 2012.
Computational methods for performance analysis of horizontal axis tidal stream
turbines. Appl. Energy 98, 512–523.
To improve the energy conversion efficiency of HATST, three hori- Li, H., Hu, Z., Chandrashekhara, K., Du, X., Mishra, R., 2014. Reliability-based fatigue life
zontal axis tidal stream turbines with winglets were designed and investigation for a medium-scale composite hydrokinetic turbine blade. Ocean. Eng.
89, 230–242.
investigated. The hydrodynamic performance was conducted using CFD Liu, Jing, Lin, Htet, Purimitla, Srinivasa Rao, 2016. Wake field studies of tidal current
and compared with that of the conventional turbine. The CFD validation turbines with different numerical methods. Ocean. Eng. 117, 383–397.
results obtained using the MRF with k-ω SST turbulence model distinctly Maughmer, M.D., 2003. The design of winglets for high-performance sailplanes. J. Aircr.
40 (6), 1099–1106.
matched with the experimental data.
Menter, F., 1994. Two-equation eddy-viscosity turbulence models for engineering
In the turbine design, the blade tip of the conventional tidal turbine applications. AIAA J. 32 (8), 1598–1605.
was equipped with winglets while other parts kept constant. A winglet Monier, A. Elfarra, Uzol, Nilay Sezer, Sinan Akmandor, I., 2014. NREL VI rotor blade:
may carry a hydrodynamic load that produces a flow field that interacts numerical investigation and winglet design and optimization using CFD. Wind
Energy 17, 605–626.
with that of the blade thereby reduces the amount of spanwise flow. The Noruzi, R., Vahidzadeh, M., Riasi, A., 2015. Design, analysis and predicting hydrokinetic
trapezoid tip turbine, the hybrid tip turbine and the triangle tip turbine performance of a horizontal marine current axial turbine by consideration of turbine
performed 3.65%, 2.01% and 4.34% increment in power coefficient at installation depth. Ocean. Eng. 108, 789–798.
Seo, Jeonghwa, Lee, Seung-Jae, Choi, Woo-Sik, Park, Sung Taek, Hyung Rhee, Shin, 2016.
the optimum TSR (TSR ¼ 5), respectively. Meanwhile, the results showed Experimental study on kinetic energy conversion of horizontal axis tidal turbine.
that energy conversion efficiency of the trapezoid tip turbine and the Renew. Energy 97, 784–797.
hybrid tip turbine could not be improved at all TSRs. With the help of the Shahsavarifard, Mohammad, Bibeau, Eric Louis, Chatoorgoon, Vijay, 2015. Effect of
shroud on the performance of horizontal axis hydrokinetic turbines. Ocean. Eng. 96,
triangle tip winglet, the tidal turbine was equipped with the winglet of 215–225.
ellipse sections, the energy conversion efficiency would be improved at Song, Museok, Kim, Moon-Chan, Do, In-Rok, Hyung Rhee, Shin, Hyun Lee, Ju,
all TSRs. In fact, this turbine yielded the most power among all the Hyun, Beom-Soo, 2012. Numerical and experimental investigation on the
performance of three newly designed 100 kW-class tidal current turbines. Inter J.
HATSTs. Moreover, with the configuration of winglets, the energy con- Nav. Archit. oc. Eng. 4, 241–255.
version efficiency was improved, but the axis thrust was increased. The Usar, D., Bal, S., 2015. Cavitation simulation on horizontal axis marine current turbines.
increase rate in thrust coefficient of the trapezoid tip turbine, the hybrid Renew. Energy 80, 15–25.
Weichao, Shi, Atlar, Mehmet, Rosli, Roslynna, Aktas, Batuhan, Norman, Rosemary, 2016.
tip turbine and the triangle tip turbine at the optimum TSR was 5.66%,
Cavitation observations and noise measurements of horizontal axis tidal turbines
2.74% and 3.97%, respectively. with biomimetic blade leading-edge designs. Ocean. Eng. 121, 143–155.
The pressure coefficient distribution on the blade span section and the Whitcomb, R.T., 1976. A design approach and selected wind-tunnel results at high
subsonic speeds for wing-tip mounted winglets. NASA TN D-8260. NASA Langley
tip vortices behind the blade tip of each HATST were well reproduced.
Research Center, p. 1976.
The results showed that the HATST with winglet could generate higher Zhang, Jisheng, Gao, Peng, Zheng, Jinhai, Wu, Xiuguang, Peng, Yuxuan, Zhang, Tiantian,
torque on the span section near the blade tip, and the winglet enhanced 2015. Current-induced seabed scour around a pile-supported horizontal-axis tidal
the energy conversion efficiency by decreasing the intensity of the stream turbine. J. Mar. Sci. Tech. 23 (6), 929–936.

tip vortex.

383

You might also like