Journal of Applied Fluid Mechanics, Vol. 9, No. 2, pp. 537-542, 2016.

Available online at www.jafmonline.net, ISSN 1735-3572, EISSN 1735-3645.

Usability of the Selig S1223 Profile Airfoil as a High Lift

Hydrofoil for Hydrokinetic Application
S. A. Oller1†, L. G. Nallim1 and S. Oller2
1 INIQUI – CONICET –Faculty of Engineering, National University of Salta, Salta, 4400, Argentina
2
UPC-CIMNE – International Center for Method in Engineering, UPC Technical University of Catalonia
(Barcelona Tech), Barcelona 08034

†Corresponding Author Email: sergio.oller@conicet.gov.ar

(Received December 7, 2014; accepted March 4, 2015)

ABSTRACT

This work presents a numerical analysis of the ability of the high lift airfoil profile Selig S1223 for working
as hydrofoil under water conditions. The geometry of the hydrofoil blade is designed through a suitable airfoil
profile and then studied carefully by means of Computational Fluid Dynamics (CFD) in order to check its
hydrodynamic behavior, i.e., including lift and drag analysis, and determinations of streamlines velocities and
pressures fields. Finally conclusions on the use of this profile in a possible application for hydrokinetic
turbine blades are detailed.

Keywords: Hydrofoil; Hydrokinetic generation; Computational Fluid Dynamics (CFD).

NOMENCLATURE

airfoil chord power

lift coefficient
drag coefficient design attack angle
lift force real angle of attack
pressure maximum aerodynamic profile’s α
Reynolds number velocity tolerance convergence error
wing surface pressure tolerance convergence error
torque fluid density
time angular velocity
absolute flow velocity

1. INTRODUCTION river depth (Singh et al. 2014).

Majority of lowlands worldwide rivers, like the
It is a fact the raise of the renewable energies ones appearing in the Major River Basins of the
requirements. Hydraulic energy is one of the more world map (Fig.1) of the Global Runoff Data Centre
powerful ones, but the extremely high economic (GRDC 2007), which is based on HYDRO1K
and environmental costs of the reservoirs system of the U.S Geological Survey (USGS),
constructions, turned the situation of these kinds of averages about 10m depth,and nearby
constructions around the world in a decreasing 1.5 m⁄s ~2 m⁄s of flow velocity (Hossein et al.
tendency. 2012).So in hydrokinetic river operation (Khan and
Hydrokinetic turbines are an easier way of Bhuyan 2009),the efficiency of the hydrodynamic
hydraulic energy usage due to the use of kinetic rotor is fundamental due to the low speed flows in
energy of current flow waters, instead of the fluvial beds, and the first efficiency step belongs to
reservoirs (Khan et al. 2008). Majority of the achieve a high performance hydrofoil´s design
available published information concerns about (Singh et al. 2014).
WCT (Water Current Turbines) under marine tidal This work is motivated by the possibility of using
work conditions (Güney et al. 2010). Unfortunately, inside water media flow, an airfoil profile capable
these kinds of rotors are useless on rivers because of of taking advantage of high lift efficiency, at low
their big size (4 – 8 times higher than a common
 M, Vol. 9, No. 2, pp. 537-542, 20016.
S. A. Olller et al. /JAFM

flow speeed operations. So the possibilitties of using

that kind of profile as the basis for hydrokinetic
h
turbine blade
b design are explored.

ary normalized coordinates

Fig. 3. Unita c of S1223
S
airfoil generic profile.

Figg. 1. Major Riveer Basins of the world. University off Illinois wind tunnel
t measuress a lift
coefficient of o c = 2.2 and a drag coeffficient
2. HYDROFOIL SELECTION c = 0.046 in n a position off α = 10∘ of anngle of
attack, also measures
m the mmaximum attack k angle
From thee viewpoint of engineering
e desig
gn, the more ofα = 15∘ . Beyond that nnumber boundaryy layer
torque (T)
( has the turbbine rotor, the more m power detachment flow will haappen, droppinng lift
(W) willl develop the tuurbine, see Eq.((1). So, it is coefficient (PPengyin et al. 2014)
2 and enorm mously
importannt to take advvantage of thee maximum Goundar et al. 2012).
increasing draag coefficient (G
possible torque and turbine’s rotor veloccity(ω) too. Aerodynamicss and Hydrodynnamics disciplinnes are
similar, but thhe study of the hhydrodynamic reequires
W=T·ω (1) taking care of ventilatiion and cavvitation
Torque and angular velocity
v of thee rotor are phenomenon. Turbine woorking in riverbed
achievedd by airfoil’s lift forces (Fig.2). Lift force operation avoiids the possibilitty of being affeccted by
depends on the change of pressures (∆ ∆P) (Eq. (2)) ventilation evvent. Cavitationn is produced in the
generateed in the airfoil surfaces,
s and theese pressures outer surfacee due to the low pressure of o the
depends on fluid densitty, airfoil shapee profile and incompressiblle fluid producedd in its neighbo orhood;
the airfo
oil angle of attackα (Eq. (3)) (Balaka and it depends on water temperatu ture, airfoil profi
file and
Rachmaan 2012). flow velocity.. As higher the velocity is, high her the
cavitation possibility.
∆P = − (2)
3. MODEL
O GEOMETR
RY DESCRIPTIO
ON

Starting out from Michael S. Selig and James

Suction zone
z Gugliemo physical moodel (Selig and
(pulling foorces) Guglielmo19997) (Fig.4), a similar three spatial
finite elemennt dimension numerical mod del is
generated (Löhner. 2001) (Figg.5).

P
Positive pressure zone
(ppushing forces)
Fig. 2.
2 Schematic rep presentation of lift forces
generatted in a hydrofooil. Negative upper surface
pressuure (suction) an
nd positive loweer surface
preessure.

If more lift is obtained by one airfoil, more torque

(T) and angular velocityy (ω) will be obttained by the
turbine´ss rotor. This co ommitment is achieved
a by
selectingg a high-lift aerrodynamic shapee profile for
the hydrofoil design. The T selected S1 1223 profile
belongs to the high lift low Reynolds profiles
p class Fig. 4. Micchael S. Selig an
nd James Guglieemo
(Selig and
a Guglielmo1997) (see Fig g.3). Under physical model sketch (Selig and
cambereed airfoils, like S1223,
S have the best ratio of Guglielmo 11997).
generatinng an extremely y high lift at a minimum
m of
flow velocity operationn, and also this high lift is
The numerical model geom metry consists on a
generateed at very low anngles of attack.
hydrofoil madde from a Selig´ss S1223 profile, which

538
S. A. Oller et al. /JAFM, Vol. 9, No. 2, pp. 537-542, 2016.

is located inside a cylindrical fluid control volume of Kratos aims to solve the Navier-Stokes equations
tunnel shape. Details and dimensions of this 0.2 m (see Eq. (3)). Instability of using linear FEM, are
chord length (Pengyin et al. 2014) real scale solved by different approaches like Fractional step
numerical model can be observed in Fig.5, which or Subgrid scale stabilization (Codina. 2002).
shows a diametral-plane of the 3-D finite element
model of the volume of control used for the Model conditions involvev = 2 flow velocity iny
numerical simulation. axis positive direction, crossing the hydrofoil in
axial form (Fig.5). Also, for all model surfaces,
ano-slip condition of null velocity is applied.
Elapsed simulation time t = 1.5s is used to ensure a
state of steady flow achievement. Result drops
unsteadily during th egap between t = 0.0s and
t = 0.4s, and beyond that point velocities and
Flow pressure stabilization occurs.
∅4
Incompressible problem type is solved using a bi-
conjugate gradient stabilized (Van der Vorst. 1992)
solver on velocity and pressure resolution.
Convergence criterion reaches a maximum of 100
4.5 iterations involving velocity convergence error
tolerance (η ) and pressureconvergence error
tolerance (ηP )ofη = 1 · 10 = ηP , using
a∆t stabilization of1 · 10 s.
A Finite Element Variational Multiscale Simulation
(FEVMS) (Hughes. 1995), (Guermond. 1999),
(Hughes et al. 2000), method is applied to solve the
1.55 grid, by the use of the general isothermal fluid
Navier-Stokes governing equation for
incompressible flow applications (Eq. (3)).
0.2 D
ρ D = ρg − ∇p + μ∇ v (3)
Fig. 5. Hydrofoil and 2-D diametral-plane One point-one million of 4 nodes linear
scheme of the Volume of Control dimension of tetrahedicalfinite element (Zienkiewicz and Taylor
the 3-D numerical model geometry. 1991) (Lewis et al. 2004)isusedin a no structured
volume mesh, and 1 · 10 cordal error is given as a
The bases of the difference between the numerical strong tolerance to hydrofoil surface mesh.
model presented, and the physical model, are the
geometry of the volume of control and the fluid
5. MODEL VALIDATION UNDER AIR
parameters. In numerical model a cylindrical
CONDITIONS
volume of control is presented instead of the
rectangular shape of the physical one, to avoid the
influences of edges and corners in the fluid For the calibration of this model and sureness of its
behavior. Also the push rod and airfoil anchors are correct behavior, numerical model is also tested
not used, so it is ensured these elements will not using air parameters instead of water ones.
affect the flow activity. Fluid parameters involve Numerical lift coefficientc obtained values are
crossing the line from Selig and Guglielmo easily comparable with the experimental Michael S.
compressible flow essay, to a non-compressible Selig and James Gugliemo (Selig and
numerical experiment, where the similitude Guglielmo1997) wind tunnel obtained values
between the lift coefficients must be ensured, but (Fig.6).
not this way the streamlines and more less the
It can be observed the similitude of the ratio curve
cavitation effect.
involving c relative toα, between experimental
and numerical results model, maintaining less than
4. MODEL CONDITIONS, MESHING 17%of average difference between both results. The
CRITERIA AND SOLVER qualitative shape of the numerical result is correct,
and the difference observed in figure 6 for the
Since this work aims to analyze the ability of the obtained result is due to the influence of the
mentioned profile operating like a hydrofoil device cylindrical volume control (Fig. 5), chosen to avoid
and not like airfoil one, fluid domain involves water the edges and corners singularities, and also the
conditions, so water parameters are given to fluid wing position that has been chosen in the numerical
variables. test, instead of the actual volume chosen by M. S.
Selig in the experimental test (Fig. 4). This features
New validated Finite Element free open source
and results, also validates the use of the numerical
multiphysics code KRATOS (“Kratos Multi-
model (FEM) presented in this work. Lift
Physics”. 2005) is used for the numerical
coefficient allows the obtaining of airfoil lift force
simulation. The Incompressible Fluid Application

539
S. A. Oller et al. /JAFM, Vol. 9, No. 2, pp. 537-542, 2016.

(Eq. (4)), and so this way curve comparison of lift

forces are presented in Fig.7.

Lift Coefficient under air conditions

CFD Experimental
3

2.5
Fig. 8. Negative pressure distribution (suction)
[Pa] generated on the hydrofoil upper surface.
2 Figure corresponding to = ∘ .
Cy
1.5

0.5

0
0 10 20
 []
Fig. 6. Comparison of lift coefficients obtained
Fig. 9. Positive pressure distribution [Pa]
by numerical results and wind tunnel.
generated on the hydrofoil lower surface. Figure
corresponding to = ∘ .
Lift Force under air conditions

CFD Experimental

100
90
80
70
Fy [N]

60
Fig. 10. Pressure [Pa] distribution in the
50
hydrofoil surrounding flow field. Figure
40 corresponding to = ∘ .
30
20 Stable streamlines along the entire Volume of
Control field can be observed (Fig.11) in post
10 process.
0
0 10 20
 []
Fig.7. Lift force generated by airfoil in air fluid
conditions.

6. HYDROFOIL RESULTS UNDER WATER

CONDITIONS

Axial flow generates uniform pressure distribution

Fig.11. Streamlines generated by flow crossing
along the hydrofoil wingspan surface (Fig.8, Fig.9)
the Volume of Control. Figure corresponding to
and is clearly visible the induced change of pressure
= ∘ . Velocity units [m/s].
∆P formed in the flow field nearby the airfoil
(Fig.10).
Starting out from the pressure (∆P) generated in the
inner and outer surface of the tested wing, is obtain
lift force (F ) thru wingspan surface (S ). Fluid
density (ρ) and flow velocity (v), complete the

540
S. A. Oller et al. /JAFM, Vol. 9, No. 2, pp. 537-542, 2016.

necessary parameters to obtain the wing lift flow turbines operates at low speed flows, their
coefficient (c ), calculated thru Eq. (4) as follows blades have no needs of avoid cavitation, as well as
is imperative to produce the highest lift possible just
F for the same reason of the low speed flow.
c = (4)
ρ · 0.5 · v · S
Numerical model results show that lift force of
profile S1223 airfoil is highly increased (Fig.12),
reaching a maximum F = 1516 Nat α = 15° in
the usage of water fluid conditions, instead of the
air fluid conditions initial designed for.
a)
Lift Force under water conditions
1600
1400
1200
1000
Force [N]

800
600
c)
Incipient flow detachment
400
200
0
0 10 20 d)
 [] Growth of the flow detachment
Fig. 12. Numerical lift for ceobtained under
water conditions.

Figure 13 remarks the well boundary layer

behavior, between the gap formed by α = 0° and
α = 5° of angle of attack (Fig.13.a, Fig.13.b); and
how detached flow becomes incipient at the angle
of attack α = 10° (Fig.13.c) with a clearly growing e) Full detached flow
tendency in α = 15° (Fig.13.d). Also, can be clearly
observed in Figure 13.e, that beyond α = 20°, full Fig. 13. Flow detachment evolution, from = °
hydrofoil detached flow occurs. Detached flow (a), = ° (b), = °(c), = ° (d), to
changes the operating principle of the turbine, from = ° (e).
lift operation to drag operation. Working on drag
ambit decreases dramatically turbine efficiency. Reaching Re = 4.0 · 10 Reynolds number, this
Detached flow also produces structural vibrations, model discloses a good behavior in water fluid
which are especially dangerous to rotating airfoils conditions, exposing right pressure distribution
with high aspect ratio. along the wingspan, and in the flow field, also
proving not entering in cavitation zone. Similar to
Results show nearby 98861 Pa of absolute pressure the original airfoil, this hydrofoil begins detaching
in the outer foil surface, and cavitation phenomena flow at the angle of attack α = 10º, reaching full
occurs below 1250 Pa in 10 º C water, so it is detachment beyond α = 20º.
possible to certify no cavitation phenomenon for the
S1223 high lift profile, working at the detailed Section 2 of this work explains the direct
operation conditions. relationship existing between high efficient
hydrofoil design achievements with the efficiency
7. CONCLUSION in hydrokinetic generation. Results achieved in
this work evidence how suitable is the S1223
profile (aerodynamic initially designed for) at
The achievement of a free flow water turbine
hydrodynamics tasks; and consequently, for the
challenges the energy extraction belonging to a very
particular use of hydrokinetic turbine blade
low speed flow.
design.
Classic water falls turbines operate at high flow
This work can be used as starting point for a future
speeds, so are not suitable for free flow stream. In
design of a high performance hydrokinetic rotor
hydrodynamics symmetrical profiles are commonly
thru their corresponding turbine blades.
used to avoid cavitation, but as hydrokinetics free

541
S. A. Oller et al. /JAFM, Vol. 9, No. 2, pp. 537-542, 2016.

ACKNOWLEDGEMENTS and the origin of stabilized methods.Comput.

Methods Appl. Mech. Eng. 127, 387–401.
This work has been supported by the Spanish Hughes, T., L. Mazzei and J. Jansen (2000).Large
Government through the Ministry of Science and Eddy simulation and the
Innovation (RECOMP project, ref. BIA2005-06952 variationalmultiscalemethod.Comput.Visual.
and DELCOM Project, ref. MAT2008-02232); Sci. 3, 47–59.
AECID (ref. A/024063/09) Spain; Barcelona Tech
(Technical University of Catalonia UPC), Spain; Khan, M.J., G. Bhuyan, M. T. Iqbal and J. E.
International Center for Numerical Method in Quaicoe (2009). Hydrokinetic energy
Engineering (CIMNE), Spain; CIMNE-Classrooms conversion systems and assessment of
of National University of Salta, Argentina; Consejo horizontal and vertical axis turbines for river
Nacional de Investigación Científica y Técnica and tidal applications: A technology status
(CONICET), Argentina, and Consejo de review.Applied Energy86, 1823-1835.
Investigación de la Universidad Nacional de Salta
Khan, M.J., M. T. Iqbal and J. E. Quaicoe (2008).
(CIUNSa), Argentina. In addition, this work was
River current energy conversion systems:
partially supported by the European Research
Progress, prospects and challenges.Renewable
Council under the Advanced Grant: ERC-2012-
and Sustainable Energy Reviews12, 2177-
AdG 320815 COMPDES-MAT. All this support is
2193.
gratefully acknowledged.
Kratos Multi-Physics, CIMNE. (2000).
REFERENCES [Online].Available
Lewis, R., P. Nithiarasuand K. Setharamu
Balaka, R. and A. Rachman (2012).Pitch angle (2004).Fundamentals of the finite element
effect for horizontal axis river current method for heat and finite element method for
turbine.Procedia Engineering, no. 50, 343- heat and fluid flow.John Wiley and sons.
353.
Löhner, R. (2001). Applied CFD Techniques: An
Global Runoff Data Centre (GRDC) (2007). Introduction Based on Finite Element
http://www.bafg.de/GRDC/EN/02_srvcs/22_gs Methods.John Wiley & Sons.
lrs/221_MRB/riverbasins_node.html .
Pengyin, L., Y. Guohua, Z. Xiaocheng and D.
Goundar, J. N., M. A. Rafiuddin and Y. H. Lee Zhaohuix (2014).Unsteady aerodynamic
(2012).Numerical and experimental studies on prediction for dynamic stall of wind turbine
hydrofoils for marine current airfoils with the reduced order modeling,
turbines.Renewable Energy42, 173-179. Renewable Energy69, 402-409.
Guermond, J. (1999). Stabilization of Galerkin Selig, M. S. and J. J. Guglielmo (1997).High-Lift
approximations of transport equations by Low Reynolds Number Airfoil Design.Journal
subgrid modeling.M2AN33, 1293–1316. of aircraft34(1), 72-79.
Güney, M.S. and K. Kaygusuz (2010). Hydrokinetic Singh, P. M.andY. D. Choi (2014). Shape design
energy conversion systems: A technology and numerical analysis on a 1 MW tidal
status review. Renewable and Sustainable current turbine for the south-western coast of
Energy Reviews 14, 2996–3004. Korea.Renewable Energy 68, 485-493.
Hossein, B., J. Woods and E. L. Bibeau (2012). Van der Vorst, H. A. (1992). Bi-CGSTAB: A Fast
Investigation of macro-turbulent flow and Smoothly Converging Variant of Bi-CG
structures interaction with a vertical for the Solution of Nonsymmetric Linear
hydrokinetic river turbine.Renewable Systems.SIAM Journal on Scientific
Energy48, 183-192. Computing 13, 631–644.
Hughes, T. (1995).Multiscale phenomena, Green’s Zienkiewicz, O. C. and L. R. Taylor (1991).The
functions, the Dirichletto-Neumann finite element method. London: McGraw-Hill.
formulation, subgrid-scale models, bubbles

542