Experimental Study on the Analysis of the Use of Forward and Rearward

Wingtip Fences 90

◦

Cant Angle on Wing Airfoil Eppler 562

Bayu Dwi Cahyo

1,2

∗

, Sutardi

and Setyo Hariyadi

1,2

Fluid Mechanic Laboratory, Mechanical Engineering Department, FTI, ITS, Surabaya, Indonesia

Aviation Polytechnic of Surabaya, Surabaya, Indonesia

Keywords:

Wingtip Fence, Eppler 562, Oil Flow Visualization, C

Abstract:

This research is a detailed experimental study on aerodynamic characteristics for wing model airfoil Eppler

562 with and without wingtip. There are two types of wingtip fences at variations that will be used in this

research that are rearward and forward wing tip fence set at cant angle 90

◦

. The chord length of the airfoil is 72

mm and the span length is 300 mm. The Reynolds (Re) number used is 2.3× 10

∞

= 10 m/s) with angle of

attack variations (α) = 0

◦

to 19

◦

. For this research, pressure distributions over the airfoil were measured using a

pressure transducer. Moreover, measurements lift and drag forces of the airfoil were obtained by using a load-

cell system. Oil ﬂow visualization method was used to represent the surface ﬂow patterns. The experimental

results showed that as the angle of attack increased, the separation and the transition points moved towards the

leading edge at all models. Furthermore, for airfoil with forward wingtip fence with cant angle 90

◦

better than for airfoil with rearward wingtip fence with cant angle 90

◦

and plain wing. Forward wingtip fence

showed the best optimum performance of α = 10

◦

settings compared to the other models.

1 INTRODUCTION

The development of aviation technology is currently

growing a rapid and promising, especially in the de-

sign of aerodynamic forms of aircraft. The outer edge

of the wing (wingtip) is one part that has an impor-

tant role in aircraft. Wingtip design will affect the

air ﬂow conditions along the wingspan which will af-

fect the lift coefﬁcient and drag coefﬁcient to improve

aerodynamic performance of aircraft(Sutrisno et al.,

2015). One important thing to note in the design of

a plane is the selection of an airfoil and its modiﬁ-

cations. The advantages of the optimal design are a

reduction in drag and increasing in lift. The main

parameter of wing performance is the C

ratio.

The wing design is considered good if it has a high

value of C

ratio. The low value of C

ratio

is caused by ﬂow separation followed by wake and

vortex occurring on wingtip. Flow separation occurs

due to excessive adverse pressure gradient (APG) and

friction effect. Next, the tip vortex is circular pattern

of air ﬂows from below the wing to the top of the wing

around the wingtip. Separation and vortex are two

phenomena that increase drag component and reduce

lift due to reduce the wingspan effectiveness. One at-

tempt to reduce the vortex is delaying the separation.

Another attempt is wingtip modiﬁcations by addition

of winglet.

Turanoguz and Alemdaroglu (2015) performed a

numerical simulation on airfoil type Eppler 562 by

varying winglet in the form of shifted downstream

type wingtip, hoerner type wingtip, and blended type

wingtip. In the research, it was showed that the

winglet was able to increase the C

value. My-

ilsamy et al. (2015) et al conducted numerical inves-

tigation on NACA 4412 airfoil coordinates for the

wing design and the winglet with the blended design.

The design of the entire wing including winglet were

examined at different cant angles of winglets vary-

ing from 0

◦

, 30

◦

and 90

◦

degrees at different angles of

attack from -2

◦

to 10

◦

. They have observed among

the cases of this study, wings with winglets produces

higher C

ratio performance than the normal air-

craft wing without winglets.

Suranto Putro et al. (2016) investigated by numer-

ical simulation on airfoil type NACA 43018 by afﬁx-

ing winglet in the form of forward wingtip fence and

rearward wingtip fence. In the study it was showed

that the addition of winglet was able to increase the

value up to 22.9% for forward wingtip fence

type at α = 2

◦

. Ristic (2007) studied experimentally

on NACA2415 airfoil by varying angle of attack from

206

Cahyo, B., , S. and Har iyadi, S.

Experimental Study on the Analysis of the Use of Forward and Rearward Wingtip Fences 90o Cant Angle on Wing Airfoil Eppler 562.

DOI: 10.5220/0008549102060214

In Proceedings of the 3rd International Conference on Marine Technology (SENTA 2018), pages 206-214

ISBN: 978-989-758-436-7

-12

◦

to 20

◦

at low Reynolds. In this study, the ﬂow

visualizations were done by using oil ﬂow technique

for qualitative analysis of the transition zone. The

experimental results showed that as the angle of at-

tack increased, the separation and the transition points

moved towards the leading edge at all Reynolds num-

bers.

The aim of this study is to evaluate aerodynamic

performance of airfoil Eppler 562 (E562) without and

with the addition of wingtip fence. There are two

types of wingtip fence variations that will be used

in this research. They are rearward wing tip fence

and forward wing tip fence with cant angle 90

◦

The

Reynolds (Re) number used is 2.3 × 104 (U

∞

= 10

m/s) with angle of attack variations (α) = 0

◦

, 2

◦

, 4

◦

, 8

◦

, 10

◦

, 12

◦

, 15

◦

, 17

◦

and 19

◦

2 RESEARCH METHODOLOGY

2.1 Wind Tunnel and Models

Experiments were conducted in the Aerodynamics

Laboratory Department of Aircraft Engineering at

Aviation Polytechnic of Surabaya The experiments

were carried out in a low-speed, suction-type wind

tunnel with a square test section dimension of 600

mm × 600 mm. The ratio of cross sectional area of

contraction cone was ± 9:1 and the side walls of the

working section were expanded with a divergence an-

gle of 0.3

◦

on each side to minimize boundary layer

effects on the working section walls, and to give a

constant static pressure. The wind tunnel could be

operated at a maximum air speed of 50 m/s. The ex-

periments were carried out at Reynolds (Re) numbers

of 2.3 × 104 based on chord length of airfoil (c) and

free-stream velocity (U∞). The experimental setup

and schematic diagram are shown in Figure 1 and 2.

The airfoil models were manufactured out of

acrylic sheet (Figure 3) that was formed according to

airfoil Eppler 562 proﬁle performed by heated. The

manufactured airfoils have a span length of 300 mm,

and a chord length of c= 72 mm. The proﬁle winglet

maximum chord forward wingtip fence from leading

edge and winglet maximum chord rearward wingtip

fence from leading edge are 72 mm and 22.62 mm re-

spectively, and the winglet minimum chord is 15.12

mm.

2.2 Pressure Measurements

For the measurement of pressure distributions of suc-

tion and pressure surfaces on the Eppler 562 airfoil,

Figure 1: Wind Tunnel WT-60 Set-up

Figure 2: Schematic Diagram of The Experimental Set-up.

a system including a pitot-static tube, a National In-

strument unit, a 32 channel pressure transducer and

32 pressure tapping of 1/16 inch in diameter, which

are ﬂush along the mid-span and tip-span alternately

of the upper and lower surfaces of the wing was used

(Figure 4). Pressure measurements were carried out

by using a computer data acquisition system. The

pressure was measured by using ni cdaq-9172 Na-

tional Instrument which the output was a voltage. The

maximum response time of the pressure transducer

was about 1 ms. Pressure signals were obtained at

a sampling rate of 100 samples data per second or

in one snapshot and a Signal Express software from

Ni-max was used to display the results of conversion

analog pressure data to digital (A/D) in ASCI format

data, and ﬁnal post-processing was complemented in

Microsoft Excel Software to calculate the mean pres-

sure distributions and created the graph. Experiments

were investigated over a range of angles of attack in

order to calculate the pressure coefﬁcient distribution

on all type airfoil.

2.3 Force Measurements

An external load-cell system was used for measuring

the lift and drag forces on the airfoil. The Eagle Tree

system data recorder (Figure 5) software used to dis-

play data from DAQ system that converted an ana-

log input to digital output of the lift and drag forces

Experimental Study on the Analysis of the Use of Forward and Rearward Wingtip Fences 90o Cant Angle on Wing Airfoil Eppler 562

207

Figure 3: Airfoil models manufactured with end-plate: a) plain wing b) rearward wingtip fence and c) forward wingtip fence.

Figure 4: Pressure Tap Location on Midspan and Tip Span

at All Models.

in gram. The calibration was performed by loading

the load cell with known weights. Calibration was re-

peated before each set of experiments to ensure con-

sistency. Sets of data chosen at random were repeated.

The force data was collected at a sampling 100 sam-

ples data over 120 s in FDR format data. Mean forces

and its coefﬁcients were calculated using Microsoft

Excel Software.

2.4 Oil Flow Visualization

The simple and effective way of observing surface

ﬂow events was used oil ﬂow visualization(Ristic,

2007). This method performed on the mat black air-

foil surface was painted before with pigmented oil

Figure 5: Experimental Set-up for The Force Measure-

ments.

was used to get the clear photograph of surface ﬂow

events (Figure 6) and the wind tunnel is run. It is im-

portant that the type of oil mixture which would work

at the speed of the wind tunnel. The mix should have

the right consistency to effectively indicate the devel-

opment of the boundary layer. The simplest mix to

make was palm oil and titanium dioxide (TiO

) the

ratio of palm oil to titanium dioxide was roughly 5:1

then stir for 15 minutes until the TiO

ﬁne granules

not visible. Furthermore, the mixture was diluted by

adding kerosene the ratio 5:1, respectively. And this

ratio of mixture was used in the present study (Figure

SENTA 2018 - The 3rd International Conference on Marine Technology

208

Figure 6: Sample of Oil-ﬂow Visualization Experiment over

Eppler 562 Aerofoil on which Oil laid.

6). However, the pigment was added in some doses

based on Reynolds number.

3 RESULTS AND DISCUSSIONS

3.1 Drag Coefﬁcient (C

)

The nature of the air which always moves from high

pressure to lower pressure towards the tip of the

wing, causes movement of air from lower side wing

which is directed outward around the outer edge of

the wing such as ”spill” and created a vortex around

the wingtip. The existence of vortex will disrupt the

air ﬂow along the wing thereby reducing the effective

area of the wing which resulted in an increase in drag.

This leads to a decrease in the average attack angle

relative to the airﬂow around the wing, too. The co-

efﬁcient of lift and the coefﬁcient of drag have been

calculated from the experimental results. Also vari-

ous graphs have been drawn to examine the measured

and calculated data nature.

Figure 7: Drag Coefﬁcient (C

) vs α on Plain Airfoil and

using Winglet.

In Figure 7, it is shown that the drag coefﬁcient

all of the aircraft wing model under test. The drag

increases slowly with increase in angle of attack to a

certain value and then it increases rapidly with further

increase in angle of attack. The initial value of drag

coefﬁcient of the plain wing at zero angle of attack

for Reynolds number 2.3 × 10

is 0.1713. The value

of the drag coefﬁcient at the transition point i.e. at

an angle of attack of 8

◦

is 0.3271. The experiments

have been done up to an angle of attack of 19

◦

. At the

maximum angle of attack of 19

◦

the drag coefﬁcient is

1.1216. The rapid increase in drag coefﬁcient, which

occurs at higher values of angle of attack, is probably

due to the increasing region of separated ﬂow over

the wing surface, which creates a large pressure drag.

The comparison drag coefﬁcient data for wingtip for

the three variations i.e. plain wing (without wingtip),

rearward wingtip fence cant angle 90

◦

, and forward

wingtip fence cant angle 90

◦

are given in Fig.7. For

all wingtip variations, a similar pattern has been ob-

served. At 10

◦

angle of attack the drag coefﬁcients for

the plain (without wingtip), rearward wingtip fence

cant angle 90

◦

, and forward wingtip fence cant angle

◦

are 0.3832, 0.3561, and 0.3468 respectively, and

began to effectively reduce the drag. It appears that

drag coefﬁcient using forward wingtip fence is lowest

than the other models.

3.2 Lift Coefﬁcient (C

In Figure 8, it is shown that the lift coefﬁcient all of

the aircraft wing model under test. The lift increases

with increase in angle of attack to a maximum value

and thereby decreases with further increase in angle

of attack. The initial value of lift coefﬁcient of the

plain wing at zero angle of attack for a chord based

Reynolds number 2.3 × 10

is 0.2897. The max-

imum value of the lift coefﬁcient is 1.542 and this

maximum values occur at an angle of attack of 12

◦

The experiments have been done up to an angle of

attack of 19

◦

. At the maximum angle of attack of

◦

the lift coefﬁcient is 0.5452. The reason for a

drop in lift coefﬁcient beyond a certain angle of at-

tack e.g. 12

◦

is probably due to the ﬂow separation,

which occurs over the wing surface instead of hav-

ing a streamlined laminar ﬂow there. This condition

is called stalling condition and the corresponding an-

gle of attack is called stalling angle. The stalling an-

gle occurs to be 12

◦

. The comparison maximum lift

coefﬁcient data for wingtip for the three variations:

plain wing (without wingtip), rearward wingtip fence

cant angle 90

◦

, and forward wingtip fence cant an-

gle 90

◦

are given in Figure 8. are 0.6075, 0.7277 and

0.8423 respectively corresponding to an angle of at-

tack of 12

◦

, 17

◦

, and 17

◦

respectively which is also

the stall angle of attack. The addition of the winglet

Experimental Study on the Analysis of the Use of Forward and Rearward Wingtip Fences 90o Cant Angle on Wing Airfoil Eppler 562

209

Figure 8: Lift Coefﬁcient (C

) vs α on Plain Airfoil and

using Winglet.

effective enough to prevent air rotation (vortex) on the

wingtips caused by the encounter of the lower air of

the high pressure wing with low pressure upper air

wing which reduces the angle of attack effectively.

From the graph, it can be concluded that lift coefﬁ-

cient for using forward wingtip fence is highest than

the other variations and it is also seen that the use of

wingtip fence can delay the occurrence of stalls.

3.3 Lift and Drag Coefﬁcient (C

)

A winglet’s main purpose is to improve performance

by reducing drag. By using the winglet, the strength

of the vortex can be reduced and the induced drag can

also be reduced. Such performance improvements

can be seen in the C

comparison graph. The

ratio is the outcome of the observations made

in the two preceding sections. It is observed from

the Fig. 9 that the C

ratio for all the conﬁgu-

rations considered increases with an angle of attack

to its maximum value and thereby it decreases with

further increase in angle of attack. As the effective

area of the wing increased, the value of the C

get in increased. Vortex tips derived from the lower

side wings contribute to reducing this effective area

and also increase the induced drag of the wing. If the

winglet can prevent or minimize air spill effectively,

it can reduce tip vortex.

Table 1: The Separation Point (X

) vs x/c obtained from

The Results of The Oil-ﬂow Visualization.

Re Number α

◦

Plain Rearward Forward

Wingtip Wingtip

Fence Fence

2.3 x 104 0 0.1 0.1 0.1

4 0.08 0.03 0.03

8 0.03 0.02 0.02

12 0.01 0.01 0.01

Figure 9: Comparison of C

between Plain Airfoil and

Airfoil with Addition of Winglet from Experimental Result.

Figure 9 shows comparison of C

ratio on

plain wing and with variations winglet in some an-

gle of attack. This indicates that with the addition

of winglet will increase the C

ratio and show an

improve trend along with the increase of angle of at-

tack. The airfoil wing model without winglet gives a

measured C

ratio of 1.762 whereas the respec-

tive values of the C

ratio for rearward wingtip

fence cant angle 90

◦

, and forward wingtip fence cant

angle 90

◦

are 2.00, and 2.188 respectively at an angle

of attack of 8

◦

Practically it is observed that the C

ratio ver-

sus angle of attack curve gives similar results for 8

to 15 degrees, for the all type of without and with

winglet. It can be said that the wing with forward

wingtip fence cant angle 90

◦

variation has the bet-

ter performance as compared to plain wing (with-

out wingtip) and rearward wingtip fence cant angle

◦

and it is giving the better C

ratio (2.188). It

also shown that the result of experimentally have quite

good agreement with Suranto Putro et al. (2018) nu-

merical research did. The graph indicated as the an-

gle of attack increase, the trend C

experimentally

have a similar to numerical simulation.

3.4 Pressure Coefﬁcient (C

)

The pressure coefﬁcients were calculated from the

pressure readings of Signal Express software that ob-

tained from the pressure tapping at many points across

the midspan and tip span airfoil surface. Figure 10

describes the evolution of the pressure coefﬁcient as

the angle of attack were increased, the graphs of pres-

sure distribution (C

) over the Eppler 562 airfoil plain

SENTA 2018 - The 3rd International Conference on Marine Technology

210

Figure 10: Pressure coefﬁcient (C

) versus the angle of attack (α) 0

◦

to 19

◦

at pressure tap location: a) plain wing mid span b)

plain wing tip span c) rearward wingtip fence mid span d) rearward wingtip fence tip span e) forward wingtip fence mid span

f) forward wingtip fence tip span.

Experimental Study on the Analysis of the Use of Forward and Rearward Wingtip Fences 90o Cant Angle on Wing Airfoil Eppler 562

211

wing and with variations of winglet at angles of α =

◦

, 4

◦

, 8

◦

, 12

◦

, 19

◦

at Re = 2.3 × 10

. In general

these plots illustrate the development of the boundary

layer as the pressure slowly increases after peak suc-

tion from the leading-edge to the trailing-edge. Figure

10a shows at the plain wing (without winglet) that the

effective midspan better than tip span area. This can

be known that the resulting plot pressure coefﬁcient

magnitude at midspan is wider than tip-span area at

all condition of angle of attack. This indicates that tip

vortex affecting characteristics air ﬂow at the wingtip

area, while midspan area not affected because it was

longer distance from the wingtip.

Figures 10c and 10e are airfoil with adding rear-

ward wingtip fence cant angle 90

◦

, and forward

wingtip fence cant angle 90

◦

, respectively. It can

be seen that forward wingtip fence has a bigger

area pressure coefﬁcient distribution than rearward

wingtip fence. This condition is caused by tip on

rearward wingtip fence model not completely cover

by the endplate. So, there is still an air spill from

lower side disrupts effective air ﬂow on upper side

especially in the leading edge wingtip area. Conse-

quently, wing performance at rearward wingtip fence

types worse than forward wingtip fence ones.

3.5 Oil Flow Visualization Results

In Figures 11 and 12, it is describes the concept

boundary layer separation process, oil ﬂow visualiza-

tion was applied to the upper surface of the airfoil at

angles of attack of 0

◦

, 4

◦

, 8

◦

, and 12

◦

. The dense

area of pigment describes the ﬂow has decelerated, it

means the point at which the pressure gradient repeal

causing separation.

As the angle of attack increases the separation

point moves towards the leading edge at all four an-

gle of attack that the vorticity magnitude of the vortex

will increase as the angle of attack increases. At an-

gle of attack 0

◦

(Fig. 11a) this occurs at 10% chord,

◦

(Figure 12d) at 8% chord, 8

◦

(Figure 13a) at 3%

chord and for 12

◦

(Figure 12d) at 1% chord. Further-

more, the separation point (X

) vs x/c obtained from

the results of the oil-ﬂow visualization and pressure

coefﬁcient experiments can be seen in Table 1. It

also seen tip vortex get in interference at upper side

wingtip has an impact to coefﬁcient of lift. The higher

the angle of attack the vortex is formed the wider. As

the vortex formation area increase the lift coefﬁcient

will be decrease. In Fig. 12 shows vortex formation

of the tip where plain wing produces widest vortex

contour.

The forward wingtip fence produces smallest vor-

tex contour than other types. It was occurs due to

the vortex formation of the tip is obstructed by the tip

of the winglet. Therefore, design endplate cover tip

from front until rear of the tip causes tendency of air-

ﬂow from the lower surface near the end of the wing

to ”jump” to the upper surface can be effectively re-

duced. As the disruption of airﬂow from lower sur-

face decrease the vortex formation area will decrease.

It follow the rising of lift coefﬁcient.

4 CONCLUSIONS

The results of this experimental investigation show

that the use of winglet can improve the performance

on airfoils Eppler 562 using variations of wingtip

fence will increase drag along with increasing angle

of attack. However, with the addition of winglet, lift

force can be improved better than plain wing. Winglet

causes the formation of vortex tip can be reduced sig-

niﬁcantly.

From the experimental study it was concluded that

the use of winglet can produce some ﬂow characteris-

tics, namely:

• Increased angle of attack will increase wingtip

vortices and drag coefﬁcient.

• Wing with the addition of forward wingtip fence

cant angel 90

◦

produces C

higher than plain

wing and rearward wingtip fence cant angel 90

◦

• To produce a high C

then the pressure re-

quired on the lower surface is much higher than

the upper surface. Forward wingtip fence cant

angel 90

◦

produces better C

than rearward

wingtip fence cant angel 90

◦

because forward

wingtip fence enlarges the formation of effective

area so that the lift can be increased. It is gen-

erally found that the addition of forward wingtip

fence reduces the induced drag than plain wing.

• On forward wingtip fence cant angel 90

◦

produce

higher performance than other wing start at α

= 6

◦

while rearward wingtip fence cant angel

◦

produce better performance start at α = 8

◦

• In the oil-ﬂow visualization and pressure distribu-

tion experiments, it was concluded that as the an-

gle of attack increased the separation point moves

towards the leading edge at wing airfoil Eppler

562 without and with winglet.

• Forward wingtip fence cant angle produce small-

est vortex formation area at tip than other types

that performed by oil-ﬂow visualization.

SENTA 2018 - The 3rd International Conference on Marine Technology

212

Figure 11: The photographs of oil-ﬂow visualization experiments over the Eppler 562 airfoil for α = 0

◦

(a) plain wing (b)

rearward wingtip fence (c) forward wingtip fence, for α = 4

◦

(a) plain wing (b) rearward wingtip fence (c) forward wingtip

fence.

Figure 12: The photographs of oil-ﬂow visualization experiments over the Eppler 562 airfoil for α = 8

◦

(a) plain wing (b)

rearward wingtip fence (c) forward wingtip fence, for α = 12

◦

(a) plain wing (b) rearward wingtip fence (c) forward wingtip

fence.

Experimental Study on the Analysis of the Use of Forward and Rearward Wingtip Fences 90o Cant Angle on Wing Airfoil Eppler 562

213

REFERENCES

Myilsamy, D., Thirumalai, Y., and P S, P. (2015). Perfor-

mance investigation of an aircraft wing at various cant

angles of winglets using cfd simulation.

Ristic, S. (2007). Flow visualisation techniques in wind tun-

nels part i – non optical methods. Scientiﬁc Technical

Review LVII N., 1.

Suranto Putro, S., Sutardi, S., and Widodo, W. (2016). Nu-

merical study of aerodynamic analysis on wing airfoil

naca 43018 with the addition of forward and rearward

wingtip fence. volume 1778.

Suranto Putro, S., Sutardi, S., and Widodo, W. (2018). Drag

reduction analysis of wing airfoil e562 with forward

wingtip fence at cant angle variations of 75

◦

and 90

◦

volume 2001, page 050003.

Sutrisno, T., Mirmanto, H., Sasongko, H., and Noor, D.

(2015). Study of the secondary ﬂow structures caused

the addition forward facing step turbulence generated:

Case study: Horseshoe vortex between 9c7/32.5c50

body and endwall. Advances and Applications in

Fluid Mechanics, 18:129–144.

Turanoguz, E. and Alemdaroglu, N. (2015). Design of a

medium range tactical uav and improvement of its per-

formance by using winglets. pages 1074–1083.

SENTA 2018 - The 3rd International Conference on Marine Technology

214