Regional Conference on Mechanical and Aerospace Technology

Bali, February 9 10, 2010

Flow characteristic around a D-type cylinder near a plane wall

Triyogi Yuwono
a)
and Wawan Aries Widodo
b)

Fluid Mechanics and Fluid Machineries Laboratory,
Department of Mechanical Engineering Faculty of Industrial Technology
Institute Technology of Sepuluh Nopember (ITS)
Kampus ITS Keputih Surabaya (60111)-Indonesia
Tel. (62) 031 5922941, Fax no. : (62) 031 5922941, e-mail :
a)
triyogi@me.its.ac.id and
b)

wawanaries@me.its.ac.id





Abstract: The flow characteristic around a D-type cylinder with cutting angle 53
o

located near a plane wall were investigated experimentally in a uniform flow at a Reynolds
Number of 5.3 x 10
4
. The pressure distributions along the surface of the cylinder and the
plane wall were measured by varying the gap-to-diameter of cylinder ratio (G/D) in the
range of 0 < G/D < 1.0. Surface oil-film techniques were used to investigate the flow
patterns on the cylinder. The results showed that for the gap-to-diameter ratio G/D < d/D,
there is no bubble separation at the lower side of the D-type cylinder 53
o
. And the
stagnation points were not observed at the cutting surface. At these gaps, the lower and
upper surface distributions are unsymetric, where the magnitude of the minimum pressure
on the upper side is much lower than on the lower surface. As the gap-to-diameter ratio
(G/D) increase, the bubble separation tends to be occured at both sides of the D-type
cylinder 53
o
. And the lower and upper surface distributions tend to become symetric. The
multi stagnation points have gradually arisen as gap ratio increase.

Keywords : D-type cylinder 53
o
, plane wall, bubble separation.



1. Introduction

There are many research efforts have been done to
invertigate the flow characteristic around a bluff
body immersed in a boundary layer, both for
academic interest and for a wide range of
engineering applications. This kind of study can
be applied to many engineering problems such as
undersea pipelines, building constructions and
tube in heat exchangers, etc. There are certainly
the larger number of studies on the flow around a
bluff body of various shape located near a plane
wall, but it may be still of important to continue
studying the flow arround a bluff body near a
plane wall, especially for a bluff body with a
special shape.

There are many discrepant points in previous
studies on a bluff body placed near a wall. For
example, the authors [1] investigated the flow
around an airfoil placed near a wall. They found
that the level of turbulence has a large effect on
the interaction between the boundary layer and
airfoil of various shape. The author [2] reported
that drag coefficient of circular cylinder was
almost unaffected by the gap-to-diameter ratio
until gap was about the same as the thickness of
the wall boundary layer, but lift coeficient
depended strongly on the gap-to-diameter ratio.
The authors [3] studied the flow past a square
cylinder placed near a wall. They reported that the
gap-to-diameter ratio has a large effect on the
sheding motion. The authors [4] confirmed that as
the gap ratio increases, the drag coeficient of the
elliptic cylinder is increased.

For special shape of bluff body, the authors [5]
and [6] have studied flow characteristics of a bluff
body cut from a circular cylinder. Two types of
test models were employed in their study. The
first one is a I-type, which was produced by
cutting both sides of the circular cylinder in
parallel with y-axis, and the second one is a D-
type in which only the front side of the circular
cylinder was cut. The authors [5] used two values
of cylinder diameter d, 20 mm and 30 mm. The
test model with d = 20 mm was used for
measurement of base pressure coefficient (-C
pb
)
and that with d =30 mm for C
D
. They used the
Regional Conference on Mechanical and Aerospace Technology
Bali, February 9 10, 2010



cutting angles (
s
q ) of 0
o
<
s
q <72.5
o
for each
type. Their results show that flow characteristics
are singular in the vicinity of
s
q =53
o
, and -C
pb
is
minimum (0.5 ~0.55) in the two cylinders, when
s
q =53
o
and R
e
>2.5 x 10
4
. In the vicinity of
s
q =
53
o
, the value of C
D
for each model is minimum
and about 50 % of that for the circular cylinder.
Their experimental results also show that for R
e
=
3.1 x 10
4
and
s
q >60
o
, the values of -C
pb
and C
D

are higher than that of the circular cylinder for D-
type. They also stated that when the curvature of
the shear flow from the normal surface to the
uniform air flow equals to that of the circular
cylinder, the shear layer separated from the
normal surface attaches to the circular surface and
change to a turbulent boundary layer along the
circular surface. The authors [6] reported the
existence of bubble separation in lower and upper
side of cylinder with
s
q =53
o
. It was occured just
after the corners of cutting surfaces.

Based on the previous works, the purpose of this
paper is to present a bluff body cut from a circular
cylinder (D-type) placed near a plane wall.
However, as far as we know, there was no study
published in the literature on the flow
characteristic of this unique shape of bluff body
(D-type cylinder) placed near a plane wall. The
aim of the study is to investigate experimentally
the wall effect on the flow characteristics arround
D-type cylinder, with cutting angle of (q
s
) 53
o
, for
various gap-to-diameter ratios.

2. Experimental Set-up

The experiments were carried out in a subsonic
open circuit wind tunnel with test section of 660
mm height, 660 mm width, and 1780 mm length.
Figure 1 shows a schematic diagram of the
cylinder configurations and different parameters
under investigation. A PVC tube was used as
model which had external diameters, D, of 60 mm
and at the two ends of the cylinders spanning the
wind tunnel test-section. It gives length to
diameter ratio and geometric blockage ratio
(diameter-to-wind tunnel height ratio) was the
same value of 9.1 %. The cylinder circular of
PVC tube is cut only at the front side in parallel
with the-y axis, with cutting angle of q
s
=53
o
. It is
called as D-type cylinder 53
o
.

As a plane wall, a
smooth acrylic flat plate 6 mm in thickness and
1160 mm in length was installed 100 mm above
the bottom surface of the test section. The leading
edge of the plate was sharp-edge with an angle of
30
o
. The upstream of cylinders arrangement was
located at 430 mm downstream from the leading
edge of the flat plate.


Figure-1. Schematic diagram of experimental
set-up (dimensions: mm).

The free stream velocities in the wind tunnel was
constantly maintained at 14 m/s, corresponding to
Reynolds number of R
e
=5.3 x 10
4
(based on
diameter of cylinder D and the free stream
velocity U
o
). The boundary layer thickness
formed over the flat plate as d =15.2 mm at the
cylinder location. The gap distance between the
bottom of the cylinders and the flat plate relative
to the cylinder diameter (G/D) were varied from 0
to 1. To measure the pressure distribution on the
cylinders surface and on the plane wall, the 72
pressure taps with interval 5
o
were installed on the
wall of circular cylinder and the 52 pressure taps
were installed on the plane wall at -5 <x/D <10,
where x/D =0 is the positon of the center of
cylinder. The pressure taps were connected to an
inclined kerosene manometer, it makes possible to
measure the pressure distribution around the
cylinder. The velocity profile behind the cylinder
was measured using a Pitot-static tube connected
to the inclined manometer. The Pitot-static tube
was placed at 30 cm at the rear of the axis of
cylinder or corresponding to x/D= 5. Room
temperature was measured using a thermometer,
and this temperature was used to calculate the
fluid viscosity and density. Surface oil-film
techniques were used to investigate the flow
patterns on the cylinder. A mixture composed of
kerosene, titanium dioxide, silicone oil with ratio
of 1:4:1 in weight was used for surface oil-flow
visualization. A circular cylinder wrapped in a
black paper of 0.2 mm in thickness was uniformly
smeared with the mixture, and then cylinders in
configuration were placed inside the wind tunnel
to obtain a surface oil-flow pattern. Finally, the
paper was carefully unwrapped from the cylinder,
and a photograph was taken with a digital camera.


Regional Conference on Mechanical and Aerospace Technology
Bali, February 9 10, 2010



3. Results and Discussion

3.1. Pressure distribution

The pressure coefficient (C
p
) distributions around
the D-type cylinder at the Reynolds number (Re)
=5.3 x 10
4
are shown in Figure 2.


(a)

(b)

(c)
Figure 2. Pressure distribution around D-type
cylinder 53
o
for R
e
=5.3 x 10
4
:
(a). 0 <G/D <0.2, (b). 0.267 <G/D <0.6, (c).
0.667 <G/D <1.0.

The figure 2(a) shows the distribution of C
p
for
G/D <0.2 compared with the distribution of C
p
at
the centerline of wind tunnel. The results show
that the distributions of C
p
are completely
asymmetric. It is important to be noted that for
G/D <0.2 or G/D <d/D (where d/D =0.253)
there is no bubble separation at the lower side of
D-type cylinder 53
o
. Whereas [6] have still
observed the existence of bubble separation
phenomenon at the both sides (the lower and
upper sides) of D-type cylinder 53
o
, and they have
also demonstrate a symmetric wake, if the
cylinder is placed at the centerline of wind tunnel
or far from the wall. This absent of the bubble
separation at the lower side of cylinder is due to
the position of cylinder being closed to the plane
wall. The blockage effect in the gap between the
cylinder and the plane wall has caused a large part
of fluids flow to the upper side of cylinder and the
shear layer separated from the edge of the flat
surface reattaches to the circular arc before
separated at about q =110
o
. A little part of fluids
flow to the lower side causes laminar separation at
q =220
o
. When the gap G/D increasing for 0.2 <
G/D < 0.6, the bubble separation phenomenon
will be gradually raised at the both sides, as
showed in the figure 2(b), but the distributions of
C
p
are still asymmetric compared with the
distribution of C
p
at the centerline of wind tunnel.
Where the C
p
min is observed still at the upper
side of cylinder. Figure 2(c) shows the
distribution of C
p
for 0.633 <G/D <1.0 compared
with the distribution of C
p
at the centerline of
wind tunnel. It seems that the distribution of Cp
tends to be symmetric when the gap is continously
increased. It is due to the blockage effect
becoming weak in accordance with the gap
increase.


Figure 3. Pressure distribution a long the plane
wall as function of the gap-to-diameter ratio
(0 <G/D <1.0) for R
e
=5.3 x 10
4
.

Figure 3 shows the pressure distribution along
the plane wall as function of the gap ratio
between D-type cylinder 53
o
and the wall
Regional Conference on Mechanical and Aerospace Technology
Bali, February 9 10, 2010



(G/D). The pressures were measured in range
of -5 <X/D <10.5. The results show that the
upstream surface pressure decrease and
minimum pressure increase as the gap ratio
increase. As the flow moves downstream, the
negative pressure recovers to static pressure.
It is also observed that as the gap ratio
increases the location of minimum pressure
moves downstream.

3.2. Drag and Lift Coefficient



Figue 4. The variation of the drag coefficient
relative to the value of single cylinder (C
Dp
/C
DP0
)
and the pressure lift coefficient (C
Lp
) with cutting
the gaps G/D for Re =5.3 x 10
4
.

Figure (4) shows the evolution of the pressure
drag coefficient of D-type cylinder 53
o
relative to
the value of the pressure drag coefficient at the
centerline of wind tunnel (C
Dp
/C
Dp0
) and the
coefficient of pressure lift (C
Lp
) for varied gaps
G/D. The result shows that the value of C
Dp
/C
Dp0

is seemly more than 1 for G/D <0.333 and it is
practically closed to 1 for G/D >0.333. For the
evolution of coefficient of pressure lift, the value
of C
Lp
decrease as the gap increase for G/D <
0.333. The coefficient of pressure lift will be
practically closed to zero for G/D > 0.333. It
means that, in general for G/D > 0.333, the
interaction beetween the plane wall and cylinder
will be weak and the flow distributions will be
symmetric as the gap increase. It will give the
value C
Dp
come near to C
Dp0
and the value of C
Lp

tends to zero. This ratio of gap to cylinder
diameter (G/D >0.333) is clearly more than the
ratio of boundary layer thickness to cylinder
diameter (d/D =0.253). It means that the position
of cylinder is in the outside of boundary layer, so
the influence of boundary layer to the flow around
the cylinder becomes weaker than if the cylinder
is closed to plan wall.

3.3. Flow Visualization

Figure 5 show the results of flow visualization at
the wall of the D-type cylinder 53
o
using the
surface oil-flow pattern method for Re =5,3 x 10
4
.

(a)



(b)


(c)

Figure 5. Flow visualization using surface oil-
flow pattern method at the wall of the D-type
Cylinder 53
o
for R
e
=5.3 x 10
4
: (a). at center line
of wind tunnel, (b). G/D =0.133, (c). G/D =0.8.
[MSt = multi stagnation points, MSp = massive
separation, BSp = bubble separation].

Figure 5(a) shows flow visualization for D-type
cylinder 53
o
placed at the center line of wind
tunnel. It seems that multi stagnation points are in



BSp

BSp

MSt

MSp

MSp





BSp

BSp

MSt

MSp

MSp


BSp
BSp
MSp
MSp
Regional Conference on Mechanical and Aerospace Technology
Bali, February 9 10, 2010



the range of 345
o
< q < 15
o
. The bubbles
separations were clearly observed at the both sides
(the lower and upper sides) of the cylinder. These
separations were approximately located in the
range of 53
o
<q <63
o
and 307
o
<q <297
o
for the
upper and the lower sides respectively. The
authors [6] called these phenomena as a turbulent
separation, where the shear layer that is separated
from the edge of the flat front surface reattaches
to the circular arc of the cylinder at q =63
o
and
297
o
for the upper and lower side respectively.
The reattachment flow transits to turbulent mode
and the separation points appear at q =93
o
and q
= 267
o
, for the upper and the lower sides,
respectively. It means that the streamlines arround
the D-type cylinder 53
o
at the center line of wind
tunnel are symetric, which also gives the symetric
wake region behind the cylinder.

Figure 5(b) shows flow visualization for D-type
cylinder 53
o
placed at the gap ratio of G/D =
0.133. It seems that there is no bubble separation
phenomenon at the lower side of cylinder. These
phenomena are also investigated by measuring
pressure distribution directly in the surface of the
large cylinder as it has been discussed in section
3.1.

Figure 5(c) shows flow visualization for D-type
cylinder 53
o
placed at the gap ratio of G/D =0.8.
As the gap-to-diameter ratio (G/D) increase, the
bubble separation tends to be occured at the both
side of the 53
o
-D type cylinder. And the flow
tends to become symetric. The multi stagnation
points have gradually arisen as gap ratio increase.

4. Conclusions

The flow characteristics around a D-type cylinder
with cutting angle 53
o
located near a plan wall
was investigated experimentally. Various gap-to-
diameter ratios (G/D) were considered in order to
study the interaction between the cylinder wake
and the boundary layer.

The results showed that for the gap-to-diameter
ratio G/D <d/D, there is no bubble separation at
the lower side of the D-type cylinder 53
o
. And the
stagnation points were not observed at the cutting
surface. At these gaps, the lower and upper
surface distributions are unsymetric, where the
magnitude of the minimum pressure on the upper
side is much lower than on the lower surface.

As the gap-to-diameter ratio (G/D) increase, the
bubble separation tends to be occured at both
sides of the D-type cylinder 53
o
. And the lower
and upper surface distributions tend to become
symetric. The multi stagnation points have
gradually arisen as gap ratio increase.

Acknowledgements

This research was funded by Research Grant
Program of Indonesia Managing Higher
Education for Relevance & Efficiency (IMHERE)
Project of Mechanical Engineering Department of
Institute Technology of Sepuluh Nopember (ITS)
Surabaya, Indonesia.

References

[1]. M.D. Zhou, L.C. Squire, The interaction
of a wake with a turbulent boundary layer,
Aeronautical J ournal 89, pp. 72-81, 1985.
[2]. Zdravkovich, M.M., Forces on circular
cylinder near a plane wall, Applied Ocean
Research, Vo. 7, No. 4, pp. 1997 201,
1985.
[3]. G. Bosch, M. Kappler, W.Rodi,
Experiments on flow past a square
cylinder placed near a wall, Experimental
Thermal and Fluid Science 13, pp. 292-305,
1996.
[4]. Choi, J .H. and Lee, S.J ., Ground effect of
Flow around an elliptic cylinder in a
turbulent boundary layer, J ournal of
Fluids and Structure, Vol. 4, pp. 697-709,
2000.
[5]. Aiba, S., and Watanabe, H., "Flow
characteristics of bluff body cut from a
circular cylinder", ASME J ournal of Fluid
Engineering, Vol. 119, pp. 453-454, 1997.
[6]. T. Igarashi, Y. Shiba, Drag reduction for
D-shape and I-shape cylinders
(Aerodynamic mechanism of reduction of
drag), J SME International J ournal Series
B, vol. 49, no. 4, pp. 1036-1042, 2006.

Appendix

C
p
coefficient of pressure =
( ) ( )
2
o
U 0.5 / p p

-
C
Dp
coefficient of pressure drag =

b
a
p
d C q q q ) cos( ) (
2
1

Regional Conference on Mechanical and Aerospace Technology
Bali, February 9 10, 2010



C
Lp
coefficient of pressure lift =

-
b
a
p
d C q q q ) sin( ) (
2
1


Notation

A cross-sectional area of the main cylinder =
(d x L) (m
2
)
D diameter of the main cylinder (m)
L length of the cylinder (m)
p surface pressure of the main cylinder
(N/m
2
)

p free stream pressure (N/m

2
)
U
o
free stream velocity (m/s)
r free stream density (kg/m
3
)
q position in angle (degree)