fundamental research on resistance reduction of surface...
TRANSCRIPT
-
ORIGINAL ARTICLE
Fundamental research on resistance reduction of surfacecombatants due to stern flaps
Atsuo Maki1 • Jun Arai1 • Tatsuhiro Tsutsumoto1 • Keisuke Suzuki1 •
Yoshiki Miyauchi1
Received: 20 April 2015 / Accepted: 20 November 2015 / Published online: 23 February 2016
� JASNAOE 2016
Abstract It is well known that a decrease in ship resis-
tance may be achieved due to the installation of a stern flap.
Therefore, so far, a considerable amount of research on
stern flaps has been conducted. Previous research has
demonstrated that the primary mechanism by which a stern
appendage reduces resistance is a change in the pressure
distribution over the aft body of the hull, and secondly
through effects on the running attitude, near and far field
wave generation, and local transom flow among other
phenomena. However, the change in pressure distribution
is influenced by the other components. Hence, there is still
room for argument about the relative contribution of each
component to the pressure distribution. Therefore, as the
first step of the research, by conducting the model exper-
iment in towing tank and CFD (Computational Fluid
Dynamics) analysis, we examined the effect of running
attitudes and wave making at the after portion of the hull
on resistance reduction. As a result, it is concluded that a
flap affects a change in the wave generated at the transom
part and it could lead to a decrease in wave-making
resistance.
Keywords Stern flaps � Resistance reduction � Destroyer �CFD
List of symbols
Fn Froude number (¼ u=ffiffiffiffiffiffiffiffiffiffiffi
gLWLp
)
g Gravitational acceleration [=9.8 (m/s2)]
LWL Waterline length (m)
Cr Residual resistance coefficient
Cw Wave-making resistance coefficient
Rr Residual resistance coefficient (N)
U Forward speed (m/s)
r Volume (m3)q Water density (kg/m3)
1 Introduction
Stern flaps have been utilized for many recent naval surface
combatants to provide increased maximum speed and to
improve the fuel efficiency [e.g., 1]. Stern flaps are rela-
tively small appendages attached to the ship’s transom.
They have now been installed on new construction vessels
or as a retrofit on over 170 US Navy and US Coast Guard
ships [e.g., 2]. In particular, the US Navy has retrofitted all
28 ships of the USS Arleigh Burke (DDG 51) Flight I/II
class guided missile destroyer hulls with a stern flap [3]. In
the JMSDF (Japan Maritime Self Defense Force), JDS
Akizuki (DD 115) class destroyers are also equipped with a
stern flap. Cave et al. [4] showed that based on powering
model tests of the guided missile frigate USS Oliver
Hazard Perry (FFG 7), a stern flap decreased the delivered
power by 8.4 % at 26 knots, meaning that a decrease in
annual fuel consumption of 3.8 percent can be achieved.
Moreover, they emphasize that the flap will produce a
larger decrease in delivered power at full scale than shown
by the model experiments. Cusanelli et al. [5] estimated,
using the pre- and post–pre trials delivered power data, that
the annual time-averaged delivered power reduction for the
stern flap of 11.7 % can be achieved for the USS Spruance
& Atsuo [email protected]
1 Naval Systems Research Center, Acquisition, Technology
and Logistics Agency, Ministry of Defence, 2-2-1
Nakameguro, Meguro, Tokyo 153-8630, Japan
123
J Mar Sci Technol (2016) 21:344–358
DOI 10.1007/s00773-015-0356-8
http://crossmark.crossref.org/dialog/?doi=10.1007/s00773-015-0356-8&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1007/s00773-015-0356-8&domain=pdf
-
(DD 963) class destroyers. The US navy has been making a
great effort to investigate the scaling effects between the
stern flap performance at real size and at model scale [6].
Moreover, it is noteworthy that stern flaps show beneficial
effects at Froude numbers of 0.8 or more for a patrol boat
with twin shaft and fixed-pitch propellers [7, 8]. Although
the present installed stern flaps are fixed, Cusanelli has
demonstrated test results for active stern flaps which aim to
improve not only the powering performance but also sea-
keeping performance [9]. Moreover, there exists an opti-
mization tool for stern flap design [10].
Cusanelli et al. [e.g., 1] explained that the primary
mechanism by which a stern appendage works is the
change in pressure distribution over the after body of the
hull, and secondly by affecting the running attitudes, near
and far field wave change, local transom flow and so on. As
experimentally shown in this paper, we also considered that
a flap improves the hull pressure distribution over the after
part, and that it is considered to be a primary mechanism of
resistance reduction. On the other hand, this change in
pressure distribution has an influence on the other sec-
ondary components. Hence there is still room for argument
about the contribution ratio of each component to change in
pressure distribution.
This study aims to clarify the point stated above through
fundamental tank testing. Although the change in running
attitude due to stern flaps can be easily observed in tank
tests, sufficient experimental data describing its effect have
not been reported. Therefore, first of all, we examine this
effect in detail through the tank tests. On the other hand,
visible large changes in the stern wave profile can be also
observed in the tank test. To investigate the effect of the
changes in the wave pattern, wave measurement tests and
pressure measurement tests are conducted for not only
several stern shapes but also stern geometries. Moreover, to
clarify the more detailed mechanism of the resistance
reduction, CFD (Computational Fluid Dynamics) is
applied.
2 Change in running attitudes
As stated in the Sect. 1, so far, it has been considered that a
primary mechanism of resistance reduction is an improve-
ment on the pressure distribution over the after part of the
hull [e.g., 1, 5]. On the other hand, it could be expected to
accumulate the knowledge for stern flap design by reviewing
the resistance reduction from several aspects other than an
improvement on the hull pressure distribution. Therefore, we
started the investigation into the resistance reduction
mechanism through fundamental model testing.
As shown in the following results, changes in trim angle
are clearly observed in resistance testing. Generally
speaking, changes in trim due to transom flaps for high-
speed planing craft affect the powering performance.
Hence, for such a ship and a craft, stern appendages are
effectively used for the purpose of changing the running
attitude [e.g., 11]. On the other hand, maximum Froude
numbers of destroyer-size ships are usually 0.45 or so, at
most. In this speed region, a ship experiences small chan-
ges in running attitude, and it appears that such a trim
change does not strongly affect the resistance reduction. So
far, this point has been examined in some research with the
conclusion that the resistance reduction is caused by a
mechanism other than trim angle [e.g., 12, 13]. However, it
is considered that sufficient data sets to explain this point
have not been shown. Therefore, as a first step of this
research, we tried to examine the trim effect on the
decrease in resistance through tank testing.
In this research, all the tests were conducted in the
Meguro basin of TRDI (Technical Research and Devel-
opment Institute). This tank has a length of 248 m, breadth
of 12.5 m and depth of 7 m. In this paper, according to the
purpose and contents of the test, three models having the
same DTMB 5415 hull form geometry are utilized. The
principal particulars and the body plan are shown in Fig. 1
and Table 1, respectively. The values shown in Table 1 are
those for the equivalent full-scale DTMB 5415. This hull
form is considered to be a good example of a benchmark
hull form for a naval surface combatant. This model has fin
stabilizers and bilge keels as appendages. For the tests
Fig. 1 Body plan of the used model (2.75 m model)
Table 1 Principal particulars ofDTMB 5415 (full scale)
Items Value
Length: Lpp 142.0
Breadth: B 19.06
Draught: d (m) 6.15
Volume: V (m3) 8424.4
Block coefficient: Cb 0.507
J Mar Sci Technol (2016) 21:344–358 345
123
-
discussed in this chapter, the model ship having the hull
form of DTMB5415 with a length of 5.779 m is utilized.
Figure 2 indicates the drawing and the picture, respec-
tively, of the model fitted with the selected stern flap, at a
10� angle relative to the centerline buttock at the stern, andwith a chord length of 1.0 % Lwl. This is a basic flap
geometry in this research, hence the authors call it flap 0.
In all the tests, test set-up is same described as follows.
The resistance is measured by load cells, and sinkage and
trim are measured using potentiometers. Two heave mea-
surement points are set at the forward and after end, and
these two values are converted to sinkage and trim. In the
wave measurement tests, a servo mechanism wave probe
was utilized.
At first, to confirm the performance of flap 0 designed
by the authors, model tests were conducted for model
conditions with and without flap 0. The obtained results are
shown in Fig. 3. In this paper, the definitions of the resis-
tance coefficients are defined by Eqs. 1 and 2:
Ct ¼ Rt=0:5qU2r2=3 ð1ÞCt ¼ Cr þ Cf ð2Þ
Cf is estimated by Schoenherr’s formula. The wave-
making resistance Cw, based on wave analysis, appearing
in this paper, is also non-dimensionalized by 0:5qU2r2=3.Sinkage is positive downward and trim is positive bow-up.
In this test, the range of Froude numbers from
Fn = 0.3–0.45 corresponds to Reynolds Numbers about
1.0 9 107–1.5 9 107. In Fig. 3, a ratio is also presented
which is calculated as the resistance reduction value divi-
ded by the resistance value for the baseline hull. In the
legend of each graph, w/o flap indicates the model condi-
tion without a stern flap, whereas w f indicates that with the
21 20 21 20
170
57.8
10
Fig. 2 The drawing and picture of the model fitted with flap 0. In this picture, white part is flap 0. Here, the vertical struts are the support guideof the flap for model testing
Fn
w/o flap
w/o flap
w f0
w f0
Fn
w/o flap
w f0
Cr
EHP
0
0.01
0.02
0.03
0.04
0.05
0.1 0.2 0.3 0.4 0.5
Cr
-0.5
0
0.5
1
1.5
2
0.1 0.2 0.3 0.4 0.5
Trim
( deg
.)
Fn
0
0.001
0.002
0.003
0.004
0.005
0.1 0.2 0.3 0.4 0.5
Sink
age /
L
-4
0
4
8
12
0.25 0.3 0.35 0.4 0.45 0.5Fn
Cha
nge i
n C
r and
EH
P (%
)
Fig. 3 Comparisons of the obtained residual resistance coefficient, the running attitude between with and without flap 0, and effect of stern flapon effective force power and residual resistance coefficient
346 J Mar Sci Technol (2016) 21:344–358
123
-
flap. The number behind w f is the utilized flap number.
This figure shows that flap 0 decreases the residual resis-
tance Cr and the effective horse power EHP at speeds more
than Froude number of about 0.3. The hump of the
decrease in Cr is located at a Froude number of about 0.35,
where EHP was reduced by about 6 % due to the flap in-
stallation. From these figures, it is confirmed that the
design direction of the base form flap, that is flap 0, is
appropriate. Moreover, in this figure, the running trim
shows visible change, whereas not so large change in
sinkage is observed. To analyze the trim effects on the
resistance reduction in depth, resistance tests for the w/o
condition model with the running attitudes of the w f0
condition, that is the trim and sinkage for the w f0 condi-
tion, were conducted. The test speeds are Froude numbers
of 0.35 and 0.45.
The obtained results are shown in Fig. 4 and Table 2.
Here, w/o R.A. indicates the condition that the running
attitudes are almost adjusted to w f0. As can be seen in
these figure and table, the running attitudes of w f0 and w/
o R.A. are well coincided, but the total resistance is only
slightly changed. The effect of running attitude on
resistance reduction is 19.94 % in Fn of 0.35 and
8.941 % in Fn of 0.45. From this fact, we can notice that
the change in running attitude improves the resistance
performance, but only accounts for at most 20 % of the
total reduction due to a flap in this test condition. This
result supports the opinion that the resistance reduction is
caused by a mechanism other than trim angle [e.g., 12,
13] with rigorous experimental data.
Secondly, the authors conducted a wave analysis.
Cusanelli also conducted a wave analysis using longitudi-
nal cuts to discuss changes in wave-making resistance [e.g.,
1]. Here, we utilized transverse cuts with a longitudinal
position of about 0.75 LWL (4.334 m) aft from the model’s
aft end. The measurements were conducted for Froude
numbers of 0.3, 0.35, 0.4 and 0.45. Figure 5 shows the
comparison of wave profiles between the w/o, w f0 and w/o
R.A. conditions. In this figure, the abscissa, Y, indicates the
transverse position from the centerline of the ship, whereas
the ordinate represents the wave height (mm). From this
figure, a noticeable change in the wave profile can be
observed over a portion relatively close to the ship cen-
terline, and the change in running attitude is observed to
not strongly affect the wave profile. It is also noticed that
the discrepancy between w/o flap and w f condition
becomes smaller with more distance from the center. This
tendency can be also found in the test data conducted for
other Froude numbers.
To investigate the contribution of the change in wave-
making resistance to resistance reduction, wave analysis
based on Thin-ship theory [e.g., 14] was conducted as
shown in Fig. 6. This figure shows a comparison of the
change in the residual and the wave-making resistance
coefficients for each condition. As can be seen, the ten-
dency of the theoretically calculated wave-making resis-
tance agrees well with that of the residual resistance
coefficient obtained from resistance tests. This fact implies
that the change in the wave-making resistance strongly
influences the total resistance reduction.
0.01
0.02
0.03
0.04
0.3 0.35 0.4 0.45 0.5
Cr
Fn
w/o flap
w f0
w/o R.A.
Fig. 4 Comparison of residual resistance coefficient between w/o, wf0 and w/o R.A. conditions, as a function of Froude number Fn
Table 2 Comparison of the test results
w/o w f0 w/o R.A.
Fn = 0.35
R (N) 87.28 83.54 86.57
h (�) -0.07067 20.1568 20.1570fG (mm) 17.00 16.40 16.41
Fn = 0.45
R (N) 220.0 213.5 219.4
h (�) 0.8836 0.6941 0.6925fG (mm) 28.25 27.68 27.59
Bolditalic values are nearly values each other
-30
-20
-10
0
10
20
30
40
50
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
Hw
(mm
)
Y (m)
w/o flapw f0w/o R.A.
Fig. 5 Comparison of wave profiles between w/o, w f0 and w/o R.A.conditions with Froude number of 0.35
J Mar Sci Technol (2016) 21:344–358 347
123
-
3 Experimental efforts from the viewpointof change in wave pattern
In the preceding chapter, it was confirmed that the wave
pattern after the hull is influenced by a stern flap. Subse-
quent to the above-mentioned work, the authors continued
the investigation from the same viewpoint as this chapter.
In the first section of this chapter, the effect of flap shape
variation on the wave-making resistance is investigated.
Therefore, stern flaps having different configurations are
tested. In the second section, the effect of stern configu-
ration on the wave-making resistance is examined. Here,
the similarity of stern flap effects to stern extensions is
examined through resistance tests and pressure measure-
ment tests using a model with a cruiser stern.
3.1 Effect of flap shape variation
At first, six stern flaps having different geometries were
examined in a resistance test. The hull form of the used
model is the same as that in chapter 2, but the model has a
length of 2.75 m. This model has fin stabilizers and bilge
keels as well as two shafts and their brackets as appen-
dages. In designing these stern flaps, the authors focused on
the flow around the transom end. For ships with a transom
stern and stern flap operating at high speed, the flow
smoothly detaches at the transom and an end hollow is
formed right behind the transom. A picture describing this
transom flow is shown in Fig. 7. Sakao et al. [15] described
this local flow as a hollow bounded by the flow detached
along the ship side part and along the ship bottom, with
water rising to a point where the two flows meet behind the
transom. During tank tests, the change in the flow behind
the transom due to stern flaps was clearly observed. This
flow change could influence the resistance reduction or
improvements in the afterbody pressure distribution.
Therefore, the authors sought to investigate the relationship
between resistance reduction and the change in generated
waves behind the transom by testing stern flaps with dif-
ferent shapes. The drawing of flaps 1–5 is shown in Fig. 8,
and their particulars are summarized in Table 3. In this
table, percentage means the length of the chord against the
ship length LWL. Further, the bottom area angle means the
value relative to the buttock at the stern, and the angle
concerning wall side area represents the value relative to
the ship side at the stern. In this test, the range of Froude
numbers from Fn = 0.3–0.45 corresponds to about Rey-
nolds numbers from 3.0 9 106–5.0 9 106.
Flap 1 and flap 2 are the flaps which extended to the side
upper part, and particularly flap 1 has an angle of attack
relative to the ship side. Flap 3 has only a side part, and is
intended to change the flow detached along the side. Flap 4
has only a center part. Flap 5 has the extended side part
similar to flap 2, but the chord length of the side part is
different. In these flaps, flap 1, flap 2, flap 3 and flap 5 aim
to deform the flap detached along the ship side.
Figure 9 shows a comparison of change in resistance
coefficients between flaps 1–5 at a Froude number of 0.45.
In this figure, concerning with resistance coefficients,
wave-making resistance obtained from wave analysis [e.g.,
14] is also plotted. Here, we utilized transverse cuts at a
longitudinal position of about 1.5 LWL (4.133 m) aft from
the ship aft end.
Here, as can be seen in Fig. 9, the correlation between
Cr and Cw in this chapter is relatively worse than the results
shown in the last chapter. In the last chapter, we utilized a
larger size model having a length of 5.779 m, whereas the
model size in this chapter is 2.75 m. The model scale could
affect the measurement accuracy of Cr and Cw. From the
wave analysis results, however, a tendency can be observed
that the flaps for which a large decrease in resistance
coefficient Cr is achieved also show a large reduction in
-0.0005
0.0000
0.0005
0.0010
0.0015
0.0020
0.0025
w f0(Fn=0.35)
w/o R.A.(Fn=0.35)
w f0(Fn=0.45)
w/o R.A.(Fn=0.45)
Cha
nge
in C
r and
Cw
Change in CrChange in Cw
Fig. 6 Comparison of change in the residual and the wave-makingresistance coefficients between w/o, w f0 and w/o R.A. conditions, as a
function of Froude number FnFig. 7 The picture describing change in the transom flow due to sternflap. Left picture indicates the transom flow for the model without
flaps whereas right picture does that for the model with flap 0
348 J Mar Sci Technol (2016) 21:344–358
123
-
wave-making resistance Cw. This implies a significant
correlation between the change in Cr and Cw due to the
stern flaps.
From Fig. 9, all flaps except flap 3 largely decrease
residual resistance Cr. Here, the flaps are categorized into
three groups by their tendency to produce a resistance
reduction. The first group consists of flap 0, flap 1 and flap
2. The second group consists of flap 4 and flap 5. The third
group includes only flap 3. Concerning the first group,
there is not a large difference between flap 1 and flap 2 in
residual resistance. From this fact, it can be inferred that
attack angle relative to the ship side does not strongly
affect the flow behind the stern, and that the resistance
reduction is not sufficiently achieved in this test condition.
However, wetness of the side part is not as large for this
case. Therefore, it could be concerning with small change.
Furthermore, the amount of reduction seems to be almost
the same as that for flap 0, so the decrease in the wave-
making resistance is established by changing the flow
detached along the ship bottom. Each flap in the second
group has a large bottom area compared with the other
Fig. 8 The drawing of the stern flaps 1–5
Table 3 Particulars of the flaps 1–5
Bottom area Wall side area
Center Side
Flap 0 1 % (10�) 1 % (10�) –Flap 1 1 % (10�) 1 % (10�) 1 % (10�)Flap 2 1 % (10�) 1 % (10�) 1 % (0�)Flap 3 – 1 % (10�) 2 % (10�)Flap 4 2 % (10�) – –Flap 5 1 % (10�) 2 % (10�) 2 % (0�)
Here, percentage means the length of the chord against the ship length
LWL. Further, the angle concerning bottom area means the value
relative to the buttock at the stern, and the angle concerning wall side
area represents the value relative to the ship side at the stern
0.000
0.001
0.002
0.003
0.004
0.005
0.006
w f0 w f1 w f2 w f3 w f4 w f5
Cha
nge
in C
r and
Cw
Change in Cr
Change in Cw
Fig. 9 Comparison of change in the residual and the wave-makingresistance coefficients between w f0, w f1 and w f2 conditions with
Froude number of 0.45
J Mar Sci Technol (2016) 21:344–358 349
123
-
groups, and it is considered that the resistance reduction for
this group could be achieved by the large bottom area.
Finally, the resistance reduction of the third group, that is
flap 3, is relatively limited in the high-speed region. It is
summarized from the above arguments that the decrease in
wave-making resistance Cw due to flaps could be achieved
by changing the flow detached along the ship bottom in this
condition.
Based on the above facts, the further tank tests were
conducted. The newly designed stern flaps here from the
viewpoint of changing the flow detached along the ship
bottom are shown in Fig. 10 and Table 4. Flap 6 and flap 7
have 10� angles relative to the centerline buttock at thestern, and a chord length of 1.0 % Lwl. Flap 6 and flap 7
have a span of seventy five percent and fifty percent of flap
0, respectively. Flap 8 has the same span as flap 0, but this
flap is at an 8� angle relative to the centerline buttock at thestern and at 12� relative to both side ends. The anglecontinuously changes in the span direction. Flap 9 has a 0�angle relative to the buttock at the stern, and this flap was a
chord length of 1.0 % Lwl, so it can be regarded as an
extension of the stern.
The hull form geometry is the same as that in the above-
mentioned tests, but the model has a length of 3.5 m. In this
test, the range of the Froude numbers from Fn = 0.3–0.45
corresponds to about Reynolds numbers between
5.0 9 106–7.0 9 106. Figure 11 denotes the comparison of
the resistance coefficients between flaps 6–9 at a Froude
number of 0.45. In this figure, wave-making resistance
obtained from wave analysis [e.g., 14] is also plotted. For
calculation of wave-making resistance, transverse cuts at a
longitudinal position of about 1.5 LWL (5.250 m) aft from
the ship aft end were used.
From this figure, in the same way as the above results, a
significant correlation between the change in Cr and Cw due
to flaps is observed. Comparing the results of flap 0, 6
Fig. 10 The drawing of the stern flaps 6–9
Table 4 Particulars of the flaps 6–9
Angle Span (%)
Center Side
Flap 0 10� 10� 100Flap 6 10� 10� 75Flap 7 10� 10� 50Flap 8 8� 12� 100Flap 9 0� 0� 100
350 J Mar Sci Technol (2016) 21:344–358
123
-
(75 % span of flap 0) and 7 (50 % span of flap 0), it is
noticed that the flap span strongly affects the resistance
reduction in these test conditions, so it is considered that
the wider span flap is more effective. Concerning with flap
8, which has the continuously changing angle relative to
the buttock, a larger decrease in resistance than the original
flap is achieved. This implies that a small adjustment of the
flap angle to span direction could lead to better perfor-
mance, and that such a design challenge should be explored
in future work. Concerning flap 9, it can be seen that the
flap at 0� shows a decrease in resistance. In the literature [5,10], the effect for flaps at 0� is mentioned, and the resultsobtained in this study correspond with the results shown in
this literature. As described above, the flap at 0� is regardedas a kind of transom extension, so this result implies that
there exists a relationship between the flap installation and
transom extension from the viewpoint of wave making at
the transom.
3.2 Research from the viewpoint of the transom
geometry
Next, the authors tried to investigate the decrease in
resistance from the viewpoint of stern geometry. Although
recent combatants usually have a transom stern which is
characterized by a near vertical ending, lower speed com-
mercial vessels and past combatants sometimes have a
cruiser stern. As shown in the literature [e.g., 16], a cruiser
stern shows better performance in the lower speed region.
On the other hand, it is usually considered that in the higher
speed region, transom sterns tend to generate less resis-
tance. In Sect. 3.1, the authors showed that the flap at a 0�angle, which can be regarded as a transom extension, is
examined, and has the potential to decrease the resistance.
It is considered that an extension of 1 % Lwl could
improve the wave behind the stern. In this section, based on
the results shown in Sect. 3.1, the authors attempt to study
the relationship between the transom extension and
installation of flaps.
The hull form of the used model is the same as that in the
above-mentioned tests, and the model has a length of 3.5 m.
A drawing is shown in Fig. 12. Here, c1, c2 and c3 have a
transom extension length of 5 % LWL, 7.5 % LWL and 10 %
LWL, respectively. In this test, not only resistance measure-
ments but also pressure measurements are carried out. Fur-
ther, Fn is calculated for the original LWL although the ship
length changes due to the transom extension. Here, pressure
taps P1 and P2 are set to 0.62 % LWL and 2.8 % LWL from the
aft end at the centerline, respectively. The tests were con-
ducted for the conditions in which the running attitudes are
free and fixed. This is because by comparing the data for
different transom shapes with the same running attitude,
dependence of the resistance and pressure on transom
geometry can be observed. In a fixed condition, the running
attitudes of each condition are set to be those of the w/o f
condition. In the following descriptions, fix means the con-
dition in which the running attitude is fixed, whereas free
means the condition in which the running attitude is free.
At first, the resistance test data for the free condition are
shown in Fig. 13. In the following, the change in the
resistance is the value based on the resistance for w/o flap
(free). As can be seen in this figure, c1 (free) shows a larger
decrease in resistance than flap 0 until about a Froude
number of 0.45.
-0.0005
0.0000
0.0005
0.0010
0.0015
0.0020
0.0025
0.0030
w f0 w f6 w f7 w f8 w f9
Cha
nge
in C
r and
Cw
Change in Cr
Change in Cw
Fig. 11 Comparison of change in the residual and the wave-makingresistance coefficients between w f0, w f6, wf7, w f8 and w f9
conditions with Froude number of 0.45
Fig. 12 The drawing of transom extension, that is c1, c2 and c3, andflap 0
J Mar Sci Technol (2016) 21:344–358 351
123
-
Figure 14 shows the change in the residual resistance
and wave-making resistance at a Froude number of 0.45.
As can be seen in this figure, the change in the resistance of
w/o (fix) is very small, so it is understood that the fixed
condition tests are successfully conducted. It can be also
noticed that the change in the resistance of w f0 (fix) is
small. This implies that the running attitude could not
strongly affect the decrease in this test condition. This fact
coincides with the results obtained in chapter 2. The
residual resistance and wave-making resistance for c1, c2
and c3 conditions exhibit a larger decrease. Here, we uti-
lized transverse cuts at a longitudinal position of about
1.5 LWL (5.250 m) aft from the ship original aft end. From
the wave profiles about c1 shown in Fig. 15, wave heights
become relatively smaller due to the transom extension, so
it contributes to the decrease in wave-making resistance.
Further, it is observed that the wave profile for w f0 (fix)
and w c1 (fix) conditions show a similar tendency in this
test condition. It could be our future task to explore the
mechanism of this point.
Figure 16 shows the pressure measurement results.
Here, the pressure is non-dimentionalized by 0:5qU2. It isunderstood from this figure that the flap installation and
transom extension lead to the increase of the pressure at the
after portion of the hull. Increase in pressure implies a
decrease in flow velocity.
Figure 17 shows the variation in pressure due to flap
installation at the P1 and P2 positions, respectively. Here,
the running attitude is free for both conditions. From these
figures, pressure of the ship after part decreases along with
an increase of ship forward speed, but the pressure is
recovered due to the installation of the flap. This recovery
at the after portion of the hull results in the decrease of
resistance. About the pressure improvement, the more
detailed explanation is shown in the next chapter.
4 CFD analysis
Through the extensive model tests in this study, the effect
of wave making on decrease in resistance is clarified.
Finally, in this chapter, in order to investigate this subject
more, CFD analysis was done using STAR-
CCM ? v8.06.007. The mesh geometry applied for CFD
-9.54E-18
0.001
0.002
0.003
0.004
0.005
0.25 0.3 0.35 0.4 0.45 0.5
Cha
nge
in C
r
Fn
w f0 (free)
w c1 (free)
Fig. 13 Comparison of change in the residual resistance between wf0 and w c1 conditions, as a function of Froude number Fn
-0.001
0.000
0.001
0.002
0.003
0.004
0.005
0.006
w/o f (fix) w f0 (free) w f0 (fix) w c1 (fix) w c2 (fix) w c3 (fix)
Cha
nge
in C
r and
Cw
Change in CrChange in Cw
Fig. 14 Comparison of change in the residual and the wave-makingresistance with Froude number of 0.45
-20
-10
0
10
20
30
40
50
60
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
Hw
(mm
)
Y (m)
w/o f (fix)w f0 (fix)w c1 (fix)
Fig. 15 Comparison of wave profiles between w/o, w f0 and w c1conditions with Froude number of 0.45
-0.05
0.00
0.05
0.10
0.15
w/o(free)
w/o(fix)
w f0(free)
w f0(fix)
w c1(fix)
w c2(fix)
w c3(fix)
P'
P1
P2
Fig. 16 Comparison of the pressure with Froude number of 0.35
352 J Mar Sci Technol (2016) 21:344–358
123
-
analysis is shown in Fig. 18. Here, by introducing the
symmetry boundary, computation is done only for half
width. The mesh was refined around the hull and free
surface to capture the Kelvin wave as shown in Fig. 18. We
also refined the mesh around the bow and transom to track
high waves occurring in these regions. The size of the mesh
refinement region was determined for each speed so that
the maximum wave height could be captured in about 20
grid points. The hull form of the used model is the same as
that in the above-mentioned tests, and the model has a
length of 5.779 m. To reduce computational costs, a bare
hull model is used here. The computational domain is
determined based on the guideline shown by ITTC [17]:
the length from fore perpendicular to inlet boundary is set
to 1.5 LWL, that from aft perpendicular to pressure
boundary is set to 3.5 LWL, and that from ship bottom and
side of the domain is set to 1.5 LWL. To prevent the wave
reflection on the boundary, VOF wave damping model is
applied in the region within 2 meter of each boundary.
The utilized setting data are shown in Tables 5 and 6. A
linear ramping of under-relaxation factors is used for tur-
bulence model and VOF model to mitigate the instability
caused by giving a uniform flow as the initial condition.
The thickness and the number of prism layers are the same
in all cases. We confirmed that the maximum y? of the
closest grid points to the ship surface was about 70 and the
wall model is valid in this range. The simulations were run
up to 45 s. However, calculated total resistance fluctuated
after 45 s calculation due to the unsteadiness. For this
reason, we took the mean value of Ct in the last fluctuation
cycle. The amplitudes of Ct in the last cycle were less than
0.5 % in most cases, except for three cases (Fn = 0.20
with flap and 0.25 with and without flap) having about 1.5
to 2 % fluctuation.
0
0.05
0.1
0.15
0.1 0.2 0.3 0.4 0.5
Cha
nge
in P
'
Fn
P1P2
Fig. 17 Change in pressure at P1 and P2, as a function of Froudenumber Fn
Fig. 18 Mesh geometry aroundthe hull (upper X–Y view, lower
X–Z view)
J Mar Sci Technol (2016) 21:344–358 353
123
-
At first, the authors validate the CFD analysis with the
tank test results. Figure 19 shows the obtained total resis-
tance coefficient by EFD and CFD. Here, CFD calculations
were done in 25 �C. Therefore, resistance coefficientsobtained by CFD were corrected to 10 �C in the samemanner as Longo and Stern [18]. Calculated resistance
coefficients are still about 2.5 to 6.1 percent lower than the
EFD results. These differences are considered to be caused
by the presence or absence of appendages. A bare hull
model was used in CFD, while a hull with fin-stabilizer and
bilge keel was used in EFD. Figure 20 shows the com-
parison of change in resistance coefficient by EFD and
CFD, and the results of CFD agree well with that of EFD. It
is noteworthy that the forward speed at which the flap starts
to work is well predicted.
Figure 21 (left) shows the longitudinal distribution of
mean pressure on the hull surface. The hull was split into
strips whose width was LWL/40. Then, pressure was inte-
grated on each strip and the integrated pressure was divided
Table 5 Dominant calculation models and common parameters
Item Model or parameter Item Model or
parameter
Time marching Implicit unsteady Under-relaxation factor (VOF) 0.8
Fluid flow solver Segregated flow Under-relaxation ramping (k, e,VOF)
Initial value 0.1
Linear up to 10
iterations.
Sub iterations 5
Surface capturing VOF model Surface mesh size on ship 50 mm
Turbulent model Realizable k-e Cell size in stern region 10 mm
Wall model Hybrid model of wall function and low Re RANS. (called as
All y? wall treatment)
Total thickness of prism layer 20 mm
Under-relaxation factor
(Pressure)
0.4 Number of prism layers, and
stretching ratio
15 and 1.1
Table 6 Calculation condition and dominant parameters
Fn Dt (s) Anisotropic refinement around free surface (over all) Anisotropic refinement around free surface (bow region)
Dz (mm) Vertical range (mm) Dz (mm) Vertical range (mm)
0.20 0.025 5 -60 to 60 9 -50 to 90
0.25 0.025 7.5 -75 to 75 10 -50 to 110
0.30 0.020 9 -90 to 90 10 -50 to 130
0.35 0.020 10 -100 to 100 10 -50 to 150
0.40 0.015 10 -130 to 130 10 -50 to 200
0.02
0.03
0.04
0.05
0.06
0.2 0.25 0.3 0.35 0.4
Ct
Fn
w f0 (CFD)
w f0 (EFD)w/o f (EFD)
w/o f (CFD)
Fig. 19 Comparison of the total resistance obtained by EFD andCFD for w/o and w f0 conditions as a function of Froude number Fn
-0.001
0.000
0.001
0.002
0.2 0.25 0.3 0.35 0.4
Cha
nge
in C
t
Fn
EFD
CFD
Fig. 20 Comparison of change in the total resistance obtained byEFD and CFD for w/o and w f0 conditions as a function of Froude
number Fn
354 J Mar Sci Technol (2016) 21:344–358
123
-
by the surface area of each strip to obtain mean pressure.
Here, the value is gauge pressure and non-dimensionalized
by the stagnation pressure. Figure 21 (right) shows the
difference in mean pressure corresponding to Fig. 21 (left).
As can be seen in these figures, the pressure on the ship
fore part is not disturbed by the stern flap, and it is only
affected on the hull before the flap. This pressure recovery
observed in the after body can be considered as direct
mechanism of decrease in resistance.
At first, to confirm the effect of the running attitudes on
the decrease, the calculations for reverse running attitude
condition (R.R.A) are carried out. R.R.A. means w/o f
condition with the running attitude of w f0 test data or w f0
condition with the running attitude of w/o f test data. Fig-
ure 22 shows the comparison of change in total resistance
due to stern flap between normal running attitude and
reverse running attitude (R.R.A). As shown in Fig. 22, an
almost equivalent order of decrease is observed even for
the R.R.A. condition, so it is concluded again that the
running attitude is not the dominant mechanism for the
resistance reduction due to installation of the flap.
Next, the authors conduct the investigation for the stern
flow using CFD analysis. As shown above, the stern wave
is affected by a flap and it could lead to the pressure
recovery in after body. Figure 23 shows the wave height
distribution with a Froude number of 0.35. In the region
soon behind the transom, contour lines for w/o f are
denser than those for w f0, and this means that the wave
steepness is larger. Generally speaking, steeper waves
may lead to the wave breaking, so stern flaps could
contribute to decrease the wave breaking soon behind the
ship. To look in more detail, Fig. 24 shows the longitu-
dinal wave cut on the center line in the vicinity of the
transom. Wave height increases just after the stern with-
out a stern flap. This makes the wave profile steeper. On
the other hand, that increases moderately with the stern
flap. However, higher maximum height appears in the
case with the stern flap. It seems that a steeper wave
profile rather than the wave height itself makes resistance
larger. From this point of view, it is a future work to
survey the relationship between these local changes of
wave profile and wave-making resistance. As Raven et al.
[19] pointed out, change in wave pattern of the stern
wave played a big role in the scale effect. Therefore, it is
desirable to consider the change of wave profile men-
tioned above from this point of view.
Finally, the comparison of stream lines is shown in
Fig. 25. As shown in this figure, the stern flap deflects and
slightly thins stream lines in the vicinity of the stern. These
deflections and thinning seem to cause pressure recovery as
shown in Fig. 21. Furthermore, the stern flap thickens
stream lines behind and makes a contracted flow. This
contracted flow seems to make a moderate wave profile as
shown in Fig. 24. It is also a future work to survey the
relationship between these local changes of flow pattern
and wave-making resistance as mentioned before.
5 Concluding remarks
The obtained primary outcomes from this research are
summarized as follows:
-0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
-0.2 0 0.2 0.4 0.6 0.8 1 1.2M
ean
pres
sure
of e
ach
stip
P'
x/L
w/o f(CFD)
w f0(CFD)
-1.0
0.0
1.0
2.0
3.0
4.0
5.0
-0.2 0 0.2 0.4 0.6 0.8 1 1.2
Diff
eren
ce in
P' [
%]
x/L
Fig. 21 Left mean pressure ofeach strip with Froude number
of 0.35. Here, concerning
x direction, 0.0 denotes fore
perpendicular, whereas 1.0
denotes aft perpendicular.
Pressure was non-
dimensionalized by the
stagnation pressure. Right
difference in mean pressure of
each strip with Froude number
of 0.35
-0.001
0.000
0.001
0.002
0.20 0.25 0.30 0.35 0.40
Cha
nge
in C
t
Fn
w f0
w f0 R.R.A.
Fig. 22 Comparison of change in total resistance due to stern flapbetween normal running attitude condition and reverse running
attitude (R.R.A)
J Mar Sci Technol (2016) 21:344–358 355
123
-
1. The decrease in resistance due to the installation of the
flap is successfully observed.
2. The effect of the running attitudes on the decrease is
examined by tank tests and CFD analysis, and then it is
concluded that the change in running attitudes is not
the primary mechanism of the resistance reduction in
the destroyer’s speed region.
3. The wave analysis results show that there is a
significant correlation between decrease in total resis-
tance and wave-making resistance. This suggests that a
flap affects a change in the wave generated at the
transom part. Further, it is confirmed that the primary
decrease in wave-making resistance could be derived
w/o f
w f0
(1) Overall
w/o f
w f0
(2) Stern region
Wave height [m]
Fig. 23 Waveheight distribution (Fn = 0.35)
-40
-20
0
20
40
60
80
100
0 0.1 0.2 0.3 0.4 0.5
Wav
e he
ight
[mm
]
Distance from stern / ship length
w f0 (CFD)
w/o f (CFD)
Fig. 24 Longitudinal wave cut on the center line in the vicinity ofstern (Fn = 0.35)
356 J Mar Sci Technol (2016) 21:344–358
123
-
by a change in flow detaching from the ship bottom
due to a flap in this test condition.
4. From the viewpoint of the stern geometry, the tank
tests are examined. Then, for a particular case, similar
tendencies of the wave profile between the stern
extension and the installation of the flap are observed.
The further exploration of this point could be our
future task.
5. The pressure recovery is observed in tank tests and
CFD analysis due to a flap. It is reconfirmed by CFD
analysis that the region in which this pressure recovery
only appears to the part of the hull just before a flap.
In this research, the authors attempt to clarify the
decrease in resistance due to the installation of a flap by
conducting several tank test measurements and CFD
analysis. As introduced in the first chapter, Cusanelli et al.
[e.g., 1] explained that the primary mechanism by which a
stern appendage works is the change in pressure distribu-
tion of the after body of the hull. Through this research, it
becomes clearer that there exists a correlation between the
hull pressure improvement and change in wave making at
the transom. Therefore, our conclusions are considered to
be one of the interpretations of the improvement on pres-
sure distribution from the aspect of wave making.
Acknowledgments First and foremost, I would like to thank Dr.Kazuyuki Yamakita (Director of Naval Systems Research center,
Ministry of defense) for his support and encouragement. The authors
are also grateful to Dr. Masahiro Tamashima from Fluid Techno Co.,
Ltd. for his technical advice for pursuing the research. Sincere
appreciation due to Mr. Hisatada Itoh, Mr. Minoru Atsumi, Mr.
Noriyuki Yokoyama, Mr. Mashiko Hidetake and Naofumi Yamato at
Meguro basin, for their assistance in model testing.
References
1. Cusanelli DS, Karafiath G (1997) Integrated wedge-flap for
enhanced powering performance. FAST 97:751–764
2. Cusanelli DS (2011) Hydrodynamic supportive structure for
gated ship sterns—amphibious ship stern flap. In: 11th Interna-
tional Conference on Fast Sea Transportation, pp 421–429
3. Cusanelli DS, Karafiath G (2012) Hydrodynamic energy saving
enhancements for DDG 51 Class Ship, ASNE Day, American
Society of Naval Engineers
4. Cave WL, Cusanelli DS (1993) Effect of stern flaps on powering
performance of the FFG-7 class. Mar Technol 30(1):39–50
5. Cusanelli DS, Hundley L (1999) Stern flap powering performance
on a Spruance Class Destroyer: ship trials and model experi-
ments. Nav Eng J 111(2):69–81
6. Cusanelli DS (2009) Scaling effects on stern flap performance,
Progress Report, NSWCCD
7. Cusanelli DS, O’Connell L (1999) US Coast Guard Island Class
110 WPB: stern flap evaluation and selection (Model 5526),
Hydromechanics Directorate Report, NSWCCD
8. Cusanelli DS, Barry CD (2002) Stern flap performance on 110 ft
Patrol Boat WPB 1345 STATEN ISLAND, Hydromechanics
Directorate Report, NSWCCD
9. Cusanelli DS, Shen YT, Bishop RC (2004) Active stern flaps—
for ship motion control, fuel savings, and speed improvement,
Carderock Division, NSWC-Technical Digest, CARDER-
OCKDIV-04/CT02, pp 50–61
w f0 (CFD)
w/o f (CFD)
Fig. 25 Stream line around the ship transom (Fn = 0.35)
J Mar Sci Technol (2016) 21:344–358 357
123
-
10. Persons GM, Singer JD, Gaal MG (2006) Multicriterion opti-
mization of stern flap design. Mar Technol 43(1):421–454
11. Savitsky D, Broun PW (1976) Procedure for hydrodynamic
evaluation of planing hulls in smooth and rough water. Mar
Technol 13(4):381–400
12. Thornhill E, Cusanelli D, Cumming D (2008) Stern flap resis-
tance reduction for displacement hulls. In: 27th symposium on
naval hydrodynamics
13. Karafuath G, Fisher CS (1987) The effect of stern wedges on ship
powering performance. Nav Eng J: 27–38, 1987
14. Sakuma S (2003) The wave analysis of a box barge using thin-
ship theory, technical report of Technical Research and Devel-
opment Institute, Japan Defense Agency, pp 1–18 (in Japanese)15. Sakao M, Shimoya N (1980) On stern waves of transom-stern
ships. J Kansai Soc Nav Archit 179:27–34 (in Japanese)
16. Tamura K (1975) A power prediction method for ships with
transom stern, Mitsubishi Heavy Industries Technical Review.
12(5):1 (in Japanese)17. Specialist Committee on CFD in Marine Hydrodynamics (2011)
Practical guide lines for ship CFD applications, ITTC 7.5-03-02-
03
18. Longo J, Stern F (2005) Uncertainty assessment for towing tank
tests with example for surface combatant DTMB Model 5415.
J Ship Res 49(1):55–68
19. Raven HC, van der Ploeg A, Starke AR (2008) Toward a CFD-
based prediction of ship performance—progress in predicting
full-scale resistance and scale effects. Int J Marit Eng. RINA
MARINE CFD conference
358 J Mar Sci Technol (2016) 21:344–358
123
-
本文献由“学霸图书馆-文献云下载”收集自网络,仅供学习交流使用。
学霸图书馆(www.xuebalib.com)是一个“整合众多图书馆数据库资源,
提供一站式文献检索和下载服务”的24 小时在线不限IP
图书馆。
图书馆致力于便利、促进学习与科研,提供最强文献下载服务。
图书馆导航:
图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具
http://www.xuebalib.com/cloud/http://www.xuebalib.com/http://www.xuebalib.com/cloud/http://www.xuebalib.com/http://www.xuebalib.com/vip.htmlhttp://www.xuebalib.com/db.phphttp://www.xuebalib.com/zixun/2014-08-15/44.htmlhttp://www.xuebalib.com/
Fundamental research on resistance reduction of surface combatants due to stern flapsAbstractIntroductionChange in running attitudesExperimental efforts from the viewpoint of change in wave patternEffect of flap shape variationResearch from the viewpoint of the transom geometry
CFD analysisConcluding remarksAcknowledgmentsReferences