November 197 1
LABORATORIUM VOOR
SCHEEPSBOUWKUNDE
TECHNISCHE HOGESCHOOL DELFT
THE EFFECT OF DESIGN MODIFICATIONS ON THE NATURAL COURSE STABILITY OF
FULL TANKER MODELS
by
Ir. C.C. Glansdorp
and
J.G.L. Pijfers
.
List of Fig4
Fig. Body plans of models.
Fig. 2A Arrangement of rudders of the original model.
Fig. 213 Arrangement of rudders of the modfied model-Free running propeller.
charteristic-Fig. 4 'Rudder forces and rudder sosonoc of ruder 1.
Budder. forces and rudder mument,...; of rudder IV.
increae of increment of sewa:ying derivative as a funtion
.of revulutions of the prs.peller.
amplitde-Yt-m', N' r r Y' N' V
'
v YI NI 6 ' 6 m' A Y etc. subscript s subscript R subscript p subscript s.T rudder area ship length ship speedvelocity of water near the rudder
revolutions per minute of ship propeller coefficient of advance
torque coefficient thrust coefficient
yaw damping derivatives sway damping derivatives
rudder derivatives dimensionless mass increment of derivatives refers to skeg refers to rudder refers to propeller refers to slinstream
swaying velocity amplitude
CL c)t
lift gradient
wake factor
6 rudder angle
Centat.,:its.
4.
erìt..t.
tha* rett!ts of the to::;t-t.
'
L.;7 :Eesi:Cits nf -fr-esor
íTro,elier tets.
4.3 Pesuits tf rt:ider
General.
(:omvariscn of t-w-
'rare
Effet. of Hkec.
5.4 Effe.::tt hf soliese st.- in
slip
.. .treaTo.5.5 Eff,*(2t
() Efft of'
t:t1,-*Summary.
The course stability characteristics of two full tanker-afterbodies with different rudder-propeller-skeg configurations are experimentally
determined.
Natural course stability characteristics are improved by adding lateral force generating devices in the afterbody.
Introduction.
The improvement of stability characteristics of full course unstable vessels can be done principally in two ways. The first is to install an autopilot or a similar electronic device aboard the ship in order to achieve directional stability of the newly created system, a course unstable vessel together with an electronic device. This solution is very attractive in an economic
sense.
The adjustment of the autopilot characteristics to make a course unstable ship directionally stable is somewhat more difficult than for a course stable ship. Nevertheless current practice shows that large tankers, which are very often course unstable, equiped with suitable autopilots have a reasonable performance in unrestricted waters and these large tankers are certainly directionally stable.
Another way to improve the stability characteristics is to modify the after-body and to improve the lateral force generation of the control devices in the afterbody by well designed new configurations of the system:
skeg-propeller- rudder. In this case the control-fixed stability or course stability is improved and certainly facilitates the achievement of
directional stability by either a helmsman or an autopilot. Some work in this field, regarding
ship born manoeuvring devices has been done by English and Steel [1]. However some of their experimental
set-ups are expected to be very expensive in practice due to the fact that they need a more detailed development
before practical application. Therefore it is questionable whether these devices will be applied
on large ships when relatively cheap electronic control devices are available which
can
2
-Design modifications to improve the stability characteristics should be such that they do not effect the building costs of ships to a perceptible extent including the costs of developing those modifications especially regarding problems of strength and vibrations.
In this paper relatively small design modifications of a parent model are considered; the linear damping derivatives combined in a course stability criterium are measured with a P.M.M. (Planar Motion Mechanism). It is shown that these small modifications produce a relatively small improvement of
course stability. None of the investigated configurations could make a very unstable naked hull course stable. But one of the better configurations
together with a well designed electronical controlling device will contribute, in the author's opinion, to a more secure and more economic directionally stable system. It is known that the performance index as proposed by Motora and Koyama [2] is a very good measure in comparing performances of different directionally stable systems exposed to random disturbances. Within
reasonable limits it is felt that the less course unstable a ship is, the better the performance index will be if we assume that the ship has a suitably adjusted autopilot on board.
3. Model particulars and test progam.
Two models of a 200.000 tons deadweight tanker were used for the experimental investigations; they were made of reinforced fibre glass. The model of an excisting ship is considered as the parent model; the afterbody of the second model has been modified. The ship particulars are given in Table 1
including the particulars of the modified model in full scale dimensions. Table 1: Ship and model particulars.
original model modified model Length between perpendiculars 310.00 m 310.00
m Breadth 47.16 m 47.16
m
Draft 18.90 m 18.90m
Displacement 235,000 m3 232,690 m3Length center of boyancy +8.73 m
+9.81 m
Propellerdiameter 8.80 m
8.80 m Pitch ratio
.700 .700
Blade area ratio .5143
.5143
Model scale 1:100
The body plans can be seen in fig. 1. The particulars of the used rudders are given in Table 2.
Table 2: Rudder dimensions.
With respect to control fixed stability the fixed portion of the rudder can be considered as being an integrated part of the rudder with respect to the calculation of the lift gradient. For turning circle calculations another lower value has to be used regarding the effect of camber at a certain rudder angle and the adverse effect of the fixed portion of the rudder.
Rudders I and II are used in connection with the original ship; rudders III, IV and V in connection with the modified model. The lift gradient mentioned in Table 2 is calculated according to a formula published
by Whicker and Fehlner [3].
A survey of the configurations of the hull and the appendages and
the
designation of these configurations for reference in section 4 where the results are presented, is given in Table 3.
original ship modified ship
I II III
*
*-IV
*-v.
balanced balanced Mariner fixed portion twin rudders fixed portion aspect ratio sweep angle taper ratio upper chord lower chord rudder height profile thick-ness ratio area L 2.87 0 1 7.25 m 7.25 m 10.40 m NACA0018 75.4 m2 3.063 2.30 0 1 10.00 m 10.00 m 11.50 m NACA0018 115 m2 2.683 3.36 100 .612 8.50 m 5.20 m 11.50 m NACA0018 78.8 m2 3.311 2.30 0 1 10.00 m 10.00 m 11.50 m NACA0018 115 m2 2.683 3.37 0 1 4.75 m 4.75 m 8.50 m NACA0018 2*38 m2 3.332 act
M(odified) ship, speed 15 knots.
_ 4 _
Table 3: Survey of testing conditions.
0(riginal) ship, speed 15 knots
When the propeller had been rotating the number of revolutions per minute is indicated in the second or third column.
The measured configurations can be seen in fig. 2A and 2B.
Swaying and yawing tests with the models were performed with the P.M.M. available in the Delft Shipbuilding Laboratory. If the propeller
ran during the P.M.M.-tests torque and thrust measurements were simultaneously carried
out.
Designation propeller rudder
01
--02
N=85 I03
- II 0 4 N=67 II05
N=85 II 0 6 N=102 IIDesignation skeg propeller rudder
Ml
M2
-yes-M3
- N= 85-m 4
yes N= 85-M5
- - IIIM6
yes - IIIMl
-N=85
IIIM8
yesN=85
IIIM9
yes - IV M10 -N=85
IV M11 yes N= 67 IV M12 yes N= 85 IV M13 yes N=102 IV M14 yes - V M15 yes N= 85 VThe free running characteristics of the model propeller have been measured in order to be able to estimate the wake in both cases.
Rudder tests have been carried out with different numbers of revolutions of the propeller and without propeller and the rudder force and rudder moment
exerted on the model were measured for all rudder types used in this investigation.
4.
Presentation of the results of the tests.4.1
Results of P.M.M.-damping derivatives.For all configurations the in-phase and quadrature components of the forces on forward and after strut of the P.M.M. are measured in both modes, swaying and yawing. Due to the relatively high dimensionless
wL ,
frequency w' = during the tests ranging up from
.8
to 2.6, one mayU
expect that frequency effects have an adverse effect on the measurements[4]. Nearly all damping forces and moments show non linearities in the
measured range, so it is difficult to distinguish between linear and non linear part of the forces and moments mainly due to scatter of the few measuring points. It was decided to adopt one fixed non-linear term
for the determination of the linear terms for all configurations belonging to one model. Although it is probable that the cross-flow introducing
non-linear damping terms will not vary very much with the change of configurations an error has been accepted in determining the linear derivatives by this assumption.
In Table 4 the results of the measurements analyzed with the aforementioned fixed non linear term are presented. For the sake of simplicity all
mass derivatives are omitted. The stability criterium used is:
N'
If
Yv-m
yl
then the model is course stable.Reference is made to table 4 where the points of application of the
-6
Table 4: Results of the P.M.M. tests with respect to damping derivatives.
It is to be noted that the dimensionless mass is for the original ship
m'
* 105
= 820and for the modified model m' 105 = 812
4.2 Results of free running propeller tests.
A free running propeller test has been carried out at approximately the same magnitude of the Reynolds number as was the case during the P.M.M.-tests. The results of the tests can be seen in fig. 3.
, ,
swaying yawing
Desig-nation Yv,.105
N'.105
Nv%Yv' (Y,'-m').105 Nr'.105 Nrt/(Yri-m')0 1 -1562 -808 .517 -1232 -255 .207 0 2 -1938 -668 .345 -1086 -314 .289 0 3 -1733 -739 .426 -1109 -302 .272 0 4 -1836 -690 .376 -1092 -307 .281 0 5 -1821 -680 .373 -1079 -318 .285 0 6 -1878 -669 .356 -1053 -331 .314 M 1 -1438 -862 .599 -1272 -232 .182 M 2 -1510 -831 .550 -1228 -256 .208 M 3 -1600 -811 .507 -1205 -250 .207 M 4 -1717 -756
440
-1149 -280 .244 M 5 -1559 -796 .511 -1209 -268 .222 M 6 -1619 -764 .472 -1124 -286 .254 M 7 -1716 -735 .428 -1105 -303 .274 M 8 -1825 -700 .386 -1083 -322 .297 M 9 ..1663 -741 .446 -1104 -297 .269 M10 -1762 -696 .395 -1086 -316 .291 Mll -1768 -692 .391 -1068 -325 .304 M12 -1813 -684 .377 -1054 -332 .315 M13 -1829 -665 .364 -1049 -338 .322 M14 -1722 -748 .434 -1162 -294 .253 M15 -1832 -696 .380 _1105 -320 .289For all rudders the rudderforces and ruddermoments exerted on the ship are measured as a function of the rudderangle. The models sailed along the center line of the towing tank without driftangle. In Table 5 the results of the linear derivatives are presented which are calculated from the experimental points using a least squares fit. In fig. 4 and
5 the results of the measurements for the rudders I and IV are given.
Table
5:
Results of rudder tests.5. Discussion of results.
5.1 General
Qualitatively it is believed that the numerical reEults are accurate
enough for drawing final conclusions. Quantitatively it appears as indicated by the least squares analysis, that the average accuracy is as given
in Table 6.
Differences between derivatives can not be considered significant when they do not exceed the values mentioned in this Table.
Y(5'105 N6'.105
rudder-type propeller propeller
without N=67 N=85 N=102 without N=67 N=85 N=102 72 163 275 372 -34 - 91 -138 -176 II 124 250 305 377 -63 -104 -152 -191 III 94 217 277 330 -45 -101 -133 -166 IV 127 275 344 411 -59 -128 -162 -206 V 57 169 208 237 -29 - 76 - 98 -121
Table 6: Average accuracy of measured derivatives.
IL
Yvl 105
25IA
Nv1.105
15IA
Yr1.105
20-
8
-5.2
Comparison of the two bare hulls (01 and M1).Both models are in bare hull condition very course unstable. The original model shows a point of application of the swaying force in front of the forward perpendicualr, thus indicating a negative damping in the afterbody, probably due to speed effects. The modified model shows an even more pronounced picture. The point of application of the swaying force is even a little bit further ahead of the forward perpendicular. Looking at the stability criterium the ship with the modified afterbody is not better than the original ship with respect to course stability; this fact was not anticipated.
5.3
Effect of skeg (M1 andM2, M3
andM4, M5
andM6, M7
andM8, M10
andM12).
In Table 7 the contributions of the skeg were calculated.
Table
7:
Contributions to derivatives of skeg.From Table 7 it can be concluded that the relation between the skeg
derivatives, indicating the situation of the skeg with respect to
the center of gravity, is almost perfect, regarding the accuracy. Comparing
the
conditions M1 and M2 and M2 and 01 we may conclude that the ship with the
modified afterbody together with a skeg has an equivalent course stability as the original version. Increasing the skeg area improves course stability as already was mentioned by many other authors and especially by
Tsakonas
[5]
and Jacobs [6].5.4
Effect of rudders not in the slipstream.(01 and 03, M2 and M6, M2 and M9, M2 and M1)).
comparing conditions
(AYv')5.105
(Al\y)s.105
(AYr')s.105(AN1,1),.105
M 2
M 1 72 3144
24
M 4
M 3117
55 5630
M 6
M 560
3285
18
M 8
M 7109
3522
-19M12 - M10
- 51
12 32-16
average of all conditions average of derivatives - 82- 82
33
41 48 41-21
21
In Table 8 the contributions of the rudders to the derivatives for the mentioned configurations are calculated and compared with the rudder derivatives Y6' and
N6'
in the same conditions.Table
8:
Contribution of rudder to the derivatives.In this case, the rudder is situated in the wake of the ship, the contribution of 1(AYv')R1 exceeds the magnitude of Ya'.
The explanation of this fact is given when the simple formulas (1) and (2) are compared:
(AYvt)R
= Y AU,
n R DCL U L2 *9a
AR DC uR2 U2 L2 a URIn this case
7
< 1 due to the wake effects, UR = U(1-4), thereforeIY6'
<(AYv')R I°
In Table 4 it can be seen that the rudder, even if it is placed in the wake of the ship has a very favourable effect on course stability.
5.5
Effect of propeller(M3 - M1).
Comparing the conditions M3 and M1 the following contributions to the
derivatives by the propeller can be calculated. Table
9:
Contributions of propeller.(2)
comparing
conditions
(AYv')R*105
(A,N ')105
v R* (AYr')ii*
105 (LNr')R*105Y6'.105
N(5'.1051
0 3 - 01
-171
+69
+123
-47
124-63
M
6- M2
-109
+67
+104
-30
94-45
M
9 - M2
-153
+90
+124
-41
127-59
M14
- M2
-212
+83
+ 66
-38
i 57
-29
condition (AY
v p
)'.105
(ANv)p'.105
(AY )'.105
r p (ANr p) ' 105
M3 - ml
averaged all derivatives-162
-118
+51+59
+67
+59
-18
-29
If the propeller distance to the center of gravity is accounted for new average derivatives can be calculated.
It is to be noted that the fin effect of the propeller (no rudder present) is nearly of the same order of magnitude as the effect of the rudder
in the wake of the ship (no propeller present). The so-called fin effect of the propeller is in fact the generation of a stabilizing transverse force set up by an oblique flow on the propeller.
5.6
Effect of the slipstream.(06, 05, 04
and03, M13, M12, Mil
and M9) In Table 10 the effect of the slipstream on the derivatives has been calculated.Table 10: Effect of slipstream on derivatives.
If we assume a fixed relation between the derivatives then the averages can be calculated. In fig. 6 the results of the average increase of AYv' in the slipstream are plotted as a function of
R.P.M.
Two major items influence the lift generation of the rudder in the
slipstream. The propeller accelerates fluid particles thus giving a higher incoming velocity of the water onto the rudder, and secondly the
transverse velocity of the obligue flow behind the propeller is decreased due to the propeller action, both phenomena introducing the straightening effect of the propeller, making the effectiveness of the rudder less. comparing
condition (AYv's.T.l' 0
(ANvs
)'.T-5
-,
in)s.0
(Ay)' T.105 s. (AN 'NT*1
0 4 - 03
_103
+49
+17
- 5
averaged_ 64
+32
+32
-16
0 5 - 03
- 88
+59
+30
-16
averaged - 82 +41 +41-20
0 6 - 03
_145
+70
+56-29
averaged_128
+64 +64-32
M11 - M9
-105
+49
+36
-28
averaged - 97+49
+49
-2 4M12 - M9
-150
+57
+50
-35
averaged-126
+63
+63
-31
M13 - M9
_166
+76
+55-41
averaged_148
+74+74
-37
Also the straightening effect of the hull plays a certain role in the nature of lift generation of the control devices in the afterbody.
The real nature of this effect and the way to measure it, is still a bit obscure [7].
5.7 Comparison of the rudders in the slipstream (02, 05, M8, M12, M15). This comparison is in fact the most important because the conditions
mentioned are more or less like the situation for the ship sailing in fully loaded condition under maximum power. In Table 11 the results are recalled.
Table 11: Comparison of derivatives under normal sailing.
The following conclusions can be drawn from the results in this table. None of the investigated configurations is course stable. The old version with the original rudder seems to be the best configuration, but with respect to the value of significant
difference of arm the differences in course stability are of minor importance. Compared to the actual performance on full scale of this configuration, reference r8] gives some evidence that this configuration was indeed course unstable.
Comparing 02 and 05 and M8 and M12 show that increasing the
rudder area not necessarily gives a better course stability. Considering an area of 115 m2 as a limit for a single rudder
configuration then it seems justified for achieving course stability to use twin rudders. They ought to have considerably more area in total than applied here during this investigation.
1(1-ea-st,l(g)--n Y-,; .105 N .105 Nv%Yv' (Y; -m')105 N; .105 t Nr AR Yr'-m' 2 0 2 -1938 -668 .345 -1086 -314 .289 75.4 m 0 5 -1821 -680 .373 -1079 -318 .285 115 m2 M 8 -1825 -700 .386 -1083 -322 .297 78.8 m2 M12 -1813 -684 .377 -1054 -332 .315 115 m2 M15 -1832 -696 .380 -1105 -320 .289 2*38 m2 signifi-cant dif-ference in arm .035 .020
-
12
-In Table 12 thecontributions of rudder and propeller are calculated for the 5 different rudders and compared with the results of the rudder tests.
Table
12:
Contributions of rudders and propellers and comparison with rudder derivatives.It can be seen in this table that the propeller contribution to the
derivatives is about 70% of the rudder contribution. Compared to the rudder derivatives Y6' the contribution of the rudder to the swaying derivatives
is about 60% of the rudder derivatives. The total measured value of the influence of rudder and propeller is of the same order of magnitude of the measured value for the rudder derivative during the rudder tests with the exception of the twin rudders.
5.8
Effect of wake.The thrust and torque measurements during the trials and the free running characteristics of the propeller could be used for calculating the wake. designation
(AYv').105
(ANIr')105
(AYr').105 (ANr')105Y6'.105
N6'.105
0 2 - 01
-376
+140
+146
-59
-
_ averaged-296
+148
+148
-74
275
-138
o 5 - 01
-259
+128
+153
-63
-
-averaged-268
+134
+134
-67
305
-152
M 8 - M2
-315
+131+145
-66
-
-averaged-283
+141 +141-71
277
-133
M12 - M2
-303
+147
+174
-76
-
-averaged-312
+156
+156
-78
344
-162
M15 - M2
-322
+135
+123
-64
-
-averaged-274
+137
+137
-68
208
- 98
propeller averaged-118
+ 59
+ 59
-29
-
-rudder I-178
+ 89
+ 89
-45
275-138
rudder II-150
+ 74
+ 74
-37
305-152
rudder III-165
+ 82
+ 82
-42
277
-133
rudder IV-194
+ 97
+ 97
-48
344
-162
rudder V-156
+ 78
+ 78
-39
208
- 98
This is done for all types of rudders; it was shown that the
wake-factor doesnot vary with yawing or swaying amplitude and the wakewake-factor was nearly 0.52 for the original version and nearly 0.55 for the modified version as can be seen in fig.
7.
These values are rather high and some scale effect in the wake must be assumed, thus influencing the incoming velocities and making the pertinent derivatives larger.6.
Conclusions.The following general conclusions can be derived.
Course stability characteristics are improved by adding lateral force devices in the afterbody.
The modified afterbody did not give better course stability characteris-tics. In fact a skeg was needed to have nearly the same performance as
the original version.
The effect on course stability of a rudder in the wake of a ship is of the same order of magnitude as the effect of a rotating propeller without rudder.
The effect of propeller slipstream is important for the improvement of course stability.
None of the investigated configurations proved to be course stable. Increasing rudder area doesnot always imply a better course stability. For this type of tankers one could hardly expect that course stability can be achieved by relatively small changes in the design, apart from the question to what extent course instability can be accepted.
The total improvement in course stability of rudder-propeller configu-rations is for the greater part due to the rudder action. However the propeller seems to have a substantial effect in both, the generation of stabilizing transverse forces and reducing the stabilizing effect of the rudder by producing a straightening slipstream, and secondly making the incoming water velocities in the rudder higher.
7.
References:[1] English, J.W. and Steele, B.N.:
"Ship-borne Manoeuvring Devices"
Paper read at the 1st International Tug Conference,
1969.
[21
Motora, S. and Koyama, T.:"Some Aspects of Automatic Steering of Ships"
Japan Shipbuilding and Manen Engineering, July
1968.
13]
Whicker, L.F. and Fehlner, L.F.:"Free Stream Characteristics of a Family of Low Aspect Ratio, All Movable Control Surfaces for Application to Ship Design"
DTMB Report
933,
May1958.
[41
Glansdorp, C.C. and Pijfers, J.G.L.:"Systematic Horizontal Oscillation Tests with Models of Series 60, Blockcoefficient
.70
with Varying Length-Breadth Ratio's"Report
322
of the Shipbuilding Laboratory of Delft. University of Technology, August1971.
Tsakonas, S.:
"Effect of Appendage and Hull Form on Hydrodynamic Coefficients of Surface Ships"
Davidson Laboratory, Report No.
740,
July1959.
Jacobs, W.R.:
"Estimation
of
Stability Derivatives and Indices ofVarious Shipforms and Comparison with Experimental Results."
Journal of Ship Research, September
1966.
f7]
Mandel, P.:"Some Hydrodynamic Aspects of Appendage Design" Transactions SNAME
1953
[8]
Glansdorp, C.C. and Buitenhek, M.:"Manoeuvring Trials with a
200.000
Tons Tanker"Report
248
of the Shipbuilding Laboratory of the Delft University of Technology, August1969.
Fig. 1
modified version
2A links
APP
///zez,/, 7,40
APP
100
2E rechts
50
APP
3
10
KQ
KT
1
0.6
0.4
0.2
Wageningen
B-series
-Experiments
---
KQ)
KT
..
....,
..
...K
....,%0I
<1...Kr".,
'
0.2
04
0.6
j
08
1.0
RUDDER I
nks
x
O
A
without propeller
200
n
=
67 R.P.M.
viS .105
n
=
85
s.n
=
102
..
1100
.
_3Ö0
.45
'''
1°
,...
s
300
if
O
O.
100
A200
1
N
-105
50
100
1°
s
Nzt
4 rechts
o
5 links
RUDDER TE
incl. skeg
200
105
100
Nsb.105
50
1
6
Yvi '1
-5
5-10
-150
oR .P. M.
40
80
120
)Rudder 1E
ORudder ISE
7