Você está na página 1de 8

DETERMINATION OF DESIGN

SLAMMING LOADS ON BOW DOORS FOR RO-RO SHIPS

Hui Wang, China Ship Scientific Research Center, Wuxi, Jiangsu, China
Xuekang Gu, China Ship Scientific Research Center, Wuxi, Jiangsu, China
Jinwei Shen, China Ship Scientific Research Center, Wuxi, Jiangsu, China
SUMMARY
With the combination of the nonlinear strip theory and Stavovy-Chuangs slamming theory, an approach of predicting
bow flare slamming pressures is presented and its practicality is demonstrated with the comparisons between test and
calculation results of slamming pressures on bow door of a Ro-Ro ship. With respect to the characteristics of slamming
pressures, a method of determining design-slamming pressures on bow door is presented as a direct calculation
approach for bow doors of Ro-Ro ships. The resultant pressures are compared with design values in rules and
regulations.

AUTHORS BIOGRAPHY

900 lives on board were lost (Sames, etc., 2001). This


tragedy induced an intensive re-examination of design
loads, design procedures and acceptance criteria of bow
doors as well as their support ing structures in
classification rules and regulations.
Sames
(2001)
performed
theoretical
and
experimental researches on impact loads of a Ro-Ro ferry.
To verify the reliability of theoretical calculations, the
ship speed of 26 knots was assumed in survival wave
conditions and model tests were performed. Pressure
sensors were installed on the bow door of the model and
the maximum measured slamming pressure was up to
400kPa in full-scale. The slamming loads were also
theoretically predicted by two approaches, one is based
on linear motion prediction and the pressure estimation
of Ochi-Motter (1973); the other is based on the method
of Zhao (1993).
Kvalsvold etc. (1996) developed a numerical method
to calculate slamming pressures. The heave and pitch
motions are calculated with nonlinear theory; the bow is
divided into strips normal to FP and slamming pressures
are calculated with 2 dimensional water-impacting theory
and integrated to form slamming force. The numerical
results were compared with the results of two model tests
and the agreements were quite well.
In this paper, to theoretically predict the slamming
pressures on bow structures, a nonlinear strip theory is
first employed to predict relative motions between ship
and waves in regular and irregular head waves, then
Stavovy-Chuang(1976)s method is used to calculate
slamming pressures. Calculated results are compared
with the test results on the bow door of a Ro -Ro model
and the agreements are generally good. Finally the design
slamming pressures were recommended for bow door
structures of the ship and compared with corresponding
values in classification rules.

Hui Wang is Ph. D. student at China Ship Scientific


Research Center (CSSRC). He received a Bachelor of
Science from Harbin Engineering University and
subsequently a Master of Engineering from CSSRC. He
has been involved in several experimental and numerical
research programs at CSSRC and now is mainly engaged
in projects of direct design calculations of ship structures.
Xuekang Gu received a Bachelor of Science from
Peking University in 1984. Subsequently he received a
Master of Engineering in 1990 and a Doctor of
Philosophy in 1999 from China Ship Scientific Research
Center. His main research interests are in wave-induced
motions of and loads on ships and marine structures,
direct calculations of design wave loads on ship
structures with theoretical and experimental tools.
Jinwei Shen graduated in 1959 and then finished his
postgraduate from Shanghai Jiaotong University. In last
forty years he has been heavily involved in many
research projects on wave-induced loads.

1.

INTRODUCTION

Sailing in severe sea conditions, ship body will


encounter wave impacts frequently because of the large
relative motions between ship and waves. These impacts
could be divided into several types as bottom impact,
flare slamming, green water and slap etc. Generally they
may lead to hull structural fatigue or local structural
deformation, and in some worst cases, may lead to
structural collapse.
Ro-Ro ferry ESTONIA sank in Baltic Sea in
September 1994. Its visor bow door was knocked out by
stormy waves and the cabins were flooded. More than

Session B1

2.

P = kV

BRIEF
DESCRIPTION
OF
THE
THEORETICAL APPROACH AND THE TESTS

where ship form, wave conditions, ship speed and


motions are inputted as given conditions.
Combin ing the two theories above, slamming
pressures on bow door could be predicted.
The principal dimensions of tested Ro-Ro ship
model
is
L B T Cb=
3.67 0.546 0.135 0.123t 0.627. Large wave heights
test were carried out in towing tank of CSSRC, and the
test conditions are showed in Tab.1. The wave spectrum
adopted was two-parameter wave spectrum by ISSC as

Generally considering, the relationship of ship


motions, especially heaving and pitching, and wave
height is assumed to be linear as ship sails in little or
middle amplitude waves. But the relationship will be
nonlinear when bow bottom emerging and green water
occurs. The nonlinear strip theory adopted in this paper is
a two-dimensional theory based on linear strip theory
(Gu X K ,1999). It considers the changing of
hydrodynamic coefficients with instantaneous draught
and the nonlinear hydrodynamic forces such as bow
impact, flare slamming, green water and dynamic
buoyancy led by real ship body under the surface of
incident waves. The ship motion equations are solved in
time domain and each motion responds in regular or
irregular waves are given, such as relative positions of
ship and wave for each point on bow surface, relative
velocities, wave profiles in corresponding positions etc.
Stavovy-Chuang (1976) theory could predict
slamming pressures at each point on hull surface. Firstly,
the normal velocity V of a hull point relative to wave
surface and an effective impact angle in the normal plane
of wave surface and ship body surface should be
determined. Then according to pressure peak coefficient
k obtained from experiments, maximum impact
pressure P could be estimated in a general form as

Wave
condition
Irregular sea

Regular sea

3.

S ( ) =

T2 5

Tab.1 Tested conditions in towing tank


Ratio of wave height
Ship speed
Course angle
to ship length
Fn
0.188~0.250
180
Hs = 1 ~ 1
(heading sea)
L
44
15

1
1
L = 44 ~ 15

496
exp 4 4
T2

where Hs is significant wave height, cross-zero wave


period is selected from statistic data of South China sea
area (Fang Z S etc., 1995).
Pressure-measuring points were arranged on port
bow flare: 9 pressure sensors on bow door, 2 accelerators
in vertical and horizontal directions at bow station 0, and
there were also wave height meter and motions recorder
that measured pitching and heaving. The positions of 9
pressure sensors are showed in Fig.1, which happened to
have the same number and arrangement as the Ro-Ro test
by Sames (2001). Three samples of typical pressures in
regular and irregular waves are shown in Fig.2 and Fig.3.

0.188~0.250

COMPARISONS BETWEEN THEORETICAL


AND EXPERIMENTAL RESULTS

180
(heading sea)

Recording time
(relative to full-scale)
No less than 1 hour at
full-scale
No less than 10
encounter periods
stable recording

some irregular phenomena may exist in the pressure


time-domain curves in regular wave tests, shown in Fig.2.
It could be estimated that bow wave making should be an
influential factor, but it is not yet considered in
theoretical predictions at present.
Generally speaking, the comparisons with test
values not only verified the theoretical method but also
shown that it could be used to estimate design slamming
pressures on bow flare of vessels.

Theoretical calculations were performed with all


tested conditions of regular waves, and these results are
compared with those of model tank tests, represented by
P2, P5 and P8, shown in Fig.4~6. Similarly, results with
irregular waves were compared by means of statistical
behaviors such as average values or significant values
shown in Fig.7~8.
From these comparisons, it could be seen that the
calculation results are larger than those of test in small
regular wave heights, but it is just the opposite in large
wave height. In irregular waves, the statistical behavior
values of pressures in calculation fits very well with test,
especially for those large statistical values.
Slamming is a very complex phenomenon of partly
hydrodynamic impact that occurs when vessels sail under
some certain wave conditions and speeds. There are
many factors to cause it, in which local relative velocity
and incident impact angle are active especially. Usually

Session B1

124 H s

4.

DETERMINATION OF DESIGN SLAMMING


LOADS ON BOW DOOR

4.1 ESTIMATION OF DESIGN SLAMMING LOADS


ON BOW BY RULES
In view of the accident of Ro-Ro ship Estonia and
the urgent requirement of operation safety for Ro-Ro ship,
IACS (International Association of Classification
Societies) published URS-8 in 1995, and then URS-9,

10

URS-15 and URS-16, which are revisions of URS-8, in


1996 and 1997. Corresponding stipulations are also set
one after another by many societies about minimum
design slamming values on bow door. CCS (China
Classification Society), based on URS-8, publishes
calculation formula about minimum values of the design
loads on bow door structures and fixed devices of Ro-Ro
ships in Chapter 9 of its classification rules. Besides,
BV(1996), DNV(1995), and GL(1998) etc. also give
similar calculation formula in their rules.
Though these design loads determined by various
rules are a little different in values (value by DNV is
more larger), they are basically similar to the requirement
of URS-8, which is minimum design pressure approved
internationally. But design load is only a premise in
structure design, for designers could present structure
components with different forms and sizes under the
same design loads, especially in design of the fixed
components on bow door. It must lead to the facts that
some may be destroyed even have accidents but some
serve safely during vessels operating. Obviously, these
practical problems are related to vessel bow structures
including fixed components on bow door. They are also
involved design thoughts especially the choices and
arrangements of fixed components. All these are beyond
the area in this paper and should not be discussed in
details.

point, the pressure value on bottom is about 1.15 times


than P5; on upper part 0.85 times than P5; on fore part
1.06 times than P5; on back part 0.79 times than P5. So,
assuming taking the pressure value and persisting period
of center point to design bow door structures, the design
loads should not be simply taken as the measured values
and its wave profiles on P5. The pressures of all
measuring points on bow door should be averaged in
values and in time domain. For example, the time
recording
procedures of P2, P5 and P8 in tested
condition
H s L = 1 19.8 , T2 = 8.654s , Fn = 0.219 are
shown in Fig.10, including the average value recording
Pave of P2, P5 and P8 in time domain. It could be found
that the time width between slamming pressure and its
followed hydrodynamic pulse is about 0.5 second for
ship model; the pressure peaks at every point dont
appear in the same instant but in turn from bottom to
upper part with water entry and slamming procedures of
bow surface. That is, there are differences not only in
values but time phases of the pressure peaks. The
distribution of slamming pressures in vertical direction
and average value of the three measuring points in three
instants is shown in Fig.9. By analyzing the recorded
slamming pressures in irregular waves, the comparisons
of each statistical behavior value are shown in Tab.2. The
statistical values of Pave are 0.8 times than that of P2, P5
and P8, 0.9 times as the measured value of P5. So if the
design loads, based on center point, take the pressure
values averaged in space and time domain, then either
according to model test recordings or theoretical
approach, strictly speaking, certain points on bow door
should be selected (for instance 9 points) and pressures
should be averaged in values and time domain. And then
according to the statistical behaviors of the average value,

4.2 DETERMINATION OF DESIGN SLAMMING


LOADS
ACCORDING
TO
DESIGN
CONDITIONS
4.2.1

ANALYSIS OF SLAMMING PRESSURES


ACTING ON BOW DOOR

Before discussing the determination of design


slamming loads, it s necessary to analyze the
characteristics of slamming pressures acting on bow door.
Its well known that the pressure peak value, emerging
time and persisting period are different for various points
on bow door. From model tests the general regular
pattern could be found that the slamming pressures acting
on bottom or fore positions on bow flare are larger than
those on upper or back parts. For example, if P5 which is
near the center of bow door is taken as representative

100 in general, comparisons and analyses with

minimum design values determined by rules should be


carried out. In the light of the analyses of model test and
theoretical calculation about slamming loads acting on
bow structures by Pegg(1988), it is suggested to be most
suitable to perform dynamic FEM analysis after the
pressures characteristics of bow door center point (value
and time procedure) are obtained.

Tab.2 Comparison between pressure peak values and average pressure in irregular waves (Pa)
(Heading sea, H s L = 1 19.8 , T2 = 8.654s , Fn = 0.219 )
Statistical values

P2

P5

P8

Pave

3Pave /( P 2 + P5 + P8)

Pave/ P5

Average values

27

19

16

15

0.725

0.789

1 highest average values


3
1 highest average values
10
1
100 highest average values

46

34

32

30

0.804

0.882

67

52

52

47

0.825

0.904

110

85

81

74

0.804

0.871

Session B1

11

4.2.2

DETERMINATION OF DESIGN SLAMMING


LOADS ACCORDING TO TWO DESIGN
CONDITIONS

be achieved:
1. The approach of combining the nonlinear strip
theory and Stavovy-Chuangs method can be
reasonably used to calculate slamming pressures on
bow doors.
2. Because of the local and transient characteristics of
slamming pressures, the averaging of pressures at
different locations should be made in time domain to
obtain same load effects in structures. The average
of one-hundredth highest pressures might be
recommended as design values.
3. The two design conditions can rationally reflect the
relations between design slamming pressures and
design ship speeds or wave heights.

The merchant ship structural rules by Russia (1991)


firstly define two loading conditions, cruising condition
and extreme survival condition, to determine design
loads. Later, Provisional Rules for Classification of
Naval Ships by LR (1999) followed this design
conception.
In this paper, according to the factual wave height
ship speed curve in vessel serving shown in Fig. 11,
based on two design conditions afforded by Kvalsvold
(1996), the design slamming loads could be determined.
Two design conditions shown as follow:
(1). Cruising condition-at 16 knots ship speed, significant
wave height to be 4m, cross-zero period to be 7.12s
(taken as serving condition)
(2). Extreme serving condition-at minimum maneuvering
speed (assumed to be 5 knots), in extreme regular wave
height under certain wave length, for example, wave
height is taken 9 meters as wave length is equal to ship
length, and wave height is taken 10 meters as wave
length is 1.2 times as ship length; or at ship speed in a
range of 6~10 knots, in 6 meters significant wave height
(taken as extreme survival condition)
All results of theoretical calculation and test show
that the design loads in extreme survival condition may
be smaller than in cruising condition, which is the same
conclusion as in the paper of Kvalsvold (1996). Here
only analyses of calculation results in irregular waves are
provided for the two designed conditions.
After theoretical calculation for 9 pressure points
and their averaging procedure, the relationship curve of
the

REFERENCES:
Fang, Z.S., Jin, C.Y and Miao, Q.M., 1995, Wave Data of
the Northwest Pacific Ocean, Press of China
Defense Industry.
Gu, X.K., 1999, Direct Design Calculation of Ship
Structures, Ph. D. Thesis, China Ship Scientific
Research Center.
Gu, X.K., Hu, J.J. and Moan, T., 2001, Design Slamming
Pressures of A High-speed Hydrofoil-assisted
Catamaran, Proceedings of the 6th International
Conference on Fast Sea Transportation (FAST01),
Southampton, UK.
Kvalsvold, J.,Svensen, T. and Hovem, L., 1996, Bow
Impact Loads on Ro-Ro Vessels , Royal Institution
of Naval Architects.
LR, 1999, Provisional Rules for the Classification of
Naval Ships.
Ochi, M.K. and Motter, L.E., 1973, Prediction of
Slamming Characteristics and Hull Responses for
Ship Design, Transaction of Society of Naval
Architects and Marine Engineers.
Pegg, N.G., Vernon, T.A., Wegner, L., and Nethercote,
W.C.E. , 1988 , Finite Element Prediction of
Measured Bow Flare Plate Stresses under Dynamic
Wave Loading , Royal Institution of Naval
Architects.
Sames, P.C., Kapsenberg, G.K. and Carrignan, P., 2001,
Prediction of Bow Door Loads in Extreme wave
Conditions, International Conference of Design and
Operation for Abnormal Conditions II, Royal
Institution of Naval Architects, London, UK.
Stavovy, A.B., and Chuang, S.L., 1976, Analytical
Determination of Slamming Pressures for High
Speed Vechicles in Waves, Journal of Ship Research,
Vol.20, No.4.
Zhao, R. and Faltinsen, O., 1993, Water Entry of
Two-Dimensional Bodies, Journal of Fluid
Mechanics, Vol.246.

100 highest average value of center point of bow

door to ship speed could be obtained, shown in Fig.12, in


which model test results are also given. From this figure,
it could be clearly seen that to sail safely in upper limit of
rough sea wave scale ( H s = 4 m ) the design slamming
pressure on bow door should be 70 Kpa. And if this
design value is adopted the ship could also sail in very
rough sea ( H s = 6 m ), but its speed could only be less
than 10 knots. In addition, design slamming loads
determined by four primary societies rules are calculated
and compared with the results by method in this paper,
shown in Fig.13. The latter one respects the nonlinear
changing of slamming loads with speed and wave height
more actually than rules.

5.

CONCULUDING REMARKS

Based on the comparison of theoretical and


experimental results, recommended and rule -based
design slamming pressures, following conclusions might

Session B1

12

P7

P8

P9

P4

P5

P6

P1

P2

P3

Calm waterline

-1

1?

Fig.1 A Ro-Ro ship bow door and pressure measuring points


4

10 0
P2 (kPa)

Wave (mm)

20 0

-10 0
-20 0

50

51

52

53

54

-2

55

50

51

52

-2

50

51

52

53

54

55

54

54

55

-2

55

50

51

52

T im e (s )

53

T im e (s )

Fig.2 Recording samples of pressures in regular waves (Heading sea,


20 0

Fn = 0.219 , L = 1.2 , H L = 1 18.4 )

5 .0

10 0

P2 (k Pa)

Wave (mm)

53

T im e (s )

P8 (kPa)

P5 (kPa)

T im e (s )

2 .5
0

-10 0

- 2 .5
-20 0

48

50

52

54

56

48

58

50

5.0

2.5

2.5

P8 (k Pa)

P5 (k Pa)

5.0

0
-2. 5

54

56

58

56

58

0
-2. 5

48

50

52

54

56

58

48

T im e ( s )

50

52

54

T im e ( s )

Fig.3 Recording samples of pressures in irregular waves (Heading sea,

Session B1

52

T im e (s )

T im e (s )

13

Fn = 0.219 , H s L = 1 18.4 )

1 60

te s t
c a lc

P2 (kPa)

1 20

80

40

10

15

20

S e r ia l N u m b e r

Fig.4 Comparisons between calculation and test results of P2 in regular waves

90

te s t
c a lc

P5 (kPa)

60

30

10

15

20

S e r ia l N u m b e r

Fig.5 Comparisons between calculation and test results of P5 in regular waves

50

te s t
c a lc

P8 (kPa)

40

30

20

10

10

15

20

S e r ia l N u m b e r

Fig.6 Comparisons between calculation and test results of P8 in regular waves

Session B1

14

140

Average

( Cal.)
( Test)

100

Ave.1/3 highest values(Cal.)

kPa

120

( Test)

80

Ave.1/10 highest values(Cal.)

60

( Test)

40

Ave.1/100 highest values(Cal.)

20

( Test)
Max. values (Cal.)

0
P2

P5

( Test)

P8

Fig.7 Comparisons between calculation and test results in irregular waves


(Heading sea,

H s L = 1 19.8 , T2 = 8.654s , Fn = 0.219 )

140

Average

(Cal.)
(Test)
Ave.1/3 highest values(Cal. )
(Test)
Ave.1/10 highest values(Cal.)
(Test)
Ave.1/100 highest values(Cal.)
(Test)
Max. values
(Cal.)
(Test)

120

kPa

100
80
60
40
20
0
P2

P5

P8

Fig.8 Comparisons between calculation and test results in irregular waves


(Heading sea,

H s L = 1 17.4 , T2 = 9.459s , Fn = 0.188 )

P 2

50

t3 = 4 .0 8 1 s

P (kPa)

40

30

20

P a ve
10

P 5
0

P 8

10

D is ta n c e f r o m b a s e lin e ( m )

P 2

50

t3 = 4 .1 4 5 s
40

P (kPa)

P 5

P a ve

30

20

P 8
10

50

10

D is ta n c e f r o m b a s e lin e ( m )

P 2
t3 = 4 .1 8 1 s

P (kPa)

40

P 5

P a ve

30

P 8
20

10

D is ta n c e f r o m b a s e lin e ( m )

10

Fig.9 Instantaneous values of pressures at representative times and their average

Session B1

15

P 2

P 5

P 8

P av e

4 .5

t (s)

5 .5

Fig.10 Instantaneous values of pressures and their average


18

V (kn)

15
12
9
6
3

Hs (m)

12

Fig.11 Possible highest speeds

90

80

P5 (kPa)

70

60

50

H s= 4 m T e s t
H s = 4 m C a lc .

40

30

H s = 6 m C a lc .
6

1 0

12

14

16

18

V (kn)

Fig.12 Average

100 highest values of pressures on bow door center

90

80

70

P (kPa)

C a lc .
DNV
60

CCS

50

BV
GL

40

30

10

12

14

16

18

V (k n)

Fig.13 Comparisons between calculation results with methods presented by several


societies and this paper

Session B1

16

Você também pode gostar