Anda di halaman 1dari 22

Ocean Engineering 28 (2001) 417–438

Technical Note
An experimental method for determining the
frequency-dependent added mass and added
mass moment of inertia for a floating body in
heave and pitch motions
*
Jong-Shyong Wu, Mang Hsieh
Institute of Naval Architecture and Marine Engineering, National Cheng-Kung University, Tainan,
701, Taiwan, ROC

Received 19 August 1999; accepted 24 November 1999

Abstract

Most of the existing relevant materials have been obtained from experiments, in which
evaluating the added mass at the resonant frequency corresponding to the peak of a frequency-
response curve obtained from the “forced” vibration analysis is the most popular technique.
In this paper, a simple experimental method was presented where the “free” vibration responses
instead of the “forced” ones were used to determine the values of mah and Iap. The main part
of the experimental system is composed of a floating body (model) and a spring–shaft shaker.
The “free” vibration of this main part was induced by imposing on it an initial displacement
(and/or an initial velocity), and from the time histories of displacements information such as
the “damped” natural frequencies, damping ratios, sectional added mass coefficients (CV and
CP) were obtained. Since the displacements of the spring–shaft shaker are “translational” and
those of the floating body due to pitch motions are “angular”, a technique for the transform-
ation between the associated parameters of the two components of the main part was presented.
 2000 Elsevier Science Ltd. All rights reserved.

Keywords: Added mass; Added mass moment of inertia; Floating body (model); Spring–shaft shaker;
Time history; Frequency-response curve

* Corresponding author. Fax: +886-6-2747-019.


E-mail address: jasck@www.nm.ncku.edu.tw (M. Hsieh).

0029-8018/01/$ - see front matter  2000 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 2 9 - 8 0 1 8 ( 0 0 ) 0 0 0 0 8 - 1
418 J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438

1. Introduction

For a single-degree-of-freedom (SDOF) oscillating system composed of a spring–


shaft shaker and a floating body (or model), if w⬘a, ka and ma represent the
“undamped” natural frequency, the effective spring constant and the effective mass
of the oscillating system with the floating body “in air”, respectively, and w⬘W, kW
and mW represent the corresponding values with the floating body “contacting with
water”, then in theory the values of w⬘a and w⬘W may be obtained from the formulas:
冑 冑
w⬘a⫽ ka/ma and w⬘W⫽ kW/mW

A comparison between w⬘a and w⬘W shows that w⬘W⬍w⬘a. Since all conditions around
the oscillating system remain unchanged except for those surrounding the floating
body, that is, w⬘a is obtained for the case of the floating body “in air” and w⬘W is
obtained for the case of the floating body “contacting with water”, it is evident that
mW⬎ma will be the main reason that wW⬘⬍wa⬘ if the effect of water buoyancy on
the spring constant kW and all the effect of damping on w⬘W have been taken into
consideration. In other words, when a floating body (e.g. a ship hull) is oscillating
in water, part of the surrounding water will oscillate with it synchronously so that
the total mass of the body seems to increase significantly. The difference between
mW and ma denotes the quantity of increment and is usually called the “added” mass
in “heave” motion, mah, i.e. mah=mW⫺ma. From the above description it is easy to
conceive that the added mass of a floating body will have something to do with a
lot of factors, such as area coefficient s, beam/draft ratio b, boundary conditions of
the water tank (or sea), slenderness ratio (L/B), oscillation frequency w⬘W etc. Besides,
since w⬘a is obtained by neglecting the damping ratio xa the value of w⬘W should also
be determined under the same condition that xW=0.
The magnitude of added mass per unit length of a floating body is usually rep-
resented by the sectional added mass coefficient CV for the vertical (heave) motion,
CH, for the horizontal (sway) motion and CT for the rotational (torsion) motion (Todd,
1961). Since 1929 when Lewis presented a paper to state the inertia of water sur-
rounding a vibration ship, a lot of researchers have devoted themselves to the studies
of the relationships between the sectional added mass coefficients (CV, CH or CT)
and area coefficient s and beam/draft ratio b (Browne et al., 1930; Kumai, 1958;
Landweber and Mcagno, 1957; Murtha, 1954). As to the effect of shallow water or
a narrow water way on the added mass, the materials concerned are much fewer
(Koch, 1949; Marwood and Johnson, 1953; Prohaska, 1947). In the above-mentioned
literature, the values of the sectional added mass coefficients were obtained under
the assumption that the length of the floating body is infinite (i.e. L→⬁) so that the
end effect may be neglected. But according to Lewis, Taylor and Townsin the three-
dimensional effect must be considered in general (Todd, 1961; Taylor, 1930; Town-
sin, 1969).
Although it is evident that the influence of the oscillating frequency w⬘W on the
added mass of a floating body should be significant, the information in this area is
still quite limited. In the existing literature, the values of CV were obtained by using
J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438 419

the resonant technique (Gerritsma, 1957; Moullin and Browne, 1928; Tanaka and
Kicagawa, 1962), where the floating body was suspended by a spring and then an
exciting force with adjustable exciting frequency we was imposed. The exciting fre-
quency we, corresponding to the peak of the frequency-response curve, was con-
sidered as w⬘W, i.e. wW⬘
⬇we. Finally, the values of sectional added mass m⬘ (and
coefficient CV) were calculated based on the obtained natural frequency w⬘W and the
relating formulas. In experiments, the resonant conditions are not easily obtained
with good accuracy, hence the implementation of the resonant technique is usually
very difficult. For this reason, a simple experimental method is presented in this
paper, where all the oscillating frequencies, upon which the added mass of the float-
ing body is evaluated, are the natural frequencies of free vibrations rather than the
exciting frequencies of forced vibrations for the oscillating system.
It is also worthy of note that the natural frequencies obtained from experiments
for the case of the floating body contacting with water are the “damped” ones wW,
while those required for calculating the values of added mass (or added mass moment
of inertia) are the “undamped” ones w⬘W. Hence it is necessary to evaluate w⬘W from
wW, by using the formula w⬘W=wW/√1−x2W, where the damping ratios xW are easily
determined from the “free” vibration response curves presented in this paper, but
this is not true for the (forced) frequency-response curves adopted by most of the
existing literature.
The experimental equipment of this study consists of a water tank (5 m long×2
m wide×1.5 m deep), a spatial framework, a planar framework, a spring–shaft shaker,
a floating body (or model) and some measuring instruments. Among all the items,
the spring–shaft shaker is most worthy of mention, since its total mass and total
effective stiffness are adjustable so that the floating body connecting with it can
perform “free” oscillation with any desired frequencies and based on which the corre-
sponding frequency-dependent added mass and added mass moment of inertia were
determined. Four models (floating bodies) with lengths L=0.6 and 1.0 m were studied
in this paper. The cross sections of two of them (Models 1(a) and 1(b)) are rectangu-
lar (with area coefficients s=1.0) and those of the other two are Lewis forms (with
area coefficients s=0.785 at draft d=0.08 m). In the existing literature, no information
about the sectional added mass coefficient obtained from pitch motions, CP, was
found, since the values of the total “added” mass moment of inertia in “pitch” motion
(Iap) were determined by using the coefficient obtained from heave motions, CV as
one may see from Eq. (4) of this paper. However, for a floating body with compli-
cated configuration, the distribution of the sectional added mass m⬘ (or coefficient
CV) for heave motion is usually not obtainable. In such a case the experimental
method presented in this paper will be available for evaluating the “total” added
mass moment of inertia of the floating body, Iap.

2. Formulation of the problem


2.1. Equations of motions
For a floating body with length L, the equations for the free heave and pitch
motions are given by Korvin-Kroukovsky and Jacobs (1957).
420 J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438

a1z̈⫹b1ż⫹c1z⫽0 (heave) (1)


A1q̈⫹B1q̇⫹C1q⫽0 (pitch) (2)
where


L

a1⫽ m⬘ dxb⫹rwⵜ, b1⫽ N⬘ dxb, c1⫽ rwgBdxb
L

L
(3)


L

A1⫽ m⬘x2b dxb⫹k 2yyrwⵜ, B1⫽ N⬘x2b dxb, C1⫽ rwgBx2b dxb
L
冕L
(4)

In the last expressions, z, ż and z̈ are the vertical displacement, velocity and acceler-
ation of the center of gravity (G) of the floating body due to heave motion; q, q̇
and q̈ are the angular displacement, velocity and acceleration of the floating body
due to pitch motion about the yb-axis, respectively (see Fig. 1); m⬘ is the sectional
added mass; rw is the mass density of water; g is the gravitational acceleration; N⬘
is the sectional damping coefficient due to surface wave generation; while B, ⵜ and
kyy are the width, volume displacement and longitudinal radius of gyration of the
floating body, respectively. The coordinate systems for the heave and pitch motions
of the floating body are shown in Fig. 1, in which xyz is a fixed coordinate system
with origin on the still water surface, xyz is a (moving) coordinate system with origin
at the center of gravity of the floating body (G) and always parallel to the xyz axis,
and xbybzb is a (moving) coordinate system fixed on the floating body with origin
at G.
From Eq. (4) one sees that the parameter A1 is the mass moment of inertia for
the pitch motion of the floating body. It is composed of two parts, one of them is
due to the sectional added mass of vertical motion (m⬘) and the other is due to the
material mass of the floating body itself (rwⵜ). Since the total added mass moment
of inertia of the floating body (Iap) in pitch motion is evaluated by using the sectional
added mass due to heave motion, m⬘, the sectional added mass coefficient obtained
from pitch motions, CP, is seldom mentioned in the existing literature.
If the cross sections are uniform along the length L of the floating body, then Eqs.
(3) and (4) reduce to

Fig. 1. Definitions of coordinate systems for the heave and pitch motions of a floating body.
J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438 421

a1⫽m⬘L⫹rwⵜ⫽(m⬘⫹rwB ds)L, b1⫽N⬘L, C1⫽rwgBL (5)


1 1 1 1
A1⫽ m⬘L3⫹k 2yyrwⵜ⫽ (m⬘⫹rwB ds)L3, B1⫽ N⬘L3, C1⫽ rwgBL3 (6)
12 12 12 12
where d and s are the draft and area coefficient of the floating body, respectively.

2.2. Natural frequencies of heave and pitch motions in still water surface

Either Eq. (1) or Eq. (2) is equivalent to the equation of free vibration for a SDOF
system, where a1 (or A1), b1 (or B1) and c1 (or C1) are equivalent to the mass,
damping coefficient and spring constant, respectively. Hence the “undamped” natural
frequencies are as given by Wu and Sheu (1996).

冪a
c1
w⬘h⫽ (heave) (7)
1

冪A
C1
w⬘p⫽ (pitch) (8)
1

2.3. Natural frequencies of heave and pitch motions for the floating body
connecting with the spring–shaft shaker

In order to make the floating body (model) perform “free” oscillation at various
frequencies, the floating body is connected with a spring–shaft shaker. The natural
frequencies of heave and pitch motions of the floating body together with the spring–
shaft shaker in such cases are derived as follows.
The sketch of Fig. 2 shows the arrangement for the floating body connecting with
the spring–shaft shaker to perform “heave” motion. If k̄h, m̄h and w̄⬘h represent the
effective spring constant, the effective mass and the “undamped” natural frequency
for the entire system, then one has

冪m̄
k̄h
w̄⬘h⫽ (9)
h

From Eqs. (5) and (7) one sees that the values of w⬘h are functions of the added
mass (m⬘L), the material mass (rw BdLs) and the buoyancy (rwgBL), while for the
values of w̄⬘h defined by Eq. (9) one is required to consider the additional contribution
of the spring–shaft shaker, hence
k̄h⫽c1⫹ks⫽rwgBL⫹ks (10)
m̄h⫽a1⫹m3⫽m⬘L⫹rwBdLs⫹ms (11)
where ks and ms are the spring constant and the mass of the spring–shaft shaker
422 J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438

Fig. 2. The floating body is connected with the spring-shaft shaker to perform heave motion.

alone, respectively. In practice, the value of ks is easily obtained from a simple


tension test and that of ms may be determined by subtracting the mass of the floating
body (=rwBdLs) from the total mass of the entire oscillating system weighing in air.
The same spring–shaft shaker may also be used to connect with the floating body
to make the latter perform “free” pitch motion with various frequencies as shown
in Fig. 3, but the connecting point is at the right end of the floating body (rather
than at the center of gravity, G) and the rigid connector (for heave motion) is also
replaced by a “hinge”. Besides, the two sides of the floating body must be supported
by pivots with rotating axis coinciding with the yb-axis and passing through the
center of gravity G (see Fig. 3(a) and (b)).
Similar to Eq. (9), the “undamped” natural frequency for the entire system takes
the form

冪m̄
k̄p
w̄⬘p⫽ (12)
p

where k̄p and m̄p are the effective spring constant and effective mass for the “pitch”
system as shown in Fig. 3 and given by
1
k̄p⫽kq⫹ks⫽ rwgBL⫹ks (13)
3
1
m̄p⫽mq1⫹mq2⫹ms⫽mq1⫹ rwB dLs⫹ms (14)
3
where ks and ms are the spring constant and the mass of the spring–shaft shaker
J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438 423

Fig. 3. The floating body is connected with the spring-shaft shaker to perform pitch motion: (a) front
view; (b) right side view.

alone, respectively; kq is the equivalent “translational” spring constant due to buoy-


ancy, while mq1 and mq2 are the equivalent “translational” added mass and material
mass of the floating body, respectively. The values of kq, mq1 and mq2 are evaluated
in the next subsection.

2.4. Transformation between parameters for angular motion and those for
translational motion

The pitch motion of the floating body is an “angular” oscillation about the yb-axis
(see Fig. 3(a)). Since the foregoing angular motion is induced by the “translational”
oscillation of the spring–shaft shaker, all the angular quantities of the floating body
must be transformed into the equivalent “translational” ones located at the hinge (at
the right end of the floating body). Then Eq. (12) may be used to calculate (or to
adjust) the “pitch” frequency of the floating body. The values of kq, mq1 and mq2
appearing in Eqs. (13) and (14) are the “translational” quantities and are derived
as follows.

2.4.1. Determination of kq
Due to buoyancy, the righting moment associated with the pitch angle q as shown
in Fig. 3(a) is given by


L/2
1
Myb⫽ rwgBx2bq dxb⫽ r gBL3q (15)
12 w
⫺L/2
424 J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438

If all the buoyancy is concentrated at the hinge and replaced by the restoring force
of a spring with spring constant kq, then the righting moment associated with the
pitch angle q is given by

Myb⫽kq 冉冊
L L 1 2
q ⫽ kLq
2 2 4 q
(16)

From the last two equations one obtains


1
kq⫽ rwgBL (17)
3

2.4.2. Determination of mq1


Let m⬘ap be the sectional “added” mass associated with the “pitch” motion, then
the added mass moment of inertia of the floating body about the yb-axis is given by


L/2
1 ⬘ 3
Iap⫽ m⬘apx2b dxb⫽ m L (18)
12 ap
⫺L/2

If all the added mass of the floating body is concentrated at the hinge and denoted
by mq1, then the added mass moment of inertia of mq1 about the yb-axis must be
equal to that defined by Eq. (18), i.e.

Iap⫽mq1 冉冊
L 2 1 ⬘ 3
2
⫽ mapL
12
(19)

Hence
1
mq1⫽ m⬘apL (20)
3

From Eqs. (12)–(14) one sees that if the value of w̄⬘p is obtained from experiment,
then the only unknown quantity mq1 appearing in Eq. (14) may be determined and
by means of Eq. (20) the frequency-dependent sectional added mass m⬘ap is also
defined.

2.4.3. Determination of mq2


Since the total mass of the floating body is equal to that of the displacement water,
the mass moment of inertia for the material of the floating body about the yb-axis is


L/2

I⬘ap⫽ rwB dsx2b dxb⫽rwB dLsk 2yy (21)


⫺L/2
J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438 425

Because a number of holes exist for the ballasting block weights, one must con-
sider the actual distribution of the holes and the weights in calculating the longitudi-
nal radius of gyration kyy. If all the material mass is concentrated at the hinge and
denoted by mq2, then by definition one has

I⬘ap⫽mq2 冉冊
L
2
2
(22)

To substitute the last equation into Eq. (21) gives


mq2⫽4rwB dsk 2yy/L (23)

2.5. Frequency-dependent sectional added mass coefficients, CV and CP

By using the equipment shown in Fig. 2 one may measure the natural frequency
(w̄⬘h) of heave motion of the floating body. Then from Eqs. (9)–(11) one may obtain
the associated sectional added mass m⬘. Similarly, by means of the arrangement
shown in Fig. 3 one may obtain the natural frequency (w̄⬘p) of pitch motion, and then
the values of mq and m⬘ap can be determined from Eqs. (12–14) and (20). Hence, the
sectional added mass coefficient obtained from heave-motion tests, CV, and that
obtained from pitch-motion tests, CP, are given by
m⬘

冉 冊
CV⫽ (24a)
1
r pb2
2 w
m⬘ap

冉 冊
CP⫽ (24b)
1
r pb2
2 w
where b is the half-beam of the floating body. It is evident that if the value of CP
is equal to that of CV approximately, then the conventional approach for calculating
the added mass moment of inertia of floating with Eq. (4) will be reasonable. Other-
wise, improvement on the existing approach should be required.

3. Experimental equipments

The experiments were conducted in a water tank 5 m long, 2 m wide and 1.5 m
deep. Over the tank a spatial framework was constructed as shown in Fig. 4. For
convenience of supporting the floating body, a planar framework was also con-
structed on the top of the tank. The oscillating frequency of the floating body is
controlled by an assemblage of a spring and a shaft, which is called the “spring–
426 J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438

Fig. 4. Sketch of the experimental system: (a) front view; (b) right-side view.

shaft shaker” in this paper. The latter is supported by the spatial framework with
two ball bearings so that it can freely oscillate in the vertical direction. The upper
end of the spring–shaft shaker is suspended under the spatial framework and the
lower end is connected with the testing model with set screws. To keep the shaft in
the vertical direction as accurately as possible, the pads of the spatial framework are
equipped with adjusting screws as one may see from Fig. 4.
The main purpose of the foregoing equipment is to make the floating body oscillate
with various frequencies. To this point, the total mass of the spring–shaft shaker is
adjusted by the total number of disk weights on a pan attached on the shaft as shown
in Figs. 2 and 3. Besides, the stiffness of the spring–shaft shaker is also changeable.
Four springs with effective spring constants ks=176.6, 329.7, 812.6, 1570.0×9.8 N/m
are available.
The measurements were made with the linear displacement transducer for the case
of displacement (d) greater than 0.003 m and with the eddy current detector for the
case of dⱕ0.003 m. The other instruments required for this experiment include a
FFT signal analyzer, two signal amplifiers, one multi-channel data recorder, etc.

4. Experimental steps

In general, one may obtain some results from an experiment, but the reliability
of the results has a close relationship with the experimental procedures. Hence the
main steps of the present experiment are described as follows.

(a) Disconnect the testing model (or the floating body) from the spring–shaft
J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438 427

shaker by loosing the set screws before performing a new experiment (cf Figs. 2
and 3). The free water surface in the tank is assumed to be at the desired level.
(b) Adjust the draft d of the model to the desired value by changing the total
number of ballasting block weights (in the holes of the model).
(c) Change the total mass ms of the spring–shaft shaker by adding (or removing)
disk weights on (or from) the pan attached on the shaft (see Figs. 2 and 3). The
spring constant ks of the spring–shaft shaker may also be changed by replacing
the original spring with a new one having an appropriate spring constant.
(d) Let the model and the spring–shaft shaker stay in their individual static equilib-
rium positions and then connect them by fastening the set screws.
(e) Give the entire oscillating system an initial displacement z0 (or q0) or an initial
velocity ż0 (or q̇0) to make the model (together with the spring–shaft shaker)
perform harmonic “free” oscillation and then record the time history of displace-
ments as shown in Fig. 5, where z0 and ż0 represent the initial displacement and
velocity for heave motion and q0 and q̇0 represent those for pitch motion, respect-
ively (see Figs. 1–3). Determine the damping ratio xW and the “damped” natural
frequency wW for the model “contacting with water” from the formulas (Clough
and Penzien, 1975)
ñn−ñn+m
xW ⫽ (25a)
2pmñn+m

wW⫽2p/TW (25b)
where TW is the period of the damped free oscillations obtained from the time
history as shown in Fig. 5(a), and ñn and ñn+m are the amplitudes of oscillation
displacements with m cycles apart. The “undamped” natural frequencies required
by Eqs. (9) and (12) are now given by
wW
w⬘W⫽
冑1−x 2
(26)
W

(f) Open the exit valve of the tank to let part of the water in the tank flow out
until the free water surface lowers to the level that free oscillation of the model
(together with the spring–shaft shaker) in air (without contacting with water) is
possible. Repeat step (e) to obtain the time history of displacements “in air” as
shown in Fig. 5(b). Determine the damping ratio xa and the “damped” natural
frequency wa from
nn−nn+m
xa ⫽ (27a)
2pmnn+m

wa⫽2p/Ta (27b)
428 J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438

Fig. 5. Time histories of vertical displacements of the oscillating system for the cases of (a) the model
“contacting with water”, and (b) the model completely “in air”.

where Ta, nn and nn+m have the same meanings as TW, ñn and ñn+m respectively,
the only difference is that the former (Ta, nn and nn+m) are obtained from the time
history (Fig. 5(b)) for the model oscillating “in air” and the latter (TW, ñn and
ñn+m) from the time history (Fig. 5(a)) for the model oscillating “contacting
with water”.

In theory, the value of the “undamped” natural frequency of the entire oscillating
system “in air” is given by
J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438 429

冪m
ka
w⬘a⫽ (28)
a

where ka and ma are the spring constant and the total mass of the entire oscillating
system “in air”. Both of them can be measured precisely, hence Eq. (27b) may be
used to check the measured values of ka and ma if the damping effect is negligible.
Since the relationship between wa and w⬘a is
wa
w⬘a⫽
冑1−x 2
(29)
a

the last formula may also be used to check the value of xa defined by Eq. (27a)
obtained from experiments.

5. Experimental results and discussions

Four models (or floating bodies) were studied here. Two of them (Models 1(a)
and 1(b)) as shown in Fig. 6(a) and (b) are parallelepiped with rectangular cross
sections, the lengths are 0.6 and 1.0 m, respectively. The other two (Models 2(a)
and 2(b)) as shown in Fig. 6(c) and (d) are uniform beams with Lewis sections, the

Fig. 6. The four models (or floating bodies) studied in this paper: (a) Model 1(a); (b) Model 1(b); (c)
Model 2(a); (d) Model 2(b).
430 J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438

lengths are also 0.6 and 1.0 m, respectively. The area coefficient for Models 1(a)
and 1(b) is s=1.0 and that for Models 2(a) and 2(b) is s=0.785 at draft d=0.08 m
as shown in Table 1.

5.1. Reliability of theory and experimental equipments

From the above-mentioned experimental steps one finds that the only difference
between step (e) and step (f) is that the model contacts with water for step (e) and
the model does not contact with water for step (f). In view of this fact, one may
infer the reliability of step (e) if the reliability of the results obtained from step (f)
is confirmed.
For example, in one of the present experiments, the spring constant of the spring–
shaft shaker was found to be ka=ks=15,386 N/m from the simple-tension test, and
the total mass of the spring–shaft shaker and the model (including the ballasting
weights) was measured to be ma=9.525 kg by using the precise scales, hence the
theoretical “undamped” natural frequency of the entire oscillating system in air is
given by w⬘a=√ka/ma=40.191 rad/s (see Eq. (28)). The experimental result for the
present case was found to be wa=2p/Ta=38.461 rad/s and xa=0.108, hence
w⬘a=wa/√1−x2a=38.642 rad/s. Since the theoretical value of w⬘a (=40.191 rad/s) is very
close to the experimental one (=38.642 rad/s), the reliability of the theory and the
experimental equipments for step (f) should be acceptable.
Now, turn back to step (e). The “undamped” natural frequency of the oscillating
system with the model contacting with water for heave motion is given by Eq. (9), i.e.

冪m̄
k̄h
w̄⬘h⫽ (30)
h

Table 1
The principal particulars for the four models studied in this paper

Models Cross sections Lengths, Beam, B Depth, D Beam/draft Area Mass


L (m) (m) (m) ratios, b=B/d coeff. s displacement,
rwⵜ (kg)

1(a) Rectangular 0.6 0.08 0.13 0.75 1.0 4.99


0.85 4.47
0.95 3.95
1(b) 1.0 0.75 8.37
0.85 7.41
0.95 6.40
2(a) Lewis form 0.6 0.08 0.13 0.75 0.785a 4.53
0.85 4.01
0.95 3.49
(b) 1.0 0.75 6.67
0.85 5.71
0.95 4.93

a
For the immersed part.
J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438 431

The influence of the buoyancy (rwgBL) on the effective spring constant k̄n is shown
in Eq. (10) and the influence of the damping ratio (xw) on w̄⬘h is given by

w̄h⫽w̄⬘h 1−x2W (31)

It is evident that the value of m̄h obtained from Eq. (30) is the effective mass of
the entire oscillating system and the difference between m̄h and ma represents the
total added mass of the system due to the water surrounding the model, i.e.
mah⫽m̄h⫺ma⫽11.480⫺9.525⫽1.955 kg (32)

Hence the sectional added mass for heave motion is


m⬘⫽mah/L⫽1.955/0.6⫽3.258 kg/m (33)
and from Eq. (24a) one obtains the sectional added mass coefficient
m⬘

冉 冊
CV⫽ ⫽1.296 (34)
1
r pb2
2 w

The last value of CV agrees with that obtained from Todd (1961) in the conditions
that area coefficient s=1.0, beam/draft ratio b=B/d=0.95 and oscillation frequency
wW=w̄h=37.165 rad/s.

5.2. Effect of damping ratio xw on the sectional added mass coefficients CV and
CP

For Model 2(b) (with area coefficient s=0.785 and length L=1.0 m) in the con-
dition of beam/draft ratio b=B/d=0.95, Table 2 and Fig. 7 show the influence of

Table 2
Effect of damping ratios xW on the sectional added mass coefficients, CV and CP, for Model 2(b) at
b=B/d=0.95

Oscillating xW is considered xW is neglected


frequencies ww
(rad/s)
CV CP CV CP

10.0 1.103 1.106 1.174 1.188


15.0 1.048 1.063 1.116 1.141
20.0 1.025 1.039 1.091 1.115
25.0 1.017 1.027 1.083 1.102
30.0 1.015 1.022 1.080 1.096
35.0 1.013 1.020 1.078 1.094
40.0 1.013 1.020 1.078 1.093
432 J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438

Fig. 7. Influence of damping ratios xW on the sectional added mass coefficients, CV and CP, for Model
2(b) at b=B/d=0.95.

damping ratio xW on the sectional added mass coefficients, CV (obtained from heave-
motion tests) and CP (obtained from pitch-motion tests), in the range of oscillating
frequencies wW=10–40 rad/s.
In Fig. 7, the dashed lines %쐌% (and · · ·) represent the relationships between
CV (and CP) and wW by “neglecting” the effect of damping ratio xW, while the solid
lines —•— (and ———) denote those by “considering” the effect of xW. Among
the four curves, two of them with black points (%쐌% and —쐌—) are for CP vs
wW. From Fig. 7, one sees that the damping ratio xW has some effect on the coef-
ficients CV and CP. For example, the values of CV and CP at oscillating frequency
wW=20 rad/s are : CV=CVd=1.025 and CP=CPd=1.039 for the case of “considering”
the damping effect; while CV=CVO=1.091 and CP=CPO=1.115 for the case of “neg-
lecting” the damping effect. For convenience, the values of CV are represented by
CVd if the damping ratio xW is considered and by CVO if xW is neglected. Similarly,
CPd and CPO represent the values of CP with xW considered and neglected, respect-
ively. From Fig. 7, one sees that the values of CV obtained by considering the effect
of damping (i.e. xW⫽0) are smaller than the corresponding ones obtained by neglect-
ing the effect of damping (i.e. xW=0). For the example just mentioned above, one has

冉 CVd−CVO
CVd 冊
⫻100%⫽ 冉
1.025−1.091
1.025 冊
⫻100%⫽⫺6.439% (35)

and

冉 CPd−CPO
CPd 冊
⫻100%⫽ 冉
1.039−1.115
1.039 冊
⫻100%⫽⫺7.315% (36)
J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438 433

It is evident that the influence of xW on CV is almost the same as that on CP.


Besides, from Fig. 7, one sees that the two dashed lines are kept parallel approxi-
mately and close to each other and so are the two solid lines. In other words, when
either the damping effect is considered or neglected, the values of CV are very close
to the associated ones of CP.

5.3. Influence of oscillating frequencies wW on the coefficients CV and CP

For the cases of b=B/d=0.75, 0.85 and 0.95, the values of CV for the four models
[Models 1(a), 1(b), 2(a) and 2(b)] as shown in Fig. 6(a)–(d) obtained from the present
experimental method are listed in Table 3(a)–(c) and the values of CP for the Models
2(a) and 2(b) are listed in Table 4. Based on the data listed in Table 3(a)–(c) one
obtains the curves of CV vs wW, as shown in Fig. 8(a) for Model 1(a) (with L=0.6
m) and Fig. 8(b) for Model 1(b) (with L=1.0 m). Similarly, the relationships for CV
vs wW and CP vs wW, as shown in Fig. 9(a) for Model 2(a) (with L=0.6 m) and Fig.

Table 3
Added mass coefficients CV for the four models

Models 1(a) 1(b) 2(a) 2(b)


Lengths, L (m) 0.6 1.0 0.6 1.0

(a) Added mass coefficients CV for the four models at b=B/d=0.75


Oscillating frequencies, ww (rad/s) 10.0 1.482 1.546 1.194 1.144
15.0 1.430 1.486 1.148 1.091
20.0 1.398 1.453 1.131 1.069
25.0 1.379 1.443 1.125 1.062
30.0 1.370 1.440 1.123 1.061
35.0 1.368 1.438 1.121 1.060
40.0 1.367 1.440 1.122 1.060

(b) Added mass coefficients CV for the four models at b=B/d=0.85


Oscillating frequencies, ww rad/s) 10.0 1.456 1.510 1.159 1.128
15.0 1.401 1.428 1.117 1.076
20.0 1.367 1.393 1.099 1.055
25.0 1.348 1.381 1.092 1.048
30.0 1.338 1.378 1.090 1.045
35.0 1.335 1.376 1.088 1.043
40.0 1.336 1.376 1.089 1.042

(c) Added mass coefficients CV for the four models at b=B/d=0.95


Oscillating frequencies, ww (rad/s) 10.0 1.414 1.423 1.111 1.103
15.0 1.362 1.342 1.065 1.048
20.0 1.328 1.304 1.046 1.025
25.0 1.307 1.289 1.039 1.017
30.0 1.298 1.284 1.036 1.015
35.0 1.295 1.283 1.033 1.013
40.0 1.296 1.284 1.035 1.013
434 J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438

Table 4
Added mass coefficients CP for the Models 2(a) and 2(b)

Models 2(a) 2(b) 2(a) 2(b) 2(a) 2(b)


Length, L (m) 0.6 1.0 0.6 1.0 0.6 1.0
b=B/d 0.75 0.85 0.95

Oscillating frequencies, ww 10.0 1.263 1.156 1.226 1.130 1.178 1.106


(rad/s) 15.0 1.217 1.111 1.184 1.091 1.131 1.063
20.0 1.198 1.088 1.164 1.069 1.110 1.039
25.0 1.191 1.078 1.155 1.058 1.101 1.027
30.0 1.187 1.073 1.151 1.053 1.098 1.022
35.0 1.183 1.071 1.149 1.050 1.094 1.020
40.0 1.182 1.070 1.148 1.051 1.092 1.020

9(b) for Model 2(b) (with L=1.0 m) were obtained by using the data of Tables 3
and 4.
From Figs. 7–9, one finds that all the curves for CV vs wW and CP vs wW, in the
range of wW=10–40 rad/s, take the form
CX⫽C0⫹C1wW⫹C2w2W⫹C3w3W⫹C4w4W (37)
where Ci (i=0–4) are coefficients of curve fitting (Nakamura, 1991) and the subscript
X=V,P. In other words, either CV or CP may be obtained from a polynomial of wW
with the power of four. For example, the curve of CV vs wW at b=B/d=0.95 as shown
in Fig. 8(a) may be represented by
CV⫽1.584⫺2.029⫻10−2wW⫹5.588⫻10−4w2W⫺4.711⫻10−6w3W⫹2.629 (38)
⫻10−9w4W

5.4. The relationship between the coefficients CV and CP

In Section 1 of this paper, it has been shown that the values of total added mass
moment of inertia of a floating body, Iap, were determined by using the values of
the sectional added mass m⬘(=12rpb2CV) in the conventional approach (see Eq. (4)).
To confirm the reliability of this conventional approach, a comparison between CV
and CP was made by means of the curves shown in Fig. 9(a) and (b). Among these
curves, the ones “without” black points (쐌) are for CV vs wW, and those “with” the
black points are for CP vs wW. Three cases with b=B/d=0.75, 0.85 and 0.95, respect-
ively, were studied.
From the two figures one sees that the curves of CV vs wW (=w̄⬘h) are very close
to the corresponding ones of CP vs wW (=w̄⬘p) in various cases (L=0.6, 1.0 m and
b=B/d=0.75, 0.85, 0.95). It is noted that both CV and CP are the sectional added
mass coefficients of a floating body, the only difference is that CV is obtained from
the heave-motion tests and CP is obtained from pitch-motion tests. Hence the conven-
J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438 435

Fig. 8. The relationship between the oscillating frequency wW(=w̄⬘h) and the sectional added mass coef-
ficients, CV at b=B/d=0.75, 0.85 and 0.95 for: (a) Model 1(a) (L=0.6 m); (b) Model 1(b) (L=1.0 m).

tional approach of calculating the values of Iap from the values of CV is reasonable.
However, for a floating body with complicated configuration, the sectional added
mass coefficient CV is a function of coordinate (x) along the length of the body
and is usually difficult to be obtained. In such cases, the conventional approach of
determining the value of Iap from the values of CV by using the theoretical integration
(or summation) is not practical and directly obtaining the value of Iap will be required.
In this situation, the experimental method presented in this paper will provide a
useful key to the problem.
436 J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438

Fig. 9. The relationship between the oscillating frequencies wW(=w̄⬘h or w̄⬘P) and the sectional added
mass coefficients, CV and CP at b=B/d=0.75, 0.85 and 0.95 for (a) Model 2(a) (L=0.6 m); (b) Model 2(b)
(L=1.0 m).

5.5. Effect of area coefficient s, model length L and beam/draft ratio b

From the existing literature one sees that the area coefficient s is an important
parameter affecting the values of CV (and CP), hence the influence of s on CV in
J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438 437

the range of wW=10–40 rad/s is shown in Fig. 10. The latter was obtained from the
experimental results in the condition that b=B/d=0.95 for the four models shown in
Fig. 6(a)–(d). It is evident that the larger the value of s the larger the value of CV
no matter whether the model length is L=0.6 or 1.0 m.
From Fig. 10 one also finds that the curves of CV vs wW for L=0.6 m are almost
coincident with the ones for L=1.0 m if the values of the area coefficients s are the
same. This is because the influence of model length L on the coefficient CV is negli-
gible for the case of slenderness ratio L/B⬎12.5 (=1.0/0.08). As to the effect of
beam/draft ratio b(=B/d), from Figs. 8 and 9 one sees that the value of CV decreases
with increasing value of b. All the above-mentioned trends for the influence of area
coefficient s and beam/draft ratio b on the sectional added mass coefficient CV agree
with those presented in the existing literature (Todd, 1961).

6. Conclusions

For the determination of total added mass (mah) and total added mass moment of
inertia (Iap) of a floating body based on its oscillating natural frequencies contacting
with water, the damping ratio xW is an important parameter that should be considered.
In such a case, the experimental method presented in this paper provides an effective
and convenient technique, particularly when the configuration of the floating body
is so complicated that the theoretical calculation for the frequency-dependent values
of mah and Iap is not available.
For a floating body, the sectional added-mass coefficient obtained either from the
heave-motion tests (CV) or from the pitch-motion tests (CP) is a function of the
oscillating frequency wW.
Because the values of CP are very close to the corresponding ones of CV for the
same floating body with uniform cross sections along its length, the conventional

Fig. 10. The influence of area coefficient s on the sectional added mass coefficient CV for the four
models shown in Fig. 6 in the condition that b=B/d=0.95.
438 J.-S. Wu, M. Hsieh / Ocean Engineering 28 (2001) 417–438

approach to calculate the added mass moment of inertia of a floating body from its
sectional added mass is reasonable. The last result may also be a significant evidence
that the experimental method and the associated theory presented in this paper should
be reliable.
For the cases studied in this paper (with L/B⬎12.5) the influence of floating-body
length (L) on CV (or CP) is negligible, where B is the beam of the floating body.

Acknowledgements

The authors wish to thank, for financial support, the Committee of Agriculture of
the Republic of China, and also, for experimental help, Messrs. W.I. Chen, W.Y.
Shu, S.J. Yuang and M.L. Huang, etc.

References

Browne, A.D., Moullin, E.B., Perkins, A.J., 1930. The added mass of prisms floating in water. Proc.
Camb. Phil. Soc. 24, 258–272.
Clough, R.W., Penzien, J., 1975. Dynamics of Structures. McGraw-Hill Inc, New York.
Gerritsma, IR.J., 1957. Experimental determination of damping added mass and added mass moment of
inertia of a ship model. I.S.P. 14 (38), 505–519.
Koch, J.J., 1949. Experimental Method for Determining the Virtual Mass for Vibrations of Ship, Taylor
Model Basin. Trans 225, Washington, DC.
Korvin-Kroukovsky, B.V., Jacobs, W.R., 1957. Pitching and heaving motions of a ship in regular waves.
Trans SNAME 65, 590–621.
Kumai, T., 1958. Added Mass Moment of Inertia Induced by Torsional Vibration of Ships, European
Shipbuilding, 6.
Landweber, L., Mcagno, M., 1957. Added mass of two dimensional forms oscillating in a free surface.
J. Ship Res. (Nov.), 20–30.
Lewis, F.M., 1929. The inertia of the water surrounding a vibrating ship. Trans. SNAME 37, 1–20.
Marwood, W.J., Johnson, A.J., 1953. Vibration Test on an Up-River Collier, with Special Reference to
the Influence of Depth of Water. North East Coast Institution of Engineers & Shipbuilders, Newcastle
on Tyne.
Moullin, E.B., Browne, A.D., 1928. On the periods of a free–free bar immersed in water. Proc. Camb.
Phil. Soc. 24, 400–413.
Murtha, J.P., 1954. Virtual Mass of Partially Submerged Bodies, Thesis. Department of Civil Engineering,
Carnegie Institute of Technology.
Nakamura, S., 1991. Applied Numerical Methods with Software. Prentice-Hall International Inc, New
York(274–280).
Prohaska, C.W., 1947. The Vertical Vibrations of Ships. ATMA.
Tanaka, H., Kicagawa, H., 1962. On the study of characteristics of ship motion by a forced oscillation
method. J. Zosen Kiokai III.
Taylor, J.L., 1930. Vibration of Ships. Institution of Naval Architects, London.
Todd, F.H., 1961. Ship Hull Vibration. Edward Arnold Ltd, London.
Townsin, R.L., 1969. Virtual mass reductions factors ‘J’ values for ship vibration calculations derived
from tests with beams including ellipsoids and ship model. Royal Institute of Naval Architects III,
385–397.
Wu, J.S., Sheu, J.J., 1996. An exact solution for a simplified model of the heave and pitch motions of a
ship hull due to a moving load and a comparison with some experimental results. J. Sound Vibration
192 (2), 495–520.

Anda mungkin juga menyukai