547

THE

DESIGN

OF

UPRIGHT

BREAKWATERS

Depart.ent of Civil Engineering

Yokoha.a National University, Yokoha.a 240, Japan

ABSTRACT
The historical development of upright breakwaters in Japan is briefly
reviewed as an introduction. Various .wave pressure formulas for vertical walls are discussed. and then the design formulas currently employed
in Japan are presented with an exa.ple of calculation. Several design
factors are also discussed.
TABLE OF CONTENTS
1. Introduction

2. Historical Development of Upright Breakwaters in Japan

2.1 Examples of Upright Breakwaters in Modern History of Japanese
Ports
2.2 Some Features of Japanese Upright Breakwaters
3. Review of Wave Pressure Formulas for Vertical Walls
3.1 Hiroi's Formula
3.2 Saifl's Frmula
3.3 Minikin's Formula and Others
4. Design Formulas of Wave Pressures for Upright Breakwaters
4.1 Proposal of Universal Wave Pressure Formulas
4.2 Design Wave
4.3 Wave Pressure. Buoyancy. and Uplift Force
4.4 Stability Analysis
4.5 Example of Wave Pressure Calculation and Stability Analysis
5. Discussion of Several Design Factors
5.1 Precautions against Impulsive Breaking Wave Pressure
5.2 Structural Aspects of Reinforced Concrete Caisson
5.3 Ar.or Units for Rubble Mound
6. Concluding Reaarks
References
19-1

548

YOSHIMI GODA

l. INTRODUCTION

An upright breakwater is defined here as a structure having an upright

section rested upon a foundation. It is of ten cal led a vertical breakwater or
composite breakwater. The former is sometimes referred to a structure directly
built on the rock foundation without layers of rubble stones. The latter on
the other hand means a breakwater functioning as a sloping-type structure when
the tide level is low but as a vertical-wall structure when the tide level is
high. Because the terminology may vary from person to person, the definition
above is given here in order to avoid further confusion.

Upright breakwaters are of quite old structural type. Old ports in the
Roman Empire or ports in even older periods had been provided .ith breakwaters with upright structures. The upright breakwaters of recent construction have the origin in the 19th century. Italian ports have many upright
breakwaters as discussed in the following lecture by Dr. L. Franco. British
ports also have a tradition of upright breakwater construct ion as exemplified
in Dover Port. The British tradition can be observed in old breakwaters of
Indian ports such as Karachi, Bombay, and Madras. Japanese ports owes this
tradition of upright breakwaters to British ports, because the J80dern breakwater construction began at Yokohama Port in 1890 under supervision of British
aray engineer, retired Major General H. S. Palmer. Since then Japan has built
a large nu.ber of upright breakwaters along her long coastline extending over
34,000 km. The total length of upright breakwaters in Japan would exceed several hundred kilometers, as the totaJ extension of break.aters is more than
1,000 km.
The present note is intended to introduce the engineering practice of
upright breakwater design to coastal and harbot engineers in the world, based
on the experience of Japanese engineers.
2. HISTORICAL DEVKLOPMKNT OF UPRIGHT BRKAKWATERS IN JAPAN
2.1 Kxamples of Upright Breakwaters in Modern History of Japanese Ports

Figure 1 illustrates typical cross sections of upright breakwaters in

Japan in tillesequences, which is taken fiom Goda [1985J. The east breakwater
of Yokohaaa Port in Fig. 1 (a) utilized the local material of soft clayey
stones for rubble foundation and minimized the use of concrete blocks in the
upright section. The stone-filled middle section was replaced by concrete
blocks fully during reconstruction after the storm da.age in 1902. The wave
condition in Yokohama was not severe with the design height of 3 m.
The structural type of upright break_aters _as adopted at a more exposed
location of Otaru Port as sho_n in Fig. 1 (b) by I. Hiroi in 1897, _ho was the
chief engineer of regional government, later became a professor of the Tokyo
Imperial University, and established the framework of Japanese barbor engineering. The first reinforced-concrete caisson breakwater in Japan was built
at Kobe in 1911, based on the successful construction of caisson-type quay_all
19-2

11IE DESIGN OF UPRiGlIT BREAKW A!ERS

YOKOHAMA PORT Eut

Breakw.ter

H-3m

(1890 - 1896)
(Unit. in m)

." IUV. L. + 2.04

1} L.II'.L.+O.OO

Soft CI.yey Stones

Orilinll

Ground

~---------------

(a)

OTARU PORT North Breakwlter

H=6m

(1897-1907)
[Unit. in m}

(b)
OTARU PORT

I.land Brelkw.ter

H=6m
Harbor Side

(1912-1917)
[Unit. in m}

Su ..ar" Siele

+ 1.67 + (1.61

En,ineerinl

FiJI

(c)

Fig. 1 (a-c) Historical develop.ent of upright breakwater in Japan

af ter Goda [1985].
19-3

549

550

YOSHIMI GODA

ONAHAMA PORT First

Breokwlter

(1929-1938)
[Units in m)

H... ,=8.5m

Horbor Side

S .. word Side

..

H.W.L.+1.28
L.W.L.O.O

(d)
YOKOHAMA PORT Outer Br .. kw.ter (1928-1943)
[Units in m)
H=3m
Concrete

+2.2

--29.0---------11--------30.0---Ce)
WAKAYAMA North Horbor-We.t
H",=6m

Bre.kwlter

(1957-19601
[Units in m]

H.W.L+2.1
L.IV.
L.+0.3 ..

r--

17.9

(0
Fig. 1 (d-f)

Historical develop.ent of upright breakwater in Japan

(continued) af ter Goda [1985].
194

1HE DESIGN OF UPRlGIIT

OFUNATO PORT T.unlmi

551

BREAKWATERS

Brukwltu

(1962-1968)
rUni.. in ml
TJunlmi:II-6m.
T"15-40min'

Wind Wav~.:II -4m.T -91

(g)
HOSOJIMA PORT Brukwlter

(1974-1985)
[Unit in m)

H.=8.3m. T.,.=U.Os
'------24.5----..j

H.. bor SKIe

Suwlrd

Side

CIP Concrete
+2.5

H.IV.L. + 2.38
L.IV.L.O.OO

0.254.5 0.5

Qulrt)' .Run Fill

.....:

(h)
ONAHAMAPORT Offshore Break","ter (1980-)
[Unit in m)
H =7.4m. H.... =13.3m. T =13.0 I
H.rbor Side
Precaste Concrete SI.b

Se.... rd Side
.. H. W.L. + 1.40
;J L.W.L.+0.10

C.isson (20.0X15.0x 19.0)

(i)

Fig. 1 (g-i)

Historical develop.ent of upright breakwater in Japan

(continued) af ter Goda [1985].
195

552

YOSHIMI GODA

at Rotterdam in 1905. Then Hiroi, immediately seeing the br ight future of

caisson breakwaters, employed the concept to an island breakwater of Otaru
Port shown in Fig. 1 (c), where the design wave was 6 m high. He carried out
various field measurements, including wave pressures on a vertical wall, for
his finalization of breakwater design. Through these efforts, he came to
propose the wave pressure formula for breakwater design, which is to be
discussed in the next section.
Hiroi's breakwater caissons were filled with concrete for durability and
stability. The work time for concrete placement was sometimes saved by the
use of precast blocks as in the example of Onahama Port in Fig. 1 (d).
Concrete filling of breakwater caisson had been a tradition before the end of
World War 11, but a pioneering construction of reinforeed concrete caisson
breakwater with sand filling was carried out in Yokohama Port during the
period of 1928 to 1943: Fig. 1 (e) shows its cross section. Af ter World War 11
the use of sand as the filler material of caisson cells gradually became a
common practica in Japan.
The breakwater of Wakayama Port shown iriFig. 1 (f) was built upon a quite
soft ground so that it was provided with a wide foundation for the purpose of
counter-balancing the weight of upright section. The breakwater of Ofunato
Port in Fig. 1 (g) was built to reduce the inflow of tsunami waves into the
bay. The water depth of 35 m below the datum level was the deepest one at the
time of construct ion in 1962, but the present record of the deepest breakwater
in Japan is held at Kamaishi Port with the depth of 60 m. Some design features
and wave pressures on this breakwater have been discussed by Tanimoto and
Goda [1991bJ. One of the widest breakwaters is that of Hosojima Port shown in
Fig. 1 (h): the widest at present is found at Hedono Port in a remote island
with the width 38m (see Tanimoto and Goda 1991a). The breakwater of Onahama
Port shown in Fig. 1 (i) is of recent design us ing Goda' s wave pressure
formulas to be discussed later.
2.2 So.e Features of Japanese Upright Breakwaters
As seen in these examples, Japanese breakwaters of upright type have a few
common features. One is the relatively low crest elevation above the high
water level. Presently, the recommendation for ordinary breakwaters is the
crest height of 0.6 81/3 above the high water level for the design condition.
For the design storm condition, this elevation is certainly insufficient to
prevent wave agitations by the overtopped waves. However, it is a way of
thinking of harbor engineers in Japan that the design waves are accompanied by
strong gale and storm winds in any case and safe mooring of large vessels
within a li.ited area of harbor basin cannot be guaranteed even if wave agitations are reduced minimu.. As the storm waves with the return period of one
year or less are .uch lower than the design wave, the above crest elevation
is thought to be sufficient for maintaining a harbor basin cal. at the ordinary stormy conditions.
Another feature of Japanese upright bFeakwaters is a relatively wide ber.
of rubble foundation and provision of two to three rows of large foot (toe)
19-6

553

TIIE DESIGN OF UPRIGHf BREAKW ATERS

protection blocks. There is no fixed rule for select ion of the ber. width and
engineers always consult with the examples of existing breakwaters in the
neighborhood or those at the location of similar wave conditions. It is sOllewhat proportional to the size of concrete caisson itseif, but the fina! decision must await good judgment of the engineer in charge. The foot protection
concrete blocks have the size ranging from 2 to 4 m in one direction and the
height of 1.5 to 2 m, weighing 15 to 50 tf. Though these blocks used to be
solid ones, recent blocks are provided with several vertical holes to reduce
the uplift force and thus to increase the stability against wave action.
A new development in upright breakwaters of Japan is the e.ployment of
various modifications to the shape of concrete caissons, such as perforated
walis, vertical slits, curved slits with circular are lIembers, dual cylindrical walls and others (see Tanimoto and Goda 1991a). These new caisson
shapes have been developed to actively dissipate wave energy and thus to
reduce wave reflection and wave pressures. A number of these breakwaters have
been built and functioning as expected.

3. REVIEW OF WAVE PRESSURE FORMULAS FOR VERTICAL WALL

3.1 Hiroi's For.ula
Prof. Hiroi published the wave pressure formula for breakwater design in
1919. It is a quite simple formula with the uniform pressure distribution
of the following intensity:
p

1. 5

Wo

(1)

where Wo denotes the spec i I ic weight of sea water and H the incident wave
height. This pressure distribution extends to the elevation of 1.25 Habove
the design water level or the crest of breakwater if the latter is lower, as
shown in Fig. 2.

Fig. 2 Wave pressure distribution by Hiroi's formula.

Prof. Hiroi explained the phenomenon of wave pressure exerted upon a vertical wall as the momentum force of illpinging jet flow of breaking waves and
gave the reasoning for its quantitative evaluation. However, he must have had
SOllegood judgment on the magnitude of wave pressure from his long experience
of harbor construction and several efforts of pressure measurements in situ.
He states that he obtained the records of wave pressure exceeding 50 tf/.2
by
19-7

YOSHIMIGODA

554

the pressure gauges set at a concrete wall in water of several meters deep.
Nevertheless, he did not incorporate such high pressures into the formuia of
breakwater design, by saying that the high wave pressure must have lasted for
only a short duration and are ineffective to cause appreciable damage to
breakwaters.
Hiroi's wave pressure for.ula was intended for use in relatively shallow
water where breaking waves are the governing factor. He also recommended to
assuae the wave height being 90% of water depth if no reliable information is
available on the design wave condition. Hiroi's wave pressure formula was
soon accepted by harbor engineers in Japan, and almast all breakwaters in
Japan had been designed by this formuIa till the mid-1980s.
The reliability of Hiroi's formuIa had been challenged thrice at least.
The first challenge was the introduction of Sainflou's formula in 1928 for
standing wave pressures. Differentiation of two formulas was made, by referring to the recommendation of PIANC in 1935, in such a way that Hiroi's for.ula was for the case of the water dep th above the rubble foundation being
less than twi~e the incident wave height, while Sainflou's formula was for the
water depth equal to or grater than twice the wave height. The second challenge was raised when the concept of significant wave was introduced in early
1950s. Whicb one of H.... , H1/10,
or HI/3 is to be used in Hiroi's oraula
was the question. A consensus was soon forlled as tbe recollmendation for tbe
use of HI/3
based on the examination of existing breakwater designs and wave
co~ditions. Tbe tbird challenge was made by Goda [1973] against tbe insensitivity of the estimated pressure intensity to tbe variations in wave period
and other factors. Hiroi's formula could not meet this challenge and is not
used presently for the design of major breakwaters.
Though the pressure formuIa by Hiroi was sa sillple, the total wave force
thus estimated was quite reliable on the average. Thanks to this characteristic, Japanese breakwaters had rarely experienced catastrophic damage despite
the very long extension around the country.
3.2 Sainflou's For.ula
As weIl known, Saiflou published a theory of trochoidal waves in front of
a vertical wall in 1928 and presented a simplified formuia for pressure estiaation. Tbe pressure distribution is sketched as in Fig. 3, and the pressure
intensities and the quantity of water level rise 00 are given as
PI
pz

00

+ 1'0 h) (H+oo)
H / cash leh
( 11:
H 2/ L ) co th leh

(pz

"0

(h +H+oo)

}
(2)

where L is the wave 1ength and Ie is the wavenumber of 211:/ L.

Sainflou [1928] presented the above formula for standing wave pressures of
nonbreaking type and the for.ula has been so utilized. The for.ula was derived for the purpose of practical application froa the standpoint of a civil
19-8

555

THE DESIGN OF UPRIGHT BREAKWATERS

engineer and it has served its objective quite weil. Just like the case of
Hiroi's formula. it was born when the concept of wave irregularity was unknown.
There seems to exist no established rule for the choice of representative wave
height to be used with Sainflou's formula. Some avocates the use of HI/3
some favors Hl/ID. and the other prefers the selection of Hl.".

Fig. 3 Wave pressure distribution by Sainflou's formula.

It was customarily in Japan to use HI/3 with Sainflou's formula but in a
modified form. Through examinations of several minor damage of breakwaters,
it had been revealed that a simple application of Sainflou's for.ula had
yielded underestimation of wave pressures under storm conditions. For the
zone extending H /2 around the design water level, the wave pressure by
Sainflou's fou.ula was replaced with that by Hiroi's for.ula. The .odified
formula was sometimes cal led the partial breaking wave pressure for.ula in
Japan, because it was aimed to introduce the effect of partial wave breaking
in relatively deep water. The dual system of Hiroi's wave pressure for.ula for
breaking waves and of modified Sainflou's formula for standing waves had been
the recoended engineering practice of breakwater design in Japan for the
period fro. around 1940 to the early 1980s.

3.3 Minikin's For.ula and Others

Although Hiroi's formula had been regarded as the most dependable for_ula
for breaking wave pressures in Japan, it remained unknown in Europe and A.erica. As the field measurement at Dieppe revealed the existence of very high
pressures caused by impinging breaking waves and the phenomenon was confir.ed
by laboratory experiments by Bagnold [1939], harbor engineers in western countries began to worry about the impact breaking wave pressures. Then in 1950,
Mini kin proposed the following for.ula for breaking wave pressures, .hieh
consisted of the dynamic pressure P. and the hydrostatic pressure P. as
sketched in Fig. 4:
Oyna.ic pressure
P.
p.

p.

101

"0

(l-2Izl/H)2
d (1 + d / h ) H / L

I z I ~ H/2 }
(3)

Hydrostatic pressure :
p. =

J 0.5
I 0.5

"0
'0

H (1 H

2z / H )

19-9

o ~
z <

z
0

<

H/2

(4)

556

YOSHIMI GODA

Because it was the first descriptive formula for breaking wave pressures,
it was immediately accredited as the design formula and listed in many textbook and engineering manuals. Even in present days, technical papers based on
Minikin's formula are published in professional journals from time to time.

Fig. 4 Wave pressure distribution by Minikin' s formula.

Minikin [1950] did not give any explanation how he derived the above formulation except for citing the experiments of Bagnold. In the light of present knowledge on the nature of impact breaking wave pressures, the formula
has several contradictory characteristics. First, the maximum intensity of
wave pressure increases as the wave steepness increases, but the laboratory
data indieates that waves with long periodicity tends to generate weIl developed plunging breakers and produce the impact pressure of high intensity. In
f'act , Bagnold earried out his experiments us ing a solitary wave.
Second, Eq. 3 yields the highest ~.. when d is equal to h or when no
rubble foundation is present. It is harbor engineers' experienee that a
breakwater .ith a high rubble mound has a larger possibility of being hitten
by strong breaking wave pressures than a breakwater with alo. rubble mound.
Third, Minikin's formula yields excessively large wave force against whieh
no rational upright breakwater could be designed. To the author's knowiedge,
no prototype breakwater has ever been cons t ruct.ed w ith the wave pressures
estimated by Minikin's for.ula. Reanalysis of the stability of prototype
breakw~ters in Japan which experieneed storm waves of high intensity, some
undamaged and others having been displaeed over a few meters, has shown that
the safety factor against sliding widely varies in the range between 0.31 and
2.06 [Goda 1973b and 1974]. The safety factors of undamaged and displaeed
breakwaters were totally mixed together and no separation was possible. Thus
tne applieability of Minikin's formula on prototype breakwater design has been
denied definitely.
There has been several proposals of wave pressure formulas for breakwater
design. Among them, those by Nagai [1968, 1969]] and Nagai and Otusbo [1968]
are most exhaustive. Nagai elassified the various patterns o~wave pressures
aeeording to the wave eonditions and the geometry of breakwater, and presented
several sets of design formuilas based on many laboratory data. However, his
system of wave pressure formulas was quite.eomplieated and these formulas gave
different prediction of wave pressures at the boundaries bet.een the zones of
19-10

THE DESIGN OF' UPRIGlIT BREAKW ATERS

557

their applications. Another problem in the use of Nagai's method is the lack
of specification for represtantaive wave height for irregular waves. There was
onlya few cases of verification of the applicability of his aethod for breakwater design using the performance data of prototype breakwaters. Because of
these reasons, the method is not used in Japan presently.
The Miche-Rundgren formula for standing wave pressure [CERC 1984J represents an effort to improve the accuracy of Sainflou's formula for engineering
application. Certainly, the formula would give better agreement with the
laboratory da ta than Sa infIou's one. However, it has not been ver ified w ith
any field data and its applicability for brekwater design is not confirmed yet.
4. DESIGN FORMULAS OF WAVE PRESSURES FOR UPRIGHT BREAKWATKRS

4.1 Proposal of Universal Wave Pressure Forauias

It is a traditional approach 1n wave pressure calculation to treat the
phenomena of the standing wave pressures and those by breaking waves separately. Casual observations of wave forms in front of a vertical wall could
lead to a belief that breaking wave pressures are much more intensive than
nonbreaking wave pressures and they should be calculated differently. The
previous practice of wave pressure calculation with the dual forauias of
Hiroi's and Sainflou's in Japan was based on such belief. The popularity of
Minikin's formula prevailing in western countries seeas to be owing to the
concept of separation of breaking and nonbreaking wave pressures.
The difference between the magnitudes of breaking and nonbreaking wave
pressures is a misleading one. The absolute magnitude of breaking wave pressures is certainly much larger than that of nonbreaking one. The height of
waves which break in front of a vertical wall, however, is also greater than
that of nonbreaking waves. The d iaens ion less pressure intensity, p /"0 H.
therefore, increases only gradua~ly with the increase of incident wave height
beyond the wave breaking limit, as dellonstrated in the extensive laboratory
data by Gda [1972J.
A practical inconvenience in breakwater design with the dual pressure formula system is evident when a breakwater is extended offshoreward over a long
distance from the shoreline. While th site of construction is in shallow
water, the wave pressures are evaluated with the breaking wave pressure foraula. In the deeper portion, the breakwater would be subject to nonbreaking
waves. SOllewhere in between, the wave pressure f'oreula aust be swi tched froa
that of breaking to nonbreaking one. At the switching section, the estiaated
wave pressures jump from one level to another. With the Japanese systea of
the combined formulas of Hiroi's and aodified Sainflou's, the juap was about
30%. To be exact with the pressure calculation, the width of upright sectiorr
aust be changed also. However, it is against the intuition of harbor engineers
who believe in saooth variation of the design section. The location of
switching section is also variabie, dependent on the design wave height. If
19-11

YOSHIMI GODA

558

the design wave height is modified by a review of storm wave conditions af ter
an experience of some damage on the breakwater, then an appreciable length of
break.ater section would have to be redesigned and reconstructed.
The first proposal of universal wave pressure formula for upright break.ater was aade by Ito et al. [1966] based on the sl id ing test of a model section of breakwaters under irregular wave actions. Then Goda [1973b, 1974]
presented another set of formulas based on extensive laboratory data and being
supported by verification with 21 cases of breakwater displacement and 13
cases of no damage under severe storm conditions. The proposed forllulas were
critically reviewed by the corps of engineers in charge of port and harbor
construction in Japan, and they were finally adopted as the recommended
Corllulas for upright breakwater design in Japan in 1980, instead of the
previous dual Cormulas of Hiroi's and modified Sainflou's.
4.2 Design Wave
The upright breakwater should be designed against the greatest force of
single wave expected during its service life. The greatest force .ould be
e~erted by the highest .ave allong a train of random waves corresponding to the
design condition on the average. Thus the wave pressure forllulas presented
herein are to be used together with the highest .ave to be discussed belo
(1) flave height
H....

I 8 Hl/3
{ ..in { ({Jo

h/ lo

+ (JI'

Ho'

h)

K. Ho'

in which the symbol

and Ho' denotes the
{Jo and others have
data of random wave
{J 0
{JI
f3

ra. x

Ho'

+ (J

h ),

{J....

Ho',

h/ lo !i,';; O. 2
}
h/ lo
0.2

(6)

s, Ho'

<
<

min{a, b, c } stands for the minimum value among a, band c,

equivalent deepwater significant height. The coefficients
empirically been formulated from the numerical calculation
breaking in shallow water as follows, af ter Goda [1975]:

0.028

oe u:

)-0.38

exp[20tanI.5B]

0.52 exp]4.2tan e ]
max {0.92, 0.32 (Ho'/lo
0.052

IJ ......

0.2
(5)

h/
min { ({J

!i,';;

l. 8 HI/3 }
lo
0.2

{J .... ' Ho',

(Ho' / t; ) -0.3.

0.63 exp[3. 8tan B]

aax (1.65,0.53 (Ho'/lo

}
(7)

)-0.29

exp[20tan

exp[2.4tanB]}
I. 5

B]

}
(8)

)-0.29

exp[2.4tanB]}

in .hich the sYllbol lIax{a, b } stands for .the larger of

notes the inclination of sea botto .
The shoaling coefficient K. is evaluated
19-12

by

a or b, and tan e de-

taking the finite

allplitude

559

THE DESIGN OF UPRIGHT BREAKWATERS

effect into consideration. Figure 5 has been prepared

on the theory of Shuto [1974J.

3.0

Ol

~~:..!#lfH
v,

:z:1:t: 2.5 I\.

11

Ia

,j 2.0

!ti
ti

IlO

.5

'\
I\.

~c9~

MLo
0.4

I-HtI1frl

111111111

1\

111

1111

J I

I I II

0"L

l\ .

I'"..
..~

0.01

002

I...,~.
...,

I/)

- . :::,......02
::::: S

1.0
f-'

0.006

1111111111

I I IJ

1.5

0.004

0.6

;;Jj().h
H/Lo:/.'lJ ..
0.005

ilo/Lo=0.0005
1
.......
~ 0
0.00
i'-

8
]

,... ~
\

..ti

0.3

~4,.;

~\

I~

0.2

0.15

for this purpose based

0.008 0.01

0.04
''"'
:'::: ~: t::-._ ...
I

0.015 0.02
0.03 0.04
Rel.tive W.ter Depth. It/Lo

0.06

0.08 0.1

Fig. 5 Diagram of nonlinear wave shoaling coefficient K.

The selection of the fixed relation H.... = 1.8 H1/3
outside the surf
zone was based on three factors of reasonining. First. the fixed ratio was
preferred to an introduction of duration-dependent relation based on the
Rayleigh distribution of wave heights. because such variability in the design
wave height would cause soae confusion in design procedures. Second, the
exallinatlOn oi prototype breakwater performance under severe stor. wave
actions yielded reasonable results of safety factor against sliding by using
the above fixed relation. Third. a possible
deviation of the ratio
H. /H'/3 froll1.8 to 2.0, say, corresponds to an increase of 11% and it
can be covered within the margin of safety factor whieh is custo.arily taken
at 1.2. However, it is a reeoendation and an engineer in charge of breakwater design can us. other criterion by his own judg.ent.
For evaluation of H by the second part of Eq. 7 or within the surf
zone, the'water depth at a distance 5 H1/3 seaward of the breakwater should be
e.ployed. This adjust.ent of water dep th has been introduced to si.ulate the
nature of breaking wave force whieh beeo.es the greatest at so.e distance
shoreward of the breaking point. For a breakwater to be built at the site of
steep sea bOttOll, the location shift for wave height evaluation by the distance 5 H1/3 'produces an appreciable increase in the aagn i tude of wave force
~nd the resultant widening of upright section.

19-13

560

YOSHIMI GODA

(2) KavePeriod
The period of the highest wave is taken as the same with the significant
wave per iod of design wave, i.e.,
(9)

TM = T'/3

The relation of Eq. 9 is valid as the ensemble mean of irregular waves. Though
individual wave records exhibit quite large deviations from this relation, the
use of Eq. 9 is ~eco.mended for breakwater design for the sake of simplicity.
(3) AngJe of Kave Inc idence to Breekes ter
Waves of oblique incidence to a breakwater exert the wave pressure smaller
than that by waves of noraal incidence, especially when waves are breaking.
The incidence ang Ie {3 is measured as that between the direction of wave approach and a line noraal to the breakwater. It is recommended to rotate the
wave direction by an amount of up to 15 toward the line normal to the breakwater froa the principal wave direction. The recommendation was originally
g.i ven by Prof. Hiroi together .ith his wave pressure formula, in consideration
of the uncertainty in the esti.ation of wave direction, which is essentially
based on the 16 points-bearing of wind direction.

'.3

Wave Pressure, Buoyancy, and Uplift Pressure

(1)

EJevation to "hieh the the "ave pressure is exerted

The exact elevation of wave crest along a vertical wall is difficult to

assess because it var ies c.onsiderably from 1. OH to more than Z.OH, depending
on the wave steepness and the relative water depth. In order to provide a
consistency in wave pressure calculation, however, it was set as in the f'ollowing siaple forauia:

Fi~.-6 Wave pressure distribution by Goda's for.ulas.

19-14

1lIE DESIGN OF UPRIGHr

n"

0.75 (1

+ cos (J)

561

BREAKW ATERS

H....

(10)

For waves of normal incidence, Eq, 10 gives the elevation of

(2)

Kave pressure exerted upon the front face of a vertical

= 1. 5 H

'fJ'

wal1

The distribution of wave pressure on an upright section is sketched in

Fig. 6. The wave pressure takes the largest intensity PI at the design water
level and decreases linearly tcvards the elevation 'fJ' and the sea bottom, at
which the wave pressure intensity is designated as pz.
The intensities of wave pressures are calculated by the following:
PI
PZ
P3

0.5 (1 + cos IJ )
PI/cosh kh
a 3 PI

(al

+a

z cosZ (J)

Wo

Hm.

(11)

}}

(12)

in which
al
az

a3

0.6 + 0.5 [2kh/sinh 2leh F

min {[(ho - d )/3h.
](H .... / d )Z, 2d/ H....
1- (h ' / h ) [1 - 1/ cosh leh ]

where h. denotes the water depth at the location at a distance

.of the breakwater.

5HI/3

seaward

The coefficient a I takes the minimum value 0.6 for deepwater waves and
the maximum value 1.1 for waves in very shallow water. It represents the effect of wave period on wave pressure intensities. The coefficient a Z is
introduced to express an increase of wave pressure intensities by the presence
of rubble mound foundation. Both coefficients a I and a z have eap ir ica lly
been formulated, based on the data of laboratory experiments on wave
pressures. The coefficient a 3 is derived by the relation of linear pressure
distribution. The above pressure intensities are assulled to re.ain the sa.e
even if wave overtopping takes place.
The effect of the incident wave angle on wave pressures is incorporated in
(1 + cos (J) and a aod if ication to the term
of az with the factor of cos" {J.

n ' and PI with the factor of 0.5

(3) Buoyancy and uplift

pressure

The upright section is subject to the buoyancy corresponding to its displacement volume in still water below the design water level. The uplift pressure acts at the bottom of the upright section, and its distribution is assumed to have a triangular distribution with the toe pressure P. given by Eq.
13.
P. = 0.5 (1 + cos (J)
a I a 3 Wo H
(13)
The toe pressure P. is set smaller than the wave pressure Pa at the lowest
point of the front wall. This artifice has been introduced to improve the ac-

19-15

562

YOSHIMI GODA

curacy of the prediction of breakwater stability, because the verification

with the data of prototype breakwater performance indicated some overestimation of wave force if P. were taken the same with P3.
When the crest elevation of breakwater he is lower than 7J' , waves are
regarded to overtop the breakwater. Both the buoyancy and the uplift pressure,
however, are assumed to be unaffected by wave overtopping.
4.4 Stability Analysis
The stability of an upright breakwater against wave action is examined for
the three modes of failure: ~e., sliding, overturning, and collapse of foundation. For the first two modes, the calculation of safety factor i.sa common practice' of examination. The safety factors against sliding and overturning are defined by the following:
Against sliding:

S.F.

Against overturning

S.F.

IJ. (I;-U)
(lit

-Nu

/P

(14)

) / Np

(15)

The notations in the above equations are defined as follows:

Np
Nu

P
t
ij

K
IJ.

moment of total wave pressure around the heel of upright section

moment of total uplift pressure around the heel of upright section
total thrust of wave pressure per unit extension of upright section
horizontal distance between the center of gravity and the heel of
upr igh t section
total uplift pressure per unit extension of upright section
weight of upright section per unit extension in still water
coefficient of friction between the upright section and the rubble
mound

The safety factors against sliding and overturning are dictated to be

equal to or greater than l. 2 in Japan. The friction coefficient between
concrete and rubble stones is usually taken as 0.6. The coefficient seems to
have a smaller value in the initial phase of breakwater installment, but it
gradually rises to the value around 0.6 through consolidation of the rubble
.ound by the oscillations of the upright section under wave actions. The fact
that most of breakwater displacments by storm waves occur during the construction period or within a few years after construction supports the above
conjecture.
The bearing capacity of the rubble mound and the sea bottom foundation was
used to be exarninedwith the bearing pressures at the heel of upright section
and at the interface between the rubble mound and the foundation. Ho.ever, a
recent practice in Japan is to make analysis of circular slips passing through
the rubble rnoundand the foundation, by utilizing the simplified Bishop aethod
(see Kobayashi et al. 1987). For the rubble mound, the apparent cohesion of c
= 2 tf/.2 and the ang le of internal friction of ~ =35 are reco..ended.
0

19-16

THE DESIGN OF UPRIGHT

4.5

563

BREAKWATERS

Exallple of Wave Pressure Calculation

An example of calculation is given here in order to facilitate the understanding of the breakwater design procedure. The design wave and site conditions are set as in the following:
Waves:
Depth etc. :
Bottom slope:

Ho'
h

tan

7.0 m,
18 m,
= 1/50

T'/3

= 11 s ,
d = 10 m,

(3

h '

= 10
= 11.5 m,

h.

4.5 m

The incident wave angle is the value af ter rotation by the amount up to 15 .
The geometry of upright breakwater is illustrated in Fig. 7.

Fig.7

Sketch of upright breakwater for stability analysis.

i) Design wave height

H.... and the maximumelevation

of wave pressure

90

The coefficients for wave height calculation are evaluated as

Lo =
{3 0 =
{3 o' =

188.8 m, Ho' / Lo = 0.0371, h / t; = 0.0953,

K.
0.1036, P I = 0.566,
11 max = min {O. 92, O.84}
O.1924, {3,'=
0.680,
{3 ..... = min {1.65, 1.39}

0.94
0.92
1.65

Then, the wave heights and the maximum elevation are obtained as
H'/3

hh

H....
",0

ij)

min {l0.91, 6.44, 6.58} = 6.4411

18.0 + 5X6.44/50
= 18.64 m
ain {l4.02, 11.55, 11.84} = 11.5511
0.75X Cl + cos10) X11.55 = 17.1911

Pressure cOlllponents

The wavelength at the depth 18

wave pressure are evaluated as
kh a,

11

is

131.5

211'.
X 18/131. 5 = 0.860
0.6 + 0.5X[2XO.860/sinh(2XO.860))2

19-17

11.

The coefficients

0.802

for

564

YOSHIMl GOpA

min {[(l8.64-10.0)/(3X18.64)]X(l1.55/10)2
2X10/11.55
)
min {O. 206, 1. 732)
0.206
1/cosh(0.860)]
1-11.5/18.0X
[1

a2
a3

Then, the intensities

of wave pressure

= 0.820

and uplift

pressure

are calculated

as
5X 0+0.9848)
X [0.802+ O. 206X (0.9848)2]
11.83 tf/m2
11. 83/ cosh(O. 860) = 8.49 tf/m2
0.820X11.83
= 9.70 tf/m2
11.83X Cl - 4.5/17.19)
= 8.73 tf/m2
0.5XO+0.9848)XO.802XO.820X1.03X11.55

PlO.
P2
Pa
P~
P.

The symbol P~ denotes

i i i)

the pressure

TotaJ pressure end uplift,

intensity

u=

0.5X18.0X7.76
(2/3) X 69. 8 X 18

i v)

StabiJity
The specific

= 7.76 tf/m2

at the top of upright

section.

end their soment.s

P = O.5X (11. 83+ 9. 70) X 11. 5+ O. 5X

/lp = 1366.2 tf-m/m

s; =

XL 03X 11. 55

ui. 83+

7. 76) X 4. 5

167.9

tf/m

69.8 tf/m
837.6 tf -Rl/m

of upright section against wave action

weight

The portion
The portion

of upright

section

above the elevation

below the elevation

is assumed as in the following:

+0.5 m
+ 0.5 m

re

= 2.3 tf/m3

r ..'= 2.1 tf /m3

The difference
in the specific
weight reflects
a current practice
of sand filling in the cells of concrete caisson.
The weight of upright
section
is calculated for the dry and in situ conditions,
respectively,
as

11. = 2.1X01.5+0.5)X18.0+2.3X(4.5-0.5)X18.0
11 = 619.2-1.03X11.5X18.0
The safety
are calculated
Against
Against

factors
against
sliding
as in the following:

sliding:
overturning:

S.F.

S.F.

619.2

tf/m

406.0 tf/m
and overturning

of the upright

0.6 X (406.0- 69. 8)/167. 9

(406.0 X 9. 0- 837. 6)/1366.2

section

1. 20
2.06

Therefore,
the upr igh t breakwater
with
the unif'or s width of B = 18.0
sketched in Fig. 7 is considered
stabie aga ins t the design wave of Ho' =7.0
and TI/3 = 11. 0 s.

19-18

11
11

1HE DESIGN OF UPRIGHT BREAKW ATERS

565

5. DISCUSSION OF SEVERAL DESIGN FACTORS

5.1

Precautions against Impulsive Breaking Wave Pressure

The universal wave pressure formulas described hereinbefore do not address

to the problem of impulsive break ing.wave pressure in a direct manner. The
coefficient a 2. however. has the characteristic of rap id increase with the
decrease of the ratio d IHm This increase roughly reflects the generation of impulsive breaking wave pressure.
Though the impact pressure of breaking waves exerted upon a vertical wall
is much feared by coastal and harbor engineers, it occurs under the limited
conditions only. If waves are obliquely incident to a breakwater, the poss ibility of impact pressure generation is slim. If a rubble aound is low, the
Rea bottom should be steep and waves be of swell type for the impact pressure
to be generated. A most probable situation under which the impact pressure is
exerted upon an upright breakwater is the case with a high rubble mound with
an appreciable berm width (see Tanimoto et al. 1987). Most of breakwater
failures attributed to the action of the impulsive breaking wave pressure are
due to the wave forces of normal magnitude, which could be estimated by the
universal wave pressure formulas described in the present lecture note.
The impact pressure of breaking waves last for a very short time duration,
which is inversely proportional to the peak pressure intensity. In other
words. the impulse of impact pressure is finite and equal to the forward
momentum of advancing wave crest which is lost by the contact with the
vertical wall. The au thor has given an estimate of the average value of the
impact pressure effective in causing sliding of an upright section, by taking
into account the elastic nature of a rubble mound and foundation [Goda 1973aJ.
Because the major part of impact is absorbed by the horizontal oscillations
and rotational motion of the upright section, the impact pressure effective
for si iding is evaluated as (2-3) 11'0 Hu .
Neveriheiess, ihe above order of pressure intensity is too great to be
taken into the design of upright breakwaters: the mean intensity of wave pressure employed for the stability analysis of the breakwater sketched in Fig. 7
is only 0.91 11'0 Hm...
Engineers in charge of breakwater design should arrange the layout and the cross section of breakwater in such way lo avoid the
danger of impact pressure generation. Ir the exertion of impulsive breaking
wave pressure on the upright section seems inevitable. a change in the type of
breakwater structure, such as a sloping-type breakwater or a vertical breakwater protected by a mound of concrete blocks, should be considered.
5.2 Structural Aspects of Reinforced Concrete Caisson
The upright section of vertical breakwater is nowadays made by reinforced
concrete caisson. The width is determined by the stability condition against
wave action. The height of caisson or the base elevation is so chosen to yield
the minimum sum of the construction cost of rubble mound and upright section.

19-19

566

YOSHIMI GODA

The length of caisson is governed by the capacity of manufacturing yard. In

March 1992, Kochi Port facing the Pacific in Shikoku, Japan, set a breakwater
caisson with the length 100 m in position. It is of hybrid structure with
steel frames and prestressed concrete.
A concrete caisson is divided into a number of inner celis. The size of
inner cells is limited to 5 m or less in ordinary design. The outer wall is 40
to 50 cm thick, the parti tion wal I 20 to 25 cm thick, and the bottom slab 50
to 70 cm thick. These dimensions are subject to the stress analysis of
reinforced concrete. As the upright breakwater withstands the wave force
mainly with its own weight, the use of prestressed concrete for breakwater
caisson is not advantageous in the ordinary situations. For the caisson of
special shapes for enhancing wave dissipation such as the caisson with circular arc members, prestressed concrete is utilized.
5.3 Ar.or Units for Rubble Mound
The berm and slope of a rubble mound needs to be protected with armor
units against the scouring by wave action. Foot-protection blocks weighing
from 15 to 50 tf are placed in front of an upright section. The rest of the
berm and slope are covered by heavy stones and/or specially-shaped concrete
blocks. The selection of armor units is left to the judgment of engineers,
with the aid of hydraulic model tests if necessary.
A formula for the weight of armor stones on the berm of rubble mound has
be~n proposed by Tanimoto et al. [1982J as the results of systematic model
tests with irregular waves. The minimum weight of armor stones can be calculated by a formula of the Hudson type:
ft=

'Y,

H,/3'

[H."

(S, -

(16)

l)"J

In which ft is the weight of armor stones, 'Y

the specific weight of armor
stones, S, the ratio of 'Y, to the specific weight of seawater, and H. the
stability number, the value of which depends on the waveconditions and mound
dimensions.
For waves of normal incidence, Tanimoto et al. [1982J
function for armor stones:
N.

lAax {I .,8

(1 ..~3

1-"

gave

Hh'
+ 1. 8 exp [- 1. 5 (1-,,)2
1/3
,,173

the following

h'
u;-;;

l) }
(17)

in which the paraaet er " is calculated by

,,=

[Uh

' / sinh 2 kh ' J sin 2 (2 1t Bil / L ')

(18)

and where h ' denotes the water depth at which ar.or stones are p1aced. L'
the wavelength at the depth h '. and Bil the berm width.

19-20

THE DESIGN OF UPRIGHT

BREAKWATERS

567

Though the stability number for concrete blocks has not been formulated,
a similar approach to the data of hydraulic model tests on concrete blocks
will enable the formulation of the stability number for respective types of
concrete blocks.
6.

CONCLUDING REKARKS

The design and construction of upright breakwaters is a weil established,

engineering practice, at least in Japan, Korea, and Taiwan. A large number of
these breakwaters have been built and will be built to protect ports and harbors. In these countries, the problem of impulsive breaking wave pressure is
rather lightly dealt with. The tradition owes to Prof. Hiroi, who established
the most reliable wave pressure formula in shallow water and showed the upright breakwaters could be successfully constructed against breaking waves.
This is not to say that no breakwaters have failed by the attack of storm
waves. Whenever a big storm hits the coastal area, several reports of breakwater damage are heard. However, the number of damaged caissons is very small
compared with the total number of breakwater caissons instalied along the
whole coastline. Probably the average rate per year would be less than 1%,
though no exact statistic is available. Most cases of breakwater damage are
attributed to the underestimation of the storm wave condition when they were
des igned.
In the past, the majority of breakwaters were constructed in relatively
shallow water with the depth up to 15 m, for example, because the vessels
calling ports were relatively small. In such shallow water, the storm wave
height is controlled by the breaking limit of the water depth. One reason
for the low rate of breakwater failure in the past could be this wave height
limitation at the locations of breakwaters.
The site of breakwater construction is moving into the deeper water in
these days. Reliable evaluation of the extreme wave condition is becoming the
most important task in harbor engineering, probably much more than the
improvement of the accuracy of wave pressure prediction.
REFKRKNCES
Bagnold, R.A. [1939]: Interim report on wave-pressure research, J. lnst. Civil
Engrs., Vol.12, pp.202-226.
CERC (Coastal Engineering Research Center, U.S. Army Corps of Engrs.) [1984]:
Shore Protection Kanual, U.S. Government Printing Office, pp.7-161--173.
Goda, Y. [1972]: Experiments on the transition from nonbreaking to postbreaking wave pressures, Coastal Engineering in Japan, Vol.15, pp.81-90.
Goda, Y. [1973a]: Kotion of composite breakwater on elastic foundation under
the action of impulsive breaking wave pressure, Nept. Port and Harbor
Nes. lost., Vol.12, No.3, pp.3-29 (in Japanese ).
Goda, Y. [1973b]:' A new method of wave pressure calculation for the design of
19-21

YOSHIMI GODA

568

composite breakwater, Nept. Port and Harbar Nes. Inst., Vol. 12, No. 3,
pp.31-70 (in Japanese ).
Goda, Y. [1974J: New wave pressure formulae for composite breakwater, Proc.
14th Int. Con~ Coastal Eng., pp.1702-1720.
Goda, Y. [1975J: lrregular wave deformation in the surf zone, Coastal Epgineering in Japan, Vol.18, pp.13-26.
Goda, Y. [1985J: NandolRSees and Design of KaritiJle Structures,
University of
Tokyo Press., pp.l08-110.
Hiroi, 1. [1919J: On a method of estimating the force of waves, KeJloirs of
Engg. Faculty, Imperial University of Tokyo, Vol. X, No.l, p.19.
lto, Y., Fujishima, H., ahd Kitatani, T. [1966J: On the stability of breakwaters, Nept. Port and Harbar Nes. Inst.,
Vol. 5, No. 14, pp.I-134 (in
Jspenese ).

Kobayashi, H., Terashi, H., and Takahashi, K. [1987J: Bearing capacity of a

rubble mound supporting a gravity structure, Nept. Port and Harbar Nes.
Inst., Vol.26, No.5, pp.215-252.
in ikin, R. R. [1950J: frinds, waves and Karitime Structures,
Griffin, London,
pp.38-39.
Hagai, S. [1968J: Pressures of partial standing waves, J. fraterways and Harbars Div., Proc. ASCE, Vol.94, No.WW3, pp.273-284.
Hagai, S. [1969J: Pressures of standing waves on a vertical wall, J. fraterways
end Harbars Div., Proc. ASCE, Vol.95, No.WW1, pp.53-76.
Hagai, S. and Otsubo, T. [1968J: Pressure by breaking waves on composite
breakwaters, Proc. llth Int. Conf. Coastal Engg, pp.920-933.
Sainflou, G. [1928J: Essai sur les digues aar i times, verticales, Annales Pants
.
et Chausses, Vol.98, Ho.4.
Shuto, N. [1974J: Honl inear long waves in a channel of variabIe section, Coastal Engineering in Japan, Vol.17, pp.I-14.
Tanimoto, K., Yagyu, T., and Goda, Y. [1982J: lrregular wave tests for composite breakwater foundation, Proc. 18th Int. Conf. Coastal Engg., pp.21442163.
Tanimoto, K., Takahashi, S., and Kimura, K. [1987J: Structures and hydraulic
characteristics of breakwaters - The state of arts of breakwater design
in Japan, Rept. Port and Harbor Res. lnst., Vol.26, No.5, pp.II-55 ..
Tani.oto, K. and Goda, Y. [1991aJ: Historical development of breakwater structures in the world, Proc. Conf. on CoastaJ Structures
end Breekws ters,
Inst. Civil Engrs., pp.153-166.
Tanimoto, K. and Goda, Y. [1991bJ: Stability of deepwater caisson breakwater
against random waves, Proc. Con~ on Coastal Structures and Breakwaters,
Inst. Civil Engrs., pp.181-193.

19-22