STABILITY O F MOUND BREAKWATER'S H E A D

AND TRUNK
By Cesar Vidal, 1 Miguel A. Losada, 2 Member, ASCE, and Raul Medina 3

ABSTRACT: The stability of a cubic armored breakwater head and trunk under
Downloaded from ascelibrary.org by Universidad de Cantabria on 06/02/15. Copyright ASCE. For personal use only; all rights reserved.

monochromatic wave trains is experimentally studied. For the head there is a sector,
roughly of 60°, where the unit stability is minimum. The stability function values
(similar to Hudson's number) are 1.3-three times higher there than those obtained
from biodimensional tested sections. Head shape, at least for the cases tested, does
not exert significant influence on the head stability. Trunk sections tested under
long crested waves with a standing longitudinal wave height variation (here called
quasi-three-dimensional [3D] tests) are less stable than two-dimensional (2D) tested
sections. Furthermore, trunk and head sections show a different behavior as the
damage grows; the radio of initiation of damage wave height to destruction wave
heights is much lower for head than for trunk sections. Head sections are more
brittle than trunk sections. Additionally, exploration was carried out into standing
longitudinal variation of wave height along the breakwater induced by different
factors, such as the reflexion of the scattered waves on the lateral boundaries. Data
from prototype failures that have occurred in Spain suggest that these wave patterns
are worthy of further research.

INTRODUCTION

In the last decades many researchers have studied the stability of mound
breakwaters: see Bruun (1985) as a general reference. Experimentation in
wave flume was the most common way to address this problem. Recently,
Kobayashi et al. (1986), Kobayashi and Otta (1987), and Kobayashi and
Wurjanto (1989) introduced numerical computation to evaluate flow on the
porous slope and the stability of units under no interlocking conditions.
However, most of the research done, either experimental or numerical,
analyzes the stability of the section bidimensionally. Very few papers con-
sidered the stability under oblique incidence, Losada and Gimenez-Curto
(1982) or the stability of the head, Argershou et al. (1982), Jensen (1984),
and Benassai et al. (1984). Still today the rule of thumb enounced by Iri-
barren and Nogales (1964) and Bruun (1985) (to avoid future problems head
units should weigh 1.5-two times more than trunk units) is used worldwide.
This paper gives new experimental information about the stability of head
units as a function of the relative size of the head to the wave length and
of the head shape. The rule of thumb is made more precise, opening new
ways to safer and cheaper designs. Further, the paper shows experimental
results of trunk sections obtained from quasi-three-dimensional (3D) tests
and compares them with two-dimensional (2D) experimental results. The
differences will be explained later.
The paper is organized in the following way. First, the damage criteria
are discussed. After describing the experimental setup test, results on head
and quasi-3D sections are presented and commented on. Finally, in a long

2
'Asst. Prof., Univ. of Cantabria, Santander, 39005, Spain.
3
Prof., Univ. of Cantabria, Santander, 39005, Spain.
Res. Asst., Univ. of Cantabria, Santander, 39005, Spain.
Note. Discussion open until April 1, 1992. To extend the closing date one month,
a written request must be filed with the ASCE Manger of Journals. The manuscript
for this paper was submitted for review and possible publication on March 5, 1990.
This paper is part of the Journal of Waterway, Port, Coastal, and Ocean Engineering,
Vol. 117, No. 6, November/December, 1991. ©ASCE, ISSN 0733-950X/91/0006-
0570/$1.00 + $.15 per page. Paper No. 26340.

570

J. Waterway, Port, Coastal, Ocean Eng. 1991.117:570-587.

 discussion section (1) A qualitative description of the flow on the head and
- trunk section induced by incident waves is given; (2) stability results are
compared with previous data obtained in a wave flume (2D experimenta-
tion); (3) plausible sources of longitudinal variation of wave height along
the breakwater are explored; and (4) some prototype damage records are
analyzed. Conclusions are highlighted with some useful design recommen-
dations.
Downloaded from ascelibrary.org by Universidad de Cantabria on 06/02/15. Copyright ASCE. For personal use only; all rights reserved.

DAMAGE CRITERIA

This paper addresses only the hydrodynamic stability of cubic armor units.
Therefore, damage criteria given herein are only adequate to evaluate this
type of failure. Other types of failure, as described by Bruun (1985), are
here disregarded. Losada et al. (1986) defined three hydrodynamic damage
criteria: (1) incipient damage; (2) Iribarren's damage; and (3) destruction.
In order to be more precise, another stage of damage is included; it is called
incipient destruction.
For a mound breakwater consisting of a core, a secondary layer and a
main layer with a lower and an upper layer of armor units, the definitions
of the four damage criteria are as follows:

1. Initiation of damage: This level of damage defines the condition attained

when a certain number of armor units are displaced from their original position
to a new one at a distance equal to or larger than a unit length. Holes larger
than average porous size are clearly appreciable. In this research we adopted
for this number a value of 2% of the displaced units required to achieve Iri-
barren's damage.
2. Iribarren's damage: This damage occurs when the extension of the failure
area on the main layer is so large that the wave action may extract armor units
placed on the lower armor layer.
3. Initiation of destruction: A small number of units, two or three, in the
lower armor layer are forced out and the waves work directly on pieces of the
secondary layer.
4. Destruction: Pieces of the secondary layer are removed. If the wave height
does not change the mound will definitely be destroyed and it will cease to give
the level of service defined in the design.

Comments
Damage criteria has always had a certain degree of subjectivism and some
of the aforementioned definitions may sound arbitrary, particularly initiation
of damage. Losada et al. (1986) showed for a given damage level that the
extraction of units may be considered as a Bernouilli experiment, each single
extraction being a binomial variable. If/? means probability of extraction
of a unit and n is the number of units then the scatter indicator has a value
= V ( l - p)l(pri). When p is small the indicator takes large values and
vice versa. In this notation, dispersion is an intrinsic property of the analyzed
phenomenon and it is therefore unavoidable. The criterion adopted here
for initiation of damage is a compromise between strictness and scatter.
Iribarren and Nogales (1964) gave the definition of Iribarren's damage
based on laboratory experimentation and field experience. They supposed
that if the damage on the upper armor units produces a hole big enough,
the lower armor units and the pieces of the secondary layer are less than
571

J. Waterway, Port, Coastal, Ocean Eng. 1991.117:570-587.

 the coefficients between the two layers of armor units, as shown by Bruun
and Johannesson (1974, 1976). Iribarren's damage is characterized by im-
portant displacements of armor units, causing a high compactness and de-
veloping an incipient s-profile. The mound remains stable as long as the
wave height does not grow.
Except for destruction, the definition of damage criterion assumes that
Downloaded from ascelibrary.org by Universidad de Cantabria on 06/02/15. Copyright ASCE. For personal use only; all rights reserved.

the breakwater stabilizes at the specified level of damage if the wave action
does not increase. However, damage criteria are only stages or points of
reference in a continuous process that do not give any information on the
way each level of damage is reached. In other words, they do not and cannot
give a continuous description of damage evolution. Furthermore, because
of the significant changes suffered by the breakwater at each damage level,
the damage evolution curve always passes through each defined level of
damage. At least this is the case for cubic armor units tested herein.

EXPERIMENTATION

In order to analyze the stability of trunk and head sections, experiments

were performed in a wave tank under monochromatic wave trains. For head
sections the dependence of damage on Iribarren's number, Ir (which is
defined later), on head shape and on section location were studied. The
longitudinal variation of damage was determined. Further, section stability
under quasi-3D tests was compared with section stability under bidimen-
sional tests (Losada et al. 1986).

Experimental Device and Technique

Tables 1,2, and 3 define the characteristics of the tested models, armor
units, and water waves, respectively. Fig. 1 shows the definition of the trunk
sections and head sectors and a typical cross section. Also this figure shows
two head shapes, called model I, Fig. 1(a), and model II, Fig. 1(b), re-
spectively. Model I is a symmetric head, whereas the model II head is turned
offshore. Both models are truncated cones and have the same radius, R. In
order to identify the damage, cone sectors on the head and sections of 50-

TABLE 1. Characteristics of Model: Armor Units

Slope a Weight P, Units

Type Shape (cotan) (cm) (g) (g/cm3) (m (2))
(1) (2) (3) (4) (5) (6) (7)
Cubes Axaxa 2.0 3.21 70 2.12 1,250

TABLE 2. Characteristics of Models: Cross Section

Thick- Grain Mean

ness size weight
Layer (cm) Material (D50 cm) (g)
(1) (2) (3) (4) (5)
Armor 6.5 Concrete units — 70
Underlayer 15.8 Limestone gravel 1.5 8
Core — Almost impermeable — —

572

I
J. Waterway, Port, Coastal, Ocean Eng. 1991.117:570-587.
 TABLE 3. Characteristics of Applied Waves
Characteristic Description
(1) (2)
Wavemaker type Piston, horizontal displacement
Type of waves Regular, long crested
Downloaded from ascelibrary.org by Universidad de Cantabria on 06/02/15. Copyright ASCE. For personal use only; all rights reserved.

Water depth (h) 0.35 m

Wave height (H) 6-16.8 cm
Wave period (T) 1.4, 1.6, 1.8 s
Wave steepness {HIL„) 0.0693-0.0484
Relative depth (hIL) 0.1133-0.1535
Ratio (h/H) 2.08-5.83
Ir 1.9-3.1 (damage range)

FIG. 1. Definition of Trunk Sections and Head Sectors for Models I and II

573

J. Waterway, Port, Coastal, Ocean Eng. 1991.117:570-587.

 cm wide on the trunk are labeled with numbers. Finally, the tested break-
waters are bounded at the nearshore end by the side wall of the wave tank.
Tests under such conditions are here called quasi-3D tests, because of the
standing longitudinal wave height variation of the incident wave trains.
Each test was performed with fixed water depth and wave period. Trunk
and head sections were completely rebuilt before each test. Slight increases
Downloaded from ascelibrary.org by Universidad de Cantabria on 06/02/15. Copyright ASCE. For personal use only; all rights reserved.

in the wave height defined test stages. At each stage the wave height was
held until an equilibrium was achieved. In order to avoid rereflexions of
waves at the paddle, wave trains were generated in cycles with a defined
number of waves per cycle depending on the wave length. Wave charac-
teristics were recorded by capacitance wave gauges.

Stability Function
Using dimensional analysis, Losada and Gimenez-Curto (1979) proposed
the following relationship for evaluating W (the weight of an armor unit as
a function of the incident wave height) H, and the specific weights of water
and armor units, 7,,, and ys, respectively

W = yw.Q.H3.Y (la)

2 = {W)
( ^

Ss^^ (lc)

where Y = a dimensionless function called stability function, which, for a

trunk section, depends on Iribarren's number, slope angle, a, damage cri-
terion, type of armor unit, unit placement, and roughness and permeability.
For a head sector it will be shown that Y also depends on RIL (where L =
the wave length and R the radius of the head cone), the head shape, and
the section position on the head. Notice that Iribarren's number (also called
surf similarity parameter) is defined by

Ir = ^ (2)

EXPERIMENTAL RESULTS

Experimental results of the head structure and trunk section stability are
given in Tables 4 and 5. Also some comments about the general features
of these results are made.

Head Structure
Fig. 2-5 show experimental results of the stability of head sections, model
I and model II, for initiation of damage, Iribarren's damage, initiation of
destruction, and destruction, respectively. Abscissa represents Iribarren's
number; the vertical axis gives the values of the stability function calculated
from (1). In each figure three sets of date are presented:
574

J. Waterway, Port, Coastal, Ocean Eng. 1991.117:570-587.

 TABLE 4. Model I

Period Initiation of Iribarren's Initiation of

Section (s) damage damage destruction Destruction
(1) (2) (3) (4) (5) (6)
1 1.4 8.6 9.7 >14.8 >14.8
Downloaded from ascelibrary.org by Universidad de Cantabria on 06/02/15. Copyright ASCE. For personal use only; all rights reserved.

1 1.6 10.3 11.9 15.1 >15.9

1 1.8 13.2 14.1 >16.3 >16.3
1 1.4 11.5 >12.8 >12.8 >12.8
2 1.4 8.6 10.3 >14.8 >14.8
2 1.6 11.3 11.9 15.1 >15.9
2 1.8 12.4 14.1 >16.3 >16.3
2 1.4 12.3 >12.8 >12.8 >12.8
3 1.4 8.6 10.3 >14.8 >14.8
3 1.6 12.7 13.4 >15.9 >15.9
3 1.8 14.1 15.3 >16.3 >16.3
3 1.4 10.9 >12.8 >12.8 >12.8
4 1.4 >14.8 >14.8 >14.8 >14.8
4 1.6 12.7 14.5 >15.9 >15.9
4 1.8 11.8 >16.3 >16.3 >16.3
4 1.4 12.3 >12.8 >12.8 >12.8
5 1.4 8.6 11.3 >14.8 >14.8
5 1.6 >15.9 >15.9 >15.9 >15.9
5 1.8 11.3 12.4 >16.3 >16.3
5 1.4 >12.8 >12.8 >12.8 >12.8
6 1.4 8.3 10.3 12.9 14.8
6 1.6 12.2 12.7 15.5 15.9
6 1.8 9.8 10.3 16.0 16.3
6 1.4 8.3 10.9 12.3 12.8
7 1.4 8.2 8.6 12.9 14.8
7 1.6 11.3 11.9 13.4 15.9
7 1.8 9.8 10.3 11.1 16.3
7 1.4 10.5 10.9 12.3 12.8
8 1.4 8.2 8.6 11.8 14.8
8 1.6 11.3 11.9 13.4 15.9
8 1.8 10.3 11.1 11.3 16.3
8 1.4 11.2 11.5 >12.8 >12.8
9 1.4 >14.8 >14.8 >14.8 >14.8
9 1.6 >15.9 >15.9 >15.9 >15.9
9 1.8 >16.3 >16.3 >16.3 >16.3
9 1.4 >12.8 >12.8 >12.8 >12.8

• The first set of data (indicated by H in the figures) shows the values of
Yfor the worst head sectors, i.e., sectors 6, 7, and 8 for model I and
sections 9, 10, and 11 for model II.
• The second set of data (indicated by T in the figures) shows the values
of Y for the trunk sections 1,2, and 3 for both models (these data will
be discussed later.
• The third set of data (indicated by Z in the figures) shows the values
Yfor the remainder head sectors, i.e., sectors 4, 5, and 9 for model I
and sectors 4-8 for model II.
575

J. Waterway, Port, Coastal, Ocean Eng. 1991.117:570-587.

 TABLE 5. Model II

Period Initiation of Iribarren's Initiation of

Section (s) damage damage damage Destruction
(1) (2) (3) (4) (5) (6)
1 1.4 10.4 11.0 >12.4 >12.4
Downloaded from ascelibrary.org by Universidad de Cantabria on 06/02/15. Copyright ASCE. For personal use only; all rights reserved.

1 1.6 10.1 11.3 14.4 15.7

1 1.8 15.1 15.4 16.4 16.8
2 1.4 10.4 11.0 >12.4 >12.4
2 1.6 10.1 11.9 14.4 15.7
2 1.8 15.1 16.4 16.4 16.8
3 1.4 10.4 12.4 >12.4 >12.4
3 1.6 13.8 13.8 >15.7 >15.7
3 1.8 13.0 15.1 16.4 16.8
4 1.4 12.0 >12.4 >12.4 >12.4
4 1.6 14.4 >15.7 >15.7 >15.7
4 1.8 12.4 15.1 15.7 16.8
5 1.4 11.2 11.5 12.2 12.4
5 1.6 11.3 13.8 >15.7 >15.7
5 1.8 12.4 14.1 15.4 15.7
6 1.4 10.4 11.8 11.8 12.0
6 1.6 11.9 13.8 14.8 15.1
6 1.8 12.4 14.1 15.2 15.4
7 1.4 10.4 11.0 11.8 12.0
7 1.6 10.1 13.0 13.5 13.8
7 1.8 13.0 13.6 15.2 15.4
8 1.4 10.2 10.4 11.8 12.0
8 1.6 13.5 13.5 13.5 13.8
8 1.8 12.8 13.0 15.2 15.4
9 1.4 9.7 9.9 10.8 11.0
9 1.6 11.9 12.2 12.2 12.4
9 1.8 12.8 13.0 15.1 15.1
10 1.4 9.2 9.7 9.9 10.4
10 1.6 8.0 8.6 9.5 10.7
10 1.8 12.2 12.4 14.1 15.1
11 1.4 9.7 9.9 10.8 11.0
11 1.6 8.0 8.6 10.1 10.7
11 1.8 10.9 12.2 12.4 15.1

From these figures some information about the behavior of the head
structure can be obtained:

1. Brittleness: It is worth noting the proximity of the stability curves for

different levels of damage. This is an indication of the brittleness of the head
sectors. Once the damage starts in one cone sector, the evolution of damage
up to destruction is achieved under a few small increases in wave height.
2. Critical stability sector: A cone sector of roughly 60° defined from the
normal to the wave ray tangent to the head toe (see Fig. 6) is less stable than
other head sectors. Jensen (1984) reported similar results for irregular waves.
Furthermore, not only does the damage always initiate there, but the destruction
does too.
576

J. Waterway, Port, Coastal, Ocean Eng. 1991.117:570-587.

 0.10
INITIATION OF DAMAGE

O.OB -
H WORST HEAD SECTORS
T TRUNK SECTIONS
Z REMAINDER HEAD SEC.
O.OB -
Downloaded from ascelibrary.org by Universidad de Cantabria on 06/02/15. Copyright ASCE. For personal use only; all rights reserved.

-MAX. HEAD VAL.

MAX. TRUNK V A L .
MAX. 2D VALUES
(LOSADA ET
Z< H" ZH AL 1986]
0.02 T T
2 r

0.00
2.0 2.E 2.4 2.6 2.8 3.0
Ir
FIG. 2. Stability Function Values as Function of Ir for Initiation of Destruction (/)

o.io
IHIBAHREN'S DAMAGE

o.oa
H WORST HEAD SECTORS
T TRUNK SECTIONS
Z REMAINDER HEAD SEC.

MAX. HEAD VAL.

0.04 - MAX. TRUNK V A L .
MAX. 2D VALUES
(LOSADA ET
AL 1986]
0.02

FIG. 3. Stability Function Values as Function of //' for Iribarren's Damage (IR)

3. Damage: Damage on upstream sectors is always damage produced by the

upstream propagation of the damage that occurred on the critical stability sector.
4. Stability function values: Comparing the maximum values of the stability
function obtained for the critical head sectors with those obtained from 2D tests,
head values 1.3-3.0 times higher are obtained, depending on the damage level.
5. Head shape: There are no significant differences in the stability of model
I and model II. Therefore the shape of the head should be defined by other
factors, mainly constructional or economic. Losada et al. (1990) showed that
there are no reasons from a harbor agitation point of view for building large
head structures.

Trunk Sections
As mentioned before, Figs. 2-5 include the stability function values ob-
tained from quasi-3D experimentation and a horizontal line that represents
577

J. Waterway, Port, Coastal, Ocean Eng. 1991.117:570-587.

 INIT. OF DESTRUCTION

o.oa- H WORST HEAD SECTORS

T TRUNK SECTIONS
Z REMAINDER HEAD SEC.
Downloaded from ascelibrary.org by Universidad de Cantabria on 06/02/15. Copyright ASCE. For personal use only; all rights reserved.

.MAX. HEAD VAL.

0.04 MAX. TRUNK VAL.
MAX. 2D VALUES
(LOSADA ET
If AL 198BI
H
0.02 *

0.00 -| , 1 , 1 1 1 1 r
2.0 2.2 2.4 2.6 2.8 3.0
Ir

FIG. 4. Stability Function Values as Function of Ir for Initiation of Destruction

(ID)

0.10

DESTRUCTION

o.oa- H WORST HEAD SECTORS

T TRUNK SECTIONS
Z REMAINDER HEAD SEC.
0.06

MAX. HEAD VAL.

0.04 MAX. TRUNK VAL.

0.02-

0.00

FIG. 5. Stability Function Values as Function of Ir for Destruction (D)

their maximum. The measured longitudinal variation or antinodal crest

height in front of the breakwater was roughly 10%. However, as usual, the
incident wave height value, measured far away from the influence of the
breakwater, was used in the computation of the stability function. Also,
figures include another horizontal line that represents the maximum value
of the 2D test results by Losada et al. (1986). From these figures the fol-
lowing features of quasi-3D tested sections may be deduced.

1. Initiation of damage and Iribarren's damage. Because of the local higher

waves, some 3D tested sections start to fail earlier than normal 2D sections.
2. Initiation of destruction. At this stage of failure trunk sections, 2D tested
sections and 3D tested sections have similar responses.
578

J. Waterway, Port, Coastal, Ocean Eng. 1991.117:570-587.

 NCI DENT WAVE
Downloaded from ascelibrary.org by Universidad de Cantabria on 06/02/15. Copyright ASCE. For personal use only; all rights reserved.

HEAD AREA WHERE DAMAGE STARTS

FIG. 6. Critical Stability Sector Definition

3. Destruction. 2D test sections reach this level of damage with lower waves
than 3D tested sections.

An Overall View
Next, Fig. 7 gives the range of variation of the stability function, Y,
obtained from Figs. 2-5. Maximum, minimum, and mean values are in-
cluded. Also, the maximum values given by Losada et al. (1986) are pre-
sented. From this figure some information about the stability function that
may be useful for breakwater design can be summarized:

1. Larger ranges of variation are encountered in the worst head sectors.

2. Maximum, minimum, and mean stability function values of the worst head
sectors are for all types of damage higher than those of the remainder head
sectors and trunk sections.
3. Scatter decreases with the evolution of damage.

ANALYSIS

The distinct response of head and breakwater sections obtained from the
experimentation under monochromatic waves merits some review of the
behavior of the waves while propagating on the head and on the trunk.
Furthermore, it would be interesting to conduct further analysis to establish
whether failures that occurred in prototypes show similar results.
In the following, the propagation of waves on a head and along a finite
breakwater is briefly reviewed. Further, field data of damage recorded at
two Spanish breakwaters, Bilbao and San Ciprian, Figs. 8(a) and 9(a) are
presented. It is suggested that the variation of the damage occurring on
laterally bounded breakwaters along the trunk may be induced by a standing
longitudinal variation of the incident wave height along the breakwater.
579

J. Waterway, Port, Coastal, Ocean Eng. 1991.117:570-587.

 Xa WORST HEAD SECTORS
m REMAINDER HEAD SECTORS
m 3D TRUNK SECTIONS
S 2D TESTS (LOSADA 1986)
0.08

^MEAN VALUE OF Y

I1
Downloaded from ascelibrary.org by Universidad de Cantabria on 06/02/15. Copyright ASCE. For personal use only; all rights reserved.

0.06

0.04 .
2
11
0.02 1

IR
1 ID
DAMAGE
D

FIG. 7. Summary of Stability Function Values

Waves on a Head and Related Unit Stability

A breakwater head refracts, diffracts, and shoals waves capturing a sig-
nificant amount of wave energy. Because of refraction, waves focus over
the submerged rounded head and eventually may break forming a jet (Pal-
mer 1960). Dalrymple et al. (1986) showed that a rounded head causes the
refraction of wave energy onto the leeside, forcing the waves to be trapped
there as edge wave modes. Very mild slopes capture all the waves that
propagate around submerged islands (Pocinki 1950), whereas steeperhead
slopes focus smaller amounts of wave energy. In either case, the wave height
varies along the head.
For nonbreaking waves, this variation is a function of head size and shape.
As the radius of the head increases the wave height around the head in-
creases as well, achieving a maximum value near the normal to the wave
ray tangent to the head toe. Its precise position depends on the head radius
and on the head slope (Losada et al. 1990).
For breaking waves, the wave height variation along the head may be
obtained from the numerical and experimental study of the breaking of a
solitary wave on a submerged cylinder accomplished by Cooker et al. (1989).
Although a wave passing over a sloped head is not exactly the previously
described case because of the tridimensional effects on the head, mainly
wave focussing, both phenomena surely have flow similarities. Those au-
thors reported that the wave invariably breaks forwards beyond the cen-
terline of the semicircle. This is contrary to the general intuition; that one
expects a wave to break as the depth is decreasing. The distinct behavior
of waves passing an obstacle or a head, depends on the ratio of the radius
of the obstacle or of the slope length, S, to the wave length, L. For small
580

J. Waterway, Port, Coastal, Ocean Eng. 1991.117:570-587.

Downloaded from ascelibrary.org by Universidad de Cantabria on 06/02/15. Copyright ASCE. For personal use only; all rights reserved.

FIG. 8. (a) Location of Bilbao's Breakwater; and (b) Length and Orientation of
Bilbao's Breakwater

or moderate values of SIL, the waves, whether they break or not, pass the
head with forward motion. For large values of SIL, the breakwater head,
following Pocinki's model, captures most of the energy forcing the waves
to break into the head structure.
Since most prototype breakwaters have a head slope cotan a > 2.0; the
model tested breakwaters were tested with a 2:1 slope too. In all these test
cases damage was produced by forward breakers. Once the units are dis-
placed from their position, they move with forward motion because of the
combined action of flow and gravity. This behavior has two nondesirable
consequences: First, the units move off the sheltered area, possibly affecting
the navigation channel or moorings; and second, these units do not create
a seaside berm that protects the damage area as the displaced trunk units
do. Because of their distinct wave patterns, head sectors are more brittle
than trunk sections.
581

J. Waterway, Port, Coastal, Ocean Eng. 1991.117:570-587.

Downloaded from ascelibrary.org by Universidad de Cantabria on 06/02/15. Copyright ASCE. For personal use only; all rights reserved.

FIG. 9. (a) Location of San Ciprian's Breakwater; and (b) Length and Orientation
of San Ciprian's Breakwater

Waves on the Trunk and Related Unit Stability

Bilbao's breakwater on the Spanish coast of the Gulf of Biscay suffered
intensive damage during two storms in March 1976 and December 1976.
Length and orientation of the breakwater are shown in Fig. 8(b). Torum
et al. (1979) reported damage on the 65-ton block breakwater obtained from
profiles recorded at 860-2,100 m from the nearshore end of the breakwater.
Fig. 10 shows the distribution of the damage percentage along the break-
water. Damage was defined as "the number of armor elements moved from
their original position during the storm in relation to the total number of
elements in the test section."
San Ciprian's breakwater, also on the coast of Spain in the Gulf of Biscay,
was built with dolos of 50 tons. Fig. 9(b) shows length and orientation of
the breakwater. It suffered general damage during successive storms that
582

J. Waterway, Port, Coastal, Ocean Eng. 1991.117:570-587.

Downloaded from ascelibrary.org by Universidad de Cantabria on 06/02/15. Copyright ASCE. For personal use only; all rights reserved.

400 900 1400 1900 2400

DISTANCE FROM THE NEARSHORE END IN M.
FIG. 10. Distribution of Damage along Bilbao's Breakwater (after Torum et al.
1979)

i i i i i i t i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i
200 400 BOO BOO
DISTANCE FROM THE NEARSHORE END I N M.
FIG. 11. Number of Broken Dolos along San Ciprian's Breakwater

occurred after its completion in December 1977. Since then extensive repairs
have been undertaken. Because of the enormous cost, only the head was
repaired with cubes of 90 tons, while the trunk sections were repaired with
remaining dolos units. Since 1986, the owner, an aluminum factory, surveys
the state of the dolos units every year, recording the broken units and their
position on the slope. Fig. 11 shows in percentage the number of broken
dolos in relation to the total number of dolos in the section along the
breakwater, after the field work finished in August 1989. Each point ac-
counts for the number of broken dolos between sections 50 m apart.
Both damage curves show an oscillating behavior, where there are alter-
nating sections almost without damage while other sections are almost com-
pletely destroyed. It was hypothesized that this damage distribution was
associated with a standing variation of wave height along the breakwater
trunk. To check this hypothesis two actions were taken. First, the wave field
in the vicinity of a breakwater with a nearshore end that is vertical and
583

J. Waterway, Port, Coastal, Ocean Eng. 1991.117:570-587.

 impermeable was numerically computed. Second, the stability of a break-
water with this specific disposition was tested in the wave tank. Numerical
results for an impermeable breakwater are given by Losada et al. (1990)
and they show an enhancement of the maximum wave amplitudes on trunk
sections spaced roughly a wave length apart. This enhancement is believed
to be a result of the reflexion of the scattered wave on the lateral wall.
Despite the few sections tested in the tank, experimental results under'
Downloaded from ascelibrary.org by Universidad de Cantabria on 06/02/15. Copyright ASCE. For personal use only; all rights reserved.

normal incident waves also show fixed sections where damage always occurs.
However, the reflexion of the scattered wave on the lateral boundary is
not the only way to have an enhancement of the wave height in some section
along the trunk. Wave refraction on uneven bottom is the most widely
known phenomenon. In the following, other possibilities for horizontal bot-
tom are explored.
Dalrymple et al. (1990) obtained the solution to the problem of waves
arriving obliquely on a vertical porous breakwater, extending the solution
given some years ago by Sollit and Cross (1972). The full solution has
evanescent modes built up to match the flow inside and outside the porous
structure. For infinite breakwater length, the evanescent modes propagate
along the breakwater with a wave length given by Ly = L*sin 0, where 9
is the wave angle incidence to the breakwater normal. However, for a
laterally bounded breakwater, the full solution contains standing evanescent
modes, which may enhance the wave height in some trunk sections. Notice
that because of their offshore decay, it is only necessary for the lateral
boundary to extend as far as the decay distance.
If the boundary extends farther off and waves break on the lateral wall
under oblique incidence, waves arrive at the breakwater with a standing
longitudinal variation, given by the following equations:
^ = e-Hkx-x + kyy-M) + CR.e-H,-kx-x + kry-a,t) (^o)

2TT
k = —- ; kx = k cos P; ky = k sin p (3b)

where kx, kv = the wave number components in the x- and y-direction,

respectively ^ where x is parallel to the breakwater and y is along the lateral
wall; p = the wave angle incidence to the normal to the lateral boundary;
w = the wave angular frequency; and CR = the reflexion coefficient at the
wall.
With regard to the Bilbao and San Ciprian breakwaters, it seems that
some of the described effects were due to the longitudinal variation of the
damage. In Bilbao these were probably evanescent modes, while in San
Ciprian the reflexion of the scattered waves could be the source of the
longitudinal variation of the number of broken dolos along the trunk.
Regardless of its origin, once the longitudinal variation of wave height
exists, the flow on the slope is primarily as in a beach cusp, capturing water
from the lateral and returning it through the center of the section. Under
such types of flow, displaced units are transported with the centered return
flow down to the toe, where a berm is built. The envelope of the berm and
of the damage area is an ^-profile. For earlier stages of damage and because
of the lateral contribution of units, quasi-3D tested sections stabilize faster
than bidimensional sections, but diffusing the damage along the trunk.

CONCLUSIONS AND RECOMMENDATIONS

From an experimentation under monochromatic waves the following con-

clusions on the stability of head and trunk sections were obtained.
584

J. Waterway, Port, Coastal, Ocean Eng. 1991.117:570-587.

 Head

1. There is a sector of 60° of minimum stability, see Fig. 6.

2. Depending on the level of damage, the weight of head units in this sector
must be 1.3-3.8 times higher than the weight of trunk units.
3. Because of the forward flow induced by breaking waves on the head,
Downloaded from ascelibrary.org by Universidad de Cantabria on 06/02/15. Copyright ASCE. For personal use only; all rights reserved.

displaced units do not create a protection berm. The head is more brittle than
the trunk.

It is recommendable to design the breakwater head reinforcing the sector

of 60° defined from the normal to the tangent wave ray to the head toe. If
more than one design wave direction exists, all of them must be considered;
then the maximum sector of all directions must be reinforced. The increment
of unit weight placed on this sector depends on the level of damage defined
in the design. The calculation of the unit weight must be performed with
the envelope value given in Figs. 2 - 5 , as justified by Losada et al. (1986).

Trunk

1. A lateral boundary such as a cliff or other breakwater may induce the

reflexion of the scattered wave, the reflexion of evanescent modes, or the re-
flexion of the main wave train. Under such conditions, standing longitudinal
variation in wave height along the breakwater is to be expected. This standing
longitudinal variation of wave height may enhance the possible energy concen-
tration induced by refraction.
2. A standing longitudinal variation of wave height induces a longitudinal
variation of the damage along the trunk.
3. Quasi-3D tested sections initiate damage earlier, but stabilize faster, than
2D tested sections. Further, it is harder to destroy a quasi-3D section than a
2D section.

It is recommended that the eventual sources of standing longitudinal

variation in the prototype be analyzed. Special attention should be paid to
the lateral boundaries in the laboratory, in order to avoid false-standing
longitudinal wave height variations. As for the head, unit weight must be
calculated by using the maximum value of the stability function.

ACKNOWLEDGMENTS

This study was funded by the M . O . P . U . , Direction General de Puertos

y Costas and by the D G I C Y T under contract No. PA85-076.

APPENDIX I. REFERENCES

Argershou, H., Lundgren, H., and Sorensen, T. (1982). Planning and design of ports
and marine terminals, John Wiley and Sons, New York, N.Y.
Benassai, E., D'Antonio, V., Grimaldi, F., and Van der Weide, J. (1984). "Study
of the transition from a rubble mound to a vertical wall breakwater." Int. Symp.
on Maritime Structures in the Mediterranean Sea, National Technical University of
Athens, 1.95-1.106.
Bruun, P. (1985). Design and construction of mounds for breakwaters and coastal
protection. Elsevier, Amsterdam, The Netherlands.
Bruun, P., and Johannesson, P. (1974). "A critical review of the hydraulics of rubble

585

J. Waterway, Port, Coastal, Ocean Eng. 1991.117:570-587.

 mound structures." Rept. No. R3-1974, Div. of Port and Ocean Engrs., Norwegian
Inst, of Tech. Trondheim, Norway.
Bruun, P., and Johannesson, P. (1977). "Parameters affecting the stability of rubble
mounds." /. Wtrwy., Harb., and Coast. Engrg. Div., ASCE, 103, 141-164.
Cooker, M. J., Peregrine, D. H., Vidal, C , andDold, J. W. (1989). "The interaction
between a solitary wave and a submerged semi-circular cylinder." /. Fluid Mech.,
215(Jun.), 1-22.
Downloaded from ascelibrary.org by Universidad de Cantabria on 06/02/15. Copyright ASCE. For personal use only; all rights reserved.

Dalrymple, R. A., Losada, M. A., and Martin, P. (1990). Reflection and transmission
from porous structures under oblique wave attack. Ctr. for Appl. Coastal Res.,
Dept. of Civ. Engrg., Univ. of Delaware, Newark, Del.
Dalrymple, R. A., Kirby, J. T., and Seli, D. J. (1986). "Wave trapping by break-
waters." 20th Int. Coastal Engrg. Conference, ASCE, 1820-1830.
Iribarren, R., and Nogales, C. (1964). Obras maritimas. S. A. Dossat, ed., Madrid,
Espafia. (in Spanish).
Jensen, O. J. (1984). A monograph on rubble mound breakwaters. Danish Hydr.
Inst., Denmark.
Kobayashi, N., Roy, I., and Otta, A. K. (1986). "Numerical simulation of wave
run-up and armor stability."OrePaper 5088, 18th Offshore Tech. Conf., 51-56.
Kobayashi, N., and Otta, A. K. (1987). "Hydraulic stability analysis of armor units."
J. of Wtrwy., Port, Coast, and O. Engrg., ASCE, 113(2), 171-186.
Kobayashi, N., and Wurjanto, A. (1989). "Armor stability on rough permeable
slopes of marine structures." XXIII Congress, IAHR, Aug., Ottawa, Canada, C,
407-414.
Losada, M. A., and Gimenez-Curto, L. A. (1979). "The joint effects of wave height
and period on the stability of rubble mound breakwaters using Iribarren's number."
Coastal Engrg., 3(Dec), 77-96.
Losada, M. A., and Gimenez-Curto, L. A. (1982). "Mound breakwaters under
oblique wave attack: A working hypothesis." Coastal Engrg., 6, 83-92.
Losada, M. A., Desire, J. M., and Alejo, L. M (1986). "Stability of blocks as
breakwater armor units." /. Struc. Engrg., ASCE, 112(11), 2392-2401.
Losada, M. A., Dalrymple, R. A., and Vidal, C. (1990). "Water waves in the vicinity
of breakwaters."/. Coastal Res., SI-7, Spring, 119-138.
Palmer, R. Q. (1960). "Breakwaters in the Hawaiian Islands." /. Wtrwy. and Harbor
Div., ASCE, 39-67.
Pocinki, L. S. (1950). "The application of conformal tansformations to ocean wave
refraction problems." Trans. Amer. Geophys. Un., 31(6) 856-866.
Sollitt, C. K., and Cross, R. H. (1972). "Wave transmission through permeable
breakwaters." 13th Coastal Engrg. Conference, ASCE, 1827-1846.
Torum, A., Mathiesen, B. J., and Escutia, R. (1979). "Reliability of breakwater
model tests." Coastal Structures '79, ASCE, 454-469.

APPENDIX II. NOTATION

The following symbols are used in this paper:

a = armor side length;

cHR = reflexion coefficient at wall;
= incident wave height;
h = depth;
Ir = Iribarren's number;
i = V-l;
k = wave number;
kx> ky = wave number components in x- and _y-direction;
L0 = wave length in deep water;
L = wave length;
n = number of units;
P = probability density;
586

«
J. Waterway, Port, Coastal, Ocean Eng. 1991.117:570-587.
 R = radius of head cone;
• S = slope length;
W = weight of armor unit;
Y = stability function;
a = slope angle;
P = wave incidence angle with normal to lateral boundary;
Downloaded from ascelibrary.org by Universidad de Cantabria on 06/02/15. Copyright ASCE. For personal use only; all rights reserved.

7„, = specific weight of water;

ys = specific weight of units;
T| = vertical water displacement;
0 = wave incidence angle with respect to breakwater normal;
ps = density of armor unit.

587

J. Waterway, Port, Coastal, Ocean Eng. 1991.117:570-587.