You are on page 1of 17

11 Composite/Vertical Wall Breakwater Design

Ref:

Shore Protection Manual, USACE, 1984


Basic Coastal Engineering, R.M. Sorensen, 1997
Coastal Engineering Handbook, J.B. Herbich, 1991
EM 1110-2-2904, Design of Breakwaters and Jetties, USACE, 1986
Breakwaters, Jetties, Bulkheads and Seawalls, Pile Buck, 1992
Coastal, Estuarial and Harbour Engineers' Reference Book, M.B. Abbot and W.A. Price,
1994, (Chapter 29)
Coastal Engineering, K. Horikawa, 1978
Coastal Engineering Manual (VI-5), USACE, 2003

Topics
Composite/Vertical Wall Breakwater Design
Wave Force Calculations
Caisson Width
Sliding and Overturning Stability
Soil Bearing Capacity Calculations
Summary of Design Procedure
--------------------------------------------------------------------------------------------------------------------Composite/Vertical Wall Breakwater Design
Wave Force Calculations
A characteristic of vertical wall breakwaters is that the kinetic energy of the wave is stopped
suddenly at the wall face. The energy is then reflected or translated by vertical motion of the
water along the wall face. The upward component of this can cause the wave crests to rise to
double their deep water height (non-breaking case). The downward component causes very high
velocities at the base of the wall and horizontally away from the wall for of a wavelength, thus
causing erosion and scour.
Many analytical and laboratory studies and field observations have been undertaken to
understand the wave pressure and develop wave pressure formulas. However, most of the
formulas are based on monochromatic regular wave of constant height and period.
Wave-generated pressures on structures are complicated functions of the wave conditions and
geometry of the structure. For this reason laboratory model tests should be performed as part of
the final design of important structures. For preliminary designs the formulae presented in this
section can be used within the stated parameter limitations and with consideration of the
uncertainties.
Two-dimensional wave forces on vertical walls.
Non-breaking waves incident on smooth, impermeable vertical walls are completely
reflected by the wall giving a reflection coefficient of 1.0. Where wales, tiebacks, or other
structural elements increase the wall surface roughness and retard the vertical water
motion at the wall, the reflection coefficient will be slightly reduced. Vertical walls built
on rubble bases will also have a reduced reflection coefficient.

The total hydrodynamic pressure distribution on a vertical wall consists of two timevarying components: the hydrostatic pressure component due to the instantaneous water
depth at the wall, and the dynamic pressure component due to the accelerations of the
water particles. Over a wave cycle, the force found from integrating the pressure
distribution on the wall varies between a minimum value when a wave trough is at the
wall to a maximum values when a wave crest is at the wall as illustrated below for the
case of non-overtopped walls or caissons.
Pressure distributions for a non-breaking wave

The resulting total hydrodynamic load when the wave trough is at the vertical wall is less
than the hydrostatic loading if waves were not present and the water was at rest. For
bulkheads and seawalls this may be a critical design loading because saturated backfill
soils could cause the wall to fail in the seaward direction. Therefore, water level is a
crucial design parameter for calculating forces and moments on vertical walls.
Pressure distribution on an overtopped wall

Wave overtopping of vertical walls provides a reduction in the total force and moment
because the pressure distribution is truncated as shown schematically above. Engineers
should consider the effect overtopping might have on land-based vertical structures by
creating seaward pressure on the wall caused by saturated backfill or ponding water.

Three Types of Wave Forces on Vertical Walls


1) Non-breaking Waves
2) Breaking (plunging) waves with almost vertical fronts
3) Breaking (plunging) waves with large air pockets

1) Non-breaking waves: Waves do not trap an air pocket against the wall. The pressure at the
wall has a gentle variation in time and is almost in phase with the wave elevation. Wave
loads of this type are called pulsating or quasi-static loads because the period is much larger
than the natural period of oscillation of the structures. (For conventional caisson breakwaters
the period is approximately one order of magnitude larger.) Consequently, the wave load can
be treated like a static load in stability calculations. Special considerations are required if the
caisson is placed on fine soils where pore pressure may build up, resulting in significant
weakening of the soil.
Non-Breaking Waves - assumes forces are essentially hydrostatic
Linear Wave Theory
standing wave (known as the "clapotis")
total reflection crest to trough excursion of the water surface = 2H

amplitude of the dynamic pressure, p d =

at the bottom (z = -h) p d =

H
cosh kh

H cosh k (h + z )
, z = 0 at surface;
cosh kh

2) Breaking (plunging) waves with almost vertical fronts: Waves that break in a plunging mode
develop an almost vertical front before they curl over (see Figure VI-5-57b). If this almost
vertical front occurs just prior to the contact with the wall, then very high pressures are
generated having extremely short durations. Only a negligible amount of air is entrapped,
resulting in a very large single peaked force followed by very small force oscillations. The
duration of the pressure peak is on the order of hundredths of a second.
3) Breaking (plunging) waves with large air pockets: If a large amount of air is entrapped in a
pocket, a double peaked force is produced followed by pronounced force oscillations as
shown in Figure VI-5-57c. The first and largest peak is induced by the wave crest hitting the
structure at point A, and it is similar to a hammer shock. The second peak is induced by the
subsequent maximum compression of the air pocket at point B, and is it is referred to as
compression shock, (Lundgren 1969). In the literature this wave loading is often called the
Bagnold type. The force oscillations are due to the pulsation of the air pocket. The double
peaks have typical spacing in the range of milliseconds to hundredths of a second. The period
of the force oscillations is in the range 0.2-1.0 sec.

Formula
Sainflou formula (modified by
Miche-Rundgen, 1958)

Wave Type
Standing

Structure Type
Impermeable vertical wall

CEM Table
VI-5-52

Goda formula
Goda formula (modified by
Takahashi, Tanimoto, and
Shimosako 1994)
Goda formula forces and
moments
Goda formula (modifed by
Tanimoto and Kimura 1985)
Goda formula (modified by
Takahashi and Hosoyamada
1994)
Goda formula (modified by
Takahashi, Tanimoto, and
Shimosako 1990)
Goda formula (modifed by
Takahashi, Tanimoto, and
Shimosako 1994)

2-D oblique
Provoked
breaking

Impermeable vertical wall


Impermeable vertical wall

VI-5-53
VI-5-54

Provoked
breaking
2-D head-on

Impermeable vertical wall

VI-5-55

Impermeable inclined wall

VI-5-56

2-D head-on

Impermeable sloping top

VI-5-57

2-D head-on

Horizontal composite structure

VI-5-58

3-D head-on

Vertical slit wall

VI-5-59

CEM Table VI-5-52 through VI-5-59 provide formulae for estimating pressure distributions
and corresponding forces and overturning moments on vertical walls due to non-breaking and
breaking waves.
o Wave pressure distributions for breaking waves are estimated using Table VI-5-54,
o corresponding forces and moments are calculated from Table VI-5-55.

Minikin's Method:
o Older breaking wave forces method of Minikin (Shore Protection Manual, 1984)
o can result in very high estimates of wave force, as much as 15 to 18 times those
calculated for non-breaking waves.
o These estimates are too conservative in most cases and could result in costly
structures. There may be rare circumstances where waves could break in just the right
manner to create very high impulsive loads of short duration, and these cases may not
be covered by the range of experiment parameters used to develop the guidance given
in Table VI-5-54.
o In addition, scaled laboratory models do not correctly reproduce the force loading
where pockets of air are trapped between the wave and wall (CEM Figure VI-5-57).
o For these reasons, it may be advisable to design vertical-front structures serving
critical functions according to Minikin's method.
Most of the methodology is based on the method presented by Goda (1974) and extended by
others to cover a variety of conditions. These formulae provide a unified design approach to
estimating design loads on vertical walls and caissons.

NOTE: All of the methods calculate the pressure distribution and resulting forces and
moments for only the wave portion of the hydrodynamic loading. The hydrostatic pressure
distribution from the SWL to the bottom is excluded.

o For a caisson structure (with water on both sides), the SWL hydrostatic forces would
exactly cancel (i.e. hydrostatic pressure on the seaside would be opposed by the
pressure on the lee-side);
o however, it will be necessary to include the effect of the SWL hydrodynamic pressure
for vertical walls tied into the shoreline or an embankment.
Non-Breaking Waves:
Sainflou's Formula (1928) modified by Miche-Rundgren (1944, 1958)
(CEM Table VI-5-52, p. VI-5-138)
Derived theoretically for regular, non-breaking waves and a vertical wall, but may be
applied to irregular waves
Uses 2nd order wave theory
Assumes linear depth-dependent pressure distribution below the water line (assumes
force is essentially hydrostatic)
cannot be used for breaking waves or overtopping

p1 = (p 2 + w h s )

p2 =

H + o
h s + H + o

wH
cosh kh s

p3 = w (H o )

Radiation stress considerations show the reflected wave causes a set-up (o) at the
vertical wall
H 2
2
o =
coth kh s , k =
L
L

Simplified formula assumes a linear pressure distribution below the water level
(conservative assumption, see reflected wave diagram)

Increase in pressure due to the standing wave:


1+ wH
p1 =

2 cosh kh s
where = wave reflection coefficient (1.0 for vertical wall with total reflection)

Pressure for calculating uplift force is p2

Breaking Waves:
Goda (1974) (CEM Table VI-5-53, p. VI-5-139)
based on model tests
design against single largest wave force in design sea state, uses highest wave in
wave group
Hdesign is estimated at a distance of 5Hs seaward of breakwater (Hdesign = 1.8Hs)
hb = water depth at 5Hs seaward of breakwater
L (or k) is calculated at hb using Ts = 1.1Tm (Tm is the average period)
modified to incorporate random wave breaking model
assumes trapezoidal shape for pressure distribution along front
Caisson is imbedded into the rubble mound
Uplift pressure distribution is assumed triangular

Note: hs
includes
wave setup

Elevation to which wave pressure is


exerted:

Pressure on Front of Vertical Wall:

= 0.75(1 + cos )H design


= direction of waves with respect to
breakwater normal (for waves approaching
normal to breakwater, = 0)
p1 = 0.5(1 + cos )(1 + cos 2 )H design

h c
p for > h c
1
p 2 = 1

0
for h c

p3 = 3p1
Buoyancy and Uplift Pressure

p u = 0.5(1 + cos )1 3 H design

(1) Effect of wave period on pressure distribution


2

2khs

1 = 0.6 + 0.5
sinh
2
kh
s

minimum = 0.6 (deep water), maximum = 1.1 (shallow)


(*) Increase in wave pressure due to shallow mound
2

= 2 = minimum of

h b d H de sin
2d

or
3h b d
H design

(3) Linear pressure distribution

h hc
1

1
3 = 1 w
h s cosh kh s
Decrease in Pressure from Hydrostatic under Wave Trough
: 0.5H design z < 0
z

p=
: z < 0.5H design
0.5H design
Tanimoto etal. (1976) added structure type modification factors (1, 2, 3) which are one for a
vertical wall (1 = 2 = 3 = 1)
= 0.75(1 + cos )1H design

p1 = 0.5(1 + cos ) 11 + 2 cos 2 H design

h c
p for > h c
1
p 2 = 1

0
for h c

p u = 0.5(1 + cos ) 31 3 H design


Takahashi, Tanimoto and Shimosako (1994) Table VI-5-54 modified the shallow mound
coefficient (*) for head-on breaking waves
Takahashi, Tanimoto and Shimosaka (1994) Table VI-5-54 modified the shallow mound
coefficient (*) for head-on breaking waves
Tanimoto and Kimura (1985) Table VI-5-56 updated the 3 modification factor to account for
inclined walls

Takahashi and Hosoyamada (1994) Table VI-5-57 developed corrections to p1, p2, p3 to account
for a structure with a sloped portion beginning just below the waterline

Takahashi, Tanimoto and Shimosako (1990) Table VI-5-58 updated the structure type
modification factors (1, 2, 3) to account for a vertical wall structure protected by a
rubble mound

Tanimoto, Takahashi and Kitatani (1981) and Takahashi, Shimosako and Sakaki (1991) Table
VI-5-59 updated the structure type modification factors (1, 2, 3) to account for
caissons with vertical slit fronts faces and open wave chambers

Force and Moment Calculations


Basic Calculation (refer to Sainflou), forces and moments are per unit length of structure
B
FG = FB - W

p1

FH
dh

hw

hs

FB
p2
pu

FU

bu

For Wave Crest


FG = [ c h w (h s + H + o )]B

FH = 12 w h s (H + o ) + 12 p 2 (h s + H + o )
=

1
2

(p 2 + w h s )(h s + H + o ) 12 w h s2

FU = 12 p u B
where pu = p2 for Sainflou (conservative) and c is the specific weight of the caisson
Overturning moments is then:
M H = FH d h

= 16 w h s (h s + H + o ) + 16 p 2 (h s + H + o ) 16 w h 3s and
2

1
6

(p 2 + w h s )(h s + H + o )2 16 w h 3s

M U = Fu ( 23 B) = 13 p u B2
Similar calculations can be made for the pressure distribution under the wave trough
Table VI-5-55, p VI-5-141 has the formulae for the Goda equations which include
biasing and uncertainty corrections

EM 1110-2-1100 (Part VI)


Proposed Publishing Date: 30 Apr 03
Table VI-5-55
Resulting Wave Induced Forces and Moments, and Related Uncertainties and Bias When Calculated From Wave Load
Equations by Goda and Takahashi

  
            
    

  
     
                 
 
           
 


#$ % %"
    !  " !   !  "

#$ % %&"
       
#$ % %'"
             










 

 


 

 



     
   
 
 



    
 


     
   
 
 



 

  

   



 
 
 

 

 


          


        
  !  " !
  
(

!   !  "  !   !  "

(

      
&

 
      







#$ % %%"
#$ % %("
#$ % %)"



     
   
 
 

    
  



     
   
 
 

 
 


 
      
 * +      
,      - .." + ..'"     - ,     / ..'"

      


  /  0  $  /
 0 *    
        
           
 
               
         




 

  


 

 

 



 








 









 

!

"#
$
$
#
#


 
#
#
#$
Fundamentals of Design

VI-5-141

Minikin's Method (1950)


Based on wave pressure records and shock press. work by Bagnold
pressure distribution with peak pressure at or near the still-water level
vertical breakwater resting on rubble mound
impact pressure decreases parabolically to zero at z = - H
generally overestimates pressures (15-18)
Dynamic Pressure:
p max = 101d(1 + d h ) H b L h
combined

2z
, z H 2
p m = p max 1

Hb
Hydrostatic Pressure:
p d = (d + 12 H b )
pmax = max dynamic pressure at SWL
pm = dynamic pressure
z = vertical distance from SWL
h = the depth of water a distance L from the
wall, h = d + Ldm
Ld = the wavelength at depth d
Lh = the wavelength at depth h
Hb = breaker height

m
1

Force under the dynamic distribution (acting at the SWL): Fm = 13 p max H b


Overturning Moment from the dynamic force: M m = 13 p max H b d
adding the hydrostatic force and moments to these gives:
2
Ftotal = 13 p max H b + 12 (d + 1 2 H b )
total force:
total overturningmoment:

M total = 13 p max H b d + 16 (d + 1 2 H b )

Minikin's Method for a wall on a low rubble mound


ps
pmax
Dynamic Pressure
0.5H

0.5H

dH
p max = 101d1 + b
h Lh
2

2z
, z 12 H b
p m = p max 1

H
b

Static Pressure
1 2z
0 z < 12 H b
H1
p s = 2 H b

1
z<0
2 H b z

Minikin's Method for a wall with top below the design breaker crest
using reduction factors (rm and a) from plots below

Hb

Dynamic force component:


Fm ' = rm ( 13 p max H b )

b'

Dynamic component of overturning moment


M m ' = 13 p max H b [d (d + a )(1 rm )]

0.9

0.9

0.8

0.8

0.7

0.7

0.6

0.6

2a/Hb

rm

= 13 p max H b [rm (d + a ) a ]

0.5

0.5

0.4

0.4

0.3

0.3

0.2

0.2

0.1

0.1
0

0
0

0.1

0.2

0.3

0.4

0.5

b'/H b

0.6

0.7

0.8

0.9

0.1

0.2

0.3

0.4

0.5

b'/H b

0.6

0.7

0.8

0.9

Caisson Width and Mound Dimensions Guidance


Caisson width:
General guidance: B = 1.7 to 2.6 H1/3 for reflective to breaking waves
Wave transmission is of primary concern.
hc

d
h

~ 5m
key stone
(scour protection)

Caisson Crest Elevation:


General guide: hc = 0.5 to 0.75 H1/3 , however design requirement become more
important:
allowed overtopping specifications

lee-side wave transmission requirements


Overtopping is less critical for structurally integrity compared to rubble
mound breakwaters (i.e. there is no armor layer vulnerable to wave attack).
However, a shorter caisson will have a shorter moment arm (see overturning
stability discussion below).
Mound Crest Elevation:
General guidance: d/h < 0.6 for breaking waves.
Scour at the base of the caisson is still a concern, especially in a breaking
wave environment. Therefore, the height of the rubble mound should be limited.
However, as seen below in the soil bearing capacity discussion, higher mounds
distribute the load more and enhance the ability of the soil to support the more
concentrated weight of the caisson. Large key stones may be placed at the base of
the caisson to reduce scour problems.
Sliding and Overturning Stability
To assess the sliding and overturning stability of the upright section, the weight
(W), buoyancy (B), the horizontal wave induced force (Fh) and uplift force (U) must be
considered. Buoyancy is the weight of the water displaced by the submerged volume of
the upright section. The dynamic uplift pressure is assumed to vary linearly from the
seaside to the lee-side.

Fh

Slip circles
dh

bu

Safety Factors (S.F.) are calculated as follows:


(1) For sliding, the friction due to the net downward forces opposes the horizontal wave
induced force

S .F . = (W B U ) Fh ,
where is the coefficient of friction between the upright section and the rubble
mound (or the bottom). For a new installation 0.5. After the initial
shakedown, 0.6.
(2) For overturning, moments are calculated about the lee-side toe

S .F . = (M W M u ) M p
for a symmetric section with no eccentricity:

M W = 0.5(W B )
M U = buU = 2 / 3U

M p = d h Fh
In designing breakwaters for harbor protection, safety factors are taken as 1.2 or higher.
"The overturning of a caisson implies very high pressures on the point of rotation.
The bearing capacity of the stone underlayer will be exceeded and the crushing of stones
at the caisson heel will take place. In reality the bearing capacity of the underlayer and
the sea-bed sets the limiting conditions. The soil mechanics methods of analyzing the
bearing capacity of a foundation when exposed to eccentric inclined loads should be
applied, i.e. slip failure or the use of bearing capacity diagrams." (Abbott and Price, p.
422)

Soil Bearing Capacity Calculations


B

Soft Soil - higher mound will


distribute the load more

Soft Soil - replace clay with


sand key

sand

Sand Key

clay

clay

sand

Generally, a rubble mound will distribute the weight of the caisson according to
its friction angle. Higher base mounds will distribute the load over a wider area and
reduce the load on the soil. Weak soil may also be replaced with a sand key which will
further distribute the load.
Guideline (D = depth of top sand layer or sand key):

D 2B only consider soil strength in sand (neglect clay below)


2B > D > 1.5B use combined strength by spreading the load
D 1.5B use clay load, sand may still be added to (1) increase drainage,
(2) help distribute load, (3) give better, more even surface

Eccentricity will shift the load as well.


e

te

Fh

W-B-U

As previously the allowable load developed from a bearing capacity


analysis must equal or exceed the actual load. The eccentricity (e) can be
calculated from the angle of the resultant force. Since the soil cannot support a
tension stress, the load must be corrected as follows:
For e 16 B :

6e W
p1 = 1 + ,
BB

For e > 16 B :

p1 =

2W
,
3 te

6e W
p 2 = 1
BB

p 2 = 0 ; where t e =

B
e
2

Soil cannot
support tension

Load on soil
p2

B'

p1

qa = allowable load
qa

1
2

( p1 + p2 )

Correction since soil


cannot support tension

p1

Summary of Design Procedure


1. Specify design conditions: design wave, water levels, etc.
2. Set rubble mound dimensions
3. Compute external wave loading
4. Perform stability analysis
a. Sliding stability
b. Overturning stability
c. Slip stability: local, translational and global
5. Perform a bearing capacity analysis
a. At mound level (i.e. at the toe of the caisson)
b. At the foundation level
6. Determine caisson stability under towing conditions and during the installation phase
7. Stress configurations
a. During towing and installation
i. Side
ii. Bottom
iii. Internal panel
b. Post installation
i. Side
ii. Bottom
iii. Internal panel
8. Structural Detailing