By Richard Silvester, M. Asce: and Ocean Engineering, Vol. 112, No. 1, January, 1986. ©ASCE, ISSN 0733-950X

USE OF GROUT-FILLED SAUSAGES
IN COASTAL STRUCTURES
By Richard Silvester, 1 M. ASCE
ABSTRACT: An alternative to rubble mound construction, with or without pre-

cast concrete armor units, is recommended which comprises flexible mem-
Downloaded from ascelibrary.org by CESAR TORRES on 01/03/13. Copyright ASCE. For personal use only; all rights reserved.
branes filled with cement mortar in situ. The placement of these for greatest
stability is suggested by maximizing surface contact of sausage type units. Re-
sults of tests to derive an economical mortar mix are described. Construction
details are outlined together with comparative costs for a marina breakwater.
INTRODUCTION
The age-old concept of constructing shoreline structures with rock or

precast concrete armor units based u p o n weight for any given wave height
has recently come into disrepute (8). The formula derived by H u d s o n
(10) h a s been used for t w o decades even t h o u g h it contains only wave
height as the force input. This m a y have been suitable in shallow water
where celerity of all waves is essentially the same, determined as it is
by water depth only. However, as structures have been taken to even
greater depths, the waves strike the structure at velocities which are de-
pendent on both wave period a n d depth. Thus, longer period compo-
nents of the spectrum approach at greater speed with resultant higher
forces for any given wave height.
There has been a need for larger a n d larger armor units, which in-
crease the cost prodigiously. If these are just d u m p e d randomly, or are
even placed with so-called interlocking action, they still only have point
contact with each other. They therefore suffer rocking motions a n d can
be plucked from their position in the sloping face by high water pres-
sures developed behind t h e m or b y outward flow w h e n the wave trough
arrives. These unreinforced units are not designed to absorb applied dy-
namic loads. Once cracked a n d broken, their weight is halved making
subsequent removal easier. They can then be thrown around by the waves
to break further units.
An aspect of breakwater design which has received little attention is
the inner core comprising smaller material. Swashing of water t h r o u g h
this rubble-type stone can dislodge it and cause settlement of the armor
units resting on it. Smith a n d Gordon (30) state: "Unfortunately it seems
that so much design attention is normally applied to the behaviour of
armor in the actual wave breaking zone that the potential hydraulic forces
that must be applied to the submerged armour tend to be largely ig-
nored."
Per Bruun and Gunback (3) explained the p h e n o m e n o n of "reso-
nance" on the slope of rubble m o u n d material as " t h e situation that
TJept. of Civ. Engrg., Univ. of Western Australia, Australia.
Note.—Discussion open until June 1, 1986. To extend the closing date one month,
a written request must be filed with the ASCE Manager of Journals. The manu-
script for this paper was submitted for review and possible publication on Sep-
tember 17, 1984. This paper is part of the Journal of Waterway, Port, Coastal
and Ocean Engineering, Vol. 112, No. 1, January, 1986. ©ASCE, ISSN 0733-950X/
86/0001-0095/$01.00. Paper No. 20317.
95
J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

occurs when run-down is in a low position and collapsing—plunging
wave breaking takes place simultaneously and repeatedly at or close to
that location." Sawaragi et al. (24) have shown this to occur when the
surf similarity parameter (tan 6 VH/L0) has values between 2 and 3,
where tan 0 is the slope of the face, H is local incident height, and L0 is
the deep-water wave length. This parameter is the same as Iribarren's
number (11) which has been related to minimal sized stones for stability
(15). Again critical values of this parameter are between 2 and 3. The
importance of this parameter in design has also been shown (4), sug-
gesting an S shape profile for the breakwater. This, however, still results
in steep underwater slopes which can readily reflect waves.
Reflection of oblique waves from the lower face of breakwaters, even
when the upper zone is being dissipated by large armor units, can effect
substantial scour of the bed, which results in a trench with side slopes
equal to the angle of repose of the sediment. This critical state cannot
countenance build-up of pore pressure without sudden failure of the
face, as near breaking waves approach the structure. The degree of re-
flection from rubble-mound slopes has been studied (16) which shows
its dependence on: (1) Porosity; (2) width of material; (3) diameter of
stones; (4) water depth; (5) wave period; and (6) incident wave height.
A graph was put in the form of Kr versus a parameter
fcw
= 2^r
tanh
L
where We = width of equivalent rubble mass; d = water depth; L = local
wave length; and L„ = deep water wave length from which the curves
of Fig. 1 have been derived. As seen in the inset of this figure an equiv-
alent submerged rectangular volume of stone has been used in the anal-
ysis which is believed to be equivalent to the trapezoidal cross section
normally existing. The reflection coefficient Kr is based upon the wave
as a whole which may involve breaking near SWL. Thus, considering
reflection for lower sections of the water column, the reflection com-
ponent for this would be even greater than indicated in Fig. 1. It should
be remembered that in many breakwaters the thickness of protective
stone layers is not great and many engineers insert fabric just beneath
them to reduce scouring of smaller size material of the core. This con-
centrates flow of the downwash from the wave which is maximum at
the trough. When a steep crest approaches simultaneously an upward
flow in this permeable mass occurs in the same region, which magnifies
the outward force on the stones and the possibility of failure.
The recommendations of the various authors quoted can be summa-
rized by the conclusions of Sawaragi et al. (25) in which they state: "De-
sign formulae in future must take the hydrodynamic forces produced by
resonance into consideration as an external disturbing force." All this
points to the need for a new approach to the design of breakwaters,
which permits large scale voids to be present throughout the whole
structure and obviates the use of small sized material to save costs.
96

Kr
0.2 0.3 0.4 0.50.6 I.O
FIG. 1.—Reflection Coefficient Kr for Given We/D and rf/L,
As noted already, scouring of the seabed adjacent to breakwaters from

waves reflecting obliquely may also be another cause of such slumping
of armor units (26,28). Waves, particularly the insidious swell, do not
necessarily approach normal to a structure, in spite of the accepted flume
tests to determine stability. When they are angled to the face, especially
the long-period components of the wave spectrum, short-crested sys-
tems are established which are very conducive to scouring along the
length of the structure. Some engineers believe this cannot be the case
in deeper water, but the vortices so generated in this system can pen-
etrate to the bed with very little attenuation (26). The associated macro-
turbulence this produces maintains sediment in suspension which is then
carried downcoast by the excessive mass-transport in this short-crested
system (9).
Thus, if any progress is to be made in the design of coastal structures
the following points should be noted:
1. Armor units should be designed which fit snugly together, giving

resistance to movement by friction over wide surfaces.
2. Core material should consist of units equally as large as other armor
blocks if they can be manufactured from cheap concrete.
3. Spaces should be provided between elongated units to permit wave
absorption, reduce reflection and attenuate upward pressures.
97

4. The bed seawards of the breakwater should be protected from
scouring by an impermeable mat for either traditional rubble mound
structures or flexible membrane designs.
HISTORICAL
Porraz et al. (19-22) have pioneered the development of flexible mem-

branes acting as formwork for cement mortar which is cast in situ. Their
purpose was to utilize local unskilled labor rather than expensive im-
ported technology in developing countries. Although it has proved use-
ful in this situation, this new concept has application in western coun-
tries where many marine structures are required in isolated sites where
infrastructure for transport of stone is not always available or is very
expensive.
This solution utilizes high strength fabrics which can be sewn to any
desired size of armor unit. Not requiring to be man-handled once con-
structed they need not have the same strength as the precast concrete
alternative. Because they will take up the shape of units already placed
beneath, their surface contact is broad which prevents rocking. Good
frictional resistance is provided which resists the pushing and pulling
action of waves. This is complemented by the undulatory nature of the
surface contact which provides a sawtooth resistance to shear forces of
each against its neighbor beneath.
The weak mortar mix need be only of sufficient compressive strength
to support the weight above plus the moment from the side force of the
waves. Since the flexible fabric is required only to hold the mixture in
place until it sets, subsequent deterioration of the membrane is of no
concern. The mortar must resist swashing by water, which perhaps con-
tains suspended sediment.
Thus, the demands as set out in items 1, 2, and 3 above can readily
be met. As in any engineering structure, costs of alternatives must be
considered. In such a comparison it is not the unit cost (e.g., cost per
m3) that is important but the overall expenses of establishment, inter-
ruption in construction, maintenance, and relative volumes of material.
It will be shown that the flexible membrane solution can be cheaper than
rubble, with or without precast armor units.
The writer believes that longer containers, in the form of sausages,
have greater advantages over the shorter alternative, which is equivalent
to large bags. The overall size of each homogeneous mass can be made
so large that it is inconceivable that any wave could disturb or remove
it. The weight of individual units can be such, and their frictional resis-
tance with each other so effective, that even the stability coefficient (10)
normally used for breakwaters becomes irrevelant. Tests are required to
prove this contention. As for precast concrete armor units, these cast in
situ monoliths would not be reinforced; but, as they do not suffer dy-
namic forces due to rocking, the need for high shear and tensile strength
is not so great.
Settlement of the structure, occurring with randomly placed rubble
material which compacts as the load grows, is not likely with mortar
filled units since they are added uniformly. Compaction of the sand be-
low the structure occurs evenly due to the rectangular nature of the cross
98

section. Should any cracking occur due to such differential settlement,
it will occur in sausages well down in the structure where wave forces
are less; but the resulting units are still massive compared to stone com-
ponents.
DESIGN OF UNITS
Analysis and verification of sausage shapes has been carried out at the
Univ. of Western Australia (13,14), from which the curves in Fig. 2 have
been obtained. The input variables are the circumference (S) of the fabric
and the pressure head {bt) of the mortar in the filled condition as mea-
sured from the base. From the ratio bJS values of height (h), width (b),
cross-sectional area (A), contact width at the base (V), and tension (T)
per unit length of sausage can be derived from the dimensionless pa-
rameters.
Values of the above variables are as listed in Table 1, assuming a mem-
brane width of 12 ft (3.66 m) (although multiples of such width can be
used with longitudinal seams). If the breaking strength of this specific
fabric in the warp direction is 70 lb/in. (1,260 kg/m), the maximum height
of sausage possible (cast underwater) is 3.28 ft (1.0 m). The increase in
cross-sectional area to the equivalent rectangular shape (A/hb) is only
4% from 2.62 to 3.28 ft (0.9 to 1.0 m) in height and hence this lower
value, with much lower tensile stress of 38.5 lb /in. (693 kg/m) is pre-
ferred. It should be noted that tensile stress for the above water casting
is almost twice that for the submerged state.
Sausages could be laid across the width of the structure with only their
ends in contact with the open sea. Except through tensional cracking
these units cannot conceivably be detached from the breakwater by waves.
As each unit should have full support beneath it and full vertical load
I Q I r—• V ' I ' ' 1 1 1 1 1 1 1 1 1 1 |03

O O! '2 o.3 0.4 O.S 0.6 Q7 OS. . 0 . 9 I.O U 1.2 1.3 1.4 IS 1.6 •
FIG. 2.—Characteristics of Flexible Membrane Units Filled with Mortar (SG = 2.0)
99

TABLE 1.—Variables in Design of Mortar-Filled Sausages
Mm) h/s bi/s Mm) b/s b(m) b'b (m)
(1) (2) (3) (4) (5) (6) (7)
1.0 0.273 0.62 2.27 0.348 1.27 0.365
0.9 0.250 0.42 1.54 0.361 1.32 0.490
0.8 0.218 0.31 1.13 0.380 1.39 0.590
0.7 0.191 0.25 0.92 0.395 1.45 0.650

0.6 0.164 0.19 0.70 0.410 1.50 0.700
0.5 0.137 0.16 0.59 0.425 1.55 0.780
a
Tension when cast submerged.
b
Tension when cast above water surface.
above it, the resulting face should be almost perpendicular and not less
than 4:1. As indicated in Fig. 3, the preferable alinement of units is 45°
to the axis in order to make each longer and heavier. The underlying
row will provide an undulating surface which the upper unit will follow,
so giving a "sawtooth" contact for good shear strength (see section BB
of Fig. 3).
Positioning of sausages in alternate layers should be carried out ac-
curately to keep them from sitting on the space beneath, which would
cause flexure in the units running normally between them. Such posi-
tioning can be effected by nailing the empty membrane to those already
in place.
At the extremity of the structure a rectangular extension should be
provided, as depicted in Fig. 4, so sausages of greater length can be
positioned at all levels. This a critical region of the structure, where vor-
tices due to wave diffraction are pronounced (6), and, hence, bed pro-
tection should be extended around to the leeward side.
FIG. 3.—Plan and Sections of Break- FIG. 4.—Enlargement of Breakwater Tip

water Trunk Using Sausage-Type Units to Permit Sausage Units of Greater
Length
100

with Circumference of 12 ft (3.66 m) (See Fig. 2)
V (m) A/lib A (m2) 2T/yS2 T (kg/m) a V (kg/m) b
(8) (9) (10) (11) (12) (13)
0.46 0.825 1.05 0.13 1,202 2,096
0.65 0.832 0.99 0.075 693 1,209
0.82 0.851 0.95 0.043 397 692
0.94 0.875 0.89 0.031 286 499

1.05 0.905 0.81 0.021 194 338
1.21 0.950 0.74 0.015 139 242
The suggested spaces would provide channels through which water

could oscillate to relieve high pressures on the seaward face. These tubes
of corrugated form would have virtually rounded orifices at each end.
Jets would discharge on the leeward side of the structure which will be
normal to each other at alternate layers. The vortices generated should
attenuate each other quickly and reduce wave transmission on the har-
bor side. The jets themselves would be reduced by the alternative pas-
sages where flow would proceed to zones of lower pressure. This par-
ticularly would be so for waves arriving obliquely to the structure. But
also, storm waves are multidirectional and, hence, wave forces and pres-
sures will not be synchronous along the whole structure. Uplift forces
on sausages will be greatly reduced by the provision of such a pressure
releasing mechanism. Tests are required to measure transmitted waves
through a structure as well as records of pressure throughout the struc-
ture.
The cross section of the breakwater will depend upon the amount of
overtopping that can be permitted during the worst storms. Overturning
moments and shear forces must be resisted at all levels, particularly at
the bed and the crown. Since the frictional forces to resist horizontal
shear are determined by the vertical weight from above, the sausages at
the crown must be carried high enough to where forces are sufficiently
reduced. This is where the diagonal alinement of units is beneficial since
the spatial force of any wave is spread along each sausage progressively
and not applied instantaneously over the complete length.
Wave forces reduce in triangular fashion from SWL to the reach of the
partial standing wave, as indicated by AB in Fig. 5. If overtopping to
the extent of BC is accepted, the horizontal force (as indicated by the
hatched area CD plus shear from the flow) must be resisted by the top
sausage layer through its contact at level D. This is provided by both
surface friction, through the weight of this layer, and the undulatory
surface with the units just below level D. Should experience indicate that
insufficient allowance has been made, new sausages can readily be cast
in any spaces created in the zone CD and others added above C, where
horizontal forces are reducing linearly.
Near the bed internal and leeward jets could promote scouring, there-
fore, it is suggested that the first 3 layers of sausages be cast without
spaces between. Those at the crown should also be poured adjacent to
101

FIG. 5.—Horizontal Forces Exerted on Structure Due to Partial Standing Wave
each other. In this case, alternate units should be cast first so that sub-
sequent sausages clasp them along their sides. This implies that they do
not act individually in being lifted by internal pressures, nor sheared
separately. A continuous undulatory surface will result at the crown on
which a concrete road can be cast for access (see Fig. 3).
As already stated, all breakwaters reflect some energy from incident
waves, particularly those of long period. The structures made from flex-
ible membranes will be steeper and, in spite of the perforations as sug-
gested, may reflect more readily. This quality could be utilized bypass-
ing sediment across harbor and river mouths by forming short-crested
systems from oblique swell waves, as suggested elsewhere (27,28). In
either type of construction, rubble-mound or otherwise, the seabed near
the structure should be protected by mattresses.
Sausages could be cast over the floor seawards of the breakwater and
along the ribbon of reflected waves. The extent of this needs further
research and experience in prototype situations. The thickness of these
units should equal the depth of the trough below SWL for standing waves
which is equivalent to the lift exerted on them. Additions to permit a
factor of safety should be considered. Alternate sausages laid with spaces
between could have later units cast which would form "tongue and
grooved" joints with them. By this means the vortices generated within
the short-crested wave pattern (26) cannot penetrate to the floor. This
is not the case with stone cover since vortices enter the voids and suck
sand from beneath the rubble elements.
DESIGN OF MORTAR MIX
Since these monoliths are not required to be handled once they are
cast, or not suffer impact forces due to rocking, they can be designed
for their major strength requirement, withstanding compressive load.
This includes the vertical weight of the structure above plus the
compression resulting from bending moments exerted by wave impact.
These can be met by mortar strengths much less than those of normal
102

concrete (7). An added requirement is resistance to wear by water with
suspended sediment. Since limestone meets these requirements satis-
factorily, a strength similar to this material could be adequate. It should
be noted that the mortar, not the flexible membrane, provides weara-
bility against water flow. The fabric is needed to give support only for
the period of mortar setting, after which it can deteriorate from UV rays
or tearing incidents of any origin.
Besides these basic strength requirements, other criteria for the mortar
are as follows:
1. It must have 100% slump, or be a slurry, in order to flow to any

part of the fabric container without vibration.
2. It must provide a tensile strength equal to that of limestone.
3. It must have a specific gravity equal to or exceeding that of lime-
stone.
4. It must gave a water wearing capability better than that of lime-
stone.
5. It must be as cheap as possible for the cementing materials avail-
able.
Tests have been carried out (7,29) which compare various ratios of
cement and admixtures to sand with the compressive and tensile strengths
of limestone. Samples of limestone from seven different sites were found
to differ greatly in these values so that one, which appeared better by
far from the rest, was used in the comparison. To derive the most eco-
nomical mix the additions to Portland cement tested were g r o u n d gran-
ulated blast-furnace slag, fly ash and diatomite. Of these only the first
improved the characteristics of the test cylinders. Since the bulk of ma-
terial on any coastal site would be beach sand this was included, in ad-
dition to sand from terrigenous sources. Also seawater was used in some
specimens to determine its influence. In order to reduce the w a t e r / c e -
ment ratio, while maintaining a high slump a n d easy flow through
p u m p i n g equipment, various percentages of detergent were added.
Measurements made on all limestone samples were compressive and
tensile strengths, specific gravity, a n d scour b y a water jet over 72 hr.
The maximum compressive strength was 1,335 psi (9.2 MPa). The ratio
of tensile/compressive was found to be a linear relationship with a con-
stant of 0.127, thus giving a tensile strength of 170 psi (1.17 MPa). The
scour as determined by the maximum depth of hole after 72 hr with a
constant jet pressure head resulted in the following expression:
compression (MPa) = 10.4 - 1.92 scour (mm)

implying that scour did not become zero until the compressive strength
exceeded 1,510 psi (10.4 MPa). For the accepted compression 1,335 psi
(9.2 MPa) the scour w a s t h u s 0.024 in. (0.62 m m ) . The specific gravity
related as follows:
compression
SG = 1.64 +
33 MPa
which for 1,335 psi (9.2 MPa) gave SG = 1.95.
103

0.4 0.6
S/(S*C)
FIG. 6.—Mortar Strength versus Sand/Cement Ratio for Various Sand Gradings
3 4
sond/ccmznt
FIG. 7.—Compressive Strength versus Blast-Furnace Slag (S) Plus Cement (C)
for Variations in Fresh Water (W) and Dispersenex F (ml/kg of S + C) for Ratio
of Laboratory Sand/(S + C) = 5.0
104

The traditional approach with mortars is to use a high cement content
and low water/cement ratio. This increases cost and the possibility of
cracking while workability lessens. The object in this case was to have
a slurry for placement without vibrators, with a compressive strength
and other characteristics similar to the best limestone sample. The ad-
mixtures were used to achieve this while minimizing cracking in w h a t
are virtually bulk castings.
Various gradations of sand in mortars have been used (1,18,31), re-
JOO JOO
FIG. 8.—Strength versus Age for: (a) Laboratory Sand and Fresh Water (—), and
Seawater; (b) Beach Sand and Seawater ( ). All Cases, S/(S + C) = 0.6 and
Sand/(S + C) = 5.0
LEGEND
Mini M i x l l
optimal strenqth o ®
2 8 d o y sJr«ngt»o •»
IOO , , % 200 300

slump (mm)
FIG. 9.—Strength versus Slump for: (a) Mix I with Laboratory Sand and Fresh
Water; (b) Mix II with Beach Sand and Seawater. All Cases, S/(S + C) = 0.6 and
Sand/(S + C) = 5.0
105

suits of which are summarized in Fig. 6. It is seen that increasing sand/
cement ratio decreases 28 day strength, as expected. The fine sand curve
(the lowest) indicates that with a ratio of 5:1 a strength of (1,450 psi) (10
MPa) is obtainable, which is slightly larger than for the limestone tested.
Since it was anticipated that this would improve with admixtures this
ratio was adopted for all tests.
Included in the data of Fig. 6 are the water/cement ratios used, whose
increase caused the strength to fall. To obviate this and to obtain a smooth
mix for pumping and placing a detergent known as Dispersenex F was
added at rates X = 0, 2.5, 5.0, and 6.76 ml per kilogram of cementing
agents (equivalent to up to 0.676% by weight, approximately).
Granulated ground-blast-furnace slag was the only additive which im-
proved strength. As seen in Fig. 7 optimum 28 day strengths came from
ratios S/S + C = 0.6 which is equivalent to slag being 150% by weight
of cement. It is also observed that X = 2.5 gave better results than zero
detergent. Extra specimens were therefore tested for these specific mixes
for beach sand and seawater. Compression tests were carried out at 7,
14 and 28 days, as indicated in Fig. 8, where it is seen by extrapolation
that the 300 day value was 7% greater than for 28 days. This age char-
acteristic has been observed elsewhere (2,23). Also apparent in this fig-
ure is the benefit of using beach sand and seawater. Reasons for this
could be the increased content of calcium carbonate in beach sand (15.5%)
compared to 0.1% in terrigenous material, plus the presence of sodium
magnesium compounds in seawater. These have all been found in Port-
land Cement (17).
Since tests were conducted on specimens with limited slump it was
necessary to extrapolate strengths for complete slump, or 12 in. (300 mm)
in this case. In Fig. 9 optimal 300 day and 28 day compressive strength
is plotted against slump. It is seen that for a 12 in. (300 mm) slump the
highest strength possible is 1,720 psi (11.8 MPa) which is higher than
the 1,335 psi (9.2 MPa) indicated for limestone.
From Fig. 10 a water/cement ratio W/S + C = 1.18 with X = 2.5 is
required for 300 mm slump. Further tests are suggested to confirm this
slump and strength as predicted in Figs. 8-10. However, as a first trial
on local sands these are worthwhile test values.
Results of indirect tensile strength (by the Brazil test) are shown in
Fig. 11, where again the highest value, obtainable for S/S + C = 0.6 and
X = 2.5 ml, was 238 psi (1.64 MPa) which is higher than the 170 psi
(1.17 MPa) from limestone. The ratio tensile/compressive was 0.129,
similar to the 0.127 for limestone. Thus, for the optimum compression
of 1,720 psi (11.8 MPa) at 300 days for 12 in. (300 mm) slump a tensile
strength of 220 psi (1.52 MPa) can be expected, which still exceeds that
for limestone.
Water scour of the cement mix, under a similar jet pressure head for
the limestone tests, showed that compressive strengths in excess of 1,020
psi (7 MPa) suffered zero scour over the 72 hr. This was an improvement
on the limestone which required 1,510 psi (10.4 MPa) for this capacity
to withstand scour.
The specific gravity of cement mixes was con-elated with 28 day strength
resulting in
106

o- ^
_^r
s^
_ ^ r _
0.4
0.2
lO 60 100 200 300
slump (mm)
FIG. 10.—Water/Cement Ratio versus Slump for Both Laboratory Sand and Fresh
Water, and Beach Sand and Seawater, with X = 2.5 ml/kg of (S + C)
FIG. 11.—Indirect Tensile Strength (by Brazil Test) versus S/(S + C), for Vari-
ations in X and W for Sand/(S + C) = 5.0
107

SG= 1.68 + -
30
which was commensurate with that of the limestone. For the accepted
strength of 1,720 psi (11.8 MPa) previously derived an SG of 2.07 is ob-
tainable which exceeds that of limestone (1.95).
From these results the following proportions are suggested for local
tests for the desired slump and other characteristics:

beach sand/(slag + cement) = 5.0
slag/(slag + cement) = 0.6 by weight
seawater/(slag + cement) = 1.18 by weight
detergent = 0.25% of (slag + cement) by weight
The above mix should give a complete slump and provide the follow-
ing values:
300 day compressive strength = 1,720 psi (11.8 MPa)

300 day tensile strength = 220 psi (1.52 MPa)
specific gravity = 2.07
This mortar when set should also provide good resistance against water
scour; certainly better than that of limestone, which is widely used in
marine structures.
COSTS
It is difficult to compare costs of a flexible membrane structure with

equivalent rubble-mound construction due to the following:
1. A specific situation has to be analyzed which may not be appro-
priate at any other site where material supply conditions are different.
2. The methods of costing for equipment used in sausage construction
can vary greatly.
3. The purchase price of cement and additives (the bulk of the cost)
is determined by local suppliers, which can increase swiftly over time.
4. The wave climate of the area is unique.
Even so, it is worth comparing the cost of a known rubble-mound
structure with a flexible membrane alternative for an assumed mode of
construction. This has been carried out (7) for a limestone breakwater in
Western Australia for which costs and design waves were known. A
structure consisting of mortar sausages was designed and costed. The
cross sections of the two alternatives are illustrated in Fig. 12.
The total length of breakwater was 4,430 ft (1,350 m) and contained
300,000 tons of limestone rock, the armor rock consisting of 12 tons. The
design wave was 9.85 ft (3 m) in height and 11 sec period in 19.7 ft (6
m) of water. The job was completed in 24 weeks and cost $1,125,000 in
1978, which excluded the access road. Inflation to 1982, when the com-
parison was made, was 2.5, giving the comparative cost of $2,800,000.
The design of the sausage structure covered sliding at the base, bear-
ing capacity of the soil, overturning of the structure and shear at SWL.
108

wt. (tonnes) "/„ by w t .
H o-8m > 7 13
W 2 • I in 5-7 2S
sausage structure A t-SSm1 rubble mound — 3-5 37
13 l a y e r s v L
'°™ ' i_ r I 3 25
FIG. 12.—Cross Section of Breakwaters Used in Cost Comparison
A protective mat was to be provided for this vertical faced breakwater,

even though bed protection was not provided for the rubble-mound al-
ternative. The sausages comprising this mat were required to be 3.08 ft
(0.94 m) thick in order to resist uplift by standing waves and to be car-
ried out for a quarter of the wave length of 65.5 ft (20 m).
A particular nylon fabric was chosen 16.4 ft (5.0 m) wide to give a
circumference of this dimension. With a filled height of 2.62 ft (0.8 m)
this gave a cross-sectional area of 16.7 sq ft (1.55 m2). For a structural
width of 23 ft (7 m) and sausages laid at 45°, the length of each is 32.5
ft (9.9 m), giving a total volume for each of 540 cu ft (15.35 m 3 ). These
would weigh over 30 tons each. Allowing 0.985 ft (0.3 m) spaces be-
tween units for 16.8 ft (5 m) of the total height of 28.6 ft (8.73 m) the
final volume of mortar, including the protective mat, was calculated as
3,370,000 cu ft (93,000 m3).
Mortar mixing equipment and accessories for two teams were priced,
including standbys in case of breakdown. This total was divided by 5
for 20% depreciation on the job. Fuel and labor costs (at current prices)
were added as follows:
nylon fabric 156,755

cement + slag 1,829,000
detergent 66,125
machinery + running cost 103,656
labor 328,227
Total $2,483,763
This was 88% of the cost for the limestone version. If the protective
mat was excluded the total was $1.79 or 64% of the rubble mound al-
ternative.
It is seen that, even with some reservations in the manner of esti-
mating, this novel approach equals or is more likely to be less than that
of the traditional method of construction. However, the final cost of any
complex should include that for the probable maintenance. In an ex-
treme storm sequence it is normally accepted that rubble-mound struc-
tures will require some rebuilding. But with the massive units used in
the mortar-sausage version, with good surface contact, no rocking or
displacement should take place. Even if it does, the addition of extra
109

material will not require as large an establishment cost as the rock al-
ternative, involving the use of trucks and cranes. Thus, it is submitted
that this new approach should be priced and compared in any designs
for breakwaters, groynes, and headlands.
LOGISTICS
One of the great advantages of this constructional concept is the sim-

plicity of the operation. Very little equipment is necessary and this is
transportable by trucks of reasonably low capacity. Certainly the cement
and additives must be delivered either in bags or in bulk. It therefore
can serve remote sites where there is little infrastructure in terms of roads
or water supply. Two concrete mixers, a slurry pump, and a small air
compressor for use by divers, is all that is necessary on the beach, and
perhaps additional units in case of breakdown. A small pump for sea-
water and cleaning is a requirement. Sand may have to be shifted short
distances for which a modest front-end loader should suffice.
Two mixers could be filled by unskilled laborers using lightweight boxes
for measuring volumes. The mixers then feed into a hopper of a slurry
pump which delivers through a 3 m (75 mm) pipe up to 656 ft (200 m)
to the point of construction. For longer delivery a second hopper and
pump may be necessary. Unskilled workmen can also connect this pipe
to inlets of successive sausage units without stopping the pump. Work
underwater must be carried out by divers with breathing masks.
The alternative of establishing a ready-mix complex on site will de-
pend upon the site conditions and magnitude of the operation. Hoppers
for sand and bulk cement and blast furnace slag (or other admixture)
would be required to feed mixer trucks which should pour continually
into the pump hopper. It is necessary with either method to maintain a
continuous flow of slurry since settling in the pipe could make restarting
of the pump difficult. From two normal mixers 1 ton of mortar could be
poured every 3 min, but longer sausages may require doubling of this
rate. There is little doubt, from experience overseas, that construction
by this method is much faster than for the traditional rubble-mound
structure. Cessation of operations due to inclement weather, shortage
of funds, or labor disputes, does not involve expensive equipment with
large overheads.
Divers are required to hold membranes underwater while the initial
sausages are poured. Once the base units are in place membranes for
upper layers can be nailed to these. As the fabric container fills, nails
pull out, causing slight tears which are of little consequence. As the
membrane becomes full, small cuts across the crown will permit excess
water to drain, leaving solid mortar to the desired thickness of sausage.
APPLICATIONS
Because of the great variety of sizes possible these armor units are
applicable to all types of marine structure, such as breakwaters, groynes,
and headlands. Since overtopping should not dislodge units at the crown,
it could be acceptable during the more severe storms. Many applications
for fabric formwork in marine and fluvial situations have been described
110

(12). They have been used to protect offshore pipelines and support them
over rocky seabed undulations. Scour holes beneath and adjacent to bridge
piers have been filled by such monoliths. Levees and embankments have
been protected from wave action in large bag or mattress form. They
have application in causeways for pipes or roads and also as dykes for
reclaimed areas. They, in fact, could be used wherever the traditional
but expensive rubble-mound alternative has been used in the past.
The major advantage of the system is that the bulk material used in
construction is available on the site. This obviates quarrying and the pro-
vision of large rock units, or the crushing and screening of smaller stone
for use in precast concrete units. This overcomes the necessity for road
networks and the consumption of expensive fuel. For developing coun-
tries, this also means less importation of technology at great expense.
Rock itself is becoming scarcer or more distant. As breakwaters are taken
into ever deeper water, the size of the armor units grows, requiring larger
trucks and cranes to handle them. Even so, the stability of some of the
precast blocks is being questioned (5).
The low strength requirement of these mortar structures calls for lean-
er mixes and less strict demands on the sand used. The smaller structure
required to achieve the hydraulic task of dissipating or reflecting waves,
rather than the larger rubble-mound alternative whose size is dictated
by construction requirements, makes for a more economical approach
overall. Interruptions for short or even longer terms does not increase
the cost greatly due to lower mobilization charges. The use of unskilled
labor could aid all economies, be they of developing or fully developed
countries.
Refs. (1,19-22) have been cited where the flexible membrane approach
has been applied to modest construction as in groynes and some off-
shore applications. Pilot studies are now required in natural settings to
show the advantages of the sausage concept as distinct from the bag-
type application so far used.
One of the requirements, that of bringing down the cost of flexible
membrane units, has been discussed above. Local sands, cements and
additives available at reasonable prices should be tested, in order to ob-
tain strengths equivalent to limestone in common usage in marine struc-
tures. The logistics of the system require attention for both metropolitan
and remote coastal sites.
There is little doubt that these various problems can be overcome if
the same research effort is applied to them as has been the case for tra-
ditional structures in the past few years. Work ort soil-cement dams and
pavements has application in the sea environment, the main difference
being in the type of slurry to be employed. These marine structures us-
ing light-weight flexible formwork must become a feature of coastal con-
struction in the future, to reduce the expense of materials and transport
of rubble-mound alternatives with or without precast concrete armor units.
CONCLUSIONS
1. The current inquiry into the stability of breakwater armor units

needs to be extended to cover new and more economical modes of con-
struction.
111

2. Such novel approaches should provide permeability within the
structures in order to minimize uplift and scouring forces within sub-
layers.
3. New design formulas should take into account the resonant phe-
nomenon when outwash coincides with the arrival of the following crest.
4. Reflection of waves, particularly long period components of the
spectrum, from steeply sloped rubble-mound structures is a fact to be
appreciated.
5. Where persistent waves arrive obliquely to a breakwater, reflec-
tion will establish a short-crested system which is conducive to bed scour,
calling for bed protection.
6. Constructional units are required which have extensive surface
contact to prevent rocking and provide good shear strength.
7. Large monolithic blocks can be built by using flexible membrane
formwork filled in situ with a mortar, using sand from the site.
8. The mortar, in the form of a slurry for easy placement, need have
only the strength of limestone to withstand the wave forces and wear
due to sloshing of water with suspended sediment.
9. Units of sausage shape are preferable to those of bag proportions
in order to provide the weight to impede uplift or horizontal displace-
ment in shear.
10. Characteristics of sausage units cast under water are available for
design purposes.
11. Sausage units laid trellis fashion across the structure distribute forces
j along each, thereby minimizing the effects of wave action.
| 12. Spacing of sausage units at mid-depths reduces the forces on them,
while leeward waves are kept to a minimum by the interaction of vor-
tices from the jets at each layer.
13. A mortar mix has been devised which could lead to economical
combinations of local sand, cement, and additives to be tested for each
site.
14. Costs of structures using sausage-type units can be less than for
the rubble-mound alternative, taking both short and long term charges
into account.
15. Logistics of flexible membrane construction require attention, but
the reduced requirements for transport and fuel, equipment and over-
heads, technical expertise to operate machinery, and overall volume of
material, must lead to reduced costs.
16. There are and have been many applications of flexible membrane
units both on the coast and offshore, hence, their expansion to larger
marine structures is inevitable.
17. The resistance inertia of the profession to novel ideas can be over-
come by pilot studies of modest character, which require research funds
to be displaced from traditional concepts that have proved less than ef-
fectual.
APPENDIX 1.—REFERENCES
1. Alvariz, J. A. M., Faradji, M., and Porraz, M., "Breakwater, Rockfill and
'Insitu' Rocks Construction with Bolsacreto System," 4th International Annual
Conference on Materials Technology, 1974, pp. 1-20.
112

2. Bamforth, P. B., "In-situ Measurements of the Effect of Particle Portland Ce-
ment Replacement using Either Fly-Ash or Ground Granulated Blast-furnace
Slag on the Performance of Mass Concrete," Proceedings of the Institution of
Civil Engineers, Part 2, 1980, pp. 777-800.
3. Bruun, P., and Gunback, A. R., "Stability of Sloping Structures in Relation
to J = tan a/VH/L 0 and Risk Criteria in Design, Journal of Coastal Engineering,
Vol. 1, 1978, pp. 287-322.
4. Bruun, P., and Johannesson, P., "Parameters Affecting Stability of Rubble
Mounds," Journal of the Waterway, Port, Coastal and Ocean Engineering Division,
ASCE, Vol. 102, No. WW2, 1976, pp. 141-164.
5. Edge, B. L., et al., "Failure of the Breakwater at Port Sines, Portugal," Coastal
Engineering Research Council, ASCE, 1982.
6. Chang, H. T., "Vortex Scour due to Wave Action at a Breakwater Tip," Pro-
ceedings 6th Australasian Conference of Hydraulics and Fluid Mechanics, 1977, pp.
348-351.
7. Goh, P. J. P., "Use of Mortar Filled Containers for Marine Structures," thesis
presented to the Univ. of Western Australia, at Nedlands, Western Australia
in 1983, in partial fulfillment of the requirements for the degree of Master
of Engineering Science.
8. Harlow, E. PL, "Large Rubble-Mound Breakwater Failures," Journal of the
Waterway, Port, Coastal and Ocean Division, ASCE, Vol. 106, No. WW2, 1980,
pp. 275-278.
9. Hsu, J. R. C , Silvester, R., and Tsuchiya, Y., "Boundary-Layer Velocities
and Mass-Transport in Short-Crested Waves," Journal of Fluid Mechanics, Vol.
99, No. 2, 1980, pp. 321-342.
10. Hudson, R. V., "Laboratory Investigation of Rubble-Mound Breakwaters,"
Journal of the Waterway, Port, Coastal and Ocean Division, ASCE, Vol. 85, No.
WW3, 1959, pp. 93-121.
11. Iribarren, R., "Formule pour le Calcul des Diques en Enrochements Naturels
ou Elements Artificiels," 21st International Naval Congress, Stockholm, Sec. II,
1965.
12. Koerner, R. M., and Welsh, J. P., "Fabric Forms Conform to any Shape,"
Concrete Construction, 1980, pp. 401-405.
13. Liu, G. S., "Mortar Sausage Units for Coastal Defense," thesis presented to
the Univ. of Western Australia, at Nedlands, Western Australia, in 1977, in
partial fulfillment of the requirements for the degree of Master of Engineer-
ing Science.
14. Liu, G. S., and Silvester, R., "Sand Sausages for Beach Defense Work," Pro-
ceedings of the 6th Australasian Conference on Hydraulics and Fluid Mechanics,
1977, pp. 340-343.
15. Losada, M. A., and Gimenez-Curto, L. A., "The Joint Effect of the Wave
Height and Period on the Stability of Rubble-Mound Breakwaters using Ir-
ibarren's Number," Journal of Coastal Engineering, Vol. 3, 1979, pp. 77-96.
16. Madsen, P. A., "Wave Reflection from a Vertical Permeable Wave Ab-
sorber," Journal of Coastal Engineering, Vol. 7, 1983, pp. 381-396.
17. Montgomery, D. C. and Dunstan, M. R. H., A Particular use of Flyash in
Concrete-Rolled Concrete Dams, Civil Engineering Transactions, The Institution
of Engineers Australia, Vol. CE23, No. 4, 1981, pp. 227-233.
18. Neville, A. M., Properties of Concrete, Pitman Press, Bath, 1973.
19. Porraz, M., Maza, J. A., and Munoz, M. L., "Low Cost Structures using
Operational Design Systems," Proceedings Coastal Sediments Conference, ASCE,
1977, pp. 672-685.
20. Porraz, M., and Medina, R., Low Cost, Labour Intensive Coastal Develop-
ment Appropriate Technology, Sea Technology, 1977, Aug., pp. 19-24.
21. Porraz, M., and Medina, R., Exchange of Low-Cost Technology between De-
veloping Countries, ECOR General Assembly, Washington DC, 1978.
22. Porraz, M., Maza, J. A., and Medina, R., "Mortar-filled Containers, Lab and
Ocean Experiences," Proceedings Conference on Coastal Structures, ASCE, 1979,
pp. 270-289.
113

23. Production of Cement from Granulated Blast-furnace Slag, Nippon Steel Com-
pany, Japan.
24. Sawaragi, T., Iwata, K., and Kobayashi, M., "Condition and Probability of
Occurrence of Resonance on Steep Slopes of Coastal Structures," Coastal En-
gineering in Japan, Vol. 25, 1982, pp. 75-90.
25. Sawaragi, T., Ryu, C , and Iwata, K., "Considerations of the Destruction
Mechanism of Rubble-Mound Breakwaters due to the Resonance Phenom-
enon," Proceedings of the 8th International Naval Congress, 1983, pp. 3/197-208.
26. Silvester, R., "Wave Reflection at Seawalls and Breakwaters," Proceedings of

the Institution of Civil Engineers, Vol. 51, 1972, pp. 123-131.
27. Silvester, R., "Sediment Transmission across Entrances by Natural Means,"
Proceedings of the 16th Congress, International Association of Hydraulics and Fluid
Mechanics, Vol. I, 1975, pp. 145-156.
28. Silvester, R., "The Role of Wave Reflection in Coastal Processes," Proceedings
of Coastal Sediments 1977, ASCE, 1977, pp. 639-654.
29. Silvester, R., "Design of In-situ Cast Mortar Filled Armor Units of Marine
Structures," Proceedings of the 6th Australian Conference Coastal and Ocean En-
gineering, 1983, pp. 289-292.
30. Smith, A. W. S., and Gordon, A. D., "Large Breakwater Tow Failure," Jour-
nal of Waterway, Port, Coastal and Ocean Engineering Division, ASCE, Vol. 109,
No. 2, 1983, pp. 253-255.
31. Yen, T., Su, C. F., and Chang, M. F., "A possibility of Increased Mortar
Strength for Ferro-Cement," International Conference on Materials Construction
for Developing Countries, Bangkok, 1978, pp. 665-683.
APPENDIX II.—NOTATION
The following symbols are used in this paper:
b = width of cast unit,

V = contact width at base of cast unit,
h = head of mortar mix from bottom of unit when full,
cd = weight of cement in mix,
= depth of water,
H = local incident wave height,
h = height of cast unit,
Kr = reflection coefficient,
kW .= parameter to derive Kr (Fig. 1),
L = local incident wave length,
h = deep-water wave length,
S = circumference of cast unit,
S = weight of ground granulated blast-furnace slag,
SG = specific gravity,
T = •
stress in flexible membrane (submerged),
T = stress in flexible membrane (above SWL),
we = width of equivalent rectangular rubble mass (Fig. 1),
W = weight of water in mix,
X = volume of Dispersenex F in mis per kg of cementing agents,
7 = specific weight of mortar when set, and
e = slope of breakwater face.
114

By Richard Silvester, M. Asce: and Ocean Engineering, Vol. 112, No. 1, January, 1986. ©ASCE, ISSN 0733-950X

Uploaded by

Copyright:

Available Formats

By Richard Silvester, M. Asce: and Ocean Engineering, Vol. 112, No. 1, January, 1986. ©ASCE, ISSN 0733-950X

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

By Richard Silvester, M. Asce: and Ocean Engineering, Vol. 112, No. 1, January, 1986. ©ASCE, ISSN 0733-950X

Uploaded by

Copyright:

Available Formats

USE OF GROUT-FILLED SAUSAGES

ABSTRACT: An alternative to rubble mound construction, with or without pre-

The age-old concept of constructing shoreline structures with rock or

J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

FIG. 1.—Reflection Coefficient Kr for Given We/D and rf/L,

As noted already, scouring of the seabed adjacent to breakwaters from

1. Armor units should be designed which fit snugly together, giving

J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

Porraz et al. (19-22) have pioneered the development of flexible mem-

J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

I Q I r—• V ' I ' ' 1 1 1 1 1 1 1 1 1 1 |03

J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

0.7 0.191 0.25 0.92 0.395 1.45 0.650

FIG. 3.—Plan and Sections of Break- FIG. 4.—Enlargement of Breakwater Tip

J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

0.94 0.875 0.89 0.031 286 499

The suggested spaces would provide channels through which water

J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

FIG. 5.—Horizontal Forces Exerted on Structure Due to Partial Standing Wave

DESIGN OF MORTAR MIX

J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

1. It must have 100% slump, or be a slurry, in order to flow to any

compression (MPa) = 10.4 - 1.92 scour (mm)

J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

Various gradations of sand in mortars have been used (1,18,31), re-

IOO , , % 200 300

J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

tests for the desired slump and other characteristics:

300 day compressive strength = 1,720 psi (11.8 MPa)

It is difficult to compare costs of a flexible membrane structure with

J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

FIG. 12.—Cross Section of Breakwaters Used in Cost Comparison

A protective mat was to be provided for this vertical faced breakwater,

nylon fabric 156,755

J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

One of the great advantages of this constructional concept is the sim-

J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

1. The current inquiry into the stability of breakwater armor units

J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

26. Silvester, R., "Wave Reflection at Seawalls and Breakwaters," Proceedings of

The following symbols are used in this paper:

b = width of cast unit,

J. Waterway, Port, Coastal, Ocean Eng. 1986.112:95-114.

You might also like