Sei sulla pagina 1di 166

This publication is number 86

in the Technical Report Series


of the Industrial Developrtlent Branch
::5.r
~ ; ; J . . 3
C!? -:3 / ~ - . .
:::;:t= J' ~ c - ;r-
FERROCEMENT FOR
CANADIAN FISHING VESSELS
-A Summary and Interpretation
of Test Results-1969 - 1974
prepared by
Arnold W. Greenius
B.C. Research
Vancouver, Canada
for
Vessels and Engineering Division
Industrial Development Branch
Fisheries and Marine Service
Environment Canada
Project Officer
G.M. Sylvester
Division Chief
H.A. Shenker
March 1975
Opinions expressed and conclusions reached
by the author are not necessarily endorsed
by the sponsors of this project
ABSTRACT
This report summarizes, integrates, and interprets the
test results obtained at B.C. Research on the properties of ferrocement
for fishing vessel construction during the five program years, 1969 to
1974. The programs were funded by the Industrial Development Branch,
Fisheries and Marine Service, Environment Canada.
The work has examined many aspects of ferrocement, e.g.
effect of sands, cements, admixtures, reinforcing rods, and reinforcing
meshes on workability, strength, and durability. It has also examined
patching, bolting, painting and other factors affecting performance.
The study is not an in-depth study of any specific
engineering property or performance but is rather a preliminary broad
survey to develop a IIfee1
11
for the engineering properties and behaviour
of ferrocement as a material of construction for fishing vessels.
TABLE OF CONTENTS
A. I NTRODUCTI ON
1. Objectives
2. Historical Background
3. Scope of the Test Work 1969-1974
4. Applicability of the Test Results
B. DIGEST OF TEST RESULTS AND
1. Mortars
2. Reinforcements
3. The Reinforced Mortar Composite
4. Bibliography
C. TEST PROCEDURES AND RESULTS
1. The Component Materials - Mortars
Page
1
1
1
3
4
6
6
8
11
15
16
and Steel Reinforcements 16
2. Evaluation of Various Mortars 21
3. Evaluation of Various Reinforcements 41
4. Evaluation of a Typical Ferrocement Construction 73
5. Bolting Tests 98
6. Design Considerations 101
7. Patching Ferrocement 101
8. Protective Coatings 104
9. Quality Assurance 113
D. BIBLIOGRAPHY OF FERROCEMENT LITERATURE
E. ACKNOWLEDGEMENTS
F. REFERENCES
G. APPENDIX
1. Development of Mathematical Model - J. D. Smith
2. Regulations for Construction of Ferrocement Boats
- Canada Transport Ministry
3. Guidelines for the Construction of Ferro-cement
Vessels - American Bureau of Shipping
4. Bibliography of Ferrocement Literature
114
115
116
122
1
FERROCEMENT FOR CANADIAN FISHING VESSELS:
A SUMMARY AND INTERPRETATION OF TEST RESULTS - 1969 to 1974
A. INTRODUCTION
1. Objectives
Since 1968 the Industrial Development Branch of Fisheries
and Marine Services, Canada has funded a series of programs
at B.C. Research, Vancouver, B.C., to evaluate some of the properties
of ferrocement for fishing vessel construction. The objectives vf this
report are to summarize and interpret the findings obtained and documented
in five major reports during the last five years. This report proides
some measure of the strengths and shortcomings of the test data obtained
from programs designed to cover a wide range of properties and character-
istics at the expense of in-depth completeness.
2. Historical Background
Ferrocement is defined most simply as a thin cement mortar
shell highly reinforced with fine reinforcement.
This definition is general enough to include the most common composites
of steel mesh and rods in a matrix of Portland cement mortar, those which
contain short steel fibres or no reinforcing rods, and those which have
a matrix of polymer mortars. In its present state of practical develop-
ment for fishing vessels, ferrocement is a Portland cement mortar
containing a widely-distributed reinforcement of steel mesh of various
kinds generally with steel rods.
Various have been stated for ferrocement as a
material of boat construction, e.g. good sound insulation, good thermal
insulation, fire-resistance, resistance to marine borers, ease of fabri-
cation, ease of repair, low cost, good abrasion resistance, good impact
strength, good corrosion resistance, and others. Some of these are
undoubtedly questionable.
Ferrocement as a boatbui1ding material has a long history
dating back some 125 years. Although the construction of ferrocement
boats in the years 1848 to 1896 by Lambot, Gabe1lini, and others is
of historical interest, the re-invention and naming of ferrocement and
the design and construction of several boats in the years 1941 to 1949
by P.l. Nervi mark the modern development of ferrocement. After a few
2
years of relative inactivity, a ferrocement renaissance in New Zealand
and England in 1961 started a world-wide adoption of the ferrocement
method of building boats. Although pleasure sailboats were of prime
1rtterest in 1961, China, Cuba, and many developing countries are now
building ferrocement fishing boats, work and harbour boats, and shelter
boats in great numbers.
In 1968, the activities of amateur and professional boat-
builders, especially on Canada1s Pacific Coast, interested the Fisheries
Service, Canada Department of Fisheries. These builders were constructing
both pleasure and work boats. The Industrial Deve1pment Branch of the
Fisheries Service commissioned B.C. Research to undertake a study of
ferrocement as a construction material for fishing vessels. The stated
aims of the first study were: l} to collect and collate the available
data on ferrocement for boatbuilding, and 2} to undertake a limited
test program to determine the properties of conventional ferrocement.
This work was reported in 1968 by Kelly and English
l
to the Fisheries
Service and by Kelly and Mouat
2
to the Conference on Fishing Vessel
Construction Materials, Montreal, Canada, October 1-3, 1968.
As a result of this first program of collecting and
collating SOffie of the available data and of performing some tests, the
Industrial Development Branch embarked on the funding of the first of
five annual programs at B.C. Research to evaluate ferrocement for fishing
vessel construction.
Since ferrocement is a non-homogeneous composite material
of great variety, it was realized that the available fundtng would allow
only a very general assessment of the properties and performance of
ferrocement. It was hoped, however, that patently unsuitable materials
and procedures would be identified and that a set of guidelines to aid
persons wishing to make a f2rrocement boat and persons charged with the
responsibility for certifying ferrocement boats would be developed. In
fact, the five reports
3
-
7
of the work submitted to the Industrial Develop-
ment Branch over the past five years have recorded the results of tests
covering a very broad range of properties. The engineering and perform-
ance data obtained are not the results of in-depth investigations of a
few properties but are rather the preliminary results of tests on a
broad front. The reports were reproduced verbatim by the Branch for
general distribution even though the complete documentation of the
test results diminished the usefulness of the results to a prospective
builder. Therefore, at the end of the fifth program, the Branch
contracted with B.C. Research to summarize and interpret all the test
results obtained so that prospective builders might more readily use
them.
This summary has attempted to consolidate the findings
on a single aspect, e.g. mortar strength, durability, and corrosion,
3
into an integrated section, to interpret the significance of the findings,
and to draw some conclusions. The surrmary has "culled out" or ignored
some results which now appear to be of little significance and has
"gleaned" some useful information, not formerly appreciated, from some
of the tests performed. And finally, it should be stated that this
is not a "How-to-build-it" book, an aspect well covered in several
practical guides.
S
-
14
3. Scope of the Test \ ~ o r k 1969-1974
The programs undertaken in the five annual contracts
1969-1970 to 1973-1974 are briefly summarized as follows:
1969-1970 - The mortar mix - various mixes of sands, cements, and
water/cement ratios were tested in standard compression
tests and tested as tiles reinforced with "standard
ll
mesh reinforcements in bending, impact, and freeze-
thaw and seawater exposure tests.
- The mesh reinforcement - various meshes were tested
in tiles with "standard
ll
mortar mix in bending and
impact.
- Repair - the use of mortar and polymers for patching
was examined.
1970-1971 - The mortar mix - the use of admixtures was examined
for improvements to workability, strength, and durability
of mortar.
- The reinforcement - various kinds of mesh and rods
were evaluated on the basis of mortar penetration,
bond strength, bending and impact strengths.
- The effect of interrupted mortaring on strength was
observed.
- The attachment of fittings by powder-activated tools
was examined.
- The preliminary development of a mathematical model
for the behaviour of ferrocement was considered.
4
1971-1972 - The behaviour of ferrocement under cyclic bending
loads of constant deflection was briefly examined.
- The bolting strength of ferrocement was tested.
- Ceatings for ferrocement were evaluated in laboratory
and marine exposure tests.
1972-1973 - More flexure data were obtained for the future develop-
ment of a mathematical model for the behaviour and
design of ferrocement.
- Additional seawater exposure tests on painted f ~ r r o
cement specimens.
1973-1974 - The effect of crack size and thickness of mortar cover
on the corrosion of the steel reinforcement was assessed.
- The behaviour of ferrocement under cyclic bending loads
of constant magnitude was observed.
- Additional tests on interrupted or multi-stage mortaring
were undertaken.
The programs over the five-year period also involved the
up-dating of a bibliographic listing and maintaining a file of ferrocement
literature at B.C. Research, advice to specific technical enquiries from
persons interested in ferrocement construction, and various miscellaneous
tests to assist experimenters, e.g. especially testing of a polymer-
aggregate composites.
4. Applicability of t ~ e Test Results
None of the annual contracts for work presupposed or
anticipated a subsequent contract. Each year's work was planned to be
complete in itself and to provide information over as broad a range of
variables as possible within the imposed limits of time and funding.
No exhaustive in-depth study of any property was intended or undertaken
and no attempt has been made to obtain an "optimum ferrocement construction".
Materials which have seemed unsuitable in the tests reported
herein may prove acceptable when treated in a different manner. For
example, expanded metal lath appeared unsuitable because of its
anisotropic properties. It was also difficult to get good penetration
5
of the mortar into several layers of the mesh by trowelling.. However,
one manufacturer uses expanded metal lath successfully by embedding it
layer by layer into successive layers of mortar sprayed into a female
mould. The potential innovative builder, therefore, will experiment
with and evaluate the available materials in terms of his proposed
construction techniques.
Although the more common relatively homogeneous materials
are not always adequately characterized by the standard test specimens
and procedures available, a reasonable correlation exists between the
behaviour of a component and the results obtained from a standard test
specimen of the material in the component. However, standard tests and
procedures are not available for ferrocement in its present state of
technical development. Each investigator of ferrocement properties has
used such specimens for strength, flexure, and impact as suits his ferro-
cement construction facilities and techniques and his testing equipment.
It is unwise to try to quantitatively correlate the results presented
herein with those of another investigator. The results are best used to
compare ferrocement constructions in which one variable is changed, ego
kind of mesh reinforcement. Satisfactory replication of tests has not
been possible. The results presented apply only to the ferrocement
constructions used in the tests. Much more work needs to be undertaken
before a mathematical model can be developed which will enable test results
obtained to be reliably applied to other ferrocement constructions.
In conclusion, the results presented herein are not "optimum"
properties but are the results obtained over the past five years from various
preliminary tests, often insufficiently replicated, performed on several
constructions of ferrocement. It is believed that the test results provide
a good "feel" for many of the properties of ferrocement, typical of the
constructions which have been and will be used by the amateur and semi-
professional builder of a ferrocement boat.
6
B. DIGEST OF TEST RESULTS AND CONCLUSIONS
Some of the test results and the conclusions drawn are
briefly presented in this digest. The reader should refer to the pertinent
sections of the main part of the report for the details of the tests performed.
This will help him to judge what degree of confidence should be placed on
the results and conclusions obtained from a test program which has not
always allowed sufficient replication.
1. Mortars
Mortars of several sands, cements, and admixtures have been
examined for the effect of the various components on workability (penetration
of the mesh reinforcement), compressive and flexure strength, absorption,
and durability in freeze-thaw and seawater environments.
The following sands, cements, and admixtures were used in the
tests:
- dry bagged concrete sand (some plus 8-mesh)
- dry bagged mortar sand (no plus 8-mesh)
- sharp Del r1onte silica sands (8-, 20-, and 3D-mesh grades}
- Type I cement (general construction)
- Type II cement (moderate sulphate resistance)
- Type III cement (better early strength)
- Type V cement (superior sulphate resistance)
- aluminous cement (improved strength)
- pozzolan (replaced 1/4 of cement)
- water-reducing agent (6.5 f1.oz/bag cement)
- air-entraining agent (3/4 f1.oz/bag cement)
- Polyviny1acetate emulsion (pva/water, 1:1.44)
(a) Workability
Panels, 30 x 3D-in., were made with several layers of mesh.
Mortars of various combinations of sands, cements, and admixtures were
placed with vibrating trowels into the horizontal form moulds containing
the mesh layers. The ease of placing the mortar and the penetration into
the mesh layers were evaluated qualitatively. All mortars tested gave
adequate penetration provided the cone slump value was greater than 3 in.
The mortars containing sands with some plus 8-mesh material were more
difficult to trowel smoothly. The various cements and admixtures appeared
to g i v ~ no significant improvements.
7
The 36-in. panels made in the upright open frame moulds
were somewhat more difficult to mortar. The vibrating trowel could
not be used. Higher slumps (3 to 6 in.) were necessary for good penetration.
(b) Strength
Compressive strengths of the various mortars were obtained
from 7-day and 28-day 2-in. cube specimens. The flexure strengths were
obtained from unreinforced portions of the panels made. No important
differences in strength were observed in the mortars made with the several
sands, cements, and admixtures.
(c) Absorption
Specimens were soaked in water and subsequently dried
to determine the voids. They absorbed about 7 percent water. The differences
in water absorption for mortars of various sands and cements were not
significant. The water-reducing agent appeared to reduce the absorption
to about 4 percent.
(d) The Hydrogen Gas Problem
Gas bubbles in the form of blisters were encountered on
panels which contained both rods and galvanized mesh. A galvanic reaction
between the two kinds of reinforcement in the presence of the cement
mortar causes hydrogen to evolve. Chromium trioxide was added to the
mortar to prevent the reaction.
(e) Durability
Unreinforced mortar coupons were subjected to 350 freeze-
thaw cycles, 10C to -4C. The kind of cements and sands used had no
apparent effect on the freeze-thaw durability. No significant deterioration
was observed after 350 cycles. Admixtures did not improve the durability.
The upper unreinforced layer of coupons reinforced with wire mesh delaminated
or disintegrated badly in a test of 76 cycles of freezing and thawing. It is
concluded that differential expansion between the reinforced and unreinforced
layers of the test coupons aggravated the stresses from freezing and thawing.
It is also concluded that thick layers of unreinforced mortar padding in
a boat hull should be avoided.
Mesh-reinforced coupons with various sands, cements and
admixtures were tested in 350 seawater cycles. No changes in
visual appearance and hardness of the mortar were observed. It is con-
cluded that all have good resistance to seawater.
8
(f) Conclusions
The tests of mortars with various sands, cements, and
admixtures showed, in general, no real advantage for one type of sand or
cement over another and no real improvement from the addition of admixtures.
It appeared best to avoid coarse sands, to use a cement with at least some
sulphate resistance, and to avoid the use of admixtures. It was therefore
decided to use a "standard" mortar for all subsequent tests as follows:
- Dry bagged mortar sand (good tro\'/elling) 2 parts
- Type II cement (moderate sulphate resistance) 1 part
- Water/cement ratio, 0.4 to 0.5 (as close to 0.4 as
possible with acceptable slump values 3 to 6 in .. )
- Chromium trioxide (Cr03) 300 ppm of water was added
to prevent hydrogen gas problem when galvanized mesh used
with ungalvanized rod reinforcement.
The following average compressive and flexural strengths
were obtained for the many "standard" mortar batches (Type II cement, dry
mortar sand) over the several years:
Comrressive Strength:
ASTMC109, 2-in. cubes
7-day
28-day
5670 psi
7840 psi
ASTMC349, 1.575-in. prisms
28-day 10,070 psi
Flexural Strength:
ASTM C348, 1.575-in. prisms
28-day 1360 psi
2. Reinforcements
(a) Strength
Several mesh and rod reinforcements have been tested
alone and as components in a matrix of "standard" mortar, i.e. in
a ferrocement composite.
The following tensile strengths of the six kinds of rods
used were obtained:
hot-rolled black C1020
galvanized C1020
bright drawn C1010 nail wire
double-drawn high-tensile A82
double-drawn extra high-strength A82
deformed double-drawn A82
70,000 psi
50,000 psi
73,000 psi
79,000 psi
101,000 psi
84,000 psi
9
It was observed that a "cross weld" with minimum heat imput
did not seriously impair the strength of the double-drawn wire although it
is possible that "kinking" at the joint may occur in hull areas with greater
curvature.
The following breaking strengths of several of the meshes
used were obtained:
1/2-in. 16-ga. galv. welded square mesh
1/2-in. 19-9a. galv. hardware cloth
3/8-in. 20-ga. black welded square mesh
1/2-in. 22-ga. galv. hexagonal mesh
- longitudinal direction
- transverse direction
2.5 1b expanded metal
1/4-in. coil 20-ga. fire screening
400 lb/in. width
140 1b/in. width
154 lb/in. width
60 lb/in. width
20 lb/in. width
not done
not done
Panels containing mesh only were tested in flexure. The
results indicated tbat the flexure strength is essentially proportional to
the weight of mesh per unit area of the panel, regardless of kind of mesh.
This is not true for a mesh tested in its weaker direction, e.g. hexagonal
mesh and expanded metal tested in transverse direction.
(b) Bond Strength
The bond strength of rods was determined from "hair pin" rod
specimens encased in mortar. Black hot-rolled, pickled hot-rolled, and
cleaned and lightly rusted double-drawn high-tensile rods gave similar
bond strengths. Bright rods, especially with the drawing lubricant (or rust
preventing), and galvanized rods gave lower bond strengths even after 4 1/2 months.
The deformed douole-drawn rod had high bond strength but split the mortar
coupons.
No formal bond tests were performed on mesh. The locking or
keying of twisted and galvanized mesh and the \'/elded joints ir. welded square
meshes overrides any tensile/bond differences.
10
(c) Specific Surface
The specific surface of the several meshes used have been
determined. The values calculated for the lengthwise wires only are
3.26 1n.-
1
for 1/2-in. 16-ga. ga1v. welded square mesh, 3.35 in.-
1
for
1/2-1n. 19-9a. ga1v. h a r d w a r ~ cloth, 4.17 in.-
1
for 1/2-in. 22-ga. galv.
hexagonal mesh, and 4.67 in. 1 for 3/8-in. 20-ga. black welded square
mesh, all with 1/16 in. mortar cover.
(d) Conformability and Costs
Other factors have been briefly considered. In general,
meshes of woven and twisted wires are more conformable to compOlmd
curvature than are welded square meshes. The difference in cost for
equal weights of the various meshes will not be an overriding factor
in the choice of mesh reinforcement.
(e) -Corros i on
Corrosion coupons cut from various test panels and specimens
damaged in flexure tests were subjected to seawater exposure. These coupons
and specimens contained galvanized reinforcements. The corrosion attack
of the mesh was qualitatively assessed.
Shaped coupons containing a single ungalvanized bare wire
embedded 0.5, 2.0, and 3.5 mm below the mortar surface were prepared. The
mortar coupons were stressed to provide controlled cracks in the mortar
0,0.05,0.1,0.2, and 0.5 mm wide. Triplicate sets of each combination
of mortar cover and crack size were immersed at mean tide at the Vancouver
Kitsilano Base of the Canadian Coast Guard. A triplicate set of each
combination was removed at 1, 2,4, and 8 months. (A set for 16 months
exposure is still exposed ct the time of writing).
The coupons were examined visually, the loss of wire section was
measured with a micrometer, and the loss of wire strength was obtained from
tensile tests. The following observations were made:
- corrosion after 8 months exposure was not severe in wires
with a crack <0.1 mm wide even with only 0.5 mm mortar cover
- corrosion after only 1 to 2 months exposure was severe if the
crack width was >O.lmm regardless of the thickness of the mortar
cover.
- a mortar cover of 2 mm protected the wire in uncracked coupons
from corrosion for at least eight months.
a mortar cover of 0.5 mm failed to protect the wire in uncracked
coupons from corrosion for over two months.
11
It is concluded that cracks >0.1 mm cannot be tolerated
and that a mortar cover of at least 2 mm should be used. It seems advisable
to use galvanized mesh for marine applications of ferrocement.
3. The Reinforced Mortar Composite
(a) Effect of Rods and Meshes on Strength
Tests were performed to assess the contribution of the kind
and spacing of rods to the flexure and impact strengths of panels of the
ferrocement composite. The presence of rods is very important, as is their
spacing. However, the kind of rods used, viz. hot-rolled, galvanized, or
double-drawn high-tensile rods appears to have only a small ef&ect on the
strength of the panels. Other considerations such as "kinkingll over stiffener
nodes and attachment to stern and bow may be of some importance.
Tests \'Jere performed to compare the f1 exure and impact
strengths of ferrocement panels containing approximately equal thicknesses
of the several kinds of mesh reinforcement. The flexure strength of the
specimens containing 1/2-in. 16-oa. welded square mesh was superior to
that of specimens containing gauge welded square meshes (1/2-in.
22-ga. and 3/8-in. 20-ga.) and markedly superior to that of specimens containing
the 1/2-in. 22-ga. hexagonal mesh. It should be mentioned, however, that
differences were almost eliminated when rods were also present.
The drop-impact performance of panels reinforced with
1/2-1n. 22-ga. hexagonal mesh when rods were also present was still
somewhat poorer than that of comparable panels of the welded square meshes.
The argument for using 1/2-in. welded square meshes in
preference to 1/2-1n. hexagonal mesh (when rods are present) is not strong.
Conformabi1ity over curved portions of the hull and fairness (waffle-effect
of mesh pressed between rods) may be the overriding consideration. High-
strength woven square meshes, extolled by some authorities, have not been
tested.
It was also observed that the flexure strengths of specimens
with the rods, mesh, or both oriented at 45 degrees to the specimen length
were only 50 to 90 percent (depending on the particular configurations) of
the flexure strengths of specimens with normal orientations.
(b) Strengths of Typical Ferrocement Constructions
Typical ferrocements have been examined in this study in somewhat
greater detail. The "typical" ferrocements considered here had a mortar
12
matrix of Type II cement and dry mortar sand reinforced with 0.225-in. double-drawn
high-tensile A-S2 rods spaced at 2-in. centres and two layers each side
(Panel 207) or three layers each side (Panel 20S) of 1/2-in. 19-9a. galvanized
hardware cloth.
A single tensile specimen prepared from each panel was tested
but each specimen held a load only equal to the sum of combined strength of
the mesh wires and the rod/mortar bond strength. (Tensile test specimens
from panels containing rods do not adequately represent the strength of a
"semi-infinite" ferrocement plate because of the lack of rod. anchoring).
The following tensile strengths and the values of elastic modulus
obtained were obtained: 207 20S
No. of layers of mesh
No. of wires in section
Area of section, sq.in.
Load at first visible crack, lb.
Max. load held, lb.
Modulus of elasticity, psi x 10
6
4
16
2
1300
1420
1.1
6
23
2
1450
1900
0.9
Specimens, 1 x 6 x 18, 24, and 36 in., were tested in third-
point flexure on an lS-in. span. A modified extensometer measured the
tensile strain in the centre portion of the span between the third-point
loads. The beam deflections at the third-points and at mid-length 'were
measured. The location of the mesh wires and the rods relative to the
specimen beam surface were obtained at the third-points and mid-length
for future work.
The flexure strengths at load of first visible crack, P
fvc
at P
fvc
f2, and at maximum load held (modulus of rupture) and the elastic mOdulus
values from beam curvature and elastic formulas at the load of first visible
crack, Pfvc' and at Pfvcf2 were calculated for six specimens from each plate.
The following average values for flexure strengths and elastic
modulus were obtained:
207 20S
No. of layers of mesh 4 6
Flexure strength, psi, at Pfvc 2600 2900
at PpVCf2
l300 1450
Modulus of rupture, psi at max 3700 4400
Elastic modulus, psi x 10
6
E = ~ M at Pfvc
1.5 1.6
at Pfvcf2
2.3 2.7

E = at Pfvc
at Pfvct2
(c) Behaviour Under Repeated Flexure Loads
0.9
0.7
1.3
1.3
Tests under cyclic flexure loadings were performed on
4 x 12 x 1 in. specimens from several panels. The first tests used
13
an apparatus which loaded the specimens to constant deflection. The
later tests used an apparatus which loaded the specimens in third-point
loading to constant magnitude.
Specimens from a panel containing three layers of 1/2-in. 19-9a.
galvanized hardware cloth on each side of 0.225 in rods (one of the "typical"
constructions already described) were loaded with repeated unilateral third-
point bending loads of 2500 psi to 1320 psi fibre stress. The specimen at
1320 psi fibre stress withstood 500,000 cycles without visible cracking.
The work needs greater replication and tests at lower fibre
stress levels to obtain an endurance limit. A true endurance limit may not
exist for (as for many other materials) but the values obtained
give a "feel
ll
for the probable behaviour of "typical
ll
ferrocement construction
under repeated bending loads.
The fracture surfaces of mesh wires in the specimens broken
under a simple single bending load and under the repeated loadings were
examined by means of the scanning electron microscope. The photomicro-
graph of the single load fracture shows the "dimpled" appearance typical of a
tensile failure \'/hereas that of the repeated load fracture shows the "striated
ll
appearance typical of a bending fatigue failure.
(d) Bolted Strength
Preliminary tests were undertaken to determine the bolted
strength of ferrocement. Specimens 4 and 8 in. wide were cut from one-inch
thick panels with 0.225-in. dia. rods spaced at 2 in. and 2 layers of
1/2-in. 16-ga. welded square mesh, 5 layers of 1/2-in. 19-9a. hardware
cloth, or 10 layers of 1/2-in. 22-ga. hexagonal mesh. The 4-in. wide
specimens carried a co-planer load, transmitted by a single 1/2-in. bolt,
of about 500 lb/in. width when the single bolt hole was 2 in. from the
free edge and the "equivalent bolt hole pitch" was 4 in. A reinforced edge
and a greater distance between bolt hole and free edge would allow the bolt
joint to possibly fail in compressive bearing.
14
(e) Design Considerations
The use of the mechanical properties developed in the program
for design has not been pursued. Readers are referred to the excellent
review - An Introduction to Design for Ferrocement Vessels by G.W. Bigg
for the Industrial Development Branch, Fisheries Service, Canada Department
of the Environment in 1972.
The preliminary mathematical model prepared by J.D. Smith
early in the program is presented in the Appendix.
(f) Patching
Patching of ferrocement in which the mesh reinforcement
has not been broken has been shown to give virtually the original flexure
strength with either cement/sand mortars or with filled epoxy-resin patching
compounds. Ferrocement with some broken mesh regained 50 to 80 percent
of its original strength. A pneumatic chipping and scaling
tool was effective in breaking up most of the loose mortar debris and in
opening up cracks without seriously damaging the mesh. The behaviour of
the patch under weathering and conditions of thermal expansion has not been
ascertained.
(g) Protective Coatings
The question of coatings has received considerable attention.
Twelve coating systems of various materials for primers and topcoats were
subjected to dry laboratory environment (control), Weather-Ometer, laboratory
wet-dry cycling in and marine tidal exposures.
The most satisfactory performance was obtained from the
following systems:
- Primer - inorganic ethyl-silicate zinc-rich paint
- Topcoat - same
- Primer - two-component clear epoxy finish
- Topcoat - two coats of vinyl resin-base anti-fouling paint
- Primer - two-component pigmented epoxy resin
- Topcoat - same
- Primer - po1yviny1ch1oride-based enamel
- Topcoat - two coats of same.
15
(h) Quality Assurance
Attention is drawn to testing of the components of the ferrocement
composite and of the composite itself. The importance of the cement/sand
and water/cement ratios, slump, compressive and flexure tests, strength of
the reinforcement materials, reinforcements and layup, and mortar placement
procedures is stressed. Testing of ferrocement specimens must realistically
represent the properties of the hull. Non-destructive testing of the hull
is not well developed. Tentative guidelines for construction of ferrocement
hulls have been issued by the certifying inspection bodies, viz. Canada
Transport Ministry, Lloyd's Register of Shipping, and American Bureau of
Shipping.
4. Bibliography
The bibliographic listing of articles pertinent to ferrocement
vessel construction has been maintained and is presented in the Appendix.
16
c. TEST PROCEDURES AND RESULTS
1. The Component a 1 s - and Steel Rei nforcements
(a) General
The two main components of the ferrocement composite are
the mortar and the steel reinforcement. The steel reinforcement,
generally rods and wire mesh, provides the skeleton on which the mortar
is applied. It resists the tensile loadings. The function of the
mortar in a ferrocement vessel is to keep water out and to provide
compressive strength to the mortar/mesh composite. In addition,
mortar must resist chemical attack by seawater and disintegration
from freezing and thawing and other forces. The mortars and reinforce-
ments have been treated singly and jointly throughout the program.
(b) The Mortar
The mortar, as generally used in ferrocement construction,
is a mixture of Portland cement, sand, and water, with or without one or
more admixtures added to give special properties. It is generally agreed
that the properties of mortar are related to the materials used, e.g.
types of sand and cement; to the proportions of the components, e.g.
cement/sand and water cement ratios; and to other variables, e.g. curing
conditions. The number of possible combinations is great. It was
necessary to limit the number of tests by maintaining some variables
constant. For example. it was assumed that many amateur and semi-
professional builders would not be able to adequately steam-cure a newly
plastered boat. Effective steam-curing requires that the surrounding
atmosphere be raised to 130F for at least 48 hours. Most builders will
cover their hull with polyethylene sheeting and hose it down regularly
for about a month. Our test panels were done in this manner. r10rtar
test specimens were wetted, wrapped in paper towels. placed in a plastic
bag for testing after 7 days or 28 days. A cement/sand ratio of 1:2,
a ratio commonly used by amateur and semi-professional builders, was
used throughout the programs. The water/cement ratio was held as close
to 0.4 as possible since a low water/cement ratio favours high strength.
The water/cement ratio was adjusted to allow penetration of the mortar
through the several layers of mesh reinforcement. The mortar variations
assessed were restricted to several types of sands and cements and to
a few admixtures. The mortars have been evaluated as an unreinforced
material and as a reinforced mortar/steel composite.
17
The sand for Portland cement mortars should be clean,
strong, hard, sharp, durable, and well-graded. However, tests
reported by the widely distributed and accepted Concrete Manual of the
Bureau of Reclamation U.S. Department of the Interior have shown that
changes in sand grading over an extreme range have no material effect
on compressive strength of mortar when the water-cement ratio and the
slump are held constant. It suggests that a sand of the size shown in
Table 1 should be suitable for corrosion-resistant mortar linings for
steel pipes.
Tests were performed to determine any significant differ-
ence in the workability,strength, and durability of mortars for ferro-
cement constructi on when vari Ol.'S sands were used. The sands exami ned
were a dry bagged concrete sand, a dry bagged mortar sand, and a blend
of three sizes of sharp Del t10nte silica sand.
Portland cements are produced in five major types (Types
I to V) for use under specific conditions. The types differ in their
proportions of the four main chemical compounds, namely tricalcium
silicate, dicalcium silicate, tricalcium aluminate, and tetracalcium
alumino-ferrite.
Type I cement is used in general construction work when
special properties are not required. Type I, with its high tricalcium
aluminate content, is not resistant to sulphate attack. Type II cement
is used where moderate sulphate conditions are present. Its relatively
low tricalcium silicate content provides good resistance to sulphate
solutions. Type III cement is used where the strength must be developed
rapidly. Type IV generates less heat then the other types and at a
lesser rate. It is used in massive structures. Type V ceme'nt is used
for c o n c r e t ~ in contact with soils and waters of high sulphate content.
The low tricalcium aluminate and tetracalcium alumino-ferrite contents
give much greater resistance to sulphates.
All types are reported to have about equal strengths after
two months but Types II, IV and V appear to surpass Types I and III after
several years. Another cement, kno\'Jn as aluminous cement, has monocalcium
aluminate as its chief component. It is strong, has good resistance to
seawater, but it appears to suffer catastrophic disintegration in some
humid climates. A description and the requirements of the several
Portland Cements may be found in ASTM C150.
18
Tests were performed to determine significant differences
in the performance of ferrocement mortars made with the several cements.
Admixtures fulfill an important function in modern concrete
technology. They can be of benefit \.,rhen used as air-entraining agents,
set-retarders, accelerators, water-reducing agents, and workability
improvers. The Division of Building Research, National Research Council,
Canada has pointed out that any benefits are contingent on proper use and
a knowledge of any possible harmful side-effects. The admixtures are
often proprietary and specific formulations may not be generally avail-
able across the country.
It was decided ~ h a t unless admixtures showed a substantial
improvement in the mortar it was best to avoid their use. A pozzolan,
a water-reducing agent, an air-entraining agent, and a polyvinylacetate
emulsion were evaluated in a limited test program.
Pozzolans are described as siliceous or siliceous and
aluminous materials, possessing little or no cementitious value,
which, in finely divided form and in the presence of moisture,
will react chemically with calcium hydroxide at ordinary temp-
eratures to form compounds with cementitious properties in
concrete. No recommendations for the optimum proportions of
pozzolans in mortar have been found.
Water-reducing agents are considered to lower the
mix water requirement which results in an increased compressive
strength. It should be possible to reduce the cement content
for a given strength. The basic ingredient of the common
water-reducing admixtures is either salts of lignosulphonic
acid or hydroxycarboxylic acids. The water-reducing agent
used in this study is described as nan aqueous solution of
metallic salts of lignin sulfonic acids which contains a
catalyst to counteract the hydration-retarding actionn.
Air-entraining agents have a reputation for producing
concretes which can resist damage by frequent wetting and by
cycles of freezing and thawing. ~ 4 a n y specifications require
concretes which contain about six percent entrained air. The
agents are generally formulated from wood resins, sulphonated
hydrocarbons, and synthetic detergents. The agent used in
this study is described as lIan aqueous solution of purified
and modified triethylamine salts of a sulfonated hydrocarbon
and which contains a catalyst to promote more rapid and
complete hydration of the Portland cement
ll

Polyvinylacetate emulsions have been used to toughen
concrete. The added polymer, in the right proportion, fills
the voids in the cement gel structure and exerts a binding
effect on this gel structure. The air spaces, formed by air
entrainment, are not filled. A 1:5 emulsion/cement ratio is
reported to improve the workability of concrete mixes and to
allow a lower water/cement ratio to be used. The cost makes
it a relatively expensive addition.
(c) The Reinforcement
19
Steel reinforcer.:ent materials in a \'/ide variety of rein-
forcements have been used by builders of ferrocement boats. Each has
undoubted advantages and disadvantages. One kind or combination of
reinforcement may be acceptable for one method of construction and
unacceptable for another. One part of the hull may require a reinforce-
ment which can conform readily to a compound curve. Another part may
accept a stiffer kind of reinforcement.
The general mode of constructing ferrocement boats is to
build a "skeleton" of steel rods on which layers of mesh are laid. The
longitudinals (stringers or horizontals) ~ x t e n d from the stem to the
stern. Transversals or diagonals mayor may not be used, depending on
the size and design of the boat. Several layers of steel mesh are laid
on each side of the rods. However, some boats, incorporating closely
spaced ribs, have been built without rods. Some builders believe that
the main purpose of rods is to act as spacers to maintain the mesh in
the outer portions of the hull section thereby increasing the section
modulus. Some believe that rods greatly add to the hull weight without
contributing much to its strength. The size and shape of the vessel
are of undoubted prime importance in deciding the need for and spacing
of rods.
Steel rods of several kinds are available for ferrocement
boat construction. Their strength, surface finish, protective coating,
size and other factors affect their performance as reinforcing members
of the composite. Rods of about 1/4 in. diameter are most commonly
used. Hulls with longitudinal and transverse rods of this size are
about one inch thick. Smaller sizes have been used, especially for
transversals. One general requirement is that a steel rod should not
kink where it passes over a frame since it will result in a hull shape
which must b ~ faired by mortar padding.
20
The following readily available kinds of reinforcing rod
material were obtained for testing purposes:
hot-rolled black (C1020) 1/4-in. rounds
galvanized (C1020) 1/4-in. rounds
br1ght drawn (C1010) 1/4-in. nail wire
double-drawn, high-tensile ASTt1 A82 0.225-in. rod
double-drawn, high-tensile ASTI1 A82 0.225-in. rod,
extra high strength
double-drawn, high tensile ASTM A82 0.225-in. rod,
defonned by passing through dimpling rolls.
The kinds of IIfinely divided steel reinforcement
ll
are
~ l m o s t limitless. The reinforcement may be steel rods of many kinds,
chopped wire or fibre, slitted and expanded mesh, woven or welded mesh.
The material may be of various mesh sizes and geometrics, e.g. hexagonal
mesh, and link mesh. The wires may be hard drawn, oil-tempered, annealed,
bright,black, galvanized, plated or coppered.
Many kinds of IImeshes
ll
have been used by tne ferrocement
boatbu11der'. Many, such as welded wire fabric and hardware cloth, hex-
agonal mesh IIchicken
ll
wire, expanded metal lath, and woven screening
are readily available in many gauges, opening sizes, and strengths.
Others are specialty items, such as IIthree-dimensional
ll
It/ire IIplanksll.
In pra'ctice, most boats have been built with a hexagonal mesh, welded
square mesh, hardware cloth, or combinations of these. Expanded metal
lath has been successfully used, at least in part of the structure, by
Fibersteel Company and its licensees. High tensile woven mesh has been
successfully used by the U.S. Department of the Navy.IS
It became evident in the early work of this program that
only a few of the several varieties of mesh available locally could be
examined. Various tests were undertaken with a view to ruling out
patently unsuitable kinds. The following kinds, mainly 1/2-in. mesh,
were chosen for examination:
25 gao 2.5 lb/sq. yd. expanded metal lath,
galvanized before slitting and expanding.
1/2-in. 16-ga. welded square mesh, galvanized
after welding.
1/2-1n. 16-ga. welded square mesh, as above with
zinc coat removed.
1/2-1n. 22-ga. hexagonal mesh, galvanized before
weaving - light coat.
1/2-in. 22-ga. hexagonal mesh, galvanized after
weaving.
1/2-1n. 19-9a. welded galvanized hardware cloth.
1/4-1n. coil 19-9a. black and oiled fire screening
(curtain) .
3/8-1n. 20-ga. welded ungalvanized square mesh.
21
A few other meshes, only slightly different from those
listed, were also examined in a preliminary way but are not worthy of
record. It was not possible to purchase loose-woven 1/2-in. mesh
screening locally at the time of the tests. Some high tensile drawn
and oil-tempered woven screens were available but these seemed too
stiff and too expensive to warrant testing at the time.
2. Evaluation of Various Mortars
(a) General Procedure
Tests were performed to evaluate the workability, compres-
sive strength, flexure strength, impact strength, porosity (permeability),
and durability (resistance to disintegration in freeze-thaw and seawater
environments) of mortars made with several kinds of sand, cement, and
admixtures. The mortar was tested alone and in combination with steel
reinforcement as ferrocement test panels.
All batches of ferrocement mortars were prepared in a
horizontal-arm mixer which was able to mix 120 lb. of dry cement and
sand without difficulty.
All batches had the following proportions, viz. cement/
sand ratio 0.5, water/cement ratio 0.4 to 0.5. One or more slump tests
(ASTM C143) were made from each batch. In general, the slumps ranged from
3 to 5 in. Compressive strength test specimens, 2-in. cubes (ASTt1 C109),
were made from each batch. In later work, compressive strengths
C349) were also determined from broken flexure prism specimens, 1.575 x
1.575 x 6.3 in., cast in steel moulds. Flexure strengths of the mortars
were determined from unreinforced portions of the test panels and, in
later work, from flexure prisms, 1.575 x 1.575 x 6.3 in. (ASTt1 C348).
Curing times of 7 and 28 days were used. The compressive strengths
obtained from the broken flexure prisms were generally higher than those
from the 2-in cubes. This is considered to be partly due to better
geometry of the moulds used.
The first test panels were made in a horizontal 30 x 3D-in.
plywood mould lined with 5 mil plastic sheeting. Later panels were made
in an open 36 x 36-in. frame mould in an upright position to nearly
simulate conditions encountered in. mortaring a hull. The mould and mould
frame are shown in Figs. 1 and 2.
Test panels, 30 x 30 in., containing 12 layers of 1/2-in.
22-ga. galvanized hexagonal mesh, cut 27 x 30 in. to leave a three-in.
unreinforced portion along one edge, were made with various sands, cements,
and admixtures.
22
Fig. 1. Mortar placement with vibrating trowel
in horizontal 30-in. form mould.
m
,::: 'rl
IF
c ;;::

c::
1.
'''':

t'
":;
,;;J, .
:; -
f"I

,-<. I..:;

1= t=

I"" t""E;
F

1-- I:f.' I'"' b
fF=
t:;:
,-
,- P' Ii:. H r ... r-l' I:":
"S' I'"' f-
" . 1
I ...
; - F'-
c
I.:;;.

E:
I';
t,::;.!S
'),

E' !im,

I;;:

if-


p.


h-
I,;;
1- t;;:; H

-
Fig. 2. Vertical 36-in. open frame mould.
The following materials were used:
dry bagged concrete sand,
dry bagged mortar sand
sharp Del r10nte silica sand (8-, 20-, 3D-mesh)
Type I 'Portland cement
Type II Portland cement
Type III Portland cement
Type V Portland cement
aluminous cement
water-reducing agent
air-entraining agent
pozzolan
polyvinyl acetate emulsion
(b) Morkability
23
The workability of a mortar is commonly measured by the
cone slump test C143) and the flow table test (ASTM C124).
Although such tests do provide a measure of the workability, the tests
must be correlated to the application. Only the slump test was used
in this study. The critical measure of the workability for ferrocement,
however, is the ability to fully penetrate mortar into the several layers
of wire mesh with a trowel.
Preliminary tests showed that a minimum slump of 3 inches
was necessary for penetration into 12 layers of hexagonal mesh in a form
mould. Mixes with cone slumps of only 3 inches were very stiff. These
mixes readily flowed into the horizontal 30 x 3D-in. moulds when a
vibrating trowel was used. However, it was impossible to obtain complete
of the 12 layers of 1/2-in. 22-ga. hexagonal mesh by hand
trowelling.
In later work, in which panels were made in a 36 x 36-;n.
upright open frame mould to more nearly duplicate conditions of hull
mortaring, it was not possible to use the vibrating trowel. The slump
of the mortar was much more critical. Slump values of 3 to 6 inches
were required.
24
In the first series of tests 26 30 x 3D-in. panels were
made using various meshes, types of cements, and types of sand. Twelve
of these panels contained a "s.tandard" reinforcement of 12 layers of
1/2-in. 22-ga. galvanized woven hexagonal mesh 27 x 30 in. (A 3-in.
unreinforced portion was left on one side of the panel.) Various cements,
viz. Types I, II, III, V, and aluminous, and three sands, viz. dry bagged
concrete sand, dry bagged mortar sand, and Del f10nte silica sand (8-,
20-, 3D-mesh) were used. The water/cement ratios ranged mainly from
0.4 to 0.5 and the slumps from 5 to 6 1/2. Compressive strength specimens
of 2 ~ i n . cubes, were made for 7- and 28-day tests.
The mortar workability, as measured by the difficulty of
penetrating the mesh, appears 'to be not significantly affected by the
type of cements or sands used ?lthough the coarser and poorly graded
sand gave some difficulty in finish trowelling. The comparisons are
shown in Table 2.
The effect of admixtures on workability was assessed by
ten 30 x 3D-in panels reinforced with 12 layers of 1/2-in. 22-ga. galvan-
ized hexagonal mesh, 27 x 3D-in. The mortar mixes for the panels contained
1 part Type II cement, 2 parts dry bagged mortar sand, a water/cement ratio
of 0.4 to 0.45.
One of the following admixtures was added to each pair
of mortar batches: pozzolan, 1/4 of cement replaced; water-reducing
agent, 6.5 fl. oz./bag cement; air-entraining agent, 3/4 fl. oz./bag
cement; and polyvinyl acetate emulsion, 1/2 of water replaced. One pair
contained no addition of an admixture. Test cubes for 7- and 28-day
compressive strengths were prepared. The construction details, admixtures,
and a description of the ease of trowelling the mortars into the panel
mesh are provided in Table 3.
The tests did not show any significant improvement in \'lOrk-
ability. The optimum proportions of the admixture in the mortars may not
have been realized. Additional work is needed to affirm or refute the
benefits of admixtures on workability.
The workability of mortar into the mesh has been discussed
in terms of the sands, cements, and admixtures. The kind of mesh being
penetrated also influences the apparent workability. Five panels of
various meshes, viz.: 7 layers of 3/8-in. 20-ga. welded square mesh,
12 layers of 1/2-in. 22-ga. galvanized hexagonal mesh, and 2 layers of
1/4-in. 20-ga. black fire-screening, were made. Although some low-slump
mortars used in some early tests did not penetrate well into one kind
of mesh reinforcement, mortar penetrated all the meshes without great
difficulty. Table 4 compares the workability of mortar into the various
mesh constr.uctions outlined above.
u.s.
Standard
Sieve
4
8
16
3D
50
100
Pan
TABLE 1. Typical Analyses of Sands Used in Ferrocement Tests.
(weight retained on sieve. percent)
U.S. Bureau Dry Bagged Dry Bagged
Del Monte Silica Sand
of Reclam- Concrete Mortar
ation * Sand Sand 8-mesh 20-mesh 30-mesh
0 1
- - -
0-5 11

1
.
-
10-20 22 8 73
.
-
20-30 22 26 26 43 1
25-40 29 39
-
52 61
15-20 11 20
.
5 28
3-7 4 7
- -
10
Fineness Modulus 2.84 2.08 3.75 2.38 1.53
* Recommended for corrosion-resistant mortar linings for steel pipes
.. Proportions of a-mesh. 20-mesh. 30-mesh Del Monte silica sand
25
1 :2:1**
.
-
18
28
42
10
2
2.50
TABLE 2. Workability and strength of Mortar with Various Cements. Sands. and Slumps.
(Panels contain 12 layers of 1/2 in. 22-ga. galvanized hexagonal mesh in a horizontal form mould.)
Panel Cement
Waterl
Slull'p
Water Con;pressive Modulus of
Sand Cement Workability Absorption Strength. psi Rupture. psi
No. Type
Ratio
in:
S (unreinforced)
7-dav 28-dav
5 II Mortar 0.40
1lt EISY 7.2 1125
. penetratfon 1100
6 II
0.40 SIt
6.0 1'42
890
14 II
0.40 5Is
7.0 6200 7500 890
.
72SO
.
16 II
0.40 6It
6.6 5830 9450 760
9875 820
17 I
0.40 5 7.3 5040 5875 10SO
6175 960
18 III
0.45 4'1 Some
.
5400 8325 1320
difficulty 7750 1280
19 III
0.47 5 Easy
-
5 1 ~ 8450 590
penetration 7750 560
22 III
0.45 5 Some
.
8000 7000
-
difficulty 7800
-
23 Y
0.41 5Ja
Easy
-
7575 10000 932
penetration 10700 843
24 Aluminous

0.36 4'1
.

-
91 SO 7250 1028
6280 792
25 II Del 0.40 lit
Tears. some
-
5700 6390 635
Monte difficulty 5780 782
26 I Mortar 0.41 lit


-
68SO 8110 855
94SO 738
26
TABLE 3. Effect of Admixtures on Workability. Absorption. and Strength of Mortars.
(MOrtar - Type II cement. dry mortar sand. cement/sand ratio 1:2)
(Panels reinforced with 12 layers 1/2-in. 22-ga. galv./hexagonal mesh.)
Panel Water/
Water Compressive Modulus of
Ad111xture
No. Cement
Slump, Workability Absorption strength. psi Rupture. psi
raUo
in. S
7-dav
(unreinforced)
1I0ne 31 0.40 21s Stiff mixes 5.2 5900 6950 1120
Penetration 5800 930
32 0.40 2 difficult 4300 6450
mo..
5100 6355 1025
Pozzolan replaced 33 0.42 21s Stiff mix. Penetra- 5.6 4550 7200 840
25 pe"cent of tion difficult. 7850 680
cement 34 0.45 lis
More workable. 3575 6225
4975
4060 6562 760
Water-reducing 35 0.35 2 Stiff mix. Penetra- 3.9 7725 5500 1100
agent, 6.5 fl.oz. Uon difficult. 4700 1015
per bag cement
I
36 0.37 3 More workable. 5825 5875
6775
5875
1055 5485
Afr-entrainfng 37 0.40 3 Good workability 6.7 4900 6100 835
agent, 3/4 fl.oz. and easier 5100 735
per bag cement 38 0.40 3 penetration 5455 6050
7200
5175 6110 785
39 0.48* 3 Stiff mixes 3500 3650
trill sion. Penetration 3775
pva/water 1:1.44 40 0.49* lis difficult. 5.8 2250 3000 850
1!QQ.
790
2875 3456 820
* includes pva
TABLE 4. Workability and Strength of Mortar in Panels Reinforced with Various Meshes.
(Panels made in horizontal form mould with Type TI cement and mortar sand.)
Panel Kind of Mesh Water/ Slump.
Water Compres s he Modulus of
No. Reinforcement Cement in.
Workability Absorption strength. psi Rupture. psi
ratio
S dry basis
7-day 28-day
(unreinforced)
10 3/8 in.-20 gao 0.40 &Is Easy Penetration 6.1 6050 7100 905
welded square mesh 705
not valv. 7 layers
11 1/2 1n.-19 gao 0.40 71s Easy Penetration 6.8
-
7420 1180
welded hardware 990
cloth. galv
9 layers
12 1/2 1n.-16 gao 0.40
7"
Easy Penetration 7.2 5400 7950 840
welded square 9875 895
lllesh, ga lv
5 layers :
13 2.5 lb expanded 0.40
5"
Easy Penetration 6.6 5950 7200 725
llletal lath. ga1v 7500 792
5 layers
14 1/2 i".-22 ga. 0.40
&Is
Easy Penetration 7.0 6200 7600 890
hexagona 1 mesh. 7250
-
galv. 12 layers
15 1/4 1n.-20 ga. 0.40
SIs
Easy Penetration 7.0 5600 8150 830
fire screening, 7675 960
2 l.yers
27
(c) Mortar strength
The compressive and flexural strengths of the mortars
(or mortars reinforced with a "standard" reinforcement) containing the
various sands, cements, and admixtures were compared in a series of test
batches and panels.
Compressive strengths were determined for the early batches
by 2-i n. cube specimens (ASTf.1 Cl 09) cured for 7 and 28 days and for the
later batches by both cubes and the broken halves of the 1.575 x 1.575
x 6.3-in flexure beam prisms (ASTf.1 C349).
Flexure strengths were obtained for the early batches and
panels on specimens from the 3-in. wide unreinforced portion of the test
panels. The specimen layout pattern is shown in Fig. 3. Mortar flexure
tests for the later batches used 1.575 x 1.575 x 6.3-in. flexure beam
prisms (ASTM C348). The former s p e c ~ m e n s were loaded at the third-points
on a 10-in. span, the latter at mid-point on a 4.685-in. span. In
addition, some flexure tests were performed on specimens reinforced with
"standard II mesh reinforcement.
Drop-impact tests were also made, where applicable, on
specimens from reinforced panels. The drop-impact strength specimens,
15 x 15-in., were cut from mesh-reinforced panels containing mortars
with various sands, cements, and admixtures. The drop-impact apparatus,
shown in Figs. 4 to 6, was built to approximately simulate the velocity
and shock of a collision between a hull and a IIdeadhead" log at about
15 knots. The 50-1b drop tup was a round-bottom aluminum air tank
loaded with steel balls. The drop height was 10 feet, giving an impact
energy of 500 ft.-lb. Details of the support and cushioning with a disc
of 1/4 in. plywood are shown in Fig. 4. Tests on 30-in. and 36-in. panels
were performed with the same drop apparatus except that the large panels
uere supported on 24 x 26-in. and 31 x 31-in. 2 x 4 'Nord frames. The
drop-impact damage was assessed qualitatively and by measuring the
dishing of the impacted panel specimen and the extent of cracking on
the bottom "convex" surface.
The mortars made in later years were considered to be
the "standard" mortar mix, viz., Type II cement/dry bagged mortar sand
ratio, 0.5; water/cement ratio, 0.4 to 0.5, chromium trioxide (to
inhibit a reaction between galvanized mesh and bare rods), 300 ppm.
The compressive and flexural strengths of these "standard" mortars
made over the five years have been statistically analyzed.
28
D
B
G
c
E Fig. 3. Layout of 30-in. panels
F
for test specimens.
A,B - drop-impact or diagonal tests
C - flexure tests on unreinforced mortar
O,E - flexure tests on reinforced panel,
longitudinal and transverse
F,G - various test, durability, exposure,
paint, etc.
50 lb t u ~ , 10 ft. above specimen .
guide frame
plywood disc, 1/4 in. x 6 in. dia
foot of frame, 1/4 in. steel plate
ferrocement specimen, 15 in. square
plywood ring, 3/4 in. x 12 in. hole.
steel support frame
support pad, 2-in wood plank
Fig. 4. Sketch of drop-impact test guide frame.
29
Figs. 5 and 6. Drop-impact apparatus.
30
The mean values are:
Compressive strength
ASTM C109, 2-in. cubes
7-day
28-day
ASTM C349, 1.575-in. prisms
5670 psi
7840 psi
28-day 10,070 psi
Flexure Strength
ASTM C348, 1.575-in. prisms
28-day 1360 psi
The number of batches, the means, and the standard deviations
from the means are presented in Table 5.
The flexural strength of the various mortars was also evaluated
from specimens cut from panels reinforced with IIstandard mesh reinforcement
ll
,
viz. 12 layers of 1/2-in. 22-ga galvanized hexagonal mesh. These flexure
specimens and specimens for impact tests were cut from the 30-in. panels
as shown in Fig. 3. The II modu1us of rupture
ll
values were obtained.
The fibre stress formula for beams loaded in flexure, S = ~ c ,
applies to stresses below the elastic limit, i.e. where the d,istribution
of stress across the section is linear. Since the formula is not rationa1
Mc
for beams loaded beyond the elastic limit, a IImodified fibre stress
ll
, s1=-1
is used for beams loaded to rupture. This value of IIfibre stress
ll
is
c a l l ~ d the modulus of rupture. Throughout this report, the term II modu1us
of rupture
ll
has been reserved for the IIfibre stress
ll
calculated for :the
maximum load held by the beam. The value is calculated not at the onset of
first visible brittle cracking of the mortar (unless otherwise stated) but
at the maximum load when mesh wires broke, rods slipped, or both.
i. The Effect of Various Sands
The strengths of mortars and panels of similar reinforcement
containing either the dry bagged mortar sand or the Del Monte silica sand
were compared. The average 7-day and 28-day compressive strengths for the
mortar sand are 6015 and 8544 ps i, respecti ve 1y, and for the Del r10nte sand
5700 and 6100 psi, respectively. Although the Del Monte sand appears to give
somewhat lower compressive strengths the values are certainly within two
standard deviations, 20, calculated for all IIstandard mortar
ll
batches as
shown in Table 5.
The average modulus of rupture values, 1 ~ ~ , for unreinforced
panels of the two mortars is 823 psi for the mortar sanS mortars and 709 psi
for the Del Monte sand mortars. This latter value is lower than the flexure
31
strength, 869 psi, obtained on 14 mortar sand mortars. Single flexure
tests on simi 1 ar1y reinforced mortars. show a1most no di ffert;!nce between the
two mortars. The specimens were obtained from panels reinforced with 12
layers of 1/2-in. 22-ga. galvanized hexagonal mesh. The beam specimens
approximately 3 x 12 in. were tested in third pOint loading on a 10-in.
span. The load at first visible crack in the mortar and the maximum load
held by the beam specimens, oriented in both longitudinal and transverse
mesh directions, were recorded. The modulus of rupture, at maximum
load was calculated.
The compressive and flexure strengths for mortars and
panels containing the dry bagged mortar sand and Del Monte silica sand
are presented in Tables 6 and 7 for comparison. The test results, within
the limitations of replication show no consistent or practical
differences between the two sands used. It is concluded on the basis of
these preliminary tests that it was probable that a "special" sand would
not be significantly better. Therefore, it was decided to make all future
panels with the dry bagged mortar sand readily available from a local
source.
ii. Effect of Various Cements
A series of panels, Nos. 16,17,18,19,22,23, and 24
were made us i ng the "s tanda rd" dry mortar sand wi th Types I, II, I II, V,
and aluminous cements. All cement/sand ratios were 0.5. The water/
cement ratios were 0.36 to 0.47. The reinforcement in all panels was 12
layers of 1/2-in. 22-ga. galvanized hexagonal mesh.
The compressive strengths of the mortar from 2-in. cubes
at 7 and 28 days, the modulus of rupture values from flexure tests on
unreinforced portions of the panels and on longitudinal and transverse
reinforced specimens, and impact tests on 15-in. tiles were obtained.
The results, presented in Table 8, showed no consistent superiority of
mortar made with one kind of cement over that made with another kind.
-Within the replication possible, it was concluded that
the type of cement used was not an important factor in the strengths of
ferrocement mortars. The type may be a factor in durability, however.
iii. Effect of Admixtures
The average compressive and flexure strengths of a series
of ten panels which contained no admixture, pozzo1an (replacing 25 percent
of the cement), a water-reducing agent (6.5 fl. oz. per bag of cement), an
air-entraining agent (3/4 fl. oz. per bag of cement), and a po1yviny1-
acetate emulsion (pva/water 1:1.44) are presented in Table 9. It was
32
1IILI 5. Compressive and Flexure Strength of Standard Mix Mortar Batches Made Between 1969 and 1974.
(-Standard Mix is 1 part Type 11 cement. 2 parts dry bagged mortar sand - no admixtures.)
Cure NUll'ber Mean Stand. Mean
Tilt Proper\), Period. of Value. DeY.a :t3CJ
days Batches psi. psi. psi.
Compressiye Strength 7 33 5666 624 3794-7538
ASTM Cl09. 2-1n. cubes 28 40 7843 898 5149-10537
CompressiYe Strength 28 14 10071 924 7299-12843
ASTM C349. 1.575 x 1.575 x
6.5 in. prisms
Flexure Strength
ASTM C348. 1.575 x 1.575 x
28 14 1371 121 1008-1734
6.5 in. prisms
TABLE 6. Summary of Strength Tests on All Mortars Made with Type II Sand and Dry Mortar Sand or Del Monte Sand.
Year
No. of Compressive Strength, psi Compressive Flexure Modulus of
Of
Kind Of Panels 2-in. cubes, C109 Strength,psi Strength, psi Rupture, psi
Test
Sand (Batches) 1.575 in. prism 1.575 in. prism (unreinforced)
Tested 7-day 28-day ASTM C349 ASTH C348
1969-70 Dry bagged 7 5838 8087
IIIOrtar sand
(-8 mesh) 14 869
Del Monte sand 1
(8-/20-/30-mesh)
5700 6100 709
1970-71 Dry bagged 23 5505 7874
IIIOrta r sand
1972-73 Dry bagged 10 5390 7680 10,090 1350
IIIOrtar sand
11169-73 Dry bagged 4300 to 6400 to 7500 to 1120 to 705 to
IIIOrtar sand 7275 9950 12.400 1730 1230
33
TABlE 7. Results of Tests on Mortars Ind Plnels Using Two Types of Sand'
(Type II cement. 12 layers of 1/2-in. 22-ga. galy. hexagonal mesh.)
'anel
Compressive Modulus of Panel Strength
Sand Strength, psi Rupture, psi
No.
(unreinforced)
Flexure
l
psi
Impact
2
7-day 28-dav dishing, 1/16 in.
5 Dry blgged
- -
1125
IIOrtar sand 1100
6
- -
742
890
14 6200 7600 890 longit. 2421 19. exposed
7250
-
transY. 1610 mesh. broken wires
16 5830 9450 760
9875 820
-- -- --
6015 8544 904
25 Del Monte 5700 6390
I
635 10ngit. 2490
1:2:1 5780 782 trln 1660
1 Simple beam, approx. 3 x 12 in., span 10 in., third-point loading
2 SO-lb tup with round base (9-in. radius), dropped 10 ft. onto 15-in. panel specimen.
TABlE 8. Results of Tests on Mortars and Panels Using Various Types of Cement
(Dry bagged mortar sand. 12 layers of 1/2-in. 22-ga. galv. hexagonal mesh.)
Compressive Modulus of
Panel Strength
Pinel Type of Strenqth psi Rupture, psi
Ho. Cement (unreinforced)
Flexure
l
.-. pst

7-day 28-day dishing, 1/16 in.
17 Type I 5040 5875 1050 2910 16, exposed
6175 960 Transv. 1665 mesh. broken wires
-- -- --
5040 6025 1005
14 Type II 6200 7600 890 10ngit: 2420 19. exposed
7250
-
trlnsv. 1610 mesh. broken wires
16 5830 9450 760
-
9875 820
-
wrn- m
18 Type III 6400 8325 1320
-
-
1280
-
19 5150 8450 590 10ngit. 2430 22. exposed
7750 560 trlnsv. 1585 mesh. broken wires
22 8000 7000
- -
7800
- -
m;
1865 m
23 Type V 7575 10000 932 10ngf.t. 2960 11. exposed
10700 843 trlnsv. 1735 mesh. broken wires
-- -- -
7575 10350 888
24 Aluminous 9150 7250 1028 longit. 3500 18. exposed
--
6280
.m.
trlnsv. 1810 mesh. broken wires
9150 6765 910
1 simple beam. approx. 3 x 12 in span 10 in., third-point loading
I 50-lb tup with round base (9-in. rldius) dropped 10 ft. onto 15-in. panel speCimen.
observed that, except for the polyvinyl acetate emulsion which had low
values of 7-day and 28-day compressive strengths, the several admixtures
had little effect on the strength of the mortars. The average 7-day and
28-day compressive strengths for the admixtures (excluding the polyvinyl-
acetate emulsion) were between 3 standard deviations of the mean for
all "standard mix" mortars 1969-1974 shown in Table 5.
Although the optimum addition rate for the above admixtures
in mortars is now known, the chance of significant improvements in strengths
did not seem to warrant further work at this time.
(d) Water Absorption and Transmission
A few of the were tested for differences inwater
absorption and transmission properties. Specimens from unreinforced panels
made with Type II cement and dry mortar sand were soaked 24 hours in water.
The specimens absorbed 6.0 to 7.3 percent water on an oven-dried basis.
The mortars containing an admixture were also tested. The mortar containing
the water-reducing agent absorbed 3.9 percent, those with the other admixtures,
5.2 to 6.7. The absorption values are given in Tables 2, 3, and 4.
It is worth mentioning that test coupons subjected to many
wet/dry exposure cycles lost only about two percent of their weight after
being oven-dried overnight. This suggests that the voids may become filled
with a gel or become otherwise clogged by immersion in water for a long time.
Other minor tests were also performed for manifestation
of the water permeability of ferrocement. A 2-ft. column of water was main-
tained for one year on a mortar test coupon '3/4-in. thick. A small grey
spot appeared on the bottom for a short time but no water was evident.
The grey spot soon disappeared as a gelling reaction at the mortar/water
interface sealed off the mortar surface. Another similar test coupon was
placed on a thin polyethylene sheet. Capillary forces failed to draw
moisture through the coupon at the grey spot.
It was concluded that a well-made mortar will provide a
water-tight hull.
(e) Durability
The durability of concretes (and mortars) has been defined
as "resistance to deteriorating influences of internal 'and external factors
to which it is exposed within the duration of life expected from the
structure". It follows that the requirements for a durable concrete will
depend on the type of structure, the type of exposure or condition,
and on the required service life. The service exposure inc1u/ weathering,
chemical action, and wear.
35
Although other characteristics may be important in tropical
climates, the characteristics of greatest importance to the performance
of mortars in ferrocement hulls of Canadian fishing vess"els operating in
Canadian waters are weather resistance and seawater resistance.
The disintegration of concretes (mortars) by weathering
is caused mainly by the disruptive action of freezing and thawing and of
expansion and contraction from variations in temperature and alternate
wetting and drying. Although laboratory tests for durability are difficult
to correlate with service performance, tests conducted over many years have
shown that freeze-thaw tests can distinguish between durable and non-
durable types of concrete.
Seawater is considered to be mildly corrosive to concretes
(mortars) mainly because of the soluble sulphate salts in the water. It
is generally recommended that concrete for use in seawater be made with
Portland cements which contain not more than 8 percent tricalcium aluminate.
Types II, IIA, IV, and V meet this requirement. A low water/cement ratio
and entrained air are claimed to increase the resistance to attack by
seawater.
Although much research and exposure testing has been under-
taken over many years on structural concrete, relatively little research
has been reported on the durability of ferrocement. A series of tests to
determine the resistance of various ferrocement mortars to freeze-thaw and
seawater exposures was therefore undertaken. The tests described herein
are not comprehensive but are intended to show any drastic sho"rtcomings in
the behaviour of the various mortars and ferrocement constructions used.
i. Freeze-thaw Exposure Tests
Since ferrocement vessel hulls are of relatively thin section,
specimens from ferrocement panels which represent the hull in section thick-
ness do not meet the specimen sizes required in ASTM specifications for
testing concrete and for brick and structural clay tile, i.e. ASTM C666-73
and A S T t ~ C67-73. C666 requires the specimens be subjected to 300 cycles
(40F to OF) of freezing and thawing (or until its relative dynamic modulus
of elasticity reaches 60 percent of its initial modulus) and C67 requires
the specimens be subjected to 50 cycles of freezing and thawing.
Twelve weighed unreinforced mortar coupons, 3 x 5 x 3/4 in.,
containing various cements, sands, and admixtures were placed in shallow
water-filled trays in an environmental chamber*. The chamber was programmed
* Environmental Chamber Model ELHH-27-MRLC-l, Associated Testing
Laboratories, Inc.
36
to give six freeze-thaw cycles per 24-hour day. The water temperature (and
coupon temperature) cycled from 10C to -4C and the air temperature from
25C to -12C. The cycling prpgram was continued for 350 cycles. The coupons
were observed regularly. At the end of the program the coupons were dried
at 200F and weighed.
The mortars tested, the visual observations, and the weight
loss from spalling are recorded in Table 10. The mortar sample which contained
the pozzolan addition disintegrated badly, that which contained the po1yviny1-
acetate emulsion showed some spa11ing and weight loss. In general, admixtures
did not confer improved resistance to freeze-thaw cycling. Fig. 7 shows four
coupons, one of which has disintegrated.
It has been shown that the presence of mesh affects the freeze-
thaw resistance of ferrocement. Eleven test coupons, 3 x 4 x 3/4-in., containing
12 layers of 1/2-in. 22-ga. galvanized hexagonal mesh reinforcement, were sub-
jected to 76 freeze-thaw cycles in a similar program. The coupons had a top
layer without reinforcement. The sawn edges of these reinforced specimens
were coated to prevent ingress of water at the sawn edges.
The specimens were examined after 36 and 76 cycles. The test
results are presented in Table 11. In general, the unreinforced portions of
the test coupons were disintegrated, often by delamination at the reinforced/
unreinforced interface, after 76 cycles of freezing and thawing; the reinforced
portions were sound. Fig. 8 shows typical specimens.
Although there were some inconsistencies in the test results,
the preliminary results indicate that mesh reinforcement (e.g. 1/2-in. 22-ga.
hexagonal mesh) improves the freeze-thaw durability of ferrocement mortars and
the kinds of cements, sands, and admixtures used appear to have little effect
on the freeze-thaw durability of ferrocement mortars.
ii. Seawater Exposure Tests
Test coupons, 3 x 4' in., were sawn from various panels containing
various mortars and meshes. The sawn edges were coated with an epoxy compound
to prevent entry of water through the sawn edges. The 24 test coupons were
inserted into slots in two wheels of an exposure apparatus. The apparatus
immersed the specimens in a bath of natural filtered seawater for one hour
and removed them for drying in front of a fan for three hours. The tests
were performed at ambient room temperature.
The condition of the test coupons after 350 cycles of seawater
immersion were evaluated by visual appearance and scratch and penetration hard-
ness compared with coupons not subjected to the wet-dry cycling.
Panel
No.
31. 32
33. 34
35. 36
37. 38
39. 40
Pinel
No.
16
17
19
23
24
25
26
31
33
35
37
4D
37
TABLE 9. Ef'ect of Admfxtures on Compressfve and Flexure Strength of Mortars.
(Mortar - Type II cement. dry bagged mortar sand. cement/sand ratto 1:2)
(Panels refnforced with 12 layers 1/2-in. 22-ga. galv. hexagonal mesh)
Average.Compressive Modulus of
Matxture
Strength. pst Rupture. pst
(unretnforced)
7-day 28-day
None 5100 6430 1025
Pozzolan-replaced 25S of cement 4060 6560 760
Water-reducing agent - 6.5 fl. oz./bag of cement 6780 5490 1060
Air-entratntng agent - 3/4 fl. oz./bag of cement 5180 6110 785
Polyvtnylacetate emulsion - pya/water is 1':1.44 2880 3460 820
TABLE 10. Freeze-thaw Durabtlity of Mortars Containing Various Cements.
Sands Ind Admixtures after 350 cycles. 10C to -4C.
.
Description of Mortar No.
Visual Appelrance
Weight loss
of
of Test Coupons
Dry Bisis.
Cement Admixture Sand
Cycles
S
11 None Dry Mortar
Sind
350 No significant change 0
I None 350
0
III None 350
0.8
V None 350
0
Aluminous None 350
0
11 None Del Monte
sf1icl sand
350 Slight powdering
0.7
I None Dry mortar 350 No Significant chlnge 1.1
Sind
II None 350
1.2
II POllohn
SO-loo Some corner crumbling
350 Crlmlbled badly. 26.2
II Water-reducing
350 No signiftcant change 1.1
agent
II Air-entraining 350
2.2
Igent
II Polyvinyl- 100-125 Sltght surface flaking
acetlte 125-150 Slight surface spalling
350 Slight surface spilltng 4.9
38
..
Fi g. 7. Condition of four unreinforced test
coupons after 350 freeze-thaw cycles.
~ J
F-4 ...
~ . "" ,.! ..: ~
-/( ............ '.: l.1. "7.
Fig. 8. Condition of mesh-reinforced coupons
after 76 freeze-thaw cycles.
39
The description of the test coupons exposed, their visual
appearance, and the penetration hardness (a Rockwell Hardness tester using
a 1/8 in. steel ball and a 60 kg. load) are presented in Table 12.
The changes in visible appearance and in scratch and pene-
tration hardness after 350 cycles of exposure were insignificant.
The weight loss of the test coupons on oven drying ranged
from 0.7 to 2.0 percent, much lower than the normal water absorption per-
centages obtained on mortars not subjected to the cyclic exposure program.
It is believed that the precipitation of salts and formation of cement gels
may clog the pores and voids after prolonged exposure.
It is tentatively concluded that all of the mortars tested
resist attack from seawater under the ambient laboratory conditions used.
The effect of various reinforcements, galvanized or baret and of seawater
attack at damaged areas is considered later.
(f) Interrupted Mortaring
A large hull may tax the ability of the plastering crew to
finish in one day. Interrupted mortaring or multi-stage cold-joint plastering
has been used to avoid the deteriorating quality of work by an overtired crew.
Several tests were performed to determine any loss of strength which may
result from interrupted mortaring. Mortar was applied to one side of a
panel with layers of 1/2-in. 19-9a. hardware cloth on 0.225-in. rods. The
other side was wetted and plastered 24 hours later. Sectioning revealed
some boundary interface sponginess in some areas, good bond in others.
Flexure testing did not cause delamination. It was c o n c 1 ~ d e d that, provided
the mortar consistency was proper and the mesh penetrability was good, a
satisfactory bond between the two mortar layers may be obtained.
Another panel was completely plastered from both sides except
for a 12-in. wide strip at one side. The rewetted panel was finished 24
hours later. A flexure test specimen with the joint at mid-span cracked at
the joint but its modulus of rupture was only slightly lower than that of
specimens containing no joint.
Another panel was similarly made with a joint down the centre.
Part of the joint was wetted and part was coated with a neat cement paste
before remortaring 24 hours later. Flexure tests produced cracks at the joint
but the flexure strengths at first visible crack were similar to those with
no joints. The modulus of rupture values (at maximum load) of joint specimens
are 60 to 80 percent of the modulus of the specimens with no joint.
40
TAILI 11. Freeze-Thaw Durability of Mesh-Reinforced Test Coupons of Various Cements, Sands, and Admixtures.
(Reinforcement 12 layers 1/2-in. 22-ga. hexagonal mesh.)
Description of Mortar
No. of Vtsual Appearance
Cement Admixture Sand Cycles
Type I n11 Dry 36 No visible change.
IIOrtar 76 Unreinforced layer partly disintegrated.
sand Reinforced layer sound.
Type U n11

36 No visible change.
76 Unreinforced layer completely disintegrated.
Reinforced layer sound.
Type III n11 36 No visible change.
76 Unreinforced layer completely disintegrated.
Reinforced layer sound.
Type Y n11

36 No visible change.
76 Unreinforced layer partly disintegrated.
Reinforced layer sound.
Ahlllfnous n11

36 No visible change.
Type II
Type II
Type II
Type II
Type II
76 No visible change.
n11 Del Monte 36 Slight flaking top surface.
s11ica 76 Unreinforced top surface flaked slightly.
Reinforced layer sound.
n11 Dry IOOrtar 36 No visible change.
sand 76


POllohn

36
76
Water-
"
36
reducing 76
agent
Air-
II
36
entraining 76
agent
TABlE 12. Assessment of Test Coupons of Various Mortars After 350 cycles
of Immersion in Seawater and Drying.
(coupons contained various mesh reinforcements)
Description of Mortar No. Visual Appearance Penetration Hardness Value
of of Coupons After
Unex osed EXDosed
Cement Sand Admixture
Coupons 350 cycles of Exposure
Top Bottom Top Bottom
I DrylOOrtar n11 2 No change. 90 95 109 96
II 10
102 102 108 103
III 2
102 90 102 103
Y 2
102 105 95 103
Aluminous 2
109 109 112 113
II Del Monte 1
96 91 100 105
s11ica
II Dry IOOrtar 1
90
-
9S 9S
II pOllolan 1
89
-
99 99
II water 1
103
-
105 106
reducing
agent
II air en- 1
85 76 111 117
training
agent
II polyvinyl 1
54 sa 65 62
acetate
41
It is concluded that the flexure strength of ferrocement made
with interrupted mortaring will be somewhat lower than that of ferrocement
with no joints. However, the difference in strength at the first visible
crack does not appear to be great. Possible leakage problems at the joint
have not been examined. However, it is believed that a jOint, properly bonded
with cement of the proper consistency or with an epoxy interface, should
not leak.
3. Evaluation of Various Reinforcements
(a) General Procedure
Tests were performed to determine the strength of the various
rod and mesh reinforcements alone and as components of a steel/mortar composite
material. The composites were tested mainly in flexure and impact. The mortar
was generally a "standard" mix of Type II cement (1 part), dry bagged mortar
sand (2 parts), water/cement ratio 0.4 to 0.5, chromic oxide Cr03 300 ppm.
(b) Rod Reinforcement
i. Kinds and Strengths
The ultimate tensile strength, the elongation (percent in
8 in.), and the reduction in area were obtained for the above rod reinforcement
materials. In addition, since it is common practice for ferrocement boat
builders to arc-weld some or all of the cross-over junctions of the longitudinal
and transverse rods, tensile tests were also performed on specimens . of double-
drawn high-tensile rods with a "typical" cross-weld. Double-drawn high-tensile
A82 rods obtain their enhanced strength from cold working. Heating, as by
welding, should diminish this strength.
The ultimate tensile strengths of the several rods tested were
as follows:
hot-rolled black (C1020) rounds
galvanized (C1020) rounds
bright drawn (C1010) nail wire
double-drawn high-tensile A82
double-drawn high-tensile A82 with arc weld
double-drawn A82 (extra high strength)
double-drawn A82 (extra high strength)
with arc weld
deformed double-drawn A82
70,000 psi
50,000 psi
73,000 psi
79,000 psi
80,000 psi
101,000 psi
89,000 psi
84,000 psi
The strengths, elongations, and the reductions in area are
presented in Table 13. The double-drawn A82 rod material is strongest. Its
strength was not seriously diminished by a small arc weld. Excessive heat
input should be avoided, . however.
42
11. Bond Strengths
Many papers have appeared in the literature showing the
merits of black vs galvanized steel reinforcement for concrete construction.
Building Research Digest 109
16
has summarized five separate investigations into
the bond strengths of steel reinforcing rods with galvanized, smooth, mill
scale, or rusted surfaces. It concludes that the bond performance of galvan-
ized bars is as good or better, on the average, than that of the smooth, mill
scale, or rusted bars. It further points out that the protection against
corrosion provided by the galvanized rods gives improved long-term bond
performance over the others in a corrosive environment. Rust scaling and
subsequent spa11ing of the concrete cover are avoided.
The effect of scale, drawing lubricant, rust, zinc
coating, and deforming of rods on the rod/mortar bond strength in ferrocement
applications was assessed in this laboratory by a series of tests. Mortar
(Type II cement, dry mortar sand) was packed in moulds around hairpin-shaped
rods of the several kinds to form duplicate specimens like those shown in
Fig. 9. The embedded length of the arms was 6 in. The specimens were cured
28 days.
The duplicate rod/mortar specimens were tested in tensile
loading. The maximum load was reached when one arm of the hairpin started to
slip. Both arms of the galvanized rod appeared to slip simultaneously. The
deformed double-drawn rod "hairpins" split the mortar blocks, presumably due
to the wedging action of the dimpled surface. Two tested mortar blocks are
shown in Fig. 10. The unit bond strengths were obtained by dividing the
maximum load held by the embedded area of both arms of the hairpin. The bond
strength of the single arm of each specimen which held was obtained after an
additional period of 3 1/2 months. The results obtained in the tests after
1 month and 4 1/2 months are shown in Table 14.
The following conclusions are drawn:
- The rod/mortar bond strength increases with time (at least
under non-corrosive conditions).
- The rod/mortar bond strength of rods pickled to remove mill
rolling scale is similar to that of rods with the mill scale
intact.
- The rod/mortar bond strength of double-drawn rods is improved
by removing the drawing or protective lubricant.
- Light rusting on double-drawn rods further improves its
rod/mortar bond strength.
- The dimpling process for double-drawn rods improves the rod/
mortar bond strength causes splitting of the mortar block.
- The galvanized rods had very poor bond strengths.
Fig. 9. Rod/mortar bond specimens.
Fig. 10. Rod/mortar bond specimens showing
splitting of mortar block by deformed
double-drawn rod.
43
TABLE 13. Tenstle Properties of Various Reinforcing Rod Materills
used tn the Ferroeement Test Progrlm.
Descriptton of Rod or Rod Simple
Dtameter. U.T.S.
tn. psf
Hot-rolled black (C1020) rounds 0.250xO.265 70,000
(oval) 69,700
Galvanfzed (C1020) rounds 0.220xO.276 50,600
(ovll) 49,300
Bright drawn (Cl010) nafl wfre 0.250 73.400
72,700
Double-drawn high-tensfle A82 rods 0.225 78,800
Double-drawn hfgh-tensfle A82 rods - arc weld
-
80,000
cross-over joint
Double-drawn high-tensile A82 rods - extra 0.225 100,000
high strength 99,000
102,000
Double-drawn high-tensile A82 rods extra
-
89,000
high strength, arc weld - cross-over Joint
i
1
Jspecfmen broke close to or outside of gauge marks.
2 SpeCimen broke at arc-weld
J Specimen broke 1/2 in. from weld.
Elong. Red. 1n
S fn 8 fn. Area.
S
23.5 67.3
24.0 66.8
24.4 71.3
25.1 69.0
(1) 63.9
8.5 64.8
10.0 63.5
7.5(2 ) 62.5
~ 1 )
49.5
1) (1)
58.5
3.0 52.0
1.25 (3)
TABlE 14. Results of ,Rod/Mortar Bond Tests on Hairpin Specfmens of Various Refnforcing Rods.
Bond Strength, lb/sq. in.
Type and Conditfon of Refnforcfng Rods of Embedded Surface Remarks
After 28 days After 4lt months
Hot rolled ,Cl020) 406
(scale fntact)
580 One arm slipped
Hot rolled (Cl020) 300 555 One arm slipped
(selle removed by picklfng)
Double-drawn A82 135 280 One arm slipped
(IS received with drawin9 lubricant)
Double-drawn A82 188 330 One Irm sl fpped
(drawing lubrfcant removed)
Double-drawn A82
(l1ghtly rusted by few days exposure) 286 518 One arm slfpped
Deformed double-drawn A82 451 660 Mortar blocks split
(clean), by wedge action
Galvanfzed (Cl020) rod 33 57 Both arms slipped
(cleln) simultaneously
Old rod (reclaimed from outside panels 480 700 Rods broke, heavy
submitted) corrosion and pit-
ting as fnstalled
45
It seems likely that the splitting action of the deformed
(dimpled) double-drawn rod could result in delamination of ferrocement. Until
it has been proven otherwise, it is recommended that this material not be
used. The markedly inferior bond strengths obtained by the galvanized rod
was unexpected. Building Research Digest 109 claims that the reaction between
zinc and the alkaline liquid in the pores of freshly placed concrete can form
bubbles of hydrogen gas which would have an unfavourable effect on the bond
strength in normal reinforced concrete. Frazier 17 of the American Hot Dip
Galvanizers Association cites tests which dispute the postulation of an earlier
investigator that a reduced bond value may have resulted from hydrogen bubbles
forming on the galvanizing which caused the adjacent concrete to become spongy.
He seems to support the finding that concrete may cure more slowly in the
presence of zinc and that the long-term bond strength of galvanized reinforce-
ment rods will be superior. It is generally agreed that chromate-treatment
of galvanized rods or the addition of chromium trioxide to the concrete (or
mortar) will cure the hydrogen problem and enhance the bond strength.
It has been suggested that some specifications bodies are already considering
the incorporation of a requirement to chromate-treat all galvanized reinforce-
ment bar. (The addition of chromium trioxide to the mortar is discussed in
Section C-3-d-iv, The Hydrogen Gas Problem.)
In conclusion, it is felt that the use of galvanized rods, if
properly treated, should be acceptable even in the short-term. Galvanized
rods may be superior in the long term to bare steel which may be attacked by
the seawater or which may be sacrificially protected by the zinc coating on
the mesh. If the zinc on the mesh is consumed sacrificially to protect bare
steel rods, it seems likely that the life of the mesh will be diminished. No
tests or post-mortem examinations of old hulls have been made to validate this
postulation.
iii. Costs
The approximate current (Feb. 1975) costs of the several
rods used are presented for the purpose of comparison:
0.225-in. dia. double-drawn A82
1/4-in. hot-rolled C1020
1/4-in. galvanized C1020
1/4-1n. C1015 bright nail wire
c/1b
40
50
63
53
$/100 ft.
5.30
8.25
10.39
8.74
It will be observed that the cost of the rod material on a per
1b basis will not be a very important factor in the kind of rod chosen. The
differences can be calculated for a specific hull design on the basis of
the cost/100 ft. making due allowance for the difference in rod diameters.
46
(c) Mesh Reinforcement
1. Kinds and Strengths
The breaking strengths of some of the meshes were determined
at various times throughout the five year program. The strengths of the dif-
ferent batches purchased from various sources over the several years varied
somewhat. Typical breaking strengths of mesh obtained from sing1e-, two-,
three-
t
four-wire and folded-layer samples of mesh are as follows:
1/2-in. 16-ga. ga1yanized welded square mesh
1/2-in. 19-9a. galvanized hardware cloth
3/8-in. 20-ga. black welde; square mesh
1/2-in. 22-ga. galvanized hexagonal mesh
- longitudinal direction
transverse direction
400 1b/in. width
140 1b/in. width
154 1b/in. width
60 lb/in. width
20 1 b/in. 'width
The strengths/wire obtained by the various techniques varied
because of the mesh geometry as well as from difference in the strength of
the steel itself. Table 15 summarizes the breaking strengths and unit tensile
strengths found.
On the basis of the simple tensile strength obtained on the
various meshes it is apparent that similar tensile strengths in the panels will
be obtained by 2 layers of 1/2-in4 l6-ga. galvanized welded square mesh, about
6 layers of 1/2-in. 19-9a. galvanized hardware cloth, 5 layers of 3/8-in.
20-ga. black welded square mesh, and 13 layers of 1/2-in. 22-ga. galvanized
hexagonal mesh (in its stronger longitudinal direction).
ii. Bond Strengths
No rigorous tests to determine the mesh/mortar bond strengths
of the various meshes have been performed. Bond strength will vary with the
geometry of the mesh (twisted wires, cross-welded wires), the kind of surface
e.g. galvanized, unga1vanized, phosphated, chromated, rusted, scaled. One ad
hoc test on 1/2-in. 16-ga. mesh with galvanized coating intact and stripped
showed no significant difference in panel modulus of rupture. Each mesh was
separated from the, mortar in the spec.i.mens wi th equal diffi cu1 ty. The grooves
left by the separated wires show a textural difference. Those of the galvanized
wire showed a spongy appearance whereas those of the stripped wire showing
a smooth appearance. The spongy appearance may be due tQ a zinc-water-cement
reaction.
The surface bond area per square foot of panel and the specific
bond area, the ratio of the surface area of the mesh to its volume (only those
portions of the wire oriented in the load stress direction) have been calculated
for equal weights of mesh reinforcement, e.g. 2 layers of 1/2-in. 16-ga.
galvanized square mesh, 5 layers of 1/2-in. 19-9a. galvanized welded hardware
47
cloth, 10 layers of 1/2-in. 22-ga. hexagonal me's'h (in both directions) and
6 layers of 3/8-in. 20-ga. black square mesh. These values are summarized
in Table 16.
111. Specific Surface
The specific surface of reinforcement, K, the total surface of
wire in contact with the mortar divided by the volume of the composite, is
often used as one of the parameters to define ferrocement. Bezukladov
19
claims that ferrocement must have a specific area of steel wire between 2.0 cm-
1
and 3.0 cm.-
1
(5.1 in.-
1
to 7.6 in.-
1
). The specific surface, K = 2TT d n
where d=wire diameter, n=number of layers of mesh, a=wire spacing, at
and t=thickness of panel specimen has been calculated for equal weights of the
above mesh reinforcements in mortar ;,pecimens with a bare mortar cover and
with 1/16-in mortar cover on each side of the mesh The values are
shown in Table 17. If rods had been present, the surface area of the rods and the
volume of mortar encasing the rods would have been ignored.
The specific surfaces ranged from about 2.6 to 3.9 cm-
1
for
a bare cover and 1.7 to 3.20 cm.-
1
for 1/16-;n. cover. It is apparent that
the hexagonal mesh has the highest specific area of the three 1/2-in mesh
materials examined. In practice, it is not possible to compress the layers
of hexagonal mesh uniformly to the "minimum" thickness and the specific
surface for this "effective" thickness will be somewhat lower.
It is a1lo apparent that the 1/2-in. 16-ga. galvanized welded
square mesh gives the least panel (hull) thickness for equal weight of mesh
reinforcement and the 1/2-in. 22-ga. galvanized hexagonal mesh the greatest
panel (hull) thickness.
1v. Workability
As shown in Table 4, the kind of mesh used, viz. 1/2-in. 16-ga.
galvanized welded square mesh, 1/2-in. 19-9a galvanized hardware cloth,
3/8-in. 20-ga. black welded square mesh, 1/2-in. 22-ga. galvanized hexagonal
mesh, 1/4-in. 20-ga. firescreening, and 2.5 lb/sq. yd. expanded metal lath,
presented no real problems of mortar penetration when the mortar was applied
into a horizontal form mould with vibration trowelling. The small holes of
the expanded metal lath used would make penetration of mortar difficult or
impossible where hand trowelling of a vertical panel was necessary.
v. Conformabi1ity
No formal tests on the conformability of the various meshes
to compound curves have been undertaken. The hexagonal mesh, because of its
48
1/2-1n.
1/2-1n.
3/8-in.
1/2-1n.
TABLE 15. Summary of Typica' Strengths of the Several Meshes Used
Description of Wire
Wire Mesh 5t -enath
Diameter,
lb/wire
in.
16-ga. galvanized welded square mesh 0.0624
- longitudinal direction 219
- transverse direction 175
19-9a. galvanized hardware cloth 0.033 58
- longitudinal direction 64
- transverse direction 63
20-ga. black welded square mesh 0.034 58
22-ga. galvanized hexagonal mesh 0.024
- longitudinal direction of twists 27
- transverse direction 27
TABLE 16. Unit and Bond Areas of Three Meshes as Equal Weight
Reinforcements in Ferrocement Panels.
lb/in.
width
418
350
116
128
126
154
60
21
Typical
U. T.5.,
psi
65,000
70,000
64,000
60,000
60,000
Mesh Weight Bond Area
Description of
Ib/ft
2
Ib/ft
2
in2/ft2 in
2
/in
3
mesh panel of panel of mesh
1/2-1n. 16-ga. galvanized welded square mesh 0.56
(2 layers of mesh) 1.12 118 62
1/2-1n. 19-9a. galvanized hardware cloth 0.24
(5 layers of mesh) 1.20 190 96
1/2-1n. 22-ga. galvanized hexaqonal mesh 0.11
layers of mesh - longitudinal direction) 1.10 300 120
10 layers of mesh - transverse direction) 1.10 195 160
3/8-1n. 20-ga. black welded square mesh 0.200
(6 layers of mesh) 1.20 252 114
49
"soft" geometry, conforms most easily. Loose.woven wire, because of its
ability to assume a diamond-shape, possesses a useful degree 'of conformability.
The 1/2-1n. 16-ga. welded square mesh has least conformabi1ity but assumes
the "fairest" shape on simple curves.
vi. Cost
The rapidly changing prices for steel make absolute costs of
doubtful use. However, the relative costs are of interest. The local prices
(Spring 1975) show that the cost of ~ / 2 - i n 16-ga. galvanized welded square
mesh is about twice that of 1/2-in. 19-9a. galvanized hardware cloth and
1/2-in. 22-ga. hexagonal mesh. (cf. 44, 23, 22/sq. ft.) (It is interesting
to compare the quoted 1971 prices of about 30, 12, and 6!) The difference
in cost of the three mesh materials is not great on an equal weight basis
(especially in 1971).
(d) Reinforcement in the Ferrocement Composite
i. Effect of Rods
I
Although the strength of the rods per se is of undoubted
importance the real measure of their usefulness is their performance in a
ferrocement hull. Panels, although they have relatively short rod/mortar
bond lengths, have been constructed to determine the contribution of rods
to the drop-impact resistance and flexural strength of the ferrocement comp-
osite and to assess the effect of the various kinds of rods and their spacing
on these properties.
The effect of using rods was established by a series of panels
which contained similar mesh reinforcements. The effect of the presence of
rods on drop-impact resistance and flexural strength was examined in panels
which contained no rods (panels 47, 48) (rods initially present to separate
mesh layers, then removed immediately after mortaring) and panels which con-
tained rods at 2-in. spacing (panels 42, 45), each with 1 layer of 1/2-in.
16-ga. weld mesh on the outer portions of the panel.
The drop-impact resistance and flexural strength test results
presented in Tables 18 and 19 show the considerable contribution of rods
to the panel strengths.
The effect of rod spacing was examined by .pane1s reinforced
with 1/2-in. 19-9a. galvanized hardware cloth on 0.225 in. double-deformed
high tensile A82 rods at 2-in. spacing (panels 41, 44) and 4-in. spacing
(panels 54, 55). The drop-impact resistance and the flexural strengths,
presented in Tables 20 and 21 show that 2-in. spacings are superior to 4-in.
spacings.
50
TABLE 17. Specific Surface of Various Meshes in Theoretical Panel
Specimens with Bare Mortar Cover and 1/16-1n Mortar Cover Each Side.
(IPProX. equal weights of mesh reinforcement and corresponding panel thickness)
Assumed
Specific Surface Specfftc Surface K
Description of Meshes
Panel Wires in long. Wires in trans. Wires in both
Thick. dirHtion
di rect 1 on"s.
tn. in -! em. -! ;n - t"m - :.t
em -1
Wtth bare mortar cover
1/2-tn. 16-ga. galv. welded square mesh 1/4 3.26 1.28 3.26 1.28 6.52 2.56
(2 layers. 1.12 lb/sq.ft. panel)
1/2-1n. 19-,a. galv. hardware cloth 3/8 3.35 1.32 3.35 1.32 6.70 2.64
(5 ayers. 1.20 lb/sq.ft. panel)
1/2-1n. 22-ga. galv. hexagonal mesh 1/2 4.17 1.64 5.83 2.30 10.00 3.94
(10 1.10 lb/sq . ft. panel)
3/8-1n. 20-ya. black welded square mesh 3/8 4.67 1.85 4.67 1.85 9. 34 3.70
(6 ayers. 1.20 lb/sq. ft. panel)
With 1/16 in. mortar cover. both sides
.
1/2-1n. 16-9a. galv. welded square mesh 3/8 2.18 0.86 2.18 0.86 4.36 1.72
(2 layers. 1.12 lb/sq. ft. panel)
1/2-1n. 19-ya. galv. hardware cloth 1/2 2.54 0.98 2.54 0.98 5.08 1.96
(5 ayers. 1.20 lb/sq. ft. panel)
1/2-1n. 22-ga. galv. hexagonal mesh 5/8 3.34 1.31 4.67 1.84 8.01 3.15
(10 layers. 1.10 lb/sq. ft . panel)
3/8-fn. 20-ga. black welded square mesh 1/2 3.51 3.51 1.38 7.02 2:76
(6 layers, 1.20 lb/sq . ft. panel)
TABLE 18. Drop-Impact Resistance of Mesh-Reinforced Panels with and without Rod Reinforcement.
(mesh reinforcement 1 layer of 1/2-in. 16 gao galvanized welded square mesh on each side)
Pinel
Rod Reinforcement Dishing
Max. circle
Mode of Failure
Ho. of cracks,
in 1/16 in.
in.
45 Double-drawn high-tensile 8 18 Fine radial cracks
A82 rods at 2-in. spacing Fine rectilinear cracks.
48 No rods 18 32 Ffne radial cracks.
One wide rectilinear cracks.
TABLE 19. Flexural Strengths of Mesh-Reinforced Panels with and without Rod Reinforcement.
(mesh reinforcement 1 layer of 1/2-in. 16-ga. galvanized welded square mesh on each side)
Panel
Rod Reinforcement
F1rst
Maximum
Modulus
Mode of Failure
1C0. Visible
Load, lb
of Rupture
crack, lb psi
42A Double-drawn high-tensile 1300 2800 6050 Hain crack at a load point
AB2 rods at 2 in. Rod slip noted
42B spacing 1950 3240 7000 Fine cracks at load points
Rod slip noted
47A 1100 1600 2800 Several fine cracks between
No rods load panels
47B 900 130D 2280 Fine cracks between
load points.
A transverse rods in tens10n side, B lengthwise rods in tension side.
51
TABLE 20. Drop-Impact Resistance of Mesh-Reinforced Panels with Rods with 2-in. and 4-in. Spacing
(Mesh reinforcements 3 layers of 1/2 in. 19-9a. galvanized hardware cloth on one side, 2 layers on other)
'Inel
Dishing
Max. circle
Rod Reinforcement of cracks, Mode of FI11 ure
No. in 1/16 in.
in.
44 Double-drawn - high-tensile A82 rods 4 20 Very fine radial cracks.
spaced at 2-in. centres One fine rectilinear crack.
55 Deformed double-drawn - high-tensile 10 24 Radial cracking.
A82 rods spaced at 4-in. centres Fine rectilinear cracks.
Centre spall, 6-in. dia.
TABLE 21. Flexure Strength of Panels with Rods with 2-in. and 4-1n. Spacing
(Mesh reinforcements 3 layers of 1/2 in. 19-91. galvanized hardware cloth on one side, 2 layers on other)
Pinel
First
Maximum
Modulus of
Rod Reinforcement Visible (Rupture, Hode of Failure
No.
Crack, lb
Load, lb
psi)
41A 0.225-in. double-drawn 1100 2240 3930 Main crack at load point.
high-tensile A82 rods spaced No rod slip apparent.
418 at 2-in. centres 1700 2900 5080
Main crack at load point.
Rod s 11 p noted.
53A As in 41A, 418 1100 1840 3230 Fine cracking.
No rod slip apparent.
538
a
1900 3080 5400 Main crack at load pOint.
No rod slip apparent.
54A 0.225-in. deformed A82 1000 1460 2560 Main crack at load point.
rods spaced at 4-in. Fine cracking between
centres load pOints.
Rod slip noted.
52
The effect of kind of rods used was examined by panels with 3
layers of 1/2-in. 19-9a. galvanized hardware cloth above and 2 layers below
the following rod reinforcements spaced on 2-in. centres:
0.225 in. dia. double-drawn high-tensile A82 rods (panels 41 t 52)
0.225 in. dia. deformed double-drawn high-tensile A82 rods (panels 53, 54)
1/4 in. dia. galvanized C1020 steel rods (panel 49)
1/4 in. dia. bright C1015 nail wire rods (panel 50)
1/4 in. hot-rolled C1020 rods (panel 51)
These were compared only under flexure loadings. The results
presented in Table 22 indicate a significant difference in the flexure strengths
of panels made with the various kinds of rods used. The hot-rolled rods gave
the best ultimate flexure strength. This would seem to be the result of
better rod/mortar bond characteristics which delayed rod slippage. In an
"infinitely 10ngll panel or in one in which the rods were cross-welded or held
as at bow and stern, the difference is not likely to be important. Further,
the diameter of the lower strength hot-rolled, galvanized, and nail wire rods
is 0.25 in. Their section modulus is about 25 percent larger than that of
the 0.225 in. dia. double-drawn rods. Also, the larger diameter forces the
layers of mesh farther from the "neutral axis" of the panel thereby making
the mesh more effective.
It is concluded that:
- rods make a significant contribution to the drop-impact resistance
and flexure strength of mesh-reinforced panels.
- rods at 2-in. spacing give better drop-impact resistance and
flexure strength to mesh-reinforced p a n e 1 ~ than rods at 4-in.
spacing.
- the kind of rods used is probably not of great importance to
the strength provided that the rods can give a we11-faired hull
(no kinks at cross-over points) and meet any other requirements,
such as corrosion-resistance, weldabi1ity, repairability, and cost.
the smaller diameter of the double-drawn high-tensile wire (for
the same breaking strength), its resilience, and cost/ft. seem to
favour its use.
ii. Effect of Mesh
The behaviour of the various meshes was evaluated by a series
of test panels reinforced only with mesh reinforcement. No rod reinforcement
was used. In the first series of tests, 30 x 30-in. panels were made in a
plywood mould lined with plastic sheeting. Slightly-staggered 29 1/2 x
26 1/2-in. layers of reinforcement mesh were laid in the mould. All layers
were oriented in the same direction. Layers were added until a reinforced
53
thickness of 1/2-in. was obtained. The overall, was about 3/4 in.
The mortar (Type II cement 1 part, dry mortar sand 2 parts) was carefully
mixed and laid into the panel form mould with a vibrating trowel. The stripped
panels were wetted regularly and cured for 28 days under plastic sheeting.
The reinforcements used in these tests were:
2.5-lb. galvanized expanded metal lath
1/2-in. l6-ga. galvanized welded square mesh
1/2-in. 22-ga. hexagonal mesh, galvanized after weaving
1/2-in. 22-ga. hexagonal mesh, galvanized before
1/4-in. 20-ga. fire screening
1/2-in. 19-9a. galvanized hardware cloth
1/2-in. l6-ga. welded square mesh, galvanized removed
3/8-in. 20-ga. welded square mesh
The number of mesh layers and weight of mesh per square foot
of panel are given in Table 23.
The panels were sectioned, as shown in Fig. 3, to yield various
test specimens.
. The panel specimens containing equal thicknesses of mesh were
tested under drop-impact and flexure load conditions. The impact test was a
50 1b weight (9-in. bottom radius) dropped 10 feet onto a 15-in. square specimen
supported on 3/4-in. plywood as described earlier and in Figs. 4 - 6.
The damage to the specimens was evaluated both qualitatively and quantitatively.
The evaluation is presented in Table 24. The visual examinations and the
measured concave deflections (dishing) 'rated the panel constructions as
,follows:
slight damage
1/2-in. l6-ga. galvanized welded square mesh
3/8-1n. 20-ga. black welded square mesh
moderate damage
1/2-in. 19-9a. galvanized hardware cloth
1/2-1n. 22-ga. galvanized hexagonal mesh
severe damage
2.5-lb expanded metal lath
1/4-in. 20-ga. fire screening
54
TAILE 22, Flexural Strength of Mesh-Reinforced Panels with Various Kinds of Rod Reinforcement
(3 layers of 1/2-1n. 19-9a.galvanized hardware cloth on one side of rods, 2 layers on other s ) ~
Panel
First Maximum Modulus
Rod Reinforcement Visible Load, of Rupture, Mode of Failure
No.
Crack, lb lb psi
41A 0.225-in. double-drawn high-tensile 1100 2200 3930 Main crack at load point.
A82 rods spaced at 2-in. centres No rod slip apparent.
418 1700 2900 5080 Main crack at load pOint.
52A 1300 1870
Rod slip noted.
3270 Uniform cracking.
528 2000 2050 3600
No rod slip apparent.
Several fine cracks.
No rod slip apparent.
54A Deformed 0.225-in. A82 rods 1000 1460 2560 Major crack at load point.
Rod slip noted.
53A 1100 1840 3230 Fine cracking.
538 1900 3080
No rod slip apparent.
5400 Main crack at load point.
No rod slip apparent.
49A 1/4-in. galvanized C1020 rods 1300 1530 2680 Main crack at load point.
No rod slip apparent.
498 1300 2500 4910 Several fine cracks.
Rod slip noted.
50A 1/2-in. bright C1015 nail wire rods 1300 2350 4120 Main and fine cracks.
508 2320 2400
Rod sHp noted.
4210 Uniform crackinq.
Rod s 11 p noted.
51A 1/4-in. hot rolled C1020 900 2900 5080 Uniform cracking.
No rod slip apparent.
518 1500 3600 6310 Main crack at load point.
No rod slip apparent.
A transverse rods in tension side, S & longitudinal rods in tension side.
TABLE 23. Description of Panels Reinforced with Mesh Only and Used for Strength Evaluation of Mesh-Reinforced
Panels
(All mortars Type II cement - 1 part, dry mortar sand - 2 parts)
Panel N(
I Weight
No.
Mesh Reinforcement
LIoyer:.
. Mesh, 1bl
sq ft panel
3, 13
2.5 lb. expanded metal lath, oalvanized
5 1.23
4, 12 1/2-1n. 16-ga.welded square mesh, galvanized 5 2.85
5. 14 l/2-in. 22-ga.hexagonal mesh, galvanized after weaving 12 1.35
6. 16 1/2-1n. 22-ga.hexagonal mesh, galvanized before weaving 12 1.29
7. 15 1/4-in. 20-ga.fine screening, oiled, cleaned 2 1.20
8. 11 1/2-1n. 19-9a.hardware cloth, galvanized 9 1.59
9 1/2-1n. 16-ga. welded square mesh, zinc removed 5 2.85
10 i 3/8-1n. 20-ga.welded square mesh, black 7 1.59
i
I
I
I
I
55
The f1ehJre strength, calculated as the modulus of r.upture or
maximum fibre stress at ultimate failure of the beam, was determined for
lengthwise and transverse 3 x 12-in". cut from the " panels containing
various mesh reinforcements. Third-point loading on a 10-in. span was used in
the manner described in ASTM C78 Flexural Strength of Concrete (using simple
beam with third-point loading). The deflections under increments of loads,
the load at the first visible crack, the maximum load held, and the mode of
failure were reCOr?ad. The "modu1us of rupture II at the maximum load held was
calculated as R = where W = load, b = breadth, d = "thickness of the
specimen. The values for the lengthwise and transverse tests are
summarized in Table 25. Fig. 11 shows typical load-deflection curves.
The results show the marked superiority of the 1/2-in. 16-ga.,
3/8-in. 20 ga., and 1/2-in. 19-9a. welded square meshes in flexure. The
hexagonal mesh and expanded metal lath are reasonably strong in the longitudinal
direction but are poor in the transverse direction.
It is concluded that the square mesh reinforcements are superior
to the others in both drop-impact resistance and flexure strength. Those mesh
reinforcements which have very poor transverse properties, i.e. metal lath
and hexagonal mesh, are deemed unsuitable except where they can be oriented
so that the transverse loading is negligible or where rod-reinforcement
carries these loads. Figs. 12 and 13 show the inferior crack distribution
of the transverse specimens.
The major shortcoming of the square meshes and the obvious
advantage of the hexagonal mesh is the conformabi1ity for compound curves.
Fire screening is deemed completely unsuitable.
The relationship between the maximum flexural strengths obtained
and the weight of reinforcement per square foot of panel, regardless of the
kind of reinforcement, appears to be linear. Fig. 14 shows the plotted strengths.
In addition to the foregoing tests which compared strengths
of panels with equal thicknesses of several mesh reinforcements, another set
of panels with equal weights of the several mesh reinforcements was also
tested in a similar manner. Four 30 x 30-in. panels were made with the
following constructions:
3 layers of 1/2-in. 22-ga. galvanized hexagonal mesh
1 layer of 1/2-in. 16-ga. galvanized welded square mesh
2 layers of 1/2-in. 22-ga. galvanized hexagonal mesh
56
',nel
110.
4.t.12
10
8.11
6.16
3.13
7.15
'anel
No.
3.13
4.9.12
5.14
6.16
7.15
8.11
10
TABLE 24. urop-Impact Resistance of Panels Reinforced with Various Meshes
(All panels have lIZ-in. thickness of reinforcement mesh)
Reinforcement
Dishing
Description
lblsq ft in
Description of Mode of Failure
1/16 in.
1/2-in. 16-ga. galvanized 2.S5 3 No visible cracks in top surface. Some fine
welded square mesh radial and rectilinear cracking to of
bottom surface.
3/S-in. 20-ga. black 1.59 6 Five ring cracks in top surface. Slightly
welded square mesh open rectilinear and radial cracking in
bottom surface.
1/2-in. 19-9a. galvanized 1.79 9 Moderate open ring crack in top surface.
hardware cloth Slightly open ring cracking in bottom
surface.
1/2-in. 22-ga. hexagonal ' 1.35 15 Large open ring cracks in top surface.
mesh, galvanized after Shear spalling, open radial cracks,
weaving and some broken wires in bottom surface.
2.5-lb galvanized 1.23 . 17 Open ring and transverse cracks in top
expanded metal lath surface. Large diagonal cracks and
torn mesh in bottom surface.
1/4-in. 20-ga. black 1.20 22 Extremely severe major rinq cracks in top
fire screening surface. Open radial cracks and mortar
crumbling in bottom surface.
TABLE 25. Flexure Strength of Panels Reinforced with Various Meshes
(All panels have 1/2-in. thickness of reinforcement mesh)
Reinforcement
Average Modulus
lb/sq ft of Rupture, psi
Mode of
Description Failure
Panel
Longit. Transv.
2.5-lb galvanized 1.23 3690 S50 Longit. specimens cracked
expanded metal lath over wide area, trans-
verse mainly single
crack. Broken mesh.
1/2-in. 16-ga. welded square mesh, 2.85 65S0 6040 Cracking at l-in.
galvanized and stripped intervals on bottom.
Top spalling.
1/2-fn. 22-ga. hexagonal mesh, 1.35 2660 1485 Lonqit. specimens
9alvanized after weaving cracked over wide area,
transverse mainly sinqle
crack.
l/2-in. 22-ga. hexagonal mesh, 1.29 3040 1275 As above.
galvanized before weaving
1/4-in. 20-ga. black fire screening 1.20 1395 950 Mortar crumbled badly
in bottom surface at one
or two major cracks.
1/2-in. 19-9a. galvanized 1.79 3560 3565 Both longitudinal and
hardware cloth transverse specimen
cracked at 1/2-in.
intervals over wide area.
3/8-in. 20-ga. black welded 1.59 5000 4460 Both lonqitudinal and
square mesh transverse specimens
cracked at 3/8-in.
intervals over wide span.
3000
2000
Load
1b
1000
o
o O. 1
welded square mesh
1/2-in. 19-9a. hardware cloth
llE
10E
3IB-in. 20-ga. welded square mesh
14D
l/2-in. 22-ga. hexagonal mesh
0.2 0.3 0.4
Def1ecti on, in.
Fig. 11. Flexure tests on mesh-reinforced beams.
57
0.5
58
..
,
' -
. ~ . ' ) l IG <J<'l.
..
r o - . ... ..., _V I ~ C
. 1
Fig. 12. Bottom (tenston) surface of flexure specimens bent with
bottom layer of mesh in transverse direction (left specimens)
or longitudinal direction (right specimens).
'{
t -
. )
\
~
-
I ~
If:

Fig. 13. Bottom (tension) surface of flexure specimens bent with
bottom layer of mesh in transverse direction (left specimens)
or longitudinal direction (right specimens).
7000
5000
Modulus
of
Rupture
psi
3000
1000
o

o
or.
<to>
ItS
r-
r-
ItS
<to>
CIJ
E
"0
CIJ
"0
c:
ItS
Q.
><
CIJ
.Q

..-
U')

.
or.
en
CIJ
E
r-
ItS
c:
0
0'1
ItS
><
CIJ
or.
.
ItS
0'1

N
N
c:
-

N
........
r-
1




.
ItSClJor.
O'ls.-en
IItSCIJ
O::SE
N CT
en
.
c:"O
.r- CIJ
"0
co r-
........ CIJ
M3:



.
ItS
0'1
I
0'1
r-
.
c:
.r-
I
N
........
r-
2
Mash Reinforcement - lb/sq.ft. panel
Fig. 14. Modulus of rupture vs. weight of mesh
reinforcement in strongest direction
. 59
.e-
en
CIJ
E
CIJ
s.-
10
::s
CT
en
"0
CIJ
"0
r-
CIJ
3:
c:
.r-
I
N
........
r-
3
60
10 layers of 1/2-in. 22-ga. galvanized hexagonal mesh
2 layers of 1/2-in. l6-ga. galvanized square mesh
6 layers of 1/2-,;n. 20-ga. galvanized hardware cloth.
The weight of the mesh reinforcement in each panel was about
1.15 lb/sq. ft. of panel. The mortar in all panels was Type II cement 1 part,
Del Monte sand 2 parts.
Specimens cut from the panels were tested as before in flexure
and drop-impact. The results are tabulated in Tables 26 and 27.
The drop-impact of the panel containing two layers
of 1/2-in. 16-ga. galvanized welded square mesh was markedly superior to the
panels containing other meshes. The flexure tests, performed as before, .
showed the panel containing the two layers of 1/2-in. 16-ga. galvanized welded
square mesh to but the failure was a single bottom crack. The
specimens from the panel containing six layers of 1/2-in. 19-9a. galvanized
hardware cloth was considerably stronger than those from the panel containing
the ten layers of 1/2-in. 22-ga. hexagonal mesh and had a somewhat superior
distribution of bottom cracking. Both constructions, however, had a better
crack distribution than that of the panel containing t.he 1/2-in. 16-ga. weld
mesh.
Hulls are most often reinforced with rods between layers of
mesh. Therefore, additional tests were undertaken to determine if rods
diminished the differences between the longitudinal and transverse strengths
of panels reinforced with the several kinds of meshes.
Six 36 x 36-in. panels were made in the free-standing panel
moulds with 0.225 in. double-drawn high-tensile A82 rods at 2-in. spacing.
The three pairs of specimens contained approximately equal weights of 1/2-in.
19-9a. galvanized hardware cloth, 1/2 in. 16-ga. galvanized welded square
mesh, or 1/2-in. 22-ga. hexagonal mesh, viz.,
Panels 41,44 3 layers 1/2-in. 19-9a. galvanized hardware
cloth on one side of rods, 2 layers on other side
42, 45 1 layer 1/2-in. 16-ga. galvanized welded square
mesh lath on each side of rods.
43. 46 5 layers of 1/2-in. 22-ga. galvanized hexagonal
mesh on each side of rods.
61
TABLE 26. Drop-Impact Resistance of Panels Reinforced with Equal Weights of Mesh
Panel
Rei nforcement,
No. Iblsq ft
D1Shing Mode of Failure
Description
Panel
1/16 in.
27 3 layers 1/2-1n. 22-ga. hexagonal mesh 1.14 14 Top - moderate r1nq
1 layer 1/2-in. 16-ga. welded square mesh Bottom - radial crack1nq over
2 layers 1/2-1n. 22-ga. hexagonal mesh 12-in. area, exposed mesh,
broken wires.
28 10 layers 1/3-1n. 22-ga. hexagonal mesh 1.15 16 Top - severe cracking.
Bottom - some sheer spalling,
radfal cracking over 12-1n.
area, broken wires.
29 2 layers 1/2-fn. 16-ga. welded square mesh 1.14 7 Top - fine part ring crack.
Bottom - moderate rectilinear
crackfnq over 3-1n. area.
30 6 layers 1/2-1n. Ig-ga. cloth 1.19 15 Top - moderate ring cracking.
Bottom - radial and recti-
linear cracking over 10-in.
area.
TABLE 27. Flexure Strengths of Panels Reinforced with Equal Weights of Mesh
Panel
Reinforcement
Modulus of
No.
Description
Ib/sq ft
Rupture, ps 1 Mode of Failure
Panel
long1t. Transv.
27 3 layers 1/2-in. 22-ga. hexagonal mesh 1.14 4130 3600 long1t. - bottom cracks
1 layer 1/2-1n. 16-ga. welded square mesh over 4-1n. span.
2 layers 1/2-1n. 22-ga. hexagonal mesh Transv. - bottom cracks
over 4-in. span.
28 10 layers 1/3-in. 22-ga. hexagonal mesh 1.15 2640 1650 longft. - bottom cracks
over 4-fn. span.
Transv. - Single major
bottom crack.
29 2 layers 1/2-1n. 16-ga. welded square mesh 1.14 5350 4520 longit. - single major
bottom crack.
Transv. - single major
bottom crack.
30 6 layers 1/2-1n. 19-9a. hardware cloth 1.19 3460 2990 long1t. - bottom cracks
over 3-fn. span.
Transv. - bottom crack
over 2-1n. span.
.62
Drop-impact tests were performed using a 31 x 31-in. supporting
frame but the same impact load device, i.e. 50 lb dropped 10 ft. The drop-
impact results, presented in Table 28, show that the panel reinforced with the
1/2-in 22-ga. galvanized hexagonal mesh on rods has a drop-impact resistance
inferior to that of the panels reinforced with similar weights of either 1/2-in.
19-9a. galvanized hardware cloth or 1/2-in 16-ga. welded square mesh.
Two 12-in. wide flexure test specimens, longitudinal and trans-
verse were cut from adjacent sides of the panels 41, 42, and 43 {mates to
panels 44, 45, and 46,respective1y}. These flexure specimens were tested
1n third-point loading on a span of 21 inches. Mid-span deflections were
measured. Loading was continued until the load dropped significantly. The
loads at first visible crack and at maximum load (modulus of rupture) were
recorded. The results are -presented in Table 29. The load deflection curves
are shown in Figs. 15 and 16.
The following observations and conclusions may be made from
these tests on panels which contain rods and approximately equal weights of
three different meshes:
- the presence of rods tends to minimize the differences in flexure
strengths obtained in panels containing the three meshes, viz.
1/2-in. 19-9a. hardware cloth, 1/2-in. 16-ga. welded square mesh, and
1/2-1n. 22-ga. hexagonal mesh.
- the rods diminished the difference in flexure strengths which existed
in the lengthwise and transverse directions of the 1/2-in. 22-ga.
hexagonal mesh.
- the flexure strength of the panels containing the 1/2-in. 22-ga.
hexagonal mesh is about equal to the strength of the panel containing
the 1/2-in. 19-9a. hardware cloth as tested but would still be inferior
if the latter panel had had the 3 layers of mesh in the tension side
and the 2 layers in the compression side.
- the flexure strength of the panels is higher when the rods in the
tensile side of the specimen are in the lengthwise direction of the
specimen than when they are in the transverse direction.
iii. Effect of Rod and Mesh Orientation on Flexure Strength
The effect of the orientation of the mesh and of the rod/mesh
assembly with respect to the flexure stress direction has been examined in
two.ways. In the first test, specimens containing either 5 layers of 1/2-in.
16-ga. galvanized welded square mesh, 12 layers of 1/2-in. 22-ga. galvanized
Pln.'
No.
44
45
46
63
TABLE 28. Drop-Impact Resistance of Rod-Reinforced Panels with
Equal Weights of Mesh Relnforcements
Ofshfng
Circle
Mesh Reinforcement on of Mode of Failure
Double-Drawn A82 Rods Spaced on 2-1n. Centres
in
Cracks
1/16 in.
in.
2 layers 1/2-in. 19-9a. galv. hardware cloth 4 20 Very fine radial crack.
3


II

.
One fine rectilinear crack.
1 layer 1/2-in. 16-ga. galv. welded square mesh 8 18 Fine radlal cracks.



. .

. .
Fine rectilinear cracks.
5 layers l/2-in. 22-ga. galv. hexagonal mesh 9 30 Radial cracking.



II . II

TABLE 29. Flexural Strength of Rod-Reinforced Panels With Equal Weights of Mesh Reinforcements
Pinel
Mesh Reinforcement First
Maximum
Modulus of
No.
(listed ln order from Visible
load Ib
Rupture Mode. of Failure
compression to tenslon sldes) Crack Ib psi
41A 3 layers 1/2-1n. 19-9a. hardware cloth 1100 2240 3930 Mafn crack at a load polnt.
lengthwise rods Several fine cracks.
transverse rods No rod slip apparent.
2 layers of 1/2-in. 19-9a. hardware cloth
418 3 layers 1/2-1n. 19-9a. hardware cloth 1700 2900 5080 Main cracks at load points.
transverse rods Rod slip noted.
lengthwl se rods
2 layers 1/2-1n. 19-9a. hardware cloth
42A 1 layer 1/2-1n. 16-ga. welded square 1300 2800 6050 Main crack at load pOint.
mesh Rod slip noted.
lengthwise rods
transverse rods
1 layer of 1/2-1n. 16-ga. welded square
mesh
428 1 layer 1/2-1n. 16-ga. welded square 1950 3240 7000 Fine cracks at load points.
mesh Rod sl1p noted.
transverse rods
lengthwise rods
1 layer 1/2-in. 16-ga. welded square
43A 5 layers of 1/2-1n. 22-ga. hexagonal 1800 2650 4740 Main crack at a load polnt.
mesh Fine crack at other polnt.
lengthwlse rods No rod slip apparent.
transverse rods
5 layers 1/2-1n. 22-ga. hexagonal
mesh
438 5 layers of 1/2-in. 22-ga. hexagonal 1400 2700 4740 Major crack at a load polnt
mesh Rod slip noted.
transverse rods
lengthwise rods
5 layers of 1/2-1n. 22-gl. hexagonal
IIesh
3000
Load
1b
2000
1000
o
64
42A 1/2-in. 16-ga. welded square
mesh
41A 1/2-in. 19-9a. hardware cloth
o
A82 rods on 2-in. centres
tension side rods-transverse
Beam - 12 x 24 x 1 in.
Span - 21 in.
Load - third-points
0.5
Deflection, in.
Fig. 15. Flexure strength of beams with equal weights
of three meshes.
1.0
3000
Load
lb
2000
1000
o
o
42B l/2-in. 16-ga. w ~ l ~ e d square mesh
43B 1/2-in. ?2-ga. hexagonal mesh
A82 rods on 2 in. centres
tension side rods - lengthwise
Beam - 12 x 24 x 1 in.
Span - 21 in.
Load - third-point
0.5
Deflection, in.
Fig. 16. Flexure strength of beams with equal weights
of three meshes.
65
1.0
66
hexagonal mesh, 9 layers of 1/2-in. 19-9a. galvanized hardware cloth, or
7 layers of 3/8-in. 20-ga. black welded square mesh were cut from the diagonal
direction of the panels for comparison with the corresponding specimens
c o n t a i n i n ~ mesh in the longitudinal and transverse directions. The flexure
strength (modulus of rupture) of the diagonal specimens ranged from about 65
to 90 percent of the longitudinal and transverse specimens, except in the
case of the transverse specimens from the hexagonal mesh panel which exhibits
very anisotropic properties. As might be expected the strength of the diagonal
specimen was appreciably higher than that of the transverse specimen. The
results are shown in Table 30.
One panel was reinforced with 3 layers of 1/2-in. 19-9a.
hardware cloth above and 2 layers below the 0.225 in. double-drawn rods at
2-in. spacing. The mesh layers were placed at 45 degrees to the rods. The
flexure strength (modulus of rupture) of a flexure specimen oriented in line
with the mesh and at 45 degrees to the rods was only about 60 percent of
the strength of a longitudinal or transverse specimen cut from a similar
normal panel. The strength of a specimen oriented in line with the rods and
at 45 degrees to the direction of the mesh was only about 50 percent of the
strength of a normal panel, as shown in Table 31.
iv. The Hydrogen Gas Problem
Two of the panels made early in the program showed an occasional
blister or eruption in the mortar surface soon after mortaring was complete.
These panels had, as reinforcement, only 1/2-in. 19-9a. hexagonal mesh,
galvanized before weaving. The hexagonal mesh was known to have an extremely
light coating of zinc. Later, the first panels with bare steel rods and
various galvanized meshes showed much more severe blistering along the lines
of the reinforcing rods and especially at intersections of the transverse and
longitudinal rods. Christensen and Williamson
19
clearly explain how a galvanic
cell is set up between the black steel rods (cathode) and the galvanized mesh
(anode) in the cement mortar paste (electrolyte). The resulting galvanic
reaction liberates hydrogen ions which form nascent hydrogen atoms and hence
hydrogen molecules. The hydrogen gas forms blisters in the setting mortar
which may lead to poor rod/mortar bond strength and,possibly, less protection
of the rod against corrosion attack. The hydrogen at the rod surface may
cause hydrogen embrittlementin high-tensile reinforcement. The problem seems
to have been of little importance in practical boat-building but the potential
seriousness warrants steps to prevent the reaction. Christensen and Williamson
have suggested a number of possible solutions to the problem, viz.
- eliminate dissimilar metals by:
- using ungalvanized rods and mesh
- using galvanized rods and mesh
Panel
No.
4
5
8
10
Panel
No.
41A
41B
5eA
58B
67
TABLE 30. Effect of Mesh Orientation on Flexure Strength of Mesh-Reinforced Panels
Modulus of Rupture
Mesh Reinforcement
Longit. Transverse Mesh at 45 deqrees
pst pst
~ s i S of Longit. -: of Transv.
1/2-in. 16-ga. galvanized 59oo 6130 4700 80 77
welded square mesh, 5 layers
1/2-in. 22-ga. galvanized hexagonal 2900 1360 1950 67 143
mesh, 12 layers
1/2-in. 19-9a. galvanized hardware 3520 3530 3140 89 89
cloth, 9 layers
3/8-in. 20-ga. welded square mesh, 5000 4460 3720 75 83
7 layers
TABLE 31. Effect of Mesh and Rod Orientation on Flexure Strength
(3 layers of 1/2-in. 19-9a. hardware cloth above 2 layers below 0.225 in
dOuble-drawn rods oriented at 45 degrees to the mesh)
Load at
Orientation of Flexure Modulus
Specimen with Respect
First Maximum
of Rupture Mode of Failure
Visible Load, lb
to Direction of Reinforcement
Crack, lb
psi
Mesh and rods in same direction as 1100 2240 3930 Main crack under load
specimen.
point.
No rod slip apparent.
~ l e s h and rods in same direction a ~ 1700 2900 5080 Main crack under load
specimen. point.
Rod slip noted.
Mesh at 45 degrees to and rods in 500 1220 2140 Fine cracks between
direction of specimen length. load points.
Rod slip noted.
Rods at 45 degrees to and mesh in 1600 180U 3150 Many cracks under load
direction of specimen length.
points.
Rod slip noted.
',-
- applying a protective insulating coating to the rods
- chemically or inhibiting the galvanic cell action
As they rightly point out, ungalvanized mesh would be threatened by rapid
corrosion in the marine environment, galvanic action might still occur between
galvanized rods and mesh, cold-drawn high-strength rods could not be galvan-
ized without some loss of strength, and the integrity of protective coatings
on rods could not be guaranteed. They showed that chromium trioxide prescribed
for passivating or inhibiting zinc in concrete was a simple effective cure
for the galvanic cell problem in ferrocement construction. The concentrations
of chromium trioxide used, 100 to 300 ppm by weight of water, effectively
passivated the zinc coating. As has already been pointed out, the Building
Research Station in the United Kingdom
16
states that 70 ppm chromium trioxide
in the cement paste (0.0035 percent in the dry cement, assuming a water/cement
ratio of 0.5) will inhibit the formation of hydrggen gas. In later tests,
Cornet et a1
20
showed that the mechanical properties of concrete, viz. compres-
sive strength and elastic modulus, were not impaired when chromium trioxide
was added to 100, 200, and 300 ppm water.
Many normal
trioxide. Others may not.
galvanized steel should be
are made to the mix.
Portland cements contain sufficient soluble chromium
It is therefore recommended that chromate-treated
used with all cements unless chromic acid additions
As a result of the problem and of the investigations performed
elsewhere, chromium trioxide (chromic oxide, Cr03) was added to the water for
the mortar mix at a concentration of 300 ppm (0.3g/1 or 0.3 lb/100 Imp. gallons
water) for all panels subsequently made in this laboratory.
v. Corrosion
Under most conditions Portland cement concrete is considered
to provide good protection of the encased steel reinforcement against corrosion.
The protection is said to be provided by the high alkalinity of the concrete.
However, even under covers of two to three inches, steel reinforcement is
known to have corroded where voids have allowed easy ingress of corrodents
and oxygen. Thick covers of cement mortars are not possible in ferrocement
boat construction. The mortar cover seldom will exceed 1/8 in. In some
areas mesh may even be exposed during the fairing of the hull. Although the
mortar has a low content of voids, microcracking may be present.
Many investigators have examined corrosion of reinforcement
bars and in concrete. Much of the work has been directed to examining
the corrosion of reinforcements in prestressed concrete water reservoirs and
fixed marine structures. Although the reservoirs are considered to be concrete,
a layer of less-permeable mortar is often applied to the prestressing wires
by pneumatic placement to prevent corrosion of the wires. Even so, catas-
trophic failures have occurred.
69
Ferrocement hulls, possibly one inch or less in thickness,
with a fine steel reinforcement covered by a thin layer of rich fine mortar
of relatively low permeability and intermittently wetted with seawater,
present a different problem which has not been adequately examined.
The mechanisms which pertain to the corrosion and protection
of steel reinforcement are briefly reviewed. Concrete or mortar normally
protects embedded steel from corrosion since the steel in concrete is
polarized anodically. This polarization results in the formation of a
passive film of gamma iron oxide on the steel. Corrosive anions, such as
chloride ions may break down this passive film in the presence of oxygen
and moisture.
21
, 22 The passive film will break down when ,sufficient
chloride ions have penetrated to the steel surface. Shalon and Raphae1
23
and Hausmann
24
have suggested that ~ threshold concentration of chloride
ions must be exceeded before corrosion occurs. Gjrv
25
explains that the
electrochemical potential of the steel becomes more negative (anodic) as
the passivity of the steel becomes partly, or completely, broken while
other portions of the steel with passive potential intact act as cathode.
Since moist concrete (mortar) is a good electrolyte, galvanic cells develop
along the steel wires. The electromotive force in such cells depends mainly
on the pH value (hydrogen ion concentration) and chloride concentration in the
moisture next to the steel and on the transfer of dissolved oxygen through
the concrete (mortar) cover.
Ferrocement hulls have achieved a reputation in several areas,
especially in corrosion resistance and durability. The existence of a boat
hull over a hundred years old is often cited as testimony. It is generally
agreed that integrity of the mortar is a most important factor in preventing
attack of reinforcement. Permeability is a measure of the capacity of the
mortar to transmit fluids. If the pores are not connected, the permeability
will be low and corrosion less likely. However, as Lewis and Copenhagen
26
, 27
point out, there is ample evidence that concrete is permeable to varying degrees
depending on such factors as water/cement ratio, aggregate/sand graduation,
richness of the mix, compaction, and curing. Since the passivating action
of concrete on steel depends on its ability to exclude inimicable ions from
the steel surface, permeability is clearly of utmost importance.
Well-made ferrocement has relatively low permeability. This
laboratory has shown no passage of water through a 3/4-in. section of
unreinforced ferrocement mortar subjected to a hydrostatic head of three feet.
Kowalski
28
cites numerous U.S.S.R. tests which maintained a hydrostatic pressure
of 230 psi on ferrocement only 20 cm thick without water percolation. However,
ferrocement contains microcracks and may, when damaged, contain macrocracks.
Walkus,29, 30 who has performed valuable work on the state of cracking in
ferrocement, defines microcracks as those cracks not exceeding 20 microns
(0.02 mm). Walkus and Kowalski
31
claim that cracks between 20 and 50 microns
develop during quasi-elastic deformation and that loads producing cracks up
70
to 50 microns wide are in the practical elastic working range. Further,
when a crack reaches 100 microns (0.1 mm), the steel reinforcement alone
carries all the tensile forces. Wa1kus and Mackiewicz
32
claim that the
material is completely watertight if the width of the cracks does not
exceed 20 microns (0.02 mm) and that corrosion does not occur if the width
of the cracks does not exceed 100 microns (0.1 mm). No explanation or
laboratory evidence to support this statement were provided in the reference
although other papers by the authors which have not been obtained may indeed
provide the evidence. It was considered necessary to carry out simple,
practical tests under known conditions of exposure.
Several preliminary tests were undertaken early in the program
period to assess the corrosion of steel reinforcements in ferrocements. The
first test specimens to assess the durability of various mortars to seawater
were also evaluated qualitatively for the condition of the encased reinforce-
ment. The panels from which the test coupons were cut were made in a horiz-
ontal panel mould lined with polyethylene sheeting. The bottom layer of
mesh could be clearly seen imprinted in the bottom surface of the panel,
i.e. there was only a thin film or no mortar covering the mesh wires. The
specimens, of the "standard" mor.tar mix (Type II cement and dry mortar sand)
contained the following meshes:
1/2-in. 16-ga. welded square mesh, galvanized
1/2-in. 16-ga. welded square mesh, zinc stripped
1/2-in. 22-ga. hexagonal mesh, galvanized after weaving
1/2-in. 22-ga. hexagonal mesh, galvanized before weaving
1/2-in. 19-9a. hardware cloth, galvanized
The specimens were immersed in seawater for 8 hours, dried at
room temperature for 16 hours. Changes ' were observed during the 15 test cycles.
The wires of mesh with no zinc coating produced red rust stains on the mortar
surface after two cycles. The corrosion became more severe as the cycles
increased. The hexagonal mesh with the light zinc coating applied before
weaving showed rust stains after ten cycles. The other meshes showed no
rust after 15 cycles.
Five 3 x 12 in. specimens used in flexure tests were exposed
in seawater at room temperature. These specimens had suffered severe damage
in the flexure test such that much of the mesh was highly exposed. Seawater
could easily penetrate deep into the specimens. A copious yellow-white
encrustation was quickly developed on the galvanized me$hes and severe red
rust staining was observed on the wires of the 3/8-in. 19-9a. coppered or
liquor-finished mesh. (This wire treatment is intended to provide only
temporary corrosion protection for shipping and storage.) .
71
The effects of crack width and thickness of mortar cover on
the corrosion attack of unga1vanized reinforcement wire encased in mortar
were examined in an extensive program of tidal seawater exposures. The tidal
seawater exposure was chosen because oDservations in the field have indicated
that the most parts of concrete structures are those parts most
exposed to intermittent wetting and drying within the splash zone. It has
been shown that if concrete becomes completely water-saturated, the corrosion
becomes less severe.
A mortar specimen with a reduced section at mid-length and
one reinforcing wire was developed which could be cracked in a
controlled manner to yield cracks of specific widths. Multiple-cavity moulds,
which allowed 50 specimens or coupons to be made from each of eight batches
of mortar mix, were constructed. The cavities had a draft allowance and
removable dividers to permit easy removal of the specimens from the mould.
Slots to allow embedment of the wires in the mortar to a predetermined mortar
coyer were located in the ends of the moulds. The size of the specimens is
6 cm. wide by 14 cm. long by 2 cm. thick (2 1/2 x 5 1/2 x 3/4-in.).
The width at the reduced neck section is 3 cm. (1 1/4-in.).
The composition of the several mortar batches used, as in all
recent work carried out in the program, is Type II cement one part, -8 mesh
dry mortar sand two parts, water/cement ratio 0.45, slump 2 1/2 to 3-in.
The mortar was placed in each cavity and trowelled smooth. A 10-in length
of 0.105 in. dia. "cut and straightened" carbon steel wire (breaking strength
of about 900 lb.) was set into the locating grooves, pressed into the mortar
coupon, and covered with small dabs of fresh mortar. The coupon was again
trowelled smooth. After curing, the mortar specimen was carefully ground
to a cover thickness in the necked section of 0.5, 2.0, 3.5 mm. before
the mid-length crack was made.
Controlling cracking of the specimens was accomplished by
forcing opposing hardened steel rollers (1 1/2-in. dia.) into the notches
each side of the specimen, i.e. into the reduced section, until a crack
of the required size was obtained. The crack widths are as follows:
o (no crack), 0.05 mm (finest crack visible to unaided eye), 0.1 mm,
mm, and 0.5 mm.
72
The crack widths were measured with a standard Brine1l
hardness test microscope containing a micrometer-scale rettcule graduated
in 0.1 mm. which allows measurements to be estimated to 0.05 mm. Specimens
with uneven cracking, or otherwise unsuitable, were discarded. Figure 17
shows a typical set of specimens after exposure to the tidal seawater
environment. Five lots, each of 45 test specimens, were prepared from the
four hundred specimens moulded for exposures of 1, 2, 4, 8, and 16 months.
(A sixth set of 45 was held in reserve or as a control set.) The specimens
of each lot were in triplicate, i.e. three of each of the crack width/mortar
cover combinations for each of the five exposure periods. The tabulation
below shows one lot of specimens for one exposure period.
Crack Width,
Mortar Cover, mm.
Total No. of
mm.
'0.5 2.0 3.5
Specimens
o (no crack) 3 3 3 9
0.05 (hairline) 3 3 3 9
0.1 3 3 3 9
0.2 3 3 3 9
0.5 3 3 .3 9
Totals 15 15 15 45
The 45 specimens comprising each lot were tied together by
means of holes drilled through each specimen. The five lots were installed
on December 3, 1973 at the dock of the Kitsilano (Vancouver) station of the
Canadian Coast Guard, Ministry of Transport at an immersed depth five feet
below high tide.
One lot of specimens was removed after 1, 2,4, and 8 months
of exposure for examination and testing. The last lot has not yet been
retrieved but the trend has been established.
Each specimen was visually examined, one half of the mortar
was removed and the crack surface was examined for rust stains; the wire
was removed and examined under the stereoscopic microscope, and the depth
of pitting or ring corrosion was measured by a micrometer with pointed
anvil and spindle. Finally, the breaking strengths of the wires were
obtained. .
73
The visual assessment and micrometer measurements and the
breaking strengths of the wires are presented in Tables 32 and 33. A set
of wires after 8 months exposure is shown in Fig. 18 and graphical present-
ations of the breaking strengths of the wires are shown in Figs. 19 to 21.
The following observations are made on the basis of the above
tests:
- corrosion of the steel wires in mortar coupons containing cracks
not wider than about 0.1 mm is not severe, even after 8 months
exposure and with a mortar cover only 0.5 mm. thick.
- corrosion of the steel wires in mortar coupons containing cracks wider
than about 0.1 mm. is severe after 1 to 2 months exposure, regardless
of the thickness of the mortar cover, viz. 0.5, 2.0, and 3.5 mm.,
- the thickness of the mortar cover used, viz. 0.5, 2.0, and 3.5 mm. has
little effect on the severity of corrosion of the steel wire if a crack
is present.
- a mortar cover of 2.0 mm. substantially prevented corrosion of the
steel wires in uncracked mortar coupons, at least for eight months.
- a mortar cover of 0.5 mm. failed to protect the steel wire completely
for two months even in uncracked mortar coupons.
- it seems advisable to use a protective coating on the mortar cover
to protect ungalvanized reinforcement from corrosion, even when no
cracks are present and the mortar cover is greater than 2 mm.
- the loss of strength of unga1vanized rods which are much heavier and
at greater depths in the mortar is likely to be small under crack
conditions not susceptible to hull leakage. However, ungalvanized
mesh nearer the surface is likely to suffer severe loss of strength
from corrosion if cracks greater than about 0.1 mm. are present. Gal-
vanized is therefore recommended.
4. Evaluation of a Typical Ferrocement Construction
The strength data reported so far pertain to ferrocement
containing a wide variety of mortars and mesh and rod reinforcements, i.e.
a wide variety of ferrocement constructions. It seemed desirable to investigate
a "typica1
11
ferrocement construction in somewhat greater depth and greater
replication so that data of greater reliability could be It was
hoped that such data could used later in the development of a mathematical
model to aid design. .
74
Exposure
No.
1
2
4
8 '
TABLE 32. Assessment of Corrosion of 0.105 in. d1a. Wires in Hortar Coupons with Cracks
o to 0.5 nrn. wide and wlth 110rtar Covers of 0.5, 2.0, and 3.5 nrn.
After Tidal Seawater Exposures (worst of trlpllcate exposures recorded).
Mortar Appearance of wire and depth of pits on ring groove.
Cover
nrn. o (No crack) H (0.05 nrn) 0.1 nrn 0.2 nrn 0.5 nrn
0.5 no attack slight etch slight etch slight etch ring groove,
pit 0.001 in. 0.002 In. deep
2.0 110 attack slight etch sllght etch slight etch rlng groove,
shallow pit 0.003 ln deep
3.5 no attack no attack slight etch rlng etch ri ng groove,
0.004 In. deep
0.5 slight etch pits, rlng pits, ring plts, ring rl ng groove,
0.001 In. deep 0.001 In. deep 0.001 in. deep 0.003 in. deep
2.0 no attack pit., rlng rlng, plts, ring, pits, rlng,
0.01" in deep 0.002 In-. deep 0.002 in. deep 0.002 in. deep
3.5 no attack slight etch pits, pits, ring, ring groove,
0.001 in. deep 0.002 in. deep 0.008 in. deep
0.5 pits, plts, rlng, pits, plt, ring groove,
0.001 in. deep 0.001 In. deep 0.001 In. deep 0.005 In. deep 0.008 In. deep
2.0 slight etch pits, plts, slight etch rl ng groove,
0.001 In. deep 0.001 In. deep 0.006 in. deep
3.5 pits, pits, rlng, pits. plts. ring, ri ng groove,
0.001 in deep 0.001 in. deep 0.001 In. deep 0.001 in. deep 0.005 in. deep
0.5 pits, rlng, rlng groove. half ring, ring groove,
0.002 in. deep 0.001 In. deep 0.004 In. deep 0.020 In. deep 0.010 in. deep
2.0 no attack no attack slight etch half rlng, half ring,
0.010 in. deep 0.020 ln deep
3.5 no attack plts,
0.001 in. deep 0.001 In. deep 0.010 in. deep 0.025 In. deep
TABLE 33 Breaklng strength of 0.105 In. dia. Wires ln Mortar Coupons with Cracks a to 0.5 nrn. wide and
with Mortar Covers of 0.5, 2.0, and 3.5 nrn. After Tidal Seawater Exposures (Average of triplicate
values).
Hortar Average breaking strength of wire, 1b
Exposure
Cover,
no.
nrn.
a (No cracks) H (0.05 nrn.) 0.1 III!I. 0.2 nrn. 0.5 nrn.
1 0.5 893 885 908 921 911
2.0 897 892 900 886 891
3.5 867 919 912 898 877
l
~ : g
91? ~ O O
~ g :
902 881
893 909 904 884
3.5 910 902 903 891 804
4 0.5 915 891 9Il 875 886
2.0 891 911 904 898 848
3.5 901 904 915 895 842
8 0.5 9!8 895
~ 1 ~
U
6 6?4
2.0 903 903 731 686
3.5 916 912 907 808 546
-
.'
Fig. 17. One set of specimens after exposure of 4 months.
Crack size (left to right) 0, 0.05, 0.1, 0.2, and 0.5 mm.
Mortar cover is 2 mm.
Fig. 18. Close-up of attack in one set of wires after exposure of
8 months. Mortar cover by groups (left to right) 0.5, 2.0,
and 3.5 mm. Crack widths in each group (left to right)
0,0.05,0.1,0.2, and 0.5 mm.
75
~
r--
~
~
~
en
c
v
~
~
~
en
c
. ~
~
~
v
~
co
76
1 0 0 0 r - - - - - ~ - - - - ~ - - - - - - - - - - - - r _ - - - - - - - - - - - - - - - - - - - - - - - - ~
700
600
500
200
100
. Crack
width, mm.
o ~ ____ ~ ______ ~ __________________________________________ ~
o 4
Exposure, months
Fig. 19. Wire strength vs Duration of exposure in tidal seawater
for various crack widths. Mortar cover 0.5 mm.
8
.0
,....
..
.r:
+"
0'1
C
f
+"
V)
0'1
C
.'-
~
ftS
Q)
s-
ec
77
1000
Crack
width, nm.
0.1
900
800
700
600
500
200
100
a L-____ ~ ____ ~ ~ __________ ~ ~ ________________________ ~
a 1 2
4
Exposure, Months
Fig. 20. Wire strength vs Duration of exposure in tidal seawater
for vari ous crack wi dths. ~ 1 o r t a r cover 2.0 mm.
8
1000
900
800
700
.0
.....
..
..c
....,
600
0'1
C
f
....,
V')
0'1
C
.,....
.:.l
ItS
500
cu
'-
c:o
200
100
o
78
o 1 2 4
Exposure, Months
Crack
width, mm.
o
0.1
0.05
Fig. 21. Wire strength vs Duration of exposure in tidal seawater
for various crack widths. Mortar cover 3.5 mrn.
8
79
The "typical" ferrocement construction chosen contained 0.225-
in. dia. double-drawn high-tensile A82 rods spaced on two-inch centres. Panels
were made with two layers of 1/2-in. 19-9a. galvanized hardware cloth on each
side of the rods and with three layers on each side. i.e.:
2 layers 1/2-in. 10-ga. galvanized hardware cloth
0.225 in. high-tensile A82 rods at 2-in. transverse
0.225 in. high-tensile A82 rods at 2-in. longitudinal
2 layers 1/2-in. 19-9a. galvanized hardware cloth
3 layers 1/2-in. 19-9a. galvanized hardware cloth
0.225 in. high-tensile A82 rods at 2-in. transverse
0.225 in. high-tensile A82 rods at 2-in. longitudinal
3 layers 1/2-in. 19-9a. galvanized hardware cloth
The mortar used for all the panels was Type II cement 1 part.
dry mortar sand 2 parts, water/cement ratio 0.40 to 0.45, chromium trioxide
300 ppm water. slump> 3 in. Compression cubes and flexure and compression
prisms were cast, cured, and tested at 7 and 28 jays.
were obtained:
Mortar
The following properties of the mortar and
Compression Strength (C109. C349)
2-in. cubes 7-day 5400 psi
28-day 7700 psi
1.575 in. prisms 28-day 10,000 psi
Flexural Strength (C348)
1.575 in. prisms 28-day 1350 psi
Reinforcement
0.225 in. double-drawn high-tensile A82 rods
Breaking strength 3,360 lb
U1t.tensile strength 86,000 psi
Yield strength, 0.2% 72,000 psi
E10ng. % in 10 in. 6.5
1/2-in. 19-9a. galvanized
Diameter, stripped
Breaking strength
U1t. tensile strength
Yield strength, approx.
Specific surface
(mesh layers only)
hardware cloth
0.038 in.
63 lb
55,000 psi
44,000 psi
7.1 in-
1
(2.80 cm.-
1
)
80
(a) Strength and Elastic Modulus in Tension
Tensile tests of ferrocement specimens reinforced with mesh
only present few problems. The specimens can be shaped, if necessary, and
are easily gripped in the testing apparatus. Ferrocement panels or specimens
containing rod reinforcement as well as mesh are difficult to test in tension.
Specimens in which the rods are not cross-welded or otherwise anchored cannot
develop the full strength. A considerable amount of testing failed to develop
a test specimen and gripping system which adequately assessed the tensile
strength of the material composite - mortar, mesh, and rods. One pair of
specimens (207-3 with 4 layers of 1/2-in. 19-9a. hardware cloth and 208-3
with 6 layers) was prepared with shoulders and a reduced central portion
(2 x 1 x 6-in. gauge 1ength) . by drilling and diamond sawing. Each specimen
had one lengthwise rod and three rods in the test section. Each
specimen was fitted with a standard 2-in. linear variable . differential trans-
former extensometer modi fed to 6-in. The stress-strain curves shown in
Fig. 22(a) and (b) were obtained. The loads at first visible crack and the
maximum loads were recorded. The single rod in the centre of the specimen
began to slip as the rod/mortar bond failed. The strengths so obtained are
lower than would be expected from very long specimens with very long rod/
mortar bond lengths or from specimens with rod cross-welds or rods somehow
anchored into the specimen. The following test results show how the strengths
developed are not much higher than the combined strength of the wires in the
mesh.
Specimen 207-3
(4 layers of 1/2-in. 19-9a. hardware cloth,
2 sq. in. of ferrocement cross-section area)
Load at first visible crack
Maximum load held
Strength of 16 wires in mesh
Rod bond (by difference)
Specimen 208-3
(6 layers of 1/2-in. 19-9a. hardware cloth, 1
2 sq. of ferrocement cross-section area)
Load at first visible crack
Maximum load held
Strength of 23 wires in mesh
Rod bond (by difference)
1 central lengthwise rod,
1300 1b
1420 1b
1010 1b
410 1b
central lengthwise rod,
1450 1b
1900 lb
1450 1b
450 1b
The values of tangent modulus of elasticity (tangent 0 to 1000
lb load) obtained from the load/strain curves for 207-3 and 208-3 are 1.2 x
10
6
and 0.97 x 10
6
psi respectively.
CI)
CI)
OJ q-.,..
S-XCI)
~ Q.
V)
CI)
CI)
OJ q- .,..
S-XCI)
~ Co.
V)
M
I
co
0
N
"0
c:
IU
M
I
r-.
0
N
CI)
c:
OJ
E
.,..
u
OJ
Q.
CI)
OJ
,...
.,...
CI)
c:
OJ
~
S-
O
I+-
CI)
OJ
>
S-
::s
u
c:
.,...
IU
S-
~
CI)
"-
CI)
CI)
OJ
S-
~
til
.a
IU
N
N
C'I
.,...
...
82
Since the potential contribution of the rod to the strength
of the ferrocement specimens has not been adequately assessed it is considered
that the above values of unit strength and elastic modulus are not a proper
measure of the real strength of the ferrocement composite. as used in a boat
hull.
(b) Strength and Elastic Modulus in Flexure
Flexure tests, third-point loading on 18-in span, were performed
on 6-in. wide test specimens from panel 207 (4 layers of 1/2-in 19-9a. , galvan-
ized hardware cloth and rods as previously described) and from panel 208
(6 layers). Pairs of 18-, 24-, and 36-in. long specimens were tested to assess
the effect of specimen length, especially rod/mortar bond 1ength,on the bending
behaviour. An extensometer modified to 6-in.gauge length was used to measure
the tensile strain in the centre-third of the specimen. The beam deflections
at midspan and third-points were recorded for increments of loads. The location
of the mesh wires, both lengthwise and transverse, of each layer of mesh and
rods in each specimen was carefully measured with respect to the tensile surface
of the specimen for possible future use in the further development of a math-
ematical model. These readings are not presented.
The values of load, strain, and deflection, shown in abridged
form in Tables 34 and 35, were used to calculate:
- flexural strength at the load at first visible crack, Pfvc
- flexural strength at load Pfvc t 2
- flexural strength at maximum load held, {called herein Modulus of Rupture
- effective modulus of elasticity in bending at the load at first visible
crack, Pfvc, and at 400 lb Pfvc t 2) based on beam curvature
formula E = (where R = radius of M = moment, I = section
I moment of inertia)
- effective modulus of elasticity in bending, E = at the load at first
visible crack, Pfvc, and at 400 lb Pfvc t 2) based on fibre
stress formula F =
These values are shown in Table 36.
83
TABLE 34. Flexure Results - load Midspan Deflection D2. Difference in
Midspan and Third-Point and Fibre Strain in 18-, 24-,
and 36-1nch specimens from Panel 207 on la-in. span. (Abridged)
Midspan
Dl;o3
' - ,
Load
Deflection, D2 -D2
Fibre Strain.
Rellllrks
1b
O.ODl in.
1n./1n.
0.001 in.
Specimen 207-1 (18 in. long)
0 0 0 0
400 37 3.5 0.00053
800 98 12.5 0.00170 First visible crack
1230
213' 25.0 Max1mum'load held,
Specimen 207-4 (18 in. long)
0 0 0 0
400 45 5.5 0.00056
800 100 14.5 0.00148 First visible crack
1100 156 23.5 Maximum load held
Specimen 207-5 (24 in. long)
0 0 0 0
400 38 4.5 0.00064
800 115 17.5 0.00217 First visible crack
1000 172 27.5 Maximum load held
Specimen 207-6 (24 in. long)
0 0 0 0
400 40 1.5 0.00045
1100 166 20 0.00252 First visible crack
1400 255 33.5 Maximum load held
SpeCimen 207-7 (36 in. long)
0 0 0 0
400 47 1.5 0.00068
900 143 13.5 0.00235 First visible crack
1100 206
. Maximum load held
Specimen 207-8 (36-1n. long)
0 0 0 0
400 56 6.5 0.00088
800 110 15.0 0.00190 First visible crack
1500 280 47.5 Maximum load held
84
TABLE 35. Flexure Results - load. Midspan Deflection 02. Difference in Midspan
and Third-point Deflections. and Fibre Strain in 18-. 24-. and 36-in.
Specimens from Panel 208 on 18-in. span. (Abridged)
Load
Midspan 0,+03
Fibre Strain.
Relllarks
lb
Deflection. 02 --- 02
1n./in.
0.001 in.
0.001 in.
Specimen 208-1 (18-in. long)
0 0 0 0
400 42 2.5 0.00066
1000 138 12.5 0.00276 First visible crack
1200 186 17.0 Maximum load held
Specimen 208-4 (18-in. long)
0 0 0 0
400 44 2.5 0.00066
1100 143 14.5 0.00266 First visible crack
1450 230 23 Maximum load held
Specimen 208-5 (24-in. long)
0 0 0 0
400 55 4 0.00090
900 140 13 0.00282 First visible crack
1700 395 42 Maximum load held
Specimen 208-6 (24-in. long)
0 0 0 0
400 41 1
-
1000 128 9 0.00060 First visible crack
1600 318 30 0.00226 Maximum load held
Specimen 208-7 (36-in. long)
0 0 0 0
400 43 0.5 0.00064
800 103 7.0 0.00152 First visible crack
1400 283 30.0 Maximum load held
Specimen 208-8 (36-in. long)
0 0 0 0
400 55 7.5 0.00080
1000 156 23.0 0.00280 First visible crack
1500 387 44 Maximum load held
Pinel
Specimen
No.
207-1
-4
-5
-6
-7
-8
Av.
208-1
-4
-5
-6
-7
-8
Av.
Panel
No.
210
205
85
TABlE 36. Flexure Strength and Effective Elastic Modulus at the Load .t
First Visible Crack (Pfvc), 400.1b Pfvct2), and Maximum Load
(Pmax) (Flexural Strength only at Pmax) for Specimens from Panels 207 & 208.
Specimen Flexure Strength, psi Effective Elastic Modulus in Bending
Length,
psi x 10
6
E1t. psi x 10
6
in.
at PPfvc at PPfvc at Pmax c'
--r
PPfvc P"400 lb Pfvc P .. 400 lb
18 2400 1200 3700 1.38 2.95 0.88 1.19
18 2400 1200 3300 1.76 2.08 1.18 1.57
24 2400 1200 3000 1.19 2.32 0.60 1.03
24 3300 1650 4200 * * 0.82 1.34
36 2700 1350 3300 * * 0.91 1.31
36 2400 1200 4500 1.52 1.96 1.14 1.20
2600 HOO 3666 1.46 2.32 0.92 1.27
18 3000 1500 3600 1.85 3.85 0.75 1.24
18 3300 1650 4350 1.75 3.70 0.79 1.21
24 2700 1350 1.57 1.35 0.58 0.94
24 3000 1500 4800 * * 0.74 1.35
36 2400 1200 4200 * * 0.92 1.42
36 3000 1500 4500 1.27 1.95 0.66 1.31
2900 1450 4425 1.61 2.71 0.74 1.25
* Specimens kinked, bending curve not uniform, radius R very large.
TABLE 37. Number of Cycles to First Visible Crack and Collapse
of Specimens under Unilateral Cyclic Flexure Loadings.
Load
Apparent No. of Cycles to
Kg
Fibre
First Visible Ultimate
Stress
psi
Crack Collapse
2 layers of 1/2-in. 19-9alvanized 380 2500 1 200
hardware cloth each side of 340 2250 5 3,000
0.225-in. double-drawn rods spaced 300 2000 12 10,000
at two inches in each direction 280 1850
- -
260 1720 50 140,000
240 1580
- -
200 1320
- -
3 layers of 1/2-in. 19-9alvanized 380 2500 6 20,000
hardware cloth each side of 340 2250 13 25,000
0.225-in. double-drawn rods spaced 300 2000 1,600 55,000
at two inches in each direction 280 1850 12,000 120,000
260 1720 60,000 240,000
240 1580 so ,000 260,000
200 1320 >500,000* >500,000*
*No visible crack after 500,000 cycles.
The average values for flexure strength, deflection, and
elastic modulus taken from Tables 34, 35, and 36 are as follows:
Panel
207 208
Layers of mesh on each
side of rods 2 3
Modulus of rupture at Pmax, psi 3700 4400
Flexure strength, psi
at load at first visible crack, Pfvc 2600 2900
at Pfvc f 2 1300 1450
Aver. midpoint deflection, in. ,
at load at first visible crack Pfvc 0.13 0.12
at P400
Elastic modulus, E = psi
0.04 0.04
at load at first visible crack Pfvc 1.46 x 10
6
1.61 X 10
6
at . Pfvc t 2, i.e. P '"' 400 1b 2.32 x 10
6
2.71 X 10
6
Elastic Modulus, E = psi
at load of first isib1e crack, Pfvc 0.92 x 10
6
1.27 x 10
6
at . Pfvc t 2, i.e. P = 400 lb 0.74 x 10
6
1.25 X 10
6
(c) Behaviour Under Repeated Flexure Stresses.
It is recognized that materials in general subjected to repeated
tensile stresses can fail by the development and incremental extension of a
crack at stress levels well below the stress required to crack the material
in a single application of. the load. On the basis of early work by C1emmer
33
Older
34
, and Hatt
35
on repeated loading of concrete cylinders in compression
and of concrete beams, both reinforced and unreinforced, in bending, it
generally agreed that repeated stresses of about fifty percent of the
static failure stress will cause failure by fatigue. Yoshimoto et a1.36
and Hi1lsdorf and Kes1er
37
, more recently, have examined the effect of repeated
loading on concrete. In 1974, Committee 215 of the American Concrete Institute
discussed considerations for design of reinforced concrete subjected to fatigue
10ading.
38
Although the report applies particularly to reinforced concrete,
the tests on slabs reinforced with welded wire fabric and on welded wire fabric
alone are' of considerable interest in that they may provide a more macroscopic
view of the location of cracking (at cross-weld joints or between them), the
disturbance due to the welded intersection, and deterioration of the bond
between the smooth wire and concrete which can be applied to ferrocement.
Romua1di
39
and Romua1di et al.
4o
have examined the fatigue
behaviour of concrete mortar reinforced with short pieces of thin steel wire.
Romualdi points out that the low tensile strength of concrete is due to the
87
presence of flaws and is not an inherent property of the material. The
presence of this reinforcement, randomly and uniformly closely-spaced
steel wires or wire mesh, provides a crack-arrest mechanism which increases
the cracking strength of the concrete or mortar. Romua1di
1
s brief review of
fatigue tests showed that the endurance limit of an unreinforced concrete
beam was about fifty-five percent of the static cracking strength, whereas
the endurance limit (without visible cracking) of the beams reinforced
with short steel wires was approximately equal to the static tensile strength.
Dinsenbacher and Brauer
41
, (Naval Ship Research and Develop-
ment Center, U.S. Department of the Navy), in their description of the
development and evaluation of ferrocement planing boats, have presented
data for NSRDC/A ferrocement construction comprising five layers of 1/2-in.
rectangular unga1vanized 19 gao s t e ~ l woven wire cloth. Their test spec-
imens were tapered to give a constant stress on a six-inch portion of
cantilevered specimens. The specimens were loaded on each cycle to constant
displacement, so it is probable that the load decreased with time. They
concluded that the endurance limit was 3,000 psi, about fifty percent of the
static flexural strength. Larsen
42
ran fatigue tests at maximum stresses equal
to about 0.5 to 2.4 times the average cracking stress of 1320 psi obtained
from static flexure tests to one million cycles.
The behaviour of ferrocement under flexure loads at both
constant deflection and constant magnitude was examined.
i. Stress-Strain Hysteresis Effect-Ten Cycles
Two 6-in. wide lB-in. long specimens from Panels 207, containing
2 layers of 1/2-in. 19-9a. galvanized hardware cloth on each side of a 0.225
in. high-tensile rods, and 20B, containing 3 layers each side, were tested
under repeated loadings in flexure. Mid-span deflections and the tensile
side strain were measured, the latter with a modified 6-in. extensometer.
The specimens were loaded to 500 1b, a load known from experience to be
lower than the load at first visible crack. The load was removed and reapplied
in a similar manner ten times. The repeated cycles of load vs strain on the
tensile side are shown in Fig. 23 (a) and (b). After the first cycle, plastic
strains of 0.0005 and 0.0003 in/in. were observed for specimens 207-2 and
20B-2 respectively. No further plastic strain was observed after ten cycles.
The effective moduli of elasticity obtained from a tangent
line at the 500 1b load after the first cycle and after ten cycles are as
follows:
Specimen 207-2
Specimen 20B-2
After First
Load Cycle
0.B6 x 10
6
psi
1.41 x 10
6
psi
After Ten
Load Cycle
1.BB x 10
6
psi
2.B2 x 10
6
psi
Load,
lb.
88
ISpec i men 207-2
900
Soecimen 208-2
~ r - 800
[350
~ r - 700
~ , 3 0 0 600
~ 500
40
3
E1cngi 'on 1 div.=0.000133 in./in.
12
1 2
a b
Fig. 23 a, b. Load/strain curves for flexure specimens 207-2
and 208-2 loaded 10 times.
89
ii. Cyclic Flexure - At Constant Deflection
An apparatus was constructed to load eight 4 x 12-in. ferro-
cement specimens in flexure simultaneously. The apparatus consi-sted of four
pairs of roller cam arms fixed to a rotating shaft. Each specimen was sub
jected to thi rd-poi nt 1 oadi ng by means of a cali brated stra'i n-gauged channe 1-
shaped load cell connected" to a quick-response Visicorder monitor. The
deflection of the specimen under load was measured by a dial gauge in a
saddle support. The number of load cycles were counted by a cam-activated
electric counter. The apparatus was simple, able to test up to four pairs of
specimens at four different loads but suffered the serious disadvantage of
being a constant-deflection rather than a constant-load device.
Four specimens were taken from Panel 51 containing 2 layers
of 1/2-in. 19-9a. hardware cloth on each side of longitudinal and transverse
1/4-in. hot rolled rod spaced at 2 in. The fibre stress in static bending
at first visible crack was 2620 psi and the modulus of rupture (maximum-load)
was 6310 psi. These were loaded to initial fibre stresses of 1930, 1990,2130,
andJ2600 psi, i.e. loads of 75 to 100 percent of the maximum f i ~ r e stress
deve10ped at the first visible crack in a static bend test. The ' specimens
were subjected to 200,000 cycles in unilateral bending. The char, cteristics
curves for load (apparent fibre stress) vs Number of cycles are shL''In in
Fig. 24. The apparent fibre stress of specimen 51-3,which developed fine
cracks durtng the initial loading, levelled off at about 400 psi. The stress
on the other specimens after 200,000 test cycles had dropped to 600 to 1000 psi.
Four 4 x 12 x 1-in. specimens were taken from a Panel 49
containing 2 layers of 1/2-in. 19-9a. galvanized hardware cloth on longitudinal
and transverse 0.225-in. high-tensile rods spaced at 2-in. The flexure
strength (fibre stress) at first visible crack was 2300 psi and the modulus
of rupture (at maximum load) was 4900 psi. The specimens were loaded to
initial fibre stresses of 1060,1340,2240, and 2550 psi, i.e. stresses from
less than 50 to 100 percent of the flexural strength at first visible crack, and
subjected to 300,000 load cycles of unilateral bending. Since the apparatus
1s a constant deflection device, the applied load diminished to a fairly
constant value as permanent set developed. The characteristic curves, Apparent
fibre stress vs Number of cycles, are shown in Fig. 25.
These initial tests suggested that ferrocement of the construction
described could withstand a fibre stress of about 1000 psi under constant
deflection for many thousands of cycles provided that the initial loading
1s low enough that 'cracking does not develop early in the test campaign.
.r-
en
Q.
..
en
en
QJ
So-
...,
V)
QJ
S0-
n
-
LL.
-
en
Q.
..
en
en
QJ
So-
...,
V)
QJ
S0-
n
.r-
LL.
90
3 0 0 0 r - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ~ - - - - - - - - - - - - - - - - - - ~
2000
Initial
Fibre
Stress, psi
1000
2600
2130
1990
1930
o ______ ..L--___ --'-____ -L-___ --L. ____ ...L.-____ ,
100,000 200,000
Number of Load Cycles
Fig. 24. Characteristic fibre stress curves for beams from Panel 51
under repeated bending.
3 0 0 0 r - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ~
2000
Initial
Fibre
Stress, psi
1000
49-7
1340
49-1
2550
49-2 1060
49-3 2240
O ~ - - - - - ~ - - - - ~ ~ - - - - ~ - - - - ~ - - - - - - - L - - - - - - ~
100,000 200,000 3.00,000
Number of Load Cycles
Fig. 25. Characteristic fibre stress curves for beams from Panel 49
under repeated bending
91
iii. Cyclic Flexure - At Constant Magnitude
Specimens, 4x12xl in. cut from Panel 210,
containing 2 layers of 1/2-in. galvanized hardware cloth above
and below 0.225 in. dia double-drawn high-tensile rods spaced at,2 in.,
and from Panel 205, containing 3 layers each side, had fibre strengths
at first visible crack of 1660 psi and 2650 psi, respectively, and
modulus of rupture values (maximum) of 3500 psi and 4250 psi, respectively.
Fig. 26 shows the load-deflection curves for the specimens. It is
interesting to note that the first visible crack in each specimen has
occurred at a mid point deflection of 0.030 in.
Specimens 4x12xl in. from the same panels were
subjected to repeated flexure in third-point loading in an Instron
(Model 1130) testing machine, equipped with a 500 kg (tensile-compression)
load cell and a meter-relay load cycling device. The repeated loads
are unilateral loads of constant magnitude. An electric counter
recorded the number of load cycles imposed. Fig 27 shows the apparatus
set-up.
Cyclic loads of 440 to 835 lb (200 to 380 kg)
resulting in apparent fibre stresses of 1320 to 2500 psi (93 to
176 kgf/cm2) were imposed on specimens from Panels 210 and 205. The
test cycling was run to ultimate collapse of the specimen (or to
500,000 cycles in the case of the lowest load). Complete collapse
quickly followed ... ,ire breakage. Fig 28 shows a collapsed specimen.
The number of cycles when the first visible crack
in the mortar on the tensile side of the specimens was seen and the
number of cycles when the layers of mesh on the tensile side failed
(resulting in gross bending of the rods and specimen collapse) are
recorded in Table 37. The number of cycles to first visible crack
to ultimate collapse were plotted on semi-logarithm scale in the
tradi ti ana 1 manner endurance 1 imit curves, Fi g. 29.
The plotted curves, subject to verification by
greater test replication, suggest that the ferrocement specimens from
Panel 205 with three layers of mesh on each side of rods can withstand
an "apparent fibre stress" of 1750 psi for about 200,000 cycles before
complete and ultimate collapse and of 1320 psi for at least 500,000
cycles without visible cracking.
The broken wires and the broken mortar matrix in
fractures of ferrocement specimens broken by both a single application
4200
3600
3000
Fibre
stress
2400
psi
1800
1200
600
o
.92
o 20
/
40
/'
/
/'
,/
~
,/
/
/
60
Fibre stress at load of
first visible crack, Pfvc
80 100
Midpoint Deflection 0.001 in.
Fig. 26. Static flexure tests for specimens from Panels 210 and 205.
120
'-
, .,'
93
Fig. 27. Cyclic flexure
loading at constant
magnitude.
I, "
Fig. 28. Flexure specimen which broke after 1800 cycles.
2000
Fibre
Stress
1500
psi
1000
500
----- ----- first visible crack
ultimate collapse
~ . no failure at 500,000 cycles
O ~ - - - - ~ ~ - L - - - - - - ~ ~ - - - - __ ~ __ ~ ______ ~ ~ ______ ~ __ ~ ____ ~ ~ ~
1 10 100 1000 10,000 100,000 1,000,000
Fig. 29. Fatigue curves for specimens from Panels 210 and 205.
of the bending load and by repeated applications of bending load
'were examined under the stereoscopic microscope a.t. magnifications
to 80 times. All the wires broken in the single-load bend test
95
showed necked fractures indicative of ductile tensile failures. Most
of the wires in the bending fatigue specimens showed brittle fractures,
i.e. fractures with no apparent necking. Because of the fineness
of the wire (0.038 in., 0.965 mm. dia), no markings characteristic
of progressive fractures were able to be observed in the fatigue
failures. Optical micrbscopy did reveal that most of the flat brittle
fractures lay in the plane of the mortar fracture, whereas the necked
fractures usually stood proud of the surface. Fig 30 and 31 show
these features.
No difference in texture between the surfaces of
the mortar fracture in specimens in static bending and those broken
under repeated bending stress was observed.
The fracture surfaces containing necked tensile
fractures and flat brittle fatigue fractures were examined under the
electron microscope at magnifications to 2,000 times. The
necked tensile and brittle fatigue fractures are shown in Fig 32 and 32b
at magnifications of 15 times. At higher magnifications, details of
the fracture surfaces are clearly evident. Fig 33a shows the dimpled
surface typical of steel which has broken in a ductile manner under
a tensile load. Fig 33b shows the striations typical of fatigue
fractures. These striations indicate the stepwise manner in which
a crack progresses or propagates across the structure. Examination
showed that the crack progressed from one side (the tensile side of
the ferrocement specimen) completely across the wire cross-section.
A preliminary examination under the scanning electron
,microscope of the mortar fractures from specimens broken under a
static single application of bending load and under repeated applic-
ations has not revealed any distinct difference in appearance between
the two fractures.
The fatigue tests reported herein are of a preliminary
nature and require greater replication and additional tests at 1m'ler
stress levels to establish the endurance limit for ferrocement of
the constructions used.
The results of the tests which imposed repeated
loads of constant show that ferrocement of the constructions
tested can withstand over 500,000 cycles at an apparent fibre stress of
96
Fig. 30. Portion of surface of specimen broken in "static" bending.
Note tensile nature of wire fractures. Mag. 3x.
Fig. 31. Portion of surface of specimen broken in cyclic bending.
Note flat brittle nature of wire fractures. Mag. 3x.
97
a b
Fig. 32. Scanning electron microscope photographs of broken wires of
(a) specimen broken in static single-cycle bending and (b)
specimen broken in cyclic (fatigue) bending. Mag. l5x.
a b
Fig. 33. Scanning electron microscope photographs of broken wires showing
(a) dimpled structure of wire fracture typical of tensile failure
(Fig. 32a above) and (b) striated structure of wire fracture
typical of fatigue failure (Fig. 32b above). Mag. 2000x.
98
1320 psi. The endurance limit at . lOG or 10
7
cycles has not yet been
established and it i$ not yet known whether exhibits a
true endurance limit.
The examination of the fractures under the scanninq
electron microscope showed that, in general, the fractures of \'Iires
in specimens that failed under repeated bending loads were in the plane
of the major mortar crack or fracture, whereas those wires in specimens
that failed under a single-cycle bending load stood proud ot the major
mortar fracture. The broken wires in these specimens have fractures
which are characteristically tensile and which show no evidence of bending.
The scanning electron microscope showed that the' broken wires from
specimens subjected to repeated of bending stress have the
striations uniquely characteristic of bending fatigue. The fatigue
developed on the "tensile" side of the wires and progressed
completely through the wire. The appearance suggests that the bond
between wires that fail in fatigue and the mortar matrix is still
intact (and probably with a bond stress distribution described by
Kar and the crack width is very small, and that the wire
under the be"riding stress behaves in a "stiff" manner. ' The appearance
of the "tensile" fracture in specimens broken in a single-cycle load
suggests that the 'bond between wire and matrix has failed, that the
mortar crack is relatively wide at failure, and that the stress in
the wire at the point of fracture is almost pure tension.
5. Bolting Tests
Various brackets, bumper strips, chain plates, and
marine fittings must be ,attached to the hull and decks of ferrocement
boats. The bolts in those fittings which load the hull or deck as a
diaphragm can generally have substantial steel washers or bearing plates
to distribute the load. However, attachments which exert in-plane loads,
especially near a free edge, may be needed in areas not especially reinforced.
A preliminary examination of the behaviour of ferrocement It/hen steel
plates or bars carrying planar loads are attached to ferrocement panel
specimens has been made. The relationship between the marginal pitch
(the distance of a bolt hole from the edge) and the bolt hole pitch in
ferrocement panels of several constructions without special reinforcement
has been considered.
The ferrocement panel specimens were prepared from one-inch
thick panels of the following constructions:
1 layer of 1/2-in. l6-ga . galvanized welded square mesh on each
99
side of 0.225-in. double-drawn high strength rod spaced at
2-in. in both directions,
2 layers of 1/2-in. 19-9a. galvanized hardware cloth on one
side and 3 layers on the other side of the same rods, and
5 layers of 1/2-in. 22-ga. galvanized hexagonal mesh on
each side of the rods. All specimens had approximately
the same weight of reinforcement per unit area.
1/2-in. bolts:
The test specimens had the hole locations for
-
Type A boH holes 2-in. from each side and end of 4 x 8-in.
specimens
-
Type B bolt holes 2-in. from each side and 4-in. from each
end of 4 x 8-in. specimens
-
Type C bolt holes 4-in. from each side and 2-in. from each
end of 8 x 8-in. specimens -
-
Type 0 bolt holes 4-in. from each side and from each end
of 8 x 12-in. specimens.
All holes were drilled in the centre of the 2-in. space
between the rods. Steel bars 5/16 in. thick x 3 in. wide were bolted to
the ferrocement specimens. A 4 x 4-in. piece of 3/4-in. plywood interfaced
one of the specimen. Tensile loads were applied to the ends of the
bars. Coplaner loads, or slightly offset loads, were therefore imposed on
the ferrocement specimens. The maximum loads held are presented in Table 38.
The average maximum loads held, 16 per in. width of specimen, are:
Mesh Type Types A,B Type C Type 0
2 layers 1/2-in. 16-ga. welded 840 525 750
square mesh
5 layers 1/2-in. 19-9a. hardware 760 450 710
cloth
10' layers 1/2-in. 22-ga. hexagonal 640 450 650
mesh
It is tentatively concluded that bolt loads of about
500 1b/in. width may be used in ferrocement of the constructions described
if the distance of the bolt hole from the edge if more than two inches
and the bolt hole pitch is about four inches. A reinforced edge and a
9reater distance between bolt hole and free edge would increase the
bolt load able to be carried, possibly allowing failure of the ferrocement
100
Panel Construction
1/2-16 ga welded
square mesh
(2 layers)
1/2-19 ga hardware
cloth
(5 layers)
1/2-22 ga hexagonal
lllesh
TABLE 38. Loads Held in Bolting Tests
(single lIZ-in. bolts in 4 in. wide (A,B)
and 8-1n. wide specimens)
Specimen Load at First Max. load
Mode of Fail ure
Type Visible Crack, lb. held, lb.
A (2400) 3300 transverse, vertical
B 2600 2800 transverse
B (2800) 3400 transverse
B 3000 3900 transverse
C 3600 4200 vertical, transverse
D 4100 6000 transverse, vertical
A (2200) 2700 transverse, vertical
A 2800 3300 transverse
A (2300) 2750 transverse
A 2100 2600 transverse
B 2800 3200 transverse
B
{2600J
2800 transverse
B 3000 3850 transverse
C 3200 3600 vertical, transverse
C 3300 3600 vertical, transverse
D (5100) 6000 transverse, vertical
D 5000 5400 transverse, vertical
A* 800* 1200* transverse
A

2600 transverse, vertical
B 1700 2500 transverse
B 1800 2600 transverse
C 3000 3600 transverse, vertical
D 4000 5200 transverse, vertical
* tensile loading transverse to direction of mesh.
101
in compressive bearing. If offset loads which impose bending stresses
are used or if the loads are cyclic, the maximum load able to be carried
will be reduced.
Some preliminary tests were performed on the placement
of powder-actuated 1/4-20 NC studs with 1/8-in. shanks to hold brackets.
Some spalling and cracking of the panels was observed. Withdrawal
loads ranged from only a few pounds to about 1000 pounds. The tests
were not successful but further work appears to be warranted.
6." Design Considerations
This study has not delved deeply into the design
theory for ferrocement. It is generally agreed that reinforced .-
concrete analysis does not quite apply to a composite containing mortar
and finely divided steel reinforcement. As part of this study, reported
1n 1-972, J. D. Smi th (4) developed a prel iminary simpl e "mathematical
model" to explain the behaviour of ferrocement. His development is
presented in the Appendix. .
In 1972, Bigg(44) prepared a comprehensive treatise
for the Industrial Development Branch, Canada Fisheries Service, which
discussed the analysis of ferrocement structures and the design criteria
for these structures with a view to establishing a rational basis for
design. In the study he reviewed the mechanical properties obtained from
m a ~ y investigators of ferrocement and found gross differences due to
various testing techniques and the "highly s"tatistical" nature of the
prope,rties of ferrocement. He cites many of the variables responsible
for this "statistical" variation. He used the work of Bezukladov(18)
as a basis for his analysis of a beam. Bigg has expressed hope that
in the future he will be able to extend the analysis to a stage where
a designer can establish the type and placement of rod and mesh
reinforcement for a given geometry and material.
7. Patching Ferrocement
Damaged ferrocement hulls must be repaired to restore
a watertight condition, the strength, and the appearance and to prevent
further deterioration of the steel reinforcement from corrosion. Damage
can vary in severity, e.g., surface scrapes, gouges, spalling, and
cracking of the mortar matrix and extensive damage to the mortar matrix
and steel reinforcement. The damaged area may be covered with oil,
slimes, or other contaminants.
102
Most authors of books on ferrocement boat construction(S
extoll the ease of repairability as "one of the most endearing aspects
of ferrocement hulls". Some ignore the problem. Others prescribe
"knocking out the loose mortar, brushing out the dust and small particles,
hammering the mesh into shape, and re-mortaring". Some prescribe
applying an epoxy bonding agent to the edges of the damaged area, re-mortaring
with a rich Portland cement mix or an epoxy suitably stiffened with either
dry cement fine sand. The American Concrete Institute r1anual . of
Concrete Practice, Part 2, 1967 (ACI specifies a procedure
for patching surface defects in concrete. A bonding grout of thick cream
consistency (1 part cement to 1 part -30 mesh sand) is brushed onto the
surface. The patch is then filled with a concrete of the same kind
being patched. Further details are ,Irovided. The American Concrete
Institute Publication SP-21 Epoxies with discusses uses,
application techniques, and characteristics of epoxy-resin compounds for
concrete. Simple cracks, after thorough drying, may be injected with
epoxy by various ingenious pressure techniques. Where hull damage is
severe the hull must be restored to its original shape by jacks or other
means. The mortar "rubble" must be cleaned out as thoroughly as possible
by apneumatic needle chipping tool, sandblasting wire brushing, or other
means. Experience in this laboratory has that it will not be
possible to remove all broken pieces from the mesh. Damaged rods and
mesh must be repaired by inserting pieces of rod, as necessary, and
overlapping mesh.
A series of test repairs was made on specimens damaged
by prior impact and flexure tests. The 30 x 3D-in. panel impact
specimens, reinforced with 1/2-in. 22-ga. hexagonal mesh only, had a
highly-fractured centre area and major cracks extending to the surface.
No mesh had been broken. A pneumatic needle gun was found to be the
most effective way to clean out fragmented mortar, to open up the cracks,
and to roughen the surface. This device is an airpowered chipping and
scaling tool which uses a small bundle of steel rods 2 to 3 mm in diameter.
The rods easily penetrated the mesh with little damage to the wire.
Cracks were opened up to expose mesh which could assist in anchoring the
patch material. Another impact specimen had suffered mesh damage in
the test. It was similarly cleaned out, the mesh was not repaired. Both
specimens were wetted and patched with a Portland cement mortar of the kind
used in the original panels. The patches were cured 21 days. Several
3 x 12-in. flexure test specimens were patched with a cement/sand mortar
and several with a two-component epoxy marine patching compound,
containing a filler, and with a blO-component epoxy floor-patching
compound requiring a sand filler. The specimens received a minimum of
preparation. The patching materials were spread on the specimens
and forced into the cracks before these specimens were straightened.
When the specimens were straightened the excess epoxy squeezed out was
removed. The specimens were cured 21 days for the mortar patches,
7 days for the epoxy patches.
Flexure specimens reinforced with 1/2-in. 16-ga.
welded square mesh and patched with cement mortar regained about 80
percent of their original modulus of rupture values. Those reinforced
with 1/2-in. 22-ga. hexagonal mesh regained 50 to 70 percent of
their original strength owing to some prior damage to the wire mesh.
The several flexure specimens containing 1/2-in.
16-ga. welded square mesh, 1/2-in. E-ga. hardware 'cloth, and 3/8-in.
20-ga. welded square mesh were patched with epoxy patching components.
All well-penetrated epoxy-patched specimens regained virtually all
their original strengths.
It is concluded that cement/sand mortar can make a
reasonably strong repair but an epoxy filler at the edges might
103
prevent spalling at the fractured edges. The epoxy materials m : ~ c
strong repairs, virtually equal to the original strength, provided the
debris was removed. The effects of contamination with oils, weathering,
thermal expansion, and other factors have not been ascertained.
104
8. Protective Coatings
(a) General
Most concrete structures are not painted and there is little
recorded experience available to support the claim that paint can prevent
corrosion of steel in concrete structures. An investigation of 40
circular wire-wound prestressed concrete water reservoirs showed that
corrosion of the steel occurred in all reservoirs except two which had
been painted \t/ith a polyvinylacetate latex coating over a flood coat of
raw linseed oil It was concluded that sealing the exterior wall
surface reduced and equalized the oxygen supply to the wires. The only
substantial reference related solely for concrete is the
comprehensive guide prepared by the American Concrete Institute in 1966
48
The report identified 20 basic film-forming materials available for
coatings manufacture. However, minor variations in formulations can
affect the performance of the coatings. The report pointed out that
surface preparation, method of application, the environmental conditions
at the time of application, and the film thickness, are of prime importance.
Two more recent books 49, 50 provi de a broad survey of the many conventi.ona 1
and more recently developed organic coating systems for most materials,
concrete.
The several authors of books on the construction of
ferrocement boats do not stress the need to paint ferrocement hulls.
Whitener 13 recommends the application of finishes on hulls in sea-
water and claims that the epoxy resin finishes are best and that anti-fouling
paint should be applied twice a year. Jackson and Sutherland 10 state
that professional New Zealand practice has been to coat the outside of
the hull with an epoxy resin finish but claim that a guaranteed 0.05 in.
of high quality mortar over the mesh eliminates the need for painting.
Cairncross 14 feels that painting is not an absolute necessity but
1s an added safeguard and more pleasing aesthetically. He claims that
epoxy paints are the most successful coatings, bonding well and giving
good protection. Primers, he states, can be coal tar-epoxy, clear epoxy
resin, or chlorinated rubber paint. He states that polyurethanes, vinyls,
and acrylics have also been successful. Some of the factors cited as
affecting the efficacy of possible coatings is the strength (e.9. epoxy
coatings may be ten times as strong in tension as cement mortar), flexibility,
elongation, coefficient of thermal expansion (e.g. epoxies may be much higher
than cement mortar); antifouling characteristics, resistance to seawater,
and changes of these with time.
105
In general, oil base paints must be avoided because the
alkali in the mortar saponifies the oil-base unless the concrete (mortar)
is very thoroughly dried over a long period so that the surface and inner
alkalinity is neutralized. The limited resistance of oil-based paints
and their permeability to moisture is said to limit their use in appli-
cations where moisture behind the film can cause flaking and blistering.
Latex (water emulsion) paints of the styrene-butadiene, polyvinyl acetate,
and acrylic types are said to be relatively insensitive to water and to
resist alkalinity well. The nature of water-based paints allows appli-
cation over damp (not wet) and unaged (not uncured) mortar. Some synthetic
resin coatings e.g. epoxy resins, polyesters, polyurethanes, and various
synthetic rubbers have been used successfully on c o n ~ r e t e intended for
severe service.
(b) Exposure Tests
The paint coatings evaluated in this study, on the basis
of exposure to various environments, included an inorganic two-component
self-curing ethyl silicate zinc-rich coating, a polyvinyl chloride-based
enamel, a two-component chemical-curing polyamide resin, a chlorinated
rubber-based enamel, and a highly-modified polyester resin coat in various
primer and top-coat combinations.
The mortar surfaces of the 15 x 30-in. panels were
prepared as follows:
- ground with flexible abrasive disc CF885F A10xite 50
Fascut resin disc) to smooth the surface and remove any
cement film.
- etched by scrubbing with 200 cc of 30-percent (or 15-percent
where particularly specified by paint manufacturer) muriatic
acid until all bubbling ceased and re-etched for 30 seconds.
Panels 13 and 14 were not etched.
- washed with much water and thoroughly dried at 100F overnight.
- dusted surface before brush application of paint according to
manufacturers direction.
- paint coatings cured 14 days at normal room temperature
before exposure.
The 12 paint systems used are described in Table 39.
106
TABLE 39. P.fnt Systems Applfed to Ferrocement Panels for Exposure Tests.
Pinel
Descrfption of P.int System
Toul
Mo. Thfckness.
.n,.
3A Primer-inorganic two-component self-curing ethyl silicate zinc-rfch.
,
Top-coat - same as primer coat.
6 Primer seal coat - Two-component clear epoxy finish. 4
Top coats - Two coats of vinyl resin - base ant i foulin9 paint.
7 Primer seal coat - chlorinated rubber-based paint cut with 15 percent thinner. 4
Top coats - Two coats chlorinated rubber-based paint.
9A Primer seal coat - Zinc silico-fluoride solution. 4
Top coat - A highly ~ d i f i e d polyester resin vehicle with metal
and metal oxide fillers.
9B Primer seal coat - none. 4
Top coat as in 9A.
llA Primer seal coat - Polyvinylchloride-based enamel cut with
Top coats - Two coats polyvinylchloride-based enamel.
15 percent thinner. S
11B Primer seal coat - none. 4
Top coats - Two coats two-component pigmented epoxy-resin.
12A Primer seal coat - Inorganic two-component self-curing ethyl silicate zinc-rich. 6
Top coat - Two-component clear chemical-cured polyamiip. resin.
12B Primer seal coat - Inorganic two-component self-curing ethyl silicate zinc-rich. 6
Top coat - Polyvinyl chloride-based enamel cut with 30 percent vinyl thinner.
14 Same as 7 except that mortar surface not etched. 4
16A Primer seal coat - Two-component clear chemical-curing polyamide resin. S
Top coat - Two-component clear chemical-curing polyamide resin .
16B Primer seal coat - Chlorinated rubber-based enamel cut with 1'5 percent thinner. 4
Top coats - Two coats chlorinated rubber-based enamel.
107
Specimens, 3 x 3 and 3 x,6-in. cut from the panels
for the several exposure procedures, viz. ambient dry laboratory conditions
(control), Weather-Ometer, laboratory wet-dry cycling in a seawater bath,
and marine tidal. The specimens were assessed on the basis of visual
appearance, {crazing, flaking, etc.}, ease of separating the paint
coatings from the mortar substrate with a razor blade, spall-resistance
when penetrated with two parallel score marks of a sharp pen-knife, and
resistance to gouging-abrasion by a pointed tool.
i. Dry Laboratory Environment {Control}
The set of 3 x 3in. specimens were assessed as above
after. four months in the ambient laboratory environment. The assessment is
presented in Table 40.
ii. Weather-Ometer Exposure
Duplicate 3 x 6-in. specimens were exposed in a standard
Weather-Ometer apparatus. Each cycle of exposure consists of 102 minutes
of ultraviolet light plus 18 minutes of ultraviolet light and a water spray.
The temperature is maintained at 140F. An exposure period of 250 hours is
reported to be equivalent to one year of outside mid-temperature climate.
The appearance of the specimens was recorded at regular intervals. The
specimens were removed after 1500 hours {6 years equivalent} and rated on
the above basis. The assessment is presented in Table 41.
iii. Laboratory Seawater Cyc1 ing
Duplicate 3 x 3-in. specimens of 12 paint systems were
inserted into a cyclic exposure apparatus which dipped the specimens in
a bath of filtered seawater for one hour and lifted the specimens for.
drying before an air fan for three hours. The specimens were observed
at regular intervals and assessed, as above, after 600 cycles. An
additional 2000 cycles of exposure produced no visible change. The
results are shown in Table 42.
iv. Marine Tidal Exposures
Two duplicate sets 3 x 3-in. specimens of the 12 paint
systems were set into shallow pans of paraffin wax and covered with a
.screen cage. One set was exposed mid-February to mid-r1ay at mean tide level
at the Vancouver Kitsi1ano Station of the Canadian Coast Guard and the other
at below low-low tide. The former received one or two immersion cycles
per day, depending on the tides. The latter was constantly immersed.
The coatings were assessed, as above, after 84 days exposure. The condition
108
Specimen
No.
3
,
7
9A
98
llA
118
12A
128
14
16A
168
Specimen
No.
3
,
-7
9A
98
11A
llB
12A
12B
14
16A
16B
TABLE 40. Assessment of Coatings After 4 months in Room Environment
(control specimens not exposed to weather or service environment)
Visual Appearance
Separation Resistance
(Lifting with razor blade)
Spall Resistance
(Scoring with penknife)
Gouge Resistance
(Pointed tool)
Fine crazing Separates eas ily.
brittle flakes.
Spalls and crumbles. Fair
No defects No separation at interfaces. Spalls Fair
No separation at interfaces. No spalling. Good
No separation at interfaces. No spalling. Good
No separation at interface. No spal1tng. Debris tenacious. Good
Separates with difficl.'lty. No spalling. Debris tenacious. Good

Separates with difficulty. No spal1tng. Debris tenacious. Good
Separates fairly eaSily, Spalling. Debris tenacious. Good
brittle.
Top coat separates from Spalls. Debris tenacious. Fair
primer: Tough.
Spalls slightly. Fair Separates with difficulty.

H
Separates with difficulty. No spal1tng. Debris tenacious Good
Top coat separates from Spall ing. Debris brittle. Poor
primer. Brittle.
TA8LE 41. Assessment of Coatings After 1500 Hr (6 yr) in Weather-Ometer
-- -
Visual Appearance
Separation Resistance
(Lifting with razor blade)
Spall Resistance
(Scoring with penknife)
~ u g e Resistance
(Pointed tool)
Fine craze cracking No separation at interface. No spalling. Debris crumbly. Good
Yellow encrustation
Fine craze cracking Separates cleanly from inter- Spalls cleanly in flakes. Fair
Yellow stain. face but with difficulty.
Fine craze cracking Separates easily and cleanly Spalls easily and cleanly. Poor
and flaking. at interface.
Fine craze cracking No separation at interface Spalls slightly. Debris Fair
Clean surface. flaking and crumbly.
Fine craze cracking No separation at interface Spalls slightly. Debris Fair
Clean surface. flaking and crumbly.
Good surface. Separates readily at inter- No spal1tng. Tough fl1m. Good
face in continuous film.
Good surface. Separates only with diffi-
culty.
No spalling'. Tough film. Good
Fine craze cracking Separates readily at inter- Spalls easily and cleanly. Fair
face with spalling.
A few craze cracks No separation at interface Spalls slightly. Debris crumbl) Fair
Fine craze cracking Separates cleanly and Spalls easily and cleanly. Poor
and flaking easily at interface.
A few fine craze Separates easily and cleanly
cracks. at interface.
No spalling. Tough film. Good
Craze cracking. Separates cleanly with
curling, film loss. difficulty at interface.
Spalls easily and cleanly. Poor
Specimen
No.
3
6
7
9A
98
llA
118
12A
128
14
16A
16B
TABlE 42. Assessment of Coatings After 2600 cycles in Seawater.
(1 hr immersed - 3 hr dr,ying in moving air)
Visual Appearance
Heavy brown bloom.
No surface defects.
No surface defects.
Yellow stain. *






Separation Resistance
(Lifting with razor blade)
No separation at interfaces.
No separation at interfaces.
No separation at interfaces.
No separation at interfaces.
No separation at fnterface.
Easy separation at mortar
interface.
Separates with difficulty.
Separates fairly easily.
Brittle.
Coating lifted at Top coat separates from
one corner primer as tough film.
No surface defects. Separates wfth difficulty.
No surface defects. Separates with difficulty.
No surface defects. Top coat separates from
primer.
Spall Resistance
(Scoring with penknife)
No spalling. Debris crumbly
Spalls. Debris slightly
cruq,ly.
No spalling. Debris
slightly crumbly.
No spalling. Debris
tenacious.
No spalling. Debris
tenacious.
Spalling. Debris
tenacious.
No spalling. Debris
tenacious.
Spalling. Debris
No spalling. Debris
tenacious.
Spalls slightly. Debris
tenacious.
No spalling. Debris
tenacious.
Spalling. Debris brittle.
* All specimens exhibited yellow stair.ing.
109
Gouge Resistance
(Pointed tool)
Good
Fafr
Fair
Fair
Fafr
Fair
Fafr
Fafr
Poor
Poor
Fair
Poor-spa 11s.
110
of the specimens except for a slime coating appeared to have changed
little. The results are reported in Table 43.
I The two lots of specimens were returned to the sea from the
end of May and to the end of October. Examination after this additional
150 days of exposure showed that the barnacles had attached themselves to
all of the specimens except those two which had the inorganic two-component
self-curing ethyl silicate zinc-rich paint primer and top coat, (specimen 3)
and those having the two-component clear epoxy finish primer with
two top coats of vinyl resin-base antifouling paint (specimen 6).
" The specimens were scraped free of barnacles and reexamined.
Scraping easily removed the shells but their bare plates remained
attached to the substrate. Close inspection revealed that the plates had
removed the paint to bare mortar on those specimens coated with the
highly-modified polyester resin vehicle containing metal and metal oxide
fillers. The barnacles did not the other coatings but scraping
to remove the base plates of the barnacles, as would be necessary on a hull,
caused some damage. The results are found in Table 44.
(c) Discussion and Conclusions
On the basis of the visual and physical assessment of the
specimens subjected to the above exposure tests the most satisfactory
coatings are deemed to be:
System on Panel 3A
Primer - inorganic ethyl-silicate zinc-rich paint
Topcoat - inorganic ethyl-silicate zinc-rich paint
System on Panel 6
Primer - two-component clear epoxy finish
"Topcoat - two coats of vinyl resin-base anti-fouling paint
System on Panel 118
Primer - two-component pigmented epoxy resin
Topcoat - two-component pigmented epoxy resin.
System on Panel 11A
Primer - polyvinyl chloride-based enamel
Topcoat - two coats of po1yvinylch1oride-based enamel.
The other systems tested were considered less acceptable.
It is considered that barnacle-free, freshwater exposures may result in a
different rating order. Thermal expansion, aesthetics of colour, and other
factors have not been considered in the rating system.
Specimen
No.
3
,
7
9A
98
llA
118
12A
128
14
16A
168
111
TABLE 43. Assessment of Coatings After 84 days Marfne Exposure at Mean Tide
(Vancouver - Kitsilano)
Visral Appearance
Separation Resistance
(Lifting with razor blade)
Slime layer - otherwise No separation at interface.
clean as origfnal.









Spalls with dffficulty.
No separation at interface.


Separates as tough film.
Separates only with difficulty
Spalls at interface.
Top coat separates from
primer as tough film.
No separation
Spalls readily.

Spall Resistance
(Scoring with penknife'
Slight spalling. Debris crumbly.
Spalls. Debrfs slightly crumbly.
No spalling. Debris slightly
crumly.

Slight spalling. Debris slightl)
crumbl_y.
Slight separation. Debris
tenacious.

Spalls fairly readily
Spalls fairly readily. Debris
crumbly.
Slight separation. Debris
tenacious.
Spalls readily.

~ o u g e Resistance
(Painted tool)
Good
Fair
Good
Good
Good
Good
Good
Good
Fair
Fair
Good
Poor
112
TAILE 44. Conditions of Specimens After Second Period of Marine Tidal Exposure - Total 234 Days at low-low Tfde
Specimen Pafnt System Used Cond1Uon
No.
3 Prime and top coat - 1i\organic two-component self-curing ethyl
silicate zinc-rich paint.
No barnacles.
6 Primer - two-component clear epoxy finish. 110 barnacles.
Top coat - two coats of vinyl resin base antj-fouling paint.
7 Prfmer - thinned chlorfnated rubber-based paint. Bases of barnacles did not damage
~ o p coat - two coats of chlorinated rubber-based paint. paint but are dffficult to scrape
off without paint damage.
9A Primer - sealed with zinc s1l1co-fluoride solution. Barnacles completely dissolved paint
Top coat - a highly modified polyester resin/metal/oxfde. around base and partly under base.
98 Primer - no seal coat. As in 9A.
Top coat - as in 9A.
llA Primer - thinned polyvinyl chloride-based enamel. As in 7.
Top coat - two coats polyvinyl chloride-based enamel.
118 Primer - none. As in 7.
Top coat - two coats of two-component pigmented epoxy resin.
12A Primer - inorganiC two-component self-curing ethyl silicate Bases did not damage paint and could
zinc-rich primer. be scraped off with little damage to
Top coat - two-component clear chemical curing polyamide resin. paint.
128 Primer - as in 12A. As in 7.
Top coat - thinned polyvinyl chloride-based enamel.
14 Primer - thinned chlorinated rubber-based paint. As in 7.
Top coat - two coats of chlorinated rubber-based paint.
16A Primer - clear chemical curing polyamide resin. As in 12A.
Top coat - two-component clear chemical curing polyamide resin.
168 Primer - thinned chlorinated rubber-based enamel. As in 7.
Top coat - two coats of chlorinated rubber-based enamel. -
9. gyality Assurance
Many areas of construction must be controlled if the
ferrocement hull is to be of good quality. Various tests to verify
the strength of the mortar and reinforcement and that of the resul-
113
tant composite have already been outlined. Sands and cements of
recognized quality are readily obtainable. Controls, such as cement/
sand and water/cement ratios, slump, compressive, and flexure tests
for the mortar are strai ghtfor'lJard. Tests for the strength of the
steel reinforcements are also straightforward. It is important that
the details of mix, placement procedure, and properties of the mortar
be adequately specified. The kinds and lay-up configurations of the
steel reinforcements used are important. Specimens to determine the
strength and performance of the ferrocement composite must realistically
represent the properties of the hull.
It is very desirable that non-destructive tests be developed
td......assist in quality assurance. It has been eVident in the five-year
program that it is quite possible to w.ake substandard ferrocement even
when the panels are simple and made in the favourable environment of
the laboratory. Sectioning revealed unsoundness in panels of good outward
appearance. The construction of hulls with extra layers of mesh and
rod reinforcement in thick sections at the keel, ribs, bow, and stern poses
serious problems of mortar penetration. Our preliminary test using
ultrasonic techniques failed to detect unsoundness in a plane specimen.
It seems likely that tapping with a hammer may be the most effective
way to test plane sections at the present time. Heavy T-rib and keel
sections pose a different problem.
8igg 44 has outlined many of the important points
requiring attention in a quality control program. He has outlined problem
areas, and suggested parameters for judging its character.
The certifying bodies, viz. Marine Services Canada Ministry
of Transport and American Bureau of Shipping,nave not had to deal with many
requests for certification. Each body laid down certain general regulations,
requirements, and guidelines for the construction of ferrocement vessels
{generally available in mimeo only) ca 1968 when interest in ferrocement
construction was high. 51,52,53. For convemence these are presented in
the Appendix.
114
D. BIBLIOGRAPHY OF FERROCEI,1ENT LITERATIIRF
Much of the literature on ferrocement, or related to
ferrocement, has been collected and annotated in a continuing
bibliographic listing during the several program years. Many of the
items were of technical significance but some of the early items were
"news i terns II
The bibliography, presented in the Appendix of this
report, has been stripped of the "news " items but those of technical
and historical significance have been retained. Undoubtedly many
worthwhile contributions to the ferrvcement technology are not found
in the list. Some of these may be present in the annotated p a ~ e r s ,
many of which have a good bibliography.
Most of the items listed are on file at B.C. Research
but interested persons should, \'/here possible, try to obtain articles
from the original source. Notification and receipt of articles on
the subject are welcomed for inclusion into any future reprintings
of the report.
115
E. ACKNOWLEDGEMENTS
B.C. Research thanks the Fisheries and Marine Services.
Environment Canada. for the privilege of undertaking this work for it during
the past six years. It particularly appreciates the cooperation enjoyed
with Mr. L.S. Bradbury, Director of the Industrial Development Branch.
Mr. H.A. Shenker. Chief of the Vessels and Engineering Division of the
Branch. and Mr. G.M. Sylvester', Vessel Technologist and Project Officer
of the Division.
It is felt that the program, although it has not exhaustively
explored any of the properties of ferrocement, has provided a much better
"feel" for a wide range of properties of the material which will facilitate
design and construction of fishing and other vessels. However, much work
is required to bring the ferrocement to the same level of technology as
that of the other common construction materials.
~ ~
/'
A. W. Greenius
Division of Engineering Physics
Head, Physics
116
f. REFERENCES
1. Kelly, A.M. and W.N. English, Ferro-Cement as a Fishing
Vessel Construction Material, B.C. Research report to
Industrial Development Service, Canada Department of
Fisheries, 1968, published in Project Report No. 42,
Ferro-Cement for Canadian Fishing Vessels, W.G. Scott,
Edit., Industrial Development Branch, Fisheries
Service, Canada Department of the Environment, Ottawa,
Aug. 1971, pp. 11.
2. Kelly, A.M. and T.W. Mouat, as a Fishing
Vessel Construction Material, Proc. of a Conference on
Fishing Vessel Construction Materials, Montreal,
Canada, Oct. 1-3, 1968. (Also published in Project
Report No. 42, Ferrocement for Canadian Fishing
Vessels, W.G. Scott, Edit., Industrial Development
Branch, Canada Department of the Environment, Ottawa,
Aug. 1971, pp. 42 incl. Appendix and Reference List.)
3. Greenius, A.W., The Development of Ferro-Cement for
Fishing Vessel Construction, Project Report No. 42,
Ferro-Cement for Canadian Fishing Vessels, W.G. Scott,
Edit., Industrial Development Branch, Fisheries
Service, Canada Department of the Environment, Ottawa,
August 1971, pp. 61 and Technical Supplement, pp. 58.
4. Greenius, A.W., Ferro-Cement for Canadian Fishing Vessels,
Vol. 2, and Smith, J.D., Part II - Development of a
Mathematical Model, Project Report No. 48, Industrial
Development Branch, Fisheries Service, Department

5 . Greenius, A.W., Ferro-Cement for Canadian Fishing Vessels,
Vol. 3, Project Report No. 55, Industrial Development
Branch, Fisheries Service, Canada Department of the
Environment, Ottawa, August 1972, pp. 55.
6. Greenius, A.W., Ferro-Cement for Canadian Fishinq Vessels,
Vol. 4, Technical Report No. 64, Industrial Jeve10pment
Branch, Fisheries and Marine Service, Canada Department
of the Environment, Ottawa, 1973, pp. 57.
7. Greenius, A.W., Ferro-Cement for Canadian Fishing Vessels,
B.C. Research report to Industrial Development Branch,
Fisheries and Servi ce, Canada Department of the
Environment, Ottawa, Oct. 1974, pp. 50. (Not pub-
lished for general distribution.)
8. Hartley, R., Boatbuilding with Hartley, Boughb/ood
Printing House, Auckland, New Zealand, 1967, 33-50.
9. Samson, J. and G. Wellens, How to Build a Ferro-Cement
Boat, Samson Marine Design Enterprises Ltd.,
Vancouver, Canada, 1968, pp. 120.
10. Jackson, G.W. and W.M. Sutherland, Concrete Boatbuilding,
Allen and Unwin, London, 1969, pp. 106.
11. Benford, J.B. and H. Husen, Practical Ferro-Cement
Boatbuilding, International Marine Publishing Co.
Camden, 1970, pp. 176.
117
12. Roberts, W.H., Guide to Ferrocement Sail and Power Boats
and Design, Rapid .Blue Print Limited, Hamilton, Canada,
1970.
13. Whitener, J.L., Ferrocement Boat Construction, Cornell
Maritime Press Inc., Cambridge, t.1d., 1971, pp. 128.
14. Cairncross, C., Ferrocement Boat Construction, Inter-
national Marine Publishing Co., Camden, Me., 1972,
pp. 192.
15. Dinsenbacher, A.L., and F.E. Brauer, Material Development,
Design, Construction, and Evaluation of a Ferrocement
Planning Boat, Marine Technology, 11(3} July 1974,
277-296.
16. Building Research Station Digest 109, Zinc-coated
Reinforcement for Concrete, Building Research Station,
Garston, Watford, U.K., Sept. 1974, pp. 8.
17. Frazier, K.S., The Protection of Reinforcing Steel in
Concrete, Paper presented at Second Annual Offshore
Technology Conference, Houston, Texas, April 1970,
pp. 8.
118
18. Bezuk1adov, V.F., Ship Hulls Made of Reintorced Concrete,
Shipbuilding Publishing House, Leningrad, 1968,
Trans. Navships Trans. 1148, CFSTI AD680042, 1968,
pp. 187.
19. Christensen, K.A. and R.B. Williamson, the
Galvanic Cell Problem in Ferro-Cement, Report No.
UC SESM 71-14, University of California, Berkeley,
July 1971, pp. 58.
20. Cornet, I.,.R.B. Williamson, B. Bresler, S. Nagarajan, and
K.A. Christensen, Chromate Admixture to Improve
Performance of Galvanized Steel in'Concrete Sea
Structures, Proc. FIP Symposium Concrete Sea Structures,
Tbilisi, Sept. 1972, published by Federation Internationale
de la Precontrainte Terminal House, London, SWIOWAU,
April 1973, 159-163.
21. Robinson, Design of Reinforced Concrete Structures
for Corrosive Environments, Materials Protection and
Performance, Vol. 11, No.3, March 1972,15-19.
22. Hausmann, D.A., Steel Corrosion in Concrete, Material
Protection, Vol. 6, No. 11, November 1967, 19-23.
23. Sha 1 on, R. and t1. Raphael, Inf1 uence of Sea Water on
Corrosion of Reinforcements, Proc. ACI Journal, Vol. 55,
1959,
24. Hausmann, D.A., Electrochemical Behaviour of Steel in
Concrete, Proc. ACI Journal, V.ol. 1964, 171-188.
25. O. E"., Durabil i ty of Concrete Structures in the
Ocean Environment, Proc. of the FIP Symposium Concrete
Sea Structures, Tbi1isi, Sept. 1972, published by the
Federation Internationa1e de la Precontrainte, .
Terminal House, London, SWIWOAU, April 1973, 141-145.
26. Lewis, D.A.and W.J. Copenhagen, The. Corrosion of Reinforcing
Steel in Concrete in Marine Atmospheres,. the South
African Industrial Chemist, Oct. 1957, 207-219.
27. Lewis, D.A. and W.J. Copenhagen, Corrosion of Reinforcing
Steel in Concrete in Marine Atmospheres, Corrosion,
Vol. No.7, July 1959, 382t-388t.
119
28. Kowalski, T.G., Ferrocement in Hong Kong, Far East Builder,
July 1971, 29-35.
29. Walkus, R.; State of Cracking and Elongation of Ferrocement
Under Axial Tensile Load (I), Bul. Inst. Po1it. Din
lasi, XIV(XVIII), 3-4, 1968, 653-664.
30. Wa 1 kus, R., Sta te of Cracking and El onga ti on of Ferrocement
Under 'Axial Tension (II), Bul. Inst. Pol it. Din lasi,
XVI(XX), 3-4, 1970, 53-60.
31. Walkus, R. and T.G. Ferrocement: a survey,
Concrete, 5 (2), Feb. 1971,48-53.
32. Walkus, B.R. and A. Mackiewicz, Application of Ferrocement
Shells in Marine Structures, mimeo, published at
Symposium on Industrialized Spatial and Shell
Structures, Kielce, Poland, June 18-23, 1973,
385-398.
Clemmer, H.F., The Fatigue of Concrete, Proc. American
Society for Testing Vol. 22,1922,
408-419.
34. Older, G., Highway Research i.n Illinois, Trans. American
Society of Civil Engineers, Vol. LXXXVII, 1924,
1180-1224, esp. 1196.
35. Hatt, W.K., Researches in Concrete, Engi-neering Bulletin
of Purdue University, 24, Nov. 1925.
36. Yoshimoto, A., S. Ogino, and M. Kawakami, Micr0cracking
Effect on Flexural Strength of Concrete after Repeated
Loading, ACI Journal, Proc. Vol. 69, No.4, April 1972,
233-240.
37. Hilsdorf, H.K. and C.E. Kesler, Fatigue Strength of Concrete
under Varying Flexural Stresses, .ACI Journal, Proc.
Vol. 63, No. 10, Oct. 1966, 1059-1076.
38. ACI Committee 215, Considerations for Design of Concrete
Structures Subject to Fatigue Loading, ACI Journal,
Proc. Vol. 71, No.3, March 1974, 97-121.
120
39. Romualdi, J.P., The Static Cracking Stress and Fatigue
Strength of Concrete Reinforced with Short Pieces
of Thin Steel Wire, The Structure of Concrete (and
its behaviour under load), Proc. International Conference,
London, Sept. 1965, Edit., A.E. Brooks and K. Newman.
40. Romualdi, J.P., M. Ramey, and S.C. Sanday, Prevention and
Control of Cracking by Use of Short Random Fibers,
Paper No. 10, Causes, Mechanism, and Control of Cracking
in Concrete, American Concrete Institute Publication
SP-20, American Concrete Institute, Detroit, Michigan,
1968, 179-203.
41. Dinsenbacher, A.L. and F . E ~ Brauer, Material Development,
Design, Construction, and Evaluation of a Ferro-Cement
Planing Boat, Marine Technology, Vol. 11, No.3,
July 1974, 277-296.
42. Larsen, H.J., Jr. (John A. Blume & Associates, San
Francisco, Ca.), Study and Evaluation of Ferro-cement
for Use in Wind Tunnel Construction, prepared for
National Aeronautics and Space Administration, Ames
Research Center, Moffatt Field, California, NASA-CR-114501,
JABE-ARC-07, July 1972, pp. 155.
43. Kar, J.N. and A.K. Pal, Strength of Fiber-Reinforced
Concrete, J. Structural Division, Proc. American
Society of Civil Engineers, Vol. 98, ST5, May 1972,
1053-1068.
44. Bigg, G.W., An Introduction to Design for Ferrocement
Vessels, Project Report No. 52, Industrial Development
Branch, Fisheries Service, Canada Department of the
Environment, Jan. 1972, pp. 224.
45. Manual of Concrete Practice, Part 2, The American Concrete
Institute, 1967, (AeI 301-66), 32.
46. Epoxies with Concrete, Publication SP-21 , The American
Concrete Institute, 1968, pp. 140.
47. Westerback, A . E ~ and L.B. Hertzberg, Testing Corrosion of
Reservoirs, Materials Protection, 6 (6), June 1967,
58-62.
48. Guide for the Protection of Concrete Against Chemi.ca1
Attack by Means of Coatings and Other Corrosion-
Resistant Materials, ACI Committee 515, J. American
Concrete Institute, 63(12), Dec. 1966, 1305-1391.
49. Paints and Protective Coatings, Technical Manual
No. TM5-618, U.S. Dept. of the Army, Navy, and
Airforce, Washington, D.C., 1969.
50. Roberts, A.G., Organic Coatings - Properties, Selection,
and Use, Building Science Series 7, U.S. Dept. of
Commerce, National Bureau of Standards, Washington,
D. C ., 1968, pp. 187.
51. Bonn, W.E., Regulatory Aspects of Traditional and New
Construction Materials, (esp. Part 6 - Requirements
Applicable .to Vessels Constructed of Ferro-cement),
paper presented to Conference On Fishing Vessel
Construction Materials, Montreal, Canada Oct. l-3,
1968, Canadian Fisheries Reports No. 12, June 1969,
73-90.
52. Anon, Tentative Requirements for the Application of
Ferrocement to the Construction of Yachts and Small
Craft, lloyd ' s Register of Shipping, london, pp. 15,
and Tech. Note: FC/REQ/1 Tentative Requirements for
the Construction of Yachts and Small Craft to Ferro-
cement, Jan. 2, 1967, pp. 10.
121
53. Anon, Guidelines for the Construction of Ferro-Cement
Vessels, American Bureau of Shipping, ca. 1969, pp. 12.
122
G. APPENDIX
1. Development of Mathematical Model - J. O. Smith
2. Regulations for of Ferrocement Boats -
Canada Transport Ministry.
3. Guidelines for the Construction of Ferro-cement Vessels -
American Bureau of Shipping.
4. Bibliography of Ferrocement Literature.
Appendix 1. Development of a Mathematical Model
(by John D. Smith, P. Eng.*)
A. INTRODUCTION.
123
The available literature on ferro-cement boat building
mainly concentrates on construction techniques developed through
experience for specific types of reinforcement, with little information
on design methods to aid the naval architect. Architects with experience
in ferro-cement design tend to regard their methods as trade secrets.
The more technical articles on ferro-cement are often qualitative
comparisons of results of tests in which various strengths of mortar,
mesh sizes and types, percentages of steel, and reinforcing rods were
used. Also, most investigators emphasize the ultimate strengths
achieved. It is only with great difficulty that the naval architect
can find design information in a form useful to him.
The naval architect needs to know:
1. What working stress can be used for designing under tension,
compression, shear, and bending and what is the effect of
combined loading.
2. What deflection will be produced by a given load. At times
the most significant criteria will be deflection rather than
stress.
3. What type of failure will occur from accidents such as grounding
(possible high local loads), or hitting a rock or log (impact
loads). If the damage will seriously affect the water-tight
integrity of the hull, he may want to allow extra material to
compensate.
One approach to determining the required design informa-
tion is to make and test a sufficient number of panels covering the range
of likely reinforcement combinations under various load conditions.
Since test results for reinforced concrete typically show a wide spread,
the number of specimens of each type must be large enough to define perform-
ance adequately and to allow the range of deviations to be determined.
It should be kept in mind that a small defect in a small specimen will
have a greater effect than the same defect in a large panel. Therefore,
a greater number of small specimens may be required.
It is evident that this approach will be lengthy and
expensive. This method could lead to greater confidence in an acceptance
of ferro-cement as a boatbuilding material but it could also retard the
evolution of new and better reinforcing materials and combinations.
*Mr. John D. Smith was formerly associated with B.C. Research but is now
at Defence Research Establishment Pacific, Esquimalt, B.C.
124
An alternative approach is to develop a mathematical
model for the behaviour of ferro-cement. This model could then be
proven or improved by the results of a limited number of test specimens.
If successful, this approach could effect a considerable saving in time
and money and allow new combinations of reinforcement materials to be
cheeked out on paper. Only the most promising combination need then be
chosen for experimental verification.
Experimental methods would still be required to investi-
gate such factors as labour requirements, the effect of joints (hull to
deck, frames to hull, etc.), ease of fabrication, and behaviour under
impact loads.
The object is to develop the simplest model for the
behaviour of ferro-cement which will give results agreeing fairly well
with observed experimental results. Since the loading conditions to be
experienced by a vessel during its lifetime can only be guessed at and
since there can be large variations in actual strength due to workman-
ship, it is doubtful that a large increase in complication to achieve a
small increase in accuracy is justified. The model and its application
methods should also indicate the effects of changes due to increasing
mortar strength and in the type, quantity, and distribution of the
reinforcement.
The development of a mathematical model can be divided
into four convenient stages:
1. A linear model in which both the mortar and steel are assumed to
behave in a linear fashion. The linear model should result in
the 'least complicated solutions. Since working loads will be
considerably lower than ultimate loads, it is probable that
non-linear effects will not be significant and that a linear
model will be adequate in this region. The analysis of natural
frequencies and resonances will be much simpler if a suitable
linear model can be used.
2. The linear model is extended to account for t4e non-linearities
of the mortar. The steel is still assumed to behave linearly.
3. The loads and deformations where yielding of the steel or
limited compressive failure of the concrete occurs is investi-
gated. It is important to know when permanent deformation
occurs and what the properties will be after a partial failure .
With suitable assumptions it is probable that a linear model can
be used to describe the behaviour of the section under low loads
after some permanent deformation has occurred.
125
4. The analysis of ultimate failure and prediction of the mode
of failure would be investigated.
In the following sections, only the first two stages
will be investigated.
B.. DEVELOPHENT OF THE MODEL.
Muhlert(l) at the University of Michigan has applied
methods used in the design of reinforced concrete beams to an analysis
of ferro-cement bend test specimens. His results have been fairly
consistent but conservative. Mowat(2) at the University of Calgary
attempted to predict the performance of ferro-cement through a non-
linear analysis. Most of his calculations were to obtain ultimate loads,
and he used the ratio of actual moment at failure to the predicted
moment as a measure of the efficiency of the reinforcement. One criticism
of his results is that in calculating his ultimate loads he has assumed
that the mortar at the compression face has reached the maximum strain
before compression failure occurs. He also computes the height of the
neutral axis and the steel strains on this basis. Therefore, if the
steel fails in tension before this occurs the efficiency calculated will
be low. A better criteria would be the ratio between the percentage of
effective steel and the ultimate load.
Both Muhlert and Mowat used relatively simple computer
programs to perform the calculations using successive approximations.
However, in this study graphical methods will be used as much as possible
to make it easier to visualize the effects of changing mortar strengths
and reinforcement quantities.
Linearized model.
All loads and s ~ r e s s e s are calculated as a-function of
the total strain (the sum of the absolute values of the strains at
the tension and compression faces) since the ratio of thickness
to the total strain is equal to the radius of curvature and
proportional to the deflection.
(1) Muhlert, H.F., Analysis of Ferro-cement in Bending, The University
of Michigan, Paper No. 043, January 1970.
(2) Mowat, D.L., Flexural Testing of Ferro-cement Planks, Thesis for M.Sc.
in Civil Engineering, University of Calgary, Calgary, Canada,
January 1970.
126
It will be assumed that the mortar can exert only a
compressive force. Shah(3) has reported that once cracking has
occurred, the modulus of elasticity in tp.nsion of ferro-cement is
close to that determined from the steel alone. ' Since fine surface
cracks have been observed on new hulls (possibly due to settling
during curing or to shrinkage stress) it is reasonable to assume
the cracked condition.
y
e
t
a strain at tension face
e
t
e .. strain at compression
c
& a total strain
h = thickness
h
c .. height of neutral axis
1
!
above compression face
c
I y = height of any element
above compression face
At any point the strain is given by:
&
e = (y - c) h' taking elongation as positive.
Also define the following dimensionless parameters:
yaY...
h
Therefore
.
,
e = (Y - C)&
face
To achieve equilibrium the sum of the forces in the steel,
Fs, and in the mortar, F
m
, must equal zero for pure bending.
(3) Shah, S.P., Ferro Cement as a New Engineering Material, College
of Engineering Report No. 70-11, University of Illinois, Dec. 1970.
127
Linear steel.
For one layer of steel reinforcement:
I i
The force on the layer is given by
and the net steel force is
where Ai a area/inch width
Y
i
D height of layer
E a modulus of elasticity
s
n (Yi - c)
Fs a it h EAiEs b f 1
n anum er 0 ayers
&E
s F a __
s h
However,
Z a
Define
where z height of the centroid
of the steel
128
The moment for the steel is given by:
2
(Y
i
- c)
Ms - I h AiEs
x moment of the steel about the neutral
axis of the beam.
If 1
s
- moment of the steel about the centroid of the
steel
Linear mortar.
E
M - [1 + (z - C)2A ] s
s s s h
The mortar is assumed to exert a force only in compression
and the force is assumed to be linear with strain.
The stress at the compressive face S is given by:
c
s - - .. E
m h m
where E - modulus of elasticity in
m compression for the mortar.
129
The mortar force is then
F
c C
D _
(h tEm)
m 2
C
2
tE h
m
-...
2
and the moment for the mortar is
M
2
.. F (- - c)
m c 3
C
3
E th
2
M
m
D
m 3
For equilibrium the sum of the steel and mortar forces
must equal zero
or
F a [Z - C] tE A
s s s
C
2
tE h
F __ --:._m_
m 2
dividing both by (th) and equating
F F
Z
- C E A
C
2
E
....!.+...!!.=
m
0
--2- =
th th
h s s
Z
- C E A
C
2
E
m
=--
h s s 2
We can then solve for C by plotting both sides of the
equation as a function of C.
It should be noted that at this stage no account has
been made of the concrete displaced by the steel on the compression
side of the neutral axis. The effect of this will be investigated
with an example later.
130
The moment equations obtained were:
1
e)2 A ]
M
s
(Z -
- [-+
eE h
8
h
2 s s
and
e
3
E eh
2
M
m

m 3
Dividing both by
2
eh we get
1 A
[ -.!. + (Z - e)
2
-..!. ] E
h3 h s
and
As an example it would be convenient to use Mowat's
results for his Series 3 Plank 9. This plank contained 13 layers
of 1/2" x 1/2" welded mesh and was strain-gauged top and bottom.
The gauges remained intact for the full range of loads allowing us
to find both the total strain and the height of the neutral axis for
each bending moment.
Table 1 presents Mowat's data and the derived total strains
and heights for the neutral axis. These are also plotted in Fig. LA.
To calculate the height of the neutral axis we must have
a value of Modulus of Elasticity for the mortar.
Mowat used a stress strain relationship.
e - e
f .. f [1 _ ( cu e)A]
e eu e
eu
where: f .. mortar stress
e
f .. ultimate mortar stress
eu
e .. ultimate mortar strain
eu
e .. mortar strain
e
).
25
..
1.25 + f (psi)
e
d f
Taking E c ___ c_
m de
c
e - 0
c
Howat used a value of 0.38% for e For
cu
f - 5000 psi, we get
cu
E a 5.26 X 10
6
psi.
m
131
Since the steel is symmetrically arranged Z a 0.50 and
Z E A
: 8 T 0.539 x 10
6
C
2
E
Z - C m
The functions h EsAs and -2-- are plotted in the top
half of Fig. 2A. The curves intersect at C = 0.293.
The curves for
and each part of
H
s
- -
h
are also plotted. Taking the values from the curves
~ - (0.045 + 0.077 + 0.045) x 10
6
h
2
6
- 0.167 x 10
for a 0.2% and h = 1.0 in.
M - 0.167 (0.002)(1)2 x 10
6
- 334 in.-lb/in.
132
This corresponds to a load of 365 lb which is much lower
than the value of 530 lb taken from Fig. LA. However, if the load vs
strain for the linear model is drawn on Fig. LA it will be seen that
the calculated strain curve is offset from the experimental load/
strain curve but nearly parallel to it.
Before commenting further, the effect of the concrete
displaced by the steel on the compression side of the neutral axis
should be considered. This can be done by replacing the area
E - E
Ai by . . . . . ; . . s ~ E - - m - Ai for those layers.
s
For this case the corrected A = 0.033 sq in.
s
Z a 0.518
A
The corrected line for (Z - C) Es: is shown in Fig. 2A
by a dashed line.
moment is:
4
The corrected I = 0.00229 in. lin. and the corrected
s
M a 326 in.-lb/in. for = 0.2%
This is a change of 5.4%, which is small compared to the
extra work involved.
Non-linear mortar.
It will be assumed that the steel behaves linearly as
in the last section. However, the effect of a non-linear stress/
strain relationship for the mortar will be investigated.
The mortar force will be given by:
Fma-af C
cu
F
"'!!c:-af C
Eh cu
a .. a stress block factor defined as the ratio of
the average mortar stress to the ultimate
stress f
cu
This is
III
The stress strain relationship
e
fe
f
cu
- 1 - (1 - R. where R - will be used.
e
eu
plotted in Fig. 3A.

o
o R 1.0
For a given R
f
a - average value of between 0 and R.
cu
shaded area

R
1 - (1 - R. + 1
This gives a = 1 - (A + 1) R and is plotted in Fig. 4A.
Now
e
c
R a --- and e - C
e c
cu
F
Therefore = - A f C is now a function of as well as
h cu
C and a separate curve can be plotted for each value of . This is
done for = 0.01%, 0.2% and 0.4% in Fig. SA. It can be seen that
the height of the neutral axis increases as the total strain increases,
which is opposite to the trend shown in Fig. 1A for Mowat's specimen 3.9.
Improved linear mortar model.
Previously it was assumed that the mortar could only exert
a force in compression. If we now assume that the mortar has a
limited tensile strength and characterize this by an ultimate tensile
134
strain e
tu
The mortar force now becomes
E
m
--2
and the mortar moment becomes
M
m
eh
2
E
m
a-
3
e 2 E
(tu) ~
& 2
]
When & e
tu
the force and moment approach those for
the previous linear model. The height of the neutral axis now
decreases for increasing strain, approaching the value given by the
earlier linear model. This is demonstrated in Fig. 6A.
c. DISCUSSION ..
The linear model resulted in relatively simple expressions
for forces and moments. ~ e graphical method employed (Fig. 2A) in
solving for the height of the neutral axis and moments clearly shows the
effect of varying parameters such as steel area, mortar modulus of
elasticity and beam thickness.
Using Mowat's Series 3 Plank 9 as an example gave calculated
moments lower than the experimental results. However, the calculated
load/strain curve was roughly parallel to the experimental. It appeared
that the calculated values could be improved by choosing a higher value
for the modulus of elasticity for the mortar. This would also lower
the calculated neutral axis height, as well as increasing the calculated
moment.
The non-linear model indicated that the height of the
neutral axis above the compression face would increase with increasing
strain as long as the steel did not yield. This result was inconsistent
with the experimental result.
135
Compared to the linear model, the non-linear mortar
approximation gave a greater neutral axis height and would give a lower
bending moment, thus resulting in an even poorer agreement with the
experiment.
The improved linear model which accounted for some tensile
strength in the mortar correctly forecast the trend for decreased neutral
axis height with increasing strain and gave a higher mortar moment than
the previous models. It is felt that with a suitable adjustment of the
parameters Em and etu a good fit with the experimental result could be
attained.
No attempt was made to fit the experimental curves more
closely since this should be done using the results of many tests rather
than one isolated example, and the values chosen were adequate to illustrate
trends.
The correction of the linear model for the volume of
concrete displaced by the steel in the compression zone did not result in
a large change in neutral axis height or calculated moment.
D. CONCLUSIONS AND RECOMMENDATIONS.
Either the linear or the improved linear model promises
to be a satisfactory basis for a design procedure for ferro-cement in the
working stress range. Further development of the method will require the
elimination of appropriate constants by comparison with experimental
results.
Strain measurements on both tension and compression faces
are an essential requirement to determine the behaviour of the test
specimens. The effects of cracking at the tension face on the strain
gauge readings must be considered.
There exists enough experimental data and information on
standard practices to establish a standard mortar to be used for design
purposes. This would be the first step in establishing an approved
design procedure. Design charts and standards could then be developed
for this standard mortar. When advances in cement technology permit
mortars of higher properties to be produced a second
standard mortar could then be established.
Using the standard mortar for test specimens, efperimengal
results could be used to establish points on the curves for -- and
Eh
136
Initially the points could be fitted to a C
2
function for
mortar forces and a C3 function for moment. As the experimental data
increased, empirical curves could be established and values of Es to be
used with basic types of reinforcement could be determined. The final
stage would be to determine whether an improved linear mortar model would
be an improvement. It is likely that the experimental data will show
a great enough spread that a more complicated model would not be justified.
Rather than establishing working stresses for design
purposes, the possibility of determining strain limits. for tension and
compressive strains under design loads should be investigated.
Most of the available data gives load/strain or load/
deflection curves for specimens loaded progressively to failure. Since
boat hulls will occasionally experience high loads and the effects ~ a y
be cumulative, it is essential to monitor specimens through repeated load
cycles to higb loads and througv cycles of alternating loads. A design
procedure must consider the strains and strengths of the material after
many years of service, not just for the new condition.
137
TABLE 1. Mowat's Data for Series 3 Plank 9.
13 layers 1/2" x 1/2" welded mesh - assumed evenly distributed.
Wire dia - 0.042 in.
-Thickness 1 in.
Width
Length
6 in.
33 in.
Total longitudinal wire = 0.00276 sq in./in./layer
a 0.0359 sq in./in.
Moment of steel about steel centroid I -
s
Mortar compressive strength = 4830 psi.
Compressive Tension
Load Defl. Strain Strain
(lb) (in. ) (in./in. ) (in./in. )
150 .035 0.010 0.02
300 .129 0.027 0.06
450 .260 0.045 0.114
600 .385 0.061 0.176
750 .529 0.080 0.241
900 .719 0.107 0.950
1050 1.19 0.157 0.642
1090 Failure
4
0.00257 in. lin.
Total
Strain Neutral Axis
(in./in. ) height/thickness
0.030 0.30
0.087 0.31
0.159 0.283
0.237 0.257
0.320 0.250
0.457 0.234
0.819 0.192
138
0.50 1000

0.40 800
c:s
'"
..
"Load for Linear Hodel
u
II)
oM
~ .0
lIS
0.30
r-I 600
.-4
..
X
lIS -t:S
"""- C for Linear Hodel
~ lIS
.&J 0
::s H
Q)
c:s
'H
0
\c
.&J
0.20 400
,d
00
'"
Q)
~
0.10 200
o
Fig. lA.
o
o 0.20 0.40 0.60 0.80
Strain, percent
The derived height of neutral axis, C = c/h and derived loads
at various strain levels compared with measured values by
Mowat (Series 3, Plank 9)*
~ o w a t , D.L., Flexural Testing of Ferro-cement Planks, Thesis
for Master of Science in Civil Engineering, University of
Calgary, Calgary, Canada, Jan. 1, 1970.
1.00
.05

.!t
_<'01
~ s:i
I ~ .10
.
~
-
\0
I
o
.-4
- .15
.20
Fig. 2A.
.2-
c
C
a
-
h
h
.3
Plot of modulus functions vs strain to obtain the value
of C and hence location of neutral axis.
140
III)
III)
..
1.0
...
41
III)
~
...
a
..
0.8
41
!
41
rot
='
......
CD
CD
U
...
0.6
41
CD
...
41
...
a
0
....
0.4
41
~

'W u
u ~ = '
0.2.
o
o
Fig. 3A.
fc
R
-r-
vs
cu
f
III 5000 psi
c
(). III 4)
0.2 0.4 0.6 0.8
e
R" ~ III Ratio mortar ' strain/ultimate mortar strain.
e
cu
Stress Ratio vs Strain Ratio R.
1.0
III
III
.,
...
U
III
.,
U
]
u
....
::J
.......
III
III
.,
...
U
III
...
CIS
u
...
g
.,
00
CIS
...
OJ
~
0
ori
u
:!

cs
141
1.0
0.8
0.6
0.4
A D 4
0.2
o ~ - - - - - - ~ - - - - - - - - ~ - - - - - - ~ ~ - - - - - - ~ - - - - - - ~
o
Fig. 4A.
0.2 0.4 0.6 0.8
e
R D ~ = Ratio mortar strain/ultimate mortar strain
e
cu
Stress ratio a vs strain ratio R.
1.0
\0
I
142
1.0
0.8
0.6
~ 0.4
tC
0.2
A
(Z - C)E s
s-
h
& a 0.01%
O l ~ ~ ~ ~ ______ ~ ______ ~ ______ ~ ____ ~ __ __
0.1 O . ~
F
C
Ca-
h
0.3
Fig. SA. & ~ vs C for Non-linear Mortar.
0.4 0.5
0.8
Assumed e
tu
~ 0.02% e
tu
0.6 a 0.2%
0.2
o
F
A
s
(Z - C)E -
s h
c
0.3
Fig. 6A. : vs C for Improved Linear Model.
0.4
143
144
Appendix 2. Extract from Regulatory Aspects of Traditional and New
Construction Materi a 1 s, Harren E. Bonn, r1arine Regul ati ons
Branch, Canada Tra'nsport Ministry. (Ref. 51.)
Part 1 - General Requirements Applicable
To All Fishing Vessels
. Notification of Construction
The owner or the builder 'of the vessel should advise
the Steamship Inspection Service of the proposed construction, size
of vessel, nature of service, type of material of which it is to be
built and the extent of the voyages for which it is required.
Submission of Plans
Prior to commencement of construction the plans and
information (required by Small Fishing Vessel Inspection Regulations
and Large Fishing Vessel Inspection Regulations of Canada Shipping
Act) should be submitted to the nearest Steamship Inspection Service
Office for approval and should the owner or .builder require any
particular information relative to requirements for the type of
construction he is proposing the Board will be pleased to provide all
possible advice and assistance within their jurisdiction.
Inspection During Construction
During construction a Steamship Inspector will carry
out regular inspections to check that the vessel is being built in
accordance with the approved plans and that the materials and workmanship
are to the required standards. In addition to the hull construction
he will witness all necessary hose testing and tank testing and will
examine and test the machinery installation, piping installations and
steering arrangements. He will also check the lifesaving, firefighting
and navigating appliances and other statutory requirements.
Sea Trials
On completion of construction the Steamship Inspector
shall be present during the sea trials to ensure that the machinery and
all essential services are functioning properly and that the vessel
is operating in a safe and satisfactory manner.
Certification
On completion of the "First Inspection" the Steamship
Inspector will issue an appropriate Inspection Certificate for the
voyages on which the vessel will be engaged. The period of validity
of the certificate will normally be
145
i) one year for vessels of more than 150 gross tons,
ii) one year for vessels that are steam driven,
regardless of their tonnage, or
iii) four years for vessels that are not steam
driven and not more than 150 gross tons.
Periodical Inspection and Certification
Periodical inspections will be carried out by a
Steamship Inspector when renewal of an Inspection Certificate is
required and in accordance with the requirements of the Large and
Small Fishing Vessel Inspection Regulations.
It should always be remembered that it is the respon-
sibility of the owner, operator or master to have his vessel inspected
and certificated in accordance with the requirements of the Canada
Shipping Act. That is to say he should advise the local Steamship
Inspection Office when the vessel is due and ready for inspection and
in the case of new construction, the builder should advise when he
wants any particular inspection or test etc. carried out.
Shipbuilder
Part 6 - Requirements Applicable to
Vessels Constructed of Ferro-Cement
Construction should be carried out at an approved
builders, where the personnel are properly trained and familiar with
the type of work which they are to perform.
Strength
The modulus of the midship section and the stresses in
the structural members should be acceptable to the Board. Calculations
for the reinforcements should be made from first principles.
146
Care should be taken to ensure that the reinforcements
form continuous strength members and that discontinuities and local
high stress areas are avoided.
Materials
The strength of the ferro-cement hull is obtained from
the homogeneous qualities of the cement and the grid re-inforcements
which bind together to form a solid structure. The requirements for
the materials are as follows:
Connections
(a) Cement - The cement although contributing
to the strength of the vessels hull has the
primary function of giving rigidity to the
re-inforcements. For the cement the Board
stipulates the following requirements:
(i) it should be of the Portland or equivalent
type, should be recommended by the manu-
facturer for marine use and should be
approved by the Board.
(ii) the water used for mixing should be clean
fresh water and free of impurities and
chemicals that may effect the concrete mix,
(iii) the aggregate of the mix should be of a
suitable type and as recommended by the
manufacturer and approved by the Board.
The water/cement ratio should be controlled
as low as poss i b 1 e to g.i ve a good qua 1 ity
and workable material.
(b) Reinforcements - The reinforcing pipes, rods,
bars and wire mesh used are to be of an approved
grade of steel for which certificates should be
available. The steel should be clean and free
of scale, oil, grease or other similar contamination.
Welding, lacing and clipping of all main hull re-
inforcements should be carried out with care and completed to the
satisfaction of the Steamship Inspector.
Construction of Tanks
Tanks for oil or water may be constructed of steel or
ferro-cement that has been treated with a suitable sealer.
way of all
possible.
reduce the
Adequate supporting structure should be provided in
tanks and through bolting should be avoided wherever
Longitudinal divisions shall be fitted in wide tanks to
effect of free surface liquids.
Machinery Seatings
Due to induced from the vibrations and weight
of the machinery special attention should be given to the design and
construction of machinery seatings. Care should be taken that hard
.notches and corners are eliminated and the continuity of strength
maintained.
Insl Procedures
During the construction regular inspections will be
car' 'ied out by a Steamship Inspector with particular attention being
given to the following stages:
147
(a) when the steelwork re-inforcement is half complete,
(b) when the steelwork re-inforcement is complete,
Testing Procedures
(c) during the application of the cement mixture,
(d) at the removal of forms or moulds,
(e) at completion of the hull prior to curing,
(f) at completion of the hull after curing, and
(g) on completion of the vessel and during the
running of the sea trials.
At the present time the Board of Steamship Inspection is
participating in a research program, instituted by the Industrial Dev-
elopment Service of the federal Department of Fisheries and being
undertaken by the British Columbia Research Council, to determine the
qualities and suitability of ferro-cement as a shipbuilding material.
hope that results will be forthcoming from this program in the
near future that will provide clear guidelines into the construction,
testing and inspection procedures which we should follow.
148
However, pending the completion of the above mentioned
research program the Board has decided that the following testing
procedures should be adopted:
(a) During the course of construction the Steamship
Inspector will carry out standard slump tests
on the concrete mix to ensure that the mixture
is a good quality and workable material.
(b) There should be prepared, concurrent with the
hull plastering, test specimens of the main
hull structures, the preparation of which
should be witnessed by the Steamship Inspector.
(c) Tests will t : ~ required for tensile, compressive,
flexural and impact strengths. These tests
should be carried out at a recognized laboratory
and witnessed by the Steamship Inspector.
For vessels built in Canada to date these
tests have been carried out at the Department
of Public Works testing laboratories in Ottawa.
(d) The number of test specimens will be decided
upon by the Board for individual vessels
generally depending upon their size, type of
construction and whether the vessel is of a
prototype design.
Provisional Requirements for Periodical Inspections and Certification
Special attention will be given to a vessel of ferro-
cement construction for the first four years of operation and
provisional inspection and certification will be as follows:
(a) the vessel wi 11 be 1 imi ted to Home Trade
Class III Voyages - i.e. not more than 20
miles' off shore and not more than 100 miles
between ports of refuge,
(b) inspection will be made of the vessel afloat
every six months, and
(c) underwater inspection will be carried out
annually.
Following the first four-year period and provided
the vessel is found in satisfactory condition, the normal inspections
will be carried out in accordance with the Large or Small Fishing
Vessel Inspection Regulations as applicable.
Appendix 3 GUIDELINES CONSTRUCTION OF VESSELS
(AMERICAN BUREAU OF SHIPPING)
1. General Conditions
1.1 Classification
149
Vessels which have been wholly or primarily built of
ferro-cement, and which have been built under the special supervision
of the Surveyors to the Bureau, in accordance with these Guidelines,
or their equivalent and other relevant sections of the Rules, will be
consiaered for classification, and where approved by the Committee,
distinguished in the Record by the symbols + Al "Annual Survey. II The
type of construction "Ferro-Cement" will be noted in the Record.
1.2 Workmanship
The Surveyor is to satisfy himself that all operators
employed in the construction of vessels to be considered for classification,
are properly qualified in the type of work proposed, and that equipment
and other facilities are such, that acceptable standards can be obtained
for the construction of the hull, superstructures and appendages
thereto, and for the installation of equipment, machinery, piping and
the electrical system.
1.3 Construction Surveys
The builder is to maintain a schedule of systematic
"inspections at regular intervals during the construction of the vessel,
and records thereof, made by qualified personnel of the yard, are to be
made available for inspection by the Surveyor. The Surveyor is to be
present at the completion of all the major stages of construction.
Additional visits will depend on the" size of the vessel, and requests
of the owner or builder.
1.4 Surveys after Construction
The hull is to be subject to an Annual Survey on drydock,
equivalent to a Special Survey. The hull is to be examined internally
and externally, and all framing, appendages, deck houses, bulkheads
etc. are to be examined. These Annual Surveys are to continue, until
sufficient experience has been acquired to determine that surveys at
longer intervals are reasonable and proper.
150
1.5 Submission of Plans
Plans showing the particulars
9
arrangements and details
of the principal parts of the hull structure of each vessel to be built
under Special Survey are to be submitted and approved before the work
of construction is commenced. These plans are to indicate clearly
the particu1ars
9
as well as the details and arrangements of the
reinforcements of the hull. A construction schedule, giving details
of materials, mixes, reinforcements, mortar application and curing
procedures, is also to be submitted. Plans should generally be
submitted in triplicate.
1.6 talcu1ations
The designer is to prepare strength calculations to
justify the strength of the hull and its components. These calculations
are to be based on the values obtained from the testing procedure, as
prescribed in Section 4, and are to be submitted with the plans, as
required by 1.5
2. Particulars and Construction
2.1 Definitions
The definition of a hull or structure built in Ferro-
Cement is as follows:
2.2 Reinforcements
A thin, highly reinforced shell of concrete,
in which the steel reinforcement is distributed
widely throughout the concrete, so that the
material, under stress, acts approximately as
homogeneous material. The strength properties
of the material are to be determined by testing
a significant number of samples of representative
panels, according to Section 4. If a similar
approval is obtained from -a recognized authority,
the Bureau may waive these requirements, upon
review of the previous tests and such check tests
as may be deemed necessary.
The steel content of the Ferro-Cement should be as high
as practicable, and arranged in such a manner as to allow adequate
151
of the mortar, and thereby result in a void-free material.
The reinforcing rods, pipes and wire mesh are to be evenly distributed
and shaped to form. Transverse frames or bulkheads are to be fitted to
provide adequate transverse strength. The reinforcement network should
be securely welded or otherwise fastened together, so that" it remains
in its original position during the application of the mortar.
Structural steel sections may be into the
Ferro-Cement structures as longitudinal strength members, floors etc.,
but care must be taken to ensure penetration of the mortar, and a
proper bonding between the framework and the mesh.
The reinforcements may be tapered towards the ends of
the structure, where the hull form becomes "finerll, but care is to
be taken to avoid discontinuities in the strength of the reinforcements,
and ends of members are to be faired into the adjoining structure.
The overlaps of the mesh layers at the keel, transom edges, etc. are to
be staggered to allow even distribution of reinforcement in those areas,
and to ensure satisfactory penetration of the mortar. Butts in
reinforcement are to be suitably staggered, to avoid discontinuities.
2.3 Formwork
In methods of construction where internal or external
forms are employed, satisfactory penetration of the mortar must be
ensured, and the reinforcements are to be secured so that distortion is
minimized on application of mortar. The formwork is to have a
surface, and is to be thoroughly cleaned before applying the
mortar.
2.4 Concrete
The methods employed for the mixing, handling, compacting
and curing of the concrete are to be consistent and result in high
quality material. The mortar should be applied as soon as possible
after mixing, and constant agitation of the mix is to be provided during
the waiting period. If any separation of water from the mix is
observed durinq the waiting period, the mortar is to be remixed before
application. Containers used to transport the mortar are to be clean.
Care must be taken during the application of the mortar, so that no void
spaces remain adjacent to the reinforcements or in corners. Vibrators,
and/or hand rodding, are to be used to compact the mortar at thicker
sections. A complete coverage of the reinforcement is to be ensured,
although the thickness of coating should be kept at a minimum, and an
excessive buildup of cement is to be avoided. Ferro-Cement structures
152
ere to be cured in a satisfactory manner. Various methods of curing
are acceptable, depending on ambient conditions, but in general, the
curing should be done by water spraying, by steam curing under a hood,
or by membrane curing. Curing should normally not commence until
about three to four hours after the mortar application, or when the
mortar has taken its first set. This period may be longer in association
with low atmospheric temperatures. The temperature during the curing
period is to be kept approximately constant. Where a form is employed,
it shall be kept in position for as long as practicable during curing.
3. Material s
3.1 Cement
The cement shall be ordinary Portfand Cement, in
accordance with a suitable approved specification, such as ASTM
C 535-67T. Other cements will also be considered providing they
offer adequate water-tightness and uniform consistency. Cement should
be stored under dry conditions, and if the application of the mortar is
done in stages, a suitable turnover of. cement stock is to be arranged
to ensure consistent freshness. Any presence of lumps in the cement
renders it for use, and it is to be sieved before mixing.
3.2 Aggregate
Aggregates are to have suitable strength and durability,
and are to be free of foreign materials, including chemical salts.
The aggregate is to normally include clean washed sand of a silicious
nature. The aggregate isto comply with a suitable specification, such
as ASTM C 330-68T.
3.3 Water
Water is to be free from foreign materials that may im-
pair the strength and resistance of the mortar. It is to be free of
salts.
3.4 Mixing
Mixing is to be done in such proportions as to consistently
give the required strength, as determined by Section 4. The proportions
of the mix are to be by weight. The water-cement ratio is to be
controlled as low as possible to give the material a consistent quality
153
and workability. Initially this is to be judged by a slump test and
practicable workability under the existing conditions. Once this
criterion is established, the mortar is to be held to a consistent
slump test standard.
3.5 Reinforcements
Reinforcements (rods, pipes, expanded metal, wire mesh)
are to have sufficient tensile and yield strength and ductility, and
other properties essential for good construction. The reinforcements
are to comply with a suitable specification, such as ASTM A 615-68,
185-64 and A 390-66. The reinforcements are to be free frommillscale,
grease and any other contamination. Light corrosion is not objectionable,
but should be brushed to remove free oxide. Black or galvanized
reinforcements are acceptable.
4. Testing
4.1 Mechanical Properties Testing
The mechanical properties tests, as listed below, are to
be performed on representative samples. Prior to commencement of
construction, preliminary tests are to be carried out on standard test
pieces, as described below, in order to determine that the proportions
of the mortar mixes and properties and arrangements of the reinforcements
will satisfy the deSign strength requirements of the vessel. The
preliminary test pieces are to be in accordance with a suitable
specification such as ASTM C 192-68, although curing may be done at an
accelerated rate. Preliminary tests are to be carried out satisfactorily
before construction is begun. During construction, test pieces, as
described below, are to be made from the same mortar batches used in
the actual hull construction, and the following tests are to be carried
out.
For each 50 cubic feet or fraction thereof, a minimum
of one each of the following tests: direct tensile, compressive,
flexural and impact tests" are to be made. At least three of each of
these tests are to be made for each.hull or structure. Hhere 1 arger
unit hulls or structures are being built and large identical mixes of
mortar are used, one set of tests per batch of 10 cubic yards
(7.65 m
3
) or fraction thereof shall be carried out. A minimum of six
sets of tests are to be made for each unit of construction. These
154
tests are to be carried out in a manner that will yield reliable
values of the tensile and strength at both cracking and
failure, as well as the modulus of rupture and elasticity and impact
strength of the reinforced samples. In the construction tests, the
curing is to be in accordance with a suitable specification such as
ASTM C 31-66.
a) Compressive Test
The compressive test is.to be carried out on un-
reinforced samples, which are to measure 4 inches in diameter and 8 inches
long. The compressive test is to conform to a suitable specification,
such as ASn., C 39-66, and the test report is to conform thereto.
b) The tensile strength is to be determined by the "split cylinder"
test, using similar testing apparatus as in the compressive test
above. ASTM C 496-66 specification described the procedure for this
test, and offers the formula to determine the "splitting tensile
strength". The test pieces are to be unreinforced and of the same
size as those in the compressive test. However, it should be noted
that the true tensile strength of the specimen lies between 50 and 70%
of the "sp1itting tensile strength".
c) Flexural Test
The flexural tests are to be carried out on slabs of
concrete approximately 4 feet long and 12 inches wide and of the same
thickness as the hull. The test pieces are to be reinforced and should
have the same pattern of reinforcement as the actual hull. The
flexural test is to conform to a suitable specification, such as ASTM
C 293-68 and ASTM C 78-64, but care must be taken to ensure that the
load application and support blocks to provide a uniform load across
the test piece. Furthermore, readings are to be .taken at both cracking
and failure.
d) Impact Test
An impact test is to be performed on representative
reinforced panels. The thicknesses and reinforcement .of the test
panels are to follow the same patterns as those of the actual hull.
The panel is to be flat and should measure about 2 feet x 2 feet, and
is to have two mutually perpendicular vee notches 1/25" (1 mm.) wide
and 1/12" (2 mm.) deep across the centroid of test panel. The notches
are to be at right angles to the edges. A drop weight type test is
155
to be employed. Failure occurs when the test panel develops a leak,
and this is to be determined by a water-hose test er equivalent.
The test report is to include, the following:
4 ~
1 1 Identification Number.
All dimensions of the specimen.
Applied load that causes failure.
Curing history, and moisture condition of specimen
at testing.
6
5) Qefects of specimen and age.
) Ambient conditions.
156
..,..tx: '4 libl10graphy Of ferracement Literature
AUTItOR. ORGANIZATION
Z Bezukladov. Y.F. et
1
3 Canby. C.B.
4 AdaIllS. R. (T.Y. Lin
and Associates)
8 Harper. R.
t Hartley. R.T.
10 Jackson. G.W. and
W.M. Sutherland
11 fIowat. D.N.
12 Muhlert. H.F.
13 Samson. J. and
G. Wellens
17 Anon.
37
40
41
42
43
Anon.
Anon. Cement 10
Concrete Association
f Australia.
and
ibson. P.
agenbach. T.M.
Hurd. H.K.
Anon. Kaiser Cement
Co Oakland. CA.
Kelly. A.M. and
T.W. Mouat
Krfstinsson. G.E.
Lachance. L. and
P. Fugere
TITLE CITATlOII
Ship Hulls Made of Refnforced Shipbufldfnq Publishing House. xxxxxx
Concrete (ferrocement). Lenfn9rad. 1968. Navshfps Trans
(1148) AD680042. pp 187
Ferro-cement - with Particullr,No. 014. The Department of Haval x x x
Reference to Marine Applicat- IIArchitecture and Marine Engineering.
toni. The University of Michigan. Ann
Arbor. Mfch. March 1969. pp 64.
IFerro Cement Panels. Yol 1. ,CR 69.008 Naval Civil Engineering x x x
l
'Experimental evaluation and ,Laboratory. Port Hueneme. CA.
protective potential. /AD850630.
I
Boltbuilding fn Ferro-cement. ;Book. The authors. Vancouver. B.C x
l
ea 1.67. pp 36. mimeo.
x
x x
,x
X
x
x
Boatbuildfng with Hartley. ,Book. Hartley's Boat Plans Ltd.. x I'X x :x
I IAuckllnd. N.Z. 1967. 33-50. !
'Concrete Boatbuilding. ;Book. Allen and Unwin. London. 1969. x xlxlx x x x
iPP 106. ' ,
:Flexural of Ferro- iM.Se. Thesis. Dept. Civil Engineeringx xlxi ,x x x
;cement Planks. j'UniV. of Calgary. Calgary. Alta. " II""
I ,1'70. I
!AnalYSis of Ferro-cement ,in :No. 043. The Department of Naval ,x x:x ,x I xi
I
IBendtn
g
:Architecture and Marine Enqineering. : i I 'I! I
jThe Untversity of t';chigan. Ann ! iii"
:Arbor. Hich. Jan. 1970. pp 86. ! I "
! I . ,
"'BHoaOWtt.o Build a Ferro-Cement 'Book. Samson. Marine Design IX x 'I ,x IX I::
X
,'II I
:Enterprises Ltd Vancouver. B.C.
p.68. pp 136. I
struction of an 18m Ferro- 'Kung-ch'eng Chien-she (Engineering Ix x I ix I" '
nt Hyperbolic Shell. 'Construction) Peiping. No.9. 15 I! ' I
:May 1960. Trans, OTS U.S, Dept. II', I I,:, I
jCommerce. Washington 25. D.C
'Ferro Cement boats.
I
I
tontrol of cracking in
concrete.
8 Nov. 1960. pp 7. I I : i
Monthly. Sept. 1967. r r r I

C.39. Cement 10 Concrete 'x I


saciatton of Australia. ca 1968. : ; I
PP I I
! ! I
,Battelle Technical Review 17. 'x jX I
jSePt/Oct 3-9 ,; !
erro-cement construction for :Fishing News International. April , 'x
hhin, vessels. :1.68. 51-55. May 1968. 39-43. 'i!,', I "
1968. 30-32.
hhboats in Ferro-Cement Fisheries. Jan. 1968. ,:!X'ix

erro-Cement Boats Conf. on Fishing Vessel' x:,';XIX
Ferro-Cement Boats.
;Ferro-Cement Sea-going
,"httO<t"" "",,,to.
Ferro-cement as a Fishing
ivessel Construction lolaterialo
,The Growing Acceptance of
Ferrocement as a First-class
Boatbuilding Material.
Construction of a Ferro-
shotcrete Motor-sailer Hull.
tonstruction Materials. Montreal. I
Pet. 1-3. 1968. Canadian Fisheries l :, I

No. 12. June 1969. 365-371.


J. Amerfcan Concrete Institute. i!X I x'
March 1968. I I
!Speci.l Report T-l9. Concrete Trends.x :
iKaiser Cement Co Oakland. CA. 'I
I
PP
8. I
;Canadfan Fisheries Report No. 12. IX
:June 1969. 135-162.
I
iSla Harvest 10 Ocean Sciences.
!Dec. 1 969/Jan. 1970. 32-34.
L'1ngenfeur. March. 1970.
x x
X X x x x
43 Lachance, L
46 Morgan, R.G.
47 Morgan, R. G.
48 Nervi, P.L.
50 Nervi, P.L.
51 Nervi, P.L.
52 Nervi, P.L.
53 Oberti, G.
54 ,Oberti, G.
62 Collen, L.D.G. and
R.W. Kirwan
63 Collen, L.D.G.
65 Anon. Portland
Cement Association,
Skokie, 111.
66 Samson, J.
67 Fraser, D.J.
68 Hedges, L. and
E. Perry
69 Gardner, J.
71 Daranandana, H.
et al
74 Collins, J.F. and
J.S. Claman
n Benford, J.R. and
H. Husen
78 Romualdi, J.P.
Ferro-shotcrete: a promising Ocean Industry, Nov. 1970. x
IIIlteria 1. 60-62.
Concrete.
Concrete as a Shipbuilding
Material.
Underwater Science and Technology ,x, x
Journal. June 1970, 74-80. I
International and Shipping :x
Conference, London. June 1969, Sect-j ,
157
x x
x x
x x x
Its Charact-
eristics and Potentialities.
ton II. Materials, 9-14. "
L'Jngegnere. No.1. 1951 ANIAI. x x x x
Italy. (English Transl.) pp 17. I'
Chap. 4, Ferro-cement.
!Precast Concrete Offers New
Possibilities for Design of
Shell Structures.
Thin Reinforced Concrete
Members form Twin Exhibition
Hulls.
The Penstock of Castelbell0
Some Conclusions about
Deformability and Resistance
lin Tension of Ferro-cement.
Book. Structures. McGraw-Hill, x x
1956, 50-62. I
J. American Concrete Institute, XI x
Feb. 1953. Title No. 49-37, "
537-548.
,Civil Engineering, V.83, Jan. 1951, xl
I
Energia Electrica. 8rochure No.5. x'x,
Vol XXX, 1953, (English Transl.) I:,'
PP 18. I
, Source unknown. Milan, Dec. 1949. Ix x!
i (English Transl.) pp 13. II
Feb. 1959, 195-196.
x
x
,
XI i
II
x
ISome Notes on the Charact-
leristics of Ferro-Cement.
I
some Experiments in Design
and Construction with Ferro-
Cement.
I Civil Engineering and Public Works 1'1: XI:' !;I x I x'I',. XIII
Review, Vol. 54, No. 632.
The Institution of Civil Engineers !X!X;X IXtx! x
: Vol. 86. No.2, II,' I I i I'
;Ferro-cement 80ats. CR OlO.OlG, Portland Cement Assoc- !x,: xIx x
tatton, Skokie, Ill. 1969, pp 13. i II I
Ferro-Cement 80at Construction Proc. Conf. on Fishing Vessel Con- IXI ix x !
struction Materials, Montreal. ' ,! I
I ! I 'I
Oct. 1-3, 19f9, Canadian Fisheries I,' I' ,I
Report No. 12, June 1964, 267-279.
Estimate Hull Work and ibid., 305-311.
Material Content for 100
Combination Fishing Vessel in
Different Materials.
IFerro-cement Fishing Vessels.
,Consultant Finds Ferro-Cement
;Has Limitation.
!Ferro-cement for Construction
"""' ..... "
!Ferro-cement for Marine
'Applications - an Engineer-
ing Evaluation.
Practical Ferro-cement
8oatbuilding.
ibid., 427-429
National Fisherman, Feb. 1970,
4-B, 5-B, 16-B.
Research Report No. 21/14. Dept.
'of Fisheries. rlinistry of Agricul t-
ure, 8angkok, Thailand, 1969,
PP 27.
Presented toNew England section.
The Society of Naval Architects and
Marine Engineers. Boston, Mass.
March 1969, pp 68.
Book, International Pub-
lishing Co., Camden, role., 1970,
pp 176.
The Static Crackinq Stress and Proc. International Conference, The
Fatigue Strength of Concrete Structure of Concrete, Sept. 1965,
Reinforced with Short Pieces London, Cement and Concrete
or Thin Steel Wire. AsSOCiation, 190-201.
x x x x x
x
IX ,x
x x x 'XiX x
x
i II
L" ,11"
: i " I '
,
x x x x: x x
"" I,
158
79 Whitener, J.R.
80 Shah, S.P. and
W.H. Key, Jr.
81 Naaman. A.E. and
S.P. Shah
82 Walkus. R.
831 Walkus. R.
84 Jagtianie. M.K.
I
85 IKar, J.N. and
I A.K. Pal.
86 ,Shah. S.P. and
B.V. Rangan
87 Lessard. Y.
88 ITurner. D.R.
89 Anon. American
Concrete Institute
90 Christensen. K.A.
and R.B. Williamson
91 Tancreto, J.E. and
H.H. Haynes
92 Bi99, G.W.
93 Sutherland. W.M.
Ferro-Cement Boat
Construction.
Book. Cornell Maritime Press, Inc., x x xl lxlx x
Cambridge, Maryland, 1971, pp 128. I'
I I
Impact Resistance of Ferro-
cement.
J. Structural Division, Proc.
American SO,ciety of Civil Engineers.
Jan. 1972. 111-123.
xxx'xxx x
,Tensile Tests of Ferrocement American Concrete Institute J.,
Title No. 68-58, Sept. \971,
693-698.
I
x x xI x
I
x
I
I
State of Cracking and Bul. Inst. Polito Din lasi. x x Xi X x, x
Elongation of Ferrocement I XIV(XVIII), 3-4. 1968, 653-664. \ I
,Under Axial Tensile Load (1) I I
:State of Cracking and ibid. x x XI X X x
I XVI(XX). 3-4. 1970, 53-60. i I I
iConcrete Boats, Barges, and ! Journal not identified, from :xix x: !X'X X x
I
:ShiPS. I GaRlllon India Ltd . Bombay 25, India" I 'i I!
1972. PP 19. I : I
,Strength of Hber-Reinforced I J. Structural Division, Proc. XIX x. !xl ,x
;Concrete. American SOCiety of Civil Engineers, ; I 'I i
i May 1972. 1053-1067. Iii i I
t I I t I I
Reinforced Concrete ! American Concrete Institute J., IX;X!Xi Ix: I 'x
:Properties. 'Title No. 68-14, Feb. 1971, I' I . I II' I
I 1126-135. I i I
IProprietes et Applications du I M.S. Thesis. Facule des Sciences, IX'X x: ix
I Universite Laval, Quebec, P.Q. II .. ; I I ; I
I Hay 1971, pp 196. I I iii :
',An Investigation into the M.E. Thesis. Faculty of Engineering'ix;xix' 'x'xi ix
Tensile Strength of Ferro Sir George Williams University, ; I , : I ! ; I
Icement. Ioklntreal, Que., April 1971, pp 71. I '\ ! , ! i ,
'State-of-the-Art Report. ACI Committee 544, American Concretelx:x x: jX ! !x
t iber Reinforced I Institute, Nov. 1971. pp 56. I iii I j ! !
'Solving the Galvanlc Cell Report tlo. UC SE SM 71-14. ,X!X',x: x:x! Ix
Problem in Ferro-cement. Structural Engineering Laboratory, !; i '! i
I
, University of Cal ifornia, Berkeley, I: I I"!
CA July 1971, pp 59. I ii, I I
! I I I
:Flexural Strength of Ferro- Technical Report R772, Naval Civil ',XiX xl X!x\ IX
,cement Panels. Engineering Laboratory. Port I I I : I'
I
Huenema. CA., Aug. 1972. pp 332. I I' I
An Introduction to Design for Project Report No. 52. Industrial :xlx Xllxix! x
I
Ferrocement Vessels. Development Bra;lch, Fisheries I, I . I I. I
Service, Canada Department of the I
I
Environment, Jan. 1972. pp 224. II i I
I ' I!
,Boats from Ferro-cement. Utilization of ShipbuildinQ and IX!x xl x xlxlx
Repair Facilities. Series No. I, i" 'i I i I I' I
United Nations Industrial Dev- I
e10pment Organizations. Vienna. I i I
United Nations, New York, 1972, i ii
I \
94 Anon. National Acad- 'Ferrocement: Applications
pp 123. , I '
in National Academy of Sciences. Ix x x 'x XIX
I
Washington, D.C., Feb. 1973, pp 89.,; , : I' I emy of Sciences. Developing Countries.
Washington. D.C.
95 Anon. The Society of References on Ferro-cement
Naval Architects and the Marine Environment.
Marine Engineers.
in Technical and Research Report R-14, i I ,x
Task Group H. S. -6-4 (Ferro-Cement) I . I
New York. N.Y.
96 Greeniu5, A.W. and
J.D. Smith
'The Development of Ferro-
for FishinQ Vessel
Construction, llarch 31, 1970,
P 61 and Technical
upplement. llay 31 1970,
pp 53.
The Society of Naval Architects and'
Marine Engineers, New York, N.Y.,
Oct. 1972. pp 32.
Ferro-Cement for Canadian Fishing
Vessels, edit. W.G. Scott, Proj.
Report No. 42. Industrial Dev-
elopment Branch. Fisheries Service,
Department of the Environment.
Ottawa. AU!1. 1971, pp 119.
,
I
x x x XiX
!
x
I
I
17 Greenius. A.W. and
J.D. Smith
IS Greenius. A.W.
19 Greenius. A.W.
10"1' ..... "'. R.G.
101 Morgan. R.G.
102IGj-'rv.0.E.
,.,1 ....
,
.". T.G.
104 Cornet.!.. R. B.
W111iamson. S.
Nogarajan, and K.A.
I Chri s tensen
10S'Haynes, H.H.
106 Kudzfs. A.
107 Kowalskf. T.G. and
B.R. Walkus'
108 Walkus. B.R. and
T.G. Kowalski
109 Kowalskf. T.G.
110 Walkus. B.R. and
A. Macldewicz
111 Walkus. B.R .
M. Kaminska and
E.
112 Dinsenbacher, A.L.
and F. Eo Brauer
113 ibid.
114 Calrncross. C.
115 Eyres. D.J.
The Development of Ferro-
cement for Fishing Vessel
Construction-II, 'lay 31, 1971
pp 113.
Ferrocement for Fishing
Vessel Construction-III,
une 1972, pp 54.
Ferrocement for Fishing
Vessel Construction-IV.
une 1973. pp 57.
Concrete Floating and Sub-
rged Structures.
Concrete Ships.
I
Durability of Concrete
Structures in An Ocean
Environment.
,Ferrocement Marine Mixes in a
Iwarm and Humid Environment.
:Chromate Admixture to Improve
Performance of Galvanized
'Steel in Concrete Sea
Structures.
!ReSearCh and Development of
IDeep-SUbmergence Structures.
I
,pre-stressed Polymer-cement
for Sea Structures.
.Concrete Technology in the
Control of Ferro-
Lement Vessels.
,Ferrocement: a survey.
Ferrocement in Hong Kong.
Ferro-cement for Canadian Fishfng
Vessels, Vol. 2. Project Report No.
48. Industrial Development Branch.
Fisheries Service. Dept. of the
EnVironment., Ottawa. Jan. 1972.
pp 113.
159
x x x
ibid., Vol. 3, Project Report No. 55 x x x, I' X x x
Aug. 1972. pp 54.
ibid Vol. 4. Technical Report. X.X x: !x x .x
No. 64. Industrial Development I i I
i I
'I Branch. Fisheries and r1arine Service I' ,',"! ,.
Environment Canada. Ottawa, 1973. pp
57.
: The Concrete Society. Terminal x:,,' "x:: ! ':
i

x
, xl x
'House. Grosvenor Gardens, London
SWIW OAJ. pp 56 and plates.
I " I
Proc. FIP Symposium Concrete Sea Ix.xix Ixi
StrUctures, Tbilisi, Sept. 1972, Iii" i I
Published by Federation Internat- 'i i I
lonale de la Precontrainte, I ! i I,
Terminal House. Grosvenor Gardens,
london. April 1973, pp 114-119. 125. I
ibid pp 141-145.
ibid pp 150-158.
I ibid . pp 159-163.
i
I ibid .
pp 180-185.
ibid pp 164-165.
ibid pp 246-251.
Concrete (London), Vol. 5. No.2.
Feb. 1971. pp 48-52.
Far East Builder, July 1971.
pp 29-35.
x
, I'
Ie X x
i
x
I
x x x
Ix
x x x
x
x
x
i x
I
I
x' x x
x
IX
I
!x
I
x
x
x
x
I !
x I,x 1. x x (
of Ferrocement Symposium on Industrial Spatial and x XIX !x'x X:X
:Shel1s in Marine Structures. Shell Structures, Kielce, Poland, !! I I ;
, June 15-23. 1973. mimeo. pp 385-398. i,' i
'Mechanized of Manu- fbfd., mimeo. pp 399-410. x x x !x x !x
;facture of Prefabricated ; I I i
(She 11 s . ! I I
I I I!
Material Development. Design, I Presented at Chesapeake Section x x x :x X Ix
,Construction, and Evaluation !Meeting, Society of Naval Architects I ' :.'.! ,i
[
' a Ferrocement Planing Boat. and I"arine Enqineers, mimeo, Naval
Shfp Research and Development I
Center. Bethesda. Md U.S,A. pp 77. !! I
I i
ibid. Harine Technolooy, Vol. II, No.3, x x xi XIX xix
\ July 1974. pp 277-296. I
Ferrocement Boat Constructfon Book, International Marine Pub1ish- x x x'x x
ing Company, Camden, Me., U.S.A., I
Survey of Ferro-Cement Fish-
ing Boats Built in New Zeal-
and. mimeo, pp 30.
1972. pp 192. I
Seminar on the Desfgn and Construc- JX X, x X X
tion of Ferro-Cement Fishing Vessels ' I
Wellington. New Zealand, Oct. 9-13.
1972, F.A.O. of United Nations.
160
116 Larsen. H.J . Jr. Study and Evaluation of Prepared by J. Blume & Associates x x x x x x x
cement for Use in Wind Tunnel for National Aeronautics and Space
Construction. Administration, Ames Research
Center. Moffatt Field. Ca.,
NASA-CR-l14501. JABE-ARC-07.
July 1972. pp 155.
117 Anon. The Concrete Ferroeement A Naval Architect. No.4 (1973),
Oct p 115 Socfety Prelfmfnary Report 'of The
Concrete Society Workfng
Party.
118 Greenius. A.W.
119 Johnston, C.D. and
D.N. Mowat
IFerrocement for Fishing
IVessel Construction, V,
10ct. 1974, -pp--.5J1.
I
iFerrocement - Material
'Behaviour in Flexure.
Not yet for distribution
by Industrial 9velopment Branch,
Fisheries and r rine Service,
Envtronment Can da. Ottawa.
x
x x x
1
I
i
I x
I
IX xix x
t
I
I
120' Roberts. W.H.
I
Guide to Ferrocement Sail &
tProc. American /Society of Civil
Engineers, J. of the Structural
Division. V.100 No. St 10,
October 1974. 2053-2069.
Book, Baycrete Marine, Hamilton
1
xxi
I
j
I !
Xi IX x x
121
1
Smith. R.B.L.
I
I
l221Greenius, A.W.
123 Surya Kumar, G.V.
I 124! Raj ase ka ran. S.,
l
et &1
1,125
1
Rajagopolan, K. and
'V.S. Parameswaran
Ipower Boats and Design.
I
:A Rationale for the Use and
,Development of Ferrocement.
I
:Behaviour of Ferrocement
!Under Repeated Stresses.
i
'An Investigation into the
Flexural Behaviour of
:Ferrocement.
:Behaviour of Ferrocement
;specimens in Bending and
l
ont. (Rapid Blue Print Limited,
:Hamilton, Ont.) 1970, pp llC.
' J. of Structural Engineering
(Roorkee) Vol. 2, No.4, Jan. 1975,
1125-128.
I
I ibid 129-136.
i
i
! ibid., 137-144.
i
i
i
! ibid., 145-154.
ICOmpression. ,
[ i
,Analysis of Ferrocement Beams. ; ibid., 155-164.
i i
,I
I
x
i
i
,
X X X X x:
I
x X x
II
)xx
x x x x t
11261 Basavarajaiah, B.S.,
I H.V. Venkatakrishna,
I and U.R. Raghotham
' Experimental Study of the 1 ibid., 193-197. x x x x x
Applicability of Ferrocement i
;for Precast Folded Plate i
IElements. I
'The Design and Construction ! ibid., 199-202.
of Two Ferrocement Canoes. ;
1127 Troitsky. M.S. and
D.R. Turner
x x
128ISo".'. P.G.
The Impact of !M.SC. Thesis, Naval Architecture x x x
I
129, Snyder. P.G. et al
Mo('lfied Ferrocement Panels. : and Marine Engineering,
!Massachusetts Institute of Tech-
jnology, Sept. 1973, pp 77.
,
The Impact Resistance of 'Report No. MITSG 74-18.
Modified Ferro-cement Panels . Massachusetts Institute of Tech-
Inology, Mar. 31, 1974, pp 74.
x x x
1301,,"", W.E., (,,", ..
Dept. of Transport.
Ottawa.?
Regulatory Aspects of Conf. on Fishing Vessel Constructio X
:nal and New Construction ; Materials, Montreal, Canada, Oct, '
esp.Part 6. 1968. Canadian Fisheries
applicable to Vessels : No.12 June 1969. 73-90
ronstructed of Ferrocement. '
1131 Anon Lloyd's 'Tentative Requirements for the mimeo, 1971,
Register of Shipping, Application of Ferrocement to . pp.13
London Construction of Yachts and!
Small Craft Hull Requirements ' j .
132 Anon American ; uidel ines for the Construction mimeo. 1969.
Bureau of Shipping 'of Ferrocement Vessels. pp.12
x
x
x
x
x
x
x
x
x
x
x

Potrebbero piacerti anche