Cement and Concrete Compositer 19 (1997) 107-122

0 1997 Elsevier Science Ltd

Printed in Great Britain. All rights reserved
0’158-9465/97/$17.u0
ELSEVlER
PI1:S0958-9465(96)00046-7

Fiber-reinforced Concrete: an Overview after 30

Years of Development
Ronald F. Zollo
Department of Civil and Architectural Engineering, University of Miami, Coral Gables, Florida, USA

(Received 16 November 1995; accepted 21 October 1996)

Abstract However, few if any of that era accurately envi-

sioned the magnitude of the impact that FRC
This paper presents a rhetorical discussion on the would have on research and commercial
subject of fiber-reinforced concrete, FRC. It is development communities world-wide.
intended as an overview of the types of commer- In the ensuing three decades, thousands of
cially available FRCs and how they work. It scientific papers have been published on the
discusses commonly applied terminology and subject. Large numbers of individuals have
models of mechanical behavior that form a basis worked their way through academic degree
for understanding material pe$ormance without requirements at all levels, Bachelors, Masters
presenting mathematical details. Historical review and Doctoral, conducting research and contri-
is intended to help build a background for what is buting to the development of FRC. Assessment
currently understood about FRC rather than as of the number of local seminars, regional sym-
historical reporting. References from both early posia and international conferences still held
and contemporary authors are included as a each year and throughout the world provides
means of tying the subject together along a time evidence that there remains a high level of
line. 0 1997 Elsevier Science Ltd. interest in FRC development. Such educational
Keywords: Fiber reinforced concrete, fibers for, programs serve the dual purpose of fostering
mechanics of, concrete mechanics, fiber con- understanding of FRC as a material of con-
crete testing, strength, toughness, energy struction and transferring the fruits of research
absorption. into commercial enterprise.

INTRODUCTION SCOPE OF DISCUSSION

The term fiber-reinforced concrete (FRC) is Intending to be as brief as is consistent with the
defined by AC1 116R, Cement and Concrete goals of the presentation, the discussion is
Terminology, as concrete containing dispersed divided into four parts: Fiber-reinforced Port-
randomly oriented fibers. Over 30 years have land cement concrete composite technology,
passed since the initiation of the modern era of Mechanics of behavior, Testing and perform-
research and development on fiber-reinforced ance evaluation, and Value-added applications.
concrete. In the early 1960s Romualdi, Batson, The breadth of FRC formulations and appli-
and Mandel published the papers’,* that cations makes clear the fact that consideration
brought FRC to the attention of academic and of either the mechanics of behavior or the per-
industry research scientists around the world. formance of FRC in a strictly generic sense, ie.
The writer can report that at that time there fibers in concrete, is an unproductive over-sim-
was a strong sense of discovery and an air of plification. Instead it is necessary to establish
excitement regarding the promise that FRC subcategories within the generic concept of
held for the future development of composite FRC and to establish associated terminologies.
materials based on Portland cement concrete. A discussion of the diverse formulations of
107
108 R. I? Zollo

FRC and associated terminology constitutes the tion as well as the fiber material type, fiber
section 3 of this presentation. geometry, fiber distribution, fiber orientation and
Even while stressing the differences among fiber concentration. Correspondingly, it is
types of FRC, both contemporary and historical expected that more than one mechanical behav-
thinking on the subject point to a somewhat ior theory could aptly model the performance
unified, if not singular, notion as to how fibers of a particular FRC formulation. Likewise,
act to improve the performance of brittle-matrix while the phrase FRC is useful generic termin-
Portland cement concrete. Accordingly, the dis- ology, most often these materials do not permit
cussion seeks to integrate seemingly diverse simple generic descriptions when seeking to
FRC historical developments as a basis for determine material properties needed for
understanding the mechanical behavior. engineering analysis and design.
Much of the discussion incorporates what has The following terminology has been widely
been reported in published state-of-the-art accepted as being useful when describing the
reports,“74 books,5-7 and special publications great variety of FRC composites.
which are available world-wide. However, the
discussion also contains concepts and a point of Fiber materials
view based on experience through long associa-
tion with the subject, especially on the matter of According to terminology adopted by the
performance testing and evaluation. American Concrete Institute (ACI) Committee
Finally, the section on applications is not 544, Fiber Reinforced Concrete, there are four
intended to be exhaustive but selected applica- categories of FRC based on fiber material type.
tions are included as a means of demonstrating These are SFRC, for steel fiber FRC; GFRC,
the utility of applying the energy concepts of for glass fiber FRC; SNFRC, for synthetic fiber
the previous discussion. A complete compilation FRC including carbon fibers; and NFRC, for
of potential applications is left to the imagina- natural fiber FRC.
tion of the reader. Table 1 lists the materials and properties of
fibers which are currently commercially avail-
FIBER-REINFORCED PORTLAND CEMENT able to FRC production.
CONCRETE COMPOSITE TECHNOLOGY

Terminology Fiber geometry

The character and performance of FRC Individual fibers are produced in an almost
changes with varying concrete binder formula- limitless variety of geometric forms including

Table 1. Selected fiber types and properties

Fiber type” Equivalent Specific Tensile Elastic
diameter gravity strength modulus
(in x IO-“) (W (hi)
Acrylic 0.5-4.1 1.16-1.18 39-145 2000-2800
Aramid I 0.47 1.44 425 9000
Aramid II 0.40 1.44 340 17000
Carbon I 0.30 1.9 250 55 100
Carbon II 0.35 1.9 380 33 400
Nylon 0.90 1.14 140 750
Polyester 0.78 1.34-1.39 33-160 2500
Polyethylene 1.040.0 0.92-0.96 1 l-85 725-17 000
Polypropylene - 0.90-0.91 20-100 500-700

Alkali-resistant 2.7-2.74 355-360 11400-l 1600

Non alkali-resistant 2.46-2.54 450-500 9400-10400

Coconut 4-16 1.12-1.15 17.4-29.0 2750-3770

Sisal 40-82.4 1880-3770
Bagasse 8-16 1.2-1.3 26.65-42.0 2175-2750

*Listed fiber types include synthetic, glass and natural fiber varieties. Steel fiber types and properties are governed by
ASTM A820. Not all types of synthetic fibers are currently used for commercial production of FRC.
1 in = 25.4 mm; 1 ksi = 6.895 MPa.
 Fiber-reinforced concrete: an overview 109

Prismatic: rounded or polygon cross-section Fiber denier

with smooth surface or deformed throughout or
only at the ends. Principally when discussing SNFRC, the term
Irregular cross-section: cross-section varies fiber denier (or just denier) is often used. This
along the length of the fiber. is terminology that evolved from the textiles
Collated: multifilament (alternatively termed industry. The denier of a fiber is defined as the
branching or fibrillated) or monofilament net- weight, in grams, of 9000 metres of fiber. Fiber
works (or bundles) that are usually designed to denier is mathematically related to equivalent
separate during FRC production (mixing). diameter for individual fibers if the specific
Present concrete mixing and placing technol- gravity of the fiber material is known. This is
ogies include: presented in the following equation:
batch mixing and placing,
112
pneumatic placement,
slip forming (moving-form extrusion),
extrusion (through a stationary die),
slurry infiltration, and where:
sheet production with or without vacuum or
d equivalent diameter
pressure forming.
f 0.0120 for d in millimetres,
Specific production techniques allow almost
f 0.0005 for d in inches,
any fiber geometry and a wide range of fiber
D fiber denier, and
amount, usually specified as a volume percent
SG fiber specific gravity.
of the total composite, to be placed.
The nomograph of Fig. 1 is a useful graphical
Equivalent diameter tool for conversion between fiber denier and
equivalent diameter.
For fibers that are not circular and prismatic in
cross-section, it is useful to determine what Pre- and post-mix denier
would be the diameter of an individual fiber if
its actual cross-section were formed as a pris- The terms pre-mix denier and post-mix denier
matic circular cross- section. The equivalent are sometimes necessary in material specifica-
diameter of a fiber is the diameter of the circle tions to reflect the fact that the form of the
having the same area as that of the average fiber may change from the time that the fibers
cross-sectional area of an actual fiber, are added to a particular concrete mixture until
Relatively small equivalent diameter fibers the time that they become dispersed within the
have correspondingly low flexural stiffness and mixture. This applies to collated or bundled
thus have a certain ability to conform to the fiber products aof all material types. For
shape of the space they occupy in the paste
phase of the concrete mixture in between aggre- 1 .o

gate particles. Relatively large equivalent

diameter fibers have greater flexural stiffness
and will have a correspondingly greater effect
on consolidation of aggregates during the pro- 0.1
cess of mixing and placement.
EQUIVALENT
FIBER

Fiber aspect ratio DIAMETER.

d

0.01
The fiber aspect ratio is a measure of the slen-
derness of individual fibers. It is computed as
fiber length divided by the equivalent fiber
diameter for an individual fiber. Fibers for FRC
0.001
can have an aspect ratio varying from approxi-
mately 40 to 1000 but typically less than 300. FIBER DENIER _ 0
This parameter is also a measure of fiber stiff- FIBER SPEC. GRAV. 7%

ness and will affect mixing and placing. Fig. 1. Fiber diameter vs denier relationship.
110 R. I! Zollo

example, some synthetic, glass, natural or steel The binder (matrix) component
fibers are introduced to concrete mixtures
either as rovings (bundles), or as collated The matrix material applicable to this discus-
(loosely connected) fibers. Thus fiber denier will sion is ordinary Portland cement concrete.
be different for bundled or collated (pre-mix) However, other brittle cement types have been
used as a host matrix for special applications.
fibers than for the individual fibers (post-mix).
Matrix materials suitable for refractory appli-
When the mixing process is sufficient to break
cations contain high-alumina cement and utilize
up fiber bundles, by mechanical shear action or
special mineralogy and mixing processes.* FRC
by dissolution chemistry, then it is the post-mix
refractory precast has enjoyed widespread
fiber denier that should be applied to calcula-
industrial acceptance.
tions involving the number and distribution of
Early work on extruded FRC9 a pressure
fibers within a particular mixture.
forming process, as well as more recent
research’” into this production technique, used
Strength and toughness
mineral additives to assist in preventing dewa-
tering and segregation during the pressure
Strength and toughness are generic terminology
forming process. Provided there is ample fiber
useful only when precisely defined and deter-
strength and ductility, FRC performance is
mined. As regards FRC, no single definition is
generally enhanced by pressure forming produc-
universally accepted. Furthermore, consensus
tion processes thus confirming theories of
on a definition does not appear to be forth-
increased fiber efficiency through enhanced
coming and in fact may not be necessary. If
fiber-to-matrix bonding.
strength is considered a stress capacity, and
Portland cement concrete with only bubble
toughness an energy capacity, then when and
voids and fiber as the aggregate phase is termed
how these are measured must be specified.
fiber-reinforced cellular concrete (FRCC). Cel-
Strength and toughness measurements are
lular concrete is produced in two ways, either by
affected by the particular testing machinery and
utilizing foaming agents or by a process using
measurement devices employed, and by the size chemical and mineral additives to achieve a
and shape of test specimens. As it is generally volume expansion within the matrix. In this
accepted that the principal benefit of fibrous form the matrix is relatively weak and brittle.
reinforcement relates to tensile stress and strain The latter system of FRCC is commercially
capacities, cracking and crack propagation are available utilizing polypropylene synthetic fib-
the failure events most often used in strength ers. 11.12 FRCC is a relatively lightweight
and toughness definitions. But these events can- concrete, about 40 lb/ft3 (640 kg/m3), having
not themselves be precisely determined, which suitable and reliable strength and toughness
makes comparison of test results among testing properties for many applications.
laboratories problematic. Nevertheless, strength To summarize as regards the matrix materials
and toughness, both in pre- and post-cracking currently used for FRC, the composite matrix
regimes of performance, are the parameters material is based on Portland cement and is
best suited for establishing design criteria for always brittle. Fibers are added to inhibit crack-
FRC. ing, control the brittle fracture process, provide
Precise determination of strength and tough- reliable post-cracking strength and, by virtue of
ness, however specified, generally requires the post-cracking strength and deformation
sophisticated and costly testing procedures. behavior, provide post-cracking toughness as
Such procedures are thought to be more applic- well.
able to R&D efforts than they are to
production and quality control testing. Meas- Regimes of fiber content
urements which are generally required for
engineering design and specification or for The concentration of fiber within a given unit
quality control should be obtained with less volume of FRC ranges from high to low relative
effort. Test methods which integrate over the to the total volume of concrete produced. It is
imprecisely defined events that make testing useful to classify FRC on the basis of fiber con-
problematic, such as cracking, are under centration (volume percentage) as this one
development and are discussed later in this text. factor is seen to significantly affect mixing, plac-
 Fiber-reinforced concrete: an overview 111

Likewise, the notion that cracks which

encounter fibers will consume energy depending
ACI FIBER TYPES upon how the crack progresses through or
around fibers is related to both the number of
fibers encountered and the surface area of these
fibers. Energy-absorbing mechanisms, debond-
ing and pull-out being two examples, depend on
the surface area of the affected fibers.
Lastly, the ability of fibers to transfer stress
across a crack will depend on the cross-sectional
area of fibers within the crack plane as well as
on the elastic properties and bond character-
istics of the fibers. Therefore, the fiber count
- CURRENTLY PRACTICAL FRC RANGES (FC), fiber specific surface (FSS), and reinforce-
1 BY BATCH MIXING AND CASTING
2 BY SLURRY INFILTRATION ment area are of particular interest.
3 BY SHOTCRETE
4 BY PRESSURE OR VACUUM FORMING

Fig. 2. Fiber types and amount used by volume per cent

Fiber count and specific suface
of matrix.
Fiber count (FC) and fiber specific surface
(FSS) are the number of fibers in a unit volume
ing, and hardened concrete performance, as of FRC and the surface area of fiber in a unit
much as any other single factor. Volume per- volume of FRC, respectively.
centage may be considered high if in the range Consider the mass of an FRC composite
3 to 12%, moderate if in the range 1 to 3%, and strictly on a volume basis. The total volume of
low if in the range 0.1 to l.O%, based on the fiber in any given unit of volume of composite,
total volume of the concrete produced. i.e. the volume fraction (or percentage if multi-
Fiber concentration affects the choice of plied by loo), may consist of only one single
FRC production technology. Figure 2 provides (large) fiber or it may be any number of smaller
a summary of the basic fiber types, their ranges individual fibers. Thus, for any given fiber
of commercial application, including the low,‘” volume, the number of fibers depends on the
moderate, and high14 regimes, and the associ- individual fiber volume. In any case, large or
ated production technologies. Notice that the small numbers of fibers, the fibers are neces-
low range of fiber addition is apparently well sarily located and distributed within the paste or
suited for batch mix preparation using conven- binder phase of the composite. The actual
tional mixing equipment and drop placement. number of fibers in a unit of volume of the
Higher fiber concentrations often require composite, FC, may be computed from any one
special mixing or placing techniques. of the following exit-essions:-

1
7.5 x DRT x lop4
Fiber parameters relating to geometry FC=
L lxd2xSG

=[y;;2v]
=[3-o;x;;lo3
Fiber parameters that relate to fiber geometry
can be usedI in evaluating fiber effectiveness.
Specifically, the number of fibers within a unit
volume of concrete, the surface area of fibers in

[ 1
a unit volume of concrete, and the cross-sec-
5.08 x Vx SG x 10”
tional area of fibers across a given plane of an =
FRC volume, appear to be the most relevant. 1 x PoMD
Assume for the moment that the presence of
where:
fibers influences fracture energy requirements
during crack propagation. Then, for any cause V total fiber volume percentage/loo,
of damage, the probability that a crack will 1 fiber length (in inches),
extend to reach critical size initiating unstable d fiber equivalent diameter (in inches),
failure will depend in some way on the fiber SG specific gravity of the fiber material,
spacing, which is related to the packing density. DRT fiber dosage rate (lb/yd3), and
112 R. l? Zollo

PoMD post-mix denier. FC, and by multiplying FC by rc, d and L, find

FSS. The results are for a unit length of fiber.
Multiplying the FC by the surface area of a
For other than unit-length fibers divide the
characteristic individual fiber gives the fiber
result by the actual fiber length.
specific surface, FSS, in the unit volume of com-
Inasmuch as the fibers must reside in the
posite. Then:
paste phase of the composite, there are trade-
FSS=FCxnxdxL offs that can affect FRC performance in the
Alternatively, FSS may be computed directly plastic (mixing and placing) production process
from any of the following expressions: as well as in the hardened state. The ability of
fibers to either conform to or displace aggregate
2.36 x DRT x 1O-3 particles is expected to be greatest in the case
FSS = ]= [?$I]
dxSG of coarse-diameter, relatively stiff and long fib-
ers and less in the case of fine-diameter, flexible

1
4.71 x DRT and short fibers which are better able to con-
form to the aggregate packing density and
= (PoMD x SG)“*

8xV/x
shape.

1
(SG)“* x lo3
= Reinforcement area
(PoMD)~‘*
It is often desirable to determine the actual
The nomograph of Fig. 3 is useful to deter- cross-sectional area of fiber that cuts across any
mine FC and FSS for fibers of specific plane of an FRC composite containing a ran-
equivalent diameter and fiber volume (diameter dom distribution of fibers. This is the
and length). To use the chart enter the volume reinforcement area, equivalent to A, in a con-
percentage of fiber at the bottom, proceed ver- ventional steel-reinforced concrete beam.
tically upwards to a chosen fiber equivalent Recognize that the desired cross-sectional
diameter, turn to the left-hand scale and find area is a planar or two-dimensional concept and
that, for uniformly distributed fibers within a
unit volume of concrete, many fibers will not
penetrate a random single plane. The implica-
tion is that the computation of reinforcement
area is affected by the length of individual fib-
ers.
For example, for any system of units chosen,
if the actual fiber length is one half of the
dimension of the side of the unit of volume of
concrete which is the depth dimension, and if
the fibers are assumed for the moment to be in
perfect (most favorable) alignment as reinforce-
ment (perpendicular to an intersecting plane),
then exactly two fibers would have to be placed
butt end to end to obtain a uniform distribution
throughout the volume. Any plane cut through
the volume would necessarily intersect only l/2
the total number of fibers in the volume.
Now add the fact that the fibers are not nor-
1L VOLUME PERCENT OF FIBER
mally perfectly aligned but can be assumed to
be randomly oriented. A number of
(1 - ASSUMING CIRCULAR CROSS SECTION
(2 - FOR OTHER THAN UNIT LENGTH FIBERS authors2,‘6-20 have proposed factors to be
MULTIPLY BY RECIPROCAL OF FIBER applied when determining the effective cross-
LENGTH
(3) - FOR SPECIFIC SURFACE MljLTlPLY FIBER section of fibers in a single plane when
COUNT BY TT AND FIBER LENGTH AND orientation effects. One such
EFFECTIVE DIAMETER
considering
Fig. 3. Fiber count and specific surface as a function of recommendation2’ is 54%, indicating that the
fiber volume and geometry. individual fiber cross-sectional area is only 54%
 Fiber-reinforced concrete: an overview 113

effective given random orientation in compari- MECHANICAL BEHAVIOR MODELS FOR

son with the most favorable alignment. FRC
Incorporating all of those factors which influ-
ence reinforcement area discussed above, the Most often energy theories are only indirectly
expression that provides the cross-sectional area considered by design engineers. This may be
of a random distribution of fiber depends only due to valid criticism based on mathematical
upon the orientation factor, the volume fraction difficulties. However, qualitative application of
of fiber, and the area of the plane of concrete energy concepts are useful in relation to model-
for which the reinforcement area is required. ing FRC behavior. By doing so the analyst can
Thus the total cross-sectional area of randomly relate energy concepts to the more familiar clas-
distributed and oriented fibers intersecting a sical methods of mechanics such as are applied
plane through a volume of FRC, Ati, is given as in the analysis of conventional reinforced con-
follows. Letting: crete.
Vf = volume fraction of fibers in the concrete,
Historical analytical development
FC = fiber count, total number of fibers in
the volume, The development of FRC that began three dec-
If = characteristic length of the individual ades ago was not based on the traditional
concept of reinforced concrete. It was based on
fiber,
a fracture mechanics concept whereby the
af = cross-sectional area of an individual strength of a brittle Portland cement concrete
fiber, could be improved by reinforcing the matrix
with closely spaced continuous wires.’ The
I,, w,, h, = length, width and height of the
concrete volume, respectively, operative term was closely spaced. There was
evidence that the improvement in strength of a
then filament-reinforced composite disappeared as
V, = (FC x lf x a,)@, x w, x h,) the filament spacing increased. Initial tests were
based on closely spaced continuous wire fila-
Solving for af: ment but further testing using discrete fibers
af = (Vf x I, x w, x h,)I(FC x If) produced the same result: no strength improve-
ment with increased spacing.2
But the total area of all fibers in the volume,& Fibers suitable for testing purposes, first steel
is: and soon after polypropylene, were not readily
Af=FCxaf= (Vfxl,xw,xh&) available in the early stages of research. Sources
were identified from diverse industries including
Now, the total area of fibers in a volume of the automobile tire manufacturing and textile
depth Z, is related to the total area of fibers industries. FRC was produced and laboratory
crossing any single plane through the volume by experiments were conducted but there was little
the relation: that could be done to control or vary the fiber
geometry component of the concrete.
The first fibers tested were relatively coarse, 6
Therefore to 10 mil (0.15 to 0.25mm) diameter, and were
A fX= V,cxw,xh, thus relatively stiff. Mixing and casting prob-
lems were evident. Coarse, stiff fibers could not
or, including the effects of random orientation, conveniently, and sometimes not even inconven-
A fX= 0.54 x V, x w, x h, iently, be distributed in a batch mix concrete
such that there would result a close enough
The resulting units for AEr are in the dimen-
spacing’ of distributed fiber to produce the
sional units of the area of concrete intersected
desired improvement in composite strength as
by the plane; the units of w, x h,.
fracture mechanics would predict. However,
For example, for 0.1~01% of fiber, in a 1 foot
when suitable mixtures could be made in the
width of 4 inch FRC slab, the total area of fiber
laboratory, a statistically significant improve-
crossing any single plane is:
ment in strength was found in both the pre- and
AfX = 0.54 x (O.l/lOO) x 12 x 4 = 0.026 in2 post-cracking behavior of the composite.
114 R. I! Zollo

The inability to conveniently, and thus eco- ineffective for other than high load-rate applica-
nomically, produce high fiber concentration tions and glass fibers had problems associated
FRC in field applications diverted the interest with durability given the rigors of batch mixing
of researchers away from attempts to improve and the effect of chemical alkaline attack.
the initial cracking strength of FRC. At this An example of how much influence historical
time research took a detour towards a concen- precedent had in directing FRC development
tration on post-cracking behavior. In retrospect was the attention given to the use of even
this was a turning point that actually delayed coarser steel fibers. Precedent required rela-
the development of FRC. tively high fiber concentration (volume per
Since the middle of the 1980s many new fiber cent) in order to obtain useful post-cracking
types and fiber geometries have been intro- strength and toughness levels but high fiber
duced. These have significantly altered FRC concentrations presented mixing and placing
production techniques and have affected FRC problems. Conventional wisdom said that rela-
strength and toughness (crack control) perform- tively high fiber volumes could be maintained
ance measures. Readily available FRC by using even larger (coarser) fibers, which of
component materials, new admixtures, and new course reduces the fiber count and thus
production techniques have each had favorable increases even further the fiber spacing.
effect on the fiber concentrations that can be Research was going in exactly the wrong direc-
achieved in field applications as well as on fiber tion.
bond efficiency; two factors which are known to A fiber geometry similar to that of steel nails
greatly influence mechanical behavior. As a of approximately 2 in (50 mm) length were
result of technological development, materials applied. This allowed field production of FRC
designers and suppliers are now readily able to without significant mixing and placement prob-
economically produce and specify a broad range lems but with significantly fewer fibers within a
of FRC formulations. unit volume of composite, a corresponding
reduction of the fiber specific surface, and even
Reinforcing and strain energy dispersion greater theoretical average fiber spacing
resulted.
Just how do fibers affect FRC performance?
With early emphasis on post-cracking behavior, The fiber spacing misnomer
which recall resulted from the inability to eco-
nomically produce FRC with suitable fiber The phrase ‘close enough as it refers to the
concentrations, the well-known concepts of spacing of fibers needs to be examined. The
reinforced concrete based in engineering phrase is most naturally understood according
strength of materials (SOM) were applied. to its literal implied spatial reference. While the
These procedures require cracked sections and fiber spacing concept was at the root of early
a relatively high modular ratio between rein- FRC development, strict reference to spacing in
forcing materials and the reinforced matrix. The its literal meaning is an over-simplification and
following discussion is intended to show how somewhat misleading.
SOM and the application of conventional rein- The term Close enough spacing’ is a mis-
forced concrete analytical methods alone is an nomer because is does not consider factors
over-simplification of the behavior of FRC. affecting fiber efficiency other than fiber spac-
SOM does not consider the important con- ing. Close enough spacing for FRC in a broader
tribution of fibrous reinforcement in energy interpretation is that which is required to guar-
absorption during crack growth. Recall history antee that a propagating crack is affected so as
once again. The earliest fibers tested were to be stopped or at least slowed in its advance-
coarse steel, glass and synthetic fibers. They ment. Close enough spacing includes the
were all relatively large diameter, relatively stiff probability that further energy input to the
fibers21 in the flexural sense. This caused mixing composite system will either be absorbed or
and placing problems especially when higher redirected to areas of lesser fracture tough-
fiber concentrations were used. These problems ness.22 This behavior is truly a form of matrix
pre-empted reaching the goal of achieving a strengthening. The matrix itself continues to
desired close fiber spacing. In addition, coarse conduct the load (stress) while on the ascending
fibers of synthetic geneses were generally (strain hardening) or the descending (strain
 Fiber-reinforced concrete: an overview 115

FIBER FAfLURE
softening) portions of the material’s load/deflec-

;:,;-I
FIBER PULL-OUT
tion response.
The crack arrest mechanism for FRC is simi-
lar to the manner in which aggregate fillers
absorb energy and are known to arrest micro-
cracks in concrete. It appears, however, that
benefits attributable to fibers accrue to the
brittle matrix in amounts exceeding that which
is predicted by the law of mixtures. In addition
to a fiber’s strength and elastic properties,
improved fiber efficiency is related to bond or
anchorage and the potential for energy absorp-
tion through the accumulated effect of the
number of fibers, their specific surface, and
their orientation within the matrix.
The number of fibers in a unit volume of
matrix material, i.e. the fiber count, FC, is
directly related to the statistical probability of
cracks encountering fibers. The fiber specific Ref: “Fracture Mechanics’ by
surface, FSS, is directly related to the amount T.L. Anderson,
1991.
CRC Press.

of energy that is absorbed in encounters Fig. 4. Energy-absorbing fiber/matrix mechanisms.

between cracks and fibers. In this way weak
brittle matrices can be improved, some would
say strengthened, to provide a measure of post- strength of materials (SOM), and a crack arrest
cracking toughness. The same is true for strong energy-absorption approach related to fracture
brittle matrices but in that case the fibers have mechanics (FM). The relative degree of import-
to achieve even greater fiber efficiency through ance given to either theory may depend on the
strength, bond, stiffness, orientation and specific formulation and technique used for the
numbers, in order to counter the effect of production of the FRC as well as the applica-
increased amounts of stored elastic strain tion. However, experience suggests that too
energy that is released and made available to little emphasis is given the energy-absorbing
drive cracks at failure. characteristics of FRC and that this perform-
The schematic diagram of Fig, 4 demon- ance benefit is under-utilized.
strates the ways in which fibers act to absorb
energy and control crack growth. Starting from FRC matrix design, maturity and production
the leftmost fiber element in the figure and pro- considerations
ceeding along the crack towards the right there
is represented fiber rupture, fiber pull-out, fiber In the case of immature concrete the favorable
bridging by tension through the fiber, and deb- modular ratio required by SOM will exist even
onding at the fiber/matrix interface, for relatively low modulus fiber types. Further-
respectively. These mechanisms do not depend more, at later ages when cracks appear or
on fiber spacing and they are effective, albeit in extend for whatever reason, the modulus within
insignificantly small amounts, even for a single the crack is effectively zero and crack opening is
fiber. It is the cumulative effect, however, of resisted by the transfer of stress through the
large numbers of fibers located in the restricted fiber for any fiber material used provided that
topography of the brittle paste phase of the (1) there is adequate bond or anchorage of the
typical concrete composite that has been shown fiber reinforcement, and (2) there is adequate
to be significant. opportunity for the fibers to participate in the
process, i.e. an adequate concentration of fiber.
Fracture mechanics vs strength of materials Attention should also be given to taking care
that mixing and placement processes2” ensure
Evidently there are at least two theoretical that the particular fiber used will not adversely
approaches to modeling FRC mechanics: the affect composite fracture toughness by becom-
conventional reinforcement theory related to ing matrix-disruptive. The use of coarse, stiff
116 R. F: Zollo

fiber may require adjustment in the aggregate engineering specifications as a means of gaining
mix design to minimize this effect. acceptance in the stream of commerce as a
material of construction, is generally readily
Qualitative summary of FRC behavioral appreciated. However, it is worthwhile to take
mechanics another brief look at history to gage the pro-
gress of standards development as regards FRC.
The chart shown in Fig. 5 summarizes the theo- The history of the development of FRC
retical bases for the two theoretical approaches, through the AC1 Committee 544 began in about
i.e. strength of materials and fracture mech- 1970. This group has been instrumental in
anics, as they apply to FRC mechanics. reporting the state-of-the-art on the subject and
their educational mission is evidenced by the
TESTING AND PERFORMANCE list of publications and symposium sponsorship.
EVALUATION Suited to their function, ASTM has recog-
nized the interest in commercial utilization of
Standards development FRC and has responded with a number of
standards regarding test methods for perform-
The development of standards as regards FRC
ance evaluation, primarily through its
has been guided by contributions from indivi-
Committee C09.42.
duals and organizations through the
Progress through consensus organizations
professional societies. The American Concrete
such as AC1 and ASTM is sometimes made
Institute (ACI) and the American Society for
difficult by divergent commercial interests.
Testing and Materials (ASTM) in North
While delay in the progress of standards
America, and their counterparts on other conti-
development is not always an unwelcome out-
nents such as the International Union of
come, history demonstrates how research
Testing and Research Laboratories for
development has been diverted along non-pro-
Materials and Structures (RILEM) or the Inter-
ductive paths apparently creating useless delay.
national Organization for Standardization
As an example consider the historical develop-
(ISO), have led in the information exchange
ment of testing standards.
process. The influence of standards publishing
organizations in shaping model codes and
Testing for flexural strength and toughness

Attempts to quantify the toughness of FRC

MECHANICS
using energy concepts related to the area under
a stress/strain diagram have been numer-
. . ous. 24-26 Criticism of these methods generally
REINFORCEMENT FRACTURE comes from one of two directions: relevance in
THEORY OUGHENING THEOR’I ( representing material performance and diffi-
(based on SOM) (based on FM)
culty of application.27’28 For example, critics
l Emphasis on l Emphasis on
have argued that the current ASTM Cl018
Bond and Energy standard, which discusses measurement and use
Anchorage Absorption
of strength and toughness performance indica-
l Requires Strong l Requires tors, suffers from indexing characteristic
Stiff Fibers “Adequate”
No’s of Fibers performance measurements to ill-defined crack-
l Cracks Must l Fibers Function in
ing events requiring the use of elaborate test
Span Over Plastic, Young & equipment and sophisticated test technicians
Fibers Mature Concrete
which are not commonly or economically avail-
l Generally Implies 0 Goal is to Maintain able within the construction industry.
a Significant Matrix Integrity
Amount of More recent research demonstrates simpler
Matrix Damage
testing and performance evaluation tech-
/ niques29 which seek to avoid the complexities
-.-
associated with the post-cracking performance
/ FIBER;o;gDUCT / [ -e evaluation procedures as described in ASTM

Fig. 5. Summary of mechanical behavior of FRC.

I C1018. These methods are, at present, termed
post-peak testing methods and are working
 Fiber-reinforced concrete: an overview 117

their way through the process of review and as the system fails by cracking in the FRC speci-
approval. The reader may wish to become men. During cracking of the FRC the load and
aware of the deliberations of AC1 544 and energy that are shed by the FRC and by the
ASTM C09.42 for more details. testing machinery and apparatus transfers to the
Post-peak testing is concerned only with the parallel-loaded steel element.
post-cracking range of performance of FRC as At a predetermined deformation of the paral-
characterized by measures termed the residual lel-loaded system the load is removed and the
strength of the material. Some high-perform- steel element which was placed to control the
ance FRC composites exhibit substantial strain energy transfer during cracking is removed from
hardening, i.e. increase in strength in a strain the system. The flexure test is then restarted
softening regime of behavior, and these are yet with only the FRC test specimen in place in the
to be considered in relation to post-peak test- same test configuration and loading apparatus
ing. However, for the low and moderate ranges as had been used to crack the specimen. A load
of fiber additions which are most common to vs deflection record is obtained up to a chosen
current commercial batch cast and pneumati- deformation such as that which is considered to
cally applied FRC production, the cracking be potentially within the useful range for a
event is usually well exhibited as a single critical chosen material application as measured by
crack and reliable post-cracking strength is average strain or crack width.
found beyond that point. Thus for FRC in The post-peak methodology is consistent with
which macro crack redistribution is not found, the notion that for most FRC applications the
reliable post-cracking residual strength is a usefulness of the material system is less often
measure of material performance which is most related to the ultimate load capacity and more
useful for design and analysis with FRC. often related to the ability to transfer stress
Although the concept has thus far only been across a crack at a given strain or deformation.
developed for the flexural testing, similar per- As discussed earlier, this is especially appro-
formance has been demonstrated in other test priate for applications in which there is some
configurations such as direct tension and shear. degree of confinement of the FRC.
Briefly, post-peak testing recognizes the fact Figure 6 depicts a typical residual strength
that, with the possible exception of high-per- load vs deflection curve for a particular type of
formance formulations, FRC most generally fiber-reinforced cellular concrete (FRCC). For
produces little influence on strength, when this application, the FRC specimen is a
defined as the maximum stress at the time of standard 4 in x 4 in x 14 in (100 mm x 100 mm
failure, beyond that of the plain unreinforced x 350 mm). The steel plate used as the parallel
matrix. Failure in this context generally occurs load element is 0.5 in (12.7 mm) in thickness.
after a sample has been caused to exhibit a The material modulus for the FRCC is quite
substantial matrix crack. But the cracking event, low, approximately 200 ksi (29 kPa). The beam
sometimes erroneously termed the first crack, is is subject to four-point loading. The initial slope
itself difficult to define and associated measure- of the load/deflection diagram indicates the
ments are sensitive to the specific type of testing stiffness of the combination of steel plate and
equipment and apparatus used. The post-peak FRC specimen. This is substantially greater
test accomplishes the failure event without the
need for a corresponding measurement and l LOAD PLATE & SPECIMEN
x UNLOAD PLATE & SPECIMEN
with some degree of control. Thereafter, meas- 1000
” RELOAD SPECIMEN
(CRACKED SECTION. ABSENT PLATE)
urements of residual strength at post-peak 800 K/I STEEL PLATE IN PARALLEL LOADING CONFIGURATION
,,
deformations are a demonstration of the effect 0
_ 600 I IP
of the fiber reinforcement.
The test begins by loading the specimen with
a parallel loading of a relatively stiff elastic ele-
ment as a means of controlling the sudden
0.02 0.04 0.06 0.08 0.10
release of elastic strain energy that is available
to extend damage to the specimen at the time DEFLECTION, in.
of failure. In the case of the flexure test, this is
accomplished by having a steel plate loaded in LOAD v. DEFLECTION IN FLEXURE
parallel with the FRC specimen until such time Fig. 6. Post-peak load vs deflection in flexure for FRCC.
118 R. E Zollo

than the stiffness of the cracked FRC specimen Two categories of applications, one low fiber
loaded in the absence of the steel plate. volume and the other involving high strain or
The post-cracking response demonstrates an loading rates, each of which emphasize the
essentially elastic-plastic behavior. The residual importance of fracture mechanics over the
strength is a reliable tensile stress capacity for a strength of materials approach to FRC analysis,
given crack width and can be used for design are chosen for further discussion.
purposes, as a measure of the post-cracking
strength in comparisons among alternative FRC Low fiber volume applications
formulations, and as a specification for quality
control and acceptance testing. Low fiber volume applications, using less than
0.5 ~01% fiber, have experienced commercial
success. Within this category performance cri-
teria will usually include a desire for a high
VALUE-ADDED APPLICATIONS degree of material integrity in the form of crack
width and area reduction, i.e. crack control, for
The most successful applications of FRC are both aesthetic and serviceability-related consid-
those which result from understanding the erations. These are applications in which the
mechanical behavior of FRC using the energy stress due to loading is less important than con-
concepts of the previous discussion. trol of the effects of volume change.3
There are many applications in which FRC The use of relatively low fiber volumes mini-
may be required to act as a primary structural mizes the effect of fibers in batching, mixing
load-carrying component, i.e. provide structural and placing operations. Also, in these applica-
integrity. However, there are many more appli- tions the energy-absorbing capabilities of fibers
cations in which the fibers are intended and their effect on matrix integrity are realized
principally to augment the integrity of the at all stages of concrete maturity; including
matrix material and in this way favorably affect from the time the concrete is placed and is still
the integrity of the structural system. In the in the plastic state.
latter case, the goal is to affect the non-struc- FRC performance in field and laboratory
tural serviceability aspects of the design. research3’T3’ at 0.1 to 0.5 vol% percentage for
Confinement of an FRC matrix can either be synthetic and steel fiber respectively, makes
from internal sources, such as conventional clear the fact that the effect of the fibers is
reinforcement, or from external structural more in the nature of energy absorption and
support. Applications in the low fiber volume crack control rather than in increased load-
regime include slab on grade and composite transfer capacity. Although the fiber volume is
deck as two examples. In these applications the relatively low the fibers are yet present in large
continued ability to transfer tensile stress, either numbers uniformly distributed throughout the
through the matrix or through fibers which concrete mass so that they are affecting crack
bridge cracks, improves the serviceability propagation in both the immature and the
aspects of design such as durability and tough- mature concrete. For these applications analysis
ness. solely by the SOM approach is not well suited.
Applications of cast in place and precast The genesis of the design as to specification
FRC in new construction and repair include of the amount of fibers is empirically based.
dam, bridge deck, mine, tunnel, canal, and Given the previous discussion on how fibers
reservoir lining, security and utility vaults, work to absorb and dissipate energy, minimum
caisson, pile and pile cap foundation elements, percentage levels of fiber addition cannot
slope stabilization, refractory castables and pre- reasonably be determined. However, most field
cast, modular panel including tilt up and sheet, experience and reported research has been at
breakwaters, mine crib block, machine bases, fiber volume loadings of 0.1~01% for synthetic
pipe, and non-structural flatwork such as for fiber and 0.5 ~01% for steel fiber. In the case of
highway, airport, composite deck, residential synthetic fiber the 0.1~01% level can be
grade slab, and industrial floors. For the wide doubled or tripled, still remaining in the low
range of applications listed there is a corre- volume regime, and corresponding improve-
sponding wide range of appropriate FRC ments in performance can be measured with the
formulations. post-peak test methodology.
 Fiber-reinforced concrete: an overview 119

Some purveyors of synthetic and steel fibers far less than that which would be required to
recommend specification of even lower fiber stretch the solid sheet by the same amount.
volume loadings than the respective 0.1 and What has changed as a result of the slitting
0.5 ~01% discussed above for the fiber they (cracking) process is the effective in-plane stiff-
market. As reported research involving such ness. The slits (cracks) do not propagate as
applications is generally not available, it appears cracks owing to the fracture toughness of the
that such reductions are based solely on eco- material (steel in the lath analogy) and yet
nomic competitive factors. In view of the earlier stress, and force, are able to be conducted
discussion one would have to expect a corre- through the matrix’ of the material.
sponding reduction in effectiveness in The analogy is shown in Fig. 7(d). The short
comparison with higher minimum fiber load- uniformly distributed lines in the schematic dia-
ings. gram are not fibers but rather are meant to
The diagram of Fig. 7 demonstrates schema- represent fine cracks, as shown in the magnified
tically the shrinkage behavior of brittle and insert. Prior to coalescence of the distributed
ductile materials. An analogy can be drawn cracks into a single through crack, the matrix
between maintenance of matrix integrity in material continues to conduct the force. The
fiber-reinforced brittle composites and the force that is conducted is of magnitude some-
manner in which steel wire lath is manufactured where between zero and the force required to
from slit sheet material. stretch the concrete by an amount 6, reflecting
The lath is produced when sheet steel is first a reduction in the stiffness of the composite.
slit, so as to provide cracks distributed through- In relation to the performance of low fiber
out the steel continuum, and then stretched volume FRC, it is also useful to consider exactly
in-plane to expand the sheet by opening the what changes volume and when such volume
cracks and thus form the lath. The force changes are likely to occur in concrete. For the
required to open the sheet which has been slit is most part shrinkage is confined to the paste

(b) F T F=AfA

(DUCTILE)

(BRITTLE)

Cd) F O<F<BA
L

\ (TOUGH)\ /
RANDOM MICRO-CRACKS
(FIBERS NOT SHOWN)

Fig. 7. Shrinkage models for ductile and brittle materials.

120 R. E Zollo

phase of Portland cement concrete. Further- set of the concrete mass. The influence of the
more, much of the volume change takes place fiber in resisting the effects of volume change
within the first few hours after concrete place- within the paste phase can best be described
ment, and continues at lesser rate and towards using the FM and energy approach to analysis
a lesser total amount after the initial chemical and the analogy discussed using Fig. 7.
reactions in the concrete have taken place and
during drying of the concrete. Impact energy application
Figure 8 is a stylized version of the results of
research on plastic and drying shrinkage in con- The ability of FRC to absorb energy trans-
crete and mortar.32 This demonstrates that mitted by high rates of loading has widely been
shrinkage strain is concentrated in the mortar reported.33”5 This capability is further demon-
phase and at an early age in the maturation strated by the fiber-reinforced cellular
process. Fibers, which recall are found in rela- composite in the photograph of Fig. 9. The
tively large number in the paste volume of photograph is itself a composite of three photo-
FRC, are effective from the time of the initial graphs showing the ability of an 8 in (20 cm)
nominal dimension FRCC wall panel to resist
ballistic impact. The photograph shows core
PLASTIC
ZONE
samples taken from a test wall which had been
a- DRYING split open after firing a handgun at the wall at a
ZONE
range of 20 ft (6.1 m). From the left to the right
side of the photograph are, respectively, the
trajectory and position of a 44 caliber magnum
bullet which penetrated approximately two-
thirds of the wall thickness, a close-up of the
same 44 caliber magnum bullet after firing, and
finally a 9 mm bullet which penetrated less than
3.5 in (90 mm) into the wall.
The same wall panel system was tested in
impact according to the Dade County, Florida,
Testing Protocol PA 201-94.36 In the subject
test37 a nominal size 2 in x4 in (50 mm x
100 mm) #2 surface-dry southern pine dimen-
sional lumber projectile, cut to between 7 and
9 ft (2.13 and 2.74 m) long so as to weigh
1 DAY 7 DAYS 365 DAYS
between 9 and 9.5 lb (4.08 and 4.31 kg), is
impacted at 80 ft/s (24 m/s), or 30 ft/s (9 m/s) in
TIME (NOT TO SCALE)
excess of the test protocol requirement. The
Fig. 8. Where and when concrete shrinks. test panel is indented approximately 0.75 in

Fig. 9. Ballistic impact resistance of FRCC.

 Fiber-reinforced concrete: an overview 121

(19 mm) with little or no damage, in the form fits accruing to the composite through the use
of tight surface cracks at some test locations, of fibers outweigh any possible disruptive effects
seen on the reverse side. Conventional nominal that the fibers may have.
8 in (200 mm) concrete block was completely The operative mechanisms that exist for
penetrated in a comparative test. The photo- improving crack control and fracture toughness
graph of Fig. 10 shows the indentations made in mature concrete are also available in
by the standard missile with area dimension of immature concrete composites containing
approximately the same size as that of the lum- fibrous reinforcement. It follows that fibers of
ber impact projectile. all types, including relatively low modulus
materials, which are capable of being placed in
fiber concentrations that ensure a high fiber
SUMMARY AND CONCLUSION
count and specific surface, will improve compo-
site fracture toughness in both immature and in
The mechanics of improving the fracture tough-
mature concrete composites.
ness of FRC composites can be summarized as
Fiber efficiency is a notion of fiber effective-
follows.
ness that considers a combination of factors
Existing as paste-phase reinforcement, fibers
including fiber material properties, bond or
can, as a practical matter, be disruptive to a
anchorage capability, fiber concentration, and
brittle concrete composite. This will be true for
production process. Achieving a desired degree
all fibers but will be especially true for high-
of fiber efficiency provides reliable pre- and
modulus, relatively stiff and coarse fibers. Mix
post-cracking strength and performance related
designs can be adjusted and mixing and placing
to toughness, as may be needed for particular
precautions taken to minimize this effect. Irreg-
applications. The effect of the variety of avail-
ularities in the paste-phase microstructure that
able fiber types on design parameters needs to
are considered disruptive include the formation
be assessed. The development of standards for
of voids and aggregate particle segregation. The
testing and design are still in progress and are
production process must ensure that the bene-
currently limiting the development of FRC.
Many FRC applications are such that the
concrete is acting with some degree of confine-
ment; such as in slabs with edge protection or
confinement, or in combination with conven-
tional and to some degree confining
reinforcement. In these applications improve-
ment in the matrix toughness has particular
significance with regard to energy absorption,
crack control and durability.
In addition to reports from basic research
there has also been noteworthy, if less scientific,
field experience gained on the behavior of FRC
as a material of construction. The developments
stemming from this type of activity have been
noteworthy and should not be overlooked by
the research community. Indeed, based on the
history of FRC, it appears that basic research
has progressed largely in response to field
experience for applications in which there was a
need or desire to use FRC.
Transfer of what has been learned regarding
FRC to the design engineering community has
suffered from a basic lack of training or interest
within the professional ranks in the field of
engineering materials science and the applica-
tion of energy methods in analysis and design.
Fig. 10. Large-missile impact test on FRCC panel. The reader is urged to consult those industry,
122 R. E Zollo

professional, and trade organizations which are 19. Aveston, J. & Kelly, A., Theory of multiple fracture
of fibrous composites. J. Mater: Sci., 8(3) (1973)
publishers of FRC research and development 352-362.
information in order to explore potential design 20. Soroushian, P. & Lee, Cha-Don, Distribution and
applications. orientation of fibers in steel fiber reinforced con-
crete. ACZ Mater: J., 87(5) (1990) 433-439.
21. Goldfein, S., Fibrous reinforcement for Portland
REFERENCES cement. Modem Plastics, 42(8) (1965) 156-160.
22. Shah, S. P., Do fibers increase the tensile strength of
1. Romualdi, J. P. & Batson, G. B., Mechanics of crack cement based mixtures? AC1 Mater: J., 88(6) (1991)
arrest in concrete. Proc. ASCE, 89 EM3 (1963) 595-602.
147-168. 23. Guide for Specifying, Proportioning, Mixing, Placing
2. Romualdi, J. P. & Mandel, J. A., Tensile strength of
and Finishing Steel Fiber Reinforced Concrete,
:oncrete affected by uniformly distributed closely
544.3R-93. AC1 Committee 544, American Concrete
spaced short lengths of wire reinforcement. AC1 J.
Institute, Detroit, MI, 10 pp.
Proc., 61(6) (1964) 657-671.
3. State of the Art Report on Fiber Reinforced Con- 24. Shah, S. P. & Rangan, B. V., Fiber reinforced con-
:rete, 544.1R-82(86). AC1 Committee 544, American crete properties. ACZJ., Proc., 68(2) (1971) 126-135.
Concrete Institute, Detroit, MI, 22 pp. 25. ASTM C1038, Standard Test Method for Flexural
4. state of the Art Report on Fiber Reinforced Shot- Toughness and First Crack Strength of Fiber Rein-
:rete, 506.1R-84(89). AC1 Committee 506, American forced Concrete. In ASTM Standards for Concrete
Concrete Institute, Detroit, MI, 13 pp. and Mineral Aggregates, Vol. 04.02. American Society
5. Balaguru, P.N. & Shah, S.P., Fiber Reinforced Cement for Testing and Materials, Philadelphia, PA, 1989, 7
Composites. McGraw Hill, Inc., New York, 1992.
6. Bentur, A. & Mindess, S., Fiber Reinforced Cementi- 26. ?&I Committee 544. Measurement of properties of
‘eous Composites. Elsevier Applied Science, New fiber reinforced concrete. ACZ Mater J., 85(6) (1988)
York, 1990. 583-593.
7. Hanant, D. J., Fibre Cements and Fibre Concretes. 27. Zollo, R. F., Fiber reinforced concrete flexural test-
John Wiley and Sons, Chichester, 1978. ing: developing standardized techniques. ACZ J.,
8. State of the Art Report on Refractory Concrete, Z’roc., 77(5) (1980) 363-368.
547R-83 (87). AC1 Committee 547, American Con- 28. Balaguru, P., Narahari, R. & Patel, M., Flexural
:rete Institute, Detroit, MI, 224 pp. toughness of steel fiber concrete. ACZ Mater J., 89(6)
9. Zollo, R. F., Fiber reinforced concrete extrusions. (1992) 541-546.
Proc. ASCE, lOl(ST12) (1975) 2573-2583.
29. Bayasi, Z. & Zeng, J., Properties of polypropylene
10. Shah, S. P., Extrusion Research. Advanced Cement
fiber concrete. ACZ Mater J., 90(6) (1993) 605-610.
Based Materials Center for Science and Technology,
Northwestern University, Evanston, IL, 1993. 30. Padron, I. & Zollo, R. F., Effect of synthetic fibers
11. Zollo, R. F. & Hays, C. D., A habitat of fiber rein- on volume stability and cracking of Portland cement
forced concrete. Concrete Znt. (ACI), 16(6) (1994) concrete and mortar. ACZ J., 87(4) (1990) 327-332.
23-26. 31. Kraii, P., A proposed test to determine the cracking
12. El-Aryan, E., An investigation of selected material potential due to drying shrinkage of concrete. Con-
properties of fiber reinforced cellular concrete. Mas- crete Construction, The Aberdeen Group, Addison,
ters thesis, Department of Civil and Architectural IL, September 1985, pp. 775-778.
Engineering, University of Miami, May 1995. 32. Preece, F. R., Concrete cracking: where do we go
13. Zollo, R. F. & Hays, C. D., Fibers vs. WWF as non- from here?. Structural Movements, No. 1. Structural
structural slab reinforcement. Concrete Znt. (ACI), Engineers Association of Northern California,
13( 11) (1991) 50-55. November 1979.
14. Naaman, A. E., Sifcon: tailored properties for struc- 33. Mindess, S. & Vondran, G., Properties of concrete
tural performance. In High Performance Fiber reinforced with fibrillated polypropylene fibres under
Reinforced Cement Composites, Proc. Int. RILEMi impact loading. Cement Concrete Res., 18 (1988)
AC1 Workshop. E and FN Span/Chapman and Hall, 109-115.
London, 1992. 34. Banthia, N., Mindess, S. & Bentur, A., Impact
15. Zollo, R. F., Fibers by the number: the role of fiber behavior of concrete beams. Mater: Strut., 20(119)
count as paste phase reinforcement. In High Peqorm- (1987) 293-302.
ante Fiber Reinforced Cement Composites, Proc. Int.
35. Gopalaratnam, V. & Shah, S. P., Properties of steel
RILEM/ACI Workshop. E and FN Span/Chapman
fiber reinforced concrete subjected to impact load-
and Hall, London, 1992.
16. Krenchel, H., Fibre Reinforcement. Akademisk For- ing. ACZJ., Proc., 83(l) (1986) 117-126.
lag, Copenhagen, 1964. 36. Protocol PA 201-94. Dade County Building Code
17. Parimi, S. R. & Rao, J. K. S., Effectiveness of ran- Compliance Office, Miami, FL.
dom fibres in fibre-reinforced concrete. In 37. Hays, C. D., & Zollo, R. F., Testing applied to the
Mechanical Behaviour of Materials, Proc. Int. Conf., evaluation of damage to FRC and other material
Kyoto, 15-20 August 1971. Vol. 5, pp. 176-186. systems caused by large missile impact to building
18. Kar, J. N. & Pal, A. K., Strength of fibre-reinforced envelopes during storm events. In Testing of Fiber
concrete. X Struct. Div., Proc. ASCE, ST5 (1972) Reinforced Concrete, SP-155. American Concrete
1053-1068. Institute, Detroit, MI, 1995, pp. 217-232.