ENGINEERED CEMENTITIOUS COMPOSITES (ECC)

TAILORED COMPOSITES THROUGH MICROMECHANICAL

MODELING*

Victor C. Li
Advanced Civil Engineering Materials Research Laboratory,
Department of Civil and Environmental Engineering,
University of Michigan, Ann Arbor, MI 48109-2125, USA

Abstract: This article provides a brief overview of several classes of fiber

reinforced cement based composites and suggests future directions in FRC
development. Special focus is placed on micromechanics based design
methodology of strain-hardening cement based composites. As example, a
particular engineered cementitious composite newly developed at the ACE-MRL at
the University of Michigan is described in detail with regard to its design, material
composition, processing, and mechanical properties. Three potential applications
which utilize the unique properties of such composites are cited in this paper, and
future research needs are identified.

*
To appear in Fiber Reinforced Concrete: Present and the Future, Eds: N. Banthia,
A. Bentur, and A. Mufti, Canadian Society of Civil Engineers, 1997.
INTRODUCTION
The application of fiber reinforced concrete (FRC) can be grouped into two
general classes: Thin sheet products and bulk structures. When FRCs are used in
thin sheet products, such as cladding walls, roofing tiles, or pipes, they typically
have the following characteristics (1): A high fiber volume fraction is used, often in
the range of 3-10% with special processing methods (e.g. Hatschek, spray up, lay-
up, extrusion and pultrusion processes) to accommodate the high fiber volume
fraction. The fibers are often in continuous and aligned form (although one of the
most successful form -- asbestos cement, has short random fibers in high volume),
to take advantage of the simple geometric shape of the thin sheet product and to
optimize the reinforcement efficiency of the fibers. As a result of this high
efficiency and fiber volume fraction, this type of FRC often exhibits excellent
mechanical performance in tension and bending, e.g. (2,3), to the extent that
primary steel reinforcement is not needed. Alternatively, one can view the thin
sheet product as one in which primary steel reinforcement is difficult to place, and
therefore the FRC must be able to serve as primary load carrier. The excellent
performance of this type of FRC is often reflected by its strain-hardening behavior
beyond first cracking in tension. Despite this excellent performance, application of
this type of FRC is limited by the simple geometric shape requirement. The pre-
cast nature and special processing needs also restrict the extension of this class of
FRC to applications in cast-in-place and bulk structures.
FRC applications in bulk structures can be subdivided into approximately
three different classes, according to their fiber volume fraction and intended
functions of the reinforcing fibers. Fibers in low fiber volume fraction (<1%) FRCs
are often used for plastic shrinkage crack control. The fibers usually do not serve
any structural functions. Moderate fiber volume fraction (between 1 and 2%) FRCs
are versatile materials which can be found in both cast-in-place and pre-cast
structural members. Because the fibers used are typically chopped and because of
the low volume percentage, regular concrete mixing and casting processes can be
employed. This type of materials are characterized by their improved modulus of
rupture (MOR), fracture toughness, fatigue resistance, impact load resistance and
other desirable mechanical properties (1,4). The fibers in such FRCs are often
regarded as secondary reinforcement used in conjunction with main reinforcing
steel. Attempts at partial replacement of shear steel reinforcement (e.g. 58) and
reinforcement for crack width control (9,10) have been successful. Even so, the
application of this type of FRC often find obstacles in cost/performance
justifications. Recent progress at integrating structural performance with tensile
characteristics of this type of FRC should place the cost/performance at an
increasing advantage. For example, in an ongoing EU project, the tensile stress-
crack width relationship of the FRC is taken into account in the design of a
continuous FRC pavement overlay (11) for handling the expected thermal and
mechanical tensile load. Similarly, in the design of tunnel linings (12), the tension-
softening curve of the FRC is taken into account.

Li. Engineered Cementitious Composites (ECC) Tailored ... 2

 In recent years, techniques at placing large amount of fibers (between 5 and
20% by volume) into bulk structures such as beams, columns and connections have
been successfully introduced. Some examples of high fiber volume fraction FRCs
include SIFCON (520 % steel fibers, slurry infiltrated, (13,14)), SIMCON (6%
steel fiber mat, slurry infiltrated (15); slurry infiltrated steel wool (16)) and CRC
matrix (510% fine steel fibers (17)). These materials have excellent mechanical
properties, including improvement in all strength properties, fracture toughness, and
sometimes even appear to exhibit strain-hardening behavior as in some thin-sheet
FRCs. Because of the significant improvement in mechanical properties, such high
fiber volume FRCs often share primary importance with the main reinforcements in
a given structural member. For example, they have been considered for providing
structural ductility in over-reinforced R/C beams (18) and in brittle carbon FRP R/C
structures (19). Despite their high performance, their wider application may be
hindered by the special processing requirement due to their high fiber volume
fractions and they are often restricted to precast members. These characteristics are
common with those of thin-sheet products. The high cost and weight associated
with the high fiber volume fraction are also important concerns. These materials
are often restricted to steel fiber type. In addition, a number of researchers have
pointed out the anisotropic properties (20) and existence of weak planes (21) in
some of these materials due to fiber orientation effect associated with the material
processing technique.
A major challenge to the research community is to develop a new type of
FRC which has the favorable characteristics of the various classes of FRCs in use
today and summarized above. These favorable characteristics should include:
1. Flexible processing can be used in pre-cast or cast-in-place applications and
no requirement of very special processing machinery.
2. Short fibers of moderate volume fraction to maintain flexible processing,
reduce cost and weight.
3. Isotropic properties no weak planes under multiaxial loading conditions in
bulk structures.
4. High performance leading to significant improvements in strength, ductility,
fracture toughness, and exhibit pseudo strain-hardening.
The fourth characteristics appear to be exclusive of the others, and current
materials satisfy some but not all of these characteristics. Conventionally, research
has focused on studying the property dependence of FRC on one or two parameters
at a time, typically the fiber volume fraction, or fiber length. However, it is now
well-known that composite properties depend on three groups of constituent
properties the fiber, matrix and interface properties (Table 1). The importance of
this is the recognition that fiber volume fraction, for example, is only one of more
than ten constituent parameters under our control for material engineering.
It is not enough to understand the individual influence of each parameter on
composite properties, which can be (at least in principle) established empirically.
Composite optimization requires that the combined influence of all relevant
parameters on composite properties be known. Composite optimization can lead to

Li. Engineered Cementitious Composites (ECC) Tailored ... 3

a composite with excellent performance with only moderate fiber volume fraction,
thus meeting the favorable characteristics of an ideal FRC described above.

Table 1: Cement Based Fiber Composite Material Constituents and Their Properties
CONSTITUENTS PROPERTIES
Fiber Elastic modulus, tensile strength,
length, diameter, volume fraction
Matrix Fracture toughness, elastic modulus, initial flaw size
Interface Bond properties, snubbing coefficient

To establish the combined influence of the constituent parameters on

composite properties, it is necessary to develop a fundamental understanding of the
micromechanisms which govern a given material property. (As a simple example,
it is well known that the unstable propagation of a material defect in the form of a
pre-existing crack governs the tensile strength of a brittle solid). Based on this
understanding, it will be possible to identify the material microstructure and
associated properties which control composite behavior. (The above example will
lead to the defect size and intrinsic fracture toughness as micro-parameters
governing the brittle material strength). Hence micromechanics serves to establish
the link between material constituents and composite properties. The resulting
information can be used to advantage for composite design. When fully developed,
micromechanics can also be utilized as a tool for material property tailoring.
In this paper, we review some results of this micromechanics based approach
(2224) to engineering an FRC which shows pseudo strain-hardening, despite the
relatively low fiber volume fraction (2% and less) and employing typical mixing
and casting techniques. We have called this class of materials Engineered
Cementitious Composites (ECC). In the following, we briefly review the
conditions for pseudo strain-hardening, emphasizing the role played by each
material constituent. Some mechanical properties of a specific ECC which should
influence structural performance are summarized. Potential applications of this type
of ECCs are indicated. Future research needs are then identified.

SYNOPSIS OF MICROMECHANICS BASED DESIGN METHODOLOGY

Micromechanical models constructed on the basis of fracture mechanics and
deformation mechanisms provide an opportunity of tailoring microparameters so as
to control the failure mode, the tensile strength, and ultimate tensile strain of the
composite. Three types of tensile failure modes have been observed in cementitious
materials (Fig. 1): brittle, quasi-brittle, and strain-hardening failure. Brittle failure
can be observed in hardened cement paste material. It is characterized by a linear
stress-strain curve (curve A) followed by a sudden drop in stress at first cracking
with an ultimate tensile strain in the order of 0.01 %. Quasi-brittle failure can be
observed in concrete and most fiber reinforced cements and concretes. It is
characterized by a linear stress-strain curve (curve B) followed by a softening tail

Li. Engineered Cementitious Composites (ECC) Tailored ... 4

after first cracking, due to the bridging action of aggregates, cement ligaments,
and/or fibers. The ultimate tensile strain of quasi-brittle materials is of

Stress
Strain-hardening
Quasi-brittle
B Brittle
A

Strain

Fig. 1: Three Type of Failure Modes Observed in Cementitious Materials

the same order of magnitude as that for brittle materials. Strain-hardening materials
are characterized by their ability to sustain increasing levels of loading after first
cracking while undergoing large deformation (curve C). The ultimate strain value
(at peak tensile load) of a strain-hardening material can be orders of magnitude
higher than that of brittle or quasi-brittle material.
One of the most important conditions for the transition of quasi-brittle to
strain-hardening failure mode is the presence of 'steady state' cracking (2225). In
fiber composites, the extension of a matrix crack is accompanied by fiber bridging
across the crack flanks. As the matrix crack extends, the bridging zone increases in
length. During crack opening, the bridging stress increases as fiber/matrix
interfaces debond and the debonded segments of fibers stretch. When the bridging
stress increases to the magnitude of the applied load, the crack flanks flatten to
maintain the constant applied stress level (26). This load level is termed the steady
state cracking stress. The crack has now gone into the steady state cracking mode,
extending without the need of further increase in applied load. Thus during steady
state cracking, the tensile load is independent of crack length. This is in contrast to
the well known Griffith residual strength concept, which relates a decreasing tensile
load to increasing crack size.
Based on a J-integral analysis of a steady state crack, Marshall and Cox (26)
showed that
ss

Jc = ss ss ( )d (1)
0

Li. Engineered Cementitious Composites (ECC) Tailored ... 5

where Jc refers to the crack tip toughness. In most fiber reinforced cementitious
composites with less than 5% fiber volume fraction, Jc can be approximated as the
cementitious matrix toughness. The steady state stress ss and the flattened crack
opening ss are related via the bridging law (). The bridging law describes the
relationship between the averaged stress carried by the fibers bridging across a
matrix crack and the opening of this crack. For steady state cracking to occur at all,
the steady state cracking stress must be less than the maximum bridging stress o in
the bridging law. That is,

ss o (2)

Eq. (1) and (2) together provide a general condition for transition from quasi-
brittle to strain-hardening failure mode. Apart from steady state cracking condition,
it is also necessary for the critical flaw size dependent first crack strength to be less
than the maximum bridging stress (22). Otherwise, the bridging fibers will not be
able to bear the tensile load shed by the matrix at first crack.
For Eq. (2) to be useful in fiber, matrix and interface tailoring, it will be
necessary to determine the bridging law specific for a given composite system. In
fiber reinforced cementitious composite in which the fibers are randomly oriented
and in which pull-out (rather than fiber rupture) are expected, Li (27) shows that the
bridging law can be derived as

o [ o o] for
2( / )1/ 2 ( / )
o

( ) = o (1 2 / L f )
2
for o L f / 2 (3)

0 for Lf / 2

where o = Lf2/(Efdf (1+)) is the crack opening corresponding to the maximum

bridging stress

1 L
o = g V f f (4)
2 df

Corresponding equations for cases where fibers can rupture and for cases
where fibers are of variable length can be found in (28, 29). In Eqs. (3) and (4), Vf,
Lf, df, and Ef are the fiber volume fraction, length, diameter and Young's Modulus,
respectively. is the fiber/matrix frictional bond strength, and the snubbing factor

2
g= 2 (1+ e )
f / 2
(5)
(4 + f )

Li. Engineered Cementitious Composites (ECC) Tailored ... 6

where f is a snubbing coefficient which must be determined experimentally for a
given fiber/matrix system. The snubbing coefficient raises the bridging stress of
fibers bridging at an angle inclined to the matrix crack plane, appropriate for
flexible fibers exiting the matrix analogous to a rope passing over a friction pulley.
Finally, = (VfEf)/(VmEm), where Vm and Em are the matrix volume fraction and
Young's Modulus, respectively.
The condition for steady state cracking expressed in Eq. (2) can now be
interpreted as a critical fiber volume fraction above which the composite will show
pseudo strain-hardening. Using (1) and (3) in (2), this critical fiber volume fraction
can be defined in terms of the fiber, matrix and interface parameters (24):

12J c

crit
Vf (6)
g (L f / d f ) o

Equation (6) is important for composite design. It provides guidelines for

tailoring the microparameters such that V crit f is minimized. Strain-hardening
composites can then be designed with the minimum fiber content. This idea is
further explained below:
1. Tailoring of Matrix Toughness: Smaller amount of fiber is needed to
make the transition from the quasi-brittle failure mode to the strain-hardening
failure mode if the matrix toughness Jc is small. This has implications on matrix
design since the toughness can be controlled by the water/cement (w/c) ratio, the
volume, size and type of aggregates and microfillers such as silica fume (30).
However, it must be noted that too low a matrix toughness will lead to a low first
crack strength, undesirable for normal service loads.
2. Tailoring of Interfacial Bond: Improving the frictional bond property ,
but without causing fiber breakage, will lead to mode transition at lower fiber
volume fraction. The ductility of the composite is associated with the inelastic
strain generated as a result of multiple cracking (22). This inelastic strain results
from the multiple crack density, and from the opening of each crack. The multiple
crack density is expected to increase with the bond strength , which controls the
rate of stress transfer from the bridging fiber into the matrix material. Fiber/matrix
interface bond properties can be altered by fiber surface modification, fiber
deformation, and transition zone modification.
3. Tailoring of Fiber Length: Eq. (6) suggests that the fiber length Lf
plays a crucial role since V crit
f scales inversely with the third power (approximately,

becomes exact only for low Vf case due to the -term in o) of Lf. (Note that o
scales with the square of Lf). While long fibers are preferred, difficulty in
processing because of poor workability puts a limit on the choice of fiber length.
Further, long fiber length can lead to fiber rupture and poor post-peak behavior. On

Li. Engineered Cementitious Composites (ECC) Tailored ... 7

the other hand, it may not be necessary to use this post-peak behavior if the material
strain-hardens which already provides structural ductility.
Once the failure mode transition is assured, the ultimate strength cu of the
strain-hardening composite can be identified with the maximum bridging stress o
given in Eq. (4). That is,

cu = o (7)

Hence the ultimate composite strength can also be controlled by tailoring the fiber
volume fraction, the interface frictional bond strength, and the fiber aspect ratio
Lf/df, according to (4) and (7). The ultimate strength is not affected by the matrix
properties.
To facilitate composite constituent selection, design charts (30) such as that
shown in Fig. 2 can be used in lieu of Eq. (6).
0.10
2
Jc = 0.02 kJ/m
0.08
2
0.015 kJ/m
0.06 0.010 kJ/m 2
Vfcrit

2
0.005 kJ/m
0.04

0.02

0.00
0.4 0.6 0.8 1.0
Bond Strength (MPa)

Fig. 2: Matrix Toughness & Interfacial Property Effect on Critical Vf. (Ef = 120
GPa, Lf = 12.7 mm, df = 0.038 mm, g = 2.0, Em = 25 GPa)

MATERIAL
ECCs can be formed with a variety of fibers, including polymeric (23), steel
(31) and carbon (29). The matrices used are mostly cement and mortar. So far,
most research has been conducted with a high modulus polyethylene fiber (Trade
name Spectra 900) in a cement matrix, and this material forms the basis of
discussions in this paper. The properties reviewed below involves Spectra ECCs
the composition of which has varied somewhat from test to test, carried out over the
last six years at the University of Michigan. Typical material composition and mix
proportions are given in Table 2. For the exact mix design, the reader is referred
back to the original publications. Fiber properties are given in Table 3. Most
mechanical properties described below are for a composite with 2% of fibers by

Li. Engineered Cementitious Composites (ECC) Tailored ... 8

volume. For comparison purposes, corresponding properties of a typical FRC
containing 1% of (ZL 30/50 hooked end) steel fibers are also shown.

Table 2: Material Mix Proportions

MATLS CEMENT SILICA SUPER- W/C AGGREGATES

FUME PLASTICISER FA/CA
ECC 1.0 0.10 0.20 0.01 0.03 0.30 0.32
FRC 1.0 0.45 1.73/1.73

Table 3: Polyethylene Fiber Dimensions and Mechanical Properties

FIBER FIBER LENGTH ELASTIC FIBER FIBER

DIAMETER (mm) MODULUS STRENGTH DENSITY
(m) (GPa) (MPa) (g/cm3)
38 12.7 120 2700 0.98

The interfacial bond property has been measured (32) both by single fiber
pull-out test as well as by back calculation based on ultimate strength measurement
(using Eq. (4)). The range of = 0.5 to 0.7 appears typical for the type of matrices
used. It is known to depend on age, matrix composition, and even fiber volume
fraction. This low bond property can be increased by a factor of two or more by
means of plasma treatment (32). The snubbing factor g has not been measured for
this material system, although polypropylene and nylon fibers in cementitious
systems show snubbing factors of 1.8 and 2.3, respectively (33). For the present
purpose, a value of g = 2 has been assumed.
The fracture toughness Km measured using LEFM techniques (34) yields a
value of 0.33 MPaP  7KH HODVWLF PRGXOXV Em has been estimated (30) using
Hirch's formula to be about 23 GPa, based on its age and w/c ratio.
Using Eq. (6) and the above parametric values, the critical fiber volume
fraction is estimated between 0.5 % and 1%. It should be understood that this is a
rough estimate, since the exact in-situ values of matrix and interface parameters has
not be measured directly. At any rate, the critical fiber volume fraction is well
below 2%. Hence a composite with 2% fiber should satisfy the condition of pseudo
strain-hardening, and exhibit high strain capacity after first cracking.
The polyethylene fibers are supplied by the manufacturer in bundle-like
form. Prior to mixing, the fibers were dispersed using air pressure for
approximately one minute. Then the amount of fibers needed for the mix was
weighted. After measuring the weight of all mix constituents, the cement was
poured into a three speed (Hobart) mixer with a planetary rotating blade. Silica
fume was slowly added to the cement when the mixer has been started. Then water
and superplasticizer were mixed together and slowly added. When all water and
superplasticizer were added and the cement paste mix became uniform, the

Li. Engineered Cementitious Composites (ECC) Tailored ... 9

dispersed fibers were slowly added by hand to the mix. The total mixing time was
between 15 to 30 minutes depending on the batch size and the amount of fibers used
(fiber volume fraction). After the mix was ready, the specimens were cast under
high frequency vibration (150 Hz) in already greased Plexiglas molds.
Subsequently, they were covered with a polyethylene sheet and allowed to harden at
room temperature for one day prior to demolding. The specimens were placed in a
water curing tank for 4 weeks following the demolding process, and then removed
from water and prepared for testing. For most specimens, a thin white coating of
lime was applied on the specimens prior to testing to better monitor the
development of cracks. The age at testing of the specimens was 30 to 60 days.

MECHANICAL PROPERTIES OF ECC

Uniaxial tensile test, compression test, flexural test and fracture test have
been carried out for ECCs at the ACE-MRL. In each case, the test conditions,
specimen and loading configuration, pertinent stress-deformation curves, and failure
pattern on the specimens are summarized here. A compendium of mechanical
properties obtained so far for the ECC described above is given in Table 4. For
comparison purpose, similar data for FRC tested under the same conditions are also
included in this table.

Table 4: Properties of ECC and FRC

Tensile Compressive Stiff- Flex- Fract-
ness ural ure
fc fc cu cu fc' c' E MOR J
(MPa) % (MPa) % (MPa) % (GPa) (MPa) (kJ/m2)
ECC 2.5 0.021 4.6 5.6 68.5 0.67 22 25 27
FRC 4.3 0.035 4.3 0.035 55 0.48 32.5 10.9 4.9

Uniaxial Tensile Behavior

The uniaxial tensile specimens were tested under displacement control
(loading rate ~ 0.005 mm/s) in a 133.5 kN capacity MTS 810 material testing
system with hydraulic wedge grips. Aluminum plates were glued onto the ends of
the tension specimens to facilitate gripping. Two linear variable differential
transducers (LVDT's) with gage length of 207 mm were used to measure tensile
deformation on the specimen surface (see Fig. 3). More details of the uniaxial
tension test set up can be found in (31).
Figure 4 shows the stress-strain curves recorded on uniaxial tension
specimens. The ECC shows clear pseudo-strain hardening behavior with average
strain at peak stress cu approximately equal to 5.6 % (about 560 time the strain
capacity of the unreinforced matrix). The first crack strength fc and strain fc are

Li. Engineered Cementitious Composites (ECC) Tailored ... 10

 LVDT Output

LVDT Output
LOAD LOAD

LVDT

Specimen
205 304.8

LVDT Holder

Epoxy
LOAD
Aluminum Plate
LOAD
12.7
76.2

All dimensions are in mm

Fig. 3: Tensile Coupon with LVDT Holder Mounted Before Test

FRC
5
ECC
4
Stress (MPa)

0
0 2 4 6 8
Strain (%)

Fig. 4: Uniaxial Tensile Stress-Strain Response of ECC with FRC

Li. Engineered Cementitious Composites (ECC) Tailored ... 11

2.5 MPa and 0.021 %, respectively. The ultimate tensile strength cu is 4.6 MPa.
For this composite, real-time observation showed that multiple cracking occurred
with many sub-parallel cracks across the specimen during strain-hardening. Beyond
peak stress, localized crack extension occurred accompanied by fiber bridging. For
comparison, the stress-strain curves for a 1% steel FRC are also shown in this
figure. This material reveals the typical quasi-brittle behavior of most FRCs. The
softening branch is not a true strain, but reflects the increasing opening of a single
crack (divided by the gage length of the LVDT).

5
(c)
4 (b)
(d)
(a)
Stress (MPa)

0
-1 0 1 2 3 4 5 6 7
Strain (%)
Fig. 5: Uniaxial Tensile Stress-Deformation Record for ECC

Figure 6 shows an example of a damage record at four different stages of

loading (see Fig. 5). As indicated in this figure, although the specimen is already
cracked at stage (a) beyond the first crack strength, the material continues to sustain
the applied load. At stage (b), at about half the maximum strain capacity, more
cracks have developed in the specimen, but the material is still capable of resisting
higher levels of loading. As further deformation is imposed on the specimen,
additional cracking takes place, with saturation approached near peak load (stage
(c)). At stage (d), a single macrocrack has already localized in the specimen and the
material has started to soften. The actual macrocrack was not recorded, but it has
occurred in a location close to the one shown at the top of Fig. 6(d).

Compressive Strength
Compression cylinders (7.62mm x 15.24mm) were tested in an Instron
Model 8000 test system with a 2500 kN capacity loading frame (30). Each cylinder
was tested under displacement control at a loading rate of 0.0254 mm/s.
Compressive stress-strain curves for the ECC and FRC are shown in Fig. 7.
The compressive strength of this ECC, about 68.5 MPa, is not significantly higher
than that of the FRC (55 MPa). The compressive strain capacity has been observed
to increase by approximately 50%-100% over normal concrete and FRCs. Post-
peak ductility of ECCs are expected to be similar to that of normal FRCs.
The modulus of ECC, as in ordinary concrete, depends on the amount of
aggregates used. However, the presence of aggregates also changes other properties

Li. Engineered Cementitious Composites (ECC) Tailored ... 12

of the matrix in the ECC composite. Careful control of aggregate content and size
is of paramount importance in the design of ECCs. The modulus of this ECC has
been measured by strain gages as 20.3 MPa. Higher modulus, but without
sacrifying tensile strain-hardening has been achieved (30).

Fig. 6: Damage Evolution on Uniaxial Tensile Specimens at

Li. Engineered Cementitious Composites (ECC) Tailored ... 13

 (a) = 0.3%, (b) = 2.2%, (c) = 4.2%, (d) ""= 5.9%

80
70 FRC

60 ECC

Stress (MPa)
50

40

30
20

10
0
0 0.5 1 1.5 2 2.5 3
Strain (%)

Fig. 7: Compression Stress-Strain Curves of ECC and FRC

Modulus of Rupture
The geometry and loading configuration of the flexural beam specimens are
shown in Fig. 8. This experimental set-up is recommended in ASTM C78-75,
standard test method of flexural strength of concrete (using simple beam with third-
point loading). The flexural tests were conducted in the same MTS testing system
as the uniaxial tensile tests. The specimens were loaded to complete failure with a
constant cross head speed (0.01 mm/s). The load, head displacement of the
machine, and deflection of the beams at the middle point were recorded in each test.
More details of the test set-up can be found in (35).

P (unit : mm)

76.2
101.6

measuring point

25.4 3@101.6 = 304.8 25.4

.

Li. Engineered Cementitious Composites (ECC) Tailored ... 14

 Fig. 8: Geometry of Bending Specimens
The flexural stress-deflection curves of the ECC are shown in Fig. 9. For
comparison, the stress-deflection curves for a 1% steel FRC are also shown. For the
steel FRC, the flexural stress increases rapidly to the peak value and then starts to
decay. The average beam deflection at peak stress is about 0.6 mm. For the ECC,
however, the flexural stress increases at a slower rate. This increase is accompanied
by the development of multiple fine cracks. The average beam deflection at peak
stress is about 7.4 mm. The flexural strength (MOR) for the ECC is determined to
be 12.5 MPa, compared to 10.9 MPa for the steel FRC. Although toughness index
has not been measured for the ECC, it is expected to be much higher than the FRC
based on the area under their flexural stress-deflection curves.

15
FRC
ECC
Flexural Stress (MPa)

10

0
0 2 4 6 8 10
Deflection (mm)

Fig. 9: Flexural Stress-Deflection Curves of ECC and FRC

The crack pattern of the ECC is distinctly different from plain concrete or
normal FRC. The first crack started inside the mid-span at the tensile face, and
multiple cracks developed from the first cracking point and spread to the outside of
the mid-span. The multiple cracks in the outside of the mid-span were inclined
cracks similar to shear cracks in steel reinforced concrete (R/C) beams. As the
MOR is approached, one of the cracks inside the mid-span started to open up after a
large damage zone had been developed. The through-thickness damage zone can
2
reach an areal dimension of 200 cm . Fig. 10 shows a typical cracking pattern that
develops in the beam middle span around the peak load.
For ideally brittle material, the MOR to tensile strength ratio is unity. For
quasi-brittle material such as concrete or FRC, this ratio lies between 1 and 3. The
upper limit describes the case of a elastic-perfectly plastic material. For the case of
ECC, this ratio can be expected to be higher than 3 due to the strain-hardening
nature after first crack. This expectation is confirmed by the test results (35, 36)
which show that the ratio is equal to 5.0 for the ECC, compared to 2.5 for the FRC.

Li. Engineered Cementitious Composites (ECC) Tailored ... 15

Fracture Energy
Fracture toughness tests of ECCs (34, 37) were conducted in the same MTS
testing system as the uniaxial tensile tests. Double cantilever beam (DCB)
specimens of size shown in Fig. 11 was used to determine the fracture energy of the
composite. A clevis and pin arrangement (similar to that recommended by ASTM

Fig. 10: Cracking Pattern in ECC Beam Mid-Span Around Peak Load

E399-78) was employed at both the top and bottom of the DCB specimens to allow
rotation as the specimens were loaded. The specimens were loaded to complete
failure with a constant cross head speed; the testing time was typically 40 minutes
for all tests. The load-line displacement L was measured using two LVDTs. The
total fracture energy was determined by means of the J-based technique described
by Li et al (38) and using a set of DCB specimens with different notch lengths.
Concurrently with the tests, damage evolution on the specimen surface was
recorded using a camera. The size of the specimen has been chosen to ensure
steady state crack growth.

P LVDT

L H
t
B
W Holder

Li. Engineered Cementitious Composites (ECC) Tailored ... 16

 Fig. 11: Geometry of Fracture Specimens
Large (W = 490 mm; H = 585 mm; B = 35 mm)
Small (W = 127 mm; H = 153 mm; B = 35 mm)

Figure 12 shows an example of load-displacement curves recorded for the

DCB fracture specimens. For the ECC material, it is seen that despite the presence
of the deep notch the material produces significant damage tolerance subsequent to
the bend-over point. Fig. 13 presents the damage evolution recorded for various
load-line deformation values indicated in Fig. 12. It is particularly noted that an
extensive microcrack damage zone spreads around the notch tip before the localized
crack starts to grow. Significant energy absorption is therefore expected from the
off-crack-plane volumetric inelastic deformation process. For the FRC material,
small non-linearity was observed prior to reaching the peak load. Then, the DCB
specimen failed in a quasi-brittle manner by opening of a single macrocrack bridged
by steel fibers. Note that the specimen size used for the FRC is smaller than that
used for the ECC. The larger specimen size used for the ECC material was
necessary to achieve steady-state off-crack-plane cracking in the specimens (34).
200
Load Per Unit Thickness (N/mm)

(c)
(b) (d)
150
(a)

100
FRC (W = 127 mm, a = 75 mm)
ECC (W = 490 mm, a = 241 mm)
50

0
0 5 10 15 20 25 30
Load Line Displacement (mm)

Fig. 12: DCB Load-Displacement Record for ECC and FRC

2
The total fracture energy measured for this ECC was 27 kJ/m , with
approximately over half of this energy consumed in the inelastic damage process
occupying an area of 1150 cm2 around the crack tip, and the rest coming from the
pull-out of fibers on the crack wake. It is noted that in normal FRC, only the pull-
out mechanism of fracture energy absorption is available. Li et al (36) pointed out
that this difference in energy absorption mechanism is responsible for the much
sharper rise in R-curve of ECCs compared to normal FRCs. The fracture energy of
the FRC is this study has been measured to be 4.9 kJ/m2.

POTENTIAL APPLICATIONS
ECCs are relatively new materials. Although further refinement and
optimization are expected, preliminary considerations suggest that the material can

Li. Engineered Cementitious Composites (ECC) Tailored ... 17

be used to advantage in a number of applications. From Table 4, it is clear that the
strength properties of this ECC are comparable to those of high strength concrete,
while the deformation, crack width control, and energy absorption capacities are
significantly improved because of the strain-hardening phenomenon. Applications
in which such properties play an important role in improved performance can take
advantage of ECCs. In the following we describe an example of each of these
classes of applications. The application of ECC as patch repair material is currently
under investigation at the University of Michigan.

Fig. 13: DCB Damage Evolution as a Function of Deformation

(a) L = 3.10 mm (b) L = 7.32 mm (c) L = 19.45 mm (d) L = 23.16 mm

Li. Engineered Cementitious Composites (ECC) Tailored ... 18

Concrete Elements Subjected to Shear
Shear failure is generally brittle in concrete structures. Examples of concrete
structural failure related to shear loading includes bridge deck punching failure (39)
corbel failure (40), anchor bolt pull-out (41) and segmental bridge shear key failure
(42). Since shear failure often involves diagonal tensile cracks, it is expected that
ECC structural members should reflect improve ductility under shear.
The Ohno shear beam configuration (43) was chosen to establish the shear
performance of ECCs (44). Performance contrast with plain concrete,
conventionally reinforced concrete, and ordinary fiber reinforced concrete was
established. The geometry and loading arrangements are indicated in Fig. 14. The
flexural steel has been designed to prevent flexural failure in the shear panel and
ensure a shear mode of failure. The flexural reinforcement layout is shown
schematically. Since a state of pure shear stress exists at the centroid of the
specimen, the Ohno shear specimen gives an estimate of shear strength that is closer
to the actual shear strength of the material as compared to the conventional two
point loading shear beam test.

Li. Engineered Cementitious Composites (ECC) Tailored ... 19

 A A-A

Shear
panel 210

A
50 50 50
185 170 185

150
540

50
50 150

50

160 25 170 135 50

Flexural
Reinforcement
2P/3 P /3
Flexural
Reinforcement

210

All dimensions are

P /3 2P/3 in millimeters .

50 330 160

Fig. 14: Ohno Shear Beam Geometry and Loading Configuration

The Ohno shear beams were tested in an Instron Model 8000 test system with
a 500 kN capacity loading frame. Tests were run under displacement control at a
loading rate of 0.381 mm/min. Total test time was approximately 10 minutes. The
loads were applied through rollers resting on 25.4 mm wide thin aluminum spreader
plates glued to the specimen. The beams were placed on roller supports.
The shear load versus deflection curve is shown in Fig. 15. Beam deflection
was measured by a LVDT located under the interior load point. After first crack
strength, the ram load continued to increase. The pseudo strain-hardening behavior
of ECC revealed itself in the form of multiple diagonal cracks (Fig. 16) with small
crack widths of less than 0.1 mm even up to ultimate load. In contrast, the FRC
beam failed shortly after first crack load with a single crack opening as the crack
width increased at continuously softening load. It is clear from Fig. 15 that the
ductility of the ECC beam is extensive both pre-peak and post-peak. Indeed Li et al

Li. Engineered Cementitious Composites (ECC) Tailored ... 20

(44) showed that the ductility of this ECC beam is even better than a similar beam
with conventional shear reinforcement in the form of a welded steel wire fabric. In
the FRC and especially in the conventionally shear reinforced beam, extensive
spalling of concrete was observed after first cracking. This does not occur in the
ECC beam, in spite of the large ductility shown in Fig. 15.

200

FRC
ECC
150
Ram Load (kN)

100

50

0
0 1 2 3 4 5
Interior Load Point Deflection (mm)

Fig. 15: Ram Load versus Deflection Curves of ECC and FRC Ohno Shear Beams

The average shear strength in the Ohno shear beams was estimated as the
shear force at the beam center line (which is one-third of the ram load) divided by
the cross-sectional area resisting the shear force. The ECC system failed at a stress
of 5.09 MPa, compared to 3.03 MPa for the FRC. The unique ductility gain in the
ECC beam is reflected by the average shear strain of 2.6% at ultimate load
compared to 0.6% for the FRC.
The ductile shear response in ECC suggests that ECCs can be utilized in
structures where intense shear loading can be expected, such as experienced by
some concrete bridge decks (45), in concrete elements connected by steel anchors,
and where conventional shear reinforcement is desired but prevented from adoption
due to reinforcement congestion or otherwise.

Crack Width Control in R/C Beam

Maalej and Li (46) proposed a new design for reinforced concrete flexural
members for the purpose of improving their durability. The design makes use of
the unique strain-hardening property of ECCs to limit the crack width. The
composite is used as a replacement for the concrete material that surrounds the main
reinforcement in a regular reinforced concrete member. With this design it was
shown that crack widths under service load conditions can be limited to values that
could never be achieved using conventional steel reinforcement and commonly used
concrete. Under these conditions, it was concluded that it would be possible to
prevent the migration of aggressive substances into the concrete or the
reinforcement. Furthermore, accelerated corrosion due to longitudinal crack or

Li. Engineered Cementitious Composites (ECC) Tailored ... 21

spalling will be reduced if not eliminated, and spalling and delamination problems
common to many of today's reinforced concrete structures will be prevented.

Fig. 16: Crack Pattern of Shear Panel, After Peak Load Reached

In the proposed design of the R/C member, a layer of ECC is substituted for
the concrete that surrounds the main flexural reinforcement (Fig. 17). This ECC
layered beam has the same cover thickness as for a control specimen, which has a
regular concrete cover. Two performance requirements are imposed on the ECC
material to serve its intended purpose: (1) the ultimate tensile strain capacity of the
ECC material should be greater than the maximum strain that can be developed in
the outermost fiber at the tensile face of the R/C beam, and (2) the crack width at
the ultimate strain capacity of the ECC material (hereafter referred to as ultimate
crack width) should be less than the maximum crack width allowed in a particular
environment. The first condition ensures that no strain localization will take place
in the ECC layer, and the second condition ensures that the crack opening in the
ECC is maintained below the allowable value. Assuming that at ultimate load, the
strain in the extreme compression fiber of the concrete is equal to 0.003, and that
plane sections remain plane, the strain in the extreme tension fiber of the ECC is
found to be equal to 0.013. Therefore, the ECC that should be selected should have
an ultimate strain capacity at least equal to 0.013. In addition, suppose that the
member is to be exposed in an environment of seawater and seawater spray under
wetting and drying. In this case, according to ACI Committee 224, the crack width
should be limited to 0.15 mm. Therefore, the ultimate crack width of ECC should
be less than 0.15 mm. The 2% Spectra ECC discussed above was found to satisfy

Li. Engineered Cementitious Composites (ECC) Tailored ... 22

the prescribed performance requirements and was selected as the material for the
ECC layer (note that the average ultimate crack width for this material is 0.14 mm).
The beams were tested in an Instron Model 8000 test system with a 500 kN
capacity loading frame. The tests were run under displacement control, and the
total test time was approximately 15 minutes (loading rate approximately .025
mm/sec). Ram load and head displacement information were recorded. Four
LVDTs were used to measure the beam deflection and curvature during loading.

114
16

=5
102
127
= 10
152

25
16
13

(unit = mm)

152

305 305 305

Fig. 17: Geometry of the R/C beam with ECC layer and reinforcement details

Figure 18 shows the test results, in the form of moment curvature and crack
width curvature diagrams, for both beams. As shown in this figure, there is no
significant difference between the moment curvature response of the two beams.
The beam with the ECC layer shows a 10 % higher load and curvature at failure.
The crack width-curvature response of the two beams is, however, significantly
different. Fig. 18 shows that the crack width in the control specimen increases
almost linearly as function of curvature. Before yielding of the reinforcement
(moment   N1-m) the width of the crack at the bottom of the beam is
maintained below 0.20 mm. After yielding, the load starts to increase at a much
slower rate while the crack width continues to increase at the same rate. At peak
load the width of the crack is approximately equal to 1.52 mm. For the ECC
layered beam, the crack width maintains a small value at all times. The crack width
first starts to increase almost linearly as function of curvature, and then actually
decelerates at higher curvatures. Before yielding of the reinforcement (moment 

Li. Engineered Cementitious Composites (ECC) Tailored ... 23

10.7 kN-m) the crack width is maintained below 0.05 mm. At ultimate load the
crack width reaches 0.19 mm. Also the strain measured in the ECC material at the
bottom of the beam was 0.026 which is smaller than the ultimate strain capacity of
the material (0.055).

Control R/C Beam

20 R/C Beam with ECC Layer
1.6

15

Crack Width (mm)

Moment (kN-m)

1.2

10
0.8

5 ACI Interior Exposure Limit 0.4

ACI Exterior Exposure Limit

0 0
0 0.05 0.1 0.15 0.2 0.25
Curvature (1/m)

Fig. 18: Moment and Crack Width-Curvature Diagrams

Fig. 19 shows two series of pictures illustrating the increase of crack width as
function of load for both the control R/C beam (series a) and the R/C beam with the
ECC layer (series b). The load level is indicated on the bottom of each picture.
These series of pictures indicate a significant difference in the cracking behavior of
the two beams. They illustrate that for a given load level, the crack width in the
ECC layered R/C beam is always smaller than that in the control specimen, and that
the rate of crack width increase as function of load is much higher for the latter.
The overall crack patterns of the beams are shown in Fig. 20. For the control
R/C beam, the commonly observed crack pattern (Fig. 20a) with tensile cracks
emanating from the concrete cover was observed. For the ECC layered beam, the
cracks in the concrete material diffused into many fine cracks when they met the
ECC layer (Fig. 20b). This phenomenon appears similar to the crack pattern
observed in Double Cantilever Beam (DCB) fracture specimens (Fig. 13). When a
large crack develops in concrete it is accompanied by a strain concentration at the
location where this crack meets the ECC material. Because of the stress transfer
capability of the reinforcing fibers in the ECC material, stress redistribution occurs
so that localized fracture is delayed. In fact, localized fracture may never develop in
the ECC layer if the maximum strain in the layer is kept below the material ultimate
strain capacity. Consequently, an expanded zone of matrix cracking must develop
in the ECC layer prior to localized fracture.

Li. Engineered Cementitious Composites (ECC) Tailored ... 24

 Tsukamoto (47) determined that fluid flow rate scales with the third power of
crack width in concrete. Based on the above results, one could conclude that the
flow of aggressive substance into the ECC layered R/C member could be
significantly reduced, thus slowing down corrosion rate. Note that this reduction in
crack width is achieved without reducing the concrete cover thickness, considered
to be important in structural durability (48).

(a) (b)
Fig. 19: Variation of Crack Width as Function of Load: (a) Control R/C Beam; (b)
R/C Beam with ECC Layer

The corrosion of a reinforcing bar is accompanied by the production of iron

oxides and hydroxides which occupy a volume larger than the original metal. As a
result, a large pressure is exerted on the surrounding concrete which results in local
radial cracks. These cracks can propagate along the bar, producing longitudinal
cracks (49), spalling of the concrete, and potential loss of strength in the R/C
member. In addition, a large crack may form parallel to the concrete surface at a
plane of bars, resulting in the delamination of the surface, a serious problem in
bridge decks. When the concrete that surrounds the reinforcement is replaced with
ECC material, the longitudinal cracks are likely to be arrested before reaching the
exposed surface, or otherwise be limited in width. In this case the control of the
width of longitudinal cracks can be very effective in reducing the corrosion of the
reinforcement. Moreover, by the use of an ECC, the spalling and delamination of

Li. Engineered Cementitious Composites (ECC) Tailored ... 25

concrete will be eliminated. This is due to the high fracture resistance of the ECC
material associated with a rapid rising R-curve. These potential benefits of ECC in
structural applications remains to be laboratory and field tested.

(a)

(b)

Fig. 20: Crack Pattern (a) Control R/C Beam; (b) R/C Beam with ECC Layer

Energy Absorption In Plastic Hinge Of Beam-Column Connection

In earthquake resistant design, the structural system performance
requirements can be specified in terms of minimum ductility ratio, number of load
cycles, sequence of application of load cycles and permissible reduction in strength
at the end of loading. On the beam-column connection component level, the
following performance are desirable: (i) ductile plastic hinge behavior under high
shear stress, (ii) no congestion of transverse reinforcement for confinement and for
shear, (iii) maintain concrete integrity under load reversals, and (iv) concrete
damage contained within a relatively short hinging zone. These performance are

Li. Engineered Cementitious Composites (ECC) Tailored ... 26

difficult to achieve with ordinary concrete, although some encouraging results have
been obtained with fiber reinforced concrete. Desirable performance of the plastic
hinge is not easy to translate directly into numerical quantities of materials property
requirement. In general, however, it may be expected that the following properties
of the concrete material in the plastic hinge should be advantageous:
(i) high compression strain capacity to avoid loss of integrity by crushing,
(ii) low tensile first cracking strength to initiate damage within the plastic hinge,
(iii) high shear and spall resistance to avoid integrity loss by diagonal fractures, and
(iv) enhanced mechanisms that increases inelastic energy dissipation.
In a recent study (50) the use of a strain-hardening ECC to achieve these
objectives instead of increased shear steel reinforcement was investigated.

10 in.

3"
12.5 in. 3" 10 in.
1"
3"

11 in.

30 in. 8 in. 30 in.

Fig. 21: Schematic of the experimental set-up

Experiment. The sub-assemblage of R/C moment resisting frame selected for

this testing program is shown schematically in Fig. 21. The test specimen
represents two half beams connected to a stub column, in a strong column-weak
beam configuration. The beams are simply supported at their ends to represent mid-
span inflection points, under lateral loading of a framed structure.
Two specimens, one using plain concrete (PC) for the entire specimen and
the other using ECC material in the plastic hinge zone and PC in the rest of the
specimen, were tested. The so called non-ductile or ordinary detailing (Fig. 22) is
used for both specimens to highlight the contribution of the ECC. For comparison,
the seismic detailing is also shown in Fig. 22. Equal tension and compression steel
( = 0.017) is provided in form of two #6 bars each at the top and bottom of the

Li. Engineered Cementitious Composites (ECC) Tailored ... 27

section. Closed shear stirrups (#2 bars) are provided at 4 in. spacing throughout the
span. Shear span of the specimen is 30 in. and effective depth is 8.63 in. (Shear
span to depth ratio = 3.5). The loading history used in this testing program consists
of simple multiple steps of symmetric cycles of increasing displacement amplitude.
The displacement controlled loading sequence used in this test is shown in Fig. 23.

Results and discussions. The load vs. deflection hysteretic behavior is shown
in Fig. 24. For the PC hinge, the displacement ductility factor defined as the ratio
of ultimate deflection (corresponding to a failure load that is about 20% lower than
the maximum load carrying capacity) to yield deflection of about 4.8. For the ECC
hinge, the displacement ductility factor increases to 6.4, with less amount of
pinching and a much reduced rate of stiffness degradation (50). The cracking
pattern (Fig. 25) was distinctly different with more cracking taking place in the
plastic hinge zone with ECC rather than the zone outside as in the case of the PC
control specimen. The damage is mostly in the form of diagonal multiple cracking
in perpendicular direction. Unlike the control specimen which fail in a
predominantly shear diagonal fracture, the ECC specimen fails by a vertical flexural
crack at the interface between ECC plastic hinge zone and the plain concrete at the
column face. No spalling was observed in the ECC hinge, whereas the concrete
cover mostly disintegrated in the control. The cumulative energy over the load
cycles for the two specimens are compared in Fig. 26 which shows that the ECC
hinge absorbs about 2.8 times as much energy as the control. The control specimen
does behave in a manner similar to the ECC hinge specimen in its range of
deflection. However the ECC specimen far out-performs the control specimen in
the deflection regime beyond 1.2".

Li. Engineered Cementitious Composites (ECC) Tailored ... 28

 6.0 in.
5/8 in.

#6

#2 d = 8.63 in.

h = 10 in.
#6
2.5 in.

5/8 in.
5/8 in.

#3
#6 #5
#2 at 4 in. c/c

10 in.
16 in.

30 in. 8 in. 30 in.

Fig. 22: Test configuration and reinforcement layout

RESEARCH NEEDS
Further research in ECCs are needed at both the material and structural
levels. On the material level, additional property characterization and
improvements are needed. These include, for example, the characterization of high
and low cycle fatigue behavior, shrinkage behavior and freeze-thaw durability.
2.0
9 y
1.5

1.0
Head Displacement (in.)

0.5
y
0.0
y
-0.5

-1.0

-1.5
9 y
-2.0
0 100 200 300 400 500
Reading No.

Fig. 23: Experimental data of loading sequence used in the test

Li. Engineered Cementitious Composites (ECC) Tailored ... 29

 (a) (b)
Fig. 24: Load vs. deflection response of (a) specimen with plain concrete plastic
hinge, and (b) specimen with ECC plastic hinge

Research have been conducted in first crack stress and elastic modulus
enhancement by addition of aggregates (30). This involves, however, careful
design of matrix properties since the addition of aggregates leads to an increase in
Jc which can rapidly increase the amount of fibers needed in satisfying the
condition for pseudo strain hardening given in Eq. 6. Hence it is necessary to
balance the various composite property needs by properly adjusting the material
constituents. The delicate balance is greatly aided by the availability of the
micromechanical models.
On a more micro-level, it is necessary to systematically investigate the
tailoring of fiber, interface and matrix, guided by micromechanical principles.
Fiber properties are continually improved, and new fiber types arise on an almost
monthly basis. However, very few commercial fibers are especially geared towards
applications in

Li. Engineered Cementitious Composites (ECC) Tailored ... 30

 cementitious composites. Manufacturers of advanced fibers tend to be more
aware of the special needs in their fibers used in polymer matrix composites (FRP),
ceramic matrix composites (CMC) and metal matrix composites (MMC). Unique
requirements for reinforcement in cement based composites, apart from cost, needs
to be researched and articulated. As an example, high fiber stiffness is much more
important in FRP than in cement based composites. In contrast, high bending
flexibility (translated into high tensile strain capacity) of the fiber is much more
important in cement based composites, especially if the fibers are to be utilized in
random orientations. Equally important are the fiber surface finish and treatment,
both for fresh state processing purpose and for the hardened state mechanical
properties. Fiber surface finish can affect the uniform dispersion of the fibers.
Surface treatment, such as that by plasma, has been shown to effectively enhance
interface bond property (30, 32).

(a)

(b)
Fig. 25: Photograph of final failure, in the plastic hinge zone of

Li. Engineered Cementitious Composites (ECC) Tailored ... 31

 (a) control, and (b) ECC specimens

600
Sp #1 (PC) 548.0

Cum. Energy Absorbed (kip-in)

500 Sp #2 (ECC)

400

300

200 196.0

100

0
0.0 0.1 0.4 0.8 1.2 1.6
Displacement (in.)

Fig. 26: Comparison of cumulative energy absorption vs. deflection

of ECC hinge vs. the control

While micromechanical models are useful in guiding the detail manipulations

of the fiber, matrix and interface properties, to achieve desirable composite
properties, it is necessary to recognize the limitations of micromechanical models as
well. For example, it is well known that the addition of second phase material in a
composite can lead to changes in the microstructure and property of the matrix
constituent (51) and these changes are sensitive to details of processing and the
amount of the second phase additions. Cement based composites in general, and
ECCs in particular, do not escape this phenomenon. Current micromechanical
models typically do not recognize this fact, and straight forward use of these models
can lead to erroneous results. Research into the quantification of the effects of fiber
addition to cement matrix property changes is required.
Research in applications of ECCs are in an embryonic stage, and is badly
needed. Field applications provide valuable feedback into the design of ECCs, both
for fundamental mechanical and physical properties, as well as for properties
needed for a particular application. Although the examples of applications of ECC
in an element under intense shear loading, in a ECC layered R/C beam, and in a
plastic-hinge zone of a beam-column connection described above demonstrate the
value of ECC in structural performance enhancement, there is no question that both
deeper studies of these applications, and broader studies of other ECC applications,
are needed. As an example, the high strain capacity of ECC may suggest that the
material may be very well suited for use as a patch material, wherein substrate
constraint may create substantial imposed tensile strain due to drying shrinkage of
repair material.

Li. Engineered Cementitious Composites (ECC) Tailored ... 32

 Continued research at the material level and structural level should be linked
together, to provide feedback on each other so that progress on the material
properties should be rapidly transferred to the structural performance level, and
progress on structural behavior understanding should be quickly identified with
specific material properties which can be affected by microstructure design. Fiber
composites in general, and ECC in particular, provides ample opportunities to tailor
material microstructure for structural performance in specific applications.
Advances in such performance driven material design will no doubt require
continued and sustained research, but the payoff can prove to be substantial also.

CONCLUDING REMARKS
This article reviews the various types of FRCs in use today, and suggest the
need for developing a new class of FRCs which has the strain-hardening property
but which can be processed with conventional equipment. It is demonstrated that
such a material, termed engineered cementitious composites or ECCs, can be
designed based on micromechanical principles. The result is a moderately low fiber
volume fraction (<2%) composite which shows extensive strain-hardening, with
strain capacity extending to several percent (up to 6% has been demonstrated), with
compressive strength in the typical high strength concrete range (about 68.5 MPa),
with fracture energy several times above typical FRCs (about 27 kJ/m2). Under
bending and shear loads, the ECC beams show extensive ductile behavior both pre-
and post-peak. In addition, the ECC material has indicated notch insensitivity in
double edged uniaxial tensile specimens, suggesting that high reliability can be
achieved in this type of composite. ECC has isotropic mechanical properties.
Further improvement in ECCs, guided by micromechanical principles, are under
investigations in the area of matrix modifications, interface tailoring, and fiber
design. Because of the processing flexibility, the material can be used for pre-cast or
cast-in-place structures, and broad classes of potential applications, such as those
which require structural ductility, energy absorption, and crack width control under
high strain imposition, are identified to be suitable to take advantage of the unique
properties of ECCs. Some additional potential applications include
1. High energy absorption structures/devices:
EQ resistant structures Columns, short span beam, beam-column connections
Seismic retrofits shear walls, dampers
Steel structures joints
Hybrid structures steel/RC connections
2. Structures subject to impact loads:
Pavements durability, reflective cracking
Building core element
Lightweight durable bridge decks
3. Large deformation structures:
Underground structures; conformability to soil deformation, leakage prevention
Concrete pipes
4. Others:

Li. Engineered Cementitious Composites (ECC) Tailored ... 33

 Permanent formwork; replacement of steel jacket of concrete columns.
Extruded products with structural capacity
Concrete cover for durability
Durable repair material
FRP reinforced concrete structure
Radio-active waste immobilization

Li. Engineered Cementitious Composites (ECC) Tailored ... 34

ACKNOWLEDGMENTS
This article is partially based on an invited lecture given at the Engineering
Foundation Conference on Advances in Cement and Concrete held in New
Hampshire, June 1994. Additional material originally served as a contribution to
the ACI/RILEM/ACBM 2nd International Workshop on High Performance Fiber
Reinforced Cementitious Composites held in Michigan, June, 1995, has been
included for completeness. Some materials presented here represent parts of
doctoral theses research carried out by M. Maalej and D.K. Mishra, under the
supervision of the author at the University of Michigan. Support of the Advanced
Civil Engineering Materials Research Laboratory (ACE-MRL) by the Conoco Inc.,
Eternit Company, Kajima Corporation, NATO, Redco Company, Rocla Company,
Shimizu Corporation, U.S. Gypsum Corporation, the U.S. National Science
Foundation, and the U.S. National Research Council, are gratefully acknowledged.

REFERENCES

1. Balaguru, P.N. and Shah, S.P., Fiber Reinforced Cement Composites, McGraw
Hill, 1992.
2. Krenchel, H., Synthetic Fibres for Tough and Durable concrete, in
Developments in Fiber Reinforced Cement and Concrete, RILEM Symposium,
R. N. Swamy, R. L. Wagstaffe and D.R. Oakley eds., 1986, pp. 333-338.
3. Krenchel, H. and Stang, H., Stable Microcracking in Cementitious Materials, in
Proceedings of 2nd Int'l Symp. on Brittle Matrix Composites, Brandt, A.M.,
and I.H. Marshall, 1988, pp. 20-33.
4. Bentur, A. and Mindess, S., Fiber Reinforced Cementitious Composites,
Elsevier Applied Science, 1990.
5. Batson, G., Jenkins, E. and Spatney, R., Steel Fibers As Shear Reinforcement In
Beams, ACI Journal, 69(10) 1972, pp. 640-644.
6. Sharma, A. K., Shear Strength Of Steel Fiber Reinforced Concrete Beams, ACI
Journal Proceedings, 83(4) 1986, pp. 624-628.
7. Swamy, R. N. and Bahia, H.M., The Effectiveness Of Steel Fibers As Shear
Reinforcement, Concrete International, 1985 pp. 35-40.
8. Li, V.C., Ward, R. and Hamza, A.M., Steel And Synthetic Fibers As Shear
Reinforcement, J. Materials, American Concrete Institute, 89(4) 1992, pp. 499-
508.
9. Stang, H. and Aarre, T., Evaluation Of Crack Width In FRC With Conventional
Reinforcement, Cement & Concrete Comp. 14(2) 1992, pp. 143-154.
10. Stang, H., Li, V.C. and Krenchel, H., Design And Structural Applications Of
Stress-Crack Width Relations In Fiber Reinforced Concrete, RILEM J. of
Materials and Structures, 28 1995, pp. 210-219.
11. Stang, H., Personal communication, 1994.
12. Horii, H. and Nanakorn, P., Fracture Mechanics Based Design Of SFRC Tunnel
Lining, in Proceeding of JCI International Workshop on Size Effect in Concrete
Structures, Oct. 31 - Nov. 2, 1993, Sendai, JAPAN, pp. 347-358.

Li. Engineered Cementitious Composites (ECC) Tailored ... 35

13. Lankard, D.R. and Lease, D.H., Highly Reinforced Precast Monolithic
Refractories, Bulletin American Ceramic Society, 61(7) 1982, pp. 728-732.
14. Schneider, B., Development Of SIFCON Through Applications, in "High
Performance Fiber Reinforced Cement Composites", Reinhardt, H.W. and A.E.
Naaman (Ed.), E&FN Spon, 1991, pp. 177-194.
15. Hackman, L.E., Farrell, M.B. and Dunham, O.O., Ultra High Performance
Reinforced Concrete, in Fiber Reinforced Concrete, Development and
Innovations, Eds. J.I. Daniel, and S.P. Shah, Published by ACI, SP-142, 1994,
pp. 235-248.
16. Bentur, A. and Cree, R., Cement Reinforced With Steel Wool, Int. J. Cem.
Comp. & Ltwt. Concr., 9 1987, pp. 217-23.
17. Bache, H.H., Introduction To Compact Reinforced Composite, CBL Reprint No.
15, Aalborg Portland, 1987.
18. Naaman, A.E., Reinhardt, H.W. and Fritz, C., Reinforced Concrete Beams with
a SIFCON Matrix, ACI Structural J., 89 (1) 1992, pp. 79-88.
19. Jeong, S.M., Naaman, A.E. and Tan, K.H., Investigation Of Beams Partially
Prestressed With Carbon Fiber Reinforced Composite Tendons, Proceedings of
the FIP XII Congress, Washington, May 1994, pp. B56-B61.
20. Homrich, J. and Naaman, A.E., Stress-strain Properties of SIFCON in
Compression, In "Fiber Reinforced Concrete, Properties and Applications", ed.
S.P. Shah & G.B. Batson ACI SP-105, ACI, Detroit, 1987, pp. 283-304.
21. vanMier, J.G.M. and Timmers, G., Shear Fracture in SIFCON, in "High
Performance Fiber Reinforced Cement Composites", Reinhardt, H.W. and A.E.
Naaman (Ed.), E&FN Spon, 1991, pp. 518-528.
22. Li, V.C. and Leung, C.K.Y., Steady State And Multiple Cracking Of Short
Random Fiber Composites, ASCE J. of Engineering Mechanics, 118(11) 1992,
pp. 2246-2264.
23. Li, V.C. and Wu, H.C., Conditions For Pseudo Strain-Hardening In Fiber
Reinforced Brittle Matrix Composites, J. Applied Mechanics Review, 45(8)
1992, pp. 390-398.
24. Li, V.C., From Micromechanics To Structural Engineering -- The Design Of
Cementitious Composites For Civil Engineering Applications, JSCE J. of
Struc. Mechanics and Earthquake Engineering, 10(2) 1993, pp. 37-48.
25. Marshall, B., Cox, B.N. and Evans, A.G., The Mechanics Of Matrix Cracking In
Brittle Matrix Fiber Composites, Acta Metall., 33(11) 1985, pp. 2013-2021.
26. Marshall, D. and Cox, B.N., A J-integral Method For Calculating Steady-State
Matrix Cracking Stress In Composites, Mechanics of Materials, 7 1988, pp.
127-133.
27. Li, V.C., Post-Crack Scaling Relations For Fiber Reinforced Cementitious
Composites, ASCE J. of Materials in Civil Engineering, 4(1) 1992, pp. 41-57.
28. Maalej, M., Li, V.C. and Hashida, T., Effect of Fiber Rupture On Tensile
Properties Of Short Fiber Composites, ASCE J. Engineering Mechanics, 121(8)
1995, pp. 903-913.

Li. Engineered Cementitious Composites (ECC) Tailored ... 36

29. Li, V.C. and Obla, K.H., Effect of Fiber Length Variation On Tensile Properties
Of Carbon Fiber Cement Composites, Int'l J. of Composites Engineering, 4 (9)
1994, pp. 947 - 964.
30. Li, V.C., Mishra, D.K. and Wu, H.C., Matrix Design For Pseudo Strain-
Hardening Fiber Reinforced Cementitious Composites, RILEM J. Materials and
Structures, 28 (183) 1995, pp. 586-595.
31. Li, V.C., H.C. Wu, M. Maalej, D.K. Mishra and T. Hashida, Tensile Behavior
of Cement Based Composites with Random Discontinuous Steel Fibers, J.
Amer. Ceramics Soc., 79 (1) 1996, pp. 74-78.
32. Li, V.C., Wu, H.C. and Chan, Y.W., Interfacial Property Tailoring For Pseudo
Strain-Hardening Cementitious Composites, in Advanced Technology on
Design and Fabrication of Composite Materials and Structures, Ed. Carpinteri
and Sih, Kluwer Acad. Publ., Netherlands, 1995, pp. 261-268.
33. Li, V.C., Wang Y. and Backer, S., Effect of Inclining Angle, Bundling, And
Surface Treatment On Synthetic Fiber Pull-Out From A Cement Matrix, J.
Composites, 21(2) 1990, pp. 132-140.
34. Maalej, M., Hashida, T. and Li, V.C., Effect Of Fiber Volume Fraction On The
Off-Crack-Plane Fracture Energy In Strain-Hardening Engineered Cementitious
Composites", 78 (12) 1995, pp. 3369-3375.
35. Maalej, M. and Li, V.C., Flexural/tensile Strength Ratio In Engineered
Cementitious Composites, ASCE J. of Materials in Civil Engineering, 6(4)
1994, pp. 513-528.
36. Li, V.C., Maalej, M. and Lim, Y.M., Fracture and Flexural Behavior In Strain-
Hardening Cementitious Composites, in "Fracture of Brittle, Disorder Materials:
Concrete, Rock and Ceramics", Eds. G. Baker and B.L. Karihaloo, 1995, pp.
101-114.
37. Li, V.C. and Hashida, T., Engineering Ductile Fracture In Brittle Matrix
Composites, J. of Materials Science Letters, 12, 1993, pp. 898-901.
38. Li, V.C., Chan, C.M. and Leung, C.K.Y., Experimental Determination Of The
Tension- Softening Curve In Cementitious Composites, Cement and Concrete
Research, 17(3) 1987, pp. 441-452.
39. Csagoly, P.F., Design of Thin Concrete Deck Slabs By The Ontario Highway
Bridge Design Code, Ontario Ministry of Transportation and Communications,
1979.
40. Fattuhi, N.I. and Huges, B.P., Ductility of Reinforced Concrete Corbels
Containing Either Steel Fibers Or Stirrups, ACI Materials J., 86(6) 1989, pp.
644-651.
41. Eligehausen, R. and Sawade, G., A Fracture Mechanics Based Description Of
The Pull-Out Behavior Of Headed Studs Embedded In Concrete, in "Fracture
Mechanics of Concrete Structures", 1989, pp. 281-299.
42. Lamberson, E.A. and Barker, J.M., Kishwaukee River Bridges, Concrete
International, August, 1981, pp. 93-101.

Li. Engineered Cementitious Composites (ECC) Tailored ... 37

43. Arakawa, T. and Ohno, K., Shear Tests Of Reinforced Concrete Beams By
Special Type Of Loading, Transactions of the Architectural Institute of Japan,
57, pp. 581 - 584. (In Japanese)
44. Li, V.C., Mishra, D.K., Naaman, A.E., Wight, J.K., LaFave, J.M., Wu, H.C. and
Inada, Y., On the Shear Behavior Of Engineered Cementitious Composites, J. of
Advanced Cement Based Materials, 1(3) 1994, pp. 142-149.
45. Mufti, A. A., L.G. Jaeger, B. Bakht and L.D. Wegner, Experimental
Investigation Of Fiber-Reinforced Concrete Deck Slabs Without Internal Steel
Reinforcement, Can. J. Civ. Eng. 20 1993, pp. 398-406.
46. Maalej, M. and Li, V.C., Introduction Of Strain Hardening Engineered
Cementitious Composites In The Design Of Reinforced Concrete Flexural
Members For Improved Durability, J. Structures, American Concrete Institute,
92(2) 1995, pp. 167-176.
47. Tsukamoto, M., Tightness of Fiber Concrete, Darmstadt Concrete, Annual
Journal on Concrete and Concrete Structures, 5, 1990, pp. 215-225.
48. Beeby, A.W., Cracking: What are Crack Width Limits For?, Concrete, 12(7)
1977-1978, pp. 31-33.
49. ACI Committee 224, Control of Cracking In Concrete Structures, ACI Manual
of Concrete Practice Part 3-1991: Use of Concrete in Buildings-Design,
Specifications, and Related Topics, American Concrete Institute, Detroit, 1991.
50. Mishra, D.K. and V.C. Li, Performance of a Ductile Plastic Hinge Designed
With An Engineered Cementitious Composite, UMCEE Report No. 95-06, Feb.,
1995.
51. Cohen, M.D., Goldman, A. and Chan, W.F., The Role Of Silica Fume In Mortar
- Transition Zone Versus Bulk Paste Modification, Cement and Concrete
Research, 24(1) 1994, pp. 95-98.

Li. Engineered Cementitious Composites (ECC) Tailored ... 38