Construction and Building Materials 111 (2016) 408–418

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

Innovative ECC jacketing for retrofitting shear-deficient RC members

Chung-Chan Hung ⇑, Yu-Syuan Chen
Department of Civil Engineering, National Cheng Kung University, No. 1, University Rd, Tainan City 701, Taiwan

h i g h l i g h t s

The ECC jacket improved the cyclic behavior of shear-deficient beams considerably.

Using ECC to replace mortar in ferrocement reduced the crack width by two times.

Cover spalling of regular ferrocement was improved by using ECC to replace mortar.

The ECC jacket with a single layer of bar meshes was the best retrofitting scheme.

a r t i c l e i n f o a b s t r a c t

Article history: Engineered Cementitious Composite (ECC) is distinguished from conventional fiber reinforced concrete
Received 7 September 2015 by its ductile tensile strain-hardening behavior and crack width control ability. This study investigates
Received in revised form 23 December 2015 the performance of ECC jacketing for retrofitting shear-deficient reinforced concrete (RC) members. Six
Accepted 17 February 2016
RC cantilever structural beams are prepared, and five of them are retrofitted with jackets. The experimen-
Available online 21 March 2016
tal parameters involve the properties of the jacket, i.e., (1) ECC or mortar as the matrix, (2) presence or
absence of steel meshes, and (3) welded wire or bar meshes. The performance of the various schemes
Keywords:
of ECC jacketing is evaluated using the test results of the cantilever beams under cyclic loading. In par-
High performance fiber reinforced concrete
Engineered Cementitious Composite
ticular, this study explores whether the appealing properties of ECC shown on the material scale can
Ferrocement translate into better performance of the ECC jacket on the structural scale. Multiple performance mea-
Jacketing sures of the beams are employed, including damage patterns, hysteretic loops, energy dissipation capac-
Retrofitting ities, rebar strain profiles, shear distortions, and failure modes. The test results show that the ECC jacket
Shear failure without steel meshes is able to improve the cyclic behavior of the original shear-deficient beam consid-
erably. The behavior of the retrofitted beam at the performance level of the ultimate limit state can be
further enhanced by reinforcing the ECC jacket with steel meshes. In addition, the multiple performance
measures suggest that the ECC jacket with a single layer of bar meshes be the best retrofitting scheme.
Based on the test results, both the advantages and disadvantages of the ECC jacketing technique are
reported.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction composite, is commonly made with a cement-based mortar rein-

RC jacketing is a popular method for retrofitting RC structural
members. This is mainly because while it can effectively enhance
the mechanical performance of structural members, it also has a
set of other advantages when compared to steel or fiber reinforced
polymer composite jackets, which include high durability, ade-
quate fire and corrosion resistance, simple construction technique,
and wide availability of construction materials [1–6].
Ferrocement jacketing is a special class of the RC jacketing
technique. Ferrocement, which is a thin laminated cement based
⇑ Corresponding author.
E-mail address: cchung@mail.ncku.edu.tw (C.-C. Hung).
forced with closely spaced layers of continuous steel meshes [7].
It is distinguished from regular reinforced concrete by its relatively
small size reinforcement and the absence of large size aggregates
in the matrix. Ferrocement has been used in numerous
applications as a tough and strong protective shell element, such
as in new terrestrial and marine structures and for the repair and
rehabilitation of existing structures [8–11]. Kondraivendham and
Pradhan [12] applied ferrocement as an external confinement to
concrete members under compression. The effect of external
ferrocement on the compressive behavior of the strengthened
members, including the ultimate compressive strength and failure
strain, was investigated. Mourad and Shannag [13] repaired and
strengthened reinforced concrete square columns using ferroce-
ment jackets containing two layers of welded wire meshes. The

http://dx.doi.org/10.1016/j.conbuildmat.2016.02.077
0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.
 C.-C. Hung, Y.-S. Chen / Construction and Building Materials 111 (2016) 408–418
experimental variable was the magnitude of the preloading axial
compression on the column specimens. Kumar et al. [14] studied
the effectiveness of ferrocement jacketing for retrofitting, and the
experimental variable was the angle of the steel wire mesh. The
evaluation was conducted based on the results of four-point bend-
ing tests on the retrofitted beams. Anugeetha and Sheela [15] also
investigated the performance of ferrocement jacketing for retrofit-
ting beams using four-point bending tests, with the experimental
variable being the number of layers of steel meshes. Shah [16] used
ferrocement for strengthening unreinforced masonry columns. The
effects of mesh spacing and water-to-cement ratio of the ferroce-
ment on the compressive property of the strengthened masonry
columns were investigated. Kaish et al. [17] used ferrocement jack-
eting for re-strengthening square RC short columns, with a focus
on the performance of various schemes for enhancing the bond
between the jacket and the original column.
When ferrocement is subjected to external loading, the contribu-
tion of the reinforcement becomes significant only after matrix
cracking. The tensile property of the matrix is therefore a critical
factor affecting the performance of ferrocement. Engineered
Cementitious Composite (ECC) is a special class of high performance
fiber reinforced cement-based composites (HPFRCCs) [18–22]. As
illustrated in Fig. 1 [18,22], it is characterized by the ductile pseudo
strain-hardening behavior under direct tension, with an ultimate
strain capacity often reported to be greater than 2%. It has been
widely shown in material tests that ECC possesses a favorable crack
width control ability that restrains the average crack width of ECC to
less than 60 lm until failure [18].
The objective of the present study is to investigate the perfor-
mance of various ECC jacketing schemes for retrofitting shear-
deficient RC members. The design variables of the jacket include:
(1) mortar or ECC as the matrix, (2) presence or absence of steel
mesh reinforcement, and (3) welded steel bar or wire meshes.
Six cantilever beams are tested under displacement reversals to
assess the feasibility of ECC jacketing. Multiple performance
parameters of the beams are computed for evaluation purpose.
2. ECC materials
The mixed proportions of the ECC material employed in this
study are summarized in Table 1. The components of ECC include
Type I ordinary Portland cement (specific gravity = 3.15), Class F
fly ash (moisture content = 7% and loss of ignition = 2.6%), silica
sand (specific gravity = 2.65, AFS = 75, water absorption = 3%, and
particle sizes ranging between 0.15 and 0.2 mm),
polycarboxylate-based superplasticizer admixture, water, and
polyvinyl alcohol (PVA) fibers. No coarse aggregate is used in
ECC. The water-to-binder ratio is 0.35. A 2% volume fraction of
PVA fibers is used in the ECC. The length and diameter of the
PVA fibers are 12 mm and 39 lm, respectively. The fibers have a
tensile strength of 1600 MPa, a density of 1300 kg/m3, an elastic
modulus of 41 GPa, and a maximum elongation ratio of 6%.
Fig. 1. Tensile stress and strain relationships for HPFRCCs.
409
Table 1
Mixed proportions of ECC (kg/m3).
Type I Portland cement Fly ash Silica sand Water HRWRa PVA fiber
548 658 548 397.5 24.6 28.6
a
High-range water reducers.
The tensile characteristics of the ECC material are determined
using uniaxial tensile testing. Fig. 2 shows the test setup, configura-
tion, and dimensions of the tensile specimen. The deformation of the
specimen is measured using two linear variable displacement trans-
ducers (LVDTs) mounted on the two sides of the specimen, with a
gauge length of 80 mm. The magnitude of the applied tensile force
is measured using a load cell attached to the actuator. The tensile
test is terminated when the applied force drops to 70% of the max-
imum force. The compressive strengths of the used cementitious
materials are obtained via compressive tests of standard cylinder
specimens (100 mm  200 mm). Both the tensile and compressive
tests are carried out using displacement control with a loading rate
of 0.5 mm/min to simulate the quasi-static loading condition.
3. Experimental program
Six identical shear-deficient RC structural members are prepared. The design
details of the prototype specimen are shown in Fig. 3. The specimen is composed
of a beam element connected to an RC stub. The cross-sectional dimensions of
the beam element are 250 mm in width, 350 mm in height, and 1200 mm in length.
The beam element is intended to represent a beam cantilevered from the RC stub.
The test setup for the specimens is shown in Fig. 4. In order to sufficiently restrain
the RC stub in all degrees of freedom, the stub is clamped by a top steel plate and a
precast RC foundation through ten steel rods. The steel rods are fixed to the strong
floor with a total prestressed force of 1250 kN. Vertical displacement reversals are
imposed on the free-end of the beam element.
3.1. Specimen designs and test setup
The reinforcement details of the prototype shear-deficient beam are shown in
Fig. 3. The beam specimen is flexurally reinforced with 4-#8(D25) steel bars. In
the transverse direction, #3(D10) steel stirrups with a loose spacing of 300 mm
are used. The names and details of the six beam specimens are summarized in
Table 2.
B-CONT, which is employed as the control specimen, is the original specimen
without retrofitting. The five other specimens are retrofitted using U-shape, i.e.,
three-sided, jacketing with a uniform thickness of 40 mm, as shown in Fig. 3c.
Before the beams are retrofitted, the concrete cover is chipped, and the surface is
roughened to increase the bond strength between the original and the new parts.
The experimental variables of the five retrofitted specimens are associated with
the materials of the jacket, i.e., (1) mortar or ECC as the matrix, (2) presence or
absence of steel reinforcement, and (3) welded steel bar or wire meshes as the rein-
forcement. Both the steel bar mesh and the steel wire mesh have square mesh
openings. The steel bar mesh has a bar diameter of 6 mm and a bar spacing of
Fig. 2. Uniaxial tensile testing on ECC materials.
410 C.-C. Hung, Y.-S. Chen / Construction and Building Materials 111 (2016) 408–418
Fig. 3. Details of the specimen design (unit: mm).
100 mm, whereas the steel wire mesh has a diameter of 1.2 mm and a wire spacing
of 25 mm. For the jackets with steel reinforcement, the amount of reinforcing steel
meshes is designed to be the same, and equal to 1.8% volume fraction of the jacket.
This leads to six layers of wire meshes or a single layer of bar meshes. Fig. 5 shows
the installation of the two types of steel meshes in the jacketing before placement
of the matrix. The nomenclature of the specimens is based on B – ‘‘matrix type” –
‘‘number of layers of steel meshes”. For the matrix type, the letter ‘‘M” or ‘‘E”
denotes that the matrix of the jacket is pure mortar or ECC. For example, B-E-6
denotes a shear-deficient beam retrofitted by an ECC jacket reinforced with six lay-
ers of steel wire meshes; and B-E denotes the beam retrofitted by an ECC jacket that
has no steel mesh reinforcement.
The material properties of the concrete, mortar, and ECC materials, as measured
on the test days, are summarized in Table 2. The uniaxial tensile stress–strain
curves of the mortar and ECC materials are shown in Fig. 6a. It can be seen that
all ECC materials exhibit pronounced pseudo strain-hardening behavior with an
ultimate failure strain above 2%. It can also be observed during the material tests
that all the ECC tensile specimens have dense arrays of narrow cracks. Fig. 6b shows
a representative multiple cracking pattern of the ECC tensile specimen. The
mechanical properties of the steel reinforcements are summarized in Table 3. The
properties of the steel rebar are obtained using direct tensile tests in the laboratory.
The properties of the steel meshes are provided by the manufacturer.
3.2. Test instrumentation and load protocol
The vertical displacement reversals imposed on the beam are applied using a
hydraulic actuator with a loading capacity of 2000 kN. The center line of the
hydraulic actuator is located at 1000 mm from the face of the RC stub. The magni-
tude of the imposed load is measured using a load cell connected to the actuator.
Fig. 7 presents the drift ratio history of the displacement reversals. Three cycles
at each target drift are applied. The drift ratio is calculated as the ratio of the deflec-
tion at the free end of the beam to the shear span. The deflection is measured using
two string potentiometers, which are attached to the two sides of the beam and
aligned to the center line of the hydraulic actuator, as shown in Fig. 4a.
The deflection measured by the string potentiometers is corrected for the pos-
sible rotation of the fixed RC stub, which is measured using a rotational gauge
installed on the top of the steel clamping plate. The corrected deflection of the beam
is calculated as D ¼ Da  Db , where Da is the average of the values measured by the


two string potentiometers; and Db ¼ L þ 12 H h is the deflection of the beam due to
the rotation of the clamped RC stub. L is the span length of the beam, H is the depth
of the RC stub, and h is the measurement of the rotation gauge. Although the com-
puted Db is very minor compared to Da , it is still taken into account to enhance the
accuracy of the test.
In order to monitor the distortion and rotation in the critical region of the
beams, three measuring blocks consisting of eight Demec points are deployed on
the beam, as shown in Fig. 4b. The lengths between the neighboring Demec points
are measured using a Demec gauge at the 1st cycle of the target displacement. The
shear deformation of each measuring block, l, can be calculated using Eq. (1)
according to [23]:
ðL21  L22 Þ  ðD21  D22 Þ
u¼ ð1Þ
4H
The notations in Eq. (1) are illustrated in Fig. 8. The rotation at the beam base is
also monitored using the measurements of two LVDTs mounted on the beam, as
shown in Fig. 4a. Moreover, a dense array of strain gauges is attached to the longi-
tudinal and transverse reinforcements to monitor the reinforcement strain
throughout the test.
4. Experimental results
Extensive data are obtained from the cyclic loading tests. The
performance of the various retrofitting schemes is evaluated care-
fully using the damage pattern, load–deflection relationship, crack
width, stiffness, energy dissipation capacity, beam rotation and
shear distortion, and reinforcement strain.
 C.-C. Hung, Y.-S. Chen / Construction and Building Materials 111 (2016) 408–418 411

Fig. 4. Details of the test setup for cantilever beams.

Table 2
Details of the cantilever beam specimens.

Specimen Jacket Original Properties of ECC/mortar in the jacket Properties of beam

beam
Matrix Steel Concrete Compressive Tensile Peak Ultimate Maximum Ultimate Enhanced Initial Failure
reinforcement compressive strength strength tensile tensile drift (%) strength strength stiffness mode
strength (MPa) (MPa) strain strain (kN) (%) (kN/m)
(MPa)
B-CONT n/a n/a 30.3 n/a n/a n/a n/a 3 130.3 (+) n/a 10.4 (+) Flexural-
130.8 () n/a 11.3 () shear
B-E ECC None 30.5 63.2 2.0 1.10 2.82 5 143.8 (+) 10.4 (+) 13.1 (+) Flexural-
142.8 () 9.2 () 14.3 () shear
B-M-6 Mortar 6 layers of 31.5 28.1 1.0 0.06 0.06 7 151.3 (+) 16.1 (+) 15.9 (+) Flexure
wire meshes 138.8 () 6.1 () 15.0 ()
B-M-1 Mortar 1 layer of bar 31.3 62.0 1.7 0.08 0.08 7 176.2 (+) 32.3 (+) 15.3 (+) Flexure
meshes 165.2 () 26.3 () 14.9 ()
B-E-6 ECC 6 layers of 30.6 49.6 1.2 0.92 2.36 7 158.3 (+) 21.5 (+) 13.9 (+) Flexural-
wire meshes 160.3 () 22.5 () 13.0 () shear
B-E-1 ECC 1 layer of bar 31.1 53.4 1.4 2.67 2.20 8 167.2 (+) 28.3 (+) 15.9 (+) Flexure
meshes 164.7 () 26.0 () 16.8 ()

4.1. Damage patterns
Fig. 9 shows the final damage patterns of the beam specimens.
The maximum drifts at which the beams reach are summarized in
Table 2.
4.1.1. B-CONT
For the control specimen B-CONT, the failure pattern is domi-
nated by the significant shear cracks at 3% drift. In addition, the
presence of the substantially extended longitudinal cracks shown
in Fig. 9b implies that the longitudinal steel rebar in B-CONT suf-
fers bond failure. While a 3% drift demand causes the test of B-
CONT to be terminated due to significant damage, the same drift
demand only results in minor concrete spalling and cracks in all
the retrofitted beams, as will be shown later.
4.1.2. B-E
In B-E, the flexural and shear cracks start to localize at 2% drift.
When the drift demand is increased to 5%, considerable concrete
spalling and crushing occur on the bottom side of the beam, as
shown in Fig. 9f. The test is consequently terminated due to sub-
stantial reduction in the strength of B-E. Although the drift capac-
ity of the original beam can be moderately enhanced through the
ECC jacketing that contains no steel reinforcement, it is noted that
the multiple cracking pattern that ECC exhibits on the material
scale in Fig. 6b is absent in B-E.
412 C.-C. Hung, Y.-S. Chen / Construction and Building Materials 111 (2016) 408–418

Fig. 5. Installation of steel meshes before jacket casting.

Fig. 7. Reversed cyclic drift pattern.

Fig. 6. Tensile behavior of the ECC materials.

4.1.3. B-M-6 and B-M-1

For B-M-6, only multiple narrow flexural and diagonal cracks
are observed prior to 3% drift. At 3% drift, the steel mesh at the
fixed-end of the beam partially fractures, which subsequently
causes an obvious localized flexural crack in B-M-6. When the drift
demand is increased to 6%, the localized flexural crack becomes
more significant, and the ferrocement jacket also suffers consider-
able spalling. The test of B-M-6 is terminated at 7% due to signifi-
cant concrete crushing and spalling in the concrete core, as shown Fig. 8. Calculation of the shear distortion.
in Fig. 9i, which reduces the effective beam depth and causes sub-
stantial degradation in the beam strength. The failure progress of 4.1.4. B-E-6 and B-E-1
B-M-1 is similar to that of B-M-6, except that the spalling of the Similar to B-M-6, the only damage in B-E-6 prior to 3% drift is
jacket in B-M-1 is less significant than that in B-M-6. the dense array of narrow flexural and shear cracks. At 3% drift, a

Table 3
Tensile properties of the reinforcing steel.

Type Size/diameter (mm) Mesh opening size (mm) Yielding strength (MPa) Ultimate strength (MPa) Elongation (%)
Wire mesh 1.2 25  25 293 346 10
Bar mesh 6 100  100 547 562 9.4
Rebar #3(D10) – 457 675 19
Rebar #4(D13) 445 651 21
Rebar #8(D25) – 438 639 21
 C.-C. Hung, Y.-S. Chen / Construction and Building Materials 111 (2016) 408–418 413

Fig. 9. Damage patterns of the specimens at the end of tests.

flexural crack at the fixed-end of the beam starts to localize. As the
drift demand is increased to 6%, a localized diagonal shear crack
suddenly occurs. At the 7% drift demand, considerable concrete
spalling occurs on the bottom side of the beam, as shown in
Fig. 9o, which causes the beam strength to drop significantly, and
the test is terminated. The damage progress of B-E-1 is similar to
that of B-E-6 prior to 6% drift, except that no localized diagonal
crack is observed in B-E-1. When B-E-1 is at 6% drift, minor con-
crete spalling takes place on the bottom side of the beam. The test
of B-E-1 is terminated at 8% drift, due to considerable strength
degradation, as a result of the combined damage of greatly local-
ized flexural cracking and concrete crushing and spalling.
4.1.5. Discussion of damage patterns
It is noted that although B-E-1 and B-E-6 are retrofitted with the
ECC jackets that contain the same amount of steel reinforcement,
B-E-1 has less shear damage than B-E-6. In order to examine the
probable cause of this, the B-E-6 specimen was broken open with
hammers after the experimental test. This revealed that the ECC
was not adequately cast within the region filled with wire meshes,
as can be seen in Fig. 10. This is primarily because casting fresh ECC
in the mold packed with six closely spaced layers of wire meshes is
substantially more difficult than is the case with a single layer of
welded bar mesh. Consequently, although the fresh ECC exhibits
adequate workability in the slump flow test (650 mm in diameter),
the placement quality and the integrity of the ECC jacket reinforced
with wire meshes are both reduced.
Comparisons of the damage patterns of B-M-1, B-M-6, B-E-1,
and B-E-6 indicate that using ECC to replace the mortar in B-M-1
and B-M-6 can effectively restrain concrete spalling and crushing.
Moreover, a comparison of B-E and other beams retrofitted with
the steel-reinforced jackets suggests that the addition of bar or
414 C.-C. Hung, Y.-S. Chen / Construction and Building Materials 111 (2016) 408–418

Table 4
Maximum crack widths at different deformation demands.

Drift (%) Maximum crack width (mm)

B-CONT B-E B-M-6 B-M-1 B-E-6 B-E-1
0.5 0.35 0.1 0.25 0.2 0.1 0.1
0.75 0.7 0.4 0.5 0.4 0.15 0.2
1 2.5 0.8 1.5 1 0.25 0.35
2 10 2.5 3.5 2 0.3 0.5
3 38 7 9.4 6 2 3
4 – 24 15 10 3.5 5.5
5 – 25 19 18 4.5 10
6 – – 20 21 8.5 17
7 – – 23 22 20 21
8 – – – – – 24

4.2. Load–drift response

Fig. 10. Deficiency in the ECC placement in B-E-6.
wire meshes in the jacket is beneficial to promoting closely dis-
tributed cracks over the beams. This thus restrains the widening
of diagonal shear cracks and prevents premature shear failure.
Fig. 11 presents the load–drift relationships of the cantilever
beams under displacement reversals. It can be seen in Fig. 11a that
the strength of the control specimen B-CONT drops sharply as soon
as the maximum strength is reached, due to the brittle shear fail-
ure. In contrast to B-CONT, all the retrofitted beams exhibit a stable
hysteretic response until failure, even for B-E, which has an ECC
jacket without steel mesh reinforcement. Overall, the beam

Fig. 11. Load–drift responses.

 C.-C. Hung, Y.-S. Chen / Construction and Building Materials 111 (2016) 408–418 415

Fig. 12. Stiffnesses of the beams under cyclic loading.

Fig. 13. Energy dissipation capacities of the beams.

strength can be enhanced by 10%, 20%, and 30% by the ECC jacket
with no steel reinforcement, the wire mesh reinforced jacket, and
the bar mesh reinforced jacket, respectively. Although the ECC
jacket with no steel reinforcement is shown to be an effective ret-
rofitting scheme, its performance is still second to that of other
steel reinforced jackets with regard to improving the strength,
deformation capacity, and strength retention ability of the original
beam. In particular, the addition of steel meshes in the jacket is
able to substantially enhance the drift capacity of the original
beam by more than two times.
Among all the retrofitting schemes, B-E-1 is the most effective
in improving the cyclic behavior of the original beam, which
includes the deformation capacity, strength retention ability, and
pinching behavior. In particular, the strength and drift capabilities
of the original beam are enhanced by 28% and 167%, respectively.
This is because the confining and bridging effects of PVA fibers in
ECC delay concrete spalling and crushing, thus effectively sustain-
ing the intactness of the beam section under displacement rever-
sals. In addition, the reinforcement of a single layer of steel bar
mesh in the jacket allows substantially easier casting work than
is the case with six layers of steel wire meshes, thus ensuring the
integrity of the matrix in the jacket and the full bond between
the steel mesh and matrix.
4.3. Crack width
The relationships between the maximum crack width and the
drift demand for the beams are summarized in Table 4. It can be
seen that all the retrofitted beams have substantially smaller crack
widths than the original beam at all stages of loading, implying
that the retrofitting schemes are able to improve the durability
of the original beam. It is worth noting that B-E, which is retrofitted
with an ECC jacket without steel mesh reinforcement, shows a bet-
ter crack width control ability than B-M-1 and B-M-6 prior to 1%
drift. It also performs as well as B-M-1 and B-M-6 at 2% and 3% drift
ratios. This implies that while B-E has a greatly simplified retrofit-
ting procedure compared to the others, because its jacket has no

Fig. 14. Shear distortion profiles in the plastic hinge regions. Fig. 15. Rotation demand profiles in the plastic hinge regions.
416 C.-C. Hung, Y.-S. Chen / Construction and Building Materials 111 (2016) 408–418

steel mesh reinforcement, it is still capable of favorable crack

width control ability in the intermediate damage limit state, which
is comparable to the beams retrofitted with mortar jackets rein-
forced with steel meshes. After 3% drift, the bridging effect of
PVA fibers in B-E significantly degrades due to pull out and fracture
failures of fibers. Consequently, the cracks in B-E widen more
rapidly than those in B-M-1 and B-M-6.
The inclusion of steel meshes in the ECC jacket (i.e., B-E-1 and
B-E-6) substantially improves the crack width control capability,
especially in the ultimate damage limit state. The resulting crack
width is narrower than that in the regular mortar-based ferroce-
ment (i.e., B-M-1 and B-M-6) by more than two times on average
at all stages of loading. In particular, B-E-6 has a crack width about
40% less than that in B-E-1 until failure, implying that the ECC
jacket that is reinforced with a smaller sized steel meshes has a
better crack width control ability.

4.4. Stiffness

The initial stiffnesses of the beams are summarized in Table 2. It

is worth mentioning that the definition of initial stiffness varies
widely in the literature for tested structural members under cyclic
loading. In this study, it is calculated as the ratio of 60% of the max-
imum beam strength to the corresponding deflection, because at
such a strength demand, all steel reinforcement remains elastic
based on the measurements of strain gauges, and no critical dam-
age is observed on all specimens. It can be seen in Table 2 that the
stiffness of the original beam can be effectively enhanced by all the
retrofitting schemes by more than 20% in both positive and nega-
tive loading directions. In particular, the initial stiffness is
increased most substantially via the B-E-1 retrofitting scheme by
about 50%.
Fig. 12 presents the stiffness versus drift demand relationships
for the tested beams under cyclic loading. As illustrated in Fig. 12a,
the stiffness at a specified target drift is calculated as the slope
between the force–drift values at the maximum and minimum tar-
get drifts. Each stiffness value is calculated at the 3rd cycle to the
specified target drift. Fig. 12b shows that all the retrofitting
schemes enhance the stiffness of the original beam at all stages
of loading in an approximately equal manner. It is worth mention-
ing that although the stiffness of B-E degrades to be slightly less Fig. 16. Strain profiles for longitudinal reinforcement.
than that of other retrofitted beams at 5% drift, due to the lack of
steel mesh reinforcement in the jacket, B-E performs as well as
other retrofitted beams prior to 4% drift. Overall, B-E-1 shows the that the inclusion of fibers in a bar mesh reinforced jacket
largest stiffness among all the retrofitting schemes throughout enhances the energy dissipation capacity of the retrofitted beam
the test. by more than 10% after 2% drift. This is because the PVA fibers in
ECC help prevent early localized damage, such as widening cracks
4.5. Energy dissipation capacity and concrete spalling and crushing. However, the addition of PVA
fibers reduces the energy dissipation capacity in the retrofitting
The performances of the various retrofitting schemes are also cases with six layers of steel wire meshes, i.e., B-E-6. This is
evaluated by comparing the energy dissipation capacities of the because the dense steel meshes in B-E-6 increase the difficulty of
beams. As illustrated in Fig. 13a, the energy dissipated within a ECC placement, thus impairing the integration of ECC and the rein-
cycle of a specified target drift is computed as the enclosed area forcing steel mesh as a composite jacket. At the end of the test, B-E-
of the hysteretic loop at the 3rd cycle to the target drift divided 1 and B-M-6 have the greatest energy dissipation capacities among
by the rectangular area enclosing the loop. Fig. 13b shows the all the retrofitted beams.
trends of the cumulative energy dissipation capacity with the
increasing beam drift response for the tested specimens. It can
be seen that all the retrofitting schemes are able to enhance the
energy dissipation capacity of the shear-deficient RC beam under 4.6. Shear distortion and rotation profiles
cyclic loading. In particular, the ECC jacket that does not have steel
meshes can also lead to a considerable improvement in the energy Fig. 14 presents the shear deformation profiles in the plastic
dissipation capacity, about two times greater than that of the orig- hinge region of the beams, and the value shown in parentheses is
inal beam. the failure drift. The results imply that the diagonal shear cracks
The addition of fibers in the steel-reinforced jacket appears to in B-CONT cause the most significant shear deformation. While
have opposite effects on the energy dissipation capacities for all the retrofitting schemes are shown to be capable of reducing
B-E-1 and B-E-6. In a comparison of B-E-1 and B-M-1, it is found the shear deformation demand in B-CONT, B-E-1 appears to be
 C.-C. Hung, Y.-S. Chen / Construction and Building Materials 111 (2016) 408–418
the most effective scheme, reducing the shear distortion demand
by about 50%.
Fig. 15 shows the beam rotation demand profiles in the plastic
hinge region. The rotation demands of all beams are concentrated
at the fixed end of the beam. This can be attributed to the presence
of the localized flexural crack at the beam base. Compared to the
retrofitted beams, the rotation demand is more uniformly
distributed in B-CONT. This is because the failure behavior of
B-CONT is dominated by shear, and the flexural cracking in
B-CONT is less significant than that seen with the other beams.
4.7. Steel reinforcement strains
Fig. 16 presents the strain profiles in the longitudinal steel rebar
at 1%, 3%, and 5% drifts. Because the data measured by the strain
gauges beyond 5% drift gradually become invalid, they are not
reported herein. Fig. 16a shows that the longitudinal steel bars in
all beams at 1% drift just enter or are about to enter the inelastic
stage. At 3% drift, the longitudinal steel bars in all beams exhibit
inelastic behavior. The results also show that all the retrofitting
Fig. 17. Strain profiles for transverse reinforcement.
417
schemes are able to reduce the strain demand of 0.0055 in the orig-
inal beam B-CONT to a value less than 0.004 at 3% drift.
The strain profiles of the transverse steel reinforcement at 1%,
3%, and 5% drifts are shown in Fig. 17. The data shown in
Fig. 17a imply that all the stirrups behave elastically when the drift
demand is 1%. At 3% drift, the transverse reinforcement in B-CONT
exhibits substantial inelastic behavior, whereas all the retrofitted
beams have a strain value less than the yielding strain of stirrups.
The retrofitting schemes of the ECC jackets reinforced with steel
meshes (i.e., B-E-1 and B-E-6) perform most satisfactorily in reduc-
ing the strain demand for the transverse reinforcement, about
three times less. In particular, B-E-1 performs more effectively than
B-E-6 at 5% drift, with a 15% smaller strain. This is because the lar-
ger mesh size used in B-E-1 makes the ECC placement easier, and
thus better integrates ECC and steel meshes as an intact composite
to resist the shear force. It is also interesting to note that the ECC
jacketing without steel (B-E) and the mortar-based ferrocement
jacketing (B-M-1 and B-M-6) perform similarly well in restraining
the strain demands for both the longitudinal and transverse steel
reinforcement.
5. Conclusions
The work presented herein investigated the performance of
innovative ECC jacketing for retrofitting shear-deficient RC mem-
bers. All the retrofitting schemes of ECC jackets were able to signif-
icantly improve the cyclic behavior of the original RC member, in
terms of the strength, stiffness, ductility, energy dissipation capac-
ity, shear distortion demand, and strain demand for the steel rebar.
The main conclusions of this study are as follows.
1. The ECC jacket with no mesh reinforcement not only had a sim-
plified construction process, but also outperformed the regular
mortar-based ferrocement jacket in arresting cracks at the level
of the intermediate damage limit state. It also considerably
increased the energy dissipation capacity of the original mem-
ber by about two times and performed as well as the other ret-
rofitting schemes in enhancing the stiffness of the member
prior to 4% drift. Moreover, it was as effective as the regular fer-
rocement jacket in restraining the strain demand for the longi-
tudinal and transverse steel reinforcement embedded in the
original structural member. Nevertheless, when the ECC jacket
was not reinforced with steel meshes, the bridging effects of
ECC shown on the material scale did not efficiently translate
into enhanced retrofitting performance with regard to improv-
ing the greatly localized shear crack pattern.
2. Although the mortar-based ferrocement jacket was able to
enhance the mechanical performance of the original beam,
the matrix cover of the jacket spalled off significantly at 6% drift.
This deficiency was satisfactorily improved by using ECC to
replace the mortar in ferrocement. Additionally, using ECC to
replace the mortar in ferrocement reduced the crack width by
an average of more than two times at all stages of loading.
While the wire mesh reinforced ECC jacket performed better
than the bar mesh reinforced ECC jacket in restraining the
growth of crack width by about 40% until failure, the dense lay-
ers of wire meshes that were required could impair the con-
struction quality, and ultimately adversely affect the
mechanical performance of the retrofitted beams.
3. The inclusion of steel meshes in the ECC jacket was a critical
factor for the retrofitted beam to behave satisfactorily at the
performance level of the ultimate damage limit state. Overall,
the ECC jacket reinforced with a single layer of steel bar meshes
showed the best retrofitting performance. Compared to the
results before retrofitting, it reduced the transverse steel strain
418 C.-C. Hung, Y.-S. Chen / Construction and Building Materials 111 (2016) 408–418
by about three times and shear distortion demand by more than
50%. It also substantially enhanced the strength and drift capac-
ities by 28% and 167%, respectively. Moreover, it leaded to a 50%
increase in the beam initial stiffness and a fivefold increase in
the energy dissipation capacity, as compared to the original
beam.
Acknowledgements
The research described herein was sponsored in part by the
Ministry of Science and Technology under Grant No. 103-2221-E-
006-268. The opinions, findings, and conclusions expressed in this
paper are those of the authors, and do not necessarily reflect those
of the sponsor.
References
[1] F. Altun, An experimental study of the jacketed reinforced concrete beams
under bending, Constr. Build. Mater. 18 (2004) 611–618.
[2] G. Martinola, A. Meda, G.A. Plizzari, Z. Rinaldi, Strengthening and repairing of
RC beams with fiber reinforced concrete, Cem. Concr. Compos. 32 (2010) 731–
739.
[3] R.H.F. Souza, J. Appleton, Flexural behavior of strengthened reinforced concrete
beams, Mater. Struct. 30 (1997) 154–159.
[4] Y.G. Diab, Strengthening of RC beams by using sprayed concrete: experimental
approach, Eng. Struct. 20 (1997) 631–643.
[5] K.G. Vandoros, S.E. Dritsos, Concrete jacket construction detail effectiveness
when strengthening R C columns, Constr. Build. Mater. 22 (2008) 264–276.
[6] A.D.G. Tsonos, Performance enhancement of R/C building columns and beam-
column joints through shotcrete jacketing, Eng. Struct. 32 (2010) 726–740.
[7] ACI Committee 549, Guide to Design and Construction of Externally Bonded
FRCM Systems for Repair and Strengthening Concrete and Masonry Structures,
American Concrete Institute, 549.4R-13, Farmington Hills, MI, 2013.
[8] A.E. Naaman, Ferrocement and Laminated Cementitious Composites, 3rd ed.,
Techno Press 3000, Ann Arbor, Michigan, USA, 2000.
[9] M.J. Shannag, T.B. Ziyyad, Flexural response of ferrocement with fibrous
cementitious matrices, Constr. Build. Mater. 21 (2007) 1198–1205.
[10] S.D. Shri, R. Thenmozhi, An experimental investigation on the flexural behavior
of scc ferrocement slabs incorporating fibers, Int. J. Eng. Sci. Technol. 5 (4)
(2012) 2146–2158.
[11] P. Paramasivam, R. Sri Ravindrarajah, Effect of arrangements of reinforcements
on mechanical properties of ferrocement, ACI Struct. J. 85 (1) (1998) 3–11.
[12] B. Kondraivendham, B. Pradhan, Effect of ferrocement confinement on
behavior of concrete, Constr. Build. Mater. 23 (2009) 1218–1222.
[13] S.M. Mourad, M.J. Shannag, Repair and strengthening of reinforced concrete
square columns using ferrocement jackets, Cem. Concr. Compos. 34 (2012)
288–294.
[14] M. Kumar, P.P. Bansal, S.K. Kaushik, Effect of wire mesh orientation on strength
of beams retrofitted using ferrocement jackets, Int. J. Eng. 2 (2008) 8–19.
[15] B. Anugeetha, S. Sheela, Study on the performance of reinforced concrete
beams retrofitted using ferrocement and GFRP, Proceedings of National
Conference on Technological Trends, 2009, pp. 143–147.
[16] A.A. Shah, Application of ferrocement in strengthening of unreinforced
masonry columns, Int. J. Geol. 5 (2011) 21–27.
[17] A.B.M.A. Kaish, M.R. Alam, M. Jamil, Improved ferrocement jacketing for
restrengthening of square RC short column, Constr. Build. Mater. 36 (2012)
228–237.
[18] V.C. Li, Tailoring ECC for special attributes: a review, Int. J. Concr. Struct. Mater.
6 (2012) 135–144.
[19] G. Fischer, V.C. Li, Effects of matrix ductility on the deformation behavior of
steel-reinforced ECC flexural members under reversed cyclic loading
conditions, ACI Struct. J. 99 (6) (2002) 781–790.
[20] A.E. Naaman, H.W. Reinhardt, Characterization of high performance fiber
reinforced cement composites, in: A.E. Naaman, H.W. Reinhardt (Eds.),
Proceedings of the 2nd International Workshop on High Performance Fiber
Reinforced Cement Composites: HPFRCC 2, RILEM, 1996, pp. 1–24.
[21] C.-C. Hung, S. El-Tawil, Seismic behavior of a coupled wall system with HPFRC
materials in critical regions, ASCE J. Struct. Eng. 137 (12) (2011) 1499–1507.
[22] C.-C. Hung, Y.-F. Su, K.-H. Yu, Modeling the shear hysteretic response for high
performance fiber reinforced cementitious composites, Constr. Build. Mater.
41 (2013) 37–48.
[23] G.J. Parra-Montesinos, P. Chompreda, Deformation capacity and shear strength
of fiber-reinforced cement composite flexural members subjected to
displacement reversals, ASCE J. Struct. Eng. 133 (2007) 421–431.