Materials and Structures (2008) 41:287–299

DOI 10.1617/s11527-007-9238-9

O R I G I N A L A RT I C L E

Repair of heat-damaged RC shallow beams using advanced

composites
Rami H. Haddad Æ M. Jamal Shannag Æ
Alaa Moh’d

Received: 10 April 2006 / Accepted: 21 February 2007 / Published online: 22 March 2007

RILEM 2007

Abstract Several repair techniques for restoring a typical flexural failure with very fine and well-
the structural capacity of heat-damaged high- distributed hairline cracks in the constant mo-
strength reinforced concrete shallow beams using ment region.
advanced composites are proposed. A series of 16
under-reinforced concrete hidden beams were
cast, heated at 600C for 3 h, repaired, and then
tested under four point-loading. Tests were con- Keywords Repair
Heat-damaged beams

ducted to study the effectiveness of externally Fiber
Concrete
applied composite materials on increasing the
flexural capacity of beams. The composites used 1 Introduction
include high strength fiber reinforced concrete
jackets; ferrocement laminates; and high-strength Use of wide or shallow reinforced concrete beams
fiber glass sheets. The beams repaired with steel is a common construction practice mainly for
and high performance polypropylene fiber rein- architectural reasons. These beams are generally
forced concrete jackets regained up to 108 and hidden within the floor, solid or ribbed slabs
99% of the control beams’ ultimate load capacity, construction systems. In high rise buildings, to
with a corresponding increase in stiffness of up to reduce the construction depth of each floor and
104 and 98%, respectively. The beams repaired hence the overall building height, wide and
with fiber glass sheets and ferrocement meshes shallow reinforced beams (i.e. width > depth) of
regained up to 126 and 99% of the control beams’ high strength concrete (HSC) are commonly
ultimate load capacity, with a corresponding used. The use of HSC offers significant economic,
increase in stiffness of up to 160 and 156%, architectural and durability advantages over nor-
respectively. Most of the beams repaired showed mal strength concrete (NSC). However, it is well
stipulated that HSC has certain shortcomings
such as high brittleness and poor fire resistance.
R. H. Haddad
A. Moh’d
Department of Civil Engineering, Jordan University When exposed to fire, reinforced HSC structural
of Science & Technology, Irbid, Jordan elements undergoes significant damage due to
loss of their structural capacity and corresponding
M. J. Shannag (&) increase in deflection; Such loss due to fire are
Department of Civil Engineering, King Saud
University, P.O. Box 800, Riyadh 11421, Saudi Arabia primarily because of the decrease in strength and
e-mail: [email protected] bond of embedded steel [1, 2]. This is referred to:
288 Materials and Structures (2008) 41:287–299

the decomposition of HSC, and the increase in its The main objective of the present investigation
porosity as well as the formation of thermal is to propose repair techniques for restoring the
cracks [3]. With relatively low water/binder ratios structural capacity of heat-damaged reinforced
and the use of pozzolanic additives, HSC becomes high strength concrete shallow beams using
denser, and thus exhibits much more spalling and advanced composite materials. The composites
degradation under heating as compared to normal used include: (i) high strength cementitious grouts
strength concrete (NSC) as a result of the high reinforced with different types of discrete, dis-
pressure created because of escaping vapors continuous steel and high performance polypro-
[4–6]. pylene fibers; (ii) ferrocement laminates; and (iii)
Limited or no data exists in the literature on high-strength fabric polymeric sheets such as glass
the flexural behavior of shallow HSC beams (GFRC) sheets. The results of this study are
when exposed to elevated temperature. It is a expected to provide a cost effective, and long
challenge for structural engineers to develop term repair and retrofit solutions that can be
efficient rehabilitation techniques that enable implemented in concrete industry.
such beams to restore their structural integrity,
after being exposed to intense fires for longer
periods of time. A few researchers have used 2 Experimental program
conventional methods for repairing fire-damaged
reinforced concrete members [7, 8]. Most of 2.1 Materials properties
these methods involved essentially removing
deteriorated concrete layers and replacing them 2.1.1 Reinforcing steel
with fresh concrete. Lin et al. [7] investigated the
behavior of reinforced concrete columns re- Grade 75 deformed steel rebars of diameters
paired with fresh cast-in-place concrete after (14 mm and 10 mm) were used in all RC beams.
being damaged by fire. The results showed that The stress–strain diagrams for the 14-mm-rebars
most of the repaired columns could reach their before and after heating to 600C and then
original stiffness and strength or even higher cooling are presented in Fig. 1. The geometric
than those of unheated columns. Lange [8] used and mechanical properties of the rebars, as
epoxy injection, shotcrete repair of damaged measured in the laboratory, are listed in Table 1.
concrete surfaces, and stiffening of structural
members to regain the structural integrity of the
St. Elizabeth Hospital in Holland after being
damaged by an intense fire. 800

Usually, repair of heat-damaged high strength

700
members would be the only solution to regain
structural capacity. A practical method of repair 600
should take into consideration, the amount of
500
Stress (MPa)

damage, the shape of the beam, materials of

repair, construction cost, time and practicality. 400
Structural repair of reinforced concrete structures
300
have become a significant part of the total cost of
construction in the world. Several repair tech- 200
niques were used to repair damaged concrete at 23C
100 at 600C
structural elements. These involve strengthening
beams and columns by epoxy bonding of steel- 0
plates, external fixing of ferrocement laminates or 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Strain (m/m)
high performance fiber reinforced concrete, and
bonding of fiber reinforced polymers sheets to Fig. 1 Stress–strain diagram for Grade 75 steel rebars
existing damaged concrete [9–19]. before and after heating then cooling
Materials and Structures (2008) 41:287–299 289

Table 1 Mechanical and physical properties of reinforcing steel used

Temperature (C) YS (MPa) US (MPa) EAF (%) RH (mm) RS (mm) FA (degrees)

23 633 760 10.6 0.53 5.4 53.8

600a 500 668 30.4 0.57 5.05 54.3
a
= Properties after cooling, YS = Yield stress, US = Ultimate stress, EAF = Elongation at failure, RH = Rip height,
RS = Rip space, FA = Face angle

2.1.2 Fibers beams. The geometric and physical properties of

the glass fabric sheets are listed in Table 4.
Two types of fibers were used in preparing the Special resin was used to create a very strong
high strength fibrous cementitious grout layers. bond between the fabric sheets and the concrete
The fibers used were a blend of equal proportions surface. The quantity of resin coat recommended
of two types of steel fibers (brass-coated (BCS), for gluing the sheets with concrete surface ranges
and hooked steel (HS)), and high performance from 0.70 kg/m3 to 1.2 kg/m3 and that for inter-
polypropylene fibers (HPP) at volumetric frac- layer and final coats was about 0.5 kg/m3. The
tions of 2, and 1.5%, respectively. The mechanical tensile strength, the strain at breaking, and
and geometrical properties of the fibers used are flexural modulus of elasticity for the resin as
listed in Table 2. provided by the manufacturer were 30 MPa,
1.5%, and 3.8 GPa, respectively.
2.1.3 Ferrocement laminates
2.1.5 Concrete mix
Two galvanized woven square steel wire meshes
of two sizes (small and medium) at two or four Crushed coarse basalt aggregates and a mixture of
layers were used in preparing the 25-mm-ferro- equal proportions of crushed fine basalt and silica
cement laminates. The geometric and mechanical sands were used with ordinary Portland cement
properties of meshes used are listed in Table 3. (Type I) in preparing the high strength concrete
mix. The crushed coarse basalt had a maximum
2.1.4 High-strength fabric sheets system aggregate size of 9.5 mm. The fineness moduli of
fine basalt and the silica sand, determined accord-
Two types of unidirectional fabric sheets of ing to ASTM-C136 [20] was found to be 3.4 and
0.13 mm thickness of glass (GFRP) fabric types 1.6, respectively. The bulk specific gravity and the
were used in repairing heat-damaged beams. absorption for coarse and fine basalt particles,
Epoxy based impregnating resin was used to glue obtained according to ASTM-C127 and C128 [20]
the sheets along the tension surfaces of the were 2.73 and 1.9%; and 2.64 and 5.2%,

Table 2 Mechanical and geometrical properties of fibers

Fiber Geometrical configuration YS (MPa) SG Diameter (mm) Length (mm) Aspect ratio

BCS 2950 7.8 0.15 6 40

HS 1172 7.8 0.5 30 60

HPP 525 0.91 (1 · 0.8)a 40 –

BCS = Brass-coated steel, HS = Hooked steel; HPP = High performance polypropylene, SG = specific gravity, YS = Yield
stress
a
rectangular cross section
290 Materials and Structures (2008) 41:287–299

Table 3 Geometric properties of the wire meshes used in ferrocement laminates

Mesh size NOL Geometric properties Mechanical properties
Opening space (mm) Diameter (mm) YS (MPa) US (MPa) USY (m/m) USR (m/m)

Small 2 or 4 2.8 0.5 295 500 0.83% 2.1%

Medium 4 6 0.7 215 384 0.87% 2.4%
NOL = Number of layers, YS = Yield stress, US = Ultimate strength, YSR = Yield strain, USR = Ultimate strain,
E = Elasticity modulus

Table 4 Physical properties of GFRP same high strength cementitious slurry mentioned
above, reinforced with steel meshes of different
Fiber type Glass
sizes at two or four layers.
Area weight 440 g/m2
Tensile strength 2276 N/mm2
Tensile E-modulus 72500 N/mm2 2.2 Specimen geometry and steel
Strain at break 4%
reinforcement detailing

respectively. The corresponding values for silica Sixteen shallow HSC beams
sand were 2.6 and 0.5% respectively. The unit (100 · 250 · 1500 mm) were designed according
weight for coarse aggregate, according to ASTM- to the ACI Code [22]. The cross sectional
C29 [20] was found to be 1560 kg/m3. geometric properties and steel detailing are pre-
The high strength concrete mix (HSC) used in sented in Fig. 2. The beams were cast in a
this study was designed using a w/c ratio of 0.35 specially designed wooden mold of 20 mm thick-
according to ACI-211 mix design procedure to ness, with a concrete cover of 25 mm to the
attain 28-days cylinder compressive strength of centroid of main reinforcement. Special steel
65 MPa and a slump of about 50 mm. The spacers were positioned inside the mold under-
contents of cement, water, coarse and fine basalt, neath the reinforcement to achieve the required
and silica sand used were 516, 215, 763, 452, and cover.
452 kg/m3, respectively. A superplasticizer was
used at 2% by weight of cement. 2.3 Mixing, casting, and curing of the beams

2.1.6 Repair layers A titling drum mixer of 0.15 m3 capacity was used
in mixing the concrete ingredients following
The slurry mix was designed to have a high ASTM-C192 [20]. The slump was measured,
compressive strength, high workability, and excel- according to ASTM-C143 [20] and was found to
lent durability. The slurry mix proportions were be about 50 mm. The casting of the beam
1:0.6:0.15:0.42:0.03 by weight of Type I cement, specimens was performed in the following se-
sand, silica fume, water and superplasticizer, quence: the wooden molds were oiled and the
respectively [21]. steel cages were placed, and followed by pouring
The fibrous grouts consist of the same slurry the concrete in two layers; each was compacted
mixture reinforced with fibers of different types using a vibrating table to achieve a satisfactory
and contents. Type I cement, sand, silica fume, compaction. The surface of the beams was
water and superplasticizer were used in preparing finished smooth by a trowel. Finally, the cast
the slurry. The silica sand had a maximum size of beams were demolded and covered with wet
about 0.2 mm whereas the silica fume used had an burlap for about 4 months. For controlling the
average of 93% silica dioxide content and a quality of the concrete, 100 mm cubes were cast
specific gravity of 2.35. The superplasticizer used in steel molds, demolded 24 h after casting, and
was Cico fluid Type ME1 with a specific gravity of cured under the same environment, the beams
nearly 1. The ferrocement layers consist of the were subjected to.
Materials and Structures (2008) 41:287–299 291

Fig. 2 Details of beams I Φ 8 mm/S=90 mm Φ 8 mm/S=50 mm

reinforcement

I
1400 mm
1500 mm

2 Φ10 mm

h=100 mm

3Φ14 mm
b=250 mm
SECTION I-I

2.4 Heating procedures 800

Furance Temprature ( °C)

After about 4 months of curing, the specimens 600
were left for a week in the laboratory air (relative
humidity 60%, temperature 23C). Beams 400
designated to undergo thermal treatment were
subjected to 600C for 3 h using the electrical
200
furnace shown in Fig. 3. The time–temperature
schedule used in this study is shown in Fig. 4.
0
0 4 8 12 16 20 24 28
2.5 Repair techniques Time (hours)

Fig. 4 Time–temperature schedule

The heat-damaged beams were strengthened at
the compression surface by 25-mm-thick plain
grout layer and on the tension surface by three Casting of the top grout layer aimed at: (i)
types of composite materials: high strength compensating for the dramatic loss in compres-
fibrous cementitious grouts; 25-mm-thick ferroce- sive strength upon heating; and (ii) increasing the
ment laminates, and two layers of GFRP sheets, resisting arm of tension layers added.
respectively. Steel studs of 8-mm were planted in The surfaces of heat-damaged beams were
regions of high shear to prevent horizontal roughened using a special hammer. This was
slippage of ferrocement laminates and the plain followed by drilling and vacuum cleaning 10-mm-
and fibrous grout layers as shown in Fig. 5. holes in the regions of high shear, through which
8-mm-steel studs were planted using a special
epoxy. Then, wooden frames were assembled and
fixed to the perimeter of beams, and lubricated to
allow for casting 25-mm-cover layers. Thereafter,
the plain grout ingredients were mixed in a
Hobart Type Laboratory Mixer of 0.06 m3 capac-
ity. The 28-day-compressive strength of the plain
grout was about 50 MPa obtained using cylindri-
cal specimens of 75 · 150 mm.

2.5.1 Fibrous grout layers

The fibrous grout layers were cast by infiltrating a

Fig. 3 Beam and cube specimens in the electric furnace blend of hooked and brass coated steel fibers into
292 Materials and Structures (2008) 41:287–299

Fig. 5 Distribution of Grout layer

steel studs in the high
shear region (dimensions Fibrous or
are in mm) ferrocement layer 80 150 150 150 150 150 150 80

Side view

80 150 150 150 440 150 150 150 80

50 75 75 50

250
1500

the grout, or infiltrating polypropylene fibers at dust cleaned by a vacuum. Tinner saturated cloths
three stages. After 24 h, the repaired zones were were placed at the cleaned surfaces to reduce the
demolded and cured using wet burlap for another moisture content. The fabric was carefully placed
27 days. onto a special resin coating, mixed of two com-
ponents, and applied to the surface by a brush at
2.5.2 Ferrocement laminates 1 kg/m2. Using a wooden roller, the fabric was
carefully worked into this resin until resin
The ferrocement laminates were prepared using squeezed out between the roving. The second
two or four layers of small or medium spacing layer of the sheets was rolled and worked into
steel wire meshes, and spaced using 5-mm-thick 0.50 kg/m2 of the resin, applied to the first layer.
steel pieces, as shown in Fig. 6. The mesh layers Finally, an additional 0.50 kg/m2 of resin was
were drilled at location where steel studs were brushed on the surface of the exterior sheet layer.
pre-planted then placed on the tension surface of
the specimens. Thereafter, the plain grout was 2.6 Flexural testing
infiltrated in the meshes and vibrated.
The simply supported beams were tested under
2.5.3 High-strength fabric sheet system four points-bending test to determine the load–
deflection curves using Dartec compression ma-
Two layers of glass sheets (250-mm by 1500-mm) chine with a full jack capacity of 2000 kN under a
were used in repairing the heat-damaged beam. loading rate of 0.02 kN/s. An LVDT, placed at
The areas of sheets were chosen so that both the the middle point of the bottom of the beam, was
existing steel and the sheets yielded prior to connected to a data acquisition system to measure
delamination of sheets or crushing of concrete mid-span deflection (Fig. 7). All beams were
[23]. The concrete surface was first smoothened tested under monotonic increasing load up to
and leveled using a diamond grinding disk, then failure.

3 Results and discussion

The flexural behavior of the control, damaged,

and repaired beams was evaluated by studying the
load–deflection diagrams, and their characteris-
tics, namely, the ultimate load capacity, corre-
sponding deflection, stiffness and toughness. The
load–deflection diagrams shown in Figs. 8–11
Fig. 6 Ferrocement layer details represent the average of two tested beams. The
Materials and Structures (2008) 41:287–299 293

Fig. 7 Details of the P

flexural test setup

Hinge support Roller support

LVDT

400 mm
1400 mm
1500 mm

slope of the linear portion of the load–deflection

diagrams represents the stiffness of the beam
while the area underneath the diagram represents
toughness. The characteristics of the load–deflec-
tion diagrams for control, heat-damaged and
repaired beams tested in this study are summa-
rized in Table 5. Cracking modes of the beams
after heating and results of flexural testing are
presented and discussed in the following articles.

3.1 Effect of heat treatment

Experimental load versus mid-span deflection

Fig. 9 Load–deflection for control, damaged, and fibrous
curves for the beams before and after heating grout repaired beams
are shown in Fig. 8. Three specific regions can be
identified: a linear region to yield, a transition
region of continuous yield, and a region of full ultimate load capacity after heating and an
plastic deformation until failure. The results increase in the corresponding deflection by about
indicated a substantial loss of about 28% in the 1.4 times than that of the control beams. The

Fig. 10 Load–deflection for control, damaged, and ferro-

Fig. 8 Load–deflection for control and damaged beams cement repaired beams
294 Materials and Structures (2008) 41:287–299

niques showed an increase in the flexural load

capacity with varying mid-span deflections. In
general, the increase in the ultimate load capacity
is due to the influence of increased moment of
inertia of the composite cross section. The stiff-
ness improved significantly, whereas the tough-
ness decreased.

3.3 Fibrous grout layers

The load–deflection curves for beams repaired

Fig. 11 Load–deflection for control, damaged, and FRP
with fibrous grout layers are shown in Fig. 9.
sheet repaired beams
Beams repaired with HBCSFRC or PPFRC
layers achieved the original load flexural capacity,
stiffness of the damaged beam was reduced by and experienced limited reduction in deflections
about 56%, due to the reduction in the flexural at ultimate loads; they achieved about 108 and
rigidity. The toughness of the damaged beams 99% of the original load flexural capacity respec-
was almost the same as that of the control ones. tively. The toughness showed a slight increase as
It should be noted that the reduction in flexural compared to that of control, whereas the stiffness
strength of the heat damaged beams is related to of beams with HBCSFRC and PPFRC layers
the decrease in concrete strength, and loss in yield were close to that of control ones.
strength of reinforcing steel, and the loss of bond
strength between steel and concrete; whereas the
increase in ductility is related to the increase in 3.4 Ferrocement laminates
elongation of reinforcing steel at failure upon
heating as shown in Table 1. Since the failure of The load versus mid-span deflection relationship
heat-damaged beams was controlled by yielding for beams repaired with the three types of
of the bottom reinforcing steel, the ductility of the ferrocement layers are illustrated in Fig. 10.
reinforced concrete beams was increased. Beams repaired with ferrocement layers showed
a significant increase in ultimate load and a
3.2 Effect of repair technique reduction in the corresponding deflection. The
curves pertaining to repaired beams exhibit sim-
Depending upon the type and properties of the ilar behavior to that of damaged and control ones.
repair materials used, all types of repair tech- The beams repaired with two and four layers of

Table 5 Load–deflection diagram characteristic for control, heat-damaged and repaired beams
Beam Pult (kN) dult (mm) Stiffness (kN/mm) Toughness (kN mm)

Control 82.3 24.82 5.36 1259

Damaged 58.8 (71.4%)a 34.38 (138.5%) 2.37 (44.2%) 1264 (100.4%)
FR2S 79.9 (97.1%) 18.2 (73.3%) 6.44 (120.2%) 1000 (79.4%)
FR4S 78.5 (95.4%) 20.7 (83.4%) 5.67 (105.8%) 1102 (87.5%)
FR4M 75.9 (92.2%) 13.9 (56%) 8.38 (156.3%) 720 (57%)
HBCSFRC 88.7 (107.8%) 23.3 (93.8%) 5.56 (103.7%) 1392.3 (110.6%)
PPFRC 81.7 (99.3%) 23 (92.6%) 5.25 (97.9%) 1282 (101.8%)
GFRP sheet 104.1 (126.5%) 15.2 (61.2%) 8.58 (160%) 929 (73.7%)
a
= Percentage of control, Pult = Ultimate load capacity at failure, dult = Ultimate deflection at mid-span corresponding to
ultimate load capacity, FR2S = Ferrocement at two layers of small mesh size, FR4S = Ferrocement at four layers of small
mesh size, FR4S = Ferrocement at four layers of medium mesh size, HBCFRC = Hooked and brass-coated fiber reinforced
grout, HPPFRC = High performance polypropylene fiber reinforced grout, GFRP = Glass fiber reinforced polymer
Materials and Structures (2008) 41:287–299 295

small and medium size wire meshes (FR2S, FR4S compared to the parent concrete as seen from
and FR4M) regained a large portion of the Tables 1 through 4. It has been shown that an
original flexural load capacity and reached about appropriate use of these composites in the form of
97, 95 and 92%, respectively. On the other hand, a jacket serve as an external reinforcement that
with respect to deflections of control beams, the provides a better confinement to the damaged
deflections were reduced by 27, 17 and 35%, concrete and leads to an increase in the resisting
respectively. The flexural toughness for beams moment arm, and thus results in a considerable
with ferrocement layer was reduced compared increase in the flexural strength of the section.
with that of the control ones. The percentage Larger increase in flexural strength can also be
reductions were about 21, 12 and 43% for (FR2S, achieved by ensuring full bonding of the jacket
FR4S and FR4M), respectively. The stiffness of with the beam surface (e.g. roughening the beam
the same beams reached as high as 156% of that surface and using concrete screws for improving
of the control ones. shear resistance).

3.5 High-strength fabric sheets system

4 Cracking patterns and failure modes
The load–deflections curves for control and dam-
aged specimens as compared to those repaired 4.1 Cracking after heating
with glass (GFRP) sheets are shown in Fig. 11.
The results reflect the advantage of using the The cracking patterns resulting from heating the
GFRP in repairing heat-damaged HSC beams. RC beams to 600C are shown in Fig. 12. The
The flexural load capacity of these beams cracks (marked) spread laterally and longitudi-
exceeded the flexural capacity of control and nally at the location of transverse and longitudi-
damaged HSC beams. The curves pertaining to nal reinforcements with widths ranging from
(GFRP) beams exhibited linear behavior before 0.1 mm to 2 mm. The actual size of these cracks,
becoming nonlinear prior to failure; indicating measured directly after heating, indicated much
yielding of main reinforcement prior to concrete larger values. This behavior is related to the
crushing. The enhancement in the ultimate load higher expansion of concrete under heating as
far exceeded that of control beams by 127%, with compared to that of embedded steel due to the
a corresponding decrease in deflections by 39%. difference in the coefficient of thermal expansions
It can be noticed that the stiffness showed [6]. Because of the relatively high exposure
improvements by about 160%, while the flexural temperature, the concrete surrounding steel bars
toughness was reduced by about 26%. suffered extensive cracking. Upon cooling of
The loss of ductility in GFRP strengthened heat-treated concrete beams, cracks sizes were
beams is due to the lower ductility of the GFRP reduced due to the significant contraction of
layers as compared to that of original reinforcing embedded longitudinal and transverse reinforcing
steel, as indicated by the ultimate strains at failure
that are considerably less than those for steel as
shown in Tables 1 and 4. Since the failure of
repaired and control concrete beams was dictated
by the failure of the extreme bottom fibers, the
beams strengthened with the GFRP layers
showed lower ductility as compared to control
(not strengthened) ones.
The causes for the increase in the flexural
strength of the repaired beams can be explained
as follows: the composites used for repair in this
investigation possess relatively high tensile Fig. 12 Cracking patterns marked on the surface of a
strength in longitudinal and transverse directions heated beam
296 Materials and Structures (2008) 41:287–299

steel. Short cracks in the shape of a spider net The cracking characteristics and failure modes
appeared also on the surface of the beams. These of the beams are listed in Table 6. The initial
were caused by the combined effects of water crack load under flexural loading for damaged
vapor pressure, and decompositions of calcium beams exhibited lower value than that of the
hydroxide and possibly calcium silicate hydrate control ones by about 50%. Relatively wider
[6]. cracks with similar crack spacing and sizes were
observed on the beams throughout the region of
4.2 Control and damaged beams uniform bending moment.

A typical ductile flexural failure was observed in 4.3 Repaired beams

under-reinforced control, and damaged beams.
The flexural cracks increased and propagated 4.3.1 Fibrous layers
upward in the region of high moment with load
increase. Upon yielding of main steel, the upward The cracking patterns at the soffit of the beams
shift of the neutral axis was significant and repaired with fibrous grouts indicated a reduction
resulted in crushing of the concrete in the in the number of cracks by about 50% and an
compression zone. Small shear cracks of minor increase in the cracks spacing by about 50% as
effect were observed throughout the shear spans. compared to those of control ones. Wider cracks
Figure 13 shows a typical failure mode for heat- spacing were observed in the beam repaired with
damaged beams under flexural loading. HBCSFRC layers compared to those repaired with

Fig. 13 Cracking pattern

for damaged beam

Table 6 Cracking loads and characteristics and mode of failure for different HSC beams
Beam Pcr (kN) Crack width (mm) Cracks number Crack spacing (cm) Mode of failure

Control 10 0.2–2 17 6 Flexural failure

Damaged 5 0.2–1 16 8 Flexural failure
FR2S 10 0.2–1 10 9 Peeling of grout in FR2S layer
FR4S 10 0.3 11 7 Peeling of grout in FR4S layer
FR4M 12 0.2–1 10 5 Yielding of wire mesh
HBCSFRC 20 0.2–5 8 15 Flexural failure
HPPFRC 15 0.2–0.4 7 13 Flexural failure
GFRP 38 <0.1 14 7 Compression failure
FR2S = Ferrocement at two layers of small mesh size, FR4S = Ferrocement at four layers of small mesh size,
FR4S = Ferrocement at four layers of medium mesh size, HBCFRC = Hooked and brass-coated fiber reinforced grout,
HPPFRC = High performance polypropylene fiber reinforced grout, GFRP = Glass fiber reinforced polymer
Materials and Structures (2008) 41:287–299 297

Fig. 14 Cracking pattern

for beam repaired with
PPFRC layer

PPFRC layer. Typical flexural cracking patterns mode of the repaired beams depends on the mesh
for the beams repaired with PPFRC layers are size used; beams repaired with FR2S and FR4S
illustrated in Fig. 14. Repairing with PPFRC and layers exhibited peeling failure of the grout from
HBCSFRC layers, resulted in an increase in the the wire mesh, whereas those repaired with
cracking loads as compared with control ones by FR4M layers showed typical flexural failure
about 50% and 100%, respectively. without peeling of the meshes, as shown in
Figs. 15 and 16. The peeling of the outer grout
4.3.2 Ferrocement layers layers in FR2S and FR4S beams is a result of the
insufficient bonding between the small size wire
Visual observations, and statistical evaluation of mesh and the repair grout caused by the inability
the cracks appeared on the tested beams indi- of the coarse sand particles in the repair grout to
cated that the ferrocement laminates modified the penetrate in between the meshes; in contrary to
cracking patterns at the soffit of the repaired the case of FR4M beams where the medium size
beams. Closer crack spacing and smaller widths of meshes allowed the grout to form a continuous
the cracks were observed in the beams repaired matrix through the entire thickness of the repair
compared to the un-repaired ones. The cracks’ layer, and thus allowed the steel mesh to yield
spacing and widths at the soffits were reduced by prior to failure without peeling.
about 50% and 25%, respectively, as shown in
Table 6. This is due to the significant increase in 4.3.3 High-strength fabric sheets system
the moment of inertia of the cross section and to
the improvement in the tension resistance of the The beams repaired with GFRP sheets showed
bottom surface of the repaired beams. The failure limited deflection, accompanied with an extensive

Fig. 15 Cracking pattern

for beam repaired with
FR2S layer
298 Materials and Structures (2008) 41:287–299

Fig. 16 Cracking pattern

for beam repaired with
FR4M layer

transverse cracks at the top surface of the upper the design of GFRP layers would overcome this
compression grout layer, prior to failure. At type of failure.
higher loads the cracks extended through the
steel studs, used to prevent the slippage of the
upper layer. The load carrying capacity of the
5 Conclusions
beams kept on increasing till the cracks at the top
and the soffit of the beams propagated in all
The following conclusions could be drawn from
directions, followed by sudden drop in the
the experimental results of the present study:
ultimate load, and thus lead to failure. The
cracking loads for the beams repaired with GFRP 1. Heating high strength reinforced concrete
sheets were significantly increased, whereas the shallow beams at 600C for 3 h caused
crack width and spacing were reduced as com- extensive cracking accompanied with a sub-
pared to those of control and damaged ones as stantial loss in flexural load capacity, and the
shown in Table 6. The GFRP sheets themselves corresponding mid-span deflection, stiffness
showed no signs of delamination or failure prior and toughness.
to main steel yielding and crushing of the upper 2. The beams repaired with fiber glass sheets
layer, indicating that the failure mode of the (GFRP) regained up to 127% of the control
beams repaired with GFRP sheets was a com- beam’s ultimate load capacity with a sub-
pression failure of the upper grout layer prior to stantial increase in stiffness, first cracking
yielding of steel or rupture of the GFRP sheets as loads, and a corresponding decrease in mid-
demonstrated in Fig. 17. This type of behavior, span deflections at ultimate load.
although undesirable, reflected the substantial 3. The beams repaired with high strength steel
contribution of GFRP to the tensile strength of fiber reinforced concrete jackets (SFRC)
the bottom surface of the beam. Optimization in regained up to 106% of the control beam’s

Fig. 17 Cracking pattern

for beam repaired with
GFRP sheet
Materials and Structures (2008) 41:287–299 299

ultimate load capacity with a significant 4. Chun S, Peng G, Chun K (1996) Comparison between
increase in stiffness, first cracking loads, and high strength concrete and normal strength concrete
subjected to high temperature. Mater Struct 29:616–619
a corresponding decrease in mid-span deflec- 5. Zoldrens N, Malhotra V, Wilson H (1960) Effect of
tions at ultimate load. high temperatures on concretes incorporating different
4. The beams repaired with high performance aggregates. Proc ASTM 60:1087–1108
polypropylene fiber reinforced concrete jackets 6. Bazant Z, Kaplan M (1996) Concrete at high temper-
ature: material properties and mathematical models,
(PPFRC) have shown a marginal improvement 1st edn. Longman Group Limited, England, ch. 2
in their flexural behavior without a noticeable 7. Lin C, Chen S, Tang C (1995) Repair of fire-damaged
regain in their original load capacity. reinforced concrete columns. ACI Struct J 92:406–411
5. The beams repaired with two and four layers of 8. Lange G (1980) Structural repair of fire damaged
concrete. Concrete Int 2(9):27–29
small and medium size wire meshes regained up 9. Altin S, Anil O, Kara M (2005) Improving shear
to 99% of the control beam’s ultimate load capacity of existing RC beams using external bonding
capacity with a significant increase in stiffness, of steel plates. Eng Struct 27:781–791
and a corresponding decrease in mid-span 10. Adhikary B, Mutsuyoshi H (2006) Shear strengthening
of RC beams with web-bonded continuous steel plates.
deflections and toughness. Constr Build Mater 20:296–307
6. Under flexural testing, control and heat- 11. Paramasivam P, Lim C, Ong K (1998) Strengthening of
damaged beams showed typical flexural RC beams with ferrocement laminates. Cement Con-
cracks in the pure bending region and shear- crete Comp 20:53–65
12. Al-kubaisy M, Jumaat M (2000) Flexural behavior of
flexure cracks in the constant shear region. reinforced concrete slabs with ferrocement tension
The repaired beams showed typical flexural zone cover. Constr Build Mater 14:245–252
failure with an increase in the first crack load. 13. Nassif H, Najm H (2004) Experimental and analytical
7. All the repair techniques proposed in this study, investigation of ferrocement–concrete composite
beams. Cement Concrete Comp 26:787–796
may be considered as a promising methods for 14. Shannag M, Al-rousan R (1999) Shear strengthening of
restoring the flexural capacity of heat damaged high-strength reinforced concrete beams using fibrous
RC members: GFRC being the most effective composite. Mag Concrete Res 56:419–428
and expensive, SFRC and PPFRC effective and 15. Krstulovic N, Al-shannag M (1999) Slurry infiltrated
mat concrete (SIMCON)-based shear retrofit of rein-
less expensive, followed by ferrocement, the forced concrete members. ACI Struct J 96:105–114
least expensive and still effective, which may be 16. Naaman A, Reinhardt H, Fritz C (1993) Reinforced con-
considered as a cheaper repair alternative in low crete beams with SIFCON matrix. ACI Struct J 89:1–15
cost construction. 17. Khalifa A, Nanni A (2002) Rehabilitation of rectan-
gular simply supported RC beams with shear defi-
ciencies using CFRP composites. Constr Build Mater
Acknowledgements The authors would like to 16:135–146
acknowledge the support received from the deanship of 18. Ashour A, El-refaie S, Garrity S (2004) Flexural
scientific research of University of Science and strengthening of RC continuous beams using CFRP
Technology at Jordan, and the assistance of the laminates. Cement Concrete Comp 26:765–775
technicians at the engineering workshop and civil 19. Barros J, Fortes A (2005) Flexural strengthening of
engineering laboratories. concrete beams with CFRP laminates bonded into
slits. Cement Concret Comp 27:471–480
20. ASTM standards (2005) Construction: concrete and
aggregates, vol 04–05. American Society for Testing
References Materials, MI, USA
21. Shannag M (2002) High-performance cementitious
grouts for structural repair. Cement Concrete Res
1. El-hawary I, Ragab A, Abd el-azim A, Elibiari S
32:803–808
(1996) Effect of fire on flexural behaviour of R.C.
22. ACI committee 318 (2005) Building code requirements
beams. Constr Build Mater 10:147–150
for structural concrete (ACI 318M-05). American
2. Haddad R, Shannis I, (2004), Post-fire behavior of
Concrete Institute, Mich., USA
bond between high strength pozzolanic concrete and
23. Grace N, Singh S (2005) Durability evaluation of car-
reinforcing steel. Constr Build Mater 18:425–435
bon fiber-reinforced polymer strengthened concrete
3. Chan Y, Luo X, Sun W (2000) Compressive strength
beams: experimental study and design. ACI Struct J
and pore structure of high performance concrete after
102:40–53
exposure to high temperature up to 800C. Cement
Concrete Res 30:247–251