1 s2.0 S0141029612000338 Main

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 11

Engineering Structures 39 (2012) 176–186

Contents lists available at SciVerse ScienceDirect

Engineering Structures
journal homepage: www.elsevier.com/locate/engstruct

Flexural and shear behaviors of reinforced ultrahigh toughness cementitious


composite beams without web reinforcement under concentrated load
Shilang Xu a,b,⇑, Lijun Hou a, Xiufang Zhang a
a
Department of Civil Engineering, Dalian University of Technology, Dalian, China
b
Institute of Advanced Engineering Structures and Materials, Zhejiang University, Hangzhou, China

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

Article history: Ultrahigh toughness cementitious composite (UHTCC) behaves strain-hardening and multiple-cracking
Received 18 January 2011 when subjected to tensile load. The current work investigates both bending and shear performances of
Revised 1 January 2012 steel reinforced UHTCC (RUHTCC) beams through the bending test under concentrated load. The failure
Accepted 11 January 2012
mode was designed from flexure to flexure-shear to shear through varying reinforcement ratio in beam.
Available online 22 March 2012
The experimental results showed that for flexural behavior, RUHTCC beams presented high load-carrying
capacity, deformation ability and numerous fine cracks around mid-span region in comparison with
Keywords:
ordinary reinforced concrete (RC) beam. Based on the internal force equilibrium, the theoretical equation
Ultrahigh toughness cementitious
composite (UHTCC)
for determining the moment of inertia of fully cracked section of RUHTCC beam is proposed, and the pre-
Concentrated load dicted maximum deflection at service load according to proposed equation based on bilinear tensile and
Flexural behavior compressive constitutive models agrees well with the tested value. For shear behavior, RUHTCC beam
Shear behavior exhibited shear strength about 1.91 times that of RC control beam and a large reserve of shear bearing
Crack pattern capacity after the shear cracking due to the diagonal multiple-cracking pattern and the stable crack
development. And for RUHTCC beam with reinforcement ratio of 2.28% failing in flexure-shear, a large
deformation comparable to that of RC beam with 0.67% reinforcement ratio was observed.
Ó 2012 Published by Elsevier Ltd.

1. Introduction cementitious matrix be improved effectively. Recently, a group of


high performance fiber reinforced cementitious composite (HPF-
In practice, RC members always work with cracks during its RCC) characterized by deflection-hardening behavior even tensile
work-life. For ensuring the durability of concrete structure in some strain-hardening behavior appeared continually [8], such as SIF-
specific environments, the maximum allowable crack width is of- CON (slurry infiltrated fiber concrete), SIMCON (slurry infiltrated
ten limited to prevent the corrosive medium from entering the mat concrete), ECC (engineered cementitious composites), and
structure freely [1]. In addition, the ductility of RC member with UHTCC (ultra high toughness cementitious composite) [9]. Similar
large reinforcement ratio is not high enough to ensure structure to ECC [10], UHTCC presents excellent tensile strain capacity up to
safety under accidental load, i.e., earthquake and impulsive action. 3%, fine crack dispersion capability with average crack width less
For other important members requiring a large shear resistance, than 0.1 mm and tight crack spacing of 1–2 mm, and outstanding
such as the beam-column joint, shear wall, beam ends and column deformation capacity [9,11].
base, large web reinforcement ratio is always applied to enhance The flexural experiment under two symmetrical concentrated
the shear capacity and compensate the brittle property of concrete load have been conducted by Xu and Zhang [12]. Their results
[2]. Consequently, to improve the ductility in complex structures showed that compared with the RC beam, RUHTCC beam, as well
or structural key positions as well as to enhance the durability in as RC/UHTCC member in which a layer of UHTCC with varied depths
harsh environments, some new materials with excellent nonlinear replaced concrete as a cover of longitudinal steel bar, not only im-
deformation and crack control abilities are desired for the practical proved load capacity, deformation and ductility capability, but
application instead of brittle concrete in future. effectively controlled the crack opening by forming a large number
For improving the flexural and shear performances of plain of fine cracks in pure-bending zone. Li and Wang [13] conducted the
concrete, one of main methods was the addition of various fibers comparison studies on flexural and shear behaviors between glass
[3–7], which makes the native brittleness property of fiber-reinforced polymer (GFRP) reinforced ECC (GRE) beams and
GRC beams. The experimental results revealed that the crack
⇑ Corresponding author at: Institute of Advanced Engineering Structures and opening can be effectively suppressed by use of ECC even if GFRP
Materials, Zhejiang University, Hangzhou, China. Tel.: +86 571 88208676. has a lower elastic modulus than steel bar. For the members with
E-mail address: slxu@zju.edu.cn (S. Xu). the same reinforcement ratio and similar compressive strength of

0141-0296/$ - see front matter Ó 2012 Published by Elsevier Ltd.


doi:10.1016/j.engstruct.2012.01.011
S. Xu et al. / Engineering Structures 39 (2012) 176–186 177

matrix, shear load capacity of GRE beam with no stirrups was close
to that of GRC beam with web reinforcement ratio of 0.77%. Li et al.
[14] carried out the Ohno shear beam to experimentally compare
the shear performances between PC (plain concrete), RC, SFRC (steel
fiber reinforced concrete), SPECC (spectra fiber ECC) and DRECC
(dramix fiber ECC). Among these materials investigated by them,
SPECC possessed tensile strain of 5.6% and ultimate strength of
4.7 MPa, which are similar to that of UHTCC mentioned above.
The experimental results indicated that the SPECC beam had a shear
strength of 5.09 MPa, about twice that of PC and even comparable
to that of RC beam with web stirrups ratio of 0.75%. Moreover,
the SPECC beam demonstrated a diagonal multiple-cracking pat-
tern and hence had the much larger shear strain than other beams.
Shimizu et al. [15] also reported that the shear behavior of PVA-ECC
beam was far superior to plain RC beam, presenting a prominent
crack control ability and about 1.6 times load-carrying capacity.
Moreover, due to the outstanding mechanical properties and Fig. 1. Details of beams tested.
durability, UHTCC can be used to repair the aged RC members.
The experimental studies [16–18] demonstrated the improved load
the depth and effective depth of cross section. For steel bar, Es
bearing capacity and ductility as well as fine cracking pattern were
and fy were respectively considered to have a value of 200 GPa
obtained through casting a layer of UHTCC in tension zone, but the
and 380 MPa referring to previous experiments. In the initial design
interface failure between UHTCC and old material can result in the
of this test, ecp, rtc and cubic compressive strength fcu of UHTCC
shear failure of repaired member [17].
were first assumed to be about 0.005, 4.0 MPa and 44 MPa [12],
In previous experimental studies, it was noticed that the flex-
respectively, according to the test data previously reported. Based
ural behavior of RUHTCC beam were investigated through sym-
on the compression experimental data with respect to UHTCC con-
metrically four point loading pattern. With respect to structural
ducted by Xu and Cai [19], it is found that the cube compressive
performance of RUHTCC beam under other loading pattern such
strength (fcu) was almost comparable to the prism compressive
as one concentrated point load at the center of beam, there were
strength (fc), which differs from the transformation relationship of
little discussions up to date. Furthermore, the shear member in
fc = 0.79fcu for concrete [20]. That means that the cube compressive
practical engineering structures appears generally in a type of
strength fcu of UHTCC can be taken as an estimate of its prism com-
beam or column member, which is different from the Ohno shear
pressive strength fc. Thus, substituting these values into the Eqs. (1)
beam presented in reference [14,15]. With these considerations,
and (2), the critical reinforcement ratio qb of roughly 4.84% was pre-
in the current paper, the experiment study on RUHTCC beams
dicted for RUHTCC beam designed in current test.
without web reinforcement is performed under a concentrated
Based on limit reinforcement ratio of 4.84%, the longitudinal
load at the center of beam to comprehensively understand their
reinforcement ratio was designed as 0.67%, 1.00%, 1.73%, 2.28%,
bending and shear responses. In this test, the failure mode from
and 3.25%. The specimens were respectively labeled as Af, Ac, Bf,
flexural failure to flexure-shear failure to shear failure was
Bc, Cf, Cc, Df, Dc, Ef, and Ec, in which A–E corresponded to five
achieved for RUHTCC beams through varying longitudinal rein-
different reinforcement ratios, f stood for RUHTCC series and c
forcement ratio.
denoted RC control series. RC beams had a target concrete com-
pressive strength of about 40 MPa comparable to that of UHTCC.
2. Experimental program Herein, it ought to be pointed out that in the design of test, in order
to achieve the 1% reinforcement ratio using given type of steel bar,
2.1. Test specimens the dimension of cross section for series B was further adjusted as
100 mm  180 mm. Fig. 1 illustrates the configuration of tested
A total of 10 beams, five different series, were tested. In each beam geometry and reinforcement details. The details of test spec-
series, two beams were fabricated, where one was RUHTCC beam imens are summarized in Table 1, including specimen dimension
and the other was RC counterpart specimen. All beams were (l  b  h), cover thickness (c), the shear span-to-effective depth
preliminary designed to have the same size of cross section ratio (a/d), reinforcement ratio (q), area of longitudinal steel bar
120  180 mm as well as cover thickness of 25 mm. Web reinforce- (As), and critical reinforcement ratio (qb).
ment was not provided for all specimens.
The main variable between different series was the longitudinal
reinforcement ratio, the up bound of which was slightly lower than
critical reinforcement value (qb) that is estimated according to Eqs. 2.2. Material properties and specimen fabrication
(1) and (2) suggested by Xu and Zhang [12] based on sectional
analysis of beam, The mix proportion of concrete used in this test was cement:-
water:sand:coarse aggregate = 1:0.535:1.483:2.755. For UHTCC
f 4 rtc rtc h material, the main compositions consisted of cement, water, fine
qb ¼ 0:85nb c þ nb  ð1Þ
fy 3 fy fy d sand, fly ash, mineral addition, and PVA fiber. Type I Portland
cement was used. The fraction of fiber was 2% in volume. The basic
0:75 parameters of PVA fiber are 40 lm in diameter, 12 mm in length.
nb ¼ f
ð2Þ
1 þ ecpyEs This type of fiber has an elastic modulus of 40 GPa and a tensile
strength of 1600 MPa. In the process of mixing, dry composites
where fc, rtc, and ecp are the axial compressive strength, the tensile were first mixed for about 2 min and then the water was added
cracking strength and the ultimate compression strain of UHTCC; nb slowly while keeping mixing until the mixture was blended into
is relative limited compression zone depth; fy and Es are the tensile a paste. Subsequently, the superplasticizer was added to obtain
yielding strength and elastic modulus of the steel bar; h and d are good workability in the fresh composite. Finally, fiber was added
178 S. Xu et al. / Engineering Structures 39 (2012) 176–186

Table 1
Details of tested specimens.

Series l  b  h (mm) c (mm) a/d q (%) As (mm2) qb (%) fy (MPa) fu (MPa) fcu (MPa) fr (MPa) rfc (MPa)
Af 1150  120  180 25 3.02 0.67 113.04 5.57 £12:347 £12:497 46.81 – 4.05
Ac 1150  120  180 25 3.02 0.67 113.04 5.16 43.25 4.74 –
Bf 1150  100  180 25 3.04 1.00 153.86 5.03 £14:384 £14:532 46.81 – 4.05
Bc 1150  100  180 25 3.04 1.00 153.86 4.50 43.25 4.74 –
Cf 1150  120  180 25 3.04 1.73 307.72 4.96 46.22 – 3.75
Cc 1150  120  180 25 3.04 1.73 307.72 4.71 45.25 5.37 –
Df 1150  120  180 25 3.06 2.28 200.96 4.99 £16:381 £16:535 46.22 – 3.75
Dc 1150  120  180 25 3.06 2.28 200.96 5.06 48.08 5.50 –
Ef 1150  120  180 25 3.08 3.25 568.34 4.87 £18:392 £18:554 46.39 – 3.95
Ec 1150  120  180 25 3.08 3.25 568.34 4.66 £20:336 £20:520 48.08 5.50 –

qb for RUHTCC beam is calculated by substituting the material parameters practically measured in the test into Eq. (1); qb for RC beams is calculated according to the equation
given in reference [2]; fcu of concrete is obtained from the 100 mm cube compressive strength multiplying a coefficient of 0.95.

Fig. 3. The test apparatus and loading set-up.

Fig. 2. The measured uniaxial tensile stress–strain curves of UHTCC.


the measurement error. Fig. 3 shows the test apparatus and loading
set-up. The beams were simply supported at a clear span of
by hand while mixing. The overall mixing was stopped when the 900 mm, which corresponded to a shear span-to-effective depth
uniform fiber dispersion was observed. ratio (a/d) of 3.02–3.08.
The beams were cast in plywood molds. Before initial setting of The present test was controlled manually by displacement. For
concrete and UHTCC, the surface polish finishing was performed. beams with yielding of reinforcement, the displacement rate was
Then the specimens were covered by plastic thin membrane to taken as 0.12–0.15 mm/min before yielding, and then a higher rate
avoid the loss of water. After the formwork was removed, the of 0.30–0.40 mm/min was employed after yielding. For beams with-
beams were cured under humid conditions. out yielding of reinforcement, the same rate as 0.12–0.15 mm/min
While casting each batch of beam, three 100 mm cubes and three was also adopted during whole loading. At the ascending stage of
70.7 mm cubes were simultaneously cast to separately evaluate the load, when the load of the integer times of 3–5 kN was reached, it
compressive strength (fcu) of concrete and UHTCC. Three thin plates was retained for about 3–5 min to observe crack development.
with dimension of 350  50  15 mm were made to test the tension The cracking pattern was then drawn along the trajectory of crack
property of UHTCC. One set of 400  100  100 mm beams of con- growth, and crack width was measured using crack observation
crete was cast to obtain modulus of rupture (fr), and the other set apparatus with a precision of 0.02 mm.
of UHTCC beams was used to measure its flexural cracking strength Four linear variable differential transformers (LVDTs) which
(rfc). The uniaxial tensile stress–strain curves of UHTCC obtained were fixed on both sides of mid-span and two supports were used
from direct tension tests are shown in Fig. 2. to measure the deformation of beam. The net deflection at mid-
The deformed steel bars were used, with a diameter of £12, span location of beam was eventually determined as the difference
£14, £16, £18, and £20. The yielding strength (fy) and ultimate between the mid-span displacement and the settlement at
tensile strength (fu) were obtained by tension test. The yielding supports. To record the strain in longitudinal reinforced steel, an
strain of these types of steel bar, which was measured experimen- electric resistance strain gauge was stuck on the surface at mid-
tally by electric resistance strain gauge stuck on their surface, was span of each steel bar. All readings were continuously collected
shown to be about 2000–2400 lm/m. For the sake of the conserva- by IMC data-acquisition systems until the failure of beam.
tive consideration, the yielding strain ey was uniformly taken as
2150 lm/m in the below calculation of this paper. The elastic mod-
3. Experimental results and discussion
ulus Es of steel bar was correspondingly calculated as fy/ey based on
the linear behavior of reinforcement before yielding. All of the
3.1. General behavior of beams
material properties measured are summarized in Table 1.
Fig. 4 shows the load–deflection curves of RUHTCC and RC
2.3. Test setup and instrumentation beams with different reinforcement ratios, in which the width of
series B was scaled up from 100 mm to 120 mm only for carrying
A concentrated load was applied at the center of the beams out a convenient comparison between them. It is well known from
using an electro-hydraulic servo-controlled testing machine with beam theory that in general, when the depth of beam keeps the
the maximum capacity of 10,000 kN. A load cell with the allowable same, the increase in width of beam would yield a proportionally
capacity of 300 kN was connected to loading machine to minimize increased load-carrying capacity but almost consistent flexural
S. Xu et al. / Engineering Structures 39 (2012) 176–186 179

(a) (b) (c)

(d) (e) (f)

Fig. 4. Load–deflection curves of beams with different reinforcement ratios: (a) series A; (b) series B; (c) series C; (d) series D; (e) series E; (f) comparison between RUHTCC
series.

deformation. Based on this, the tested load for series B was corre- where l is the length of span; Ig is the moment of inertia of the gross
spondingly scaled up to 1.2 times while keeping the deflection un- section; Mcr is cracking moment and can be calculated according to
changed. From this figure, for the series A and B of RC control ACI code [21]; and yt is the distance from the extreme tension fiber
specimens, the apparent yielding deformation stage was pre- to the neutral axis. For RUHTCC beams, fr is replaced by the flexural
sented, whereas the brittle shear failure took place for series C, cracking strength rfc. For calculating the cracking moment more
D, and E of RC control specimens. By contrast, only Ef beam with accurately, Ig in Eq. (3) is replaced with moment of inertia of un-
the reinforcement ratio of 3.25% suffered a typical shear failure cracked transformed section (Iut). In calculation of Iut, the elastic mod-
while the remaining series of RUHTCC beams all showed apparent ulus of UHTCC (Eu) was taken as 4.012(fcu)0.40 based on the studies on
deflection hardening stage after yielding. Fig. 4f shows a compari- the tension and compression properties of UHTCC [19,22].
son between RUHTCC beams with different reinforcement ratios. It Comparison between the experimental values and the com-
can be seen that the RUHTCC beam with heavy reinforcements had puted ones, as summarized in Table 2, indicated that the measured
a higher load-carrying ability but a lower ultimate deflection. values were larger than the calculated ones for RUHTCC beams
Three failure modes were presented according to Fig. 4 and the whereas the contrary tendency was noticed for RC beams. The pos-
experimental observation. The first was typical flexural failure, sible reason was that the micro-cracks that appeared in RUHTCC
characterized by the crushing of matrix material in compression beams at the early loading stage were not wide enough to be found
zone after the reinforcement yields, i.e., beams Ac, Af, Bf, and Cf. by crack observing apparatus.
The second was typical shear failure with inclined shear principal
crack but without yielding of reinforcement, i.e., beams Cc, Dc, Ec
and Ef. The last was flexure-shear failure, i.e., beams Bc and Df, pre- 3.2.2. Maximum deflection at service load
senting that the finial failure occurred by the propagation of prin- Deflection is considered as a main control parameter at normal
cipal diagonal crack although the marked yielding plateau and the service stage [2]. The short-term deflection have been extensively
apparent propagation of flexural cracks after yielding were experi- investigated with respect to RC beam [23,24], reinforced FRC beam
enced. Therefore, the evolution process from typical flexural failure [25], and concrete beam reinforced by FRP rebar [26], etc. In these
to flexural-shear failure to shear failure was presented for both RC studies, many theoretical methods for estimating the short-term
beams and RUHTCC beams. deflection at service load were proposed, in which the tension stiff-
ness effect of concrete between cracks was taken into account.
3.2. Flexural behaviors of RUHTCC beams In general, the structures always work under the normal service
condition at which the maximum load level is often limited to
3.2.1. Flexural cracking load about 50–70% ultimate load. According to this limitation, the max-
Flexural cracking load was both experimentally and theoreti- imum load level under service condition was considered to be
cally determined in this analysis. For the experimental flexural equal to the tested ultimate load Pu divided by a factor of 1.7 in
cracking load Pfcr,exp, it was defined as the level of load at which the reference [27]. Thus, the experimental maximum mid-span
the first bending crack was found by crack observation apparatus. deflection dser,exp corresponding to this maximum load level at ser-
For the theoretical flexural cracking load Pfcr,cal, it was predicted vice stage can be determined for RUHTCC beams failing in flexure
using the following expression, and flexure-shear, as listed in Table 3.
Further, the maximum allowable deflection at service stage can
4 4 fr Ig be analytically calculated based on elastic bending deformation
Pfcr;cal ¼ M cr ¼  ð3Þ
l l yt equation as follows:
180 S. Xu et al. / Engineering Structures 39 (2012) 176–186

Table 2
Summary of experimental results.

Series Pfcr,cal Pfcr,exp Pscr Yielding load Py Yielding deflection dy Ultimate load Pu Ultimate deflection du Ductility Failure
(kN) (kN) (kN) (kN) (mm) (kN) (mm) index mode
Af 12.56 14.42 – 43.53 2.27 44.10 22.68 9.99 Flexure
Ac 14.36 13.51 – 33.18 1.43 39.10 11.28 7.89 Flexure
Bf 13.24 16.65 – 62.40 2.65 69.73 19.12 7.22 Flexure
Bc 15.61 14.43 45.62 43.42 1.75 49.97 8.21 4.19 Flexure-
shear
Cf 13.21 16.18 51.50 85.42 3.00 95.40 18.66 6.22 Flexure
Cc 17.56 16.21 51.54 – – 51.54 1.53 – Shear
Df 13.78 18.55 55.50 110.12 4.37 116.77 10.25 2.35 Flexure-
shear
Dc 18.55 17.27 56.14 – – 66.84 3.89 – Shear
Ef 15.37 18.62 53.00 – – 133.20 6.49 – Shear
Ec 19.51 18.74 61.40 – – 69.58 3.36 – Shear

Table 3
Maximum deflection at service load for RUHTCC beams failing in flexure and flexure-shear.

Ie1 (mm4) Ie2 (mm4) dser,exp (mm) dser,1 (mm) dser,1/dser,exp dser;1 ðmmÞ dser;1 =dser;exp dser,2 (mm) dser,2/dser,exp

Af 1.956E+07 1.956E+07 1.175 1.181 1.005 1.479 1.259 1.181 1.005


Bf 2.012E+07 2.012E+07 1.405 1.380 0.982 1.463 1.041 1.380 0.982
Cf 3.233E+07 2.774E+07 1.622 1.420 0.875 1.440 0.888 1.655 1.020
Df 3.704E+07 2.846E+07 1.883 1.515 0.805 1.526 0.810 1.972 1.047
Mean 0.917 1.000 1.014
C.V. 0.102 0.198 0.027

2
Ma l in which, Ec is the elastic modulus of concrete; ec(x) is compressive
dser;cal ¼ ð4Þ strain distribution along the beam depth; etop is the compressive
12EIe
strain at extreme compression fiber. Based on the plane section
in which Ma is the applied maximum (mid-span) moment at the assumption, the following relationship can be obtained,
time when the deflection is calculated; E is the elastic modulus of
matrix material (concrete or UHTCC); Ie is the effective moment etop c
¼ ð10Þ
of inertia specified in ACI code and is written as es d  c
 3  3  3 # "
M cr M cr M cr Substituting Eq. (10) into Eq. (9) and removing the same factor,
Ie ¼ Icr þ ðIg  Icr Þ ¼ Ig þ 1  Icr ð5Þ Eq. (9) can then be transformed as Eq. (7). That is, the neutral axis
Ma Ma Ma
depth determined using Eq. (7) which is derived based on the equi-
in which Ig is the moment of inertia of gross section as mentioned librium of moment of area between compression zone and tension
above; Icr is the moment of inertia of fully cracked section. Herein, zone is actually equivalent to that determined using Eq. (9) which
Ig is replaced with the uncracked transformed moment of inertia is obtained based on the internal force equilibrium together with
(Iut) to consider the influence of reinforcing bar. For RC beams, the the linear compressive stress distribution along the beam depth.
traditional theoretical definition of Icr can be expressed as [23] In the case of RUHTCC beam, UHTCC in tension region can still
3 carry the tensile stress before the formation of the local main crack.
bc
Icr ¼ þ ns As ðd  cÞ2 ð6Þ Therefore, the tension contribution of UHTCC cannot be ignored.
3 According to previous analysis with respect to RC beam, the
2 neutral axial depth of RUHTCC beam can also be derived through
bc the equilibrium of either moment of area or internal force at cross
þ ns As ðc  dÞ ¼ 0 ð7Þ
2 section. In the following analysis, the simplified bilinear tension
in which c is the neutral axis depth at mid-span section that corre- constitutive relationship, as shown in Fig. 5a, is suggested for
sponds to the distance from the extreme compression fiber to the UHTCC to model its tensile behavior. The compressive stress of
neutral axis; ns is the ratio of elastic modulus of reinforcement to UHTCC in the compression zone is still considered to vary in a
that of concrete. When neglecting the tension contribution of con- linear manner along beam depth, which is similar to that used
crete, c can be obtained by Eq. (8) according to the internal force in internal force analysis for RC beam previously. Besides, the
equilibrium at the cross section constitutive relationship of steel bar is simplified as classical elas-
Z c tic–plastic model, as shown in Fig. 5c. Based on the equilibrium of
rc ðxÞbdh ¼ es Es As ð8Þ moment of area, the modified equation for determining the neutral
0 axial depth of RUHTCC beam is derived as
in which, Es is the elastic modulus of steel bar; rc(x) is the compres-
2 2
sive stress distribution along the beam depth; es is the strain of rein- bc bt n1 b
¼ ns As ðd  cÞ þ þ ðh  c  tÞðh  c þ tÞ ð11Þ
forcing bar. If assuming that there is a linear distribution of rc(x) at 2 2 2
full cracked section, Eq. (8) can be further rewritten as
Z c
in which, n1 is the ratio of E1u to Eu, E1u is the slope corresponding to
bcEc etop the tension hardening stage of UHTCC in Fig. 5a; t is the distance
Ec ec ðxÞbdh ¼ ¼ es Es As ð9Þ
0 2 from the neutral axis to end point of elastic stress stage in tension
S. Xu et al. / Engineering Structures 39 (2012) 176–186 181

(a) 5 (b) (c)


50
fy
4
40
( cu, fc)

stress /MPa
(4.2%, 4.33)

stress /MPa
3
30
2 (3/Eu, 3) 20 ( cu/3, 2fc/3)
1 10
0 0
0 1 2 3 4 5 0.0 0.1 0.2 0.3 0.4 0.5 0.6
strain /% strain /% y u

Fig. 5. Constitutive relationship of UHTCC and steel bar: (a) UHTCC in tension; (b) UHTCC in compression; (c) steel bar in tension.

region, and it can be expressed by Eq. (12) based on the plane sec- According to these theoretical analyses, corresponding to any
tion assumption given tensile strain of steel bar, the values of their neutral axis
t etc depth c and the moment of inertia I can be worked out through
¼ ð12Þ Eqs. (13) and (14). In calculation, typical RUHTCC beam failing in
d  c es
flexure are chosen, i.e., beam Af with smaller reinforcement ratio
in which, etc is the tensile cracking strain computed by rtc/Eu, where of 0.67% and beam Cf with larger reinforcement ratio of 1.73%.
rtc is cracking tensile stress and has a value of 3 MPa, as shown in Fig. 6 shows that the variations of neutral axis depth c and moment
Fig. 5a. But, based on the internal force equilibrium at cross section, of inertia I with the increase in tensile strain of steel bar for the
the equations for obtaining the neutral axial depth can be further beams Af and Cf. For observing the influence of tensile contribution
expressed as of UHTCC on c and I of RUHTCC beam, Fig. 6 also displays the re-
2 2 sults of c and I calculated by Eqs. (6), (7), and (11). It is clearly no-
bc bt n1 b ticed from Fig. 6a that the value of c obtained based on Eq. (13) is
¼ ns As ðd  cÞ þ þ btðh  c  tÞ þ ðh  c  tÞ2 ð13Þ
2 2 2 far larger than that calculated by Eq. (7), whereas Eq. (11) almost
Compared Eq. (11) with Eq. (13), it is revealed that the last item gives the identical value of c compared to Eq. (7) when the tensile
in Eq. (11) is not equivalent to the last two items in Eq. (13). In strain is above 1250 lm/m. This phenomenon demonstrates that
other words, the neutral axis depth obtained by Eq. (11) is different the tensile bearing capacity of UHTCC after cracking greatly affects
from that calculated by Eq. (13). It is well-known that the physical the neutral axis depth of cross section, and Eq. (11) derived from
significance of neutral axis location means the intersected line the equilibrium of moment of area cannot reflects such influence
between cross section and neutral layer in which no strain is actually. From Fig. 6b, it is found that moment of inertia obtained
presented. This implies that the resultant force above and below based on equilibrium of moment of area is almost the same each
the neutral axis of the cross section is equal to internal force car- other after the tensile strain of steel bar is beyond 1000 lm/m.
ried at this section. Therefore, Eq. (13) obtained based on internal However, for RUHTCC beam with smaller reinforcements, i.e.,
force equilibrium is rational in physical meaning and can be used beam Af, there is apparent difference between moment of inertia
to solve the neutral axis depth of RUHTCC beam. And then the I attained according to equilibrium of internal force and that at-
moment of inertia (I) of RUHTCC beam for the case of linear com- tained according to equilibrium of area moment.
pressive stress distribution along the depth can be obtained as Fig. 6b reveals that the moment of inertia I is always well close
3 3
to each other regardless of whether the contribution of UHTCC in
bc bt tension region to flexural stiffness of beam is taken into consider-
I¼ þ ns As ðd  cÞ2 þ
3 3 ation or not when es is larger than 750 lm/m. That implies that the
" #
ðh  c þ tÞ2 ðh  c  tÞ3 contribution of UHTCC in tension zone to flexural stiffness of beam
þ n1 b ðh  c  tÞ þ ð14Þ can be neglected and the calculation of moment of inertia can be
4 12
further simplified without big loss of accuracy.
If the influence of UHTCC in tension region on the moment of Moreover, it is also found that both c and I vary little when ten-
inertia of cracked section is small enough to be neglected, then sile strain of steel bar is beyond 2000 lm/m, presenting that based
Eq. (14) can be simplified as Eq. (6), in which c varies with the on either Eqs. (13) and (14) or Eqs. (13) and (6), the maximum
change of strain of steel bar. ratios of c and I at strain of 2400 lm/m to those at 2000 lm/m

(a) (b)

Fig. 6. Change of c and I with the increase in tensile strain of reinforcement: (a) c; (b) I.
182 S. Xu et al. / Engineering Structures 39 (2012) 176–186

are about 1.033 and 1.023, respectively. It can be therefore said of internal force equilibrium, the simplified bilinear model for
that the neutral axis depth and moment of inertia almost remain compressive stress–strain behavior of ECC proposed by Maalej
constant near the yielding stain of reinforcing bar, and yielding and Li [29], as shown in Fig. 5b, is adopted. In this model, the strain
state can be considered as fully cracked state for RUHTCC beam, (ecc) and the stress (rcc) at the turning point are assumed as ecu/3
as indicated by Ashour [23] for RC beam. and 2fc/3, respectively. The ultimate compressive strain ecu is taken
From these points, the moment of inertia of fully cracked sec- as 0.5% [12]. Therefore, based on the bilinear tensile and compres-
tion (Icr) for RUHTCC beams can be taken as moment of inertia at sive constitutive relationships of UHTCC, the moment of inertia of
yielding of beam without consideration of contribution of UHTCC fully cracked section can be obtained by following group of
in tension zone. Thus, Icr for RUHTCC beam is further simplified as equations,

3 8 2
bcy ðc uÞ2 2
Icr ¼ þ ns As ðd  cy Þ2 ð15Þ >
>
>
>
bu
2
þ buðcy  uÞ þ n2 b y 2 ¼ ns As ðd  cy Þ þ bt2 þ btðh  cy  tÞ þ n21 b ðh  cy  tÞ2
3 >
> h i
< I ¼ bu3 þ n b ðc  uÞ ðcy þuÞ2 þ ðcy uÞ3 þ n A ðd  c Þ2
cr 3 2 y 4 12 s s y

in which, cy is the neutral axis depth computed by Eq. (13) when the >
> etc t
>
>
> ey ¼ dcy
strain of steel bar reaches its yielding strain, i.e., es = ey. >
: ecc ¼ u
ey dcy
With the calculated Icr using Eq. (15) together with Eq. (13), the
ð16Þ
deflection of RUHTCC beams at service load can be further predicted
through Eq. (4) after obtaining the effective moment of inertia Ie
given by Eq. (5). The calculated maximum mid-span deflection in which, u is the distance from the neutral axis to the turning point
(dser,1) at service load is listed in Table 3. For RUHTCC beams with of compressive stress; n2 is the ratio of E2u/Eu, E2u is the slope cor-
failure mode of flexure or flexure-shear, the mean and C.V. of their responding to the stage of the second straight line in simplified
dser,1/dser,exp were 0.907 and 10.2%, respectively, which indicated bilinear compression model; ecc is the compressive strain at the
these equations based on linear compression behavior underesti- turning point in simplified bilinear stress–strain model, as shown
mated deflection of beam. Furthermore, the degree of underestima- in Fig. 5b.
tion of deflection becomes more and more evident as the The calculated effective moment of inertia (Ie2) and deflection
reinforcement ratio increased. These observations can be explained (dser,2) at maximum service load are summarized in Table 3. It is
from the following two aspects. For the cracked beam, the decrease noticed that the computed deflection dser,2 agrees well with the
in stiffness (EI) is associated with the reduction of both the moment experimental value and the mean and C.V. of dser,2/dser,exp are
of inertia and the elastic modulus. However, the change with respect merely 1.014 and 2.27%, respectively. Compared with Ie1 obtained
to elastic modulus of UHTCC, due to its nonlinear behavior of com- based on the linear elastic compression behavior, as shown in
pressive stress–strain curve, is not considered in the calculation of Table 3, Ie2 of beams Cf and Df reduced apparently due to taking
deflection according to Eq. (4). On the other hand, for the beam sub- into account the decrease in compressive elastic modulus at the
jected to one point loading, the shear deformation also contributes second linear stage. Correspondingly, the calculated deflection
to mid-span deflection to some extent and this contribution be- dser,2 of beams Cf and Df is well close to the tested value. Besides,
comes more apparent with the increase in shear force. the calculated deflection dser,2 of beams Af and Bf is still the same
Further, a full cracked beam is assumed in the calculation of the as dser,1 because the compressive stress is at the first linear stage.
deflection of RUHTCC beams, which means the effective moment of Further, the deflection at load lower than the yielding load is
inertia of a cracked beam Ie is equal to the moment of inertia of also calculated to examine the whole deflection process before
fully cracked beam Icr (Ie = Icr) [28]. The calculated deflection dser;1 the yielding of beam. The calculated curve for RUHTCC beams Af,
according to this assumption is listed in Table 3. As shown in Table Bf, and Cf that failed in flexure are shown in Fig. 7. It is shown that
3, the mean of dser;1 =dser;exp is 1.0, whereas the dispersion degree the computed deflection curves intersect with the experimental
becomes more apparent with C.V. up to 20%. Besides, it is found deflection curves and the load corresponding to this intersection
that the computed deflection dser;1 of beam Af with reinforcement point is discovered to be near the maximum service load. The com-
ratio of 0.67% was obviously larger than dser,1, while there is slight puted deflection is firstly larger than the tested value before the
difference between these two calculated deflections for the intersection point, whereas that is less than the tested value after
remaining beams. The possible reason is that the ratio of the crack- this point with the maximum error of about 15% at yielding load.
ing moment to the mid-span moment at maximum service load Therefore, the theoretical equations for calculating Icr of RUHTCC
(Mcr/Ma) for beam Af is far larger than that of other beams with beam, which are proposed based on internal force equilibrium to-
larger reinforcement ratio, and thus the second term containing gether with assumed bilinear tensile and compressive constitutive
(Ig–Icr) in Eq. (5) has a larger effect on Ie for beam Af with lower models, can predict the deflection before yielding of RUHTCC beam
reinforcement ratio than for other beams. Accordingly, such
assumption greatly weakens the stiffness of beam with low rein-
forcement ratio but has little influence on the stiffness of beams
with high reinforcement ratio.
In fact, the elastic modulus of UHTCC in compression region
gradually reduced with the increase in compressive strain. In par-
ticular, such drop is more apparent when the compressive strain is
higher than ecu/3 [29]. At yielding state, the compressive strain at
extreme compression fiber for RUHTCC beams with high reinforce-
ment ratio may be higher than ecu/3, which may result in an elastic
modulus lower than the initial value. Therefore, the fully cracked
moment of inertia Icr and the resulting effective moment of inertia
Ie are both greatly overestimated.
It is necessary to introduce the rational compressive constitu-
tive relationship to reflect the reduction of elastic modulus of Fig. 7. Comparison between calculated deflection and experimental value before
UHTCC in compression region. For the convenience in the analysis yielding.
S. Xu et al. / Engineering Structures 39 (2012) 176–186 183

with good accuracy, particularly to the deflection at maximum ser- increased in Af beam, the maximum crack opening was merely
vice load. 0.05 mm at the yielding state, which was far lower than 0.4 mm
in Ac beam. In practice, such fine cracks in RUHTCC beam can be
3.2.3. Flexural behavior after yielding considered to be harmless and therefore can adequately meet the
Comparing the results of RUHTCC beams with those of RC con- width requirement in durability design even in harsh environment.
trol beams shown in Table 2, it is seen that beams Af and Bf respec- For tension strain-hardening FRCC such as ECC and UHTCC, the
tively presented 31.2% and 41.8% higher yielding load than the past investigations [31–33] had shown that no stress concentration
corresponding control beams Ac and Bc and, their corresponding occurred on steel bar while undergoing large nonlinear deforma-
yielding deflections increased by 43.2% and 51.4%, respectively. tion because of the native strain-hardening behavior and large
In addition, the 13% and 40% improvement of beams Af and Bf over strain capacity. In other words, UHTCC was able to compatibly de-
beams Ac and Bc were shown for the ultimate loads and the corre- form with the reinforcement. Therefore, the number of fine cracks
sponding ultimate deflection also had about one time increase. was expected to increase to remain the continuous strain distribu-
Such improvements in both load carrying capacity and deforma- tion between UHTCC and reinforcement. While for concrete matrix,
tion capacity are associated with advantageous material properties once the flexural cracking initiated, the concrete would no longer
of UHTCC such as tensile strain-hardening behavior, large deforma- undertake tensile stress and the stress previously carried by con-
tion ability, high compression toughness capacity, and outstanding crete would be transferred to reinforcing bar. This stress transfer
deformation compatibility with reinforcement. From these experi- possibly resulted in a stress concentration on steel bar at the vicin-
mental results, it can be concluded that the ductile UHTCC can re- ity of crack such that an obvious jump in strain eventually oc-
place a small amount of steel bar to improve ductility of structural curred. As the load was further applied, such deformation
members while remaining similar ultimate load capacity. incompatibility between steel bar and concrete resulted in an in-
It can be also observed from Table 2 that, the RUHTCC beam crease in the crack width.
with heavy reinforcement ratio had higher yielding load, yielding With the increase in deformation after yielding, the crack
deflection and ultimate load but lower ultimate deflection and opened rapidly in RC beam, while a large number of fine cracks
ductility. With respect to displacement ductility index [27] that with a tight spacing appeared around the mid-span zone in
is the ratio of deflection at ultimate state (du) to that at yielding RUHTCC beam, as shown in Fig. 8c. Moreover, it was observed in
state (dy), RUHTCC beams failing in flexure showed a value ranging RUHTCC beam that the crack width hardly increased during the
from 6.22 to 9.99 which is far larger than the target value of 3–5 stage from yielding to the formation of main crack, with the max-
required in seismic design [30]. However, for RUHTCC beam Df that imum value of about 0.1 mm. After the beam reached yielding, the
showed flexure-shear failure mode, although apparent deflection strain in reinforcing bar at the maximum moment location drasti-
hardening stage was presented, its displacement ductility index cally increased. Due to deformation compatibility between UHTCC
was only 2.35, which is lower than 3. From this point, it is advised and steel bar, the same amount of strain increment occurred in
that in practical use some web reinforcements ought to be installed UHTCC. This resulted in the appearance of more fine cracks around
to effectively restrain the shear deformation such that the the mid-span region of RUHTCC beam. Consequently, the multiple-
improved ultimate deformation capacity is obtained to ensure cracking behavior under bending load is an intrinsic property of
the ductility requirement in seismic design. UHTCC, which is independent of the loading manner.
For the RUHTCC beam that used low reinforcement ratio, like After the main bending crack formed, it was observed for Ac
beams Af and Bf, the load was seen to slightly go up after yielding beam that, when the width of the first main crack grew up to about
as deflection was further imposed (Fig. 4a and b). When the main 5 mm, the second main crack rapidly opened while keeping the
flexural crack formed and went through the whole bottom surface constant width for first main crack. After the second major crack
of beam, a slight drop in load occurred. Subsequently, as the major extended into the compression region, horizontal layered cracks
crack slowly extended upwards toward the compression region of appeared, resulting in the crushing in compression region. But
beam, the load either again went up slightly or kept a constant up for RUHTCC beams Af, Bf and Cf, once the main bending crack
to the crushing in compression zone. However, for other RUHTCC formed, it was just this crack that propagated locally and no new
beams with larger reinforcement ratio, like beams Cf and Df, no main crack developed until the crushing in compression region
drop in load was observed even when main flexural cracking occurred. In this stage, the UHTCC didn’t deform compatibly with
formed after yielding. The possible reason is that for RUHTCC steel bar, and the stress carried by UHTCC was partly transferred
beams containing a great amount of reinforcement, the moment to the reinforcement. And thus the yielding reinforcing bar with
at a cross section is dominated by longitudinal steel bar in tension rather low elastic modulus produced a large deformation locally,
region and UHTCC in compression region, and thus during propa- which in turn led to the further development of only one main
gation of main bending crack, the gradual loss of tension capacity crack.
of UHTCC has little influence on load bearing capacity.
3.3. Shear behaviors of RUHTCC beams
3.2.4. Crack development in flexure
The flexural crack development during loading was similar for 3.3.1. Shear cracking load
different reinforcement ratio of RUHTCC beams that failed in flex- For RC slender beam (a/d P 2.5), it is generally accepted that
ure, i.e., beams Af, Bf and Cf. For investigating the superior crack the shear cracking always appears suddenly above the longitudinal
control capacity of UHTCC to plain concrete, the cracking patterns steel bar and near the outmost flexural crack [34]. This crack ex-
of RUHTCC beam Af and RC beam Ac that had the same reinforce- tends upwards along the inclined flexural crack and at the same
ment ratio of 0.67% were carefully compared. Fig. 8 shows their time extends downwards along the longitudinal steel bar, which
typical crack development at yielding, together with crack devel- is different from an ordinary flexural-shear diagonal crack in shear
opment in beam Af when approaching the appearance of main span zone. But for RUHTCC member, the first inclined shear crack
flexural crack. It is evident that the cracking behaviors between occurred either in the middle part of shear-span region or near
them differed significantly. At the yielding state, only 3 visible the flexural cracks, and extended both upwards towards the load-
cracks developed around mid-span region in Ac beam, whereas ing point and downwards towards support point.
there were about 12 cracks in Af beam due to the multiple cracking The experimental shear cracking load Pscr is listed in Table 2. A
property of UHTCC. Because the number of cracks markedly similar amount of shear cracking load was observed between RC
184 S. Xu et al. / Engineering Structures 39 (2012) 176–186

Fig. 8. Typical flexural crack pattern around the mid-span region: (a) Ac beam at yielding; (b) Af beam at yielding; (c) Af beam when approaching the formation of main
bending cracking.

beams and RUHTCC beams. And, it was noticed that the shear with steel bar. During loading, the stress concentration on rein-
cracking load increased slightly for RC beams with the increase forcement near the crack was eliminated, which prevented the for-
in reinforcement ratio while there were unappreciated variation mation and development of horizontal crack along the longitudinal
for RUHTCC beams. It was also observed that, for RC beams such steel bar that was considered as the major reason for the unstable
as beams Cc, Dc and Ec, the shear cracking load was close to the development of shear crack in RC slender beam [34]. Moreover, the
ultimate load, but a large reserve of shear carrying capacity after bridge action of fiber restricted the opening of shear crack and re-
shear cracking was exhibited for RUHTCC beam (see beam Ef). tarded the damage of the beam, and therefore greatly improved its
For slender RC beams without stirrup, the shear crack always integrity.
unsteadily developed shortly after it formed and rapidly pene- Comparison between the shear load capacities of RC beams fail-
trated into the compression region to result in a diagonal tension ing in shear indicated that the ultimate load increased slightly with
failure. In contrast, for slender RUHTCC beams, the numerous fine the increase in reinforcement ratio. Beam Ec had the maximum
diagonal cracks progressively appeared after the shear cracking ultimate shear load of 69.58 kN. However, it was found from Table
and subsequently developed in a stable manner. No sudden open- 2 that, compared to beam Ec, beam Ef had improved ultimate load
ing of diagonal cracks took place during the stable development. up to 133.20 kN, about 1.91 times that of beam Ec, and the corre-
sponding deformation in Ef beam rose by 93%. These experimental
3.3.2. Ultimate load capacity and deformation results revealed that UHTCC presented higher shear resistance
Comparing the load–deflection curves shown in Fig. 4, beams capacity than plain concrete with similar compressive strength,
Dc and Ec both underwent three load peaks, whereas beam Cc which further confirmed the findings by Li et al. [14]. This implied
had only one load peak. This indicated that a more serious diagonal that for the tension strain-hardening material such as UHTCC, the
tension failure took place in beam Cc. During testing, the principal conventional relationship between the compressive strength and
shear crack only at one side of beam Cc was observed to extend shear strength suggested in ACI code was no longer suitable to pre-
into the compression region, while the shear cracks in beams Dc dict its shear strength.
and Ec were observed to develop alternately in right and left shear But for beams failing in flexure-shear, their ultimate load level
span before the critical diagonal crack formed. Fig. 4f revealed that was closely related to the yielding of reinforcing bar and therefore
RUHTCC beam Ef also produced a shear load–deflection curve with cannot reflect the actual shear capacity of matrix material. Besides,
only one load peak. But, the failure process in RUHTCC beam Ef dis- a large mid-span deformation was usually undergone due to the
tinguished in essence from that in beam Cc. In contrast to the sud- yielding of reinforcement. For example, as listed in Table 2, beams
den separation into two parts occurred in beam Cc, the stable crack Df and Bc separately experienced the mid-span deflection about
development mode was presented in beam Ef from the first shear 91% and 73% that of beam Ac that failed in flexure.
cracking to the ultimate failure. It can be attributed to the substan- From these points, the application of UHTCC instead of concrete
tial ductile deformation of UHTCC, the bridging action of fiber in structural member is expected to be able to save a part of web
across crack in matrix and its excellent deformation compatibility reinforcements required in some key locations of structures, i.e.,
S. Xu et al. / Engineering Structures 39 (2012) 176–186 185

Fig. 9. The cracking pattern in beam Df at yielding.

(a)

(b) Fig. 11. The local shear cracking pattern in Ef beam when reaching ultimate load.

Fig. 10. The typical shear crack development in RUHTCC (Ef) and RC (Ec) beam: (a)
RUHTCC beam; (b) RC beam.

beam-column joint, column base, and beam ends etc., and thus to
well ensure casting quality of matrix material.

3.3.3. Diagonal cracking pattern


Fig. 9 shows the crack pattern of Df beam at yielding state. It
was found that lots of diagonal cracks developed in shear-span
zone, with maximum crack width of roughly 0.13 mm. After yield-
ing, the speed of extension of inclined cracks slowed down,
whereas the number of flexural cracks continued to increase rap- Fig. 12. The evolution of diagonal crack width in beams Ec and Ef.
idly while keeping an almost unchanged width. With the further
increase in deformation, a local bulge appeared under the loading
point, resulting in a fast propagation of diagonal crack. At last, a 4. Conclusions
principal diagonal crack rapidly propagated in the middle of sec-
tion and then penetrated into the compression region. In this paper, the experimental research on flexural and shear
Fig. 10 shows the typical shear crack development in RC beam behaviors of steel reinforced ultrahigh toughness cementitious
and RUHTCC beam. It was found that a single shear crack, consist- composites (RUHTCC) beams was performed through the bending
ing of an inclined branch towards loading point and a horizontal test of beam with five different longitudinal steel ratios under
branch along the longitudinal steel bar, appeared in each shear one concentrated load. With the increase in reinforcement ratio,
span for RC beam. However, multiple fine diagonal cracks that three failure modes were presented for both RC and RUHTCC
almost distributed evenly in the overall shear span zone was beams, including flexure, flexure-shear, and shear. Load carrying
exhibited for RUHTCC beam. Fig. 11 shows the local diagonal crack capacity, deformation and cracking pattern were discussed and
pattern observed in Ef beam when the ultimate load was reached. compared between RUHTCC and RC beam. The following conclu-
A total of 25 cracks almost evenly distributed within the region of sions can be drawn:
about two grids (about 150 mm), which further showed excellent
crack dispersion capacity of UHTCC. (1) The flexural performances of RUHTCC beam under one point
It was also observed that, unlike RC beam, critical diagonal loading at mid-span were demonstrated to be comparable to
crack in RUHTCC beam developed almost along the line from the that under two symmetrical loading, presenting about
support to the loading point. It seems for RUHTCC beam that the 0.05 mm in crack width at yielding and the multiple cracking
splitting failure along inclined section connecting the support loca- pattern around mid-span region after yielding. This means
tion and the loading point location is main mechanism of shear that the flexural behavior of RUHTCC beams is independent
failure. of loading manner. Compared with the corresponding RC
Fig. 12 presents the maximum diagonal crack width at each control beam, the load-carrying capacity, cracking pattern,
load increment for Ec and Ef beams. It was noticed that the deformation and ductility ability of RUHTCC beams were
maximum crack width in Ec beam was only 0.14 mm when markedly improved.
approaching first peak load, but it suddenly went up to about (2) Based on the internal force equilibrium, the theoretical
0.4 mm when the shear cracking occurred. Then, the diagonal crack equations for calculating the moment of inertia of fully
opened rapidly up to 1.2 mm in width, accompanied by the up and cracked section (Icr) of RUHTCC beam is proposed, in which
down fluctuation of load in this process. For Ef beam, it was found the contribution of UHTCC in tension zone to the moment
from Fig. 12 that the crack opened in a stable manner. The crack of inertia is neglected without apparent loss of accuracy.
width first rapidly grew up to about 0.08 mm at the load of According to the proposed expressions based on bilinear
68 kN and then slowly increased to 0.18 mm at the ultimate load. tensile and compression constitutive model, the predicted
At the load level which is the same as the ultimate load of beam Ec, maximum deflection at service load agrees well with tested
the crack width was only 0.08 mm. value for RUHTCC beams failing in flexure and flexure-shear.
186 S. Xu et al. / Engineering Structures 39 (2012) 176–186

(3) In the case of beams suffering flexure-shear failure, the yield- [10] Li VC, Wang S, Wu C. Tensile strain-hardening behavior of PVA-ECC. ACI Mater
J 2001;98(6):483–92.
ing of reinforcing bar occurred prior to the principal diagonal
[11] Xu SL, Li HD. A review on the development of research and application of ultra
cracking, which yielded a large deformation in both RC beam high toughness cementitious composites. China Civil Eng J 2008;41(6):45–60
and RUHTCC beam. But, the deformation of RUHTCC beam [in Chinese].
still is far larger than that in RC beam. Particularly, RUHTCC [12] Xu SL, Zhang XF. Theoretical analysis and experimental investigation on
flexural performance of steel reinforced ultrahigh toughness cementitious
beam with 2.28% reinforcement ratio presented a deforma- composite (RUHTCC) beams. Sci China Ser E – Technol Sci
tion capacity close to that of RC beam with 0.67% longitudinal 2009;52(4):1068–89.
steel bar ratio. And, multiple diagonal cracks were observed [13] Li VC, Wang S. Flexural behaviors of glass fiber-reinforced polymer (GFRP)
reinforced engineered cementitious composite beams. ACI Mater J
in RUHTCC beam. 2002;99(1):11–21.
(4) With respect to RUHTCC beam failing in shear, the shear car- [14] Li VC, Mishra DK, Naaman AE, Wight JK, et al. On the shear behavior of
rying load and shear deformation were respectively about engineered cementitious composites. Adv Cement Base Mater
1994;1(3):142–9.
1.91 and 1.93 times those of RC comparison beam although [15] Shimizu K, Kanakubo T, Kanda T, Nagai S. Shear behavior of steel reinforced
they had similar matrix strength. A large reserve in shear PVA-ECC beams. In: 13th World conference on earthquake engineering,
strength after the shear cracking was observed and numer- Vancouver, BC, Canada, paper no. 704; 2004.
[16] Li QH, Xu SL. Experimental investigation and analysis on flexural performance
ous diagonal cracks with a very narrow spacing appeared of functionally graded composite beam crack-controlled by ultrahigh
in shear-span zone, too. The horizontal crack along the rein- toughness cementitious composites. Sci China Ser E – Technol Sci
forcement observed in reference RC beam was effectively 2009;52(6):1648–64.
[17] Kim JJ, Lim YM, Won J, Park H, Lee K. Shear capacity and failure behavior of
eliminated in tested RUHTCC beams due to the tensile
DFRCC repaired RC beams at tensile region. Eng Struct 2007;29(1):121–31.
strain-hardening behavior of UHTCC and the deformation [18] Maalej M, Lin VWJ, Nguyenb MP, Quek ST. Engineered cementitious
compatibility between UHTCC and the steel bar, which composites for effective strengthening of unreinforced masonry walls. Eng
ensured the stable extension of shear cracks in RUHTCC Struct 2010;32(8):2432–9.
[19] Xu SL, Cai XR. Experimental study and theoretical models on compressive
beam and prevented jump in load and crack width during properties of ultra high toughness cementitious composites. ASCE J Mater Civil
loading. Eng 2010;22(10):1067–77.
[20] Cheng WX, Yan DH, Wang TC, Jiang JJ. Reinforced concrete
structure. Beijing: China Architecture & Building Press; 2008.
[21] ACI 318-02. Building code requirements for reinforced
concrete. Detroit: American Concrete Institute; 2002.
Acknowledgements
[22] Li HD. Experimental research on ultra high toughness cementitious
composites. Doctoral dissertation, Dalian University of Technology, Dalian,
This present work was financially supported by the Key Science China; 2008.
and Technology Innovation Team of Zhejiang Province (No. [23] Ashour SA. Effect of compressive strength and tensile reinforcement ratio on
flexural behavior of high-strength concrete beams. Eng Struct
2010R50034), the Research Fund for the Doctoral Program of 2000;22(5):413–23.
Higher Education (No. 20100101120058), and the Program of the [24] Stramandinolia RSB, La Rovereb HL. An efficient tension-stiffening model for
National Natural Science Foundation of China for Youth (No. nonlinear analysis of reinforced concrete members. Eng Struct
2008;30(7):2069–80.
50908029). [25] Ashour SA, Wafa FF, Kamal MI. Effect of the concrete compressive strength and
tensile reinforcement ratio on the flexural behavior of fibrous concrete beams.
Eng Struct 2000;22(9):1145–58.
References [26] Issa MS, Metwally IM, Elzeiny SM. Influence of fibers on flexural behavior and
ductility of concrete beams reinforced with GFRP rebars. Eng Struct
[1] GB/T 50476-2008. Code for durability design of concrete 2011;33(5):1754–63.
structures. Beijing: China Architecture & Building Press; 2008 [in Chinese]. [27] Rashid MA, Mansur MA. Reinforced high-strength concrete beams in flexure.
[2] GB 50010-2002. Code for design of concrete structures. Beijing: China ACI Struct J 2005;102(3):462–71.
Architecture & Building Press; 2002 [in Chinese]. [28] Paster JA, Nilson AH, Slate FO. Behavior of high-strength concrete beams.
[3] Narayanan R, Darwish IYS. Use of steel fibers as shear reinforcement. ACI Struct Research report No. 84-3, Cornell University, Ithaca, NY (USA); 1984. p. 311.
J 1987;84(3):216–27. [29] Maalej M, Li VC. Flexural/tensile-strength ratio in engineered cementitious
[4] Kwak Y, Eberhard MO, Kim W, Kim J. Shear strength of steel fiber-reinforced composites. ASCE J Mater Civil Eng 1994;6(4):513–28.
concretes beams without stirrups. ACI Struct J 2002;99(4):530–8. [30] Shuaib HA, Ray B. Flexural behavior of reinforced high-strength lightweight
[5] Qian CX, Patnaikuni I. Properties of high-strength steel fiber-reinforced concrete beams. ACI Struct J 1991;88(1):69–77.
concrete beams in bending. Cem Concr Compos 1999;21(1):73–81. [31] Fischer G, Li VC. Influence of matrix ductility on tension-stiffening behavior of
[6] Kim DJ, Naaman AE, El-Tawil S. Comparative flexural behavior of four fiber steel reinforced engineered cementitious composites (ECC). ACI Struct J
reinforced cementitious composites. Cem Concr Compos 2008;30(10):917–28. 2002;99(1):104–11.
[7] Blunt JD, Ostertag CP. Deflection hardening and workability of hybrid fiber [32] Otsuka K, Mihashi H, Kiyota M, Mori S, Kawamata A. Observation of multiple
composites. ACI Mater J 2009;106(3):265–72. cracking in hybrid FRCC at micro and Meso levels. J Adv Concr Technol
[8] Andrzej MB. Fiber reinforced cement-based (FRC) composites after over 40 2003;1(3):291–8.
years of development in building and civil engineering. Compos Struct [33] Zhang XF, Xu SL, Hou LJ. Improvement on flexural and cracking behavior of RC
2008;86:3–9. beam using ultra-high toughness cementitious composite II: experimental
[9] Li HD, Xu SL, Leung CKY. Tensile and flexural properties of ultra high toughness study. China Civil Eng J 2009;42(10):53–66 [in Chinese].
cementitious composite. J Wuhan Univ Technol – Mater Sci Ed [34] Kim W, White RN. Shear-critical cracking in slender reinforced concrete
2009;24(4):677–83. beams. ACI Struct J 1999;96(5):757–65.

You might also like