Download as pdf or txt
Download as pdf or txt
You are on page 1of 14

Engineering Structures 199 (2019) 109576

Contents lists available at ScienceDirect

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

Static and fatigue performance of reinforced concrete beam strengthened T


with strain-hardening fiber-reinforced cementitious composite

Bo-Tao Huanga, Qing-Hua Lia, , Shi-Lang Xua, Li Zhangb
a
Institute of Advanced Engineering Structures and Materials, Zhejiang University, Hangzhou 310058, China
b
North China Power Engineering Co., Ltd, 110120 Beijing, China

A R T I C LE I N FO A B S T R A C T

Keywords: The static and fatigue performance of reinforced concrete beams strengthened by strain-hardening fiber-re-
Reinforced concrete inforced cementitious composite is investigated. Two series of strengthened beam specimens are prepared with
Engineered cementitious composite (ECC) different thicknesses of the enhancement layer (40 mm and 50 mm), and three fatigue stress levels (0.9, 0.8, and
Strain-hardening cementitious composite 0.7) are tested. The fatigue life, mid-span deflection, and crack mode of the tested specimens are analyzed.
(SHCC)
Emphasis is placed on the fatigue response of the strain-hardening fiber-reinforced cementitious composite layer
Fatigue
Fiber reinforcement
and longitudinal reinforcements. A simplified method is proposed to model the fatigue performance of the
Composite beam composite beam. The mechanism of the fatigue enhancement of the strengthened beam compared to a con-
ventional reinforced concrete beam is as follows: (1) the enhancement layer physically contributes by taking part
of the stress in the tension zone, and (2) the enhancement layer can lower the strain localization and stress
concentration of the longitudinal reinforcements. Several methods for further improvement in the fatigue per-
formance of reinforced concrete beams are suggested.

1. Introduction using Weibull function was developed for plain concrete, fiber-re-
inforced concrete and UHTCC [14].
Reinforced concrete structures are the most widely used in modern Because of the superior fatigue performance of UHTCC compared to
infrastructures. In recent decades, with an increase in demand for high- that of conventional concrete materials, it is promising to apply UHTCC
speed transportation systems (e.g., high-speed railways and highways), to the parts of concrete structures that will undergo high fatigue stress
the concrete structures have to sustain a larger amount of traffic and strain. Zhang and Li [15] investigated the fatigue behavior in the
loadings compared to conventional transportation infrastructures. bending of UHTCC in an overlay system, and found that reflective
Hence, knowledge of fatigue performance of concrete structures is cracking failure in pavement overlays can be eliminated by using
getting more and more attention. It has been found that the inclusion of UHTCC. The static and fatigue behavior of UHTCC-layered concrete
fibers is beneficial to the fatigue performance of concrete [1], and such beams without reinforcements was investigated in Refs. [16,17]. It was
effect is much more significant in strain-hardening fiber-reinforced found that a UHTCC layer can significantly improve the bending fatigue
cementitious composites [2–7]. These micromechanically-designed ce- life of plain concrete beams. A UHTCC-layered concrete beam can
mentitious materials [8] are called engineered cementitious composites sustain fatigue loading at a larger deflection than plain concrete beam.
(ECCs) [9], strain-hardening cementitious composites (SHCCs) [10], or This is related to the effectiveness of the UHTCC layer in controlling the
ultra-high toughness cementitious composites (UHTCCs) [11]. The cracks formed in the concrete layer [16]. To date, investigations of the
flexural fatigue performance of UHTCC was first investigated in Refs. fatigue behavior of UHTCC-layered concrete structures have focused on
[2,12,13]. It was found that fatigue life and deformation capacity of unreinforced concrete members of small scale.
UHTCC are significantly larger than those of conventional concrete, For reinforced concrete members, the use of UHTCC was first pro-
polymer cement mortar, and steel-fiber-reinforced concrete. The com- posed for enhanced durability [20]. It is known that UHTCC can ef-
pressive fatigue performance of UHTCC is also superior to that of fectively control the crack width (generally below 50 μm [11]), whose
conventional plain and fiber-reinforced concrete [4,5]. For tensile fa- effect on both water permeability and chloride diffusivity can be
tigue performance, the cracked UHTCCs can still sustain a large amount minimal [18,19]. Hence, UHTCC can be employed as a protective layer
of loading cycles [6]. Recently, a succinct fatigue deformation model for steel reinforcements to achieve a high durability of reinforced


Corresponding author.
E-mail addresses: botaohuang@zju.edu.cn (B.-T. Huang), liqinghua@zju.edu.cn (Q.-H. Li), slxu@zju.edu.cn (S.-L. Xu), zhangli@ncpe.com.cn (L. Zhang).

https://doi.org/10.1016/j.engstruct.2019.109576
Received 6 April 2019; Received in revised form 11 August 2019; Accepted 21 August 2019
0141-0296/ © 2019 Elsevier Ltd. All rights reserved.
B.-T. Huang, et al. Engineering Structures 199 (2019) 109576

concrete structures [22–25]. In transportation infrastructures, re-


inforced concrete structures are widely used, which have to sustain a
large amount of fatigue loadings. To apply UHTCC in such structures, it
is important to understand the effect of UHTCC on the fatigue behavior
of UHTCC-layered reinforced concrete elements. However, to date,
there is little work on this issue. Only a few investigations (e.g.,
[26,27]) have been carried out regarding the fatigue performance of
reinforced SFRC beams. It was found that the steel fiber prolonged the
fatigue life of the beams by taking some of the stresses in the tension
zone. Considering the superior performance of UHTCC compared to
conventional fiber-reinforced concrete, the fatigue performance of re-
inforced concrete members strengthened by a UHTCC layer is antici-
pated. In addition, knowledge of the response of UHTCC in structural
reinforced concrete members under fatigue loading is crucial for the
application of this material in practice.
The flexural fatigue behavior is important for reinforced concrete
structures subjected to cyclic loading (e.g., pavement, bridge decks, and
railway sleepers). In this study, the static and fatigue performance of
reinforced concrete beams strengthened by UHTCC layers (short for RC-
UHTCC beams) is experimentally investigated using a four-point
bending test. Two series of RC-UHTCC beam specimens are prepared
with different thicknesses of UHTCC layer. Three fatigue stress levels
(i.e., ratio of the maximum fatigue load to the ultimate static load of the
beam specimen) are tested. Conventional reinforced concrete beams Fig. 2. Photographs of the test setup: (a) entire test setup and locations of
with the same dimensions and reinforcements are also prepared and LVDT-1 & 3, (b) locations of LVDT-1 & 2, (c) location of LVDT-4.
tested. The fatigue life, mid-span deflection, and crack mode of the
tested specimens are analyzed. Emphasis is placed on the fatigue be- achieve a more economical concrete component. The distance between
havior of the UHTCC layer and the longitudinal reinforcements. A the center of the longitudinal reinforcements and the bottom of the
simplified method is also introduced to model the fatigue performance beam specimen is 30 mm (see in Fig. 1). In this case, the interface of the
of the RC-UHTCC beam. In addition, the mechanism of the fatigue UHTCC and concrete is close to the longitudinal reinforcements. In the
enhancement of the UHTCC layer is discussed, and several methods for following parts, “UH40” and “UH50” stand for the beam specimens
further improvement in the fatigue performance of RC-UHTCC beams with 40-mm and 50-mm UHTCC layers, respectively. As shown in
are suggested. Finally, relevant conclusions are drawn. Fig. 1, the thickness of UHTCC was set to zero for the RC beams, which
have the same dimensions and reinforcements as the RC-UHTCC beams.
The photographs of the test setup and the locations of the LVDTs used in
2. Experimental program
the tests are shown in Fig. 2.
The fabrication process for the RC-UHTCC beams was as follows.
2.1. Specimen preparation
First, the reinforcements were fixed in the wood formworks. Second,
concrete layers with required thicknesses (i.e., 160 mm for UH40, and
For static and fatigue tests, two series of RC-UHTCC beams and one
150 mm for UH50) were cast. Third, after about 1 h, fresh UHTCC was
series of RC beams were prepared. Fig. 1 presents the details of the
cast to fill the upper parts of the formworks. Finally, all specimens were
beams used in the tests. All the beam specimens are of the same di-
demolded after 48 h, cured for 28 days, and laid in an ambient en-
mensions and reinforcements. The only difference between these three
vironment for three months before testing. For concrete materials, the
series of beam specimens is the thickness of the UHTCC layer. For RC-
strength increase may have an influence on the fatigue test results. In
UHTCC beams, the thicknesses of UHTCC layers of the two series spe-
this study, the specimens were kept for three months after curing, so
cimens are 40 mm and 50 mm. These two thicknesses were selected for
that the strength increase of UHTCC and concrete during fatigue test
the following reasons. It was reported that compared to a conventional
would be small and its impact on the fatigue test results can be reduced.
RC beam, an RC beam with a UHTCC layer around the longitudinal bars
It should be pointed out that for a real strengthening practice, sprayable
had a higher resistance against corrosion [21,23], and the failure pat-
UHTCC [28] can be utilized to fabricate the RC-UHTCC beams, instead
tern of an RC beam with a 50-mm UHTCC layer was insignificantly
of the above fabrication process.
changed after accelerated corrosion [23]. The thickness of the UHTCC
layer was thus set to 50 mm. Because the cost of an RC-UHTCC com-
ponent would be higher with an increase in the thickness of the UHTCC
layer, the other thickness of UHTCC in this study was set to 40 mm to

Fig. 1. Specimen details and test setup for RC and RC-UHTCC beams.

2
B.-T. Huang, et al. Engineering Structures 199 (2019) 109576

2.2. Material properties

In this study, the proportion of the UHTCC matrix was cementitious


binders:water:fine sand = 1:0.24:0.6 (by weight) [4,29]. Polyvinyl al-
cohol (PVA) fibers (12 mm in length and 40 μm in diameter) in the
matrix were 2% of the UHTCC volume. The average tensile strength of
the UHTCC (i.e., σ0-UH) was 2.8 MPa, with an average strain capacity
over 2%. The proportion of concrete in this study was as follows: ce-
ment:water:sand:coarse aggregate = 366:205:676:1151 (by weight).
The Portland cement P•O 42.5 was utilized, corresponding to the Chi-
nese standard GB175-2007. Cube concrete specimens
(150 mm × 150 mm × 150 mm) were prepared. The curing condition
and testing age of the concrete specimens were the same with those of
the UHTCC specimens. The average compressive strength of these
concrete specimens was 49.9 MPa.
As shown in Fig. 1, the diameters of the longitudinal reinforcements Fig. 3. Applied fatigue loading.
and stirrup were 12 mm and 6.5 mm, respectively. For the longitudinal
reinforcements, the yield strength was 386 MPa, the ultimate strength
was 572 MPa and the measured elastic modulus is 190 GPa. For the load in the following fatigue tests.
stirrups, the yield strength was 280 MPa, and the ultimate strength was In this study, constant-amplitude fatigue loads were used in the
480 MPa. fatigue tests, and the applied fatigue loading is presented in Fig. 3. Note
that the load ratio R (i.e., ratio of the minimum fatigue load to the
maximum fatigue load) was set to 0.2 for all the fatigue tests. The
2.3. Testing methods loading frequency was set to 2 Hz. Before the fatigue tests, twice the
preloading (0 – Pmean) was applied before each specimen was tested.
It can be seen in Fig. 1 that a four-point bending test was used to For the UH40 and UH50 beam specimens, three stress levels (i.e.,
evaluate the static and fatigue performance of the beams. Figs. 1 and 2 S = 0.9, 0.8, and 0.7) were used in the tests. The stress level is the ratio
show the test setup, and the tests were performed in a 1000-kN IN- of the maximum fatigue load to the static load capacity. For the RC
STRON system. A pair of LVDTs (LVDT-3 & 4) were used to measure the specimens, one stress level (i.e., S = 0.8) was considered.
mid-span deflection of the beam specimens. In addition, another two For the fatigue tests, the detailed loading condition, specimen ID,
LVDTs at a 320-mm gauge length (see Figs. 1 and 2) were used to and fatigue life are listed in Table 2. For each loading condition, one
measure the local deformations of the specimens. One LVDT (LVDT-1) beam specimen was tested. The “S” in the specimen ID stands for the
was attached to the bottom of the beam specimens and the other fatigue stress level. The load ranges are determined by the stress level,
(LVDT-2) was attached at the same height (i.e., 30 mm) as the long- load ratio, and static load capacity of each series of beam specimens.
itudinal reinforcements. To measure the strains of the concrete in the Note that the load ranges of UH40 and UH50 are quite close for each
compression zone, 100-mm electronic strain gauges were attached at stress level because the load capacities of UH40 and UH50 are close (see
the top of the specimens. Moreover, strain gauges were also used to Table 1). In addition, the load range of RC-S0.8 is close to those of
measure the strains of longitudinal reinforcements at mid-span (see UH40-S0.7 and UH50-S0.7.
Fig. 1), but most of these gauges failed during the fatigue tests. The
procedure for placing strain gauges on steel reinforcements was as 3. Results and discussion
follows: (1) a small area of the reinforcement was ground and cleaned;
(2) a gauge was attached at the ground area using cyanoacrylate ad- 3.1. Static tests
hesive; (3) insulating adhesive and electric tape were used to protect
the gauge and line; (4) a piece of textile with uncured epoxy was used to The load-vs.-mid-span deflection curves of the three series of beam
wrap this area; (5) sand was used to coat the wrapped area to improve specimens are shown in Fig. 4. Fig. 4 shows that the average yield loads
the bond; (6) the epoxy was cured. of Series RC, UH40, and UH50 are 53.5 kN, 62.4 kN, and 64.6 kN,
The static tests for the three series of beam specimens (i.e., RC, respectively. This phenomenon is related to the tensile strain-hardening
UH40, and UH50) were performed under a displacement control of behavior of UHTCC. The tensile strength of cracked UHTCC is con-
0.2 mm/min. Two beam specimens were tested for each series to de- sidered for beams under monotonic loading [22,24,25], while that of
termine the ultimate load capacity. The specimen IDs and load capacity plain concrete is always neglected. Thus, the yield load of the beam
of the beam specimens in the static tests are listed in Table 1. Note that specimen increases as the thickness of the UHTCC increases. Similarly,
“S1” in the specimen ID stands for the static load (i.e., stress level equal the static ultimate loads of Series UH40 and UH50 are higher than that
to 1). The average load capacities were used to determine the fatigue
Table 2
Table 1 Specimen ID, loading condition, and fatigue life of specimens in fatigue tests.
Specimen ID and load capacity of specimens in static tests.
Test series Stress level, S Specimen ID Load range (kN) Fatigue life, N
Test series Specimen IDs Load capacity (kN)
Range ΔP
Capacity Average
RC 0.8 RC-S0.8 49.2–9.8 39.4 343,835
RC RC-S1-1 61.4 61.5
UH40 0.9 UH40-S0.9 62.9–12.6 50.3 16,278
RC-S1-2 61.5
0.8 UH40-S0.8 55.9–11.2 44.7 266,629
UH40 UH40-S1-1 67.7 69.9 0.7 UH40-S0.7 48.9–9.8 39.1 626,966
UH40-S1-2 72.0
UH50 0.9 UH50-S0.9 62.2–12.4 49.8 109,635
UH50 UH50-S1-1 67.9 69.1 0.8 UH50-S0.8 55.3–11.1 44.2 485,149
UH50-S1-2 70.3 0.7 UH50-S0.7 48.4–9.7 38.7 903,286

3
B.-T. Huang, et al. Engineering Structures 199 (2019) 109576

Fig. 4. Load-vs.-mid-span deflection curves and crack patterns of beam specimens in static tests.

of Series RC. It should be noted that the capacities of specimens UH40- (measured by strain gauges) and UHTCC/concrete at the same height
S1 and UH50-S1 are similar. It may related to the fact that the increased (measured by attached LVDTs). It can be found that the strains of steel
10-mm UHTCC layer is closer to the natural axis compared to the 40- bars and UHTCC/concrete are very similar before the yielding point.
mm layer and the corresponding strength change is limited. Similar After the point, the strains are also roughly similar for each specimen. It
phenomenon was also observed in the former literature [30]. In future indicates that the bond between the steel reinforcements and UHTCC/
investigation, test results on more number of static specimens is needed concrete is good under static loading. It should be noted that most of
to arrive at a more reliable phenomenon and conclusion. the strain gauges are not working when the strain value is relatively
In Fig. 4, the crack patterns of each series of specimens are also large.
presented. The main cracks are marked in red. In the four bending tests,
the failure planes of all specimens were located in the constant moment
region. It can be observed from Series UH40 and UH50 that cracks of 3.2. Fatigue tests
the concrete layer turned into fine cracks in UHTCC. It should be no-
ticed that although the ultimate load capacities of Series UH40 and 3.2.1. Fatigue life
UH50 are close, the mid-span deflection of Series UH50 is larger than The fatigue lives of all tested specimens are listed in Table 2 and are
that of Series UH40, and the number of fine cracks in the UHTCC of plotted in Fig. 6. The stress-level-vs.-fatigue-life relations are shown in
Series UH50 is much larger than that of Series UH40. This phenomenon Fig. 6(a). It can be seen that the fatigue life of Series UH50 is longer
might be related to the location of longitudinal reinforcements in the than that of Series UH40 at the same stress level. At a stress level of 0.8,
UHTCC layer. The locations of the steel bars of Series UH50 are close to the fatigue life of Series UH50 is the longest compared to Series UH40
the middle of the UHTCC layer, while those of Series UH40 are close to and RC. This means that the layer of UHTCC may improve the fatigue
the top of the UHTCC layer (near the concrete-UHTCC interface). For performance of reinforced concrete beams. A similar phenomenon was
Series UH40, the location of the longitudinal reinforcement might affect reported in [16]. That is, the UHTCC layer significantly improved the
the bonding behavior of UHTCC, and thus, fewer fine cracks are ob- fatigue life of a concrete beam without reinforcement. It should be re-
served. Further investigation of this phenomenon is needed. membered that the load ranges of UH40 and UH50 are quite close for
Fig. 5 presents the comparison of the strains of longitudinal bars each stress level, and the load range of RC-S0.8 is close to those of
UH40-S0.7 and UH50-S0.7. Thus, the load-range-vs.-fatigue-life

Fig. 5. Comparison of the static strains of longitudinal bars (measured by strain gauges) and UHTCC/concrete at the same height (measured by LVDTs).

4
B.-T. Huang, et al. Engineering Structures 199 (2019) 109576

Fig. 6. Fatigue lives of specimens: (a) stress level vs. fatigue life; (b) load range vs. fatigue life.

relations are plotted in Fig. 6(b), in which the fatigue enhancement of seen that the deflections of the specimens increase as the number of
the RC-UHTCC beam can be more clearly observed. The fatigue life of load cycles increases. This reflects the fatigue degradation of the stiff-
Series UH50 is longer than that of Series UH40 in a similar load range. ness. Fig. 7(a) presents the mid-span deflection vs. ratio of load cycles
In addition, the fatigue life of the beam specimen increases with an for Series RC, UH40, and UH50. An S-shaped three-stage deflection can
increase in the thickness of the UHTCC layer in a load range of about 39 be observed, including rapid developing, stable developing, and failure
kN. stages. It can be seen in the first window of Fig. 7(a) that the mid-span
All tested beam specimens failed by the rupture of longitudinal re- deflection of the beam specimens at the same cycle ratio increases as
inforcements in the constant moment region. The flexural fatigue life of the thickness of the UHTCC layer increases. This indicates that an RC-
an under-reinforced concrete beam is governed by the longitudinal UHTCC beam can sustain fatigue loading at a larger deflection without
reinforcement [31]. For RC-UHTCC beams, the tensile strength of the failure, compared to a conventional RC beam. A similar phenomenon
cracked UHTCC is considered for beams under monotonic loading, was observed in the UHTCC-layered concrete beam without reinforce-
because of its tensile-strain-hardening characteristic. The UHTCC layer ment in Ref. [16].
can thus reduce the stress in the longitudinal bars under fatigue For Series RC, the average static failure deflection (23.3 mm, RC-S1)
loading, which results in the fatigue enhancement of the RC-UHTCC is larger than the fatigue failure deflection (8.9 mm, RC-S0.8). For
beams. Series UH40 and UH50, the average static failure deflections are
22.3 mm and 29.5 mm, respectively. The fatigue failure deflections
shown in Fig. 7 of Series UH40 and UH50 are also smaller than the
3.2.2. Mid-span deflection and cracking modes
static ones. As the stress level decreases, the fatigue deflections of the
The mid-span deflections of the specimens in the fatigue tests are
beam specimen decrease significantly for Series UH40 and UH50
shown in Fig. 7. In this study, n/N represents the normalized cycles,
(shown in Fig. 7). This phenomenon is related to the fact that the
where n is the current fatigue cycle and N is the fatigue life. It can be

Fig. 7. Mid-span deflections of specimens under fatigue loads vs. (a) ratio of load cycles and (b) number of load cycles.

5
B.-T. Huang, et al. Engineering Structures 199 (2019) 109576

Fig. 8. Crack patterns of beam specimens under static and fatigue loads.

fatigue deformations of steel and UHTCC are smaller than the static the interface, while the one in the shear region propagates along the
ones, and the deformations also decrease as the stress range decreases direction of the bottom support. The influence of the interface dela-
[6,7,32]. mination on the mechanical and durability performance of UHTCC-
The crack patterns of the beam specimens under static and fatigue layered reinforced concrete elements needs further investigation. Also,
loads are shown in Fig. 8, and all fatigue specimens failed by the rup- methods to improve interface bonding under both static and fatigue
ture of longitudinal reinforcements. It can be seen that the numbers of loadings should be paid more attention in following studies.
cracks in both the fatigue and static specimens increase as the thickness
of the UHTCC layer increases. This phenomenon coincides with the 3.2.3. Strain profiles
previous observation that the mid-span deflection of specimens in- The strain profiles of the beam specimens under fatigue loads can be
creases as the thickness of the UHTCC layer increases. In addition, the drawn based on the data measured by the strain gauge on the extreme
number of cracks in the UHTCC layer of the fatigue specimen is gen- compressive face and the LVDTs attached to the UHTCC layer. The
erally lower than that of the static specimen. This is related to the fact strain profiles of beam specimens at various cycle ratios (i.e., n/
that the multi-cracking characteristic of UHTCC is less pronounced N = 0.01, 0.25, 0.50, 0.75, and 0.99) are shown in Fig. 10. It needs to
under tensile cyclic loading [6,7]. It should be noted that the interface be pointed out that the LVDT at the bottom of UH50-S0.8 dropped
cracking is an important issue for reinforced concrete beam strength- during the fatigue tests, and the strain gauge at the top of UH50-S0.7
ened with strain-hardening cementitious composite layer [22,24,28]. In was not stable. Thus, the corresponding strains were not obtained, and
this study, the static and fatigue failure modes of the beam specimens the strain profiles of UH50-S0.8 and UH50-S0.7 are not given. All
were bending failure. However, the interface cracks did occur in both strains in Fig. 10 were obtained at the maximum fatigue loads of the
bending and shear regions. Fig. 9 presents two photographs to show the corresponding cycles.
typical interface cracks in bending region (Window ① in Fig. 8) and Fig. 10 shows that during the fatigue tests, the strain profiles of all
shear region (Window ② in Fig. 8) of the RC-UHTCC beams and the presented specimens exhibited a linear variation in general, which
interface cracking can be clearly observed. It is known that for concrete means an assumption of plane sections can be used in the subsequent
beams strengthened by a bonded plate, delamination cracks along the analysis. The fatigue degradation of stiffness can be observed with the
interface are caused by the shear stress concentration owing to the increasing of the cycle ratio. As the stress level decreases, the de-
opening of cracks in the concrete layer [33]. In this study, the re- gradation of stiffness of the RC-UHTCC beam becomes slower, which
inforced UHTCC layer can be considered as the bonded plate and the coincides with the evolution of the mid-span deflection mentioned
delamination cracks are occurred after the interface shear stress con- above. Additionally, it can be found that the behavior of UH40-S0.8 is a
centration caused the bending and shear cracks in Fig. 9. The delami- little bit different from the other ones. It might be related to the var-
nation crack in the bending region propagates in both directions along iation of cracking locations of concrete and UHTCC layer. In the

Fig. 9. Typical interface cracks in (a) bending region (Window ① in Fig. 8) and (b) shear region (Window ② in Fig. 8).

6
B.-T. Huang, et al. Engineering Structures 199 (2019) 109576

Fig. 10. Strain profiles of beam specimens under fatigue loads.

following study, more measure points, as well as more specimens, are of the fatigue life. It can be seen that the steel strain ranges of RC-S0.8
needed to arrive at a more reliable conclusion on this phenomenon. and UH50-S0.7 are a little larger than the strain ranges of concrete and
UHTCC. For specimens UH50-S0.8 and UH40-S0.7, the results of steel
3.2.4. Strain range of longitudinal bar and UHTCC bars and UHTCC layers are quite similar. Although the cracking modes
Remember that strain gauges were used to measure the strains of of concrete and UHTCC layers and interface debonding may have an
the longitudinal bars, but many of them failed during the first half of influence on the measurement of LVDTs, generally speaking, the strain
the total fatigue life. Fig. 11 presents the comparison of the fatigue estimated by the LVDT at the same height as the steel bars can be
strain ranges of steel bars (measured by strain gauges) and UHTCC/ roughly used as the strain of the steel bars for the fatigue tests with the
concrete at the same height (measured by LVDTs). Note that authors stress levels of 0.8 and 0.7. For the fatigue specimens under the stress
present all the values of the strain gauges that worked longer than half level of 0.9, the strain gauges ran out at the early stage, which may be

Fig. 11. Comparison of the fatigue strain ranges of longitudinal bars (measured by strain gauges) and UHTCC/Concrete at the same height (measured by LVDTs).

7
B.-T. Huang, et al. Engineering Structures 199 (2019) 109576

Fig. 12. Steel strain range vs. (a) ratio of load cycles and (b) number of load cycles.

related to the relatively large deformation development. Hence, there is RC-S0.8, UH40-S0.7, and UH50-S0.7 are very close (see Table 2).
not enough data to evaluate the bonding performance of steel bars for However, the strain range of RC-S0.8 is wider than those of UH40-S0.7
specimen UH40-S0.9 and UH50-S0.9 and more attention should be paid and UH50-S0.7 at the same number of load cycles [see Fig. 12(b)].
on this in future study. Moreover, the strain range of UH40-S0.7 is wider than that of UH50-
The strain range of the longitudinal bar is analyzed in the following, S0.7. This indicates that the UHTCC layer reduces the strain in the
since it is known that the strain range is one of the determining factors longitudinal reinforcements, and this effect is more pronounced as the
of the fatigue life of steel. For the specimens RC-S0.8, UH40-S0.7 and thickness of the UHTCC layer increases.
UH50-S0.7, the strain ranges of the steel reinforcements were measured On the basis of the measured strains and the plane section as-
by strain gauge (see Fig. 11). For the other fatigue specimens, the stains sumption, the strains of UHTCC for all fatigue specimens can also be
of the longitudinal bars are estimated from the strain of the LVDT at- estimated (shown in Fig. 13). It should be noted that the strain of
tached at the same height. The measured or estimated steel strain UHTCC is the strain at the center axis of the UHTCC layer. Thus, the
ranges of all fatigue specimens are shown in Fig. 12. An S-shaped three- plotted strain ranges of UHTCC in Fig. 13 are wider than those of the
stage characteristic can be observed in Fig. 12(a). The stable developing steel bars in Fig. 12.
stage is about 90% of the fatigue life, and the increase in the strain Note that the LVDT-1 of UH50-S0.8 dropped during the fatigue
range is limited during this stage. This indicates that the strain range of tests, hence the UHTCC strain of this specimen is estimated using the
the longitudinal bars may be considered as a constant during the fatigue strains of concrete and LVDT-2 based on plane section assumption.
test. The average value of the steel strain range for each fatigue spe- Fig. 13 shows that the strain range of UHTCC decreases as the stress
cimen is listed in Table 3. In general, the fatigue life increases as the level decreases, and the strain range of Series UH40 is wider than that
steel strain range decreases. For each stress level, Fig. 12(b) shows that of Series UH50 at the same stress level. The average values of the
the strain range of Series UH40 is larger than that of Series UH50 at the UHTCC strain ranges for Series UH40 and UH50 are listed in Table 3,
same number of load cycles. This phenomenon explains why the fatigue and will be used to analyze the fatigue degradation of the UHTCC layer
life of Series UH50 is higher than that of Series UH40. in the following section.
It can be seen in Tables 2 and 3 that the fatigue life of RC-S0.8 is
higher than those of UH50-S0.9, UH40-S0.9 and UH40-S0.8. This 3.2.5. Fatigue degradation of UHTCC
phenomenon can be explained as follows. The static load capacity of RC It can be found in Fig. 13 that the strain range of UHTCC may be
beam is lower than that of the UHTCC-layered ones. Hence, the applied considered as a constant during the fatigue test since the increase in the
fatigue load range for RC-S0.8 is lower than those of the UHTCC- strain range during the stable developing stage is limited. To simplify,
layered beams under the stress levels of 0.9 and 0.8 (see Table 2). It also the fatigue loadings applied on the UHTCC layer of the beam specimen
results in the lower steel strain range of RC-S0.8 (see Table 3). The might be considered as a strain-controlled fatigue load. In Ref. [34], a
longer fatigue life of specimen RC-S0.8 can hence be explained by the fitting equation of the bridging stress degradation of UHTCC under
relatively lower steel stress range. Additionally, it should be pointed out strain-controlled fatigue loading is proposed with the idea of a Weibull
that only one specimen was tested for each condition in this study and distribution. Considering that the simplified fatigue loading condition
test results on more number of specimens can help to arrive at a more of the UHTCC layer in this study is similar to that in Ref. [34], the
reliable conclusion on this in future study. The fatigue load ranges of results of the proposed fitting equation will be used to qualitatively
analyze the fatigue degradation of UHTCC in the following.
Table 3 The fitting equation of the bridging stress degradation of UHTCC in
Average strain range of steel and UHTCC in fatigue specimens. Ref. [34] is as follows:
Specimen ID Fatigue life, N Average strain range (103 με) σn − UH
σ0 − UH
Steel UHTCC m
log n ⎞ ⎞
= (1 − a1 log n) exp ⎜⎛−⎜⎛ ⎟ ⎟ + (b2 − a2 log n )
RC-S0.8 343,835 1.58 / log n0 ⎠ ⎠
UH40-S0.9 16,278 1.87 2.02
⎝ ⎝
m
UH40-S0.8 266,629 1.68 1.77 ⎛ ⎛ ⎛ log n ⎞ ⎞ ⎞
UH40-S0.7 626,966 1.34 1.51 ⎜1 − exp ⎜−⎜ log n ⎟ ⎟ ⎟
UH50-S0.9 109,635 1.98 2.13 ⎝ ⎝ ⎝ 0⎠ ⎠
⎠ (1)
UH50-S0.8 485,149 1.45 1.51
UH50-S0.7 903,286 1.16 1.09 where σn-UH is the bridging stress of UHTCC at the n load cycle; σ0-UHis
th

the average tensile strength of UHTCC (i.e., 2.8 MPa); m and n0 are the

8
B.-T. Huang, et al. Engineering Structures 199 (2019) 109576

Fig. 13. Strain range of UHTCC vs. (a) ratio of load cycles and (b) number of load cycles.

Fig. 14. Bridging stress degradation curves of UHTCC under tensile fatigue loads plotted on (a) logarithm scale and (b) linear scale.

Table 4 UH40-S0.7 and UH50-S0.8, and the curve of Strain 0.15% is used to
Parameters of Eq. (1). describe the stress evolution of the UHTCC layer in UH50-S0.7.
Maximum strain levels (%) m n0 a1 a2 b2
For Series UH40 and UH50 in the fatigue tests, the mid-span de-
flections of the specimens and the stress degradation of the UHTCC
0 8 39,811 1/15 0 0.2 layers are plotted in Fig. 15. Note that each bridging stress degradation
0.15 5.7 63,096 1/15 0 0.2 curve of UHTCC in Fig. 15 is plotted using the same range of the
0.20 8 31,623 1/20 0 0.33
number of load cycles with the corresponding fatigue specimen. Fig. 15
shows that the stresses of UHTCC are decreasing during the entire fa-
shape parameter and scale parameter of the Weibull function, respec- tigue life for UH40-S0.9 and UH50-S0.9. This might be one of the
tively; and a1, a2, and b2 are the fitting coefficients. In Ref. [34], Eq. (1) reasons why the mid-span deflections of UH40-S0.9 and UH50-S0.9
was applied in the curve-fitting of the bridging stress degradation increase faster than the others. The yield of the longitudinal re-
curves of the tensile fatigue test under strain control at maximum strain inforcements may also contribute to the larger mid-span deflections
levels of 0.10%, 0.15%, and 0.20%. The corresponding degradation (see Fig. 10), which are larger than the yield strain of steel bars (ap-
curves fitted by Eq. (1) are shown in Fig. 14. The values of the para- proximately 0.2%). For UH40-S0.8 and UH50-S0.8, the stress de-
meters for Eq. (1) are listed in Table 4, which are references to the gradation of the UHTCC layer during the second half of the fatigue life
tensile fatigue results of UHTCC in Refs. [34,35]. It can be seen that the is much slower. In this case, the increase in the mid-span defection may
residual bridging stress of UHTCC at the same number of load cycles be related to the yield of the steel bars, since the strain at the height of
increases as the maximum fatigue strain decreases. In addition, a sharp the longitudinal bars increases significantly at n/N = 0.75 compared to
decrease in the bridging stress can be observed at a number of load that at n/N = 0. 50 (see Fig. 10). For UH40-S0.7 and UH50-S0.7, the
cycles between 103 and 105. After about 2 × 105 load cycles, the increase in the mid-span deflection during the early stage of the fatigue
bridging stress becomes stable. life may be caused by the stress degradation of the UHTCC. After that,
The average strain range of the UHTCC layer of each fatigue spe- the stress of the UHTCC layer is stable during the major part of the
cimen is listed in Table 3. It can be found that the fatigue loadings fatigue life, and hence the corresponding mid-span deflection is also
applied to the UHTCC layer are similar to the strain-controlled fatigue stable until the failure stage.
loadings in Fig. 14. For simplicity, the degradation curve of Strain
0.20% is used to describe the stress evolution of the UHTCC layers in 3.2.6. Simplified mechanical modeling
UH40-S0.9, UH40-S0.8, and UH50-S0.9, since their average strain In this section, a simplified mechanical modeling method is pro-
ranges are close to 0.20%. Similarly, the degradation curve of Strain posed for UHTCC-layered reinforced concrete beam to estimate its
0.15% is used to describe the stress evolution of the UHTCC layers in maximum fatigue bending moment at each cycle, which may be helpful

9
B.-T. Huang, et al. Engineering Structures 199 (2019) 109576

Fig. 15. Mid-span deflection of beam specimens and degradation of UHTCC layers under fatigue loads plotted on (a) linear scale and (b) logarithm scale.

for the structural design of this composite beams. Through the com- strain and stress of UHTCC layer in tension at nth fatigue cycle, re-
parison of the applied and estimated fatigue bending moments, the spectively. For the sake of simplification, the tensile stress of concrete is
applicability of this method will be discussed. not considered and the tensile stresses of the UHTCC layer at various
On the basis of the plane section assumption, simplified distribu- heights are considered as the same. Additionally, the compressive stress
tions of strain and stress for UHTCC-layered reinforced concrete beam is assumed as a linear distribution. There are two reasons for this
under fatigue loading are presented in Fig. 16. Here, h and b are the simplification. First, the compressive strain of concrete is not such large
height and width of the section, respectively; a is the thickness of for most of the fatigue specimens (see Fig. 10). Second, in the following
UHTCC layer; m is the location of the longitudinal bars in tension; cn-con method, the compressive force of concrete is calculated by the tensile
is the location of the natural axis at nth fatigue cycle; εn-con and σn-con are forces of steel bars and UHTCC based on force-balance equation. The
the strain and stress of concrete in compression at nth fatigue cycle, assumed distribution of concrete stress only slightly affects the location
respectively; εn-st and σn-st are the strain and stress of longitudinal bars of the equivalent action point of resultant compressive force. In the
in tension at nth fatigue cycle, respectively; and εn-UH and σn-UH are the following, the proposed method based on these simplifications will be

10
B.-T. Huang, et al. Engineering Structures 199 (2019) 109576

accceptable for fatigue moment estimation. The estimated result for


UH50-S0.9 is much larger than the applied moment. In future study,
further investigations on this method should be carried out. Moreover,
numerical modeling is also an important issue for fatigue of UHTCC-
layered reinforced concrete beam and it should be done in following
work to contribute further to the design of this composite beam.

3.2.7. Fatigue strength of longitudinal bar


In this study, all tested beam specimens failed by the rupture of
longitudinal reinforcements in the constant moment region. The stress-
range-vs.-fatigue-life relation (i.e., S-N relation) is always used to esti-
Fig. 16. Simplified distributions of the strain and stress along the specimen mate the fatigue behavior of steel bars. The average stress range of each
section at the maximum load of nth fatigue cycle.
fatigue specimen can be obtained based on the corresponding strain
range in Table 3, and the stress-range-vs.-fatigue-life relation of the
evaluated and discussed though the comparison of the applied and tested fatigue specimens is shown in Fig. 19. However, the strain of
calculated fatigue bending moments. UH40-S0.9 and UH50-S0.9 are larger than the yield strain of the steel
According to Fig. 16, the equilibrium equation of the force (N) can bar (see Fig. 10). Thus, the estimation of the steel stress range of these
be obtained as follows: two specimens may not be accurate, and the corresponding results are
not plotted in Fig. 19.
∑N = 0 ⇒ fn − c = abσn − UH + Ast σn − st (2) The fib Model Code for Concrete Structures 2010 [36] gives a design
where fn-c is the resultant compressive force of concrete and Ast is the curve for the steel bar, which is also presented in Fig. 19. It can be seen
section areas of the longitudinal bars in tension. Here, σn-UH can be that most of the data points are located at the upper side of the curve in
obtained using Eq. (1); σn-st is equal to EsΔεn/(1 − R), which is calcu- Model Code 2010. It should be remembered that the load ranges of
lated based on the load ratio R, the strain range (Δεn) at nth fatigue cycle Series RC-S0.8, UH40-S0.7, and UH50-S0.7 are close to each other. The
and elastic modulus of steel bars (Es). The elastic modulus of steel bars steel stress ranges of these three specimens decreases with a decrease in
is assumed as not changed during the fatigue test. the thickness of the UHTCC layer. This indicates that the UHTCC layer
For Fig. 16, the equilibrium equation of the moment (M) can be reduced the fatigue stress in the longitudinal bars, and hence the fatigue
obtained as follows: life of the beam specimen was prolonged. In this study, only one spe-
cimen was tested for each condition. To obtain a reliable S-N relation
∑M = 0 ⇒ Mn − est for the RC-UHTCC beams, more fatigue tests on the RC-UHTCC speci-
= abσn − UH (cn − a 2) + Ast σn − st (cn − m) + fn − c [2(h − cn ) 3] mens are required.

(3) 3.3. Discussion


where Mn-est is the estimated fatigue bending moment; fn-c can be ob-
tained using Eq. (2); and cn-con can be calculated using the strain pro- On the basis of the above analysis, the mechanism of the fatigue
files and plane section consumption assumption. On the basis of Eqs. (2) enhancement of the RC-UHTCC beam compared to the conventional RC
and (3), the fatigue bending moment at nth fatigue cycle for the speci- beam may be as follows. On the one hand, because of the stain-hard-
mens in this study can be estimated based on the measured strains of ening behavior and high tensile strain capacity of UHTCC, the UHTCC
concrete, UHTCC and longitudinal bars. Also, this method can be ap- layer physically contributes by taking part of the stress in the tension
plied to reinforced concrete beam without UHTCC by setting a to zero. zone (see Fig. 20). As a result, the stress range of the steel bars de-
Fig. 17 presents the estimated and actual applied fatigue bending creases. It can be found in the log-scaled S-N curve provided by Model
moments for the reinforced concrete beams with and without UHTCC. Code 2010 (shown in Fig. 19) that a small decrease in the steel range
Note that the strain gauge at the top of UH50-S0.7 was not stable, hence will result in a significant increase in the fatigue life. On the other hand,
its bending moment could not be estimated. Generally speaking, for the the UHTCC layer can lower the strain localization and stress con-
specimens UH40-S0.9, UH40-S0.8, UH40-S0.7, UH50-S0.8 and RC- centration of the longitudinal reinforcements. As illustrated in
S0.8, the estimated moments show good agreement with the applied Fig. 20(b), because of the multi-cracking behavior of the UHTCC layer,
moments. It indicates that the proposed method should be applicable. the crack formed in the concrete layer can turn into many fine cracks in
However, for specimen UH50-S0.9, the estimated moment is larger than the UHTCC layer, while the crack width in the RC beam is much larger.
the applied one. It should be pointed out that the strain gauge on the With a smaller crack width, the strain localization and stress con-
steel bars did not work since the early stage of its fatigue life and hence centration of the steel bars in RC-UHTCC will be delayed.
its steel strain was estimated based on the LVDT at the same height. The The above mechanism of the fatigue enhancement of the RC-UHTCC
mid-span deflection of UH50-S0.9 is the largest one among the fatigue beam may inspire some methods for further improvement of its fatigue
specimens and the debonding between the steel bars and UHTCC may performance. It is found that although the ultimate static strengths of
happen. Hence, the steel strain estimated by LVDT may be larger than Series UH40 and UH50 are very close, the fatigue life of Series UH50 is
the actual steel strain and the tension forced of steel bars may be longer than that of Series UH40. This means the thicker UHTCC layer
overestimated, which results in the larger estimated moment in Fig. 17. may result in a lower crack-induced stress concentration. Thus, the
The other reason of this phenomenon may be related to the assumed thickness of the UHTCC can be increased to improve the fatigue per-
linear distribution of concrete stress. Owing to the larger deflection of formance. Moreover, the optimum thickness of the UHTCC should be
UH50-S0.9 compared to the other ones, the compressive strain of investigated. The lower steel stress in the RC-UHTCC beam also benefits
concrete in this cases is larger and its distribution is different, which the fatigue performance. Hence, using strain-hardening cement-based
will affect the estimated bending moment. The above explanation material with higher tensile strength (e.g., high-strength, high-ductility
should be further investigated in future study. concrete in Refs. [37,38], and ultra-high-performance fiber-reinforced
For each specimen, Fig. 18 shows the average value of the estimated concrete in Refs. [39,40]), the stress in the longitudinal bars can be
moment and applied moment. It can be seen that most of the points more significantly decreased, and the corresponding fatigue perfor-
(except UH50-S0.9) are located close to the equality lines, which also mance may be further improved.
indicates that the stability of the proposed method should be It needs to be pointed out that although most of the observed cracks

11
B.-T. Huang, et al. Engineering Structures 199 (2019) 109576

Fig. 17. Applied and estimated fatigue bending moments for reinforced concrete beams with and without UHTCC layer.

Fig. 18. Comparison of the applied and estimated fatigue bending moments. Fig. 19. Comparison of fatigue lives in test and Model Code 2010.

in the beam specimens of the static and fatigue tests are bending and investigated using a four-point bending test. Two series of RC-UHTCC
shear cracks (see Fig. 8), several delamination cracks along UHTCC- beam specimens are prepared with different thicknesses of UHTCC
concrete interface can be found for the RC-UHTCC beams. In the failure layer (i.e., 40 mm and 50 mm), and three fatigue stress levels (i.e.,
process of Series UH40 and UH50, shear stresses were induced along S = 0.9, 0.8, and 0.7) are tested. The following conclusions can be
UHTCC-concrete interface, which were caused by the tension in the drawn.
longitudinal bars and UHTCC. As mention in Section 3.2.2, for concrete In the static test, RC-UHTCC beam specimens present a higher load
beams strengthened by a bonded plate, delamination cracks along the capacity and deflection compared to the RC beam, while an increase in
interface are caused by the shear stress concentration owing to the the thickness of the UHTCC layer (i.e., from 40 mm to 50 mm) has little
opening of cracks in the concrete layer [33]. This effect is more pro- effect on the load capacity of the beam specimen. In the fatigue test, it is
nounced once strain-hardening cement-based materials with higher found that with an increase in the thickness of UHTCC, the fatigue life
tensile strength are used instead of UHTCC. Hence, in following in- of the RC-UHTCC beam at the same stress level increases, and the mid-
vestigation, method to improve the interface bonding should be con- span deflection at the same cycle ratio also increases. It is proven that
sidered. the UHTCC layer can reduce the strain in the longitudinal reinforce-
ments of a reinforced concrete beam. The fatigue degradation of the
4. Conclusions UHTCC layer and the yield of the steel bars are responsible for the in-
crease in the fatigue deflection of the RC-UHTCC beam. On the basis of
In this study, the static and fatigue behavior of RC-UHTCC beams is the measured fatigue strains of concrete, UHTCC and steel bars, a

12
B.-T. Huang, et al. Engineering Structures 199 (2019) 109576

Fig. 20. Fatigue failure process of (a) RC beam and (b) RC-UHTCC beam.

simplified method is introduced to model the fatigue performance of [6] Müller S, Mechtcherine V. Fatigue behaviour of strain-hardening cement-based
the RC-UHTCC beam and the estimated fatigue bending moments show composites (SHCC). Cem Concr Res 2017;92:75–83.
[7] Huang BT, Li QH, Xu SL, Zhou BM. Tensile fatigue behavior of fiber-reinforced
good agreement with the actual applied moments. cementitious material with high ductility: experimental study and novel P-S-N
The mechanism of the fatigue enhancement of an RC-UHTCC beam model. Constr Build Mater 2018;178:349–59.
compared to an RC beam may be as follows: (1) the UHTCC layer [8] Li VC, Leung CK. Steady-state and multiple cracking of short random fiber com-
posites. ASCE J Eng Mech 1992;118(11):2246–64.
physically contributes by taking part of the stress in the tension zone, [9] Li VC. From micromechanics to structural engineering-the design of cementitious
and thus, the stress range of the steel bars is lowered; and (2) the composites for civil engineering applications. J Struct Mech Earthq Eng
UHTCC layer can lower the strain localization and stress concentration 1993;10(2):37–48.
[10] Lin X, Yu J, Li H, Lam JY, Shih K, Sham IM, et al. Recycling polyethylene ter-
of the longitudinal reinforcements. In addition, two methods to im- ephthalate wastes as short fibers in strain-hardening cementitious composites
prove the fatigue performance of the RC-UHTCC beam are suggested: (SHCC). J Hazard Mater 2018;357:40–52.
(1) the thickness of UHTCC can be increased to improve the fatigue [11] Li H, Xu S, Leung CKY. Tensile and flexural properties of ultra high toughness ce-
mentitious composite. J Wuhan Univ Technol – Mater Sci Ed 2009;24(4):677–83.
performance, and the optimum thickness of UHTCC should be in-
[12] Matsumoto T, Suthiwarapirak P, Kanda T. Mechanisms of multiple cracking and
vestigated; and (2) by using strain-hardening cement-based material fracture of DFRCC under fatigue flexure. J Adv Concr Technol 2003;1(3):299–306.
with higher tensile strength, the stress in the longitudinal bars can be [13] Suthiwarapirak P, Matsumoto T. Fiber bridging degradation based fatigue analysis
more significantly decreased, and the corresponding fatigue perfor- of ECC under flexure. J Appl Mech 2003;6:1179–88.
[14] Huang BT, Li QH, Xu SL. Fatigue deformation model of plain and fiber-reinforced
mance may be further improved. concrete based on weibull function. ASCE J Struct Eng 2019;145(1):04018234.
Finally, it should be pointed that in this study, only two beams were [15] Zhang J, Li VC. Monotonic and fatigue performance in bending of fiber-reinforced
considered for each static test and one for each fatigue test. The num- engineered cementitious composite in overlay system. Cem Concr Res
2002;32(3):415–23.
bers of the specimens in both static and fatigue tests are not enough to [16] Leung CK, Cheung YN, Zhang J. Fatigue enhancement of concrete beam with ECC
reflect the variability of test results. The test results can describe the layer. Cem Concr Res 2007;37(5):743–50.
tendency and help to obtain these qualitative results. In following work, [17] Liu W, Xu S, Li Q. Flexural behaviour of UHTCC-layered concrete composite beam
subjected to static and fatigue loads. Fatigue Fract Eng Mater Struct
at least three specimens should be tested in each category to obtain 2013;36(8):738–49.
more reliable static and fatigue performance of UHTCC-layered re- [18] Wang K, Jansen D, Shah SP, Karr A. Permeability study of cracked concrete. Cem
inforced concrete beam. Concr Res 1997;27(3):381–93.
[19] Takewaka K, Yamaguchi T, Maeda S. Simulation models for deterioration of con-
crete structures due to chloride attack. J Adv Concr Technol 2003;1(2):139–46.
Declaration of Competing Interest [20] Maalej M, Li VC. Introduction of strain-hardening engineered cementitious com-
posites in design of reinforced concrete flexural members for improved durability.
ACI Struct J 1995;92(2):167–76.
The authors declared that there is no conflict of interest.
[21] Maalej M, Ahmed SF, Paramasivam P. Corrosion durability and structural response
of functionally-graded concrete beams. J Adv Concr Technol 2003;1(3):307–16.
Acknowledgements [22] Leung CKY, Cao Q. Development of pseudo-ductile permanent formwork for dur-
able concrete structures. Mater Struct 2010;43(7):993–1007.
[23] Xu S, Cai X. Corrosion resistance of reinforced concrete beams with cover replaced
The authors would like to acknowledge the financial support pro- by UHTCC. China Civ Eng J 2011;44(5):79–85. [in Chinese].
vided by the National Natural Science Foundation of China under Grant [24] Huang BT, Li QH, Xu SL, Li CF. Development of reinforced ultra-high toughness
Nos. 51622811 and 51678522. The authors also thank Mr. Xiao-Hua Ji cementitious composite permanent formwork: experimental study and Digital
Image Correlation analysis. Compos Struct 2017;180:892–903.
at the College of Civil Engineering and Architecture of Zhejiang [25] Li QH, Huang BT, Xu SL. Development of assembled permanent formwork using
University for his support in the experiments. The authors appreciate ultra high toughness cementitious composites. Adv Struct Eng 2016;19(7):1142–52.
the efforts of the anonymous reviewers to improve the quality of this [26] Kormeling HA, Reinhardt HW, Shah SP. Static and fatigue properties of concrete
beams reinforced with continuous bar and with fibers. ACI J Proc
study. 1980;77(1):36–43.
[27] Parvez A, Foster SJ. Fatigue behavior of steel-fiber-reinforced concrete beams. ASCE
References J Struct Eng 2014;141(4):04014117.
[28] Huang BT, Li QH, Xu SL, Zhou B. Strengthening of reinforced concrete structure
using sprayable fiber-reinforced cementitious composites with high ductility.
[1] Lee MK, Barr BIG. An overview of the fatigue behaviour of plain and fibre reinforced Compos Struct 2019;220:940–52.
concrete. Cem Concr Compos 2004;26(4):299–305. [29] Huang BT. Fatigue performance of strain-hardening fiber-reinforced cementitious
[2] Suthiwarapirak P, Matsumoto T, Kanda T. Flexural fatigue failure characteristics of composite and its functionally-graded structures. Ph. D. thesis. Department of Civil
an engineered cementitious composites and polymer cement mortars. J Mater Engineering, Zhejiang University; 2018.
Concrete Struct Pavements 2002;57:121–34. [30] Li Q, Xu S. Experimental investigation and analysis on flexural performance of
[3] Suthiwarapirak P, Matsumoto T, Kanda T. Multiple cracking and fiber bridging functionally graded composite beam crack-controlled by ultrahigh toughness ce-
characteristics of engineered cementitious composites under fatigue flexure. ASCE J mentitious composites. Sci China Ser E: Tech Sci 2009;52(6):1648–64.
Mater Civ Eng 2004;16(5):433–43. [31] Zanuy C, Albajar L, De la Fuente P. Sectional analysis of concrete structures under
[4] Huang BT, Li QH, Xu SL, Zhou BM. Frequency effect on the compressive fatigue fatigue loading. ACI Struct J 2009;106(5):667.
behavior of ultrahigh toughness cementitious composites: experimental study and [32] Ahlström J, Karlsson B. Fatigue behaviour of rail steel—a comparison between
probabilistic analysis. ASCE J Struct Eng 2017;143(8):04017073. strain and stress controlled loading. Wear 2005;258(7–8):1187–93.
[5] Huang BT, Li QH, Xu SL, Liu W, Wang HT. Fatigue deformation behavior and fiber [33] Leung CKY. Delamination failure in concrete beams retrofitted with a bonded plate.
failure mechanism of ultra-high toughness cementitious composites in compression. ASCE J Mater Civ Eng 2001;13(2):106–11.
Mater Design 2018;157:457–68. [34] Matsumoto T, Wangsiripaisal K, Hayashikawa T, He X. Uniaxial

13
B.-T. Huang, et al. Engineering Structures 199 (2019) 109576

tension–compression fatigue behavior and fiber bridging degradation of strain [37] Ranade R, Li VC, Stults MD, Rushing TS, Roth J, adn Heard WF. Micromechanics of
hardening fiber reinforced cementitious composites. Int J Fatigue high-strength, high-ductility concrete. ACI Mater J 2013;110(4):375–84.
2010;32(11):1812–22. [38] Yu KQ, Yu JT, Dai JG, Lu ZD, Shah SP. Development of ultra-high performance
[35] Matsumoto T, Chon P, Suthiwarapirak P. Effect of fiber fatigue rupture on bridging engineered cementitious composites using polyethylene (PE) fibers. Constr Build
stress degradation of fiber reinforced cementitious composites. In: Li VC, Leung Mater 2018;158:217–27.
CKY, Willam KJ, Billington SL, editors. FraMCoS-5: proceedings of the fifth inter- [39] Wille K, El-Tawil S, Naaman AE. Properties of strain hardening ultra high perfor-
national conference on fracture mechanics of concrete and concrete structures, vol. mance fiber reinforced concrete (UHP-FRC) under direct tensile loading. Cem Concr
2; 2004. p. 653–60. Compos 2014;48:53–66.
[36] Model Code 2010. fib Model Code for Concrete Structures 2010. International [40] Yoo DY, Banthia N. Mechanical properties of ultra-high-performance fiber-re-
Federation for Structural Concrete (fib), Lausanne, Switzerland; 2010. inforced concrete: a review. Cem Concr Compos 2016;73:267–80.

14

You might also like