- Email: [email protected]
Monotonic and hysteretic pullout behavior of superelastic SMA fibers with different anchorages
Accepted Manuscript Monotonic and hysteretic pullout behavior of superelastic SMA fibers with different anchorages Eunsoo Choi, Dongkyun Kim, Jong-Han Lee, Gum-Sung Ryu PII:
S1359-8368(16)30738-7
DOI:
10.1016/j.compositesb.2016.09.080
Reference:
JCOMB 4572
To appear in:
Composites Part B
Received Date: 18 May 2016 Revised Date:
16 September 2016
Accepted Date: 28 September 2016
Please cite this article as: Choi E, Kim D, Lee J-H, Ryu G-S, Monotonic and hysteretic pullout behavior of superelastic SMA fibers with different anchorages, Composites Part B (2016), doi: 10.1016/ j.compositesb.2016.09.080. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Monotonic and hysteretic pullout behavior of superelastic SMA fibers
RI PT
with different anchorages
Hongik University, Seoul 04066, Korea Fax
:
82-0-332-1244
E-mail s: [email protected]
M AN U
Phone : 82-2-320-3060;
SC
Eunsoo Choi (Corresponding author): Associate Professor, Department of Civil Engineering,
Dongkyun Kim : Assistant Professor, Department of Civil Engineering, Hongik University, Seoul 04066, Korea
TE D
E-mail s: [email protected]
Gyeongsan , Korea
EP
Jong-Han Lee: Assistant Professor, Department of Civil Engineering, Daegu University,
AC C
Email: [email protected]
Gum-Sung Ryu: Senior Researcher, Structural Engineering Research Institute, KICT, Gyeonggi, Korea
Email: [email protected]
1
ACCEPTED MANUSCRIPT
Monotonic and hysteretic pullout behavior of superelastic SMA fibers
RI PT
with different anchorages
Abstract
The aim of this study is to assess the pullout resistance of superelastic shape memory alloy
SC
(SMA) short fibers with different end shapes, which provide different anchoring actions, through monotonic and hysteretic pullout tests. For this study, NiTi superelastic SMA wire
M AN U
with a diameter of 1.0 mm was prepared and cut by a length of 40 mm to make SMA short fibers. Four types of SMA fibers with different end shapes were manufactured namely: 1) straight end shape, 2) crimped end, 3) bended end with L-shape, and 4) spearhead end. The straight end-shaped fiber was one without any anchoring action on the end part. The crimped fiber had grooves on the two sides manufactured by crimping an end part of 5 mm. The end-bended fiber had an L-shaped end with a 30° bending angle. The fiber with a
TE D
spearhead was manufactured by pressing the end part of 5 mm. The pullout tests of the fibers from the mortar matrix were first monotonically conducted with displacement control, and the hysteretic pullout behavior was obtained by 4 cyclic loadings. The Lshaped fibers increased the pullout resistance significantly compared with the straight and
EP
crimped fibers. However, they did not reach the upper plateau stress of the SMA fiber to induce state transformation. Only the fibers with a spearhead end exceeded the stress for
AC C
the state transformation because they provided sufficient anchoring resistance due to the spearhead. The maximum pullout resistance of the spearhead fiber was 3.74 times that of the L-shaped fiber. Moreover, they showed flag-shaped behavior during the hysteretic tests.
Keywords: SMA fibers; superelasticity; bond; flag-shaped behavior; cement composites
2
ACCEPTED MANUSCRIPT
1. Introduction Superelastic or pseudo-elastic shape memory alloys (SMAs) can provide a self-centering
RI PT
capacity as well as additional energy dissipation due to flag-shaped behavior because of superelasticity. Thus, they have been used for self-centering devices for systems or structures [Gao et al., 2015; Ozbulut and Hurlebaus, 2011; Lafortune et al., 2007; McCormick et al., 2006].
SC
The superelastic behavior of SMAs, which is different from shape memory effect, can be induced internally without any external action, such as heating to induce the shape memory
M AN U
effect. Thus, superelastic SMAs are considered effective materials to close cracks developed in concrete members [Abdulridha et al., 2013; Wierschem and Andrawes, 2010; Kuang and Ou, 2008; Di et al., 2009; Sakai et al., 2003]. Most of the superelastic SMA applications for concrete structures have been conducted with SMA wires or bars. Recently, SMA cables were developed [Reedlunn et al., 2013a, 2013b], but their applications are still rare. The applications
TE D
of superelastic SMA wires or bars were usually SMA beam-column connections [Wang et al., 2015; Speicher et al., 2011; Youssef et al., 2008; Ma et al., 2007], bottom couplers of concrete columns [Billah and Alam, 2012; Saiidi et al., 2009; Sadiidi and Wang, 2006] and SMA
EP
reinforcements of concrete beams [Sun et al., 2013; Billah and Alam, 2012; Wierschem and
AC C
Andrawes, 2010;]. These superelastic SMA elements increased the deformation-recovering capacity of the members after large deformation and recovered the deflection of beams to close the cracks. For the cases of concrete beams, most of the SMA applications were long elements that passed through from one end to the other. Applications of superelastic SMA short fibers for cement composites, however, have been so scarce that previous studies seemed not to provide sufficient information of superelastic SMA fibers. In this study, “short” means that the length of SMA fiber is much shorter than the length of the member. Shajil et al. (2013) provided experimental results of pullout tests for superelastic SMA fibers with an end-hook to
3
ACCEPTED MANUSCRIPT
increase pullout resistance. They reported that the SMA fibers with an end-hook provided a much greater pullout resistance than the fibers without the end-hook. The end-hook was similar to that of conventional steel fibers; thus, they were not considered proper for mass production
RI PT
since it is not easy to bend superelastic SMA wires like the end-hook. Shajil et al. (2016) also conducted tests for the self-centering and ductility of concrete beam-column joints reinforced with the same superelastic SMA fibers. However, they did not discuss the flag-shape behavior
SC
of superelastic SMA fibers induced by the anchoring resistance of the fiber.
Recently, there have been studies that suggested a new concept of SMA fibers using the shape
M AN U
memory effect [Choi et al., 2014; Kim et al., 2014]. When a martensite SMA wire with prestrain recovered its deformed length by heating due to the shape memory effect, the wire also recovered its thickness because of Poisson’s effect. Thus, by heating only the end parts of prestrained martensitic SMA fibers, the end parts bulged and their shape became like a dog-
TE D
bone. It was verified through experiments that the dog-bone shaped SMA fibers showed a much larger pullout resistance and energy dissipation capacity than prismatic SMA fibers because of the anchoring action at the end parts. Choi et al. (2015) also conducted bending tests for mortar
EP
beams with the martensitic SMA fibers to close or repair cracks. In these cases, the SMA fibers were heated by a flame or hot air from external devices to induce the shape memory effect; thus,
AC C
martensitic SMA fibers did not provide any self-centering capacity in concrete beams. The dogbone shaped fiber was globally straight without any bending part, which is different from the fibers with an end-hook. SMA fibers with hooked ends or other similar shapes, such as a loop or star, at the end parts would increase the pullout resistance while they may be inconvenient for manufacturing and not be appropriate for mass-production. Therefore, like the dog-boneshaped SMA fiber in the martensite phase, superelastic SMA fibers, which can provide sufficient pullout resistance, are needed while they do not have a bended shape at the end. The
4
ACCEPTED MANUSCRIPT
martensitic SMA fiber with the dog-bone shape was applied for direct tensile tests of the cement composite to improve the tensile behavior of cement composites [Kim et al., 2016]. For superelastic SMA fiber, Dawood et al. (2015) investigated the embedded length of the SMA
RI PT
fiber embedded between carbon sheets to induce upper stress plateau of the SMA fiber according to diameter of the SMA fiber.
The aims of this study were to suggest superelastic SMA fibers with different end shapes and
SC
to investigate their pullout behaviors according to the end shape. For the purpose, this study prepareed four types of SMA fibers including end-bended and end-bulged-shaped fibers. This
M AN U
study focused on comparing the pullout resistance of the fibers with the two end shapes. This study planned to conduct monotonic and hysteretic pullout tests for superelastic SMA fibers. The hysteretic test was expected to demonstrate the superelastic effect of the SMA fibers.
Hysteretic tensile behavior
TE D
2. Superelastic SMA short fibers
This study prepared superelastic SMA wires of NiTi alloys with a diameter of 1.0 mm and cut
EP
them into fibers with a length of 40 mm. Cyclic tensile stress-strain curves of the superelastic
AC C
SMA wire is provided in Figure 1. The superelastic effect was observed when austenitic SMA was deformed by external loading. Upon loading, the SMA deformed elastically followed by phase transformation to a detwinned martensite phase. Upon unloading, the detwinned martensite returned back to the austenite phase, and, during this process, energy dissipation occurred because of the difference between the loading and unloading paths. The tensile test was conducted by increasing first until the 2% strain and then releasing the load. Then, 2% strain was added to the previous strain up to 10% strain. The secant elastic modulus of the SMA wire was 45.5 GPa with 0.5% strain, and the SMA started state transformation at a strain
5
ACCEPTED MANUSCRIPT
of 1.23%; after this strain, the SMA did not stay in pure austenite state and it was mixed with martensitic state because of beginning state transformation. The corresponding stress was 470.0 MPa, which was the upper plateau stress, and increased slightly with an increasing
RI PT
deformation. The lower plateau stress during unloading from a 2% strain was 230.8 MPa. The residual strain of 0.17% remained due to unloading from the 2% strain; the residuals strain increased slightly with an increase in unloading deformation. The upper and lower plateau
SC
stress decreased with an increasing number of loading cycles; this was the training effect. The transformation to martensite appeared to be completed at a strain of 7.97%, and the elastic and
M AN U
plastic deformation occurred after this. Thus, the loading path to a 10% strain showed hardening behavior after the 8% strain, which was stress-induced-martensite (SIM) hardening. The stress beginning for the SIM hardening was 550 MPa. When the unloading occurred beyond the transformation completion, elastic deformation recovered while plastic deformation
TE D
did not recover. Thus, the residual strain after unloading from a 10% strain was 0.925%, which was much higher than the other residual strains induced from the upper plateau stress. The room temperature of the test was approximately 25℃, which was higher than the Af
AC C
the test condition.
EP
temperature of 18.0℃ of the SMA. Thus, the SMA wire experienced the SIM hardening under
Types of SMA fibers
This study prepared four types of superelastic SMA fibers, and Figure 2 shows the general shapes and dimensions of the four types of the SMA fibers. The figure shows the embedded half of the fibers. As a reference, the first one was prismatic and straight (SF-RF) with a length of 40 mm. The second type had a crimped surface from two directions (SF-CR) manufactured by crimping the end part of 5 mm by a press. Crimping for the ‘SF-CR’ fiber was conducted using
6
ACCEPTED MANUSCRIPT
a special device with a saw-toothed surface. Thus, the fiber had a series of hatches at both sides of the end part, and the distance between the hatches was 1.0 mm with the width and depth of the hatch being 0.1 mm, respectively. The third one has a cramped end part and looked like a
RI PT
spearhead (SF-SH). For the ‘SF-SH’ fibers, a length of 5 mm at the end part was cramped using jack-cramping, and its shape looked like a spear. The thickness and width of the spearhead was 0.5 mm and 1.5 mm, respectively. The above three types of SMA fibers were globally straight;
SC
this means that they did not have a bended end part. The last one had L-shaped end with a length of 5 mm (SF-LS), and this type represented the shape of an end-hook. Their bending
M AN U
angle was approximately 30°. The fibers were initially bent with a right angle, but they recovered most of the bending because of the superelasticity. The four types of fibers were
TE D
expected to provide different anchorage methods for pullout tests.
3. Pullout tests
A set of experiments was designed to investigate the monotonic and hysteretic pullout behavior
EP
of superelastic SMA fibers based on the anchorage methods. For this purpose, a series of single-
AC C
fiber pullout tests was planned, and six specimens were prepared for each type of fibers; three specimens were for monotonic tests and the other three were for hysteretic tests.
Specimens and test set-up For pullout tests of SMA short fibers, specimens of a single fiber embedded in mortar were prepared (see Figure 3). In the specimen, the end part that looked like a dog-bone was held
7
ACCEPTED MANUSCRIPT
against the pulling force. The composition of the mortar matrix is given in Table 1, and its compressive strength was 50 MPa. Silica sand with average diameter of 0.22 mm was used, and
RI PT
the sand-to cement ratio of the mortar was 1.0. Each fiber had a length of 40 mm, and half of its length was embedded in mortar to provide pullout resistance. The mold in Figure 3a
SC
manufactured 10 specimens at once, and Figure 3b shows a completed specimen in which 20
M AN U
mm length of the fiber was protruded. For the straight fiber (SF-RF), the embedded length was 15 mm because its bond resistance should be compared with that of the straight part of the other three fibers with anchoring parts. The specimens had a square section at the top with a width of 25 mm, and the bottom part looked like a dog-bone; its total length was 71.5 mm. Figure 4
TE D
shows the setup for the pullout test of a single fiber. A half-circular holder gripped the bottom of the specimen, and the fiber was pulled out by an actuator. Applied force was measured by a load
EP
cell placed on the top of the machine cross head, and the slip between the fiber and the mortar
AC C
was measured by a displacement transducer mounted on the specimen. Displacement control with a speed of 1.0 mm/min was applied, and the sampling rate was 5.0 Hz.
4. Results of pullout tests 4.1 Monotonic behavior Figures 5-8 show the pullout behaviors of the SMA fibers, and the pullout stress indicates the stress developed in the fiber that was obtained by dividing the pullout force by the cross-
8
ACCEPTED MANUSCRIPT
sectional area of the fiber. In the graphs, the x-axis included the slip as well as the deformation of the fiber. This study used two quantitative values, namely maximum τ max and equivalent
expressed by the following equations [Kim et al., 2014]:
τ eq =
2W p
πd f L2em
(1)
SC
Pmax πd f Lem
(2)
M AN U
τ max =
RI PT
bond strength τ eq , to compare the pullout resistance of the SMA fibers; the two parameters are
where, Pmax and W p are the maximum pullout force and pullout energy during a pullout test,
TE D
respectively, and d f and Lem are the diameter and embedded length of the fiber, respectively. Thus, the maximum bond strength is a function of the peak pullout force while the equivalent bond strength is a function of the pullout energy during the fiber’s pulling out. Figure 9 shows
EP
the calculated maximum and equivalent bond strengths of the fibers and their averages, which
AC C
are shown in Table 2.
SF-RF fiber
The bond stress of a fiber consists of three contributions namely; (1) chemical adhesion, (2) frictional resistance, and (3) anchoring action [Choi et al., 2016]. For the SF-RF fibers, the anchorage action did not work since their cross-section was prismatic. Thus, the chemical adhesion and frictional resistance worked simultaneously to produce the maximum pullout force. After the chemical adhesion was broken down, only frictional resistance existed to
9
ACCEPTED MANUSCRIPT
provide the pullout force. The average Pmax of the fibers was 51.99 N, and the corresponding τ max and τ eq were 1.11 and 0.68 MPa, respectively. The average slip
RI PT
corresponding to the average Pmax was 0.20 mm. For the SF-RF-2 and -3 fibers, the anchoring action seemed to work since the pullout force was stable, even with an increasing pullout length. In Figure 5, the case of SF-RF-1 shows typical behavior without any
SC
contribution from the anchoring action. The chemical adhesion disappeared after the maximum pullout force, and the friction resistance decreased with an increasing pullout
M AN U
length since the frictional force was proportional to the embedded length of the fiber. In the SF-RF-1 fiber, the maximum pullout force was 64.05 N, and, after the loss of chemical adhesion, the force was dropped to 37.8 N, which was the contribution of frictional resistance. Thus, the contribution of the chemical adhesion was 26.25 N. The bond stress
(3)
EP
τ fric πd f Lem = Pfric
TE D
τ fric due to the frictional force Pfric can be expressed as shown below:
Thus, the estimated frictional bond stress of the fiber was 0.8 MPa. Similarly, the chemical
AC C
bond stress can be estimated as 0.57 MPa. The length of the SF-RF fibers was shorter by 5 mm than the other types of fibers. Thus, this short length partially contributed to their relatively small maximum pullout force Pmax . However, in calculating maximum bond strength τ max , the embedded length of the fiber was used; this means that the τ max was calculated per an unit length. Thus, the maximum bond strength τ max of the SF-RF fibers can be compared with those of other types of fibers
10
ACCEPTED MANUSCRIPT
regardless to fiber length.
SF-CR fiber
RI PT
The crimping of the SF-CR fibers did not increase significantly the Pmax compared with that of the SF-RF fibers; the increment was only 8.2%. However, the crimping provided anchoring action at the end part of the fiber; thus, it increased the τ eq by 36.8% compared
SC
with the SF-RF fiber. In addition, the slip corresponding to the Pmax was 0.33 mm, which
M AN U
was delayed much more than that of the SF-RF fibers due to the anchoring action of the crimping. After breaking chemical adhesion, the pullout force was stable, which indicated the existence of the anchoring action; the anchoring action was constant, regardless of the
SF-LS fiber
TE D
embedded length of the fiber since it acted on the end part of the fiber.
The SF-LS fiber increased the Pmax (=166.1 N) significantly and delayed the corresponding
EP
slip δ max (=1.81 mm) compared with those of the SF-RF and SF-CR fibers because the Lshaped end provided considerable anchoring action. The pure pullout force due to the
AC C
chemical adhesion and friction was estimated as 51.99 N, which was from the SF-RF fiber. Thus, the increased pullout force in the SF-LS fiber due to the anchoring action was 114.11 N. The L-shaped end provided pullout resistance when it got straightened during pulling out through the duct in the mortar. The anchoring action almost disappeared when the L-shaped part was positioned in the vertical duct, and the corresponding slip was 5-6 mm, which was almost the same as the length of the L-shaped part. This behavior was different from that of the SF-CR fiber that showed stable anchoring action. A test-completed specimen of the SF-
11
ACCEPTED MANUSCRIPT
LS fiber is shown in Figure 10a, where the L-shaped part still remained even though the fiber passed through a straight duct. After loss from the anchoring action, the L-shaped part appeared to not have influenced the frictional resistance since the frictional pullout force of
RI PT
the SF-LS fiber was similar to that of the SF-RF fiber.
SF-SH fiber
SC
The average Pmax of the SF-SH fibers was 621.3 N, which is 12 times larger than that of the SF-RF fibers. Thus, it can be conjectured that the pullout force of the SF-SH fiber was
M AN U
mainly from the anchoring action, and the SF-SH fibers did not show a decrease in pullout resistance as like the other types of fibers. The slopes of the curves were softened around the slip of 0.8-1.0 mm, which was similar to the length of the transition part of the spearhead. Then, a softened increase of the pullout force was followed till the slip of 5 mm, which was
TE D
the length of the spearhead. A similar phenomenon was observed in the pullout test of steel reinforcing bars; the pullout resistance increased until the slip of the length of the transition part of a rib (Choi et al., 2014; Lee et al., 2012; Choi et al., 2011).
EP
The pullout stresses of the SF-SH-1 fibers reached about 570 MPa, which exceeded the
AC C
beginning stress for the state transformation and was just below the stress for beginning the SIM hardening of the fiber (see Figure 1). For the SF-SH-3 fiber, the anchoring resistance appeared to be sufficient to increase the pullout stress of the fiber beyond the upper plateau stress of the SMA; thus, the fiber experienced SIM hardening. Therefore, during this stage, the fiber was slipped accompanied by deformation. The specimen of the SF-SH-3 fiber was fractured because of the tensile fracture of the mortar (see Figure 10c); this was splitting bond failure. The splitting bond failure occurred when the thickness of the mortar was not sufficient and did not provide appropriate confinement for the fiber. For the SF-SH-2 fiber, 12
ACCEPTED MANUSCRIPT
however, its anchoring resistance seemed not to be sufficient to induce full SIM hardening in the fiber; the specimen was fractured at the pullout stress of 670 MPa. In this case, the fiber was slipped through the duct without any additional large deformation of the fiber. For the
RI PT
SF-SH-1 fiber, the fiber was broken at the transition point of the spearhead (see Figure 10b); it appears that stress concentration occurred at the point. The pullout stress at the fracturepoint was about 570 MPa, and the SIM hardening was not observed. From the above
SC
observations, it was found that the superelastic SMA fiber with the spearhead exceeded the upper plateau stress of the SMA for state transformation. It also appears that the fiber can
M AN U
show full SIM hardening behavior if the mortar provides sufficient confinement with appropriate thickness.
4.2 Hysteretic behavior
TE D
Figures 11 and 12 show the hysteretic pullout behaviors of the fibers. During the tests, unloading and reloading were conducted 4 times before the test’s completion, and the increase in the slip for unloading was determined based on the monotonic curves. In Figure
EP
11, the three fibers of the SF-RF, -CR, and -LS showed vertical unloading paths. This indicates that the fibers were in elastic range and did not reach the state transformation. It
AC C
also intimates that their pullout displacements were mainly caused by the slip of the fibers. However, for the SF-SH fibers in Figure 12, they showed the flag-shaped behavior during the unloading and reloading procedures and recovered a part of the applied slip plus deformation. This verifies that the SF-SH fibers experienced the state transformation and reached the stress for beginning the upper plateau stress of the fiber. In Figures 12a, b, and c, the slip+ deformations of the first unloading were almost not recovered. This indicates that the fibers were in elastic range and the movement was mainly caused by the slip. The other
13
ACCEPTED MANUSCRIPT
noticeable thing is that the stresses of the fibers at the first unloading point did not get fully into the upper plateau stress. However, in Figure 12b, the fiber at the first unloading point showed a stress of 540 MPa, which exceeded the beginning of the upper plateau stress, and,
RI PT
thus, it recovered a part of the slip plus deformation and showed the flag-shaped behavior. The total movement of the fiber at the first unloading point was 2.09 mm, and 0.59 mm was recovered during the unloading. Thus, the slip at the first unloading point was 1.50 mm, and
SC
the deformation of the fiber was 0.59 mm, which was the recovered movement. The total movement from the first reloading to the second unloading was 2.78 mm, and 1.89 mm was
reloading process was 0.89 mm.
M AN U
recovered during the unloading, which was the deformation of the fiber. Thus, the slip in the
The beginning stress of the state transformation of the fiber in the pullout tests was a little higher than that of the tensile test in Figure 1, and the reason seems to be the
TE D
contribution of friction around the fiber. For the SF-SH fibers, their pullout behavior included the slip and deformation of the fibers, and, thus, except for the slip, the behavior was similar to that of the direct tensile test. The slip of the SF-SH fibers initially occurred
EP
due to the transitional shape of the spearhead, and mortar crushing around the spearhead
AC C
subsequently induced the slip; the white part of the mortar specimen in Figure 13 indicates the slip due to the crushing of the mortar. Therefore, the anchoring action of a superelastic SMA fiber should be greater than the upper plateau stress of the SMA to induce the flagshaped behavior and to recover the deformation. The slip is related to the shape of the spearhead and strength of the mortar. If a higher strength of the mortar was used and the mortar was not crushed, the slip of the SF-SH fiber would be restricted within the initial slip of the spearhead. Otherwise, if the upper plateau stress of the superelastic SMA was lower, then the fiber could easily get into the state transformation and deform before the slip
14
ACCEPTED MANUSCRIPT
occurring due to the crushing of mortar. Therefore, for the practical use of superelastic SMA fibers, a relatively low upper plateau stress is desirable for use with normal concrete.
RI PT
5. Discussion of results
In the monotonic tests; the results were compared in Figure 14, the fibers with a straight shape and crimped end showed basic behavior of fibers in mortar or concrete. The L-shaped fibers
SC
increased the pullout resistance and energy dissipation compared with the above two fibers. However, the fiber also showed a decreased pullout resistance after a slip of 5 mm. The fibers
M AN U
with the spearhead showed totally different pullout behavior from the conventional fibers for mortar or concrete. They showed a hardening behavior after the initial slip. Moreover, the developed stress in the fiber due to the pullout resistance reached the stress for the state transformation of the SMA. This indicates that the spearhead provided sufficient anchoring
TE D
action to induce the state transformation in the fiber.
The main purpose of using superelastic SMAs is to provide self-centering capacity for a system, a member, or an element. As shown in Figures 11 and 12, only the SMA fiber with the
EP
spearhead showed a flag-shape in hysteretic behavior; this provided self-centering capacity. For the other three fibers, the displacements of the fibers were mainly composed of the slip in
AC C
mortar matrix; this was not recovered during the unloading. Their deformations were elastic, which were recovered; however, the amount of the recovery was insignificant. For the SF-SH fiber with the spearhead, the displacement of the fiber consisted of the slip and deformation of the fiber; the deformation included elastic deformation as well as deformation of the stat transformation. Thus, the deformation of the SF-SH fiber was totally recovered due to the superelasticity. Slip of the fiber occurs when the pullout force is greater than the anchoring resistance; the pullout force is mainly dependent on the stress of the fiber. Therefore, if the
15
ACCEPTED MANUSCRIPT
anchoring resistance of the SF-SH fiber is greater than the upper plateau stress of the SMA, the slip of the fiber will not occur. In that case, the hysteretic behavior would be similar to the hysteretic tensile behavior in Figure 1. Or, if the upper plateau stress of the SMA is lower than
RI PT
the anchoring resistance of the fiber, the pullout force would be smaller than the anchoring action, and the fiber’s slip would not occur.
In this study, the SF-LS fiber with the bended end failed to reach the upper plateau stress and to
SC
induce the state transformation. However, Shajil et al. (2013) used superelastic SMA fibers with a double bended end; thus, the end of the fibers looked like an N-shape. The fibers with the N-
M AN U
shaped end exceeded the stress for state transformation; however, they did not conduct a hysteretic test, and the flag-shape was not observed. The bending at the end was not easy, in particular for the double bending. Thus, the cramped spearhead is appeared to be more effective
6. Conclusions
TE D
in providing sufficient anchoring action than the double bended end.
This study investigated the pullout resistance of the NiTi superelastic SMA fibers with different
EP
end shapes; four types of end shapes, namely, straight end shape, crimped end, bended end with L-shape, and spearhead end, were employed. Monotonic and hysteretic pullout tests were
AC C
conducted to understand the pullout behavior of the superealstic SMA fibers. The following conclusions can be drawn from the experimental study:
● Pullout resistance of the superelastic SMA fiber should be larger than the upper plateau stress of the SMA to show the flag-shaped behavior and provide self-centering capacity.
● Spearhead shape of the SMA fiber was appropriate to provide sufficient anchoring action to
16
ACCEPTED MANUSCRIPT
reach the upper plateau stress of the superelastic SMA.
● The SMA fiber with the spear head showed flag-shaped behavior during hysteretic tests; this
RI PT
indicates that the spearhead shaped SMA fiber has the capacity of self-centering.
action and to induce the state transformation of the SMA.
SC
● End-bended with L-shape of the SMA fiber was not effective to provide sufficient anchoring
M AN U
In this study, the mortar specimens with the spearhead shaped fibers experienced splitting failure; thus, the ultimate pullout behavior of the fiber was not observed. Therefore, further studies with a sufficiently large mortar specimen are necessary to induce the pullout failure of the SMA with the spearhead. In addition, further studies should investigate an appropriate
Acknowledgments
TE D
upper plateau stress for state transformation to avoid the fiber’s slip in a mortar matrix.
EP
This research was supported by the Basic Science Research Program through the National
AC C
Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (Project No. 2015-041523)
17
ACCEPTED MANUSCRIPT
References Gao N, Jeon JS, Hodgson DE and DesRoches R. An innovative seismic bracing system based
RI PT
on a superelastic shape memory alloy ring, Smart Materials and Structures 2015; 25(5): 055030. Ozbulut OE and Hurlebaus S. Re-centering variable friction device for vibration control of
SC
structures subjected to near-field earthquakes, Mechanical Systems and Signal Processing 2011;
M AN U
25:2849-2862.
Lafortune P, McCormick J, DesRoches R and Terriault P. Testing of superelastic recentering pre-strained braces for seismic restraint design, Journal of Earthquake Engineering 2007; 11:383-399.
TE D
McCormick J, DesRoches R, Fugazzaa D and Auricchio F. Seismic vibration control using superelastic shape memory alloys, Journal of Engineering Materials and Technology 2006;
EP
128:294-301.
AC C
Abdulridha A, Palermo D, Foo S and Vecchio FJ. Behavior and modeling of superelastic shape memory alloy reinforced concrete beams, Engineering Structures 2013; 49:893-904. Wierschem N and Andrawes B. Superelastic SMA-FRP composite reinforcement for concrete structures, Smart Materials and Structures 2010: 19:025011.
18
ACCEPTED MANUSCRIPT
Kuang Y and Ou J. Self-repairing performance of concrete beams strengthened using superealstic SMA wires in combination with adhesives released from hollow fibers, Smart
RI PT
Materials and Structures 2008; 17:025020.
Di S, Ji S, Hua W, Li H and Steve Z. Constraint superelastic SMA based experimental study of
SC
self-monitoring and repairing of concrete beam crack, Second International Conference on
M AN U
Smart Materials and Nanotechnolgy in Engineering, Porc. of SPIE 2009;7493:74930N-1. Sakai Y, Kitagawa Y, Fukita T and Iiba M. Experimental study on enhancement of selfrestoration of concrete beams using SMA wire, Smart Systems and Nondustructive Evaluation, Proc.of SPIE 2003; 5057.
TE D
Reedlunn B, Daly S and Shaw J. Superelastic shape memory alloy cables: Part I – Isothermal tension experiments, International Journal of Solid and Structures 2013a; 50:3009-3026.
EP
Reedlunn B, Daly S and Shaw J. Superelastic shape memory alloy cables: Part II –
AC C
Subcomponent isothermal responses, International Journal of Solid and Structures 2013b; 50:3027-3044.
Wang W, Chan TM, Shao H and Chen Y. Cyclic behavior of connections equipped with NiTi shape memory alloy and steel tendons between H-shaped beam to CHS column, Engineering Structures 2015; 88:37-50.
19
ACCEPTED MANUSCRIPT
Speicher MS, DesRoches R and Leon RT. Experimental results of a NiTi shape memory alloybased recentering beam-column connection, Engineering Structures 2011; 33:2448-2457.
RI PT
Youssef MA, Alam MS and Nehdi M. Experimental Investigation on the Seismic Behavior of Beam-Column Joints Reinforced with Superelastic Shape Memory Alloys, Journal of
SC
Earthquake Engineering 2008; 12:1205-1222.
M AN U
Ma H, Wilkinson T and Cho C. Feasibility study on a self-centering beam-to-column connection by using the superelastic behavior of SMAs, Smart Materials and Structures 2007; 16:1555-1563.
Billah AHMM and Alam SM, Seismic performance of concrete columns reinforced with hybrid
TE D
shape memory alloy (SMA) and fiber reinforced polymer (FRP) bars, Construction and Building Materials 2012; 28:730-742.
EP
Saiidi MS, O’Brien M, Sadrossadat-Zadeh M. Cyclic response of concrete bridge columns
AC C
using superelastic Nitinol and bendable concrete, ACI Structural Journal 2009; 160(1):69-77. Saiidi MS and Wang H. Exploratory study of seismic response of concrete columns with shape memory alloys reinforcement, ACI Structural Journal 2006; 103(3):436-443.
Sun L, Liang D, Gao Q and Zhou J. Analysis of factors affecting the self-repair capability of SMA wire concrete beam, Mathematical Problems in Engineering 2013; 2013:138162.
20
ACCEPTED MANUSCRIPT
Billah AHM M and Alam MS. Seismic performance of concrete columns reinforced with hybrid shape memory alloy (SMA) and fiber reinforced polymer (FRP) bars, Construction and
RI PT
Building Materials 2012; 28:730-742.
Shajil N, Srinicasan SM and Santhanam M. Self-centering of shape memory alloy fiber
SC
reinforced cement mortar members subjected to strong cyclic loading, Materials and Structures
M AN U
2013; 46(4):651-61
Shajil N, Srinicasan SM and Santhanam M. An experimental study on self-centering and ductility of pseudo-elastic shape memory alloy (PESMA) fiber reinforced beam and beamcolumn joint specimens, Materials and Structures 2016; 49:783-793.
TE D
Choi E, Kim DJ, Chung YS and Nam TH. Bond-slip characteristics of SMA reinforcing fibers obtained by pull-out tests, Materials Research Bulletin 2014; 58:28-31.
EP
Kim DJ, Kim HA, Chung YS and Choi E. Pullout resistance of straight NiTi shape memory
AC C
alloy fibers in cement mortar after cold drawing and heat treatment, Composites Part B: Engineering 2014; 67:588-594. Choi E, Kim DJ, Youn H and Nam TH. Repairing cracks developed in mortar beams reinforced by cold-drawn NiTi or NiTiNb SMA fibers, Smart Materials and Structures 2015; 24:125010. Kim MK, Kim DJ, Chung YS and Choi E. Direct tensile behavior of shape memory alloy fiber reinforced cement composites, Construction and Building Materials 2016; 102:462-470.
21
ACCEPTED MANUSCRIPT
Choi E, Kim DJ, Jeon C and Gin S. New SMA short fibers for cement composites manufacturing by cold drawing, Journal of Materials Science Research 2016; 5(2):68-81.
RI PT
Choi E, Cho BS, Jeon JS and Yoon SJ. Bond behavior of steel deformed bars embedded in concrete confined by FRP wire jackets, Construction and Building Materials 2014; 68:716-725.
SC
Lee H, Choi E, Cho SC, Park T. Bond and splitting behavior of reinforced concrete confined by
M AN U
steel jackets without grouting, Magazine of Concrete Research 2012; 64(3):225-237. Choi E, Chung YS, Kim YW and Kim JW. Monotonic and cyclic bond behavior of confined concrete using NiTiNb SMA wires, Smart Materials and Structures 2011; 20:075016. Dawood M, El-Tahan MW, and Zhang B. Bond behavior of superelastic shape memory alloys
EP AC C
247.
TE D
to carbon fiber reinforced polymer composites. Composites Part B: Engineering 2015; 77:238-
22
RI PT
ACCEPTED MANUSCRIPT
SC
List of Tables
Table 1. Composition of matrix mixtures by weight ratio and compressive strength
M AN U
Table 2. Averages of maximum and equivalent bond strength of the fibers
List of Figures
TE D
Figure 1. Hysteretic tensile stress-strain curve of superelastic SMA wire Figure 2. Shapes and dimensions of four types of SMA fibers Figure 3. Mold and specimen with dimension for pullout test
EP
Figure 4. Test set-up for single-fiber pullout test Figure 5. Pullout behavior of SF-RF fibers
AC C
Figure 6. Pullout behavior of SF-CR fibers Figure 7. Pullout behavior of SF-LS fibers Figure 8. Pullout behavior of SF-SH fibers Figure 9. Maximum and equivalent bond strength of the fibers Figure 10. Test completed specimens of the SF-LS and –SH fibers Figure 11. Hysteretic pullout behavior of SF-RF, -CR and –LS fibers Figure 12. Hysteretic pullout behavior of SF-SH fibers
23
ACCEPTED MANUSCRIPT
Figure 13. Fractured surface of mortar specimen with SF-SH fiber
AC C
EP
TE D
M AN U
SC
RI PT
Figure 14. Comparison of pullout behavior of the four SMA fibers
24
RI PT
ACCEPTED MANUSCRIPT
Table 1. Composition of matrix mixtures by weight ratio and compressive strength
(Type 3)
0.15
High-range water-
Sand
reducing admixture
1.00
0.009
Compressive
Water
AC C
EP
TE D
M AN U
1.00
Fly ash
Silica
SC
Cement
0.35
strength (MPa) 50
ACCEPTED MANUSCRIPT
Table 2. Averages of maximum and equivalent bond strength of the fibers Wp
τ max
τ eq
(mm)
(MPa)
(N·mm) 254.6 266.9 197.7 239.7 521.9 710.8 508.0 580.2 1015.8 1142.9 1093.2 1084.0 860.3 5165.9 6450.6 4158.9
(MPa)
(MPa)
82.37 70.57 47.65 66.86 95.54 93.40 99.70 96.21 204.7 225.6 210.4 213.6 574.2 677.5 1145.5 799.1
AC C
1.37 1.17 0.79 1.11 1.20 1.16 1.24 1.20 2.55 2.81 2.62 2.66 7.14 8.43 14.3 9.96
SC
0.18 0.24 0.18 0.20 0.46 0.37 0.16 0.33 1.88 1.62 1.93 1.81 2.59 11.64 10.43 8.22
RI PT
σ max
M AN U
64.05 54.87 37.05 51.99 74.70 72.63 77.53 74.95 159.2 175.4 163.6 166.1 446.5 526.8 890.7 621.3
δ max
EP
SF-RF-1 SF-RF-2 SF-RF-3 Average SF-CR-1 SF-CR-2 SF-CR-3 Average SF-LS-1 SF-LS-2 SF-LS-3 Average SF-SH-1 SF-SH-2 SF-SH-3 Average
Pmax (N)
TE D
Fiber type
0.72 0.76 0.56 0.68 0.84 1.14 0.81 0.93 1.63 1.83 1.75 1.74 1.38 8.26 10.32 6.65
1000
550 MPa
SC
470 MPa
600
400
200
0 0
1
2
3
M AN U
Stess (MPa)
800
RI PT
ACCEPTED MANUSCRIPT
4
5
6
7
8
9
TE D
Strain (%)
AC C
EP
Figure 1. Hysteretic tensile stress-strain curve of superelastic SMA wire
10
11
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
(a) Pictures of the four SMA fibers and spear-head
(b) SF-SH fiber
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
(d) SF-LS fiber
EP
TE D
(c) SF-CR fiber
Figure 2. Shapes and dimensions of four types of SMA fibers
RI PT
ACCEPTED MANUSCRIPT
Unit : mm
25
M AN U
41.3
(a) Mold
(b) Specimen
(c ) Dimension
AC C
EP
TE D
Figure 3. Mold and specimen with dimension for pullout test
71.5
SC
25
25
ACCEPTED MANUSCRIPT
PULLOUT LOAD
RI PT
LOAD CELL
SC
FIBER GRIP SYSTEM
M AN U
FIBER
SPECIMEN
TE D
SPECIMEN GRIP SYSTEM
AC C
EP
Figure 4. Test setup for single-fiber pullout test
ACCEPTED MANUSCRIPT 90
110
(a) SF-RF-1
80
100
70 50
60
40
50
30
40 30
20
20
10
RI PT
Pullout force (N)
80
60
Pullout stress (MPa)
90
70
10
0
0 0
2
4
6
8
10
12
14
16
90
110
(b) SF-RF-2
100
M AN U
80
80
60
70
50
60
40
50 40
30
30
20
20
TE D
10 0 0
2
4
6
8
10
12
14
Pullout stress (MPa)
90
70
Pullout force (N)
SC
Slip + deformation (mm)
10 0 16
Slip + deformation (mm)
(c) SF-RF-3
80 70
AC C
60
90
50
60
40
50
30
40 30
20
20
10
10
0
0 0
2
4
6
8
10
12
Slip + deformation (mm)
Figure 5. Pullout behavior of SF-RF fibers
14
Pullout stress (MPa)
70
Pullout force (N)
100
EP
80
ACCEPTED MANUSCRIPT 140
(a) SF-CR-1
100
100
60
80 60
40 40 20
20
0
0 0
2
4
6
8
10
12
14
16
18
20
22
140
(b) SF-CR-2
100
Pullout force (N)
80
100 80
60
60
40
40
20
Pullout stress (MPa)
M AN U
120
SC
Slip + deformation (mm)
0
2
4
6
8
TE D
20
0
10
12
14
16
18
20
0 22
Slip + deformation (mm)
EP
140
(c) SF-CR-3
100
AC C
80
120
80
60
60
40
40
20
20
0
0 0
2
4
6
8
10
12
14
16
18
Slip + deformation (mm)
Figure 6. Pullout behavior of SF-CR fibers
20
22
Pullout stress (MPa)
100
Pullout force (N)
RI PT
80
Pullout stress (MPa)
Pullout force (N)
120
ACCEPTED MANUSCRIPT 210
(a) SF-LS-1
240
180
120
150
90
120 90
60
60 30
30
0
0 0
2
4
6
8
10
12
14
16
18
20
220
(b) SF-LS-2
270
40
M AN U
200
20
30
210
160
180
140 120
150
100
120
80
90
60
0 0
2
4
6
8
TE D
60
10
12
14
16
18
Pullout stress (MPa)
240
180
Pullout force (N)
SC
Slip + deformation (mm)
0 20
Slip + deformation (mm)
EP
(c) SF-LS-3 180
210 180
AC C
150
240
120
150
90
120 90
60
60
30
30
0
0 0
2
4
6
8
10
12
14
16
Slip + deformation (mm)
Figure 7. Pullout behavior of SF-LS fibers
18
20
Pullout stress (MPa)
210
Pullout force (N)
RI PT
Pullout force (N)
180
Pullout stress (MPa)
210 150
ACCEPTED MANUSCRIPT 600
(a) SF-SH-1
700
500 500
300
400 300
200 200 100
100
0
0 0
1
2
3
4
5
700
(b) SF-SH-2
800
M AN U
600
SC
Slip + deformation (mm)
Pullout force (N)
600
400
500 400
300
300
200
200
0 0
2
4
6
8
TE D
100
10
12
14
16
18
Pullout stress (MPa)
700
500
100 0 20
Slip + deformation (mm)
EP
1200
(c) SF-SH-3
1000
AC C
800
800
600
600
400
400
200
200
0
0 0
2
4
6
8
10
12
14
16
Slip + deformation (mm)
Figure 8. Pullout behavior of SF-SH fibers
18
20
Pullout stress (MPa)
1000
Pullout force (N)
RI PT
400
Pullout stress (MPa)
Pullout force (N)
600
ACCEPTED MANUSCRIPT
14
(a)
Average
12
RI PT
10 8 6
2 0
SF-RF
SF-CR
SF-LS
Specimen
12
Average
10
TE D
8 6 4 2
SF-RF
SF-CR
AC C
0
EP
Equivalent bond strength (MPa)
(b)
SF-SH
SC
4
M AN U
Maximum bond strength (MPa)
16
SF-LS
SF-SH
Specimen
Figure 9. Maximum and equivalent bond strength of the fibers
(b) SF-SH-1
EP
(a) SF-LS fiber
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
(c) SF-SH-2 & 3
AC C
Figure 10. Test completed specimens of the SF-LS and -SH fibers
ACCEPTED MANUSCRIPT 80
100
(a) SF-RF-hysteresis curve
40 20 20
0
0 0
2
4
6
8
10
12
14
16
100 120
M AN U
(b) SF-CR-hysteresis curve 80
80
60
60
40
40
20
20
0 0
2
4
6
8
10
12
14
16
18
Pullout stress (MPa)
100
TE D
Pullout force (N)
SC
Slip + deformation (mm)
RI PT
Pullout force (N)
60 40
Pullout stress (MPa)
80
60
0 20
Slip + deformation (mm)
EP
300
(c) SF-LS-hysteresis curve
250
AC C
Pullout force (N)
200
200
150
150
100
100
50
50
0
Pullout stress (MPa)
250
0 0
2
4
6
8
10
12
14
16
18
20
Slip + deformation (mm)
Figure 11. Hysteretic pullout behavior of SF-RF, -CR, and –LS fibers
ACCEPTED MANUSCRIPT 800
1000
(a) SF-SH-1-hysteresis curve
700
900
600
400
500
300
400 300
200
200 100
100
0
0 0
1
2
3
4
5
6
7
8
9 10 11 12 13 14
700
900
(b) SF-SH-2-hysteresis curve
500
600
400
500 400
300
300
200
200
Slip
Deformation
TE D
100 0 0
1
2
3
4
5
6
7
Pullout stress (MPa)
700
Total movement
Pullout force (N)
800
M AN U
600
SC
Slip + deformation (mm)
100 0 8
Slip + deformation (mm)
(c) SF-SH-3-hysteresis curve
500
AC C
Pullout force (N)
600
900 800 700 600
400
500
300
400 300
200
200
100
100
0
0 0
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15
Slip + deformation (mm)
Figure 12. Hysteretic pullout behavior of SF-SH fiber
Pullout stress (MPa)
EP
700
RI PT
Pullout force (N)
700 500
Pullout stress (MPa)
800
600
SC
RI PT
ACCEPTED MANUSCRIPT
M AN U
Slip
AC C
EP
TE D
Figure 13. Fractured surface of mortar specimen with SF-SH fiber
600
SF-SH
SC
400 300
SF-LS
200
M AN U
Pullout force (N)
500
RI PT
ACCEPTED MANUSCRIPT
SF-CR
100
SF-RF
0 0
2
4
6
8
10
12
TE D
Slip + deformation (mm)
AC C
EP
Figure 14. Comparison of pullout behavior of the four SMA fibers
14
16
S1359-8368(16)30738-7
DOI:
10.1016/j.compositesb.2016.09.080
Reference:
JCOMB 4572
To appear in:
Composites Part B
Received Date: 18 May 2016 Revised Date:
16 September 2016
Accepted Date: 28 September 2016
Please cite this article as: Choi E, Kim D, Lee J-H, Ryu G-S, Monotonic and hysteretic pullout behavior of superelastic SMA fibers with different anchorages, Composites Part B (2016), doi: 10.1016/ j.compositesb.2016.09.080. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Monotonic and hysteretic pullout behavior of superelastic SMA fibers
RI PT
with different anchorages
Hongik University, Seoul 04066, Korea Fax
:
82-0-332-1244
E-mail s: [email protected]
M AN U
Phone : 82-2-320-3060;
SC
Eunsoo Choi (Corresponding author): Associate Professor, Department of Civil Engineering,
Dongkyun Kim : Assistant Professor, Department of Civil Engineering, Hongik University, Seoul 04066, Korea
TE D
E-mail s: [email protected]
Gyeongsan , Korea
EP
Jong-Han Lee: Assistant Professor, Department of Civil Engineering, Daegu University,
AC C
Email: [email protected]
Gum-Sung Ryu: Senior Researcher, Structural Engineering Research Institute, KICT, Gyeonggi, Korea
Email: [email protected]
1
ACCEPTED MANUSCRIPT
Monotonic and hysteretic pullout behavior of superelastic SMA fibers
RI PT
with different anchorages
Abstract
The aim of this study is to assess the pullout resistance of superelastic shape memory alloy
SC
(SMA) short fibers with different end shapes, which provide different anchoring actions, through monotonic and hysteretic pullout tests. For this study, NiTi superelastic SMA wire
M AN U
with a diameter of 1.0 mm was prepared and cut by a length of 40 mm to make SMA short fibers. Four types of SMA fibers with different end shapes were manufactured namely: 1) straight end shape, 2) crimped end, 3) bended end with L-shape, and 4) spearhead end. The straight end-shaped fiber was one without any anchoring action on the end part. The crimped fiber had grooves on the two sides manufactured by crimping an end part of 5 mm. The end-bended fiber had an L-shaped end with a 30° bending angle. The fiber with a
TE D
spearhead was manufactured by pressing the end part of 5 mm. The pullout tests of the fibers from the mortar matrix were first monotonically conducted with displacement control, and the hysteretic pullout behavior was obtained by 4 cyclic loadings. The Lshaped fibers increased the pullout resistance significantly compared with the straight and
EP
crimped fibers. However, they did not reach the upper plateau stress of the SMA fiber to induce state transformation. Only the fibers with a spearhead end exceeded the stress for
AC C
the state transformation because they provided sufficient anchoring resistance due to the spearhead. The maximum pullout resistance of the spearhead fiber was 3.74 times that of the L-shaped fiber. Moreover, they showed flag-shaped behavior during the hysteretic tests.
Keywords: SMA fibers; superelasticity; bond; flag-shaped behavior; cement composites
2
ACCEPTED MANUSCRIPT
1. Introduction Superelastic or pseudo-elastic shape memory alloys (SMAs) can provide a self-centering
RI PT
capacity as well as additional energy dissipation due to flag-shaped behavior because of superelasticity. Thus, they have been used for self-centering devices for systems or structures [Gao et al., 2015; Ozbulut and Hurlebaus, 2011; Lafortune et al., 2007; McCormick et al., 2006].
SC
The superelastic behavior of SMAs, which is different from shape memory effect, can be induced internally without any external action, such as heating to induce the shape memory
M AN U
effect. Thus, superelastic SMAs are considered effective materials to close cracks developed in concrete members [Abdulridha et al., 2013; Wierschem and Andrawes, 2010; Kuang and Ou, 2008; Di et al., 2009; Sakai et al., 2003]. Most of the superelastic SMA applications for concrete structures have been conducted with SMA wires or bars. Recently, SMA cables were developed [Reedlunn et al., 2013a, 2013b], but their applications are still rare. The applications
TE D
of superelastic SMA wires or bars were usually SMA beam-column connections [Wang et al., 2015; Speicher et al., 2011; Youssef et al., 2008; Ma et al., 2007], bottom couplers of concrete columns [Billah and Alam, 2012; Saiidi et al., 2009; Sadiidi and Wang, 2006] and SMA
EP
reinforcements of concrete beams [Sun et al., 2013; Billah and Alam, 2012; Wierschem and
AC C
Andrawes, 2010;]. These superelastic SMA elements increased the deformation-recovering capacity of the members after large deformation and recovered the deflection of beams to close the cracks. For the cases of concrete beams, most of the SMA applications were long elements that passed through from one end to the other. Applications of superelastic SMA short fibers for cement composites, however, have been so scarce that previous studies seemed not to provide sufficient information of superelastic SMA fibers. In this study, “short” means that the length of SMA fiber is much shorter than the length of the member. Shajil et al. (2013) provided experimental results of pullout tests for superelastic SMA fibers with an end-hook to
3
ACCEPTED MANUSCRIPT
increase pullout resistance. They reported that the SMA fibers with an end-hook provided a much greater pullout resistance than the fibers without the end-hook. The end-hook was similar to that of conventional steel fibers; thus, they were not considered proper for mass production
RI PT
since it is not easy to bend superelastic SMA wires like the end-hook. Shajil et al. (2016) also conducted tests for the self-centering and ductility of concrete beam-column joints reinforced with the same superelastic SMA fibers. However, they did not discuss the flag-shape behavior
SC
of superelastic SMA fibers induced by the anchoring resistance of the fiber.
Recently, there have been studies that suggested a new concept of SMA fibers using the shape
M AN U
memory effect [Choi et al., 2014; Kim et al., 2014]. When a martensite SMA wire with prestrain recovered its deformed length by heating due to the shape memory effect, the wire also recovered its thickness because of Poisson’s effect. Thus, by heating only the end parts of prestrained martensitic SMA fibers, the end parts bulged and their shape became like a dog-
TE D
bone. It was verified through experiments that the dog-bone shaped SMA fibers showed a much larger pullout resistance and energy dissipation capacity than prismatic SMA fibers because of the anchoring action at the end parts. Choi et al. (2015) also conducted bending tests for mortar
EP
beams with the martensitic SMA fibers to close or repair cracks. In these cases, the SMA fibers were heated by a flame or hot air from external devices to induce the shape memory effect; thus,
AC C
martensitic SMA fibers did not provide any self-centering capacity in concrete beams. The dogbone shaped fiber was globally straight without any bending part, which is different from the fibers with an end-hook. SMA fibers with hooked ends or other similar shapes, such as a loop or star, at the end parts would increase the pullout resistance while they may be inconvenient for manufacturing and not be appropriate for mass-production. Therefore, like the dog-boneshaped SMA fiber in the martensite phase, superelastic SMA fibers, which can provide sufficient pullout resistance, are needed while they do not have a bended shape at the end. The
4
ACCEPTED MANUSCRIPT
martensitic SMA fiber with the dog-bone shape was applied for direct tensile tests of the cement composite to improve the tensile behavior of cement composites [Kim et al., 2016]. For superelastic SMA fiber, Dawood et al. (2015) investigated the embedded length of the SMA
RI PT
fiber embedded between carbon sheets to induce upper stress plateau of the SMA fiber according to diameter of the SMA fiber.
The aims of this study were to suggest superelastic SMA fibers with different end shapes and
SC
to investigate their pullout behaviors according to the end shape. For the purpose, this study prepareed four types of SMA fibers including end-bended and end-bulged-shaped fibers. This
M AN U
study focused on comparing the pullout resistance of the fibers with the two end shapes. This study planned to conduct monotonic and hysteretic pullout tests for superelastic SMA fibers. The hysteretic test was expected to demonstrate the superelastic effect of the SMA fibers.
Hysteretic tensile behavior
TE D
2. Superelastic SMA short fibers
This study prepared superelastic SMA wires of NiTi alloys with a diameter of 1.0 mm and cut
EP
them into fibers with a length of 40 mm. Cyclic tensile stress-strain curves of the superelastic
AC C
SMA wire is provided in Figure 1. The superelastic effect was observed when austenitic SMA was deformed by external loading. Upon loading, the SMA deformed elastically followed by phase transformation to a detwinned martensite phase. Upon unloading, the detwinned martensite returned back to the austenite phase, and, during this process, energy dissipation occurred because of the difference between the loading and unloading paths. The tensile test was conducted by increasing first until the 2% strain and then releasing the load. Then, 2% strain was added to the previous strain up to 10% strain. The secant elastic modulus of the SMA wire was 45.5 GPa with 0.5% strain, and the SMA started state transformation at a strain
5
ACCEPTED MANUSCRIPT
of 1.23%; after this strain, the SMA did not stay in pure austenite state and it was mixed with martensitic state because of beginning state transformation. The corresponding stress was 470.0 MPa, which was the upper plateau stress, and increased slightly with an increasing
RI PT
deformation. The lower plateau stress during unloading from a 2% strain was 230.8 MPa. The residual strain of 0.17% remained due to unloading from the 2% strain; the residuals strain increased slightly with an increase in unloading deformation. The upper and lower plateau
SC
stress decreased with an increasing number of loading cycles; this was the training effect. The transformation to martensite appeared to be completed at a strain of 7.97%, and the elastic and
M AN U
plastic deformation occurred after this. Thus, the loading path to a 10% strain showed hardening behavior after the 8% strain, which was stress-induced-martensite (SIM) hardening. The stress beginning for the SIM hardening was 550 MPa. When the unloading occurred beyond the transformation completion, elastic deformation recovered while plastic deformation
TE D
did not recover. Thus, the residual strain after unloading from a 10% strain was 0.925%, which was much higher than the other residual strains induced from the upper plateau stress. The room temperature of the test was approximately 25℃, which was higher than the Af
AC C
the test condition.
EP
temperature of 18.0℃ of the SMA. Thus, the SMA wire experienced the SIM hardening under
Types of SMA fibers
This study prepared four types of superelastic SMA fibers, and Figure 2 shows the general shapes and dimensions of the four types of the SMA fibers. The figure shows the embedded half of the fibers. As a reference, the first one was prismatic and straight (SF-RF) with a length of 40 mm. The second type had a crimped surface from two directions (SF-CR) manufactured by crimping the end part of 5 mm by a press. Crimping for the ‘SF-CR’ fiber was conducted using
6
ACCEPTED MANUSCRIPT
a special device with a saw-toothed surface. Thus, the fiber had a series of hatches at both sides of the end part, and the distance between the hatches was 1.0 mm with the width and depth of the hatch being 0.1 mm, respectively. The third one has a cramped end part and looked like a
RI PT
spearhead (SF-SH). For the ‘SF-SH’ fibers, a length of 5 mm at the end part was cramped using jack-cramping, and its shape looked like a spear. The thickness and width of the spearhead was 0.5 mm and 1.5 mm, respectively. The above three types of SMA fibers were globally straight;
SC
this means that they did not have a bended end part. The last one had L-shaped end with a length of 5 mm (SF-LS), and this type represented the shape of an end-hook. Their bending
M AN U
angle was approximately 30°. The fibers were initially bent with a right angle, but they recovered most of the bending because of the superelasticity. The four types of fibers were
TE D
expected to provide different anchorage methods for pullout tests.
3. Pullout tests
A set of experiments was designed to investigate the monotonic and hysteretic pullout behavior
EP
of superelastic SMA fibers based on the anchorage methods. For this purpose, a series of single-
AC C
fiber pullout tests was planned, and six specimens were prepared for each type of fibers; three specimens were for monotonic tests and the other three were for hysteretic tests.
Specimens and test set-up For pullout tests of SMA short fibers, specimens of a single fiber embedded in mortar were prepared (see Figure 3). In the specimen, the end part that looked like a dog-bone was held
7
ACCEPTED MANUSCRIPT
against the pulling force. The composition of the mortar matrix is given in Table 1, and its compressive strength was 50 MPa. Silica sand with average diameter of 0.22 mm was used, and
RI PT
the sand-to cement ratio of the mortar was 1.0. Each fiber had a length of 40 mm, and half of its length was embedded in mortar to provide pullout resistance. The mold in Figure 3a
SC
manufactured 10 specimens at once, and Figure 3b shows a completed specimen in which 20
M AN U
mm length of the fiber was protruded. For the straight fiber (SF-RF), the embedded length was 15 mm because its bond resistance should be compared with that of the straight part of the other three fibers with anchoring parts. The specimens had a square section at the top with a width of 25 mm, and the bottom part looked like a dog-bone; its total length was 71.5 mm. Figure 4
TE D
shows the setup for the pullout test of a single fiber. A half-circular holder gripped the bottom of the specimen, and the fiber was pulled out by an actuator. Applied force was measured by a load
EP
cell placed on the top of the machine cross head, and the slip between the fiber and the mortar
AC C
was measured by a displacement transducer mounted on the specimen. Displacement control with a speed of 1.0 mm/min was applied, and the sampling rate was 5.0 Hz.
4. Results of pullout tests 4.1 Monotonic behavior Figures 5-8 show the pullout behaviors of the SMA fibers, and the pullout stress indicates the stress developed in the fiber that was obtained by dividing the pullout force by the cross-
8
ACCEPTED MANUSCRIPT
sectional area of the fiber. In the graphs, the x-axis included the slip as well as the deformation of the fiber. This study used two quantitative values, namely maximum τ max and equivalent
expressed by the following equations [Kim et al., 2014]:
τ eq =
2W p
πd f L2em
(1)
SC
Pmax πd f Lem
(2)
M AN U
τ max =
RI PT
bond strength τ eq , to compare the pullout resistance of the SMA fibers; the two parameters are
where, Pmax and W p are the maximum pullout force and pullout energy during a pullout test,
TE D
respectively, and d f and Lem are the diameter and embedded length of the fiber, respectively. Thus, the maximum bond strength is a function of the peak pullout force while the equivalent bond strength is a function of the pullout energy during the fiber’s pulling out. Figure 9 shows
EP
the calculated maximum and equivalent bond strengths of the fibers and their averages, which
AC C
are shown in Table 2.
SF-RF fiber
The bond stress of a fiber consists of three contributions namely; (1) chemical adhesion, (2) frictional resistance, and (3) anchoring action [Choi et al., 2016]. For the SF-RF fibers, the anchorage action did not work since their cross-section was prismatic. Thus, the chemical adhesion and frictional resistance worked simultaneously to produce the maximum pullout force. After the chemical adhesion was broken down, only frictional resistance existed to
9
ACCEPTED MANUSCRIPT
provide the pullout force. The average Pmax of the fibers was 51.99 N, and the corresponding τ max and τ eq were 1.11 and 0.68 MPa, respectively. The average slip
RI PT
corresponding to the average Pmax was 0.20 mm. For the SF-RF-2 and -3 fibers, the anchoring action seemed to work since the pullout force was stable, even with an increasing pullout length. In Figure 5, the case of SF-RF-1 shows typical behavior without any
SC
contribution from the anchoring action. The chemical adhesion disappeared after the maximum pullout force, and the friction resistance decreased with an increasing pullout
M AN U
length since the frictional force was proportional to the embedded length of the fiber. In the SF-RF-1 fiber, the maximum pullout force was 64.05 N, and, after the loss of chemical adhesion, the force was dropped to 37.8 N, which was the contribution of frictional resistance. Thus, the contribution of the chemical adhesion was 26.25 N. The bond stress
(3)
EP
τ fric πd f Lem = Pfric
TE D
τ fric due to the frictional force Pfric can be expressed as shown below:
Thus, the estimated frictional bond stress of the fiber was 0.8 MPa. Similarly, the chemical
AC C
bond stress can be estimated as 0.57 MPa. The length of the SF-RF fibers was shorter by 5 mm than the other types of fibers. Thus, this short length partially contributed to their relatively small maximum pullout force Pmax . However, in calculating maximum bond strength τ max , the embedded length of the fiber was used; this means that the τ max was calculated per an unit length. Thus, the maximum bond strength τ max of the SF-RF fibers can be compared with those of other types of fibers
10
ACCEPTED MANUSCRIPT
regardless to fiber length.
SF-CR fiber
RI PT
The crimping of the SF-CR fibers did not increase significantly the Pmax compared with that of the SF-RF fibers; the increment was only 8.2%. However, the crimping provided anchoring action at the end part of the fiber; thus, it increased the τ eq by 36.8% compared
SC
with the SF-RF fiber. In addition, the slip corresponding to the Pmax was 0.33 mm, which
M AN U
was delayed much more than that of the SF-RF fibers due to the anchoring action of the crimping. After breaking chemical adhesion, the pullout force was stable, which indicated the existence of the anchoring action; the anchoring action was constant, regardless of the
SF-LS fiber
TE D
embedded length of the fiber since it acted on the end part of the fiber.
The SF-LS fiber increased the Pmax (=166.1 N) significantly and delayed the corresponding
EP
slip δ max (=1.81 mm) compared with those of the SF-RF and SF-CR fibers because the Lshaped end provided considerable anchoring action. The pure pullout force due to the
AC C
chemical adhesion and friction was estimated as 51.99 N, which was from the SF-RF fiber. Thus, the increased pullout force in the SF-LS fiber due to the anchoring action was 114.11 N. The L-shaped end provided pullout resistance when it got straightened during pulling out through the duct in the mortar. The anchoring action almost disappeared when the L-shaped part was positioned in the vertical duct, and the corresponding slip was 5-6 mm, which was almost the same as the length of the L-shaped part. This behavior was different from that of the SF-CR fiber that showed stable anchoring action. A test-completed specimen of the SF-
11
ACCEPTED MANUSCRIPT
LS fiber is shown in Figure 10a, where the L-shaped part still remained even though the fiber passed through a straight duct. After loss from the anchoring action, the L-shaped part appeared to not have influenced the frictional resistance since the frictional pullout force of
RI PT
the SF-LS fiber was similar to that of the SF-RF fiber.
SF-SH fiber
SC
The average Pmax of the SF-SH fibers was 621.3 N, which is 12 times larger than that of the SF-RF fibers. Thus, it can be conjectured that the pullout force of the SF-SH fiber was
M AN U
mainly from the anchoring action, and the SF-SH fibers did not show a decrease in pullout resistance as like the other types of fibers. The slopes of the curves were softened around the slip of 0.8-1.0 mm, which was similar to the length of the transition part of the spearhead. Then, a softened increase of the pullout force was followed till the slip of 5 mm, which was
TE D
the length of the spearhead. A similar phenomenon was observed in the pullout test of steel reinforcing bars; the pullout resistance increased until the slip of the length of the transition part of a rib (Choi et al., 2014; Lee et al., 2012; Choi et al., 2011).
EP
The pullout stresses of the SF-SH-1 fibers reached about 570 MPa, which exceeded the
AC C
beginning stress for the state transformation and was just below the stress for beginning the SIM hardening of the fiber (see Figure 1). For the SF-SH-3 fiber, the anchoring resistance appeared to be sufficient to increase the pullout stress of the fiber beyond the upper plateau stress of the SMA; thus, the fiber experienced SIM hardening. Therefore, during this stage, the fiber was slipped accompanied by deformation. The specimen of the SF-SH-3 fiber was fractured because of the tensile fracture of the mortar (see Figure 10c); this was splitting bond failure. The splitting bond failure occurred when the thickness of the mortar was not sufficient and did not provide appropriate confinement for the fiber. For the SF-SH-2 fiber, 12
ACCEPTED MANUSCRIPT
however, its anchoring resistance seemed not to be sufficient to induce full SIM hardening in the fiber; the specimen was fractured at the pullout stress of 670 MPa. In this case, the fiber was slipped through the duct without any additional large deformation of the fiber. For the
RI PT
SF-SH-1 fiber, the fiber was broken at the transition point of the spearhead (see Figure 10b); it appears that stress concentration occurred at the point. The pullout stress at the fracturepoint was about 570 MPa, and the SIM hardening was not observed. From the above
SC
observations, it was found that the superelastic SMA fiber with the spearhead exceeded the upper plateau stress of the SMA for state transformation. It also appears that the fiber can
M AN U
show full SIM hardening behavior if the mortar provides sufficient confinement with appropriate thickness.
4.2 Hysteretic behavior
TE D
Figures 11 and 12 show the hysteretic pullout behaviors of the fibers. During the tests, unloading and reloading were conducted 4 times before the test’s completion, and the increase in the slip for unloading was determined based on the monotonic curves. In Figure
EP
11, the three fibers of the SF-RF, -CR, and -LS showed vertical unloading paths. This indicates that the fibers were in elastic range and did not reach the state transformation. It
AC C
also intimates that their pullout displacements were mainly caused by the slip of the fibers. However, for the SF-SH fibers in Figure 12, they showed the flag-shaped behavior during the unloading and reloading procedures and recovered a part of the applied slip plus deformation. This verifies that the SF-SH fibers experienced the state transformation and reached the stress for beginning the upper plateau stress of the fiber. In Figures 12a, b, and c, the slip+ deformations of the first unloading were almost not recovered. This indicates that the fibers were in elastic range and the movement was mainly caused by the slip. The other
13
ACCEPTED MANUSCRIPT
noticeable thing is that the stresses of the fibers at the first unloading point did not get fully into the upper plateau stress. However, in Figure 12b, the fiber at the first unloading point showed a stress of 540 MPa, which exceeded the beginning of the upper plateau stress, and,
RI PT
thus, it recovered a part of the slip plus deformation and showed the flag-shaped behavior. The total movement of the fiber at the first unloading point was 2.09 mm, and 0.59 mm was recovered during the unloading. Thus, the slip at the first unloading point was 1.50 mm, and
SC
the deformation of the fiber was 0.59 mm, which was the recovered movement. The total movement from the first reloading to the second unloading was 2.78 mm, and 1.89 mm was
reloading process was 0.89 mm.
M AN U
recovered during the unloading, which was the deformation of the fiber. Thus, the slip in the
The beginning stress of the state transformation of the fiber in the pullout tests was a little higher than that of the tensile test in Figure 1, and the reason seems to be the
TE D
contribution of friction around the fiber. For the SF-SH fibers, their pullout behavior included the slip and deformation of the fibers, and, thus, except for the slip, the behavior was similar to that of the direct tensile test. The slip of the SF-SH fibers initially occurred
EP
due to the transitional shape of the spearhead, and mortar crushing around the spearhead
AC C
subsequently induced the slip; the white part of the mortar specimen in Figure 13 indicates the slip due to the crushing of the mortar. Therefore, the anchoring action of a superelastic SMA fiber should be greater than the upper plateau stress of the SMA to induce the flagshaped behavior and to recover the deformation. The slip is related to the shape of the spearhead and strength of the mortar. If a higher strength of the mortar was used and the mortar was not crushed, the slip of the SF-SH fiber would be restricted within the initial slip of the spearhead. Otherwise, if the upper plateau stress of the superelastic SMA was lower, then the fiber could easily get into the state transformation and deform before the slip
14
ACCEPTED MANUSCRIPT
occurring due to the crushing of mortar. Therefore, for the practical use of superelastic SMA fibers, a relatively low upper plateau stress is desirable for use with normal concrete.
RI PT
5. Discussion of results
In the monotonic tests; the results were compared in Figure 14, the fibers with a straight shape and crimped end showed basic behavior of fibers in mortar or concrete. The L-shaped fibers
SC
increased the pullout resistance and energy dissipation compared with the above two fibers. However, the fiber also showed a decreased pullout resistance after a slip of 5 mm. The fibers
M AN U
with the spearhead showed totally different pullout behavior from the conventional fibers for mortar or concrete. They showed a hardening behavior after the initial slip. Moreover, the developed stress in the fiber due to the pullout resistance reached the stress for the state transformation of the SMA. This indicates that the spearhead provided sufficient anchoring
TE D
action to induce the state transformation in the fiber.
The main purpose of using superelastic SMAs is to provide self-centering capacity for a system, a member, or an element. As shown in Figures 11 and 12, only the SMA fiber with the
EP
spearhead showed a flag-shape in hysteretic behavior; this provided self-centering capacity. For the other three fibers, the displacements of the fibers were mainly composed of the slip in
AC C
mortar matrix; this was not recovered during the unloading. Their deformations were elastic, which were recovered; however, the amount of the recovery was insignificant. For the SF-SH fiber with the spearhead, the displacement of the fiber consisted of the slip and deformation of the fiber; the deformation included elastic deformation as well as deformation of the stat transformation. Thus, the deformation of the SF-SH fiber was totally recovered due to the superelasticity. Slip of the fiber occurs when the pullout force is greater than the anchoring resistance; the pullout force is mainly dependent on the stress of the fiber. Therefore, if the
15
ACCEPTED MANUSCRIPT
anchoring resistance of the SF-SH fiber is greater than the upper plateau stress of the SMA, the slip of the fiber will not occur. In that case, the hysteretic behavior would be similar to the hysteretic tensile behavior in Figure 1. Or, if the upper plateau stress of the SMA is lower than
RI PT
the anchoring resistance of the fiber, the pullout force would be smaller than the anchoring action, and the fiber’s slip would not occur.
In this study, the SF-LS fiber with the bended end failed to reach the upper plateau stress and to
SC
induce the state transformation. However, Shajil et al. (2013) used superelastic SMA fibers with a double bended end; thus, the end of the fibers looked like an N-shape. The fibers with the N-
M AN U
shaped end exceeded the stress for state transformation; however, they did not conduct a hysteretic test, and the flag-shape was not observed. The bending at the end was not easy, in particular for the double bending. Thus, the cramped spearhead is appeared to be more effective
6. Conclusions
TE D
in providing sufficient anchoring action than the double bended end.
This study investigated the pullout resistance of the NiTi superelastic SMA fibers with different
EP
end shapes; four types of end shapes, namely, straight end shape, crimped end, bended end with L-shape, and spearhead end, were employed. Monotonic and hysteretic pullout tests were
AC C
conducted to understand the pullout behavior of the superealstic SMA fibers. The following conclusions can be drawn from the experimental study:
● Pullout resistance of the superelastic SMA fiber should be larger than the upper plateau stress of the SMA to show the flag-shaped behavior and provide self-centering capacity.
● Spearhead shape of the SMA fiber was appropriate to provide sufficient anchoring action to
16
ACCEPTED MANUSCRIPT
reach the upper plateau stress of the superelastic SMA.
● The SMA fiber with the spear head showed flag-shaped behavior during hysteretic tests; this
RI PT
indicates that the spearhead shaped SMA fiber has the capacity of self-centering.
action and to induce the state transformation of the SMA.
SC
● End-bended with L-shape of the SMA fiber was not effective to provide sufficient anchoring
M AN U
In this study, the mortar specimens with the spearhead shaped fibers experienced splitting failure; thus, the ultimate pullout behavior of the fiber was not observed. Therefore, further studies with a sufficiently large mortar specimen are necessary to induce the pullout failure of the SMA with the spearhead. In addition, further studies should investigate an appropriate
Acknowledgments
TE D
upper plateau stress for state transformation to avoid the fiber’s slip in a mortar matrix.
EP
This research was supported by the Basic Science Research Program through the National
AC C
Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (Project No. 2015-041523)
17
ACCEPTED MANUSCRIPT
References Gao N, Jeon JS, Hodgson DE and DesRoches R. An innovative seismic bracing system based
RI PT
on a superelastic shape memory alloy ring, Smart Materials and Structures 2015; 25(5): 055030. Ozbulut OE and Hurlebaus S. Re-centering variable friction device for vibration control of
SC
structures subjected to near-field earthquakes, Mechanical Systems and Signal Processing 2011;
M AN U
25:2849-2862.
Lafortune P, McCormick J, DesRoches R and Terriault P. Testing of superelastic recentering pre-strained braces for seismic restraint design, Journal of Earthquake Engineering 2007; 11:383-399.
TE D
McCormick J, DesRoches R, Fugazzaa D and Auricchio F. Seismic vibration control using superelastic shape memory alloys, Journal of Engineering Materials and Technology 2006;
EP
128:294-301.
AC C
Abdulridha A, Palermo D, Foo S and Vecchio FJ. Behavior and modeling of superelastic shape memory alloy reinforced concrete beams, Engineering Structures 2013; 49:893-904. Wierschem N and Andrawes B. Superelastic SMA-FRP composite reinforcement for concrete structures, Smart Materials and Structures 2010: 19:025011.
18
ACCEPTED MANUSCRIPT
Kuang Y and Ou J. Self-repairing performance of concrete beams strengthened using superealstic SMA wires in combination with adhesives released from hollow fibers, Smart
RI PT
Materials and Structures 2008; 17:025020.
Di S, Ji S, Hua W, Li H and Steve Z. Constraint superelastic SMA based experimental study of
SC
self-monitoring and repairing of concrete beam crack, Second International Conference on
M AN U
Smart Materials and Nanotechnolgy in Engineering, Porc. of SPIE 2009;7493:74930N-1. Sakai Y, Kitagawa Y, Fukita T and Iiba M. Experimental study on enhancement of selfrestoration of concrete beams using SMA wire, Smart Systems and Nondustructive Evaluation, Proc.of SPIE 2003; 5057.
TE D
Reedlunn B, Daly S and Shaw J. Superelastic shape memory alloy cables: Part I – Isothermal tension experiments, International Journal of Solid and Structures 2013a; 50:3009-3026.
EP
Reedlunn B, Daly S and Shaw J. Superelastic shape memory alloy cables: Part II –
AC C
Subcomponent isothermal responses, International Journal of Solid and Structures 2013b; 50:3027-3044.
Wang W, Chan TM, Shao H and Chen Y. Cyclic behavior of connections equipped with NiTi shape memory alloy and steel tendons between H-shaped beam to CHS column, Engineering Structures 2015; 88:37-50.
19
ACCEPTED MANUSCRIPT
Speicher MS, DesRoches R and Leon RT. Experimental results of a NiTi shape memory alloybased recentering beam-column connection, Engineering Structures 2011; 33:2448-2457.
RI PT
Youssef MA, Alam MS and Nehdi M. Experimental Investigation on the Seismic Behavior of Beam-Column Joints Reinforced with Superelastic Shape Memory Alloys, Journal of
SC
Earthquake Engineering 2008; 12:1205-1222.
M AN U
Ma H, Wilkinson T and Cho C. Feasibility study on a self-centering beam-to-column connection by using the superelastic behavior of SMAs, Smart Materials and Structures 2007; 16:1555-1563.
Billah AHMM and Alam SM, Seismic performance of concrete columns reinforced with hybrid
TE D
shape memory alloy (SMA) and fiber reinforced polymer (FRP) bars, Construction and Building Materials 2012; 28:730-742.
EP
Saiidi MS, O’Brien M, Sadrossadat-Zadeh M. Cyclic response of concrete bridge columns
AC C
using superelastic Nitinol and bendable concrete, ACI Structural Journal 2009; 160(1):69-77. Saiidi MS and Wang H. Exploratory study of seismic response of concrete columns with shape memory alloys reinforcement, ACI Structural Journal 2006; 103(3):436-443.
Sun L, Liang D, Gao Q and Zhou J. Analysis of factors affecting the self-repair capability of SMA wire concrete beam, Mathematical Problems in Engineering 2013; 2013:138162.
20
ACCEPTED MANUSCRIPT
Billah AHM M and Alam MS. Seismic performance of concrete columns reinforced with hybrid shape memory alloy (SMA) and fiber reinforced polymer (FRP) bars, Construction and
RI PT
Building Materials 2012; 28:730-742.
Shajil N, Srinicasan SM and Santhanam M. Self-centering of shape memory alloy fiber
SC
reinforced cement mortar members subjected to strong cyclic loading, Materials and Structures
M AN U
2013; 46(4):651-61
Shajil N, Srinicasan SM and Santhanam M. An experimental study on self-centering and ductility of pseudo-elastic shape memory alloy (PESMA) fiber reinforced beam and beamcolumn joint specimens, Materials and Structures 2016; 49:783-793.
TE D
Choi E, Kim DJ, Chung YS and Nam TH. Bond-slip characteristics of SMA reinforcing fibers obtained by pull-out tests, Materials Research Bulletin 2014; 58:28-31.
EP
Kim DJ, Kim HA, Chung YS and Choi E. Pullout resistance of straight NiTi shape memory
AC C
alloy fibers in cement mortar after cold drawing and heat treatment, Composites Part B: Engineering 2014; 67:588-594. Choi E, Kim DJ, Youn H and Nam TH. Repairing cracks developed in mortar beams reinforced by cold-drawn NiTi or NiTiNb SMA fibers, Smart Materials and Structures 2015; 24:125010. Kim MK, Kim DJ, Chung YS and Choi E. Direct tensile behavior of shape memory alloy fiber reinforced cement composites, Construction and Building Materials 2016; 102:462-470.
21
ACCEPTED MANUSCRIPT
Choi E, Kim DJ, Jeon C and Gin S. New SMA short fibers for cement composites manufacturing by cold drawing, Journal of Materials Science Research 2016; 5(2):68-81.
RI PT
Choi E, Cho BS, Jeon JS and Yoon SJ. Bond behavior of steel deformed bars embedded in concrete confined by FRP wire jackets, Construction and Building Materials 2014; 68:716-725.
SC
Lee H, Choi E, Cho SC, Park T. Bond and splitting behavior of reinforced concrete confined by
M AN U
steel jackets without grouting, Magazine of Concrete Research 2012; 64(3):225-237. Choi E, Chung YS, Kim YW and Kim JW. Monotonic and cyclic bond behavior of confined concrete using NiTiNb SMA wires, Smart Materials and Structures 2011; 20:075016. Dawood M, El-Tahan MW, and Zhang B. Bond behavior of superelastic shape memory alloys
EP AC C
247.
TE D
to carbon fiber reinforced polymer composites. Composites Part B: Engineering 2015; 77:238-
22
RI PT
ACCEPTED MANUSCRIPT
SC
List of Tables
Table 1. Composition of matrix mixtures by weight ratio and compressive strength
M AN U
Table 2. Averages of maximum and equivalent bond strength of the fibers
List of Figures
TE D
Figure 1. Hysteretic tensile stress-strain curve of superelastic SMA wire Figure 2. Shapes and dimensions of four types of SMA fibers Figure 3. Mold and specimen with dimension for pullout test
EP
Figure 4. Test set-up for single-fiber pullout test Figure 5. Pullout behavior of SF-RF fibers
AC C
Figure 6. Pullout behavior of SF-CR fibers Figure 7. Pullout behavior of SF-LS fibers Figure 8. Pullout behavior of SF-SH fibers Figure 9. Maximum and equivalent bond strength of the fibers Figure 10. Test completed specimens of the SF-LS and –SH fibers Figure 11. Hysteretic pullout behavior of SF-RF, -CR and –LS fibers Figure 12. Hysteretic pullout behavior of SF-SH fibers
23
ACCEPTED MANUSCRIPT
Figure 13. Fractured surface of mortar specimen with SF-SH fiber
AC C
EP
TE D
M AN U
SC
RI PT
Figure 14. Comparison of pullout behavior of the four SMA fibers
24
RI PT
ACCEPTED MANUSCRIPT
Table 1. Composition of matrix mixtures by weight ratio and compressive strength
(Type 3)
0.15
High-range water-
Sand
reducing admixture
1.00
0.009
Compressive
Water
AC C
EP
TE D
M AN U
1.00
Fly ash
Silica
SC
Cement
0.35
strength (MPa) 50
ACCEPTED MANUSCRIPT
Table 2. Averages of maximum and equivalent bond strength of the fibers Wp
τ max
τ eq
(mm)
(MPa)
(N·mm) 254.6 266.9 197.7 239.7 521.9 710.8 508.0 580.2 1015.8 1142.9 1093.2 1084.0 860.3 5165.9 6450.6 4158.9
(MPa)
(MPa)
82.37 70.57 47.65 66.86 95.54 93.40 99.70 96.21 204.7 225.6 210.4 213.6 574.2 677.5 1145.5 799.1
AC C
1.37 1.17 0.79 1.11 1.20 1.16 1.24 1.20 2.55 2.81 2.62 2.66 7.14 8.43 14.3 9.96
SC
0.18 0.24 0.18 0.20 0.46 0.37 0.16 0.33 1.88 1.62 1.93 1.81 2.59 11.64 10.43 8.22
RI PT
σ max
M AN U
64.05 54.87 37.05 51.99 74.70 72.63 77.53 74.95 159.2 175.4 163.6 166.1 446.5 526.8 890.7 621.3
δ max
EP
SF-RF-1 SF-RF-2 SF-RF-3 Average SF-CR-1 SF-CR-2 SF-CR-3 Average SF-LS-1 SF-LS-2 SF-LS-3 Average SF-SH-1 SF-SH-2 SF-SH-3 Average
Pmax (N)
TE D
Fiber type
0.72 0.76 0.56 0.68 0.84 1.14 0.81 0.93 1.63 1.83 1.75 1.74 1.38 8.26 10.32 6.65
1000
550 MPa
SC
470 MPa
600
400
200
0 0
1
2
3
M AN U
Stess (MPa)
800
RI PT
ACCEPTED MANUSCRIPT
4
5
6
7
8
9
TE D
Strain (%)
AC C
EP
Figure 1. Hysteretic tensile stress-strain curve of superelastic SMA wire
10
11
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
(a) Pictures of the four SMA fibers and spear-head
(b) SF-SH fiber
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
(d) SF-LS fiber
EP
TE D
(c) SF-CR fiber
Figure 2. Shapes and dimensions of four types of SMA fibers
RI PT
ACCEPTED MANUSCRIPT
Unit : mm
25
M AN U
41.3
(a) Mold
(b) Specimen
(c ) Dimension
AC C
EP
TE D
Figure 3. Mold and specimen with dimension for pullout test
71.5
SC
25
25
ACCEPTED MANUSCRIPT
PULLOUT LOAD
RI PT
LOAD CELL
SC
FIBER GRIP SYSTEM
M AN U
FIBER
SPECIMEN
TE D
SPECIMEN GRIP SYSTEM
AC C
EP
Figure 4. Test setup for single-fiber pullout test
ACCEPTED MANUSCRIPT 90
110
(a) SF-RF-1
80
100
70 50
60
40
50
30
40 30
20
20
10
RI PT
Pullout force (N)
80
60
Pullout stress (MPa)
90
70
10
0
0 0
2
4
6
8
10
12
14
16
90
110
(b) SF-RF-2
100
M AN U
80
80
60
70
50
60
40
50 40
30
30
20
20
TE D
10 0 0
2
4
6
8
10
12
14
Pullout stress (MPa)
90
70
Pullout force (N)
SC
Slip + deformation (mm)
10 0 16
Slip + deformation (mm)
(c) SF-RF-3
80 70
AC C
60
90
50
60
40
50
30
40 30
20
20
10
10
0
0 0
2
4
6
8
10
12
Slip + deformation (mm)
Figure 5. Pullout behavior of SF-RF fibers
14
Pullout stress (MPa)
70
Pullout force (N)
100
EP
80
ACCEPTED MANUSCRIPT 140
(a) SF-CR-1
100
100
60
80 60
40 40 20
20
0
0 0
2
4
6
8
10
12
14
16
18
20
22
140
(b) SF-CR-2
100
Pullout force (N)
80
100 80
60
60
40
40
20
Pullout stress (MPa)
M AN U
120
SC
Slip + deformation (mm)
0
2
4
6
8
TE D
20
0
10
12
14
16
18
20
0 22
Slip + deformation (mm)
EP
140
(c) SF-CR-3
100
AC C
80
120
80
60
60
40
40
20
20
0
0 0
2
4
6
8
10
12
14
16
18
Slip + deformation (mm)
Figure 6. Pullout behavior of SF-CR fibers
20
22
Pullout stress (MPa)
100
Pullout force (N)
RI PT
80
Pullout stress (MPa)
Pullout force (N)
120
ACCEPTED MANUSCRIPT 210
(a) SF-LS-1
240
180
120
150
90
120 90
60
60 30
30
0
0 0
2
4
6
8
10
12
14
16
18
20
220
(b) SF-LS-2
270
40
M AN U
200
20
30
210
160
180
140 120
150
100
120
80
90
60
0 0
2
4
6
8
TE D
60
10
12
14
16
18
Pullout stress (MPa)
240
180
Pullout force (N)
SC
Slip + deformation (mm)
0 20
Slip + deformation (mm)
EP
(c) SF-LS-3 180
210 180
AC C
150
240
120
150
90
120 90
60
60
30
30
0
0 0
2
4
6
8
10
12
14
16
Slip + deformation (mm)
Figure 7. Pullout behavior of SF-LS fibers
18
20
Pullout stress (MPa)
210
Pullout force (N)
RI PT
Pullout force (N)
180
Pullout stress (MPa)
210 150
ACCEPTED MANUSCRIPT 600
(a) SF-SH-1
700
500 500
300
400 300
200 200 100
100
0
0 0
1
2
3
4
5
700
(b) SF-SH-2
800
M AN U
600
SC
Slip + deformation (mm)
Pullout force (N)
600
400
500 400
300
300
200
200
0 0
2
4
6
8
TE D
100
10
12
14
16
18
Pullout stress (MPa)
700
500
100 0 20
Slip + deformation (mm)
EP
1200
(c) SF-SH-3
1000
AC C
800
800
600
600
400
400
200
200
0
0 0
2
4
6
8
10
12
14
16
Slip + deformation (mm)
Figure 8. Pullout behavior of SF-SH fibers
18
20
Pullout stress (MPa)
1000
Pullout force (N)
RI PT
400
Pullout stress (MPa)
Pullout force (N)
600
ACCEPTED MANUSCRIPT
14
(a)
Average
12
RI PT
10 8 6
2 0
SF-RF
SF-CR
SF-LS
Specimen
12
Average
10
TE D
8 6 4 2
SF-RF
SF-CR
AC C
0
EP
Equivalent bond strength (MPa)
(b)
SF-SH
SC
4
M AN U
Maximum bond strength (MPa)
16
SF-LS
SF-SH
Specimen
Figure 9. Maximum and equivalent bond strength of the fibers
(b) SF-SH-1
EP
(a) SF-LS fiber
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
(c) SF-SH-2 & 3
AC C
Figure 10. Test completed specimens of the SF-LS and -SH fibers
ACCEPTED MANUSCRIPT 80
100
(a) SF-RF-hysteresis curve
40 20 20
0
0 0
2
4
6
8
10
12
14
16
100 120
M AN U
(b) SF-CR-hysteresis curve 80
80
60
60
40
40
20
20
0 0
2
4
6
8
10
12
14
16
18
Pullout stress (MPa)
100
TE D
Pullout force (N)
SC
Slip + deformation (mm)
RI PT
Pullout force (N)
60 40
Pullout stress (MPa)
80
60
0 20
Slip + deformation (mm)
EP
300
(c) SF-LS-hysteresis curve
250
AC C
Pullout force (N)
200
200
150
150
100
100
50
50
0
Pullout stress (MPa)
250
0 0
2
4
6
8
10
12
14
16
18
20
Slip + deformation (mm)
Figure 11. Hysteretic pullout behavior of SF-RF, -CR, and –LS fibers
ACCEPTED MANUSCRIPT 800
1000
(a) SF-SH-1-hysteresis curve
700
900
600
400
500
300
400 300
200
200 100
100
0
0 0
1
2
3
4
5
6
7
8
9 10 11 12 13 14
700
900
(b) SF-SH-2-hysteresis curve
500
600
400
500 400
300
300
200
200
Slip
Deformation
TE D
100 0 0
1
2
3
4
5
6
7
Pullout stress (MPa)
700
Total movement
Pullout force (N)
800
M AN U
600
SC
Slip + deformation (mm)
100 0 8
Slip + deformation (mm)
(c) SF-SH-3-hysteresis curve
500
AC C
Pullout force (N)
600
900 800 700 600
400
500
300
400 300
200
200
100
100
0
0 0
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15
Slip + deformation (mm)
Figure 12. Hysteretic pullout behavior of SF-SH fiber
Pullout stress (MPa)
EP
700
RI PT
Pullout force (N)
700 500
Pullout stress (MPa)
800
600
SC
RI PT
ACCEPTED MANUSCRIPT
M AN U
Slip
AC C
EP
TE D
Figure 13. Fractured surface of mortar specimen with SF-SH fiber
600
SF-SH
SC
400 300
SF-LS
200
M AN U
Pullout force (N)
500
RI PT
ACCEPTED MANUSCRIPT
SF-CR
100
SF-RF
0 0
2
4
6
8
10
12
TE D
Slip + deformation (mm)
AC C
EP
Figure 14. Comparison of pullout behavior of the four SMA fibers
14
16