- Email: [email protected]
Effectiveness of shrinkage-reducing admixture in reducing autogenous shrinkage stress of ultra-high-performance fiber-reinforced concrete
Cement and Concrete Composites 64 (2015) 27e36
Contents lists available at ScienceDirect
Cement and Concrete Composites journal homepage: www.elsevier.com/locate/cemconcomp
Effectiveness of shrinkage-reducing admixture in reducing autogenous shrinkage stress of ultra-high-performance fiber-reinforced concrete Doo-Yeol Yoo a, Nemkumar Banthia a, Young-Soo Yoon b, * a b
Department of Civil Engineering, The University of British Columbia, 6250 Applied Science Lane, Vancouver, BC V6T 1Z4, Canada School of Civil, Environmental and Architectural Engineering, Korea University, 145 Anam-ro, Seongbuk-gu, Seoul 136-713, Republic of Korea
a r t i c l e i n f o
a b s t r a c t
Article history: Received 14 July 2015 Received in revised form 9 September 2015 Accepted 20 September 2015 Available online 25 September 2015
This study describes the effect of shrinkage-reducing admixture (SRA) on free and restrained autogenous shrinkage behaviors of ultra-high-performance fiber-reinforced concrete (UHPFRC). To investigate the cracking potential, tensile strength development was experimentally obtained and predicted on the basis of the degree of hydration model. Three different SRA to cement weight ratios of 0, 1, and 2% and three different reinforcement ratios of 1.3, 2.9, and 8.0% were considered. A higher SRA content contributed to a slightly higher tensile strength and a lower autogenous shrinkage. In addition, a higher SRA content and a lower reinforcement ratio resulted in better restrained autogenous shrinkage behaviors, such as lower autogenous shrinkage stress and cracking potential. Therefore, it can be concluded that the use of SRA or a lower reinforcement ratio is favorable for improving the restrained shrinkage behaviors of UHPFRC. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Ultra-high-performance fiber-reinforced concrete Autogenous shrinkage Time-zero Shrinkage-reducing admixture Reinforcement ratio Cracking potential
1. Introduction Recently, in order to overcome several drawbacks of conventional concrete such as low tensile strength, low ductility, and low strength to weight ratio, ultra-high-performance fiber-reinforced concrete (UHPFRC) has been developed with excellent strength, i.e., compressive strength > 150 MPa and tensile strength > 8 MPa, ductility, energy absorption capacity, and durability [1]. These outstanding properties are attributed to a low water-to-binder ratio (W/B) equal to 0.2, an optimization of particle sizes of compositions, and high volume contents of steel fibers. Thus, this material has been attractive to thin-plate structures, i.e., long-span bridge decks, roofs, and thin-walls, and field-cast joints for precast bridge decks [2,3]. However, owing to its high early-age autogenous shrinkage and small cross-sectional area for thin-plate structures, UHPFRC is highly vulnerable to premature shrinkage cracking [2]. Cracking and residual stress that occur in concrete structures by the restraint of shrinkage are the main concerns with respect to
* Corresponding author. Tel.: þ82 2 3290 3320; fax: þ82 2 928 7656. E-mail address: [email protected] (Y.-S. Yoon). http://dx.doi.org/10.1016/j.cemconcomp.2015.09.005 0958-9465/© 2015 Elsevier Ltd. All rights reserved.
durability. The restrained shrinkage behavior of concrete is quite complex, since it is affected by many factors, such as the rate and magnitude of free shrinkage, developments of strength and elastic modulus, stress relaxation and creep, degree of restraint, and geometry of element [4]. In particular, because of stress relaxation and creep characteristics of concrete, free shrinkage measurement alone is insufficient to predict the cracking potential due to the restraint of shrinkage. A conceptual view of the restrained shrinkage behavior of concrete with an internal reinforcing bar (rebar) is shown in Fig. 1. If no rebar exists, concrete will be freely deformed in a direction of Dεþ by shrinkage εsh (b). On the contrary, if the shrinkage of concrete is restrained by the internal rebar, and if no creep effect is assumed, the strain can be divided into two categories: elastic restraint strain by rebar εe and elastic rebar stain εe,r (c). However, in reality, owing to the tensile creep of concrete εcr, the strain obtained in rebar (or in concrete) by shrinkage decreases to εr (d). Therefore, the free shrinkage strain εsh is equal to the summation of εe, εcr, and εr. In order to mitigate the early-age shrinkage of UHPFRC, the use of shrinkage-reducing admixture (SRA) has been investigated by Soliman and Nehdi [5]. Their studies were focused on investigating the implication of adding SRA on the free shrinkage, and the
28
D.-Y. Yoo et al. / Cement and Concrete Composites 64 (2015) 27e36
a W/B of 0.2 was used. Smooth steel fibers with a diameter of 0.2 mm and a length of 13 mm were incorporated at a 2% volume fraction within a mortar matrix. The detailed mechanical and geometrical properties of the steel fibers are given in Table 2. A high-performance water-reducing agent, polycarboxylate superplasticizer (SP) with a density of 1.06 g/cm3, was also added to provide suitable fluidity. In addition, to investigate the implication of SRA contents on shrinkage behaviors of UHPFRC under both free and restrained conditions, three different SRA to cement weight ratios (0, 1, and 2%) were considered (Table 1) using the glycolbased SRA (METOLAT® P 860) produced in Germany. 2.2. Experimental setup and procedure
Fig. 1. Conceptual view of restrained shrinkage behavior of concrete with internal rebar.
effectiveness of SRA on reducing the shrinkage strains has been reported. However, although steel rebars are generally included in UHPFRC structures [6,7], no published study exists on the effects of using SRA on the shrinkage behavior of UHPFRC restrained by internal steel rebars. This restricts the practical use of SRA to the UHPFRC structures in spite of its numerous advantages. Accordingly, in this study, the effect of SRA on both the free shrinkage and restrained shrinkage (by internal steel rebar) behaviors of UHPFRC was investigated. Owing to its extremely high autogenous shrinkage but insignificant drying shrinkage [2], autogenous shrinkage tests were conducted according to the recommendation by Japan Concrete Institute (JCI) [8]. Correspondingly, three different SRA to cement weight ratios were used, and deformed steel rebars with three different diameters were applied in restrained autogenous shrinkage tests in order to provide various degrees of restraint. In addition, tensile strength development of UHPFRC was measured and predicted to evaluate the cracking potential. 2. Experimental program 2.1. Materials and mix proportions The used mix proportions are presented in Table 1. Type 1 Portland cement with a specific surface area of 3413 cm2/g and a density of 3.15 g/cm3 and silica fume with a specific surface area of 200,000 cm2/g and a density of 2.10 g/cm3 were used as cementitious materials. The chemical and physical properties of the cementitious materials can be found elsewhere [2]. Sand with grain size smaller than 0.5 mm and silica flour with a diameter of 2 mm including 98% SiO2 were added to the mixture. For all test series,
2.2.1. Flow and direct tensile tests In order to quantitatively measure flowability, a flow table test was carried out according to ASTM C 1437 [9]. The average flow was calculated by averaging the maximum flow diameter and the corresponding perpendicular diameter. The average flow values are listed in Table 1. The fluidity was slightly increased by adding SRA, even though identical W/B and amount of SP were used. To evaluate the cracking potential, the tensile strength development of UHPFRC with age was investigated. Dog-bone-shaped specimens with a middle cross-section of 50 mm 100 mm were fabricated and cured at a room with a temperature of (23 ± 1) C and a humidity of (60 ± 5)%. The details of the geometry and test setup are shown in Fig. 2. A pin-fixed end condition was used to avoid secondary flexural stress and to ensure a centric-loading condition. The alignment of the specimen was also carefully checked using a plumb before testing. Three specimens for each variable were used at each age, and the load was applied through displacement control using a universal testing machine (UTM) with maximum load capacity of 250 kN. The displacement was increased at a rate of 0.4 mm/min and the applied load was measured using a load cell attached to the bottom of the crosshead. 2.2.2. Autogenous shrinkage tests under free and restrained conditions Prismatic specimens with cross-sectional dimension of 100 mm 100 mm and length of 1000 mm were used for measuring free autogenous shrinkage. In order to precisely measure the autogenous shrinkage from a very early age, a dumbbellshaped strain gage that has a nearly zero stiffness and a coefficient of thermal expansion (CTE) of 11 mε/ C e similar to that of hardened UHPFRC e and a thermocouple were set horizontally in the middle of the mold before concrete casting, as shown in Fig. 3. Teflon sheet and polyester film were placed on the mold to eliminate the friction between the mold and the concrete. After concrete casting, the top surface of each specimen was covered with a polyester film to prevent the moisture evaporation. In order to estimate restrained autogenous shrinkage stress of UHPFRC, prismatic specimens that have a dimension identical to that used in free autogenous shrinkage tests were prepared. Deformed steel rebars with three different nominal diameters of
Table 1 Mix proportions. Name
UH-S0 UH-S1 UH-S2
Relative weight ratios to cement Cement
Water
Silica fume
Silica sand
Silica flour
SP
SRA
1.00
0.25
0.25
1.10
0.30
0.018
0.00 0.01 0.02
[Note] SP ¼ superplasticizer and SRA ¼ shrinkage-reducing admixture.
Steel fiber (Vf)
Flow (mm)
2%
235 245 240
D.-Y. Yoo et al. / Cement and Concrete Composites 64 (2015) 27e36
29
Table 2 Properties of smooth steel fibers. df (mm)
Lf (mm)
Aspect ratio (Lf/df)
Density (g/cm3)
ft (MPa)
Es (GPa)
0.2
13.0
65.0
7.9
2500
200
Image
[Note] df ¼ diameter of fiber, Lf ¼ length of fiber, ft ¼ tensile strength of fiber, and Es ¼ elastic modulus of fiber.
Fig. 2. Direct tensile test; (a) geometry of dog-bone specimen (unit: mm), (b) test setup.
12.7 mm, 19.1 mm, and 31.8 mm, which lead to three different reinforcement ratios of approximately 1.3%, 2.9%, and 8.0%, were used to restrain the volume change of UHPFRC by autogenous shrinkage. The detailed properties of the rebars used are summarized in Table 3. In order to provide identical autogenous shrinkage stress at the center of the specimen, the ribbed edges of the rebars were lathed to within 150 mm from the center and they were covered with a Teflon sheet to prevent friction between the
Fig. 3. Dumbbell-shaped embedded strain gage and thermocouple for free autogenous shrinkage measurement.
concrete and rebar according to the recommendation of the technical committee on autogenous shrinkage of concrete of the JCI [8], as shown in Fig. 4. In addition, based on previous studies [10,11], an embedment length of 350 mm was applied at both ends to provide full autogenous shrinkage stress instead of 600 mm recommend by the JCI [8] for reducing the specimen size. A strain gage and a thermocouple were attached to the center of the rebar before covering it with a Teflon sheet to measure the strain and temperature. A Teflon sheet and a polyester film were also placed inside the mold to minimize the frictional force between the mold and concrete, and the top surface was covered with a polyester film immediately after concrete casting to prevent the moisture evaporation. For both free and restrained autogenous shrinkage tests, all test specimens were demolded after 24 h from concrete casting and were immediately sealed with aluminum adhesive tape. In addition, the shrinkage measurement began just after concrete casting and was performed in a room with a temperature of (23 ± 1) C and a humidity of (60 ± 5)%.
30
D.-Y. Yoo et al. / Cement and Concrete Composites 64 (2015) 27e36
Table 3 Properties of deformed steel rebar. Reinforcement ratio
dr (mm)
Ar (mm2)
Er (GPa)
fy (MPa)
εy (mm/mm)
fu (MPa)
r ¼ 1.3% r ¼ 2.9% r ¼ 8.0%
12.7 19.1 31.8
126.7 286.5 794.2
200.0
522.7 508.5 509.0
0.00261 0.00254 0.00255
627.6 625.2 601.1
[Note] r ¼ reinforcement ratio, dr ¼ nominal diameter of rebar, Ar ¼ area of rebar, Er ¼ elastic modulus of rebar, fy ¼ yield strength of rebar, εy ¼ yield strain of rebar, and fu ¼ ultimate strength of rebar.
Fig. 4. Test setup for autogenous shrinkage stress measurement (unit: mm) (modified from JCI recommendation [8]).
3. Experimental results and discussion
by Janasson [16], as follows:
3.1. Tensile strength development
o n ft ðtÞ ¼ ft28 exp l1 ½lnð1 þ ðt t0 ÞÞk1
Fig. 5 shows the tensile strength development of UHPFRC at various SRA contents. Tensile strength slightly increased with an increase in the SRA content. For example, the 28-day tensile strengths were found to be 7.4 MPa for UH-S0, 7.8 MPa for UH-S1, and 8.0 MPa for UH-S2. In a similar way, the improvement of 28day flexural strength of high-strength steel-fiber-reinforced concrete (SFRC) with the addition of SRA was observed by Wang et al. [12]. However, these results differed from the previous tensile test results of the steam cured UHPFRC at 90 C [13]: lower tensile strength was obtained for the specimens including SRA for the case of steam curing. This discrepancy is due to the different curing conditions, and further investigation must be conducted to accurately prove this observation. Since the tensile strengths were only obtained at specific dates, an equation for predicting the tensile strength development of UHPFRC with age needs to be suggested to analyze the cracking potential. In accordance with a previous research [14], the strength development of UHPFRC exhibited almost linear relationship with the degree of hydration. Therefore, in order to predict the tensile strength development of UHPFRC, the following equation (Eq. (1)) was adopted [15] based on the degree of hydration model proposed
(1)
where ft28 is the tensile strength after 28 days, t0 is the time when the shrinkage stress first develops, and l1 and k1 are the regression coefficients, summarized in Table 4. This Eq. (1) includes the 28-day tensile strength, which is similar to the Graybeal's formula [17], and as shown in Fig. 5, the predicted values are in good agreement with the experimental results. 3.2. Autogenous shrinkage behavior To obtain pure autogenous shrinkage, the thermal dilation needs to be compensated from the measured strain using Eq. (2), as expressed by
εas ¼ εm a DT
(2)
where εas is the autogenous shrinkage strain, εm is the measured strain, a is the CTE of concrete, and DT is the temperature variation. In general, the CTE of concrete varies with age at an early age [18]. However, not only is measuring the early age CTE of concrete difficult, but also the effect of the CTE variation on autogenous shrinkage is insignificant because of its small cross-sectional area
D.-Y. Yoo et al. / Cement and Concrete Composites 64 (2015) 27e36
31
Fig. 5. Tensile strength development of UHPFRC with various SRA contents; (a) UH-S0, (b) UH-S1, (c) UH-S2.
Table 4 Non-linear regression coefficients for tensile strength development in Eq. (1). Name
l1
k1
R2
UH-S0 UH-S1 UH-S2
0.240 0.272 0.332
1.067 1.241 1.096
0.93046 0.93942 0.91867
(100 mm 100 mm) of the prismatic specimen used in the current study. For these reasons, a constant CTE of 11 mε/ C, as recommended by AFGC/SETRA recommendations [19], was adopted. According to the technical committee on autogenous shrinkage of concrete of the JCI [8], the purpose of measuring autogenous shrinkage is to predict cracking response of concrete. Thus, they suggested the use of the initial point of autogenous shrinkage measurement, called “time-zero”, as the time of the initial setting, in order to exclude the volume change that occurs when concrete is still fresh. However, several researchers have reported [20,21] that the use of setting time as a time-zero for the autogenous shrinkage measurement is erroneous, because the start point of selfdesiccation, which is the main mechanism of autogenous shrinkage, is influenced by bleeding water and the setting time cannot account for this effect. For this reason, the authors used the start time of the shrinkage stress development in concrete as the time-zero, according to a previous study [15].
Fig. 6 shows the comparison of a 30-day autogenous shrinkage strains measured at different zeroing points. The highest strain was obtained as the start time of the shrinkage stress development was determined as time-zero. It is interesting to note that autogenous shrinkage strains measured from the start time of shrinkage stress development exhibited very similar values to those measured from the deviation point of shrinkage strain and internal temperature, as reported by Yoo et al. [15]. Therefore, the deviation point of measured strain and temperature at a very early age can be used as the time-zero for UHPFRC. On the other hand, significantly lower autogenous shrinkage strains were obtained when the initial and final setting times were used as time-zero. This was because UHPFRC showed a very steep increase in autogenous shrinkage at early age. The 30-day autogenous shrinkage strains measured from the initial and final sets were respectively 25% and 55% lower, on average, than those measured from the start time of the shrinkage stress development. From these observations, the use of the initial and final setting times as the time-zero point appears to be inappropriate for UHPFRC. Fig. 7(a) shows the autogenous shrinkage of UHPFRC measured from the start time of shrinkage stress development, which is equal to the time when the rebars start to deform as a result of concrete shrinkage. Regardless of the SRA contents, autogenous shrinkage rapidly increased just after time-zero. After certain points (i.e., 15 h for UH-S0, 17 h for UH-S1, and 18 h for UH-S2), the increase rate of
32
D.-Y. Yoo et al. / Cement and Concrete Composites 64 (2015) 27e36
Fig. 6. Comparison of 30-day autogenous shrinkage strains measured from start time of shrinkage stress development ðt0 * Þ, deviation time of strain and temperature ðt0 y Þ, initial setting time (ti), and final setting time (tf).
Fig. 8. Comparison of 30-day autogenous shrinkage of UHPFRC with different SRA contents.
Fig. 7. Autogenous shrinkage and temperature responses of UHPFRC with different SRA contents; (a) autogenous shrinkage, (b) temperature.
shrinkage strain suddenly decreased and slight expansion was observed. The main reason for sudden decrease of shrinkage increase rate was that the chemical shrinkage and the volume contraction from the negative pressure in the internal voids were self-restrained by the hardening of concrete. The point where the increase rate of shrinkage strain abruptly decreased was delayed with the inclusion of SRA. It is interesting to note that a significant reduction of autogenous shrinkage with the addition of SRA was
observed at a very early age (before 18 h). In addition, as shown in Fig. 8, the additions of 1% and 2% SRA resulted in 15.2% and 28.4% reductions of autogenous shrinkage after 30 days, respectively. The 30-day autogenous shrinkage of UH-S0 was found to be approximately 760 mε, which is much larger than values reported for normal- and high-strength concretes [22] and often sufficient to cause cracking. With the addition of SRA, the maximum temperature decreased and the time when the maximum temperature reached was delayed, as shown in Fig. 7(b). The maximum temperature of UH-S0 was found to be 34.6 C after nearly 0.67 days, which is 4% and 13% higher than the maximum temperature values of UH-S1 and UH-S2, respectively. 3.3. Restrained shrinkage behavior 3.3.1. Autogenous shrinkage stress The responses of strain and temperature in the rebars are shown in Fig. 9. Based on the restrained UHPFRC slab test results performed by Yoo et al. [23], the steel rebar strain and the strain in the UHPFRC near the steel rebar showed very similar values to each other owing to the restraint of shrinkage. Therefore, the slip deformation between the rebar and UHPFRC was ignored in this
D.-Y. Yoo et al. / Cement and Concrete Composites 64 (2015) 27e36
33
Fig. 9. Restrained shrinkage behaviors of UHPFRC with different SRA contents and reinforcement ratios; (a) UH-S0 (r ¼ 1.3%), (b) UH-S0 (r ¼ 2.9%), (c) UH-S0 (r ¼ 8.0%), (d) UH-S1 (r ¼ 1.3%), (e) UH-S1 (r ¼ 2.9%), (f) UH-S1 (r ¼ 8.0%), (g) UH-S2 (r ¼ 1.3%), (h) UH-S2 (r ¼ 2.9%), (i) UH-S2 (r ¼ 8.0%).
study. To obtain pure strains in the rebar from concrete shrinkage, the thermal dilation needed to be compensated by adopting an appropriate CTE. The CTE of the deformed steel rebar was used with a value of 11.7 mε/ C [24]. As shown in Fig. 9, a higher strain was obtained with a smaller diameter of rebar regardless of the SRA content. The strains without thermal dilation yielded higher values at early age than the measured strains because of the hydration heat. However, since the thermal strain returned from expansion to shrinkage during the cooling stage, the strains without thermal dilation became very similar to the measured strains. The increase of strains in the rebars gradually decreased with age and finally converged to a stable value after 30 days. This was attributed to the stabilization of both the autogenous shrinkage and the elastic modulus development. The JCI committee on autogenous shrinkage [8] suggested an equation to calculate the autogenous shrinkage stress sc in concrete as follows:
sc ¼
Er εr Ar Ac
(3)
where Er is the elastic modulus of rebar, εr is the strain obtained in rebar excluding the thermal dilation, and Ar and Ac are the areas of the rebar and concrete, respectively. Fig. 10 shows the comparison of tensile strengths and autogenous shrinkage stresses at various SRA contents and reinforcement ratios. The tensile strengths calculated by Eq. (1) were much higher than the autogenous shrinkage stresses for all test series. Therefore, no shrinkage crack was observed during testing. Regardless of the SRA content, a higher autogenous shrinkage stress was obtained with a higher reinforcement ratio. The autogenous shrinkage stress increased more gradually at an early age compared with the free autogenous shrinkage strain. The free autogenous shrinkage strains at a very early age (at nearly 18 h) were found to be approximately 78% (on average) of the 30-day autogenous shrinkage strains. On the other hand, the restrained autogenous shrinkage stresses at nearly 18 h were found to be approximately 60% (on average) of the
34
D.-Y. Yoo et al. / Cement and Concrete Composites 64 (2015) 27e36
Fig. 10. Comparison of autogenous shrinkage stress and tensile strength; (a) UH-S0, (b) UH-S1, (c) UH-S2.
3.3.2. Degree of restraint If the steel rebar has high enough stiffness to perfectly restrain the shrinkage of concrete, it will not be deformed. However, as shown in Fig. 9, the steel rebar was deformed by the shrinkage of concrete, and thus, the restraint strain εrs considering tensile creep effect was calculated by subtracting the strain in the rebar from the autogenous shrinkage strain, as expressed by
εrs ¼ εas εr
(4)
Therefore, the degree of restraint j is determined by the ratio of the restraint strain and the autogenous shrinkage strain, as given by Eq. (5):
j¼ Fig. 11. Comparison of 30-day autogenous shrinkage stresses according to reinforcement ratio and SRA content.
30-day restrained autogenous shrinkage stresses. This was mainly attributed to the low stiffness of UHPFRC at early age. As shown in Fig. 11, the 30-day autogenous shrinkage stress decreased with an increase in the SRA content. For example, the 30day autogenous shrinkage stress of UH-S0 with r ¼ 1.3% was found to be 1.3 MPa, approximately 11% and 22% higher than those of UHS1 and UH-S2 at an identical reinforcement ratio, respectively.
εrs εr ¼1 εas εas
(5)
Fig. 12 shows the degree of restraint at different SRA contents and reinforcement ratios. The degree of restraint decreased as time passed, owing to the increases in the shrinkage strain and elastic modulus of concrete. A higher degree of restraint was obtained at a higher reinforcement regardless of the SRA content. This was because a lower strain in rebar was obtained for the specimen with a higher reinforcement ratio. On the other hand, the degree of restraint was insignificantly influenced by the SRA content, as shown in Fig. 13. This was caused by the fact that both the strain in the rebar and autogenous shrinkage decreased with the addition of SRA. Thus, it was concluded that the degree of restraint was
D.-Y. Yoo et al. / Cement and Concrete Composites 64 (2015) 27e36
35
Fig. 12. Degree of restraint; (a) UH-S0, (b) UH-S1, (c) UH-S2.
marginally affected by the SRA content, but was significantly affected by the reinforcement ratio. For example, the degree of restraint for UH-S0 with r ¼ 8.0% was 0.68 after 30 days, approximately 40% and 94% higher than those of UH-S0 with r ¼ 2.9% and r ¼ 1.3%, respectively. 3.3.3. Cracking potential If there is no shrinkage cracking, the investigation of cracking potential is imperative to compare the restrained shrinkage
Fig. 13. Comparison of degree of restraints according to reinforcement ratio and SRA content.
behavior. The cracking potential is defined by dividing the autogenous shrinkage stress by the tensile strength, as expressed by
Qcr ¼
sc ft
(6)
where Qcr is the cracking potential. For all test series, the cracking potentials were much lower than 100%, as shown in Fig. 14.
Fig. 14. Comparison of cracking potentials according to reinforcement ratio and SRA content.
36
D.-Y. Yoo et al. / Cement and Concrete Composites 64 (2015) 27e36
Therefore, no shrinkage crack was observed during testing. The cracking potential increased with an increase in the reinforcement ratio, whereas it decreased with the inclusion of SRA. The highest cracking potential was found to be 63% for UH-S0 with r ¼ 8.0%, whereas the lowest cracking potential was found to be 14% for UHS2 with r ¼ 1.3%. This indicates that even though shrinkage crack is not observed, the structures made of UHPFRC at a high reinforcement ratio are likely to be cracked under a low tensile stress state, owing to their high residual stress as a result of the restraint of shrinkage. Yoo and Yoon [25] recently reported that the first cracking load of steel bar-reinforced UHPFRC beams decreased with an increase in the reinforcement ratio due to the higher residual stress. 4. Conclusions This study investigated the effects of SRA content and reinforcement ratio on autogenous shrinkage behaviors of UHPFRC under free and restrained conditions. Based on the test results, the following conclusions can be drawn: 1) The 28-day tensile strength of non-steam-cured UHPFRC slightly increased with an increase in the SRA content up to 2%. 2) The time-zero of autogenous shrinkage measurement was suggested as the start time of shrinkage stress development. The deviation point between measured strain and temperature at early age was also proper as the time-zero for UHPFRC, whereas the initial and final setting times were inappropriate. 3) The autogenous shrinkage strain decreased with an increase in the SRA content. A significant reduction of autogenous shrinkage by SRA was observed at a very early age (before 18 h). After 30 days, the inclusion of 1% and 2% SRA led to approximately 15.2% and 28.4% reductions of autogenous shrinkage strains, respectively. 4) For all test series, the tensile strengths were higher than the autogenous shrinkage stresses. Thus, no shrinkage crack was observed during testing. In addition, autogenous shrinkage stress decreased at lower reinforcement ratios and higher SRA contents. 5) The degree of restraint decreased with age due to the increase of shrinkage and elastic modulus in concrete. A higher degree of restraint was obtained with a higher reinforcement ratio, but it was marginally influenced by the SRA content. 6) The cracking potentials for all test series were much lower than 100%, and thus, no shrinkage crack was detected. In addition, lower cracking potentials were obtained with lower reinforcement ratios and higher SRA contents. Acknowledgments This research was supported by a grant (13SCIPA01) from Smart Civil Infrastructure Research Program funded by Ministry of Land, Infrastructure and Transport (MOLIT) of Korea government and Korea Agency for Infrastructure Technology Advancement (KAIA).
References [1] P. Richard, M. Cheyrezy, Composition of reactive powder concretes, Cem. Concr. Res. 25 (7) (1995) 1501e1511. [2] D.Y. Yoo, H.O. Shin, J.Y. Lee, Y.S. Yoon, Enhancing cracking resistance of ultrahigh-performance concrete slabs using steel fibers, Mag. Concr. Res. 67 (10) (2015) 487e495. [3] V. Perry, W. Gary, Innovative field cast UHPC joints for precast bridge decksedesign, prototype testing and projects, in: International Workshop on UHPFRC, Marseille, France, 2009. [4] W.J. Weiss, W. Yang, S.P. Shah, Shrinkage cracking of restrained concrete slabs, J. Eng. Mech. 124 (7) (1998) 765e774. [5] A.M. Soliman, M.L. Nehdi, Effect of drying conditions on autogenous shrinkage in ultra-high performance concrete at early-age, Mater. Struct. 44 (5) (2011) 879e899. [6] M.A. Saleem, A. Mirmiran, J. Xia, K. Mackie, Ultra-high-performance concrete bridge deck reinforced with high-strength steel, ACI Struct. J. 108 (5) (2011) 601e609. [7] D.Y. Yoo, N. Banthia, S.W. Kim, Y.S. Yoon, Response of ultra-high-performance fiber-reinforced concrete beams with continuous steel reinforcement subjected to low-velocity impact loading, Compos Struct. 126 (2015) 233e245. [8] Japan Concrete Institute, in: E. Tazawa (Ed.), Committee report, Autogenous Shrinkage of Concrete, E&FN Spon, 1999. [9] ASTM, C 1437-07 Standard Test Method for Flow of Hydraulic Cement Mortar, American Society of Testing and Materials, West Conshohocken, PA, 2007, p. 2. [10] T. Nakagawa, Y. Ohno, Investigation of test method of self stress due to autogenous shrinkage in concrete, Proc. Jpn. Concr. Inst. 20 (2) (1998) 751e756 (in Japanese). [11] D.Y. Yoo, Performance Enhancement of Ultra-high-performance Fiber-reinforced Concrete and Model Development for Practical Utilization (PhD thesis), Korea University, Seoul, Korea, 2014, p. 586. [12] J.Y. Wang, N. Banthia, M.H. Zhang, Effect of shrinkage reducing admixture on flexural behaviors of fiber reinforced cementitious composites, Cem. Concr. Compos 34 (4) (2012) 443e450. [13] D.Y. Yoo, S.T. Kang, J.H. Lee, Y.S. Yoon, Effect of shrinkage-reducing admixture on tensile and flexural behaviors of UHPFRC considering fiber distribution characteristics, Cem. Concr. Res. 54 (2013) 180e190. , E. Brühwiler, Thermal effects on physico-mechanical [14] A. Kamen, E. Denarie properties of ultra-high-performance fiber-reinforced concrete, ACI Mater. J. 104 (4) (2007) 415e423. [15] D.Y. Yoo, J.J. Park, S.W. Kim, Y.S. Yoon, Influence of reinforcing bar type on autogenous shrinkage stress and bond behavior of ultra high performance fiber reinforced concrete, Cem. Concr. Compos 48 (2014) 150e161. [16] J.E. Jonasson, Slipform Construction-calculations for Assessing Protection against Early Freezing, Fo 4: 84, Swedish Cement and Concrete Institute, Stockholm, 1985, pp. 1e13. [17] B.A. Graybeal, Compressive behavior of ultra-high-performance fiber-reinforced concrete, ACI Mater. J. 104 (2) (2007) 146e152. [18] Ø. Bjøntegaard, E.J. Sellevold, Interaction between thermal dilation and autogenous deformation in high performance concrete, Mater. Struct. 34 (5) (2001) 266e272. nie Civil, Ultra high performance fibre-reinforced [19] Association Française de Ge concretes, in: Interim Recommendations, AFGC/SETRA, Bagneux, France, 2002, p. 152. [20] T.A. Hammer, H. Justnes, Ø. Bjøntegaard, E.J. Sellevold, 8.2 Suggestions on the terminology and the test methods proposed by JCI, in: E. Tazawa (Ed.), Autogenous Shrinkage of Concrete. Proceedings of the International Workshop, E&FN Spon, 1998, pp. 397e399. [21] E. Holt, Early Age Autogenous Shrinkage of Concrete, VTT Publication 446, Technical Research Centre of Finland, 2001, pp. 1e184. [22] K.M. Lee, H.K. Lee, S.H. Lee, G.Y. Kim, Autogenous shrinkage of concrete containing granulated blast-furnace slag, Cem. Concr. Res. 36 (7) (2006) 1279e1285. [23] D.Y. Yoo, K.Y. Kwon, J.M. Yang, Y.S. Yoon, Effect of cover depth and rebar diameter on shrinkage behavior of ultra-high-performance fiber-reinforced concrete slabs, Struct. Eng. Mech. (2015) (under review). [24] H. Abdalla, Concrete cover requirements for FRP reinforced members in hot climates, Compos Struct. 73 (1) (2006) 61e69. [25] D.Y. Yoo, Y.S. Yoon, Structural performance of ultra-high-performance concrete beams with different steel fibers, Eng. Struct. 102 (2015) 409e423.
Contents lists available at ScienceDirect
Cement and Concrete Composites journal homepage: www.elsevier.com/locate/cemconcomp
Effectiveness of shrinkage-reducing admixture in reducing autogenous shrinkage stress of ultra-high-performance fiber-reinforced concrete Doo-Yeol Yoo a, Nemkumar Banthia a, Young-Soo Yoon b, * a b
Department of Civil Engineering, The University of British Columbia, 6250 Applied Science Lane, Vancouver, BC V6T 1Z4, Canada School of Civil, Environmental and Architectural Engineering, Korea University, 145 Anam-ro, Seongbuk-gu, Seoul 136-713, Republic of Korea
a r t i c l e i n f o
a b s t r a c t
Article history: Received 14 July 2015 Received in revised form 9 September 2015 Accepted 20 September 2015 Available online 25 September 2015
This study describes the effect of shrinkage-reducing admixture (SRA) on free and restrained autogenous shrinkage behaviors of ultra-high-performance fiber-reinforced concrete (UHPFRC). To investigate the cracking potential, tensile strength development was experimentally obtained and predicted on the basis of the degree of hydration model. Three different SRA to cement weight ratios of 0, 1, and 2% and three different reinforcement ratios of 1.3, 2.9, and 8.0% were considered. A higher SRA content contributed to a slightly higher tensile strength and a lower autogenous shrinkage. In addition, a higher SRA content and a lower reinforcement ratio resulted in better restrained autogenous shrinkage behaviors, such as lower autogenous shrinkage stress and cracking potential. Therefore, it can be concluded that the use of SRA or a lower reinforcement ratio is favorable for improving the restrained shrinkage behaviors of UHPFRC. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Ultra-high-performance fiber-reinforced concrete Autogenous shrinkage Time-zero Shrinkage-reducing admixture Reinforcement ratio Cracking potential
1. Introduction Recently, in order to overcome several drawbacks of conventional concrete such as low tensile strength, low ductility, and low strength to weight ratio, ultra-high-performance fiber-reinforced concrete (UHPFRC) has been developed with excellent strength, i.e., compressive strength > 150 MPa and tensile strength > 8 MPa, ductility, energy absorption capacity, and durability [1]. These outstanding properties are attributed to a low water-to-binder ratio (W/B) equal to 0.2, an optimization of particle sizes of compositions, and high volume contents of steel fibers. Thus, this material has been attractive to thin-plate structures, i.e., long-span bridge decks, roofs, and thin-walls, and field-cast joints for precast bridge decks [2,3]. However, owing to its high early-age autogenous shrinkage and small cross-sectional area for thin-plate structures, UHPFRC is highly vulnerable to premature shrinkage cracking [2]. Cracking and residual stress that occur in concrete structures by the restraint of shrinkage are the main concerns with respect to
* Corresponding author. Tel.: þ82 2 3290 3320; fax: þ82 2 928 7656. E-mail address: [email protected] (Y.-S. Yoon). http://dx.doi.org/10.1016/j.cemconcomp.2015.09.005 0958-9465/© 2015 Elsevier Ltd. All rights reserved.
durability. The restrained shrinkage behavior of concrete is quite complex, since it is affected by many factors, such as the rate and magnitude of free shrinkage, developments of strength and elastic modulus, stress relaxation and creep, degree of restraint, and geometry of element [4]. In particular, because of stress relaxation and creep characteristics of concrete, free shrinkage measurement alone is insufficient to predict the cracking potential due to the restraint of shrinkage. A conceptual view of the restrained shrinkage behavior of concrete with an internal reinforcing bar (rebar) is shown in Fig. 1. If no rebar exists, concrete will be freely deformed in a direction of Dεþ by shrinkage εsh (b). On the contrary, if the shrinkage of concrete is restrained by the internal rebar, and if no creep effect is assumed, the strain can be divided into two categories: elastic restraint strain by rebar εe and elastic rebar stain εe,r (c). However, in reality, owing to the tensile creep of concrete εcr, the strain obtained in rebar (or in concrete) by shrinkage decreases to εr (d). Therefore, the free shrinkage strain εsh is equal to the summation of εe, εcr, and εr. In order to mitigate the early-age shrinkage of UHPFRC, the use of shrinkage-reducing admixture (SRA) has been investigated by Soliman and Nehdi [5]. Their studies were focused on investigating the implication of adding SRA on the free shrinkage, and the
28
D.-Y. Yoo et al. / Cement and Concrete Composites 64 (2015) 27e36
a W/B of 0.2 was used. Smooth steel fibers with a diameter of 0.2 mm and a length of 13 mm were incorporated at a 2% volume fraction within a mortar matrix. The detailed mechanical and geometrical properties of the steel fibers are given in Table 2. A high-performance water-reducing agent, polycarboxylate superplasticizer (SP) with a density of 1.06 g/cm3, was also added to provide suitable fluidity. In addition, to investigate the implication of SRA contents on shrinkage behaviors of UHPFRC under both free and restrained conditions, three different SRA to cement weight ratios (0, 1, and 2%) were considered (Table 1) using the glycolbased SRA (METOLAT® P 860) produced in Germany. 2.2. Experimental setup and procedure
Fig. 1. Conceptual view of restrained shrinkage behavior of concrete with internal rebar.
effectiveness of SRA on reducing the shrinkage strains has been reported. However, although steel rebars are generally included in UHPFRC structures [6,7], no published study exists on the effects of using SRA on the shrinkage behavior of UHPFRC restrained by internal steel rebars. This restricts the practical use of SRA to the UHPFRC structures in spite of its numerous advantages. Accordingly, in this study, the effect of SRA on both the free shrinkage and restrained shrinkage (by internal steel rebar) behaviors of UHPFRC was investigated. Owing to its extremely high autogenous shrinkage but insignificant drying shrinkage [2], autogenous shrinkage tests were conducted according to the recommendation by Japan Concrete Institute (JCI) [8]. Correspondingly, three different SRA to cement weight ratios were used, and deformed steel rebars with three different diameters were applied in restrained autogenous shrinkage tests in order to provide various degrees of restraint. In addition, tensile strength development of UHPFRC was measured and predicted to evaluate the cracking potential. 2. Experimental program 2.1. Materials and mix proportions The used mix proportions are presented in Table 1. Type 1 Portland cement with a specific surface area of 3413 cm2/g and a density of 3.15 g/cm3 and silica fume with a specific surface area of 200,000 cm2/g and a density of 2.10 g/cm3 were used as cementitious materials. The chemical and physical properties of the cementitious materials can be found elsewhere [2]. Sand with grain size smaller than 0.5 mm and silica flour with a diameter of 2 mm including 98% SiO2 were added to the mixture. For all test series,
2.2.1. Flow and direct tensile tests In order to quantitatively measure flowability, a flow table test was carried out according to ASTM C 1437 [9]. The average flow was calculated by averaging the maximum flow diameter and the corresponding perpendicular diameter. The average flow values are listed in Table 1. The fluidity was slightly increased by adding SRA, even though identical W/B and amount of SP were used. To evaluate the cracking potential, the tensile strength development of UHPFRC with age was investigated. Dog-bone-shaped specimens with a middle cross-section of 50 mm 100 mm were fabricated and cured at a room with a temperature of (23 ± 1) C and a humidity of (60 ± 5)%. The details of the geometry and test setup are shown in Fig. 2. A pin-fixed end condition was used to avoid secondary flexural stress and to ensure a centric-loading condition. The alignment of the specimen was also carefully checked using a plumb before testing. Three specimens for each variable were used at each age, and the load was applied through displacement control using a universal testing machine (UTM) with maximum load capacity of 250 kN. The displacement was increased at a rate of 0.4 mm/min and the applied load was measured using a load cell attached to the bottom of the crosshead. 2.2.2. Autogenous shrinkage tests under free and restrained conditions Prismatic specimens with cross-sectional dimension of 100 mm 100 mm and length of 1000 mm were used for measuring free autogenous shrinkage. In order to precisely measure the autogenous shrinkage from a very early age, a dumbbellshaped strain gage that has a nearly zero stiffness and a coefficient of thermal expansion (CTE) of 11 mε/ C e similar to that of hardened UHPFRC e and a thermocouple were set horizontally in the middle of the mold before concrete casting, as shown in Fig. 3. Teflon sheet and polyester film were placed on the mold to eliminate the friction between the mold and the concrete. After concrete casting, the top surface of each specimen was covered with a polyester film to prevent the moisture evaporation. In order to estimate restrained autogenous shrinkage stress of UHPFRC, prismatic specimens that have a dimension identical to that used in free autogenous shrinkage tests were prepared. Deformed steel rebars with three different nominal diameters of
Table 1 Mix proportions. Name
UH-S0 UH-S1 UH-S2
Relative weight ratios to cement Cement
Water
Silica fume
Silica sand
Silica flour
SP
SRA
1.00
0.25
0.25
1.10
0.30
0.018
0.00 0.01 0.02
[Note] SP ¼ superplasticizer and SRA ¼ shrinkage-reducing admixture.
Steel fiber (Vf)
Flow (mm)
2%
235 245 240
D.-Y. Yoo et al. / Cement and Concrete Composites 64 (2015) 27e36
29
Table 2 Properties of smooth steel fibers. df (mm)
Lf (mm)
Aspect ratio (Lf/df)
Density (g/cm3)
ft (MPa)
Es (GPa)
0.2
13.0
65.0
7.9
2500
200
Image
[Note] df ¼ diameter of fiber, Lf ¼ length of fiber, ft ¼ tensile strength of fiber, and Es ¼ elastic modulus of fiber.
Fig. 2. Direct tensile test; (a) geometry of dog-bone specimen (unit: mm), (b) test setup.
12.7 mm, 19.1 mm, and 31.8 mm, which lead to three different reinforcement ratios of approximately 1.3%, 2.9%, and 8.0%, were used to restrain the volume change of UHPFRC by autogenous shrinkage. The detailed properties of the rebars used are summarized in Table 3. In order to provide identical autogenous shrinkage stress at the center of the specimen, the ribbed edges of the rebars were lathed to within 150 mm from the center and they were covered with a Teflon sheet to prevent friction between the
Fig. 3. Dumbbell-shaped embedded strain gage and thermocouple for free autogenous shrinkage measurement.
concrete and rebar according to the recommendation of the technical committee on autogenous shrinkage of concrete of the JCI [8], as shown in Fig. 4. In addition, based on previous studies [10,11], an embedment length of 350 mm was applied at both ends to provide full autogenous shrinkage stress instead of 600 mm recommend by the JCI [8] for reducing the specimen size. A strain gage and a thermocouple were attached to the center of the rebar before covering it with a Teflon sheet to measure the strain and temperature. A Teflon sheet and a polyester film were also placed inside the mold to minimize the frictional force between the mold and concrete, and the top surface was covered with a polyester film immediately after concrete casting to prevent the moisture evaporation. For both free and restrained autogenous shrinkage tests, all test specimens were demolded after 24 h from concrete casting and were immediately sealed with aluminum adhesive tape. In addition, the shrinkage measurement began just after concrete casting and was performed in a room with a temperature of (23 ± 1) C and a humidity of (60 ± 5)%.
30
D.-Y. Yoo et al. / Cement and Concrete Composites 64 (2015) 27e36
Table 3 Properties of deformed steel rebar. Reinforcement ratio
dr (mm)
Ar (mm2)
Er (GPa)
fy (MPa)
εy (mm/mm)
fu (MPa)
r ¼ 1.3% r ¼ 2.9% r ¼ 8.0%
12.7 19.1 31.8
126.7 286.5 794.2
200.0
522.7 508.5 509.0
0.00261 0.00254 0.00255
627.6 625.2 601.1
[Note] r ¼ reinforcement ratio, dr ¼ nominal diameter of rebar, Ar ¼ area of rebar, Er ¼ elastic modulus of rebar, fy ¼ yield strength of rebar, εy ¼ yield strain of rebar, and fu ¼ ultimate strength of rebar.
Fig. 4. Test setup for autogenous shrinkage stress measurement (unit: mm) (modified from JCI recommendation [8]).
3. Experimental results and discussion
by Janasson [16], as follows:
3.1. Tensile strength development
o n ft ðtÞ ¼ ft28 exp l1 ½lnð1 þ ðt t0 ÞÞk1
Fig. 5 shows the tensile strength development of UHPFRC at various SRA contents. Tensile strength slightly increased with an increase in the SRA content. For example, the 28-day tensile strengths were found to be 7.4 MPa for UH-S0, 7.8 MPa for UH-S1, and 8.0 MPa for UH-S2. In a similar way, the improvement of 28day flexural strength of high-strength steel-fiber-reinforced concrete (SFRC) with the addition of SRA was observed by Wang et al. [12]. However, these results differed from the previous tensile test results of the steam cured UHPFRC at 90 C [13]: lower tensile strength was obtained for the specimens including SRA for the case of steam curing. This discrepancy is due to the different curing conditions, and further investigation must be conducted to accurately prove this observation. Since the tensile strengths were only obtained at specific dates, an equation for predicting the tensile strength development of UHPFRC with age needs to be suggested to analyze the cracking potential. In accordance with a previous research [14], the strength development of UHPFRC exhibited almost linear relationship with the degree of hydration. Therefore, in order to predict the tensile strength development of UHPFRC, the following equation (Eq. (1)) was adopted [15] based on the degree of hydration model proposed
(1)
where ft28 is the tensile strength after 28 days, t0 is the time when the shrinkage stress first develops, and l1 and k1 are the regression coefficients, summarized in Table 4. This Eq. (1) includes the 28-day tensile strength, which is similar to the Graybeal's formula [17], and as shown in Fig. 5, the predicted values are in good agreement with the experimental results. 3.2. Autogenous shrinkage behavior To obtain pure autogenous shrinkage, the thermal dilation needs to be compensated from the measured strain using Eq. (2), as expressed by
εas ¼ εm a DT
(2)
where εas is the autogenous shrinkage strain, εm is the measured strain, a is the CTE of concrete, and DT is the temperature variation. In general, the CTE of concrete varies with age at an early age [18]. However, not only is measuring the early age CTE of concrete difficult, but also the effect of the CTE variation on autogenous shrinkage is insignificant because of its small cross-sectional area
D.-Y. Yoo et al. / Cement and Concrete Composites 64 (2015) 27e36
31
Fig. 5. Tensile strength development of UHPFRC with various SRA contents; (a) UH-S0, (b) UH-S1, (c) UH-S2.
Table 4 Non-linear regression coefficients for tensile strength development in Eq. (1). Name
l1
k1
R2
UH-S0 UH-S1 UH-S2
0.240 0.272 0.332
1.067 1.241 1.096
0.93046 0.93942 0.91867
(100 mm 100 mm) of the prismatic specimen used in the current study. For these reasons, a constant CTE of 11 mε/ C, as recommended by AFGC/SETRA recommendations [19], was adopted. According to the technical committee on autogenous shrinkage of concrete of the JCI [8], the purpose of measuring autogenous shrinkage is to predict cracking response of concrete. Thus, they suggested the use of the initial point of autogenous shrinkage measurement, called “time-zero”, as the time of the initial setting, in order to exclude the volume change that occurs when concrete is still fresh. However, several researchers have reported [20,21] that the use of setting time as a time-zero for the autogenous shrinkage measurement is erroneous, because the start point of selfdesiccation, which is the main mechanism of autogenous shrinkage, is influenced by bleeding water and the setting time cannot account for this effect. For this reason, the authors used the start time of the shrinkage stress development in concrete as the time-zero, according to a previous study [15].
Fig. 6 shows the comparison of a 30-day autogenous shrinkage strains measured at different zeroing points. The highest strain was obtained as the start time of the shrinkage stress development was determined as time-zero. It is interesting to note that autogenous shrinkage strains measured from the start time of shrinkage stress development exhibited very similar values to those measured from the deviation point of shrinkage strain and internal temperature, as reported by Yoo et al. [15]. Therefore, the deviation point of measured strain and temperature at a very early age can be used as the time-zero for UHPFRC. On the other hand, significantly lower autogenous shrinkage strains were obtained when the initial and final setting times were used as time-zero. This was because UHPFRC showed a very steep increase in autogenous shrinkage at early age. The 30-day autogenous shrinkage strains measured from the initial and final sets were respectively 25% and 55% lower, on average, than those measured from the start time of the shrinkage stress development. From these observations, the use of the initial and final setting times as the time-zero point appears to be inappropriate for UHPFRC. Fig. 7(a) shows the autogenous shrinkage of UHPFRC measured from the start time of shrinkage stress development, which is equal to the time when the rebars start to deform as a result of concrete shrinkage. Regardless of the SRA contents, autogenous shrinkage rapidly increased just after time-zero. After certain points (i.e., 15 h for UH-S0, 17 h for UH-S1, and 18 h for UH-S2), the increase rate of
32
D.-Y. Yoo et al. / Cement and Concrete Composites 64 (2015) 27e36
Fig. 6. Comparison of 30-day autogenous shrinkage strains measured from start time of shrinkage stress development ðt0 * Þ, deviation time of strain and temperature ðt0 y Þ, initial setting time (ti), and final setting time (tf).
Fig. 8. Comparison of 30-day autogenous shrinkage of UHPFRC with different SRA contents.
Fig. 7. Autogenous shrinkage and temperature responses of UHPFRC with different SRA contents; (a) autogenous shrinkage, (b) temperature.
shrinkage strain suddenly decreased and slight expansion was observed. The main reason for sudden decrease of shrinkage increase rate was that the chemical shrinkage and the volume contraction from the negative pressure in the internal voids were self-restrained by the hardening of concrete. The point where the increase rate of shrinkage strain abruptly decreased was delayed with the inclusion of SRA. It is interesting to note that a significant reduction of autogenous shrinkage with the addition of SRA was
observed at a very early age (before 18 h). In addition, as shown in Fig. 8, the additions of 1% and 2% SRA resulted in 15.2% and 28.4% reductions of autogenous shrinkage after 30 days, respectively. The 30-day autogenous shrinkage of UH-S0 was found to be approximately 760 mε, which is much larger than values reported for normal- and high-strength concretes [22] and often sufficient to cause cracking. With the addition of SRA, the maximum temperature decreased and the time when the maximum temperature reached was delayed, as shown in Fig. 7(b). The maximum temperature of UH-S0 was found to be 34.6 C after nearly 0.67 days, which is 4% and 13% higher than the maximum temperature values of UH-S1 and UH-S2, respectively. 3.3. Restrained shrinkage behavior 3.3.1. Autogenous shrinkage stress The responses of strain and temperature in the rebars are shown in Fig. 9. Based on the restrained UHPFRC slab test results performed by Yoo et al. [23], the steel rebar strain and the strain in the UHPFRC near the steel rebar showed very similar values to each other owing to the restraint of shrinkage. Therefore, the slip deformation between the rebar and UHPFRC was ignored in this
D.-Y. Yoo et al. / Cement and Concrete Composites 64 (2015) 27e36
33
Fig. 9. Restrained shrinkage behaviors of UHPFRC with different SRA contents and reinforcement ratios; (a) UH-S0 (r ¼ 1.3%), (b) UH-S0 (r ¼ 2.9%), (c) UH-S0 (r ¼ 8.0%), (d) UH-S1 (r ¼ 1.3%), (e) UH-S1 (r ¼ 2.9%), (f) UH-S1 (r ¼ 8.0%), (g) UH-S2 (r ¼ 1.3%), (h) UH-S2 (r ¼ 2.9%), (i) UH-S2 (r ¼ 8.0%).
study. To obtain pure strains in the rebar from concrete shrinkage, the thermal dilation needed to be compensated by adopting an appropriate CTE. The CTE of the deformed steel rebar was used with a value of 11.7 mε/ C [24]. As shown in Fig. 9, a higher strain was obtained with a smaller diameter of rebar regardless of the SRA content. The strains without thermal dilation yielded higher values at early age than the measured strains because of the hydration heat. However, since the thermal strain returned from expansion to shrinkage during the cooling stage, the strains without thermal dilation became very similar to the measured strains. The increase of strains in the rebars gradually decreased with age and finally converged to a stable value after 30 days. This was attributed to the stabilization of both the autogenous shrinkage and the elastic modulus development. The JCI committee on autogenous shrinkage [8] suggested an equation to calculate the autogenous shrinkage stress sc in concrete as follows:
sc ¼
Er εr Ar Ac
(3)
where Er is the elastic modulus of rebar, εr is the strain obtained in rebar excluding the thermal dilation, and Ar and Ac are the areas of the rebar and concrete, respectively. Fig. 10 shows the comparison of tensile strengths and autogenous shrinkage stresses at various SRA contents and reinforcement ratios. The tensile strengths calculated by Eq. (1) were much higher than the autogenous shrinkage stresses for all test series. Therefore, no shrinkage crack was observed during testing. Regardless of the SRA content, a higher autogenous shrinkage stress was obtained with a higher reinforcement ratio. The autogenous shrinkage stress increased more gradually at an early age compared with the free autogenous shrinkage strain. The free autogenous shrinkage strains at a very early age (at nearly 18 h) were found to be approximately 78% (on average) of the 30-day autogenous shrinkage strains. On the other hand, the restrained autogenous shrinkage stresses at nearly 18 h were found to be approximately 60% (on average) of the
34
D.-Y. Yoo et al. / Cement and Concrete Composites 64 (2015) 27e36
Fig. 10. Comparison of autogenous shrinkage stress and tensile strength; (a) UH-S0, (b) UH-S1, (c) UH-S2.
3.3.2. Degree of restraint If the steel rebar has high enough stiffness to perfectly restrain the shrinkage of concrete, it will not be deformed. However, as shown in Fig. 9, the steel rebar was deformed by the shrinkage of concrete, and thus, the restraint strain εrs considering tensile creep effect was calculated by subtracting the strain in the rebar from the autogenous shrinkage strain, as expressed by
εrs ¼ εas εr
(4)
Therefore, the degree of restraint j is determined by the ratio of the restraint strain and the autogenous shrinkage strain, as given by Eq. (5):
j¼ Fig. 11. Comparison of 30-day autogenous shrinkage stresses according to reinforcement ratio and SRA content.
30-day restrained autogenous shrinkage stresses. This was mainly attributed to the low stiffness of UHPFRC at early age. As shown in Fig. 11, the 30-day autogenous shrinkage stress decreased with an increase in the SRA content. For example, the 30day autogenous shrinkage stress of UH-S0 with r ¼ 1.3% was found to be 1.3 MPa, approximately 11% and 22% higher than those of UHS1 and UH-S2 at an identical reinforcement ratio, respectively.
εrs εr ¼1 εas εas
(5)
Fig. 12 shows the degree of restraint at different SRA contents and reinforcement ratios. The degree of restraint decreased as time passed, owing to the increases in the shrinkage strain and elastic modulus of concrete. A higher degree of restraint was obtained at a higher reinforcement regardless of the SRA content. This was because a lower strain in rebar was obtained for the specimen with a higher reinforcement ratio. On the other hand, the degree of restraint was insignificantly influenced by the SRA content, as shown in Fig. 13. This was caused by the fact that both the strain in the rebar and autogenous shrinkage decreased with the addition of SRA. Thus, it was concluded that the degree of restraint was
D.-Y. Yoo et al. / Cement and Concrete Composites 64 (2015) 27e36
35
Fig. 12. Degree of restraint; (a) UH-S0, (b) UH-S1, (c) UH-S2.
marginally affected by the SRA content, but was significantly affected by the reinforcement ratio. For example, the degree of restraint for UH-S0 with r ¼ 8.0% was 0.68 after 30 days, approximately 40% and 94% higher than those of UH-S0 with r ¼ 2.9% and r ¼ 1.3%, respectively. 3.3.3. Cracking potential If there is no shrinkage cracking, the investigation of cracking potential is imperative to compare the restrained shrinkage
Fig. 13. Comparison of degree of restraints according to reinforcement ratio and SRA content.
behavior. The cracking potential is defined by dividing the autogenous shrinkage stress by the tensile strength, as expressed by
Qcr ¼
sc ft
(6)
where Qcr is the cracking potential. For all test series, the cracking potentials were much lower than 100%, as shown in Fig. 14.
Fig. 14. Comparison of cracking potentials according to reinforcement ratio and SRA content.
36
D.-Y. Yoo et al. / Cement and Concrete Composites 64 (2015) 27e36
Therefore, no shrinkage crack was observed during testing. The cracking potential increased with an increase in the reinforcement ratio, whereas it decreased with the inclusion of SRA. The highest cracking potential was found to be 63% for UH-S0 with r ¼ 8.0%, whereas the lowest cracking potential was found to be 14% for UHS2 with r ¼ 1.3%. This indicates that even though shrinkage crack is not observed, the structures made of UHPFRC at a high reinforcement ratio are likely to be cracked under a low tensile stress state, owing to their high residual stress as a result of the restraint of shrinkage. Yoo and Yoon [25] recently reported that the first cracking load of steel bar-reinforced UHPFRC beams decreased with an increase in the reinforcement ratio due to the higher residual stress. 4. Conclusions This study investigated the effects of SRA content and reinforcement ratio on autogenous shrinkage behaviors of UHPFRC under free and restrained conditions. Based on the test results, the following conclusions can be drawn: 1) The 28-day tensile strength of non-steam-cured UHPFRC slightly increased with an increase in the SRA content up to 2%. 2) The time-zero of autogenous shrinkage measurement was suggested as the start time of shrinkage stress development. The deviation point between measured strain and temperature at early age was also proper as the time-zero for UHPFRC, whereas the initial and final setting times were inappropriate. 3) The autogenous shrinkage strain decreased with an increase in the SRA content. A significant reduction of autogenous shrinkage by SRA was observed at a very early age (before 18 h). After 30 days, the inclusion of 1% and 2% SRA led to approximately 15.2% and 28.4% reductions of autogenous shrinkage strains, respectively. 4) For all test series, the tensile strengths were higher than the autogenous shrinkage stresses. Thus, no shrinkage crack was observed during testing. In addition, autogenous shrinkage stress decreased at lower reinforcement ratios and higher SRA contents. 5) The degree of restraint decreased with age due to the increase of shrinkage and elastic modulus in concrete. A higher degree of restraint was obtained with a higher reinforcement ratio, but it was marginally influenced by the SRA content. 6) The cracking potentials for all test series were much lower than 100%, and thus, no shrinkage crack was detected. In addition, lower cracking potentials were obtained with lower reinforcement ratios and higher SRA contents. Acknowledgments This research was supported by a grant (13SCIPA01) from Smart Civil Infrastructure Research Program funded by Ministry of Land, Infrastructure and Transport (MOLIT) of Korea government and Korea Agency for Infrastructure Technology Advancement (KAIA).
References [1] P. Richard, M. Cheyrezy, Composition of reactive powder concretes, Cem. Concr. Res. 25 (7) (1995) 1501e1511. [2] D.Y. Yoo, H.O. Shin, J.Y. Lee, Y.S. Yoon, Enhancing cracking resistance of ultrahigh-performance concrete slabs using steel fibers, Mag. Concr. Res. 67 (10) (2015) 487e495. [3] V. Perry, W. Gary, Innovative field cast UHPC joints for precast bridge decksedesign, prototype testing and projects, in: International Workshop on UHPFRC, Marseille, France, 2009. [4] W.J. Weiss, W. Yang, S.P. Shah, Shrinkage cracking of restrained concrete slabs, J. Eng. Mech. 124 (7) (1998) 765e774. [5] A.M. Soliman, M.L. Nehdi, Effect of drying conditions on autogenous shrinkage in ultra-high performance concrete at early-age, Mater. Struct. 44 (5) (2011) 879e899. [6] M.A. Saleem, A. Mirmiran, J. Xia, K. Mackie, Ultra-high-performance concrete bridge deck reinforced with high-strength steel, ACI Struct. J. 108 (5) (2011) 601e609. [7] D.Y. Yoo, N. Banthia, S.W. Kim, Y.S. Yoon, Response of ultra-high-performance fiber-reinforced concrete beams with continuous steel reinforcement subjected to low-velocity impact loading, Compos Struct. 126 (2015) 233e245. [8] Japan Concrete Institute, in: E. Tazawa (Ed.), Committee report, Autogenous Shrinkage of Concrete, E&FN Spon, 1999. [9] ASTM, C 1437-07 Standard Test Method for Flow of Hydraulic Cement Mortar, American Society of Testing and Materials, West Conshohocken, PA, 2007, p. 2. [10] T. Nakagawa, Y. Ohno, Investigation of test method of self stress due to autogenous shrinkage in concrete, Proc. Jpn. Concr. Inst. 20 (2) (1998) 751e756 (in Japanese). [11] D.Y. Yoo, Performance Enhancement of Ultra-high-performance Fiber-reinforced Concrete and Model Development for Practical Utilization (PhD thesis), Korea University, Seoul, Korea, 2014, p. 586. [12] J.Y. Wang, N. Banthia, M.H. Zhang, Effect of shrinkage reducing admixture on flexural behaviors of fiber reinforced cementitious composites, Cem. Concr. Compos 34 (4) (2012) 443e450. [13] D.Y. Yoo, S.T. Kang, J.H. Lee, Y.S. Yoon, Effect of shrinkage-reducing admixture on tensile and flexural behaviors of UHPFRC considering fiber distribution characteristics, Cem. Concr. Res. 54 (2013) 180e190. , E. Brühwiler, Thermal effects on physico-mechanical [14] A. Kamen, E. Denarie properties of ultra-high-performance fiber-reinforced concrete, ACI Mater. J. 104 (4) (2007) 415e423. [15] D.Y. Yoo, J.J. Park, S.W. Kim, Y.S. Yoon, Influence of reinforcing bar type on autogenous shrinkage stress and bond behavior of ultra high performance fiber reinforced concrete, Cem. Concr. Compos 48 (2014) 150e161. [16] J.E. Jonasson, Slipform Construction-calculations for Assessing Protection against Early Freezing, Fo 4: 84, Swedish Cement and Concrete Institute, Stockholm, 1985, pp. 1e13. [17] B.A. Graybeal, Compressive behavior of ultra-high-performance fiber-reinforced concrete, ACI Mater. J. 104 (2) (2007) 146e152. [18] Ø. Bjøntegaard, E.J. Sellevold, Interaction between thermal dilation and autogenous deformation in high performance concrete, Mater. Struct. 34 (5) (2001) 266e272. nie Civil, Ultra high performance fibre-reinforced [19] Association Française de Ge concretes, in: Interim Recommendations, AFGC/SETRA, Bagneux, France, 2002, p. 152. [20] T.A. Hammer, H. Justnes, Ø. Bjøntegaard, E.J. Sellevold, 8.2 Suggestions on the terminology and the test methods proposed by JCI, in: E. Tazawa (Ed.), Autogenous Shrinkage of Concrete. Proceedings of the International Workshop, E&FN Spon, 1998, pp. 397e399. [21] E. Holt, Early Age Autogenous Shrinkage of Concrete, VTT Publication 446, Technical Research Centre of Finland, 2001, pp. 1e184. [22] K.M. Lee, H.K. Lee, S.H. Lee, G.Y. Kim, Autogenous shrinkage of concrete containing granulated blast-furnace slag, Cem. Concr. Res. 36 (7) (2006) 1279e1285. [23] D.Y. Yoo, K.Y. Kwon, J.M. Yang, Y.S. Yoon, Effect of cover depth and rebar diameter on shrinkage behavior of ultra-high-performance fiber-reinforced concrete slabs, Struct. Eng. Mech. (2015) (under review). [24] H. Abdalla, Concrete cover requirements for FRP reinforced members in hot climates, Compos Struct. 73 (1) (2006) 61e69. [25] D.Y. Yoo, Y.S. Yoon, Structural performance of ultra-high-performance concrete beams with different steel fibers, Eng. Struct. 102 (2015) 409e423.