Internal relative humidity and creep of concrete with modified admixtures

Internal relative humidity and creep of concrete with modified admixtures Zhi-hai HE1, 2, Chun-xiang QIAN1, 2 1. College of Materials Science and Engineering, Southeast University, Nanjing 211189, China; 2. Jiangsu Key Lab of Construction Materials, Nanjing 211189, China Received 21 June 2011; Accepted 14 September 2011 Abstract: The creep characteristics and internal relative humidity (IRH) of concrete with shrinkage reducing admixtures (SRA), poly vinyl alcohol fiber (PVAF) and a mixture of fly ash and slag (MFS) were investigated by the self-made loading device. The pore structure was tested, and the relationship of creep and IRH of concrete was discussed. The results indicate that MFS and SRA reduce the total creep and delay the change of IRH. However, the effect of PVAF is contrary, relative to MFS and SRA. The total creep depends largely on drying creep. The basic creep of concrete is similar, which is significantly influenced by volume fraction of pores whose diameter is between 3.5 nm and 50 nm to the total pore (VP), and IRH of concrete is connected with porosity. It is easy to obtain a good linear relationship predicting the creep of concrete according to the change of IRH. Key words: modified admixture; internal relative humidity; total creep; drying creep; basic creep; porosity

1 Introduction Concrete creep affects various performance parameters of concrete structures directly or indirectly [1, 2]. Creep is the most important part of concrete volume stability, also one of the important reasons for the significant increasing of prestress loss, long-term deformation and internal stress redistribution of large span prestressed concrete bridges [3]. So the less terminal value of concrete creep is required to ensure the engineering safety, which can stabilize as soon as possible. Water has a powerful and pervasive influence on the performance of concrete [4]. Water, called “evaporable water”, is in fact present in concrete in-service, where it contributes to the properties of hardened concrete such as strength and elastic modulus, and plays a significant role in creep [5, 6]. In general, moisture change in concrete is represented by internal relative humidity (IRH), whose reduction will increase the creep deformation. In recent years, there has been an increasing interest in the use of fiber and shrinkage reducing admixtures (SRA) to combat shrinkage cracking in concrete elements [7−11], while the effect of them on concrete

creep has not been adequately studied till now. As for IRH, many researchers tended to concentrate on shrinkage, and the relationship between IRH and shrinkage was obtained analytically [12−14]. However, concrete creep was seldom analyzed by IRH quantitatively. In the case of concrete creep, the research mainly focused on the mechanism and model correction and prediction [3, 15−18]. However, the above research on creep was under the single factor control condition leading to concrete strength change at fixed ages, while the strength should be constant according to the structure design for practical engineering. Therefore, the above research achievements can not guide directly in practical engineering construction. Considering the requirements of concrete commonly used for main girder of prestressed concrete bridges in high-speed railway that the concrete strength grade reaches C50, elastic modulus 35.5 GPa at 7 days in order to prestress, the objective of this work is to evaluate IRH and creep of concrete with poly vinyl alcohol fiber (PVAF), SRA and a mixture of fly ash and slag (MFS) and to investigate the relationship between IRH and creep, so as to optimize concrete design and to provide the reasonable technical guidance for similar engineering.

Foundation item: Project (2009CB623200) supported by the National Basic Research Program of China; Project (2010Y01) supported by the Traffic Scientific Research Program of Jiangsu Province, China Corresponding author: Chun-xiang QIAN; E-mail: [email protected]

Zhi-hai HE, et al/Progress in Natural Science: Materials International 21(2011) 426−432

427

concrete was also prepared for comparative analysis. The mix proportions of concrete are shown in Table 1.

2 Experimental 2.1 Raw materials The cement used in this study was Grade-52.5 ordinary Portland cement. Fly ash was Grade-I from power plant with apparent density of 2.26 g/cm3. Slag was Grade-S95 blast furnace slag superfine powder with specific surface area of 400 m2/kg. Fiber was PVAF with length of 6 mm, tensile elastic modulus of 40 GPa, and fracture elongation of 13%. SRA was JM-SRA with viscosity of 50 Pa·s and surface tension of 41.7 mN/m. Polycarboxylic superplasticizer admixture with water reducing ratio of 25.3% was used. River sand with fineness modulus of 2.50 and apparent density of 2.61 g/cm3 was used as fine aggregate, and crushed limestone with size of 5−25 mm and apparent density of 2.71g/cm3 was used as coarse aggregate. 2.2 Mix proportions The requirement of 7-day same compressive strengths of concrete with modified admixtures was reached by adjusting water-binder ratio (w/b), remaining cementitious material amount and sand coarse aggregate ratio constant. The slumps were around 180 mm by adjusting superplasticizer admixture, and the controlled

2.3 Test methods The self-made device was used to test the creep and IRH of concrete, as shown in Fig. 1. The sample whose dimension was 130 mm×130 mm×400 mm with center hole of 30 mm was prepared by wooden mold, and a string was composed of 4 samples which were separated by steel clapboards, so called “gourd string samples”. The gourd string samples containing a poly vinyl chloride (PVC) pipe with an outer diameter of 30 mm and an inner diameter of 25 mm to pass prestressed steel bar with the diameter of 25 mm, were used for the loaded samples. The data collector containing digital humidity sensor and vibration string type strain gauge for testing IRH and strain of concrete respectively, was embedded in the sample through the PVC pipe during preparing concrete. The samples were demoulded after 7 days of moist curing and prestressed by the steel bar combined with a jack. The stress was controlled to 40% of prism body compressive strengths by the vibration string type pressure sensor, and the environment was controlled at temperature of (20±1) °C, relative humidity of (60±5)% during the test. In order to obtain creep strain that was calculated by

Table 1 Mix proportions and compressive strengths of concrete cured for 7 days Mix proportions of concrete/(kg·m−3) Cement

Fly ash

Slag

Sand

Gravel

Water

PVAF

SRA

Superplasticizer

Compressive strength/MPa

A

480

−

−

708

1 062

150

−

−

3.6

53.2

B

384

96

−

708

1 062

135

1.3

−

5.52

52.9

C

336

96

48

708

1 062

137

−

−

4.32

53.4

D

384

96

−

708

1 062

135

−

9.6

3.84

52.8

Sample

Fig. 1 Schematic representation of IRH and creep of concrete test (Unit: mm): (a) Test diagram; (b) Sample dimension; (c) Clapboard A (130×130×20); (d) Clapboard B (130×130×10); (e) Clapboard C (100×100×20)

Zhi-hai HE, et al/Progress in Natural Science: Materials International 21(2011) 426−432

subtracting shrinkage strain from the total longitudinal strain of loaded samples, shrinkage strain of non-loaded samples with the same dimension having PVC pipe without load was tested at the same time. The samples were sealed by epoxy resin and coated with plastic film to prevent moisture exchange with the external during basic creep test. Drying creep is calculated by subtracting basic creep from total creep. Mercury intrusion porosimetry (MIP) was used to test the pore structure of hardened samples. The mercury porosimeter is capable of generating 0−420 MPa pressure and measuring 3.5 nm−400 μm pore diameter. The samples for MIP test were broken into 3−5 mm pieces and stored in ethanol solution for the hydration termination. Before MIP test, the samples were dried and stored in sealed containers.

3 Results 3.1 Total creep The total creep of samples at different ages after loading is presented in Fig. 2.

428

3.2 Basic creep Figure 3 shows the basic creep of samples with modified admixtures in sealed state.

Fig. 3 Effect of modified admixtures on basic creep

The development trend of the basic creep of samples is similar to that of the total creep. However, the basic creep of different samples is very similar, with the magnitude ranging from 97.1×10−6 to 119.9×10−6 at 450 d after loading, and the sequence of basic creep of samples is A>B>D>C, which is different from that of the total creep. Combining Fig. 2 with Fig. 3, the proportion of basic creep in total creep is obtained. The proportions of the controlled sample A, the sample B with PVAF and the sample C with MFS are very close, 28.1%, 24.1% and 28.1% respectively, while that of the sample D with SRA is slightly greater, about 33.4%, after becoming stable. 3.3 Drying creep The contribution of drying creep to total creep can be seen in Fig. 4, which is given by Fig. 2 minus Fig. 3.

Fig. 2 Effect of modified admixtures on total creep

The total creep of the concrete samples with modified admixtures develops rapidly. The value at 100 d after loading reaches about 90% of that at 450 d, then becomes stable. It is evident that SRA and MFS can significantly reduce concrete creep, while PVAF increases it. The total creep of the sample B with PVAF slightly increases by 9.9%, while that of the sample C with MFS and sample D with SRA fall by 19.3% and 26.5% respectively, comparing with the creep of 427.5×10−6 of the controlled sample A at 450 d after loading. The sequence of total creep of samples is B>A>C>D from loading up to 450 d.

Fig. 4 Effect of modified admixtures on drying creep

Zhi-hai HE, et al/Progress in Natural Science: Materials International 21(2011) 426−432

The development trend and the sequence of drying creep of the samples are similar to those of the total creep. It is clear that the drying creep of samples is the leading position in total creep, because drying creep represents greater part of the total creep and the observation does not seem to be influenced by modified admixtures. 3.4 Internal relative humidity Figure 5 shows IRH of samples from loading after 7 d of moist curing.

Fig. 5 Effect of modified admixtures on IRH

IRH of the samples with modified admixtures decreases continuously with the increase of loading age, and becomes stable after loading of 100 d. From Fig. 5, MFS and SRA delay the variation of IRH of samples, showing the strong moisturizing ability, while PVAF accelerates the decrease of IRH. The value sequence of IRH of samples is D>C>A>B (88.5%, 86.6%, 81.3% and 81.1% respectively), at 450 d after loading.

4 Discussion 4.1 IRH, creep and pore structure It is well known that the w/b and modified

429

admixtures will change the pore structure, meanwhile, the change of micro-pore structure will lead to the variation of IRH and macro volume of cement-based materials, and creep is one of the macro volume changes. JIANG et al [12, 19] found that the w/b was a chief factor that affected autogenous relative humidity change of cement pastes with silica fume and ground blast-furnace slag. And the lower the w/b of paste was, the more reduction the autogenous relative humidity was. Therefore, to study the pore structure of the hardened samples with modified admixtures, coarse aggregate was removed from samples to carry out MIP test, and the results are shown in Table 2. It is evident that the porosity of the samples decreases with the increase of loading ages. Cementbased materials hydrate to increase the amount of the hydration products which can offer the hardened samples a stronger structure and higher resistance to creep. With the increase of loading ages, the reduction range of porosity gets smaller and the creep of the hardened samples becomes slowly stable. The modified admixtures such as PVAF, MFS and SRA have a slight retarding effect on the hydration of cement-based materials at early age [20−22]. Thus, the w/b of samples with the modified admixtures is reduced to remain 7-day same compressive strengths, compared with the w/b of the controlled sample A. Though the w/b of the sample B with PVAF is smaller, the porosity is greater than that of the controlled sample A. The result is generally explained by the relatively rough surface structure of PVAF [21, 23], which weakens the interface structure to lead to the fast reduction of IRH. The greatest porosity of the sample B with PVAF is in conformity with the lowest IRH. MFS and SRA reduce the porosity of samples, the better effect with MFS. The reasons can be attributed to the lower w/b of samples and their different hydration activities [20, 22]. However, the maximum IRH of the sample D with SRA is not in accordance with the smallest porosity. Porosity plays an important role in controlling the change of IRH, which also is influenced by the physical and chemical character of the modified admixtures [12, 19]. Generally, SRA

Table 2 Pore parameters of samples with modified admixtures at different ages after loading Sample

Porosity/%

VP/%

7d

28 d

100 d

360 d

450 d

7d

28 d

100 d

360 d

450 d

A

15.2

14.8

13.5

12.5

11.9

64.1

70.2

74.2

75.7

77.2

B

15.6

15.2

14.2

12.9

12.4

66.3

72.4

76.4

79.4

81.7

C

14.5

13.9

12.1

11.3

10.8

68.3

76.5

87.2

89.2

92.2

D

14.8

14.3

13.1

11.8

11.4

67.1

73.7

84.5

86.9

88.5

VP: Volume fraction of pores whose diameter is between 3.5 nm and 50 nm to the total pore.

Zhi-hai HE, et al/Progress in Natural Science: Materials International 21(2011) 426−432

significantly reduces the magnitude of capillary stresses and shrinkage strains by reducing the surface tension of concrete pore fluid [10]. At the same time, SRA also increases the viscosity of concrete pore fluid to show a delayed setting [20], thus testifying a slower IRH change. The maximum IRH of the sample D is associated with the lower porosity and the effect of SRA on pore fluid. A common view is that both creep and shrinkage should be considered together and as a manifestation of the same phenomenon [6]. In addition, the porosity especially pore structure parameter plays a significant role in the overall performance of concrete with respect to water transport [24]. LI et al [25] revealed that the relationship between autogenous shrinkage of cement paste with mineral admixtures and the volume fraction of pores whose diameter is between 5 nm and 50 nm was obviously proportional. Therefore, the volume fraction of pores whose diameter is between 3.5 nm and 50 nm to the total pore (VP) of samples is obtained, and the results are shown in Table 2. It is clear that all modified admixtures increase the VP of samples because of the lower w/b. On the other hand, fly ash of the samples with modified admixtures can fill in the gaps at first due to micro aggregate effect, and then generate hydration products which can fill in the pores and split the original large pore into many smaller pores because of activity effect [25]. The comprehensive effect leads to the increase of VP, compared with that of the controlled sample A. The maximum VP of the sample C with MFS results from that the complementary synergies of fly ash and slag improve remarkably the compactness of concrete to decrease porosity [25, 26]. VP of the sample D with SRA agrees with the result [27] that SRA reduces the harmful pore volume but increases the harmless pore volume. The VP value sequence is C>D>B>A all along. Creep can be divided into two categories: basic creep and drying creep. Generally speaking, the basic creep is caused due to the lagging elastic deformation and viscous deformation by the compaction effect [28], which should be connected with porosity. In contrast with the basic creep and VP of samples, the value sequence is in accordance respectively. The greater VP accords with the lower basic creep. 4.2 Relation between IRH and creep The understanding of creep and shrinkage in concrete is still limited because of the complexity. Therefore, there still exists a gap between the simplified prediction and the real fact [3]. However, it is generally believed that the deformation of concrete is related to moisture change in concrete [4−6]. JIANG et al [12] and HU et al [29] observed that there exists a good linear correlation between IRH and autogenous deformation of

430

cement-based materials. Thus, based on the test results and theory of IRH and creep, there should also be a certain relationship. The linear relationship between IRH and creep of samples can be obtained by analyzing the test results, which is presented in Fig. 6.

Fig. 6 Linear relationship between creep and IRH of samples: (a) Total creep and IRH; (b) Basic creep and IRH; (c) Drying creep and IRH

The linear regression equation of IRH and creep of samples is ε=a+b×H. Where ε is creep, including total creep, basic creep and drying creep, respectively; H is

Zhi-hai HE, et al/Progress in Natural Science: Materials International 21(2011) 426−432

internal relative humidity. ε and H should be the measured values at the same time respectively. a and b is coefficient; the correlation coefficient R2 is rather acceptable, which are about 0.9. a is the intercept of the equation, which indicates how much is ε when IRH is 0. However, actually IRH is not likely to reduce 0, so ε is considered together with IRH at different ages after loading, whose change is different from the research result [30] that the variation law of IRH with age can be described by a water-vapor saturated stage with 100% IRH (Stage I) followed by a stage that IRH gradually decreases (Stage II). The reason can be attributed to the different w/b and loading condition, especially for the different started measurement time. IRH in Fig. 5 was measured after 7 d moist curing. Therefore, IRH has accomplished Stage I and experiences Stage II directly. b is the slope of the equation, whose absolute value indicates the change speed of ε with the decrease of IRH. Drying creep represents greater part of total creep and basic creep of samples is very close. Thus, total creep depends largely on drying creep. In the case of the linear relationship of total creep and IRH and drying creep and IRH, showed in Fig. 6(a) and Fig. 6(c) respectively, a of the sample B with PVAF is maximum, and the absolute value of b is also maximum, showing ε of the sample B with PVAF is the most sensitive to IRH. a and b of other samples are close, respectively. The analysis results and the actual test phenomena match each other preferably. Therefore, it appears possible to predict creep according to IRH of concrete.

[2]

[3]

[4]

[5] [6]

[7]

[8]

[9]

[10]

[11]

[12]

5 Conclusions [13]

1) PVAF, MFS and SRA have different influence on creep and IRH of concrete with 7-day same compressive strengths. MFS and SRA reduce the total creep and delay the change of IRH, of which SRA has the better effect. However, the effect of PVAF on the total creep and IRH is contrary, relative to MFS and SRA. 2) The proportion of drying creep in total creep is further greater than that of basic creep. Thus, the total creep depends largely on the drying creep. The basic creep of concrete is close, which is significantly influenced by VP, and IRH of concrete is connected with the porosity. 3) There exists a good linear relationship between the creep and IRH of concrete, and the correlation coefficient R2 is rather acceptable. It is possible to predict creep of concrete according to the change of IRH.

[14]

[15]

[16]

[17]

[18]

[19]

References [1]

WANG Y F, MA Y S, ZHOU L. Creep of FRP-wrapped concrete

[20]

431

columns with or without fly ash under axial load [J]. Construction and Building Materials, 2011, 25: 697−704. ASAMOTO S, OHTSUKA A, KUWAHARA Y, et al. Study on effects of solar radiation and rain on shrinkage, shrinkage cracking and creep of concrete [J]. Cement and Concrete Research, 2011, 41: 590−601. FRANCIS T K A, SI X T. Accurate time-dependent analysis of concrete bridges considering concrete creep, concrete shrinkage and cable relaxation [J]. Engineering Structures, 2011, 33: 118−126. KENNETH C H. The influence of water on the performance of concrete [J]. Construction and Building Materials, 2011, 25: 3003−3013. MEHTA P K, MONTEIRO P J M. Concrete: microstructure, properties, and materials [M]. New York: McGraw Hill, 2006: 659. BASILE T T, JAMES J B. Basic creep of hardened cement paste a re-examination of the role of water [J]. Cement and Concrete Research, 2000, 30: 1465−1475. FERNANDO P, ALMIR B S N, HENRIETTE L R, et al. Effect of the addition of synthetic fibers to concrete thin slabs on plastic shrinkage cracking [J]. Construction and Building Materials, 2010, 24: 2171−2176. MESSAN A, IENNY P, NECTOUX D. Free and restrained early-age shrinkage of mortar: Influence of glass fiber, cellulose ether and EVA (ethylene-vinyl acetate) [J]. Cement and Concrete Composites, 2011, 33: 402−410. CORINALDESI V, MORICONI G. Characterization of selfcompacting concretes prepared with different fibers and mineral additions [J]. Cement and Concrete Composites, 2011, 33: 596−601. SALIBA J, ROZIERE E, GRONDIN F, et al. Influence of shrinkage-reducing admixtures on plastic and long-term shrinkage [J]. Cement and Concrete Composites, 2011, 33: 209−217. MOHAMMED S M, MASAHIRO S, RYOICHI S. Influence of a combination of expansive and shrinkage-reducing admixture on autogenous deformation and self-stress of silica fume highperformance concrete [J]. Construction and Building Materials, 2011, 25: 239−250. JIANG Z W, SUN Z P, WANG P M. Autogenous relative humidity change and autogenous shrinkage of high-performance cement pastes [J]. Cement and Concrete Research, 2005, 35: 1539−1545. RYU D W, KO J W, NOGUCHI T. Effects of simulated environmental conditions on the internal relative humidity and relative moisture content distribution of exposed concrete [J]. Cement and Concrete Composites, 2011, 33: 142−153. KAGIMOTO H, KAWAMURA M. Measurements of strain and humidity within massive concrete cylinders related to the formation of ASR surface cracks [J]. Cement and Concrete Research, 2011, 41: 808−816. BENBOUDJEMA F, MEFTAH F, TORRENTI J M. Interaction between drying, shrinkage, creep and cracking phenomena in concrete [J]. Engineering Structures, 2005, 27: 239−250. MAZLOOM M. Estimating long-term creep and shrinkage of high-strength concrete [J]. Cement and Concrete Composites, 2008, 30: 316−326. BAZANT Z P. Prediction of concrete creep and shrinkage: past, present and future [J]. Nuclear Engineering and Design, 2001, 203: 27−38. ANDREAS L, PIETRO L, ROMAN L. Shrinkage and creep of SCC–The influence of paste volume and binder composition [J]. Construction and Building Materials, 2011, 25: 2283−2289. JIANG Z W, SUN Z P, WANG P M. Internal relative humidity distribution in high-performance cement paste due to moisture diffusion and self-desiccation [J]. Cement and Concrete Research, 2006, 36: 320−325. RAJABIPOUR F, SANT G, WEISS J. Interactions between shrinkage

Zhi-hai HE, et al/Progress in Natural Science: Materials International 21(2011) 426−432

[21]

[22]

[23]

[24]

[25]

reducing admixtures (SRA) and cement paste’s pore solution [J]. Cement and Concrete Research, 2008, 38: 606−615. FELEKOGLU B, TOSUN K, BARADAN B. Effects of fiber type and matrix structure on the mechanical performance of self-compacting micro-concrete composites [J]. Cement and Concrete Research, 2009, 39: 1023−1032. WANG X Y, LEE H S. Modeling the hydration of concrete incorporating fly ash or slag [J]. Cement and Concrete Research, 2010, 40: 984−996. SHEN B, HUBLER M, PAULINO G H, et al. Functionally-graded fiber-reinforced cement composite: processing, microstructure, and properties [J]. Cement and Concrete Composites, 2008, 30(8): 663−673. DEO O, NEITHALATH N. Compressive behavior of pervious concretes and a quantification of the influence of random pore structure features [J]. Materials Science and Engineering A, 2010, 528: 402−412. LI Y, BAO J L, GUO Y L. The relationship between autogenous

[26]

[27]

[28]

[29]

[30]

432

shrinkage and pore structure of cement paste with mineral admixtures [J]. Construction and Building Materials, 2010, 24: 1855−1860. LI G Y, ZHAO X H. Properties of concrete incorporating fly ash and ground granulated blast-furnace slag [J]. Cement and Concrete Composites, 2003, 25(3): 293−299. RIBEIRO A B, GONCALVES A, CARRAJOLA A. Effect of shrinkage reducing admixtures on the pore structure properties of mortars [J]. Materials and Structures, 2006, 39: 179−187. LEE Y, YI S T, KIM M S, KIM J K. Evaluation of a basic creep model with respect to autogenous shrinkage [J]. Cement and Concrete Research, 2006, 36: 1268−1278. HU S G, WU J, YANG W, et al. Relationship between autogenous deformation and internal relative humidity of high-strength expansive concrete [J]. Journal of Wuhan University of Technology-Materials Science, 2010, 25(3): 504−508. ZHANG J, QI K, HUANG Y. Calculation of moisture distribution in early-age concrete [J]. Journal of Engineering Mechanics-ASCE, 2009, 8: 871−880.