Effect of key mixture parameters on flow and mechanical properties of reactive powder concrete

Construction and Building Materials 99 (2015) 73–81

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Effect of key mixture parameters on flow and mechanical properties of reactive powder concrete Shamsad Ahmad ⇑, Ahmed Zubair, Mohammed Maslehuddin King Fahd University of Petroleum and Minerals, Dhahran 31261, Saudi Arabia

h i g h l i g h t s
The effect of key mixture parameters on mechanical properties of RPC evaluated statistically.
The modulus of rupture and modulus of elasticity of RPC correlated to its compressive strength.
Information presented in the paper can be used to produce optimum mixtures of RPC.

a r t i c l e

i n f o

Article history: Received 5 August 2014 Received in revised form 8 August 2015 Accepted 11 September 2015 Available online 19 September 2015 Keywords: Reactive powder concrete (RPC) Mixture parameters Mixture proportions Mechanical properties Compressive strength Modulus of rupture Modulus of elasticity

a b s t r a c t The main objective of the study presented in this paper was to examine the effect of key factors, which affect the performance of reactive powder concrete (RPC) mixtures. Firstly, an optimum sand grading was selected based on the maximum compressive strength and acceptable flow of a typical RPC mixture keeping the proportions of its ingredients constant. Then, keeping the sand grading and fiber content constant at their optimum levels, a total of 27 mixtures of RPC were selected for study by considering three levels of the three key factors namely water-to-binder ratio, cement content and silica fume content, according to a 33 factorial experiment design. The dosage of superplasticizer for each mixture was optimized to keep the flow in the desirable range of 180–220 mm. The performance of the selected mixtures of RPC was evaluated in terms of compressive strength, modulus of rupture and modulus of elasticity. Statistical analysis of the experimental data indicated the significant effect of sand grading, water-tobinder ratio, cement content and silica fume content on flowability and mechanical properties of RPC. The regression equations were obtained for mechanical properties of RPC mixtures in terms of the key mixture parameters, which can be utilized to optimize the proportions of RPC mixtures within the ranges of the mixture variables considered in this study. Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction It has been a challenging task to produce a very high performance concrete due to the lack of good quality coarse aggregates in many parts of the world. In a conventional mixture of concrete, the coarse aggregate particles weaker than the surrounding mortar are crushed before mortar phase besides the presence of transition zone between the coarse aggregate and mortar matrix, which is often the source of micro cracks in concrete reducing the strength and durability [1,2]. Recently, advances in concrete technology have been reported in literature leading to the development of the reactive powder concrete (RPC), also known as ultra-high performance concrete (UHPC), is relatively new generation of concrete produced as a ultra-dense mixture of water, Portland cement, silica ⇑ Corresponding author. E-mail address: [email protected] (S. Ahmad). http://dx.doi.org/10.1016/j.conbuildmat.2015.09.010 0950-0618/Ó 2015 Elsevier Ltd. All rights reserved.

fume, fine quartz sand, quartz powder, superplasticizer and steel fibers. The RPC mixtures are optimized at the nano and microscale to provide superior mechanical and durability properties compared to conventional and high performance concretes. The quality requirement of RPC are achieved through: limiting the water-to-cementitious materials ratio to less than 0.20, optimizing particle packing, eliminating coarse aggregate, and implementing special curing regimes. Short fibers are added to enhance the material’s tensile and flexural strength, ductility, and toughness [2]. The RPC mixtures are designed to have flow in the desirable range of 180 ± 220 mm. The performance of the hardened RPC is evaluated in terms of mechanical properties such as compressive strength, modulus of elasticity, flexural tensile strength, and fracture toughness. From the literature survey [3–15], the ranges of mechanical properties of RPC at the age of 28 days are found as follows: compressive strength – 130–260 MPa; flexural tensile strength – 30–60 MPa; split-tensile strength – 6–8 MPa; modulus

74

S. Ahmad et al. / Construction and Building Materials 99 (2015) 73–81

of elasticity – 40–70 GPa; and fracture energy – 20–40 kJ/m2. The requirements of the constituent materials and their proportions for producing RPC mixtures are summarized in the following paragraphs. From the point of view of chemical composition, cements with low C3A content (for reducing the water demand) give better results [3]. Use of cements with a high fineness should be avoided due to its high water demand. The best cement in terms of rheological characteristics and mechanical performance is high silicamodulus cement. However, this type of cement has the disadvantage of a very slow setting rate, preventing its certain applications. Conventional quick-setting high performance cement offers very similar mechanical performance, despite a higher water demand. Silica fume is one of the main constituents of RPC. Silica fume in RPC has the three main functions, which greatly improve the properties of RPC. These are as follows: filling the voids in the next larger granular class, namely cement; enhancing lubrication of the mixture due to the perfect sphericity of the basic particles; and production of secondary hydrates by the pozzolanic reaction with the Ca(OH)2 from primary hydration of cement [3,16]. In typical Portland cement based concrete, 18% silica fume, by weight of the cementitious materials, is enough for total consumption of Ca (OH)2 released from cement hydration [17]. However, considering the filler effect the optimal share of the silica fume may increases up to 30% of cement [3]. Therefore, the silica fume content in RPC is normally kept in the range of 25–30% of the cementitious material. Typically, the silica fume/cement ratio used for RPC is 0.25. This ratio corresponds to optimum filling performance and it is close to the dosage required for complete consumption of the lime resulting from total hydration of cement. However, cement hydration is incomplete in an RPC, and the available quantity of silica fume is more than that required by the pozzolanic reaction. Utilization of fly ash (FA) and ground granulated blast furnace slag (GGBFS) as an alternative to silica fume in RPC has been reported in the literature [4,11]. For enhancing the homogeneity of RPC, coarse aggregate is replaced by fine quartz sand. The maximum size of sand is recommended to be 600 lm for use in RPC [3]. Sand constitutes the largest portion of RPC with about 41% by weight of RPC. To obtain a highly homogeneous matrix as well as minimum void, RPC contains finely graded sand between 150 lm and 600 lm. Sand particle sizes below 150 lm are avoided for preventing interference with the largest cement particles (80–100 lm). Sand with a mean particle size of about 250 lm is preferred. Crushed crystalline quartz powder in the size range of 10– 15 lm is used as filler in RPC. Since quartz powder is a reactive material, it acts as an excellent paste-aggregate interface filler [3]. For cases where heat-treatment is employed, quartz powder demonstrates even higher reactivity. Maximum reactivity during heat-treatment is obtained for a mean particle size of between 5 and 25 lm. The mean particle size of the crushed quartz used for an RPC is l0 lm, and is therefore in the same granular class as the cement [3]. Since RPC uses a small water/binder ratio, superplasticizer is needed to achieve its required flowability. High performance superplasticizers having either polycarboxylate, Naphthalene Sulfonate or Melamine Sulfonate (MS) enable in producing dense and highly homogeneous RPC mixtures, which can be poured without segregation. The most efficient superplasticizers are polyacrylate-based dispersing agents, but it exhibits a retarding characteristic, which can pose a problem for practical applications. RPC without fibers is also strong but very brittle, consequently fibers are included to increase the tensile capacity and improve its ductility. Studies using different fiber materials, contents, sizes, and shapes have been conducted by various researchers [18]. Dimensionally, the largest constituent in the mix are the steel

fibers. Given the relative sizes of the sand and the fibers, the steel fibers are able to reinforce the concrete matrix on a micro level [8]. The addition of steel fibers helps in preventing the propagation of micro-cracks and macro-cracks and thereby limits crack width and permeability. Because of its size relative to the other constituents, it reinforces the concrete on the micro level and eliminates the need for secondary reinforcement in prestressed bridge girders [19]. An optimum dosage of 6.2% (by weight of RPC) of steel fibers is recommended by DuctalÒ [20]. Sobolev [21] has presented the following approach for optimizing RPC mixture using the rheological and strength models. First, the optimal silica fume (SF) content and superplasticizer (SP) dosage are selected according to the strength model of modified mortars: for optimal performance, SF content is specified within 10–15% and SP dosage is set to be 10% of SF. Second, the aggregates are optimized to fit a specific grading curve. Then water to cement ratio is selected using the strength model. The parameters considered in the design of mixture of RPC are mainly, water to binder ratio, cement content, micro silica to cement ratio, total cementitious material content, total fine aggregate content, and fiber content. From literature survey [3–15], the ranges for these parameters are found to be as follows. Water to total binder ratio: 0.15–0.24 (by weight); cement content: 800–1100 kg/m3; silica fume content: 150–300 kg/m3; silica fume to cement ratio: 0.15– 0.35 (by weight); cement and micro-silica (i.e., binder or cementitious materials) content: 950–1400 kg/m3; sand and quartz content: 1000–1400 kg/m3; steel fiber content: 190–250 kg/m3; and steel fiber to total binder ratio: 0.15–0.30 (by weight). From the literature review presented above, it can be noted that a series of the mixtures of RPC can be produced with different combinations of the levels of the key factors within their ranges of variation. This exercise would help in determining optimum proportioning of the constituents of RPC based on the minimum unit cost of RPC satisfying the flowability and mechanical properties. In the present study, an attempt was made first to select the optimum sand grading and then selecting a total number of 27 mixtures of RPC considering three levels of water/binder ratio, cement content and silica fume content according to a 33 factorial experiment design. For all 27 mixtures, optimum dosages of superplasticizer were determined based on the required flowability followed by evaluation of their performance in terms of mechanical properties. Statistical analysis of the experimental data was carried out to examine the significance of each factor and finally correlation equations were obtained which could be utilized for optimizing the mixture proportioning. 2. Experimental program 2.1. Materials 2.1.1. Cement and silica fume Type I cement (ordinary Portland cement) conforming to ASTM C150 [22] with a specific gravity of 3.15 and chemical composition, as shown in Table 1, was used in all the mixtures of RPC. The chemical composition of the silica fume used is also shown in Table 1. 2.1.2. Fine dune sand Fine dune sand obtained from the deserts of Saudi Arabia, with water absorption of 0.5% and specific gravity of 2.53, was used as aggregate. The natural grading of sand, used in all the mixtures, is shown in Table 2. 2.1.3. Superplasticizer and steel fibers A liquid superplasticizer (commercial name: Glenium 51) was used to obtain the desired flow. Glenium 51 is a polycarboxylic ether (PCE) based superplasticizer, which does not contain chlorides and complies with ASTM C494 [23] Types A and F. The specific gravity of Glenium 51 was 1.095% with 65% water content by weight. Varying dosage of this superplasticizer was used to obtain a flow of 200 ± 20 mm for all the mixtures. Steel fibers of 0.22 mm diameter and 13 mm length with tensile strength greater than 2850 MPa were used in all the mixtures.

S. Ahmad et al. / Construction and Building Materials 99 (2015) 73–81

75


After the fibers have been added, continue running mixer for further 3 min to ensure that the fibers are well dispersed.

Table 1 Chemical composition of cement and silica fume. Constituent

Portland cement (wt. %)

Silica fume (wt. %)

CaO SiO2 Al2O3 Fe2O3 K2O MgO Na2O Equivalent alkalis (Na2O + 0.658K2O) SO3 Loss on ignition

64.35 22.0 5.64 3.80 0.36 2.11 0.19 0.33

0.41 86.75 0.41 2.12 0.67 0.18 0.17 0.62

2.10 0.70

0.73 3.35

The entire mixing process takes about 20–25 min and is specific to the constituents of the mixture and the mixer used, as shown in Fig. 1. Mixing of the RPC requires special attention to have uniform consistency. The prepared RPC mixture was poured into the molds. Consolidation was done using a vibrating table. As soon as mixing was completed, RPC mix was tested for flowability. Standard test method for measuring flow of hydraulic cement according to ASTM C1437 [24] was used to determine flowability of the RPC mixtures. In the flowability test, a mini slump cone was filled with RPC mixture and then the cone was removed slowly to allow the RPC to flow evenly on the table and then the flow table was dropped 25 times and its average diameter is recorded. The average flow diameter of RPC mix ranged from 180 to 220 mm.

2.5. Preparation and curing of specimens

Table 2 Dune sand grading. Sieve opening (mm)

Cumulative % retained

% Passing

4.75 2.4 1.2 0.6 0.3 0.15 0.075

0 0 0 3.8 38.6 78.1 99.0

100 100 100 96.2 61.4 21.9 1.0

Batching of each mixture was done by weight. After mixing, the flow was measured and RPC was poured in the molds. The molds were then vibrated until complete consolidation was achieved. RPC specimens were prepared and cured to carry out various tests planned under this study. A total of 405 cubical specimens of size 50 mm were cast for compressive strength test (27 mixtures  5 test ages  3 replicates). A total of 81 prisms of size 40  40  160 mm were cast for modulus of rupture (MOR) test and a similar number of 75 mm  150 mm cylindrical specimens were cast for modulus of elasticity test. After casting, the specimens were covered with plastic sheet for 24 h and placed in the laboratory environment (22 ± 3 °C) to minimize loss of water from the mixtures. After 24 h, the specimens were demolded and placed in a curing tank till the time of test.

2.6. Testing of specimens 2.2. Optimization of sand grading Trial mixtures of plain RPC (i.e., without fibers) with different sand grading were prepared to identify the optimum grading of the dune sand used in this study based on the best performance of the RPC mixtures. For this purpose, the proportions of constituent materials for each trial mixture were kept same (water/binder ratio = 0.20, cement content = 1000 kg/m3, silica fume content = 150 kg/m3, water content = 230 kg/m3, and sand content = 977 kg/m3, the mixture of cement and silica fume is termed in this paper as ‘binder’). Different trial sand grading, which were used are as follows: (i) natural grading; (ii) sand passing 600 lm sieve and retained on 150 lm sieve; (iii) sand passing 600 lm sieve; (iv) sand passing 300 lm sieve; (v) sand passing 150 lm sieve; and (vi) 1/3rd sand passing 600 lm sieve, 1/3rd sand passing 300 lm sieve and 1/3rd sand passing 150 lm sieve. This way a total of six trial mixtures of plain RPC were prepared and tested for flow and compressive strength after 28 days of water curing. 2.3. Proportions of RPC mixtures Three key mixture variables (water/binder ratio, cement content and silica fume content) were considered with their three levels for studying the effect of these mixture parameters on the performance of a total of 27 RPC mixtures as per 33 factorial experiment design. The three levels of each variables, selected within their practical ranges for producing RPC, are given below: Water/binder (W/B) ratio: 0.15, 0.175, and 0.20. Cement content (kg/m3): 1000, 1100, and 1200. Silica fume content (% of cement): 15%, 20%, and 25%. Steel fiber content of 157 kg/m3, corresponding to an optimal dosage of 6.2% of the mass of fresh RPC [20], was used in all the mixtures. Absolute volume method was used to design the mixtures. The weights of constituents determined for producing one cubic meter of each of the RPC mixtures are presented in Table 3. 2.4. Mixing procedure and flow test Since RPC is composed of very fine materials, the conventional mixing method is not appropriate. The following sequence in mixing of RPC was followed based on the previous studies [12–14], and as well as from the experience gained after several trials. The mixing procedure adopted is as follows:
Dry mixing the powders (including cement, sand and silica fume) for about 3 min with a low speed of about 140 revolutions/min.
Addition of half volume of water containing half amount of superplasticizer.
Mixing for about 3 min with a high speed of about 285 revolutions/min.
Addition of the remaining water and superplasticizer.
Mixing for about ten min with a high speed of about 285 revolutions/min.
Finally, adding steel fibers in small amounts over the course of the next 2 min into the mixture.

2.6.1. Compressive strength The compressive strength was determined on 50 mm cube specimens according to ASTM C109 [25]. The specimens were tested using a digital compression-testing machine after 3, 7, 14, 28 and 90 days of water curing. Three specimens were tested at each age and the average values were reported.

2.6.2. Flexural tensile strength/modulus of rupture (MOR) test The standard four-point flexural test conforming to ASTM C78 [26] was used for determining flexural tensile strength, i.e., the modulus of rupture (MOR). This test involved the four-point flexural loading of small-scale concrete prism specimens measuring 40  40  160 mm. During the test, the load and the mid-span deflection of the prism were recorded. These data were used to determine the MOR.

2.6.3. Modulus of elasticity As specified in ASTM C469 [27], which covers standard test method for static modulus of elasticity and Poisson’s ratio of concrete in compression, the elastic portion of the compressive stress–strain curve up to 40% of the ultimate compressive strength was used to determine the modulus of elasticity. Three 75 mm diameter and 150 mm height cylindrical specimens were utilized to determine the modulus of elasticity. The test setup included a specially designed axial deformation gauge. The two parallel rings were rigidly attached to the cylinder with a 75 mm gage length between the attachment points. The lower ring held two LVDTs whose ends were connected to the upper ring. The axial deformation of the cylinder was measured from initiation of loading through failure. The load and the output from the three LVDTs were digitally recorded throughout the test using a data logger. The testing of each cylinder was completed in a single constant load application from start to failure. Proper seating of the cylinder was ensured by monitoring the load-deformation response during the test. The modulus of elasticity was calculated based on the average LVDT-based deformation measurements and the load reading.

3. Results and discussion 3.1. Optimum sand grading The flow and compression tests results for six trial mixtures of RPC, prepared using various sand grades, are presented in Table 4. It can be seen from Table 4 that the natural sand grading gave the best results, as it required least amount of superplasticizer and had highest compressive strength with the flow in desirable range of 200 ± 20 mm. Hence, natural sand grading was considered to be the optimum one and used for the preparation of 27 RPC mixtures for detailed study.

76

S. Ahmad et al. / Construction and Building Materials 99 (2015) 73–81

Table 3 Weights of the constituents of 27 RPC mixtures (for 1 m3). Mix ID

W/B ratio

Cement (kg)

Silica fume (%)

Silica fume (kg)

Water (kg)

Fiber (kg)

SP (%)

SP (kg)

Sand (kg)

M1 M2 M3 M4 M5 M6 M7 M8 M9 M10 M11 M12 M13 M14 M15 M16 M17 M18 M19 M20 M21 M22 M23 M24 M25 M26 M27

0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.175 0.175 0.175 0.175 0.175 0.175 0.175 0.175 0.175 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20

1000 1000 1000 1100 1100 1100 1200 1200 1200 1000 1000 1000 1100 1100 1100 1200 1200 1200 1000 1000 1000 1100 1100 1100 1200 1200 1200

15 20 25 15 20 25 15 20 25 15 20 25 15 20 25 15 20 25 15 20 25 15 20 25 15 20 25

150 200 250 165 220 275 180 240 300 150 200 250 165 220 275 180 240 300 150 200 250 165 220 275 180 240 300

172.5 180 187.5 189.75 198 206.25 207 216 225 201.25 210 218.75 221.375 231 240.625 241.5 252 262.5 230 240 250 253 264 275 276 288 300

157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157

3.55 3.55 3.55 3.55 3.55 3.55 3.55 3.55 3.55 2 2 2 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1 1 1 1 1 1

40.83 42.6 44.38 44.91 46.86 48.81 48.99 51.12 53.25 23 24 25 18.98 19.8 20.63 20.7 21.6 22.5 17.25 18 18.75 12.65 13.2 13.75 13.8 14.4 15

976.81 897.51 818.21 826.55 739.32 652.09 676.29 581.13 485.97 945.25 864.58 783.91 806.45 718.35 630.25 654.37 558.26 462.15 885.8 802.55 719.29 741.06 650.11 559.17 583.03 483.81 384.60

superplasticizer was in the range of 1–1.5% in the RPC mixtures with a W/B ratio of 0.20. As expected, it is evident that the requirement of superplasticizer increased with a decrease in the W/B ratio. 3.3. Effect of key factors on compressive strength Table 6 lists the results of compressive strength test conducted on all the 27 RPC mixtures after 3, 7, 14, 28 and 90 days of water curing. The values of compressive strength of all 27 mixtures at each curing period were plotted as shown in Fig. 2. It can be observed from Fig. 2 that, like normal concrete, there is an increase in compressive strength with increase in the curing period. The variation of compressive strength at each curing period is due to the effect of mixture variables considered in this study. The maximum effect of mixture variables at a given curing period can be realized by the difference between minimum and maximum values of compressive strengths for that curing period. The percentage differences between minimum and maximum values of compressive strengths for each curing period indicate that the effect of the mixture variables is highest for 3-days curing period and lowest for 28-days curing period as the difference in minimum and maximum values for 3-days and 28-days curing periods are 31%

Fig. 1. Planetary Mixer (MIKRONS) used for mixing the constituents of RPC.

3.2. Optimum dosages of superplasticizer Several trials were carried out to optimize the superplasticizer dosage for each of the 27 RPC mixtures to meet the targeted flow of 200 ± 20 mm in each case. Results showing the optimum superplasticizer dosages and corresponding flow for all 27 mixtures are presented in Table 5. From Table 5, it is evident that the optimum superplasticizer dosage was 3.6% (by the cementitious material) for all mixtures with a W/B ratio of 0.15. For mixtures with W/B ratio of 0.175 optimum super plasticizer dosage was in the range of 1.5–2% of the cementitious material. The optimum dosage of

Table 4 Flow and compressive strength of RPC specimens prepared with different sand grading. Trial mix # Sand grading

SP Flow 28-Day (%) (mm) compressive strength (MPa)

1 2

1.5 200 1.7 180

126 117

1.8 2.0 2.1 1.9

113 109 101 64

3 4 5 6

Natural, as shown in Table 2 Passing 600 lm sieve and retained on 150 lm Passing 600 lm sieve Passing 300 lm sieve Passing 150 lm sieve Mixed (1/3rd from each of stocks passing through 600, 300 and 150 lm sieves)

160 215 175 195

77

S. Ahmad et al. / Construction and Building Materials 99 (2015) 73–81 Table 5 Optimum dosages of superplasticizer for all 27 RPC mixtures to meet flow criteria of 200 ± 20 mm. Mix #

Optimum SP dosage (%)

Flow (mm)

Mix #

Optimum SP dosage (%)

Flow (mm)

Mix #

Optimum SP dosage (%)

Flow (mm)

M1 M2 M3 M4 M5 M6 M7 M8 M9

3.60 3.60 3.60 3.60 3.60 3.60 3.60 3.60 3.60

180 180 180 200 210 190 200 205 205

M10 M11 M12 M13 M14 M15 M16 M17 M18

2.00 2.00 2.00 1.50 1.50 1.50 1.50 1.50 1.50

220 200 220 185 205 190 225 200 200

M19 M20 M21 M22 M23 M24 M25 M26 M27

1.50 1.50 1.50 1.00 1.00 1.00 1.00 1.00 1.00

220 215 220 190 188 190 210 200 190

Table 6 Compressive strength test results. Mix #

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

Average compressive strength (MPa) 3 days

7 days

14 days

28 days

90 days

92.0 94.0 107.0 100.8 103.7 110.0 99.9 104.1 108.0 98.7 101.3 104.0 108.2 108.6 110.0 90.1 91.6 98.4 91.0 99.9 100.5 83.8 98.8 101.0 90.3 99.7 102.0

104.0 106.0 120.7 108.6 113.2 119.1 107.2 110.6 115.0 111.6 119.6 121.0 117.2 119.8 121.0 107.0 112.1 113.2 105.6 108.0 109.8 102.0 104.5 107.0 102.5 108.6 110.2

118.0 122.7 129.4 119.1 125.2 131.0 119.4 123.0 126.0 125.4 129.3 133.0 119.3 123.4 127.0 112.4 120.2 121.9 115.7 119.5 122.0 114.5 115.5 117.0 111.1 115.7 117.1

132.0 135.6 136.9 132.0 136.3 138.9 132.8 135.0 137.0 129.4 133.3 135.0 128.4 130.0 133.0 130.0 134.0 136.0 121.5 126.2 128.0 123.3 125.7 128.3 128.0 132.2 135.0

140.0 145.4 150.7 136.7 147.0 154.4 136.4 140.0 143.8 142.0 146.3 149.0 140.0 147.1 148.5 133.0 136.7 154.5 130.3 140.1 143.8 130.9 133.3 136.6 138.3 142.8 144.0

Fig. 3. Variation of 28-day compressive strength of RPC mixtures with a cement content of 1000 kg/m3.

Fig. 4. Variation of 28-day compressive strength of RPC mixtures with a cement content of 1100 kg/m3.

Fig. 2. Compressive strength development for all the 27 RPC mixtures.

and 14%, respectively. Effect of mixture variables for 28-days curing period is almost half of that for 3-days cuing period. For other curing durations (7, 14, and 90 days) the difference in minimum and maximum values is almost same (around 20%). Considering 28-day curing as standard one, the effect of the mixture variables

on 28-day compressive of the RPC mixtures are discussed as follows. The variation of 28-day compressive strength of the RPC mixtures with silica fume content and W/B ratio at constant cement contents of 1000, 1100 and 1200 kg/m3 are shown in Figs. 3–5, respectively. Increase in the compressive strength of RPC is noted with increase in the silica fume content and decrease in the W/B ratio. However, the effect of W/B ratio is negligible at a higher cement content of 1200 kg/m3 and a silica fume content of 25%, as evident from Fig. 5. The analysis of variance (ANOVA) conducted for 28-day compressive strength data, as shown in Table 7, indicated that the W/B ratio has most significant effect followed by silica fume content with least effect of cement content. The variation of the compressive strength with the selected mixture parameters indicates the possibility of optimizing the mixtures of RPC if a

78

S. Ahmad et al. / Construction and Building Materials 99 (2015) 73–81 Table 8 28-Day MOR of RPC mixtures. Mix #

Average MOR (MPa)

Mix #

Average MOR (MPa)

Mix #

Average MOR (MPa)

M1 M2 M3 M4 M5 M6 M7 M8 M9

33.0 34.4 36.6 27.8 29.8 32.9 26.0 30.9 34.8

M10 M11 M12 M13 M14 M15 M16 M17 M18

32.4 34.0 34.9 24.5 25.8 28.0 24.9 26.5 27.5

M19 M20 M21 M22 M23 M24 M25 M26 M27

27.9 30.7 31.5 23.0 25.7 27.1 25.0 26.0 26.0

Fig. 5. Variation of 28-day compressive strength of RPC mixtures with a cement content of 1200 kg/m3.

regression equation correlating compressive strength with the mixture parameters is obtained utilizing the experimental data generated through the present work. The regression equation relating the 28-day compressive strength to the W/B ratio, cement content, and silica fume content is given as follows: 0

f c ¼ 133  151W=B þ 0:0123C þ 0:564S ½R2 ¼ 0:99

ð1Þ

where, f0 c = 28-day compressive strength (MPa). W/B = water/binder ratio (by mass). C = cement content (kg/m3). S = silica fume content (% of cement content).

Fig. 6. Variation of 28-day MOR of RPC mixtures with a cement content of 1000 kg/m3.

Eq. (1) indicates that the compressive strength increases with the cementitious materials content (C and S) but it decreases with increase in the W/B ratio. This is in line with the behavior of normal concrete mixtures. 3.4. Effect of key factors on modulus of rupture (MOR) 28-day MOR values obtained for 27 RPC mixtures are presented in Table 8. The variation of 28-day MOR of the RPC mixtures with silica fume content and W/B ratio at constant cement contents of 1000, 1100 and 1200 kg/m3 are shown in Figs. 6–8, respectively. Like compressive strength, the MOR of RPC increased with increase in the silica fume content and decrease in the W/B ratio. It can be observed from Figs. 6–8 that at lower cement content (1000 kg/ m3), the effect of lower W/B ratios (0.15 and 0.175) on MOR are

not that significant. At higher cement contents (1100 and 1200 kg/m3), the effect of higher W/B ratios (0.175 and 0.2) on MOR are negligible. This indicates that the significant increase in MOR by decreasing W/B ratio occurs at higher cement content. ANOVA conducted for 28-day MOR data, as shown in Table 9, indicated that the cement content has most significant effect followed by W/B ratio with least effect of silica fume content. The regression equation relating the 28-day MOR to the W/B ratio, cement content, and silica fume content is given below:

MOR ¼ 69:6  85:6W=B  0:0293C þ 0:332S ½R2 ¼ 0:98

ð2Þ

where, MOR = 28-day modulus of rupture (MPa). W/B = water/binder ratio (by mass). C = cement content (kg/m3). S = silica fume content (% of cement content).

Table 7 Results of the ANOVA of the 28-day compressive strength data. Factor

Type

Levels

Values

W/B (water/binder ratio) C (cement content) S (SF content)

Fixed Fixed Fixed

3 3 3

0.150, 0.175, 0.200 (by mass) 1000, 1100, 1200 kg/m3 15%, 20%, 25% of the cement content

Analysis of variance (ANOVA), using adjusted SS for tests Source Degree of freedom (DF) Sum squares (SS) W/B C S W/B * C C*S W/B * S Error Total

2 2 2 4 4 4 8 26

260.90 39.69 145.25 52.77 1.55 0.69 4.45 505.32

Adjusted SS

Adjusted mean sum squares (MS)

260.90 130.45 39.69 19.85 145.25 72.63 52.77 13.19 1.55 0.39 0.69 0.17 4.45 0.56 – – S = 0.75 R-Sq = 99.12% R-Sq(adj) = 97.14%

F-ratio

Probability of no-significance (P)

234.40 35.67 130.50 23.71 0.70 0.31 – –

0.00 0.00 0.00 0.00 0.62 0.86 – –

79

S. Ahmad et al. / Construction and Building Materials 99 (2015) 73–81 Table 10 28-Day modulus of elasticity (E) of RPC mixtures. Mix #

Average E (GPa)

Mix #

Average E (GPa)

Mix #

Average E (GPa)

M1 M2 M3 M4 M5 M6 M7 M8 M9

51.0 52.5 53.0 44.8 46.5 47.6 45.7 47.8 48.4

M10 M11 M12 M13 M14 M15 M16 M17 M18

48.3 49.2 50.0 42.9 44.4 45.0 41.2 42.4 45.7

M19 M20 M21 M22 M23 M24 M25 M26 M27

40.3 43.2 44.5 41.6 42.0 43.0 40.0 41.0 42.5

Fig. 7. Variation of 28-day MOR of RPC mixtures with a cement content of 1100 kg/m3.

Fig. 9. Variation of 28-day modulus of elasticity of RPC mixtures with a cement content of 1000 kg/m3.

Fig. 8. Variation of 28-day MOR of RPC mixtures with a cement content of 1200 kg/m3.

Eq. (2) indicates that the MOR decreases with increase in W/B ratio and cement content but it increases with the silica fume content.

3.5. Effect of key factors on modulus of elasticity Table 10 lists the average values of secant modulus of elasticity of RPC mixtures after 28 days of water curing. The variation of 28day modulus of elasticity of the RPC mixtures with silica fume con-

tent and W/B ratio at constant cement contents of 1000, 1100 and 1200 kg/m3 are shown in Figs. 9–11, respectively. Like compressive strength and MOR, modulus of elasticity of RPC mixtures also increased with increase in silica fume content and decrease in the W/B ratio. It was observed from the ANOVA conducted for modulus of elasticity, as shown in Table 11, that the W/B ratio affected the modulus of elasticity most significantly followed by cement content with least effect of the silica fume content. The regression equation relating the 28-day modulus of elasticity to the W/B ratio, cement content, and silica fume content is given below:

E ¼ 85:8  132W=B  0:0206C þ 0:266S ½R2 ¼ 0:99

ð3Þ

E = 28-day modulus of elasticity (GPa). W/B = water/binder ratio (by mass).

Table 9 Results of the ANOVA of the 28-day MOR data. Factor

Type

Levels

Values

W/B (water/binder ratio) C (cement content) S (SF content)

Fixed Fixed Fixed

3 3 3

0.150, 0.175, 0.200 (by mass) 1000, 1100, 1200 kg/m3 15%, 20%, 25% of the cement content

Analysis of variance (ANOVA), using adjusted SS for tests Source Degree of freedom (DF) Sum squares (SS) W/B C S W/B * C C*S W/B * S Error Total

2 2 2 4 4 4 8 26

107.45 180.88 67.71 13.25 1.19 9.63 9.63 390.05

Adjusted SS

Adjusted mean sum squares (MS)

107.45 53.87 180.88 90.44 67.71 33.86 13.25 3.31 1.19 0.30 9.63 2.41 9.63 1.21 – – S = 1.09 R-Sq = 97.53% R-Sq(adj) = 91.97%

F-ratio

Probability of no-significance (P)

44.72 75.08 28.10 2.75 0.25 2.00 – –

0.00 0.00 0.00 0.10 0.90 0.19 – –

80

S. Ahmad et al. / Construction and Building Materials 99 (2015) 73–81

qffiffiffiffi 0 MOR ¼ 3:08 f c ½R2 ¼ 0:86

ð4Þ

The equation correlating MOR with compressive strength of normal concrete, as given by ASTM C78 [26] is as follows qffiffiffiffi 0 MOR ¼ 0:99 f c . By comparing the MOR of RPC mixtures with that of normal concrete at same compressive strength, it is found that the MOR of RPC is more than three times of the MOR of normal concrete. The equation correlating modulus of elasticity with compressive strength, obtained utilizing 28-day compressive strength and modulus of elasticity data of the 27 RPC mixtures is given below:

qffiffiffiffi 0 E ¼ 4:36 f c ½R2 ¼ 0:85 Fig. 10. Variation of 28-day modulus of elasticity of RPC mixtures with a cement content of 1100 kg/m3.

ð5Þ

The equation correlating modulus of elasticity with compressive strength of concrete having strengths in the range of 80 and 140 MPa, as proposed by Kakizaki et al.[28], is as follows: qffiffiffiffi 0 E ¼ 3:65 f c . By comparing the modulus of elasticity of RPC mixtures with that of other concrete with same compressive strength, it is found that the modulus of elasticity of RPC is more by around 20% of the modulus of elasticity of other high strength concrete. 3.7. Utilization of the derived regression models for f0 c, MOR and E

Fig. 11. Variation of 28-day modulus of elasticity of RPC mixtures with a cement content of 1200 kg/m3.

C = cement content (kg/m3) S = silica fume content (% of cement content).

The regression models obtained for f0 c, MOR and E of RPC mixtures, as expressed by Eqs. (1)–(3), respectively, can be utilized for the purpose of obtaining an optimum mixture proportion satisfying the targeted values of f0 c, MOR and E. The mixture optimization can be carried out using simply the Excel-Solver by considering the unit cost of the RPC mixture as objective function and considering the models for f0 c, MOR and E as constraints. The total unit cost of a RPC mixture would be taken as the sum of the costs of all ingredients of RPC including the cost of superplasticizer. The mixture satisfying the targeted f0 c, MOR and E and having the minimum unit cost would be taken as the optimal one. 4. Conclusions

Eq. (3) indicates that the modulus of elasticity decreases with increase in W/B ratio and cement content but it increases with the silica fume content. 3.6. Correlation between compressive strength and other mechanical properties The equation correlating MOR with compressive strength, obtained utilizing 28-day compressive strength and MOR data of the 27 RPC mixtures, is given below:

Based on the findings of the present study, following conclusions can be made: 1. The natural grading of dune sand available in Saudi Arabia performed better than the modified sand grading. 2. The dosage of superplasticizer was affected mainly by the W/B ratio and to some extent by the cement content. The optimum superplasticizer dosages were typically found as 3.6%, 1.5–2%, and 1–1.5% by the weight of binder at W/B ratios of 0.15, 0.175, and 0.2, respectively.

Table 11 Results of the ANOVA of the 28-day modulus of elasticity data. Factor

Type

Levels

Values

W/B (water/binder ratio) C (Cement content) S (SF content)

Fixed Fixed Fixed

3 3 3

0.150, 0.175, 0.200 (by mass) 1000, 1100, 1200 kg/m3 15%, 20%, 25% of the cement content

Analysis of variance (ANOVA), using adjusted SS for tests Source Degree of freedom (DF) Sum squares (SS) W/B C S W/B * C C*S W/B * S Error Total

2 2 2 4 4 4 8 26

194.85 94.20 31.90 30.44 1.39 0.56 4.44 357.78

Adjusted SS

Adjusted mean sum squares (MS)

194.85 97.43 94.20 47.10 31.90 15.95 30.44 7.61 1.39 0.35 0.56 0.14 4.44 0.56 – – S = 0.74 R-Sq = 98.76% R-Sq(adj) = 95.97%

F-ratio

Probability of no-significance (P)

175.64 84.91 28.75 13.72 0.62 0.25 – –

0.00 0.00 0.00 0.00 0.66 0.90 – –

S. Ahmad et al. / Construction and Building Materials 99 (2015) 73–81

3. All the three key mixture variables considered in the present study were found affecting f’c, MOR and E. The statistical analysis of the experimental data using ANOVA indicated that while the W/B ratio affected compressive strength and modulus of elasticity of RPC most significantly, the cement content had major effect on the modulus of rupture of RPC. It is observed from the regression equations that while the f0 c increases with increase in cement content the MOR and E decrease. 4. The equations showing correlation between compressive strength and MOR and compressive strength and modulus of elasticity for RPC mixtures have indicated that for the same compressive strength value, the RPC has three times more MOR than a normal concrete and around 20% more modulus of elasticity than other high strength concrete. 5. The regression equations obtained for f0 c, MOR and E of the RPC mixtures can be utilized to carry out optimum design of the RPC mixtures within the ranges of mixture variables considered in this study.

Acknowledgments The authors gratefully acknowledge the financial support provided by King Fahd University of Petroleum & Minerals, Dhahran, Saudi Arabia under the research grant (Project No. RG1005-1 and RG1005-2). The logistical support of the Department of Civil & Environmental Engineering and the Research Institute, King Fahd University of Petroleum & Minerals, Dhahran, Saudi Arabia, is also acknowledged with appreciation. References [1] M. Suzuki, M.S. Meddah, R. Sato, Use of porous ceramic waste aggregates for internal curing of high-performance concrete, Cem. Concr. Res. 39 (5) (2009) 373–381. [2] V.Y.G. Yanni, Multi-Scale Investigation of Tensile Creep of Ultra High Performance Concrete for Bridge Applications (Ph.D. thesis), Georgia Institute of Technology, 2009. [3] P. Richard, M. Cheyrezy, Composition of reactive powder concretes, Cem. Concr. Res. 25 (7) (1995) 1501–1511. [4] G. Long, X. Wang, Y. Xie, Very-high-performance concrete with ultrafine powders, Cem. Concr. Res. 32 (4) (2002) 601–605. [5] J. Ma, H. Schneider, Properties of ultra-high-performance concrete, LACER 7 (2002) 25–32. [6] Y.W. Chan, S.H. Chu, Effect of silica fume on steel fiber bond characteristics in reactive powder concrete, Cem. Concr. Res. 34 (7) (2004) 1167–1172. [7] J. Kaufmann, F. Winnefeld, D. Hesselbarth, Effect of the addition of ultrafine cement and short fiber reinforcement on shrinkage, rheological and mechanical properties of Portland cement pastes, Cement Concr. Compos. 26 (5) (2004) 541–549.

81

[8] B.A. Graybeal, Material Property Characterization of Ultra-High Performance Concrete, Report No. FHWA-HRT-06-103, Federal Highway Administration, US Department of Transportation, 2006. [9] H. Yazici, H. Yig˘iter, A.S. Karabulut, B. Baradan, Utilization of fly ash and ground granulated blast furnace slag as an alternative silica source in reactive powder concrete, Fuel 87 (12) (2008) 2401–2407. [10] J. Liu, S. Song, L. Wang, Durability and micro-structure of reactive powder concrete, J. Wuhan Univ. Technol. Mater. Sci. Ed. 24 (3) (2009) 506–509. [11] H. Yazici, M.Y. Yardimci, H. Yig˘iter, S. Aydin, S. Türkel, Mechanical properties of reactive powder concrete containing high volumes of ground granulated blast furnace slag, Cem. Concr. Compos. 32 (8) (2010) 639–648. [12] O. Mazanec, D. Lowke, P. Schiel, Mixing of high performance concrete: effect of concrete composition and mixing intensity on mixing time, Mater. Struct. 43 (3) (2010) 357–365. [13] K.M. Ng, C.M. Tam, V.W.Y. Tam, Studying the production process and mechanical properties of reactive powder concrete: a Hong Kong study, Mag. Concr. Res. 62 (9) (2010) 647–654. [14] C.M. Tam, V.W.Y. Tam, K.M. Ng, Optimal conditions for producing reactive powder concrete, Mag. Concr. Res. 62 (10) (2010) 701–716. [15] C.M. Tam, V.W.Y. Tam, K.M. Ng, Assessing drying shrinkage and water permeability of reactive powder concrete produced in Hong Kong, Constr. Build. Mater. 26 (1) (2012) 79–89. [16] T. Vandevoort, M. Suleiman, S. Sritharan, Design and Performance Verification of Ultra-High Performance Concrete Piles for Deep Foundations, Report No. IHRB Project TR-558, Center for Transportation Research and Education, Iowa State University, USA, 2008. [17] G.P. Vagelis, Experimental investigation and theoretical modeling of silica fume activity in concrete, Cem. Concr. Res. 29 (1) (1999) 79–86. [18] M. Skazlic´, D. Bjegovic´, Toughness testing of ultra-high performance fibre reinforced concrete, Mater. Struct. 42 (2009) 1025–1038. [19] B.A. Graybeal, Characterization of the Behavior of Ultra-High Performance Concrete (Ph.D. thesis), Faculty of the Graduate School of the University of Maryland, 2005. [20] DUCTALÒ, Ultra-High Performance Concrete: A State-of-the-Art Report for the Bridge Community, Chapter 2: Materials and Production, Federal Highway Administration, FHWA-HRT-13-060, 2013. [21] K. Sobolev, The development of a new method for the proportioning of highperformance concrete mixtures, Cem. Concr. Compos. 26 (7) (2004) 901–907. [22] American Society for Testing and Materials, Standard Specification for Portland Cement, ASTM C150, ASTM International, West Conshohocken, PA, USA, 2012. [23] American Society for Testing and Materials, Standard Specification for Chemical Admixtures for Concrete, ASTM C494, ASTM International, West Conshohocken, PA, USA, 2013. [24] American Society for Testing and Materials, Standard Test Method for Flow of Hydraulic Cement Mortar, ASTM C1437, ASTM International, West Conshohocken, PA, USA, 2013. [25] American Society for Testing and Materials, Standard Test Method for Compressive Strength of Hydraulic Cement Mortars (Using 2-in. or 50-mm Cube Specimens), ASTM C109, ASTM International, West Conshohocken, PA, USA, 2013. [26] American Society for Testing and Materials, Standard Test Method for Flexural Strength of Concrete (Using Simple Beam With Third-Point Loading), ASTM C78, ASTM International, West Conshohocken, PA, USA, 2010. [27] American Society for Testing and Materials, Standard Test Method for Static Modulus of Elasticity and Poisson’s Ratio of Concrete in Compression, ASTM C469, ASTM International, West Conshohocken, PA, USA, 2010. [28] M. Kakizaki, H. Edahiro, T. Tochigi, T. Niki, Effect of Mixing Method on Mechanical Properties and Pore Structure of Ultra-High Strength Concrete, ACI SP-132.1992, pp. 997–1016.