- Email: [email protected]
Experimental investigations on the influence of paste composition and content on the properties of Self-Compacting Concrete
Construction and Building Materials 23 (2009) 3443–3449
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Experimental investigations on the influence of paste composition and content on the properties of Self-Compacting Concrete Prakash Nanthagopalan, Manu Santhanam * Department of Civil Engineering, Indian Institute of Technology Madras, India
a r t i c l e
i n f o
Article history: Received 1 February 2009 Received in revised form 24 April 2009 Accepted 18 June 2009 Available online 22 July 2009 Keywords: Self-Compacting Concrete Paste optimisation Packing density VMA
a b s t r a c t The aim of this paper is to investigate the influence of paste composition and paste volume on the fresh and hardened properties of Self-Compacting Concrete. Nineteen SCC mixtures were investigated for different paste composition and paste volume. Fresh concrete tests such as slump flow, J ring, and V funnel test were performed; hardened concrete tests were limited to compressive strength. The results revealed that slump flow and J ring flow increased with increase in paste volume. A simple empirical equation was proposed for the determination of the paste volume for the required slump flow of SCC. Compressive strength of the different SCC mixtures ranged between 20 MPa and 70 MPa. Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction Self-Compacting Concrete (SCC) is a highly flowable concrete which does not segregate and can spread into place, fill the formwork with heavily congested reinforcement, and encapsulate the reinforcement without any mechanical vibration [1]. Further, it holds good in areas of difficult accessibility and complicated structural forms while maintaining the homogeneity of the mixture [2]. As flow is a distinct characteristic of SCC, understanding the behaviour of SCC in the fresh state is essential. The flow behaviour of SCC is governed by the paste composition, paste volume, and particle size distribution of aggregates [3]. The factors governing the behaviour of SCC are determined by appropriate mixture design methods. Many mixture design methods are available for SCC. The first mixture design for SCC was developed by Okamura in Japan [4]. In this method, the coarse and fine aggregate contents are kept constant so that self-compactibility can be achieved easily by adjusting the water/powder ratio and superplasticiser dosage only. Petersson et al. [5] developed a mixture design method for SCC based on the blocking criterion, void content and paste volume along with rheological studies. A mixture design procedure for SCC was developed by Sedran et al. [6] based on characterisation of the rheological behaviour of concrete by using a new type of rheometer and adopting a numerical model called compressible packing model for determining a compact aggregate system with minimum voids, taking into account wall effect and the degree of confine-
* Corresponding author. Tel.: +91 44 22574283. E-mail address: [email protected] (M. Santhanam). 0950-0618/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.conbuildmat.2009.06.029
ment exerted by the form and the reinforcement to the mixture. Many other mixture design methods are also available [3,7–9]. Additionally, mixture design guidelines have also been formulated for the production of SCC [10,11]. Most of the above mentioned methods aimed in maximising the packing density of the aggregates and minimising the paste volume for enhanced performance of SCC. The concept of particle packing has always been a key element in the mixture proportioning of concrete. Research has shown that the packing density has significant influence on the fresh and hardened properties of concrete [12]. Experiments have shown that the packing density of concrete mixtures and the flow properties of the corresponding fresh concrete are related. The optimal flow properties are obtained for mixture compositions close to the maximum packing density [13]. In 1907, Fuller and Thompson [14] experimentally investigated the importance of the size distribution of the aggregates and the properties of the concrete on the basis of packing of constituent materials. Later, research studies were devoted for developing models for proportioning particles to attain densest packing by using the concept of particle packing. A review of particle packing theories can be found elsewhere [15]. The generally agreed theory is that the paste which is in excess after completely filling the voids of the aggregate will govern the workability of concrete. While particle packing has a significant influence on the properties of concrete, which contains different sizes of particulate inclusions, the paste properties are also affected by the interaction between the cementitious particles. It has been shown that the improvement in the packing density of the cementitious materials by blending cement with fine materials plays a major role in enhancement of the properties of the mortar produced [16].
3444
P. Nanthagopalan, M. Santhanam / Construction and Building Materials 23 (2009) 3443–3449
The essential characteristics for a successful SCC are filling ability, passing ability and segregation resistance. Many different test methods have been developed to characterise the properties of SCC. So far no single method or combination of methods has achieved universal approval. Similarly no single method has been found which characterises all the relevant workability aspects; so, each mix design should be tested by more than one test method in order to obtain different workability parameters [10]. For site quality control, two test methods are generally sufficient to monitor production quality of SCC. A combination of slump flow and Jring test is sufficient for the filling ability and passing ability of SCC respectively. In the present study, a systematic approach was followed for the optimisation of the paste and aggregate phases. The particle packing concept was used for the optimisation of powder and the aggregates separately. For the optimised combination of aggregates, the paste volume was varied along with the paste composition to investigate its effects on the fresh and hardened concrete properties of SCC. 2. Materials In the present investigation, Ordinary Portland Cement – 53 grade [17] and fly ash (Class F) from North Chennai Power Station, India, were used. A third generation Polycarboxylic ether based superplasticiser (SP) with an active solids content of 33%
100
Percentage Passing (%)
Cement Fly ash
80
60
40
20
0
1E-4
1E-3
0.01
0.1
1
Sieve size (mm)
Table 2 Physical properties of cement and fly ash. Property
Cement
Fly ash
Specific gravity Mean diameter (lm) Specific surface area (Blaine’s method) (m2/kg)
3.15 16.74 355
2.00 38.16 151
Table 3 Physical properties of aggregates. Properties
River sand
12.5 mm
20 mm
Specific gravity Bulk density(kg/m3) Void content (%) Water absorption (%)
2.57 1677 34.95 0.50
2.80 1528 45.43 0.24
2.78 1644 41.41 0.20
Table 4 Particle size distribution of aggregates. Sieve size (mm)
River sand (% passing)
12.5 mm (% passing)
20 mm (% passing)
25 20 16 12.5 10 6.3 4.75 2.36 1.18 0.6 0.3 0.15 0.075
100.00 100.00 100.00 100.00 100.00 100.00 99.00 95.90 78.40 39.50 17.70 3.20 0.70
100.00 100.00 99.82 88.81 37.69 1.89 0.74 0.00 0.00 0.00 0.00 0.00 0.00
100.00 94.89 57.57 13.37 3.41 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
was used. A microbial polysaccharide was used as viscosity modifying agent (VMA). The particle size distributions of the cement and fly ash were determined by using a laser diffraction particle size analyser and the results are shown in Fig. 1. The chemical composition and the physical properties of the cement and fly ash are given in Tables 1 and 2. Two different size of coarse aggregates (Crushed Granite – maximum sizes 12.5 mm and 20 mm) and fine aggregate (River sand) were used for the investigation. The physical properties and the particle size distribution of the aggregates are given in Tables 3 and 4 respectively. Potable water at a temperature of 28 ± 1 °C was used. Different water to powder (w/p) ratios 0.8, 0.9, 1.0, 1.1, and 1.2 by volume) were chosen for the study. This range of w/p ratios was selected in order to be able to produce SCC.
Fig. 1. Particle size distribution of cement and fly ash.
3. Experimental investigations Table 1 Chemical composition of cement and fly ash.
3.1. Optimisation of paste composition
Chemical composition
Cement (% by mass)
Fly ash-class F (% by mass)
CaO SiO2 Al2O3 Fe2O3 MgO SO3 Loss of ignition Total chloride content Na2O K2O Insoluble residue
61.18 20.01 4.98 4.88 1.78 2.36 2.18 0.03 0.20 0.60 1.23
1.41 60.56 32.67 4.44 0.23 0.02 0.21 0.01 0.02 0.03 0.46
Bogue compound composition of cement Compound C3S C2S C3A C4AF
% by mass 49.82 19.78 4.94 14.84
The powder (cement:fly ash) combination was selected based on the particle packing concept by using Puntke test [18]. The principle and the procedure of the Puntke test was explained in detail elsewhere [19]. The combination of 60:40 (cement:fly ash) by volume, resulting in maximum packing density, was selected for further investigations [20]. With optimum combination of powder, the superplasticiser dosage was optimised by using mini-slump cone test [21] for different w/p ratio. The dosage corresponding to 170 ± 10 mm without bleeding in cementitious paste was identified as the optimum SP dosage [20]. The optimum dosage of SP for different w/p ratio was validated in SCC keeping the paste volume constant (388 l/m3). The results are shown in Table 5. The optimum dosage of SP for different w/p in SCC resulted in a slump flow of 550–600 mm. VMA was also used in the investigations, only for the segregated SCC mixtures (though the w/p ratio was nominal, the segregation resulted due to high water content –
3445
P. Nanthagopalan, M. Santhanam / Construction and Building Materials 23 (2009) 3443–3449 Table 5 Paste and concrete properties of SCC at optimum SP dosage. w/p Ratio
0.8 0.9 1.0 1.1 1.2 a
Cementitious paste
Concrete properties (paste volume – 388 l)
Optimum dosage (% bwopa)
Mini-slump spread (mm)
Slump flow (mm)
T500 (s)
0.21 0.18 0.16 0.14 0.12
178 170 169 166 174
555 600 600 590 580
6.0 5.0 4.0 1.9 1.0
bwop – by weight of powder.
packing density of different combination of aggregates (fine aggregate, coarse aggregates 12.5 mm maximum size and 20 mm maximum size). Initially, the aggregates were mixed in a plastic tray for homogenisation of each aggregate combination. Then, the aggregates were filled in a cylindrical container of known volume. The container diameter (250 mm) was more than 10 times the diameter of the maximum size of aggregates used (20 mm) to eliminate the wall effect. Knowing the mass of the individual aggregate type added and the volume of the container, the void content was calculated. The packing density of the aggregates was calculated from the void content. The equations for calculating the void content and packing density are as follows:
Void content ¼ ðV c ððM1 =S1 Þ þ ðM2 =S2 Þ þ ðM 3 =S3 ÞÞÞ=V c Table 6 Paste and concrete properties of SCC with VMA. w/p Ratio
0.9 1.0 1.1 1.2 a
Cementitious paste
ð1Þ
where Vc is the volume of the container, M1, M2, M3 are mass of each aggregate type, and S1, S2, S3 are the specific gravity of corresponding aggregate type.
Concrete properties
Opt. SP dosage (% bwopa)
VMA dosage (% bwopa)
Water content (l/m3)
Sieve segregation (%)
Slump flow (mm)
0.18 0.16 0.14 0.12
0.007 0.011 0.014 0.021
217 242 225 245
19.97 19.80 19.87 19.68
790 740 690 680
Packing density ¼ 1 void content
ð2Þ
The aggregate combination of 50:20:30 (fine aggregate:coarse aggregate 12.5 mm maximum size:Coarse aggregate 20 mm maximum size) by volume resulted in maximum packing density (0.68), and was used in all the experiments. This indicates a void content of 0.32 (or 320 l) of the total volume of concrete.
bwop – by weight of powder.
4. Results and discussion more than 215 l). Appropriate dosages of VMA (reported in [22]) were added to the segregated mixes for different w/p ratio. These dosages were selected based on segregation percentage (less than 20% which corresponds to class SR1 of EFNARC guidelines) of sieve segregation test [10]. The results are shown in Table 6. VMA dosages lower than the selected dosages were not sufficient enough to control the segregation in concrete. On the other hand, higher VMA dosages resulted in reduced slump flow of concrete. In this manner, the paste composition for different w/p ratio was selected for the investigations. 3.2. Optimisation of aggregate combination The aggregate combination was selected based on the particle packing concept. Experiments were conducted to determine the
In the present investigation, experiments were carried out using different powder contents (350–650 kg/m3) and w/p ratios (0.7 to 1.7 by volume) with corresponding variation in the paste volume to investigate the influence on the fresh and hardened properties of SCC. Table 7 depicts the results of fresh concrete tests such as slump flow, T500, J ring, and V funnel. The data is presented in the order of increasing slump flow in Table 7. 4.1. Fresh concrete properties of SCC The influence of paste content and composition is depicted in the plots in Figs. 2–4. From Fig. 2, it is observed that for a given powder content, with increase in w/p ratio, the slump flow increased and for a given w/p ratio, the slump flow increased with
Table 7 Fresh concrete properties of SCC. Slump flow range (mm)
w/p
Powder (kg/m3)
Water (l/m3)
Paste (l/m3)
550–650
0.80 1.70 1.20 0.70 1.00 1.30
550 350 450 600 500 450
164 221 201 156 186 218
388 371 388 399 392 405
650–750
0.90 0.70 1.00 1.10
550 650 525 500
184 169 195 204
750–850
1.00 1.10 0.75 0.90 0.80
550 525 650 600 650
1.10 1.20 0.90 1.00
With VMA 650–750 750–850
Slump flow (mm) A
T500 (s)
V funnel time (s)
J ring flow (mm) B
Blocking assessment, A–B (mm)
68 51 68 79 72 85
555 560 580 610 620 625
6.07 1.85 1.31 35.80 1.28 1.03
135.84 2.07 15.34 blocked 31.16 5.25
510 510 530 570 580 580
45.00 50.00 50.00 40.00 40.00 45.00
408 431 410 410
88 111 90 90
700 715 720 730
1.72 8.28 1.60 1.50
60.00 41.53 15.60 10.00
670 670 680 700
30.00 45.00 40.00 30.00
205 215 181 201 193
429 430 443 444 455
109 110 123 124 135
800 810 820 830 850
1.12 1.50 2.62 2.20 1.43
4.07 9.50 30.12 12.50 55.00
790 790 755 810 840
10.00 20.00 50.00 20.00 10.00
550 550
225 245
449 470
129 150
690 680
1.56 0.44
10.6 6.44
670 660
20.00 20.00
650 650
217 242
479 503
159 183
790 820
2.09 0.63
26.91 11.13
780 810
10.00 10.00
Excess paste volume (l)
3446
P. Nanthagopalan, M. Santhanam / Construction and Building Materials 23 (2009) 3443–3449
1000 350 powder content 450 powder content 550 powder content 650 powder content
Slump flow (mm)
900 800 700 600 500 400
0.6
0.8
1.0
1.2
1.4
1.6
1.8
w/p ratio (by volume) Fig. 2. Relationship between the slump flow and w/p ratio (by volume).
Without VMA With VMA
850 y 2= 4.068x - 980.58 R = 0.8948
Slump flow (mm)
800 750 700 650
brought out in Fig. 3, which shows that the slump flow increases with increase in paste volume and approximately 370–380 l of paste is essential for achieving slump flow of 550 mm, with the given combination of aggregates having a packing density of 0.68. In this study, the paste volume was varied from 371 l/m3 to 503 l/m3 (which includes 20 l of air), and the resultant slump flow was in the range 550 mm to 850 mm. A plot was made between the water content and the slump flow (see Fig. 4), which indicates that there is no significant correlation between them. From this, it is evident that the paste volume has a predominant influence on the slump flow of SCC than powder or water content individually (for a given combination of aggregates). Mixtures with slump flow more than 850 mm resulted in segregation. In the segregated mixtures, appropriate dosages of VMA were added and fresh concrete tests were conducted. The results are included in Table 7. From the data in Table 7 and the plot in Fig. 3, it may be noted that higher paste contents are required to produce an equivalent slump flow for the mixtures with VMA. This is expected, since the VMA would increase the yield stress and plastic viscosity of the paste. The results were further analysed to estimate the required amount of paste volume in excess of void content to achieve SCC. Fig. 5 shows a plot between the excess paste volume and the slump flow. Fig. 5 includes the experimental data points as well as data from literature [23–25]. From Fig. 5, it is evident that a minimum of 50–70 l of excess paste volume, over and above the paste volume corresponding to the void content, is required for achieving the minimum slump flow of 550 mm. An empirical relation was proposed based on the regression analysis of the data. The regression analysis of the data does not include the results of the SCC mixtures with VMA.
Paste volume required ¼ Void volume þ ðTarget slump flow 321Þ=4:068 ð3Þ
600 550
360
380
400
420
440
460
480
500
Paste volume (litres/m3) Fig. 3. Relationship between the slump flow and paste volume.
850
Without VMA With VMA
This empirical relationship can be used for the determination of paste volume for required slump flow. Though this equation is empirical, this will give the user a starting point, instead of doing many trials. Apart from slump flow test, experiments were conducted to determine the T500, V funnel time, and J ring flow. The T500 time is the time required to reach 500 mm slump flow. It indirectly indicates the viscosity of the concrete – higher the time to reach 500 mm, higher the viscosity. As per the results in Table 7, T500 was mostly within 2 s. In one case, an exceptionally high time
750
Data of author Literature
850
700
800 y = 4.068x + 321.17 2
650 600 550
160
180
200
220
240
Water Content (litres/m3) Fig. 4. Relationship between water content and slump flow.
Slump flow (mm)
Slump flow (mm)
800
R = 0.8948
750 700 650 600 550
40 increase in powder content. The reason for the above observations could be attributed to the fact that the paste volume increased when the w/p ratio or the powder content increased. This is clearly
60
80
100
120
140 3
Excess Paste Volume (litres/m ) Fig. 5. Relationship between excess paste volume and slump flow.
3447
P. Nanthagopalan, M. Santhanam / Construction and Building Materials 23 (2009) 3443–3449
Without VMA With VMA
140
850
Without VMA With VMA
y = 4.3812x - 1146.1 2 R = 0.8687
800
J ring flow (mm)
(35.8 s) with slump flow of 610 mm was observed, which could be attributed to the fact that the paste had very low water content (156 l) with high powder content (600 kg/m3) that caused the SCC to be highly viscous. The V funnel flow time also indirectly indicates the viscosity of SCC. A plot between the paste volume and the V funnel time (see Fig. 6) indicates that the effect of paste volume on the V funnel time is not clear. The reason for inconsistencies in the V funnel test results could be attributed to the local arching effect of the coarse aggregates nearer to the bottom opening of the V funnel. Besides that, rheologically speaking, the paste volume has a predominant effect on the yield stress of SCC in comparison with the viscosity. On the other hand, the viscosity of the SCC is controlled by viscosity of the paste, which in turn depends on the water content. Hence, a plot was made between the water content and the V funnel time (see Fig. 7). From Fig. 7, it can be seen that with increase in water content, the V funnel time decreased indicating a decrease in the viscosity of the SCC. For assessing the passing ability of SCC, J-ring test was conducted for all the mixtures according to [26]. From Fig. 8, it was observed that the J ring flow (slump flow with J ring) increased with increase in paste volume. This could be attributed to the fact that with increase in paste volume, the aggregates are dispersed efficiently and hence the concrete passes through the ring without
750 700 650 600 550 500
360
380
400
420
440
460
480
500
3
Paste Volume (litres/m ) Fig. 8. Relationship between paste volume and the J ring flow.
congestion of the aggregates. The results of J-ring test are shown in Table 7. The blocking assessment was calculated as the difference between the slump flow and J ring flow. The results indicated that the difference in flow varied between 10 mm and 50 mm – corresponding to no visible blocking to minimal blocking respectively. These results include SCC mixtures without VMA and SCC mixtures with VMA.
120
V-Funnel time (s)
4.2. Compressive strength
100 80 60 40 20 0
360
380
400
420
440
460
480
500
Paste Volume (litres/m 3) Fig. 6. Variation of V funnel time with paste volume.
140
80
Without VMA With VMA
Compressive strength (MPa)
120
V funnel time (s)
For SCC, achieving high strengths is not difficult, due to the presence of high powder content. However, achieving low and medium strength SCC is a difficult task. Therefore, in this investigation, the main focus was to achieve low and medium strength SCC. Studies were conducted to investigate the influence of w/p ratio on the compressive strength of SCC. The average of the compressive strength of three cubes (150 mm 150 mm 150 mm) at 28 days, plotted against the w/p ratio in Fig. 9 indicates a good correlation between the compressive strength and w/p ratio. The compressive strength range varied from 20 MPa to 70 MPa. From this graph, it is possible to determine the w/p ratio for a specific strength. Though this graph is material specific (for SCC produced with 60:40 combination of cement:fly ash by volume), this will reduce the number of experimental trials to arrive at the required strength.
100 80 60 40 20 0
160
180
200
220
240
Water content (litres/m3) Fig. 7. Relationship between the V funnel time and water content.
70 60 50 40 30 20
0.6
0.8
1.0
1.2
1.4
1.6
1.8
w/p ratio Fig. 9. Relationship between the compressive strength and the w/p ratio.
3448
P. Nanthagopalan, M. Santhanam / Construction and Building Materials 23 (2009) 3443–3449
Table 8 Matrix of strengths and slump flows obtained in the study. Range
Compressive strength (MPa) 20–30
30–40
40–50
50–60
60–75
33.2–1.2 (450, 201, 388) 32.6–1.3 (450, 218, 405)
44.0–1.0 (500, 186, 392)
55.1–0.8 (550, 164, 388)
70.0–0.7 (600, 156, 399)
650–750
38.0–1.1 (500, 204, 410) 35.7–1.2 (550, 245, 470)
45.0–1.0 (525, 195, 410) 41.2–1.1 (550, 225, 449)
52.2–0.9 (550, 184, 408)
72.6–0.7 (650, 169, 431)
750–850
39.0–1.1 (525, 215, 430)
45.9–1.0 (550, 205, 429) 48.0–1.0 (650, 242, 503)
56.0–0.9 (600, 201, 444) 58.0–0.9 (650, 217, 479)
60.4–0.8 (650, 193, 455) 66.8–0.75 (650, 181, 443)
Slump flow (mm) 550–650 26.0–1.7 (350, 221, 371)
The results of slump flow and compressive strength obtained for all mixtures in the study are collectively presented in Table 8. These results are classified according to the range of slump flow and compressive strength obtained. The numbers presented in each cell are (in order): the specific compressive strength, water to powder ratio (by volume), powder content (kg/m3), water content (kg/m3), and paste volume (litres). The shaded cells represent mixtures with VMA. The results indicated that it was possible to achieve SCC with a slump flow of 560 mm for powder content as low as 350 kg/m3. From Table 8, it is possible to select the paste composition for a required combination of slump flow and compressive strength. 5. General discussions Unlike conventional concrete, for SCC numerous devices such as U box, L box, and V funnel have been proposed in the literature to test the fresh concrete properties [27]. However, they are either cumbersome to perform (large samples are required and there is very limited practical application) or more research is required to make them representative of the real confinement of structures in the case of very dense reinforcements [6]. It is preferable to use minimum number of fresh tests that addresses essential fresh properties instead of performing many tests. The intention of the above argument should not be viewed as criticism of the other available test methods, as better understanding of the behaviour of fresh SCC is possible by using such test methods. Instead, the authors’ view is that these test methods are still being perfected and cannot expect to yield predictions of desired accuracy. In the current study, slump flow and J ring were found to be sufficient to characterise the fresh concrete behaviour of SCC, since enough care was exercised to avoid segregation. In general, the fresh concrete characterisation of SCC should include a segregation test in addition to the slump flow and J-ring tests. The SCC mixtures that segregated either had water content more than 215 l or fine aggregate fraction less than 40% of the total aggregate. With regard to the present study, the segregation phenomenon is reduced by maximising the packing density; the choice of high packing density increases the range of aggregate particles, which reduces segregation by the lattice effect (grid of stable small and mid-size particles prevent the sinking of big particles). By using appropriate dosage of VMA, it was possible to obtain SCC without segregation and without compromising on the slump flow. Hence, it is suggested that VMA can be used to have a firm control over the stability and robustness of the mix as well as to achieve SCC at the same powder content as normal concrete. As a result, the potential for creep and shrinkage of SCC is also reduced. 6. Conclusions The influence of the paste composition and paste volume on the fresh and hardened concrete properties of SCC was discussed.
Experimental results revealed that paste volume had a predominant effect on the fresh concrete properties in comparison with water or powder content individually (for a given combination of aggregates). A minimum of 50–70 l of excess paste over and above the void content of the aggregates was found essential for achieving SCC with a slump flow of 550 mm. An empirical relationship was established for the determination of the paste volume for the required slump flow. The effect of paste volume on the V funnel flow time was not clear due to the blocking of aggregates nearer to the V funnel opening. The passing ability was studied by using the J-ring test. The results revealed that the difference between slump flow and the J ring flow varied between 10 and 50 mm. From the results, it is apparent that for understanding the fresh concrete behaviour of SCC, it is sufficient to conduct slump flow (flow ability) and J-ring test (passing ability). As expected, the w/p ratio had a good correlation with the compressive strength of SCC for a given cement/fly ash ratio. The compressive strength of mixes developed in this study ranged from 20 MPa to 70 MPa. It was possible to achieve SCC with powder content as low as 350 kg/m3. A matrix combination of strength and the slump flow was established based on the experimental results from which one can use the paste composition and paste content for the desired combination of slump flow and compressive strength.
References [1] ACI 237 R. Self-Consolidating Concrete. American Concrete Institute; 2007. p. 30. [2] Okamura H, Ouchi M. Self-Compacting Concrete. Development, present use and future. Proceedings of first international symposium on Self-Compacting Concrete. Stockholm; 1999. p. 3–14. [3] Saak AW, Jennings HM, Shah SP. New methodology for designing SelfCompacting Concrete. ACI Mater J 2001;98(6):429–39. [4] Okamura H, Maekawa K, Ozawa K. High performance concrete. Tokoyo: Gihodo Publishing; 1993 [in Japanese]. [5] Petersson O, Billberg P, Van BK. A model for self-compacting concrete. In: Proceedings of production methods and workability of concrete, London; 1996. p. 483–92. [6] Sedran T, de Larrard F, Hourst F, Contamines C. Mix design of self-compacting concrete. In: Proceedings of production methods and workability of concrete, London; 1996. p. 439–51. [7] Gomes PCC, Gettu R, Agullo L, Bernad C. Experimental optimisation of high strength Self-Compacting Concrete. In: Proceedings of second International RILEM symposium on self-compacting concrete, Japan; 2001. p. 377–86. [8] Su N, Hsu KC, Chai HW. A simple mix design method for self-compacting concrete. Cem Concr Res 2001;31(12):1799–807. [9] Smeplass S, Mortsell E. The particle matrix model applied on SCC. In: Proceedings of the second international RILEM symposium on selfcompacting concrete, Japan; 2001. p. 267–76. [10] EFNARC. The European guidelines for self-compacting concrete: specification, production and use; 2005. p. 68. [11] JSCE. Recommendation for construction of self-compacting concrete. In: International workshop on self-compacting concrete. Japan: Japanese Society of Civil Engineers; 1998. p. 77. [12] Glavind M, Pedersen EJ. Packing calculations applied for concrete mix design. In: Proceedings of creating with concrete. Dundee;1999. p. 1–10. [13] Johansen V, Andersen PJ. Particle packing and concrete properties. Material Science of Concrete II. Ohio: The American Ceramic Society; 1991. p. 111–47. [14] Fuller WB, Thompson SE. The laws of proportioning concrete. Transactions, American Society of Civil Engineers; 1907. p. 67–172.
P. Nanthagopalan, M. Santhanam / Construction and Building Materials 23 (2009) 3443–3449 [15] Senthilkumar V, Santhanam M. Particle packing theories and their application in concrete mixture proportioning: a review. The Indian Concrete Journal 2003:1324–31. [16] Lange F, Mortel H, Rudert V. Dense packing of cement pastes and resulting consequences on mortar properties. Cem Concr Res 1997;27(10):1481–8. [17] IS 12269. Specification for 53 grade ordinary portland cement. Indian Standards; 2004. [18] Puntke W. Wasseranspruch von feinen Kornhaufwerken (in German). Beton 2002;52(5):242–8. [19] Nanthagopalan P, Santhanam M. Investigation on the influence of granular packing on the flow properties of cementitious suspensions. Cem Concr Compos 2008;30:763–8. [20] Nanthagopalan P, Santhanam M. A new approach to optimisation of paste composition in self-compacting concrete. Ind Concr J 2008;82:11–8. [21] Kantro DL. Influence of water-reducing admixtures on properties of cement paste – a miniature slump test. Cem Concr Agg 1980;2(2):95–102.
3449
[22] Nanthagopalan P, Santhanam M. A new empirical test method for the optimisation of viscosity modifying agent dosage in Self-Compacting Concrete. Mater Struct (in press). [23] Rajayogan V, Santhanam M, Sharma BS. Evaluation of hydroxyl propyl starch as a viscosity-modifying agent for Self-Compacting Concrete. In: Proceedings of the third international RILEM symposium on Self-Compacting Concrete, Iceland; 2003. p. 386–94. [24] Kumar P, Haq MA, Kaushik SK. Early age strength of SCC with large volumes of fly ash. Ind Concr J 2004;78:25–9. [25] Bui VK, Montgomery D, Hinczak I, Turner K. Rapid testing method for segregation resistance of self-compacting concrete. Cem Concr Res 2002;32(9):1489–96. [26] ASTM C 1621. Standard test method for passing ability of self-compacting concrete by J ring. ASTM standards; 2006. [27] Okamura H, Ouchi M. Self-compacting concrete. J Adv Concr Technol 2003;1(1):5–15.
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Experimental investigations on the influence of paste composition and content on the properties of Self-Compacting Concrete Prakash Nanthagopalan, Manu Santhanam * Department of Civil Engineering, Indian Institute of Technology Madras, India
a r t i c l e
i n f o
Article history: Received 1 February 2009 Received in revised form 24 April 2009 Accepted 18 June 2009 Available online 22 July 2009 Keywords: Self-Compacting Concrete Paste optimisation Packing density VMA
a b s t r a c t The aim of this paper is to investigate the influence of paste composition and paste volume on the fresh and hardened properties of Self-Compacting Concrete. Nineteen SCC mixtures were investigated for different paste composition and paste volume. Fresh concrete tests such as slump flow, J ring, and V funnel test were performed; hardened concrete tests were limited to compressive strength. The results revealed that slump flow and J ring flow increased with increase in paste volume. A simple empirical equation was proposed for the determination of the paste volume for the required slump flow of SCC. Compressive strength of the different SCC mixtures ranged between 20 MPa and 70 MPa. Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction Self-Compacting Concrete (SCC) is a highly flowable concrete which does not segregate and can spread into place, fill the formwork with heavily congested reinforcement, and encapsulate the reinforcement without any mechanical vibration [1]. Further, it holds good in areas of difficult accessibility and complicated structural forms while maintaining the homogeneity of the mixture [2]. As flow is a distinct characteristic of SCC, understanding the behaviour of SCC in the fresh state is essential. The flow behaviour of SCC is governed by the paste composition, paste volume, and particle size distribution of aggregates [3]. The factors governing the behaviour of SCC are determined by appropriate mixture design methods. Many mixture design methods are available for SCC. The first mixture design for SCC was developed by Okamura in Japan [4]. In this method, the coarse and fine aggregate contents are kept constant so that self-compactibility can be achieved easily by adjusting the water/powder ratio and superplasticiser dosage only. Petersson et al. [5] developed a mixture design method for SCC based on the blocking criterion, void content and paste volume along with rheological studies. A mixture design procedure for SCC was developed by Sedran et al. [6] based on characterisation of the rheological behaviour of concrete by using a new type of rheometer and adopting a numerical model called compressible packing model for determining a compact aggregate system with minimum voids, taking into account wall effect and the degree of confine-
* Corresponding author. Tel.: +91 44 22574283. E-mail address: [email protected] (M. Santhanam). 0950-0618/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.conbuildmat.2009.06.029
ment exerted by the form and the reinforcement to the mixture. Many other mixture design methods are also available [3,7–9]. Additionally, mixture design guidelines have also been formulated for the production of SCC [10,11]. Most of the above mentioned methods aimed in maximising the packing density of the aggregates and minimising the paste volume for enhanced performance of SCC. The concept of particle packing has always been a key element in the mixture proportioning of concrete. Research has shown that the packing density has significant influence on the fresh and hardened properties of concrete [12]. Experiments have shown that the packing density of concrete mixtures and the flow properties of the corresponding fresh concrete are related. The optimal flow properties are obtained for mixture compositions close to the maximum packing density [13]. In 1907, Fuller and Thompson [14] experimentally investigated the importance of the size distribution of the aggregates and the properties of the concrete on the basis of packing of constituent materials. Later, research studies were devoted for developing models for proportioning particles to attain densest packing by using the concept of particle packing. A review of particle packing theories can be found elsewhere [15]. The generally agreed theory is that the paste which is in excess after completely filling the voids of the aggregate will govern the workability of concrete. While particle packing has a significant influence on the properties of concrete, which contains different sizes of particulate inclusions, the paste properties are also affected by the interaction between the cementitious particles. It has been shown that the improvement in the packing density of the cementitious materials by blending cement with fine materials plays a major role in enhancement of the properties of the mortar produced [16].
3444
P. Nanthagopalan, M. Santhanam / Construction and Building Materials 23 (2009) 3443–3449
The essential characteristics for a successful SCC are filling ability, passing ability and segregation resistance. Many different test methods have been developed to characterise the properties of SCC. So far no single method or combination of methods has achieved universal approval. Similarly no single method has been found which characterises all the relevant workability aspects; so, each mix design should be tested by more than one test method in order to obtain different workability parameters [10]. For site quality control, two test methods are generally sufficient to monitor production quality of SCC. A combination of slump flow and Jring test is sufficient for the filling ability and passing ability of SCC respectively. In the present study, a systematic approach was followed for the optimisation of the paste and aggregate phases. The particle packing concept was used for the optimisation of powder and the aggregates separately. For the optimised combination of aggregates, the paste volume was varied along with the paste composition to investigate its effects on the fresh and hardened concrete properties of SCC. 2. Materials In the present investigation, Ordinary Portland Cement – 53 grade [17] and fly ash (Class F) from North Chennai Power Station, India, were used. A third generation Polycarboxylic ether based superplasticiser (SP) with an active solids content of 33%
100
Percentage Passing (%)
Cement Fly ash
80
60
40
20
0
1E-4
1E-3
0.01
0.1
1
Sieve size (mm)
Table 2 Physical properties of cement and fly ash. Property
Cement
Fly ash
Specific gravity Mean diameter (lm) Specific surface area (Blaine’s method) (m2/kg)
3.15 16.74 355
2.00 38.16 151
Table 3 Physical properties of aggregates. Properties
River sand
12.5 mm
20 mm
Specific gravity Bulk density(kg/m3) Void content (%) Water absorption (%)
2.57 1677 34.95 0.50
2.80 1528 45.43 0.24
2.78 1644 41.41 0.20
Table 4 Particle size distribution of aggregates. Sieve size (mm)
River sand (% passing)
12.5 mm (% passing)
20 mm (% passing)
25 20 16 12.5 10 6.3 4.75 2.36 1.18 0.6 0.3 0.15 0.075
100.00 100.00 100.00 100.00 100.00 100.00 99.00 95.90 78.40 39.50 17.70 3.20 0.70
100.00 100.00 99.82 88.81 37.69 1.89 0.74 0.00 0.00 0.00 0.00 0.00 0.00
100.00 94.89 57.57 13.37 3.41 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
was used. A microbial polysaccharide was used as viscosity modifying agent (VMA). The particle size distributions of the cement and fly ash were determined by using a laser diffraction particle size analyser and the results are shown in Fig. 1. The chemical composition and the physical properties of the cement and fly ash are given in Tables 1 and 2. Two different size of coarse aggregates (Crushed Granite – maximum sizes 12.5 mm and 20 mm) and fine aggregate (River sand) were used for the investigation. The physical properties and the particle size distribution of the aggregates are given in Tables 3 and 4 respectively. Potable water at a temperature of 28 ± 1 °C was used. Different water to powder (w/p) ratios 0.8, 0.9, 1.0, 1.1, and 1.2 by volume) were chosen for the study. This range of w/p ratios was selected in order to be able to produce SCC.
Fig. 1. Particle size distribution of cement and fly ash.
3. Experimental investigations Table 1 Chemical composition of cement and fly ash.
3.1. Optimisation of paste composition
Chemical composition
Cement (% by mass)
Fly ash-class F (% by mass)
CaO SiO2 Al2O3 Fe2O3 MgO SO3 Loss of ignition Total chloride content Na2O K2O Insoluble residue
61.18 20.01 4.98 4.88 1.78 2.36 2.18 0.03 0.20 0.60 1.23
1.41 60.56 32.67 4.44 0.23 0.02 0.21 0.01 0.02 0.03 0.46
Bogue compound composition of cement Compound C3S C2S C3A C4AF
% by mass 49.82 19.78 4.94 14.84
The powder (cement:fly ash) combination was selected based on the particle packing concept by using Puntke test [18]. The principle and the procedure of the Puntke test was explained in detail elsewhere [19]. The combination of 60:40 (cement:fly ash) by volume, resulting in maximum packing density, was selected for further investigations [20]. With optimum combination of powder, the superplasticiser dosage was optimised by using mini-slump cone test [21] for different w/p ratio. The dosage corresponding to 170 ± 10 mm without bleeding in cementitious paste was identified as the optimum SP dosage [20]. The optimum dosage of SP for different w/p ratio was validated in SCC keeping the paste volume constant (388 l/m3). The results are shown in Table 5. The optimum dosage of SP for different w/p in SCC resulted in a slump flow of 550–600 mm. VMA was also used in the investigations, only for the segregated SCC mixtures (though the w/p ratio was nominal, the segregation resulted due to high water content –
3445
P. Nanthagopalan, M. Santhanam / Construction and Building Materials 23 (2009) 3443–3449 Table 5 Paste and concrete properties of SCC at optimum SP dosage. w/p Ratio
0.8 0.9 1.0 1.1 1.2 a
Cementitious paste
Concrete properties (paste volume – 388 l)
Optimum dosage (% bwopa)
Mini-slump spread (mm)
Slump flow (mm)
T500 (s)
0.21 0.18 0.16 0.14 0.12
178 170 169 166 174
555 600 600 590 580
6.0 5.0 4.0 1.9 1.0
bwop – by weight of powder.
packing density of different combination of aggregates (fine aggregate, coarse aggregates 12.5 mm maximum size and 20 mm maximum size). Initially, the aggregates were mixed in a plastic tray for homogenisation of each aggregate combination. Then, the aggregates were filled in a cylindrical container of known volume. The container diameter (250 mm) was more than 10 times the diameter of the maximum size of aggregates used (20 mm) to eliminate the wall effect. Knowing the mass of the individual aggregate type added and the volume of the container, the void content was calculated. The packing density of the aggregates was calculated from the void content. The equations for calculating the void content and packing density are as follows:
Void content ¼ ðV c ððM1 =S1 Þ þ ðM2 =S2 Þ þ ðM 3 =S3 ÞÞÞ=V c Table 6 Paste and concrete properties of SCC with VMA. w/p Ratio
0.9 1.0 1.1 1.2 a
Cementitious paste
ð1Þ
where Vc is the volume of the container, M1, M2, M3 are mass of each aggregate type, and S1, S2, S3 are the specific gravity of corresponding aggregate type.
Concrete properties
Opt. SP dosage (% bwopa)
VMA dosage (% bwopa)
Water content (l/m3)
Sieve segregation (%)
Slump flow (mm)
0.18 0.16 0.14 0.12
0.007 0.011 0.014 0.021
217 242 225 245
19.97 19.80 19.87 19.68
790 740 690 680
Packing density ¼ 1 void content
ð2Þ
The aggregate combination of 50:20:30 (fine aggregate:coarse aggregate 12.5 mm maximum size:Coarse aggregate 20 mm maximum size) by volume resulted in maximum packing density (0.68), and was used in all the experiments. This indicates a void content of 0.32 (or 320 l) of the total volume of concrete.
bwop – by weight of powder.
4. Results and discussion more than 215 l). Appropriate dosages of VMA (reported in [22]) were added to the segregated mixes for different w/p ratio. These dosages were selected based on segregation percentage (less than 20% which corresponds to class SR1 of EFNARC guidelines) of sieve segregation test [10]. The results are shown in Table 6. VMA dosages lower than the selected dosages were not sufficient enough to control the segregation in concrete. On the other hand, higher VMA dosages resulted in reduced slump flow of concrete. In this manner, the paste composition for different w/p ratio was selected for the investigations. 3.2. Optimisation of aggregate combination The aggregate combination was selected based on the particle packing concept. Experiments were conducted to determine the
In the present investigation, experiments were carried out using different powder contents (350–650 kg/m3) and w/p ratios (0.7 to 1.7 by volume) with corresponding variation in the paste volume to investigate the influence on the fresh and hardened properties of SCC. Table 7 depicts the results of fresh concrete tests such as slump flow, T500, J ring, and V funnel. The data is presented in the order of increasing slump flow in Table 7. 4.1. Fresh concrete properties of SCC The influence of paste content and composition is depicted in the plots in Figs. 2–4. From Fig. 2, it is observed that for a given powder content, with increase in w/p ratio, the slump flow increased and for a given w/p ratio, the slump flow increased with
Table 7 Fresh concrete properties of SCC. Slump flow range (mm)
w/p
Powder (kg/m3)
Water (l/m3)
Paste (l/m3)
550–650
0.80 1.70 1.20 0.70 1.00 1.30
550 350 450 600 500 450
164 221 201 156 186 218
388 371 388 399 392 405
650–750
0.90 0.70 1.00 1.10
550 650 525 500
184 169 195 204
750–850
1.00 1.10 0.75 0.90 0.80
550 525 650 600 650
1.10 1.20 0.90 1.00
With VMA 650–750 750–850
Slump flow (mm) A
T500 (s)
V funnel time (s)
J ring flow (mm) B
Blocking assessment, A–B (mm)
68 51 68 79 72 85
555 560 580 610 620 625
6.07 1.85 1.31 35.80 1.28 1.03
135.84 2.07 15.34 blocked 31.16 5.25
510 510 530 570 580 580
45.00 50.00 50.00 40.00 40.00 45.00
408 431 410 410
88 111 90 90
700 715 720 730
1.72 8.28 1.60 1.50
60.00 41.53 15.60 10.00
670 670 680 700
30.00 45.00 40.00 30.00
205 215 181 201 193
429 430 443 444 455
109 110 123 124 135
800 810 820 830 850
1.12 1.50 2.62 2.20 1.43
4.07 9.50 30.12 12.50 55.00
790 790 755 810 840
10.00 20.00 50.00 20.00 10.00
550 550
225 245
449 470
129 150
690 680
1.56 0.44
10.6 6.44
670 660
20.00 20.00
650 650
217 242
479 503
159 183
790 820
2.09 0.63
26.91 11.13
780 810
10.00 10.00
Excess paste volume (l)
3446
P. Nanthagopalan, M. Santhanam / Construction and Building Materials 23 (2009) 3443–3449
1000 350 powder content 450 powder content 550 powder content 650 powder content
Slump flow (mm)
900 800 700 600 500 400
0.6
0.8
1.0
1.2
1.4
1.6
1.8
w/p ratio (by volume) Fig. 2. Relationship between the slump flow and w/p ratio (by volume).
Without VMA With VMA
850 y 2= 4.068x - 980.58 R = 0.8948
Slump flow (mm)
800 750 700 650
brought out in Fig. 3, which shows that the slump flow increases with increase in paste volume and approximately 370–380 l of paste is essential for achieving slump flow of 550 mm, with the given combination of aggregates having a packing density of 0.68. In this study, the paste volume was varied from 371 l/m3 to 503 l/m3 (which includes 20 l of air), and the resultant slump flow was in the range 550 mm to 850 mm. A plot was made between the water content and the slump flow (see Fig. 4), which indicates that there is no significant correlation between them. From this, it is evident that the paste volume has a predominant influence on the slump flow of SCC than powder or water content individually (for a given combination of aggregates). Mixtures with slump flow more than 850 mm resulted in segregation. In the segregated mixtures, appropriate dosages of VMA were added and fresh concrete tests were conducted. The results are included in Table 7. From the data in Table 7 and the plot in Fig. 3, it may be noted that higher paste contents are required to produce an equivalent slump flow for the mixtures with VMA. This is expected, since the VMA would increase the yield stress and plastic viscosity of the paste. The results were further analysed to estimate the required amount of paste volume in excess of void content to achieve SCC. Fig. 5 shows a plot between the excess paste volume and the slump flow. Fig. 5 includes the experimental data points as well as data from literature [23–25]. From Fig. 5, it is evident that a minimum of 50–70 l of excess paste volume, over and above the paste volume corresponding to the void content, is required for achieving the minimum slump flow of 550 mm. An empirical relation was proposed based on the regression analysis of the data. The regression analysis of the data does not include the results of the SCC mixtures with VMA.
Paste volume required ¼ Void volume þ ðTarget slump flow 321Þ=4:068 ð3Þ
600 550
360
380
400
420
440
460
480
500
Paste volume (litres/m3) Fig. 3. Relationship between the slump flow and paste volume.
850
Without VMA With VMA
This empirical relationship can be used for the determination of paste volume for required slump flow. Though this equation is empirical, this will give the user a starting point, instead of doing many trials. Apart from slump flow test, experiments were conducted to determine the T500, V funnel time, and J ring flow. The T500 time is the time required to reach 500 mm slump flow. It indirectly indicates the viscosity of the concrete – higher the time to reach 500 mm, higher the viscosity. As per the results in Table 7, T500 was mostly within 2 s. In one case, an exceptionally high time
750
Data of author Literature
850
700
800 y = 4.068x + 321.17 2
650 600 550
160
180
200
220
240
Water Content (litres/m3) Fig. 4. Relationship between water content and slump flow.
Slump flow (mm)
Slump flow (mm)
800
R = 0.8948
750 700 650 600 550
40 increase in powder content. The reason for the above observations could be attributed to the fact that the paste volume increased when the w/p ratio or the powder content increased. This is clearly
60
80
100
120
140 3
Excess Paste Volume (litres/m ) Fig. 5. Relationship between excess paste volume and slump flow.
3447
P. Nanthagopalan, M. Santhanam / Construction and Building Materials 23 (2009) 3443–3449
Without VMA With VMA
140
850
Without VMA With VMA
y = 4.3812x - 1146.1 2 R = 0.8687
800
J ring flow (mm)
(35.8 s) with slump flow of 610 mm was observed, which could be attributed to the fact that the paste had very low water content (156 l) with high powder content (600 kg/m3) that caused the SCC to be highly viscous. The V funnel flow time also indirectly indicates the viscosity of SCC. A plot between the paste volume and the V funnel time (see Fig. 6) indicates that the effect of paste volume on the V funnel time is not clear. The reason for inconsistencies in the V funnel test results could be attributed to the local arching effect of the coarse aggregates nearer to the bottom opening of the V funnel. Besides that, rheologically speaking, the paste volume has a predominant effect on the yield stress of SCC in comparison with the viscosity. On the other hand, the viscosity of the SCC is controlled by viscosity of the paste, which in turn depends on the water content. Hence, a plot was made between the water content and the V funnel time (see Fig. 7). From Fig. 7, it can be seen that with increase in water content, the V funnel time decreased indicating a decrease in the viscosity of the SCC. For assessing the passing ability of SCC, J-ring test was conducted for all the mixtures according to [26]. From Fig. 8, it was observed that the J ring flow (slump flow with J ring) increased with increase in paste volume. This could be attributed to the fact that with increase in paste volume, the aggregates are dispersed efficiently and hence the concrete passes through the ring without
750 700 650 600 550 500
360
380
400
420
440
460
480
500
3
Paste Volume (litres/m ) Fig. 8. Relationship between paste volume and the J ring flow.
congestion of the aggregates. The results of J-ring test are shown in Table 7. The blocking assessment was calculated as the difference between the slump flow and J ring flow. The results indicated that the difference in flow varied between 10 mm and 50 mm – corresponding to no visible blocking to minimal blocking respectively. These results include SCC mixtures without VMA and SCC mixtures with VMA.
120
V-Funnel time (s)
4.2. Compressive strength
100 80 60 40 20 0
360
380
400
420
440
460
480
500
Paste Volume (litres/m 3) Fig. 6. Variation of V funnel time with paste volume.
140
80
Without VMA With VMA
Compressive strength (MPa)
120
V funnel time (s)
For SCC, achieving high strengths is not difficult, due to the presence of high powder content. However, achieving low and medium strength SCC is a difficult task. Therefore, in this investigation, the main focus was to achieve low and medium strength SCC. Studies were conducted to investigate the influence of w/p ratio on the compressive strength of SCC. The average of the compressive strength of three cubes (150 mm 150 mm 150 mm) at 28 days, plotted against the w/p ratio in Fig. 9 indicates a good correlation between the compressive strength and w/p ratio. The compressive strength range varied from 20 MPa to 70 MPa. From this graph, it is possible to determine the w/p ratio for a specific strength. Though this graph is material specific (for SCC produced with 60:40 combination of cement:fly ash by volume), this will reduce the number of experimental trials to arrive at the required strength.
100 80 60 40 20 0
160
180
200
220
240
Water content (litres/m3) Fig. 7. Relationship between the V funnel time and water content.
70 60 50 40 30 20
0.6
0.8
1.0
1.2
1.4
1.6
1.8
w/p ratio Fig. 9. Relationship between the compressive strength and the w/p ratio.
3448
P. Nanthagopalan, M. Santhanam / Construction and Building Materials 23 (2009) 3443–3449
Table 8 Matrix of strengths and slump flows obtained in the study. Range
Compressive strength (MPa) 20–30
30–40
40–50
50–60
60–75
33.2–1.2 (450, 201, 388) 32.6–1.3 (450, 218, 405)
44.0–1.0 (500, 186, 392)
55.1–0.8 (550, 164, 388)
70.0–0.7 (600, 156, 399)
650–750
38.0–1.1 (500, 204, 410) 35.7–1.2 (550, 245, 470)
45.0–1.0 (525, 195, 410) 41.2–1.1 (550, 225, 449)
52.2–0.9 (550, 184, 408)
72.6–0.7 (650, 169, 431)
750–850
39.0–1.1 (525, 215, 430)
45.9–1.0 (550, 205, 429) 48.0–1.0 (650, 242, 503)
56.0–0.9 (600, 201, 444) 58.0–0.9 (650, 217, 479)
60.4–0.8 (650, 193, 455) 66.8–0.75 (650, 181, 443)
Slump flow (mm) 550–650 26.0–1.7 (350, 221, 371)
The results of slump flow and compressive strength obtained for all mixtures in the study are collectively presented in Table 8. These results are classified according to the range of slump flow and compressive strength obtained. The numbers presented in each cell are (in order): the specific compressive strength, water to powder ratio (by volume), powder content (kg/m3), water content (kg/m3), and paste volume (litres). The shaded cells represent mixtures with VMA. The results indicated that it was possible to achieve SCC with a slump flow of 560 mm for powder content as low as 350 kg/m3. From Table 8, it is possible to select the paste composition for a required combination of slump flow and compressive strength. 5. General discussions Unlike conventional concrete, for SCC numerous devices such as U box, L box, and V funnel have been proposed in the literature to test the fresh concrete properties [27]. However, they are either cumbersome to perform (large samples are required and there is very limited practical application) or more research is required to make them representative of the real confinement of structures in the case of very dense reinforcements [6]. It is preferable to use minimum number of fresh tests that addresses essential fresh properties instead of performing many tests. The intention of the above argument should not be viewed as criticism of the other available test methods, as better understanding of the behaviour of fresh SCC is possible by using such test methods. Instead, the authors’ view is that these test methods are still being perfected and cannot expect to yield predictions of desired accuracy. In the current study, slump flow and J ring were found to be sufficient to characterise the fresh concrete behaviour of SCC, since enough care was exercised to avoid segregation. In general, the fresh concrete characterisation of SCC should include a segregation test in addition to the slump flow and J-ring tests. The SCC mixtures that segregated either had water content more than 215 l or fine aggregate fraction less than 40% of the total aggregate. With regard to the present study, the segregation phenomenon is reduced by maximising the packing density; the choice of high packing density increases the range of aggregate particles, which reduces segregation by the lattice effect (grid of stable small and mid-size particles prevent the sinking of big particles). By using appropriate dosage of VMA, it was possible to obtain SCC without segregation and without compromising on the slump flow. Hence, it is suggested that VMA can be used to have a firm control over the stability and robustness of the mix as well as to achieve SCC at the same powder content as normal concrete. As a result, the potential for creep and shrinkage of SCC is also reduced. 6. Conclusions The influence of the paste composition and paste volume on the fresh and hardened concrete properties of SCC was discussed.
Experimental results revealed that paste volume had a predominant effect on the fresh concrete properties in comparison with water or powder content individually (for a given combination of aggregates). A minimum of 50–70 l of excess paste over and above the void content of the aggregates was found essential for achieving SCC with a slump flow of 550 mm. An empirical relationship was established for the determination of the paste volume for the required slump flow. The effect of paste volume on the V funnel flow time was not clear due to the blocking of aggregates nearer to the V funnel opening. The passing ability was studied by using the J-ring test. The results revealed that the difference between slump flow and the J ring flow varied between 10 and 50 mm. From the results, it is apparent that for understanding the fresh concrete behaviour of SCC, it is sufficient to conduct slump flow (flow ability) and J-ring test (passing ability). As expected, the w/p ratio had a good correlation with the compressive strength of SCC for a given cement/fly ash ratio. The compressive strength of mixes developed in this study ranged from 20 MPa to 70 MPa. It was possible to achieve SCC with powder content as low as 350 kg/m3. A matrix combination of strength and the slump flow was established based on the experimental results from which one can use the paste composition and paste content for the desired combination of slump flow and compressive strength.
References [1] ACI 237 R. Self-Consolidating Concrete. American Concrete Institute; 2007. p. 30. [2] Okamura H, Ouchi M. Self-Compacting Concrete. Development, present use and future. Proceedings of first international symposium on Self-Compacting Concrete. Stockholm; 1999. p. 3–14. [3] Saak AW, Jennings HM, Shah SP. New methodology for designing SelfCompacting Concrete. ACI Mater J 2001;98(6):429–39. [4] Okamura H, Maekawa K, Ozawa K. High performance concrete. Tokoyo: Gihodo Publishing; 1993 [in Japanese]. [5] Petersson O, Billberg P, Van BK. A model for self-compacting concrete. In: Proceedings of production methods and workability of concrete, London; 1996. p. 483–92. [6] Sedran T, de Larrard F, Hourst F, Contamines C. Mix design of self-compacting concrete. In: Proceedings of production methods and workability of concrete, London; 1996. p. 439–51. [7] Gomes PCC, Gettu R, Agullo L, Bernad C. Experimental optimisation of high strength Self-Compacting Concrete. In: Proceedings of second International RILEM symposium on self-compacting concrete, Japan; 2001. p. 377–86. [8] Su N, Hsu KC, Chai HW. A simple mix design method for self-compacting concrete. Cem Concr Res 2001;31(12):1799–807. [9] Smeplass S, Mortsell E. The particle matrix model applied on SCC. In: Proceedings of the second international RILEM symposium on selfcompacting concrete, Japan; 2001. p. 267–76. [10] EFNARC. The European guidelines for self-compacting concrete: specification, production and use; 2005. p. 68. [11] JSCE. Recommendation for construction of self-compacting concrete. In: International workshop on self-compacting concrete. Japan: Japanese Society of Civil Engineers; 1998. p. 77. [12] Glavind M, Pedersen EJ. Packing calculations applied for concrete mix design. In: Proceedings of creating with concrete. Dundee;1999. p. 1–10. [13] Johansen V, Andersen PJ. Particle packing and concrete properties. Material Science of Concrete II. Ohio: The American Ceramic Society; 1991. p. 111–47. [14] Fuller WB, Thompson SE. The laws of proportioning concrete. Transactions, American Society of Civil Engineers; 1907. p. 67–172.
P. Nanthagopalan, M. Santhanam / Construction and Building Materials 23 (2009) 3443–3449 [15] Senthilkumar V, Santhanam M. Particle packing theories and their application in concrete mixture proportioning: a review. The Indian Concrete Journal 2003:1324–31. [16] Lange F, Mortel H, Rudert V. Dense packing of cement pastes and resulting consequences on mortar properties. Cem Concr Res 1997;27(10):1481–8. [17] IS 12269. Specification for 53 grade ordinary portland cement. Indian Standards; 2004. [18] Puntke W. Wasseranspruch von feinen Kornhaufwerken (in German). Beton 2002;52(5):242–8. [19] Nanthagopalan P, Santhanam M. Investigation on the influence of granular packing on the flow properties of cementitious suspensions. Cem Concr Compos 2008;30:763–8. [20] Nanthagopalan P, Santhanam M. A new approach to optimisation of paste composition in self-compacting concrete. Ind Concr J 2008;82:11–8. [21] Kantro DL. Influence of water-reducing admixtures on properties of cement paste – a miniature slump test. Cem Concr Agg 1980;2(2):95–102.
3449
[22] Nanthagopalan P, Santhanam M. A new empirical test method for the optimisation of viscosity modifying agent dosage in Self-Compacting Concrete. Mater Struct (in press). [23] Rajayogan V, Santhanam M, Sharma BS. Evaluation of hydroxyl propyl starch as a viscosity-modifying agent for Self-Compacting Concrete. In: Proceedings of the third international RILEM symposium on Self-Compacting Concrete, Iceland; 2003. p. 386–94. [24] Kumar P, Haq MA, Kaushik SK. Early age strength of SCC with large volumes of fly ash. Ind Concr J 2004;78:25–9. [25] Bui VK, Montgomery D, Hinczak I, Turner K. Rapid testing method for segregation resistance of self-compacting concrete. Cem Concr Res 2002;32(9):1489–96. [26] ASTM C 1621. Standard test method for passing ability of self-compacting concrete by J ring. ASTM standards; 2006. [27] Okamura H, Ouchi M. Self-compacting concrete. J Adv Concr Technol 2003;1(1):5–15.