This research investigates the fresh behaviour and mechanical properties of self-compacting concrete (SCC) containing high volume of limestone, metakaolin, silica fume, zeolite, and viscosity modifying admixture. Two fine aggregates with different fineness modulus were also utilized to evaluate the effect of sand’s gradation on the mechanical and flow properties of SCC containing inert and pozzolanic powder. Slump flow, V-funnel for fresh concrete and 5-minute-old concrete, J-ring, Orimet with and without J-ring, and L-box and U-box tests were performed on all 14 fresh concrete mixtures to examine the fresh properties of self-compacting concrete. Compressive strength of hardened specimens was measured at 7 and 35 days and tensile strength was also determined at the age of 28 days. The results show that sand grading significantly affects the fresh properties of SCC. It is also concluded that high volume of active powders including metakaolin, zeolite, and silica fume could not improve both the flow and mechanical properties of SCC at the same time. Limestone can be effectively used as filler in SCC in high volume content. A new set of limits for the L-box and U-box tests for SCC containing silica fume is also recommended as the existing criteria are not satisfactory.
Self-compacting concrete (SCC) is a well-known construction material developed in the last two decades to address the engineer requests demanding more workable concrete [
The common practice to obtain self-consolidation behaviour in SCC is the limitation of the coarse aggregate content, reduction of maximum size of aggregates, and use of superplasticizer [
Due to the considerable impact of powder, specifically high volume content, on the SCC properties, several researchers have started evaluating the effect of powders on fresh and mechanical properties of self-compacting concrete. Mnahoncakova et al. [
The effects of limestone and chalk powders on SCC are investigated by Zhu and Gibbs [
In addition to the fresh behavior and mechanical properties of SCC, its microstructure has been also subjected to investigation [
The focus of the research presented in the above paragraphs mostly concentrates on the impact of fly ash and limestone powders on the SCC properties as these powders are easily available across the world and therefore they can be cheaply provided. There are limited published researches addressing the effect of other mineral powders including silica fume, metakaolin, and zeolite on the properties of SCC [
Another goal of this study was to examine the influence of sand’s fineness on the flow and mechanical properties of SCC. Each powder or VMA was mixed with two different types of sand (i.e., low and high modulus of fineness) to consider the effect of sand gradation and fineness on SCC properties as well. The current study investigated if incorporation of VMA or powders could improve the fresh properties of mixes made with the courser sand. Most accepted/recommended tests [
As mentioned in the previous section, in this study, the effect of three pozzolanic powders, two inert powders, two fine aggregates, and two VMAs on the properties of SCC was investigated. The following paragraphs describe the properties of these additives, cement, and sands used in the current study.
Portland cement conforming to ASTM type I with specific gravity and Blaine fineness of 3.1 and 320 m2/kg was used, respectively.
Two limestone-based natural types of sand were used. The nominal size and the specific gravity of the sand were measured as 4.75 mm and 2.6, respectively. The fineness modulus of the courser and finer sands was calculated as 2.65 and 3.30, respectively.
Coarse aggregate displayed the maximum nominal size and bulk density of 19 mm and 1500 kg/m3. Figure
(a) Particle size distribution of sand and (b) gravel.
Polycarboxylic based superplasticizer confirming to ASTM C494 was introduced to mixes. Two polysaccharide-based viscosity modifying admixtures (VMAs) with two different solid contents were used to improve the stability of SCC mixes containing no filler. Total dissolved solids in VMA1, VMA2, and superplasticizer were 17.8, 18.8, and 357.4 g/L, respectively. The specific gravity and PH were reported by the producers as 5.7 and 1.18 for VMA1 and 5.9 and 1.19 for VMA2.
Silica fume with specific surface (measured by the nitrogen absorption technique) and specific gravity of 16000 m2/kg and 2.2, respectively, was used as a pozzolanic powder.
Zeolite was another pozzolanic powder employed in this research. Natural zeolite as volcanic or volcano-sediment materials has a unique crystal structure and is classified as a hydrated alumino silicate of alkali and alkaline earth cations [
High reactive metakaolin (HRM) was selected as the third active powder. HRM, one of the newest supplementary cementitious materials to prove its merit in field application, has been used in concrete to offer an increase in compressive strength and a reduction in permeability while offering good workability [
XRD pattern of metakaolin used in this study [
Two inert powders including limestone 1 and limestone 2 were introduced to few SCC mixes in this study. Limestone 1 exhibited coarser size compared to limestone 2. The former was produced as the by-product of a stonework factory in Esfahan where the latter was regularly made as an original product in a chemical factory. The density of limestone 1 and limestone 2 was measured as 2600 kg/m3 and 2700 kg/m3 and their residue on 45
Chemical analysis of the cement and various powders in use.
Material | Chemical analysis (% mass) | ||||||||
---|---|---|---|---|---|---|---|---|---|
SiO2 | Al2O3 | Fe2O3 | CaO | MgO | SO3 | Na2O | K2O | Ignition lsoss | |
Cement | 20.03 | 4.53 | 3.63 | 60.25 | 3.42 | 2.23 | — | — | 1.37 |
Silica fume | 98.78 | 0.27 | 0.52 | 0.2 | — | — | 0.1 | 0.01 | 0.07 |
Zeolite | 67.79 | 13.66 | 1.44 | 1.68 | 1.2 | 0.5 | 2.04 | 1.42 | 10.23 |
Metakaolin | 51.85 | 43.78 | 0.99 | 0.2 | 0.18 | — | 0.01 | 0.12 | 0.57 |
Limestone 1 | 2.74 | 0.25 | 0.34 | 50.98 | 1.40 | — | 0.12 | 0.42 | 43.00 |
Limestone 2 | 1.36 | 0.10 | 0.20 | 50.96 | 2.60 | — | 0.11 | 0.40 | 44.41 |
As shown in Table
Mixture proportion.
No. | Mix | Powder or VMA type | Powder (kg/m3) | Sand type | Aggregate (kg/m3) | Super plasticizer (by powder mass) | W/C | W/P |
---|---|---|---|---|---|---|---|---|
1 | PL1 | — | — | 1 | 1700 | 1% | 0.47 | 0.47 |
2 | PL2 | — | — | 2 | 1700 | 1% | 0.37 | 0.37 |
3 | VMA1 | VMA-B | — | 2 | 1700 | 1% | 0.45 | 0.45 |
4 | VMA2 | VMA-F | — | 2 | 1700 | 1% | 0.43 | 0.43 |
5 | ME1 | Metakaolin | 250 | 1 | 1450 | 1% | 0.61 | 0.41 |
6 | ME2 | Metakaolin | 250 | 2 | 1450 | 1% | 0.62 | 0.41 |
7 | SF1 | Silica fume | 250 | 1 | 1450 | 1% | 0.68 | 0.45 |
8 | SF2 | Silica fume | 250 | 2 | 1450 | 1% | 0.65 | 0.43 |
9 | ZE1 | Zeolite | 250 | 1 | 1450 | 1% | 0.96 | 0.63 |
10 | ZE2 | Zeolite | 250 | 2 | 1450 | 1% | 0.77 | 0.51 |
11 | LS1 | Limestone 1 | 250 | 1 | 1450 | 1% | 0.52 | 0.35 |
12 | LS2 | Limestone 1 | 250 | 2 | 1450 | 1% | 0.43 | 0.29 |
13 | CC1 | Limestone 2 | 250 | 1 | 1450 | 1% | 0.46 | 0.31 |
14 | CC2 | Limestone 2 | 250 | 2 | 1450 | 1% | 0.39 | 0.26 |
The powders were introduced to the mixture at 50 percent of cement mass when the cement mass was kept constant at 500 kg/m3 in all mixtures. Therefore, in the mixtures containing powder (all mixes except PL1, PL2, VMA1, and VMA2), the total amount of cement and powder reached 750 kg/m3. The weight of aggregate in PL1, PL2, VMA1, and VMA2 mixes was designed as 1700 Kg/m3. Since the total weight of sand and powder should be constant, in other mixtures, the weight of aggregate dropped to 1450 kg/m3. In the VMA1 and VMA2 mixes, viscosity modifying admixture was introduced at 5 kg/m3.
One percent superplasticizer was added to all mixtures in addition to the sufficient water added to obtain the slump flow value of 650–750 mm. The water required for each SCC mix to reach to the desired slump value varied as powders displayed different water absorption. Figure
(a) W/P and (b) water demand.
The ZE1 and ZE2 mixtures made with zeolite required greater amount of water to reach the specified slump flow value as zeolite shows high water absorption compared to other powders. Figure
Concrete mixtures were mixed in a pan shear mixer. After flow properties of fresh concrete were investigated, it was cast into the 300 × 150 mm cylindrical molds. The specimens were removed from the molds after one day and they were cured in a controlled temperature (
Conforming to the EFNARC (2005) recommendations, the following tests were performed on the fresh self-compacting concrete.
The compressive strengths of the specimens were determined in accordance with ASTM C39 [
The results of slump flow, V-funnel, and Orimet tests showing the flowability of SCC are presented in Table
Results of slump flow, V-funnel, and Orimet.
Mixture | Slump flow (mm) | Slump flow classification | V-funnel (sec) | V-funnel classification | Orimet (sec) |
---|---|---|---|---|---|
PL1 | 665 | SF2 | 4.8 | VF1 | 2.7 |
PL2 | 675 | SF2 | 23.5 | VF2 | 6 |
VMA1 | 650 | SF1 | 3.4 | VF1 | 0.9 |
VMA2 | 665 | SF2 | 8 | VF2 | 13.3 |
ME1 | 635 | SF1 | 2.7 | VF1 | 2.5 |
ME2 | 540 | — | 1.9 | VF1 | 2 |
SF1 | 650 | SF1 | 1.3 | VF1 | 0.9 |
SF2 | 640 | SF1 | 1.8 | VF1 | 1 |
ZE1 | 715 | SF2 | 1.2 | VF1 | 0.6 |
ZE2 | 655 | SF2 | 1.3 | VF1 | 1.1 |
LS1 | 755 | SF3 | 4 | VF1 | 3 |
LS2 | 755 | SF3 | 5.4 | VF1 | 2.5 |
CC1 | 705 | SF2 | 2.4 | VF1 | 1.5 |
CC2 | 660 | SF2 | 7.9 | VF1 | 7.5 |
As determined, the slump flow of all mixtures ranged from 650 to 750 mm. The high value of the Orimet test in PL2, VMA2, and CC2 mixes indicates the high potential of segregation and blocking. The utilization of the coarser sand in these mixtures led the segregation though the incorporation of VMA and limestone powder. The results of the V-funnel test were similar to the results of the Orimet test as both are representative of the same concept. One may conclude that in the PL2, VMA2, and CC2 mixes, blocking occurred based on the long discharging time in the V-funnel test. However, the mixes containing VMA showed lower discharging time in the V-funnel test compared to the reference mixes (PL1 and Pl2). A decrease in the discharging time in mixes incorporating VMA was also reported by Lachemi et al. [
The results of the V-funnel test also show that the ME1, ME2, SF1, SF2, ZE1, and ZE2 mixes (mixes with metakaolin, silica fume, and zeolite powders) did not experience either segregation or instability with time. Analyzing the results of the V-funnel and Orimet tests for the LS2 and CC2 mixes suggests that limestone 1 made the SCC more flowable compared to limestone 2. This is in agreement with a study performed by Poppe and De Schutter [
In Table
The results of L-box, J-ring, and U-box tests.
Mixture | L-box end difference (cm) | L-box classification |
|
|
J-ring (cm) | U-box (cm) |
---|---|---|---|---|---|---|
PL1 | 0.5 | PA1 | 1.2 | 2.5 | 0 | 1 |
PL2 | 1.4 | PA1 | 2.6 | 4.5 | 1.9 | 27.5 |
VMA1 | 1 | PA1 | 0.5 | 0.7 | 0.5 | 25.5 |
VMA2 | 2 | PA1 | 1.8 | 3.3 | — | 25 |
ME1 | 2 | PA1 | 0.8 | 1.2 | 0.5 | 3 |
ME2 | 1.8 | PA1 | — | — | 0.4 | 17 |
SF1 | −2 | — | — | 0.5 | 0.7 | −1 |
SF2 | −2.5 | — | — | 1 | 0.5 | −1 |
ZE1 | −1.6 | — | — | 0.6 | 0.4 | −4.6 |
ZE2 | −4 | — | — | 0.5 | 1.1 | 7 |
LS1 | 0.4 | PA1 | — | 1.4 | 0 | 0 |
LS2 | 0.2 | PA1 | — | — | 0.4 | 1.8 |
CC1 | 0.5 | PA1 | — | 1.6 | 0 | 0 |
CC2 | 0.5 | PA1 | 1.2 | 1.9 | 0.4 | 0.5 |
LS1 mixture in the U-box.
The result of the L-box test shows the negative value for the SF1, SF2, ZE1, and ZE2 mixes containing silica fume and zeolite. It means that the level of concrete at the end of the box exceeded the level at the beginning of the box. The negative value in U-box test was also observed for these mixtures. Low viscosity leading to high initial speed attributes to the negative value in these mixes. Other mixes, except VMA1 and VMA2, showed the reasonable end difference satisfying the EFNARC [
The same conclusion made from the U-box test can be drawn from the results of the L-box test. It can be concluded that VMA did not help to improve the stability of SCC in the VMA1 and VMA2 mixes. The results of
The results of 5-minute V-funnel, visual stability index (VSI), and Orimet with J-ring are presented in Table
V-funnel after 5 minutes, visual stability index (VSI), and Orimet with J-ring test results.
Mixture | V-funnel (sec) | VSI | VSI classification | Orimet with J-ring (cm) |
---|---|---|---|---|
PL1 | 6.9 | 1 | Acceptable | 1.9 |
PL2 | 29.1 | 3 | Unacceptable | 2.4 |
VMA1 | 11.5 | 2 | Needs improvement | 0.5 |
VMA2 | 14.5 | 2 | Needs improvement | 2.1 |
ME1 | 3.7 | 1 | Acceptable | 1.1 |
ME2 | 2.5 | 2 | Needs improvement | — |
SF1 | 1.4 | 3 | Unacceptable | 0 |
SF2 | 1.9 | 3 | Unacceptable | 0 |
ZE1 | 1.2 | 2 | Needs improvement | 0 |
ZE2 | 1.6 | 3 | Unacceptable | — |
LS1 | 5.3 | 0 | Excellent | — |
LS2 | 6.7 | 0 | Excellent | 0.5 |
CC1 | 4.5 | 0 | Excellent | 0.5 |
CC2 | 20.1 | 0 | Excellent | 2.1 |
The result of the Orimet test with J-ring helps to evaluate the dynamic stability of SCC. The difference in height of two sides of J-ring after concrete drops from the Orimet was measured for the PL1, PL2, VMA2, and CC2 mixes. The results show that high dynamic segregation occurred in the plain concrete. Meanwhile, the VMA and limestone powder were not effective in eliminating the dynamic segregation. The results of the V-funnel test after 5 minutes can show the stability characteristic of SCC after certain time. The values show that high segregation, blocking, and instability occurred in the PL2 and CC2 mixes. Short discharging time in the V-funnel test in ME1, ME2, ZE1, ZE2, SF1, and SF2 indicates that no segregation and instability happened. Also, one can expect that in these mixes the viscosity of concrete did not increase after 5 minutes. However, Şahmaran et al. [
According to the VSI values, the mixtures made with limestone powder, LS1, LS2, CC1, and CC2, can be definitely considered as SCC. In these mixtures, no segregation was observed in the slump flow test. On the other hand, the VSI value of 3 was assigned to the PL2, SF1, SF2, and ZE2 mixes, indicating the high level of instability.
The authors conclude that if sand with a lower fineness modulus is utilized in the SCC mixture, it is possible to make a stable and homogenous SCC without incorporating any additive. Finer sand helps particles to move easily and also parts of the sand can behave as filler. Meanwhile, the coarser sand results in the segregated and unstable SCC as suggested by the U-box test.
An increase in the discharge time of the Orimet test indicates that by adding VMA to the SCC mixes the viscosity of concrete can slightly improve. Nonetheless, the incorporated VMA did not significantly affect the flow properties of SCC and could not compensate the adverse effect of improper sand gradation. Therefore, one may expect consequential segregation and bleeding in SCC made with VMA and coarse sand.
The results of SCC containing metakaolin reveal that metakaolin could not introduce significant effects on the fresh properties of SCC. The sand gradation shows a stronger effect compared to metakaolin. The results of the U-box test clearly demonstrate that the metakaolin could not enhance the fresh properties of concrete made with coarse sand.
Silica fume could improve the flow properties of SCC. Self-compacting concrete made with silica fume did flow at high speed causing the concrete to expel from horizontal part of L-box, without segregation, as shown in Figure
SCC containing silica fume.
It is important to note that although the mix containing silica fume meets the criteria proposed for fresh properties of SCC, its low viscosity makes it unusable in concrete structures. Therefore, it is recommended to restrict the limitations offered by EFNARC [
The same results observed for concrete containing metakaolin could be observed when zeolite powder was added to the mix. The results of the J-ring test exhibit that the viscosity of concrete decreased significantly and concrete did quickly spread when metakaolin and zeolite powders were introduced to the mixes. Thus, after the flow of concrete stopped, a “petal-shape” pattern was formed. This type of concrete could not be classified as self-compacting as it is not able to fill the entire free space of the form due to its weak filling ability. Figure
SCC distributed as (a) uniform and (b) petal-shape in the J-ring test.
Utilization of limestone powder as inert powder could improve the workability of SCC without changing the viscosity. Yahia et al. [
The compressive strength results of hardened concrete are shown in Figure
Compressive strenght at 7 and 35 days.
In the ME1, ME2, ZE1, and ZE2 mixes, the finer sand results in the higher compressive strength. The values of 35-day compressive strength indicate that silica fume did not change the compressive strength of reference mix that is in agreement with results of a study performed by Kwan and Ng [
In the plain concrete and concrete containing VMA, where segregation and bleeding occurred, complete consolidation could not be obtained. Poor compaction led to porous concrete as well as to low compressive strength. The results also show that VMA did not improve the 7-day compressive strength of SCC. The same conclusion was also drawn by Lachemi et al. [
Comparison of the values of compressive strength for PL1 and LS1 at 7 days suggests that the incorporation of limestone increased the compressive strength by 100%. The same conclusion was observed by Zhu and Gibbs [
The results of tensile strength of SCC mixtures are shown in Figure
Tensile strength at 28 days.
The effect of different powder materials incorporated at high volume and fine aggregate with different fineness modulus on the flow and mechanical properties of SCC was investigated. Based on the results of the current research, the following conclusions can be drawn. If fine aggregate exhibits a low fineness modulus and proper gradation, there is no need to use any powder as filler to help the flow properties. The best value of the fineness modulus for fine aggregate is recommended to be 2.65. VMA cannot significantly improve the flow properties of SCC when it is utilized in a mix containing coarse sand. When coarse sand is incorporated in the mixture, the addition of VMA shows no effect on the stability, including bleeding and segregation. Metakaolin powder cannot improve the flow properties of SCC. Silica fume compensates the poor sand quality. However, it results in SCC with low viscosity that may attribute to some problems in practice, although SCC made with silica fume can successfully pass the tests. Limestone powder greatly increases the stability and homogeneity of SCC. These fillers are relatively of low cost and are widely available. Due to the high water absorption of zeolite and silica fume, the mechanical strength of SCC containing these powders deteriorates. In comparison with the reference mix, compressive and tensile strength of SCC exhibits a decrease of about 50% and 60%, respectively, when zeolite powder is used.
The authors declare that there is no conflict of interests regarding the publication of this paper.