- Email: [email protected]
Hydraulic-lime based concrete: Strength development using a pozzolanic addition and different curing conditions
Construction and Building Materials 23 (2009) 2107–2111
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Hydraulic-lime based concrete: Strength development using a pozzolanic addition and different curing conditions Ana L. Velosa a, Paulo B. Cachim b,* a b
Department of Civil Engineering/MIA, University of Aveiro, 3810-193 Aveiro, Portugal Department of Civil Engineering/LABEST, University of Aveiro, 3810-193 Aveiro, Portugal
a r t i c l e
i n f o
Article history: Received 13 March 2008 Received in revised form 30 July 2008 Accepted 25 August 2008 Available online 5 October 2008 Keywords: Hydraulic lime Curing Pozzolan Mechanical properties
a b s t r a c t Concrete is a major worldwide building material, in which Portland cement is the usual binder. Taking into account environmental factors in cement production, especially concerning CO2 emissions and energy consumption, this work aims at the development of concrete with a hydraulic-lime binder; in order to increase mechanical strength, pozzolanic materials were added. In this preliminary study, compositions with different percentages of hydraulic lime were tested and a pozzolanic material, a residue from expanded clay production, was used. Variations in percentage of pozzolan and conditioning were carried out. Concrete specimens were tested for mechanical strength at various ages and a pozzolanic index was determined in order to evaluate the influence of the pozzolanic material on attained mechanical strength. This paper presents the results of this testing campaign, concluding on the influence of pozzolanic additions and curing conditions on the strength development of this material. Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction The protocol of Kyoto has created demanding goals in terms of decrease in greenhouse gas emissions. Furthermore, global warming and climate change effects felt throughout the world have created a growing concern on energy consumption and efforts towards the use of eco-friendly materials, with a lower environmental impact, are being undertaken. Cement industry generates around 5% of global CO2 emissions, due to carbonate decomposition (about 50%), combustion of fuels in the kiln (about 40%). Electricity generation and transportation, ranking as the third largest carbon emitting industry in the EU [1,2]. It is estimated that each tonne of cement produces approximately 1 tonne of CO2, mainly from the burning of fossil fuels and from the de-carbonation of limestone [1]. Although the cement industry is taking measures in order to ensure a reduction in greenhouse gas emissions, using alternative fuels or changing cement composition [3,4], this could be further achieved using substitute materials whenever possible. As a binder, Portland cement contributes towards high concrete strength, but other binders could be used in constructions with lower structural strength demands. Both air lime – together with pozzolanic additives – and hydraulic lime were used successfully for construction purposes throughout the ages, with an adequate performance and in some cases, a great durability.
* Corresponding author. Tel.: +351 234370049; fax: +351 234370094. E-mail address: [email protected] (P.B. Cachim). 0950-0618/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.conbuildmat.2008.08.013
Portugal has two cement production companies and high per capita cement consumption, partially due to investment in newbuild versus rehabilitation. However, hydraulic lime (NHL5) is also produced, but there is no tradition in its use and it has been particularly applied in the still small market of building conservation practice. The use of hydraulic lime as a binder in concrete is a possibility, especially in applications where there is a need for a moderate versus high mechanical strength, although this characteristic may be enhanced by the use of pozzolanic materials, currently regarded as a valid possibility in an attempt to produce binders with lower CO2 emissions. These materials are characterized by the ability to react with lime (calcium hydroxide) in the presence of water, forming calcium silicate hydrates. Pozzolanic materials, of natural or artificial origin, must contain a high percentage of amorphous silica and a high specific surface in order to generate a pozzolanic reaction. Currently, the re-use of waste materials with pozzolanic properties is a growing reality as cementitious materials are widely applied and provide a suitable application possibility with evident advantages (mitigation of AAR, increase in mechanical strength, among others). Among these, products deriving from clay calcination, such as metakaolin, are starting to be applied in Portugal. The residue of expanded clay production used in this study is a similar product, resulting from clay calcinations at temperatures surrounding 1200 °C. Collected as a fine powder, or grinded, this material is a strong possibility for use in concrete and mortars. Added environmental value is given by the use of waste products or pozzolanic materials that create less environmental impacts, by lower calcinations temperatures and/or
2108
A.L. Velosa, P.B. Cachim / Construction and Building Materials 23 (2009) 2107–2111
lower energy consumption, among others. The expanded clay residue used for this study derived from industrial waste. 2. Characterization of expanded clay residue Obtained during the process of expanded clay production, this residue is a fine material with the same composition as expanded clay. It was characterized by X-ray diffraction (XRD), using a Philips X’Pert PW 3040/60 using Cu Ka radiation, with operational conditions of 30 mA and 50 kV, and speed registry 1°/2h/min, with data acquisition by Philips X’Perta Data Collector v1.2., in terms of mineral composition and by X-ray fluorescence (XRF) using a Philips PW 1400 X-ray Fluorescence Spectrometer for the determination of chemical composition. Expanded clay residue, ECR, is mainly composed of quartz (Q), spinel (SP), calcite (C) and feldspars (A, E). It has a small but evident band ranging from 20° to 30°, indicating the presence of amorphous material (Fig. 1). Silicates and aluminates are predominant in terms of chemical composition (Table 1) that also indicates the presence of iron, calcium and basic elements (sodium and potassium) in small quantities. Expanded clay residue is a very fine material, with a specific surface of 9100 cm2/g and it was sieved in order to ensure a maximum particle size of 75 lm, enhancing pozzolanic reactivity. This characteristic was measured following NP EN 196-5: Methods of testing cement. Pozzolanicity test for pozzolanic cement. Although this testing procedure is intended for application to pozzolanic cements, it has proven adequate for the measurement of the pozzolanic reactivity of pozzolanic materials as it measures Ca(OH)2 consumption in a solution with a standard quantity of cement and pozzolan. When applied to the expanded clay residue, this testing procedure classified the material as an active pozzolan.
used in low demanding structural applications such as urban equipment, cycling roads, blocks, among others. Aggregates used in this study were natural siliceous sand and calcareous coarse aggregate. Coarse aggregates were divided into two groups; one in the range 5–10 mm (CA1) and the other in the range 10–25 mm (CA2). 3.2. Concrete composition Concrete composition was studied using a modified Faury method that takes into account the differences between cement (that served as basis of the method) and hydraulic lime. Aggregates were used in a saturated condition. The water/cement ratio was 0.45. Hydraulic lime was replaced by 20% (composition M2) and 30% (composition M3) of expanded clay residue, by weight. Grading of aggregates is presented in Fig. 2 that also shows the resulting final grading curve. The final mix proportions for all types of studied eco-concrete are shown in Table 2 including hydraulic-lime percentage in the binder (p). 3.3. Methodology The developed experimental program was designed to assess the effect of ECR addition to hydraulic-lime concrete and its behaviour under different curing conditions. Three different curing conditions at atmospheric pressure were used where the relative humidity, RH, of the environment was changed. Concrete specimens were cured immersed in water, at 95% RH and at 65% RH. The temperature was kept constant at 20 °C. Fresh properties of concrete such as density and workability were measured. Workability of fresh concrete was measured using
3. Materials and methodology 3.1. Materials All the materials used in this study were commonly available in the central region of Portugal. The binder used was hydraulic lime (NHL5) that is currently the only hydraulic lime produced in Portugal. Additionally, its compressive strength of 5 MPa at the age of 28 days is sufficient to be improved by the addition of a pozzolan, allowing the achievement of concrete with strength in the range 15–20 MPa at age 90 days. Concrete with this strength could be
Fig. 2. Aggregate grading curves.
Table 2 Mixture composition including hydraulic-lime percentage in the binder (p) Name
M0 M2 M3
Fig. 1. XRD of expanded clay residue.
Constituents (kg/m3)
p (%)
CA1
CA2
Sand
NHL5
ECR
Water
619 619 619
455 455 455
321 321 321
550 440 385
– 110 165
247.5 247.5 247.5
100 80 70
Table 1 Chemical composition of expanded clay residue Oxides
CaO
SiO2
Al2O3
Fe2O3
K2O
MgO
Na2O
TiO2
P2O5
MnO
LOI
Percentage in weight
3.92
56.52
19.50
8.05
4.58
3.97
0.33
0.95
0.18
0.14
0.70
A.L. Velosa, P.B. Cachim / Construction and Building Materials 23 (2009) 2107–2111
the slump test (NP EN 12350-2:2002). Hardened concrete properties investigated were compressive strength (tests on 15 cm cubes were performed at 7, 28 and 90 days according NP EN 123903:2003 standard), tensile splitting strength (tests performed on 15 cm diameter cylinders according to NP EN 12390-3:2003). Pozzolanic strength indexes were calculated following the work of Yu et al. [5]. The specific strength ratio, R, is defined as contributions to concrete strength from unit hydraulic lime and unit mineral admixture and is defined by
R¼
fc p
ð1Þ
where fc is the concrete compressive strength and p is the cement or mineral admixture percentage of the cementitious materials (see Table 2). RC expresses the contribution of unit cement to concrete strength without any mineral admixture, RM expresses the contribution of unit mineral admixture to concrete strength, and RP is the contribution of the pozzolanic effect to concrete strength due to mineral admixture, expressed by
RP ¼ R M RC
ð2Þ
The index of specific strength, K, is the ratio of RM to RC, and the contribution of pozzolanic effect to strength, P, can be assessed by the percentage value of the contribution of pozzolanic effect to concrete strength, which can be written as
P ¼ 100ðRP =RM Þ
ð3Þ
The pozzolanic index, P, relates to the favourable contribution of the pozzolan relatively to the usual binder. 4. Results and discussion 4.1. Fresh concrete Slump results are 1.5 cm, 0.9 cm and 0.5 cm, respectively, for series M0, M2 and M3 (class S1, following NP EN 206-1:2007) for all mixes with and without substitution of hydraulic lime, that indicates a rather small workability of concrete. Since the aim of this study was the investigation of the effect of expanded clay residue and of curing conditions, this lack of workability wasn’t considered a very important issue at this stage. Nevertheless, this is an important aspect that must be accounted for in the subsequent development of the study and that may be solved by the addition of a plasticizer. Density of fresh concrete was approximately constant at 2290 kg/m3. 4.2. Hardened concrete 4.2.1. Strength The results obtained for compressive strength at 7, 28 and 90 days are shown in Table 3 together with the corresponding stan-
dard deviation. It is apparent from the Table 3 that reference limecrete, mixture M0, has higher strength until 28 days if cured at normal conditions (20 °C and 65% RH) than in a saturated environment. However, this improved strength at 65% RH is attenuated with time and at 90 days all the curing conditions lead to approximately the same strength. The presence of expanded clay residue alters this situation. As ECR is added to limecrete composition, the presence of water during curing becomes fundamental. For mixture M3 the effect of curing conditions is quite apparent, especially between 28 and 90 days, where nearly 100% increase in strength can be observed if specimens are cured in 95% RH or immersed. The results observed for mixture M2 show transition behaviour between mixtures M0 and M3, as can be easily seen in Table 3. Thus, it seems clear that when expanded clay residue is used, curing in 95% RH and in water is better. The addition of expanded clay residue only has a decisive beneficial effect on strength only at 90 days and in saturated curing conditions or under water, as both these curing conditions favour pozzolanic reaction. Fig. 3 shows results of concrete at age 28 days as a function of curing conditions. Although further studies with other pozzolanic materials must be executed in order to assess the influence of the addition of pozzolans in the strength development of limecrete, the behaviour of hydraulic-lime concrete with expanded clay residue is significantly improved in saturated conditions or under water when compared with limecrete with no pozzolanic additions in the same situation. The results obtained show that the influence of curing conditions at early ages is small but becomes evident at age 90 days: concretes with expanded clay residue perform better in saturated curing conditions or under water. This may be explained by water demand of the pozzolanic reaction, which is slower than hydraulic reaction but will only take place in the presence of available water. This slower reaction will only produce visible results at later ages, accounting for the differences between results of concretes stored in saturated conditions or under water at age 90 days. This effect can clearly be observed in Fig. 4 where the strength evolution in time, normalised by the respective 28 days strength is plotted. Results for tensile strength closely follow the compressive strength results. The relation between compressive and tensile splitting strength is shown in Fig. 5. The results are compared with the relation proposed by the European concrete design code, EC2 [6], that relates tensile and compressive strength by a power law
fct;sp ¼ 0:27fc2=3
ð4Þ
Comparison of EC2 equation with experimental data (see Fig. 5) shows that almost all the points are under the EC2 proposed curve given by Eq. (4). Because EC2 curve is intended to be for mean values, the tensile strength of limecrete seemed to be comparatively smaller than that of Portland cement concrete.
Table 3 Compressive strength and standard deviation at 7, 28 and 90 days Series
M0-65 M0-95 M0-W M2-65 M2-95 M2-W M3-65 M3-95 M3-W
Compressive strength (MPa)
Standard deviation (MPa)
7 days
28 days
90 days
7 days
28 days
90 days
6.6 5.0 3.3 6.0 5.5 4.0 4.9 4.6 4.9
8.9 8.3 5.6 9.4 11.1 8.8 8.6 10.8 9.6
10.2 11.1 10.5 13.2 18.4 14.5 11.5 26.6 18.5
0.16 0.24 0.15 0.40 0.14 0.09 0.26 0.21 0.30
0.21 0.57 0.21 0.47 0.29 0.08 0.88 0.10 0.33
0.81 0.27 0.01 0.78 0.53 0.56 0.75 0.92 0.56
2109
Fig. 3. Compressive strength at 28 days.
2110
A.L. Velosa, P.B. Cachim / Construction and Building Materials 23 (2009) 2107–2111 Table 4 Pozzolanic indexes RM, RP, K and P
Fig. 4. Strength evolution in time normalised by the respective 28 days strength.
Days of curing
Mixture
7
M0-65 M0-95 M0-W M2-65 M2-95 M2-W M3-65 M3-95 M3-W
28
90
fc (MPa)
RM
RP
K
P
6.6 5.0 3.3 6.0 5.5 4.0 4.9 4.6 4.9
0.066 0.050 0.033 0.075 0.068 0.050 0.070 0.066 0.070
0.0 0.0 0.0 0.0098 0.0187 0.0174 0.0044 0.0164 0.0370
1.0 1.0 1.0 1.15 1.38 1.53 1.07 1.33 2.13
0.0 0.0 0.0 13.0 27.3 34.8 6.3 24.9 53.1
M0-65 M0-95 M0-W M2-65 M2-95 M2-W M3-65 M3-95 M3-W
8.9 8.3 5.6 9.4 11.1 8.8 8.6 10.8 9.6
0.089 0.083 0.056 0.118 0.138 0.110 0.122 0.154 0.138
0.0 0.0 0.0 0.0292 0.0557 0.0541 0.0338 0.0716 0.0813
1.0 1.0 1.00 1.33 1.67 1.96 1.38 1.87 2.44
0.0 0.0 0.0 24.8 40.2 49.0 27.6 46.4 59.1
M0-65 M0-95 M0-W M2-65 M2-95 M2-W M3-65 M3-95 M3-W
10.2 11.1 10.5 13.2 18.4 14.5 11.5 26.6 18.5
0.102 0.111 0.105 0.165 0.230 0.181 0.165 0.380 0.264
0.0 0.0 0.0 0.0630 0.1189 0.0760 0.0628 0.2693 0.1590
1.0 1.0 1.0 1.62 2.07 1.72 1.62 3.43 2.51
0.0 0.0 0.0 38.2 51.8 41.9 38.1 70.9 60.2
Fig. 5. Relation between compressive and tensile splitting strength, compared with EC2 curve.
A regression equation was fitted to experimental data that is given by Eq. (5) with a correlation coefficient r2 = 0.735 (also shown in Fig. 5).
fct;sp ¼ 0:154fc0:74
ð5Þ
The ratio tensile strength/compressive strength is lower for this kind of concrete in relation to ordinary Portland cement concrete and a possible explanation for this is the amount of water that was used, with no aid of super plasticizers. Although this may indicate a greater cracking susceptibility, no cracking was observed in the samples. 4.2.2. Pozzolanic index Pozzolanic indexes RM, RP, K and P are shown in Table 4, for each mixture, together with the compressive strength achieved in every case, at ages 7, 28 and 90 days. At the age of 7 days there is a decrease in P when pozzolanic substitution increases from 20% to 30%, except for the case of water immersion. This behaviour differs from that of latter ages. For this reason, it is not possible to conclude on the behaviour of limecrete with easy age testing. For ages over 28 days and adequate curing conditions (95% RH and water immersion) the variation in pozzolan substitution from 20% to 30% implies a greater contribution of this material towards final strength (Fig. 6).
Fig. 6. Comparison of pozzolanic index P between 20% and 30% of cement substitution with different curing conditions and at various ages (7, 28 and 90 days).
Pozzolanic index P, related to the same substitution percentage and the same curing conditions always increases in time, with the sole exception of 20% ECR substitution, water immersed, between 28 and 90 days (Fig. 6). Conditioning at 95% RH or 100% (in water) positively influences the pozzolanic index, as expected. Water conditioning is generally more favourable at early ages. In the long run, as can be seen by results at 90 days, for pozzolanic substitution percentages of 20% and 30%, conditioning at high RH favours the pozzolanic index. Only with a 95% RH conditioning there is a continuously positive contribution of the pozzolanic addition, both in terms of substitution percentage and in terms of long term behaviour. In the case of 65% RH, the increase in pozzolan percentage from 20% to 30% does not produce a significant or consistent effect in
A.L. Velosa, P.B. Cachim / Construction and Building Materials 23 (2009) 2107–2111
2111
Fig. 7. Pozzolanic index, P.
terms of achieved strength, Table 3, and in terms of pozzolanic reaction contribution (Figs. 6 and 7). The contribution of the pozzolan when specimens are immersed in water is also positive but not as effective as 95% RH conditions. The pozzolanic index P is very sensitive to curing conditions. The performance of concretes with hydraulic lime and pozzolanic additions will improve in high RH conditions.
5. Conclusions Expanded clay residue is a suitable material for application in concrete with hydraulic-lime binder as a pozzolanic addition. A substitution of hydraulic lime by 20% and 30% of expanded clay residue produced increased compressive strength at adequate curing conditions, especially in the latter case. Therefore, these replacement percentages are satisfactory for current use. As results indicate, curing conditions produce significant changes in mechanical strength, but not at early ages, due to the development of chemical reactions in time. High relative humidity conditions are generally more favourable towards the increase in mechanical strength. Testing at early ages (7 days) does not allow attaining knowledge on future behaviour of limecrete, contrarily to current cement concrete. In fact testing procedures for current Portland cement concrete may not be applicable as hydration reaction is faster than carbonation and pozzolanic reactions that also contribute towards the hardening of hydraulic-lime based concrete.
The use of both the pozzolanic index P and mechanical strength results and their combined analyses, shows the real contribution of the pozzolanic material towards the mechanical properties of limecrete. Hydraulic-lime concrete was developed on this study without the aid of super plasticizers and therefore with a great amount of water. This was done in order to verify changes in behaviour and the material’s basic potential. However, the addition of super plasticizers may improve characteristics significantly and therefore enlarge the application possibilities for this material. This paper opens the possibility towards the use of hydraulic lime in concretes in which compressive strength needs are not very high, promoting the use of a more sustainable material. Concretes based on Portland cement will provide higher strength and have different application aims. Alternative pozzolanic materials to expanded clay residue may be tested in a similar fashion. References [1] Rehan R, Nehdi M. Carbon dioxide emissions and climate change: policy implications for the cement industry. Environ Sci Policy 2005;8(2):105–14. [2] Szabó L, Hidalgo I, Ciscar JC, Soria A. CO2 emission trading within the European Union and Annex B countries: the cement industry case. Energy Policy 2006;34(1):72–87. [3] Gäbel K, Tillman A. Simulating operational alternatives for future cement production. J Clean Prod 2005;13(13/14):1246–57. [4] Gartner E. Industrially interesting approaches to ‘‘low-CO2” cements. Cement Concrete Res 2004;34:1489–98. [5] Yu LH, Ou H, et al. Investigation on pozzolanic effect of perlite powder in concrete. Cement Concrete Res 2003;33(1):73–6. [6] CEN. Eurocode 2: design of concrete structures – Part 1-1: general rules and rules for buildings. Brussels: CEN; 2004.
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Hydraulic-lime based concrete: Strength development using a pozzolanic addition and different curing conditions Ana L. Velosa a, Paulo B. Cachim b,* a b
Department of Civil Engineering/MIA, University of Aveiro, 3810-193 Aveiro, Portugal Department of Civil Engineering/LABEST, University of Aveiro, 3810-193 Aveiro, Portugal
a r t i c l e
i n f o
Article history: Received 13 March 2008 Received in revised form 30 July 2008 Accepted 25 August 2008 Available online 5 October 2008 Keywords: Hydraulic lime Curing Pozzolan Mechanical properties
a b s t r a c t Concrete is a major worldwide building material, in which Portland cement is the usual binder. Taking into account environmental factors in cement production, especially concerning CO2 emissions and energy consumption, this work aims at the development of concrete with a hydraulic-lime binder; in order to increase mechanical strength, pozzolanic materials were added. In this preliminary study, compositions with different percentages of hydraulic lime were tested and a pozzolanic material, a residue from expanded clay production, was used. Variations in percentage of pozzolan and conditioning were carried out. Concrete specimens were tested for mechanical strength at various ages and a pozzolanic index was determined in order to evaluate the influence of the pozzolanic material on attained mechanical strength. This paper presents the results of this testing campaign, concluding on the influence of pozzolanic additions and curing conditions on the strength development of this material. Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction The protocol of Kyoto has created demanding goals in terms of decrease in greenhouse gas emissions. Furthermore, global warming and climate change effects felt throughout the world have created a growing concern on energy consumption and efforts towards the use of eco-friendly materials, with a lower environmental impact, are being undertaken. Cement industry generates around 5% of global CO2 emissions, due to carbonate decomposition (about 50%), combustion of fuels in the kiln (about 40%). Electricity generation and transportation, ranking as the third largest carbon emitting industry in the EU [1,2]. It is estimated that each tonne of cement produces approximately 1 tonne of CO2, mainly from the burning of fossil fuels and from the de-carbonation of limestone [1]. Although the cement industry is taking measures in order to ensure a reduction in greenhouse gas emissions, using alternative fuels or changing cement composition [3,4], this could be further achieved using substitute materials whenever possible. As a binder, Portland cement contributes towards high concrete strength, but other binders could be used in constructions with lower structural strength demands. Both air lime – together with pozzolanic additives – and hydraulic lime were used successfully for construction purposes throughout the ages, with an adequate performance and in some cases, a great durability.
* Corresponding author. Tel.: +351 234370049; fax: +351 234370094. E-mail address: [email protected] (P.B. Cachim). 0950-0618/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.conbuildmat.2008.08.013
Portugal has two cement production companies and high per capita cement consumption, partially due to investment in newbuild versus rehabilitation. However, hydraulic lime (NHL5) is also produced, but there is no tradition in its use and it has been particularly applied in the still small market of building conservation practice. The use of hydraulic lime as a binder in concrete is a possibility, especially in applications where there is a need for a moderate versus high mechanical strength, although this characteristic may be enhanced by the use of pozzolanic materials, currently regarded as a valid possibility in an attempt to produce binders with lower CO2 emissions. These materials are characterized by the ability to react with lime (calcium hydroxide) in the presence of water, forming calcium silicate hydrates. Pozzolanic materials, of natural or artificial origin, must contain a high percentage of amorphous silica and a high specific surface in order to generate a pozzolanic reaction. Currently, the re-use of waste materials with pozzolanic properties is a growing reality as cementitious materials are widely applied and provide a suitable application possibility with evident advantages (mitigation of AAR, increase in mechanical strength, among others). Among these, products deriving from clay calcination, such as metakaolin, are starting to be applied in Portugal. The residue of expanded clay production used in this study is a similar product, resulting from clay calcinations at temperatures surrounding 1200 °C. Collected as a fine powder, or grinded, this material is a strong possibility for use in concrete and mortars. Added environmental value is given by the use of waste products or pozzolanic materials that create less environmental impacts, by lower calcinations temperatures and/or
2108
A.L. Velosa, P.B. Cachim / Construction and Building Materials 23 (2009) 2107–2111
lower energy consumption, among others. The expanded clay residue used for this study derived from industrial waste. 2. Characterization of expanded clay residue Obtained during the process of expanded clay production, this residue is a fine material with the same composition as expanded clay. It was characterized by X-ray diffraction (XRD), using a Philips X’Pert PW 3040/60 using Cu Ka radiation, with operational conditions of 30 mA and 50 kV, and speed registry 1°/2h/min, with data acquisition by Philips X’Perta Data Collector v1.2., in terms of mineral composition and by X-ray fluorescence (XRF) using a Philips PW 1400 X-ray Fluorescence Spectrometer for the determination of chemical composition. Expanded clay residue, ECR, is mainly composed of quartz (Q), spinel (SP), calcite (C) and feldspars (A, E). It has a small but evident band ranging from 20° to 30°, indicating the presence of amorphous material (Fig. 1). Silicates and aluminates are predominant in terms of chemical composition (Table 1) that also indicates the presence of iron, calcium and basic elements (sodium and potassium) in small quantities. Expanded clay residue is a very fine material, with a specific surface of 9100 cm2/g and it was sieved in order to ensure a maximum particle size of 75 lm, enhancing pozzolanic reactivity. This characteristic was measured following NP EN 196-5: Methods of testing cement. Pozzolanicity test for pozzolanic cement. Although this testing procedure is intended for application to pozzolanic cements, it has proven adequate for the measurement of the pozzolanic reactivity of pozzolanic materials as it measures Ca(OH)2 consumption in a solution with a standard quantity of cement and pozzolan. When applied to the expanded clay residue, this testing procedure classified the material as an active pozzolan.
used in low demanding structural applications such as urban equipment, cycling roads, blocks, among others. Aggregates used in this study were natural siliceous sand and calcareous coarse aggregate. Coarse aggregates were divided into two groups; one in the range 5–10 mm (CA1) and the other in the range 10–25 mm (CA2). 3.2. Concrete composition Concrete composition was studied using a modified Faury method that takes into account the differences between cement (that served as basis of the method) and hydraulic lime. Aggregates were used in a saturated condition. The water/cement ratio was 0.45. Hydraulic lime was replaced by 20% (composition M2) and 30% (composition M3) of expanded clay residue, by weight. Grading of aggregates is presented in Fig. 2 that also shows the resulting final grading curve. The final mix proportions for all types of studied eco-concrete are shown in Table 2 including hydraulic-lime percentage in the binder (p). 3.3. Methodology The developed experimental program was designed to assess the effect of ECR addition to hydraulic-lime concrete and its behaviour under different curing conditions. Three different curing conditions at atmospheric pressure were used where the relative humidity, RH, of the environment was changed. Concrete specimens were cured immersed in water, at 95% RH and at 65% RH. The temperature was kept constant at 20 °C. Fresh properties of concrete such as density and workability were measured. Workability of fresh concrete was measured using
3. Materials and methodology 3.1. Materials All the materials used in this study were commonly available in the central region of Portugal. The binder used was hydraulic lime (NHL5) that is currently the only hydraulic lime produced in Portugal. Additionally, its compressive strength of 5 MPa at the age of 28 days is sufficient to be improved by the addition of a pozzolan, allowing the achievement of concrete with strength in the range 15–20 MPa at age 90 days. Concrete with this strength could be
Fig. 2. Aggregate grading curves.
Table 2 Mixture composition including hydraulic-lime percentage in the binder (p) Name
M0 M2 M3
Fig. 1. XRD of expanded clay residue.
Constituents (kg/m3)
p (%)
CA1
CA2
Sand
NHL5
ECR
Water
619 619 619
455 455 455
321 321 321
550 440 385
– 110 165
247.5 247.5 247.5
100 80 70
Table 1 Chemical composition of expanded clay residue Oxides
CaO
SiO2
Al2O3
Fe2O3
K2O
MgO
Na2O
TiO2
P2O5
MnO
LOI
Percentage in weight
3.92
56.52
19.50
8.05
4.58
3.97
0.33
0.95
0.18
0.14
0.70
A.L. Velosa, P.B. Cachim / Construction and Building Materials 23 (2009) 2107–2111
the slump test (NP EN 12350-2:2002). Hardened concrete properties investigated were compressive strength (tests on 15 cm cubes were performed at 7, 28 and 90 days according NP EN 123903:2003 standard), tensile splitting strength (tests performed on 15 cm diameter cylinders according to NP EN 12390-3:2003). Pozzolanic strength indexes were calculated following the work of Yu et al. [5]. The specific strength ratio, R, is defined as contributions to concrete strength from unit hydraulic lime and unit mineral admixture and is defined by
R¼
fc p
ð1Þ
where fc is the concrete compressive strength and p is the cement or mineral admixture percentage of the cementitious materials (see Table 2). RC expresses the contribution of unit cement to concrete strength without any mineral admixture, RM expresses the contribution of unit mineral admixture to concrete strength, and RP is the contribution of the pozzolanic effect to concrete strength due to mineral admixture, expressed by
RP ¼ R M RC
ð2Þ
The index of specific strength, K, is the ratio of RM to RC, and the contribution of pozzolanic effect to strength, P, can be assessed by the percentage value of the contribution of pozzolanic effect to concrete strength, which can be written as
P ¼ 100ðRP =RM Þ
ð3Þ
The pozzolanic index, P, relates to the favourable contribution of the pozzolan relatively to the usual binder. 4. Results and discussion 4.1. Fresh concrete Slump results are 1.5 cm, 0.9 cm and 0.5 cm, respectively, for series M0, M2 and M3 (class S1, following NP EN 206-1:2007) for all mixes with and without substitution of hydraulic lime, that indicates a rather small workability of concrete. Since the aim of this study was the investigation of the effect of expanded clay residue and of curing conditions, this lack of workability wasn’t considered a very important issue at this stage. Nevertheless, this is an important aspect that must be accounted for in the subsequent development of the study and that may be solved by the addition of a plasticizer. Density of fresh concrete was approximately constant at 2290 kg/m3. 4.2. Hardened concrete 4.2.1. Strength The results obtained for compressive strength at 7, 28 and 90 days are shown in Table 3 together with the corresponding stan-
dard deviation. It is apparent from the Table 3 that reference limecrete, mixture M0, has higher strength until 28 days if cured at normal conditions (20 °C and 65% RH) than in a saturated environment. However, this improved strength at 65% RH is attenuated with time and at 90 days all the curing conditions lead to approximately the same strength. The presence of expanded clay residue alters this situation. As ECR is added to limecrete composition, the presence of water during curing becomes fundamental. For mixture M3 the effect of curing conditions is quite apparent, especially between 28 and 90 days, where nearly 100% increase in strength can be observed if specimens are cured in 95% RH or immersed. The results observed for mixture M2 show transition behaviour between mixtures M0 and M3, as can be easily seen in Table 3. Thus, it seems clear that when expanded clay residue is used, curing in 95% RH and in water is better. The addition of expanded clay residue only has a decisive beneficial effect on strength only at 90 days and in saturated curing conditions or under water, as both these curing conditions favour pozzolanic reaction. Fig. 3 shows results of concrete at age 28 days as a function of curing conditions. Although further studies with other pozzolanic materials must be executed in order to assess the influence of the addition of pozzolans in the strength development of limecrete, the behaviour of hydraulic-lime concrete with expanded clay residue is significantly improved in saturated conditions or under water when compared with limecrete with no pozzolanic additions in the same situation. The results obtained show that the influence of curing conditions at early ages is small but becomes evident at age 90 days: concretes with expanded clay residue perform better in saturated curing conditions or under water. This may be explained by water demand of the pozzolanic reaction, which is slower than hydraulic reaction but will only take place in the presence of available water. This slower reaction will only produce visible results at later ages, accounting for the differences between results of concretes stored in saturated conditions or under water at age 90 days. This effect can clearly be observed in Fig. 4 where the strength evolution in time, normalised by the respective 28 days strength is plotted. Results for tensile strength closely follow the compressive strength results. The relation between compressive and tensile splitting strength is shown in Fig. 5. The results are compared with the relation proposed by the European concrete design code, EC2 [6], that relates tensile and compressive strength by a power law
fct;sp ¼ 0:27fc2=3
ð4Þ
Comparison of EC2 equation with experimental data (see Fig. 5) shows that almost all the points are under the EC2 proposed curve given by Eq. (4). Because EC2 curve is intended to be for mean values, the tensile strength of limecrete seemed to be comparatively smaller than that of Portland cement concrete.
Table 3 Compressive strength and standard deviation at 7, 28 and 90 days Series
M0-65 M0-95 M0-W M2-65 M2-95 M2-W M3-65 M3-95 M3-W
Compressive strength (MPa)
Standard deviation (MPa)
7 days
28 days
90 days
7 days
28 days
90 days
6.6 5.0 3.3 6.0 5.5 4.0 4.9 4.6 4.9
8.9 8.3 5.6 9.4 11.1 8.8 8.6 10.8 9.6
10.2 11.1 10.5 13.2 18.4 14.5 11.5 26.6 18.5
0.16 0.24 0.15 0.40 0.14 0.09 0.26 0.21 0.30
0.21 0.57 0.21 0.47 0.29 0.08 0.88 0.10 0.33
0.81 0.27 0.01 0.78 0.53 0.56 0.75 0.92 0.56
2109
Fig. 3. Compressive strength at 28 days.
2110
A.L. Velosa, P.B. Cachim / Construction and Building Materials 23 (2009) 2107–2111 Table 4 Pozzolanic indexes RM, RP, K and P
Fig. 4. Strength evolution in time normalised by the respective 28 days strength.
Days of curing
Mixture
7
M0-65 M0-95 M0-W M2-65 M2-95 M2-W M3-65 M3-95 M3-W
28
90
fc (MPa)
RM
RP
K
P
6.6 5.0 3.3 6.0 5.5 4.0 4.9 4.6 4.9
0.066 0.050 0.033 0.075 0.068 0.050 0.070 0.066 0.070
0.0 0.0 0.0 0.0098 0.0187 0.0174 0.0044 0.0164 0.0370
1.0 1.0 1.0 1.15 1.38 1.53 1.07 1.33 2.13
0.0 0.0 0.0 13.0 27.3 34.8 6.3 24.9 53.1
M0-65 M0-95 M0-W M2-65 M2-95 M2-W M3-65 M3-95 M3-W
8.9 8.3 5.6 9.4 11.1 8.8 8.6 10.8 9.6
0.089 0.083 0.056 0.118 0.138 0.110 0.122 0.154 0.138
0.0 0.0 0.0 0.0292 0.0557 0.0541 0.0338 0.0716 0.0813
1.0 1.0 1.00 1.33 1.67 1.96 1.38 1.87 2.44
0.0 0.0 0.0 24.8 40.2 49.0 27.6 46.4 59.1
M0-65 M0-95 M0-W M2-65 M2-95 M2-W M3-65 M3-95 M3-W
10.2 11.1 10.5 13.2 18.4 14.5 11.5 26.6 18.5
0.102 0.111 0.105 0.165 0.230 0.181 0.165 0.380 0.264
0.0 0.0 0.0 0.0630 0.1189 0.0760 0.0628 0.2693 0.1590
1.0 1.0 1.0 1.62 2.07 1.72 1.62 3.43 2.51
0.0 0.0 0.0 38.2 51.8 41.9 38.1 70.9 60.2
Fig. 5. Relation between compressive and tensile splitting strength, compared with EC2 curve.
A regression equation was fitted to experimental data that is given by Eq. (5) with a correlation coefficient r2 = 0.735 (also shown in Fig. 5).
fct;sp ¼ 0:154fc0:74
ð5Þ
The ratio tensile strength/compressive strength is lower for this kind of concrete in relation to ordinary Portland cement concrete and a possible explanation for this is the amount of water that was used, with no aid of super plasticizers. Although this may indicate a greater cracking susceptibility, no cracking was observed in the samples. 4.2.2. Pozzolanic index Pozzolanic indexes RM, RP, K and P are shown in Table 4, for each mixture, together with the compressive strength achieved in every case, at ages 7, 28 and 90 days. At the age of 7 days there is a decrease in P when pozzolanic substitution increases from 20% to 30%, except for the case of water immersion. This behaviour differs from that of latter ages. For this reason, it is not possible to conclude on the behaviour of limecrete with easy age testing. For ages over 28 days and adequate curing conditions (95% RH and water immersion) the variation in pozzolan substitution from 20% to 30% implies a greater contribution of this material towards final strength (Fig. 6).
Fig. 6. Comparison of pozzolanic index P between 20% and 30% of cement substitution with different curing conditions and at various ages (7, 28 and 90 days).
Pozzolanic index P, related to the same substitution percentage and the same curing conditions always increases in time, with the sole exception of 20% ECR substitution, water immersed, between 28 and 90 days (Fig. 6). Conditioning at 95% RH or 100% (in water) positively influences the pozzolanic index, as expected. Water conditioning is generally more favourable at early ages. In the long run, as can be seen by results at 90 days, for pozzolanic substitution percentages of 20% and 30%, conditioning at high RH favours the pozzolanic index. Only with a 95% RH conditioning there is a continuously positive contribution of the pozzolanic addition, both in terms of substitution percentage and in terms of long term behaviour. In the case of 65% RH, the increase in pozzolan percentage from 20% to 30% does not produce a significant or consistent effect in
A.L. Velosa, P.B. Cachim / Construction and Building Materials 23 (2009) 2107–2111
2111
Fig. 7. Pozzolanic index, P.
terms of achieved strength, Table 3, and in terms of pozzolanic reaction contribution (Figs. 6 and 7). The contribution of the pozzolan when specimens are immersed in water is also positive but not as effective as 95% RH conditions. The pozzolanic index P is very sensitive to curing conditions. The performance of concretes with hydraulic lime and pozzolanic additions will improve in high RH conditions.
5. Conclusions Expanded clay residue is a suitable material for application in concrete with hydraulic-lime binder as a pozzolanic addition. A substitution of hydraulic lime by 20% and 30% of expanded clay residue produced increased compressive strength at adequate curing conditions, especially in the latter case. Therefore, these replacement percentages are satisfactory for current use. As results indicate, curing conditions produce significant changes in mechanical strength, but not at early ages, due to the development of chemical reactions in time. High relative humidity conditions are generally more favourable towards the increase in mechanical strength. Testing at early ages (7 days) does not allow attaining knowledge on future behaviour of limecrete, contrarily to current cement concrete. In fact testing procedures for current Portland cement concrete may not be applicable as hydration reaction is faster than carbonation and pozzolanic reactions that also contribute towards the hardening of hydraulic-lime based concrete.
The use of both the pozzolanic index P and mechanical strength results and their combined analyses, shows the real contribution of the pozzolanic material towards the mechanical properties of limecrete. Hydraulic-lime concrete was developed on this study without the aid of super plasticizers and therefore with a great amount of water. This was done in order to verify changes in behaviour and the material’s basic potential. However, the addition of super plasticizers may improve characteristics significantly and therefore enlarge the application possibilities for this material. This paper opens the possibility towards the use of hydraulic lime in concretes in which compressive strength needs are not very high, promoting the use of a more sustainable material. Concretes based on Portland cement will provide higher strength and have different application aims. Alternative pozzolanic materials to expanded clay residue may be tested in a similar fashion. References [1] Rehan R, Nehdi M. Carbon dioxide emissions and climate change: policy implications for the cement industry. Environ Sci Policy 2005;8(2):105–14. [2] Szabó L, Hidalgo I, Ciscar JC, Soria A. CO2 emission trading within the European Union and Annex B countries: the cement industry case. Energy Policy 2006;34(1):72–87. [3] Gäbel K, Tillman A. Simulating operational alternatives for future cement production. J Clean Prod 2005;13(13/14):1246–57. [4] Gartner E. Industrially interesting approaches to ‘‘low-CO2” cements. Cement Concrete Res 2004;34:1489–98. [5] Yu LH, Ou H, et al. Investigation on pozzolanic effect of perlite powder in concrete. Cement Concrete Res 2003;33(1):73–6. [6] CEN. Eurocode 2: design of concrete structures – Part 1-1: general rules and rules for buildings. Brussels: CEN; 2004.