- Email: [email protected]
Cementing efficiency of electric arc furnace dust in mortars
Construction and Building Materials 157 (2017) 141–150
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Cementing efficiency of electric arc furnace dust in mortars Margareth da Silva Magalhães a, Flora Faleschini b,⇑, Carlo Pellegrino b, Katya Brunelli c a
Dept. of Civil Construction and Transport, Universidade do Estado do Rio de Janeiro, Rua São Francisco Xavier, 524, 5005-A, CEP 20550-900 Rio de Janeiro, RJ, Brazil Dept. of Civil, Environmental and Architectural Engineering, University of Padua, Via Marzolo 9, 35131 Padua, Italy c Dept. of Industrial Engineering, University of Padua, Via Marzolo 9, 35131 Padua, Italy b
h i g h l i g h t s Electric arc furnace dust is used as SCM until 10% of replacement ratio. Cementing efficiency of EAFD is estimated for varying w/b ratios and dust content. Flexural and compressive strength were taken into account to asses EAFD k-value. EAFD efficiency depends also on curing time.
a r t i c l e
i n f o
Article history: Received 19 June 2017 Received in revised form 1 September 2017 Accepted 14 September 2017
Keywords: Blended cement Cementing efficiency Compressive strength Electric arc furnace dust k-Value Supplementary cementing materials
a b s t r a c t The aim of this study is to evaluate the cementing efficiency (k-value) of an electric arc furnace dust (EAFD), used in as-received conditions, as a supplementary cementing material (SCM). The investigated variables are the amount of EAFD in the mix, water/binder ratio and curing age. Fifteen mixtures are manufactured to estimate the k-values, both considering compressive and flexural strength. In both cases, k-values depend on replacement percentage, and decrease raising the EAFD content in the mix. Water/ binder ratio and hydration time significantly influence EAFD efficiency, too. Additionally, results show that flexural k-values are greater than compressive k-values in the analyzed mortars. Ó 2017 Elsevier Ltd. All rights reserved.
1. Introduction Research about mineral additions, such as pozzolana, limestone, fly ash, silica fume and blast furnace slag as supplementary cementing materials (SCMs) has been transferred and implemented in several standards, which regulate the use of such materials in concrete [1,2]. The benefit of using blended cements, besides CO2 emissions abatement and resources conservation, is that the additions improve certain properties, such as workability, strength development and durability [3]. Accordingly, time-variant structural reliability of aging reinforced concrete elements subject to environmental deterioration can also be improved [4]. In recent years, many studies attempted to develop new SCMs, aiming to solve environmental problems linked to the disposal of solid waste. The re-use in construction industry is seen as an ⇑ Corresponding author. E-mail addresses: [email protected] (M. da Silva Magalhães), flora. [email protected] (F. Faleschini), [email protected] (C. Pellegrino), [email protected] (K. Brunelli). http://dx.doi.org/10.1016/j.conbuildmat.2017.09.074 0950-0618/Ó 2017 Elsevier Ltd. All rights reserved.
attractive alternative to landfill, because it allows to encapsulate potential toxic elements in a stable matrix. This is the case of electric arc furnace dust (EAFD), a metal-rich waste formed during steel-making process in the electric arc furnaces. This solid waste has high density, and it is typically constituted by fine particles, with a heterogeneous distribution [5,6]. Apart from iron, EAFD contains high quantity of zinc [7,8], and in lower amount, also other heavy metals, e.g. cadmium, lead, and chromium [9–11]. Due to its chemical composition, many works studied already immobilization techniques of EAFD heavy metals [12], stabilization in cementitious matrixes [13,14] and leachability of harmful compounds [15,16]. EAFD has been treated and used in different ways, such as: source for the recovery of valuable metals [17–19], additive to ceramic industry [20,21], in asphalt cement [22], in geopolymers [23,24] and polymer composites [25], or addition to cement paste, mortar and concrete [26–30]. When EAFD is used in cement-based materials, two main effects on the blended paste are observed during the hydration. First, the dust reacts as a set retarder due to the
142
M. da Silva Magalhães et al. / Construction and Building Materials 157 (2017) 141–150
presence of zinc oxide [26–32]; indeed, it is well-recognized that Zn2+ delays the early hydration of C3S. Then, in a longer time, the hydration degree is promoted, due to several concurring causes, including the increase of pH and the presence of some metals, such as Cu2+, Pb2+ and Cr3+. The former allows calcium zincate dissolution, and the latter may ease C3S hydration [33]. Some studies reported that replacing cement with EAFD until 5% does not worsen concrete or mortar mechanical strength [26,33,34]. However, when the substitution ratio raises up to 15%, a reduction of about 20% in compressive strength is noticed [34]. Furthermore, several works demonstrated how benefits could be obtained in terms of durability-related properties [9,11,28], and on slump retention [9], when EAFD is used in limited quantity. The Abrams law, which was originally formulated for proportioning concretes with cement as the only binder, and valid also to design mortars [35], is not directly applicable to this new generation of concretes. This formula cannot directly capture the influence of chemical and mineral admixtures currently used to design concretes, which may influence mechanical strength, at the same level than water/cement (w/c) ratio does. If the total amount of binder, being cement plus the SCM, simply substitutes the amount of cement, the Abrams law provides an incorrect estimate of concrete compressive strength. This occurs because the cementing efficiency of alternative binders might be not the same as that of cement, and additionally it is influenced by different variables, e.g. curing conditions, hydration time, type of cement, strength grade and quantity of the admixture itself [36,37]. Many studies were carried out to quantify the so-called cementing efficiency, or k-value, for different types of mineral admixtures, to assess how they affect concrete compressive strength [38–44]. According to Papadakis and Tsimas [38], the kvalue is defined as the part of SCM which can be considered as equivalent to Portland cement in a pozzolanic concrete, having the same properties under consideration. Hence, the efficiency factor corresponds to the number of parts of cement that may be replaced by one part of SCM, without changing the property under investigation (e.g. concrete compressive strength) [37]. The quantity of SCM can be multiplied by the k-value to estimate the equivalent cement content in the mixture, which should be added to the cement amount for calculating the w/c ratio. Obviously, if k is equal to 1, the SCM has the same efficiency as cement. It is worth to recall that the k-value can be applied to many properties, other than compressive strength. Indeed, many researchers have already derived k-values related to costs, durability, maturity, lime consumption and workability of SCMs [36,37,40–46]. Results demonstrated that, generally, the k-value resulting from compressive strength cannot be considered as a proxy-criterion for other properties, unless otherwise demonstrated. Mineral addition may have different effects on the above-mentioned properties, which should be separately analyzed. Several research works investigated already cementing efficiency of fly ash, silica fume and blast furnace slag, used at varying dosages. However, no studies were carried out to assess EAFD kvalue yet. Accordingly, in this research, the efficiency of an EAFD, used as SCM, on the strength of mortar mixtures is estimated. The analyzed variables are the amount of binder replacement ratio (5% and 10%), the water/binder (w/b) ratio (0.35, 0.4, 0.5, 0.6 and 0.7) and the curing age (7 and 28 days). In particular, k-values are evaluated considering both compressive and flexural strength, aiming to investigate whether compressive strength can be considered as a proxy criterion for flexural properties. 2. Background about cementing efficiency models for SCMs The concept of the cementing efficiency factor, or k-value, was initially introduced by Smith in 1967 [47] for a fly ash concrete,
defining that a mass of fly ash f can be considered as equivalent to a mass of cement kf, for its ability in influencing the strength development of the concrete. The rational model by Smith assumes that the strength and workability of two concretes with the same effective water/binder ratio, should be equal. In other words, if w/(c + kf) in a fly ash concrete is equal to w/c of the reference mix, the two mixtures should have the same strength. In general, the k-value approach is based on the assumption that a relationship exists between the tested property (typically, 28-days compressive strength) and a compositional parameter, which is the w/c ratio, if Abram’s law [44] or Bolomey’s equation [48] are considered for the reference Portland cement concrete. However, also other approaches could be applied, e.g. comparing the strength of two mixes with the same workability [49–51]. For instance, Schiessl and Hardtl [49] estimated the difference between the w/c ratio of a reference concrete, and the water/binder content w/(c + kf) of a fly ash concrete, necessary to achieve the same strength, and named this as the Dw concept. Bijen and Van Selst [50] used the same mathematical approach of Smith, and they found that the k-value depends on w/c ratio, cement type, fly ash quality, and concrete age. They reported that the cementing efficiency value of fly ash tends to decrease, as the w/c ratio increases. This dependence was higher in rapid setting Portland cement than in Portland blast furnace slag cement. Lastly, Babu et al. [51] separated the efficiency factor into two parameters, the general efficiency factor ke, and the percentage efficiency factor kp. The ke value is assumed constant for each SCM, whereas the kp value depends on the amount of the SCM inside the mix. The overall efficiency factor k is calculated by multiplying these two factors. Babu used the Dw concept [49] to assess the efficiency of concretes with pozzolana [51], fly ash [39], silica fume [52] and ground granulated blast furnace slag [53]. Those studies revealed that the overall efficiency factor may change with type and amount of cement, age and concrete curing conditions. Hassaballah and Wenzel [54] proposed a method based on a comparison of the compressive strength between a fly ash concrete and a control, with the same workability. Fly ash contribution to compressive strength was calculated as the strength difference between the two mixtures, thus assessing the pozzolanic efficiency factor. According to this method, positive values of k indicate a strength increase, while negative values indicate a strength loss. Wong and Razak [41] used an alternative approach to determine the efficiency of calcined kaolin and silica fume, calculating k as (RsC C0 P). Rs is the relative compressive strength between the concrete with and without SCM; C is the cement quantity in the reference concrete; C’ is the amount of cement in the SCM concrete; and lastly, P is the amount of the mineral addition. 3. Experimental program 3.1. Materials The materials used to prepare the mortar mixes are: Portland cement; electric arc furnace dust (EAFD); sand; and tap water. The cement used is a rapid setting ordinary Portland cement type I 52.5R, as defined in BS-EN 197-1 [2], with a density of 3080 kg/m3. EAFD is used in as-received condition from the dust collection system of a carbon steelmaking factory in Italy, and has a density of 3488 kg/m3. The natural sand has a 2 mm maximum particles size, apparent density of 2.64 g/ cm3 and saturated surface-dried (s.s.d.) density of 2.76 g/cm3. Fig. 1 shows particles size distribution of cement, EAFD and sand, obtained using laser diffraction technique. EAFD has a wide range of particle sizes, with coarse particles having a diameter up to 478 lm. The characteristics of the distribution are: d10 equal to 2.57 lm; median particle size d50 of 10 lm and d90 is 67.58 lm. Cement is characterized instead by d10 = 1.35 lm, d50 = 10.68 lm and d90 = 32.75 lm, being finer than the EAFD used in this work. Fig. 2 shows instead a scanning electron microscope (SEM) image of EAFD, taken in the secondary electron mode; it is possible to observe the dust morphology, made by ultrafine, spherically shaped particles. Table 1 shows the chemical composition of Portland cement and EAFD, obtained with X-ray Fluorescence (XRF) analyses. As it is possible to see, the composition of this dust is rich in Zn, Fe and Ca oxides, with few quantity of Si, Mg, Mn, Pb and Cl
143
M. da Silva Magalhães et al. / Construction and Building Materials 157 (2017) 141–150
Table 1 Major chemical compounds in ordinary Portland cement and EAFD (assessed through XRF analysis).
Fig. 1. Particles size distribution of cement, EAFD and sand.
Compound
Cement (wt%)
EAFD (wt%)
SiO2 Al2O3 Fe2O3 CaO SO3 MgO K2O MnO Na2O ZnO PbO2 CdO NiO Cr2O3 CuO Cl
21.61 5.77 2.67 61.39 6.11 2.58 0.9796 0.064 2.00 – – – – – – –
3.89 – 35.92 13.32 1.07 2.52 – 3.92 – 31.34 1.72 0.09 0.44 0.85 0.26 2.06
oxides. Also limited amount of harmful heavy-metals is present, such as Cr, Cd and Ni. The X-ray Diffraction (XRD) pattern, shown in Fig. 3 and obtained on a representative dust sample, highlights the presence of five main phases, which are magnetite (Fe3O4), franklinite (ZnFe2O4), zincite (ZnO), calcium ferrite (CaFe2O4) and lime (CaO).
evaluated for compressive and flexural strength, taking into account EAFD amount in the mixes, and the range of w/b ratios considered, between 0.35 and 0.7.
3.2. Mixture details
4.1. Compressive strength evolution
In this research, fifteen mortar mixtures are casted, with varying water/binder (w/b) ratios, between 0.35 and 0.7. Five mortars are without EAFD (labelled as Ref), five mortars contain 5% of EAFD (labelled as M01) as SCM, and lastly five mortars contain 10% of EAFD as cement substitute (labelled as M02). Mixture proportions are listed in Table 2. Binder to sand ratio is maintained constant in all the mixes, and is equal to 1:3. Substitution ratios are expressed in weight percentage. A superplasticizer based on sulfonated naphthalene is added to some mixtures, to allow a good workability for all the mixtures, including the ones with the lowest w/b ratio. In all the cases, flow diameter value ranges between 155 mm and 267 mm. All the mortars specimens are casted in steel molds, and after 24 h from the mixing they are demolded and cured until testing in 100% relative humidity (RH) conditions, at 20 ± 2 °C temperature. 3.3. Experimental methods Flexural and compressive strength of hardened mortar specimens are determined according to [55]. For the former, a 3PBT (three point bending test) is carried out on 160 40 40 mm prismatic specimens; for the latter, compressive strength is determined on cubic specimens with 40 mm sides. Compressive and flexural tests are carried out at the ages of 7 and 28 days.
4. Experimental results and discussion In this research work, a practical approach to assess the effect of EAFD on the strength of Portland cement-based mortars is used, applying the concept of SCM efficiency factor. Hence, k-values are
Fig. 4 shows the dependence of compressive strength on the amount of EAFD and on w/b ratio, both at 7 and 28 days. Compressive strength of the tested mortars depends on the w/b ratio, following the Abrams’ law. According to Neville [3], concrete strength at a given age depends mainly on w/c ratio, as well as on the degree of compaction. Here, the use of the superplasticizer in the mixtures with the lowest water amount was fundamental to ensure that specimens could be considered as fully-compacted, similarly to the other mixtures. Looking at Fig. 4, it is possible to observe that the increase of w/b from 0.35 to 0.7 lowers compressive strength of both reference and EAFD mortars (M01 and M02). The decrease ranges between 34% and 37% at 7 days, and between 33% and 34% at 28 days of curing. The porosity of the mortars increases in the hardened state with the w/b ratio, thus it directly affects the mechanical strength of the mortars. Furthermore, data shown in Fig. 4 highlight that, in most the cases, compressive strength increases when 5% of EAFD is used as SCM, whereas it decreases at 10% of substitution, regardless the w/b ratio and the age of curing. The maximum observed strength gain in M01 mortars with 5% of EAFD is 5.2% at 7 days, and 2.8% at 28 days; conversely, the highest loss in M02 mortars with 10% of EAFD is 3% at 7 days, and 4.3% at 28 days. It is worth to note that even at 7 days of curing, the retardant effect in strength development, which is frequently reported when this dust is used [26], is not observed at the lowest EAFD dosage. Fig. 5 shows the regression equations obtained from the average values of compressive strength, shown in the previous Fig. 4. The relationships are derived to predict the 7 and 28 days compressive strength of mortars, as a function of w/b ratio, for each group of mortar. At this stage, these relationships are obtained without considering the EAFD efficiency factor. The regression equations with the highest correlation coefficients are power functions of w/b ratio (Eq. (1)):
rc ¼ g1 ðw=bÞg2
Fig. 2. SEM image of EAFD.
ð1Þ
where rc is the compressive strength of the mortars, w is the water content, b is the theoretical total binder content (c + EAFD), g1 and g2 are constant coefficients for a given age, and their value is shown in Table 3. R2 values are also included in Table 3 per each equation. It is worth to note that, excepting in one case, R2 values are at least
144
M. da Silva Magalhães et al. / Construction and Building Materials 157 (2017) 141–150
Fig. 3. XRD pattern of EAFD.
Table 2 Mixture proportions of mortars. Mix
Cement (%)
EAFD (%)
w/b
Ref-035 M01-035 M02-035 Ref-040 M01-040 M02-040 Ref-050 M01-050 M02-050 Ref-060 M01-060 M02-060 Ref-070 M01-070 M02-070
100 95 90 100 95 90 100 95 90 100 95 90 100 95 90
– 5 10 – 5 10 – 5 10 – 5 10 – 5 10
0.35 0.35 0.35 0.40 0.40 0.40 0.50 0.50 0.50 0.60 0.60 0.60 0.70 0.70 0.70
Quantities (kg/m3) Cement
EAFD
Sand
Water
SP
567.7 539.8 512.0 552.0 524.9 497.8 523.1 497.4 471.7 497.1 472.7 448.2 473.6 450.3 427.0
– 28.4 56.9 – 27.6 55.3 – 26.2 52.4 – 24.9 49.8 – 23.7 47.4
1703.0 1704.8 1706.6 1656.0 1657.7 1659.4 1569.3 1570.9 1572.5 1491.3 1492.7 1494.1 1420.7 1422.0 1423.3
198.7 198.9 199.1 220.8 221.0 221.3 261.6 261.8 262.1 298.3 298.5 298.8 331.5 331.8 332.1
6.81 6.82 6.83 5.52 5.53 5.53 – – – – – – – – –
EAFD: electric arc furnace dust; w/b: water/binder ratio; SP: superplasticizer.
Fig. 4. Experimental compressive strength of tested mortars, with varying w/b ratios, at 7 and 28 days (b: overall binder content).
equal to 0.98. This means that the proposed relationships based on Eq. (1) predict well the compressive strength of the mixes, both for reference and EAFD mortars. According to Hedegaard and Hansen [56], when dealing with cement-based materials, which are intrinsically heterogeneous materials, a coefficient of correlation R2 above 0.95 indicates a good correlation in a statistical regression analysis,
between a hypothetical model and corresponding experimental data. Bolomey’s Equation (Eq. (2)) is also used to estimate compressive strength from the knowledge of b/w ratio, as it is linear, and therefore easier to use in the practice:
rc ¼ Aðb=wÞ þ B
ð2Þ
145
M. da Silva Magalhães et al. / Construction and Building Materials 157 (2017) 141–150
Fig. 5. Relationships between compressive strength and w/b ratio for the tested mortars, at 7 and 28 days (b: overall binder content).
Table 3 Regression equation parameters for compressive strength, and R2 values. Mix
Power’s equation parameters 7 days
Ref. M01 M02
Bolomey’s equation parameters 28 days
7 days
28 days
g1
g2
R2
g1
g2
R2
A
B
R2
A
B
R2
27.47 28.13 26.99
0.64 0.63 0.65
0.982 0.994 0.996
35.09 35.46 33.98
0.55 0.55 0.56
0.985 0.985 0.975
13.99 13.51 13.63
14.68 16.25 14.80
0.986 0.993 0.996
13.83 14.01 13.49
23.15 23.49 22.73
0.986 0.983 0.967
where A and B are constant parameters, and their values are listed in Table 3, for each mortar group, with the corresponding R2 value. In this experimental campaign, the value of the parameter A, which is the slope of the curve defined by Eq. (2), is almost the same at the two considered ages; the parameter B, instead, varies considerably with age. However, their values do not change significantly between each mortar group. 4.2. EAFD efficiency factors for compressive strength In this research, the k-values of EAFD are estimated initially with the Dw concept, which is described in detail in the work conducted by Babu and Rao [39] and Schiessl and Hardtl [49]. This procedure attempts to bring the w/b ratio of the EAFD mortar close to the w/c ratio of the reference, through the application of EAFD kvalue at any strength. The principle of the calculation is illustrated in Fig. 6. The k-value is defined in such a way that the w/c ratio of the reference mortar, and the water to effective cementitious materials ratio [w/(c + kD)] of EAFD mix are the same, for the same strength level. Here, w, c and D are respectively the amount of water, cement and dust in the EAFD mortar. Comparing the curves shown in Fig. 6, Dw can be expressed by Eq. (3):
Dw ¼ ðw=cÞ ðw=bÞ
ð3Þ
In the above relation, the water to effective cementitious materials ratio w/(c + kD) should substitute the w/c term, and the water to overall binder content w/(c + D) substitutes w/b. Then, the kvalue can be calculated with Eq. (4):
w c k¼ w D D Dw þ ðcþDÞ
ð4Þ
In this research, a k-value that depends on EAFD quantity (percentage efficiency factor) is calculated, because the use of a single k-value did not lead to a good correlation for each replacement ratio of the SCM. Fig. 7 shows the variation of k-values as a function of w/b ratio, for M01 and M02 mortars, both at 7 and 28 days of curing. It is
Fig. 6. Principle for calculating the efficiency factor with Dw concept.
possible to observe how the w/b ratio affects the k-values. For M01, k-values are always higher than 1, regardless the w/b ratio used, indicating that EAFD has a higher cementing efficiency than cement, when it is used as a replacement ratio of 5%. However, in M02, when EAFD is used at 10% replacement ratio, it is less efficient than cement in terms of compressive strength contribution, because k-values are less than 1, also in this case regardless the w/b ratio. In this study, however, the w/b ratio plays a very important role in affecting efficiency values. Results indicate that, when w/b ratio changes from 0.35 to 0.70, the k-values calculated for M01 mortars decrease from 1.40 to 1.20 (D = 14.5%) at 7 days, and from 1.62 to 1.30 (D = 19.4%) at 28 days. Conversely, this trend changes for the M02 mortars; when the w/b ratio changes from 0.35 to 0.70, the k-values increase from 0.72 to 0.86 (D = +19%) at 7 days, and from 0.61 to 0.80 (D = +31.3%) at 28 days. This indicates that, when higher quantity of EAFD is used, more water is needed to increase the cementing efficiency of this dust. This finding is consistent with the work conducted by Bijen and Van Selst [50].
146
M. da Silva Magalhães et al. / Construction and Building Materials 157 (2017) 141–150
Fig. 7. EAFD k-values calculated with the Dw concept, as a function of the w/b ratio at 7 and 28 days, for compressive strength (b: overall binder content).
Considering the k-values for different w/b ratios in Fig. 7, the average k-values for 5% and 10% EAFD at 7 and 28 days of curing are calculated and listed in Table 4. Standard deviation (SD) is also reported in parenthesis. For M01, the average k-values increase with mortar age. However, this effect is not observed in M02; the average values of k decrease from about 1.29 to 0.80 at 7 days, and from 1.45 to 0.71 at 28 days, when the replacement percentage varies from 5% to 10% of EAFD. Accordingly, the efficiency of EAFD is reduced with the percentage replacement. This is consistent with the results shown in Fig. 5. The k-values evaluated for this EAFD are comparable, or even higher, than the efficiency factor of high-calcium fly ash (k: 0.9 with SCM content of 10–20% [48]) and blast furnace slag (k: 0.69 with SCM content of 70% [44]) at 28 days, but lower than silica fume efficiency (k: 2.1–3.1 with SCM content of 5–15% [41]). It is important to highlight that this comparison is performancebased only, because the efficiency of SCM on mortars strength varies significantly with their properties, content inside the mixture,
and tests conditions, i.e. curing age, type of cement and strength grade [36,37]. Lastly, considering the calculated average k-values for each EAFD replacement ratio, the variation of compressive strength with water to effective cementitious materials ratio w/(c + kD) is represented in Fig. 8. This figure clearly shows that the predictions present an excellent agreement with the experimental results, and the R2 value is 0.99 at 7 days, and 0.981 at 28 days. Those values are similar or even better than ones obtained for the reference mortar, listed in Table 3 (R2 = 0.982 at 7 days and R2 = 0.985 at 28 days). The values of the coefficients g1 and g2 are similar to those obtained for reference mortar. An alternative approach is used in this research for the evaluation of EAFD cementing efficiency, which has been used in many other works in literature [44,48]. This simple methodology uses the same equation of the Portland cement mortar (reference), estimated through the empirical Eq. (1), and the experimentally measured values of compressive strength of the EAFD mortars, given in Fig. 4. The advantage of this method is that only two mixtures are required to determine the k-value for a specific SCM. In the case of mortars produced with EAFD, and considering the concept of kvalue, Eq. (1) can be re-written as:
rc ¼ n1
w cþkD
n2
ð5Þ
where g1 and g2 are the same parameters estimated in the reference mortar, and reported in Table 3. Using this equation, with the experimental values of compressive strength of the EAFD mortars, and knowing the quantities of w, c and D (given in Table 2), the k-values can be easily calculated. Results are shown in Table 4, and labelled as ‘‘alternative approach”, to be compared with the ones derived with the Dw concept. In general, the average k-values calculated with the two methods are similar; however, the SD values are higher with the alternative approach than with the Dw concept. This is due to the different w/b ratios used during the experimental campaign to cast the specimens; the clear trend about k-value dependency on w/b
Table 4 Average efficiency factors (k-values) calculated with the Dw concept and an alternative approach.
Dw concept
Mortar property
Compressive strength, 7 days Compressive strength, 28 days Flexural strength, 7 days Flexural strength, 28 days
Alternative approach
5% EAFD
10% EAFD
5% EAFD
10% EAFD
1.29 1.45 1.87 2.28
0.80 0.71 0.71 0.85
1.44 1.50 2.00 2.21
0.79 0.60 0.62 0.80
(0.08) (0.13) (0.22) (0.28)
(0.06) (0.08) (0.08) (0.04)
(0.46) (0.73) (0.58) (1.04)
Fig. 8. Relationships between compressive strength and w/(c + kD) ratio for the tested mortars, at 7 and 28 days.
(0.18) (0.45) (0.25) (0.82)
M. da Silva Magalhães et al. / Construction and Building Materials 157 (2017) 141–150
147
Fig. 9. Experimental flexural strength of tested mortars, with varying w/b ratios, at 7 and 28 days (b: overall binder content).
emphasize that, when this method is used, the k-value calculated with one only w/b ratio cannot represent the most appropriate kvalue. 4.3. Flexural strength evolution
Fig. 10. Relation between flexural and compressive strength, at different ages.
ratio, observed applying the Dw concept method, cannot be captured with the alternative approach. For instance, in M01 mortars, the k-values at 7 days are 1.11 (w/b: 0.35), 1.62 (w/b: 0.4), 1.69 (w/ b: 0.5), 0.84 (w/b: 0.6) and 1.96 (w/b: 0.7), and at 28 days they are 2.16 (w/b: 0.35), 0.47 (w/b: 0.4), 1.74 (w/b: 0.5), 2.09 (w/b: 0.6) and 1.04 (w/b: 0.7). Instead, for M02 mortars, the k-values at 7 days are 1.00 (w/b: 0.35), 0.59 (w/b: 0.4), 0.71 (w/b: 0.5), 0.97 (w/b: 0.6) and 0.69 (w/b: 0.7), and at 28 days they are 0.71 (w/b: 0.35), 0.17 (w/b: 0.4), 1.24 (w/b: 0.5), 0.72 (w/b: 0.6) and 0.15 (w/b: 0.7). These results indicate that there is a strong influence of the w/b ratio in this methodology of evaluation. Accordingly, it is important to
The 7 and 28 days-flexural strength values are presented graphically in Fig. 9. As in the case of compressive strength, and as expected, flexural strength values increase with the decrease of w/b ratio. Flexural strength values of reference, 5% EAFD (M01) and 10% EAFD (M02) mortars, cured for 7 days, increase from about 6 to 8.6 MPa, from 6.2 to 8.3 MPa and from 5.8 to 8.3 MPa, respectively, when the w/b ratio is reduced from 0.70 to 0.35. Increasing the curing time to 28 days, the strength of these mixtures further improves, until values of 9 MPa for the reference, 9.2 MPa for M01 and 8.9 MPa for M02 mortars. A slight positive effect is observed when 5% of EAFD is used; however, reduction of up to 4% on the flexural strength is noticed in 10% EAFD mortars. Fig. 10 indicates that the flexural to compressive strength ratio is not significantly influenced by the addition of EAFD inside the mixtures, and a linear relationship between these parameters can be obtained. However, at 28 days, the flexural strength increases more slowly than the compressive strength. This is consistent with the general tendency of this ratio to decrease, when compressive strength value increases [3]. Power empirical expressions (Eq. (6)) are derived from Fig. 11, to predict the flexural strength of the analyzed mixtures (reference, and mortars with 5% and 10% EAFD) at 7 and 28 days.
rf ¼ k1 ðw=bÞk2
Fig. 11. Relationships between 7 and 28 days flexural strength and w/b ratio for the tested mortars (b: overall binder content).
ð6Þ
148
M. da Silva Magalhães et al. / Construction and Building Materials 157 (2017) 141–150
Table 5 Regression equation parameters for flexural strength, and R2 values. Mix
Power’s equation parameters 7 days
Ref. M1 M2
Linear equation parameters 28 days
7 days
28 days
k1
k2
R2
k1
k2
R2
E
F
R2
E
R2
F
5.02 5.11 4.90
0.51 0.52 0.52
0.998 0.988 0.981
6.80 6.80 6.74
0.28 0.30 0.28
0.991 0.954 0.945
1.80 1.97 1.73
3.50 3.35 3.49
0.999 0.995 0.981
1.09 1.21 1.08
6.01 5.89 5.97
0.969 0.972 0.923
Fig. 12. EAFD k-values calculated with the Dw concept, as a function of the w/b ratio at 7 and 28 days, for flexural strength (b: overall binder content).
In Eq. (6), rf is the flexural strength of the mortars, w is the water content, b is the theoretical total binder content, k1 and k2 are constant parameters, to be derived from the experimental results, and which value is listed in Table 5. High correlation coefficients are obtained, as shown in Table 5. Also in this case, a linear expression (Eq. (7)) is developed to predict flexural strength of mortars; parameters E and F are respectively the slope and the intercept of the curves, and their value is listed in Table 5, with the resulting R2 values.
rf ¼ Eðb=wÞ þ F
ð7Þ
4.4. EAFD efficiency factors for flexural strength In this study, EAFD efficiency on the flexural strength is investigated, too. The k-values of EAFD considering flexural strength are
estimated with the same methods used for assessing compressive strength efficiency. Hence, the Dw concept methodology described in Section 4.2. and further illustrated in Fig. 6, is used also in this case, just considering the Dw obtained by comparing the curves in Fig. 11. The k-values are given in Fig. 12, considering both curing times, the two substitution ratios and the varying w/b ratios. The average k-values for M01 and M02 mortars, at 7 and 28 days of curing, are listed in Table 4. Standard deviation (SD) is also reported in parenthesis. Cementing efficiency values related to flexural strength are higher than k-values obtained for the compressive strength of EAFD mortars. As expected, and similarly to the case of compressive strength, the average k-values increase with the time, for both the replacement ratios, because of the strength gain due to aging. Concerning the influence of EAFD amount inside the mortars, the same trend seen for the compressive strength target property is obtained. The average k-value of M01 mortars is greater than 1 (1.87 at 7 days, and 2.28 at 28 days), whereas the average k-value of M02 mortars is lower than 1 (0.71 at 7 days, and 0.85 at 28 days). This indicates that the efficiency of EAFD on the flexural strength is reduced with the increase of the replacement ratio. It should be noted also that when w/b ratio is increased from 0.35 to 0.70, the k-values based on the flexural strength decrease from 2.33 to 1.64 at 7 days, and from 2.79 to 1.66 at 28 days, in M01 mortars (Fig. 12). However, M02 mortars display a reverse trend (Fig. 12). When the w/b ratio changes from 0.35 to 0.70, the k-values increase from 0.60 to 0.80 at 7 days, and from 0.79 to 0.90 at 28 days. Lastly, the variation of flexural strength with water to effective cementitious materials ratio w/(c + kD) is shown in Fig. 13. The prediction equations have high correlation coefficients (R2 = 0.982 and 0.973), and the coefficients k1 and k2 are similar to those obtained for the reference mortar (Table 5). These results confirm the reliability of the proposed method for the estimation of the k-values also from flexural tests.
Fig. 13. Relationships between flexural strength and w/(c + kD) ratio for the tested mortars, at 7 and 28 days.
M. da Silva Magalhães et al. / Construction and Building Materials 157 (2017) 141–150
The k-values are evaluated also with the alternative approach described in Section 4.2., which uses the power equation derived for the Portland cement mortar. Hence, k-values are estimated by the empirical Eq. (8):
rf ¼ k 1
k2 w cþkD
ð8Þ
149
Acknowledgments The authors acknowledge the Quality Improvement Program for Universities of the Ministry of Education and Culture (CAPES) for the financial support, Acciaierie Venete SpA and Italcementi SpA for supplying respectively the EAFD and the cement. References
where c, k and D were already defined, k1 and k2 are the parameters of the reference mortar listed in the Table 5. Using this equation, with the measured values of the flexural strength given in Fig. 9 and the w, c and D contents given in Table 2, the k-values for the EAFD used in this work in two replacement ratios can be derived, and their values are listed in Table 4. The average k-values calculated with this method are, generally, close to those estimated with the Dw concept method. However, the SD of the results for varying w/b ratios is noticeable higher than the corresponding SD value, obtained with Dw concept method. For instance, in the 5% EAFD mortar, the k-values at 7 days vary from null to 2.58, and at 28 days vary from 0.57 to 3.31. In 10% EAFD mortars the k-values at 7 days vary between 0.35 and 0.86, and at 28 days, vary between 0.25 and 1.74. Furthermore, also in this case the dependency of the k-value on the w/b ratio cannot be observed.
[1] [2] [3] [4]
[5]
[6]
[7] [8]
[9]
5. Conclusions [10]
The experimental campaign carried out in this work provides preliminary information about the efficiency of an EAFD on the strength of mortars, when it is used as a SCM. According to the experimental results observed in this study, the following conclusions about the cementing efficiency of EAFD can be derived. The cementing efficiency depends on the quantity of EAFD used. The cementing efficiency of EAFD at 5% replacement ratio is higher than when it is used at 10% replacement ratio. Additionally, at 5% replacement ratio, the k-value of EAFD exceeds the unit. The k-value calculated using only one water/binder ratio cannot represent the most appropriate k-value for EAFD, and more generally speaking, for a SCM. It is demonstrated that k-value depends on the w/b ratio used. In the specific case, when the w/b ratio increases, there is a decline in cementing efficiency for 5% EAFD; conversely, the trend is opposite when 10% EAFD is used. Cementing efficiency values raise with the increase of hydration time. Flexural strength k-values are greater than compressive strength k-values. This result demonstrates that EAFD addition influences more the flexural than the compressive strength. This research is a first attempt to investigate cementing efficiency of EAFD, thus further laboratory investigation is necessary to establish the reliability of the proposed method, particularly with respect to the design of blended cements with other cement types and EAFD amounts. A special care should be paid when managing such kind of solid waste, which contains substances potentially harmful for the environment; hence, future research lines will be focused also on the leachability of heavy metals from the cement-based materials where EAFD will be incorporated. About this topic, it is worth to mention that the authors are currently studying a hydrometallurgical pre-treatment to partially leach zincite, thus limiting the source of Zn2+ in the matrix.
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18] [19] [20]
[21]
[22]
[23]
[24]
[25]
[26]
ASTM C595 / C595M – 16, Standard Specification for Blended Hydraulic Cements, ASTM International, West Conshohocken, PA, 2016. EN 197-1. Cement Part 1: Composition, Specifications and Conformity Criteria for Common Cements, Brussels, Belgium (2011). A.M. Neville, Properties of Concrete, 5th edition., Pearson, United Kingdom, 2011. P. Castaldo, B. Palazzo, A. Mariniello, Effects of the axial force eccentricity on the time-variant structural reliability of aging r.c. cross-sections subjected to chloride-induced corrosion, Eng. Struct. 130 (2017) 261–274. J.G.M.S. Machado, F.A. Brehm, C.A.M. Moraes, C.A. Santos, A.C.F. Vilela, J.B.M. Cunha, Chemical, physical, structural and morphological characterization of the electric arc furnace dust, J. Hazard. Mater. 136 (3) (2006) 953–960. N. Menad, J.N. Ayala, F. Garcia-Carcedo, E. Ruiz-Ayúcar, A. Hernández, Study of the presence of fluorine in the recycled fractions during carbothermal treatment of EAF dust, Waste Manage. 23 (6) (2003) 483–491. G. Orhan, Leaching and cementation of heavy metals from electric arc furnace dust in alkaline medium, Hydrometallurgy 78 (2005) 236–245. G. Salihoglu, V. Pinarli, Steel foundry electric arc furnace dust management: stabilization by using lime and Portland cement, J. Hazard. Mater. 153 (2008) 1110–1116. M. Maslehuddin, F.R. Awan, M. Shameem, M. Ibrahim, M.R. Ali, Effect of arc furnace dust on the properties of OPC and blended cement concretes, Constr. Build. Mater. 25 (2011) 308–312. O. Ruiz, C. Clemente, M. Alonso, F.J. Alguacil, Recycling of an electric arc furnace flue dust to obtain high grade ZnO, J. Hazard. Mater. 141 (1) (2007) 33–36. R.Z. Al-Zaid, F.H. Al-Sugair, A.I. Al-Negheimish, Investigation of potential uses of electric-arc furnace dust (EAFD) in concrete, Cem. Concr. Res. 27 (2) (1997) 267–278. M.G. Mitrakas, C.A. Sikalidis, T.P. Karamanli, Immobilization of EAFD heavy metals using acidic materials, J. Environ. Sci. Health Part A Toxic/Hazard. Subst. Environ. Eng. 42 (4) (2007) 535–541. E.F. Ledesma, J.R. Jiménez, J. Ayuso, J.M. Fernández, J. de Brito, Experimental study of the mechanical stabilization of electric arc furnace dust using fluid cement mortars, J. Hazard. Mater. 326 (2017) 26–35. C.A.D. De Souza, A.T. Machado, L.R.P. De Andrade Lima, R.J.C. Cardoso, Stabilization of electric arc furnace dust in concrete, Mater. Res. 13 (4) (2010) 513–519. G. Laforest, J. Duchesne, Stabilization of electric arc furnace dust by the use of cementitious materials: ionic competition and long-term leachability, Cem. Concr. Res. 36 (9) (2006) 1628–1634. M.G. Sebag, C. Korzenowski, A.M. Bernandes, A.C. Vilela, Evaluation of environmental compatibility of EAFD using different leaching standards, J. Hazard. Mater. 166 (2–3) (2009) 670–675. M. Al-Harahsheh, M. Aljarrah, F. Rummanah, K. Abdel-Latif, S. Kingman, Leaching of valuable metals from electric arc furnace dust— Tetrabromobisphenol A pyrolysis residues, J. Anal. Appl. Pyrolysis 125 (2017) 50–60. K. Brunelli, M. Dabalà, Ultrasound effects on zinc recovery from EAF dust by sulfuric acid leaching, Int. J. Min. Metallurgy Mater. 22 (4) (2015) 353–362. X. Xuefeng, T. Yuhong, Application of electric arc furnace dust in cement production, Iron Steel (Peking) 33 (6) (1998) 61–64. C. Sikadilis, M. Mitrakas, Utilization of electric arc furnace dust as raw material for the production of ceramic and concrete building products, J. Environ. Sci. Health Part A 41 (9) (2007) 1943–1954. V.N. Stathopoulos, A. Papandreou, D. Kanellopoulou, C.J. Stournaras, Structural ceramics containing electric arc furnace dust, J. Hazard. Mater. 262 (2013) 91– 99. M.A.T. Alsheyab, T.S. Khedaywi, Effect of electric arc furnace dust (EAFD) on properties of asphalt cement mixture, Resour. Conserv. Recycl. 70 (2013) 38– 43. I. Nikolic´, D. Crossed, D. Signurovic, D. Blecic, R. Zejak, Karanovic Lj, S. Mitsche, V.R. Radmilovic, Geopolymerization of coal fly ash in the presence of electric arc furnace dust, Min. Eng. 49 (2013) 24–32. M.C. Arnold, A.S. de Vargas, Bianchini, Study of electric-arc furnace dust (EAFD) in fly ash and rice husk ash-based geopolymers, Adv. Powder Technol. 28 (9) (2017) 2023–2034. M. Niubó, A.I. Fernández, J.M. Chimenos, L. Haurie, A possible recycling method for high grade steels EAFD in polymer composites. J. Hazard. Mater. 171 (1–3) 1139–1144. M.S. Magalhães, F. Faleschini, C. Pellegrino, K. Brunelli, Effects of electric arc furnace dust (EAFD) addition on setting and strength evolutions of cement pastes and mortars, Eur. J. Environ. Civ. Eng. 1 (2017) 1–13.
150
M. da Silva Magalhães et al. / Construction and Building Materials 157 (2017) 141–150
[27] G. Fares, R.Z. Al-Zaid, A. Fauzi, A.M. Alhozaimy, A.I. Al-Negheimish, M.I. Khan, Performance of optimized electric arc furnace dust-based cementitious matrix compared to conventional supplementary cementitious materials, Constr. Build. Mater. 112 (1) (2016) 210–221. [28] F.M. Al Mutlaq, C.L. Page, Effect of electric arc furnace dust on susceptibility of steel to corrosion in chloride-contaminated concrete, Constr. Build. Mater. 39 (2013) 60–64. [29] Chen Qy, M. Tyrer, C.D. Hills, X.M. Yang, P. Carey, Immobilisation of heavy metal in cement-based solidification/stabilisation: a review, Waste Manage. 29 (1) (2009) 390–403. [30] G. Arliguie, J. Grandet, Influence de la composition d’un ciment Portland sur son hydratation em presence de zinc, Cem. Concr. Res. 20 (3) (1990) 517–524. [31] M. Murat, F. Sorrentino, Effect of large additions of Cd, Pb, Cr, Zn, to cement raw meal on the composition and the properties of the clinker and the cement, Cem. Concr. Res. 26 (1996) 377–385. [32] A. Balderas, H. Navarro, L.M. Flores-Velez, O. Dominuez, Properties of Portland cement pastes incorporating nanometer-sized franklinite particles obtained from electric arc furnace dust, J. Am. Ceram. Soc. 84 (12) (2001) 2909–2913. [33] Q.Y. Chen, C.D. Hills, M. Tyrer, I. Slipper, H.G. Shen, A. Brough, Characterisation of products of tricalcium silicate hydration in the presence of heavy metals, J. Hazard. Mater. 147 (3) (2007) 817–825. [34] A. Vargas, A.B. Masuero, A.C.F. Vilela, Investigations on the use of electric arc furnace dust (EAFD) in pozzolan-modified Portland cement I (MP) pastes, Cem. Concr. Res. 36 (10) (2006) 1833–1841. [35] G.A. Rao, Generalization of Abrams’ law for cement mortars, Cem. Concr. Res. 31 (2001) 495–502. [36] M.A. Sanjuán, A. Piñeiro, O. Rodríguez, Ground granulated blast furnace slag efficiency coefficient (k value) in concrete, Appl. Limits Mater. Constr. 61 (302) (2011) 303–313. [37] R. Siddique, M.I. Khan, Supplementary Cementing Materials, Springer Science & Business Media, 2011. pp. 288. [38] V.G. Papadakis, S. Tsimas, Greek supplementary cementing materials and their incorporation in concrete, Cem. Concr. Compos. 27 (2005) 223–230. [39] K.G. Babu, G.S.N. Rao, Efficiency of fly ash in concrete, Cem. Concr. Compos. 15 (4) (1993) 223–229. [40] V.G. Papadakis, Effect of supplementary cementing materials on concrete resistance against carbonation and chloride ingress, Cem. Concr. Res. 30 (2) (2000) 291–299. [41] H.S. Wong, Razak H. Abdul, Efficiency of calcined kaolin and silica fume as cement replacement material for strength performance, Cem. Concr. Res. 35 (2005) 696–702.
[42] E. Karim, E.H. Kadri, R. Siddique, Investigation of slag cement quality through the analysis of its efficiency coefficient, Eur. J. Environ. Civ. Eng. 15 (10) (2011) 1393–1411. [43] E. Gruyaert, M. Maes, N. De Belie, Performance of BFS concrete: k-value concept versus equivalent performance concept, Constr. Build. Mater. 47 (2013) 441–455. [44] F. Lollini, E. Redaelli, L. Bertolini, A study on the applicability of the efficiency factor of supplementary cementitious materials to durability properties, Constr. Build. Mater. 120 (2016) 284–292. [45] D.F. Aponte, M. Barra, E. Vazquez, Durability and cementing efficiency of fly ash in concretes, Constr. Build. Mater. 30 (2012) 537–546. [46] European Committee for Standardization, European Standard EN 206: Concrete-specification, performance, production and conformity. Brussels (2013). [47] I.A. Smith, Design of fly ash concretes, Proc. Inst. Civ. Eng. 36 (1967) 769–790. [48] V.G. Papadakis, S. Tsimas, Supplementary cementing materials in concrete, Part I Efficiency Des. Cem. Concr. Res. 32 (10) (2002) 1525–1532. [49] P. Schiessl, R. Hardtl, Efficiency of fly ash in concrete evaluation of ibac test results 1991, Technical Report of Institut fiir Bauforschung, RWTH, Aachen. [50] J. Bijen, R. Van Selst, Cement equivalence factors for fly ash, Cem. Concr. Res. 23 (1993) 1029–1039. [51] K.G. Babu, G.S.N. Rao, P.V.S. Prakash, Efficiency of pozzolans in cement composites, in: R.K. Dhir, M.R. Jones (Eds.), Proceedings, Concrete 2000 Economic and Durable Construction through Excellence, University of Scotland, Dundee, Design, Materials and Construction, Vol. 1, E & FN Spon, London, 1993, pp. 497–509. [52] K.G. Babu, P.V.S. Prakash, Efficiency of silica fume in concrete, Cem. Concr. Res. 25 (6) (1995) 1273–1283. [53] K.G. Babu, V.S.R. Kumar, Efficiency of GGBS in concrete, Cem. Concr. Res. 30 (2000) 1031–1036. [54] A. Hassaballah, T.H. Wenzel, A strength definition for the water to cementitious materials ratio, in: V.M. Malhotra (Ed.), Proceedings, ACI Fifth International Conference Fly Ash, Silica Fume and Natural Pozzolans in Concrete, ACI, Detroit, USA, 1995, pp. 417–437. [55] EN 1015-11, Methods of Test for Mortar for Masonry: Part 11: Determination of Flexure and Compressive Strength of Hardened Mortar, Comité Européen de Normalisation, Brussels, Belgium, 2007. [56] S.E. Hedegaard, T.C. Hansen, Modified water/cement ratio law for compressive strength of fly ash concretes, Mater. Struct. 25 (1992) 273–283.
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Cementing efficiency of electric arc furnace dust in mortars Margareth da Silva Magalhães a, Flora Faleschini b,⇑, Carlo Pellegrino b, Katya Brunelli c a
Dept. of Civil Construction and Transport, Universidade do Estado do Rio de Janeiro, Rua São Francisco Xavier, 524, 5005-A, CEP 20550-900 Rio de Janeiro, RJ, Brazil Dept. of Civil, Environmental and Architectural Engineering, University of Padua, Via Marzolo 9, 35131 Padua, Italy c Dept. of Industrial Engineering, University of Padua, Via Marzolo 9, 35131 Padua, Italy b
h i g h l i g h t s Electric arc furnace dust is used as SCM until 10% of replacement ratio. Cementing efficiency of EAFD is estimated for varying w/b ratios and dust content. Flexural and compressive strength were taken into account to asses EAFD k-value. EAFD efficiency depends also on curing time.
a r t i c l e
i n f o
Article history: Received 19 June 2017 Received in revised form 1 September 2017 Accepted 14 September 2017
Keywords: Blended cement Cementing efficiency Compressive strength Electric arc furnace dust k-Value Supplementary cementing materials
a b s t r a c t The aim of this study is to evaluate the cementing efficiency (k-value) of an electric arc furnace dust (EAFD), used in as-received conditions, as a supplementary cementing material (SCM). The investigated variables are the amount of EAFD in the mix, water/binder ratio and curing age. Fifteen mixtures are manufactured to estimate the k-values, both considering compressive and flexural strength. In both cases, k-values depend on replacement percentage, and decrease raising the EAFD content in the mix. Water/ binder ratio and hydration time significantly influence EAFD efficiency, too. Additionally, results show that flexural k-values are greater than compressive k-values in the analyzed mortars. Ó 2017 Elsevier Ltd. All rights reserved.
1. Introduction Research about mineral additions, such as pozzolana, limestone, fly ash, silica fume and blast furnace slag as supplementary cementing materials (SCMs) has been transferred and implemented in several standards, which regulate the use of such materials in concrete [1,2]. The benefit of using blended cements, besides CO2 emissions abatement and resources conservation, is that the additions improve certain properties, such as workability, strength development and durability [3]. Accordingly, time-variant structural reliability of aging reinforced concrete elements subject to environmental deterioration can also be improved [4]. In recent years, many studies attempted to develop new SCMs, aiming to solve environmental problems linked to the disposal of solid waste. The re-use in construction industry is seen as an ⇑ Corresponding author. E-mail addresses: [email protected] (M. da Silva Magalhães), flora. [email protected] (F. Faleschini), [email protected] (C. Pellegrino), [email protected] (K. Brunelli). http://dx.doi.org/10.1016/j.conbuildmat.2017.09.074 0950-0618/Ó 2017 Elsevier Ltd. All rights reserved.
attractive alternative to landfill, because it allows to encapsulate potential toxic elements in a stable matrix. This is the case of electric arc furnace dust (EAFD), a metal-rich waste formed during steel-making process in the electric arc furnaces. This solid waste has high density, and it is typically constituted by fine particles, with a heterogeneous distribution [5,6]. Apart from iron, EAFD contains high quantity of zinc [7,8], and in lower amount, also other heavy metals, e.g. cadmium, lead, and chromium [9–11]. Due to its chemical composition, many works studied already immobilization techniques of EAFD heavy metals [12], stabilization in cementitious matrixes [13,14] and leachability of harmful compounds [15,16]. EAFD has been treated and used in different ways, such as: source for the recovery of valuable metals [17–19], additive to ceramic industry [20,21], in asphalt cement [22], in geopolymers [23,24] and polymer composites [25], or addition to cement paste, mortar and concrete [26–30]. When EAFD is used in cement-based materials, two main effects on the blended paste are observed during the hydration. First, the dust reacts as a set retarder due to the
142
M. da Silva Magalhães et al. / Construction and Building Materials 157 (2017) 141–150
presence of zinc oxide [26–32]; indeed, it is well-recognized that Zn2+ delays the early hydration of C3S. Then, in a longer time, the hydration degree is promoted, due to several concurring causes, including the increase of pH and the presence of some metals, such as Cu2+, Pb2+ and Cr3+. The former allows calcium zincate dissolution, and the latter may ease C3S hydration [33]. Some studies reported that replacing cement with EAFD until 5% does not worsen concrete or mortar mechanical strength [26,33,34]. However, when the substitution ratio raises up to 15%, a reduction of about 20% in compressive strength is noticed [34]. Furthermore, several works demonstrated how benefits could be obtained in terms of durability-related properties [9,11,28], and on slump retention [9], when EAFD is used in limited quantity. The Abrams law, which was originally formulated for proportioning concretes with cement as the only binder, and valid also to design mortars [35], is not directly applicable to this new generation of concretes. This formula cannot directly capture the influence of chemical and mineral admixtures currently used to design concretes, which may influence mechanical strength, at the same level than water/cement (w/c) ratio does. If the total amount of binder, being cement plus the SCM, simply substitutes the amount of cement, the Abrams law provides an incorrect estimate of concrete compressive strength. This occurs because the cementing efficiency of alternative binders might be not the same as that of cement, and additionally it is influenced by different variables, e.g. curing conditions, hydration time, type of cement, strength grade and quantity of the admixture itself [36,37]. Many studies were carried out to quantify the so-called cementing efficiency, or k-value, for different types of mineral admixtures, to assess how they affect concrete compressive strength [38–44]. According to Papadakis and Tsimas [38], the kvalue is defined as the part of SCM which can be considered as equivalent to Portland cement in a pozzolanic concrete, having the same properties under consideration. Hence, the efficiency factor corresponds to the number of parts of cement that may be replaced by one part of SCM, without changing the property under investigation (e.g. concrete compressive strength) [37]. The quantity of SCM can be multiplied by the k-value to estimate the equivalent cement content in the mixture, which should be added to the cement amount for calculating the w/c ratio. Obviously, if k is equal to 1, the SCM has the same efficiency as cement. It is worth to recall that the k-value can be applied to many properties, other than compressive strength. Indeed, many researchers have already derived k-values related to costs, durability, maturity, lime consumption and workability of SCMs [36,37,40–46]. Results demonstrated that, generally, the k-value resulting from compressive strength cannot be considered as a proxy-criterion for other properties, unless otherwise demonstrated. Mineral addition may have different effects on the above-mentioned properties, which should be separately analyzed. Several research works investigated already cementing efficiency of fly ash, silica fume and blast furnace slag, used at varying dosages. However, no studies were carried out to assess EAFD kvalue yet. Accordingly, in this research, the efficiency of an EAFD, used as SCM, on the strength of mortar mixtures is estimated. The analyzed variables are the amount of binder replacement ratio (5% and 10%), the water/binder (w/b) ratio (0.35, 0.4, 0.5, 0.6 and 0.7) and the curing age (7 and 28 days). In particular, k-values are evaluated considering both compressive and flexural strength, aiming to investigate whether compressive strength can be considered as a proxy criterion for flexural properties. 2. Background about cementing efficiency models for SCMs The concept of the cementing efficiency factor, or k-value, was initially introduced by Smith in 1967 [47] for a fly ash concrete,
defining that a mass of fly ash f can be considered as equivalent to a mass of cement kf, for its ability in influencing the strength development of the concrete. The rational model by Smith assumes that the strength and workability of two concretes with the same effective water/binder ratio, should be equal. In other words, if w/(c + kf) in a fly ash concrete is equal to w/c of the reference mix, the two mixtures should have the same strength. In general, the k-value approach is based on the assumption that a relationship exists between the tested property (typically, 28-days compressive strength) and a compositional parameter, which is the w/c ratio, if Abram’s law [44] or Bolomey’s equation [48] are considered for the reference Portland cement concrete. However, also other approaches could be applied, e.g. comparing the strength of two mixes with the same workability [49–51]. For instance, Schiessl and Hardtl [49] estimated the difference between the w/c ratio of a reference concrete, and the water/binder content w/(c + kf) of a fly ash concrete, necessary to achieve the same strength, and named this as the Dw concept. Bijen and Van Selst [50] used the same mathematical approach of Smith, and they found that the k-value depends on w/c ratio, cement type, fly ash quality, and concrete age. They reported that the cementing efficiency value of fly ash tends to decrease, as the w/c ratio increases. This dependence was higher in rapid setting Portland cement than in Portland blast furnace slag cement. Lastly, Babu et al. [51] separated the efficiency factor into two parameters, the general efficiency factor ke, and the percentage efficiency factor kp. The ke value is assumed constant for each SCM, whereas the kp value depends on the amount of the SCM inside the mix. The overall efficiency factor k is calculated by multiplying these two factors. Babu used the Dw concept [49] to assess the efficiency of concretes with pozzolana [51], fly ash [39], silica fume [52] and ground granulated blast furnace slag [53]. Those studies revealed that the overall efficiency factor may change with type and amount of cement, age and concrete curing conditions. Hassaballah and Wenzel [54] proposed a method based on a comparison of the compressive strength between a fly ash concrete and a control, with the same workability. Fly ash contribution to compressive strength was calculated as the strength difference between the two mixtures, thus assessing the pozzolanic efficiency factor. According to this method, positive values of k indicate a strength increase, while negative values indicate a strength loss. Wong and Razak [41] used an alternative approach to determine the efficiency of calcined kaolin and silica fume, calculating k as (RsC C0 P). Rs is the relative compressive strength between the concrete with and without SCM; C is the cement quantity in the reference concrete; C’ is the amount of cement in the SCM concrete; and lastly, P is the amount of the mineral addition. 3. Experimental program 3.1. Materials The materials used to prepare the mortar mixes are: Portland cement; electric arc furnace dust (EAFD); sand; and tap water. The cement used is a rapid setting ordinary Portland cement type I 52.5R, as defined in BS-EN 197-1 [2], with a density of 3080 kg/m3. EAFD is used in as-received condition from the dust collection system of a carbon steelmaking factory in Italy, and has a density of 3488 kg/m3. The natural sand has a 2 mm maximum particles size, apparent density of 2.64 g/ cm3 and saturated surface-dried (s.s.d.) density of 2.76 g/cm3. Fig. 1 shows particles size distribution of cement, EAFD and sand, obtained using laser diffraction technique. EAFD has a wide range of particle sizes, with coarse particles having a diameter up to 478 lm. The characteristics of the distribution are: d10 equal to 2.57 lm; median particle size d50 of 10 lm and d90 is 67.58 lm. Cement is characterized instead by d10 = 1.35 lm, d50 = 10.68 lm and d90 = 32.75 lm, being finer than the EAFD used in this work. Fig. 2 shows instead a scanning electron microscope (SEM) image of EAFD, taken in the secondary electron mode; it is possible to observe the dust morphology, made by ultrafine, spherically shaped particles. Table 1 shows the chemical composition of Portland cement and EAFD, obtained with X-ray Fluorescence (XRF) analyses. As it is possible to see, the composition of this dust is rich in Zn, Fe and Ca oxides, with few quantity of Si, Mg, Mn, Pb and Cl
143
M. da Silva Magalhães et al. / Construction and Building Materials 157 (2017) 141–150
Table 1 Major chemical compounds in ordinary Portland cement and EAFD (assessed through XRF analysis).
Fig. 1. Particles size distribution of cement, EAFD and sand.
Compound
Cement (wt%)
EAFD (wt%)
SiO2 Al2O3 Fe2O3 CaO SO3 MgO K2O MnO Na2O ZnO PbO2 CdO NiO Cr2O3 CuO Cl
21.61 5.77 2.67 61.39 6.11 2.58 0.9796 0.064 2.00 – – – – – – –
3.89 – 35.92 13.32 1.07 2.52 – 3.92 – 31.34 1.72 0.09 0.44 0.85 0.26 2.06
oxides. Also limited amount of harmful heavy-metals is present, such as Cr, Cd and Ni. The X-ray Diffraction (XRD) pattern, shown in Fig. 3 and obtained on a representative dust sample, highlights the presence of five main phases, which are magnetite (Fe3O4), franklinite (ZnFe2O4), zincite (ZnO), calcium ferrite (CaFe2O4) and lime (CaO).
evaluated for compressive and flexural strength, taking into account EAFD amount in the mixes, and the range of w/b ratios considered, between 0.35 and 0.7.
3.2. Mixture details
4.1. Compressive strength evolution
In this research, fifteen mortar mixtures are casted, with varying water/binder (w/b) ratios, between 0.35 and 0.7. Five mortars are without EAFD (labelled as Ref), five mortars contain 5% of EAFD (labelled as M01) as SCM, and lastly five mortars contain 10% of EAFD as cement substitute (labelled as M02). Mixture proportions are listed in Table 2. Binder to sand ratio is maintained constant in all the mixes, and is equal to 1:3. Substitution ratios are expressed in weight percentage. A superplasticizer based on sulfonated naphthalene is added to some mixtures, to allow a good workability for all the mixtures, including the ones with the lowest w/b ratio. In all the cases, flow diameter value ranges between 155 mm and 267 mm. All the mortars specimens are casted in steel molds, and after 24 h from the mixing they are demolded and cured until testing in 100% relative humidity (RH) conditions, at 20 ± 2 °C temperature. 3.3. Experimental methods Flexural and compressive strength of hardened mortar specimens are determined according to [55]. For the former, a 3PBT (three point bending test) is carried out on 160 40 40 mm prismatic specimens; for the latter, compressive strength is determined on cubic specimens with 40 mm sides. Compressive and flexural tests are carried out at the ages of 7 and 28 days.
4. Experimental results and discussion In this research work, a practical approach to assess the effect of EAFD on the strength of Portland cement-based mortars is used, applying the concept of SCM efficiency factor. Hence, k-values are
Fig. 4 shows the dependence of compressive strength on the amount of EAFD and on w/b ratio, both at 7 and 28 days. Compressive strength of the tested mortars depends on the w/b ratio, following the Abrams’ law. According to Neville [3], concrete strength at a given age depends mainly on w/c ratio, as well as on the degree of compaction. Here, the use of the superplasticizer in the mixtures with the lowest water amount was fundamental to ensure that specimens could be considered as fully-compacted, similarly to the other mixtures. Looking at Fig. 4, it is possible to observe that the increase of w/b from 0.35 to 0.7 lowers compressive strength of both reference and EAFD mortars (M01 and M02). The decrease ranges between 34% and 37% at 7 days, and between 33% and 34% at 28 days of curing. The porosity of the mortars increases in the hardened state with the w/b ratio, thus it directly affects the mechanical strength of the mortars. Furthermore, data shown in Fig. 4 highlight that, in most the cases, compressive strength increases when 5% of EAFD is used as SCM, whereas it decreases at 10% of substitution, regardless the w/b ratio and the age of curing. The maximum observed strength gain in M01 mortars with 5% of EAFD is 5.2% at 7 days, and 2.8% at 28 days; conversely, the highest loss in M02 mortars with 10% of EAFD is 3% at 7 days, and 4.3% at 28 days. It is worth to note that even at 7 days of curing, the retardant effect in strength development, which is frequently reported when this dust is used [26], is not observed at the lowest EAFD dosage. Fig. 5 shows the regression equations obtained from the average values of compressive strength, shown in the previous Fig. 4. The relationships are derived to predict the 7 and 28 days compressive strength of mortars, as a function of w/b ratio, for each group of mortar. At this stage, these relationships are obtained without considering the EAFD efficiency factor. The regression equations with the highest correlation coefficients are power functions of w/b ratio (Eq. (1)):
rc ¼ g1 ðw=bÞg2
Fig. 2. SEM image of EAFD.
ð1Þ
where rc is the compressive strength of the mortars, w is the water content, b is the theoretical total binder content (c + EAFD), g1 and g2 are constant coefficients for a given age, and their value is shown in Table 3. R2 values are also included in Table 3 per each equation. It is worth to note that, excepting in one case, R2 values are at least
144
M. da Silva Magalhães et al. / Construction and Building Materials 157 (2017) 141–150
Fig. 3. XRD pattern of EAFD.
Table 2 Mixture proportions of mortars. Mix
Cement (%)
EAFD (%)
w/b
Ref-035 M01-035 M02-035 Ref-040 M01-040 M02-040 Ref-050 M01-050 M02-050 Ref-060 M01-060 M02-060 Ref-070 M01-070 M02-070
100 95 90 100 95 90 100 95 90 100 95 90 100 95 90
– 5 10 – 5 10 – 5 10 – 5 10 – 5 10
0.35 0.35 0.35 0.40 0.40 0.40 0.50 0.50 0.50 0.60 0.60 0.60 0.70 0.70 0.70
Quantities (kg/m3) Cement
EAFD
Sand
Water
SP
567.7 539.8 512.0 552.0 524.9 497.8 523.1 497.4 471.7 497.1 472.7 448.2 473.6 450.3 427.0
– 28.4 56.9 – 27.6 55.3 – 26.2 52.4 – 24.9 49.8 – 23.7 47.4
1703.0 1704.8 1706.6 1656.0 1657.7 1659.4 1569.3 1570.9 1572.5 1491.3 1492.7 1494.1 1420.7 1422.0 1423.3
198.7 198.9 199.1 220.8 221.0 221.3 261.6 261.8 262.1 298.3 298.5 298.8 331.5 331.8 332.1
6.81 6.82 6.83 5.52 5.53 5.53 – – – – – – – – –
EAFD: electric arc furnace dust; w/b: water/binder ratio; SP: superplasticizer.
Fig. 4. Experimental compressive strength of tested mortars, with varying w/b ratios, at 7 and 28 days (b: overall binder content).
equal to 0.98. This means that the proposed relationships based on Eq. (1) predict well the compressive strength of the mixes, both for reference and EAFD mortars. According to Hedegaard and Hansen [56], when dealing with cement-based materials, which are intrinsically heterogeneous materials, a coefficient of correlation R2 above 0.95 indicates a good correlation in a statistical regression analysis,
between a hypothetical model and corresponding experimental data. Bolomey’s Equation (Eq. (2)) is also used to estimate compressive strength from the knowledge of b/w ratio, as it is linear, and therefore easier to use in the practice:
rc ¼ Aðb=wÞ þ B
ð2Þ
145
M. da Silva Magalhães et al. / Construction and Building Materials 157 (2017) 141–150
Fig. 5. Relationships between compressive strength and w/b ratio for the tested mortars, at 7 and 28 days (b: overall binder content).
Table 3 Regression equation parameters for compressive strength, and R2 values. Mix
Power’s equation parameters 7 days
Ref. M01 M02
Bolomey’s equation parameters 28 days
7 days
28 days
g1
g2
R2
g1
g2
R2
A
B
R2
A
B
R2
27.47 28.13 26.99
0.64 0.63 0.65
0.982 0.994 0.996
35.09 35.46 33.98
0.55 0.55 0.56
0.985 0.985 0.975
13.99 13.51 13.63
14.68 16.25 14.80
0.986 0.993 0.996
13.83 14.01 13.49
23.15 23.49 22.73
0.986 0.983 0.967
where A and B are constant parameters, and their values are listed in Table 3, for each mortar group, with the corresponding R2 value. In this experimental campaign, the value of the parameter A, which is the slope of the curve defined by Eq. (2), is almost the same at the two considered ages; the parameter B, instead, varies considerably with age. However, their values do not change significantly between each mortar group. 4.2. EAFD efficiency factors for compressive strength In this research, the k-values of EAFD are estimated initially with the Dw concept, which is described in detail in the work conducted by Babu and Rao [39] and Schiessl and Hardtl [49]. This procedure attempts to bring the w/b ratio of the EAFD mortar close to the w/c ratio of the reference, through the application of EAFD kvalue at any strength. The principle of the calculation is illustrated in Fig. 6. The k-value is defined in such a way that the w/c ratio of the reference mortar, and the water to effective cementitious materials ratio [w/(c + kD)] of EAFD mix are the same, for the same strength level. Here, w, c and D are respectively the amount of water, cement and dust in the EAFD mortar. Comparing the curves shown in Fig. 6, Dw can be expressed by Eq. (3):
Dw ¼ ðw=cÞ ðw=bÞ
ð3Þ
In the above relation, the water to effective cementitious materials ratio w/(c + kD) should substitute the w/c term, and the water to overall binder content w/(c + D) substitutes w/b. Then, the kvalue can be calculated with Eq. (4):
w c k¼ w D D Dw þ ðcþDÞ
ð4Þ
In this research, a k-value that depends on EAFD quantity (percentage efficiency factor) is calculated, because the use of a single k-value did not lead to a good correlation for each replacement ratio of the SCM. Fig. 7 shows the variation of k-values as a function of w/b ratio, for M01 and M02 mortars, both at 7 and 28 days of curing. It is
Fig. 6. Principle for calculating the efficiency factor with Dw concept.
possible to observe how the w/b ratio affects the k-values. For M01, k-values are always higher than 1, regardless the w/b ratio used, indicating that EAFD has a higher cementing efficiency than cement, when it is used as a replacement ratio of 5%. However, in M02, when EAFD is used at 10% replacement ratio, it is less efficient than cement in terms of compressive strength contribution, because k-values are less than 1, also in this case regardless the w/b ratio. In this study, however, the w/b ratio plays a very important role in affecting efficiency values. Results indicate that, when w/b ratio changes from 0.35 to 0.70, the k-values calculated for M01 mortars decrease from 1.40 to 1.20 (D = 14.5%) at 7 days, and from 1.62 to 1.30 (D = 19.4%) at 28 days. Conversely, this trend changes for the M02 mortars; when the w/b ratio changes from 0.35 to 0.70, the k-values increase from 0.72 to 0.86 (D = +19%) at 7 days, and from 0.61 to 0.80 (D = +31.3%) at 28 days. This indicates that, when higher quantity of EAFD is used, more water is needed to increase the cementing efficiency of this dust. This finding is consistent with the work conducted by Bijen and Van Selst [50].
146
M. da Silva Magalhães et al. / Construction and Building Materials 157 (2017) 141–150
Fig. 7. EAFD k-values calculated with the Dw concept, as a function of the w/b ratio at 7 and 28 days, for compressive strength (b: overall binder content).
Considering the k-values for different w/b ratios in Fig. 7, the average k-values for 5% and 10% EAFD at 7 and 28 days of curing are calculated and listed in Table 4. Standard deviation (SD) is also reported in parenthesis. For M01, the average k-values increase with mortar age. However, this effect is not observed in M02; the average values of k decrease from about 1.29 to 0.80 at 7 days, and from 1.45 to 0.71 at 28 days, when the replacement percentage varies from 5% to 10% of EAFD. Accordingly, the efficiency of EAFD is reduced with the percentage replacement. This is consistent with the results shown in Fig. 5. The k-values evaluated for this EAFD are comparable, or even higher, than the efficiency factor of high-calcium fly ash (k: 0.9 with SCM content of 10–20% [48]) and blast furnace slag (k: 0.69 with SCM content of 70% [44]) at 28 days, but lower than silica fume efficiency (k: 2.1–3.1 with SCM content of 5–15% [41]). It is important to highlight that this comparison is performancebased only, because the efficiency of SCM on mortars strength varies significantly with their properties, content inside the mixture,
and tests conditions, i.e. curing age, type of cement and strength grade [36,37]. Lastly, considering the calculated average k-values for each EAFD replacement ratio, the variation of compressive strength with water to effective cementitious materials ratio w/(c + kD) is represented in Fig. 8. This figure clearly shows that the predictions present an excellent agreement with the experimental results, and the R2 value is 0.99 at 7 days, and 0.981 at 28 days. Those values are similar or even better than ones obtained for the reference mortar, listed in Table 3 (R2 = 0.982 at 7 days and R2 = 0.985 at 28 days). The values of the coefficients g1 and g2 are similar to those obtained for reference mortar. An alternative approach is used in this research for the evaluation of EAFD cementing efficiency, which has been used in many other works in literature [44,48]. This simple methodology uses the same equation of the Portland cement mortar (reference), estimated through the empirical Eq. (1), and the experimentally measured values of compressive strength of the EAFD mortars, given in Fig. 4. The advantage of this method is that only two mixtures are required to determine the k-value for a specific SCM. In the case of mortars produced with EAFD, and considering the concept of kvalue, Eq. (1) can be re-written as:
rc ¼ n1
w cþkD
n2
ð5Þ
where g1 and g2 are the same parameters estimated in the reference mortar, and reported in Table 3. Using this equation, with the experimental values of compressive strength of the EAFD mortars, and knowing the quantities of w, c and D (given in Table 2), the k-values can be easily calculated. Results are shown in Table 4, and labelled as ‘‘alternative approach”, to be compared with the ones derived with the Dw concept. In general, the average k-values calculated with the two methods are similar; however, the SD values are higher with the alternative approach than with the Dw concept. This is due to the different w/b ratios used during the experimental campaign to cast the specimens; the clear trend about k-value dependency on w/b
Table 4 Average efficiency factors (k-values) calculated with the Dw concept and an alternative approach.
Dw concept
Mortar property
Compressive strength, 7 days Compressive strength, 28 days Flexural strength, 7 days Flexural strength, 28 days
Alternative approach
5% EAFD
10% EAFD
5% EAFD
10% EAFD
1.29 1.45 1.87 2.28
0.80 0.71 0.71 0.85
1.44 1.50 2.00 2.21
0.79 0.60 0.62 0.80
(0.08) (0.13) (0.22) (0.28)
(0.06) (0.08) (0.08) (0.04)
(0.46) (0.73) (0.58) (1.04)
Fig. 8. Relationships between compressive strength and w/(c + kD) ratio for the tested mortars, at 7 and 28 days.
(0.18) (0.45) (0.25) (0.82)
M. da Silva Magalhães et al. / Construction and Building Materials 157 (2017) 141–150
147
Fig. 9. Experimental flexural strength of tested mortars, with varying w/b ratios, at 7 and 28 days (b: overall binder content).
emphasize that, when this method is used, the k-value calculated with one only w/b ratio cannot represent the most appropriate kvalue. 4.3. Flexural strength evolution
Fig. 10. Relation between flexural and compressive strength, at different ages.
ratio, observed applying the Dw concept method, cannot be captured with the alternative approach. For instance, in M01 mortars, the k-values at 7 days are 1.11 (w/b: 0.35), 1.62 (w/b: 0.4), 1.69 (w/ b: 0.5), 0.84 (w/b: 0.6) and 1.96 (w/b: 0.7), and at 28 days they are 2.16 (w/b: 0.35), 0.47 (w/b: 0.4), 1.74 (w/b: 0.5), 2.09 (w/b: 0.6) and 1.04 (w/b: 0.7). Instead, for M02 mortars, the k-values at 7 days are 1.00 (w/b: 0.35), 0.59 (w/b: 0.4), 0.71 (w/b: 0.5), 0.97 (w/b: 0.6) and 0.69 (w/b: 0.7), and at 28 days they are 0.71 (w/b: 0.35), 0.17 (w/b: 0.4), 1.24 (w/b: 0.5), 0.72 (w/b: 0.6) and 0.15 (w/b: 0.7). These results indicate that there is a strong influence of the w/b ratio in this methodology of evaluation. Accordingly, it is important to
The 7 and 28 days-flexural strength values are presented graphically in Fig. 9. As in the case of compressive strength, and as expected, flexural strength values increase with the decrease of w/b ratio. Flexural strength values of reference, 5% EAFD (M01) and 10% EAFD (M02) mortars, cured for 7 days, increase from about 6 to 8.6 MPa, from 6.2 to 8.3 MPa and from 5.8 to 8.3 MPa, respectively, when the w/b ratio is reduced from 0.70 to 0.35. Increasing the curing time to 28 days, the strength of these mixtures further improves, until values of 9 MPa for the reference, 9.2 MPa for M01 and 8.9 MPa for M02 mortars. A slight positive effect is observed when 5% of EAFD is used; however, reduction of up to 4% on the flexural strength is noticed in 10% EAFD mortars. Fig. 10 indicates that the flexural to compressive strength ratio is not significantly influenced by the addition of EAFD inside the mixtures, and a linear relationship between these parameters can be obtained. However, at 28 days, the flexural strength increases more slowly than the compressive strength. This is consistent with the general tendency of this ratio to decrease, when compressive strength value increases [3]. Power empirical expressions (Eq. (6)) are derived from Fig. 11, to predict the flexural strength of the analyzed mixtures (reference, and mortars with 5% and 10% EAFD) at 7 and 28 days.
rf ¼ k1 ðw=bÞk2
Fig. 11. Relationships between 7 and 28 days flexural strength and w/b ratio for the tested mortars (b: overall binder content).
ð6Þ
148
M. da Silva Magalhães et al. / Construction and Building Materials 157 (2017) 141–150
Table 5 Regression equation parameters for flexural strength, and R2 values. Mix
Power’s equation parameters 7 days
Ref. M1 M2
Linear equation parameters 28 days
7 days
28 days
k1
k2
R2
k1
k2
R2
E
F
R2
E
R2
F
5.02 5.11 4.90
0.51 0.52 0.52
0.998 0.988 0.981
6.80 6.80 6.74
0.28 0.30 0.28
0.991 0.954 0.945
1.80 1.97 1.73
3.50 3.35 3.49
0.999 0.995 0.981
1.09 1.21 1.08
6.01 5.89 5.97
0.969 0.972 0.923
Fig. 12. EAFD k-values calculated with the Dw concept, as a function of the w/b ratio at 7 and 28 days, for flexural strength (b: overall binder content).
In Eq. (6), rf is the flexural strength of the mortars, w is the water content, b is the theoretical total binder content, k1 and k2 are constant parameters, to be derived from the experimental results, and which value is listed in Table 5. High correlation coefficients are obtained, as shown in Table 5. Also in this case, a linear expression (Eq. (7)) is developed to predict flexural strength of mortars; parameters E and F are respectively the slope and the intercept of the curves, and their value is listed in Table 5, with the resulting R2 values.
rf ¼ Eðb=wÞ þ F
ð7Þ
4.4. EAFD efficiency factors for flexural strength In this study, EAFD efficiency on the flexural strength is investigated, too. The k-values of EAFD considering flexural strength are
estimated with the same methods used for assessing compressive strength efficiency. Hence, the Dw concept methodology described in Section 4.2. and further illustrated in Fig. 6, is used also in this case, just considering the Dw obtained by comparing the curves in Fig. 11. The k-values are given in Fig. 12, considering both curing times, the two substitution ratios and the varying w/b ratios. The average k-values for M01 and M02 mortars, at 7 and 28 days of curing, are listed in Table 4. Standard deviation (SD) is also reported in parenthesis. Cementing efficiency values related to flexural strength are higher than k-values obtained for the compressive strength of EAFD mortars. As expected, and similarly to the case of compressive strength, the average k-values increase with the time, for both the replacement ratios, because of the strength gain due to aging. Concerning the influence of EAFD amount inside the mortars, the same trend seen for the compressive strength target property is obtained. The average k-value of M01 mortars is greater than 1 (1.87 at 7 days, and 2.28 at 28 days), whereas the average k-value of M02 mortars is lower than 1 (0.71 at 7 days, and 0.85 at 28 days). This indicates that the efficiency of EAFD on the flexural strength is reduced with the increase of the replacement ratio. It should be noted also that when w/b ratio is increased from 0.35 to 0.70, the k-values based on the flexural strength decrease from 2.33 to 1.64 at 7 days, and from 2.79 to 1.66 at 28 days, in M01 mortars (Fig. 12). However, M02 mortars display a reverse trend (Fig. 12). When the w/b ratio changes from 0.35 to 0.70, the k-values increase from 0.60 to 0.80 at 7 days, and from 0.79 to 0.90 at 28 days. Lastly, the variation of flexural strength with water to effective cementitious materials ratio w/(c + kD) is shown in Fig. 13. The prediction equations have high correlation coefficients (R2 = 0.982 and 0.973), and the coefficients k1 and k2 are similar to those obtained for the reference mortar (Table 5). These results confirm the reliability of the proposed method for the estimation of the k-values also from flexural tests.
Fig. 13. Relationships between flexural strength and w/(c + kD) ratio for the tested mortars, at 7 and 28 days.
M. da Silva Magalhães et al. / Construction and Building Materials 157 (2017) 141–150
The k-values are evaluated also with the alternative approach described in Section 4.2., which uses the power equation derived for the Portland cement mortar. Hence, k-values are estimated by the empirical Eq. (8):
rf ¼ k 1
k2 w cþkD
ð8Þ
149
Acknowledgments The authors acknowledge the Quality Improvement Program for Universities of the Ministry of Education and Culture (CAPES) for the financial support, Acciaierie Venete SpA and Italcementi SpA for supplying respectively the EAFD and the cement. References
where c, k and D were already defined, k1 and k2 are the parameters of the reference mortar listed in the Table 5. Using this equation, with the measured values of the flexural strength given in Fig. 9 and the w, c and D contents given in Table 2, the k-values for the EAFD used in this work in two replacement ratios can be derived, and their values are listed in Table 4. The average k-values calculated with this method are, generally, close to those estimated with the Dw concept method. However, the SD of the results for varying w/b ratios is noticeable higher than the corresponding SD value, obtained with Dw concept method. For instance, in the 5% EAFD mortar, the k-values at 7 days vary from null to 2.58, and at 28 days vary from 0.57 to 3.31. In 10% EAFD mortars the k-values at 7 days vary between 0.35 and 0.86, and at 28 days, vary between 0.25 and 1.74. Furthermore, also in this case the dependency of the k-value on the w/b ratio cannot be observed.
[1] [2] [3] [4]
[5]
[6]
[7] [8]
[9]
5. Conclusions [10]
The experimental campaign carried out in this work provides preliminary information about the efficiency of an EAFD on the strength of mortars, when it is used as a SCM. According to the experimental results observed in this study, the following conclusions about the cementing efficiency of EAFD can be derived. The cementing efficiency depends on the quantity of EAFD used. The cementing efficiency of EAFD at 5% replacement ratio is higher than when it is used at 10% replacement ratio. Additionally, at 5% replacement ratio, the k-value of EAFD exceeds the unit. The k-value calculated using only one water/binder ratio cannot represent the most appropriate k-value for EAFD, and more generally speaking, for a SCM. It is demonstrated that k-value depends on the w/b ratio used. In the specific case, when the w/b ratio increases, there is a decline in cementing efficiency for 5% EAFD; conversely, the trend is opposite when 10% EAFD is used. Cementing efficiency values raise with the increase of hydration time. Flexural strength k-values are greater than compressive strength k-values. This result demonstrates that EAFD addition influences more the flexural than the compressive strength. This research is a first attempt to investigate cementing efficiency of EAFD, thus further laboratory investigation is necessary to establish the reliability of the proposed method, particularly with respect to the design of blended cements with other cement types and EAFD amounts. A special care should be paid when managing such kind of solid waste, which contains substances potentially harmful for the environment; hence, future research lines will be focused also on the leachability of heavy metals from the cement-based materials where EAFD will be incorporated. About this topic, it is worth to mention that the authors are currently studying a hydrometallurgical pre-treatment to partially leach zincite, thus limiting the source of Zn2+ in the matrix.
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18] [19] [20]
[21]
[22]
[23]
[24]
[25]
[26]
ASTM C595 / C595M – 16, Standard Specification for Blended Hydraulic Cements, ASTM International, West Conshohocken, PA, 2016. EN 197-1. Cement Part 1: Composition, Specifications and Conformity Criteria for Common Cements, Brussels, Belgium (2011). A.M. Neville, Properties of Concrete, 5th edition., Pearson, United Kingdom, 2011. P. Castaldo, B. Palazzo, A. Mariniello, Effects of the axial force eccentricity on the time-variant structural reliability of aging r.c. cross-sections subjected to chloride-induced corrosion, Eng. Struct. 130 (2017) 261–274. J.G.M.S. Machado, F.A. Brehm, C.A.M. Moraes, C.A. Santos, A.C.F. Vilela, J.B.M. Cunha, Chemical, physical, structural and morphological characterization of the electric arc furnace dust, J. Hazard. Mater. 136 (3) (2006) 953–960. N. Menad, J.N. Ayala, F. Garcia-Carcedo, E. Ruiz-Ayúcar, A. Hernández, Study of the presence of fluorine in the recycled fractions during carbothermal treatment of EAF dust, Waste Manage. 23 (6) (2003) 483–491. G. Orhan, Leaching and cementation of heavy metals from electric arc furnace dust in alkaline medium, Hydrometallurgy 78 (2005) 236–245. G. Salihoglu, V. Pinarli, Steel foundry electric arc furnace dust management: stabilization by using lime and Portland cement, J. Hazard. Mater. 153 (2008) 1110–1116. M. Maslehuddin, F.R. Awan, M. Shameem, M. Ibrahim, M.R. Ali, Effect of arc furnace dust on the properties of OPC and blended cement concretes, Constr. Build. Mater. 25 (2011) 308–312. O. Ruiz, C. Clemente, M. Alonso, F.J. Alguacil, Recycling of an electric arc furnace flue dust to obtain high grade ZnO, J. Hazard. Mater. 141 (1) (2007) 33–36. R.Z. Al-Zaid, F.H. Al-Sugair, A.I. Al-Negheimish, Investigation of potential uses of electric-arc furnace dust (EAFD) in concrete, Cem. Concr. Res. 27 (2) (1997) 267–278. M.G. Mitrakas, C.A. Sikalidis, T.P. Karamanli, Immobilization of EAFD heavy metals using acidic materials, J. Environ. Sci. Health Part A Toxic/Hazard. Subst. Environ. Eng. 42 (4) (2007) 535–541. E.F. Ledesma, J.R. Jiménez, J. Ayuso, J.M. Fernández, J. de Brito, Experimental study of the mechanical stabilization of electric arc furnace dust using fluid cement mortars, J. Hazard. Mater. 326 (2017) 26–35. C.A.D. De Souza, A.T. Machado, L.R.P. De Andrade Lima, R.J.C. Cardoso, Stabilization of electric arc furnace dust in concrete, Mater. Res. 13 (4) (2010) 513–519. G. Laforest, J. Duchesne, Stabilization of electric arc furnace dust by the use of cementitious materials: ionic competition and long-term leachability, Cem. Concr. Res. 36 (9) (2006) 1628–1634. M.G. Sebag, C. Korzenowski, A.M. Bernandes, A.C. Vilela, Evaluation of environmental compatibility of EAFD using different leaching standards, J. Hazard. Mater. 166 (2–3) (2009) 670–675. M. Al-Harahsheh, M. Aljarrah, F. Rummanah, K. Abdel-Latif, S. Kingman, Leaching of valuable metals from electric arc furnace dust— Tetrabromobisphenol A pyrolysis residues, J. Anal. Appl. Pyrolysis 125 (2017) 50–60. K. Brunelli, M. Dabalà, Ultrasound effects on zinc recovery from EAF dust by sulfuric acid leaching, Int. J. Min. Metallurgy Mater. 22 (4) (2015) 353–362. X. Xuefeng, T. Yuhong, Application of electric arc furnace dust in cement production, Iron Steel (Peking) 33 (6) (1998) 61–64. C. Sikadilis, M. Mitrakas, Utilization of electric arc furnace dust as raw material for the production of ceramic and concrete building products, J. Environ. Sci. Health Part A 41 (9) (2007) 1943–1954. V.N. Stathopoulos, A. Papandreou, D. Kanellopoulou, C.J. Stournaras, Structural ceramics containing electric arc furnace dust, J. Hazard. Mater. 262 (2013) 91– 99. M.A.T. Alsheyab, T.S. Khedaywi, Effect of electric arc furnace dust (EAFD) on properties of asphalt cement mixture, Resour. Conserv. Recycl. 70 (2013) 38– 43. I. Nikolic´, D. Crossed, D. Signurovic, D. Blecic, R. Zejak, Karanovic Lj, S. Mitsche, V.R. Radmilovic, Geopolymerization of coal fly ash in the presence of electric arc furnace dust, Min. Eng. 49 (2013) 24–32. M.C. Arnold, A.S. de Vargas, Bianchini, Study of electric-arc furnace dust (EAFD) in fly ash and rice husk ash-based geopolymers, Adv. Powder Technol. 28 (9) (2017) 2023–2034. M. Niubó, A.I. Fernández, J.M. Chimenos, L. Haurie, A possible recycling method for high grade steels EAFD in polymer composites. J. Hazard. Mater. 171 (1–3) 1139–1144. M.S. Magalhães, F. Faleschini, C. Pellegrino, K. Brunelli, Effects of electric arc furnace dust (EAFD) addition on setting and strength evolutions of cement pastes and mortars, Eur. J. Environ. Civ. Eng. 1 (2017) 1–13.
150
M. da Silva Magalhães et al. / Construction and Building Materials 157 (2017) 141–150
[27] G. Fares, R.Z. Al-Zaid, A. Fauzi, A.M. Alhozaimy, A.I. Al-Negheimish, M.I. Khan, Performance of optimized electric arc furnace dust-based cementitious matrix compared to conventional supplementary cementitious materials, Constr. Build. Mater. 112 (1) (2016) 210–221. [28] F.M. Al Mutlaq, C.L. Page, Effect of electric arc furnace dust on susceptibility of steel to corrosion in chloride-contaminated concrete, Constr. Build. Mater. 39 (2013) 60–64. [29] Chen Qy, M. Tyrer, C.D. Hills, X.M. Yang, P. Carey, Immobilisation of heavy metal in cement-based solidification/stabilisation: a review, Waste Manage. 29 (1) (2009) 390–403. [30] G. Arliguie, J. Grandet, Influence de la composition d’un ciment Portland sur son hydratation em presence de zinc, Cem. Concr. Res. 20 (3) (1990) 517–524. [31] M. Murat, F. Sorrentino, Effect of large additions of Cd, Pb, Cr, Zn, to cement raw meal on the composition and the properties of the clinker and the cement, Cem. Concr. Res. 26 (1996) 377–385. [32] A. Balderas, H. Navarro, L.M. Flores-Velez, O. Dominuez, Properties of Portland cement pastes incorporating nanometer-sized franklinite particles obtained from electric arc furnace dust, J. Am. Ceram. Soc. 84 (12) (2001) 2909–2913. [33] Q.Y. Chen, C.D. Hills, M. Tyrer, I. Slipper, H.G. Shen, A. Brough, Characterisation of products of tricalcium silicate hydration in the presence of heavy metals, J. Hazard. Mater. 147 (3) (2007) 817–825. [34] A. Vargas, A.B. Masuero, A.C.F. Vilela, Investigations on the use of electric arc furnace dust (EAFD) in pozzolan-modified Portland cement I (MP) pastes, Cem. Concr. Res. 36 (10) (2006) 1833–1841. [35] G.A. Rao, Generalization of Abrams’ law for cement mortars, Cem. Concr. Res. 31 (2001) 495–502. [36] M.A. Sanjuán, A. Piñeiro, O. Rodríguez, Ground granulated blast furnace slag efficiency coefficient (k value) in concrete, Appl. Limits Mater. Constr. 61 (302) (2011) 303–313. [37] R. Siddique, M.I. Khan, Supplementary Cementing Materials, Springer Science & Business Media, 2011. pp. 288. [38] V.G. Papadakis, S. Tsimas, Greek supplementary cementing materials and their incorporation in concrete, Cem. Concr. Compos. 27 (2005) 223–230. [39] K.G. Babu, G.S.N. Rao, Efficiency of fly ash in concrete, Cem. Concr. Compos. 15 (4) (1993) 223–229. [40] V.G. Papadakis, Effect of supplementary cementing materials on concrete resistance against carbonation and chloride ingress, Cem. Concr. Res. 30 (2) (2000) 291–299. [41] H.S. Wong, Razak H. Abdul, Efficiency of calcined kaolin and silica fume as cement replacement material for strength performance, Cem. Concr. Res. 35 (2005) 696–702.
[42] E. Karim, E.H. Kadri, R. Siddique, Investigation of slag cement quality through the analysis of its efficiency coefficient, Eur. J. Environ. Civ. Eng. 15 (10) (2011) 1393–1411. [43] E. Gruyaert, M. Maes, N. De Belie, Performance of BFS concrete: k-value concept versus equivalent performance concept, Constr. Build. Mater. 47 (2013) 441–455. [44] F. Lollini, E. Redaelli, L. Bertolini, A study on the applicability of the efficiency factor of supplementary cementitious materials to durability properties, Constr. Build. Mater. 120 (2016) 284–292. [45] D.F. Aponte, M. Barra, E. Vazquez, Durability and cementing efficiency of fly ash in concretes, Constr. Build. Mater. 30 (2012) 537–546. [46] European Committee for Standardization, European Standard EN 206: Concrete-specification, performance, production and conformity. Brussels (2013). [47] I.A. Smith, Design of fly ash concretes, Proc. Inst. Civ. Eng. 36 (1967) 769–790. [48] V.G. Papadakis, S. Tsimas, Supplementary cementing materials in concrete, Part I Efficiency Des. Cem. Concr. Res. 32 (10) (2002) 1525–1532. [49] P. Schiessl, R. Hardtl, Efficiency of fly ash in concrete evaluation of ibac test results 1991, Technical Report of Institut fiir Bauforschung, RWTH, Aachen. [50] J. Bijen, R. Van Selst, Cement equivalence factors for fly ash, Cem. Concr. Res. 23 (1993) 1029–1039. [51] K.G. Babu, G.S.N. Rao, P.V.S. Prakash, Efficiency of pozzolans in cement composites, in: R.K. Dhir, M.R. Jones (Eds.), Proceedings, Concrete 2000 Economic and Durable Construction through Excellence, University of Scotland, Dundee, Design, Materials and Construction, Vol. 1, E & FN Spon, London, 1993, pp. 497–509. [52] K.G. Babu, P.V.S. Prakash, Efficiency of silica fume in concrete, Cem. Concr. Res. 25 (6) (1995) 1273–1283. [53] K.G. Babu, V.S.R. Kumar, Efficiency of GGBS in concrete, Cem. Concr. Res. 30 (2000) 1031–1036. [54] A. Hassaballah, T.H. Wenzel, A strength definition for the water to cementitious materials ratio, in: V.M. Malhotra (Ed.), Proceedings, ACI Fifth International Conference Fly Ash, Silica Fume and Natural Pozzolans in Concrete, ACI, Detroit, USA, 1995, pp. 417–437. [55] EN 1015-11, Methods of Test for Mortar for Masonry: Part 11: Determination of Flexure and Compressive Strength of Hardened Mortar, Comité Européen de Normalisation, Brussels, Belgium, 2007. [56] S.E. Hedegaard, T.C. Hansen, Modified water/cement ratio law for compressive strength of fly ash concretes, Mater. Struct. 25 (1992) 273–283.