Journal Information
Vol. 8. Issue 3.
Pages 2786-2795 (May - June 2019)
Share
Share
Download PDF
More article options
Visits
387
Vol. 8. Issue 3.
Pages 2786-2795 (May - June 2019)
Original Article
DOI: 10.1016/j.jmrt.2019.01.030
Open Access
Comparison of Kambara reactor slag with blast furnace slag for Portland cement industry applications
Visits
387
Ana Carolina de Oliveira Diegueza, Samantha Luchi Nascimento Oliveirab, Georgia Serafim Araújoa, André Gustavo de Sousa Galdinoa,
Corresponding author
andregsg@ifes.edu.br

Corresponding author.
a Instituto Federal de Educação, Ciência e Tecnologia do Espírito Santo (Federal Institute of Education, Sciences and Technology of Espírito Santo), Av. Vitória, 1729, Jucutuquara, Vitória, ES, 29040-780, Brazil
b Faculdade Estácio de Sá de Vitória, Av. Dr. Herwan Modenese Wanderley, 1001 – Jardim Camburi, Vitória, ES, 29092-095, Brazil
This item has received
387
Visits

Under a Creative Commons license
Article information
Abstract
Full Text
Bibliography
Download PDF
Statistics
Figures (14)
Show moreShow less
Tables (9)
Table 1. Percent composition of CP III cement used.
Table 2. Techniques used in chemical analyses.
Table 3. Characteristics of the BB10 Mill.
Table 4. Results of the GBFS and KRS chemical analyses.
Table 5. Chemical analysis of the CP III 40 cement.
Table 6. Limits of the granulometric distribution of the fine aggregates [33].
Table 7. Values used to determine the unit weight of GBFS and KRSs.
Table 8. pH analysis results for the GBFS and KRS.
Table 9. Grinding tests of the GBFS and KRS by-products. Each grinding test is described because of the different conditions used in each test.
Show moreShow less
Abstract

This work was aimed at comparing the physical and chemical properties of the Kambara Reactor Slag (KRS) and a Granulated Blast Furnace Slag (GBFS) for achieving partial substitution of GBFS for the KRS. The KRS and GBFS were characterized by chemical, mineralogical, thermal, granulometric, visual, and microscopic analyses, which included the determination of the unit weight, pH, and ability to be ground. The KRS had lower concentrations of SiO2, Al2O3, and MgO and higher percentages of total Fe and Fe° than the GBFS. Moreover, compared to the GBFS, it was crystalline (the GBFS was amorphous); had a larger specific area with a final fraction of 150μm, higher unit weight, and predominantly spherical particles; and was more alkaline. Thus, the KRS is a by-product that may be useful for Portland cement manufacturing; however, the thermal, mechanical, or chemical activations will be needed to attain the cement requirements.

Keywords:
Kambara reactor slag
Granulated Blast Furnace Slag
Portland cement
Steel making
Full Text
1Introduction

Steel is one of the most widely used materials worldwide owing to its characteristics and properties. The steel industry generates large amounts of waste at each stage of steel production, including reduction and refining. During reduction, when a blast furnace is used, the by-product is pig iron. Refining includes the desulfurization of steel and consists of the reduction of the sulfur present in the metal by the formation of stable sulfides. These sulfides are removed from the liquid pig iron by skimming or scraping the slag formed on the surface, preventing the reaction product from being reversed and new solubilization of the sulfur to occur in the bath. In both stages, large quantities of by-products are formed, resulting in an environmental problem for the steel industry.

One alternative use of blast furnace slags has been in the manufacture of Portland cement. Compared to fly ash, silica fume, and pozzolanas, cementitious materials must have relatively constant chemical compositions, low heat of hydration, high sulfate and acid resistance, better workability, and higher ultimate strength [1]. Many studies have investigated Ground Granulated Blast Furnace Slag (GGBFS) for use in different applications, including manufacturing of alternatives to Portland cement [2–4] as a substitute for concrete cement [5–13] and production of geopolymeric pastes and mortars [14,15].

The use of Portland–GGBFS cement allows reducing the temperature rise of the concrete and improves the resistance to early cracking [16]. Moreover, when GBFS is added to a mortar that has very small particles, it increases the resistance to sulfates and improves the compressive strength with a low-dimensional expansion [17]. Other waste material from steelmaking is used alone or with a blast furnace slag to manufacture cement [18–20]. However, few studies have examined the use of the Kambara reactor slag as a raw material (such as GBFS, for example) for the manufacture of Portland cement [9,21].

This work aimed at the comparison of the physical and chemical properties of the Kambara Reactor Slag (KRS) and a GBFS to achieve partial substitution of the blast furnace slag for the KRS.

2Materials and methods2.1GBFS and KRS

The KRS, as shown in Fig. 1, was provided by Arcelor Mittal Tubarão, located in Serra, Espírito Santo, Brazil. One-hundred kg of the material with granulometry between 0 and 5mm was made available for performing the proposed tests.

Fig. 1.

Slags: (a) GBFS and (b) KRS.

(0.21MB).
2.2Portland cement type III

Portland cement type III (CP III) was used as a base to which the KRS was added; CP III was selected because this type of cement is already a constituent of clinker, gypsum, granulated slag, and carbonate materials. Table 1 lists the mass percentages of each cement raw material determined from the data provided by the manufacturer.

Table 1.

Percent composition of CP III cement used.

Clinker  GBFS  Plaster  Filler 
30.0%  66.5%  1.9%  1.6% 
2.3Chemical analysis

Chemical analyses of the slags were performed at the ArcelorMittal Tubarão Chemical Laboratory using the techniques and equipment listed in Table 2.

Table 2.

Techniques used in chemical analyses.

Chemical  Analytical method  Equipment/model 
Al2O3  X-ray fluorescence (XRF)  THERMO ARL9900 
CaO  XRF  THERMO ARL9900 
MgO  XRF  THERMO ARL9900 
MnO  XRF  THERMO ARL9900 
P2O5  XRF  THERMO ARL9900 
SiO2  XRF  THERMO ARL9900 
C total  Combustion  LECO CS444 
S total  Combustion  LECO CS444 
Fe°  Volumetry  – 
FeO  Volumetry  – 
Fe total  Volumetry  – 
MgO  ICP OES  THERMO iCAP 7000 
ZnO  ICP OES  THERMO iCAP 7000 
Na2ICP OES  THERMO iCAP 7000 
K2ICP OES  THERMO iCAP 7000 

In addition to the tests cited in Table 2, the percentage of loss to fire by gravimetry, a technique that allows determining a substance via mass change, was also recorded.

2.4Mineralogical analysis

X-Ray diffraction (XRD) was performed to verify the crystalline state of the GBFS by-products and KRS in addition to identifying the phases present in each sample. The tests were conducted with a Bruker D8 phaser diffractometer, from the Materials Characterization Laboratory of the IFES campus Vitória. The settings used for all the tests were 40kV and 10mA, with 0.02° step, and an observation range of 10°2θ100°.

2.5Thermogravimetry (TG) and differential scanning calorimetry (DSC)

Thermogravimetry (TG) measures the mass variation of a sample resulting from physical or chemical transformations when subjected to a controlled temperature gradient. In comparison, differential scanning calorimetry (DSC) measures the energy difference between a sample and thermally inert reference material required to keep both at the same temperature. The tests were performed at the Materials Characterization Laboratory of the IFES campus Vitória using a Netzch model STA 449 F3 Jupiter. The samples were subjected to a temperature ramp to 1000°C, with a heating rate of 10°C/min in air.

2.6Granulometric analysis

To determine the particle size distribution of the KRS and GBFS slags, the samples were quarantined with a Jones divider. The granulometric distribution of each slag was determined according to the ABNT NBR NM 248:2003 standard [22], using a PRODUTEST sieve agitator with a 6.3mm intermediate-series sieve and set of normal-series sieves of 4.8mm to 0.15mm, with a cover and bottom for each.

Both the quarrying and granulometry tests were performed at the Concrete Laboratory of the Federal Institute of Espírito Santo Vitória Campus. In addition, the volumetric distribution of the GBFS and KRS particles at 150-μm-throughput was determined using the Mastersizer Hydro 2000 Hydro MU from Malvern, housed at the Laboratory of Ceramics on the IFES campus Vitória.

2.7Determination of the unit weight

The KRS unit weight was determined by Kaeme Engenharia, according to ABNT NBR NM 45:2006 [23].

2.8Visual and microscopic analysis

Scanning electron microscopy (SEM) was performed for the GBFS and KRS to verify the morphologies and distributions of their particles. Zeiss model EVO MA10 at the Laboratory of Electronic Microscopy and Microanalysis of the IFES Campus Vitória was used. In addition, Energy Dispersive Spectroscopy (EDS) was performed on the GBFS and KRS samples to spectrally evaluate their elemental composition and identify their constituent elements. These spectra can provide a semi-quantitative characterization of the samples.

2.9pH determination

The pH of the slags was determined according to ABNT NBR 10.004 [24] in the Environmental Laboratory of ArcelorMittal Tubarão. Two determinations were made for each material and their arithmetic averages were calculated.

2.10Grinding

To grind the KRS, empirical tests were performed using a Marconi ball mill of the Laboratory of Ceramics of the IFES campus Vitória. The slags were placed in a steel jug internally coated with high alumina, with an internal diameter of 17cm and a length of 23cm, and the grinding bodies (balls) used were also made of high alumina. The time, ball size, and rotation speed were varied to allow the slag fineness to satisfy the criterion established in ABNT NBR 16697 [25], according to which at most 8% of the residue may remain in a 75-μm-sieve. After each milling test, the slag was sieved at 150μm and 75μm to evaluate the grinding efficiency.

A second test step was performed in the laboratory mill of an Engines Segor model BB10 of a cement factory located in southeast Brazil, and the characteristics obtained are described in Table 3.

Table 3.

Characteristics of the BB10 Mill.

CharacteristicsBB10 Mill 
Rotation speed (RPM)    52 
Internal diameter (mm)    400 
Length (mm)    120 
Steel ball bearing (kg)  50mm  4.5±0.1 
  30mm  3.0±0.1 
  25mm  2.5±0.1 
Input particle size (mm)    <3.15 
Input load (kg)   
3Results and discussion3.1Chemical analysis

The results of the KRS and GBFS chemical analyses are presented in Table 4.

Table 4.

Results of the GBFS and KRS chemical analyses.

Substance  GBFS (wt.%)  KRS (wt.%) 
Total CaO  42.8  47.8 
Loss on fire  –  18.3 
Total Fe  0.79  16.0 
SiO2  37.0  12.3 
Fe°  0.51  6.74 
FeO  0.54  6.66 
Al2O3  10.6  4.5 
Total C  0.17  4.2 
MgO  5.91  3.8 
Total S  1.04  1.25 
MnO  0.37  1.04 
P2O5  –  0.24 
Na20.29  0.12 
K20.34  0.034 
ZnO  0.0019  0.0031 

Based on the results, the GBFS is basically composed of CaO, SiO2, Al2O3, and MgO, representing 96.31% of the mass, with residual contents of other oxides in addition to a low percentage of total iron. In case of the KRS, 75.06% of the mass is composed of CaO, SiO2, Al2O3, and MgO, with an additional 16% total iron, of which 6.74% is Fe°, and a fire loss of 18.3%. The amounts of SiO2, Al2O3, and MgO are much lower in the KRS than in the GBFS, whereas the CaO content is higher in the former. Based on the CaO/SiO2 ratio, the slag is classified as basic (>1) or acid (<1). For the GBFS, the CaO/SiO2 ratio is 1.16, and for the KRS it is 2.99, and thus, both are basic.

Although the GBFS and KRS constituents are practically the same, they differ in content because of the type of manufacturing process: the GBFS is a by-product of pig iron production (in a blast furnace), whereas the KRS is a by-product of pig iron desulfurization (in the Kambara reactor).

The chemical analysis of the CP III 40 cement purchased for the execution of the research is presented in Table 5.

Table 5.

Chemical analysis of the CP III 40 cement.

Constituent  CP III (%) 
Al2O3  8.29 
Total CaO  50.5 
K20.35 
MgO  4.77 
SiO2  31.7 
Fe2O3  1.71 
SO3  2.20 
3.2Mineralogical analysis by X-ray diffraction

The GBFS X-ray diffractogram is shown in Fig. 2. The GBFS is predominantly amorphous, exhibiting only a crystalline peak at 2θ=30° relative to Ca(OH)2; in comparison, at other diffraction angles (2θ) it behaves amorphously.

Fig. 2.

GBFS X-ray diffractogram.

(0.07MB).

Fig. 3 shows the KRS X-ray diffractogram and reveals that the KRS is crystalline, with primary peaks of Ca(OH)2, graphite, CaCO3, and Ca2SiO4. This crystalline structure reduces the hydraulic activity [12,26,27].

Fig. 3.

KRS X-ray diffractogram.

(0.09MB).
3.3Thermogravimetry (TG) and differential scanning calorimetry (DSC)

Figs. 4–6 show the TG and DSC results of the KRS, GBFS, and CP III cement, respectively.

Fig. 4.

Thermogravimetric analysis and differential scanning calorimetry of the KRS.

(0.08MB).
Fig. 5.

Thermogravimetric analysis and differential scanning calorimetry of the GBFS.

(0.09MB).
Fig. 6.

Thermogravimetric analysis of cement CP III.

(0.05MB).

Fig. 4 shows the differential thermal analysis exhibiting three endothermic peaks in the thermolysis curve associated with mass losses. The first is near 100°C, characterizing loss of humidity, the second is between 350°C and 500°C, indicating the dehydration of Ca(OH)2, and the third peak is between 600°C and 700°C, which characterizes CaCO3 decomposition into CaO and CO2. Based on the mass loss percentage, the percentages of Ca(OH)2 and CaCO3 present in the sample were calculated, as done in [21].

Considering the stoichiometry of the dehydration reaction of Ca(OH)2, presented in Eq. [1], and the molar masses of the reagent and products, for each 74.1g of Ca(OH)2, 18g H2O was released, with the Ca(OH)2 mass being 4.11 times higher than the mass of water released. Thus, the percentage of the mass loss between 350°C and 500°C was multiplied by a factor of 4.11 to determine the Ca(OH)2 content.

Ca(OH)2CaO+H2O
%Ca(OH)2=3.14%×4.11=12.91%

Similarly, based on the mass loss between 700°C and 800°C and decomposition reaction of calcium carbonate (CaCO3) (Eq. [3]), the percentage of CaCO3 present in the sample was calculated (Eq. [4]).

CaCO3CaO+CO2
%CaCO3=4.72%×2.27=10.71%

Other calcium compounds likely present in the sample were CaSO4 owing to the desulfurization of the pig iron, and free CaO, as a function of the excess lime used in the desulfurization process.

Analyses of the thermolysis curve in Fig. 5 reveal that the GBFS slag has a slight and constant mass gain associated with the energy released throughout the test, which is more pronounced at 800°C.

According to Fig. 6, cement CP III suffers from a small mass loss at approximately 100°C and 600°C, whereas in other instances it slightly gains mass. We can attribute the mass gain to the presence of the GBFS in the CP III composition because the GBFS exhibits a similar behavior in Fig. 5.

3.4Granulometric analysis

The granulometric distribution results of the GBFS and KRSs are shown in Figs. 7–10.

Fig. 7.

Determination of the KRS particle size before grinding.

(0.21MB).
Fig. 8.

Determination of the GBFS granulometry before milling.

(0.22MB).
Fig. 9.

Granulometric distribution of the KRS through 150μm.

(0.12MB).
Fig. 10.

Granulometric distribution of the GBFS through 150μm.

(0.13MB).

Figs. 7 and 8 exhibit that the KRS has 21.16% particles that are smaller than 150μm, which is much larger than the value of 3.00% for the GBFS. In comparison, 85% of the GBFS particles are smaller than 1.18mm, whereas slightly more than 70% of the KRS particles are in this range.

The fineness modulus was calculated according to ABNT NBR NM 248:2003 [22], adding the retained mass percentages accumulated in the normal-series sieves and dividing by 100. The results were 2.34 and 2.54 for KRS and GBFS, respectively. Comparing the results in Table 6, both the KRS and GBFS fall in the optimal range of the granulometric distribution of fine aggregates for concrete.

Table 6.

Limits of the granulometric distribution of the fine aggregates [33].

Sieve with mesh opening  Retained mass percentage
  Lower limitsUpper limits
  Usable zone  Optimal zone  Optimal zone  Usable zone 
9.5mm 
6.3mm 
4.75mm  10 
2.36mm  10  20  25 
1.18mm  20  30  50 
600μm  15  35  55  70 
300μm  50  65  85  95 
150μm  85  90  95  100 

Notes: (1) The fineness modulus of the optimal zone varies from 2.20 to 2.90; (2) The fineness modulus of the lower usable zone ranges from 1.55 to 2.20; (3) The fineness modulus of the upper usable zone ranges from 2.90 to 3.50.

The maximum characteristic size corresponds to an opening in millimeters of the mesh, in which the accumulated retained percentage of the material is ≤5%. For the KRS, the maximum characteristic size obtained is 4.75mm, and for GBFS it is2.36mm.

For the KRS analysis, according to Fig. 8, d0.1=15.854μm, d0.5=68.198μm, and d0.9=156.193μm, implying that 10% of the fraction smaller than 150μm is less than 15.854μm, 50% lower than 68.198μm, and 90% lower than 156.193μm. The specific area of this fraction of the sample is 0.307m2/s.

The analysis of the fractions in 150μm of the GBFS (Fig. 10) yield values of d0.1=36.575μm, d0.5=119.932μm, and d0.9=267.780μm, in addition to a specific area of 0.0917m2/s. The d0.1, d0.5, and d0.9 of the GBFS are larger than those of the KRS, which corroborates with the smaller specific area.

Considering that the cement industry utilizes particle sizes smaller than 75μm, it is necessary to perform mechanical activation through grinding. Because the slag is comminuted, silicon chains will be destroyed, activating the oxygen ions bound to them, and thus, accelerating the dissolution rate of the slag, followed by precipitation of the hydrated products [28–30].

3.5Unit weight

The unit weight of the GBFS and KRS was determined by Kaeme Engenharia following the requirements of the ABNT NBR NM 45 standard [23]. The values used to determine the unit weight are described in Table 7. The mass value of the most aggregated vessel refers to the average of three weights. The KRS is denser than the GBFS, an expected result because the KRS is composed of a higher percentage of iron.

Table 7.

Values used to determine the unit weight of GBFS and KRSs.

VariablesKRS  GBFS 
mar  Mass of the most aggregated container (kg)  5.2947  4.4453 
mr  Empty container mass (kg)  0.2360  0.2360 
V  Container volume (m30.0035  0.0035 
pap  Aggregate unit mass (kg/m31437.47  1196.12 
3.6Visual and microscopic analysis

The SEM results for the GBFS and KRS are shown in Figs. 11 and 12, respectively.

Fig. 11.

Scanning electron microscopy of the GBFS.

(0.12MB).
Fig. 12.

Scanning electron microscopy of the KRS.

(0.15MB).

The KRS by-product shown in Fig. 12 has smaller particles, offering a larger specific area and causing the particles to aggregate, than the GBFS by-product particles (Fig. 11). Morphologically, the GBFS particles are polygonal and spherical (Fig. 11), whereas the KRS by-product particles are more spherical shapes, with a larger number of aggregates than in the former (Fig. 12).

In addition, EDS is performed for the GBFS and KRS samples to spectrally evaluate their elemental composition to identify their elements, thus providing a quantitative characterization of the samples. Fig. 13 illustrates the locations selected for the EDS and spectra 2, 3, and 4 of the GBFS by-product.

Fig. 13.

SEM of the selected sites for the EDS spectra generation of the GBFS by-product.

(0.19MB).

By EDS, O, Ca, Si, Al, and Mg are detected in the GBFS by-product, with O, Ca, and Si having the highest percentages. This corroborates the chemical analysis tests.

A similar analysis was performed for the KRS. The sites selected for the EDS and spectra 8 and 9 for this by-product are shown in Fig. 14. Because the KRS sample is more agglomerated, only two points are used when conducting the EDS test.

Fig. 14.

SEM of the selected sites for the generation of the KRS EDS spectra.

(0.19MB).

For the KRS, the chemical elements present in spectra 8 and 9 are O, Ca, Si, and Al, with O and Ca present in the highest percentages. These values corroborate the chemical analysis and X-ray diffraction tests for the by-product.

3.7pH

The slag pH was determined according to ABNT NBR 10004 [24] in the ArcelorMittal Tubarão Environmental Laboratory, and the results are provided in Table 8. The final pH results are the arithmetic averages of two measurements per slag type.

Table 8.

pH analysis results for the GBFS and KRS.

Slag  Average pH 
GBFS  10.310 
KRS  12.251 

The results show that the KRS is more alkaline than the GBFS. The higher pH of the KRS implies more OH ions are available when added to water, corroborating the TG result that showed 12.91% Ca(OH)2 in the KRS. The hydration of the GBFS in water occurs very slowly owing to the formation of a film lacking Ca2+, which inhibits the advance of hydration [31]. In contrast, the presence of hydroxyl facilitates the elution of Ca2+, Si4+, and Al3+ ions from the GBFS, contributing to the advancement of hydration [21]. Thus, the KRS might contribute to GBFS activation in the cement hydration process, so that they could be used together to produce Portland cement.

3.8Grinding

As mentioned above, CP III cement was used as a base for preparing the blends. Because the GBFS is already a constituent of CP III cement and is already widely used and standardized for such applications, the grinding test was performed only for the KRS.

For the grinding of the KRS, empirical tests were performed using a Marconi ball mill from the Ceramics Laboratory of the IFES Campus Vitória. A steel jug internally coated with high alumina was used, with an internal diameter of 17cm and a length of 23cm. The grinding balls were also made of high alumina. The time, ball size, and speed of rotation were varied to ensure the ground slag fineness conformed to the criterion established in ABNT NBR 16697 [25], according to which at most 8% of the retained product in sieve 200 with opening 75μm, can be retained. The grinding conditions of the tests are provided in Table 9.

Table 9.

Grinding tests of the GBFS and KRS by-products. Each grinding test is described because of the different conditions used in each test.

Test   
Balls mass (g)32mm  3397  3400 
19mm  1799  1800  1800  1800  3384  3387 
13mm  1583  1583  1580  1580  3380  3383 
Sample (g)1064  1064  1065  520  1064  1065 
Frequency (Hz)30  30  45  45  30  30 
Rotation (RPM)234  234  351  351  234  234 
Time (h)10 
Held in #0.15mm (%)–  49.5  44.5  56.8  38.5  39.2 
Held in #0.075mm (%)–  25.9  5.0  25.1  23.7  23.4 
Accumulated retention (%)–  75.4  49.5  81.9  62.2  62.6 

Calculations of the total ball mass, ball size distribution, and sample mass were performed according to the Herbst–Fuerstenau model [32].

All the beads were placed in a jar together with the sample to perform each of the grindings denoted as #1, #2, #3, and #4. At the end of the first grinding, the sample was still visually coarse, so that sieving was not performed to evaluate the fineness. After three millings while varying the time and frequency, a significant amount of material was retained in the 0.15-mm and 0.075-mm-sieves. The fourth test was performed by reintroducing a part of the test #3 sample for another 4h of milling, but the result was not satisfactory because the sum of the percentages retained in the 0.15-mm and 0.075-mm-sieves was higher than that for test 3. Thus, there was some error in the sifting process. In tests 5 and 6, milling was performed in three steps. At each step, the total mass of the single-sized balls was determined, starting with the larger 32-mm-balls, followed by the 19-mm and 13-mm-balls, respectively. Each milling lasted for 1h in test 5 and 2h in test 6. Milling in stages favored the reduction in the percentage of KRS retained in the 0.15-mm-sieve, but the sum of the total retained percentage remained high, and the best result was achieved in test 3, with 49.5% of slag higher than the percentage for the 0.075-mm-sieve.

After various millings under different conditions, the equipment used at laboratory scales was not efficient for grinding the KRS. Thus, a partner company, a cement manufacturer, was contacted to help in the grinding of this by-product.

The KRS was crushed prior to milling to ensure the input particles were smaller than 3.5mm, and then 1kg of the KRS was inserted into the mill along with the total ball load. The grinding process was performed using clinker grinding parameters to achieve a maximum of 11% retention in sieve 325 (0.045mm), and maintaining a fixed rotation speed, 4000 turns were performed. The slag was then sieved and 40% was retained in the 325 mesh, whereas the 200-mesh retained 23%.

The results of the last milling test showed an evolution relative to the previous results presented in Table 1, probably owing to the use of steel grinding bodies, which have a higher density than that of alumina, improving comminution. However, the result did not meet the cement standard requirement when evaluating the KRS alone, but considering that the additions are at most 30% KRS, it is expected that mixtures using this type of slag can be framed with the standard.

4Conclusions

In this study, the KRS had a considerably lower concentration of SiO2, Al2O3, and MgO and higher percentage of total Fe and Fe° compared to those of the GBFS. In addition, the KRS had a fire loss of 18.3%, a behavior not identified for the GBFS.

The GBFS was predominantly amorphous, whereas the main phases present in the KRS were Ca(OH)2, graphite, CaCO3, and Ca2SiO4.

Based on the TG, it was estimated that the KRS had 12.91% Ca(OH)2 and 10.71% CaCO3. However, the GBFS showed a slight and constant mass gain associated with the release of energy throughout the test, possibly owing to crystallization and oxidation, and some decomposition reactions occurred that were exothermic.

The KRS had a maximum characteristic particle size of 4.75mm, whereas the GBFS particles were at most 2.36mm.

Based on the fineness modulus (2.34 and 2.54 for the KRS and GBFS, respectively), both the KRS and GBFS fineness values were classified in the optimal zone according to ABNT NBR 7211:2005 [33].

The KRS had a specific area of 0.307m2/s and the GBFS had 0.0917m2/s, in the final fraction of 150μm (21.16% of the KRS and 3.00% of the GBFS).

It was necessary to perform mechanical activation by grinding the slag because the cement industry works with particle sizes smaller than 75μm.

The KRS had a unit weight larger than that of the GBFS (1437.47kg/m3 and 1196.12kg/m3, respectively), which was already expected owing to the higher total iron content of the KRS.

In the SEM analyses, the predominantly spherical KRS particles aggregated more than the GBFS particles, which had polygonal and spherical morphologies.

The KRS was more alkaline than the GBFS, which may have contributed the OH ions in the cement hydration process.

The grinding of the KRS using the Marconi equipment with alumina balls was not efficient possibly owing to the lower density of the balls compared to that of the steel balls.

The grinding of the KRS using the same clinker milling conditions (BB10 mill with steel balls) was more efficient, reaching 23% of material retention in a 75-μm-sieve. However, the fineness modulus required for the clinker was not reached for the KRS.

Thus, it was concluded that KRS is a by-product that offers the possibility of use for the manufacture of Portland cement, although it will be necessary to perform thermal, mechanical, or chemical activations so that the cement requirements are reached.

Conflicts of interest

The authors declare no conflicts of interest.

Acknowledgements

The authors would like to thank the Federal Institute of Espírito Santo for the financial contribution intended for the revision of the article, and Arcelor Mittal Tubarão for the Kambara Reactor scrap donation.

References
[1]
M. Öner, K. Erdoğdu, A. Günlü.
Effect of components fineness on strength of blast furnace slag cement.
Cem Concr Res, 33 (2003), pp. 463-469
[2]
O. Burciaga-Díaz, I. Betancourt-Castillo.
Characterization of novel blast-furnace slag cement pastes and mortars activated with a reactive mixture of MgO–NaOH.
Cem Concr Res, 105 (2018), pp. 54-63
[3]
L. Courard, F. Michel.
Limestone fillers cement based composites: effects of blast furnace slags on fresh and hardened properties.
Constr Build Mater, 51 (2014), pp. 439-445
[4]
S. Miyazawa, T. Yokomuro, E. Sakai, A. Yatagai, N. Nito, K. Koibuchi.
Properties of concrete using high C3S cement with ground granulated blast-furnace slag.
Constr Build Mater, 61 (2014), pp. 90-96
[5]
K.H. Mo, U.J. Alengaram, M.Z. Jumaat.
Utilization of ground granulated blast furnace slag as partial cement replacement in lightweight oil palm shell concrete.
Mater Struct, 48 (2015), pp. 2545-2556
[6]
B.S. Divsholi, T.Y.D. Lim, S. Teng.
Durability properties and microstructure of ground granulated blast furnace slag cement concrete.
Int J Concr Struct Mater, 8 (2014), pp. 157-164
[7]
J. Lou, K. Bhalerao, A.B.O. Soboyejo, W.O. Soboyejo.
An investigation of the effects of mix strength on the fracture and fatigue behavior of concrete mortar.
J Mater Sci, 41 (2006), pp. 6973-6977
[8]
E. Özbay, M. Erdemir, H.İ. Durmuş.
Utilization and efficiency of ground granulated blast furnace slag on concrete properties – A review.
Constr Build Mater, 105 (2016), pp. 423-434
[9]
B. Cho, H. Choi.
Physical and chemical properties of concrete using GGBFS-KRS-gypsum binder.
Constr Build Mater, 123 (2016), pp. 436-443
[10]
T. Proske, M. Rezvani, S. Palm, C. Müller, C.A. Graubner.
Concretes made of efficient multi-composite cements with slag and limestone.
Cem Concr Compos, 89 (2018), pp. 107-119
[11]
E. Crossin.
The greenhouse gas implications of using ground granulated blast furnace slag as a cement substitute.
J Clean Prod, 95 (2015), pp. 101-108
[12]
Y. Jiang, T.C. Linga, C. Shi, S.Y. Pan.
Characteristics of steel slags and their use in cement and concrete – A review.
Resour Conserv Recycl, 136 (2018), pp. 187-197
[13]
G.C. Ulubeyli, R. Artir.
Sustainability for blast furnace slag: use of some construction wastes.
Procedia Soc Behav Sci, 195 (2015), pp. 2191-2198
[14]
R.A. Robayo-Salazar, M. Gutiérrez, F. Puertas.
Study of synergy between a natural volcanic pozzolan and a granulated blast furnace slag in the production of geopolymeric pastes and mortars.
Constr Build Mater, 157 (2017), pp. 151-160
[15]
Z. Li, S. Li.
Carbonation resistance of fly ash and blast furnace slag based geopolymer concrete.
Constr Build Mater, 163 (2018), pp. 668-680
[16]
A. Markandeya, N. Shanahan, D.M. Gunatilake, K.A. Riding, A. Zayed.
Influence of slag composition on cracking potential of slag-Portland cement concrete.
Constr Build Mater, 164 (2018), pp. 820-829
[17]
M.M. Lópeza, Y. Pineda, O. Gutiérrez.
Evaluation of durability and mechanical properties of the cement mortar added with slag blast furnace.
Procedia Mater Sci, 9 (2015), pp. 367-376
[18]
J. Zhao, D. Wang, P. Yan.
Design and experimental study of a ternary blended cement containing high volume steel slag and blast-furnace slag based on Fuller distribution model.
Constr Build Mater, 140 (2017), pp. 248-256
[19]
S.Z. Carvalho, F. Vernilli, B. Almeida, M. Demarco, S.N. Silva.
The recycling effect of BOF slag in the Portland cement properties.
Resour Conserv Recycl, 127 (2017), pp. 216-220
[20]
S.Z. Carvalho, F. Vernilli, B. Almeida, M.D. Oliveira, S.N. Silva.
Reducing environmental impacts: the use of basic oxygen furnace slag in Portland cement.
J Clean Prod, 172 (2018), pp. 385-390
[21]
B. Cho.
Hydration characteristics of cement-free binder using Kambara reactor slag.
Mag Concr Res, 68 (2016), pp. 1143-1154
[22]
Associação brasileira de normas técnicas.
NBR NM 248:2003 aggregates – sieve analysis of fine and coarse aggregates.
(2003),
[23]
Associação brasileira de normas técnicas.
NBR NM 45:2006 aggregates – determination of the unit weight and air-void contents.
(2006),
[24]
Associação brasileira de normas técnicas.
NBR 10004:2004 – solid waste – classification.
(2004),
[25]
Associação brasileira de normas técnicas.
NBR 16697:2018 Portland cement – requirements.
(2018),
[26]
S. Kourounis, S. Tsivilis, P.E. Tsakiridis, G.D. Papadimitriou, Z. Tsibouki.
Properties and hydration of blended cements with steelmaking slag.
Cem Concr Res, 37 (2007), pp. 815-822
[27]
H.S. Kim, K.S. Kim, S.S. Jung, J.I. Hwang, J.S. Choi, I. Sohn.
Valorization of electric arc furnace primary steelmaking slags for cement applications.
Waste Manage, 41 (2015), pp. 85-93
[28]
R. Kumar, S. Kumar, S. Badjena, S.P. Mehrotra.
Hydration of mechanically activated granulated blast furnace slag.
Metall Mater Trans, 36B (2005), pp. 873-883
[29]
I.G. Richardson.
Tobermorite/jennite- and tobermorite/calcium hydroxide-based models for the structure of C–S–H: applicability to hardened pastes of tricalcium silicate, h-dicalcium silicate, Portland cement, and blends of Portland cement with blast-furnace slag, metakaolin, or silica fume.
Cem Concr Res, 34 (2004), pp. 1733-1777
[30]
I.G. Richardson, J.G. Cabrera.
The nature of C–S–H in model slag-cements.
Cem Concr Compos, 22 (2000), pp. 259-266
[31]
S.-D. Wang, K.L. Scrivener, P.L. Pratt.
Factors affecting the strength of alkali-activated slag.
Cem Concr Res, 24 (1994), pp. 1033-1043
[32]
J.A. Herbst, D.W. Fuerstenau.
Scale-up procedure for continuous grinding mill design using population balance models.
Int J Miner Process, 7 (1980), pp. 1-31
[33]
Associação brasileira de normas técnicas.
NBR 7211:2005 Agregados para concreto – especificação.
(2005),
Copyright © 2019. Brazilian Metallurgical, Materials and Mining Association
Journal of Materials Research and Technology

Subscribe to our newsletter

Article options
Tools
Cookies policy
To improve our services and products, we use cookies (own or third parties authorized) to show advertising related to client preferences through the analyses of navigation customer behavior. Continuing navigation will be considered as acceptance of this use. You can change the settings or obtain more information by clicking here.