Professional Documents
Culture Documents
Abstract: Steel slag, one of the by-products from the production of steel, is crushed, sized properly, and treated in relation to its volume
Downloaded from ascelibrary.org by East Carolina University on 07/16/18. Copyright ASCE. For personal use only; all rights reserved.
swelling potential to be used as a highly resistant aggregate called steel aggregate or steel gravel. This study analyzes the swelling mech-
anisms of steel aggregate. By laboratory testing, the volumetric swelling potential in samples of steel aggregate and samples of lateritic soil
and steel aggregate mixtures was determined in two different proportions: 90% soil þ10% steel aggregate (M9010); and 80% soil þ20% steel
aggregate, by weight (M8020). This analysis is part of a project that assesses the feasibility of the utilization of this coproduct in the con-
struction industry in Brazil. Average swelling values of 0.92, 0.73, and 0.61% were obtained from the steel aggregate samples and from the
M9010 and M8020 mixture samples, respectively. According to Brazilian and Japanese standard volumetric swelling limits, all tested ma-
terials are suitable for the use of paving as an alternative to the use of natural aggregates. DOI: 10.1061/(ASCE)EE.1943-7870.0001435.
© 2018 American Society of Civil Engineers.
Introduction All of those characteristics allow its use for extended periods with
minimal maintenance (Rohde 2002).
In the United States, approximately 21 million tons of steel industry One of the most significant difficulties, however, in the use of
slag are produced each year. Also, there are many productive com- steel slag as aggregate in layers of pavements is controlling its ex-
mercial uses of slag that have been developed, such as road bed and pansive potential, which can reach 10.0% (Silva 2001; Castelo
fill material (Proctor et al. 2000; USGS 1997). In Germany and the Branco 2004). That swelling generates internal stresses that cause
other European countries, the steel industry is concentrated on in- cracks in the structural layers of the roads and decrease their
creasing the utilization rate of industrial coproducts and recycling strength and durability. Also, steel slag is rich in calcium oxide
materials (Motz and Geiseler 2001; Krass et al. 1999). Slag repre- (CaO) content, whose uncontrolled seepage has been reported to
sents 29% of the total coproducts and wastes generated by the steel impair the quality of receiving surface waters and aquifers
industry in Brazil (IABR 2014). It is a material that presents fun- (Koryak et al. 2002; Roadcap et al. 2005; Mayes et al. 2008),
damental functions in the steel production process to ensure good but there are methods to mitigate this problem (Özkök et al.
quality of the final product. In the recent past, slag was considered a 2015), and there is a concern regarding the leaching of heavy met-
waste of little or no economic value. That point of view has als present in steel slag (arsenic, barium, cadmium, chromium, and
changed, and slag is already considered a by-product capable of lead) into the groundwater, however, their concentration levels may
conferring important added value to the materials in which it is ap- be tolerable (Yildirim and Prezzi 2015).
plied (Graffitti 2002). Expansive reactions involve specific compounds present in steel
Many studies have been performed using steel slag in pave- slag. Among those volumetrically unstable compounds, CaO and
ments (Timothy 1992; Xue et al. 2006; Wu et al. 2007; Maharaj magnesium oxide (MgO) are the main compounds responsible for
et al. 2017). In Brazil, steel slag has been used as small and large the weakening due to the molar volume difference in their reactions
aggregate in road paving, being particularly suitable for its hard- (Machado 2000). Steel slag contains a significant amount of steel or
ness, durability, free drainage, and high resistance to crushing. iron particles that get incorporated into the slag when oxygen is
introduced. Those particles can contribute to the volumetric insta-
1
Researcher, Dept. of Civil and Environmental Engineering, Univ. of bility of the slag through the generation of corrosion and oxidation
Brasilia, CEP 70910-900 Brasília, Brazil (corresponding author). Email: products (Machado 2000). Another reaction that determines the
victorh.souzaoliveira@gmail.com
2 swelling of the steel slag is the allotropic transformation of the di-
Adjunct Professor, Dept. of Civil and Environmental Engineering,
Universidad de la Costa, Calle 58, #55-66, 080002 Barranquilla, Atlántico,
calcium silicate or calcium orthosilicate (2CaO · SiO2 or Ca2 SiO4 )
Colombia; formerly, Researcher, Dept. of Civil and Environmental which favor the worsening of the effects of the expansibility caused
Engineering, Univ. of Brasilia, CEP 70910-900 Brasília, Brazil. Email: by CaO and MgO. Despite the necessity of adopting curing tech-
ntarazon2@cuc.edu.co niques for the use of this material, the possibility of the execution of
3
Associate Professor, Dept. of Civil and Environmental Engineering, thinner structural layers for the same load resistance and durability
Univ. of Brasilia, CEP 70910-900 Brasília, Brazil. Email: lmartins@ contributes to the economic advantages of using slag as aggregates
unb.br in the construction of pavements. Table 1 shows the chemical char-
4
Associate Professor, Dept. of Civil and Environmental Engineering, acteristics of steel slag used in this research obtained by X-ray dif-
Univ. of Brasilia, CEP 70910-900 Brasília, Brazil. Email: abrasil@unb.br
Note. This manuscript was submitted on December 9, 2016; approved
fraction (XRD) analyses.
on March 30, 2018; published online on June 25, 2018. Discussion period Crushing, granulometric stabilization, and reduction of the
open until November 25, 2018; separate discussions must be submitted swelling potential of the steel slag result in the stabilized and stand-
for individual papers. This paper is part of the Journal of Environmental ardized coproduct known as steel gravel or steel aggregate. In this
Engineering, © ASCE, ISSN 0733-9372. context, this study aims to evaluate the feasibility of using this
Rutile TiO2
Sulfuric (Sulfur) S
Table 3. Physical properties of the soil
Property Value
wL (%) 68.50
coproduct based on the technical specifications related to the pro- wP (%) 47.16
cess of aggregate swelling in layers of road pavements. Thus, the Plasticity index (%) 21.34
objective of this paper is to evaluate the different applicability of Unified soil classification system (USCS) Inorganic Silts
steel reinforcement in pavements replacing the natural aggregate by γ s (kN=m3 ) 29.51
means of the comparative analysis between the volumetric swelling wot (%) 31.0 (SCE)
of the steel gravel and of the mixtures composed of steel gravel and 26.5 (MCE)
soil in different proportions. γ dmax (kN=m3 ) 14.00 (SCE)
15.50 (MCE)
Note: MCE = modified compaction energy; and SCE = standard
Materials and Methods compaction energy.
80.0
pressed by the extensometers with the submerged samples at a
temperature of 69 3°C were taken. The measurements were re-
60.0
corded daily for 7 days, 2 h after the addition of water, which
should be sufficient for the sample to remain completely sub-
40.0 merged. After that period, the water was partially removed from
the vessel at the beginning of the second stage of the test, which
20.0 with chemical breakdown consists of maintaining the samples in a saturated, not submerged,
no chemical breakdown
condition for another 7 days. To guarantee the saturation of the
0.0 samples, 500 mL of water were added daily. After the 2 weeks
0.001 0.010 0.100 1.000 10.000 100.000 of swelling evaluation, the samples were removed from the mold
Particle Size (mm) and dismantled. The volumetric swelling is expressed as a percent-
age, given by the reading after 14 days, over the initial height of the
Fig. 1. Particle-size distribution curves.
sample, according to Eq. (1)
Table 4 presents the main values of the volumetric swelling of the Time (Days)
pure steel gravel samples obtained from the results of the test pro-
posed by ME 113 (DNIT 2009). Fig. 2. Cumulative expansion values of the steel gravel over the test
Day 1 represents the reading 24 h after the start of the test in the time.
submerged condition of the samples at a temperature of 69°C.
Day 7 corresponds to half the test time, when the samples were
removed from the submerged condition, and their behavior was
evaluated in the saturated, not submerged, condition. Day 14 rep-
resents the final reading, with the maintenance of the samples in the
saturated, not submerged, condition.
NBR 16364 (ABNT 2015), JIS A 5015 (JSA 2013), and ASTM
2940-98 (ASTM 1998) establish the maximum limit of swelling of
the average of the three samples at 3, 2, and 0.5%, respectively. At
the end of 14 days of the test, the maximum value of 0.96% was
obtained in the compacted sample with moisture content 2% above
the OMC. The OMC sample presented the smallest swelling among
the samples analyzed, with a value of 0.87%. The average obtained
from the three results was 0.92%, considerably below the limits
established by the Brazilian and Japanese standards. Fig. 2 shows
the swelling variation described in Table 4 as a function of the
test time.
The final swelling presented by SG-02 corresponds to the opti-
mal moisture condition, and it was the smallest of the three samples Fig. 3. Steel gravel sample after expansion test.
tested. Its rate of swelling (cumulative swelling value divided by
time) was the highest during the first 24 h of the test, gradually
decreasing until the end of the immersion stage. In the saturation
slope decreased until the end of the submersion stage. The transi-
stage, not the submerged condition, that behavior was continued
tion between stages, submerged to nonsubmerged condition, re-
and a second-degree polynomial could approximate it until the
sulted in the contraction of the sample, represented in the
12th day, when the rate became constant.
inflection observed between Days 7 and 8. From Day 8, the swell-
The sample with moisture content 2% above the OMC (SG-03)
ing rate remained constant until the end of the test time.
presented the highest accumulated swelling among the three analy-
Despite the close swelling values presented by the samples at
ses. Its rate of swelling at the beginning of the trial was the lowest,
the end of the 14 days, there are differences in the behavior of their
and it turned out to be almost constant during both stages of the
curves over the analyzed period. That fact can be explained by the
experiment.
saturation process that occurs inside each sample. Since the SG-03
The material compacted in the sample with moisture content 2%
sample was more saturated, its swelling rate showed the highest
below the OMC (SG-01) had the lowest swelling rate during the
constancy among them. Similarly, the less-saturated specimen
first 24 h of the test. That value increased between the first and
(SG-01) showed higher sensitivity to saturation conditions, with
second days and became constant until the fourth day, when this
more pronounced swelling rate variations.
At the end of the swelling test, the samples were carefully de-
molded and dismantled as indicated in the ME 113 test procedures
Table 4. Cumulative expansion values of the steel gravel over the test time (DNIT 2009). The absence of crystals on the surface of the steel
Sample gravel was observed, as shown in Fig. 3.
SG-01 SG-02 SG-03
Moisture content (%) M8020 Mixture
6.7 8.7 (OMC) 10.7 Table 5 shows the volumetric swelling values of the samples of the
Condition Day Cumulative expansion (%)
M8020 mixture obtained from the results of the test proposed by
ME 113 (DNIT 2009).
Immersed 1 0.03 0.21 0.07 Day 1 represents the reading 24 h after the start of the test in the
7 0.77 0.74 0.52 submerged condition at a temperature of 69°C. Day 7 corresponds
Saturated 14 0.92 0.87 0.96
to half of the test time, at which time the samples were removed
0.80 able for the use in pavement structures from the point of view of
M80-02 - 11.49% (OMC)
volumetric stability. Fig. 6 shows the cumulative swelling values
0.60 M80-03 - 13.49% described in Table 6 as a function of the test time. As expected
due to the increase of the soil content in the mixtures was observed.
Fig. 6. Cumulative expansion values of the M9010 mixture over the That fact results from the absorption of the swelling due to the in-
test time. crease of the presence of the fines, which are part the granulometry
of the natural soil.
All the samples fit the limit of 3% of the volumetric swelling for
use in structural layers of road pavements established by NBR
and in line with the results obtained from the tests of the steel gravel 16364 (ABNT 2015) and 2% established by JIS A 5015 (JSA
samples and the M8020 mixture, the saturation conditions inside 2013), although it was a little bigger than the limit of 0.5% estab-
each sample of the M9010 mixture determine the behavior of its lished by ASTM D2940-98 (ASTM 1998). The fact denotes effi-
accumulated swelling curves. ciency in the steel slag process of volumetric stabilization for its
Up to the fourth day, the compacted sample with moisture con- transformation into the steel gravel by-product. All tested materials
tent 2% below OMC (M90-01) presented the lowest cumulative are suitable for the use of paving as an alternative to the use of
swelling rate, which increased abruptly, resulting in the inflection natural aggregates from the point of view of volumetric stability.
observed on the fifth day. From this point, the saturation level of the The determination of the ideal proportion of steel gravel and soil
sample results in a cumulative swelling rate similar to the one pre- depends, therefore, on the availability of the components of the
sented by the compacted sample in the optimum moisture condi- mixture at the construction site of the highway.
tion, which became practically zero after Day 11. Therefore, the
curve shows a behavior governed by a third-degree polynomial.
The cumulative swelling rate observed in the compacted sample Acknowledgments
in the optimum moisture condition (M90-02) was higher until the
fifth day, decreasing gradually until the 10th day, at which point it The authors acknowledge the support of the following agencies: the
became zero. Similar to that observed in the steel gravel and Brazilian Council for Scientific and Technological Development
M8020 mixture analysis, the predominant behavior of the curve (CNPq) (Grant No. 304721/2017-4), the Coordination for the
could be approximated by a second-degree polynomial. The sample Improvement of Higher Level Personnel (CAPES), the Support
compacted with moisture content 2% greater than OMC (M90-03) Research of the Federal District Foundation (FAP-DF), the
showed a constant swelling rate during the entire period of the test, Brazilian Transport Infrastructure Department (DNIT), the Brazil
resulting in a graph with a linear trend, possibly explained by the Steel Institute (BSI), and the University of Brasilia (UnB) for fund-
saturation level of the sample, which was the highest of the ana- ing this research.
lyzed samples.
After the samples were demolded and dismantled, the absence
of crystals on the surface of the grains was observed, as shown References
in Fig. 7.
ABNT (Brazilian Association of Technical Standards). 1986. Soil samples:
Preparation for compaction tests and characterization tests. NBR
6457-86. Rio de Janeiro, Brazil: ABNT.
ABNT (Brazilian Association of Technical Standards). 1987.
“Soil—California bearing ratio.” NBR 9895-87. Rio de Janeiro, Brazil:
ABNT.
ABNT (Brazilian Association of Technical Standards). 2015. Execution of
sub-base and base stabilized granulometrically with steel aggregate for
road paving. NBR 16364. Rio de Janeiro, Brazil: ABNT.
ASTM. 1998. Standard specification for graded aggregate material
for bases or subbases for highways or airports. ASTM D2940-98.
West Conshohocken, PA: ASTM.
Castelo Branco, V. T. 2004. “Characterization of asphalt mixtures with
the use of steel slag as aggregate.” Master’s dissertation, Dept. of Civil
Engineering, Federal Univ. of Rio de Janeiro.
da Silva, E. A. 1991. “Use of steel slag in road paving.” In 28th Annual
Paving Meeting. Rio de Janeiro, Brazil: Brazilian Paving Association.
DNIT (National Department of Transport Infrastructure). 2009. Road
paving: Artificial aggregate—Evaluation of the expansion potential
Fig. 7. Sample of the M9010 mixture after expansion test.
of steel slag. ME 113. Rio de Janeiro, Brazil: DNIT.
Pennsylvania stream.” J. Freshwater Ecol. 17 (3): 461–465. https://doi Federal Univ. of Rio Grande do Sul.
.org/10.1080/02705060.2002.9663921. Silva, P. 2001. “Concrete reinforcement study of portland cement at the
Krass, K., C. Kellermann, and M. Rohleder. 1999. “Abfall, Aufbereitung experimental circular road of the institute of highway research.”
und Verwertung von Recycling-Baustoffen und industriellen Nebenpro- Ph.D. thesis, Dept. of Civil Engineering, Federal Univ. of Rio de
dukten im Wirtschaftsjahr 1997-Teil 2: Industrielle Nebenprodukte.” Janeiro.
Strasse und Autobahn 50 (9): 473–478. Timothy, R. 1992. Steel slag aggregates in Bituminous mixtures. Final
Machado, A. T. 2000. “Comparative study of test methods for evaluation Rep. Springfield, VA: Bureau of Bridge and Roadway Technology,
of steel slag expandability.” Master’s dissertation, Dept. of Civil Pennsylvania Dept. of Transportation.
Engineering, Univ. of São Paulo Polytechnic School. USGS. 1997. Mineral industry surveys: Slag-iron and steel, annual
Maharaj, C., D. White, R. Maharaj, and C. Morin. 2017. “Re-use of steel review—1997. Reston, VA: USGS.
slag as an aggregate to asphaltic road pavement surface.” Cogent Eng. Wu, S., Y. Xue, Q. Ye, and Y. Chen. 2007. “Utilization of steel slag as
4 (1): 1416889. https://doi.org/10.1080/23311916.2017.1416889. aggregates for stone mastic asphalt (SMA) mixtures.” Build Environ.
Mayes, W. M., P. L. Younger, and J. Aumônier. 2008. “Hydrogeochemistry 42 (7): 2580–2585. https://doi.org/10.1016/j.buildenv.2006.06.008.
of alkaline steel slag leachates in the U.K.” Water Air Soil Pollut. Xue, Y., S. Wu, H. Hou, and J. Zha. 2006. “Experimental investigation
195 (1–4): 35–50. https://doi.org/10.1007/s11270-008-9725-9. of basic oxygen furnace slag used as aggregate in asphalt mixture.”
Motz, H., and J. Geiseler. 2001. “Products of steel slags an opportunity to J. Hazard. Mater. 138 (2): 261–268. https://doi.org/10.1016/j
save natural resources.” Waste Manage. 21 (3): 285–293. https://doi.org .jhazmat.2006.02.073.
/10.1016/S0956-053X(00)00102-1. Yildirim, I. Z., and M. Prezzi. 2015. “Geotechnical properties of fresh and
Özkök, E., A. P. Davis, and A. H. Aydilek. 2015. “Treatment methods for aged basic oxygen furnace steel slag.” J. Mater. Civ. Eng. 27 (12):
mitigation of high alkalinity in leachates of aged steel slag.” J. Environ. 04015046. https://doi.org/10.1061/(ASCE)MT.1943-5533.0001310.