Construction and Building Materials 283 (2021) 122624

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

Roller compacted geopolymer concrete using recycled concrete

aggregate
Sk Syfur Rahman, Mohammad Jamal Khattak ⇑
Department of Civil Engineering, University of Louisiana at Lafayette, Lafayette, LA 70504, United States

h i g h l i g h t s

Roller compacted geopolymer concrete was successfully developed using recycled concrete aggregates.

Ambient temperature curing of 56d gained 80% of the maximum strength of 48-hr oven curing at 60 °C.

Roller compacted geopolymer concrete showed higher mechanical properties than similar cement concrete.

Microstructure analysis showed the presence of geo-polymeric compounds.

a r t i c l e i n f o a b s t r a c t

Article history: In this research mechanical characteristics of roller compacted geopolymer concrete (RCGPC) and roller
Received 15 September 2020 compacted cement concrete (RCC) using 100% recycled concrete aggregate (RCA) were investigated and
Received in revised form 14 December 2020 compared. The surface morphology and formation of chemical compounds of the selected RCGPC com-
Accepted 3 February 2021
posite mixtures were also analyzed using Scanning electron microscopy (SEM), X-Ray diffraction
Available online 28 February 2021
(XRD), and Fourier-transform infrared spectroscopy (FTIR). It was revealed from the investigation that
the compressive strength of the RCGPC material was dependent on the mixture variables like the molar
Keywords:
concentration of sodium hydroxide (NaOH), sodium silicate (Na2SiO3) to NaOH ratio (alkali activator),
RCA
RCC
curing time and temperature, and dosage of fly ash. The experimental results depicted that the compres-
Geopolymer sive strength, modulus of elasticity, and flexural strength of RCGPC composites varied in the range of 7.5
Roller compacted concrete to 27.6 MPa, 16 to 46.8 GPa, and 2.1 to 3.8 MPa, respectively. While RCC using 12% ordinary cement
Concrete yielded a compressive strength, modulus, and flexural strength of 18.9 MPa, 32.7 GPa, and 2.1 MPa,
GPC respectively. SEM, XRD, and FTIR analysis also revealed the formation of geopolymer compounds and
SEM their relative presence in the new composite based on the mixture variables. It is believed that the devel-
FTIR oped RCA based roller compacted geopolymer concrete could be used as an alternative material for pave-
XRD
ment construction.
Microstructure
Ó 2021 Elsevier Ltd. All rights reserved.

1. Introduction and background crete, buildings, and other concrete structures. RCA generated from
C&D is getting more and more attention nowadays since its use
1.1. Introduction could contribute to the solution of the mass storage problems, help
to reduce environmental pollution, save natural resources, and
Each year an enormous quantity of construction and demolition eventually reduce construction costs [2–4]. Many researchers have
(C&D) wastes are produced in the United States of America (USA). investigated the mechanical and durability properties of RCA,
Estimation by the Environment Protection Agency (EPA) showed which revealed the low strengths and durability as compared to
that around 600 million tons of C&D wastes were generated in the natural aggregates [5–8]. Because of the low strength and dura-
2018, only [1]. The majority of these wastes are from the construc- bility, these are normally used in smaller quantities in the new
tion and rehabilitation of structures, which consist of concrete, concrete production. Other suitable uses of RCA are landfill, low
asphalt, and masonry. Recycle concrete aggregate (RCA) is mainly strength base material, etc. RCA has been successfully used in
the C&D wastes reclaimed from demolished highways with con- cement concrete production as a replacement of a smaller percent-
age of natural aggregate without much compromising the strength
⇑ Corresponding author. [5]. Portland cement concrete (PCC) has been widely used as a con-
E-mail address: khattak@louisiana.edu (M.J. Khattak). struction material for a long time. The increase in the use of PCC

https://doi.org/10.1016/j.conbuildmat.2021.122624
0950-0618/Ó 2021 Elsevier Ltd. All rights reserved.
Sk Syfur Rahman and Mohammad Jamal Khattak Construction and Building Materials 283 (2021) 122624

demands more cement production, which requires more and more 0, 5, 10, and 15% were incorporated. The density and compressive
raw materials extraction and at the same time, more CO2 is being strengths were investigated at 28 days of curing. Optimum mixture
produced which increased environmental concerns. It is reported variables were also determined, and they found that 10 M NaOH,
that to manufacture one ton of Portland cement, 1.6 tons of raw Na2SiO3 to NaOH ratio of 1, 10% OPC content by weight of fly
materials and about 6.5 million BTUs of energy are needed, conse- ash, liquid/solid binder ratio of 1.4, recycled lightweight concrete
quently about one ton of CO2 is released into the atmosphere aggregate (RLCA)/solid binder (fly ash + OPC) ratio of 1.8 and curing
[9,10]. Due to these issues with cement, there is a recent need to temperature of 60 °C were the optimum mixture constituents.
reduce its usage in the construction industry. Enormous efforts They also concluded that the incorporation of OPC increased the
have been made throughout the world to reduce the use of Port- density of the composite. Nuaklong et al. [18] evaluated the
land cement in concrete. These include the utilization of waste mechanical and durability properties of geopolymer concrete con-
by-product materials and the development of alternative binders taining recycled concrete aggregate. OPC and nano-SiO2 were also
to Portland cement [9,11]. Geopolymer technology could be con- incorporated to enhance mechanical and durability characteristics.
sidered as a potential alternative to address this issue. The test results revealed that the addition of OPC increased the
Geopolymer promises to ensure cost-effective and advanced compressive strength and minimized water absorption, porosity,
solutions, which could improve the engineering and environmen- and chloride penetration. They also found a 38% higher strength
tal performance of pavement structures. Geopolymer-based prod- than natural aggregate geopolymer concrete (NAGC) when using
ucts are not only environmentally friendly but also require only 15% OPC-fly ash in the mixture. The addition of nano-SiO2 also
modest energy to produce. Geopolymer technology can be a poten- slightly improved the compressive strength relative to the NAGC.
tial alternative to ordinary Portland cement (OPC) and minimize Koushkbaghi et al. [19] investigated the properties of
CO2 emission by around 80% and utilize potentially hazardous metakaolin-based geopolymer concrete containing recycled con-
industrial waste as a valuable construction material [13]. Industrial crete aggregate by partially replacing natural aggregate. They tried
by-products such as blast furnace slag, fly ash, rice husk ash, and to reveal the effect of Na2SiO3 to NaOH ratio and the percentage of
clayey materials, which are rich in aluminum–silicate products, recycle concrete aggregate used on the mechanical and durability
are the right candidates for producing geopolymer binders [12]. properties of the newly formed composite. Although the incorpora-
Such materials are activated using the alkaline solutions including tion of recycled aggregate reduced compressive strength, they con-
sodium hydroxide/sodium silicate to create three-dimensional cluded the increasing trend of compressive strength as sodium
polymerized molecular chains and network that ultimately devel- silicate to sodium hydroxide ratio increases, but decrement of
ops hardened binder. The use of RCA in a combination with fly ash- chloride ion permeability and absorption. Morphology investiga-
based geopolymer binder could be more beneficial because both tion using scanning electron microscopy (SEM) also revealed that
waste products, fly ash and RCA will be utilized which will result as the Na2SiO3 to NaOH ratio increases, the debonding width at
in both cost-saving and reduced CO2 emission. the interfacial transition zone between RCA and binder reduces
and the final geopolymer product becomes more uniform and
homogenous in terms of monomer ratios. They found higher den-
1.2. Geopolymer concrete containing RCA sity and lower porosity at Na2SiO3 to NaOH ratio of 3.

Researchers are already using geopolymer cement in the stabi- 1.3. Research need and scope
lized base or subbase incorporating soil, natural, and recycle aggre-
gates [14-20]. Posi et al. [14] explored lightweight geopolymer In the above-described studies, researchers have found the use
concrete containing recycled aggregate from the lightweight block. of RCA in smaller amounts yielded acceptable mechanical proper-
Fine, medium, and coarse aggregates were generated by crushing ties for both conventional and geopolymer concretes. However, its
recycled blocks. Different mix variables like alkali to fly ash ratio, utilization for higher dosages compromised the strength and stiff-
Na2SiO3 to NaOH ratio, the molar concentration of NaOH, aggregate ness. This study focuses on the potential improvement of mechan-
to fly ash ratio were varied in the study. The results revealed that ical properties of 100% RCA by incorporating the idea of roller-
28-day strength and density was satisfactory. They also investi- compacted geopolymer concrete. Roller compacted concrete
gated water absorption, porosity, and modulus of elasticity of the (RCC) is a different type of concrete that is dryer than conventional
composite and suggested the material could be used for wall and concrete and must be compacted after placement like hot mix
partition. Sata et al. [15] investigated pervious geopolymer con- asphalt (HMA) or soils. It contains the same material as conven-
crete containing high calcium fly ash and recycled aggregate. They tional concrete except it has a reduced cement content (around
evaluated not only the compressive strength and the split tensile 200 kg/m3) and a zero-slump requirement [21]. High durability
strength but also the total void ratio, the water permeability coef- and high compressive strength (±60 MPa in 7 days) can be
ficient, and compared the results with the natural aggregate expected from a well-designed RCC [21]. RCC is less susceptible
geopolymer, and found acceptable properties. Nuaklong et al. to cracking due to drying shrinkage and sets fast which makes it
[16] evaluated geopolymer concrete containing recycled concrete more suitable for use in road and dam construction. RCC is consid-
aggregate as a coarse aggregate. They used different molar concen- ered more economical as compared to conventional concrete as it
trations (8, 12, and 16 M) of sodium hydroxide and achieved satis- requires less amount of labor and can be rapidly installed [22].
factory results in compressive strength in the range of 30–38 MPa The most beneficial aspect of RCC is its reduced cement require-
for high molar concentration. They also compared the strength and ment (100 kg/m3 for dams) [21], although for road construction
other mechanistic and fresh properties of the RCA geopolymer with it may yield a higher quantity of cement and lower water to
limestone geopolymer and found that although the strength is cement ratio relative to dams [22,23].
slightly lower but still in a higher range to be used as road con- The main goal of this study was to develop roller-compacted
struction material. geopolymer concrete (RCGPC) and roller compacted cement con-
Posi et al. [17] investigated the mechanistic properties of light- crete (RCC) using 100% RCA and compare their strength and stiff-
weight geopolymer concrete using recycle lightweight concrete ness characteristics. The effects of various mixture variables
aggregate. For this study, they used a mixture of high calcium fly including the molar concentration of sodium hydroxide, Na2SiO3
ash and OPC. Alkali activator was prepared by mixing NaOH and to NaOH ratio, fly ash content, aggregate gradation, curing time,
Na2SiO3. OPC was used as a percentage of fly ash and dosages of and temperature on the mechanical characteristics were investi-
2
Sk Syfur Rahman and Mohammad Jamal Khattak Construction and Building Materials 283 (2021) 122624

gated. Also, microstructures, morphology, and analyses of geopoly- Table 1

mer compounds were studied using Scanning electron microscopy Chemical composition of FA and RCA.

(SEM), X-ray diffraction (XRD), and Fourier-transform infrared Compounds (%) FA RCA
spectroscopy (FTIR). SiO2 36.40 17.78
Al2O3 14.82 2.52
Fe2O3 12.17 12.96
2. Materials and methods
CaO 13.87 59.68
MgO 0.79 0.40
2.1. Material system SO3 0.57 0.76
K2O 1.59 0.90
Class F fly ash (FA) used in this study is a residue of coal burning
in a thermal power plant (Fig. 1a and 1b). The chemical composi-
tions of FA are shown in Table 1. Energy-dispersive X-ray spec- content (OAC) for different RCA-RCGPC mixtures. Freshly mixed
troscopy (EDS) and X-ray diffraction (XRD) were used to perform RCA-RCGPC mixtures were placed in 152.4 mm diameter and
the chemical analysis of FA. Apart from other oxides, the FA is com- 116.4 mm high mold. The mixture was compacted in 5 equal layers
posed of 36.40% silicon dioxide and 14.82% aluminum oxide, which using 25 blows per layer with a 2.5 kg hammer and a height of fall
are considered essential elements for the occurrence of a geopoly- of 45.7 cm. For every mixture, 4 alkali contents were chosen at an
mer reaction. interval of 3%. For each of that alkali content, the RCA geopolymer
RCA was collected from a local aggregate recycling plant (Fig. 1c mixture was prepared, compacted and the density of the com-
and 1d). The specific gravity of the RCA was found to be 2.12 and pacted specimens was determined. A graph of density versus alkali
the absorption capacity was about 9.5%. The chemical composition content was plotted to determine the MDD and OAC of the RCA-
of RCA is reported in Table 1. RCA contains a large amount of cal- RCGPC mixtures.
cium (59.68%) which is originated from calcite and calcium silicate
hydrates from the cement matrix [24]. It also contains 17.78% sil- 2.2.2. Sample preparation
icon dioxide and 2.52% aluminum oxide. NaOH solution of known molarity was prepared 24 h before use.
Sodium hydroxide (NaOH) solution with different molarities Alkali activator comprised of NaOH and Na2SiO3 of the needed
was prepared in the laboratory using NaOH powder and distilled ratio was prepared at least 30 min before mixture preparation.
water. Readily available technical grade sodium silicate (Na2SiO3) Required quantities of RCA, fly ash and alkali activator was then
liquid was used in this study. The composition of Na2SiO3 was obtained and mixed thoroughly first by hand and then using the
solids 37% (SiO2 28%, Na2O 9%), water 63% by weight. mechanical mixture. The prepared mixture was then molded in
cast iron cylindrical mold (100 mm diameter and 200 mm height)
2.2. Test methods using a vibratory hammer according to ASTM C1435 [26]. The
specimen was compacted into 3 layers using a vibratory hammer
2.2.1. Maximum dry density and optimum moisture content and tamping plate. The mass of the hammer was 10 ± 0.2 kg and
The Modified proctor test (ASTM D1557 [25]) was used to could produce 2000 impacts per minute. The duration for com-
determine the maximum dry density (MDD) and optimum alkali paction was 60 s (3 layers, 20 s per layer). The compacted specimen

Fig. 1. (a) Fly ash (b) SEM of Fly ash (c) RCA (d) SEM of RCA used in the present study.

3
Sk Syfur Rahman and Mohammad Jamal Khattak Construction and Building Materials 283 (2021) 122624

was then placed in plastic mold and sealed with duct tape and a 2.2.4. Unconfined compressive strength (UCS)
ziplock bag before placing it in the oven for curing. After curing In this experiment, the UCS of the mixtures was determined per
the samples were placed at room temperature for 24 h before test- ASTM C39 [27]. Using the Material Testing System (MTS) continu-
ing for unconfined compressive strength (UCS). ous compressive load was applied at an axial rate of 0.51 mm/min
For the flexural strength test, beam specimens of until the ultimate failure of the specimen was observed (Fig. 3). To
100 mm  100 mm  400 mm size were prepared using the mold facilitate the capture of the entire stress–strain response of the
as shown in Fig. 2(b). The mixture was placed in the mold and specimen, continuous real-time load, and deformation data acqui-
using a top plate and vibratory hammer the beam specimen was sition were performed for triplicate specimens. The following
compacted. The compacted specimen was taken out from the equation was used to compute the compressive strength at failure.
mold, sealed properly, and placed in the oven for curing at the
specified curing temperature (Fig. 2d). Pc
UCS ¼ ð1Þ
A
Where UCS denotes unconfined compressive strength, Pc repre-
sents compressive load at failure, A is a cross-sectional area of the
2.2.3. Experimental design specimen surface on which compressive load was applied. The
In the RCA based RCGPC synthesis, different mixture variables slope of the straight-line portion of the stress–strain curve was
were varied to see the effect on the geopolymer concrete formed determined as Elastic modulus (E) of the concrete specimens pre-
(Table 2 and Table 3). Three molarity levels of Sodium hydroxide pared using different mixtures.
(6 M, 8 M, 10 M) were used. Sodium silicate to sodium hydroxide
ratio (Na2SiO3/NaOH) by weight was varied as 0, 1, and 2. Fly ash
dosages of 10%, 15%, and 20% by weight of total dry mass were uti- 2.2.5. Dynamic modulus (E*)
lized in mixture preparation. Three gradations of the RCA including Dynamic modulus (E*) specimens were prepared similarly as
coarse, medium, and fine were also evaluated. Upper and lower that of UCS, and then tested using a materials testing system
gradation levels based on ACPA standards are also listed in Table 2. (MTS) machine under sinusoidal loading waveform at four differ-
To test the effect of mixture variables, all specimens were cured ent frequencies; 25, 10, 5, and 1 Hz. The test was conducted follow-
at 60 °C for 48 hrs (accelerated curing) and then tested after 24 hr ing ASTM D3497 [28]. The used amplitude of dynamic compressive
of the rest period. It was revealed from the preliminary testing that stress was kept 20% of the ultimate strength of the mixture. Strain
the unconfined compressive strength (UCS) of specimens plateau corresponding to the applied load was measured by attaching both
after 48 hrs of curing at 60 °C. The effect of curing temperature LVDTs and an extensometer at the middle third of the specimen
was also evaluated at 25 °C, 40 °C, 60 °C, and cured for 48 hr. Ambi- height. Time, Load, and strain were recorded for each frequency.
ent temperature curing of 14, 28, and 56 days was also a part of the For each mixture, triplicate specimens were tested and the E*
experiment design. Triplicate specimens were prepared for each was calculated using the following equation.
mixture. A total of 9 mixtures was prepared and tested to find
r
out the effect of all these variables on the strength and modulus E ¼ ð2Þ
of the RCA-RCGPC mixtures. Additionally, for each mixture at least e
4 different alkali contents were used to determine the MDD and E* is the Dynamic modulus (Pa), r is the amplitude stress (Pa), e
OAC using a modified proctor test (Table 3). is the amplitude strain (mm/mm).

Fig. 2. (a) Compression mold (b) Flexure beam specimen mold with vibratory hammer (c) Compression specimen (d) Flexure strength specimen used in the present study.

4
Sk Syfur Rahman and Mohammad Jamal Khattak Construction and Building Materials 283 (2021) 122624

Table 2
Aggregate gradation for recycle concrete aggregate (percent passing).

Size (mm) Lower (%) Upper (%) GR 1–Lower ACPA Bound (Coarse) (%) GR 2–Middle ACPA (medium) (%) GR 3–Upper ACPA Bound (fine) (%)
19.00 100.00 – 100.00 100.00 100.00
12.50 81.00 100.00 78.60 88.99 95.11
9.50 71.00 91.00 65.00 82.00 92.00
4.75 49.00 70.00 43.32 54.65 61.32
2.36 33.00 54.00 34.64 43.70 49.03
1.18 24.00 40.00 28.13 35.49 39.82
0.60 15.00 30.00 21.94 27.68 31.05
0.30 10.00 25.00 10.80 13.62 15.29
0.15 2.00 16.00 5.72 7.22 8.10
0.08 0.00 8.00 3.55 4.48 5.03

ACPA- American Concrete Pavement Association.

Table 3
Experimental design matrix.

No. Mixture NaOH Molarity (M) Na2SiO3/NaOH Ratio Fly ash Content (%) Gradation Curing Temperature (oC) Curing Period (hrs (days))
1 M1 6 1 15 1 60 48(2)
2 M2 8 1 15 1 60 48(2)
3 M3 10 1 15 1 60 48(2)
4 M4 8 0 15 1 60 48(2)
5 M2 8 1 15 1 60 48(2)
6 M5 8 2 15 1 60 48(2)
7 M6 8 1 10 1 60 48(2)
8 M2 8 1 15 1 60 48(2)
9 M7 8 2 20 1 60 48(2)
10 M2 8 1 15 1 60 48(2)
11 M8 8 2 15 2 60 48(2)
12 M9 8 2 15 3 60 48(2)
13 M2 8 1 15 1 25 48(2)
14 M2 8 1 15 1 45 48(2)
15 M2 8 1 15 1 Ambient 14
16 M2 8 1 15 1 Ambient 28
17 M2 8 1 15 1 Ambient 56
18 M2 8 1 15 1 60 12(0.5)
19 M2 8 1 15 1 60 24(1)
20 M2 8 1 15 1 60 72(3)

Fig. 3. (a) Compression test (b) Failed specimen under compressive load.

2.2.6. Flexural strength R = modulus of rupture (MPa), P = applied load (N), L = span
The flexural strength test of the materials was performed fol- length (mm), b = width of the specimen (mm), d = depth of the
lowing the ASTM C78 [29] test procedure. Beam specimens of specimen (mm).
100 mm  100 mm  400 mm dimensions were used. Third point
loading was applied with a span length of 300 mm. The specimen
was loaded at a displacement rate of 0.51 mm/min. until complete 2.2.7. Scanning electron microscopy (SEM), X-Ray diffraction (XRD),
fracture (Fig. 4). The modulus of rupture or flexural strength of the and Fourier-transform infrared spectroscopy (FTIR)
specimen was calculated using the following equation. The surface morphology of the prepared mixtures was exam-
ined by an SEM (Hitachi S-3000 N). SEM analysis of the material
PL was done by cutting multiple thin slices of the material and putting
R¼ 2
ð3Þ
bd it on a metal stub using conductive graphite adhesive. The speci-
5
Sk Syfur Rahman and Mohammad Jamal Khattak Construction and Building Materials 283 (2021) 122624

Fig. 4. (a) Flexural strength test (b) Failed flexural beam specimen.

mens were also sputter-coated with gold using a sputter coater For 8 M NaOH, three different Na2SiO3/NaOH ratios (0, 1, 2)
(EMS150R, Electron Microscopy Sciences) to make it more conduc- were used. The experimental results indicated increased OAC for
tive and avoid the accumulation of electrons while analyzing. the mixtures as more silicate was involved. Sodium silicate has
MiniFlex600 XRD machine was used in the XRD analysis. The 37% solids and 63% water in it. The alkali solutions containing
diffraction angle 2h was varied from 5° to 90° range. Cary 630 FTIR 8 M NaOH and 0, 1, and 2 Na2SiO3/NaOH ratios had 24.24%,
Spectrometer was used for FTIR analysis using wavenumber ranges 30.62%, and 32.79% solids and correspondingly 75.76%, 69.38%,
from 600 to 4000 (cm1). Pulverized material was used for both and 67.21% of water were present, respectively. The amount of
XRD and FTIR analysis. Geopolymer mastic samples were prepared water decreased as more amount of silicate was incorporated into
and pulverized. The fine powder was then passed through the the alkali activator solution. Therefore, an increased amount of
#200 sieve (0.075 mm) to make sure only the finest part comes alkali was needed, as more water is required to gain the optimum
as a sample. For each geopolymer mixture, 3 samples were tested. condition for MDD of the mixtures. The MDD value had increased
The coefficient of variation of those 3 samples was less than 2%. when the Na2SiO3/NaOH ratio changed from 0 to 1 but after that, it
decreased for Na2SiO3/NaOH ratio of 2.
3. Results and discussions Similarly, three different fly ash contents of 10%, 15%, and 20%
were used. The addition of fly ash increased the dosage of fine par-
3.1. Optimum alkali content and maximum dry density ticles in the gradation of the mixtures. From the relationship of
MDD and alkali content, it was observed that, as more fly ash
For different mixture variables, OAC (optimum alkali contents) was added to the mixture, OAC increased since more water was
were determined for the RCA-RCGPC mixture using a modified required to lubricate the fine particles and fill up the voids. MDD
proctor test (ASTM D1557 [25]). Table 4 shows the relationship increased when the fly ash content increased from 10% to 15%,
between dry unit weights, void ratio, porosity, air, and alkali con- beyond that it dropped when the fly ash content increased from
tent of the different RCA-RCGPC mixtures. The experimental 15 to 20%. Three different gradations, coarse (GR1), medium
results have shown the changes in the optimum OAC and maxi- (GR2), and fine (GR3) were also used to evaluate the MDD and
mum dry density (MDD) based on different mixture variables. As OAC. As the gradation of the mixture became finer the OAC
the molarity of the NaOH solution in the alkali mixture increased increased slightly because of the increase in fine particles which
the OAC also increased. 6 M, 8 M, and 10 M NaOH were used and needed more alkali content for lubrication. The MDD values
their solid constituents were 28.18%, 30.62%, and 32.79% and the slightly decreased as gradation became finer.
water content was 71.82%, 69.38%, and 67.21%, respectively. As
with the increase in molarity the amount of water reduced, more 3.2. Accelerated curing
alkali solution was required to obtain the optimum condition for
the mixture to lubricate all the solid particles which increased The advantage of the geopolymer product is that their curing
the OAC of the mixtures. At the same time, there was an increase process can be accelerated by elevated temperature which facili-
in the MDD of the mixtures which were observed as more alkali tates faster laboratory investigation and data collection for mixture
solids were incorporated in the mixtures. optimization. Fig. 5 illustrates the effect of curing time at 60 °C on

Table 4
OAC and MDD for different RCA RCC mixtures.

Mixture OAC (%) Water/binder (FA + alkali solids) Maximum Dry Density, (c) (Kg/m3) Void ratio, (e) Porosity, n (%) Air content, a(%)
M1 13.6 0.51 1934 0.337 25.20 6.40
M2 14.1 0.50 1946 0.329 24.75 5.87
M3 14.5 0.49 1964 0.326 24.57 5.65
M4 12.9 0.54 1883 0.372 27.10 8.44
M5 14.9 0.62 1813 0.313 23.84 4.86
M6 13.4 0.53 1859 0.311 23.74 7.70
M7 14.8 0.66 1879 0.313 23.87 6.78
M8 14.4 0.67 1900 0.317 24.06 5.41
M9 14.8 0.51 1878 0.318 24.13 4.94

Note: 4 different alkali contents were used for each mixture for density -alkali content relationship to find OAC & MDD.

6
Sk Syfur Rahman and Mohammad Jamal Khattak Construction and Building Materials 283 (2021) 122624

the unconfined compressive strength (UCS) of RCA-RCGPC mix- One can observe from the experimental results that, the UCS
tures. It was observed that UCS increased with the increase of cur- increased by 59% as the Na2SiO3/NaOH ratio increased from 0 to
ing time up to 48 h, beyond that it plateaus. The rate of strength 1. However, when the ratio increased from 1 to 2 UCS decreased.
gain for 12-hr, 24-hr, and 48-hr was 0.86, 0.44, and 0.24 MPa/hour. The same trend was seen for the elastic moduli of the mixtures.
Hence, oven curing for 48-hour at 60 °C was considered as the So, the Na2SiO3/NaOH ratio of 1 was considered to be an optimum
accelerated curing method for this study. Accelerated curing was value. The reason behind the high unconfined compressive
performed for all the geopolymer mixtures. However ambient tem- strength and elastic modulus for the mixture is the dissolution pro-
perature curing was also performed on the selected geopolymer cess of alumina and silica tends to be higher at an equal mass ratio
mixture (M2) to (i) show the practicability of the study (ii) prove of Na2SiO3 and NaOH to accelerate the development of strength.
that similar strength will be achievable in the ambient and acceler- Hence, the bonds are strong enough to enhance the resistance to
ated curing. microcrack development.
The above discussion can further be supported by analyzing
total chemical constituents as well as the density and voids ratio
3.3. Unconfined compressive Strength, elastic modulus, Density, and of the developed RCGPC mixtures as discussed below.
void ratio
3.3.2. Effect of total chemical constituents
3.3.1. Effect of NaOH molarity and Na2SiO3/NaOH ratio Fig. 7 shows the effect of different chemical constituents on the
Fig. 6 depicts the effect of NaOH molarity and Na2SiO3/NaOH UCS of the RCGPC mixtures. The total percentage of each chemical
ratio on unconfined compressive strength (UCS) and modulus of constituent was calculated based on the chemical composition of
elasticity (E) of RCGPC mixture containing 100% RCA cured at fly ash, RCA, and alkali activator solutions (NaOH and Na2SiO3).
60 °C. To better understand the effect of sodium hydroxide molar- Fig. 7(a) illustrates the effect of Na2O (%) on the UCS. The results
ity, three different molar concentrations namely 6, 8, and 10 M show that if the Na2O (%) increases the strength also increases. The
sodium hydroxides were used with the Na2SiO3/NaOH ratio of 1. increment follows a linear trend. Fig. 7(b) depicts the effect of SiO2
It can be seen from Fig. 6 that as the NaOH molarity increased (%) on the UCS, which indicates as SiO2 (%) increases the UCS
the UCS and E also increased. The mixtures made with 8 M and increases to a maximum value and then starts decreasing.
10 M exhibited an increase in UCS of about 61% and 100%, respec-
tively, relative to 6 M mixtures. A similar trend was exhibited by 3.3.3. Effect of density and voids ratio
elastic moduli with improvements of 52% and 80%, respectively. Since the production of more aluminosilicate gel contributed to
The strength development of geopolymer products depends on reducing the voids, it improved the density of the composite with
the aluminosilicate gel formation in the chemical reaction which the increment of molarity [30]. As shown in Fig. 8(a) dry density
greatly depends on hydroxide ion (OH) produced from alkali was found to be 1934.3, 1945.5, and 1964.35 kg/m3 for 6, 8, and
(NaOH), thus increasing alkali concentration enhances the devel- 10 M NaOH, respectively which showed a slightly incremental
opment of strength. However, aluminate gel can precipitate at an trend. As the density increased, the void ratio decreased as
early stage in the presence of excess hydroxide ion if very high expected which can be observed in Fig. 8(b). Increased density
NaOH concentration is used, which were not observed up to and reduced void contributed to the strength gain as well.
10 M NaOH used for this experiment. The dry densities of the mixtures with different Na2SiO3 to
As the increase in compressive strength was observed linear NaOH are also shown in Fig. 8(a). The dry densities were found
with the increase in NaOH molarity, it is expected that the use of to be 1850, 1945.5, 1813 kg/m3 for 0, 1, 2 ratios of Na2SiO3 to NaOH
NaOH with higher molar concentration may result in higher which follows a similar trend with UCS. Void ratios for the mix-
strength, however that may not be suitable from the construction tures are shown in Fig. 8(b) which indicates that as the density
safety point of view. So, there will always be a chosen optimum increased void ratio decreased. The 8 M NaOH and Na2SiO3 contain
NaOH molarity based on the strength requirement. In the current 75.76% and 63% water based on their constituents, their combina-
study 8 M, sodium hydroxide was chosen as the optimum. When tion at ratios of 0, 1, and 2 yields 75.76%, 69.38%, and 67.21% water
the heat condition is the same, the geo-polymerization reaction content, respectively. At a higher Na2SiO3/NaOH ratio less water
between precursor and alkali activator is mostly responsible for content is available in the mixture to lubricate the particles than
the strength development in geopolymer concrete. To find out lower Na2SiO3/NaOH ratios, which may reduce the density at Na2-
the effect of Na2SiO3/NaOH ratio on compressive strength, their SiO3/NaOH ratio of 2. Besides, the pH condition of the reaction
different ratios of 0, 1, and 2 were used with 8 M NaOH (Fig. 6). environment is also affected by the concentration and amount of
Na2SiO3 and NaOH, and their ratios, which in turn affects the
geo-polymerization process, density, voids, and strength develop-
ment of the mixture [31].
25
3.3.4. Effect of fly ash content and RCA gradation
20 Fig. 9 depicts the effect of fly ash content and RCA gradation on
the unconfined compressive strength (UCS) and elastic modulus (E)
15 of the RCA RCGPC mixtures. To find the effect of fly ash content,
UCS (MPa)

mixtures were prepared with 10%, 15%, and 20% fly ash with 8 M
10 NaOH, Na2SiO3/NaOH ratio of 1. The mixtures were cured at
60 °C for 48 h. It can be seen from the results that with the increase
5 of fly ash content, no significant gain in compressive strength was
observed, however at 15% fly ash content the modulus increased to
0 40% as compared to 10% fly ash content and beyond that, the mod-
0 12 24 36 48 60 72 84
ulus slightly decreased.
Curing Time (hours)
At 10%, 15%, and 20% fly ash content the alkali contents were
Fig. 5. Effect of curing time on the compressive strength of RCGPC using 100% RCA 13.35%, 14.07%, and 14.79%, respectively which resulted in the
at 60 °C. alkali to fly ash ratios of 0.32, 0.29, 0.25, respectively. The reason
7
Sk Syfur Rahman and Mohammad Jamal Khattak Construction and Building Materials 283 (2021) 122624

50 UCS E UCS E

UCS (MPa), E (GPa)

40

30

20

10

0
6M 8M 10M 0 1 2
NaOH molarity (M) Na2SiO3/NaOH rao

Fig. 6. Effect of NaOH molarity and sodium silicate to NaOH (8 M) ratio on the compressive strength and modulus of RCGPC using 100% RCA.

30 30
(a) (b)
25 25

20 20
UCS (MPa)

UCS (MPa)
y = 4.1178x - 53.17
15 R² = 0.5251 15

10 10
y = -2.2558x2 + 313.93x - 10896
5 5
R² = 0.9245
0 0
15 17 19 21 66 68 70 72 74
Na2O (%) SiO2 (%)

Fig. 7. Effect of (a) Na2O (b) SiO2 content in the mixture on the UCS of RCGPC using RCA.

for the strength and modulus decrease can be concluded as, in the ity, and compressive strength were slightly higher than other ones
case of 10% fly ash, although the alkali solids were more, there was (Figs. 9, 10).
not enough fly ash to react with all the supplied alkali. And in the
case of 20% fly ash there were not enough alkali solids to react with 3.3.5. Effect of curing temperatures
all the fly ash supplied. In the case of 15% fly ash alkali solid and fly Fig. 11 exhibits the effect of ambient temperature curing and
ash content was optimum, which slightly increased the strength oven curing on the unconfined compressive strength (UCS) and
and significantly improved the modulus value. elastic modulus (E) of RCGPC mixtures. The mixtures were con-
Fig. 10 (a) illustrates the relationship between UCS and density structed with 15% fly ash, 8 M NaOH, and Na2SiO3/NaOH ratio of
for mixtures with different fly ash content. It was observed from 1. Three different oven curing temperatures (25 °C, 40 °C, 60 °C)
the test results that the densities of the mixtures also follow the as well as 3 different curing days (14, 28, 56) at ambient tempera-
same trend as compressive strength, at 10% fly ash content the tures were investigated. Please note that the average day and
density was 1900 kg/m3, and it increased to 1945.5 kg/m3 when nighttime specimen temperatures were measured around 34.4 ± 5.
fly ash content was 15% and beyond that further increment of fly 3 °C and 28.8 ± 2.1 °C, respectively. From the test results, it was
ash to 20% the density dropped to 1920 kg/m3. At 15% fly ash con- found that as the temperature increased, the strength and modulus
tent density was high because of more geo-polymerization reac- also increased. The increase in UCS and modulus is about 3 and 2.5
tion and aluminosilicate gel formation could happen with the times, respectively at 60 °C as compared to the 25 °C. The increase
maximum amount of fly ash being utilized. The trend of void ratio in strength and modulus is because of the accelerated and
shown in Fig. 10(b) also supports the density and compressive enhanced geo-polymerization reaction, which happened at higher
strength. So, 15% of fly ash can be considered optimum for this temperatures [32-34].
RCA gradation. Since in actual practice it is difficult and expensive to cure pave-
Three different RCA gradations coarse (GR1), medium (GR2), ment at elevated temperature, the effect of ambient temperature
and fine (GR3) were utilized in the experimentation. Mixtures curing was also investigated at 14, 28, and 56 days of curing
were prepared using 8 M NaOH, Na2SiO3/NaOH ratio of 1, and (Fig. 11). For ambient temperature curing, the specimens were
cured at 60 °C for 48 h. These mixtures yielded compressive sealed and coated on the top surface with a tack coat to simulate
strength and elastic modulus in the range of 19.21 to 21.36 MPa, the actual pavement field condition. In this experiment, UCS was
and 36.89 to 39.31GPa, respectively. It can be seen that there is found to increase gradually with the increase in curing days. It
only a marginal difference in strength and modulus values was discovered that most of the strength gain was in the first
amongst various gradations. However, for the coarse gradation 14 days (13.1 MPa) which increase to 17.5 MPa on the 56th day
(GR1), the average values of density, void ratio, modulus of elastic- of curing. The mixture gained about 82% of designed maximum
8
Sk Syfur Rahman and Mohammad Jamal Khattak Construction and Building Materials 283 (2021) 122624

increase of 40% due to the addition of Na2SiO3 in the alkali solution.

Also, when the NaOH molarity was increased from 8 M to 10 M,
flexural strength increased from 2.84 MPa to 3.73 MPa, which is
another increment of 32%. The flexural strength of RCC with 12%
cement content was found to be 2.18 MPa, which is slightly higher
than the flexural strength of the RCGPC mixture with 8 M NaOH
with no Na2SiO3. However, the flexural strength of RCGPC mixtures
with 8 M and 10 M NaOH and Na2SiO3 with a 1:1 ratio is about 30%
and 71% higher than the RCA RCC mixture.
The flexural strength increment in the RCGPC mixtures can also be
explained by observing the improvement in the adhesion bond
between RCA and geopolymer matrix (Fig. 13) due to the addition
of Na2SiO3 and higher molarity of NaOH. Fig. 13 shows the fracture
surface of different RCGPC and RCC mixtures. Fig. 13 (a) depicts the
RCGPC mixture with 8 M NaOH only. In this case, the peeling of
RCA aggregate is visible, and some RCA matrix is also fractured.
Fig. 13 (b) shows an RCGPC mixture with 8 M NaOH and Na2SiO3:
NaOH ratio of 1, which displays some broken RCA aggregate and
matrix indicating a stronger bond than the previous mixture without
Na2SiO3. Fig. 13 (c) demonstrates the RCGPC mixture with 10 M NaOH
and Na2SiO3: NaOH ratio 1 where there was no visible peeled off RCA
aggregate, numerous RCA aggregates were seen broken with some
fractured RCA mortars. This indicates a stronger bond than the two
previous RCGPC mixtures described. Fig. 13(d) shows the fracture sur-
face of the RCC mixture. Here the fracture surfaces look like the RCGPC
mixture with 8 M NaOH only. peeled off RCA aggregates, broken RCA
matrix are also visible which indicates less strong bond than RCGPC
mixtures containing 8 and 10 M NaOH and Na2SiO3: NaOH ratio of 1.
Fig. 8. Correlation between UCS and (a) density (b) void ratio for different NaOH
molarity and NaOH: Na2SiO3 ratio of RCGPC using 100% RCA.
3.5. Dynamic modulus (E*)

compressive strength based on 60 °C oven temperature curing. Fig. 14 depicts the effect of NaOH molarity and Na2SiO3/NaOH
Similar was true for the elastic modulus which was found to be ratio on the dynamic modulus of different RCA-RCGPC mixtures.
28.8 GPa on the 14th day and 38 GPa on the 56th day of ambient Four different loading frequencies (1, 5, 10, and 15 Hz) were used
temperature curing. in the evaluation. In general, all the mixture exhibited a similar
trend of change of dynamic modulus with respect to change in
loading frequency. When the loading frequency increased from 1
3.4. Flexural strength to 5 Hz the change in dynamic modulus was not significant. How-
ever, when the loading frequency increased to 10 and 25 Hz the
The flexural strength of the selected RCGPC and RCC mixtures dynamic modulus decreased significantly. Increased loading fre-
containing 100% RCA was also investigated as illustrated in quency created permanent damage in the specimen which
Fig. 12. Similar to the compressive strength, the flexural strength increased the strain and decreased modulus. It is observed from
exhibits a similar trend of strength increment. For instance, the Fig. 14a that when the NaOH molarity increased from 6 to 8, the
flexural strengths of RCGPC mixtures with 8 M NaOH only and with average dynamic modulus increased by about 89.1%, and for 8 M
Na2SiO3 are 2.02 and 2.84 MPa, respectively. This displays an to 10 M increment, it increased by only 16.9%.

50
UCS E UCS E

40
UCS (MPa), E (GPa)

30

20

10

0
10 15 20 GR 1 GR 2 GR 3
Fly ash (%) Gradaon
Fig. 9. Effect of fly ash percentage and gradation on the compressive strength and modulus of RCGPC using 100% RCA with 8 M sodium hydroxide and Na2SiO3/NaOH ratio of
1.

9
Sk Syfur Rahman and Mohammad Jamal Khattak Construction and Building Materials 283 (2021) 122624

22.0 l ¼ tanðuÞ ð4Þ

(a) Gradation Fly ash 15%
Fig. 15 displays a Mohr-Coulomb failure envelop from which a
21.0
relationship can be established between the tensile strength (rt),
unconfined compressive strength (rc) obtained from laboratory
20% GR1

testing at confining strength (r3) of zero with u and so. From that
UCS (MPa)

20.0 10% relationship, the u and so values can be computed as follows.

   
GR2 rt r2t
U ¼ sin1 ; so ¼ ðx þ rt ÞtanðuÞ; andx ¼ ð5Þ
19.0 2x þ rt rc  rt
GR3 Using the Eqs. (4) and (5), cohesion (so) and friction (m) were
calculated using unconfined compressive strength and flexural
18.0
1870 1880 1890 1900 1910 1920 1930 1940 1950 strength for different RCGPC and RCC mixtures. The results were
Density (Kg/m3) shown in Fig. 16. The results indicated that the cohesion for the
24.0 RCGPC mixture with 8 M NaOH only was 1.5 MPa. When Na2SiO3
(b) Gradation Fly ash was added to the mixture the cohesion value increased to
2.21 MPa which is a 47.33% increment. Again, when the molarity
22.0 15% of NaOH increased from 8 M to 10 M the cohesion value increased
20% to 2.85 MPa, which is another 28.96% increase. On the other hand,
RCC mixtures using 12% OPC showed a cohesion value of 1.79 MPa
UCS (MPa)

GR1
20.0 10%
which is about 80% and 62.8% of the cohesion value of RCGPC mix-
tures with 8 and 10 M NaOH and Na2SiO3: NaOH of 1, respectively.
GR2
GR3 The friction value on the other hand did not change much for RCA-
18.0
RCGPC mixtures but shows a smaller increment in the case of the
RCA-RCC mixture. For bound material, cohesion is more important
16.0 than friction. Friction normally comes into action after the micro-
0.28 0.29 0.30 0.31 0.32 0.33 0.34 0.35 cracking has been developed. Although cement has a little bit more
Void ratio
friction it does not have enough cohesion to increase the overall
Fig. 10. Correlation between UCS and (a) density (b) void ratio for different strength of the mixture. These results strongly support the fact that
Gradation and fly ash content of RCGPC using 100% RCA. geopolymer cement created better bonding with RCA than ordi-
nary Portland cement and that is the reason behind the improve-
In the case of the change of Na2SiO3/NaOH ratio (Fig. 14b) from 0 ments in the strength and moduli value of the mixtures.
to 1, the average dynamic modulus increased by 35.5%. However,
when the Na2SiO3/NaOH ratio changed from 1 to 2 the average
3.7. Relationship of mechanical and volumetric properties with
dynamic modulus decreased by 35.5%. Hence, one can conclude that
different mixture constituents
in general, RCGPC mixture with higher UCS and modulus also exhibits
higher dynamic modulus. Which makes sense as mixtures with
To establish a relationship between the mechanical properties
higher UCS and modulus has stronger geopolymer bond formed as
(UCS, E), and dry density (c) with mixture constituents such as
discussed earlier. Such improvements enhanced the capability of
NaOH molarity, percentage of Na2SiO3 in alkali solution, fly ash
the mixture to resist deformation and damage under repeated
content, curing temperature, and dry density, regression analysis
loadings.
was performed. Such an investigation can also help in the sensitiv-
ity analysis and optimization of RCGPC mixtures for further assess-
3.6. Cohesion and friction of RCA mixtures containing geopolymer and ment. Synopsis of regression analysis for UCS, E, and c are
Portland cement binder presented as follows. Various forms of models for UCS, E, and c
were analyzed using the Microsoft Excel data analysis tool. The
It can be assumed that the coefficient of friction (m) is a function coefficient of correlation, R2, ANOVA, and Hypothesis testing using
of the angle of internal friction (u) of the concrete particle and can the p-value approach was performed to finalize the models. In the
be related to each other by the following equation. model selection process if the model had a p-value for any coeffi-

50 UCS E UCS E
UCS (MPa), E (GPa)

40

30

20

10

0
25 40 60 14 28 56
Temperature (0C) Ambient temperature curing (days)
Fig. 11. Effect of curing temperature and ambient temperature curing days on the UCS and modulus of RCGPC using 100% RCA with 8 M NaOH and Na2SiO3: NaOH ratio of 1.

10
Sk Syfur Rahman and Mohammad Jamal Khattak Construction and Building Materials 283 (2021) 122624

4.0
Flexural Strength (MPa)

3.0

2.0

1.0

0.0
RCGPC 8M 0:1 RCGPC 8M 1:1 RCGPC 10M 1:1 RCC (12% OPC)

Fig. 12. Flexural strength of selected RGGPC and RCC mixtures containing RCA.

cient greater than 5% that model was rejected, and new variables
were compared. Details of regression analysis are presented below
in Eqs. (6–8).
UCS
¼ 1220 þ 35:52ðNaOHÞ þ 460:26ðNaSiÞ
Pa
2
 517:58ðNaSiÞ þ 3:71ðFAÞ þ 96:65ðT=T O Þ
þ 385:75ðc=cwÞ; ð6Þ
2
R = 0.87, F = 29.73, n = 33
E 2
¼ 3603307:75 þ 72922:31ðNaOHÞ þ 653942:25ðNaSiÞ
Pa
þ 112413:41ðFAÞÞ  2862:49ðFAÞ2 þ 180274:44ðT=T O Þ
þ 942084:32ðc=cwÞ; ð7Þ Fig. 14. Effect of (a) molarity (b) silicate on the dynamic modulus of RCGPC using
2 100% RCA.
R = 0.81, F = 17.76, n = 33
c 2
¼ 120:87 þ 0:90ðNaOHÞ þ 44:14ðNaSiÞ  115:83ðNaSiÞ
cw
 0:80ðFAÞ þ 0:38ðCuÞ; ð8Þ
2
R = 0.54, F = 12.76, n = 61
Here, Pa = atmospheric pressure, NaOH = Sodium hydroxide
Molarity (M), NaSi = sodium silicate (% of total alkali), c = unit
weight, cw = unit weight of water, FA = fly ash (% of dry mixture),
T = curing temperature, To = 25 °C, Cu = coefficient of uniformity.

(a) (c)

Pilled aggregates

Fig. 15. Determination of angle of internal friction (u) and shear strength (s0).

RCA aggregates

(b) Pilled aggregates (d)

RCA matrix

Fig. 13. Fracture surface of RGGPC with (a) 8 M NaOH: Na2SiO3 ratio 0:1 (b) 8 M
NaOH: Na2SiO3 ratio 1:1 (c) 10 M NaOH: Na2SiO3 ratio 1:1 (d) RCC with 12% OPC. Fig. 16. the cohesion of different RCGPC and RCC mixtures containing RCA.

11
Sk Syfur Rahman and Mohammad Jamal Khattak Construction and Building Materials 283 (2021) 122624

40 70 (b)
(a)
35 R² = 0.8728 60 R² = 0.8039

Predicted E (GPa)
Predicted UCS (MPa) 30 50
25 40
20
30
15
20
10
5 10

0 0
0 10 20 30 40 0 10 20 30 40 50 60 70
Measured UCS (MPa) Measured E (GPa)

2200
(c) R² = 0.53
2100
Predicted density (kg/m3)

2000

1900

1800

1700

1600
1600 1800 2000 2200
Measured density (kg/m3)
Fig. 17. Predicted Vs measured for (a) UCS (b) E (c) Density model.

The predicted versus measured values for the newly developed relationship for conventional concrete (see, Fig. 18). It was
model for UCS, E, and c are plotted and shown in the following observed from the results that the modulus of elasticity values of
Fig. 17. The R2 value for UCS, E, and c models were found to be RCGPC and RCC are higher as compared to the conventional con-
0.87, 0.81, and 0.54, respectively. The p-values for each variable crete as expressed by ACI and BS codes. According to the test data,
in each model were also significant at the 0.05 levels. In all these the developed relationship between modulus of elasticity (E) and
three models the predicted and measured values fall on and evenly UCS is E = 8.41 (UCS)0.5, where E and UCS are in MPa.
scattered around the line of equality indicating the models are pre-
dicting fairly accurately. 3.9. Correlation between unconfined compressive strength (UCS) and
flexural strength
3.8. Correlation between unconfined compressive strength and elastic
modulus Fig. 19 demonstrates the relationship between flexural strength
and unconfined compressive strength (UCS) for RCGPC and RCC
An empirical relationship between modulus of elasticity (E) and mixtures. It also displays the comparison of the results from this
unconfined compressive strength (UCS) is developed and com-
pared with existing equations (ACI318 & BS EN 1992) on such a

60
RCA-RCGPC
E= 8.41(UCS)0.5
Modulus of Elasticity (GPa)

50 RCA-RCC

40

30 BS EB 1992-1-1

20 ACI 318

10

0
0 5 10 15 20 25 30
UCS (MPa)
Fig. 19. Relationship between UCS and Flexural strength of selected RGGPC with
Fig. 18. Relationship between UCS and E of RCGPC using 100% RCA. RCA.

12
Sk Syfur Rahman and Mohammad Jamal Khattak Construction and Building Materials 283 (2021) 122624

study with the ACI code and several previous studies performed for 3.10. X-ray diffraction (XRD) analysis
conventional concrete. It was observed from the comparison that,
RCGPC produced higher flexural strength for similar UCS than the Figs. 20 and 21 illustrate the XRD analyses of RCA-RCGPC with
ACI code and some other developed equations for conventional different mixture variables. It was observed from the analysis that
concrete using natural aggregate. At the same time, it also indi- Quartz, Mullite, and Calcium compounds present in the fly ash and
cates lower flexural strength than the equation developed by RCA reacted with alkali activator and formed different geopolymer
Chorn et al, 2018 for RCC using natural aggregate which is reason- (aluminosilicate) compounds. For 6 M NaOH with 1:1 Na2SiO3:
able since the use of 100% RCA will exhibit less strength. However, NaOH, Nepheline (NaAlSiO4), and Svyatoslavite (Ca (Al2Si2O8))
as the test results presented that both the UCS and flexural were formed. On the other hand, in the case of 8 M NaOH Can-
strength increased with the increase in NaOH molarity, flexural crinite (Na6Ca2Al6Si6O24(CO3)2) was formed. There was some unre-
strength could be further improved by using a higher molar con- acted Quartz and Mullite compounds still visible in the XRD
centration of NaOH. graphs, which also agrees with the SEM micrograph of the mixture.
In the case of 10 M NaOH, all the Mullites present in the mixture
were utilized and Nepheline, Svyatoslavite, and Cancrinite were
developed which indicated the presence of a stronger geo-
polymeric reaction. When only NaOH was used as an alkali activa-
(a) Q - Quartz tor lower peaks of aluminosilicate compounds were detected. The
Q M - Mullite same trend was noticed at low temperature curing as well, which
explained low strength and modulus values for such RCA-RCGPC
N - Nepheline mixtures.

S - Svyatoslavite
3.11. Fourier-transform infrared spectroscopy (FTIR) analysis
Ca - Cancrinite
Fig. 22 displays the FTIR analysis of fly ash, RCA, and RCA-RCGPC
N for different mixture variables and curing conditions. Observation
Q from the FTIR graphs reveals that RCA-RCGPC composites using
N
M M NS A S S S N different mixture variables and curing conditions had characteris-
tics X-O bond (X = Al or Si) located between 900 and 1100 cm1.
There are some small peaks between 600 and 800 cm1 which
5 15 25 35 45 55 are most likely Al-O-Si or Si-O-Si bonds originating from the source
material fly ash and RCA (quartz, mullite). Peaks that can be seen
(b) Q
Q - Quartz
Q
M - Mullite
(a)
N - Nepheline

N C - Calcite
Ca - Cancrinite
Q Ca
N N M N Ab Ab Ca Q
N
N
N C
Ca M
N N QN C N
N
5 15 25 35 45 55
5 15 25 35 45 55

(c) Q
Q
(b)

Ca
Q Ca Q
N S N N
S A S N N Ab M N M N M
S N N N N
C

5 15 25
2θ0 35 45 55 5 15 25
2θ0
35 45 55

Fig. 20. XRD of (a) 6 M NaOH with 1:1 Na2SiO3 (b) 8 M NaOH with 1:1 Na2SiO3 (c) Fig. 21. XRD of (a) 8 M NaOH cured at 60 °C and (b) 8 M NaOH with 1:1 Na2SiO3/
10 M NaOH with 1:1 Na2SiO3 cured at 60 °C. NaOH ratio cured at 25 °C.

13
Sk Syfur Rahman and Mohammad Jamal Khattak Construction and Building Materials 283 (2021) 122624

Fig. 22. FTIR of (a) Fly ash (b) RCA (c) 6 M NaOH with 1:1 Na2SiO3 (d) 8 M NaOH with 1:1 Na2SiO3 (e) 10 M NaOH with 1:1 Na2SiO3 (f) 8 M NaOH cured at 60 °C (g) 8 M NaOH
with 1:1 Na2SiO3s cured at 25 °C.

between 1600 and 1650 cm1 are water molecules present in the mortar. After the geo-polymerization, this band shifted to
materials. In the geo-polymerization process different geo- 3352 cm-1. Carbonate peaks at 1425, 872, and 777 cm1 were also
polymeric compounds were formed due to the chemical reaction observed in RCA FTIR, which could be attributed to reactions of
between the fly ash, RCA, and alkali solution. As a result, new peaks atmospheric CO2 with calcium compounds (i.e., carbonation) pre-
were formed, peaks were shifted, and their transmittance value sent in RCA. Due to the geo-polymerization reactions, these peaks
changed. When the FTIR graph of fly ash and RCA-RCGPC contain- were also shifted as the chemical structure of existing compounds
ing 8 M NaOH and Na2SiO3 were compared, it was found that the were altered to form new compounds in RCA-RCGPC. The FTIR of
peak of the X-O band shifted from 935 to 950 cm1. New peaks geopolymer concrete containing NaOH only (no Na2SiO3), different
of H-O–H and O–H was formed between 1400 and 1650 cm1 NaOH molarity (8 M, 10 M), and low-temperature (25 °C) curing
and around 3352 cm1. If we compare the FTIR graph of RCA and was also investigated and shown in Fig. 22 (b, c). It was observed
RCA-RCGPC with 8 M NaOH and Na2SiO3, it will be observed that from those graphs that, in all cases, similar geo-polymeric bonds
the strong band at 961 cm1 was due to the Si-O asymmetric were formed but their peaks and transmittance values shifted
stretching vibration of the calcium silicate compounds present in slightly. However, transmittance shifts were in the range of 0 to
old cement mortar in RCA. 5% and the peaks were within the range of bond wavenumbers.
After geo-polymerization, this band was shifted from 961 cm1 This could be attributed to the heterogeneous nature of the
to 950 cm1. In RCA, the small band around 3600 cm1 is supposed geopolymer concrete. XRD and SEM also agree with the formation
to be coming from portlandite Ca (OH)2 originated from old cement of similar geopolymer compounds in these geopolymer mixtures.

14
Sk Syfur Rahman and Mohammad Jamal Khattak Construction and Building Materials 283 (2021) 122624

lite), RCA (calcium composites), and alkali solution (NaOH), differ-

(a) Unreacted fly ash Parally reacted fly ent geo-polymeric
compounds are formed. The SEM images revealed the presence
of Nepheline (NaAlSiO4) and Cancrinite (Na6Ca2Al6Si6O24(CO3)2)
[35–38]. Calcium compounds present in RCA are believed to be
contributed to the production of Cancrinite. Fig. 24 displays the
SEM micrographs of RCA-RCGPC mixtures containing 8 M NaOH
and Na2SiO3 with a ratio of 1:1. Due to the equal mass ratio of
NaOH and Na2SiO3 more fly ash was consumed by the alkali solu-
tion this time. The SEM micrographs revealed very few unreacted
and partially reacted fly ash particles than the previous mixture
containing only NaOH as an alkali activator which explains the
(y) higher strength of this mixture. The presence of nepheline and can-
(x) crinite were also visible in the SEM micrographs.
Fig. 25 shows the SEM micrographs of RCA-RCGPC mixtures
containing 10 M NaOH and Na2SiO3 at a ratio of 1:1. Due to the
presence of a higher concentration of OH and Na2SiO3 the geo-
polymerization reaction was more accelerated and involved more
fly ash in the reaction process. The SEM micrographs showed
(b) almost no unreacted fly ash. Therefore, the mixture with 10 M
Nepheline NaOH yielded higher compressive strength as compared to 8 M
Cancrinite NaOH. Synthesized nepheline and cancrinite were also visible in
a significant proportion in the SEM micrographs.

4. Conclusions and recommendations

In this research mechanical and microstructural characteristics

of roller compacted geopolymer concrete (RCGPC) and roller com-
pacted cement concrete (RCC) using 100% recycled concrete aggre-
gate (RCA) were investigated. Based on the analysis of results and
discussion the following conclusions and recommendations were
drawn.

The higher molarity of sodium hydroxide produced higher

strength and modulus, hence based on strength requirement
different molar concentrations could be chosen.
(c)
Na2SiO3/NaOH of 1 alkali solution presented maximum strength
and modulus than other ratios for both compression and flexu-
ral strength tests.
Cancrinite

Although there is no significant difference in strength based on
fly ash dosages, 15% of fly ash had shown increased modulus
values and no increases thereafter.

For oven curing at 60 °C, 48 h of oven curing followed by 24 h of
rest period at room temperature was found to be practical
accelerated oven curing condition.

The compressive strength gradually increased with curing days
at ambient summer temperatures. Ambient curing of 56 days
gained around 82% of the maximum strength (60 °C) and most
of which occurred during the first 14 days.

Roller compacted geopolymer concrete exhibited higher
mechanical properties than Roller compacted concrete using
12% OPC containing 100% RCA.

Roller compacted geopolymer concrete with 8 M and 10 M
Fig. 23. (a) SEM of RCA-RCGPC with 8 M NaOH cured at 600C. (b) zoomed image of NaOH displayed higher cohesion than Roller compacted con-
‘‘x” (c) zoomed image of ‘‘y” crete using12% OPC which indicated a stronger bond between
geopolymer cement with RCA than OPC based concrete.

Microstructure and morphology analysis had revealed the rela-
3.12. Scanning electron microscopy (SEM) tive presence of aluminum–silicate compounds that are known
to be formed in geo-polymeric products based on the mixture
Fig. 23 shows the SEM micrographs of the RCA-RCGPC mixture variables.
containing 8 M NaOH only. In the micrographs, it was observed
that there are some partially and unreacted fly ash particles remain The compressive strength, elastic modulus, flexural strength,
in the mixture which explains the overall low strength of the mix- and density of the produced RCA-RCGPC were in the range of
ture as compared to the mixtures with Na2SiO3 (Fig. 24). Due to the 7.53–27.6 MPa, 15.97–46.77 GPa, 2.1 – 3.8 MPa, and 1813–
chemical reactions between constituents of fly ash (quartz, mul- 1964 kg/m3. Based on the mechanical properties, the newly syn-
15
Sk Syfur Rahman and Mohammad Jamal Khattak Construction and Building Materials 283 (2021) 122624

(a) (b)
Cancrinite Cancrinite

Nephelin

Fig. 24. SEM of RCA-RCGPC with 8 M NaOH and Na2SiO3 cured at 60 °C.

Fig. 25. SEM of RCA-RCGPC with 10 M NaOH and Na2SiO3 cured at 60 °C.

thesized RCA-RCGPC can be suitable to use as a strong pavement [4] R. Dachowskia, P. Kostrzewaa, The Use of Waste Materials in the Construction
Industry, World Multidisciplinary Civil Engineering-Architecture-Urban
base with minimal hot mix asphalt (HMA) overlay on top. It can
Planning Symposium 161 (2016) 754–758.
also be used in parking lots, walkways, and low volume roads as [5] S.W. Tabsh, A.S. Abdelfatah, Influence of recycled concrete aggregates on
wearing coarse without any HMA overlay. strength properties of concrete, Constr. Build. Mater. 23 (2009) 1163–1167.
Although a geopolymer is advantageous for its durability, the [6] S.K. De Juan, P.A. Gutiérrez, Study on the influence of attached mortar content
on the properties of recycled concrete aggregate, Constr. Build. Mater. 23
RCA-RCGPC needs to be checked for durability in sulfate and acid (2009) 872–877.
attacks as well as freeze–thaw cycles. [7] M. Casuccio, M.C. Torrijos, G. Giaccio, R. Zerbino, Failure mechanism of
recycled aggregate concrete, Constr. Build. Mater. 22 (2008) 1500–1506.
[8] Faiz Uddin Ahmed Shaikh, Hoang Long Nguyen, Properties of concrete
containing recycled construction and demolition wastes as coarse
Declaration of Competing Interest
aggregates, Journal of Sustainable Cement-Based Materials 2 (3–4) (2013)
204–217.
The authors declare that they have no known competing finan- [9] V.M. Malhotra, Global warming, and role of supplementary cementing
cial interests or personal relationships that could have appeared materials and superplasticizers in reducing greenhouse gas emissions from
the manufacturing of Portland cement, International Journal of Structural
to influence the work reported in this paper. Engineering 1 (2010) 116–130.
[10] T.R. Naik, Sustainability of Cement and Concrete Industries. Achieving
Sustainability, Construction (2015).
Acknowledgments [11] T. Salloum, Effect of Fly ash Replacement on Alkali and Sulphate Resistance of
Mortars, MS Thesis, Department of Building, Civil and Environmental
Engineering, Concordia University Montreal, Quebec, Canada, 2007.
The authors wish to express sincere thanks to the University of [12] B.V. Rangan, Fly Ash-Based Geopolymer Concrete. Research Report GC-4, Faculty
Louisiana at Lafayette for providing financial support and facility of Engineering, Curtin University of Technology, Perth, Australia, 2008.
for the research. A special thanks are also extended to Mr. Mark [13] J. Davidovits, Geopolymer chemistry and Applications, 2nd edition., Institute
Geopolymer, Saint-Quentin, France, 2008.
LeBlanc for assisting in experimentation. [14] P. Posi, C. Teerachanwit, C. Tanutong, S. Limkamoltip, S. Lertnimoolchai, V.
Sata, P. Chindaprasirt, Lightweight geopolymer concrete containing aggregate
from recycle lightweight block, Mater. Des. 52 (2013) 580–586.
References [15] V. Sata, A. Wongsa, P. Chindaprasirt, Properties of pervious geopolymer
concrete using recycled aggregates, Constr. Build. Mater. 42 (2013) 33–39.
[1] Advancing Sustainable Materials Management: 2018 Fact Sheet. [16] P. Nuaklong, V. Sata, P. Chindaprasirt, Influence of recycled aggregate on fly ash
Environmental Protection Agency, Washington, D.C, USA. https://www. geopolymer concrete properties, J. Cleaner Prod. 112 (2016) 2300–2307.
epa.gov/sites/production/files/2020-11/documents/2018_ff_fact_sheet.pdf. [17] P. Posi, P. Thongjapo, N. Thamultree, P. Boontee, P. Kasemsiri, P. Chindaprasirt,
Accessed December 5, 2020. Pressed lightweight fly ash-OPC geopolymer concrete containing recycled
[2] Sustainable Management of Construction and Demolition Materials. lightweight concrete aggregate, Constr. Build. Mater. 127 (2016) 450–456.
Environmental Protection Agency, Washington, D.C, USA. www.epa.gov/ [18] P. Nuaklong, V. Sata, A. Wongsa, K. Srinavin, P. Chindaprasirt, Recycled
smm/sustainable-management-construction-and-demolition- aggregate high calcium fly ash geopolymer concrete with inclusion of OPC and
materials#America. Accessed June 30, 2019. nano-SiO2, Constr. Build. Mater. 174 (2018) 244–252.
[3] D. Kralj, M. Markic, Building Materials Reuse and Recycle, Constr. Build. Mater. [19] Mahdi Koushkbaghi, Pedram Alipour, Behzad Tahmouresi, Ehsan Mohseni,
4 (2008) 409–418. Ashkan Saradar, Prabir Kumar Sarker. Influence of different monomer ratios

16
Sk Syfur Rahman and Mohammad Jamal Khattak Construction and Building Materials 283 (2021) 122624

and recycled concrete aggregate on mechanical properties and durability of [29] ASTM C78 / C78M-18, Standard Test Method for Flexural Strength of Concrete
geopolymer concretes. Construction and Building Materials. 2019. 205: 519-528. (Using Simple Beam with Third-Point Loading), ASTM International, West
[20] S.S. Rahman, M.J. Khattak, B. Adhikari, S. Adhikari, Discrete Element Modeling Conshohocken, PA, 2018, www.astm.org
of Bonded Soil Mixtures Under Uniaxial Compression and Indirect Tension [30] W.K. Lee, van Deventer, The effects of inorganic salt contamination on the
Test, Transp. Geotech. (2020). strength and durability of geopolymer, Colloids Surf. A 211 (2002) 115–126.
[21] F. Debieb, L. Courard, S. Kenai, R. Degeimbre, Roller compacted concrete with [31] Satoru Shimodaa, Elisabeth Smela, The effect of pH on polymerization and
contaminated recycled aggregates, Constr. Build. Mater. 23 (2009) 3382–3387. volume change in PPy(DBS), Electrochim. Acta 44 (1998) 219–238.
[22] M.I. Abu-Khashaba, I. Adam, A. El-Ashaal, Investigating the possibility of [32] S. Lee, M.-D. Seo, Y.-J. Kim, H.-H. Park, T.-N. Kim, Y. Hwang, Unburned
constructing low cost roller compacted concrete dam, Alexandria Engineering carbon removal effect on compressive strength development in a
Journal (2014.53:) 131–142. honeycomb briquette ash-based geopolymer, Int. J. Miner. Process. 97
[23] C. Chhorn, Seong Jae Hong and Seung-Woo Lee. A study on performance of roller- (2010) 20–25.
compacted concrete for pavement, Constr. Build. Mater. 153 (2017) 535–543. [33] B. Joseph, G. Mathew, Influence of aggregate content on the behavior of fly ash
[24] Emiliano Moreno-Pérez, Juan Hernández-Ávila, Yamile Rangel-Martínez, based geopolymer concrete, Scientia Iranica 19 (2012) 1188–1194.
Eduardo Cerecedo-Sáenz, Alberto Arenas-Flores, Ma. Isabel Reyes- [34] M. Olivia, H. Nikraz, Properties of fly ash geopolymer concrete designed by
Valderrama and Eleazar Salinas-Rodríguez.Chemical and Mineralogical Taguchi method, Mater. Des. 36 (2012) 191–198.
Characterization of Recycled Aggregates from Construction and Demolition [35] YOUJUN DENG; FLURY, Markus; HARSH, James B; FELMY, Andrew R; QAFOKU,
Waste from Mexico City. Minerals, 2018.8: 237 Odeta. Cancrinite and sodalite formation in the presence of cesium, potassium,
[25] ASTM D1557-12e1, Standard Test Methods for Laboratory Compaction magnesium, calcium and strontium in Hanford tank waste simulants, Applied
Characteristics of Soil Using Modified Effort (56,000 ft-lbf/ft3 (2,700 kN-m/ geochemistry, 2006. 21(12):2049-2063.
m3)), ASTM International, West Conshohocken, PA, 2012, www.astm.org. [36] De Ju WANG; Yi Tang; An Gang DONG; Ya Hong ZHANG; Ya Jun WANG. Hollow
[26] ASTM C1435 / C1435M-20, Standard Practice for Molding Roller-Compacted Cancrinite Zeolite Spheres in situ Transformed from Fly Ash Cenosphere,
Concrete in Cylinder Molds Using a Vibrating Hammer, ASTM International, Chinese chemical letters. 14(12):1299-1302.
West Conshohocken, PA, 2020, www.astm.org. [37] S Horpibulsuk; M Hoy; P Witchayaphong; R Rachan; A Arulrajah. Recycled
[27] ASTM C39 / C39M-20, Standard Test Method for Compressive Strength of asphalt pavement – fly ash geopolymer as a sustainable stabilized pavement
Cylindrical Concrete Specimens, ASTM International, West Conshohocken, PA, material. IOP Conference Series: Materials Science & Engineering, 2017. Vol.
2020, www.astm.org 273 Issue 1, p1-1, 1p.
[28] ASTM D3497-79(2003), Standard Test Method for Dynamic Modulus of [38] M.I. Martı´na, F. Andreolab, L. Barbierib, F. Bondiolib, I. Lancellottib, J. Ma.
Asphalt Mixtures (Withdrawn 2009), ASTM International, West Rincóna, M. Romeroan, Crystallisation and microstructure of nepheline–
Conshohocken, PA, 1979, www.astm.org forsterite glass-ceramics, Ceram. Int. 39 (2013) 2955–2966.

17