Construction and Building Materials 346 (2022) 128491

Contents lists available at ScienceDirect

Construction and Building Materials

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

Review

A critical review of engineered geopolymer composite: A low-carbon

ultra-high-performance concrete
Nouran Elmesalami, Kemal Celik *
Division of Engineering, New York University Abu Dhabi, Abu Dhabi, P.O. Box 129188, United Arab Emirates

A R T I C L E I N F O A B S T R A C T

Keywords: “Bendable concrete” or, more formally, engineered cementitious composite (ECC) is a special type of ultra-high-
Engineered geopolymer composites performance concrete that possesses strain-hardening properties with high tensile ductility. Despite their
Strain-hardening geopolymer composites improved performance, ECCs contain about two to three times higher ordinary Portland cement (OPC) content
Mechanical properties
than conventional concrete. Given the high carbon footprint of OPC, the environmental impact of ECCs is sig­
Micromechanical models
nificant. Although copious research has been conducted on replacing OPC in ECCs with low-carbon binders, e.g.,
blended cements, such materials can only partially replace OPC in ECCs. One promising alternative that can fully
replace OPC in ECCs is geopolymer, an alkali-activated cement. Research on replacing OPC with geopolymers in
ECCs started about a decade ago, and such composites have become known in the literature as Engineered
Geopolymer Composites (EGCs) or Strain-Hardening Geopolymer Composites (SHGCs). This paper provides an
extensive literature review of all published studies on EGCs to date. The paper starts with a scientometric analysis
using science mapping to provide a holistic overview of the current research progress on EGCs. Comparisons
between EGCs and ECCs in terms of their pseudo-strain-hardening behavior and mechanical properties are
provided. The effects of varying parameters related to the matrix, fibers, and curing conditions, on the perfor­
mance of EGCs, are also discussed. Finally, durability, self-healing, and environmental impact, among other
aspects of EGCs, are highlighted. The conducted review revealed the promising potential of EGCs as low-carbon
and ultra-high-performance concretes. However, research on EGCs is still developing, and further investigations
are needed before the full potential of EGCs can be realized.

1. Introduction and from the mid-1960s, increased progress has been achieved in the
field of fiber-reinforced concrete (FRC) [6]. In the 1970s, ultra-high-
Concrete is one of the most consumed materials on Earth after water strength cement pastes having low porosities and around 240 to 510
[1,2]. This is attributed to its versatility, high load-bearing capacity, and MPa compressive strengths were first developed by Yudenfreund et al.
durability, as well as the wide availability of its constituent materials [7] and Roy et al. [8]. In the early 1980s, two different types of ultra-
[2,3]. Despite its ubiquity, the ideal performance of concrete is actually high-strength and low porosity concretes, namely macro defect-free
limited by some properties, i.e., low tensile strength, brittle response, (MDF) and densified small particles (DSP) concretes, were developed
and large environmental footprint. Concrete strength is generally by Birchall et al. [9] and Bache [10], respectively. Such improvements
limited by several factors: water-to-cementitious material ratio, type of finally led to the development of high-performance fiber-reinforced
binders, size, shape, surface texture, grading and mineralogy of aggre­ concrete (HPFRC), which is prepared by using a low water-to-binder
gates, curing conditions, ages, and so on. In the 1930s, attempts to ratio (maximum of 0.25), a high content of supplementary cementi­
improve the properties of concrete included applying pressure to fresh tious materials (SCMs), admixtures, steel or synthetic fibers, and fine
concrete during setting to improve its strength. Also, during the 1960s, aggregates, without the use of coarse aggregates [11,12]. A modification
high concrete compressive strength values of 650 MPa were obtained by to HPFRCs is the ultra-high-performance fiber-reinforced concrete
compressing concrete samples and curing them under heat and high (UHPFRC) which was first introduced in the mid-1990 s by Richard and
pressures [4]. Other attempts to overcome the limitations of conven­ Cheyrezy [4] as reactive powder concrete (RPC). UHPFRCs show even
tional concrete included the use of fibers in the concrete matrix [2,5], higher strengths and durability than HPFRCs, with compressive

* Corresponding author.
E-mail address: [email protected] (K. Celik).

https://doi.org/10.1016/j.conbuildmat.2022.128491
Received 27 May 2022; Received in revised form 12 July 2022; Accepted 12 July 2022
Available online 21 July 2022
0950-0618/© 2022 Elsevier Ltd. All rights reserved.
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

strengths exceeding 150 MPa [12]. EGCs or SHGCs are proposed to tackle the problem of the high brittleness
Despite these advancements in concrete technologies, concrete of geopolymers while reducing the carbon footprint of ECCs.
structures still exhibit limitations in terms of their resilience, durability, Although many publications report on partially replacing OPC with
and sustainability, mainly attributed to the tendency of concrete to crack high contents of SCMs or other low-carbon binders in ECCs, there is
in tension [13]. To address this issue, engineered cementitious com­ limited research, and accordingly, few review papers [44], in the liter­
posite (ECC) or strain-hardening cementitious composite (SHCC) was ature on EGCs or SHGCs. Therefore, this paper aims to comprehensively
developed in the early 1990s by Victor Li. at the University of Michigan review published research on engineered geopolymer composites. This
[14,13]. It comes under the broad class of fiber-reinforced concrete and paper presents a holistic overview of available studies on EGCs, sum­
is considered a relatively newer group of UHPFRC. However, ECC is marizes the properties of EGCs in tables and charts, and includes ana­
different from UHPFRC in that it possesses high ductility with a tensile lyses of EGC properties compared with ECC properties, as well as
strain capacity exceeding 2 %, making it informally known as ‘bendable analyses of the environmental impact of EGCs. The paper starts with a
concrete.’ Additionally, while UHPFRCs are designed on the basis of scientometric analysis of the collected literature to graphically represent
dense particle packing, ECCs are designed by deliberately engineering the interrelationships between publication sources, authors, and coun­
the fiber, matrix, and fiber/matrix interface to achieve multiple cracking tries which provides an overview of the research trends on EGCs. A brief
in tension. Engineering of the microstructural components is based on a explanation of the micromechanical model of ECC that is necessary for
body of knowledge known as ECC micromechanics, which enables the preparing ECC mix designs is then provided, followed by comparisons
tailoring of ECC properties, such as compressive strength, tensile between the micromechanical model parameters of EGCs and ECCs.
strength, and ductility, depending on the requirements of each structure Detailed discussions on the effects of different parameters on the pseudo
[13]. strain hardening behavior and the mechanical properties of EGCs are
Despite their superior performance, a major limitation of ECC ma­ presented. Finally, we discuss findings on other aspects of EGCs,
terials is their high cement content. ECCs contain two to three times including their durability, self-healing performance, and environmental
higher ordinary Portland cement content than conventional concrete impact.
because of the elimination of coarse aggregates from the mix design
[15]. As a result, ECCs are costly and have a high carbon footprint. For 2. Scientometric analysis
every ton of cement produced, about 0.7 to 1.0 tons of carbon dioxide
are emitted [16,17], with cement production alone accounting for 7 % of Scientometrics is defined as the “quantitative study of science,
global carbon emissions [18–20]. Nearly half of the carbon emissions communication in science and science policy” [45]. In this paper, a
from cement arise from fuel burning, and the remaining are produced scientometric analysis is performed to provide an overall picture of the
from the calcination of limestone [17,21]. Over the past years, methods current research progress on EGCs and the active researchers in this
of reducing carbon emissions from cement production have been area. Through this scientometric analysis, important interpretations can
investigated, including energy efficiency improvements, use of alterna­ be drawn from the scientific literature, including main research themes
tive fuels and/or raw materials, replacement of clinker with SCMs, and in relation to EGCs, current gaps in research, and required future studies
development of alternative binders to OPC. Partially substituting OPC on the topic. To ensure complete coverage of all the currently published
with SCMs, such as ground granulated blast furnace slag (GGBFS) literature on this topic, several search engines were used to conduct the
[22,23], fly ash (FA), natural pozzolan (NP) [24,25], and calcined clay literature review, such as Scopus, Web of Science, and Google Scholar. The
[26,27], is deemed the most readily available method of reducing car­ results of this extensive search revealed that the first published study on
bon emissions from cement production [28]. But the high recycling rate EGCs was in 2012 by Lee et al. [46], followed by more publications in
of steel and the divergence from coal to cleaner energy sources will later years, amounting to a total of 55 publications. Fig. 1 illustrates the
constrain the long-term availability of GGBFS and FA [29]. Hence, the annual number of publications on EGCs from 2012 to 2021. A few
development of alternative binders to OPC has the greatest promise for a studies were published in the years from 2012 to 2015, followed by an
considerable CO2 emissions cut. increase in the publications till 2017. Thereafter, a brief decline in the
One of the promising alternative binders to OPC is geopolymer. number of published studies was observed until 2019, with a significant
Geopolymers are amorphous to semi-crystalline silico-aluminate mate­ increase occurring from 2020 onwards. It should be noted that this
rials with various applications ranging from building materials and literature review was conducted up to December 2021; thus, any studies
thermal insulations to radioactive waste containment and archeology. published after that date are not covered in this review. Based on the
They are synthesized from two main constituent materials: the source current trend in publications on EGCs and the information about
material, and the alkaline activator. The source material, or precursor, currently active projects on EGCs mentioned in those publications, a
should contain a high amorphous alumina and silica content, which can further increase in the number of publications is expected in the coming
be obtained from natural minerals such as kaolinite, metakaolin, or NP, years.
or industrial by-products such as FA or GGBFS [11 12]. The alkaline Science mapping is one method of performing bibliometric or sci­
activator can be any strong alkaline solution, such as calcium hydroxide entometric analysis which aids in presenting the structural and dynamic
or sodium hydroxide, which breaks down the covalent bonds of the aspects of scientific research. It provides a spatial representation of how
source material triggering the formation of the binding structure fields, publications, or research groups are related to each other
[30–32]. This process of creating cements by activating aluminosilicate [47–49]. Numerous software tools are available for performing science
materials with alkaline solutions is thought to have started in the 1940 s, mapping, each having its advantages and drawbacks [50,51]. Using such
and such materials are referred to as alkali-activated binders [33,34]. software tools helps provide statistical summaries of the different as­
The name ‘Geopolymers’ was first coined by Joseph Davidovits in the pects of published research, which are very difficult and impractical to
1970 s [30–32], and geopolymers are considered a subset of the larger obtain through manual literature reviews. In this study, the program
group of alkali-activated binders [35]. Compared to OPC production, VOSviewer [52] was used owing to its improved visualization and
geopolymer production requires 30 % less energy [36,37] and results in graphical representations compared to other science mapping software
up to 80 % lower carbon dioxide emissions, making geopolymers a [50,51]. The collected bibliographic data was first imported from Scopus
promising low-carbon alternative binder to OPC [36–40]. It is also found into VOSviewer, then used to construct maps displaying items, such as
that fly ash-based geopolymer concrete shows comparable or superior authors, journals, countries, or keywords, from the imported bibliog­
compressive strength, tensile strength, elastic properties, and durability raphy. The maps display those items in clusters connected by links based
to OPC concrete [30,35]. Despite these advantages, geopolymers are on authors’ co-citations or keywords co-occurrence, with each cluster
inherently brittle, making them more prone to cracking [41–43]. Hence, representing a group of related studies. The sizes of fonts and circles on

2
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

Fig. 1. Summary of annual publications on EGCs till December 2021.

the maps reflect the significance of the items in terms of the number of Materials is the most influential journal in the area of EGCs, with several
publications for authors, journals, and countries and in terms of fre­ and stronger links to other journals in this field. Fig. 3 shows the active
quency of occurrence for keywords; the more prominent an item is, the researchers on the topic of EGC. Zoomed-out views of the clusters are
larger the size. Also, the thickness of a link resembles the strength of the shown on the map for clarity. From the 53 publications, a total of 113
connection between two items, with thicker links referring to stronger authors were identified in VOSviewer and were further refined into au­
connections. In addition, VOSviewer adopts the ‘distance-based’ type of thors who have at least three publications and one citation resulting in a
bibliometric maps in which the distances between items relate to the total of 20 authors. The clusters and connection lines between the au­
strength of the connections between the items, with smaller distances thors on the map are related to the number of co-authored documents.
reflecting stronger connections [52]. It should be noted that the science The map indicates that some of the most productive authors on EGCs
mapping was conducted for 53 out of the 55 collected studies since the include Nematollahi B., Sanjayan J., Li V., Lee B., and so on. The same
remaining two studies are PhD dissertations that are not detected by authors are also found to have established strong research collabora­
scientific databases such as Scopus or Web of Science. tions, as indicated by the links between the authors’ clusters. Further­
The publication sources of the collected studies and reciprocal cita­ more, the map shows that scholars such as Zhang M., Zahid M., Shafiq
tions among them are demonstrated in Fig. 2. The result indicates that N., and so forth contribute to research on EGCs with few connections to
the journals of Construction and Building Materials and Cement and Con­ other researchers in the same field. The collaborations between the
crete Composites have been highly mutually cited compared to publica­ authors are also highlighted in the network map of countries shown in
tions from the other journals. It should be noted that journals that are Fig. 4. A total of 19 countries are identified in VOSviewer, restricting the
more interrelated based on mutual citations are grouped into clusters of results to countries with a minimum number of publications and cita­
the same color. The map also shows that Construction and Building tions of 3 and 1, respectively, resulting in 12 countries displayed on the

Fig. 2. Science mapping of sources of publications on EGCs. Note: The full names of journals cannot be displayed in the VOSviewer maps.

3
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

Fig. 3. Science mapping of active authors in the field of EGC. Note: Some clusters are shown zoomed out for clarity.

Fig. 4. Science mapping of countries of publications. Note: Some clusters are shown zoomed out for clarity.

map. The countries are clustered and connected based on the number of articles, as shown in Fig. 5. In this map, the links between the clusters
co-authored papers. It can be seen from the map that the largest number relate to the number of documents in which the keywords occur
of publications on EGCs are produced in China and the United States, together, with the minimum number of occurrences of a keyword set to
followed by Australia, Denmark, and Brazil. Also, the map shows that seven. Generic terms, such as “strain-hardening,” “fiber-reinforced
international research collaborations on EGCs are scattered globally, as composites,” “cementitious composites,” etc., were eliminated from the
evident from the map clusters. For instance, strong research collabora­ analysis. The map in Fig. 5 shows that some of the most commonly
tions are identified between the United States and Australia, and be­ occurring keywords in the EGC studies include “fly ash,” “slags,” and
tween China and Denmark, with no identified collaboration between the “PVA fibers.” In addition, the most commonly occurring terms con­
United States and China. The results from this science mapping provide cerning experimental tests include “tensile strength,” “tensile strain,”
the big picture of the current research directions on EGC and would “compressive strength,” and “micromechanics,” among other related
allow researchers from various geographical locations to identify po­ terms. The results of this science mapping analysis show that the most
tential collaborations and share ideas in this area. commonly investigated materials for use in EGCs include fly ash, slags,
To obtain an overall picture of the different studied aspects of EGCs, and PVA fibers. The results also show that the most researched aspects in
science mapping analysis was conducted based on the keywords of relation to EGCs are their mechanical properties and micromechanical

4
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

Fig. 5. Science mapping based on keywords.

parameters, while there are research gaps in relation to their durability,

self-healing, and dynamic behavior. These interpretations are further
reinforced by additional analyses conducted in later sections of this
paper. Based on the results of the conducted science mapping analysis
and the later discussion in this paper, some recommendations for future
studies are provided towards the end of the paper.

3. ECC micromechanical model

The strain-hardening behavior of engineered cementitious compos­

ites is achieved through the synergistic interactions of fibers, matrix, and
fiber–matrix interface. The micromechanical model of ECC was devel­
oped as an effective tool for establishing the links between the com­
posite’s microstructure and macro-behavior. Li and Leung [53] first
introduced the principles of micromechanics in 1992, which were later
developed significantly by various researchers in the 1990 s [54]. The
ECC micromechanical model developed by Yang et al. [55] is the one
currently being used by researchers. In simple terms, the ECC micro­ Fig. 6. Typical σ-δ curve for the strain-hardening behavior of ECC. The hatched
mechanical model is based on two main criteria that need to be satisfied area represents the complementary energy (J’b), and the shaded area represents
to achieve strain-hardening in the composites. These criteria are set up the matrix crack tip toughness (Jtip) [13].
in terms of a fiber bridging stress versus crack opening relationship (σ-δ),
shown in Fig. 6 [13], which models the average load carried by indi­ and on the experimentation required for determining the model pa­
vidual fibers bridging a crack at different embedment lengths and rameters can be found in the literature [13].
various orientation angles. This requires determining a model that de­ The two criteria for achieving pseudo-strain-hardening (PSH) in the
scribes the behavior of a single fiber as it is pulled out of the matrix and composites are referred to as the strength-based criterion, illustrated by
undergoes interfacial debonding and sliding: a P(δ) model. Such a model Equation (1), and the energy-based criterion, represented by Equation
can be obtained in terms of the ECC micromechanical parameters listed (2). The strength-based criterion requires that the first cracking strength
in Table 1. of the matrix, σ fc, does not exceed the maximum fiber bridging strength,
All these micromechanical parameters can be obtained from exper­ σ o. This ensures that the initiation of a new crack from a preexisting flaw
imental tests. Detailed information on the ECC micromechanical model occurs at a tensile stress level that is lower than the fiber bridging

5
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

Table 1 parameters that affect the properties of EGCs include fiber-related pa­
ECC micromechanical model parameters. rameters and parameters related to the curing conditions. The fiber-
Parameter Symbol related parameters include fiber content, type, and geometry, while
the parameters related to curing include curing method, temperature,
Fiber-related parameters Diameter df
Length Lf and duration. A few studies have investigated the feasibility of using
Elastic modulus Ef geopolymers as alternative binders to OPC in ECC. These studies have
Tensile strength σf also investigated the effects of the previously mentioned parameters on
Volume fraction vf EGC pseudo-strain-hardening, mechanical behavior, and durability
Matrix-related parameters Matrix fracture toughness Km
Matrix elastic modulus Em
properties.
Fiber/matrix interfacial parameters Frictional bond strength τo Table 2 summarizes the mix constituents and proportions used in all
Chemical bond strength Gd the collected studies on EGC from the literature. The studies are grouped
Slip-hardening coefficient β based on the type of precursor used. The collected literature shows that
Snubbing coefficient f
the most common materials used to produce EGCs include fly ash or
Fiber strength reduction factor f’
GGBFS for the precursors, sodium silicate and sodium hydroxide solu­
tions for the alkali activators, and polyvinyl alcohol (PVA) for the fibers.
capacity, and thus avoids failure by the formation of a single crack and Among the various aluminosilicate materials available, fly ash is the
ensures the occurrence of multiple cracking. The energy-based criterion most commonly used precursor in EGCs mainly because of its wide
maintains that the steady-state flat crack propagation mode prevails availability as a by-product of coal-fired power plants and its low water
over the Griffith-type crack propagation. Unlike the Griffith crack requirement for high workability [60–63]. A major problem with the use
propagation mode, the steady-state crack propagation mode results in of fly ash in geopolymers is the requirement for heat curing to achieve
flat cracks with a constant crack opening under constant load. If one of desirable composite properties [60,64–67]. For this reason, GGBFS has
these two criteria is not met, the composite response defaults to the been used to fully or partially replace fly ash in EGCs owing to the high
tension-softening typical of ordinary fiber-reinforced concrete [13]. calcium oxide content of GGBFS, which results in higher reactivity and
increased early strength gain of the geopolymer composite without the
σ fc ≤ σo (1)
need for heat curing [60,63,66,67]. Various types of fibers have been
investigated in fiber-reinforced concretes, such as steel, glass, basalt,
(2)
′
Jtip ≤ Jb
polyethylene (PE), polypropylene (PP), PVA, natural fibers, etc. Among
In Equation (2), Jtip represents the matrix toughness at the crack tip, these fibers, PVA fibers are reported to be the most favorable for use in
which approaches K2m/Em for small fiber content, and J’b is the com­ high-performance FRC, including ECC, owing to their higher tensile
plementary energy determined from the graph of the fiber bridging strength and elastic modulus compared to several other synthetic fibers,
stress versus crack opening relationship (σ-δ). The ratios of σo to σ fc and as well as their high stability in alkaline environments [61,62,68]. Micro
J’b to Jtip are referred to as the stress-performance index (PSHstress) and PVA fibers of lengths from 8 to 12 mm and diameters of about 40 µm
the energy-performance index (PSHenergy), respectively [56]. Kanda and were the most commonly used in the studied EGC mixes. Moreover, like
Li [57] demonstrated from experimental data that PSHstress greater than typical ECC M45, most of the studies on EGCs incorporated sand as fine
1.3 and PSHenergy greater than 2.7 are required to achieve saturated PSH aggregates in the mix. Further discussions on the effects of each of these
behavior – the state at which no new micro-cracks can be produced parameters, as well as other mix constituents and their proportions on
under tensile load. PSH behavior, mechanical performance, durability, and self-healing
In the first generation of ECCs, the binder was prepared of 100 % properties of EGCs, are provided in the following sections. Also, Fig. 7
OPC dictated by the need to control matrix toughness to achieve strain provides a summary of the most important conclusions and outcomes
hardening and the need to control rheology for ease of fiber dispersion from the conducted literature on the effects of each of the parameters on
[13,15]. Greener versions of ECC have been developed by partially the PSH behavior and mechanical performance of EGCs, as well as rec­
replacing OPC with industrial by-products, such as fly ash and GGBFS. ommendations for future studies.
One such example is a composite developed by Wang and Li [15] that
incorporated polyvinyl alcohol (PVA) fibers and a relatively high fly ash-
to-OPC ratio of 1.2 while maintaining the desired PSH behavior. This 4.1. Pseudo strain-hardening behavior
composite has become known as ECC M45 and is reported to be the most
studied and the most commonly used ECC in the field [55,58]. There­ Only a few studies have investigated the application of the current
fore, ECC M45 is considered the benchmark composite in this paper for ECC micromechanical model for designing EGCs or SHGCs. The design
comparison with EGCs. of EGC is more complicated than the design of ECC since there are many
more parameters that affect the properties of EGC compared to ECC,
4. Properties of engineered geopolymer composites particularly parameters related to the matrix and interface [58,77,94].
Some studies investigated the effects of various factors related to the
In geopolymers, the alkaline activator dissolves the silico-aluminate matrix, fibers, and curing conditions on the micromechanical model
precursor material releasing aluminate and silicate species which parameters of EGCs. The investigated factors include precursor type,
become incorporated into the aqueous solution and form a gel structure. activator type, and molarity, sand content and size, water-to-
Further reorganization and polymerization of the system occur, and the geopolymer solids ratio (W/GP), fiber type and geometry, as well as
final geopolymer structure forms and hardens. The matrix composition curing temperature and period.
of geopolymers and the chemistry of the mix constituents greatly affect
these geopolymerization reactions, which in turn affect the final prop­ 4.1.1. Effects of matrix-related parameters
erties of the geopolymer concrete [59]. Therefore, the performance of
EGCs is highly dependent on parameters related to their geopolymer 4.1.1.1. Precursor. As previously mentioned, fly ash is the most
matrices. These matrix-related parameters can be classified into pre­ commonly used precursor in geopolymers. Still, it is being fully or
cursor type, alkali-activator type and chemical composition, ratio of partially replaced by GGBFS or slag since the high calcium oxide content
alkali-activator to precursor, ratio of water to precursor, sand details of slag increases reactivity without the need for heat curing
(content, type, and size), and sand-to-precursor ratio. In addition, other [60,63,66,67]. Some studies examined the effects of replacing fly ash
with slag in EGCs. In Nematollahi et al. [64], it was found that the

6
 N. Elmesalami and K. Celik
Table 2
Summary of mix constituents from all EGC studies collected from literature.
References Precursor Alkali activator Activator Activator Activator Activator Water-to- Sand type Sand-to- Fiber type Fiber Fiber Curing Total
type type NaOH Na2SiO3 Na2SiO3/ solution- geopolymer binder ratio (and content) length diameter condition curing
molarity modulus NaOH to- solids ratio* (mm) (μm) period
(SiO2/ precursor (days)
Na2O) ratio

[46;69;70;71;72] GGBFS Ca(OH)2; Ca – 0.94 – 0.093 to 0.26 to 0.45 0.4 PVA (0.25 % 8 to 18 12 to 180 Ambient 28 and
(OH)2 and 0.12 No sand; fine − 2 % vol.); temperature 30
Na2SO4 powder; silica sand PE curing
Ca(OH)2 and 1.5 % − 1.75
Na2SiO3 powder; % vol.);
Na2SiO3 powder UHMPE (2 %
vol.); copper-
coated steel
(2 % vol.)
[41;42;73;74;75;76;77;63;66;78;43;79;80] Class F fly Na2SiO3 solution; 7 M to 12 2 to 3.2 2.5 to 4.5 0.014 to 0.1 to 0.4 No sand; fine 0.03 to 0.6 PVA (1 %, 8 to 25 12 to 40 Ambient 1; 3; 7;
ash NaOH and M 0.6 silica sand; 1.5 % and 2 % temperature 28; 56;
Na2SiO3 solution; coarse silica vol.); PE curing; heat 70; 120
NaOH solution; sand; micro (1.75 % − 2 curing
KOH and K2SiO3 silica sand; % vol.); PP (2 followed by
solution; Ca lightweight % vol.); PET ambient
(OH)2 and expanded glass; (2 % vol.) temperature
Na2SiO3 powder; hollow ceramic curing
NaOH and microspheres;
Na2SiO3⋅5H2O expanded
powder perlite; natural
river sand
[61;62;68] Class C fly NaOH and 2 1.68 to 0.36 to 0.2 to 0.26 Micro silica sand 0 to 0.8 PVA (2 % 8 to 24 40 to 200 Heat curing 28
7

ash Na2SiO3 solution 8 M to 16 2.5 0.45 vol.) only; heat

M curing
followed by
ambient
temperature
curing
[81;82] POFA NaOH and 8 M to 14 3.3 2.5 0.42 to 0 to 0.1 Dune sand 1.8 PVA (2 % 8 40 Heat curing 3; 7;
Na2SiO3 solution M 0.55 vol.) followed by 14; 28
ambient
temperature
curing

Construction and Building Materials 346 (2022) 128491

[83;84;85;86;87;88;89;90] Metakaolin NaOH and 6 M to 14 1.25–3 3 0.8 to 1.85 0.26 No sand; River 0.4 to 0.5 Natural 6 to 12 13 to 40 Ambient 7; 14;
Na2SiO3 solution; M sand; fine quartz Curauá, sisal temperature 16; 28
NaOH solution in sand; silica sand and jute (10 curing
water and fumed % wt.); PVA
silica; KOH (2 % vol.); PE
solution in water (2 % vol.);
and fumed silica UHMWPE (1
%− 2%
vol.); PBO (2
% vol.)
[65;91] Ternary K2SiO3 – – – 0.12 0.25 Silica sand 1.36 Micro steel 6 to 50 15 to Ambient 28
blend of fly (1–3 % vol.); 1000 temperature
ash, GGBFS macro steel curing
and silica (1 % vol.);
fume PVA (1–2 %
(continued on next page)
 N. Elmesalami and K. Celik
Table 2 (continued )
References Precursor Alkali activator Activator Activator Activator Activator Water-to- Sand type Sand-to- Fiber type Fiber Fiber Curing Total
type type NaOH Na2SiO3 Na2SiO3/ solution- geopolymer binder ratio (and content) length diameter condition curing
molarity modulus NaOH to- solids ratio* (mm) (μm) period
(SiO2/ precursor (days)
Na2O) ratio

vol.); Glass
(1 % vol.)
[64;92;70;63;66;93;94;67;95;96] Blends of Na2SiO3 powder; 8 M to 12 0.8 to 2 to 2.5 0.071 to 0.2 to 0.45 No sand; natural 0 to 0.5 PVA (1 % − 2 5 to 20 12 to 180 Ambient 3; 28;
Class F fly NaOH and M 3.15 0.6 river sand; fine % vol.); PE temperature 70
ash and Na2SiO3 solution silica sand (0.5 % − 2 % curing; heat
GGBFS vol.); curing
UHMWPE (2 followed by
% vol.); ambient
copper- temperature
coated steel curing
(2 % vol.);
recycled tyre
steel (RTS)
(0.25 % and
0.5 % vol.);
recycled tyre
polymer
(RTP) (0.25
% − 1 %)
[97;98] Blend of NaOH and (not 3.2 to 3.3 4.6 0.3 0.23 Fine and micro 0.26 to PVA (1.1 % 12 39 Heat curing 3; 7; 28
Class C and Na2SiO3 solution reported) silica sand 0.276 wt.; 1.5 % followed by
Class F fly vol.) ambient
ash and temperature
8

metakaolin curing
[99] Blends of NaOH and (not 3.3 (not 0.31 0.24 Micro silica sand 0.3 PVA (1.1 % 12 39 Heat curing 28
Class F and Na2SiO3 solution reported) reported) wt.) followed by
Class C fly ambient
ash temperature
Blends of curing
zeolite, Class
F fly ash, and
Class C fly
ash
[60;100] Blends of NaOH and (not 3 (not (not clear 0.35 Fine silica sand (not clear PVA (0.15 %, 12 50 Ambient 3; 7; 28
class F fly Na2SiO3 solution reported) reported) from data) from data) 0.2 %, 0.25 % temperature
ash, silica vol.) curing

Construction and Building Materials 346 (2022) 128491

fume, MWCNTs 0.02 – 0.01 –
metakaolin, (0.05 %, 0.1 0.03 0.02
and GGBFS %, 0.15 % by
mass)
[101] Blends of NaOH and (not 3.2 4.5 0.3 0.22 (not reported) (not PE (1.1 % 12 24 Heat curing 28
silica fume, Na2SiO3 solution reported) reported) wt.) followed by
GGBS and ambient
IFA temperature
curing
[102] Blends of fly NaOH and (not 1.5 9.5 0.4 0.1 No sand – PVA (1 % 12 40 Ambient 7; 28
ash and steel Na2SiO3 solution reported) vol.) temperature
slag curing; heat
curing
followed by
ambient
(continued on next page)
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

friction bond strength was higher for EGC prepared with 50 % fly ash
and 50 % typical slag (containing 2 % gypsum) than for EGC with 50 %

and 28
period
curing

(days)
Total

7, 21
fly ash and 50 % gypsum-free slag. In contrast, no particular trend was

28

*Considered as the ratio of the total mass of water contained in the activator and the mass of any extra added water to the total mass of solids in the precursor and activator, as defined by Hardjito et al.[105].
observed for the chemical bond and the slip-hardening coefficient for
the different slag types. Moreover, the matrix elastic modulus and

temperature

temperature

temperature
followed by

followed by
Heat curing

Heat curing
condition

fracture toughness were about 12–29 % and 17–21 % higher, respec­

ambient

ambient
Curing

curing

curing

curing
tively, for typical slag than for gypsum-free slag. Consequently, EGC
prepared with the typical slag showed superior PSH behavior than EGC
prepared with the gypsum-free slag, in which the PSHstress and PSHenery
diameter

indices were around 29–30 % and 15–21 % higher, respectively [64]. In

Fiber

(μm)

24

18
another study, it was found that the matrix elastic modulus, fracture
toughness, and Jtip were slightly lower for EGC prepared with 100 % slag
length
(mm)
Fiber

as compared to EGC prepared with 50 % slag and 50 % fly ash [70]. The
12

6 differences in the PSH behavior of EGCs prepared with different types of

binder ratio (and content)

and 3 % vol.)
PP (1 %, 2 %
precursors were attributed to the differences in the microstructures of
PE (0.95 %
Fiber type

the EGC mixes [64,70]. A study by Kan et al. [99] investigated the PSH
behavior of EGC prepared with fly ash as the precursor with partial
mentioned) wt.)

replacement by zeolite, which is a type of aluminosilicate mineral ma­

terial. The results showed that EGC prepared with 3 % fly ash replace­
Sand-to-

ment by zeolite exhibited a 65 % higher chemical bond and three times

(not

the frictional bond value of EGC with no zeolite. Also, EGC prepared
1.5

with 3 % zeolite showed 25 % and 11 % lower Km and Jtip, respectively,

(not mentioned)

compared to EGC prepared with no zeolite. Overall, EGC prepared with

Natural river

3 % zeolite showed improved PSHenergy index in comparison to EGC with

Sand type

no zeolite indicating the higher ductility of the former [99]. In another

sand

study, Kan et al. [101] examined the performance of EGC prepared with
silica fume and GGBFS as the precursor with partial replacement by
solids ratio*
geopolymer

incineration fly ash (IFA), which is a by-product of municipal solid waste

Water-to-

incineration. The study revealed that the EGC mixes containing IFA
0.18

showed larger values of PSHenergy, PSHstress, τo, and β, than the control
0.2

EGC mixes without IFA [101]. The higher values of τo and β were
precursor
Activator

attributed to the increase in fracture toughness with IFA replacement

solution-

which increases surface roughness and enhances the fiber/matrix con­

ratio

0.35
0.3
to-

tact areas [101]. Furthermore, Zhao et al. [103] studied the performance
of EGC prepared with a precursor comprising fly ash, GGBFS, and partial
Na2SiO3/
Activator

replacement with steel slag (SS), a by-product of the steel manufacturing

NaOH

4.5

2.5

industry. The results showed that EGC mixes prepared with 5 %-25 % SS
replacement met the criteria for saturated strain hardening with PSHe­
Activator

modulus
Na2SiO3

nergy and PSHstress values exceeding 6 and 1.2, respectively [103].

(SiO2/
Na2O)

3.2

4.1.1.2. Alkaline activator. Some researchers studied the effects of

reported)
Activator

different types of alkaline activators on the PSH behavior of EGCs

molarity
NaOH

12 M

[42,76,85,94]. As previously highlighted, the alkaline activator most

(not

commonly used in studies on EGCs is a sodium hydroxide and sodium

silicate solution. Other types of activators have also been studied, such
Na2SiO3 solution

Na2SiO3 solution
Alkali activator

as lime-based activators, composed of calcium hydroxide and sodium

Blends of fly NaOH and

NaOH and

sulfate or sodium silicate solutions, and potassium-based activators,

composed of potassium hydroxide and potassium silicate solutions. It
type

was found that alkaline activators composed of sodium hydroxide and

sodium silicate solutions result in a higher matrix elastic modulus of
and steel slag

Bagasse ash
ash, GGBS

geopolymers than potassium-based activators [42,76,85]. Some studies

Precursor

showed that sodium-based activators resulted in higher Km and Jtip as

type

compared to potassium-based activators [42,76], whereas another study

showed that potassium-based activators resulted in higher Km and Jtip of
geopolymer matrices than sodium-based activators [85]. The differences
in the matrix fracture properties are thought to be a result of the
different microstructures, and varying compositions of the aluminosili­
cate gels owing to the different types of activators used [42,76]. For the
fiber–matrix interface parameters, it was found that sodium-based ac­
tivators result in a higher frictional bond and slip-hardening coefficient
Table 2 (continued )

in comparison with potassium-based activators [76]. In contrast, higher

chemical bond strengths are attained in EGCs prepared with potassium-
based activators than in EGCs prepared with sodium-based activators
References

[76,85]. It is thought that the chemical bond between the potassium-

[103]

[104]

based activator geopolymer matrix and the hydroxyl group on the

9
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

Fig. 7. Framework figure summarizing the most important conclusions, outcomes, and future recommendations from the collected literature on the pseudo-strain-
hardening behavior and mechanical performance of EGC.

fiber surface could be stronger than with sodium-based activators fiber–matrix interface parameters up to a certain limit. The parameters
resulting in the higher Gd of potassium-based EGCs [76]. Building from of τo and Em increase up to a silicate modulus value of 1.2, beyond which
the attained micromechanical model parameters, the PSHstress and their values start to decrease. Similarly, Gd, β, and Km parameters in­
PSHenery indices were found to be higher for sodium-based EGCs than for crease up to a silicate modulus of 1.0 and then decrease with a further
potassium-based EGCs SPS:refid::bib76_bib85[76,85], wherein some increase in silicate modulus [94]. A study by Zahid et al. [68] analyzed
cases, the PSHenery index could not be obtained for potassium-based the PSH behavior of EGCs prepared with different molarities of sodium
EGCs owing to the dominance of the chemical bond between the fiber hydroxide in the activator. The study showed that EGCs prepared with 8
and the matrix [85]. Similar relations in terms of PSH behavior and the M NaOH activator showed about 20 % lower Em than EGCs prepared
micromechanical model parameters were observed when comparing with 12 M and 16 M NaOH activators. On the other hand, the samples
sodium-based activators with lime-based activators in EGCs, in which formulated with 8 M NaOH exhibited higher Km values and higher
sodium-based EGCs exhibited superior PSH behavior compared to lime- PSHenery index than the ones with 12 M and 16 M NaOH [68]. Addi­
based EGCs [42]. Furthermore, Zhang et al. [94] studied the effects of tionally, Nguyen et al. [79] investigated the effects of varying the ratio
alkaline activator silicate modulus (Ms = SiO2/Na2O molar ratios) on the of activator sodium metasilicate pentahydrate (SMP) to NaOH on the
PSH characteristics of EGCs. The results showed that increasing the performance of EGC and showed that all mixes achieved ultra-ductile
silicate modulus improves the matrix fracture properties and the behavior, with the mix prepared with an SMP-to-NaOH ratio of 1.5

10
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

exhibiting the highest PSHstress and PSHenery indices of 1.84 and 322, PSHstress and PSHenery indices by 25 % and 6 %, respectively, compared
respectively. to EGCs with PE fibers [92]. Another study evaluated the friction bond
strength between geopolymer matrices and three types of fibers, namely
4.1.1.3. Fine aggregates. Other matrix-related parameters that affect the PVA, ultra-high molecular weight polyethylene (UHMWPE), and poly(p-
PSH behavior of EGCs are the sand content and size. Studies have found phenylene-2,6-benzobisoxazole) (PBO). At the same fiber volume frac­
that adding sand to the EGC matrix increases both the matrix elastic tion of 2 %, PBO fibers exhibited about 3 times and 2.5 times higher
modulus and fracture toughness [70,73,85]. It was found that doubling frictional bond strength than UHMWPE and PVA fibers, respectively
the sand-to-binder ratio from 0.3 to 0.6 increased Em and Km by 14–36 % [86]. Furthermore, Zhong and Zhang [96] showed that partially
and 16–33 %, respectively [73]. While one study showed that the replacing PVA fibers with recycled tyre polymer (RTP) fibers reduced
addition of sand to EGC increased Jtip [70], another study showed that both the PSHstress and PSHenery indices of EGCs, with an increasing RTP
sand inclusion slightly reduced Jtip compared to EGC with no sand [85]. replacement ratio resulting in further reduction of PSH behavior. In
Moreover, adding sand to the EGC matrix increased the fiber–matrix addition, increasing the PVA fiber content was found to improve both
interface parameters τo, Gd and β. The effect of sand incorporation on the the PSHstress and PSHenery indices [71,96]. In a study by Zahid et al. [68],
PSHenery index depended on the type of activator used. Adding sand to the effects of using coarse PVA fibers (0.2 mm in diameter and 18 or 24
EGC prepared with a sodium-based activator increased the PSHenery mm in length) were compared against the effects of using fine PVA fibers
index. At the same time, it resulted in a shift to strain-softening behavior (0.04 mm in diameter and 8 or 12 mm in length) on the PSH behavior of
with a potassium-based activator [85]. Also, one study showed that the EGC. The results revealed that fine PVA fibers yielded higher matrix
matrix fracture toughness was 7–22 % lower for EGC containing coarse fracture toughness than coarse PVA fibers, which was attributed to the
sand (particle size between 600 μm and 1.18 mm) than for EGC with fine larger number of fine fibers per unit volume of the matrix resulting in
sand (maximum particle size of 212 μm). This was ascribed to the close dispersion of the fine fibers allowing them to arrest more cracks.
smaller number of coarse sand particles in the matrix compared to fine Also, the study showed that using coarse PVA fibers in combination with
sand, resulting in less tortuous crack propagation. It was concluded that activators composed of high molar NaOH solution resulted in PSHenery
a high content of fine sand and the inclusion of coarse sand in EGC indices lower than 3, which do not satisfy the PSH criteria [68]. Another
significantly decreases the two PSH indices hindering saturated PSH study by Nematollahi et al. [76] assessed the effects of fiber surface oil
behavior [73]. coating on the PSH behavior of EGCs. The study found that surface oil
coating of PVA fibers resulted in lower Gd and β by 31–65 % and 43–93
4.1.1.4. Water-to-geopolymer solids ratio. Another important parameter %, respectively, depending on the activator type. This was speculated to
affecting the PSH behavior of EGCs is the ratio of the water-to- result from the reduced hydrophilic nature of PVA fibers due to the oil
geopolymer solids. Studies on EGCs have adopted the definition of the coating. Conversely, fiber surface oil coating resulted in a 35 % lower τo
water-to-geopolymer solids (W/GP) ratio proposed by Hardjito et al. with a potassium-based activator but caused about a 51 % increase in τo
[105], which defines the W/GP ratio as the ratio of the total mass of with a sodium-based activator [76].
water used to prepare the activator plus the mass of any extra water
added in the mix to the total mass of precursors and other solids in the 4.1.3. Effects of curing conditions
activator. Increasing the W/GP ratio from 0.2 to 0.23 was found to As previously highlighted, fly ash is the most commonly used pre­
decrease both the matrix elastic modulus and fracture toughness of EGC cursor in EGCs. However, one drawback of using fly ash in geopolymers
matrices by 24–45 % and about 38 %, respectively [73,76]. Similar re­ is that it requires heat curing to attain comparable physical, mechanical,
sults were reported by Choi et al. [71], in which a W/GP ratio of 0.4 and durability properties to conventional concrete. This heat curing is
resulted in higher Km and Jtip than for a W/GP ratio of 0.45 by about 6 % costly and makes large-scale in-situ applications of EGCs difficult
and 13.6 %, respectively. This effect was found to be somehow similar to [60,64–67]. Therefore, a few studies investigated the effect of utilizing
the well-known effect of the water-to-cement ratio on the properties of ambient temperature curing on the PSH behavior of fly ash-based EGCs
cement-based matrices [73]. Regarding the effect on fiber–matrix [64,92]. It was generally found that ambient temperature curing results
interface parameters, increasing the W/GP ratio was found to reduce the in slightly higher Km and Jtip of EGC compared to heat curing. This was
chemical bond strength [71,76]. However, opposing results are reported thought to be mainly a result of the different geopolymer microstruc­
for the frictional bond strength, in which one study showed that tures of the EGCs cured under different temperatures. Also, no signifi­
increasing the W/GP ratio increased τo [76], whereas another study cant effect was observed on the Em value under different curing
showed that it reduced τo [71]. No considerable effect of the W/GP ratio temperatures. Thus, it was concluded that ambient temperature curing
on the slip-hardening coefficient was observed. The lower chemical decreases the brittleness of EGCs. Also, the frictional bond strength was
bond was attributed to the reduced density of the matrix microstructure found to be higher with ambient temperature curing than with heat
near the fibers and the lower pH resulting from the higher W/GP ratio curing. This was attributed to the higher matrix fracture toughness of the
[71,76]. In general, reducing the W/GP ratio resulted in a higher ambient temperature cured EGC, which increased the fracture surface
PSHenergy index [71]. roughness and thus the contact surface in the fiber–matrix interfacial
zone during slippage, leading to a higher frictional bond [64,92].
4.1.2. Effects of fiber-related parameters Although ambient temperature curing was found to result in higher Gd
A few studies examined the effects of fiber-related parameters, and β values as compared to heat curing [64,92], the opposite trend was
including fiber type, geometry, and oil coating, on the PSH behavior of observed when gypsum-free slag was used instead of typical slag in the
EGCs. One study compared the effects of PVA fibers against those of PE precursor in which ambient temperature curing resulted in lower Gd and
fibers on the PSH behavior of EGCs [92]. It was found that PVA fibers β values [64]. Again, this was attributed to the different microstructures
resulted in considerably higher frictional bond strength and slip- of the geopolymers in EGCs prepared with different precursors [64].
hardening coefficient than PE fibers. Also, PVA fibers exhibited a Contradicting results are reported with regards to the PSHstress and
chemical bond with the EGC matrix, whereas no chemical bond occurred PSHenery indices. One study reported that heat curing resulted in 5 % and
with PE fibers. The higher τo and Gd values were attributed to the hy­ 21 % higher PSHstress and PSHenery indices, respectively, as compared to
drophilic nature of PVA fibers. In contrast, the higher β value was ambient temperature curing [92], whereas another study reported that
attributed to the lower elastic modulus of PVA fibers which results in the PSHstress index was comparable for both curing conditions, while the
fiber damage and a jamming effect during slippage. This led to lower PSHenery index was 14–20 % higher for ambient temperature curing than
complementary energy of EGCs with PVA fibers, resulting in lower for heat curing [64]. A study by Kan et al. [43] investigated the effect of
the curing period on the PSH behavior of EGCs. It was found that the

11
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

values of τo, β, PSHstress, and PSHenery indices increase with curing age up their maximum values after 7 days of curing, beyond which their values
to a certain limit. The frictional bond strength reaches a maximum value stabilize [43].
after 56 days of curing, whereas β, PSHstress, and PSHenery indices reach

Fig. 8. Micromechanical model parameters of EGCs collected from literature including (a) τo, (b) Gd, (c) β, (d) Em, (e) Km, (f) Jtip, (g) PSHstress and (h) PSHenery.

12
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

Fig. 8. (continued).

4.1.4. Comparison with ECC [58,77,94]. Additionally, it can be seen that the average PSHstress and
Micromechanics-based investigations and the resulting micro­ PSHenergy indices for EGCs are about 1.5 times and 5 times higher,
mechanical model parameters assist in evaluating the PSH behavior of respectively, than for typical ECC M45. This agrees with results from the
engineered cementitious composites. Such investigations have been literature in which it was reported that the PSHstress index [42,80] and
conducted to assess the PSH behavior of EGCs prepared with the low- the PSHenergy index [42,80,99] of EGC are comparable to higher than for
carbon geopolymers compared to that of ECCs prepared with OPC. ECC. This is most probably a result of the weaker matrix fracture
Values of the micromechanical model parameters are collected from all toughness of EGCs, which is beneficial for satisfying the criteria for PSH
literature on EGCs and are presented in Fig. 8. Also, the maximum, behavior [75,77]. It should be noted that the higher end of the PSHenergy
minimum, and average values of each of the micromechanical model index values for EGCs is for studies by Li et al. [60] and Nguyễn et al.
parameters are determined from all studies on EGCs and are indicated by [79], which reported PSHenergy indices of 43 to 45 and 151 to 322,
straight horizontal lines in Fig. 8. For comparison purposes, values of the respectively. In these two studies, EGCs exhibiting ultra-high tensile
micromechanical model parameters of a typical ECC M45 mix are also ductility were developed through the incorporation of functionalized
shown in Fig. 8. The results show that the matrix elastic modulus and multi-walled carbon nanotubes (MWCNTs) in EGCs [60] and through
matrix fracture toughness of EGCs are about 53 % and 45 % lower, the use of a new alkaline activator based on dissolving SMP powder in
respectively, than for typical ECC M45. Similar results were reported in sodium hydroxide solution [79]. In general, it was found that the
the literature, in which it was found that EGCs exhibit lower matrix micromechanical model developed by Yang et al. [55] for ECC is
elastic modulus and lower to comparable matrix fracture toughness to applicable for evaluating the tensile performance of EGC [58,75,77,94].
ECCs [42,58,75,77,80]. However, an exception to this is a study by Khan
and Ayub [80], which showed that EGC exhibited about 33 % higher
4.2. Fresh concrete properties and mechanical performance
matrix elastic modulus than ECC. The higher brittleness of EGC matrices
compared to ECC matrices is speculated to be a result of the differences
A total of 45 studies investigated the mechanical behavior of EGCs.
in the microstructures of both composites [42]. Regarding the fiber­
The studies examined the effects of different parameters on the me­
–matrix interface properties, the data shows that the chemical bond
chanical properties, i.e., tensile stress–strain response, compressive and
strength of EGCs is around 29 % higher than for ECC M45, while the slip-
flexural strengths, and the fresh properties of EGCs, i.e., workability and
hardening coefficient is about 67 % lower for EGC than for ECC M45.
density. Values of each of these properties are collected from the liter­
Similar results were also reported in the literature in which the chemical
ature on EGCs and are presented in Fig. 9. Also, the maximum, mini­
bond of EGC was found to be 1.5 to 5 times higher than that of ECC,
mum, and average values of each of the properties are determined from
whereas the slip-hardening coefficient was found to be 2 to 3 times
all studies on EGCs and are indicated by straight horizontal lines in
lower than that of ECC [58,77,94]. Moreover, Fig. 8 shows that the
Fig. 9. For comparison purposes, values of the mechanical and fresh
average value of frictional bond strength of EGCs is about 45 % lower
properties of a typical ECC M45 mix are also shown in Fig. 9.
than for typical ECC M45. However, contradicting results are reported in
the literature regarding the difference in frictional bond strength be­
4.2.1. Effects of matrix-related parameters
tween EGCs and ECCs. Some studies reported that ECCs showed 1.5 to
Various parameters affect the properties of EGCs, and each study
2.5 times higher τo than EGCs [58,77], while another study showed that
investigated the effects of different parameters on the properties of fresh
the τo of EGCs is 2 to 3 times higher than for ECCs [94]. Nevertheless, it
and hardened EGCs. The matrix-related parameters investigated include
is evident that there are significant differences between EGCs and ECCs
precursor type, activator type, molarity of NaOH in the activator, in­
in the fiber–matrix interface properties. While there is no clear expla­
clusion of sand in the mix, sand size, sand-to-binder ratio, and water-to-
nation for these differences, it is believed that the differences in the
geopolymer solids ratio.
chemical compositions of the reaction products at the interfaces
partially contribute to this, for which further research is needed
4.2.1.1. Precursor. Replacing fly ash with more than 40 % slag in the

13
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

Fig. 9. Properties of fresh and hardened EGCs collected from literature, including (a) density, (b) flow/workability, (c) compressive strength, (d) tensile strength, (e)
tensile strain, and (f) flexural strength.

precursor was found to reduce the workability of EGC [66,70], which steel bars increased with the replacement of fly ash with slag compared
was attributed to the ball-bearing effect of fly ash particles [70], and the to the mix with no slag.
setting time dropped substantially with 20 % or more replacement of fly A study by Nematollahi et al. [64] showed no significant effect on the
ash with slag [66]. Also, it was found that replacing fly ash with slag compressive strength of EGC from two types of slags (typical slag con­
improved the tensile strength of EGCs [63,66,70]. Alrefaei and Dai [70] taining 2 % gypsum and gypsum-free slag), whereas the EGC prepared
showed that 50 % replacement of fly ash with slag increased both the with typical slag showed slightly higher tensile strength and 14–38 %
tensile strength and the tensile strain of EGC. In other studies, it was higher tensile strain than the EGC prepared with gypsum-free slag [64].
found that replacing fly ash with slag in the precursor improved the Owing to their high content of reactive silica and alumina, metakaolin
tensile strength of EGC but reduced the tensile strain capacity as and zeolite have been extensively studied for improving the properties
compared to EGC prepared with 100 % fly ash [63,66]. The enhance­ of concrete [98,99]. Some studies investigated the effects of partially
ment in the tensile strengths of EGCs with the incorporation of slag is replacing fly ash with metakaolin [98] and zeolite [99] on the me­
attributed to the formation of C–S–H gel from slag hydration and chanical properties of EGC. It was found that replacing fly ash in the
polymerization combined with the formation of C-A–S–H gel from fly precursor with a small amount of metakaolin of 2.9 % increased the
ash activation [63,66]. Various results were reported for the effect of fly tensile strength of EGC by 24 % at 7 days and 17 % at 28 days, with a
ash replacement with slag on the compressive strength of EGCs. One similar trend observed for the tensile strain [98]. Furthermore, replacing
study reported that there was no significant difference in the compres­ fly ash with 3 % and 6 % zeolite improved the tensile strength and strain
sive strengths of EGC samples prepared with either 100 % fly ash or 50 % of EGC at different curing ages compared to the EGC with no zeolite
fly ash and 50 % slag [70]. Other studies showed that replacing fly ash replacement. However, increasing the zeolite replacement to 9 % of the
with slag increased the compressive strength of EGCs as compared to the total precursor resulted in reduced tensile strength and strain of EGC
EGC prepared with 100 % fly ash [63,66]. In addition, Ling et al. [63] compared to the EGC with only fly ash and no zeolite. It was found that
showed that both the flexural strength of EGC and its bond strength with the optimum EGC mix is the one with 3 % zeolite replacement as it

14
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

Fig. 9. (continued).

showed higher tensile strength and strain and higher compressive Other materials obtained as by-products of some manufacturing
strength and strain as compared to the reference EGC with no zeolite processes were investigated for use as partial replacements of precursors
[99]. in EGCs. Kan et al. [101] studied the effects of replacing silica fume and

15
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

Fig. 9. (continued).

GGBFS in the precursor with IFA at replacement ratios of 10 % to 60 % the mechanical performance of EGC prepared with a new type of acti­
by weight of the binder. The study showed that the compressive and vator, SMP and NaOH solution, showed that increasing the SMP-to-
tensile strengths of EGC generally decrease with an increase in IFA NaOH ratio from 1 to 2 improved both the tensile and compressive
replacement ratio, with the maximum compressive and tensile strengths strengths of EGC, beyond which further increases in SMP-to-NaOH ratio
attained at replacement ratios of 10 % and 20 %, respectively [101]. reduced the strengths. The SMP and NaOH activator resulted in ultra-
Other studies investigated the effects on EGC properties of partial ductile behavior of EGC with tensile strain capacities exceeding 10 %,
replacement of the precursor with steel slag [102,103]. Even though the in which an SMP-to-NaOH ratio of 1.5 yielded the highest strain capacity
effect of steel slag replacement on the compressive strength and the of 13.7 % [79].
tensile strength and strain of EGC did not generally follow a particular The properties of EGCs are not only affected by the type of alkaline
trend [102,103], one study reported that the early compressive strength activator but also by the chemical composition of the activator. The
(7 days) of EGC increased gradually with increase in steel slag alkalinity of the activator solution affects the degree of geo­
replacement ratio from 0 % to 28 % [102]. It can be inferred that the polymerization, which in turn affects the aluminosilicate gel formed
optimum steel slag replacement ratio in EGC prepared with fly ash or and, ultimately, the properties of the geopolymer [59]. In addition, the
blends of fly ash and GGBFS as precursors range from 15 % to 21 %, with ratio of Na2SiO3/NaOH in activators composed of NaOH and Na2SiO3
further increases in replacement ratio resulting in the deterioration of solutions significantly affects the strength development of geopolymers
EGC mechanical properties [102,103]. [105]. It was found that the initial setting time of EGC increased when
the molarity of NaOH increased from 4 M to 16 M, and the Na2SiO3/
4.2.1.2. Alkaline activator. Moreover, it is evident that the type of NaOH ratio increased from 1 to 2.5 in activators composed of NaOH and
alkaline activator plays a vital role in the geopolymerization reactions Na2SiO3 solutions. Also, the final setting time of EGC was found to in­
and significantly affects the properties and strength of geopolymers crease with an increase in NaOH molarity and a decrease in Na2SiO3/
[59,106]. It was generally found that sodium-based activators result in NaOH ratio [61]. Additionally, increasing NaOH molarity was found to
higher compressive strengths, tensile strengths, and tensile strains of increase the tensile strength [61,68], the compressive strength
EGCs than lime-based or potassium-based activators [42,76,85]. Also, [61,68,88], the flexural strength [61,88], and the elastic modulus [61]
EGCs prepared with sodium-based activators were found to have higher of EGC up to a certain limit beyond which further increase in NaOH
elastic moduli [85] and flexural strengths [85,90] than EGCs prepared molarity causes the EGC properties to deteriorate. Conversely, higher
with potassium-based activators. The superior performance of EGCs tensile strain capacities were observed with lower NaOH molarity
prepared with sodium-based activators over other types of activators [61,89]. On the other hand, one study found that the compressive and
was attributed to the higher degree of geopolymerization in the sodium- tensile strengths of EGC gradually increased with an increase in NaOH
based EGC matrices, which led to higher density and lower porosity molarity from 6 M to 14 M [89]. The increase in strength with an in­
resulting in improved mechanical performance [85,90]. Also, sodium- crease in NaOH molarity was attributed to the acceleration of the geo­
based EGCs showed higher PSH parameters, owing to their lower polymerization reaction with the increase in alkalinity [68,88,89]. With
chemical bond strength with fibers, resulting in their better tensile strain the increase in Na2SiO3/NaOH ratio, the compressive strength, elastic
performance compared to EGCs prepared with other alkaline activators modulus, and tensile strength were found to decrease, while the flexural
[42,76,85]. For these reasons, and in addition to their relatively lower strength and tensile strain slightly increased up to a limit and then
costs, sodium-based activators were recommended for use in EGCs over started to decrease [61]. The optimum limit of NaOH molarity and/or
other types of activators [42,76]. It should be noted that sodium-based Na2SiO3/NaOH ratio depends on the combination of precursor type and
activators composed of both sodium hydroxide and sodium silicate so­ activator composition of the particular EGC.
lutions were found to bring about more desirable mechanical and strain-
hardening behaviors than sodium-based activators prepared with only 4.2.1.3. Fine aggregates. The content and size of sand in the mix also
sodium hydroxide solution [42]. Additionally, a study on the effects on affect the mechanical properties of EGCs. It is generally found that

16
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

adding sand to EGC mixes reduces workability [62,67,70], which to-cement ratio on the properties of cement-based composites. Studies
further decreases with an increase in the sand content [62,67]. This was showed that increases in the W/GP ratio increase the workability of EGC
attributed to increased water adsorption and increased shear resistance [62], while reductions in the W/GP ratio increase the compressive and
to flow with the addition of sand [67]. Marginal reduction in setting tensile strengths [46,69,71,73,76]. The effect on the tensile strength
time was also observed with the inclusion of sand in EGC [67]. Some could be linked to the fiber–matrix interfacial bond strength in which a
studies showed that the inclusion of sand in EGC improved the stiffness higher W/GP ratio results in lower elastic modulus of the matrix and
and compressive strength [62,70,85], which further improved with an increased porosity at the interfacial zone, which leads to lower frictional
increase in sand content [62], while other studies showed that the in­ bond strength and, in turn, lower tensile strength [71,73]. One study
clusion of sand reduced the compressive strength of EGC with greater showed that the compressive strength of EGC increased with the
strength reductions observed for higher sand contents [67]. Compared decrease in W/GP ratio up to a W/GP ratio of 0.24, beyond which the
to EGC with no sand, it was found that fine sand, with a maximum size of strength started to decrease. This was attributed to the reduced work­
212 μm, marginally improved the compressive strength of EGC, which ability at lower W/GP ratios, which caused macropores in EGC, leading
further increased with an increase in fine sand content from 30 % to 60 to an abrupt drop in the compressive strength [62]. In addition, a lower
% of precursor weight, while coarse sand, of size range between 600 μm W/GP ratio was found to result in higher tensile strain of EGC [69,73,76]
and 1.18 mm, slightly reduced the compressive strength [73]. A sug­ and higher flexural strength [62]. A study by Salami et al. [81]
gested reason for the reduction in compressive strength is the influence addressed the effects of added water versus the effects of a naphthalene-
of the additional sand in EGC on the distribution and orientation of fi­ based superplasticizer on the properties of palm oil fuel ash-based EGC.
bers and the entrapped voids in the composite, which affect crack The use of the superplasticizer was found to be minimally beneficial
initiation and propagation, which in turn affect the compressive strength since adding the same amount of water was adequate for improving
[67,73]. Furthermore, the addition of sand was found to reduce the workability while achieving higher compressive strength [81].
tensile strain capacity of EGCs [70,73,85], with the reduction being
more pronounced in EGCs prepared with coarse sands [73] and in EGCs 4.2.2. Effects of fiber-related parameters
prepared with potassium-based activators as they exhibited a strain- The properties of the fibers used in an EGC mix have a direct effect on
softening behavior instead of a strain-hardening one [85]. This was the properties of the composite as they dictate the fiber–matrix inter­
attributed to the increase in matrix fracture toughness due to sand facial properties and affect the PSH parameters of the EGC. The fiber-
addition [70,73]. These results agree well with findings on conventional related parameters include fiber type, content, and geometry.
ECCs, which advised that the sand content in the matrix has to be limited
to ensure the tensile ductility of the composites [107,108]. Contra­ 4.2.2.1. Fiber content and aspect ratio. Among the different fibers
dicting results are reported in the literature on the effect of sand addition investigated for use in EGCs, PVA fibers were the most commonly
on the tensile strength of EGCs, in which one study revealed that sand studied because of the reasons previously mentioned. In ECCs, a low
addition increased the tensile strength of EGC [70], while another study fiber volume fraction of up to 2 % is usually used to allow the occurrence
showed that it reduced the tensile strength [85]. Additionally, the effect of strain-hardening. A few studies have been published on the effect of
on tensile strength was found to depend on the sand size, in which fine PVA fiber content on the properties of EGCs. Studies showed that
sand increased the tensile strength, which further increased with an increasing PVA fiber volume fraction in the range of 1 % to 2 % results in
increase in fine sand content from 30 % to 60 % of precursor weight, up to 22 % reduction in flowability [71,93,96] and delays in initial and
whereas coarse sand reduced the tensile strength [73]. Moreover, it was final setting times by 15 % and 8 %, respectively [93]. Various results
found that the flexural strength of EGC increases with the increase of are reported regarding the effect of PVA fiber content on compressive
sand content [62,67], with one study reporting that the flexural strength strength. Some studies reported that increasing PVA fiber volume frac­
starts decreasing at sand-to-binder ratios exceeding 0.2 [67]. tion in the range of 1 % to 2 % led to around 7 % to 22 % reduction in
In one study by Nematollahi et al. [74], the effects on the properties compressive strengths [93,96], while other studies reported negligible
of EGCs of using lightweight aggregates were investigated. Three types to up to 7.5 % increase in compressive strengths [65,71]. The reduction
of lightweight aggregates, including expanded glass aggregates, white in compressive strength with an increase in PVA fiber content is thought
hollow ceramic microspheres, and expanded perlite, were studied as full to result from increased entrapped air during mixing, which results in
replacement of micro silica sand at the same volume percentage. The increased porosity [93,96]. Also, Al-Majidi et al. [65] showed that
results showed that the EGCs prepared with expanded glass, ceramic increasing PVA fiber content did not have a significant effect on the
microspheres, and expanded perlite showed 24 %, 18 %, and 15 % lower flexural strength of EGC, whereas Wang et al. [93] reported that
compressive strength, respectively, than the EGC prepared with micro increasing PVA fiber content from 1.5 % to 2 % resulted in about 40 %
silica sand. This was attributed to the generally weaker strength of increase in flexural strength of EGC. However, both studies reported that
lightweight aggregates compared to micro silica sand particles. increasing PVA fiber content significantly improved the flexural
Although EGCs prepared with the lightweight aggregates showed rela­ deflection capacity of EGC by 1.35 to 2.5 times [65,93]. Moreover,
tively high ultimate tensile strengths in the range of 3.4 to 5 MPa, their increasing the PVA fiber volume fraction in the range of 1 % to 2 % was
tensile strengths values were lower than those of EGCs prepared with found to increase the tensile strength of EGC by about 30 % to 43 %
micro silica sand by 24 %, 32 %, and 14 % for expanded glass, ceramic [65,71,96] and the tensile strain by up to 1.4 times [71,96].
microspheres and expanded perlite, respectively. On the other hand, the Zahid et al. [68] studied the effects of different PVA fiber aspect
tensile strain capacities of EGCs prepared with the lightweight aggre­ ratios, including fine PVA fibers of 0.04 mm diameter and 8 mm or 12
gates were comparable to those of the EGCs prepared with micro silica mm lengths and coarse PVA fibers of 0.2 mm diameter and 18 mm or 24
sand, in which the tensile strains attained were in the range of 3.5 to 3.7 mm lengths, on the properties of EGCs at a constant fiber volume frac­
%. This was attributed to the comparable values of stress-performance tion of 2 %. It was found that coarse PVA fibers resulted in slightly
indices obtained for all the EGCs, which were in the range of 1.3 to higher compressive strength than fine PVA fibers. This was attributed to
1.5 [74]. the larger number of fine fibers per unit volume of the matrix, which
resulted in the lumping of the fibers and the introduction of pores at the
4.2.1.4. Water-to-geopolymer solids ratio. The ratio of water-to- fiber–matrix interface. However, EGC containing fine PVA fibers
geopolymer solids is another major parameter that affects the proper­ exhibited higher tensile strengths and strains than EGC containing
ties of EGCs. The effects of the ratio of the water-to-geopolymer solids on coarse PVA fibers, which showed less pronounced strain-hardening
the properties of EGCs are somewhat similar to the effects of the water- behavior. The lower number of coarse fibers caused the fibers to

17
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

disperse with a wider gap within the matrix and, thus, could not arrest an increase in temperature up to 200˚C as compared to PVA fibers owing
all cracks leading to lower tensile strength and strain [68]. Al-Majidi to the lower melting point of polyethylene fibers [90]. In addition, a
et al. [65] investigated the effects of different aspect ratios of micro study compared the mechanical properties of EGCs prepared with PVA,
steel fibers on the properties of EGCs. The results showed that incor­ PP, and PET fibers and showed that PVA fibers yielded the highest
porating steel fibers in EGCs improved the compressive strength by mechanical properties, followed by PET and then PP fibers. PVA fibers
around 15 to 25 MPa. The compressive strength increased with the in­ resulted in higher tensile strengths, tensile strains, compressive
crease of fiber content of micro steel fibers with 13 mm length and 0.16 strengths, and flexural strengths by about 47 %, 25 %, 5 %, and 6 %,
mm diameter, whereas the opposite trend was observed with micro steel respectively, as compared to PET fibers, and by around 83 %, 150 %, 7
fibers of 6 mm length and 0.16 mm diameter. This was attributed to the %, and 10 %, respectively, as compared to PP fibers [80].
higher number of the shorter fibers at high volume fractions which One study explored the effects of using natural fibers (jute, sisal, and
caused redistribution of the voids and weakening of the fiber–matrix curauá) as opposed to synthetic PVA fibers in EGCs [83]. The results
interface under compression. Moreover, increasing steel fiber content showed that using the natural fibers of jute, sisal, and curauá at 10 %
and aspect ratios significantly improved tensile strength, flexural weight content increased the tensile strength of EGC by about 2.5 times,
strength, and flexural deflection [65]. 2 times, and 1.1 times, respectively, compared to PVA fibers at 2 %
Very limited research is currently available on the effect of PE fiber volume fraction. However, the tensile strain values of EGCs prepared
content on the properties of EGCs. It was generally found that increasing with jute, sisal, and curauá fibers were more than 1.2 times, 3.5 times,
PE fiber content improves the tensile strength and strain of EGCs and 5 times lower, respectively, than those prepared with the PVA fibers
[72,95]. While one study showed that increasing the PE fibers volume [83].
fraction from 1.5 % to 1.75 % increased the tensile strength and strain by
around 15 % and 6 %, respectively [72], another study showed that the 4.2.2.3. Fiber hybridization. Various types and aspect ratios of fibers
tensile strength and strain increased up to a PE fiber content of 1 %, have been studied for use in EGCs, yielding different composite prop­
beyond which the tensile strength and strain started to decrease [95]. erties. The resulting composite properties depend on the elastic modulus
Furthermore, the addition of a 1 % volume fraction of PE fibers to EGC and strength of the fibers used. In addition, some fibers are more costly
increased the flexural strength by 6 times as compared to without fibers than others. To maximize the advantages of the different fibers, fiber
[87], and increasing PE fiber content from 0.5 % to 1.5 % improved the hybridization has been investigated for EGCs. Wang et al. [93] showed
flexural strength of EGC by up to 70 % [95]. On the contrary, various that using a hybrid combination of PVA and recycled tyre steel (RTS)
results are reported regarding the effect of PE fiber content on the resulted in reduced flowability compared to EGC, with only PVA fibers
compressive strength of EGC. Some studies showed that the addition of at the same total fiber volume fraction of 2 %. This was attributed to the
PE fibers or increase in their content in EGC had no significant effect on non-uniform geometry of RTS fibers which increased shear resistance.
compressive strength [72,87]. Another study showed that adding PE The hybrid combination also reduced setting time and resulted in about
fibers reduced the compressive strength by about 20 % to 50 % 14 % to 25 % higher compressive strength of EGC at 28 days as
compared to without fibers while increasing PE fiber content from 0.5 % compared to EGC with only PVA fibers at the same total fiber content.
to 1.5 % improved the compressive strength by up to 58 % [95]. In This was attributed to the high stiffness and hydrophilic nature of RTS
addition, a study by Akbar et al. [104] showed that increasing PP fiber fibers. On the contrary, the hybrid fiber EGC showed about 20 % lower
volume fraction in EGC from 1 % to 3 % slightly reduced the compres­ flexural strength than EGC with only PVA fibers since the fiber bridging
sive strength by up to 6 %, while the tensile strength was increased by up effect was mainly provided by the PVA fibers because the RTS fibers
to 20 % at different curing ages. On the other hand, the flexural strength showed irregular geometry caused by the recycling process [93].
slightly increased by up to 2.8 % when PP fiber content increased from 1 Another study investigated the behavior of EGC prepared with various
% to 2 %, and then slightly decreased by up to 3.3 % with a further hybrid combinations of PVA and RTP fibers at a total volume fraction of
increase in PP fiber content to 3 %. This was attributed to the poor fiber 2 %. The results showed that increasing RTP fibers content from 0.25 %
dispersion resulting from the reduced workability at a higher fiber to 1 % reduced flowability, tensile strength, and tensile strain by up to
dosage [104]. 15 %, 47 %, and 99 %, respectively. In addition, the compressive
strength decreased by up to 8.5 % when RTP fibers content increased
4.2.2.2. Fiber type. Comparing the performance of EGCs prepared with from 0.25 % to 0.75 % but increased for RTP fiber content by 1 %. The
different types of fibers, Alrefaei and Dai [70] showed that steel fibers tensile strength and strain of EGC prepared with 2 % PVA fibers were
resulted in 2.5 times higher workability, about 20 % higher compressive higher than for the hybrid EGC by more than 10 % and 33 %, respec­
strength, and around 9 % to 26 % higher tensile strength than PE fibers tively, while the compressive strength was lower by up to 9 %, except for
at the same fiber volume fraction of 2 %. On the other hand, PE fibers EGC with 1.25 % PVA and 0.75 % RTP which showed 4 % lower
resulted in about 3.5 times higher tensile strain of EGC than steel fibers compressive strength than EGC with 2 % PVA fibers. The inferior me­
at the same fiber content [70]. Another study showed that steel fibers at chanical properties observed with the increase in RTP fibers content
a 3 % volume fraction resulted in about 35 % higher compressive and were attributed to the lower stiffness and lower bridging stress of RTP
flexural strengths than PVA fibers at a 2 % volume fraction [91]. fibers and their weaker fiber–matrix interaction compared to PVA fibers
Moreover, PE fibers were found to result in around 16 % higher tensile [96]. In another study, fiber hybridization of 0.25 % PVA and 1.5 % PE
strains but slightly lower tensile and compressive strengths by about 8 % by volume had no significant effect on the compressive strength of EGC
to 9 % with EGCs as compared to PVA fibers [92]. Conversely, another as compared to 1.5 % or 1.75 % volume fraction of only PE fibers. This
study found that PE fibers resulted in higher flexural and tensile combination of hybrid fibers yielded similar EGC tensile strength
strengths than PVA fibers by about 23 % and 5 %, respectively, while compared to a 1.5 % volume fraction of PE fibers owing to the lower
comparable tensile strains were reported for both fiber types. The same tensile strength and chemical bonding of PVA fibers. However, the
study reported that PBO fibers resulted in up to 30 % and 14 % higher hybrid fiber composite showed a 6.5 % higher tensile strain capacity
flexural and tensile strengths, respectively, than both PE and PVA fibers than EGC with a 1.5 % volume fraction of PE fibers and comparable
but resulted in about 28 % lower tensile strains. This was attributed to tensile strain to EGC with 1.75 % PE fibers [72]. A study by Li et al.
the superior mechanical properties of PBO fibers in comparison to both [100] showed that EGC prepared with different hybrid combinations of
PVA and PE fibers [86]. In addition, UHMWPE fibers resulted in about 0.05 % to 0.15 % MWCNTs and 0.15 % to 0.25 % PVA fibers, by volume,
20 % to 30 % higher flexural strength of EGC than PVA fibers. However, achieved 35 to 38 MPa compressive strengths and 6 to 8 MPa flexural
UHMWPE fibers showed greater degradation in flexural properties with strengths at 28 days. The optimum hybrid combination was found to be

18
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

0.1 % MWCNTs and 0.2 % PVA, which attained compressive and flexural results showed that the mechanical properties of the EGC increased
strengths of 38.43 MPa and 7.75 MPa, respectively [100]. gradually up to a certain age beyond which they stabilized. The
Al-Majidi et al. [65] showed that EGC prepared with a hybrid of 1 % compressive strength reached a maximum value of about 100 MPa at 28
macro and 1 % micro steel fibers showed about 14 % higher compressive days, while the tensile strength and tensile strain increased to 6 MPa and
strength than EGC with 2 % micro steel fibers. Conversely, EGC incor­ 6.1 %, respectively, at 56 days [43]. Another study showed that with an
porating 1 % macro and 2 % micro steel fibers exhibited around 11 % increase in curing age of 7 days to 28 days at ambient temperature, the
lower compressive strength than EGC with 3 % micro steel fibers. compressive and tensile strengths of EGC increased by 54 % and 35 %,
However, both hybrid combinations resulted in higher compressive respectively, while the tensile strain decreased by 8 % [102].
strength than EGC with only 1 % macro steel fibers. Additionally, EGCs A study by Farooq et al. [66] examined the effects of 70 days of air
incorporating hybrid micro and macro steel fibers at total fiber volume curing and water curing at ambient temperature on the properties of
fractions of 2 % and 3 % showed lower flexural strengths and deflections EGCs. It was found that air curing resulted in marginally higher
than EGCs with only micro steel fibers at the same total fiber contents. compressive strength of EGCs than water curing. This was ascribed to
Still, they showed improved flexural response over EGCs with only the generally slow reaction kinetics of fly ash geopolymers and the
macro steel fibers at a 1 % volume fraction [65]. Alrefaei and Dai [70] leaching out of unreacted alkaline activator into the water leading to
investigated the effect on the properties of EGC by using a hybrid of incomplete geopolymerization. Conversely, water-cured EGCs incorpo­
straight copper-coated steel fibers and UHMWPE fibers at different rating both fly ash and slag showed higher first cracking strength and
dosages of each fiber type while maintaining a constant total fiber vol­ ultimate tensile strength than their air-cured counterparts, except for the
ume fraction of 2 %. It was found that increasing the UHMWPE fiber EGC with 100 % fly ash, in which air curing resulted in marginally
volume fraction in EGCs with hybrid fibers reduced workability. Addi­ higher tensile strength than water curing. The lower first cracking and
tionally, increasing the steel fiber volume fraction increased the EGC tensile strengths with air curing were attributed to the occurrence of
compressive strength while reducing the steel fiber content and early-age drying shrinkage cracks in the air-cured specimens leading to a
increasing the UHMWPE fiber content made the compression failure larger distribution of flaws [66].
mode more ductile. In general, it was found that increasing the steel
fiber content while reducing the UHMWPE fiber content increased the 4.2.4. Comparison with ECC
first cracking strength and the ultimate tensile strength of EGC. This was A few studies compared the properties of their developed EGCs with
attributed to the higher elastic modulus of steel fibers compared to PE the properties of ECC. Some of those studies compared with ECC prop­
fibers. However, this trend also depended on the type of precursor used, erties reported in the literature, while others prepared ECC mixes in
in which the composite with only PE fibers showed higher tensile their studies for comparison. It was found that EGCs generally exhibit
strength than the hybrid composites when a combination of fly ash and comparable to higher tensile strains than ECCs [41,42,46,69,85,95,96],
slag was used for the precursor. It was also found that increasing the PE and comparable to lower tensile strengths [41,80,85,95]. However,
fiber content while reducing the steel fiber content increased the tensile some studies showed that the tensile strength and strain of EGC can
strain capacity of EGC [70]. depend on the activator type and molarity [42,89]. Nematollahi et al.
(b) [42] showed that EGCs prepared with sodium-based activators
4.2.3. Effects of curing conditions showed higher tensile strengths than ECC. In contrast, EGCs prepared
Owing to the requirement for heat curing with fly ash, some studies with potassium-based and lime-based activators showed lower tensile
investigated the potential of utilizing ambient temperature curing for fly strengths than ECC [42]. Another study showed that EGC exhibited
ash-based EGCs and examined the effects of different curing conditions about 17 % to 31 % lower tensile strength and up to 67 % higher tensile
on the properties of EGCs [41,64,66,92,98,102]. Studies showed that strain than ECC at low NaOH molarities of 6 M to 8 M. The opposite
heat curing improved the tensile strength and strain of EGCs incorpo­ trend was observed for higher NaOH molarities of 10 M to 14 M, in
rating fly ash [41,98]. Also, it was found that heat curing improved the which EGC showed about 6 % to 30 % higher tensile strength and about
compressive strength of EGC [41,102], which was further improved 33 % to 60 % lower tensile strain than ECC [89]. One study found that
with an increase in heat curing duration from 4 h to 8 h [41]. Studies by EGC prepared with GGBFS as a precursor, calcium hydroxide and so­
Nematollahi et al. [64,92] investigated the properties of EGCs prepared dium sulfate as the alkali activator, and 1.75 % volume content of PE
with 50 % replacement of fly ash with slag and cured under different fibers exhibited high ultimate tensile strength and strain values of 13
conditions of either ambient temperature curing in water for 28 days or MPa and 7.5 %, respectively. This tensile strength value is reported to be
heat curing at 60˚C for 24 h followed by ambient temperature curing in more than double that of ECC prepared with PE fibers and almost three
the air for 3 days. The 3 days curing period for the heat curing condition times that of ECC prepared with PVA fibers [69].
adopted in these studies was justified by previous studies, which indi­ In addition, the compressive strength of ECC was found to range from
cated that once the heat curing period is complete, curing age would about 7 % to up to 4 times higher than the compressive strength of EGC
have no significant effect on the strength of geopolymers [64,92]. The [42,89,95]. Despite this, EGC exhibited superior compressive strength to
studies revealed no significant effect on the compressive strength of ECC in some cases. For instance, studies showed that EGCs prepared
EGCs from both curing conditions, with the ambient temperature curing with alkaline activator solutions comprising sodium hydroxide and so­
condition resulting in slightly higher compressive strength of EGCs in dium silicate showed about 6 % higher compressive strengths than ECC
some cases [64,92]. Additionally, it was found that the ambient tem­ at a NaOH molarity of 8 M [42,80] and comparable to up to 12 % higher
perature curing condition resulted in 15–33 % higher first crack strength compressive strengths than ECC at NaOH molarity of 10 M to 14 M [88].
of EGCs compared to the counterpart composites subjected to heat Also, EGC prepared with a 1.5 % volume fraction of PE fibers showed
curing [64,92]. On the other hand, contradicting results were reported about 36 % higher compressive strength than its counterpart ECC [95].
for the tensile response of EGCs. Some studies showed that ambient On the other hand, the flexural strength of EGC seems to be highly
temperature curing improved the tensile strain capacity of EGCs by affected by the alkaline activator type and molarity. For EGCs prepared
around 17 % to 76 %. In comparison, a negligible difference of up to 8 % with alkaline activator solutions of sodium silicate and sodium hy­
reduction was obtained in ultimate tensile strength [54,92]. In contrast, droxide, one study showed that flexural strengths were lower than for
another study found that ambient temperature curing reduced the ten­ ECC by around 30 % to 50 % at NaOH molarity of 8 M [95], while
sile strain of EGC by about 11 % and improved the ultimate tensile another study showed that EGCs attained about 31 % lower flexural
strength by about 27 % [92]. The effects of curing age on the strengths than ECC at NaOH molarity of 6 M, but comparable to up to 50
compressive and tensile responses of EGC prepared with GGBFS and fly % higher flexural strengths at NaOH molarities of 8 M to 14 M [88].
ash as precursors and PE fibers were investigated in Kan et al. [43]. The Another study showed that the flexural strength of EGC prepared with a

19
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

sodium-based activator was close to that of ECC, while it was lower for specimens showed a significant reduction in flexural strength after
EGC prepared with a potassium-based activator [85]. Furthermore, the exposure. The strength degradation with ECC was attributed to the loose
flexural strength of EGC apparently depends on fiber type, in which EGC PVA fibers, which was not the case with EGC. Moreover, some samples
prepared with PVA fibers had about 16 % lower flexural strength than its were preloaded prior to exposure and then tested in bending. The results
counterpart ECC. In contrast, comparable flexural strengths were ob­ showed that the combination of preloading and acid exposure resulted
tained for EGC and ECC prepared with PP and polyethylene tere­ in significant degradation in the bending stiffness of EGC specimens,
phthalate (PET) fibers [80]. while ECC samples showed much less degradation [110].
As can be seen from Fig. 9, the density of EGCs is comparable to the In another study by Salami et al. [82], the durability of EGC prepared
density of ECC M45. Both densities are lower than the typical density of with palm oil fuel ash (POFA) was assessed by measuring the residual
normal weight concrete (2500 kg/m3) but higher than the maximum compressive strength and weight of specimens exposed for nine months
density required to be specified as lightweight concrete according to the to three different sulfate solutions, including 5 % Na2SO4, 5 % MgSO4,
definition of ACI Committee 213 [109]. The data also shows that the and 2.5 % Na2SO4 + 2.5 % MgSO4. The results revealed that the EGC
tensile and compressive strengths of typical ECC M45 are higher than specimens exposed to 5 % Na2SO4 exhibited the largest loss in
the average values of EGCs by 21 % and 39 %, respectively. However, compressive strength and weight, followed by the specimens exposed to
the average tensile strain of EGCs is 41 % higher than the tensile strain of 2.5 % Na2SO4 + 2.5 % MgSO4, whereas the lowest loss in strength and
typical ECC M45. The results of this analysis are in agreement with weight was with the samples exposed to 5 % MgSO4. The study also
conclusions made in the literature that EGCs generally exhibit lower investigated the effect of different NaOH molarities of the alkaline
compressive strengths [42,89,95] and comparable to lower tensile activator on the sulfate resistance of EGC. It was found that the loss in
strengths than ECCs [41,80,85,95], but show comparable to higher compressive strength and weight increased with the increase in NaOH
tensile strain capacities [41,42,46,69,85,95,96]. molarity from 10 M to 14 M. This was ascribed to the increase in the
amount of aluminosilicate gel as the molarity increases, which increases
4.3. Durability the composite susceptibility to sulfate attack due to depolymerization.
At the end of the exposure period, the maximum and minimum per­
Since research on engineered geopolymer composites is still in its centage weight losses recorded for all specimens were 13 % and 6 %,
infancy, most of the published studies on the topic have only focused on respectively [82].
the mechanical properties and the PSH behavior of EGCs. Nevertheless,
durability is a very important aspect that needs to be thoroughly 4.4. Self-healing
investigated before the practical application of EGCs is possible.
Currently, very limited studies are published on the durability of EGCs. Self-healing is a term that describes the ability of materials to
Owing to the high content of cement and the low content of fine ag­ perform self-repair without any external interference. The development
gregates, ECCs typically exhibit very high drying shrinkage [96]. This of concrete with self-healing abilities will bring about obvious advan­
problem is more pronounced in EGCs since geopolymer matrices tages to the construction industry, such as reductions in maintenance
generally exhibit higher shrinkage than cementitious matrices [91,96]. costs and improvements in the durability and sustainability of concrete
Studies showed that increasing the fiber content of EGCs resulted in structures. Self-healing concrete is currently an active area of research,
reduced drying shrinkage at all curing ages [91,96]. This was ascribed to and several self-healing methods have been proposed by researchers
the physical restraint provided by the fibers in the geopolymer matrix over the years, such as bacteria-based biological self-healing, use of
[91]. In addition, increasing RTP fibers content in EGC prepared with expansive agents, chemical encapsulation, etc. [111]. ECCs have been
hybrid PVA and RTP fibers resulted in larger reductions in drying reported to possess a high potential for self-healing due to their low
shrinkage as compared to EGC prepared with only PVA fibers. This was water-to-cement ratio, which results in the occurrence of unhydrated
attributed to the release of entrapped liquid from the surfaces of RTP cement grains in the composite. When cracks form, external water ac­
fibers and the attached rubber particles. Despite this, the drying cesses the concrete and reacts with the unhydrated cement allowing
shrinkage of EGC was still higher than that of conventional ECC [96]. In further hydration and formation of C–S–H binders which fill in the
contrast, one study reported that increasing PVA fiber volume fraction open cracks. Carbonation with carbon dioxide in the atmosphere also
from 1.5 % to 2 % resulted in about a 52 % increase in drying shrinkage, allows for self-healing through the formation of calcium carbonate in
which was attributed to the weakened fiber–matrix interface due to the concrete [13,111]. Unlike ECCs, for which several studies have been
oil coating of the PVA fibers. However, fiber hybridization of PVA and published on their self-healing capability, very limited studies are
RTS reduced drying shrinkage with increasing RTS replacement content available on the self-healing of EGCs.
resulting in even more evident reductions in drying shrinkage, which In geopolymer materials, self-healing can be achieved through
was attributed to the high elastic modulus of RTS fibers and their strong further geopolymerization reactions of unreacted precursor particles in
fiber–matrix interaction [93]. Furthermore, drying shrinkage was found the composite [97]. A study by Kan et al. [97] showed that EGC samples
to decrease with the addition of sand [62,67], with further reductions exhibited good strain-hardening behavior after self-healing and ach­
attained with an increase in sand content [62]. Also, increases in the W/ ieved more than 3.2 % tensile strain capacity. In addition, the ultimate
GP ratio were found to increase drying shrinkage [62]. tensile strengths of most of the self-healed EGC specimens exceeded
A study by Ohno and Li [110] compared the durability of EGC and those of the control specimens, which were not preloaded. This was
ECC M45 under sulfuric acid exposure. Cycles of 6-day immersion in attributed to the improved fiber–matrix interfacial bonding due to
acid followed by 1-day immersion in water were repeated 15 times. further geopolymerization reactions in the self-healed specimens. Also,
Durability was then evaluated by weight loss measurements and residual after self-healing, the total number of cracks significantly decreased, and
compressive and flexural strengths after exposure. The results showed the maximum crack width was below 25 μm for all tested samples. It was
that after 15 cycles of exposure, the weight loss of ECC was more than concluded that laboratory air exposure could be more favorable for the
twice that of EGC. Also, it was observed that PVA fibers on the surface of occurrence of self-healing in the studied EGC system than exposure to
ECC specimens were loose and easily scratched off after the exposure wet-dry cycles [97]. However, contradicting results were reported in a
period. However, this phenomenon was much less prominent with EGCs, study by Ohno et al. [78], in which water curing resulted in improved
indicating a more durable fiber–matrix interface for EGCs. After the acid self-healing capabilities of EGC over air curing, as evidenced by the
exposure, no significant degradation in compressive strength was formation of more self-healing products and the higher residual tensile
observed for both EGC and ECC specimens. Moreover, there was no stiffness with water curing [78]. It was speculated that the conclusions
significant flexural strength degradation for EGC, whereas the ECC reached by Kan et al. [97] on the more improved self-healing with air

20
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

curing could be a result of using metakaolin in EGC, which allowed for 4.5. Environmental impact
long-term pozzolanic reactions and contributed to self-healing [78].
Through chemical and microstructural analyses, studies showed that the Although the production of geopolymers has been proven to have a
self-healing products of EGC were primarily amorphous aluminosilicate much lower carbon footprint and energy consumption than the pro­
phases [78,97,102]. Conversely, one study showed that the self-healing duction of OPC [36,37], the overall environmental impact of EGCs
products in EGCs could possibly depend on the type of fibers used, in compared to ECCs needs to be further investigated. Despite the lack of
which the healing products with PE fibers were CaCO3 and a small life cycle assessment studies on EGCs, a few studies have investigated the
amount of C-A–S–H, whereas the dominant healing product with environmental impact of EGCs as determined by material sustainability
hybrid PE-PVA fibers was CaCO3. It was also found that EGCs prepared indices (MSIs) in terms of embodied energy and CO2 emissions
with hybrid PE-PVA fibers at a total volume fraction of 2 % exhibited [64,74,77,96]. In these studies, values of embodied energy and CO2
superior self-healing performance over EGCs prepared with 1.5 % or emissions for the composites mix ingredients were collected from life
1.75 % volume fraction of PE fibers. This was evidenced by the higher cycle inventory data available in the literature. Using the collected data
post-healing crack width reduction, the higher stiffness recovery, and and the mix proportions of the composites, the total embodied energy
the more densely packed healing products of the former. The PE fibers and CO2 emissions were then calculated for each composite and are
content was also found to affect the self-healing performance of EGCs in shown in Fig. 10 and Fig. 11, respectively. It should be noted that results
which the higher volume fraction of 1.75 % resulted in a larger post- from the study by Zhong and Zhang [96] are not added to Fig. 10 and
healing recovery in tensile properties of EGC as compared to the 1.5 Fig. 11 since not enough data is reported in the study to allow for the
% volume fraction [72]. calculation of the embodied energy and CO2 emissions of each com­
Another study by Guo et al. [112] assessed the effects of curing age posite mix ingredient. The results showed that the embodied energy of
on the self-healing ability of EGC using a new non-destructive evaluation heat-cured EGCs is up to 17 % lower than that of typical ECC M45
method based on ultrasonic pulse velocity (UPV). Recovery of tensile [64,74,77]. However, the embodied energy of an ambient temperature
properties and residual crack characteristics were also evaluated to cured EGC developed by Nematollahi et al. [64] was found to be about
verify the results from UPV testing. The results revealed that the samples 35 % lower than that of typical ECC M45. This is due to the relatively
cured for 7 days showed the best self-healing performance in terms of large energy consumption of heat curing, which accounts for about 3 %
the crack characteristics and UPV results. The provided reason for this is to 13 % of the total embodied energy of EGCs, as can be seen in Fig. 10
that 3 days is a too short curing age for the advantages of fibers to be [64]. Also, the CO2 emissions of EGC were found to be significantly
realized which leads to larger crack widths, while at 14 and 28 days, lower, by around 50 % to 76 %, than those of typical ECC M45
fewer unreacted particles would be present, reducing the potential for [64,74,77]. Similar results were reported in Zhong and Zhang [96], in
self-healing [112]. Similar to the results reported in Kan et al. [97], it which the CO2 emissions of EGCs were at least 47 % lower than those of
was found that the ultimate tensile strengths of the self-healed EGC typical ECC M45. Ohno and Li [77] also compared the MSIs of EGC with
samples cured for 3 and 7 days exceeded those of the control samples, those of a greener version of ECC M45, high volume fly ash (HVFA) ECC.
which were not preloaded [112]. Also, test results from the recovered Although the embodied energy of HVFA ECC was about 9 % to 21 %
tensile properties and residual cracks matched the results from UPV lower than for heat-cured EGCs, the CO2 emissions of heat-cured EGCs
testing verifying the effectiveness of the UPV method for evaluating the were significantly lower, by around 25 % to 30 % [77]. The much higher
self-healing performance of EGCs [112]. Overall, the limited studies CO2 emissions of ECCs are associated with OPC, which is responsible for
available showed that EGCs have good self-healing capabilities about 34 % to 48 % of the total CO2 emissions of ECCs, as shown in
[72,78,97,102,112]. Yet, self-healing in EGC was found to be limited Fig. 11. In general, it can be seen that the overall environmental impact
compared to ECC due to the limited formation of healing products in of EGCs is much lower compared to ECCs [64,74,77,96].
EGC, which are also of lower tensile strength than those in ECC [78]. The results also showed that the largest contributors to energy con­
sumption in EGCs are PVA fibers, followed by alkaline activators and
heat curing, in which their associated embodied energies ranged from

Fig. 10. Embodied energy of different ECC and EGC composites reproduced from [66,74,77].

21
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

Fig. 11. CO2 emission of different ECC and EGC composites reproduced from [66,74,77].

38 % to 72 %, from 28 % to 44 %, and from 3 % to 13 %, respectively, of geopolymerization of unreacted metakaolin particles in EGCs under
the total embodied energy of EGCs. Moreover, the largest contributors to water vapor at higher temperatures [88]. The impact resistance of EGCs
CO2 emissions in EGCs ranked in the order of alkaline activators, PVA was found to improve with an increase in NaOH molarity under different
fibers, and heat curing, in which their respective CO2 emissions ranged impact drop heights and temperatures, with a recommended threshold
from 38 % to 56 %, from 22 % to 61 %, and from 10 % to 14 % of the molarity of 10 to 12 mol/L [88,89].
total CO2 emissions of EGCs [64,74,77]. Similar results were reported in Al-Majidi et al. [91] studied the use of EGC for strengthening plain
Zhong and Zhang [96], in which PVA fibers contributed about 27 % of concrete prisms. The results showed that concrete prisms strengthened
total CO2 emissions, about 42 % of total embodied energy, and around with EGCs reinforced with PVA and steel fibers showed 19 and 7 times,
70 % of the total material cost of EGCs. It was found that reducing PVA respectively, higher deflections at peak load than the non-strengthened
fiber content by 0.5 % within the range from 2 % to 1 % reduced both the control prisms. The strengthened prisms also exhibited about 3 times
total embodied energy and the CO2 emissions of EGC by up to 11 % and higher capacity under flexural load than the control prisms [91].
7 %, respectively [77,96]. Furthermore, ambient temperature curing for Furthermore, Pang et al. [87] studied the use of asphalt emulsion (AE), a
EGCs was found to reduce both the total embodied energy and CO2 self-cleaning material used in pavement construction, to improve the
emissions by around 22 % and 49 %, respectively, compared to heat rheological performance of EGC. It was found that improved rheological
curing [64]. properties, workability, and ductility were attained when AE was added
to EGC, with 8 % AE content by mass yielding the optimum performance
of EGCs in terms of rheological properties and mechanical behavior
4.6. Other aspects of EGCs
[87]. Additionally, a study by Yaswanth et al. [113] explored the
feasibility of utilizing soft computing techniques for predicting the
Most published studies on EGCs investigated their PSH behavior
compressive strength of EGCs based on a database collected from the
and/or mechanical properties, with a few studies on their durability,
literature. The study showed that Artificial Neural Networks and
self-healing, and environmental impact. Yet, some studies investigated
Response Surface Methodology models were effective in predicting the
other miscellaneous aspects of EGCs. Trindade et al. [84] studied the
compressive strength of EGC with accuracies of 93 % and 95 %,
behavior of EGCs under impact tensile loading and showed that EGCs
respectively [113].
exhibited improved mechanical performance, except for strain capacity,
under dynamic loading as compared to static loading. EGCs prepared
5. Conclusions
with PVA fibers yielded static and dynamic strain capacities of 2.5 % and
1.8 %, respectively, whereas EGCs incorporating PE fibers yielded static
This paper provided a state-of-the-art review of engineered geo­
and dynamic strain capacities of 3.6 % and 2.6 %, respectively. Under
polymer composites, which are promising low-carbon ultra-high-per­
dynamic loading, PE fibers resulted in higher bond strength, mechanical
formance concretes. The conducted scientometric analysis and extensive
performance, deformations, and energy dissipation than PVA fibers,
literature review revealed that research publications on EGCs started in
showing that PE fibers are more suitable for impact-resistance applica­
2012 and increased during the past five years, reaching 55 publications
tions of EGCs. Furthermore, EGCs were found to exhibit superior per­
up to the end of 2021. Analyses and reviews of these 55 publications
formance over ECCs under dynamic tensile loading [84]. In other
yielded the following conclusions about EGCs:
studies, the impact behaviors of EGCs were investigated under different
impact drop heights [89] and different elevated temperature conditions
• The scientometric analysis showed that Construction and Building
[88]. It was found that EGCs exhibited comparable to higher impact
Materials is the most influential journal in the field of EGCs, and some
resistance than ECC under different impact drop heights, which was
of the most productive authors in this area include Nematollahi B.,
ascribed to the stronger bond between PVA fibers and geopolymer ma­
Sanjayan J., Li V., and Lee B. Countries with the largest number of
trix, allowing for a more effective transfer of impact load between them
publications were found to include China and the United States,
[89]. Moreover, EGCs exhibited better impact behaviors than ECCs at
followed by Australia, Denmark, and Brazil. Also, the most
high temperatures of 100˚C to 150˚C, whereas ECCs showed better impact
commonly occurring terms in relation to the mix design of EGCs
behaviors at a lower temperature of 50˚C. This was ascribed to the further

22
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

include “fly ash,” “slags,” and “PVA fibers,” while keywords such as • Most studies have focused on EGCs reinforced with micro PVA fibers,
“tensile strength,” “tensile strain,” “compressive strength,” and with very limited studies on other types of fibers. More studies are
“micromechanics,” appeared the most in relation to experimental needed on the performance of EGCs reinforced with different types of
tests on EGCs. fibers, such as PE, PP, steel, glass, basalt, or natural fibers, different
• The most commonly used materials in EGC mixes include fly ash and fibers geometries and contents, as well as various fiber hybridiza­
GGBFS for the precursors, solutions of sodium silicate and sodium tions to assess the full spectrum of fibers and to identify the optimum
hydroxide for the alkaline activators, and PVA for the fibers. ones for EGCs.
• EGC matrices are more brittle than ECC matrices as the former • From the very limited publications available, EGCs show promising
generally exhibit lower matrix elastic modulus and fracture durability and self-healing capabilities potential. Thus, there is a
toughness. need for more research in these areas.
• There are significant differences in the fiber–matrix interface pa­ • Since PVA fibers and alkaline activators were identified as some of
rameters between EGCs and ECCs, which are thought to result from the largest contributors to the energy consumption and CO2 emission
differences in the chemical compositions of the reaction products at of EGCs, methods to reduce their environmental impact need to be
the interfaces. further investigated to develop greener EGCs. Also, research into
• EGCs can exhibit improved pseudo-strain-hardening behavior alternative precursors to fly ash is required to reduce the need for
compared to ECCs, owing to the weaker matrix fracture toughness of heat curing and eventually reduce the overall environmental foot­
EGCs. print of EGCs.
• The micromechanical model developed by Yang et al. [55] for ECCs • The limited published studies on the environmental impact of EGCs
was proven to be applicable for evaluating the tensile performance of adopted a simple approach based on MSIs for evaluating the partial
EGCs. life cycle analysis in terms of the composites’ raw materials only.
• The types of precursors and alkaline activators used in EGCs affect However, to obtain a true assessment of the sustainability of EGCs,
the overall pseudo-strain-hardening response of the composite due to full life cycle analysis studies are required. Further, other impacts of
the formation of different microstructures. EGCs, such as their economic impact, need to be investigated.
• Similar to findings on conventional ECCs, adding sand to EGC
matrices increases the matrix elastic modulus and fracture toughness Declaration of Competing Interest
and generally improves compressive strength but reduces tensile
strain capacity, with more adverse effects observed with coarse sands The authors declare that they have no known competing financial
over fine sands. It was advised to limit the sand content of EGCs, with interests or personal relationships that could have appeared to influence
fine sands recommended over coarse sands. the work reported in this paper.
• Like the well-known effects of the water-to-cement ratio, increasing
the W/GP ratio reduces matrix elastic modulus, matrix fracture Data availability
toughness, compressive strength, tensile strength, and tensile strain.
• PVA fibers result in significantly higher chemical and frictional Data will be made available on request.
bonds and slip-hardening coefficient than PE fibers due to the hy­
drophilic nature and the high elastic modulus of PVA fibers. References
• EGCs were found to exhibit equivalent densities, comparable to
higher tensile strains, comparable to lower tensile strengths, and [1] K. Sakai, T. Noguchi, The Sustainable Use of Concrete, CRC Press, 2012.
lower compressive strengths than ECCs. However, higher tensile and [2] P.K. Mehta, P.J. Monteiro, Concrete: Microstructure, Properties, and Materials,
McGraw-Hill Education, 2014.
compressive strengths can be achieved with EGCs prepared with
[3] P. Domone, J. Illston, Construction Materials: Their Nature and Behaviour, CRC
certain combinations of activators, precursors, and fibers. Press, 2010.
• Sodium-based activators are recommended for use in EGCs over [4] P. Richard, M. Cheyrezy, Composition of reactive powder concretes, Cem. Concr.
Res. 25 (1995) 1501–1511.
lime-based or potassium-based activators since the former result in
[5] A.E. Naaman, Fiber reinforcement for concrete, Concr. Int. 7 (1985) 21–25.
improved pseudo-strain-hardening behavior and higher compressive [6] A.E. Naaman, K. Wille, The path to ultra-high performance fiber reinforced
strengths, tensile strengths, and tensile strains of EGCs owing to their concrete (UHP-FRC): five decades of progress, Proc. Hipermat (2012) 3–15.
higher degree of geopolymerization. [7] M. Yudenfreund, I. Odler, S. Brunauer, Hardened portland cement pastes of low
porosity I. Materials and experimental methods, Cem. Concr. Res. 2 (1972)
• Replacing fly ash with slag reduces the need for heat curing while 313–330.
improving the tensile strengths of EGCs. Contradicting results are [8] D.M. Roy, G.R. Gouda, A. Bobrowsky, Very high strength cement pastes prepared
obtained regarding the effect of replacing fly ash with slag on the by hot pressing and other high pressure techniques, Cem. Concr. Res. 2 (1972)
349–366.
tensile strain and compressive strengths of EGCs. [9] J.D. Birchall, A.J. Howard, K. Kendall, Flexural strength and porosity of cements,
• Curing temperature and duration affect the geopolymer micro­ Nature 289 (1981) 388–390.
structure of EGCs, bringing about different pseudo-strain-hardening [10] H.H. Bache, Densified cement ultra-fine particle-based materials. 1981.
[11] V. Afroughsabet, L. Biolzi, T. Ozbakkaloglu, High-performance fiber-reinforced
behaviors and mechanical properties. concrete: a review, J. Mater. Sci. 51 (2016) 6517–6551.
• The limited studies available on the environmental impact of EGCs [12] D. Yoo, N. Banthia, Mechanical properties of ultra-high-performance fiber-
show that the embodied energy and CO2 emissions of EGCs are up to reinforced concrete: a review, Cem. Concr. Compos. 73 (2016) 267–280.
[13] V.C. Li, Engineered Cementitious Composites (ECC): Bendable Concrete for
35 % and 76 % lower, respectively, than for typical ECC M45. The
Sustainable and Resilient Infrastructure, Springer, 2019.
largest contributors to the energy consumption and CO2 emissions of [14] V.C. Li, From micromechanics to structural engineering the design of
EGCs are PVA fibers, alkaline activators, and heat curing. cementitious composites for civil engineering applications, Doboku Gakkai
Ronbunshu 1993 (1993) 1–12.
[15] S. Wang, V.C. Li, Engineered cementitious composites with high-volume fly ash,
6. Recommendations for Future Studies ACI Mater. J. 104 (2007) 233.
[16] K. Celik, Development and characterization of sustainable self-consolidating
As evident from the literature review conducted in this paper, concrete containing high volume of limestone powder and natural or calcined
pozzolanic materials. : University of California, Berkeley, 2015.
research in the field of EGCs is still in its infancy. Therefore, to better [17] A. Bosoaga, O. Masek, J.E. Oakey, CO2 capture technologies for cement industry,
understand the performance of EGCs and before their practical appli­ Energy Procedia 1 (2009) 133–140.
cations are possible, further investigations are necessary as follows: [18] T. Fransen, K. Lebling, D. Weyl, K. Kennedy, Toward a Tradable Low-Carbon
Cement Standard: Policy Design Considerations for the United States. 2021.
[19] K. Celik, C. Meral, A.P. Gursel, P.K. Mehta, A. Horvath, P.J. Monteiro, Mechanical
properties, durability, and life-cycle assessment of self-consolidating concrete

23
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

mixtures made with blended portland cements containing fly ash and limestone [52] N.J. Van Eck, L. Waltman, Software survey: VOSviewer, a computer program for
powder, Cem. Concr. Compos. 56 (2015) 59–72. bibliometric mapping, Scientometrics 84 (2010) 523–538.
[20] K. Celik, R. Hay, C.W. Hargis, J. Moon, Effect of volcanic ash pozzolan or [53] V.C. Li, C.K. Leung, Steady-state and multiple cracking of short random fiber
limestone replacement on hydration of Portland cement, Constr. Build. Mater. composites, J. Eng. Mech. 118 (1992) 2246–2264.
197 (2019) 803–812. [54] J. Li, J. Qiu, S. He, E. Yang, Micromechanics-based design of strain hardening
[21] K. Celik, M.D. Jackson, M. Mancio, C. Meral, A. Emwas, P.K. Mehta, et al., High- cementitious composites (SHCC). 2017:12-27.
volume natural volcanic pozzolan and limestone powder as partial replacements [55] E. Yang, S. Wang, Y. Yang, V.C. Li, Fiber-bridging constitutive law of engineered
for portland cement in self-compacting and sustainable concrete, Cem. Concr. cementitious composites, J. Adv. Concr. Technol. 6 (2008) 181–193.
Compos. 45 (2014) 136–147. [56] T. Kanda, V.C. Li, Interface property and apparent strength of high-strength
[22] K.L. Scrivener, B. Lothenbach, N. De Belie, E. Gruyaert, J. Skibsted, R. Snellings, hydrophilic fiber in cement matrix, J. Mater. Civ. Eng. 10 (1998) 5–13.
et al., TC 238-SCM: hydration and microstructure of concrete with SCMs, Mater. [57] T. Kanda, V.C. Li, Practical design criteria for saturated pseudo strain hardening
Struct. 48 (2015) 835–862. behavior in ECC, J. Adv. Concr. Technol. 4 (2006) 59–72.
[23] J. Li, W. Zhang, K. Garbev, P.J. Monteiro, Coordination environment of Si in [58] M. Ohno, Green and Durable Geopolymer Composites for Sustainable Civil
calcium silicate hydrates, silicate minerals, and blast furnace slags: A XANES Infrastructure. Green and Durable Geopolymer Composites for Sustainable Civil
database, Cem. Concr. Res. 143 (2021), 106376. Infrastructure, 2017.
[24] J. Li, W. Zhang, C. Li, P.J. Monteiro, Eco-friendly mortar with high-volume [59] P. Duxson, A. Fernández-Jiménez, J.L. Provis, G.C. Lukey, A. Palomo, J.S. van
diatomite and fly ash: Performance and life-cycle assessment with regional Deventer, Geopolymer technology: the current state of the art, J. Mater. Sci. 42
variability, J Clean Prod 261 (2020), 121224. (2007) 2917–2933.
[25] K. Celik, C. Meral, M. Mancio, P.K. Mehta, P.J. Monteiro, A comparative study of [60] F. Li, Z. Yang, A. Zheng, S. Li, Properties of modified engineered geopolymer
self-consolidating concretes incorporating high-volume natural pozzolan or high- composites incorporating multi-walled carbon Nanotubes (MWCNTs) and
volume fly ash, Constr. Build. Mater. 67 (2014) 14–19. granulated blast furnace Slag (GBFS), Ceram. Int. 47 (2021) 14244–14259.
[26] K. Scrivener, F. Martirena, S. Bishnoi, S. Maity, Calcined clay limestone cements [61] M. Zahid, N. Shafiq, M.H. Isa, L. Gil, Statistical modeling and mix design
(LC3), Cem. Concr. Res. 114 (2018) 49–56. optimization of fly ash based engineered geopolymer composite using response
[27] R. Hay, L. Li, K. Celik, Shrinkage, hydration, and strength development of surface methodology, J Clean Prod 194 (2018) 483–498.
limestone calcined clay cement (LC3) with different sulfation levels, Cem. Concr. [62] M. Zahid, N. Shafiq, Effects of sand/fly ash and the water/solid ratio on the
Compos. 104403 (2022). mechanical properties of engineered geopolymer composite and mix design
[28] E. Benhelal, E. Shamsaei, M.I. Rashid, Challenges against CO2 abatement optimization, Minerals 10 (2020) 333.
strategies in cement industry: a review, J. Environ. Sci. 104 (2021) 84–101. [63] Y. Ling, K. Wang, W. Li, G. Shi, P. Lu, Effect of slag on the mechanical properties
[29] K.L. Scrivener, R.J. Kirkpatrick, Innovation in use and research on cementitious and bond strength of fly ash-based engineered geopolymer composites, Compos. B
material, Cem. Concr. Res. 38 (2008) 128–136. Eng. 164 (2019) 747–757.
[30] B. Nematollahi, Investigation of Geopolymer as a Sustainable Alternative Binder [64] B. Nematollahi, J. Sanjayan, J. Qiu, E. Yang, Micromechanics-based investigation
for Fiber-Reinforced Strain-Hardening Composites. Doctoral dissertation 2017. of a sustainable ambient temperature cured one-part strain hardening
[31] J. Davidovits, Geopolymers: inorganic polymeric new materials, J. Therm. Anal. geopolymer composite, Constr. Build. Mater. 131 (2017) 552–563.
Calorim. 37 (1991) 1633–1656. [65] M.H. Al-Majidi, A. Lampropoulos, A.B. Cundy, Tensile properties of a novel fibre
[32] J. Davidovits, Geopolymer chemistry and applications, Geopolymer Institute reinforced geopolymer composite with enhanced strain hardening characteristics,
(2008). Compos. Struct. 168 (2017) 402–427.
[33] D.M. Roy, Alkali-activated cements opportunities and challenges, Cem. Concr. [66] M. Farooq, A. Bhutta, N. Banthia, Strain-hardening ambient-cured eco-friendly
Res. 29 (1999) 249–254. ductile geopolymer composites, ACI Mater. J. 117 (2020) 181–189.
[34] F. Pacheco-Torgal, J. Castro-Gomes, S. Jalali, Alkali-activated binders: A review: [67] Y. Wang, Y. Wang, M. Zhang, Effect of sand content on engineering properties of
Part 1. Historical background, terminology, reaction mechanisms and hydration fly ash-slag based strain hardening geopolymer composites, J. Build. Eng. (2020)
products, Constr. Build. Mater. 22 (2008) 1305–1314. 101951.
[35] J.L. Provis, V. Deventer, J.S. Jakob, Geopolymers: structures, processing, [68] M. Zahid, N. Shafiq, S.N.A. Razak, R.F. Tufail, Investigating the effects of NaOH
properties and industrial applications, : Elsevier (2009). molarity and the geometry of PVA fibers on the post-cracking and the fracture
[36] J. Davidovits, Geopolymer cements to minimize carbon dioxide greenhouse behavior of engineered geopolymer composite, Constr. Build. Mater. 265 (2020),
warming, Ceram. Trans. 37 (1993) 165–182. 120295.
[37] P. Duxson, J.L. Provis, G.C. Lukey, J.S. Van Deventer, The role of inorganic [69] J. Choi, B.Y. Lee, R. Ranade, V.C. Li, Y. Lee, Ultra-high-ductile behavior of a
polymer technology in the development of ‘green concrete’, Cem. Concr. Res. 37 polyethylene fiber-reinforced alkali-activated slag-based composite, Cem. Concr.
(2007) 1590–1597. Compos. 70 (2016) 153–158.
[38] D.A. Salas, A.D. Ramirez, N. Ulloa, H. Baykara, A.J. Boero, Life cycle assessment [70] Y. Alrefaei, J. Dai, Tensile behavior and microstructure of hybrid fiber ambient
of geopolymer concrete, Constr. Build. Mater. 190 (2018) 170–177. cured one-part engineered geopolymer composites, Constr. Build. Mater. 184
[39] S.H. Teh, T. Wiedmann, A. Castel, J. de Burgh, Hybrid life cycle assessment of (2018) 419–431.
greenhouse gas emissions from cement, concrete and geopolymer concrete in [71] S. Choi, J. Choi, J. Song, B.Y. Lee, Rheological and mechanical properties of fiber-
Australia, J. Clean. Prod. 152 (2017) 312–320. reinforced alkali-activated composite, Constr. Build. Mater. 96 (2015) 112–118.
[40] R. Bajpai, K. Choudhary, A. Srivastava, K.S. Sangwan, M. Singh, Environmental [72] J. Choi, H.H. Nguyễn, S. Park, R. Ranade, B.Y. Lee, Effects of fiber hybridization
impact assessment of fly ash and silica fume based geopolymer concrete, J Clean on mechanical properties and autogenous healing of alkali-activated slag-based
Prod 254 (2020), 120147. composites, Constr. Build. Mater. 310 (2021), 125280.
[41] M. Ohno, V.C. Li, A feasibility study of strain hardening fiber reinforced fly ash- [73] B. Nematollahi, J. Sanjayan, F.U.A. Shaikh, Matrix design of strain hardening
based geopolymer composites, Constr. Build. Mater. 57 (2014) 163–168. fiber reinforced engineered geopolymer composite, Compos. B Eng. 89 (2016)
[42] B. Nematollahi, J. Sanjayan, F.U. Ahmed Shaikh, Tensile strain hardening 253–265.
behavior of PVA fiber-reinforced engineered geopolymer composite, J. Mater. [74] B. Nematollahi, R. Ranade, J. Sanjayan, S. Ramakrishnan, Thermal and
Civ. Eng. 27 (2015) 04015001. mechanical properties of sustainable lightweight strain hardening geopolymer
[43] L. Kan, F. Wang, Z. Zhang, W. Kabala, Y. Zhao, Mechanical properties of high composites, Arch. Civ. Mech. Eng. 17 (2017) 55–64.
ductile alkali-activated fiber reinforced composites with different curing ages, [75] B. Nematollahi, J. Qiu, E. Yang, J. Sanjayan, Micromechanics constitutive
Constr. Build. Mater. 306 (2021), 124833. modelling and optimization of strain hardening geopolymer composite, Ceram.
[44] B. Lyu, C. Ding, L. Guo, B. Chen, A. Wang, Basic performances and potential Int. 43 (2017) 5999–6007.
research problems of strain hardening geopolymer composites: A critical review, [76] B. Nematollahi, J. Qiu, E. Yang, J. Sanjayan, Microscale investigation of fiber-
Constr. Build. Mater. 287 (2021), 123030. matrix interface properties of strain-hardening geopolymer composite, Ceram.
[45] D.J. Hess, Science studies: An advanced introduction. : NYU press, 1997. Int. 43 (2017) 15616–15625.
[46] B.Y. Lee, C. Cho, H. Lim, J. Song, K. Yang, V.C. Li, Strain hardening fiber [77] M. Ohno, V.C. Li, An integrated design method of Engineered Geopolymer
reinforced alkali-activated mortar–a feasibility study, Constr. Build. Mater. 37 Composite, Cem. Concr. Compos. 88 (2018) 73–85.
(2012) 15–20. [78] M. Ohno, Self-healing Capacity of Strain-Hardening Fiber Reinforced
[47] M.J. Cobo, A.G. López-Herrera, E. Herrera-Viedma, F. Herrera, An approach for Geopolymer, Composites. (2020).
detecting, quantifying, and visualizing the evolution of a research field: A [79] H.H. Nguyễn, Q. Lương, J. Choi, R. Ranade, V.C. Li, B.Y. Lee, Ultra-ductile
practical application to the fuzzy sets theory field, J. Informetrics 5 (2011) behavior of fly ash-based engineered geopolymer composites with a tensile strain
146–166. capacity up to 13.7%. Cem. Concr. Compos. 2021:104133.
[48] H. Small, Visualizing science by citation mapping, J. Am. Society Information Sci. [80] S.U. Khan, T. Ayub, PET fiber-reinforced engineered geopolymer and
50 (1999) 799–813. cementitious composites, J. Mater. Civ. Eng. 34 (2022) 06021010.
[49] R.J. Tijssen, A.F. Van Raan, Mapping changes in science and technology: [81] B.A. Salami, M.A. Megat Johari, Z.A. Ahmad, M. Maslehuddin, Impact of added
bibliometric co-occurrence analysis of the R&D literature, Eval. Rev. 18 (1994) water and superplasticizer on early compressive strength of selected mixtures of
98–115. palm oil fuel ash-based engineered geopolymer composites, Constr. Build. Mater.
[50] M.J. Cobo, A.G. López-Herrera, E. Herrera-Viedma, F. Herrera, Science mapping 109 (2016) 198–206.
software tools: review, analysis, and cooperative study among tools, J Am Soc Inf [82] B.A. Salami, M.A.M. Johari, Z.A. Ahmad, M. Maslehuddin, Durability
Sci Technol 62 (2011) 1382–1402. performance of palm oil fuel ash-based engineered alkaline-activated
[51] J.A.M. Muñoz, E.H. Viedma, A.L.S. Espejo, M.J. Cobo, Software tools for cementitious composite (POFA-EACC) mortar in sulfate environment, Constr.
conducting bibliometric analysis in science: an up-to-date review, El profesional Build. Mater. 131 (2017) 229–244.
de la información 29 (2020) 4.

24
N. Elmesalami and K. Celik Construction and Building Materials 346 (2022) 128491

[83] A.C.C. Trindade, P.H.R. Borges, S.F. de Andrade, Mechanical behavior of strain- [98] L. Kan, W. Wang, W. Liu, M. Wu, Development and characterization of fly ash
hardening geopolymer composites reinforced with natural and PVA fibers, Mater. based PVA fiber reinforced Engineered Geopolymer Composites incorporating
Today:. Proc. 8 (2019) 753–759. metakaolin, Cem. Concr. Compos. 108 (2020), 103521.
[84] A.C.C. Trindade, A.A. Heravi, I. Curosu, M. Liebscher, S.F. de Andrade, [99] L. Kan, L. Zhang, Y. Zhao, M. Wu, Properties of polyvinyl alcohol fiber reinforced
V. Mechtcherine, Tensile behavior of strain-hardening geopolymer composites fly ash based Engineered Geopolymer Composites with zeolite replacement,
(SHGC) under impact loading, Cem. Concr. Compos. 113 (2020), 103703. Constr. Build. Mater. 231 (2020), 117161.
[85] A.C.C. Trindade, I. Curosu, M. Liebscher, V. Mechtcherine, S.F. de Andrade, On [100] F. Li, Z. Yang, D. Chen, Y. Lu, S. Li, Research on mechanical properties and micro-
the mechanical performance of K-and Na-based strain-hardening geopolymer mechanism of Engineering Geopolymers Composites (EGCs) incorporated with
composites (SHGC) reinforced with PVA fibers, Constr. Build. Mater. 248 (2020), modified MWCNTs, Constr. Build. Mater. 303 (2021), 124516.
118558. [101] L. Kan, R. Shi, Y. Zhao, X. Duan, M. Wu, Feasibility study on using incineration fly
[86] A.C. Trindade, I. Curosu, M. Liebscher, V. Mechtcherine, A. Silva Fd, Mechanical ash from municipal solid waste to develop high ductile alkali-activated
Evaluation of Na-Based Strain-Hardening Geopolymer Composites (SHGC) composites, J Clean Prod 254 (2020), 120168.
Reinforced with PVA, UHMWPE, and PBO Fibers. 2021:857-67. [102] X. Guo, J. Yang, Intrinsic properties and micro-crack characteristics of ultra-high
[87] Y. Pang, X. Zhu, M. Yang, J. Yu, Tailoring rheological–strength–ductility toughness fly ash/steel slag based geopolymer, Constr. Build. Mater. 230 (2020),
properties of self-cleaning geopolymer composites with asphalt emulsion, Constr. 116965.
Build. Mater. 308 (2021), 124997. [103] Y. Zhao, T. Shi, L. Cao, L. Kan, M. Wu, Influence of steel slag on the properties of
[88] J. Cai, J. Pan, J. Han, Y. Lin, Z. Sheng, Impact behaviours of engineered alkali-activated fly ash and blast-furnace slag based fiber reinforced composites,
geopolymer composite exposed to elevated temperatures, Constr. Build. Mater. Cem. Concr. Compos. 116 (2021), 103875.
312 (2021), 125421. [104] A. Akbar, F. Farooq, M. Shafique, F. Aslam, R. Alyousef, H. Alabduljabbar,
[89] J. Cai, J. Pan, J. Han, Y. Lin, Z. Sheng, Low-energy impact behavior of ambient Sugarcane bagasse ash-based engineered geopolymer mortar incorporating
cured engineered geopolymer composites, Ceram. Int. (2021). propylene fibers, J. Build. Eng. 33 (2021), 101492.
[90] A.C.C. Trindade, M. Liebscher, I. Curosu, S.F. de Andrade, V. Mechtcherine, [105] D. Hardjito, S.E. Wallah, D.M. Sumajouw, B.V. Rangan, On the development of fly
Influence of elevated temperatures on the residual and quasi in-situ flexural ash-based geopolymer concrete, Mater. J. 101 (2004) 467–472.
strength of strain-hardening geopolymer composites (SHGC) reinforced with PVA [106] A. Fernández-Jiménez, A. Palomo, Composition and microstructure of alkali
and PE fibers, Constr. Build. Mater. 314 (2022), 125649. activated fly ash binder: Effect of the activator, Cem. Concr. Res. 35 (2005)
[91] M.H. Al-Majidi, A. Lampropoulos, A.B. Cundy, Strengthening of plain concrete 1984–1992.
beams using Strain Hardening Geopolymer Composites (SHGC), layers. 109 [107] V.C. Li, D.K. Mishra, H. Wu, Matrix design for pseudo-strain-hardening fibre
(2017) 2707–2714. reinforced cementitious composites, Mater. Struct. 28 (1995) 586–595.
[92] B. Nematollahi, J. Sanjayan, J. Qiu, E. Yang, High ductile behavior of a [108] V.C. Li, S. Wang, C. Wu, Tensile strain-hardening behavior of polyvinyl alcohol
polyethylene fiber-reinforced one-part geopolymer composite: a micromechanics- engineered cementitious composite (PVA-ECC), Materials Journal 98 (2001)
based investigation, Arch. Civ. Mech. Eng. 17 (2017) 555–563. 483–492.
[93] Y. Wang, C.L. Chan, S.H. Leong, M. Zhang, Engineering properties of strain [109] ACI Committee 213. Guide for Structural Lightweight-Aggregate Concrete. 2014.
hardening geopolymer composites with hybrid polyvinyl alcohol and recycled [110] OHNO M, LI VC. Sulfuric acid resistance of strain hardening fiber reinforced
steel fibres, Constr. Build. Mater. 261 (2020), 120585. geopolymer composite. Indian Concrete J 2019.
[94] S. Zhang, V.C. Li, G. Ye, Micromechanics-guided development of a slag/fly ash- [111] G. Yıldırım, Ö.K. Keskin, S.B. Keskin, M. Şahmaran, M. Lachemi, A review of
based strain-hardening geopolymer composite, Cem. Concr. Compos. 109 (2020), intrinsic self-healing capability of engineered cementitious composites: recovery
103510. of transport and mechanical properties, Constr. Build. Mater. 101 (2015) 10–21.
[95] F.U.A. Shaikh, A. Fairchild, Strain hardening behaviour of polyethylene fibre [112] X. Guo, G. Xiong, H. Zhang, In-situ evaluation of self-healing performance of
reinforced ambient air cured geopolymer composite. 2017:162-71. Engineered Geopolymer Composites (EGC) by ultrasonic method, Mater. Lett. 280
[96] H. Zhong, M. Zhang, Effect of recycled tyre polymer fibre on engineering (2020), 128546.
properties of sustainable strain hardening geopolymer composites, Cem. Concr. [113] K.K. Yaswanth, J. Revathy, P. Gajalakshmi, Soft computing techniques for the
Compos. 122 (2021), 104167. prediction and analysis of compressive strength of alkali-activated Alumino-
[97] L. Kan, J. Lv, B. Duan, M. Wu, Self-healing of Engineered Geopolymer Composites silicate based strain-hardening Geopolymer composites, Silicon (2021) 1–24.
prepared by fly ash and metakaolin, Cem. Concr. Res. 125 (2019), 105895.

25