Download as pdf or txt
Download as pdf or txt
You are on page 1of 13

Cement and Concrete Composites 133 (2022) 104687

Contents lists available at ScienceDirect

Cement and Concrete Composites


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

Enhancing the properties of foam concrete 3D printing using


porous aggregates
Kirubajiny Pasupathy *, Sayanthan Ramakrishnan, Jay Sanjayan
Centre for Sustainable Infrastructure and Digital Construction, School of Engineering, Swinburne University of Technology, Hawthorn, Victoria, 3122, Australia

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

Keywords: The lightweight concrete (density <1000 kg/m3) is generally attained by introducing a large amount of air voids
Foam concrete into fresh concrete for making so-called foam concrete. Such foam concrete is quite challenging in 3D concrete
3D printing printing due to the high flowability of fresh mixes affecting the printability and foam stability during extrusion
Rheological properties
process. To overcome these limitations, this study investigates a combination of lightweight aggregates and
Expanded perlite
Porosity
premade foam in foam concrete to attain a density below 1000 kg/m3 for 3D printing applications. The expanded
Compressive strength perlite (EP) aggregate was used as a replacement for fine sand that substantially reduced the foam content in the
mix. The effect of EP on the fresh state properties such as rheology and printability as well as hardened properties
including, mechanical properties, porosity and pore size distribution were investigated. It was demonstrated that
the introduction of combined lightweight aggregate and foam has significantly improved the fresh and hardened
properties of produced lightweight concrete. For instance, fresh foam concrete containing EP aggregate displayed
high yield strength and apparent viscosity compared to the foam concrete without EP at similar densities. The
compressive strength of 3D printed specimens containing EP was determined as 12.95 MPa, 15.5 MPa and 10.6
MPa in the perpendicular, longitudinal, and lateral directions respectively, compared to 5.5 MPa, 8.4 MPa and
4.2 MPa for the sand group at the similar density. Moreover, fine and regular pore size distribution was observed
for 3D printed foam concrete with EP aggregate.

1. Introduction 3D model of the structure [8].


In the recent past, the extrusion-based 3DCP has been successfully
3D concrete printing (3DCP) is an emerging technology, trans­ used to build large scale structures including the construction of an of­
forming the construction industry by adopting construction automation fice building in Dubai by WinSun [9,10], a house built by Apis Cor [11],
and digital manufacturing technologies [1]. Compared to the conven­ full-scale bridges made by TU Einthoven [12] and the Hebei University
tional construction process, 3DCP offers a significant reduction in con­ of Technology [13]. These outcomes suggest the rapid growth of 3DCP
struction wastes, labour cost, time and construction-related accidents as and the potential to transform the current construction industry.
well as enables architectural freedom to construct artistically intricate Meanwhile, the demand for lightweight and thermally enhanced
structures [2,3]. 3DCP can be in the form of extrusion-based 3DCP, buildings are increasing and the opportunity to use 3DCP to construct
powder-bed 3DCP, contour crafting, etc [4]. Among these various these structures remains a key challenge. In particular, the modern
methods, extrusion-based 3DCP is most commonly used for constructing construction methods urge the use of lightweight and thermally insu­
large-scale structures [5–8]. This is mainly due to the inherent merits of lative building elements such as sandwich panels, vacuum insulated
the extrusion method including, the ease of implementation in panels (VIPs) and structural insulated panels (SIPs) due to their merits of
large-scale 3D printing, availability of sophisticated robots or gantry high insulation capacity, airtightness and ease of installation in build­
systems for printing, rapid printing process, etc. In the extrusion ings [14]. Therefore, the introduction of 3DCP in the construction of
method, 3D printed structures are built by extruding and orienting a thin lightweight elements or buildings could be a great venue to explore for
layer of fresh concrete filaments following the shape of the extrusion modern buildings.
nozzle. The extrusion nozzle is generally mounted on a robotic arm or The most common approach to constructing lightweight structures
gantry system that moves on the pre-defined path generated through the with 3DCP is to use a lightweight concrete mix for printing. The

* Corresponding author.
E-mail address: kpasupathy@swin.edu.au (K. Pasupathy).

https://doi.org/10.1016/j.cemconcomp.2022.104687
Received 31 January 2022; Received in revised form 20 May 2022; Accepted 12 July 2022
Available online 16 July 2022
0958-9465/© 2022 Elsevier Ltd. All rights reserved.
K. Pasupathy et al. Cement and Concrete Composites 133 (2022) 104687

lightweight concrete mix containing porous aggregates is utilised with The primary challenge of 3D printing conventional foam concrete is
extrusion-based or powder bed 3D printing methods. For instance, the low yield stress at the fresh state, causing poor shape retention and
Weger et al. [15] used expanded glass beads aggregate as a replacement buildability. Such low yield stress is essential in conventional foam
for sand in powder bed 3D printing. Authors have reported that the concrete to minimise bubble destruction and collapse during the mixing
lightweight concrete showed an approximate density of 1015 kg/m3 process. High yield stress of base mix (the mix before blending with
with a strength loss of up to 50%. In another study, the fresh state foam) can cause significant destruction of bubbles due to the high shear
properties of 3D printed lightweight concrete with expanded clay ag­ stress during the mixing. To overcome this limitation and to adapt the
gregates were assessed in the extrusion 3DCP method [16]. The volume foam concrete for extrusion 3D printing, Cho et al. [27] have introduced
substitution of expanded clay aggregates for up to 30% of sand has 2% of nanopowder and 10% of calcium sulfoaluminate cement for
shown good extrudability with further increase in substitution resulting improved thixotropy in the mix, resulting in up to 15 deposited filament
in phase separation and blockages. It was also noted that lightweight layers without collapse during the buildability test. However, the me­
concrete has increased both elastic modulus and strength at an early age. chanical properties of 3D printed elements have not been reported. More
Apart from utilising a lightweight concrete mix, the design of light­ recently, Liu et al. [29] reported that incorporating hydroxypropyl
weight building elements to introduce voids or pores during the printing methylcellulose (HPMC) and silica fume as a viscosity modifier and
has also been explored. Ramakrishnan et al. [8] assessed the lightweight thixotropic agents in foam concrete improves the foam stability and
3DCP using a hollow-core filament extrusion process. The presence of increased the rheological properties. The authors managed to achieve a
voids in the filament has reduced the bulk density of the 3D printed compressive strength above 20 MPa for the density of 1815 kg/m3.
elements to 1369.6 kg/m3 while showing the compressive strength of However, the mixes containing HPMC and silica fume were affected by
19.3 MPa at 28 days. In another study [17], a combination of light­ the extrudability and buildability with the rest time. In addition, the
weight concrete containing ceramsite sand and lattice structures has performance of these additives for foam concrete at low density (density
performed 90% of the strength of solid 3D printing using 53.3% of <1000 kg/m3) remains unknown due to high foam content. Moreover,
materials, showing a significant reduction in density. From these the type of foaming agent also influences the fresh stage properties of
studies, it can be said that the utilisation of lightweight concrete with foam concrete. More recently, Falliano et al. [32] studied the effect of
3DCP could be a promising method for attaining lightweight and ther­ different types of foaming agents on the fresh properties of foam con­
mally enhanced building elements. However, the final density and crete and determined that the protein-based foaming agent displayed
insulation properties of this method are very limited due to the limita­ better workability at the low density in the order of 100 kg/m3.
tions in attaining further reduction in density with lightweight aggre­ It is therefore understood that the rheological properties of foam
gates. In this regard, aerated concrete or foam concrete has been studied concrete, particularly at low density, remains a major challenge for
to attain the density of concrete below 1000 kg/m3. 3DCP applications. This is mainly due to the high foam content (foam
Foam concrete is a class of lightweight concrete formulated by content is above 15% of cement content for density below 1000 kg/m3)
introducing large amounts of air voids/pores in the fresh cement matrix in these mixes causing low yield stress at the fresh state and hence, the
[18]. Here, air voids can be introduced by either mechanical foaming low shape retention and buildability [20]. A possible solution to manage
(pre-foaming method) or chemical foaming method. While the former the contradicting requirements of low density and yield stress is by
method introduces pre-made foam at a designated ratio into the fresh introducing lightweight aggregates to replace the sand in 3D printable
concrete mix, the latter method uses a gas releasing element (Al powder, mixes. The lightweight aggregates would substantially reduce the
Silica fume, and H2O2) mixed with the fresh mix that subsequently amount of foam content to attain a similar density [33,34]. The previous
generates voids in the concrete. Among these two methods, the studies showed that replacing sand with lightweight aggregates,
pre-foaming method is easier and more convenient as it allows precise including, expanded perlite [35], pumice [33] and polystyrene [34] has
control of the final density and pore properties of produced aerated substantially reduced the foam content and improved the mechanical
concrete [19,20]. Foam concrete has many advantages over other properties of traditional lightweight concrete. However, the influence of
lightweight concrete production methods, including the possibility of the lightweight aggregate with premade foam on the stability, rheo­
attaining ultra-low density (below 300 kg/m3), high thermal insulation logical properties, and printability of 3D printable foam concrete has not
properties, acoustic and fire-resistant properties and cost-effectiveness been explored.
[14,21]. Therefore, the application of foam concrete is emerging in In this study, a combination of lightweight aggregates and premade
modern construction methods such as façade construction, barrier walls, foam is investigated to study the effect of lightweight aggregates on the
ceiling and external cladding [22–24]. Moreover, in building construc­ fresh rheological properties and hardened properties of 3D printable
tion, foam concrete is widely used for non-load-bearing building ele­ foam concrete. The expanded perlite (EP) was chosen as the lightweight
ments to improve thermal insulation properties. With these aggregate due to its enhanced mechanical and thermal properties in
considerations, the 3DCP of foam concrete can be a great potential to foam concrete as reported in the authors’ previous study [35]. Five foam
construct complex architectural structures, particularly in façade con­ concrete mixes with a wet density ranging from 800 to 1355 kg/m3 were
struction, providing lightweight performance and high thermal insu­ considered. The effect of EP aggregate on the extrudability and fresh
lation [25]. state rheological properties was first studied. The fresh rheological pa­
So far, only a few studies have assessed the feasibility of 3DCP of rameters including, static yield stress, dynamic yield stress, plastic vis­
foam concrete [22,26–29]. Falliano et al. [30] have compared the 3D cosity and viscosity recovery were determined with the EP content for
printable lightweight foam concrete with classical lightweight foam various density mixes. In addition, mechanical properties, such as
concrete and reported that the 3D printable foam concrete has enhanced compressive strength and interlayer bond strength, were also studied
the mechanical properties without much deformation in a fresh state, and compared with the sand-based foam concrete at the similar density
compared to the classical foam concrete. Moreover, they have observed ranges. Finally, the porosity and microstructural characterisation of 3D
that increasing the mixing intensity from 1200 RPM to 3000 RPM has printed foam concrete was evaluated with the replacement of EP
resulted in an increase in compressive strength by up to 70%. Markin content.
et al. [26] have demonstrated the 3DCP of foam concrete for the density
range of 1100–1580 kg/m3 with the compressive strength achieved 2. Methodology
above 10 MPa for optimum mixture composition. In a subsequent study
[31], authors have developed four different printable foam concrete 2.1. Materials
mixes with the density ranging from 800 to 1200 kg/m3 and respective
compressive strength of 4.2–8.3 MPa at 35–38 days age. General-purpose (Type GP) Portland cement, complying with AS

2
K. Pasupathy et al. Cement and Concrete Composites 133 (2022) 104687

3972 [36], was used as a binder to formulate the foam concrete. The maximum printing dimensions of 1800 mm (L) × 1600 mm (W) × 1800
hydrophobic surface-modified expanded perlite (EP), supplied by Fil­ mm (H). The extruder is an auger-type and the diameter of the circular
chem Australia Pty Ltd, was chosen as the lightweight aggregates. The nozzle was 30 mm. A computer program was used to control the navi­
EP particles were hydrophobic coated to minimise the water affinity and gation and the speed of the extruder.
hence the water absorption from the mix. The supplier reports 0% of The samples were printed with the dimensions of each layer as 30
hydroscopic moisture of the expanded perlite [37]. The authors mm (W) × 20 mm (H) × 300 mm (L) at a constant printing speed and
attempted to measure the water absorption of the perlite particles, feed rate of 6 mm/s and 0.216 l/min respectively. At the same time,
however, the expanded perlite particles do not mix with the water. foam concrete mix was also cast in the 50 × 50 × 50 mm3 cubic moulds
Therefore, the water absorption of the expanded perlite can be consid­ for the comparison of strength properties. After printing, all the samples
ered as 0%. The fine sand with a median diameter (D50) of 600 μm was were covered with the plastic film for 24 hr to avoid moisture loss from
used. The particle size distributions of the fine sand and EP were studied the surface. After 24 hr, the specimens were removed from the platform
and published in the author’s previous study [35]. MasterGlenium® SKY and stored in a sealed container with a covered wet cloth until the test
8379 polycarboxylate ether-based superplasticiser with a solid content date.
of 34.9%, supplied by BASF, Australia, was used to adjust the work­
ability of the mixes. The foaming agents are commonly available in the 3. Experimental methods
form of protein-based agents or surfactants. In this study, a
surfactant-based foaming agent, known as sodium dodecyl sulfate (SDS), 3.1. Flow-table test
was used due to its availability and wide use for commercial foam
concrete production. The premade foam was produced by diluting SDS The flowability of fresh foam concrete was determined in accordance
(30% concentration) with water at the ratio of 1:60 and blending in a with the ASTM C1437 [38] using a flow table. During the experiment, a
high shear mixer operated at 2000 RPM for 10 mins. mini-slump cone was filled with fresh foam concrete. Then the cone was
lifted, and the fresh concrete was allowed to reach stability. The spread
2.2. Sample preparation diameter was determined in two perpendicular directions and the
average flow diameter was determined. Thereafter, the flow table was
Two foam concrete mixes containing fine sand as aggregates (M1 and dropped 25 times. The final flow diameter of the mixes was measured in
M2), and three mixes by replaced fine sand with EP (M3, M4 and M5) two perpendicular directions and the average flow diameter was
were formulated. A volumetric replacement method is followed when determined.
replacing fine sand with EP in the M3, M4 and M5 mixes. The mix
composition details are given in Table 1. Here, the water to cement ratio 3.2. Rheological properties measurement
was kept at 0.32 for the mixes containing fine sand (M1 and M2), and
0.52 for mixes designed with EP (M3, M4 and M5). It should be noted that A Viskomat XL rheometer was employed to determine the rheolog­
the EP has a high specific surface area, and therefore, the water demand ical properties of foam concrete, including static yield stress, dynamic
is high to maintain the same workability with sand-based mixes. The yield stress, plastic viscosity and viscosity recovery of the foam mixes.
mixing procedure was as follows: Initially, the dry ingredients such as For the rheological measurement, a six-blade vane probe with the radius
cement and fine sand/EP were mixed using a Hobart mixer for 2 mins at and height of each blade equal to 34.5 mm and 69 mm was used.
a low speed. Thereafter, the required amount of water and super­ It is important to determine the static yield stress of fresh foam
plasticiser were added with dry ingredients, and the mixing was concrete as it governs the buildability characteristics, which is a key
continued for another 3 mins at the intermediate speed. Meanwhile, the challenge in the 3DCP of foam concrete. Fig. 2 (a) shows the rheological
foam is generated by the mechanical foaming method using the high testing protocol used to determine the static yield stress of foam con­
shear mixer. A solution of water and surfactant at the ratio of 1:60 was crete. Before starting the measurement, the mix was initially pre-sheared
used for producing the foam. The density of the premade foam was 50 at 60 rpm for 60 s to de-flocculate the mix and to ensure a uniform
kg/m3. After the pre-made foam is produced, it was blended with starting point for all the tests. The mix was then kept at rest for 60 s,
cement sand slurry at different mass ratios to achieve the corresponding followed by a constant low shear rate (0.6 rpm) applied to determine the
density of mixes, as reported in Table 1. The mixing was continued for static yield stress properties. Fig. 2 (b) shows a typical response curve to
another 2 mins at a low speed to achieve a homogenous mixing of foam determine the static yield stress at the control shear rate.
and cement sand slurry. The wet density of the fresh mixes was imme­ The flow curve test was used to measure the dynamic yield stress and
diately determined by measuring the weight of the mix with a graduated the plastic viscosity of foam concrete. Fig. 3 (a) shows the rheological
measuring cylinder. The wet density of the fresh foam concrete mixes testing protocol for the dynamic yield stress and the plastic viscosity
was also determined after the extrusion process at the time intervals of measurements. During the flow curve test, the rotational velocity line­
0,10, and 20 mins from extrusion to determine the collapse of foam arly increased from 0 to 60 rpm in 100 s and then decreased from 60 to
during the extrusion process. 0 rpm in 100 s, as shown in Fig. 3 (a). The shear stress versus shear rate
graph was plotted, and a linear fitting line was drawn by considering the
2.3. 3D printing process fresh foam mix as Bingham fluid. Fig. 3 (b) illustrates the typical flow
curve used to determine the dynamic yield stress and the plastic vis­
Fig. 1 illustrates the 3D printer and the extruder type used in the cosity according to the Bingham model. From this model, dynamic yield
printing process for this study. It is a gantry-type 3D printer with the stress can be determined by the intercept of the fitting line on the

Table 1
Mix compositions of 3D printable foam concrete.
Mix No Cement (kg) Fine sand (kg) EP (kg) Water (kg) Sp (kg) Foam content (% of cement) Measured Wet density (kg/m3)

M1 1 0.856 – 0.32 0.0016 5 1244


M2 1 0.856 – 0.32 0.0016 11 840
M3 1 – 0.133 0.52 0.0016 2.5 1355
M4 1 – 0.133 0.52 0.0016 5 1050
M5 1 – 0.133 0.52 0.0016 7.5 800

3
K. Pasupathy et al. Cement and Concrete Composites 133 (2022) 104687

Fig. 1. (a) Gantry type 3D printer used for this study, (b) Extruder, (c) Printing process.

Fig. 2. (a) Rheological testing protocols for static yield stress measurements, (b) Typical hysteresis curve for static yield stress measurement.

vertical axis [39], whereas the slope of the linear fitting represents the stated in previous rheological test methods, the fresh foam concrete used
plastic viscosity. It is important to note that the fresh foam concrete was for this test was also pre-sheared at 60 rpm for 60 s before starting the
initially pre-sheared at 60 rpm for 60 s to de-flocculate and maintain the test.
constant starting point for all mixes.
The viscosity recovery characteristics of foam mixes were deter­ 3.3. Extrudability and buildability of foam concrete
mined by following a three-stage rheological testing protocol as shown
in Fig. 4, suggested by Ramakrishnan et al. [8] and Muthukrishnan et al. To evaluate the extrudability of the foam concrete mix, five single
[7]. In the first stage, the mix was sheared at a constant shear rate of 0.6 layers with 250 mm long filaments were printed for each mix at the same
rpm, followed by shearing at a maximum rate of 60 rpm for 30 s in the feed rate and printing speed. The good extrudability of the mix is
second stage to de flocculate the mix. During the last stage, the mix was considered as the printed layers without any changes in the layer
again sheared at 0.6 rpm for 60 s. The viscosity of the mix was assessed thickness, breakage and segregation or blockage.
during the first stage and the last stage to determine the viscosity re­ The buildability of foam concrete mixes was evaluated with three
covery of the mix after complete de-flocculation in the second stage. As layers of stacked printing filaments. The buildability criteria were

4
K. Pasupathy et al. Cement and Concrete Composites 133 (2022) 104687

Fig. 3. (a) Rheological testing protocols for dynamic yield stress measurements, (b) Typical Bingham model fitting.

precision of 0.01 kN. Three samples from each mix were used for each
direction, and the average values were reported. In addition to the 3D
printed specimens, the compressive strength of mould cast specimens
was also determined at 28 days to compare the variation in the strength
due to the extrusion process. For the mould cast specimens, 50 × 50 ×
50 mm3 cubical samples were prepared in a plastic mould, and
compressive strength values were determined at the same loading rate.

3.5. Interlayer bond strength

The interlayer bond strength assessment is crucial for the foam


concrete due to the lightweight nature of filaments and the presence of
bubbles or air voids at the interface that may affect the bond strength
between the printed layers. To determine the interlayer bond strength,
30 × 30 × 40 mm3 specimens were extracted from the 3D printed ele­
ments, and the interlayer bond strength was measured by using the test
setup shown in Fig. 6. The test specimen was clamped using metallic
brackets on the top and bottom of the specimen, and the tapered claws of
the metallic brackets were aligned at the interlayer of the specimen. This
Fig. 4. Rheological testing protocols for viscosity recovery measurements. arrangement ensures the uniform stress distribution at the interlayer. In
addition, a 4 mm depth notch was made at the interlayer of the specimen
to induce the failure at the interface of the printed samples. The test was
assessed by comparing the deformation of the bottom layer due to the
conducted using the MTS machine at the displacement control mode of
weight of the stacked filaments.
1 mm/min. A similar testing technique was also used in Refs. [8,40] to
determine the interlayer bond strength properties of 3D printed
3.4. Compressive strength test concrete.

The compressive strength of 3D printed concrete specimens was


determined after 7 days and 28 days of printing. To conduct the test, 30
× 30 × 30 mm3 specimens were extracted from the 300 (L) × 60 (H) ×
30 (W) mm3 printed layer and the strength test was performed in three
different directions such as longitudinal, lateral and perpendicular di­
rections (Fig. 5) to evaluate the anisotropic behaviour of printed speci­
mens. The test was conducted at the displacement control mode of 1
mm/min using the Universal Mechanical Testing Systems (MTS) with a

Fig. 5. Schematic illustration of compressive strength test specimens extracted


from printed samples and testing directions. Fig. 6. Interlayer bond strength measurement.

5
K. Pasupathy et al. Cement and Concrete Composites 133 (2022) 104687

3.6. Apparent porosity and total porosity measurements coated samples were observed through the SEM with a working distance
(WD) of 24–27 mm.
The total porosity of the printed sample was determined using the
compaction method proposed by Gao et al. [41]. A similar testing pro­ 4. Results and discussion
cedure was also followed in past research studies to determine the
porosity measurement of traditional foam concrete [35]. To perform the 4.1. Fresh properties
test, 30 × 30 × 30 mm3 specimens were extracted from the 3D printed
elements and the volume of the specimens (Vo) was determined by 4.1.1. Foam stability in the extrusion process
measuring the linear dimensions using a Vernier calliper. The samples The stability of foam is crucial in aerated concrete due to the possible
were then crushed into powder and compressed to a solid disc using a 25 destruction of bubbles, especially during the extrusion process of 3DCP.
mm diameter die-cast set and hydraulic jack operated at a constant In addition, the instability of foam may also occur with time since the air
pressure of 16 MPa. The volume of the compressed disc (V1) was bubbles tend to coalesce due to the pressure difference [18]. To assess
measured using the Vernier calliper. By using the volume measure­ these phenomena, the foam stability of fresh foam concrete before and
ments, the total porosity (Ø) was calculated by the following formula: after the extrusion process as well as with time was evaluated by
measuring the wet density of mixes. Density measurement provides a
Vo − V1
Ø= × 100% (1) direct indication of the foam stability in aerated concrete, as the density
Vo
increase suggests the foam collapse in foam concrete [42]. Fig. 7 (a) and
The apparent porosity of 3D printed foam concrete was determined (b) illustrate the density of various mixes with time along with the
in accordance with the ASTM C830 standard. To conduct the test, 30 × density increase percentage with time. As can be seen from Fig. 7 (a) &
30 × 30 mm3 sizes of samples were extracted from 3D printed elements (b), while there is a minor increase in the fresh density with time, the
after 28 days of the curing period. Initially, the samples were oven-dried difference between the final density (i.e. fresh density at 20 mins) and
at 105 ◦ C for 24 hr, and the weight of the samples was measured (D). The the initial density before extrusion is insignificant. Therefore, it can be
samples were then subjected to a vacuum saturation process in water for said that the extrusion process has not affected the foam stability and all
24 hr, and the saturated weight of the sample was determined (W) fol­ mixes can be considered as extrudable for 3DCP applications.
lowed by measuring the suspended weight of the specimens (S). The Considering the sand-based groups, the fresh density of M1 increases
apparent porosity (P) can be determined by using the following from 1244 to 1253 kg/m3 during the extrusion and reaches 1267 kg/m3
equation: at 20 mins, showing an overall increase of 1.8% in 20 mins. The cor­
W− D responding density changes observed in the M2 mix were 840–853 kg/
P= × 100% (2) m3 followed by 870 kg/m3 at 20 mins with an increase of 3.6%. This
W− S
observation suggests that the higher the foam content will be the high
It should be noted that the apparent porosity is widely used to level of foam collapse. Similar behaviour has also been observed in EP
determine the open porosity and it does not accurately account for the based groups as well. The increasing foam content in Mix M5 displayed
closed pores in foam concrete. However, the total porosity measurement greater density variation compared to the mixes with low foam content
with the compaction method considers the open and closed pores in (M3 and M4). However, compared to sand-based groups, the variations
foam concrete. Therefore, both types of porosity measurements have are less in EP based foam concrete at a similar density. At 20 mins, the
been carried out in this study. M3, M4 and M5 mixes showed a density variation of 0.7%, 1.5% and
2.5% compared to 1.8% and 3.6% observed for M1 and M2. This
behaviour attributes to the low amount of foam used in EP based foam
3.7. Micro-morphology analysis at the interlayer
concretes due to the lightweight properties of EP.

A scanning Electron Microscopy (SEM, ZEISS Supra 40 VP) was


4.1.2. Flowability of 3D printable foam concrete mixes
employed to assess the micromorphology characteristics of 3D printed
Fig. 8 illustrates the slump-flow measurements of fresh foam con­
specimens at the interlayer. To conduct the test, a 20 mm size sample
crete mixes before and after the flow table test. It should be noted that
was extracted at the interface and samples were gold coated to a
the slump measurement indicates the buildability and shape retention
thickness of 15 nm using a K975X vacuum coating system. The gold

Fig. 7. (a) Density of fresh foam concrete after the extrusion with time; (b) Percentage of density variation.

6
K. Pasupathy et al. Cement and Concrete Composites 133 (2022) 104687

the M3, M4 and M5 mixes were 167.5 mm, 193.5 mm and 201.5 mm,
respectively. The foam content in the respective mixes was 2.5%, 5%,
and 7.5% of cement content. The decrease in flowability caused by the
EP and the flowability increment with foam content could be associated
with the variation in rheological properties of foam concrete. Therefore,
a detailed discussion on the variation in the rheological parameters with
the EP and foam contents is provided in the following section.

4.1.3. Rheological properties


Fig. 9 presents the apparent viscosity recovery results of all five foam
concrete mixes. As can be seen in Fig. 9, the apparent viscosity of the
mixes was high at stage I (when the mixes were kept at rest with the
shearing rate of 0.6 rpm), and then the viscosity values were dropped
significantly at Stage II when the mixes were sheared at the rate of 60
rpm to achieve maximum de-flocculation. It is worth mentioning here
that the deflocculation stage for 3D printable mixes involves a high
shearing rate in the order of 80 rpm. However, in this study, the shearing
rate was limited to 60 rpm to prevent the possible destruction of bubbles
at a high shearing rate. It can be seen that the mixes have been
completely deflocculated at this shearing rate as the apparent viscosity
Fig. 8. Spread diameter of fresh foam concrete mixes.
recorded at stage II was in the order of 101–102 Pa s. Once the mix is
completely deflocculated at stage II, the mixes were kept at rest in stage
characteristics of 3D printable mixes. On the other hand, the slump-flow
III. At this stage, all the mixes reached most of their initial apparent
measurement describes the pumpability properties [39]. 3D printable
viscosity. It is interesting to note that the mix with higher density
mixes with high buildability and good pumpability should exhibit low
exhibited a significantly higher viscosity than the lower density mix. For
slump value and high slump-flow diameter. From Fig. 9, all mixes
instance, the mix M3 (wet density of 1355 kg/m3) showed an apparent
showed a flow diameter of approximately 100 mm before dropping the
viscosity of ~8,000 Pa s at stage III, while the M5 (density of 800 kg/m3)
flow table, which indicates the mixes have zero slumps. It should be
showed 1,500 Pa s. This implies that the deformability of foam concrete
noted that zero slump is beneficial for 3D printing since it would have
is increased at low density and therefore, it is important to consider
good shape retention ability during the extrusion [43]. On the other
viscosity-modifying admixtures to enhance the viscosity at low density
hand, the flow diameter varies between different mixes after dropping
(especially the density range of 600 kg/m3). Our future study will
the flow table. The mixes containing sand (M1 and M2) has higher
explore the suitable additives to the ultra-lightweight foam concrete to
flowability compared to the mixes with EP components. This indicates
increase the viscosity and thixotropy effect without causing foam
that the flow properties and corresponding workability of the mix re­
instability.
duces with the incorporation of EP, whereas mixes containing sand
The comparison of EP and sand groups indicates that the EP group
showed high flowability. This could be attributed to the high specific
mixes with lower density show higher viscosity than the sand group mix
surface area of EP, which increases the water demand and reduces the
with higher density. For instance, the M1 mix with sand aggregates
flowability of the fresh mix. In addition, the foam content in the mix also
showed an apparent viscosity of ~2,500 Pa s at the density of 1244 kg/
contributed to the variation in the workability of the fresh mixes. As
m3. Meanwhile, the M4 mix with the EP aggregates showed a higher
indicated in Fig. 8, the spread diameter after the flow table drop is
apparent viscosity of ~4000 Pa s at a significant low density of 1050 kg/
increased with the increase of foam content in the mix. For instance,
m3. On the other hand, the M3 mix containing EP aggregates showed an
considering the sand groups, the spread diameter of the M1 was 213.5
apparent viscosity of ~8,000 Pa s at a density of 1355 kg/m3. However,
mm, whereas the M2 mix with higher foam content exhibited a spread
all these mixes showed 30–100 Pa s at the deflocculation stage (stage II).
diameter of 226.6 mm. In the case of EP groups, the spread diameter of
Similarly, the apparent viscosity of the M5 (wet density of 800 kg/m3)
mix was determined as ~1500 Pa s, and ~20 Pas in the de-flocculation
stage and resting period, respectively. At the same time, the M2 (wet
density of 840 kg/m3) mix from the sand group only displayed ~500 Pa
s and 7 Pa s for the corresponding stages. This implies that EP based
foam concrete has the ability to retain the shape of 3D printed elements
due to high viscosity, thus the printability of the foam mix is enhanced
[44], meanwhile ensuring better shape retention during the extrusion
process [45]. This effect is crucial for 3D concrete printing at a low
density.
The increased apparent viscosity in EP based foam concrete mixes,
compared to sand-based mixes, is due to the reduction in foam content
to achieve a similar density. As can be seen from Table 1, the M1 mix
with sand aggregates requires 5% of foam content to attain a wet density
of 1244 kg/m3. The M3 mix with EP based foam concrete only requires
2.5% of foam content to achieve a higher density of 1355 kg/m3.
Similarly, the M2 and M5 mixes use 11% and 7.5% of foam content to
achieve the density of 840 kg/m3 and 800 kg/m3 respectively. This in­
dicates that the foam content is significantly reduced when the sand is
replaced with the EP aggregates. This is mainly due to the lightweight
properties of EP aggregates, compared to sand particles and therefore,
the density of paste surrounding the air bubbles is reduced. This leads to
Fig. 9. Viscosity recovery of fresh foam concrete mixes. a reduced foam content to achieve the same density. The foam (i.e. air

7
K. Pasupathy et al. Cement and Concrete Composites 133 (2022) 104687

bubbles) has a round, flexible and smooth surface and their presence in
the mix makes it flowable leading to low viscosity. The replacement of
foam with rough, irregular and rigid EP particles makes the mix less
flowable and hence, shows high viscosity.
Fig. 10 shows the static yield stress measurements of all five mixes
immediately after the mixing process. Here, the EP group has higher
static yield stress than sand-based groups at the same density range. For
instance, the M3 mix from the EP group showed 325 Pa, while M1 from
the sand group displayed 195 Pa. Moreover, similar to the slump flow
measurement, the static yield stress of foam concrete is reduced with the
increase of foam content in the same group. Compared to the M3 mix, the
static yield stress of the M4 and M5 mixes is reduced by 54.3% and
69.2%. It is noted that the static yield stress is important for the build­
ability and shape-retention of the 3D printable mixes [46,47], where
high static yield stress of the mix implies high shape retention and
buildability. This is in agreement with the shape retention behaviour of
foam concrete reported in Section 4.2.1. As observed in Section 4.2.1,
the sand-based 3D printed specimen showed high deformability at the
bottom layer for low density, whereas the EP group filaments did not
show any such effect during the printing stage.
Fig. 11 illustrates the fitted results after applying the Bingham model Fig. 11. Bingham model fitting results of fresh foam concrete mixes.
for foam concrete to the data collected between the shearing time of
100–200s. As similar to other rheological properties, the dynamic yield the mix with EP, due to its high workability. In addition, the extruded
stress of EP group foam concrete is slightly higher than that of sand filament becomes wider when the foam content increase in the mix. The
group foam concrete at the same density levels. Compared to M1 mix increasing width of the filament can be related to the dynamic yield
from the sand group, M3 type from the EP group showed a 53% incre­ stress of the foam concrete mixes. As explained in section 4.1.3, the
ment whereas, M5 type from the EP group displayed a 99% increment dynamic yield stress is reduced with the increase of foam content in the
compared to M2 type from the sand group in the same order of density. mix. Therefore, low dynamic yield stress has caused wider elements
Similar behaviour has been observed for the plastic viscosity values as during the extrusion process. It should be noted that the diameter of the
well. Therefore, based on the observation of all the rheological proper­ nozzle used in this printing process was 30 mm and this indicates, that
ties, the incorporation of EP in foam concrete has shown beneficial ef­ the width of all types of printed filaments is slightly higher than the
fects on the printing properties, which would enhance the printability of nozzle size.
foam concrete at a low density. Similar behaviour was also observed in Fig. 13 shows three stacked printed layers of the foam concrete mixes
the foam concrete with nano clay and hydroxypropyl methylcellulose to assess the buildability by comparing the deformation of the bottom
additives [29,48]. layer. The comparison was conducted for mixes with approximately the
same density range (M1 and M3; M2 and M5). It must be underlined that
4.2. 3D concrete printing of foam concrete the buildability assessment of three stacked layers may not represent the
buildability failure criteria in these specimens. However, the purpose of
4.2.1. Extrudability and buildability of foam concrete this test is to compare the stability of printed layers during the loading
The extrudability of foam concrete is important to ensure the shape and determining the failure load due to buildability constraints is not
of the filament remains in its original state without significant de­ investigated in the current study. By considering three stacked layers in
formations or failure. Fig. 12 illustrates five single layers for each group all specimens, all groups of specimens experience the same loading at
printed at 250 mm in length. As shown, all the foam concrete mixes the bottom layer, thus allowing the comparison of deformation between
exhibited no tearing, dis-continuation and segregation during the different groups. Moreover, increasing the number of layer may lead to
extrusion process, which indicates good extrudability of foam concrete. significant deformation/collapse in lower density groups, invalidating
In addition, the average width of each filament is also marked in the the comparison between the different groups.
same figure as 41 mm, 45 mm, 37 mm, 40 mm and 41 mm for M1, M2, As per Fig. 13, there is neither distortion in the vertical direction nor
M3, M4 and M5 mixes respectively. This indicates the width of an failure in the filament was observed in all types of mixes. However,
extruded filament of foam concrete with sand (M1 and M2) is wider than compared to the EP group, the mixes from the sand group showed

Fig. 10. Static yield stress measurements of fresh foam concrete mixes.

8
K. Pasupathy et al. Cement and Concrete Composites 133 (2022) 104687

Fig. 12. Extrudability test results of the printable foam concrete; (a) M1, (b) M2, (c) M3, (d) M4 and (e) M5.

Fig. 13. Buildability of the printable foam concrete; (a) M1, (b) M3, (c) M2, (d) M5.

excessive deformation of the bottom layers; especially, at low density fresh mix governs the buildability of 3D printed layers. The buildability
(Fig. 13 (c)). This indicates that the buildability of foam concrete is of the 3D printed mix increases with the static yield stress increment
critical at low density achieved by increasing the foam content. Mean­ [49].
while, achieving the low density at reduced foam content, by using the
EP aggregates, could reduce the deformation of bottom layers. There­ 4.3. Hardened properties
fore, the replacement of sand with a lightweight aggregate like EP in­
creases the buildability and reduces the deformation and failure of the 4.3.1. Compressive strength and density of foam concrete
3D printed layers. It is worth mentioning that the static yield stress of the Fig. 14 (a) and (b) show the average compressive strength of the 3D

9
K. Pasupathy et al. Cement and Concrete Composites 133 (2022) 104687

specimens in all three directions. Similar behaviour has also been


observed in past studies [53].
Furthermore, regardless of the testing direction, the compressive
strength of the EP group was higher than the sand-based group at the
same density order for both 7 days and 28 days periods. In the density
ranges of 1130–1250 kg/m3, the 28-day compressive strength of the M3
type showed 135%, 152% and 84.5% enhancement in perpendicular,
lateral and longitudinal directions, compared to the M1 type. Similarly,
compared to M2 type, 113.5%, 124% and 96% of increments were
observed in M5 type for the corresponding directions, respectively. The
variation in compressive strength among the different groups can be
explained by the amount of foam content used in the two groups to
achieve a similar density. While higher foam content was used in the
sand-based group, the required amount of foam was reduced with the
incorporation of EP in the EP based groups. Therefore, the substantial
reduction in foam with the incorporation of EP improved the mechanical
properties [35]. In addition, increasing the foam content attributes
higher porosity and porous structure, resulting in lower strength prop­
erties than the EP group. This can be confirmed by the porosity test
results reported in Section 4.3.3. Fig. 14 also confirms the compressive
strength reduction with decreasing density of the samples between the
same group of mixes. Once again, this is due to the increased foam
content to achieve the lower density, which attributes to lower strength
parameters as stated above.
The compressive strength derived from different 3D printable foam
concrete mixes can be compared with the 3D printable foam concrete
mixes published in the past. For example, Markin et al. [31] have re­
ported a compressive strength of 4.2, 4.8 and 8.3 MPa at 35–38 days of
age for the 3D printed foam concrete with the dry density of 932, 907,
and 985 kg/m3. This comparison reveals that the incorporation EP dis­
played a similar range of compressive strength at slightly lower density.
In this study, the M4 type mix showed a compressive strength of 4.4 MPa
at 28 days with a density of 820 kg/m3. However, the direct comparison
is also difficult due to the same density levels are not available in the
past research works.

4.3.2. Interlayer bond strength of foam concrete


Fig. 14. Compressive strength and density measurement of the 3D printed Fig. 15 presents the average inter-layer bond strength of the printed
foam concrete; (a) 7 days, (b) 28 days. foam concrete samples at 28 days of the curing period. The error bar
represents + or - one standard deviation from the average strength
printed filaments at 7 days and 28 days of curing. To counteract the values. The images showing the interface of the 3D printed foam con­
anisotropic behaviour of 3D concrete printing, the strength properties crete after the bond strength test are given in Fig. 16. Fig. 16 ensures that
are measured in three directions. The 28-day compressive strength the failure occurred at the interface between two layers in all types of
properties of mould cast elements are also included in Fig. 14 (b) to samples. As per Fig. 15, the interlayer bond strength is reduced with the
compare the strength variation due to the printing effect. In addition, the reducing density between mixes. Higher bond strength was obtained for
dry density values of 3D printed filaments are also included in the same the foam concrete with higher density, and the bond strength is reduced
plots. The error bar represents + or - one standard deviation from the when the foam content is increased in the mix. Among all mixes, the M3
average strength values. It should be noted that the dry density values
are calculated based on the weight and the volume of the tested samples.
Here, the dry density of printed M1 and M2 types were 1134 kg/m3 and
653 kg/m3, whereas the EP group mixes showed 1245 kg/m3, 820 kg/
m3 and 643 kg/m3 for the mix types of M3, M4 and M5.
As can be seen in Fig. 14, regardless of the variation in mix compo­
sitions, the 3D printed elements displayed an anisotropic behaviour on
the compressive strength values in three directions at both the 7 and 28
days. In all types, the compressive strength in the longitudinal direction
(printing direction) was the highest, which is attributed to the higher
compaction in the printing direction due to the pressure from extrusion.
This is common for the 3D printed specimens as a result of the formation
of voids between adjacent filaments in the stacking process [50,51].
Similar behaviour was also identified in the previous studies as well [2,
8,52]. The comparison between the other two directions suggests that
the perpendicular direction showed high strength followed by the lateral
direction in most of the tested samples. Moreover, the 28 days
compressive strength of mould cast specimens is higher than the printed
Fig. 15. Interlayer bond strength values after 28 days.

10
K. Pasupathy et al. Cement and Concrete Composites 133 (2022) 104687

Fig. 16. Interface of the 3D printed foam concrete layers after failure, (a) M1, (b) M2,(c) M3, (d) M4 and (e) M5.

showed the highest bond strength of about 0.15 MPa, whereas the bond total porosity measurement. It is also observed that the porosity of 3D
strength was reduced by 56% and 72% for M4 and M5, respectively. The printed specimens increases with the decreasing density of the samples
bond strength of the EP group is slightly greater than the sand group at due to the increasing air content (i.e. foam content) in the low-density
similar density ranges. This could be due to the effect of foam content in mixes. Moreover, compared to sand group 3D printed specimens, the
both groups. Compared to the EP group, the foam content is high in sand EP integrated specimens showed reduced porosity at similar or higher
groups, which generated more voids at the interface. This can be density, indicating the air void reduction with the addition of EP in the
confirmed with the SEM images observed at the interface of the 3D mix. For instance, the total porosity of the M2 mix (653 kg/m3) from the
printed specimens. sand group is 59.3%, whereas the porosity of M5 (643 kg/m3) from the
Fig. 17 depicts the morphology of hardened 3D printed foam con­ EP group was 53.5%. The reason for the porosity reduction is due to the
crete at the interface observed through SEM. The SEM images dictate a variation in the foam content between the sand groups and EP groups to
clear difference in air void sizes and their distribution among different achieve a similar density. A similar observation has also been observed
samples at the interface. In the case of EP group, the pore size and the in a previous study [35]. It should be noted that the porosity of the 3D
number of pores are increased with the decreasing density. Compared to printed foam concrete influences the mechanical properties. In general,
M3 type, the amount of pore and sizes are high in M5 due to the higher the compressive strength of foam concrete reduces with the increasing
foam content. A similar observation has also been observed in the sand porosity as the air void increases, which reduces the matrix strength
group (the M2 type showed higher numbers of pores with a larger size [55]. This is consistent with the test results observed in the compressive
than the M1 type). Moreover, a comparison of pore morphology between strength and interlayer bond strength tests. For instance, the M5 mix had
the sand and EP groups at the same density reveals that the sand group higher compressive and interlayer bond strength than that of M2 type,
mixes have a large number of coarse pores. The M1 type from the sand following the same trend of porosity measurement.
group has a coarse air void distribution compared to M3 type from the EP
group, while the dry density of M1 and M3 mixes are 1134 kg/m3 and 5. Conclusion
1245 kg/m3.
A similar trend in pore size variation can be observed for the low- In lightweight concrete 3D printing, the rheological properties of
density mixes. M5 mix has fine bubble distribution (from EP group) fresh concrete are crucial for printability and hardened properties. This
compared to the coarse bubble distribution observed in M2 specimens study investigated the influence of lightweight porous aggregate on the
(from the sand group). In addition, M2 type showed that many pores are rheological properties, printability and hardened properties of foam
merged together to form large pores. It should be noted that the dry concrete for 3D printing applications. The introduction of EP as a
density of M2 and M5 mixes were almost the same (653 kg/m3 and 643 replacement for fine aggregates (i.e. sand) in 3D printable foam concrete
kg/m3); however, the pore size distributions are varied due to different substantially reduced the foam content to reach the same density, while
amounts of foam content among these two mixes. With the increasing enhancing the fresh and hardened properties of 3D printed foam con­
foam content in M2, the number of bubbles increases at the interface, crete. The following conclusions can be drawn as major outcomes of the
and these bubbles tend to coalesce and form larger voids [54]. On the present study:
other hand, the required amount of foam content to achieve the same
density in the M5 type is reduced due to the presence of EP in the mix. As • The extrusion method used in the present work does not affect the
a result, the number of bubbles is reduced, and the pore wall becomes foam stability in 3D printed foam concrete. The wet density mea­
thicker and stronger than compared to M2 type. This has resulted in the surements revealed negligible variation in density before and after
increase in interlayer bond strength as well as compressive strength the extrusion and with time.
values for M5. It is also interesting to note that there are many • The rheological parameters such as static yield stress, dynamic yield
micro-cracks observed at the pore wall in M1 type, which indicates that stress, and plastic viscosity of the EP-based foam concrete are
the pore wall is weak in the sand-based group at the same density level. considerably higher than those in sand-based foam concrete mixes at
similar densities.
4.3.3. Porosity analysis of foam concrete • Compared to the sand group specimens, the compressive strength of
Table 2 depicts the apparent porosity and total porosity of 3D printed the EP group 3D printed concrete was high in all three tested
specimens after 28 days of the curing period. Accoding to that the directions.
apparent porosity is lower than the total porosity for all groups. This • The interlayer bond strength of sand-based foam concrete is lower
could be due to the variation in the test methods. Although the samples than the EP group at similar density ranges. This is because of the
were subjected to a vacuum saturation process during the apparent reduction of effective contact area between the filaments due to the
porosity measurements, the water might not fill the closed pores or presence of more air voids.
artificial pores in the samples. Therefore, the apparent porosity mea­ • Porosity analysis and the SEM image analysis indicate that the
surement underestimates the porosity of the samples, compared to the porosity and the pore size are reduced with the incorporation of EP.

11
K. Pasupathy et al. Cement and Concrete Composites 133 (2022) 104687

Fig. 17. Microscopic images of 3D printed foam concrete interface; (a) M1, (b) M2, (c) M3, (d) M4 and (e) M5.

Increasing the foam content in the sand group led to the coalescence
Table 2
and bubble merging, which was attributed to coarse bubbles at the
Porosity values of 3D printed foam concrete.
interlayer.
Samples Apparent porosity (%) Total porosity (%)

M1 37.3 42.7
M2 47.3 59.3 Declaration of competing interest
M3 32.4 37.8
M4 36.9 45.7 The authors declare that they have no known competing financial
M5 40.4 53.5
interests or personal relationships that could have appeared to influence

12
K. Pasupathy et al. Cement and Concrete Composites 133 (2022) 104687

the work reported in this paper. [27] S. Cho, et al., Rheology and application of buoyant foam concrete for digital
fabrication, Compos. B Eng. 215 (2021), 108800.
[28] E. Lublasser, et al., Robotic application of foam concrete onto bare wall elements-
Acknowledgement Analysis, concept and robotic experiments, Autom. ConStruct. 89 (2018) 299–306.
[29] C. Liu, et al., Influence of hydroxypropyl methylcellulose and silica fume on
The authors acknowledge the Swinburne University of Technology stability, rheological properties, and printability of 3D printing foam concrete,
Cement Concr. Compos. 122 (2021), 104158.
and Australian Research Council through DE190100646 and [30] D. Falliano, et al., 3D-printable lightweight foamed concrete and comparison with
DP170103521 for supporting this work. classical foamed concrete in terms of fresh state properties and mechanical
strength, Construct. Build. Mater. 254 (2020), 119271.
[31] V. Markin, et al., 3D-printing with foam concrete: from material design and testing
References to application and sustainability, J. Build. Eng. 43 (2021), 102870.
[32] D. Falliano, et al., Stability, flexural behavior and compressive strength of ultra-
[1] F. Craveiro, et al., An automated system for 3D printing functionally graded lightweight fiber-reinforced foamed concrete with dry density lower than 100 kg/
concrete-based materials, Addit. Manuf. 33 (2020), 101146. m3, J. Build. Eng. 51 (2022), 104329.
[2] Y. Zhang, et al., Rheological and harden properties of the high-thixotropy 3D [33] Ş. Kilincarslan, M. Davraz, M. Akça, The effect of pumice as aggregate on the
printing concrete, Construct. Build. Mater. 201 (2019) 278–285. mechanical and thermal properties of foam concrete, Arabian J. Geosci. 11 (11)
[3] P. Wu, J. Wang, X. Wang, A critical review of the use of 3-D printing in the (2018) 1–6.
construction industry, Autom. ConStruct. 68 (2016) 21–31. [34] A.A. Sayadi, et al., Effects of expanded polystyrene (EPS) particles on fire
[4] B. Furet, P. Poullain, S. Garnier, 3D printing for construction based on a complex resistance, thermal conductivity and compressive strength of foamed concrete,
wall of polymer-foam and concrete, Addit. Manuf. 28 (2019) 58–64. Construct. Build. Mater. 112 (2016) 716–724.
[5] C. Gosselin, et al., Large-scale 3D printing of ultra-high performance concrete–a [35] K. Pasupathy, S. Ramakrishnan, J. Sanjayan, Enhancing the mechanical and
new processing route for architects and builders, Mater. Des. 100 (2016) 102–109. thermal properties of aerated geopolymer concrete using porous lightweight
[6] M. Hambach, M. Rutzen, D. Volkmer, Properties of 3D-printed fiber-reinforced aggregates, Construct. Build. Mater. 264 (2020), 120713.
Portland cement paste, in: 3D Concrete Printing Technology, Elsevier, 2019, [36] A. Standard, General Purpose and Blended Cements. Standard, Standard Australia,
pp. 73–113. 2010.
[7] S. Muthukrishnan, S. Ramakrishnan, J. Sanjayan, Effect of alkali reactions on the [37] Sil-Cell. Sil-Cell.
rheology of one-part 3D printable geopolymer concrete, Cement Concr. Compos. [38] A. C, Standard Test Method for Flow of Hydraulic Cement Mortar, 2001.
116 (2021), 103899. Philadelphia, PA.
[8] S. Ramakrishnan, et al., Concrete 3D printing of lightweight elements using [39] Y.W.D. Tay, Y. Qian, M.J. Tan, Printability region for 3D concrete printing using
hollow-core extrusion of filaments, Cement Concr. Compos. 123 (2021), 104220. slump and slump flow test, Compos. B Eng. 174 (2019), 106968.
[9] M. Starr, World’s first 3D-printed apartment building constructed in China, CNet [40] T. Marchment, J. Sanjayan, M. Xia, Method of enhancing interlayer bond strength
(2015) [cited 2022 28-01-2022]; Available from: www.cnet.com/news/worlds- in construction scale 3D printing with mortar by effective bond area amplification,
first-3d-printed-apartment-building-constructed-in-china. Mater. Des. 169 (2019), 107684.
[10] F. Bos, et al., Additive manufacturing of concrete in construction: potentials and [41] H. Gao, et al., A novel inorganic thermal insulation material utilizing perlite
challenges of 3D concrete printing, Virtual Phys. Prototyp. 11 (3) (2016) 209–225. tailings, Energy Build. 190 (2019) 25–33.
[11] M. Sakin, Y.C. Kiroglu, 3D printing of buildings: construction of the sustainable [42] K. Dhasindrakrishna, et al., Effect of yield stress development on the foam-stability
houses of the future by BIM, Energy Proc. 134 (2017) 702–711. of aerated geopolymer concrete, Cement Concr. Res. 138 (2020), 106233.
[12] T.A. Salet, et al., Design of a 3D printed concrete bridge by testing, Virtual Phys. [43] A.R. Arunothayan, et al., Development of 3D-printable ultra-high performance
Prototyp. 13 (3) (2018) 222–236. fiber-reinforced concrete for digital construction, Construct. Build. Mater. 257
[13] G. Bai, et al., 3D printing eco-friendly concrete containing under-utilised and waste (2020), 119546.
solids as aggregates, Cement Concr. Compos. 120 (2021), 104037. [44] X. Guo, J. Yang, G. Xiong, Influence of supplementary cementitious materials on
[14] S. Ramakrishnan, K. Pasupathy, J. Sanjayan, Synthesis and properties of thermally rheological properties of 3D printed fly ash based geopolymer, Cement Concr.
enhanced aerated geopolymer concrete using form-stable phase change composite, Compos. 114 (2020), 103820.
J. Build. Eng. 40 (2021), 102756. [45] B. Panda, et al., Improving the 3D printability of high volume fly ash mixtures via
[15] D. Weger, et al., Lightweight concrete 3D printing by selective cement the use of nano attapulgite clay, Compos. B Eng. 165 (2019) 75–83.
activation–investigation of thermal conductivity, strength and water distribution, [46] N. Roussel, Rheological requirements for printable concretes, Cement Concr. Res.
in: RILEM International Conference on Concrete and Digital Fabrication, Springer, 112 (2018) 76–85.
2020. [47] Q. Yuan, et al., A feasible method for measuring the buildability of fresh 3D
[16] A. Rahul, M. Santhanam, Evaluating the printability of concretes containing printing mortar, Construct. Build. Mater. 227 (2019), 116600.
lightweight coarse aggregates, Cement Concr. Compos. 109 (2020), 103570. [48] Z. Zhi, et al., Effect of competitive adsorption between polycarboxylate
[17] L. Wang, et al., Mechanical behaviors of 3D printed lightweight concrete structure superplasticizer and hydroxypropylmethyl cellulose on rheology of gypsum paste,
with hollow section, Arch. Civ. Mech. Eng. 20 (1) (2020) 1–17. J. Mater. Civ. Eng. 30 (7) (2018), 04018141.
[18] K. Dhasindrakrishna, et al., Collapse of fresh foam concrete: mechanisms and [49] Y. Chen, et al., Improving printability of limestone-calcined clay-based
influencing parameters, Cement Concr. Compos. 122 (2021), 104151. cementitious materials by using viscosity-modifying admixture, Cement Concr.
[19] G. Masi, et al., A comparison between different foaming methods for the synthesis Res. 132 (2020), 106040.
of light weight geopolymers, Ceram. Int. 40 (9) (2014) 13891–13902. [50] G. Ma, et al., Mechanical characterization of 3D printed anisotropic cementitious
[20] K. Ramamurthy, E.K. Nambiar, G.I.S. Ranjani, A classification of studies on material by the electromechanical transducer, Smart Mater. Struct. 27 (7) (2018),
properties of foam concrete, Cement Concr. Compos. 31 (6) (2009) 388–396. 075036.
[21] Z. Zhang, et al., Geopolymer foam concrete: an emerging material for sustainable [51] D. Asprone, et al., Rethinking reinforcement for digital fabrication with concrete,
construction, Construct. Build. Mater. 56 (2014) 113–127. Cement Concr. Res. 112 (2018) 111–121.
[22] P. Bedarf, et al., Foam 3D printing for construction: a review of applications, [52] T.T. Le, et al., Hardened properties of high-performance printing concrete, Cement
materials, and processes, Autom. ConStruct. 130 (2021), 103861. Concr. Res. 42 (3) (2012) 558–566.
[23] N. Narayanan, K. Ramamurthy, Structure and properties of aerated concrete: a [53] T. Ding, et al., Hardened properties of layered 3D printed concrete with recycled
review, Cement Concr. Compos. 22 (5) (2000) 321–329. sand, Cement Concr. Compos. 113 (2020), 103724.
[24] H.-K. Kim, J. Jeon, H.-K. Lee, Workability, and mechanical, acoustic and thermal [54] Z. Zhang, H. Wang, The pore characteristics of geopolymer foam concrete and their
properties of lightweight aggregate concrete with a high volume of entrained air, impact on the compressive strength and modulus, Front. Mater. 3 (2016) 38.
Construct. Build. Mater. 29 (2012) 193–200. [55] K. Pasupathy, S. Ramakrishnan, J. Sanjayan, Formulating eco-friendly geopolymer
[25] H. Alghamdi, N. Neithalath, Synthesis and characterization of 3D-printable foam concrete by alkali-activation of ground brick waste, J. Clean. Prod. 325
geopolymeric foams for thermally efficient building envelope materials, Cement (2021), 129180.
Concr. Compos. 104 (2019), 103377.
[26] V. Markin, et al., Material design and performance evaluation of foam concrete for
digital fabrication, Materials 12 (15) (2019) 2433.

13

You might also like