Case Studies in Construction Materials

Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

Case Studies in Construction Materials 19 (2023) e02419

Contents lists available at ScienceDirect

Case Studies in Construction Materials


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

Effect of the superplasticizer addition time on the fresh properties


of 3D printed limestone calcined clay cement (LC3) concrete
Paulo Ricardo de Matos a, *, Tuani Zat b, Marcelo Moraes Lima b,
José da Silva Andrade Neto c, Marcelo Tramontin Souza d, *, Erich David Rodríguez b
a
Departmento de Engenharia Civil, Centro de Ciências Tecnológicas (CCT), Universidade do Estado de Santa Catarina (UDESC), Joinville, Brazil
b
Departmento de Estruturas e Construção Civil, Centro de Tecnologia (CT), Universidade Federal de Santa Maria (UFSM), Santa Maria, Brazil
c
Departmento de Arquitetura e Urbanismo, Centro de Educação Superior da Região Sul (CERES), Universidade do Estado de Santa Catarina
(UDESC), Laguna, Brazil
d
Departmento de Engenharias e Computação, Universidade Estadual de Santa Cruz (UESC), Ilhéus, Brazil

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

Keywords: Although 3D concrete printing (3DCP) is considered one of the most significant innovations in the
Limestone calcined clay cement construction sector, its large cement consumption remains an environmental concern. In
LC3 response, limestone calcined clay cements (LC3) have emerged as a promising alternative to
3D printing
Portland cement (PC) due to their potential for reducing environmental impact. While previous
3DCP
Rheology
studies have shown the influence of addition timing of superplasticizer (SP) on the fresh per­
Superplasticizer formance of PC, its impact on LC3, particularly in the context of 3DCP, remains less understood.
This study assessed the effect of the SP addition time (either directly with the mixing water or
with a 10-min delay) on the rheology and printing-related properties of LC3 and PC. Rotational
rheometry and 3DP-related tests were conducted over two hours, besides hydration assessment
through thermogravimetric analysis (TGA). The results showed that adding SP directly to LC3 led
to significantly higher yield stress (1.2–2.5 times) and viscosity (14–59 % higher) values
compared to delayed SP addition within the same time frame. However, LC3 exhibited slower
rates of yield stress and viscosity increase over time in contrast to PC (with Athix structuration
rate approximately 92 % lower for LC3). These tendencies influenced printability performance.
While direct SP addition reduced PC’s open time from 100 to 30 min, it had no impact on LC3 ’s
open time, which remained at 80 min. TGA analysis yielded additional insights, highlighting that
hydrating LC3 samples required a greater free water content to achieve the same yield stress as PC
and this difference was attributed to the presence of calcined clay. These findings make contri­
butions to the technical and scientific communities by enhancing comprehension of LC3 cement
fresh behavior and solidifying 3DCP as a disruptive construction method.

1. Introduction

Digital fabrication using 3D printing (3DP) for construction materials is rapidly expanding due to its numerous benefits, including
the elimination of formworks and the ability to create complex-shaped elements [1]. However, the extensive use of cement in 3D
concrete printing (3DCP) raises significant environmental concerns [2]. Consequently, there is an urgent need to explore eco-friendly

* Corresponding authors.
E-mail addresses: paulo.matos@udesc.br (P.R. de Matos), mtsouza@uesc.br (M.T. Souza).

https://doi.org/10.1016/j.cscm.2023.e02419
Received 11 July 2023; Received in revised form 17 August 2023; Accepted 21 August 2023
Available online 22 August 2023
2214-5095/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
P.R. de Matos et al. Case Studies in Construction Materials 19 (2023) e02419

alternatives, primarily focusing on replacing Portland cement [3]. In this regard, limestone calcined clay cements (LC3) have emerged
as a highly promising solution [4].
Previous studies have documented significant advancements in the production of LC3 -based 3DP elements. Notably, Chen and
colleagues conducted comprehensive research on the influence of mix design on the fresh and hardened properties of 3DP LC3 ma­
terials. Their findings shed light on the optimization of various factors in the production process. In Refs. [5–7], the authors assessed
the effect of the grade level of calcined clay and the clinker-calcined clay-limestone ratio on the fresh and hardened performance of
3DCP. The authors observed that by increasing the content of calcined clay and limestone in the mixtures, the buildability and stiffness
development improved significantly. However, it was noted that this increase led to a reduction in flowability. On the other hand, a
high replacement of cement with calcined clay + limestone had a negative impact on both compressive strength and hydration. As for
increasing the metakaolin content in calcined clay (i.e., the grade level), it had several positive effects such as enhancing extrusion
pressures, green strength, initial setting time, and compressive strength at different ages. These findings suggest that using
higher-grade calcined clay has the potential to greatly enhance the performance of printable cementitious materials. Long et al. [8]
evaluated the effect of using silica fume on the rheological properties and compressive strength of 3DP LC3 mortars. The presence of
silica fume in 5 wt% improved the yield stress, viscosity and structural recovery of the material. Tao et al. [9] proposed the use of
perforated fly ash cenospheres as a sand replacement material to improve the fresh and hardener performance of LC3 mortars. The
authors found that incorporating cenospheres resulted in a substantial reduction of 80 % in shrinkage strain. This reduction in
shrinkage strain was attributed to the cenospheres acting as internal curing agents within the mortars. Shantanu et al. [10] demon­
strated that the rheological properties of the LC3 matrix played a more critical role in the first few minutes of printing (e.g., 5 min),
while the matrix-aggregate interaction became the key factor at later ages (>30 min). Besides mix design, the effect of printing pa­
rameters on the performance of LC3 -based 3DCP was assesses by Chen et al. [11]. The authors varied the time interval between layers
deposition (from 20 s to 10 min) and the nozzle standoff distance (from 0 to 10 mm), evaluating the effects on the mechanical strength
and microstructure of 3DCP. According to the authors, reducing the time interval between layers was found to enhance bond strength
and reduce porosity. Conversely, the impact of increasing the nozzle standoff distance on bond strength was found to be limited.
Despite the growing interest in LC3 -based 3DCP, there are still technical issues that must be tackled. One of the major challenges for
the application of LC3 is controlling its fresh properties [12]. To address this challenge, polycarboxylate-based (PCE) superplasticizers
(SP) are commonly used as effective dispersing agents in both Portland and clay-containing cementitious systems [13]. The timing of
SP addition, whether simultaneous with mixing water or delayed, plays a crucial role in the rheology and hydration of PC [14]. Von
Daake and Stephan [15] and Uchikawa et al. [16] have shown that adding SP after a few minutes delays the hydration kinetics of PC to
a greater extent. Pott et al. [17] showed that adding a high content of SP directly (with the mixing water) can induce the undesired
formation of high ettringite contents within the first minutes of hydration. Valentini et al. [18] and de Matos et al. [19] showed that the
direct SP addition retards the conversion of the intermediate short-range ordered C-(A)-S-H to long-range ordered foil-like C-(A)-S-H,
while delayed SP addition prevents such retardation. However, the precise influence of these effects on LC3 is still not comprehensively
understood.
In our previous work [19], it was established that the timing of SP addition did not have a substantial impact on the hydration of
LC3 within the first 48 h, unlike PC. However, the effect of the SP addition time on the rheology of LC3 materials, especially for 3DCP
applications, remains unknown to the best of our knowledge. Thus, the motivation of the current work is to understand the effect of SP
addition time on the fresh properties of LC3 -based 3DCP materials. For this purpose, 3DCP mixes were produced with LC3 and PC and
SP was added either simultaneously with the mixing water or with a 10-min delay; the rheological and printing-related properties of
the mixes were assessed over the first two hours.

2. Materials and methods

2.1. Materials

The experimental materials employed in this work were:


(i) commercial Portland cement (PC) with a median diameter (Dv50) measuring 10.16 µm, a density of 3080 kg/m3, and a BET
specific surface area (SSA) of 2.56 m2/g;.
(ii) limestone filler (LF) with a Dv50 of 17.03 µm, a density of 2750 kg/m3, and a BET SSA of 1.83 m2/g;.
(iii) calcined clay (CC) with a Dv50 of 8.47 µm, a density of 2440 kg/m3, and a BET SSA of 11.99 m2/g, obtained by heating a natural
kaolinitic clay at 600 ◦ C for 3 h in a laboratory furnace;.
(iv) analytic grade gypsum (Dinâmica®; >98 % purity);.
(v) natural quartz sand as aggregate with a particle size below 600 µm and a density of 2640 kg/m3 ;.
(vi) commercial PCE SP (ADVA Cast 525, GCP), with a density of 1050 kg/m3 and a solid content of 30 wt%.
The mineralogical composition of the raw materials was determined through X-ray diffraction (XRD) and Rietveld quantitative
phase analysis. The PC consisted of 90.80 wt% clinker + ACn (amorphous or crystalline non-quantified), 3.09 wt% gypsum, 0.33 wt%
bassanite, and 5.77 wt% limestone filler. The LF comprised 78.20 wt% calcite, 14.42 wt% dolomite, and 7.38 wt% minor phases (e.g.,
quartz and illite). The CC was composed of 91.11 wt% ACn and 8.89 wt% crystalline phases, such as quartz, illite, and albite. For a
more detailed and comprehensive characterization of the materials, please refer to the complete information provided in Ref. [19].

2
P.R. de Matos et al. Case Studies in Construction Materials 19 (2023) e02419

Table 1
Mix proportions of the 3DCP produced.
Mix Portland cement (g) Extra calcium sulfate (g) Calcined clay (g) Limestone filler* (g) Sand (g) Water (g) Superplasticizer (g)

LC3 57.31 1.0 30.0 11.69 100 40 0.3


PC 99.00 1.0 – – 100 30 0.3
*
Amount added to sum 15 wt% in LC3 considering the limestone in PC.

2.2. Mix design and sample preparation

Two binders were used in this work: (i) an LC3 with a proportion of (clinker + calcium sulfate):CC:LF of 55:30:15 by weight; and (ii)
a commercial PC. In our previous study [19], it was observed that replacing both Portland cement (PC) and limestone calcined clay
cements (LC3) with an additional 1 wt% of gypsum resulted in the optimal sulfate content for these compositions. Therefore, in the
present study, the same additional gypsum content was utilized in order to maintain consistency with the previous findings and ensure
comparable experimental conditions. To produce the 3DCP, the water/binder (w/b) ratios of 0.4 and 0.3 (by weight) were selected for
LC3 and PC samples, respectively, to ensure comparable applicability. A binder:sand proportion of 1:1 by weight and a SP content of
0.3 wt% over the binder weight were fixed for all samples. Table 1 shows the detailed composition of the 3DCP mixes produced.
The 3DCP samples were produced in batches of 1 L using a 5 L mortar mixer with a capacity of 285 rpm. Each sample was prepared
with the SP added at two distinct moments: (i) together with the mixing water (referred to as “direct”); and (ii) with a 10-min delay
(referred to as “delayed”). For the “direct” method, the dry materials and 90 % of the water + SP were mixed for 1 min. After 10 min,
the remaining water was added, and the sample was mixed for another 1 min. For the “delayed” method, the dry materials and 90 % of
the water were mixed for 1 min. After 10 min, the remaining water + SP were added, and the sample was mixed for another 1 min. In
both cases, the SP was pre-mixed with the respective amount of water added together with it.
For cement hydration tests (described in Section 2.3.3), pastes with the same composition were prepared without sand. The same
mixing procedures were followed using a mortar mixer with a capacity of 650 rpm in 50 ml batches. The mixtures were prepared in
plastic containers and sealed until hydration was stopped at 20 and 120 min. To stop hydration, approximately 1 g of each paste was
immersed in 50 ml of isopropanol and stirred for 30 min. Due to the soft consistency of the pastes, there was no need for a grinding
process. Subsequently, the samples were filtered using a paper filter under low vacuum for about 10 min. After filtration, the samples
were dried in an oven at 40 ◦ C for 10 min. The dried samples were then stored in a desiccator containing silica gel until testing.

2.3. Testing methods

2.3.1. Rheological tests


The rheological tests were conducted using a Viscotester iQ Air rheometer (HAAKE, Thermo Scientific) equipped with a 22.0 mm
diameter vane spindle at 25 ◦ C. The yield stress was determined by increasing the shear rate from 0 to 0.1 s− 1 in 180 s, similar to that
conducted by Refs [20,21]. Applying a constant shear rate for static yield stress measurement is a common practice for cementitious
materials [22]. However, using a single shear rate may not suit a set of samples with a broad range of yield stress values [23] as in our
case (see Section 3), so this increasing shear rate approach was preferred. The highest recorded value was taken as the (static) yield
stress. After that, flow curves were obtained by pre-shearing the sample at a shear rate of 100 s− 1 for 30 s. Subsequently, the shear rate
was gradually decreased from 100 to 0.1 s− 1 in eight steps of 20 s each. The plastic viscosity (µp) was determined by fitting the linear
portion of the flow curve (here, in the 5–100 s− 1 range) as previously conducted by Jiao et al. [24], using the Bingham model given by
Eq. (1):
τ = τ0,d + µp . γ̇ (1)

where τ is the shear stress (Pa); τ0,d is the dynamic yield stress (Pa); and γ̇ is the shear rate (s− 1).
The measurements were taken from 20 to 120 min every 20 min. For each measurement, a fresh portion from the same batch was
placed in the rheometer container 5 min before each measurement while the spindle was introduced into the sample 1 min before each
test begins; then, the static yield stress and flow curve measurements were conducted. While the flow curves provide information about
the dynamic yield stress, we specifically chose to focus on the static yield stress for the purpose of 3DPC studies, which is considered
more representative in this context [25–29]. The structural build-up of the 3DCP samples was assessed by determining the structur­
ation rate over time (Athix) through two models: the linear model proposed by Roussel and Cussigh [30] shown in Eq. (2), and the
non-linear exponential model proposed by Perrot et al. [31] shown in Eq. (3):
τ0(t) = τ0,0 + Athix .t (2)

(3)
t
τ0(t) = τ0,0 + Athix .tc (etc − 1)

where τ0(t) is the yield stress (Pa) at a given elapsed resting time t (min); τ0,0 is the yield stress prior to flocculation (Pa), i.e., at t = 0;
and tc is the characteristic time obtained experimentally (min).

3
P.R. de Matos et al. Case Studies in Construction Materials 19 (2023) e02419

Fig. 1. 3DP setup used.

Fig. 2. Shear stress vs. time curves for yield stress determination: LC3 mixes (w/b = 0.4) with direct SP addition in (a) and delayed SP addition in
(b); PC mixes (w/c = 0.3) with direct SP addition in (c) and delayed SP addition in (d). Note: measurements were recorded from 20 to 120 min every
20 min; the grey dashed line in (c) represents the shear stress capacity of the rheometer.

2.3.2. 3D printing: buildability and printability tests


3DP tests were conducted using a Potterbot Micro 10 (3D Potter) ram extrusion printer, as illustrated in Fig. 1. The nozzle diameter
and the layer height were 6.0 and 3.5 mm, respectively. The printing speed was 15 mm/s. Further information about the equipment
used can be found in Ref. [32]. The buildability of the printed structure was evaluated based on the number of consecutive layers that
could be deposited before experiencing global fresh collapse or failing to correctly deposit the subsequent layer; pictures of the
buildability test are shown in Fig. S1. The test was conducted 20 min after the contact between water and the dry materials. The
printability was assessed by printing a continuous filament every 10 min, starting at 30-min mark. The open time was defined as the

4
P.R. de Matos et al. Case Studies in Construction Materials 19 (2023) e02419

Fig. 3. Yield stress evolution of the 3DCP samples over time. (a) linear fit from Roussel and Cussigh [30]; (b) exponential fit from Perrot et al. [31].
Note: *corresponds to a torque overload measurement, before reaching the yield stress of the sample.

Table 2
Structural build-up parameters of the 3DCP samples obtained by the linear and exponential models used.
Sample PC direct PC delayed LC3 direct LC3 delayed

Model Linear Exponential Linear Exponential Linear Exponential Linear Exponential

τ0,0 (Pa) 1936 2699 58 797 347 778 -3.1 231


Athix (Pa/min) 42 10.7 28 5.5 12.6 0.9 6.7 0.4
tc (min) – 39 – 48 – 30 – 29
R2 0.94 0.99 0.91 0.99 0.82 0.99 0.82 0.99

interval during which it was possible to print a continuous filament measuring 200 mm in length, as proposed by Ma et al. [33].

2.3.3. Cement hydration through thermogravimetric analysis


Thermogravimetric analysis (TGA) was conducted in a TGA55 (TA Instruments) equipment under a N2 atmosphere, from room
temperature to 1000 ◦ C at a heating rate of 10 ◦ C/min. About 10 mg of sample and a platinum crucible were used for the analysis. The
TGA results were used to calculate the chemically bound water (CBW) content using Eq. (4). Since the portlandite content formed
within the first 120 min of hydration was considered negligible, only the CBW content was assessed in this work.
w40◦ C − w600◦ C
CBW = [ ] (4)
w600◦ C 1 + wb

where w40 ◦ C and w600 ◦ C are the weight of the sample at 40 and 600 ◦ C, respectively (in wt%).
The free water (FW) was calculated using Eq. (5), where w/b is the water/binder ratio (0.3 for PC and 0.4 for LC3) and CBW is the
value obtained from Eq. (4). Note that both CBW and FW were normalized to 100 g of fresh paste.
[ ]
w/b
FW = − CBW (5)
1 + w/b

3. Results and discussion

Fig. 2 shows the shear stress vs. time curves of the 3DCP samples over time, while Fig. 3 shows the yield stress evolution of the 3DCP
pastes over time. Both datasets were fitted with the linear model given by Eq. (2) and the exponential model given by Eq. (3). The
results obtained from the data fitting process are summarized in Table 2. Notably, the exponential model demonstrated a significant
better fit (Fig. 3), especially for the PC samples. This is supported by the higher R2 values (See Table 2) and more consistent τ0,0 values.
For example, the linear fit yielded a τ0,0 of 58 Pa while the exponential fit yielded a τ0,0 of 797 Pa for the PC delayed sample. The first
recorded τ0 value (at 20 min) was 984 Pa, aligning more closely with the exponential model fit. In addition, the linear model yielded a
negative τ0,0 for the LC3 delayed sample, which lacks physical meaning. It is generally accepted that the τ0 increases linearly within the
first ~60 min of hydration [22,34,35] while the evaluation of longer time intervals requires non-linear models, as confirmed by our
findings.
Owing that the exponential model led to more coherent structural build-up parameters, those are discussed next. Although the LC3
and PC samples exhibited different rheological properties over time, the timing of SP addition had a similar impact on their structural
build-up parameters. Direct SP addition significantly increased the “non-flocculated” yield stress (τ0,0) in both systems, reaching a
value 2.4 times higher than that of the respective delayed SP samples for both binders. Additionally, the effect of the SP addition timing
on the structuration rate (Athix) was comparable for LC3 and PC, resulting in 1.2- and 1.0-times higher values, respectively, for direct SP

5
P.R. de Matos et al. Case Studies in Construction Materials 19 (2023) e02419

Fig. 4. Correlation between the free water of pastes and the yield stress of 3DCP samples. Note: The "PC direct – 120 min" data is not shown since its
yield stress could not be measured.

addition compared to the delayed SP addition.


The role of the Athix parameter on the yield stress evolution becomes evident when comparing the PC delayed and LC3 direct mixes
(see Fig. 3b). Initially, both samples exhibited similar τ0 values at shorter resting periods (20 and 40 min). However, from 60 min
onwards, the PC delayed displayed ~60 % higher τ0 values than LC3 direct sample. This notable difference can be attributed to the PC
sample having a 5-times higher Athix value (Table 2). Finally, the characteristic time (tc) of the samples fell within the range of
29–48 min, in line with the findings reported by Rahul et al. [36] and Perrot et al. [31] for 3DCP samples.
The disparity in τ0,0 values between samples with direct or delayed SP addition can be attributed to the interaction between the
chemical admixture molecules and the cement particles, as well as the hydrated products. Flatt and Houst [14] and Plank et al. [37]
reported that a portion of the SP added can be intercalated into hydration products, especially AFm (a phase quickly formed when
aluminates come into contact with water). This results in the formation of an organo-mineral phase (OMP) around the cement particles
during the early stages of hydration, which effectively traps the admixture molecules and hampers their dispersion. In this situation,
the SP is less effective. This phenomenon is more likely to occur with direct SP addition since it requires the simultaneous presence of
calcium sulfates, aluminates, and SP polymers. Besides, this process is particularly significant in the initial minutes when calcium
sulfoaluminate hydrates have the highest precipitation rate. In turn, when SP is added with a delay (e.g., 10 min), most of the hy­
dration products are already ettringite crystals by the time the polymer is added. Recent findings by Pott et al. [17] through XRD and
scanning electron microscopy measurements in hydrating PC samples support this statement. They observed a substantial presence of
AFm at 10 and 30 min with direct SP addition, while no AFm was observed up to 60 min with delayed SP samples, only observing
ettringite on the cement particles.
Plank et al. [37] suggested that the formation of OMP is more likely to occur in undersulfated systems where AFm precipitation is
favored over ettringite. However, for such OMP formation occur, the AFm precipitation must take place at very early ages, within the
first couple of minutes. In a previous work [38] we demonstrated that the sulfate source and its solubility play a more significant role in
the very early stages of hydration when the organo-mineral phase is formed, whereas the total sulfate content becomes more relevant
for renewed C3A dissolution. It is important to note that the sulfate contents of both the LC3 and PC systems evaluated in this study
were previously optimized. Finally, the complexation of C-S-H and polymers has been previously reported in [39,40], which could also
contribute to the trapping of a portion of SP. The precipitation of C-(A)-S-H is not expected to occur within the first two hours of
hydration in the systems assessed in this study [19].
Other studies have also investigated the impact of SP addition timing on the rheology of cement pastes. Von Daake and Stephan
[15] conducted a study specifically focused on investigating the impact of addition timing on the rheology and early hydration of PC
pastes with different superplasticizers (melamine, naphthalene and PCE). The authors reported that the saturation dosage of the SPs
sharply decreased by 40–60 % with delayed SP addition. This decrease indicated a more efficient dispersion of the particles achieved
through the chemical admixtures. Based on the findings mentioned above, it is coherent that the delayed SP addition in the samples
examined in this study resulted in higher flowability (i.e., lower yield stress) compared to direct SP addition.
Fig. 4 shows the correlation between the CBW obtained by TGA and the yield stress measured by rheometry. The free water was
determined as the difference between the initial amount of water and the CBW of the paste. The specific CBW and free water values can
be found in Table S1 (see Supplementary Material). The correlation coefficient R2 did not indicate a strong correlation, which could be
attributed to several factors. Firstly, the TGA was conducted in pastes while 3D printing was performed in mortars. Additionally, the
relatively low number of samples tested and the inherent variability in the testing process could have contributed to the observed weak
correlation. Nonetheless, certain trends were observed and found to be consistent. The LC3 samples exhibited a higher demand for free

6
P.R. de Matos et al. Case Studies in Construction Materials 19 (2023) e02419

Fig. 5. Plastic viscosity of the 3DCP samples over time.

Fig. 6. Printing-related results of the 3DCP samples.

water to achieved a given yield stress compared to the PC samples. This finding aligns with previous reports [6,13,41] which show that
the very high SSA of calcined clays (here, 11.99 m2/g for CC vs. 1.83–2.56 m2/g for the other powder materials) leads to an increased
water requirement for achieving the target flowability. These results are also consistent with those reported by Jakob et al. [42], who
demonstrated that the gradual consumption of free water during cement hydration contributes to the loss of flowability over time,
along with flocculation [43]. It is worth warning that additional techniques are desirable to confirm the phase assemblage (and CBW)
of the samples, such as X-ray diffraction.
Fig. 5 shows the variation in plastic viscosity of the 3DCP pastes over time. The general trend for viscosity among the samples was
consistent with that of the yield stress. Specifically, the PC samples exhibited higher viscosity values compared to the LC3 samples, and
the direct SP addition led to higher viscosities when comparing samples produced with the same binder. In addition, the viscosity
values felt within the range of 4.0 ± 1.3 Pa.s for the LC3 mixes and 6.5 ± 2.2 Pa.s for the PC mixes. These values are comparable to
those reported by Zhang et al. [44] and Long et al. [8] for 3DCP samples. Although the presence of calcined clay (with high SSA) may
increase the viscosity of the mixture, the higher w/b ratio used for the LC3 samples compensated for it. In turn, the viscosity growth
over time did not follow the exponential trend of the yield stress growth especially for the LC3 mixes. Instead, the viscosity exhibited a
steadier growth. These results agree with the time-dependent plastic viscosity values reported by Long et al. [8] for PC and LC3 based
3DCP. The slower viscosity growth over time of LC3 becomes evident when comparing the LC3 direct and PC delayed samples. At
20 min, both samples had equivalent viscosities, but the latter displayed ~40 % higher viscosity values than the former for longer
resting times.
Fig. 6 shows the results of the 3D printing-related test, i.e., the buildability and printability of the 3DCP samples Fig. 7 provides
visual illustration of the samples during the printability test, while this illustration is shown in Fig. S1 for the buildability test.

7
P.R. de Matos et al. Case Studies in Construction Materials 19 (2023) e02419

Fig. 7. Photographs of 3D printed filaments of samples in the printability test.

Regarding the LC3 samples, the timing of addition time did not have a significant impact on their printability. In turn, for the PC
samples, the direct SP addition significantly reduced the printability compared to delayed addition. from 90 to 30 min, When both
samples achieved a viscosity of ~7 Pa.s (as shown in Fig. 5), the direct SP addition reduced the printability from 90 to 30 min. On the
other hand, the specific viscosity threshold for printability was not clearly defined for the LC3 samples. From a practical standpoint, the
direct SP addition in LC3 3DCP was advantageous as it effectively doubled the buildability of the sample without harming the
printability (open time kept at 70 min). As for PC, direct SP addition further improved the buildability (from 10 to 34 layers) but came
at the cost of reducing the open time to one-third of the value found for delayed SP addition.
One can expect that increasing the materials’ yield stress would increase its buildability (i.e., the maximum achievable height that
can be printed with it), which did not occur to our samples. The plastic collapse theoretically occurs when the compressive load (σ),
given by σ = ρgh (where ρ is the material’s density, g is the acceleration due to gravity, and h is the height of the printed object),
overcomes the material’s yield stress. However, recent reports [45] showed divergences of 17–55 %. between the number of layers
predicted from rheometry results and measured in printed elements. In this scenario, a collective effort is being made to define what
is/are the most suitable test/s to assess the fresh properties of 3DCP, and ideally predicting its buildability [46].

4. Conclusion

In this work, the effect of the SP addition timing on the rheological and printing-related properties of LC3 and PC 3DCP was
investigated. It was found that the direct addition of SP resulted in higher yield stress and viscosity values compared to delayed SP
addition, which was observed for both LC3 and PC samples. However, LC3 exhibited lower yield stress and viscosity increase rates over
time compared to PC. Additional care must be taken in interpreting the results since PC and LC3 systems required different water
contents to achieve comparable applicability.
The SP addition time also influenced on the printability performance of the 3DCP samples. In PC samples, the direct SP addition
reduced the open time from 90 to 30 min in PC compared to delayed addition, while did not affect it for LC3 (which remained at
70 min). Regarding buildability, the direct addition of SP increased the number of successive layers compared to delayed addition. For
LC3 , the direct SP addition resulted in 20 successive layers, whereas for PC, it increased the number of layers to 34. Coupling TGA and
rheometry, it was observed that LC3 samples required a greater content of free water than PC samples to achieve a given yield stress.
This observation is consistent with the higher water demand of LC3 due to the presence of calcined clay.

Declaration of Competing Interest

The author declare that he has no known competing financial interests or personal relationships that could have appeared to in­
fluence the work reported in this paper.

Data Availability

Data will be made available on request.

8
P.R. de Matos et al. Case Studies in Construction Materials 19 (2023) e02419

Acknowledgments

The authors acknowledge the financial support from the following Brazilian governmental agencies: National Council for Scientific
and Technological Development – CNPq (grant number 403563/2021-6) and Coordination for the Improvement of Higher Education
Personnel – CAPES. The Multi-User Facility Infrastructure CMU/CCT/UDESC and Ms. Vanessa B. dos Santos are acknowledged for the
TGA.

Appendix A. Supporting information

Supplementary data associated with this article can be found in the online version at doi:10.1016/j.cscm.2023.e02419.

References

[1] R.J. Flatt, T. Wangler, On sustainability and digital fabrication with concrete, Cem. Concr. Res. 158 (2022), 106837, https://doi.org/10.1016/j.
cemconres.2022.106837.
[2] M.P. Tinoco, É.M. de Mendonça, L.I.C. Fernandez, L.R. Caldas, O.A.M. Reales, R.D.Toledo Filho, Life cycle assessment (LCA) and environmental sustainability of
cementitious materials for 3D concrete printing: a systematic literature review, J. Build. Eng. 52 (2022), https://doi.org/10.1016/j.jobe.2022.104456.
[3] I.H. Shah, S.A. Miller, D. Jiang, R.J. Myers, Cement substitution with secondary materials can reduce annual global CO2 emissions by up to 1.3 gigatons, Nat.
Commun. 13 (2022) 1–11, https://doi.org/10.1038/s41467-022-33289-7.
[4] K. Scrivener, F. Martirena, S. Bishnoi, S. Maity, Calcined clay limestone cements (LC3), Cem. Concr. Res. 114 (2018) 49–56, https://doi.org/10.1016/j.
cemconres.2017.08.017.
[5] Y. Chen, S. He, Y. Zhang, Z. Wan, O. Çopuroğlu, E. Schlangen, 3D printing of calcined clay-limestone-based cementitious materials, Cem. Concr. Res. 149
(2021), https://doi.org/10.1016/j.cemconres.2021.106553.
[6] Y. Chen, C. Romero Rodriguez, Z. Li, B. Chen, O. Çopuroğlu, E. Schlangen, Effect of different grade levels of calcined clays on fresh and hardened properties of
ternary-blended cementitious materials for 3D printing, Cem. Concr. Compos. 114 (2020), https://doi.org/10.1016/j.cemconcomp.2020.103708.
[7] Y. Chen, Z. Li, S.C. Figueiredo, O. Çopuroğlu, F. Veer, E. Schlangen, Limestone and calcined clay-based sustainable cementitious materials for 3D concrete
printing: a fundamental study of extrudability and early-age strength development, Appl. Sci. 9 (2019), https://doi.org/10.3390/app9091809.
[8] W.J. Long, C. Lin, J.L. Tao, T.H. Ye, Y. Fang, Printability and particle packing of 3D-printable limestone calcined clay cement composites, Constr. Build. Mater.
282 (2021), 122647, https://doi.org/10.1016/j.conbuildmat.2021.122647.
[9] J.L. Tao, C. Lin, Q.L. Luo, W.J. Long, S.Y. Zheng, C.Y. Hong, Leveraging internal curing effect of fly ash cenosphere for alleviating autogenous shrinkage in 3D
printing, Constr. Build. Mater. 346 (2022), 128247, https://doi.org/10.1016/j.conbuildmat.2022.128247.
[10] B. Shantanu, J. Smrati, S. Manu, Criticality of binder-aggregate interaction for buildability of 3D printed concrete containing limestone calcined clay, Cem.
Concr. Compos. 136 (2023), 104853, https://doi.org/10.1016/j.cemconcomp.2022.104853.
[11] Y. Chen, K. Jansen, H. Zhang, C. Romero Rodriguez, Y. Gan, O. Çopuroğlu, E. Schlangen, Effect of printing parameters on interlayer bond strength of 3D printed
limestone-calcined clay-based cementitious materials: an experimental and numerical study, Constr. Build. Mater. 262 (2020), https://doi.org/10.1016/j.
conbuildmat.2020.120094.
[12] M. Sharma, S. Bishnoi, F. Martirena, K. Scrivener, Limestone calcined clay cement and concrete: a state-of-the-art review, Cem. Concr. Res. 149 (2021), 106564,
https://doi.org/10.1016/j.cemconres.2021.106564.
[13] R. Sposito, M. Maier, N. Beuntner, K. Thienel, Physical and mineralogical properties of calcined common clays as SCM and their impact on flow resistance and
demand for superplasticizer, Cem. Concr. Res. 154 (2022), 106743, https://doi.org/10.1016/j.cemconres.2022.106743.
[14] R.J. Flatt, Y.F. Houst, A simplified view on chemical effects perturbing the action of superplasticizers, Cem. Concr. Res. 31 (2001) 1169–1176, https://doi.org/
10.1016/S0008-8846(01)00534-8.
[15] H. Von Daake, D. Stephan, Setting of cement with controlled superplasticizer addition monitored by ultrasonic measurements and calorimetry, Cem. Concr.
Compos. 66 (2016) 24–37, https://doi.org/10.1016/j.cemconcomp.2015.11.004.
[16] H. Uchikawa, D. Sawaki, S. Hanehara, Influence of kind and added timing of organic admixture on the composition, structure and property of fresh cement
paste, Cem. Concr. Res. 25 (1995) 353–364, https://doi.org/10.1016/0008-8846(95)00021-6.
[17] U. Pott, C. Jakob, D. Jansen, J. Neubauer, D. Stephan, Investigation of the incompatibilities of cement and superplasticizers and their influence on the
rheological behavior, Materials 13 (2020), https://doi.org/10.3390/ma13040977.
[18] L. Valentini, M. Favero, M.C. Dalconi, V. Russo, G. Ferrari, G. Artioli, Kinetic model of calcium-silicate hydrate nucleation and growth in the presence of PCE
superplasticizers, Cryst. Growth Des. 16 (2016) 646–654, https://doi.org/10.1021/acs.cgd.5b01127.
[19] P.R. de Matos, J.S. Andrade Neto, R.D. Sakata, C.E.M. Campos, A.P. Kirchheim, E.D. Rodríguez, Effect of superplasticizer addition time and metakaolin source
on the early-age hydration of limestone calcined clay cement (LC3), Mater. Struct. Constr. 55 (2022), https://doi.org/10.1617/s11527-022-02049-w.
[20] M.P. Tinoco, L. Gouvêa, K. de Cássia Magalhães Martins, R. Dias Toledo Filho, O. Aurelio Mendoza Reales, The use of rice husk particles to adjust the rheological
properties of 3D printable cementitious composites through water sorption, Constr. Build. Mater. 365 (2023), https://doi.org/10.1016/j.
conbuildmat.2022.130046.
[21] M.D.M. Paiva, L.D. Fonseca Rocha, L.I. Castrillon Fernandez, R.D. Toledo Filho, E.C.C.M. Silva, R. Neumann, O.A. Mendoza Reales, Rheological properties of
metakaolin-based geopolymers for three-dimensional printing of structures, Acids Mater. J. 118 (2021) 177–187, https://doi.org/10.14359/51733122.
[22] Q. Yuan, D. Zhou, K.H. Khayat, D. Feys, C. Shi, On the measurement of evolution of structural build-up of cement paste with time by static yield stress test vs.
small amplitude oscillatory shear test, Cem. Concr. Res. 99 (2017) 183–189, https://doi.org/10.1016/j.cemconres.2017.05.014.
[23] P.R. de Matos, R. Pilar, C.A. Casagrande, P.J.P. Gleize, F. Pelisser, Comparison between methods for determining the yield stress of cement pastes, J. Braz. Soc.
Mech. Sci. Eng. 42 (2020) 24, https://doi.org/10.1007/s40430-019-2111-2.
[24] D. Jiao, C. Shi, Q. Yuan, Time-dependent rheological behavior of cementitious paste under continuous shear mixing, Constr. Build. Mater. 226 (2019) 591–600,
https://doi.org/10.1016/j.conbuildmat.2019.07.316.
[25] D. Marchon, S. Kawashima, H. Bessaies-Bey, S. Mantellato, S. Ng, Hydration and rheology control of concrete for digital fabrication: Potential admixtures and
cement chemistry, Cem. Concr. Res. 112 (2018) 96–110, https://doi.org/10.1016/j.cemconres.2018.05.014.
[26] B. Panda, C. Unluer, M.J. Tan, Extrusion and rheology characterization of geopolymer nanocomposites used in 3D printing, Compos. Part B Eng. 176 (2019),
107290, https://doi.org/10.1016/j.compositesb.2019.107290.
[27] M.T. Souza, I.M. Ferreira, E.G. de Moraes, L. Senff, A.P.N. de Oliveira, 3D printed concrete for large-scale buildings: an overview of rheology, printing
parameters, chemical admixtures, reinforcements, and economic and environmental prospects, J. Build. Eng. 32 (2020), 101833, https://doi.org/10.1016/j.
jobe.2020.101833.
[28] M.T. Souza, I.M. Ferreira, E.G. de Moraes, L. Senff, S. Arcaro, J.R.C. Pessôa, M.J. Ribeiro, A.P.N. de Oliveira, Role of chemical admixtures on 3D printed Portland
cement: Assessing rheology and buildability, Constr. Build. Mater. 314 (2022), 125666, https://doi.org/10.1016/j.conbuildmat.2021.125666.

9
P.R. de Matos et al. Case Studies in Construction Materials 19 (2023) e02419

[29] M.T. Souza, L. Simão, E.G. de Moraes, L. Senff, J.R.C. Pessôa, M.J. Ribeiro, A.P.N. de Oliveira, Role of temperature in 3D printed geopolymers: evaluating
rheology and buildability, Mater. Lett. 293 (2021), 129680, https://doi.org/10.1016/j.matlet.2021.129680.
[30] N. Roussel, F. Cussigh, Distinct-layer casting of SCC: The mechanical consequences of thixotropy, Cem. Concr. Res. 38 (2008) 624–632, https://doi.org/
10.1016/j.cemconres.2007.09.023.
[31] A. Perrot, D. Rangeard, A. Pierre, Structural built-up of cement-based materials used for 3D-printing extrusion techniques, Mater. Struct. Constr. 49 (2016)
1213–1220, https://doi.org/10.1617/s11527-015-0571-0.
[32] P. de Matos, T. Zat, K. Corazza, E. Fensterseifer, R. Sakata, G. Mohamad, E. Rodríguez, Effect of TiO2 nanoparticles on the fresh performance of 3D-printed
cementitious materials, Materials 15 (2022) 1–18, https://doi.org/10.3390/ma15113896.
[33] G. Ma, Z. Li, L. Wang, Printable properties of cementitious material containing copper tailings for extrusion based 3D printing, Constr. Build. Mater. 162 (2018)
613–627, https://doi.org/10.1016/j.conbuildmat.2017.12.051.
[34] U. Pott, D. Stephan, Penetration test as a fast method to determine yield stress and structural build-up for 3D printing of cementitious materials, Cem. Concr.
Compos. 121 (2021), 104066, https://doi.org/10.1016/j.cemconcomp.2021.104066.
[35] U. Pott, C. Ehm, C. Jakob, D. Stephan, Investigation of the Early Cement Hydration with a New Penetration Test, Rheometry and In-Situ XRD, RILEM Bookseries.
23 (2020) 246–255. https://doi.org/10.1007/978–3-030–22566-7_29.
[36] A.V. Rahul, M. Santhanam, H. Meena, Z. Ghani, 3D printable concrete: mixture design and test methods, Cem. Concr. Compos. 97 (2019) 13–23, https://doi.
org/10.1016/j.cemconcomp.2018.12.014.
[37] J. Plank, D. Zhimin, H. Keller, F. v Hössle, W. Seidl, Fundamental mechanisms for polycarboxylate intercalation into C3A hydrate phases and the role of sulfate
present in cement, Cem. Concr. Res. 40 (2010) 45–57, https://doi.org/10.1016/j.cemconres.2009.08.013.
[38] J.S. Andrade Neto, P.R. de Matos, A.G. de la Torre, C.E.M. Campos, S.M. Torres, P.J.M. Monteiro, A.P. Kirchheim, Hydration and interactions between pure and
doped C3S and C3A in the presence of different calcium sulfates, Cem. Concr. Res. 159 (2022), 106893, https://doi.org/10.1016/j.cemconres.2022.106893.
[39] H. Matsuyama, J.F. Young, The formation of C-S-H / polymer complexes by hydration of reactive β -dicalcium silicate, Concr. Sci. Eng. 1 (1999) 66–75.
[40] A. Moshiri, D. Stefaniuk, S.K. Smith, A. Morshedifard, D.F. Rodrigues, M.J.A. Qomi, K.J. Krakowiak, Structure and morphology of calcium-silicate-hydrates
cross-linked with dipodal organosilanes, Cem. Concr. Res. 133 (2020), 106076, https://doi.org/10.1016/j.cemconres.2020.106076.
[41] T.R. Muzenda, P. Hou, S. Kawashima, T. Sui, X. Cheng, The role of limestone and calcined clay on the rheological properties of LC3, Cem. Concr. Compos. 107
(2020), 103516, https://doi.org/10.1016/j.cemconcomp.2020.103516.
[42] C. Jakob, D. Jansen, N. Ukrainczyk, E. Koenders, U. Pott, D. Stephan, J. Neubauer, Relating ettringite formation and rheological changes during the initial
cement hydration: a comparative study applying XRD analysis, rheological measurements and modeling, Materials 12 (2019), https://doi.org/10.3390/
ma12182957.
[43] N. Roussel, G. Ovarlez, S. Garrault, C. Brumaud, The origins of thixotropy of fresh cement pastes, Cem. Concr. Res. 42 (2012) 148–157, https://doi.org/
10.1016/j.cemconres.2011.09.004.
[44] Y. Zhang, Y. Zhang, G. Liu, Y. Yang, M. Wu, B. Pang, Fresh properties of a novel 3D printing concrete ink, Constr. Build. Mater. 174 (2018) 263–271, https://doi.
org/10.1016/j.conbuildmat.2018.04.115.
[45] I. Ivanova, E. Ivaniuk, S. Bisetti, V.N. Nerella, V. Mechtcherine, Comparison between methods for indirect assessment of buildability in fresh 3D printed mortar
and concrete, Unpublished 156 (2021) 1–28, https://doi.org/10.1016/j.cemconres.2022.106764.
[46] R. Nicolas, B. Richard, D. Nicolas, I. Irina, K.J. Temitope, L. Dirk, M. Viktor, M. Romain, P. Arnaud, P. Ursula, R. Lex, S. Dietmar, W. Timothy, W. Rob,
Z. Wenqiang, Assessing the fresh properties of printable cement-based materials: High potential tests for quality control, Cem. Concr. Res. 158 (2022), 106836,
https://doi.org/10.1016/j.cemconres.2022.106836.

10

You might also like