Construction and Building Materials 145 (2017) 639–647

Contents lists available at ScienceDirect

Construction and Building Materials

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

Cementitious materials for construction-scale 3D printing: Laboratory

testing of fresh printing mixture
Ali Kazemian a,b,⇑ , Xiao Yuan c, Evan Cochran a, Behrokh Khoshnevis a,c
a
The Sonny Astani Department of Civil and Environmental Engineering, University of Southern California, Los Angeles, CA, USA
b
Department of Computer Science, University of Southern California, Los Angeles, CA, USA
c
Department of Industrial and Systems Engineering, University of Southern California, Los Angeles, CA, USA

highlights

● A framework for performance-based laboratory testing of printing concrete was proposed.

● A laboratory-scale linear concrete printer was built for experimental study of printing mixtures.
● Four printing mixtures were designed to study effects of Nano-clay, silica fume and fiber inclusion.
● Layer settlement and cylinder stability tests were developed to study shape stability.
● Inclusion of silica fume and Nano-clay enhance shape stability of fresh printing mixture.

articleinfo abstract

Article history: In this study, a framework for performance-based laboratory testing of cementitious mixtures for
Received 15 January 2017 construction-scale 3D printing is developed, where workability of a fresh ‘‘printing mixture” is studied
Received in revised form 31 March 2017 in terms of print quality, shape stability, and printability window. Print quality is described using mea-
Accepted 5 April 2017
sures of surface quality and dimensions of printed layers. Details of two proposed test methods for eval -
uation of shape stability, namely, ‘‘layer settlement” and ‘‘cylinder stability” are also provid ed.
Experimental study of four different mixtures revealed that inclusion of silica fume and Nano -clay signif-
Keywords:
icantly enhance shape stability. The results of five conventional test methods, as well as four proposed
Construction-scale 3D printing
tests are used to discuss the performance of mixtures.
Concrete
Additive manufacturing © 2017 Elsevier Ltd. All rights reserved.
Shape stability
Cementitious material
Contour crafting

1. Introduction scaling up additive manufacturing techniques for automated build-

ing construction has been topic of research for several years. Addi-
After facing criticism for limited adoption of cutting edge tech- tive manufacturing is defined as ‘‘a process of joining materials to
nologies for several decades, the construction industry is undergo- make objects from 3D model data, usually layer upon layer, as
ing some profound changes now. Building Information Modelling opposed to subtractive manufacturing methodologies [10]”. It
(BIM) is gradually becoming ubiquitous, in which the essential should be noted that additive manufacturing technologies have
building design and project data are generated and managed in previously been used for concept modelling in architecture [11].
digital format throughout the building’s life cycle [1–5]. Sensing However, full-scale automated building construction would be a
automation technology is also being used in buildings to improve revolution in construction industry. A well-developed automated
energy efficiency and occupants’ satisfaction [6,7]. Use of drone construction process presents numerous advantages including
technology in construction sites is also reported, where drones design freedom, superior construction speed, and higher degree
are adopted for applications such as safety inspection [8] and 3D of customization. Some of the well-known processes for
modelling of construction sites [9]. Moreover, the novel idea of construction-scale 3D printing include Contour Crafting, Concrete
Printing, and D-shape, which are invented in the U.S., U.K., and
Italy, respectively [12,13].
⇑ Corresponding author at: The Sonny Astani Department of Civil and Environ-
Invented by Dr. Behrokh Khoshnevis of University of Southern
mental Engineering, University of Southern California, Los Angeles, CA, USA.
E-mail address: [email protected] (A. Kazemian).
California, Contour Crafting (CC) is an additive fabrication

http://dx.doi.org/10.1016/j.conbuildmat.2017.04.015
0950-0618/© 2017 Elsevier Ltd. All rights reserved.
640 A. Kazemian et al. / Construction and Building Materials 145 (2017) 639–647

technology that uses computer control to exploit the surface- experimentally measured tf values were highly correlated with val-
forming capability of troweling in order to create smooth and accu- ues calculated based on proposed expressions. In another study, Le
rate planar and free-form surfaces out of extruded materials [14]. et al. [23] developed a high-performance fiber reinforced printing
CC is commonly recognized as the first viable construction-scale mixture for concrete printing process. A high cementitious materi-
additive manufacturing process for building construction [15–17]. als content of 827 kg/m3 and low water/cementitious materials
Some of important advantages of CC include unprecedented surface ratio of 0.26 was employed to achieve desirable fresh state proper-
quality of printed elements, increased fabrication rate, and a vast ties and a 28-day compressive strength of 110 MPa. A shear vane
choice of materials [18]. The prototype CC machine has work envel- apparatus was used to describe workability of printing concrete,
ope dimensions of 5 m×8 m ×3 m, corresponding to a 120 m 2 and 0.3–0.9 kPa was suggested as the acceptable shear strength
printing zone, and is presented in Fig. 1-a. Furthermore, in the range for a printable mixture. Finally, the suitability of proposed
long-term plan to colonize the moon and Mars, CC technology has mixture was validated by construction of a 2 m-long concrete
been considered as a viable way for building immediate infrastruc- bench, consisting of 128 filaments of 6 mm thickness.
tures on the surface of these planetary objects [19,20]. While few past studies have provided an initial understanding
Another well-known construction-scale 3D printing system of some of the desirable properties of printing concrete, extensive
which was developed by researchers at Loughborough University research and experimental data is still needed. In specific, charac-
is called ‘‘concrete printing”. The concrete printing machine is terization of fresh state behavior of a printing mixture requires
composed of a 5.4 m ×4.4 m ×5.4 m frame and a printing head deeper investigation. It seems that the traditional definition for
on a mobile horizontal-beam, which moves in the y and z direc- workability of fresh concrete as ‘‘a measure of ease by which fresh
tions while the printing head only moves in the x-direction [21]. concrete can be placed and compacted [25]” is not accurate enough
The printing head is able to move up to 83 mm/s depending on and new measures should be developed for describing the fresh
the curvature of a printing path. Concrete printing process mainly state behavior of a printing mixture.
includes data preparation, material preparation, delivering, and
printing. Interestingly, several companies also have been recently
founded based on the idea of large-scale 3D printing for construc- 2. Research objectives
tion. For instance, Apis CorTM is a newly established company which
has developed a crane concrete printer with 132 m 2 printing zone Perceiving ‘‘printing concrete” as the latest special concrete,
(Fig. 1-b). Based on manufacturer, the compact dimensions of prin- there exists no relevant guideline or proposed procedure for eval-
ter are 4 m ×1.6 m 1.5×m and it weighs 2 tons [22]. uating mixtures and new materials, or any set of well-defined
A review of ongoing projects and research works reveals that acceptance criteria for this type of concrete. As mentioned before,
Portland cement concrete is the most viable option as the material few previous studies have focused on specific properties of printing
for widespread use in automated construction processes in near mixtures such as shape stability (also called shape retention and
future [22–24,16]. The major reasons are the unique fresh and green strength) [24,23,26,27]. However, a comprehensive list of
hardened properties of concrete which are well-understood, as performance requirements and test methods for a printing mixture
well as an extensive variety of readily available admixtures to cus- has not yet been suggested.
tomize its performance. The goal of current study is to present and examine a frame-
Limited research has been carried out on properties of printing work for performance-based laboratory testing and evaluation of
concrete. In 2016, Perrot et al. [24] studied the time-dependent printing mixtures. It should be noted that only fresh properties
structural build-up of cementitious materials in layer-wise con- of a printing mixture are considered herein, while further research
struction. This is important since during the layer-by-layer con- is needed to investigate the structural requirements for hardened
struction process, the previously deposited layers should be able printing mixture. Development of a comprehensive framework
to withstand the load caused by following layers. Based on com- for laboratory testing of printing concrete would be a starting point
parison of vertical stress acting on the first printed layer with the for systematic investigation on this special concrete by researchers,
critical stress related to the plastic deformation, a theoretical and a basis for future specifications and guidelines. Establishing
framework was proposed. Assuming linear evolution of yield stress universal acceptance criteria for printing concrete would be possi-
over time, these researchers defined a critical failure time (tf) as a ble only after a large number of relevant studies have been carried
function of concrete specific weight, concrete yield stress with no out, and a reasonable amount of data is available on performance
time at rest, structuration rate, construction rate and a geometric of different printing mixtures used in actual construction projects.
factor (ageom). Finally, layer-wise construction of a 70 mm diame- In this paper, the proposed framework for laboratory testing of
ter column with building rates of 1.1–6.2 m/h was used to validate fresh printing concrete is initially introduced and some relevant
the findings. Except for the smallest building rate, 1.13 m/h, the details are provided. Then, an experimental program which was

Fig. 1. (a) Contour Crafting machine (b) Crane printer developed by Apis Cor TM [22].
 A. Kazemian et al. / Construction and Building Materials 145 (2017) 639–647 641

carried out to elaborate on the proposed framework, is discussed. longest time when a mixture can be printed with acceptable print
The mixture proportions of four printing mixtures are also pre- quality) and blockage limit (the longest time when a mixture can
sented. These mixtures were designed to study effects of Nano- remain in the nozzle before the blocking happens) should be mea-
clay, silica fume and fiber inclusion. Finally, the results of several sured and reported for a specific mixture. Finally, when the labora-
conventional test methods, as well as proposed tests, are presented tory testing is aimed at developing a mixture for a specific
and a brief discussion is provided with respect to the performance construction project, the proposed framework suggests a verifica-
of different mixtures. tion test. It should be carried out using the full-size printer in a
similar ambient temperature and humidity as the intended project.
3. Experimental program Using the same concrete batching, mixing, and transporting equip-
ment as the actual project is highly recommended. The main func-
3.1. Proposed framework tion of the proposed verification test is to subject the designed
mixture to actual jobsite-based assessment.
Fig. 2 presents the proposed framework for laboratory testing of In following sections, the experimental program which was car-
printing concrete in fresh state. This testing procedure was ried out based on the proposed framework will be presented and
designed such that it is applicable to different concrete 3D printing the details of proposed requirements and test methods for print
systems, as it focuses on properties of printed layers rather than quality, shape stability and printability window will be discussed.
employed pumping or extrusion mechanism.
Based on past experience and current literature [23,28,29], it
seems that printing concrete could be characterized by high pow- 3.2. Concrete printing setup
der content, no coarse aggregate, increased paste fraction, and use
of VMA. In fact, reported mixture proportions of successful printing In order to simplify laboratory testing of various mixtures, a lin-
mixtures could be used as a starting point. After designing an ini- ear concrete printing machine was constructed (Fig. 3-a). Com-
tial mixture, fulfilling print quality requirements is the first step in pared to implementing a full-size printer (e.g. Contour Crafting
the proposed iterative mixture assessment and modification pro- machine), the main advantages of using such a laboratory-scale
cess. Three requirements are proposed with respect to print qual- printing machine include savings in time, cost, and material. The
ity, namely, surface quality, squared edges and dimension developed system is capable of printing up to ten 1.2 m long con-
conformity and consistency. All these three requirements must crete layers. The nozzle uses an extrusion mechanism similar to
be satisfied by a mixture with acceptable print quality. the Contour Crafting machine to print 38.1 mm by 25.4 mm
Next, shape stability of a mixture must be examined and rele- (1.5 in. by 1.0 in.) concrete layers. The control system was devel-
vant adjustments should be made accordingly. In this regard, cylin- oped using a combination of Arduino Mega (based on
der stability test is proposed herein for rapid evaluation and ATmega1280) and Arduino Uno (based on ATmega328) microcon-
comparison of influence of different materials (additives or admix- trollers. The machine is able to print concrete at different linear
tures) on shape stability. However, acceptance or rejection decision speeds and deposition rates. For the purposes of this study, a linear
must be based on layer settlement experiment, where concrete printing speed of 60 mm/s (2.36 in./s) was selected.
layers are printed on top of each other with the same extrusion In addition, Bluetooth communication with the printing setup
mechanism as the full-scale concrete printer. No visible deforma- enables the user to conveniently control the machine using a
tions should occur when target interlayer time gap is used. smartphone or tablet. An Android application (Fig. 3-b) was devel-
The third step in laboratory testing of a printing mixture refers oped to facilitate the remote control of the setup. This application
to printability window of a mixture, where printability limit (the enables the user to select the parameters (such as linear speed and

Fig. 2. Proposed framework for laboratory testing of printing mixture in fresh state.
642 A. Kazemian et al. / Construction and Building Materials 145 (2017) 639–647

Table 2
Properties of used chemical admixtures, fiber and silica fume.

Property HRWRA VMA Polypropylene fiber Silica fume

Specific gravity 1.1 1.02 0.91 2.2

pH 5.6 6 – –
Aspect ratio – – 29 –

A polycarboxylate-based high-range water reducing admixture

(HRWRA) was used to achieve the required flowability for the mix-
tures. In addition, to increase the plastic viscosity and cohesion of
printing mixtures, a commercially available viscosity modifying
admixture (VMA) for anti-washout concrete was used. Polypropy-
lene fiber was also used as a shrinkage reinforcement for a printing
mixture. Addition of this fiber inhibits and controls the formation
of plastic and drying shrinkage cracking in concrete [30]. This is
highly important for printing concrete since a higher rate of water
evaporation is anticipated for a printed structure, given that there
is no formwork covering the surface of freshly printed elements.
The length of used fiber is 6 mm and tensile strength is 415 MPa.
Furthermore, densified silica fume was used as a supplementary
cementitious material (SCM). It is well established that silica fume
addition improves cohesion of fresh concrete, and mechanical
strength and impermeability of hardened concrete [31,32]. Some
of the properties of mentioned materials are presented in Table 2.
A highly-purified attapulgite clay with average particle length
of 1.75 mm and average particle diameter of 3 nm was used in this
study. The specific gravity of this clay is 2.29 and the average
length divided by average diameter of clay particles is 583, indicat-
ing a high aspect ratio [33]. Therefore, they may form a highly
entangled gel even at a small addition rate, provided proper disper-
sion. This Nano-clay is commercially available and several studies
[33–35] have investigated the influence of its addition on different
properties of other special concretes, such as formwork pressure of
self-consolidating concrete.
Fig. 3. (a) Linear concrete printing machine for laboratory testing of printing
Mixture proportions of different printing mixtures are pre-
mixtures. (b) Android application developed for remote control of concrete printing sented in Table 3. These mixtures were designed to demonstrate
setup. possibilities with respect to wide choice of materials in concrete
3D printing. For comparison purposes, total cementitious materials
content and water/cementitious materials ratio were kept constant
deposition rate) for concrete printing process, and also provides in all mixtures at 600 kg/m3 and 0.43, respectively. With respect to
few pre-defined settings for convenient operation. mixing procedure, a drum mixer was used to mix the ingredients
for 8 min. For NCPM mixture preparation, Nano-clay was initially
3.3. Materials and mixture proportions mixed with water and introduced as a suspension to the mixer.

In this study, ASTM C150 Type II Portland cement was used to

produce printing mixtures. A commercially available manufac- 4. Testing and results
tured sand with nominal maximum aggregate size of 2.36 mm
was used as fine aggregate. The sieve analysis results and other In this section, different test methods which were used to char-
physical properties of used sand are presented in Table 1. acterize the printing mixtures and the obtained results are pre-
sented. Initially, some conventional test methods and the
corresponding results will be discussed. These conventional tests
Table 1
Sieve analysis and other properties of fine aggregate. serve as standard characterization of mixtures and provide the
opportunity for other researchers to compare these results with
Sieve Percent passing
other studies. Then, workability of printing mixtures will be
Number Size (mm) described and evaluated in terms of print quality, shape stability,
4 4.75 100 and printability window.
8 2.36 83
16 1.18 62
30 0.6 38 4.1. Conventional mixture characterization
50 0.3 19
100 0.15 8 In order to characterize the printing mixtures per conventional
200 0.075 3.5
concrete technology, unit weight, flow, and compressive strength
at 7 and 28-day ages were measured. Flow of each mixture was
Specific gravity 2.6
determined using a flow table (ASTM C1437-15 [36]), where ini-
Absorption capacity (%) 1.3
Fineness modulus 2.9 tially a mold is filled with mortar and compacted. Then the mold
is lifted away from mortar and the table is immediately dropped
 A. Kazemian et al. / Construction and Building Materials 145 (2017) 639–647 643

Table 3
Mixture proportions of printing mixtures (20 L of entrapped air per cubic meter concrete is assumed).

Mixture ID Fine aggregate (SSD) Portland cement Free Water Silica fume Fiber Nano-clay HRWRA VMA
kg/m3 kg/m3 kg/m3 kg/m3 kg/m3 %a %a %a

PPM 1379 600 259 0 0 0 0.05 0.11

SFPM 1357 540 259 60 0 0 0.16 0
FRPM 1379 600 259 0 1.18 0 0.06 0.10
NCPM 1379 600 259 0 0 0.30 0.15 0
a
The percentages are reported by cementitious materials mass.

25 times in 15 s. The flow is the resulting increase in average base

diameter of the mortar mass, expressed as a percentage of the orig-
inal base diameter. Compressive strength was measured according
to ASTM C109 [37], using 50 mm cubes (cured in water tank) at a
loading rate of 1200 N/s. The results are presented in Table 4.
As observed in Table 4, flow of the developed mixtures is in the
range 113–119%, and the minimum flow was measured for mix-
ture with Nano-clay (NCPM). In fact, a considerable increase in vis-
cosity and cohesion of mixture was observed as a result of small
addition of Nano-clay (1.8 kg/m3). With respect to compressive
strength, all mixtures except SFPM demonstrated almost similar
Fig. 4. Discontinuity (tearing) in the printed layer due to excessive stiffness of the
strength values after 7 and 28 days of water curing. Compared to
mixture.
PPM, a 7% increase in 7-day strength and a 12% increase in 28-
day strength was measured as a result of silica fume addition. This
could be attributed to the pozzolanic behavior of silica fume,
where formation of secondary hydration products (mainly C-S-H)
around the silica fume particles fills the large capillary pores with
45mm
a microporous, low-density material [31,38].

4.2. Print quality

39mm
In this study ‘‘print quality” refers to the properties of a printed
layer such as surface quality and dimensional conformity/consis-
tency, when using a specific printing mixture. Based on three pro-
(a)
posed criteria, the print quality of a mixture could be evaluated
and acceptance decision could be made. A printing mixture could
be considered acceptable when the three following requirements
are satisfied: 48mm

1. The printed layer must be free of surface defects, including any

39mm
discontinuity due to excessive stiffness and inadequate
cohesion.
2. The layer edges must be visible and squared (versus round
edges). (b)
3. Dimension conformity and dimension consistency must be sat-
Fig. 5. (a) Variations in width of printed layer using different mixtures at the same
isfied by printed layer. printing speed (dimension conformity). (b) Variations in width of a single layer
(dimension consistency).
Fig. 4 presents a case where a mixture with poor workability is
rejected by the first print quality requirement, due to observation
of discontinuity. With respect to the third print quality require-
ment, ‘‘dimension conformity” guarantees that the dimensions of are set for fresh concrete layers and do not consider variations
the printed layers are within an acceptable range of the target caused by shrinkage over time. In this study, a width of 38.1 mm
dimensions (Fig. 5-a), while ‘‘dimension consistency” refers to (1.5 in.) was designed for each layer and after running a large
changes in width of a printed layer and acceptable variations number of experiments, it was concluded that 10% error in the tar-
(Fig. 5-b). It should be noted that these dimensional limitations get width is a reasonable bound for accepting or rejecting printed

Table 4
Unit weight, flow and strength results for printing mixtures.

Mixture ID Unit weight Flow 7-Day compressive strength 28-Day compressive strength
kg/m3 % MPa MPa

PPM 2250 119 32.9 [0.7]a 44.7 [1.3]

SFPM 2210 116 35.2 [1.6] 49.9 [1.3]
FRPM 2265 118 31.0 [1.9] 45.1 [1.1]
NCPM 2250 113 31.8 [1.2] 45.9 [1.5]
a
Values in brackets are standard deviations of the strength measurements for each set of three cube specimens (in MPa).
644 A. Kazemian et al. / Construction and Building Materials 145 (2017) 639–647

layers. In other words, the width of printed layers using all the out to study the shape stability of printing concrete. The test pro-
accepted printing mixtures was in the range 38.1–42 mm (1.5– cedures and obtained results are provided in following paragraphs.
1.65 in.). It should be mentioned that five measurements were The ‘‘layer settlement” test was developed to enable quantita- tive
done along each printed layer to check the dimension conformity, evaluation of shape stability. In this test, two concrete layers were
and for each mixture the experiment was carried out four times printed on top of each other with a specific time gap. A cam- era
(four replicates of each mixture). The print quality of a mixture was placed in front of printed layers, a ruler was placed next to the
was considered ‘‘acceptable” only if the three requirements were layers as a scale and photos were taken both before and after the
satisfied by all four mixture replicates. The printing mixtures pre- second layer was printed. Then, ImageJ software which is a public
sented in Table 3 were selected based on mentioned print quality domain, Java-based image processing program [41], was used to
requirements. A trial and error approach was adopted for this pur- analyze the photos and to measure layer settlement. The average of
pose, as currently there is no guideline or suggested procedure for five readings for a printed layer was reported as a test result, while
design and testing of printing mixtures. the average of three tests (three printed layers) was used as the final
result for a printing mixture. For each experiment, the bottom layer
4.3. Shape stability was printed as soon as the 8-min mixing proce-
dure was complete.
A critical property of fresh printing concrete, often referred to as The layer settlement test results for both scenarios, namely, 0
shape stability, is the ability to resist deformations during layer- and 19-min time gaps, are presented in Table 5. As anticipated,
wise concrete construction. More specifically, there are three main shape stability of a printed layer improves over time and smaller
sources of deformation which apply to a deposited layer: self- deformations were measured for all mixtures when time gap was
weight, weight of following layer(s) which will be printed on top 19 min. The results indicate that when there is a 19-min time
of it, and the extrusion pressure. Based on observations, a mixture gap between layers, all mixtures except PPM possess high shape
with acceptable print quality. Based on observations, a mixture stability, denoted by no visible deformations of bottom layer. An
with acceptable print quality, as defined by three aforementioned average 1.5 mm deformation, equivalent to 5.9% of layer height,
requirements, produces a layer with no visible deformations due was measured for the layers printed with PPM.
to self-weight. However, the two latter parameters could possibly For PPM mixture with no time gap (see Fig. 6-Top), considerable
lead to undesirable deformations when following layer(s) are deformations happened after second layer was deposited. Consid-
printed. This highlights the importance of laboratory testing of ering the significant changes in both width and height of the layer
shape stability of printing concrete during mixture design stage. in this case, the test result was simply reported as ‘‘collapse”. For 0-
An earlier study on hardened properties of high-performance min time gap, the lowest deformations were measured for SFPM
printing concrete [39] investigated the influence of printing time and NCPM mixtures, where the average layer height reduced by
gap between consecutive layers on the mechanical bond strength approximately 6.7%. Considering the standard deviation of the
of layers. The results indicated that longer interlayer time gap leads obtained results, there is no statistically significant difference
to lower bond strength, which is undesirable in terms of structural between shape stability of these two mixtures. Another finding is
properties of printed structure or element. For instance, between a that acceptable print quality, as defined in this study, does not
30-min and 7-day time gap the average bond strength of printed guarantee high shape stability, as the four printing mixtures with
specimens was 53% and 77% lower than the conventionally cast acceptable print quality showed different levels of shape stability.
specimens, respectively [39]. This signifies the importance of As such, this property has to be separately evaluated during mix-
reducing interlayer time gap, in order to minimize the bond ture design.
strength loss. However, this is practical only if previously depos- It should be noted that yield stress of fresh concrete is the main
ited layers possess high shape stability. Ideally, a printed layer parameter which determines the shape stability before setting.
should withstand the load resulting from following layers right Yield stress increases over time in absence of agitation and shear
after it is deposited, which means no deformations should occur stress [24,42]. This is due to the nucleation of cement grains at
when time gap is zero. their contact point by C-S-H formation during the dormant period
In this study, in order to obtain a realistic notion of the time gap before the setting time [43]. Several researchers have reported a
required between deposition of consecutive layers, the layer-by- linear increase in yield stress during the dormant period
layer construction of a one-story building is considered. This is a [24,44,42], which suggests the possibility of a corresponding linear
108-square-meter house with two bedrooms and one bathroom, increase of shape stability with time. Another important consider-
designed by an architectural and design company [40]. Based on ation for a fundamental study of shape stability is thixotropy,
the house plan, the nozzle traveling distance for each layer was defined as build-up and breakdown of internal 3D structure within
measured as 67 meters. Considering the linear printing speed of cementitious paste. This phenomenon happens due to flocculation
60 mm/s, an interlayer time gap of 19 min was obtained. The calcu- or coagulation and dispersion of cement particles which, in turn,
lated 19-min time gap, as well as the worst-case scenario of 0-min are resulting from inter-particle forces and chemical connections
time gap, were used for shape stability assessment of the devel- [45,46]. Build-up and breakdown of internal structure cause
oped mixtures. Two different test methods, namely, ‘‘layer settle- increase and reduction in viscosity of fresh paste, respectively
ment test” and ‘‘cylinder stability” were developed and carried [47]. Considering the process of concrete 3D printing, where the

Table 5
Layer settlement test results (mm).

Mixture ID Time gap: 0 Time gap: 19 min

Test 1 Test 2 Test 3 Average reading Test 1 Test 2 Test 3 Average reading

PPM Collapse Collapse Collapse – 1.9 1.1 1.6 1.5 [0.3]

SFPM 2.2 1.8 1.5 1.8 [0.3]a 0 0 0 0
FRPM 2.8 3.3 2.5 2.9 [0.3] 0 0 0 0
NCPM 2.0 1.1 1.6 1.6 [0.4] 0 0 0 0
a
Values in brackets are standard deviations of the deformation measurements for each set of three layers (in mm).
 A. Kazemian et al. / Construction and Building Materials 145 (2017) 639–647 645

Interface of two layers

Interface of two layers

Fig. 6. A double layer specimen printed with zero time gap, reported as ‘‘collapse” (Top) and another specimen with a 19-min time gap and no visible deformations (Bottom).

cementitious mixture undergoes considerable shear stress and agi- ratus includes a frame, two semi-cylinder shells (with sealed verti-
tation before being deposited as a layer, these changes in internal cal joints), a tamping rod, two loading guides and a container for
structure and the consequent influence on shape stability need to uniform application of load to fresh concrete specimen (Fig. 8).
be considered. For example, Kawashima et al. [34] reported higher These parts were designed and 3D printed using ABS plastic. Cylin-
build-up rate of internal structure after shear-induced breakdown der stability test procedure is as follows:
for Nano-clay incorporated cementitious mixtures, especially at
early ages. This faster ‘‘structuration at rest” of Nano-clay included (1) The semi-cylinders are fixed in place and locked, and a con-
mixtures could be a likely reason for enhanced shape stability of crete layer of 40 mm (1.57 in.) is placed.
NCPM mixture in this study. Further fundamental research on rhe- (2) Using the tamping rod, the layer is consolidated by rodding
ology of printing concrete and more experimental data are needed 15 times evenly distributed around the layer.
for a deeper understanding of shape stability of printing mixture. (3) The same procedure is repeated for second layer and exces-
Additionally, in order to examine the scalability of obtained results, sive concrete is removed from the top (a well-compacted
five-layer specimens (Fig. 7) were printed using NCPM and SFPM fresh concrete cylinder with total height of 80 mm
mixtures, and an interlayer time gap of 19 min. For both mixtures, (3.14 in.) is achieved).
no visible deformations were observed during printing (4) The two semi-cylinders are unlocked and gently removed.
process. Any possible change in height as a result of self-weight is
‘‘Cylinder stability” test was developed as another test method measured and recorded.
for shape stability evaluation of different mixtures. The test appa- (5) A load of 5.5 kg (equivalent to a 4.77 kPa stress) is applied
and the resulting deformation in the fresh concrete cylinder
is measured in terms of change in height.

The main advantage of this test (compared to layer settlement

test) is eliminating the need to print concrete layers, leading to
savings in time during mixture design phase. On the other hand,
the measured deformations in this test method would be reliable
for comparison purposes primarily, while the layer settlement test
result is a realistic indicator of mixture performance in layer based
construction. It should be noted that Perrot et al. [24] used a sim-
ilar idea to simulate the load acting on the first deposited layer in
layer-wise construction. In their experiment, a 35 mm by 60 mm
fresh concrete cylinder was placed between two metal plates,
Fig. 7. A five-layer printed specimen with interlayer time gap of 19 min. and the upper plate was then loaded in 1.5 N increments. The com-

Fig. 8. (a) 3D printed parts for cylinder stability test. (b) Cylinder stability test.
646 A. Kazemian et al. / Construction and Building Materials 145 (2017) 639–647

Table 6
Cylinder stability test results (mm).

Mixture ID Test 1 Test 2 Test 3 Average reading

PPM 41 37 38 38.7 [1.7]a

SFPM 15 15 14 14.7 [0.5]
FRPM 34 29 31 31.3 [2.1]
NCPM 12 15 11 12.7 [1.7]
a
Values in brackets are standard deviations of measurements from three repli-
cate tests (in mm).

bination of surface fracture occurrence and upper plate displace-

ment was used as indications of plastic deformations of concrete
cylinder over time.
In this study, the cylinder stability test was carried out after the Fig. 9. Concrete penetrometer readings for the three mixtures.
mixing procedure was finished. The obtained results are presented
in Table 6. Based on penetrometer readings, the initial setting time of PPM,
The largest deformation in fresh concrete cylinder was observed PPM1%CaCl, PPM2%CaCl, and PPM3%CaCl was measured as 335,
for PPM mixture, followed by FRPM. NCPM mixture exhibited high- 237, 181, and 163 min, respectively. As anticipated, adding calcium
est shape stability, while SFPM mixture resulted in a slightly larger chloride resulted in accelerated hydration reaction and shorter set-
deformation. Considering the similar performance of SFPM and ting times. The results for printability window parameters, as well
NCPM, the experimental results indicate that cylinder stability test as initial setting time are presented in Table 7. In order to deter-
ranks the shape stability of mixtures in a similar order as layer set- mine the printability limit, a single layer was printed every
tlement test. This implies that this simple test can be suggested for 5 min, starting from 20 min after initial water-cement contact.
comparison of shape stability of different printing mixtures, as well Then, the earliest time when print quality criteria were not satis-
as studying influence of addition of new materials and chemical fied, was recorded as the printability limit. Similarly, the earliest
admixtures. Further research and experimental data is required time when no concrete could be extruded out of nozzle, was
to investigate any correlation between the values given by the recorded as blockage limit.
two test methods. The results indicate that nozzle blockage could happen long
before the initial setting time of each mixture, and setting time
4.4. Printability window could not be used as an alternative indicator. Nozzle blockage
can result in significant time loss, nozzle damage and extra cost
Printability window is the third property of a concrete mixture during construction. As such, measuring the blockage limit for each
which is considered in this study. It refers to the timespan during mixture is recommended during mixture design and laboratory
which the printing mixture could be extruded by the nozzle with testing.
an acceptable quality, considering the workability loss that hap- Finally, in order to investigate the relationship between changes
pens over time. This is highly important in terms of the timing of in workability over time and proposed time limits, workability loss
material delivery to the nozzle and operation of a full-size building of the four mixtures was studied. Flow at different times, from 8 to
printer such as Contour Crafting machine. In this study, two time 90 min after initial water-cement contact, were measured for dif-
limits are introduced to define the printability window of a mix- ferent mixtures and the results are presented in Table 8. The
ture: (1) Printability limit: The time when the quality of printed results indicate that the influence of calcium chloride addition on
layer is affected as a result of workability loss, recognized by triple workability loss appears only after 60 min, and it becomes distinct
‘‘print quality” requirement (2) Blockage limit: The time when the at longer times. This finding justifies the similar printability limits
concrete cannot be guided out of printing nozzle at all, and further which were measured for the three CaCl2 incorporated mixtures
delay would result in mixture solidification and damage to the (Table 7). Also, increasing the accelerator dosage from 2% to 3%
nozzle. was shown to have insignificant influence on the workability loss
An important related concept is initial setting time of concrete. during first 90 min.
Per ASTM C125 [48], initial setting time is defined as the elapsed As for conventional concrete, setting time and workability loss
time after cement-water contact, required for the mortar sieved measurements could be used to study the effects of different
from the concrete to reach a penetration resistance of 3.5 MPa chemical admixtures on fresh state behavior of printing concrete.
(500 psi). In this study, a concrete penetrometer (based on ASTM However, it seems that neither setting time measurements nor
C403 [49]) was used to measure the initial setting time of devel- workability loss measurements could replace the direct measure-
oped mixtures. Since the setting times of the four mixtures were ment of printability limit and blockage limit. Also, both proposed
similar (325–350 min), three other mixtures (based on PPM) were limits are directly dependent on the specific extrusion mechanism
developed and studied. Three different dosages of analytical grade (auger, piston, etc.) used by the concrete printer. Therefore, finding
CaCl2 (1%, 2%, and 3% of Portland cement mass) were added to PPM a printer-independent relationship between conventional parame-
mixture and the resulting mixtures were labeled as PPM1%CaCl, ters such as setting time and printability window limits seems
PPM2%CaCl, and PPM3%CaCl, respectively. It should be noted that unlikely.
calcium chloride is commonly known as an effective accelerator
for cementitious mixtures. It acts like a catalyst for C 3S reactions, Table 7
Initial setting time, printability limit and blockage limit of mixtures (min).
which also results in larger hydration heat liberation [31]. The
measurements using concrete penetrometer are presented in Mixture ID Initial setting time Printability limit Blockage limit
Fig. 9, and the red dashed line shows the reading corresponding PPM 335 55 85
to the initial setting time. For each mixture, the measurements PPM1%CaCl 237 40 75
were carried out three times and the variations in readings are PPM2%CaCl 181 40 60
PPM3%CaCl 163 45 55
shown on the graph using error bars.
 A. Kazemian et al. / Construction and Building Materials 145 (2017) 639–647 647

Table 8 [17] R. Wolfs, 3D Printing of Concrete Structures, Eindhoven University of

Flow of printing mixtures at different times (%). Technology, 2015.
[18] B. Khoshnevis, M.P. Bodiford, K.H. Burks, E. Ethridge, D. Tucker, W. Kim, H.
Mixture ID Time (min) Toutanji, M.R. Fiske, Lunar contour crafting—a novel technique for ISRU-based
habitat development, in: 43rd AIAA Aerospace Sciences Meeting and Exhibit,
8 20 40 60 70 80 90
Reno, Nevada, 2005.
PPM 119 111 100 79 69 61 55 [19] B. Khoshnevis, X. Yuan, B. Zahiri, Z. Jing, B. Xia, Construction by contour
PPM1%CaCl 112 110 102 63 52 48 42 crafting using sulfur concrete with planetary applications, Rapid Prototyping J.
PPM2%CaCl 114 113 97 54 46 37 31 22 (5) (2016) 848–856.
PPM3%CaCl 107 115 95 47 45 34 28 [20] B. Khoshnevis, M. Thangavelu, X. Yuan and J. Zhang, Advances in contour
crafting technology for extraterrestrial settlement infrastructure buildup, in:
AIAA SPACE 2013 Conference and Exposition, San Diego, CA, 2013.
[21] S. Lim, R.A. Buswell, T.T. Le, R. Wackrow, S.A. Austin, A.G. Gibb, T. Thorpe,
5. Conclusions Development of a viable concrete printing process, in: International
Association for Automation and Robotics in Construction (IAARC), 2011.
[22] Apis Cor, Apis Cor, [Online]. Available: http://apis-cor.com/en/3d-printer
In this study a framework was proposed for performance-based (Accessed 11 November 2016).
laboratory testing of fresh printing mixtures. An iterative labora- [23] T.T. Le, S.A. Austin, S. Lim, R.A. Buswell, A.G.F. Gibb, T. Thorpe, Mix design and
fresh properties for high-performance printing concrete, Mater. Struct. 45
tory testing procedure was described to assess and modify the
(2012) 1221–1232.
three suggested workability aspects of a printing mixture, namely, [24] A. Perrot, D. Rangeard, A. Pierre, Structural built-up of cement-based materials
print quality, shape stability and printability window. This testing used for 3D-printing extrusion techniques, Mater. Struct. 49 (4) (2016) 1213–
procedure was designed such that it is applicable to different con- 1220.
[25] Specification and Guidelines for Self-Compacting Concrete, European
crete 3D printing systems, as it focuses on properties of printed Federation of National Associations Representing producers and applicators
layers rather than employed pumping or extrusion mechanism. of specialist building products for Concrete (EFNARC), 2002.
To realize the proposed framework, a laboratory-scale linear con- [26] S. Shah, Design and Application of Low Compaction Energy Concrete for Use in
Slip-Form Concrete Paving, Infrastructure Technology Institute, Northwestern
crete printer capable of printing 38.1 mm by 25.4 mm (1.5 in. by University, 2008.
1.0 in.) layers was constructed. Four different printing mixtures [27] T. Voigt, J. Mbele, K. Wang, S. Shah, Using fly ash, clay, and fibers for
were also developed and the results for a combination of conven- simultaneous improvement of concrete green strength and consolidatability
for slip-form pavement, J. Mater. Civ. Eng. 22 (2) (2010) 196–206.
tional and new test methods were reported. In specific, for evalua- [28] V. Nerella, M. Krause, M. Nather, V. Mechtcherine, Studying printability of
tion of shape stability of printing concrete two different test fresh concrete for formwork free concrete on-site 3D printing technology
methods, namely, ‘‘layer settlement” and ‘‘cylinder stability” were (CONPrint3D), in: 25th Conference on Rheology of Building Materials,
Regensburg, 2016.
proposed. Experimental data revealed that inclusion of silica fume [29] L. Anell, Concrete 3D Printer, Lund University, Sweden, 2015.
and Nano-clay (a highly-purified attapulgite clay) enhance shape [30] BASF, [Online]. Available: https://www.master-builders-solutions.basf.us/en-
stability of fresh printing mixture, while minor improvement was us/products/masterfiber/1649 (Accessed 01 11 2016).
[31] P.K. Mehta, P.J.M. Monteiro, Concrete: Microstructure, Properties, and
observed as a result of polypropylene fiber addition.
Materials, fourth ed., McGraw-Hill Education, 2013.
[32] A.A. Ramezanianpour, A. Kazemian, M. Nikravan, M.A. Moghadam, Influence of a
low-activity slag and silica fume on the fresh properties and durability of high
References performance self-consolidating concrete, in: International Conference on
Sustainable Construction Materials & Technologies (SCMT3), Kyoto, Japan, 2013.
[1] S. Kubba, Handbook of Green Building Design and Construction, second ed., [33] S. Kawashima, M. Chaouche, D.J. Corr, S.P. Shah, Influence of purified
Butterworth-Heinemann, 2017. attapulgite clays on the adhesive properties of cement pastes as measured
[2] H. Wei, S. Zheng, L. Zhao, R. Huang, BIM-based method calculation of auxiliary by the tack test, Cem. Concr. Compos. 48 (2014) 35–41.
[34] S. Kawashima, M. Chaouche, D.J. Corr, S.P. Shah, Rate of thixotropic rebuilding
materials required in housing construction, Autom. Constr. 78 (2017) 62–82.
[3] H. Penttilä, Describing the changes in architectural information technology to of cement pastes modified with highly purified attapulgite clays, Cem. Concr.
understand design complexity and free-form architectural expression, Res. 53 (2013) 112–118.
[35] J. Kim, M. Beacraft, S.P. Shah, Effect of mineral admixtures on formwork
Electron. J. Inf. Technol. Constr. 11 (2006) 395–408.
[4] O. Davtalab, Benefits of real-time data driven BIM for FM departments in pressure of self-consolidating concrete, Cem. Concr. Compos. 32 (9) (2010) 665–
operations control and maintenance, in: ASCE International Workshop on 671.
[36] ASTM C1437-15: Standard Test Method for Flow of Hydraulic Cement Mortar,
Computing in Civil Engineering (IWCCE 2017), Seattle, 2017.
American Society of Testing and Materials (ASTM), 2015.
[5] R.A. Buswell, A.G.F. Gibb, R.C. Soar, S.A. Austin, A. Thorpe, Applying future
[37] American Society of Testing and Materials (ASTM), [Online]. Available: https://
industrialised processes to construction, in: CIB World Building Congress
www.astm.org/Standards/C109.htm (Accessed 08 2016).
‘Construction for Development’, Cape Town, 2007.
[38] A.A. Ramezanianpour, A. Kazemian, M. Sarvari, B. Ahmadi, Use of natural
[6] S. Ahmadi-Karvigh, B. Becerik-Gerber, L. Soibelman, A framework for allocating
zeolite to produce Self-consolidating concrete with low Portland cement
personalized appliance-level disaggregated electricity consumption to daily
content and high durability, J. Mater. Civ. Eng. 25 (5) (2013).
activities, Energy Build. 111 (2016) 337–350.
[39] T.T. Le, S.A. Austin, S. Lim, R.A. Buswell, A.G.G.R. Law, T. Thorpe, Hardened
[7] F. Jazizadeh, S. Ahmadi-Karvigh, B. Becerik-Gerber, L. Soibelman,
properties of high-performance printing concrete, Cem. Concr. Res. 42 (3)
Spatiotemporal lighting load disaggregation using light intensity signal,
(2012) 558–566.
Energy Build. 69 (2014) 572–583.
[40] FreeGreen, [Online]. Available: https://www.houseplans.com/plan/1160-
[8] J. Irizarry, M. Gheisari, B.N. Walker, Usability assessment of drone technology
as safety inspection tools, J. Inf. Technol. Constr. 17 (2012) 194–212. square-feet-2-bedroom-1-bathroom-0-garage-modern-39050 (Accessed 15
06 2016).
[9] J. Irizarry, D.B. Costa, Exploratory study of potential applications of unmanned
[41] ImageJ, [Online]. Available: https://imagej.nih.gov/ij/.
aerial systems for construction management tasks, J. Manage. Eng. 32 (3)
[42] L. Josserand, O. Coussy, F.D. Larrard, Bleeding of concrete as an ageing
(2016).
consolidation process, Cem. Concr. Res. 36 (9) (2006) 1603–1608.
[10] ASTM F2792-12a (withdrawn), American Society of Testing and Materials,
2012. [43] N. Roussel, G. Ovarlez, S. Garrault, C. Brumaud, The origins of thixotropy of
[11] R.A. Buswell, R.C. Soar, A.G. Gibb, A. Thorpe, Freeform construction application fresh cement pastes, Cem. Concr. Res. 42 (1) (2012) 148–157.
research, in: Advances in Engineering Structures, Mechanics & Construction, [44] N. Roussel, A thixotropy model for fresh fluid concretes: theory, validation and
applications, Cem. Concr. Res. 36 (2006) 1797–1806.
2006, pp. 773–780.
[45] J. Wallevik, Thixotropic investigation on cement paste: experimental and
[12] S. Lim, T. Le, J. Webster, R. Buswell, A. Austin, A. Gibb, Fabricating construction
components using layered manufacturing technology, in: Global Innovation in numerical approach, J. Nonnewton. Fluid Mech. 132 (2005) 86–99.
[46] J. Wallevik, Rheological properties of cement paste: thixotropic behavior and
Construction, Loughborough University, Leicestershire, UK, 2009.
structural breakdown, Cem. Concr. Res. 39 (1) (2009) 14–29.
[13] P. Feng, X. Meng, J. Chen, L. Ye, Mechanical properties of structures 3D printed
with cementitious powders, Constr. Build. Mater. 93 (2015) 486–497. [47] G. Heirman, Modelling and Quantification of the Effect of Mineral Additions on
[14] B. Khoshnevis, Automated construction by contour crafting- related robotics the Rheology of Fresh Powder Tupe Self-Compacting Concrete, Arenberg
and information technologies, Autom. Constr. 13 (1) (2004) 5–19. Doctoral School of Science, Engineering and Technology, 2011.
[48] ASTM C125-15b: Standard Terminology Relating to Concrete and Concrete
[15] R. Naboni, I. Paoletti, Advanced Customization in Architectural Design and
Aggregates, American Society of Testing and Materials (ASTM), 2015.
Construction, Springer, 2015.
[16] T. Wangler, E. Lloret, L. Reiter, N. Hack, F. Gramazio, M. Kohler, M. Bernhard, B. [49] ASTM C403/C403M-08: Standard Test Method for Time of Setting of Concrete
Mixtures by Penetration Resistance, American Society of Testing and Materials
Dillenburger, J. Buchli, N. Rousel, R. Flatt, Digital concrete: opportunities and
challenges, RILEM Techn. Lett. 1 (2016) 67–75. (ASTM), 2008.