Content uploaded by Stefan Chaves Figueiredo
Author content
All content in this area was uploaded by Stefan Chaves Figueiredo on Mar 07, 2019
Content may be subject to copyright.
Accepted Manuscript
An approach to develop printable strain hardening cementitious
composites
Stefan Chaves Figueiredo, Claudia Romero Rodríguez, Zeeshan
Y. Ahmed, D.H. Bos, Yading Xu, Theo M. Salet, Oğuzhan
Çopuroğlu, Erik Schlangen, Freek P. Bos
PII: S0264-1275(19)30088-7
DOI: https://doi.org/10.1016/j.matdes.2019.107651
Article Number: 107651
Reference: JMADE 07651
To appear in: Materials & Design
Received date: 23 October 2018
Revised date: 8 February 2019
Accepted date: 9 February 2019
Please cite this article as: S. Chaves Figueiredo, C. Romero Rodríguez, Z.Y. Ahmed, et al.,
An approach to develop printable strain hardening cementitious composites, Materials &
Design, https://doi.org/10.1016/j.matdes.2019.107651
This is a PDF file of an unedited manuscript that has been accepted for publication. As
a service to our customers we are providing this early version of the manuscript. The
manuscript will undergo copyediting, typesetting, and review of the resulting proof before
it is published in its final form. Please note that during the production process errors may
be discovered which could affect the content, and all legal disclaimers that apply to the
journal pertain.
ACCEPTED MANUSCRIPT
An approach to develop printable strain hardening
cementitious composites
Stefan Chaves Figueiredoa,∗
, Claudia Romero Rodr´ıgueza, Zeeshan Y.
Ahmedb, D. H. Bosb, Yading Xua, Theo M. Saletb, O˘guzhan C¸ opuro˘glua,
Erik Schlangena, Freek P. Bosb
aMicrolab, Faculty of Civil Engineering and Geosciences, Delft University of Technology,
Delft 2628, CN, The Netherlands
bDepartment of the Built Environment, Eindhoven University of Technology, Eindhoven,
The Netherlands
Abstract
New additive manufacturing methods for cementitious materials hold a high
potential to increase automation in the construction industry. However,
these methods require new materials to be developed that meet perfor-
mance requirements related to specific characteristics of the manufacturing
process. The appropriate characterization methods of these materials are
still a matter of debate. This study proposes a rheology investigation to
systematically develop a printable strain hardening cementitious composite
mix design. Two known mixtures were employed and the influence of sev-
eral parameters, such as the water-to-solid ratio, fibre volume percentage
and employment of chemical admixtures, were investigated using a ram ex-
truder and Benbow-Bridgwater equation. Through printing trials, rheology
parameters as the initial bulk and shear yield stress were correlated with
variables commonly employed to assess printing quality of cementitious ma-
terials. The rheology properties measured were used to predict the number
of layers a developed mixture could support. Selected mixtures had their
mechanical performance assessed through four-point bending, uni-axial ten-
sile and compressive strength tests, to confirm strain hardening behaviour
was obtained. It was concluded that the presented experimental and the-
oretical framework are promising tools, as the bulk yield stress seems to
predict buildability, while shear yield stress may indicate a threshold for
pumpability.
∗Corresponding author
Email address: S.ChavesFigueiredo@tudelft.nl (Stefan Chaves Figueiredo)
Preprint submitted to Materials & Design February 15, 2019
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Keywords:
3D printing, Rheology, Strain hardening, Additive manufacturing
1. Introduction
Over the last decades, the employment of automation at construction
sites has seen substantial achievements to enhance the productivity of the
sector [1, 2]. Management tools have allowed tasks and activities to become
more specialized and dynamic, thus providing an environment for shortening
of construction time and decreasing construction errors. The employment of
machinery to execute tasks on this industry was specially devoted to heavy
duties nevertheless, some other activities still rely on the skills of humans.
A relatively recent development has been the introduction of Additive
Manufacturing (AM) to the construction industry, also popularly known
as 3D printing. AM is a classification of manufacturing technologies that
fabricate objects by controlled, often layer-wise, addition of material, rather
than by removal of material from a larger piece of bulk material. Generally,
robotized equipment is applied that manufacture an object directly from
digital design input [3, 4]. For construction, advances are being made AM
of polymers [5], foams [6], glass [7], timber [8] and steel [9].
Developments have been particularly rapid for AM of concretes and
other cementitious materials (AMoC) for construction. Technologies under
development include the Stereolithography (STL) based D-shape process
[10], Contour Crafting (CC) [11], Concrete Printing (CP) [12], 3D Concrete
Printing (3DCP) [13], as well as the vertical extrusion based Smart Dynamic
Casting [14] and Mesh Mould [15], in which a mesh reinforcement acts as
the mould. Leaving the phase of showcasing the potential behind, described
in a range of publications [16–23], AMoC (also referred to as Digitally Fab-
ricated Concrete (DFC), to emphasize the automated production method)
has now entered in a period of first real uses [24], and commercial initiatives
abound (e.g. Contour Crafting, Total Kustom, WinSun, CyBe, Apis Cor,
XtreeE, Incremental 3D, COBOD, and others). Alternative techniques are
also available in which a mould of insulation material is printed and the
concrete is cast inside [25].
This requires the development of a whole new generation of materials
to meet both the manufacturing requirements (printability) as well as the
mechanical and durability demands of a long-lasting. Recent researches have
shown the development of cementitious composites with different aggregate
particle sizes [16]. In some studies, fibres have been incorporated to stabilize
2
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
the mixture at fresh state or to minimize the occurrence of cracks due to
shrinkage [26].
However, as has been pointed out by other researchers as well, the dif-
ferent technologies and materials under development still suffer from a ma-
jor drawback that forms an important obstacle for them to achieve their
potential which includes form freedom, reduced material use and labour,
decreased CO2emissions, and construction speed [27–31]. The print ma-
terials in AMoC, in general, have a low tensile strength compared to their
compressive strength. Furthermore, they are usually brittle and thus fail
relatively suddenly without large deformations. For structural use in con-
struction, this is unacceptable, as the conventional concepts, such as the use
of steel reinforcement bars, are either incompatible with AMoC or eliminate
its advantages. A number of innovative approaches have been presented to
overcome this problem, like the ones described by [32–34]. Further research
is needed to be able to fully assess their potential and applicability.
Including fibres in the print material is an obvious solution strategy,
too. It has been explored by Panda et al. [35], who compared glass fibres of
different lengths (3, 6 and 8 mm) and varying volume percentage of fibres.
Both studies reported a significant increase in flexural tensile strength, as
well as an orientation effect of the fibres in the direction of the filament
flow, but neither discussed the effects on ductility. Moreover, the use of
PVA fibres in printable cementitious mortars exposed to fire attack was
studied in [36]. In this case the authors have shown the advantages of using
fibres to enhance ductility but also to minimize the occurrence of spoling.
And recently the use of steel fibres to reinforce a printable mortar was also
explored in [37] .
Over the last decades, cementitious composites have been developed that
exhibit strain hardening behaviour [38, 39]. Their performance is based on
an optimized matrix composition, fibre performance, and matrix-to-fibre
bond, and are known as Engineered Cementitious Composites (ECCs) or
Strain Hardening Cementitious Composites (SHCCs). Significant plastic
deformations can be achieved, as well as a high tensile strain, strength, and
multiple crack development [40]. Resistance to quasi-static and dynamic
loading is generally high [41]. Jointly, this results in favourable structural
performance.
First results on the development of printable ECCs have been published
by Soltan & Li [42]. Based on considerations of extrudability (indicating
the ability of the mixture to pass through a printing system) and buildabil-
ity (indicating the ability of a mixture to remain stable after depositioning
and during printing), that together define the printability, they developed
3
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
several mixtures with polyvinylalcohol (PVA) fibres. The influence of sev-
eral ingredients on fresh state workability and processing parameters were
investigated. This resulted in at least one mixture that seems printable and
shows strain hardening failure behaviour. However, the assessment of fresh
state properties was based on the flow factor according to ASTM C1437 and
ASTM C230, which is not a true rheological property. Also, the real print-
ability was not truly yet established as only several layers were deposited
with a manual piston. Pending a more extensive publication, a brief descrip-
tion of results about the development of a printable SHCC with high-density
polyethylene (HDPE) fibres were given by [43].
The development of printable mix-designs is different from that of castable
concretes. The challenges are not restricted to the hardened properties:
competing requirements for extrudability and buildability have to be met as
well. Globally speaking, the material should be fluid enough to pass through
a print system without the use of excessive pressure and the occurrence of
ruptures and/or void, while exhibiting sufficient strength and stiffness af-
ter depositioning to avoid failure during printing or excessive geometrical
deformations. When both these requirements are met, the material can be
considered printable.
In order to evaluate the printability, different tests have been proposed,
such as the cylinder test [26], the slump of the fresh mixture in a shape of
a cylinder [44] or the slump of the printed layer itself [45]. In general, these
tests consist in measuring the slump of the fresh material with or without
a certain weight on top of it. However, such empirical tests do not result
in true physical properties that describe the rheological or mechanical be-
haviour of the material. Only recently, first attempts to analyse printability
in terms of physical rheology or mechanics properties have been presented
[46]. Suitable methodologies are still under development.
The requirements for materials employed in the process of manufacturing
an object in AMoC through a processes like 3DCP, are similar to those for
extrusion manufacturing, a process that is commonly used for several types
of concrete products. Although this has been acknowledged in some reports,
the mix-design often does not follow the procedure that is usually suggested
in the field of extrusion research [47]. Extruded cementitious mixtures can
be considered as solid suspensions. These highly concentrated suspensions
usually show dough-like texture. Therefore, the employment of conventional
shear-based rheometers is not always suitable. Slippage and plug-forming
of the evaluated mixture may lead to unreliable results [48]. Alternatively,
a ram extrusion rheometer can be used. Using the pressure measurements
from this device, true rheology parameters can be determined through the
4
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Benbow-Bridgwater rheology model [49–51]. A more detailed explanation
of the model and the apparatus will be given in Sections 2 and 3.2.
As extrusion techniques are of great importance for the construction
industry, the ram extruder proposed by Benbow-Bridgwater was extensively
employed to quantify rheological parameters of different materials, amongst
which a vast number of ingredients for cementitious composites. Particularly
relevant to this study are reports on fibre reinforced mixtures, like the ones
found in [48, 52–54], with PVA fibres, or in [55, 56] with natural fibres, or
even with nano-fibres in [57]. The fresh state properties depend on several
factors, like the volume of liquid employed, the particle size distribution,
the volume of fibre reinforcement, time, and so on. Furthermore, rheology
modifiers are commonly reported to have been used in studies on extruded
fibre reinforced cementitious composites.
As SHCCs demonstrate enhanced mechanical performance in compari-
son to standard concrete or mortar, as well as superior durability properties
[58, 59], the systematic (i.e. based on true properties) development of print-
able SHCC mixtures can move AMoC forward. Therefore, this research
aimed to develop a printable SHCC mix-design based on the rheology prop-
erties measured with a ram extruder and determined through the Benbow-
Bridgwater model. To understand the meaning of the physical properties on
the flowability, visual inspections were performed on the extruded compos-
ite. A printability trial was subsequently conducted in a large scale 3DCP
facility, on selected mixtures to assess pumpability, extrudability, and build-
ability. After hardening, several mechanical properties were determined to
show the printable mixtures do indeed result in objects with strain-hardening
failure behaviour. Only a limited number of tensile test results are shown
here. The full results of the study on mechanical properties in the hardened
state will be subject of a future publication.
2. Theoretical background
Behaving rheologically as a Bingham fluid, cementitious materials can
have their yield stress measured. Measuring rheological properties of highly
concentrated particle suspensions through conventional shear based rheome-
ters has been shown as not the best approach [48, 60], and alternatives were
given in [49–51]. The Benbow-Bridgwater model, used commonly to study
fluids such as molten plastics, clay suspensions and prefabricated cementi-
tious material, is especially suitable for this work.
In order to evaluate the rheology of composites at the fresh state a ram
extruder was employed. The ram extruder is commonly composed of an
5
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
upper barrel, where the material is introduced first and a connected die
land, with smaller diameter, from which the material is extruded at last.
A piston, moving downwards, pushes the material from the upper barrel
through the die land. Coupling the total pressure drop in the die measured
with this device alongside the Benbow-Bridgwater model, a description of
the fluid rheology can be obtained. The total pressure drop in the die is
composed by the pressure drop of the fluid on the die entry and the pressure
drop on the die land. The pressure drop does not take in consideration any
pressure drop in the barrel, as it is neglectable [61]. Therefore, the total
pressure drop is given by the equation 1:
P=P1+P2= 2 ln( D
d)(σ0+αV ) + 4L
d(τ0+βV ) (1)
where:
P = Total pressure drop [kPa]
P1= Pressure drop in the die entry [kPa]
P2= Pressure drop in the die land [kPa]
σ0= Bulk yield stress [kPa]
α= Parameter characterizing speed in the die entry [kPa.s/mm]
V = Extrusion speed in the die land [mm/s]
D = Barrel diameter [mm]
d = Die diameter [mm]
τ0= Shear yield stress [kPa]
β= Parameter characterizing speed in the die land [kPa.s/mm]
L = Die length [mm]
For the case of extruded paste developing pseudo-plastic behaviour the
influence of the extrusion speed on the total pressure drop is not linear.
For such cases, the Benbow-Bridgwater model is further enriched with the
coefficients m and n, as shown in equation 2 [61].
P=P1+P2= 2 ln( D
d)(σ0+αV m) + 4L
d(τ0+βV n) (2)
Rheological characterization of dough-like pastes are especially inter-
esting for extruded materials, as they must keep their shape after being
extruded. The resulting rheological properties are also very interesting for
printed cementitious composites. Initial shear and bulk yield stresses are
physical properties which can help quantifying important parameters for
the AM with counter craft technology. The shear yield stress quantifies the
friction of the material moving through the die while the bulk yield stress is
6
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
an intrinsic material property. These quantities can be related to the main
mixture requirements like shape stability and printability. Moreover, these
properties can also be useful to estimate the amount of layers the material
is able to support.
3. Experimental methods
3.1. Materials and sample preparation
Two SHCC mixtures from [62, 63] were chosen as a departure point for
the development of the SHCC mixtures for 3D printing. Both mixtures were
reinforced with 2% by volume of polyvinyl alcohol (PVA) fibres. The first
mixture matrix was composed by ordinary Portland cement (OPC), blast
furnace slag (BFS), and limestone powder (LP), while the second reported
mixture is composed by OPC, fly ash (FA) and sand. In order to increase the
amount of fines used in the second SHCC, additional LP was used. In table
2 and table 3 the composition of each mixture are detailed. Initially, the
rheology of their matrices (the composite without fibres) were studied. The
influence of viscosity modifying agent (VA), superplasticizer (SP), water-to-
solid ratio, PVA fibre volume and sand grain size on the fresh state properties
were investigated.
The chemical composition of powder materials and their loss on ignition
(LOI) can be found in table 1. They were assessed by X-ray fluorescence
analysis (XRF) and thermogravimetric analysis performed at 10 K/min un-
der Argon atmosphere. The LOI was calculated using the loss of mass
between 45 and 1000 ◦C. The particle size distribution of the raw materials
can be found on figure 1.
VA with viscosity 201000 mPa.s was provided by Shanghai Ying Jia
Industrial Development Co., Ltd. SP used was a Glenium 51 obtained from
BASF with solid concentration of 35%.
For the rheology tests, a volume of 0.5 litres was mixed in a planetary
mixer according to the following procedure:
•All dry materials were mixed for two minutes at low speed (speed 1 -
60 rpm);
•While mixing at speed 1, during approximately one minute, water
mixed with SP was added;
•The wet powders were mixed for the next two minutes at speed 1.
In this phase it was possible to observe a significant change in the
mixture’s viscosity. A dough like consistence was achieved;
7
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Figure 1: Particle size distribution
•At moderate speed (speed 2 - 124 rpm), the dough like mixture was
further mixed. At this phase the dough opens inside the mixing bowl,
and the fibres get dispersed.
3.2. Rheology measurements
In order to obtain the four parameters (σ0,α,τ0,β) describing the paste
flow a ram extruder was built. The design was based on the equipment
reported by [48, 50, 64] and can be found in figure 2. Three dies were
applied with an internal diameter of 12.8 mm and length-to-diameter (L/d)
ratios of 1, 4, and 8. The diameter of the piston (38.3 mm) was designed to
minimize friction with the internal walls of the barrel (D = 38.4 mm) and
to fit on an servo-hydraulic press (Instron 8872). Besides that, a Fluon R
(polytetrafluoroethylene) ring was used as the end of the piston to seal the
gap between walls and minimizing friction. During the tests this region was
always lubricated with a silicone release compound (Dow Corning 7, Dow
Corning R
). For each new experiment the piston and the ring were removed
from the Instron actuator and washed with tap water and soap.
The ram extruder barrel was filled with the mixture under evaluation.
For each portion placed inside the barrel, compaction with the help of a 30
mm diameter steel rod was done. Compaction of the paste inside of the
8
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Barrel
Die
Piston
Instron
actuator
Stand
frame
Figure 2: Ram extruder and components.
9
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Table 1: Chemical composition of powder raw materials
Compound CEM I
42.5 N [%]
Fly Ash
[%]
Blast Furnace
Slag [%]
Limestone
Powder [%]
CaO 69.53 5.30 42.00 55.80
SiO215.6 53.23 30.73 0.28
Fe2O33.84 7.77 0.54 0.03
Al2O33.09 26.67 13.30 -
SO32.6 0.81 1.45 -
MgO 1.67 1.27 9.44 0.14
K2O 0.55 1.42 0.34 -
TiO20.31 1.22 1.01 -
P2O50.14 0.25 - -
Rest 0.53 0.52 0.62 0.03
Loss on Ignition 2.14 1.55 0.57 43.71
barrel is important to avoid big pockets of air which would result in drastic
drop on the pressure during the extrusion experiment. As soon as the barrel
was filled, the piston was attached to the Instron actuator. Four different
speeds of the piston were used by controlling the displacement rate of the
Instron actuator while the reaction force to the imposed displacement of the
fluid was measured by a load cell. This load was used to calculate the total
pressure applied to the fluid. In figure 3 an example of the output data
from the experiment is given. The ram extruder experiment was performed
four times for each die. The first extrusion was not considered for the test,
as its only function was to aid with the appropriate filling of the die. Hence,
an average of the pressure at each extrusion speed of the three repetitions
was used in the calculation.
From equation 2 a linear relation between the total pressure applied on
the fluid and L/d ratios was obtained. Curve fitting employing a least square
method was used for each of the curves in order to obtain the rheology
parameters that characterize the fluid. As the experiment was done for
four different extrusion speeds and three L/d ratio, an average of each of
the components (σ0,α, m, τ0,βand n) could be obtained. Figure 4
exemplifies the linear curve of total pressure drop versus L/d obtained from
the experiment.
10
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Table 2: Mix design summary of X series in [kg/m3]
Mixtures
nomenclature
CEM I
42.5
Blast
Furnace
Slag
Limestone
powder PVA VA Superplasticizer
(Glenium 51) [g] Water
XVA1 265.2 618.9 884.2 0 1.8 17.7 353.7
XVA2 264.9 618.0 882.9 0 3.6 17.7 353.2
XVA3 264.5 617.2 881.7 0 5.2 17.6 352.7
XVA4 264.1 616.3 880.5 0 7.0 17.6 352.2
XVA3SP1 266.7 622.2 888.9 0 5.3 8.9 355.5
XVA3SP3 262.4 612.3 874.7 0 5.2 26.2 349.9
XVA3W1 224.9 524.7 749.6 0 4.5 15.0 449.7
XVA3PVA10 261.9 611.0 872.9 13.0 5.2 17.5 349.2
XVA3PVA15 260.6 608 868.5 19.5 5.2 17.4 347.4
XVA3PVA20 259.2 604.9 864.1 26 5.1 17.3 345.6
XVA4PVA20 258.9 604 862.9 26 6.9 17.3 345.2
11
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Table 3: Mix design summary of Y series in [kg/m3]
Mixtures
nomenclature
CEM I
42.5
Fly
ash
Limestone
powder
Sand (125
- 250)µm
Sand (250
- 500)µm
Sand (500
- 1000)µmPVA VA Superplasticizer
(Glenium 51) [g] Water
YVA1 492.0 581.4 111.8 110.4 174.3 207.3 0 1.7 13.3 335.4
YVA2 491.4 580.7 111.7 110.3 174.1 207 0 3.3 13.3 335.0
YVA3 490.7 579.9 111.5 110.1 173.8 206.7 0 5 13.3 334.5
YVA4 490.0 579.1 111.4 110.0 173.6 206.4 0 6.7 13.3 334.1
YVA3SP1 493.7 583.5 112.2 110.8 174.9 208 0 5 6.6 336.6
YVA3SP3 487.8 576.5 110.9 109.5 172.8 205.5 0 5 19.8 332.5
YVA3W1 420.4 496.8 95.5 94.4 148.9 177.1 0 4.3 11.4 429.9
YVA3PVA10 485.8 574.1 110.4 109 172.1 204.7 13.0 4.9 13.2 331.2
YVA3PVA15 483.3 571.2 109.8 108.5 171.2 203.6 19.5 4.9 13.1 329.5
YVA3PVA20 480.9 568.3 109.3 107.9 170.4 202.6 26 4.9 13.0 327.8
YVA3PVA20-S05 480.9 568.4 109.3 186.5 294.4 0 26 4.9 13.0 327.9
YVA4PVA20-S05 480.2 567.6 109.1 186.3 294 0 26 6.5 13.0 327.4
12
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Figure 3: An example of the output data from one of the rheology experiments employing
the ram extruder.
Figure 4: An example of the curve fitting employing a least square method of the total
pressure drop and L/d ratio.
13
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
3.3. Printing trials
After the rheological characterization, printing trials were performed to
assess the actual printability of the material. First, an initial test of the
pumpability and extrudability was performed on 6 of the developed mixtures
that were expected to show sufficient buildability based on their rheological
characterization, as assessed both through visual inspection and their quan-
titative properties. The purpose of this trial was to establish whether the de-
veloped mixtures were compatible with the equipment, particularly whether
the fibres would not cause blockage in the linear displacement pump, which
features narrow cavities. Based on the observations in this initial trial, one
mixture was subsequently selected for an object printing experiment.
For the preceding initial trials, mixed batches of the selected mixtures
were fed to the pumping unit of the mixer-pump that is part of the 3DCP
print facility of the Eindhoven University of Technology (TU/e) as described
by [16]. The mixer unit of the mixer-pump was bypassed as the extent
of mixing provided by this unit is insufficient for the developed mixtures.
Therefore, batches were mixed using the procedure that was also applied
for the rheological tests, and material from the mixed batch was inserted
into the pumping unit of the mixer pump. The pump was connected to a 5
m, ø 2.5 cm hose. It was observed whether the material would be pumped
without clogging, and whether the material could be transported through
the hose.
In the object printing experiment, the TU/e 3DCP print facility, shown
in Figure 5, was used in its entirety (except, again, for the mixing unit of
the mixer-pump). The mixer-pump was connected to the print head with
the standard 10 m, ø 2.5 cm hose. The standard print nozzle with a 40 ×
10 mm mouth opening was used. Cylinder shapes with a print path (heart
line) diameter of 500 mm were printed until failure. The appropriate print
speed was established as 5000 mm/min (or 83.3 mm/s). The print time of
a single layer, thus, was approximately 0.31 min (or 19 s). This geometry
has been used previously by [65] to study buildability of another mixture.
Overall behaviour was visually recorded and the number of stacked layers
before failure counted.
3.4. Mechanical tests
The composites reinforced with 2% by volume of PVA fibres were also
evaluated mechanically, to verify whether strain hardening failure behaviour
had indeed be obtained. For these cases, a volume of 3 litres was mixed
following the same procedure described earlier. Four-point bending, and
14
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Figure 5: 3D Concrete Printing facility of the Eindhoven University of Technology (TU/e)
(taken from: [16])
compressive tests were performed to evaluate the performance of the com-
posite. Motivated by the outcomes of the research at the rheology measure-
ment stage, only YVA4PVA20-S05 and XVA3PVA20 mixtures were chosen
to have their tensile behaviour tested.
The samples were cast and kept sealed in their moulds for three days.
Afterwards, they were demoulded and cured in a curing room at (20 ±2) ◦C
and relative humidity of (98 ±2)%. The compressive strength was measured
at 14 and 28 days on 35 mm cubes, sawn from 40×40×160 mm beams.
The samples for four-point bending test were sawn from 180×180×10 mm
slabs, with approximate dimensions of 180×40×10 mm and tested at 28
days of curing. Finally, the samples for direct tensile test were sawed from
240×60×10 mm slabs, in the end reaching final dimensions of 150×40×10
mm and tested at 35 days of curing. The compressive test was done at
loading rate of 2 kN/s. The four-point bending, with a test spam of 12 mm,
and tensile tests were performed on the same servo-hydraulic Instron 8872
machine in which the extrusion tests were done. The employed deflection
rate for the four-point bending was 0.01 mm/s and the tensile elongation rate
was 1 µm/s, controlled by linear variable differential transformer (LVDT)
sensors. It is important to observe that the specimens tested on uni-axial
tensile test had upper and lower side glued on steel plates. The lower side
was glued inside the servo-hydraulic machine to avoid bending while testing.
Due to the high viscosity of the mixtures obtaining a homogeneous thick-
15
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
ness while casting the samples was difficult, especially for the four-point
bending specimens. Therefore, prior to the mechanical tests each individual
specimen was measured at several locations.
Specimens undergoing tensile test had their frontal surface prepared for
employment of digital image correlation (DIC). Their surface was painted
white and randomly distributed black dots were made with a permanent
marker. This pattern helps enhancing the contrast needed for the DIC soft-
ware to calculate the displacements during test. The open source software
Ncorr2 was employed for the DIC [66]. A Cannon camera model EOS 6D
with Tamron aspherical 28 - 75mm lens were employed to obtain one pic-
ture each two seconds of test. An approximate resolution of 48µm/pixel was
obtained.
4. Results
4.1. Rheology
The six parameters-approach according to equation 2 was chosen to char-
acterize all mixtures, as a non-linear behaviour was identified. Figure 6
illustrates the increase of total pressure of different extrusion speeds. In the
following subsections, the influence of each of the mixture variables (viscos-
ity modifier agent, water content, superplasticizer, and fibres) is detailed. A
summary of all results is shown in table 4.
4.1.1. Effect of VA content
The effect of VA was investigated on matrix level. Four dosages of
methylcellulose were employed: 0.1, 0.2, 0.3 and 0.4% of the total solids
weight. The water-to-solid ratio and superplasticizer added were kept con-
stant at 0.2 and 2% by total powder weight, respectively.
For both matrices, the increase of VA content directly influenced the
initial bulk and shear yield stresses, as is visually shown in figure 7, and
quantitatively compared in figure 8. The increase in these rheological pa-
rameters have direct effect on the shape stability of the printed material.
Therefore, the employment of VA can contribute positively to the develop-
ment of printable mix designs. Greater rheological parameters values were
obtained for the X matrix, indicating that solid suspensions with smaller
liquid-to-particles surface area ratio are more vulnerable to changes in the
viscosity of the liquid phase. This result is important to show how sensi-
tive highly concentrated solid solutions, like the ones obtained for SHCCs
or macro defect free cementitious composites, are to the adjustment of VA
content.
16
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Table 4: Summary with all measured rheology parameters
Mixtures α[kPa.s/mm] β[kPa.s/mm] σ0[kPa] τ0[kPa] m n
XVA1 0.88 ±0.09 0.14 ±0.02 8.74 ±0.90 0.42 ±0.21 0.26 ±0.02 0.24 ±0.09
XVA2 0.93 ±0.17 0.39 ±0.10 9.33 ±1.66 2.03 ±0.41 0.28 ±0.04 0.69 ±0.05
XVA3 1.30 ±0.42 0.69 ±0.01 13.06 ±4.18 3.44 ±0.06 0.35 ±0.08 0.81 ±0.01
XVA4 2.10 ±0.44 0.87 ±0.05 21.07 ±4.38 4.36 ±0.27 0.49 ±0.08 0.92 ±0.03
XVA3SP1 2.43 ±0.41 0.94 ±0.06 24.11 ±4.28 4.68 ±0.30 0.55 ±0.07 0.96 ±0.04
XVA3SP3 0.20 ±0.12 0.75 ±0.04 1.99 ±1.21 3.74 ±0.18 0.12 ±0.03 0.85 ±0.02
XVA3W1 0.78 ±0.03 0.15 ±0.01 7.77 ±0.26 0.50 ±0.12 0.24 ±0.01 0.28 ±0.05
XVA3PVA10 3.09 ±0.54 0.42 ±0.10 30.90 ±5.42 2.07 ±0.48 0.68 ±0.09 0.63 ±0.06
XVA3PVA15 3.23 ±0.58 0.46 ±0.11 32.25 ±5.76 2.32 ±0.56 0.70 ±0.10 0.66 ±0.07
XVA3PVA20 3.45 ±0.51 0.51 ±0.10 34.52 ±5.06 2.53 ±0.52 0.74 ±0.08 0.69 ±0.06
XVA4PVA20 3.93 ±0.80 0.97 ±0.05 39.03 ±8.31 4.87 ±0.26 0.81 ±0.14 0.98 ±0.03
YVA1 0.41 ±0.07 0.18 ±0.02 4.08 ±0.72 0.74 ±0.20 0.17 ±0.02 0.37 ±0.09
YVA2 0.61 ±0.05 0.23 ±0.02 6.04 ±0.54 1.13 ±0.12 0.21 ±0.01 0.51 ±0.07
YVA3 0.70 ±0.06 0.34 ±0.05 7.03 ±0.63 1.85 ±0.12 0.23 ±0.01 0.71 ±0.10
YVA4 0.79 ±0.08 0.50 ±0.01 7.92 ±0.83 2.52 ±0.07 0.25 ±0.02 0.70 ±0.01
YVA3SP1 1.33 ±0.10 0.23 ±0.01 13.26 ±1.02 1.25 ±0.06 0.36 ±0.02 0.59 ±0.04
YVA3SP3 0.82 ±0.06 0.21 ±0.02 8.17 ±0.63 0.99 ±0.19 0.25 ±0.01 0.47 ±0.08
YVA3W1 0.70 ±0.07 0.100 ±0.001 6.88 ±0.73 0.102 ±0.004 0.23 ±0.01 0.101 ±0.002
YVA3PVA10 1.65 ±0.15 0.28 ±0.03 16.46 ±1.47 1.40 ±0.15 0.43 ±0.03 0.56 ±0.01
YVA3PVA15 2.09 ±0.23 0.32 ±0.03 20.88 ±2.28 1.59 ±0.14 0.51 ±0.04 0.57 ±0.02
YVA3PVA20 2.92 ±0.18 0.19 ±0.03 29.22 ±1.83 0.90 ±0.18 0.65 ±0.03 0.46 ±0.04
YVA4PVA20 3.60 ±0.19 0.50 ±0.02 35.96 ±1.90 2.51 ±0.08 0.77 ±0.03 0.69 ±0.01
YVA3PVA20S05 2.56 ±0.19 0.43 ±0.02 25.63 ±1.89 2.16 ±0.09 0.59 ±0.03 0.65 ±0.01
YVA4PVA20S05 3.43 ±0.20 0.50 ±0.02 34.26 ±1.98 2.49 ±0.08 0.74 ±0.03 0.69 ±0.01
17
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Figure 6: An example total pressure on the fluid measured for different extrudate veloci-
ties.
4.1.2. Effect of water content
The influence of extra water in the mixtures was investigated at matrix
level. By keeping the superplasticizer and methylcellulose constant, the
effect of increasing the water-to-solid ratio from 0.2 to 0.3 was investigated.
As expected, a higher volume of water in the solution changes signif-
icantly the flowability of mixtures where the amount of liquid to wet the
surfaces of the particles is already limited. Therefore, the decrease of X
series’ bulk yield stress was considerably larger than the one observed for
SMCE, as reported on figure 10. Anyhow, as the amount of liquid to lubri-
cate the movement of the particles against each other is higher, the largest
influence of the increase of the amount of water on the mixture can be ob-
served on the shear yield stress. The influence of the water-to-solid ratio on
the paste fluidity can be observed in figure 9.
4.1.3. Effect of superplasticizer content
A subsequent investigation targeted the effect of the superplasticizer
dosage at 1, 2, and 3% of the total powder phase. Keeping constant the
percentages of methylcellulose and water-to-solid ratio it could be noticed
that the influence on the rheology properties were not as remarkable as the
18
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
(a) XVA1 (b) XVA2 (c) XVA3 (d) XVA4
(e) YVA1 (f ) YVA2 (g) YVA3 (h) YVA4
Figure 7: Visual inspection of extruded material with different amounts of VA.
19
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Figure 8: A summary of the effect of VA content on initial bulk and shear yield stress.
(a) XVA3 (b) XVA3W1 (c) YVA3 (d) YVA3W1
Figure 9: Visual inspection of extruded material for different water-to-solid ratio.
20
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Figure 10: A summary of the effect of water-to-solid ratio on initial bulk and shear yield
stress.
one measured when the water-to-solid ratio was investigated.
The initial bulk yield stress increased or decreased whenever the amount
of superplasticizer was changed from 1 to 3% (figure 12). However, a de-
crease of approximately 32% of the initial shear yield stress when 1% of
superplasticizer was employed on Y matrix was measured. This decrease
might be correlated with the excess of liquid present in this solid suspen-
sion, as the particle size went up to 1 mm and there is a considerable high
usage of fly ash. Visually, the influence of the amount of superplasticizer
can be observed in figure 11.
4.1.4. Effect of PVA fibre reinforcement
As described in the introduction, the goal of this research was to de-
velop a printable SHCC mix design. Therefore, the influence of 1.0, 1.5 and
2.0% by total volume of fibre reinforcement on the above described matrices
was studied. For both matrices (X and Y) the content of superplasticizer,
methylcellulose and water-to-solid ratio were chosen to be 2%, 0.3% and 0.2
respectively.
For both mixtures, the fibre reinforcement increased considerably the
initial bulk and shear yield stresses. The values of σ0at least doubled when
the fibre reinforcement was incorporated, as can be seen on figure 14. Zhou
21
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
(a) XVA3SP1 (b) XVA3 (c) XVA3SP3 (d) YVA3SP1
(e) YVA3 (f) YVA3SP3
Figure 11: Visual inspection of extruded material for different SP concentrations.
22
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Figure 12: A summary of the effect of SP content on initial bulk and shear yield stress.
et al. 2005, explained this behaviour by attributing this increase to the raise
of friction between fibres and particles in the matrix while the mixture is
being extruded.
However, when the rheological properties of both mixtures were ob-
served with the increasing volume of fibres, they demonstrated a different
behaviour. The increased volume of fibres did not significantly change σ0
and τ0for the X composites. On the other hand, the enlarged volume of
reinforcement in the Y matrix significantly increased σ0. These dissimilar
effects can be attributed to the different particle size distributions of the
respective composites. The X matrix was tailored to minimize the space
between all the composing matrix particles, including the fibres. The Y ma-
trix, on the other hand, presents larger gaps between the aggregates, where
the increasing volume of fibres can be allocated. Visual inspections can be
done with the help of figure 13. There the shape stability of the extruded
filaments as well as the influence of the volume of fibre reinforcement and
0.4% of VA can be assessed. Comparing the rheological results obtained
and summarized in figure 14 with the images in figure 13, provides a clear
correlation between the shape stability and the increase values of initial bulk
yield stress.
23
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
(a) XVA3PVA10 (b) XVA3PVA15 (c) XVA3PVA20 (d) XVA4PVA20
(e) YVA3PVA10 (f) YVA3PVA15 (g) YVA3PVA20 (h) YVA4PVA20
Figure 13: Visual inspection of extruded material for different levels of PVA fibre rein-
forcement.
24
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Figure 14: A summary of the effect of PVA fibres volume on initial bulk and shear yield
stress.
4.1.5. Effect of sand maximum grain size on Y reinforced by 2% of PVA
fibres
Although the TU/e 3DCP facility [32] is capable of processing mixtures
with a particle size of up to 2 mm, it was observed in preliminary trials that
the probability of blockage in the pump system significantly decreased when
a less viscous mixture was used or the maximum grain size was reduced.
However, during the trials only with the pump, the probability of blockage
was considerably higher when aggregates up to 1 mm were employed. This
risk was only decreased when a less viscous mixture was employed at the
expense of worsening the shape stability. As reported above, if more water
or superplasticizer is added to this mixture the consequences would be the
decrease on the initial bulk and shear yield stresses. Hence, those changes
would lead to losses on the shape stability culminating with a less stable
mixture which could segregate while the composite is pumped.
Therefore, the influence of the maximum grain size of sand used on Y
composites was evaluated through the rheological parameters. Mixtures em-
ploying 0.3 and 0.4 wt.% of the total solid content of VA, with the maximum
sand grain size of 0.5 mm, and keeping 2% of PVA fibres by volume, 2%
of superplasticizer by total solid weight and 0.2 water-to-solid ratio were
evaluated.
25
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
(a) YVA3PVA20 (b) YVA4PVA20 (c) YVA3PVA20S05 (d) YVA4PVA20S05
Figure 15: Visual inspection of extruded material for sand maximum grain size of 0.5 mm
and 1 mm.
The results showed that by decreasing the maximum grain size of the
sand slightly lowered the bulk yield stress of composites with 0.3 wt.% of
VA, as it can be seen on figure 16. However, when employing 0.4 wt.% of
VA, the values of σ0remained in the same range as the ones observed for
composites employing 1 mm sand. Furthermore, the τ0of composites with
0.3% of VA increased considerably in comparison with composites with 1
mm sand. This result demonstrates that the use of smaller maximum grain
size for sand contributes to the development of a more packed composite
with an initial bulk yield stress comparable to what was obtained for the X
mixtures. In figure 15, the influence of the sand grain size and the amount
of VA can be observed.
4.2. Printing experiments
4.2.1. Initial trial
Based on the visual assessment and quantitative rheology properties,
six mixtures were selected for the initial trial with the pump and a 5 m
hose. The results are summarized in table 5. Two mixtures could not be
pumped as they led to blockage of the linear displacement pump. It could
therefore not be established whether they could be extruded through the 5
m hose. The blockages were likely caused by the maximum grain size of the
respective mixtures that turned out to be incompatible with the pumping
system. Excessive bulk and shear yield stresses of these mixtures were not
the cause, as they were in the same range as those of the other mixtures.
One other (XVA4PVA20) could pass through the pump, but generated
too much friction in the hose to be transported through it. This seems to
26
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Figure 16: A summary of the effect on PVA fibres volume.
Table 5: Results of initial pumpability and extrudability trials.
Mixture σ0[kPa] τ0[kPa] mixture can pass through...
pump 5 m hose
XVA3PVA20 34.74 ±5.20 4.41 ±0.18 X X
XVA4PVA20 39.03 ±8.31 4.87 ±0.26 X×
YVA3PVA20 29.22 ±1.83 0.90 ±0.18 ×(n/a)
YVA4PVA20 35.96 ±1.90 2.51 ±0.08 ×(n/a)
YVA3PVA20S05 25.63 ±1.89 2.16 ±0.09 X X
YVA4PVA20S05 34.26 ±1.98 2.49 ±0.08 X X
27
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
correspond to a limit of shear yield stress having been exceeded for the sys-
tem to which the mixture was applied, as it is higher than that of three
mixtures that could both be pumped and transported. Further experiments
are required to further elucidate the apparent relations between the rheo-
logical parameters bulk and shear yield stress on one hand, and mixture
behaviour in the printing process.
Considering these results, for printing, the mixture with the highest bulk
yield stress that still fulfils the pumpability and extrudability requirement
(i.e. does not exceed the shear yield stress limit), should be selected, as it
should result in optimal buildability. Thus, two mixtures (XVA3PVA20 and
YVA4PVA20-S05) seem to be comparably suitable. Their shape stability is
also visually apparent from the rheology tests, as shown in figure 13(c) and
15(d). For practical purposes, the YVA4PVA20-S05 mixture was selected
for the object printing experiment. The third mixture (YVA3PVA20-S05),
whilst being pumpable, was expected to have lower buildability due to the
lower bulk yield stress, and was thus disregarded.
4.2.2. Object printing experiment
Analysis methods to predict the buildability are still under development.
Wolfs et al. [65] have presented a solid mechanics based approach consider-
ing both stability and material failure effects, whereas others (Roussel, 2018
[67]) have proposed a rheology based failure criterion. As the print material
develops from a highly viscous to a solid state after deposition, both ap-
proaches have merit. An extensive discussion of this issue falls outside the
scope of this study. For now, a rheology based approximation of the expected
buildability was calculated to be 17 layers, based on the measured bulk yield
stress of (34.26 ±1.98) kPa, an assumed mass density of 2,000 kg/m3, and
an average layer height of 0.01 m. The mixture has an excessive open time
of more than 12 hours. Therefore, structural build-up during printing was
ignored in this estimation. The progress of the object printing experiment
is shown in figure 17. The object collapsed during printing of the 14th layer.
The calculated 17 layers apparently is a considerable overestimation, but it
is nevertheless in the same order of magnitude. The deviation is likely due
to stability effects that depend on the 3D geometry, density variations, and
dynamically changing loads caused by the deposition of the print filament.
It may nevertheless be concluded that the mixture is printable, and further
adaptations of it should be considered to improve buildability.
28
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
(a) 3rd layer (b) 6th layer (c) 9th layer
(d) 13th layer (e) 14th layer
Figure 17: Cylinder printing test.
4.3. Mechanical properties
Among all evaluated mixtures, only those reinforced by 2 vol.% of fibres
were evaluated mechanically. The influence of 0.3 and 0.4% by solid weight
of VA, as well as the maximum grain size of the sand, were evaluated through
compressive strength and four-point bending tests.
All tested mixtures delivered flexural hardening and developed a ductile
behaviour, as given in Figure 18. Nevertheless, multiple cracks were more
often observed in X series and in mixtures in which the smaller grain size
of the sand was employed. This behaviour was expected since the size of
aggregates influences the number of cracks and the crack pattern, as demon-
strated by [68]. Additionally, larger aggregates can also make the dispersion
of fibres more difficult, decreasing the number of fibres effectively bridg-
ing the cracks [69]. Figure 19 summarizes the results from the four-point
bending test.
Meanwhile, the compressive strength values of all analysed mixtures,
listed in table 6, were significantly higher for X’s, in comparison with the
Ys. The amount of VA employed in the composites did not significantly
influenced the compressive strength or the flexural behaviour of the analysed
composites, with exception of the ones with sand up to 0.5 mm. Only in
this case, composites with 0.4 wt.% of VA delivered higher performance.
As demonstrated, the mechanical performance was not influenced by the
amount of VA employed on X mixtures and 0.4 wt.% improves Y-S05 series.
Moreover, as discussed in the previous section, only XVA3PVA20 and
YVA4PVA20-S05 were suitable for printing. Therefore, only these two mix-
29
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Figure 18: An example from the obtained flexural hardening behaviour obtained from
mixture XVA3PVA20
.
Figure 19: Average performance on four point bending test of selected mixtures.
30
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Table 6: Compressive strength performance [MPa]
Mix design Load perpendicular to
the casting direction
Load parallel to the
casting direction
XVA3 41.12 ±4.18 41.71 ±3.21
XVA4 38.26 ±5.48 37.76 ±3.45
YVA3 15.89 ±0.95 14.37 ±1.81
YVA4 14.49 ±1.48 14.78 ±1.3
YVA3S05 8.39 ±1.82 8.49 ±0.5
YVA4S05 15.47 ±0.65 15.84 ±1.58
tures were chosen to be investigated employing an uni-axial tensile test, in
order to confirm their strain hardening and multiple crack behaviour. In
figures 20 and 21 the ”tensile stress versus strain” curve of both mixtures
are plotted. It could be confirmed that the modified composites mix-design
also showed high ductility and strain hardening behaviour.
In figures 22, 23, 24, 25 the last picture from the DIC analyses are
shown. In pictures depicting the vertical displacements only elongations
were shown. Regions in blue suffered less deformation than regions with
colours closer to red, where the cracks are. On pictures showing the hori-
zontal displacements elongations and compressions are shown. There, zero
displacements were demonstrated with yellow colours with elongations been
demonstrated in red and compressions in blue. Through the DIC analysis
the horizontal displacements could also be captured. Where the cracks were
concentrated it was possible to observe areas with compressive values and
some others in tension. These regions are believed to have fibres oriented
with different angles, which emphasizes the importance of fibre dispersion
in this type of composite. It is possible to observe that all samples devel-
oped at least two cracks, and specimens from the X series resulted in higher
tensile performances and considerably greater number of cracks.
5. Conclusions
Through the experimental procedure carried out in this study, a method-
ology was presented to develop printable cementitious composite mix-designs
based on fundamental rheological properties. The influence of rheology mod-
ifiers on the fresh and hardened state were evaluated for application in high
performance cementitious composites. Visual inspection together with rhe-
ology parameters evaluation were employed to obtain an optimized SHCC
31
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Figure 20: Tensile performance of X series.
Figure 21: Tensile performance of Y series.
32
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
(a) Specimen 1 (b) Specimen 2 (c) Specimen 3 (d) Specimen 4 (e) Specimen 5
Figure 22: Crack pattern obtained from DIC analysis on X samples - Horizontal Defor-
mations.
(a) Specimen 1 (b) Specimen 2 (c) Specimen 3 (d) Specimen 4 (e) Specimen 5
Figure 23: Crack pattern obtained from DIC analysis on X samples - Vertical Deforma-
tions.
(a) Specimen1 (b) Specimen 2 (c) Specimen 3 (d) Specimen 4 (e) Specimen 5
Figure 24: Crack pattern obtained from DIC analysis on Y samples - Horizontal Defor-
mations.
33
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
(a) Specimen 1 (b) Specimen 2 (c) Specimen 3 (d) Specimen 4 (e) Specimen 5
Figure 25: Crack pattern obtained from DIC analysis on Y samples - Vertical Deforma-
tions.
mixture in terms of printability, shape stability and strain hardening be-
haviour. Printing experiments were conducted to compare the pumpability
and extrudability of various mixtures. One mixture was used to print an ob-
ject and evaluate buildability. Mechanical tests were performed to confirm
the strain hardening behaviour of the developed mixtures. In summary the
following conclusions can be drawn:
•Ram extruder with the Benbow-Bridgwater model are appropriate
tools to develop printable cementitious composites. However, it is im-
portant to notice that the method has limitations. One example was
the inability to predict the blockage of some mixtures in the pump;
•The employment of rheology modifiers is crucial for the development
of high ductility cementitious composites, with dough-like consistency
in the fresh state;
•The liquid-to-solid ratio of the solid suspension is more relevant to the
shape stability of printable mixtures, than the amount of superplasti-
cizer;
•For the development of a printable cementitious mixtures, the particle
size distribution and the liquid-to-total surface area of all solids is more
relevant than the employment of rheology modifiers;
•The amount of rheology modifiers employed in the mix did not signif-
icantly influence the mechanical properties of most evaluated compos-
ites.
34
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
•XVA3PVA20 and YVA4PVA20-S05 are composites which have proven
to have high mechanical performance and superior printing quality and
therefore, must be considered for further developments in the construc-
tion printing industry.
6. Acknowledgements
This study was performed as part of the 2017 4.TU Lighthouse project
3D Concrete Printing for Structural Applications, that was performed jointly
by the Eindhoven and Delft Universities of Technology. The support of the
4.TU Federation is gratefully acknowledged
In addition, the first author would like to acknowledge the funding from
Science Without Borders from the National Council for Scientific and Tech-
nological Development of Brazil (201620/2014-6). The second author ac-
knowledges the financial support from the Construction Technology Re-
search Program funded by the Ministry of Land, Infrastructure and Trans-
port of the Korean Government under the grant 17SCIP-B103706-03. The
fifth author acknowledges the financial support from China Scholarship
Council (CSC) under the grant CSC No.201708110187.
The concrete printing was performed at the Eindhoven University of
Technology (TU/e) 3D Concrete Printing (3DCP) Facility. The TU/e re-
search program on 3DCP is co-funded by a partner group of enterprises
and associations, that on the date of writing consisted of (alphabetical or-
der) Ballast Nedam, BAM Infraconsult bv, Bekaert, Concrete Valley, CRH,
Cybe, Saint-Gobain Weber Beamix, SGS Intron, SKKB, Van Wijnen, Ver-
hoeven Timmerfabriek, and Witteveen+Bos. Their support is gratefully
acknowledged.
Finally, the authors acknowledge the support of Kuraray by providing
the PVA fibres used in this study.
7. Data availability
The raw/processed data required to reproduce these findings cannot be
shared at this time as the data also forms part of an ongoing study.
8. References
[1] B. E. Allmon, C. T. Haas, J. D. Borcherding, and P. M. Goodrum,
“U. S. Construction Labor Productivity Trends, 1970-1998,”
Journal of construction engineering and management, vol. 126,
no. April, pp. 97–104, 2000.
35
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
[2] R. R. Ramrez, L. F. C. Alarc´on, and P. Knights, “Benchmarking
System for Evaluating Management Practices in the Construction
Industry,” Journal of Management in Engineering, vol. 20, no. 3,
pp. 110–117, 2004.
[3] H. Kodama, “Automatic method for fabricating a three-
dimensional plastic model with photo-hardening polymer,” Review
of Scientific Instruments, vol. 52, no. 11, pp. 1770–1773, 1981.
[4] E. Sachs, M. Cima, J. Cornie, D. Brancazio, J. Bredt, A. Curodeau,
T. Fan, S. Khanuja, A. Lauder, J. Lee, and S. Michaels, “Three-
Dimensional Printing: The Physics and Implications of Addi-
tive Manufacturing,” CIRP Annals - Manufacturing Technology,
vol. 42, no. 1, pp. 257–260, 1993.
[5] A. C. De Leon, Q. Chen, N. B. Palaganas, J. O. Palaganas,
J. Manapat, and R. C. Advincula, “High performance polymer
nanocomposites for additive manufacturing applications,” Reac-
tive and Functional Polymers, vol. 103, pp. 141–155, 2016.
[6] L. E. Murr, S. M. Gaytan, F. Medina, E. Martinez, J. L. Martinez,
D. H. Hernandez, B. I. Machado, D. A. Ramirez, and R. B. Wicker,
“Characterization of Ti-6Al-4V open cellular foams fabricated by
additive manufacturing using electron beam melting,” Materials
Science and Engineering A, vol. 527, no. 7-8, pp. 1861–1868, 2010.
[7] J. Luo, H. Pan, and E. C. Kinzel, “Additive Manufacturing
of Glass,” Journal of Manufacturing Science and Engineering,
vol. 136, no. 6, p. 061024, 2014.
[8] J. Willmann, M. Knauss, T. Bonwetsch, A. A. Apolinarska, F. Gra-
mazio, and M. Kohler, “Robotic timber construction - Expanding
additive fabrication to new dimensions,” Automation in Construc-
tion, vol. 61, pp. 16–23, 2016.
[9] D. Herzog, V. Seyda, E. Wycisk, and C. Emmelmann, “Additive
manufacturing of metals,” Acta Materialia, vol. 117, pp. 371–392,
2016.
[10] D-Shape, “What is D-Shape.”
[11] B. Khoshnevis, “Automated construction by contour crafting - Re-
lated robotics and information technologies,” Automation in Con-
struction, vol. 13, no. 1, pp. 5–19, 2004.
36
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
[12] S. Lim, R. A. Buswell, T. T. Le, S. A. Austin, A. G. Gibb, and
T. Thorpe, “Developments in construction-scale additive manu-
facturing processes,” Automation in Construction, vol. 21, no. 1,
pp. 262–268, 2012.
[13] C. Gosselin, R. Duballet, P. Roux, N. Gaudilli`ere, J. Dirrenberger,
and P. Morel, “Large-scale 3D printing of ultra-high performance
concrete - a new processing route for architects and builders,” Ma-
terials and Design, vol. 100, pp. 102–109, 2016.
[14] E. Lloret Fritschi, L. Reiter, T. Wangler, F. Gramazio, M. Kohler,
and R. J. Flatt, “SMART DYNAMIC CASTING SLIPFORMING
WITH FLEXIBLE FORMWORK - INLINE MEASUREMENT
AND CONTROL,” in HPC/CIC Tromsø 2017. Eleventh High Per-
formance Concrete (11th HPC) and the Second Concrete Innova-
tion Conference (2nd CIC), (Tromsø), Norwegian Concrete Asso-
ciation, 2017.
[15] N. Hack and W. V. Lauer, “Mesh-Mould: Robotically Fabricated
Spatial Meshes as Reinforced Concrete Formwork,” Architectural
Design, vol. 84, no. 3, pp. 44–53, 2014.
[16] F. Bos, R. Wolfs, Z. Ahmed, and T. Salet, “Additive manufac-
turing of concrete in construction: potentials and challenges of 3D
concrete printing,” Virtual and Physical Prototyping, vol. 11, no. 3,
pp. 209–225, 2016.
[17] Y. W. D. Tay, B. Panda, S. C. Paul, N. A. N. Mohamed, M. J. Tan,
and K. F. Leong, “3d printing trends in building and construction
industry: a review,” Virtual and Physical Prototyping, vol. 12,
no. 3, pp. 261–276, 2017.
[18] D. Delgado, P. Clayton, W. J. O. Brien, C. Seepersad, M. Juenger,
R. Ferron, and S. Salamone, “Applications of additive manufactur-
ing in the construction industry A forward-looking review,” Au-
tomation in Construction, vol. 89, no. August 2017, pp. 110–119,
2018.
[19] H. A. Jassmi, F. A. Najjar, and A.-h. I. Mourad, “Large-Scale 3D
Printing : The Way Forward Large-Scale 3D Printing : The Way
Forward,” in IOP Conf. Series: Materials Science and Engineer-
ing, p. 324, IOP Publishing, 2018.
37
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
[20] B. Nematollahi, M. Xia, and J. Sanjayan, “Current Progress of
3D Concrete Printing Technologies,” in 34th International Sympo-
sium on Automation and Robotics in Construction (ISARC 2017),
(Taipei, Taiwan), 2017.
[21] S. Hamidreza, J. Corker, and M. Fan, “Additive manufactur-
ing technology and its implementation in construction as an
eco-innovative solution,” Automation in Construction, vol. 93,
no. May, pp. 1–11, 2018.
[22] R. A. Buswell, W. R. L. D. Silva, S. Z. Jones, and J. Dirrenberger,
“Cement and Concrete Research 3D printing using concrete extru-
sion : A roadmap for research,” Cement and Concrete Research,
vol. 112, no. May, pp. 37–49, 2018.
[23] M. A. Guowei, W. Li, and J. U. Yang, “State-of-the-art of 3D print-
ing technology of cementitious material An emerging technique for
construction,” Science China Technological Sciences, vol. 61, no. 4,
pp. 475–495, 2018.
[24] Agence France-Presse, “World’s first 3D-printed bridge opens to
cyclists in Netherlands,” 2017.
[25] K. Subrin, T. Bressac, S. Garnier, A. Ambiehl, E. Paquet,
and B. Furet, “Improvement of the mobile robot location ded-
icated for habitable house construction by 3d printing,” IFAC-
PapersOnLine, vol. 51, no. 11, pp. 716 – 721, 2018. 16th IFAC
Symposium on Information Control Problems in Manufacturing
INCOM 2018.
[26] A. Kazemian, X. Yuan, E. Cochran, and B. Khoshnevis, “Cemen-
titious materials for construction-scale 3D printing : Laboratory
testing of fresh printing mixture,” Construction and Building Ma-
terials, vol. 145, pp. 639–647, 2017.
[27] R. A. Buswell, R. C. Soar, A. G. F. Gibb, and A. Thorpe,
“Freeform Construction: Mega-scale Rapid Manufacturing for con-
struction,” Automation in Construction, vol. 16, no. 2, pp. 224–
231, 2007.
[28] P. Wu, J. Wang, and X. Wang, “A critical review of the use of 3-D
printing in the construction industry,” Automation in Construc-
tion, vol. 68, pp. 21–31, 2016.
38
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
[29] G. D. Schutter, K. Lesage, V. Mechtcherine, V. Naidu, G. Habert,
and I. Agusti-juan, “Cement and Concrete Research Vision of 3D
printing with concrete Technical , economic and environmental
potentials,” Cement and Concrete Research, vol. 112, no. August,
pp. 25–36, 2018.
[30] V. Nerella, V. N., Krause, M., N¨ather, M., Mechtcherine, “3D
printing technology for on-site construction,” Concr. Plant Int,
pp. 36–41, 2016.
[31] V. N. Schach, R.; Krause, M.; N¨ather, M.; Nerella, “CONPrint3D:
3D concrete-printing as an alternative for masonry,” Bauingenieur,
vol. 9, pp. 355–363, 2017.
[32] F. P. Bos, Z. Y. Ahmed, E. R. Jutinov, and T. A. Salet, “Exper-
imental exploration of metal cable as reinforcement in 3D printed
concrete,” Materials, vol. 10, no. 11, 2017.
[33] D. Asprone, F. Auricchio, C. Menna, and V. Mercuri, “3D printing
of reinforced concrete elements: Technology and design approach,”
Construction and Building Materials, vol. 165, pp. 218–231, 2018.
[34] R. Duballet, O. Baverel, and J. Dirrenberger, “Space Truss
Masonry Walls With Robotic Mortar Extrusion,” Structures,
no. November, pp. 1–7, 2018.
[35] B. Panda, S. Chandra Paul, and M. Jen Tan, “Anisotropic me-
chanical performance of 3D printed fiber reinforced sustainable
construction material,” Materials Letters, vol. 209, pp. 146–149,
2017.
[36] Y. Weng, M. Li, Z. Liu, W. Lao, B. Lu, D. Zhang, and M. J. Tan,
“Printability and fire performance of a developed 3d printable fibre
reinforced cementitious composites under elevated temperatures,”
Virtual and Physical Prototyping, vol. 0, no. 0, pp. 1–9, 2018.
[37] F. P. Bos, E. Bosco, and T. A. M. Salet, “Ductility of 3d printed
concrete reinforced with short straight steel fibers,” Virtual and
Physical Prototyping, vol. 14, no. 2, pp. 160–174, 2019.
[38] V. C. Li, “On engineered cementitious composites (ECC). A review
of the material and its applications,” Journal of Advanced Concrete
Technology, vol. 1, no. 3, pp. 215–230, 2003.
39
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
[39] S. M¨uller and V. Mechtcherine, “Fatigue behaviour of strain-
hardening cement-based composites (SHCC),” Cement and Con-
crete Research, vol. 92, pp. 75–83, 2017.
[40] V. C. Li, H. Horii, P. Kabele, T. Kanda, and Y. Lim, “Repair
and retrofit with engineered cementitious composites,” Engineer-
ing Fracture Mechanics, vol. 65, no. 2-3, pp. 317–334, 2000.
[41] F. Altmann and V. Mechtcherine, “Durability design strategies
for new cementitious materials,” Cement and Concrete Research,
vol. 54, pp. 114–125, 2013.
[42] D. G. Soltan and V. C. Li, “A self-reinforced cementitious com-
posite for building-scale 3d printing,” Cement and Concrete Com-
posites, vol. 90, pp. 1 – 13, 2018.
[43] V. N. Nerella, H. Ogura, and V. Mechtcherine, “Incorporating rein-
forcement into digital concrete construction,” in Proceedings of the
IASS Symposium 2018 Creativity in Structural Design, (Boston),
2018.
[44] M. Rubio, M. Sonebi, and S. Amziane, “3D PRINTING OF FIBRE
CEMENT-BASED MATERIALS: FRESH AND RHEOLOGICAL
PERFORMANCES,” in 2nd International Conference On Bio-
Based Building Materials (S. AMZIANE and M. SONEBI, eds.),
(Clermont Ferrand), RILEM, 119 (PRO 119), 2017.
[45] G. Ma, Z. Li, and L. Wang, “Printable properties of cementitious
material containing copper tailings for extrusion based 3D print-
ing,” Construction and Building Materials, vol. 162, pp. 613–627,
2018.
[46] R. J. Wolfs, F. P. Bos, and T. A. Salet, “Early age mechani-
cal behaviour of 3D printed concrete: Numerical modelling and
experimental testing,” Cement and Concrete Research, vol. 106,
no. February, pp. 103–116, 2018.
[47] M. Hambach and D. Volkmer, “Properties of 3D-printed fi ber-
reinforced Portland cement paste,” Cement and Concrete Com-
posites, vol. 79, pp. 62–70, 2017.
[48] X. Zhou and Z. Li, “Characterization of rheology of fresh fiber rein-
forced cementitious composites through ram extrusion,” Materials
and Structures, vol. 38, no. February, pp. 17–24, 2005.
40
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
[49] J. J. Benbow, “The dependence of output rate on die shape dur-
ing catalyst extrusion,” Chemical Engineering Science, vol. 26,
pp. 1467–1473, 1971.
[50] J. J. Benbow and J. Bridgwater, “The Influence of Formulation
on Extrudate Structure and Strength,” Chemical Engineering Sci-
ence, vol. 42, pp. 753–766, 1987.
[51] J. J. Benbow and E. W. Oxley, “THE EXTRUSION MECHAN-
ICS OF PASTES- THE INFLUENCE OF PASTE FORMULA-
TION EXTRUSION PARAMETERS,” Chemical Engineering Sci-
ence, vol. 42, no. 1967, pp. 1467 – 1473, 1987.
[52] S. Srinivasan, D. Deford, and P. Shah, “The use of extrusion
rheometry in the development of extrudate fibre-reinforced cement
composites,” Concrete Science and Engineering, vol. 1, no. 11,
pp. 26–36, 1999.
[53] R. Nath Das, C. D. Madhusoodana, and K. Okada, “Rheological
studies on cordierite honeycomb extrusion,” Journal of the Euro-
pean Ceramic Society, vol. 22, no. 16, pp. 2893–2900, 2002.
[54] X. Zhou, Z. Li, M. Fan, and H. Chen, “Cement & Concrete Com-
posites Rheology of semi-solid fresh cement pastes and mortars
in orifice extrusion,” Cement and Concrete Composites, vol. 37,
pp. 304–311, 2013.
[55] R. Teixeira, G. Tonoli, S. Santos, J. Fiorelli, H. Savastano Jr., and
F. R. Lahr, “Extruded Cement Based Composites Reinforced with
Sugar Cane Bagasse Fibres,” Key Engineering Materials, vol. 517,
pp. 450–457, 2012.
[56] S. F. Santos, R. Schmidt, A. E. F. S. Almeida, G. H. D. Tonoli, and
H. Savastano, “Supercritical carbonation treatment on extruded
fibre cement reinforced with vegetable fibres,” Cement & Concrete
Composites, vol. 56, pp. 84–94, 2015.
[57] V. da Costa Correia, S. F. Santos, R. Soares Teixeira, and H. Savas-
tano Junior, “Nanofibrillated cellulose and cellulosic pulp for rein-
forcement of the extruded cement based materials,” Construction
and Building Materials, vol. 160, pp. 376–384, 2018.
41
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
[58] S. F. U. Ahmed and H. Mihashi, “A review on durability prop-
erties of strain hardening fibre reinforced cementitious compos-
ites (SHFRCC),” Cement and Concrete Composites, vol. 29, no. 5,
pp. 365–376, 2007.
[59] M. ahmaran and V. C. Li, “Durability properties of micro-cracked
ECC containing high volumes fly ash,” Cement and Concrete Re-
search, vol. 39, no. 11, pp. 1033–1043, 2009.
[60] J. Zheng, W. B. Cadson, and J. S. Reed, “Flow Mechanics on
Extrusion through a Square-Entry Die,” Journal of the American
Ceramic Society, no. 75, pp. 3011–3016, 1992.
[61] J. J. Benbow, S. H. Jazayeri, and J. Bridgwater, “The flow of
pastes through dies of complicated geometry,” Powder Technology,
vol. 65, pp. 393–401, 1991.
[62] J. Zhou, S. Qian, M. G. Sierra Beltran, G. Ye, K. Breugel, and
V. C. Li, “Development of engineered cementitious composites
with limestone powder and blast furnace slag,” Materials and
Structures, vol. 43, no. 6, pp. 803–814, 2010.
[63] P. Wang, F. H. Wittmann, P. Zhang, E. Lehmann, and T. Zhao,
“Durability and Service Life of Elements Made With Shcc Under
Imposed Strain,” in SHCC3 - 3rd International RILEM Confer-
ence on Strain Hardening Cementitious Composites, pp. 33–41,
2014.
[64] Y. J. M. Soto, Adequa¸c˜ao de formula¸c˜oes para a produ¸c˜ao de placas
de fibrocimento por extrus˜ao. PhD thesis, Escola Polit´ecnica da
Universidade de S˜ao Paulo, 2010.
[65] R. Wolfs, F. Bos, and T. Salet, “Early age mechanical behaviour of
3d printed concrete: Numerical modelling and experimental test-
ing,” Cement and Concrete Research, vol. 106, pp. 103 – 116, 2018.
[66] J. Blaber, B. Adair, and A. Antoniou, “Ncorr: Open-Source 2D
Digital Image Correlation Matlab Software,” Experimental Me-
chanics, vol. 55, no. 6, pp. 1105–1122, 2015.
[67] N. Roussel, “Rheological requirements for printable concretes,”
Cement and Concrete Research, no. April, pp. 1–10, 2018.
42
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
[68] X. Huang, R. Ranade, W. Ni, and V. C. Li, “Development of green
engineered cementitious composites using iron ore tailings as aggre-
gates,” Construction and Building Materials, vol. 44, pp. 757–764,
2013.
[69] G. Song and G. V. Zijl, “Tailoring ECC for Commercial Ap-
plication,” Proceedings of the 6th RILEM symposium on fiber,
no. September, pp. 1391–1400, 2004.
43
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
CRediT author statement
Stefan Chaves Figueiredo: Conceptualization; Methodology; Software; Validation; Formal
Analysis; Investigation; Data Curation; Writing – Original Draft; Writing – Review & Editing;
Claudia Romero Rodríguez: Conceptualization; Methodology; Validation; Investigation; Writing
– Original Draft; Writing – Review & Editing;
Zeeshan Y. Ahmed: Validation; Investigation; Resources
D. H. Bos: Investigation
Yading Xu: Validation; Investigation; Writing – Review & Editing;
Theo M. Salet: Resources; Supervision;
Oğuzhan Çopuroğlu: Resources; Writing – Review & Editing; Supervision;
Erik Schlangen: Resources; Writing – Review & Editing; Supervision;
Freek P. Bos: Investigation; Resources; Writing – Original Draft; Writing – Review & Editing;
Supervision;
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Graphical abstract
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Highlights
A quantitative methodology based on rheological parameters to develop printable
cementitious composites is presented;
The methodology was successfully applied to the development of strain hardening
cementitious composites, thereby addressing the key issue of lack of ductility found in
most concrete printing processes
A correlation between shape stability and buildability with the initial bulk yield stress was
found. Therefore, a quantitative analysis for these parameters might be possible;
The use of viscosity modifier admixtures and the total liquid-to-solid ratio are key factors
to control the mixture stability and the fibre dispersion of a cementitious composite;
Rheological properties measured with the ram extruder coupled with the Benbow-
Bridgwater model can indicate the achievable build height of an object being printed.
ACCEPTED MANUSCRIPT
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 10
Figure 11
Figure 12
Figure 13
Figure 14
Figure 15
Figure 16
Figure 17
Figure 18
Figure 19
Figure 20
Figure 21
Figure 22
Figure 23
Figure 24
Figure 25
