A B S T R A C T
Concrete digital fabrication is an innovative construction approach where infrastructural elements can be built
additively without using formwork. This represents a significant advantage, but also introduces materials en-
gineering challenges, as the requirements normally fulfilled by the formwork are now imposed on the concrete.
In this paper, it is discussed how admixtures can be employed to achieve the rheological and hydration prop-
erties necessary for printable concrete. An overview of various admixtures currently implemented in standard
practice is presented. Then, the main required concrete states for extrusion and deposition processes are ana-
lyzed with respect to required performances and potential admixtures. Finally, possible side effects and in-
compatibilities are discussed, as well as how they could be unconventionally used for printable concrete pur-
poses. The main objective is to demonstrate how admixtures will be critical in the development of concrete
systems to realize digital fabrication, and to ultimately motivate investigation in the key areas discussed.
1. Introduction allow structures to be built additively and thereby eliminate the need
for formwork. This mainly concerns extrusion printing, although other
The use of concrete in construction has primarily been limited to techniques, such as slip casting or spraying, could also benefit from the
static and standardized formwork casting, where fresh concrete is concepts presented in this analysis.
poured into formwork, often vibrated to allow sufficient placement, and Building without formworks introduces a number of advantages,
left to set. In standard practice, it is expected that concrete reaches namely savings in cost, time, and materials associated with formwork
sufficient strength a day later, at which time the formwork is stripped construction. However, at the same time, it implies some significant
and the hardening concrete is cured to sustain hydration. The proper- materials engineering challenges as all the requirements that are nor-
ties of the fresh mix, which are workability, compactibility, and re- mally fulfilled by the formwork are now directly imposed on the mix
sistance to segregation, are generally controlled or fine-tuned through design of the concrete and the way it is deposited. Therefore, in absence
the use of admixtures to facilitate the casting process. On the other of formwork, controlling the hydration and rheological properties be-
hand, admixtures also allow the control of the hydration kinetics of the comes even more critical for successful execution [4]. Hydration ki-
mix: first to avoid premature setting during transport or placement, and netics must be delayed and accelerated in relatively extreme manners
second for the formwork to be removed on time. In general, admixtures so that the material does not set during the printing process but, in-
have contributed to numerous essential developments in concrete stead, right after deposition in order to support its own weight and that
technology [1] and will continue to do so in novel approaches that are of subsequently deposited layers of material [8,9]. Concerning
part of the emerging concrete digital fabrication [2,3]. rheology, there must be a balance between flowability during printing
Digital fabrication with concrete is a family of novel fabrication and rate of structuration immediately after deposition. In addition,
processes that offers greater freedom of shape by combining computer- given the time-sensitive nature of the technique due to the continuous
aided design with additive, subtractive or forming manufacturing progression of hydration, prediction of flow rate is important to control
[3–7]. Many approaches exist, but this paper focuses on those that the printing speed.
⁎
Corresponding author.
E-mail address: dmarchon@berkeley.edu (D. Marchon).
https://doi.org/10.1016/j.cemconres.2018.05.014
Received 20 December 2017; Received in revised form 15 May 2018; Accepted 15 May 2018
0008-8846/ © 2018 Elsevier Ltd. All rights reserved.
Please cite this article as: Marchon, D., Cement and Concrete Research (2018), https://doi.org/10.1016/j.cemconres.2018.05.014
D. Marchon et al. Cement and Concrete Research xxx (xxxx) xxx–xxx
2.1.1. Superplasticizers
Dispersing agents and superplasticizers, such as lignosulfonate (LS),
polynaphthalene sulfonate (PNS), and polycarboxylate ether (PCE), are Fig. 2. Schematic representation of a PCE superplasticizer (left) with its de-
polymeric dispersants used in cementitious materials to reduce yield scriptive structural parameter, C/E, n and P (right). n represents the number of
stress and viscosity at a constant solids content (Fig. 1) [30–32]. The side chains/repeat units in one molecule. C/E represents the number of car-
enhancement of the material workability is essential for mixing, casting boxylate functions C per ester group E through which a side chain with P
and extrusion in concrete digital fabrication. Moreover, it allows for a monomers is grafted. Adapted from [29].
2
D. Marchon et al. Cement and Concrete Research xxx (xxxx) xxx–xxx
contact between particles through steric hindrance, reducing attractive the deformation of the deposited material under its own weight, making
forces [38,39]. The successful use of these comb copolymers comes it the most suitable VMA for printable concrete.
from their molecular structure (described in Fig. 2) and composition,
which can be readily tailored to provide a broad range of properties and 2.1.3. Clays and inclusions as thixotropic agents
performances [30,31]. Inorganic additions, namely clays, and organic inclusions, such as
The adsorption of these superplasticizers on cement surfaces is the polymeric fibers, can also be effective rheological modifiers. Clays
fundamental process involved in their mechanism of action [22,40]. themselves exhibit shear thinning (i.e. decrease of the apparent visc-
Furthermore, it is important to keep in mind that these admixtures can osity by increasing the shear rate) due to their opposing surface
perturb the hydration of cement particles and alter the performance of charges, which gives rise to a house of cards structure that can form at
the material in the hardened state, as will be discussed in Section 2.2.1 rest and break down under shear [49]. They have a wide range of ap-
[28]. plications as rheological modifiers, including paints and drilling fluids
[50]. Among the different families of clays, attapulgite (or paly-
gorskite), bentonite (montmorillonite-based), kaolinite, sepiolite and
2.1.2. Viscosity modifiers contaminated clays have been studied as rheological modifiers for ce-
Viscosity modifying agents (VMAs) are commonly used in industrial ment-based materials [25,26,51–57].
practice to control water transport and porous structure in both the In nature, clay particles are micron in size and consist of multilayer
fresh and hardened state [23,24,41]. The most commonly used VMAs in stacks of crystalline layers or fibers with nanometer thickness [58]. This
cementitious systems are cellulose-ether derivatives, such as welan gum is the form in which most of the clays in the aforementioned studies
and diutan gum. Anionic polyacrylamides are also used in the con- were in, i.e. layered or fibrous particles. Clay surfaces can be modified,
struction field for their high flocculating efficiency [30,42]. Such ad- by either intercalation or exfoliation, so that plates and fibers can be
mixtures are essential for controlling the risk of segregation of highly discretely dispersed as nano-clays in cementitious matrices [59].
flowable concretes, improving the water retention, and in some cases Through this modification, clay particles are then reduced in size down
for promoting particle flocculation, such as for shotcrete applications. to 100 nm (nanoclays) [60]. For instance, bulk attapulgite, a family of
VMAs can enhance the yield stress of cementitious materials and ef- fibrous hydrous magnesium silicates, can be chemically exfoliated to
fectively reduce material deformation under its own weight, which is remove impurities and swelling clays and produce rod-like particles
critical for additive manufacturing concrete. [61]. These resulting particles have an average diameter of 30 nm and
Most VMAs are able to adsorb simultaneously on several cement average length of 1.5 μm. Nanoclays can be dispersed from bulk clays
particles and bridge them, which increases the macroscopic yield stress through the use of surfactants such as sodium dodecylsulphate (SDS)
of the suspension to a certain extent [42,43]. Their efficiency as floc- and octadecyltrimethylammonium bromide (ODTMA) [62,63]. They
culants depends on their molecular weight and their affinity with the act by adsorbing onto the surfaces of the clay particles and preventing
cement surface (Fig. 3). For example, low molecular weight VMAs are re-agglomeration or restacking of the clay layers onto each other in
unable to adsorb on multiple particles and can be at the origin of a yield solution, with different affinities for different clay types [64]. It is also
stress decrease through steric hindrance [35,44–46]. According to Ta- worth mentioning that clays can be contaminated by organics or heavy
dros [47], the particle surface covered by the polymer is a key para- metals that may influence the inherent properties of the mix [65–67].
meter for controlling polymer efficiency and, specifically for adsorbing Due to the great diversity and non-consistency of the material, prop-
VMAs, the optimum bridging flocculation is attained at half surface erties of contaminated clays can vary from batch to batch and material
coverage. Also, Bessaies-Bey et al. [42] demonstrated that poly- to material, rendering them less suitable for digital fabrication appli-
acrylamide adsorption is enhanced by its anionicity, whereas cellulose cations without further processing or identification of the material.
ether adsorption decreases when the average number of substituted Generally, clays are hydrophilic, which facilitates their incorpora-
hydroxyl groups per glucose unit (i.e. degree of substitution) increases tion in a concrete mix, but can also thicken it because of the water
[42,48]. In both cases studied in cement pastes, the yield stress in- absorption and the flocculated structure they may adopt. Their corre-
creases with the molecular weight of the VMA. Furthermore, by com- sponding aggregation behavior, and thereby rheological effect, is af-
paring the flow onset of cement pastes containing polyacrylamide and fected by pH, which changes during the progression of hydration [68].
cellulose ethers, pastes containing polyacrylamide were found to ex- A number of studies have demonstrated the potential of clays to tailor
hibit a higher elastic modulus [42,48]. Therefore, polyacrylamide limits the rheology of fresh concrete mixes to facilitate extrusion [25], re-
duction in self-compacting concrete (SCC) formwork pressure [26] and
slipform paving [27], where a balance between flowability during
casting and stiffening at rest is desired. Fundamental studies on the
fresh state microstructure have tied these properties to an increase in
flocculation induced by clays [51,69]. In particular, a chemically ex-
foliated attapulgite nanoclay was found to be effective at modest
amounts (0.3% by mass of cement). Many approaches have been ap-
plied to quantify their influence on thixotropy, including steady-state
protocols [52], creep protocols [53,54], and large amplitude oscillatory
shear [55]. Generally, these studies have shown that attapulgite na-
noclay can significantly accelerate the rebuilding kinetics and rate of
increase of static yield stress over time, which is translated to a fresh-
state stiffening of the overall system. Similar effects on rebuilding ki-
netics were shown using bentonite, although not as pronounced, likely
due to their lower external specific surface area [56]. This demonstrates
the potential of utilizing clays, particularly nanoclays, for additive
manufacturing, since they can provide rapid fresh-state stiffening to
control shape stability immediately after deposition with no adverse
Fig. 3. Relative yield stress (i.e. the ratio between the yield stress of the paste effects on the flowability during printing. In fact, some early results
with polymer and the yield stress of the reference paste) as a function of VMA were presented by the Contour Crafting group at University of Southern
dosage. Reproduced from [35]. California, where attapulgite nanoclays were found to show beneficial
3
D. Marchon et al. Cement and Concrete Research xxx (xxxx) xxx–xxx
4
D. Marchon et al. Cement and Concrete Research xxx (xxxx) xxx–xxx
Once the sulfate depletion point is reached, C3A hydration is no longer 2.2.2. Accelerators
controlled and its dissolution accelerates. This increases the aluminum This section is not intended as a comprehensive review of accel-
ions concentration in the pore solution and leads to the precipitation of erators commercially available, as this has been more deeply described
aluminate hydrate phases. If this happens before the acceleration of C- elsewhere [112,113]. Its scope is rather to describe the main accel-
S-H precipitation in a case where the optimum sulfate content is not eration mechanisms, with some related examples of admixtures, that
controlled, the flash set occurs and the silicate hydration is then are of interest for printable concretes.
strongly delayed, ultimately adversely impacting the early strength. Acceleration of cement hydration is achieved in different ways that
This is why in OPC the sulfate content is well controlled, in order to can be divided into physical and chemical actions. The first physical
have the sulfate depletion point occur after the main precipitation of C- action that we may consider regarding the previous Section 2.2.1 in-
S-H. volves the desorption of retarders from the reacting surface, as in the
case of sucrose in the presence of portlandite. Indeed, it was demon-
2.2.1. Retarders strated that the molecules of sucrose tend to desorb from the cement
Retardation or acceleration of cement hydration can be achieved by surface where they originally cause retardation and diffuse in the so-
the addition of chemical admixtures. Among the organic admixtures lution to preferentially adsorb onto the portlandite, for which sucrose
that have the sole purpose of retarding hydration, carbohydrates are has the highest affinity [28,114]. It has to be mentioned that this use
very effective retarders [28,95,96]. They prolong the induction period could be limited by the high amount of portlandite needed to effectively
by delaying the onset, but once the onset occurs, they usually accelerate uptake the sucrose for a significant acceleration of hydration. The same
the reaction rate in the acceleration period. This family of sugar deri- kind of uptake phenomenon, explained in more detail in Section 4.4,
vatives offers a certain flexibility in structural variations, i.e. functio- occurs with PCEs when a certain type of clay is added.
nalization and degree of polymerization. Therein, we more often A second physical mode of acceleration was observed by Thomas
identify glucose, sodium or calcium gluconate, sucrose and sugar al- et al. [115] after the addition of C-S-H seeds in cement paste. Specifi-
cohols [97–100]. Other organics have also shown retarding abilities, cally, this addition strongly increases the surface available for further
such as carboxylic acids, more specifically citric acid and tartaric acid, hydrate precipitation, which leads to an earlier onset and a faster ac-
or phosphonates that induce more extensive retardation [96]. Concrete celeration period. Based on this principle, a new generation of accel-
dispersants may also show a retarding action on cement hydration. This erators based on C-S-H seeds suspensions has been recently introduced
is the case for low range water reducing agents, such as LSs, but also for on the market [116].
high range water reducers, such as PCEs, that have been described in The list of most common products that chemically accelerate the
Section 2.1.1. hydration of cement is quite significant and is generally divided into
The way through which most of these compounds induce retarda- two different types: soluble inorganic salts and organic compounds
tion is still a subject of debate. Different mechanisms have been pro- [96,112]. Among the inorganic salts, calcium chloride (CaCl2) is the
posed, including complexation of ions in pore solution [101–104], in- most efficient. However, despite having been widely used, its exact
hibition of dissolution of anhydrous phases [29,91,105], or inhibition mechanism of acceleration is not fully understood, although the addi-
of nucleation and/or growth of hydrates [97,106]. For most of the tion of calcium ions must clearly affect nucleation and growth of hy-
admixtures described above, their impact on cement hydration is not drates [117]. One mechanism described in the literature is the forma-
trivial and important aspects are still not well understood. In general, in tion of a more permeable C-S-H with a higher specific surface area,
the case of sugar derivatives, the retarding ability strongly differs de- which enhances the diffusion of ions. Furthermore, higher rate of hy-
pending on their reducing or non-reducing nature, degradation degree dration observed in the induction period would also mean that the
in alkaline environment, complexing capacity, or adsorption affinity on calcium chloride enhances C-S-H nucleation [96]. However, the nega-
silicate or aluminate surfaces [28,99,100,107–109]. For example, su- tive point of using CaCl2 is the addition of chloride ions in the system,
crose, which is one of the most used and effective retarders, does not responsible for pitting corrosion of steel that compromises the long-
suffer from chemical modification through ring opening or degradation term durability of reinforced concrete [118]. This is why, depending on
at high pH, as observed in the case of glucose. Therefore, it has a lower the country and its standard, CaCl2 use is limited or absolutely pro-
calcium binding capacity than glucose, which shows a lower retarding hibited for reinforced concrete [112]. Other salts, such as bromides,
ability. This fact proves that ionic complexation alone cannot be at the carbonates, nitrates, are also common accelerators for concrete with
origin of the retardation and that one or both of the other mechanisms lower performance than CaCl2. Most commercial accelerators also
must be accounted for [99,100,110]. It also means that for any re- contain sodium/calcium nitrates or nitrites [113]. Calcium nitrate, (Ca
tarding admixture, the adsorption of molecules on the anhydrous (NO3)2), for example, increases the concentration of calcium ions
phases, hydrates or both plays an important role. In the case of comb- leading to a faster supersaturation with respect to the silicate hydrates,
shaped superplasticizers, recent results indeed showed that PCEs can and also interacts with the aluminates reaction by decreasing the sulfate
strongly inhibit dissolution of C3S, most likely through adsorption on content, reducing ettringite formation and accelerating the sulfate de-
active dissolution sites [29]. Moreover, a quantified correlation, re- pletion. This latter effect would generally be a negative side effect but,
presented by the master curve in Fig. 5 (left), was found between the in the case of 3D printing, could represent an advantage, as presented in
induced retardation of an OPC and molecular structure of PCEs Section 4.2. However, from a durability point of view, although calcium
[29,111]. More specifically, it was found that the retardation of the nitrate is known to be an effective inhibitor against chloride induced
silicate hydration depends on the dosage of the polymer and on the corrosion, there is a possibility that it might enhance corrosion through
local spatial density of functional groups (i.e. the carboxylates that have carbonation [119, 194].
the anchoring function in the adsorbed state) in the vicinity of the At this point, it has to be noted that accelerators targeting mainly
surface. In other words, the correlation is the product between the total silicate hydration might show some limitations as it would require very
number of repeat units (Fig. 2) introduced in the system and the high quantities to achieve sufficiently fast strength gain for additive
blocking capacity of each repeat unit, which depends on the electric manufacturing, which could interfere with rheological or durability
field induced by their charges on the surface, as illustrated in Fig. 5 properties. Therefore, a category of accelerators that presents inter-
(right). Combining this result with an empirical analysis of the depen- esting features for printable concrete is shotcrete accelerators
dence of the rheological properties on the molecular structure allowed, [112,120]. For sprayed concrete, they are usually added in the nozzle of
for the first time, the development of a molecular optimization tool the equipment during spraying, allowing a very quick setting as soon as
conciliating the main dispersive property of the PCEs with their most the concrete is sprayed and applied on the wall surface. Silicate and
important side effects on cement hydration [29,111]. aluminum salts are among the most commonly used types of shotcrete
5
D. Marchon et al. Cement and Concrete Research xxx (xxxx) xxx–xxx
Fig. 5. Master curve of retardation of the maximum rate of hydration, Δt, as a function of PCE structure and dosage obtained for an ordinary Portland cement (left)
and schematic representation of sub-molecular scale characteristics of an adsorbed PCE molecule (right, reproduced from [29]). The retardation is proportional to the
number of repeat units cPCE , where cPCE is the dosage in mass of PCE and MRU is the mass of one repeat unit, and to a function of the polymer structure given by
MRU
3/2
( C/E
C/E+1 ) , where C/E represents the number of carboxylate charges per repeat unit (Fig. 2). This function is proportional to the linear charge density of the
C / E + 1 1/2
backbone, C/E
C/E+1
, and inversely proportional to an average distance of the backbone to the surface of cement, ( C/E ) [29].
accelerators [120]. Silicates were first used but they have a lower facilitate on-site applications. Apis Cor performed 3D concrete printing
performance than aluminum-based products. The latter are divided into within a temperature-controlled tent due to cold temperatures – similar
two classes: alkali and alkali-free products. They are both composed of approaches can be taken to control water loss [126]. Finally, in utilizing
salts of aluminum but the alkali-free accelerator is simply an acidic admixtures and alternate binders to control rheology and hydration
solution of these salts, which has the advantage of not being hazardous. aspects for additive manufacturing, it will be essential to ultimately
With their high concentration in aluminum and sulfate, alkali-free ac- consider their effect on water transport properties [127–129].
celerators promote the precipitation of ettringite, causing a fast floc-
culation of the system and allowing the placement of thick layers of
3. Properties evolution for printable concrete
shotcrete on tunnel walls [121]. Therefore, this very basic principle is
of great interest for additive manufacturing, in cases where very fast
The aim of this section is to describe the main stages of the rheo-
setting is needed. These admixtures, as well as potassium carbonate, are
logical state of concretes used for digital fabrication. The focus is
also known to promote the flash-setting of C3A at moderate or high
mainly related to extrusion printing, which represents the most in-
dosages, something that might present interesting features for extruded
vestigated and implemented digital fabrication technique [4]. How-
concrete (see Section 4.2).
ever, as stated by others [4,8,10], the fundamental questions about the
control of the material properties are shared by other approaches, e.g.
2.3. Curing stage slip forming [4]. Fig. 6 schematically represents the expected evolution
of the yield stress of a printable concrete, dividing it into four stages,
In digitally fabricated concretes, transport properties of the freshly while Table 1 summarizes the admixtures properties needed for the
placed concrete must be carefully controlled since no fixed formwork different stages. The principle described below is mostly focusing on
can prevent water loss. Indeed, early water loss can limit the hydration batch by batch material preparation with an activation of hydration.
degree, in particular at low w/c ratio. This can affect adhesion between Continuous mixing is considered as well, but it is more challenging to
deposited layers [122] and induce plastic shrinkage cracking to ad- ensure a constant quality in material delivery. On this subject, a more
versely affect final performance. The specific aspect of water loss for detailed discussion is given by Reiter et al. [8].
additive manufacturing has not yet been investigated, as there is a clear A generic and simplified depiction of the desired performance of a
change with respect to conventional concrete applications where water cement-based system for extrusion printing is presented in Fig. 6, where
loss at very early ages does not occur due to the presence of formwork. the evolution of fluidity, expressed as yield stress (Pa), is divided into
However, it is expected that there will be parallels to the casting of four main stages. It is necessary to clarify that dynamic yield stress is
horizontal slabs or flooring systems since they also exhibit relatively the stress necessary to terminate flow under shear, while static yield
high exposed surface areas. Thus, important knowledge may be gained stress is the stress necessary to initiate flow from rest and thixotropy
from experience in such applications. For example, additive manu- accounts for the difference between the two. For extrusion printing,
facturing may benefit from using saturated lightweight aggregates for static yield stress is expected to be critical for load bearing capacity and
internal curing or chemical compounds, such as expansive agents and shape stability [130] and thus it is depicted in Fig. 6. While the trend
shrinkage-reducing admixtures [123–125]. shown here might appear intuitive, it conceals the effect of ongoing
Still, challenges and solutions are expected to be unique for additive hydration reactions on the rheological properties, that only recently
manufacturing. For instance, curing compounds can be utilized but have been quantitatively proven in the presence of PCEs [83,111].
their impact on the adhesion properties between printed layers must be Depending on the mix constituents, conditions, and other parameters,
considered for extrusion-based techniques. Water loss due to increased some of which will be discussed, the evolution of fluidity will change.
wind speed and/or temperature during printing must be addressed to Concretes with low initial yield stress values are typically needed for
6
D. Marchon et al. Cement and Concrete Research xxx (xxxx) xxx–xxx
Table 1
Summary of the requirements of the main stages described in Fig. 6 and potential admixtures to reach the properties.
Stage Concrete properties Admixtures
7
D. Marchon et al. Cement and Concrete Research xxx (xxxx) xxx–xxx
Fig. 7. a) Ternary diagram of CaO-Al2O3-SiO2 representing the typical composition of Portland cement, CAC, CSA and most common supplementary cementitious
materials, such as slag, fly ash, silica fume, natural pozzolan and clays. b) Ternary diagram of calcium aluminate cement, ordinary Portland cement and calcium
sulfate representing the main zones of composition of calcium aluminate cement used in practice (in dark) and the zone of composition where uncontrolled expansion
is expected.
admixed cement pastes that have been remixed with an equivalent structures affect not only the initial fluidity, but also the timeframe
rotational speed before spread flow measurements [83]. This procedure available for the pumping and extrusion processes (stage 0), and in
mimics the usual operative conditions on job sites before placement, turn, the extent to which covered cement particles interact. Being very
since concretes are subject to high shear rates when pumped [17]. The fundamental, this mechanism likely represents a basic principle that
almost constant yield stress measured in this stage evolves in- may be used to rationalize, in a broader sense, the yield stress gain of
dependently of hydration reactions and corresponds to the induction concretes used for additive manufacturing.
period observed in isothermal calorimetry experiments [83]. When
using copolymerized PCEs, the duration of this stage was found to in-
crease linearly with the number of added functional groups [83], with a 4. Strategies for the control of hydration and rheology and
slightly more complex relation for grafted PCEs [29,111]. Besides the possible limitations
retarding effect induced by superplasticizers, the duration of the open
time may be tuned with the combination of retarders and accelerators, 4.1. Potential alternative candidates for the binding matrix
as proposed in the Smart Dynamic Concrete production process [142].
The sudden increase of the yield stress shown in stage 3 (Fig. 6) The previous sections describe the properties expected from a con-
represents a crucial point for the construction of 3D printing structures. crete used in some of the new technologies of digital fabrication and
After this point, it is known that the material is prone to supporting potential admixtures that can be used to target them. Another aspect to
more weight due to rapid strength gain but, at the same time, the consider is the choice of the binding material to easily reach the ex-
casting should be sufficiently fast, or at least precisely controlled, to pected properties. As an example, for the transition from stages 2 to 3
prevent the formation of surface defects, if not deep cracks. For this (Fig. 6), OPC can show limitations in terms of rate of surface devel-
reason, it becomes necessary to further investigate the properties of the opment for fast structuration. The seemingly slow structuration build-
concrete system in this timeframe. Various authors have reported an up of OPC that would lead to a too long stage 2 can be balanced by
exponential increase in yield stress over time after the period of con- complementary rapid setting cements [148], such as calcium aluminate
stant fluidity, as indicated in stage 3 of Fig. 6 [143–147]. In this con- cement (CAC) [149,150] and calcium sulfo-aluminate cement (CSA)
text, a valuable new approach uses the relation between yield stress and [151–153]. Both CSA and CAC are commonly employed in situations
heat rate, as shown by isothermal calorimetry to determine an onset where fast placement, shrinkage compensation and resistance to che-
point. Beyond this point, the yield stress increases exponentially with mical attacks are needed [154,155]. Their property is driven by the
the heat rate and in an irreversible way. This discriminates the two inherently rich amounts of calcium and aluminum, as depicted in
distinct processes governing stages 2 and 3, namely, reversible and ir- Fig. 7a. Due to the high solubility of the components, swift dissolution
reversible structural build-up [83]. of the Ca and Al ions occurs, with the subsequent formation of calcium
While the initial fluidity can be varied in a simple way according to aluminate hydrates and ettringite (due to the presence of sulfates in
the required performances, the subsequent evolution over time is con- CSA). This causes a rapid surface development and exponential struc-
sidered to be a greater challenge for practical applications. Indeed, turation that is translated to both quick set/stiffening and rapid gain in
when the yield stress starts to increase, the control of the system be- early strength, typically observed in CAC and/or CSA systems
comes problematic. Mantellato was, however, able to rationalize much [150–152]. The hydration of these binders is greatly dependent on the
of this behaviour in relation to the formation of new surfaces by on- moisture, temperature, duration of reaction, and presence of nucleating
going hydration and the ability of the admixture type and dosage to agents [156–158]. When blended with OPC, the CSA and/or CAC ac-
modify it [83]. More specifically, she found that it is the combination of celerate the setting time progressively with increasing OPC replacement
additional contact points and reduction in surface coverage that ap- level and can considerably vary it, depending on the composition of the
pears to account for the rapid increase in yield stress observed in stage 3 OPC, as well as amount and type of sulphates. Furthermore, both al-
(Fig. 6). Although pumpable concretes are often necessary for digitally ternative binders are compatible with most chemical admixtures, in-
designed structures, the depiction proposed appears valid for a broader cluding superplasticizers, retarders and accelerators [159]. One aspect
range of initial yield stress values and therefore for different admixtures that cannot be neglected when dealing with blends of OPC with CSA,
contents, such as superplasticizers. Indeed, different PCE dosages and however, is the possibility of undesired expansion. This depends on the
ratio between the cements and components used, and more particularly
8
D. Marchon et al. Cement and Concrete Research xxx (xxxx) xxx–xxx
on the available amount of calcium sulfate in the system [155,160]. The example, it decreases the duration of stage 2 in Fig. 6 and allows earlier
dashed areas in Fig. 7b shows the composition of the calcium aluminate placement of the next layer faster in 3D extrusion. As mentioned earlier,
cements used in practice, whereas the dashed line represents the limit shotcrete accelerators are known to promote the flash-setting of the C3A
above which an appropriate binder composition might induce an un- phase at moderate or high dosages. In the application of extruded
controlled expansion. concrete, however, their actions may be too rapid unless their dosage is
The use of supplementary cementitious materials (SCMs) is quite strongly reduced, their addition is located very close to the outlet and/
common in concrete. When partially replacing cement, they can help to or they are used in combination with a retarder. In regard to late
achieve the rheological and packing properties needed for stage 1 and 2 strength, it should be noted that alkali free shotcrete accelerators are
in Fig. 6, as mentioned earlier. However, their slower reaction rate may considered to have little to no negative effect at normal dosages, while
show some limitations for the development of binders used in additive their alkali counterparts do. This effect is strongly dosage dependent, so
manufacturing. Geopolymers intrinsically require some form of che- in utilizing shotcrete accelerators for 3D extrusion, specific effects will
mical activation, which in a certain sense make them good candidates have to be examined since the dosages would clearly be lower than for
for the setting on demand needed in additive manufacturing. However, shotcrete applications.
the level of chemical activation needed might be quite high. Therefore, Furthermore, similar action in shortening stage 2 by a rapid stif-
it may not be trivial to efficiently deliver the needed activator without fening could be obtained by accelerating the hydration of a system with
excessively diluting the system and subsequently compromising CAC cement right at the exit of the nozzle. Finally, although all these
strength. On the other hand, initial and partial activation of a system applications represent interesting features and advantages for printing
that contains a limited amount of SCMs could be considered. However, concretes, stability of the cementitious systems with respect to hydra-
as discussed in more details in Section 4.3, PCEs suffer from competitive tion and mechanical performances are still very crucial and need more
adsorption in the presence of some activators and therefore cannot be investigation.
used optimally.
Another possible area of interest for geopolymers is in powder bed 4.3. Competitive adsorption and the related important physical aspects
printing [4,161,162]. In this approach, the activator would be delivered
by the spray head. An advantage over OPC would be that the powder is One of the most important aspects coming from this analysis is that
less sensitive to humidity and likely easier to recycle. However, there a system needed for digital fabrication applications tends to be very
are issues that must be addressed. The activators typically used are complex with different kinds of binding materials, as well as several
NaOH, KOH, Na2SiO3, K2SiO3, Na2CO3 and Na2SO4 [163]. It is worth sources of admixtures added and acting at different times. From such
recalling that geopolymers activated with silicate solutions have a high systems, undesired side effects, such as strong interactions among dif-
viscosity, which has been mainly attributed to the activator solution ferent admixtures, are also expected.
[164]. So, for powder bed printing, the high viscosity of the activator The mechanism of action of most of the chemical admixtures used in
solution may be difficult to dispense. In addition, for extrusion-based the production of concrete is very sensitive to their adsorption behavior
printing, the high viscosity of the activated geopolymer should be ac- on the surface of cement particles. The blending of these chemical ad-
counted for, which cannot be resolved by using other admixtures, such mixtures can generate a competitive adsorption on surface sites of
as superplasticizers. particles, which will influence their performances. As shown in the
literature, competitive adsorption can be an advantage leading to sy-
4.2. Al-Si-S chemical balance nergetic effects between adsorption species [171] but, on the other
hand, can in some cases reduce the desired performances [171–174].
One of the most promising solutions to improve the structuration Many examples of competitive adsorption between a superplasticizer
build-up right after deposition is to activate or enhance the aluminate and sulfates ions [175,176], hydroxides [177,178], citrates [172] and
hydration reactions. This can be done by adding shotcrete-type accel- other organic admixtures [172–174] have been studied in the literature
erators or replacing part of the cement with a calcium (sulfo) aluminate of cementitious materials. By adsorbing on the surface of cement par-
cement, as seen in the previous sections. However, playing with the ticles, these displacers restrict the adsorption of polycarboxylate-type
aluminate hydration by either promoting ettringite nucleation or superplasticizers and therefore affect their dispersion efficiency from a
mixing a calcium aluminate cement with a normal Portland cement can macroscopic point of view, as shown in Fig. 8 in presence of sodium
present some risks as this might strongly influence the sulfate con- gluconate. It was also shown in [171] that superplasticizers can, in
sumption and silicate hydration [165]. As introduced in Section 2.2 and some cases, limit the adsorption of a cellulose ether and inhibit its
demonstrated already decades ago [93,94], the sulfatation level of a flocculating behavior.
cement is one of the most important aspects that controls the reaction To control the outcome of competitive adsorption, many works
balance between the silicate, aluminate and sulfate phases. If the studied the possibility of modifying the chemical structure of the mo-
amount of the sulfate phase in a cement is too low or its consumption is lecules [167,179–181]. Furthermore, it should be kept in mind that
too fast, the sulfate depletion point occurs before the main hydration of competitive adsorption is not only a matter of the admixture chemical
the silicate and a flash set occurs. This results in a very rapid stiffening structure or surface affinity, but is also very dependent on the avail-
and subsequent loss of fluidity, which could be detrimental if it occurs ability of sites for potential adsorption on the surface of the binder. In
before extrusion. Furthermore, the normal hydration of the aluminate the literature dealing with polymers at interfaces, three main adsorp-
phase reduces the silicate hydration [89,96,111,166]. Some ad- tion regimes have been identified where competitive adsorption is more
mixtures, such as PCEs or retarders, are also well known to favor et- or less pronounced. At low surface coverage, competitive adsorption is
tringite nucleation in the first few minutes of hydration and effectively not observed and all organic admixtures can adsorb on the surface
increase its specific surface area [29,111,167–169]. This leads to a [182]. In contrast, getting closer to surface saturation (high surface
faster consumption of sulfates with potential important consequences coverage), organic admixtures start to compete for adsorption sites.
with respect to mechanical strength if the sulfate availability is not well Under the latter condition, the competitor with the highest affinity
controlled and its depletion occurs too early [170]. In particular, the would preferentially adsorb, whereas the other species may have only
silicate hydration becomes strongly delayed, the early mechanical access to residual surface sites. Based on this principle, adsorption re-
strength decreases drastically, and this can also ultimately affect late gimes, governing the competition among species, might be varied by
mechanical performances. With this in mind, and for applications tuning the polymer dosage, adsorbed conformation or adsorption en-
where high strength gain is not required, being able to control a pro- ergy. It should, moreover, be kept in mind that some adsorbed species
moted “flash set” right after deposition would be interesting since, for could leave an abundance of unoccupied adsorption sites in the area
9
D. Marchon et al. Cement and Concrete Research xxx (xxxx) xxx–xxx
Fig. 8. Desorption of a superplasticizer (PCE) in the presence of a retarder (sodium gluconate) (left) and the rheological consequences (right). Adapted from [171].
they cover, which may be occupied by the other admixture without any from solution possible and are therefore considered as anionic ex-
desorption. In addition, we would expect that the continuous formation changers with properties very similar to those of clays [184].
of adsorption sites over time due to ongoing chemical reactions may Anionic polymers such as PCEs have been shown to intercalate into
significantly affect the extent of these adsorption regimes. From this these LDHs to form layered organo-mineral phases (OMPs) [186–188].
point of view, the use of blends such as OPC/CAC/CSA can also provide The degree of intercalation is generally proportional to the charge
a rapid growth in specific surface area owing to the inherently high density of the LDHs and inversely correlated to the molecular weight of
reactivity of these systems and the high affinity of aluminate hydrate the incoming anionic polymers [189–191]. Commercial PNS are also
surfaces for PCEs. Consequently, the distribution of admixtures may be highly susceptible to intercalation in LDH, whereas LSs are usually in-
changed between silicate surfaces and aluminates, the balance of sensitive due to their steric sizes. Under normal paste conditions, it is
competitive adsorption would be shifted and the overall dispersion of expected that the polymer-LDH phases will be exfoliated and dispersed.
the system reduces, leading to yield stress growth. In any case, the presence of these layers provides for a high consump-
Furthermore, it must be kept in mind that adsorption and desorption tion potential of anionic organic compounds, modifying their impact on
kinetics could be at the origin of the evolution of material reactivity and rheology and hydration kinetics.
rheological behavior over time, which could be interesting for concrete On the other hand, clays are hydrous aluminum silicates usually
digital fabrication. In some cases, the exchange could be rapid and defined as “phyllosilicate minerals”. Most clays possess a mica-type
reversible, as with the desorption of a superplasticizer by sulfate [47] or structure, while clay flakes consist of stacked platelets where a single
sodium gluconate [46] necessary for the rapid structural build up in platelet is a unit layer [192]. As mentioned earlier, the immediate
stage 1 (Fig. 6), for example. However, the desorption of an already impact of clays is a strong reduction in workability due to an important
adsorbed polymer could also, in some cases, be a very slow process in retention of water by its high surface area, and an additional major
the order of hours, days or even years, as it requires the simultaneous impact on yield stress owing to their flocculation. Furthermore, clays,
desorption of the anchor groups of the macromolecule [183]. particularly smectites, attract and consume PCE superplasticizers. The
dominating driving force is governed by chemisorption of the PEO side
4.4. Depletion of chemical admixtures through intercalation chains in the PCE structure into the clay layered structure via hydrogen
bonds, along with water molecules anchored by the silanol groups be-
Aluminate phases, and more particularly layered structures, such as tween the aluminosilicate layers (Fig. 10) [193]. This contrasts with the
layered double hydroxides (LDHs), and clays may involve another type mobilization of PCEs by LDH phases, which is driven by backbone
of disturbing interaction with some admixtures. LDHs are a class of adsorption and for which the side chains should be expected to provide
layered materials with a chemical composition expressed by the general steric hindrance. In the case of clays, the primary consumption rate is
formula [M2+ 3+
1-x Mx (OH)2]
x+
[An- x-
x/n · yH2O] , where M
2+
and M3+ are proportional to the length of the side chain. Surface adsorption of the
divalent and trivalent metal cations and An- is a counter anion or an- PCE onto the clay surfaces through electrostatic attraction accounts
ionic polymer possessing a charge of n- [77]. When hydrating, C3A can only for a minor portion of PCEs consumed by clays.
form different layered phases: aluminate hydrates, C4AH13 and C4AH19, The sequestration of organics, such as PCEs, by aluminate hydrates
and AFm phases, such as Ca4Al2(OH)12SO4 ⋅ 6H2O and and clays depletes the dispersants from the pore solution, thereby re-
3CaO ⋅ Al2O3 ⋅ CaCO3 ⋅ 11H2O in the presence of sulfates and carbo- ducing their effectiveness. When this is not mitigated by some means,
nates, respectively. Likewise, similar layered structures such as C2AH8 i.e. the addition of sulfates that stabilize calcium aluminate sulfates or
(Fig. 9) are formed during CAC and CSA hydration. These unique the addition of sacrificial compounds as PEG polymers to be adsorbed
layered structures make specific chemisorption of anionic molecules on the clays, this phenomenon can become a major drawback in more
10
D. Marchon et al. Cement and Concrete Research xxx (xxxx) xxx–xxx
Fig. 10. Conceptual illustration of PCE side chain intercalation between the aluminosilicate layers of montmorillonite. Reproduced from [193] with permission.
11
D. Marchon et al. Cement and Concrete Research xxx (xxxx) xxx–xxx
phase based on interparticle interactions, Cem. Concr. Res. 104 (2018) 94–104, N. Roussel, Cellulose ethers and water retention, Cem. Concr. Res. 53 (2013)
http://dx.doi.org/10.1016/j.cemconres.2017.11.004. 176–184, http://dx.doi.org/10.1016/j.cemconres.2013.06.010.
[22] G. Gelardi, R.J. Flatt, 11 - working mechanisms of water reducers and super- [49] H. van Olphen, P.H. Hsu, An introduction to clay colloid chemistry, Soil Sci. 126
plasticizers, in: P.-C. Aïtcin, R.J. Flatt (Eds.), Sci. Technol. Concr. Admix., (1978) 59.
Woodhead Publishing, 2016, pp. 257–278 http://www.sciencedirect.com/ [50] H.H. Murray, Traditional and new applications for kaolin, smectite, and paly-
science/article/pii/B9780081006931000114. gorskite: a general overview, Appl. Clay Sci. 17 (2000) 207–221, http://dx.doi.
[23] K.H. Khayat, N. Mikanovic, 8 - Viscosity-Enhancing Admixtures and the Rheology org/10.1016/S0169-1317(00)00016-8.
of Concrete, in: N. Roussel (Ed.), Underst. Rheol. Concr., Woodhead Publishing, [51] R.D. Ferron, S. Shah, E. Fuente, C. Negro, Aggregation and breakage kinetics of
2012, pp. 209–228, , http://dx.doi.org/10.1533/9780857095282.2.209. fresh cement paste, Cem. Concr. Res. 50 (2013) 1–10, http://dx.doi.org/10.1016/
[24] M. Palacios, R.J. Flatt, 20 - Working mechanism of viscosity-modifying ad- j.cemconres.2013.03.002.
mixtures, in: P.-C. Aïtcin, R.J. Flatt (Eds.), Sci. Technol. Concr. Admix, Woodhead [52] Z. Quanji, G.R. Lomboy, K. Wang, Influence of nano-sized highly purified mag-
Publishing, 2016, pp. 415–432 http://www.sciencedirect.com/science/article/ nesium alumino silicate clay on thixotropic behavior of fresh cement pastes,
pii/B9780081006931000205. Constr. Build. Mater. 69 (2014) 295–300, http://dx.doi.org/10.1016/j.
[25] K.G. Kuder, S.P. Shah, Rheology of extruded cement-based materials, ACI Mater. J. conbuildmat.2014.07.050.
104 (2007) 283–290. [53] S. Kawashima, M. Chaouche, D.J. Corr, S.P. Shah, Rate of thixotropic rebuilding of
[26] J.H. Kim, M. Beacraft, S.P. Shah, Effect of mineral admixtures on formwork cement pastes modified with highly purified attapulgite clays, Cem. Concr. Res. 53
pressure of self-consolidating concrete, Cem. Concr. Compos. 32 (2010) 665–671, (2013) 112–118, http://dx.doi.org/10.1016/j.cemconres.2013.05.019.
http://dx.doi.org/10.1016/j.cemconcomp.2010.07.018. [54] Y. Qian, S. Kawashima, Use of creep recovery protocol to measure static yield
[27] Voigt Thomas, Mbele Jean-Juste, Wang Kejin, P. Shah Surendra, Using fly ash, stress and structural rebuilding of fresh cement pastes, Cem. Concr. Res. 90 (2016)
clay, and fibers for simultaneous improvement of concrete green strength and 73–79, http://dx.doi.org/10.1016/j.cemconres.2016.09.005.
consolidatability for slip-form pavement, J. Mater. Civ. Eng. 22 (2010) 196–206, [55] T. Conte, M. Chaouche, Rheological behavior of cement pastes under large am-
http://dx.doi.org/10.1061/(ASCE)0899-1561(2010)22:2(196). plitude oscillatory shear, Cem. Concr. Res. 89 (2016) 332–344, http://dx.doi.org/
[28] D. Marchon, R.J. Flatt, 12 - Impact of Chemical Admixtures on Cement Hydration, 10.1016/j.cemconres.2016.07.014.
in: P.-C. Aïtcin, R.J. Flatt (Eds.), Sci. Technol. Concr. Admix, Woodhead [56] A. Kaci, M. Chaouche, P.-A. Andréani, Influence of bentonite clay on the rheolo-
Publishing, 2016, pp. 279–304 http://www.sciencedirect.com/science/article/ gical behaviour of fresh mortars, Cem. Concr. Res. 41 (2011) 373–379, http://dx.
pii/B9780081006931000126. doi.org/10.1016/j.cemconres.2011.01.002.
[29] D. Marchon, P. Juilland, E. Gallucci, L. Frunz, R.J. Flatt, Molecular and sub- [57] R. Siddique, J. Klaus, Influence of metakaolin on the properties of mortar and
molecular scale effects of comb-copolymers on tri-calcium silicate reactivity: to- concrete: a review, Appl. Clay Sci. 43 (2009) 392–400, http://dx.doi.org/10.
ward molecular design, J. Am. Ceram. Soc. 100 (2017) 817–841, http://dx.doi. 1016/j.clay.2008.11.007.
org/10.1111/jace.14695. [58] F. Uddin, Clays, nanoclays, and montmorillonite minerals, Metall. Mater. Trans. A.
[30] G. Gelardi, S. Mantellato, D. Marchon, M. Palacios, A.B. Eberhardt, R.J. Flatt, 9 - 39 (2008) 2804–2814, http://dx.doi.org/10.1007/s11661-008-9603-5.
chemistry of chemical admixtures, in: P.-C. Aïtcin, R.J. Flatt (Eds.), Sci. Technol. [59] F. Sanchez, K. Sobolev, Nanotechnology in concrete – a review, Constr. Build.
Concr. Admix, Woodhead Publishing, 2016, pp. 149–218 http://www. Mater. 24 (2010) 2060–2071, http://dx.doi.org/10.1016/j.conbuildmat.2010.03.
sciencedirect.com/science/article/pii/B9780081006931000096. 014.
[31] R. Flatt, I. Schober, 7 - superplasticizers and the rheology of concrete, in: [60] NSTC, The National Nanotechnology Initiative – Strategic Plan, Executive office of
N. Roussel (Ed.), Underst. Rheol. Concr, Woodhead Publishing, 2012, pp. 144–208 the President of the United States, 2007.
http://www.sciencedirect.com/science/article/pii/B9780857090287500078. [61] W.L. Haden Jr., Attapulgite: properties and uses, Clays Clay Min. 10 (1963)
[32] N. Spiratos, M. Pagé, N.P. Mailvaganam, V.M. Malhotra, C. Jolicoeur, 284–290.
Superplasticizers for concrete: fundamentals, technology, and practice, (2003). [62] L.S. Sonon, M.L. Thompson, Sorption of a nonionic Polyoxyethylene lauryl ether
[33] P.-C. Aı ̈tcin, Cements of yesterday and today: Concrete of tomorrow, Cem. Concr. surfactant by 2:1 layer silicates, Clays Clay Miner. 53 (2005) 45–54, http://dx.doi.
Res. 30 (2000) 1349–1359, http://dx.doi.org/10.1016/S0008-8846(00)00365-3. org/10.1346/CCMN.2005.0530106.
[34] N. Roussel, A. Lemaître, R.J. Flatt, P. Coussot, Steady state flow of cement sus- [63] M.J. Sánchez-Martín, M.C. Dorado, C. del Hoyo, M.S. Rodríguez-Cruz, Influence of
pensions: a micromechanical state of the art, Cem. Concr. Res. 40 (2010) 77–84, clay mineral structure and surfactant nature on the adsorption capacity of sur-
http://dx.doi.org/10.1016/j.cemconres.2009.08.026. factants by clays, J. Hazard. Mater. 150 (2008) 115–123, http://dx.doi.org/10.
[35] H. Bessaies-Bey, Polymères et propriétés rhéologiques d'une pâte de ciment: une 1016/j.jhazmat.2007.04.093.
approche physique générique, Paris Est, http://www.theses.fr/2015PEST1042, [64] E. Sabah, M. Turan, M.S. Çelik, Adsorption mechanism of cationic surfactants onto
(2015) , Accessed date: 6 October 2017. acid- and heat-activated sepiolites, Water Res. 36 (2002) 3957–3964, http://dx.
[36] J.E.W. Scripture, Indurating composition for concrete, Google Patents, 1937. doi.org/10.1016/S0043-1354(02)00110-0.
https://www.google.com/patents/US2081643. [65] J.B. Croft, The problem in predicting the suitability of soils for cementitious sta-
[37] M. Gruber, 3D printing/concrete extrusion – a technology made possible with bilization, Eng. Geol. 2 (1968) 397–424, http://dx.doi.org/10.1016/0013-
green Lignosulfonates as admixture, Munich, Germany, 2016. 7952(68)90018-5.
[38] J.A. Lewis, H. Matsuyama, G. Kirby, S. Morissette, J.F. Young, Polyelectrolyte [66] S. Joshi, S. Goyal, M.S. Reddy, Influence of nutrient components of media on
effects on the rheological properties of concentrated cement suspensions, J. Am. structural properties of concrete during biocementation, Constr. Build. Mater. 158
Ceram. Soc. 83 (2000) 1905–1913, http://dx.doi.org/10.1111/j.1151-2916.2000. (2018) 601–613, http://dx.doi.org/10.1016/j.conbuildmat.2017.10.055.
tb01489.x. [67] K.A. Saeed, K.A. Kassim, H. Nur, N.Z.M. Yunus, Strength of lime-cement stabilized
[39] R.J. Flatt, Y.F. Houst, A simplified view on chemical effects perturbing the action tropical lateritic clay contaminated by heavy metals, KSCE J. Civ. Eng. 19 (2015)
of superplasticizers, Cem. Concr. Res. 31 (2001) 1169–1176, http://dx.doi.org/10. 887–892, http://dx.doi.org/10.1007/s12205-013-0086-6.
1016/S0008-8846(01)00534-8. [68] D. Plee, F. Lebedenko, F. Obrecht, M. Letellier, H. Van Damme, Microstructure,
[40] D. Marchon, S. Mantellato, A.B. Eberhardt, R.J. Flatt, 10 - adsorption of chemical permeability and rheology of bentonite — cement slurries, Cem. Concr. Res. 20
admixtures, in: P.-C. Aïtcin, R.J. Flatt (Eds.), Sci. Technol. Concr. Admix, (1990) 45–61, http://dx.doi.org/10.1016/0008-8846(90)90115-E.
Woodhead Publishing, 2016, pp. 219–256 http://www.sciencedirect.com/ [69] N.A. Tregger, M.E. Pakula, S.P. Shah, Influence of clays on the rheology of cement
science/article/pii/B9780081006931000102. pastes, Cem. Concr. Res. 40 (2010) 384–391, http://dx.doi.org/10.1016/j.
[41] J. Pourchez, B. Ruot, J. Debayle, E. Pourchez, P. Grosseau, Some aspects of cel- cemconres.2009.11.001.
lulose ethers influence on water transport and porous structure of cement-based [70] A. Kazemian, X. Yuan, E. Cochran, B. Khoshnevis, Cementitious materials for
materials, Cem. Concr. Res. 40 (2010) 242–252, http://dx.doi.org/10.1016/j. construction-scale 3D printing: laboratory testing of fresh printing mixture,
cemconres.2009.09.028. Constr. Build. Mater. 145 (2017) 639–647, http://dx.doi.org/10.1016/j.
[42] H. Bessaies-Bey, R. Baumann, M. Schmitz, M. Radler, N. Roussel, Effect of poly- conbuildmat.2017.04.015.
acrylamide on rheology of fresh cement pastes, Cem. Concr. Res. 76 (2015) [71] B. Chen, J. Liu, Contribution of hybrid fibers on the properties of the high-strength
98–106, http://dx.doi.org/10.1016/j.cemconres.2015.05.012. lightweight concrete having good workability, Cem. Concr. Res. 35 (2005)
[43] C. Brumaud, R. Baumann, M. Schmitz, M. Radler, N. Roussel, Cellulose ethers and 913–917, http://dx.doi.org/10.1016/j.cemconres.2004.07.035.
yield stress of cement pastes, Cem. Concr. Res. 55 (2014) 14–21, http://dx.doi. [72] P.-C. Aïtcin, 6 - Entrained air in concrete: rheology and freezing resistance, Sci.
org/10.1016/j.cemconres.2013.06.013. Technol. Concr. Admix, Woodhead Publishing, 2016, pp. 87–95, , http://dx.doi.
[44] L. Patural, J.-P. Korb, A. Govin, P. Grosseau, B. Ruot, O. Devès, Nuclear magnetic org/10.1016/B978-0-08-100693-1.00006-0.
relaxation dispersion investigations of water retention mechanism by cellulose [73] R. Gagné, 17 - air entraining agents, Sci. Technol. Concr. Admix, Woodhead
ethers in mortars, Cem. Concr. Res. 42 (2012) 1371–1378, http://dx.doi.org/10. Publishing, 2016, pp. 379–391, , http://dx.doi.org/10.1016/B978-0-08-100693-1.
1016/j.cemconres.2012.06.002. 00017-5.
[45] L. Patural, P. Marchal, A. Govin, P. Grosseau, B. Ruot, O. Devès, Cellulose ethers [74] L. Ducloué, O. Pitois, J. Goyon, X. Chateau, G. Ovarlez, Rheological behaviour of
influence on water retention and consistency in cement-based mortars, Cem. suspensions of bubbles in yield stress fluids, J. Non-Newtonian Fluid Mech. 215
Concr. Res. 41 (2011) 46–55, http://dx.doi.org/10.1016/j.cemconres.2010.09. (2015) 31–39, http://dx.doi.org/10.1016/j.jnnfm.2014.10.003.
004. [75] P.-C. Nkinamubanzi, S. Mantellato, R.J. Flatt, 16 - superplasticizers in practice, in:
[46] K.M.A. Hossain, M. Lachemi, Performance of volcanic ash and pumice based P.-C. Aïtcin, R.J. Flatt (Eds.), Sci. Technol. Concr. Admix, Woodhead Publishing,
blended cement concrete in mixed sulfate environment, Cem. Concr. Res. 36 2016, pp. 353–377 http://www.sciencedirect.com/science/article/pii/
(2006) 1123–1133, http://dx.doi.org/10.1016/j.cemconres.2006.03.010. B9780081006931000163.
[47] T.F. Tadros, Dispersion of Powders in Liquids and Stabilization of Suspensions, [76] S. Mantellato, A.B. Eberhardt, R.J. Flatt, 15 - formulation of commercial products,
John Wiley & Sons, 2012. in: P.-C. Aïtcin, R.J. Flatt (Eds.), Sci. Technol. Concr. Admix, Woodhead
[48] C. Brumaud, H. Bessaies-Bey, C. Mohler, R. Baumann, M. Schmitz, M. Radler, Publishing, 2016, pp. 343–349 http://www.sciencedirect.com/science/article/
12
D. Marchon et al. Cement and Concrete Research xxx (xxxx) xxx–xxx
pii/B9780081006931000151. [105] L. Nicoleau, M.A. Bertolim, Analytical model for the Alite (C3S) dissolution to-
[77] H.F.W. Taylor, Cement Chemistry, Thomas Telford, 1997. pography, J. Am. Ceram. Soc. 99 (2016) 773–786, http://dx.doi.org/10.1111/
[78] D. Marchon, R.J. Flatt, 8 - mechanisms of cement hydration, in: P.-C. Aïtcin, jace.13647.
R.J. Flatt (Eds.), Sci. Technol. Concr. Admix, Woodhead Publishing, 2016, pp. [106] M.C. Garci Juenger, H.M. Jennings, New insights into the effects of sugar on the
129–145 http://www.sciencedirect.com/science/article/pii/ hydration and microstructure of cement pastes, Cem. Concr. Res. 32 (2002)
B9780081006931000084. 393–399, http://dx.doi.org/10.1016/S0008-8846(01)00689-5.
[79] K.L. Scrivener, A. Nonat, Hydration of cementitious materials, present and future, [107] J.M. de Bruijn, A.P.G. Kieboom, H. van Bekkum, Alkaline degradation of mono-
Cem. Concr. Res. 41 (2011) 651–665, http://dx.doi.org/10.1016/j.cemconres. saccharides III. Influence of reaction parameters upon the final product compo-
2011.03.026. sition, Recl. Trav. Chim. Pays-Bas 105 (1986) 176–183, http://dx.doi.org/10.
[80] K.L. Scrivener, P. Juilland, P.J.M. Monteiro, Advances in understanding hydration 1002/recl.19861050603.
of Portland cement, Cem. Concr. Res. 78 (2015) 38–56 (Part A), https://doi.org/ [108] B.J. Smith, A. Rawal, G.P. Funkhouser, L.R. Roberts, V. Gupta, J.N. Israelachvili,
10.1016/j.cemconres.2015.05.025. B.F. Chmelka, Origins of saccharide-dependent hydration at aluminate, silicate,
[81] G.H. Tattersall, P.F. Banfill, The Rheology of Fresh Concrete, Pitman, London, and aluminosilicate surfaces, Proc. Natl. Acad. Sci. 108 (2011) 8949–8954, http://
1983. dx.doi.org/10.1073/pnas.1104526108.
[82] N. Roussel, G. Ovarlez, S. Garrault, C. Brumaud, The origins of thixotropy of fresh [109] B.J. Smith, L.R. Roberts, G.P. Funkhouser, V. Gupta, B.F. Chmelka, Reactions and
cement pastes, Cem. Concr. Res. 42 (2012) 148–157, http://dx.doi.org/10.1016/j. surface interactions of saccharides in cement slurries, Langmuir 28 (2012)
cemconres.2011.09.004. 14202–14217, http://dx.doi.org/10.1021/la3015157.
[83] S. Mantellato, Flow Loss in Superplasticized Cement Pastes, Doctoral dissertation [110] P. Suraneni, R.J. Flatt, Micro-reactors to study Alite hydration, J. Am. Ceram. Soc.
ETH Zurich, 2017. 98 (2015) 1634–1641, http://dx.doi.org/10.1111/jace.13472.
[84] W.A. Corstanje, H.N. Stein, J.M. Stevels, Hydration reactions in pastes C3S [111] D. Marchon, Controlling Cement Hydration though the Molecular Structure of
+C3A+CaSO4.2aq+H2O at 25°C.I, Cem. Concr. Res. 3 (1973) 791–806, http:// Comb Copolymer Superplasticizers, Doctoral dissertation ETH, 2016.
dx.doi.org/10.1016/0008-8846(73)90012-4. [112] P.-C. Aïtcin, 19 - Accelerators, Sci. Technol. Concr. Admix, Woodhead Publishing,
[85] E. Breval, C3A hydration, Cem. Concr. Res. 6 (1976) 129–137, http://dx.doi.org/ 2016, pp. 405–413, , http://dx.doi.org/10.1016/B978-0-08-100693-1.00019-9.
10.1016/0008-8846(76)90057-0. [113] R. Myrdal, Accelerating Admixtures for Concrete, State of the Art, SINTEF,
[86] R.J. Myers, G. Geng, J. Li, E.D. Rodríguez, J. Ha, P. Kidkhunthod, G. Sposito, Norway, 2007.
L.N. Lammers, A.P. Kirchheim, P.J.M. Monteiro, Role of adsorption phenomena in [114] L. Reiter, M. Palacios, T. Wangler, R.J. Flatt, Putting concrete to sleep and waking
cubic tricalcium aluminate dissolution, Langmuir 33 (2017) 45–55, http://dx.doi. it up with chemical admixtures, Spec. Publ. 302 (2015) 145–154.
org/10.1021/acs.langmuir.6b03474. [115] J.J. Thomas, H.M. Jennings, J.J. Chen, Influence of nucleation seeding on the
[87] J. Skalny, M.E. Tadros, Retardation of Tricalcium aluminate hydration by sulfates, hydration mechanisms of tricalcium silicate and cement, J. Phys. Chem. C 113
J. Am. Ceram. Soc. 60 (1977) 174–175, http://dx.doi.org/10.1111/j.1151-2916. (2009) 4327–4334, http://dx.doi.org/10.1021/jp809811w.
1977.tb15503.x. [116] L. Nicoleau, The Acceleration of Cement Hydration by Seeding: Influence of the
[88] P. Meredith, A.M. Donald, N. Meller, C. Hall, Tricalcium aluminate hydration: Cement Mineralogy, (2012).
microstructural observations by in-situ electron microscopy, J. Mater. Sci. 39 [117] L. Nicoleau, E. Schreiner, A. Nonat, Ion-specific effects influencing the dissolution
(2004) 997–1005, http://dx.doi.org/10.1023/B:JMSC.0000012933.74548.36. of tricalcium silicate, Cem. Concr. Res. 59 (2014) 118–138, http://dx.doi.org/10.
[89] A. Quennoz, K.L. Scrivener, Interactions between alite and C3A-gypsum hydra- 1016/j.cemconres.2014.02.006.
tions in model cements, Cem. Concr. Res. 44 (2013) 46–54, http://dx.doi.org/10. [118] U. Angst, B. Elsener, C.K. Larsen, Ø. Vennesland, Critical chloride content in re-
1016/j.cemconres.2012.10.018. inforced concrete — a review, Cem. Concr. Res. 39 (2009) 1122–1138, http://dx.
[90] F. Begarin, S. Garrault, A. Nonat, L. Nicoleau, Hydration of alite containing alu- doi.org/10.1016/j.cemconres.2009.08.006.
minium, Adv. Appl. Ceram. 110 (2011) 127–130, http://dx.doi.org/10.1179/ [119] I. Pepenar, An overview of nitrates-induced corrosion of reinforced concrete
1743676110Y.0000000007. structures: case studies, laboratory investigations and corrosion mechanisms, Civil
[91] P. Suraneni, R.J. Flatt, Use of micro-reactors to obtain new insights into the factors Eviron. Res. 5 (2013).
influencing tricalcium silicate dissolution, Cem. Concr. Res. 78 (2015) 208–215 [120] L.R. Prudêncio, Accelerating admixtures for shotcrete, Cem. Concr. Compos. 20
(Part B), https://doi.org/10.1016/j.cemconres.2015.07.011. (1998) 213–219, http://dx.doi.org/10.1016/S0958-9465(98)80007-3.
[92] E. Pustovgar, R.K. Mishra, M. Palacios, J.-B. d'Espinose, T. de Lacaillerie, [121] C. Paglia, F. Wombacher, H. Böhni, The influence of alkali-free and alkaline
A.S. Matschei, H. Andreev, R. Heinz, R.J. Flatt Verel, Influence of aluminates on shotcrete accelerators within cement systems: I. Characterization of the setting
the hydration kinetics of tricalcium silicate, Cem. Concr. Res. 100 (2017) behavior, Cem. Concr. Res. 31 (2001) 913–918, http://dx.doi.org/10.1016/
245–262, http://dx.doi.org/10.1016/j.cemconres.2017.06.006. S0008-8846(01)00509-9.
[93] N. Tenoutasse, The Hydratation Mechanism of C3A and C3S in the Presence of [122] E. Keita, H. Bessaies-Bey, W. Zuo, P. Belin, N. Roussel, Weak bond strength be-
Calcium Chloride and Calcium Sulphate, CRIC, 1969. tween successive layers in extrusion-based additive manufacturing: measurement
[94] W. Lerch, The Influence of Gypsum on the Hydration and Properties of Portland and physical origin, Cem. Concr. Res.. submitted (n.d.).
Cement Pastes, Portland Cement Association, 1946, http://www.asocem.org.pe/ [123] R. Gagné, 23 - shrinkage-reducing admixtures, Sci. Technol. Concr. Admix,
bivi/sa/dit/icem/rx012.pdf , Accessed date: 29 December 2015. Woodhead Publishing, 2016, pp. 457–469, , http://dx.doi.org/10.1016/B978-0-
[95] J.F. Young, A review of the mechanisms of set-retardation in Portland cement 08-100693-1.00023-0.
pastes containing organic admixtures, Cem. Concr. Res. 2 (1972) 415–433, http:// [124] R. Gagné, 22 - Expansive agents, Sci. Technol. Concr. Admix, Woodhead
dx.doi.org/10.1016/0008-8846(72)90057-9. Publishing, 2016, pp. 441–456, , http://dx.doi.org/10.1016/B978-0-08-100693-1.
[96] J. Cheung, A. Jeknavorian, L. Roberts, D. Silva, Impact of admixtures on the hy- 00022-9.
dration kinetics of Portland cement, Cem. Concr. Res. 41 (2011) 1289–1309, [125] A.B. Eberhardt, R.J. Flatt, 13 - working mechanisms of shrinkage-reducing ad-
http://dx.doi.org/10.1016/j.cemconres.2011.03.005. mixtures, in: P.-C. Aïtcin, R.J. Flatt (Eds.), Sci. Technol. Concr. Admix, Woodhead
[97] N.L. Thomas, J.D. Birchall, The retarding action of sugars on cement hydration, Publishing, 2016, pp. 305–320 http://www.sciencedirect.com/science/article/
Cem. Concr. Res. 13 (1983) 830–842, http://dx.doi.org/10.1016/0008-8846(83) pii/B9780081006931000138.
90084-4. [126] The first on-site house has been printed in Russia. http://apis-cor.com/en/about/
[98] L. Zhang, L.J.J. Catalan, R.J. Balec, A.C. Larsen, H.H. Esmaeili, S.D. Kinrade, news/first-house.
Effects of saccharide set retarders on the hydration of ordinary Portland cement [127] M.D. Cohen, J. Olek, W.L. Dolch, Mechanism of plastic shrinkage cracking in
and pure Tricalcium silicate, J. Am. Ceram. Soc. 93 (2010) 279–287, http://dx. Portland cement and Portland cement-silica fume paste and mortar, Cem. Concr.
doi.org/10.1111/j.1551-2916.2009.03378.x. Res. 20 (1990) 103–119, http://dx.doi.org/10.1016/0008-8846(90)90121-D.
[99] C. Nalet, A. Nonat, Effects of functionality and stereochemistry of small organic [128] M. Wyrzykowski, P. Trtik, B. Münch, J. Weiss, P. Vontobel, P. Lura, Plastic
molecules on the hydration of tricalcium silicate, Cem. Concr. Res. 87 (2016) shrinkage of mortars with shrinkage reducing admixture and lightweight ag-
97–104, http://dx.doi.org/10.1016/j.cemconres.2016.06.002. gregates studied by neutron tomography, Cem. Concr. Res. 73 (2015) 238–245,
[100] C. Nalet, A. Nonat, Ionic complexation and adsorption of small organic molecules http://dx.doi.org/10.1016/j.cemconres.2015.03.013.
on calcium silicate hydrate: relation with their retarding effect on the hydration of [129] A. Leemann, P. Nygaard, P. Lura, Impact of admixtures on the plastic shrinkage
C3S, Cem. Concr.Res. 89 (2016) 97–108, http://dx.doi.org/10.1016/j.cemconres. cracking of self-compacting concrete, Cem. Concr. Compos. 46 (2014) 1–7, http://
2016.08.012. dx.doi.org/10.1016/j.cemconcomp.2013.11.002.
[101] R. Heeb, S. Lee, N.V. Venkataraman, N.D. Spencer, Influence of salt on the aqueous [130] A. Perrot, D. Rangeard, A. Pierre, Structural built-up of cement-based materials
lubrication properties of end-grafted, ethylene glycol-based self-assembled used for 3D-printing extrusion techniques, Mater. Struct. 49 (2015) 1213–1220,
monolayers, ACS Appl. Mater. Interfaces 1 (2009) 1105–1112, http://dx.doi.org/ http://dx.doi.org/10.1617/s11527-015-0571-0.
10.1021/am900062h. [131] M. Choi, N. Roussel, Y. Kim, J. Kim, Lubrication layer properties during concrete
[102] K. Tasaki, Poly(oxyethylene)–cation interactions in aqueous solution: a molecular pumping, Cem. Concr. Res. 45 (2013) 69–78, http://dx.doi.org/10.1016/j.
dynamics study, Comput. Theor. Polym. Sci. 9 (1999) 271–284, http://dx.doi.org/ cemconres.2012.11.001.
10.1016/S1089-3156(99)00015-X. [132] D. Feys, K.H. Khayat, R. Khatib, How do concrete rheology, tribology, flow rate
[103] Y. Masuda, T. Nakanishi, Ion-specific swelling behavior of poly(ethylene oxide) gel and pipe radius influence pumping pressure? Cem. Concr. Compos. 66 (2016)
and the correlation to the intrinsic viscosity of the polymer in salt solutions, 38–46, http://dx.doi.org/10.1016/j.cemconcomp.2015.11.002.
Colloid Polym. Sci. 280 (2002) 547–553, http://dx.doi.org/10.1007/s00396-002- [133] S. Ng, H. Justnes, Influence of plasticizers on the rheology and early heat of hy-
0651-x. dration of blended cements with high content of fly ash, Cem. Concr. Compos. 65
[104] J. Plank, B. Sachsenhauser, Experimental determination of the effective anionic (2016) 41–54, http://dx.doi.org/10.1016/j.cemconcomp.2015.10.005.
charge density of polycarboxylate superplasticizers in cement pore solution, Cem. [134] S. Ng, H. Justnes, Influence of dispersing agents on the rheology and early heat of
Concr. Res. 39 (2009) 1–5, http://dx.doi.org/10.1016/j.cemconres.2008.09.001. hydration of blended cements with high loading of calcined marl, Cem. Concr.
13
D. Marchon et al. Cement and Concrete Research xxx (xxxx) xxx–xxx
14
D. Marchon et al. Cement and Concrete Research xxx (xxxx) xxx–xxx
aluminum hydrates, Inorg. Chim. Acta 359 (2006) 4901–4908, http://dx.doi.org/ [192] F. Bergaya, G. Lagaly, Chapter 1 general introduction: clays, clay minerals, and
10.1016/j.ica.2006.08.038. clay science, in: F. Bergaya, B.K.G. Theng, G. Lagaly (Eds.), Dev. Clay Sci, Elsevier,
[190] J. Plank, D. Zhimin, H. Keller, F.v. Hössle, W. Seidl, Fundamental mechanisms for 2006, pp. 1–18, , http://dx.doi.org/10.1016/S1572-4352(05)01001-9.
polycarboxylate intercalation into C3A hydrate phases and the role of sulfate [193] S. Ng, J. Plank, Interaction mechanisms between Na montmorillonite clay and
present in cement, Cem. Concr. Res. 40 (2010) 45–57, http://dx.doi.org/10.1016/ MPEG-based polycarboxylate superplasticizers, Cem. Concr. Res. 42 (2012)
j.cemconres.2009.08.013. 847–854, http://dx.doi.org/10.1016/j.cemconres.2012.03.005.
[191] S. Ng, E. Metwalli, P. Müller-Buschbaum, J. Plank, Occurrence of intercalation of [194] M. Stefanoni, U. Angst, B. Elsener, Corrosion challenges and opportunities in di-
PCE superplasticizers in calcium aluminate cement under actual application con- gital fabrication of reinforced concrete, 1st RILEM International Conference on
ditions, as evidenced by SAXS analysis, Cem. Concr. Res. 54 (2013) 191–198, Concrete and Digital Fabrication, 2018 Zurich, Switzerland.
http://dx.doi.org/10.1016/j.cemconres.2013.09.002.
15