Katholieke Universiteit Leuven Faculteit Ingenieurswetenschappen
Departement Burgerlijke Bouwkunde Norwegian University of Science
and Technology Faculty of Natural Sciences and Technology
Department of Materials Science and Engineering
Combining Plasticizers/Retarders And Accelerators
E2006 Promotor: prof. dr. H. Justnes prof. dr. ir. D. Van
Gemert
Klaartje De Weerdt Dirk Reynders
Katholieke Universiteit Leuven Faculteit Ingenieurswetenschappen
Academiejaar: 2005-2006 Departement: Burgerlijke Bouwkunde Adres en
telefoon: Kasteelpark Arenberg 40 3001 Heverlee 016/32 16 54 Naam
en voornaam studenten: De Weerdt Klaartje Reynders Dirk Titel
eindwerk: Combineren van plastificeerders/vertragers en versnellers
Korte inhoud eindwerk: De combinatie van
plastificeerders/vertragers en versnellers werd bestudeerd met drie
mogelijke toepassingen in het achterhoofd: 1) het tegengaan van het
vertragend effect van plastificeerders zonder de reologie sterk te
wijzigen, 2) de activatie van vertraagd beton op de werf na veilig
transport in warme streken of steden met onvoorspelbaar verkeer en
3) het oververtragen van overschotten aan vers beton gevolgd door
activatie na n of meerdere dagen. De experimenten werden
grotendeels uitgevoerd op cementpasta. Een Paar-Physica MCR 300
rheometer werd gebruikt ter bepaling van de reologie en een TAM Air
isotherme calorimeter ter bepaling van de hydratiecurves. Er werd
vastgesteld voor toepassing 1) dat calciumnitraat het vertragend
effect van natrium en calcium lignosulfonaat sterk terugschroeft en
in het geval van polyacrylaat zelfs volledig wegneemt terwijl de
combinaties werken als plastificeerders, voor toepassing 2) dat de
combinatie natriumgluconaat/calciumnitraat een mogelijk werkend
systeem is en voor toepassing 3) dat de combinatie
citroenzuur/calciumnitraat het hergebruik van overschotten aan vers
beton op een later tijdstip mogelijk maakt.
Promotor: prof. dr. ir. D. Van Gemert prof. dr. H. Justnes
Assessoren: prof. dr. ir. L. Vandewalle ir. G. Heirman
Katholieke Universiteit Leuven Faculteit Ingenieurswetenschappen
Year: 2005-2006 Department: Burgerlijke Bouwkunde Address en tel.:
Kasteelpark Arenberg 40 3001 Heverlee 016/32 16 54 Name and surname
students: De Weerdt Klaartje Reynders Dirk Title of thesis:
Combining plasticizers/retarders and accelerators Summary of
thesis: The combination of plasticizers/retarders with accelerators
has been studied in view of three potential concrete applications:
1) counteracting retardation of plasticizers without negatively
affecting rheology too much, 2) activating retarded concrete at
site after safe transport in hot climate or cities with
unpredictable traffic and 3) over-retarding residual fresh concrete
one day and activating it next day or after several days. The
experimental work is largely carried out on cement paste using a
Paar-Physica MCR 300 rheometer to determine flow curves and gel
strength and a TAM Air isothermal calorimeter for determination of
heat of hydration curves. It has been found for application 1) that
calcium nitrate strongly reduces retardation of sodium and calcium
lignosulphonates and even cancels retardation of polyacrylates,
whereas the blend also has plasticizing effects, for 2) that sodium
gluconate/calcium nitrate is a potentially effective system and for
3) that citric acid/calcium nitrate may facilitate later use of
residual fresh concrete.
Promotor: prof. dr. ir. D. Van Gemert prof. dr. H. Justnes
Assessors: prof. dr. ir. L. Vandewalle ir. G. Heirman
Table of Contents1 2 Introduction Background on cement, cement
hydration, rheology and admixtures 2.1 2.2 2.3 2.4 2.5 3 Cement
..................................................................................................................
Cement hydration
..................................................................................................
Rheology
...............................................................................................................
Plasticizers/retarders
.............................................................................................
Calcium nitrate
......................................................................................................
1 4 4 5 9 13 22 24
Materials and apparatus
3.1
Materials................................................................................................................
24 3.2 Apparatus
..............................................................................................................
27 4 Counteracting plasticizer retardation 4.1 4.2 4.3 4.4 5
Introduction
...........................................................................................................
Calorimetric and rheological
measurements.........................................................
Mortar
measurements............................................................................................
General
conclusion................................................................................................
34 34 35 75 80 81 81 81 95 98 101 110 111 111 111 114 116 120 126
127
Long transport of fresh concrete 5.1 5.2 5.3 5.4 5.5 5.6
Introduction
...........................................................................................................
Sodium
lignosulphonate........................................................................................
Citric acid
..............................................................................................................
Lead
nitrate............................................................................................................
Sodium
gluconate..................................................................................................
General
conclusion................................................................................................
6
Reutilizing residual fresh concrete 6.1 6.2 6.3 6.4 6.5 6.6
Introduction
...........................................................................................................
Phase I Screening of
retarders............................................................................
Phase II Determination of required retarder dosage
.......................................... Phase III Activation
using calcium nitrate
......................................................... Phase IV
Strength
measurements.......................................................................
General
conclusion................................................................................................
7
Conclusions
Chapter 1 IntroductionThis thesis continues a long tradition of
Erasmus exchanges between the Katholieke Universiteit Leuven
(Belgium) and the Norges Teknisk-Naturvitenskapelige Universitet i
Trondheim (Norway). For many years students have been studying
advanced aspects of cementitious materials. Thys, A. and Vanparijs,
F. ([1]) studied the longterm performance of concrete with calcium
nitrate, Ardoullie, B. and Hendrix, E. ([2]) focused on the
chemical shrinkage of cementitious pastes and mortars, Clemmens, F.
and Depuydt, P. ([3]) investigated early hydration of Portland
cements, the thesis of Van Dooren, M. ([4]) concerned the factors
influencing the workability of fresh concrete, and Brouwers, K.
([5]) studied a number of cold weather accelerators. In this thesis
the combination of plasticizers/retarders and accelerators has been
investigated in view of three different potential concrete
applications. The first application, which made up the major part
of this study, focused on the fact that plasticizers that are used
to increase flow for cementitious materials at equal
water-to-cement ratio also to a variable extent retard setting as a
side effect. The objective was to find an accelerator that at least
partially would counteract this retardation without negatively
affecting the rheology too much. Whereas earlier studies on this
topic focused on plastic viscosity at high shear rate (i.e.
relevant for mixing) and relatively low dosages of plasticizer, the
study reported here focused on the lower shear rate range (i.e.
relevant for pouring concrete) and higher dosages of plasticizer.
The results of this study are presented in Chapter 4. These results
are valuable elements in evaluating the combined use of
plasticizers and accelerators, as it was e.g. applied during
construction of Statoils Troll platform (Figure 1.1), a huge gas
platform located 80 km north-west of Bergen (Norway) that reaches
303 m below the surface of the sea. During the construction of its
350 m tall base an accelerator has been used to speed
1
Chapter 1: Introduction
2
up the slip forming process of the plasticized concrete as
construction works were behind schedule. The second application
concerns long transport of fresh concrete. The preliminary study
was largely carried out on paste. It was investigated if a concrete
mix from a ready mix plant after being deliberately over-retarded
for long transport in for instance hot climate or cities with
unpredictable traffic (e.g. traffic jam) could be activated by
adding an accelerator in the revolving drum close to the
construction site before pumping the concrete in place. Results are
discussed in Chapter 5. The third potential application, presented
in Chapter 6, concerns the search for a system to preserve residual
fresh concrete for a few days (e.g. over a weekend) followed by
activation before use. However, it might also be used as an
overnight concept. Whereas recently a freezing preservation
technique has been proposed as method for reutilizing left-over
concrete, this study concentrated on a technique consisting of
over-retardation of residual fresh concrete followed by later
activation using an accelerator.
Figure 1.1 Troll gas platform (1996)
Chapter 1: Introduction
3
The necessary background on cement, cement hydration, rheology
and admixtures is given in Chapter 2. Chapter 3 introduces and
describes the materials and the apparatus that have been used
throughout this work.
Chapter 2 Background on cement, cement hydration, rheology and
admixtures2.1 CementCement chemists use in general a short hand
notation, C = CaO, S = SiO2, A =Al2O3, F = Fe2O3 and S = SO3, for
the main elements in the chemical analyses of cement, inaddition to
H = H2O to describe hydration processes. The elements are
determined by X-ray fluorescence or analytical chemistry and given
as the corresponding oxides. Assuming that the only minerals in the
cement are alite (C3S), belite (C2S), aluminate phase (C3A),
ferrite phase (C4AF) and anhydrite ( C S ) the content of these
minerals may be calculated through mass balances. The first four
minerals are formed during equilibrium conditions in the burning of
the cement clinker, while the latter mineral (or gypsum, C S H 2 )
is added to the mill when clinker is ground to cement. In
specification sheets, the content of other oxides is also given: N
(Na2O), K (K2O) and M (MgO). Free lime is the content of free CaO
due to insufficient burning or due to the decomposition of C3S into
C2S and free lime if the cooling rate is too low. The specific
surface area (m2/kg) of cement is commonly determined directly by
an air permeability method called the Blaine method. In addition to
the specific area, the particle size is of importance for the
hydration rate of cement, since the hydration takes place at the
interface between the cement grain and the water phase. However, it
is important to realise that the surface of a cement grain is
inhomogeneous. The distribution of C3S/C2S- and C3A/C4AF-domains
are determined by the milling process and the difference in
resistance against fracture. Since cement grains are composite
grains with possibly all 4 major phases in one grain, efforts to
simulate
4
Chapter 2: Background
5
cement by adding corresponding amounts of individual minerals
will therefore fail. (Justnes, H., [6], p.10)
2.2. Cement hydrationIn the discussion of rheology of cement
paste and the interaction with plasticizing admixtures and
retarders, it is of importance to know something about the
hydration until setting. It is sometimes believed that no hydration
takes place in the so-called dormant period between water addition
and initial setting, while actually a substantial growth of
hydration products takes place on the surface of the cement grains.
(Justnes, H., [6], p.10)
2.2.1 The interstitial phases C3A/C4AFIn the absence of calcium
sulphates the first hydration product of C3A which appears to grow
at the C3A surface is gel-like. Later this material transforms into
hexagonal crystals corresponding to the phases C2AH8 and C4AH19.
The formation of the hexagonal phases slows down further hydration
of C3A as they function as a hydration barrier. Finally the
hexagonal phases convert to the thermodynamically stable cubic
phase C3AH6 disrupting the diffusion barrier, after which the
hydration proceeds with a fairly high speed. The overall hydration
process may thus be written as 2 C 3 A + 27 H C 2 AH 8 + C 4 AH19 2
C 3 AH 6 + 15 H (hexagonal phases) (cubic phase)
In the presence of calcium sulphate (as in a Portland cement)
the amount of hydration of C3A in the initial state of hydration is
distinctly reduced when compared to that consumed in the absence of
C S . Needle-shaped crystals of ettringite are formed as the main
hydration product:
C3 A + 3 CSH 2 + 26 H C6 AS3 H32 Minor amounts of the
monosulphate C 4 A S H 12 or even C 4 AH19 may also be formed if an
imbalance exists between the reactivity of C3A and the dissolution
rate of2 calcium sulphate, resulting in an insufficient supply of
SO 4- - ions.
Then ettringite formation is accompanied by a significant
liberation of heat. After a rapid initial reaction, the hydration
rate is slowed down significantly. The length of this dormant
period may vary and increases with increasing amounts of calcium
sulphate in the original paste.
Chapter 2: Background
6
A faster hydration, associated with a second heat release
maximum, gets under way after all the available amount of calcium
sulphate has been consumed. Under these conditions the ettringite,
formed initially, reacts with additional amounts of tricalcium
aluminate, resulting in the formation of calcium aluminate
monosulphate hydrate (monosulphate): C 6 A S3 H 32 + 2 C 3 A + 4 H
3 C 4 A S H 12 As ettringite is gradually consumed, hexagonal
calcium aluminate hydrate ( C 4 AH19 ) also starts to form. It may
be present in the form of a solid solution with C 4 A S H 12 or as
separate crystals. The origin of the dormant period, characterised
by a distinctly reduced hydration rate, is not obvious and several
theories have been forwarded to explain it. The theory most widely
accepted assumes the build-up of a layer of ettringite at the
surface of C3A that acts as a barrier responsible for slowing down
the hydration. Ettringite is formed in a through-solution reaction
and precipitates at the surface of C3A due to its limited
solubility in the presence of sulphates. The validity of this
theory has been questioned arguing that the deposited ettringite
crystals are not dense enough to account for the retardation of
hydration. The four proceeding alternative theories have been
proposed: i) The impervious layer consists of water-deficient
hexagonal hydrate2 stabilised by incorporation of SO 4- . It is
formed on the surface of C3A and
ii)
becomes covered by ettringite. C3A dissolves incongruently in
the liquid phase, leaving an aluminate rich layer on the surface.
Ca2+ - ions are adsorbed on it, thus reducing the number of active
dissolution sites and thereby the rate of C3A dissolution. A
subsequent adsorption of sulphate ions results in a further
reduction of the dissolution rate.
iii)
2 SO 4- - ions are adsorbed on the surface of C3A forming a
barrier. Contrary
to this theory it has been found that C3A is not slowed down if
the calcium iv) sulphate is replaced by sodium sulphate. Formation
of an amorphous layer at the C3A surface that acts as an osmotic
membrane and slows down the hydration of C3A.
The termination of the dormant period appears to be due to a
breakdown of the protective layer, as the added calcium sulphate
becomes consumed and ettringite is converted to monosulphate. In
this through-solution reaction both C3A and ettringite dissolve and
monosulphate is precipitated from the liquid phase in the
matrix.
Chapter 2: Background
7
The composition of the calcium aluminoferrite phase (ferrite
phase), usually written as C4AF, may vary between about C4A1.4F0.6
and C4A0.6F1.4. Under comparable conditions the hydration products
formed in the hydration of the ferrite phase are in many aspects
similar to those formed by the hydration of C3A although the rates
differ and the aluminium in the products is partially substituted
by ferric ions. The reactivity of the ferrite may vary over a wide
range, but seems to increase with increasing A/F ratio.
2.2.2 The main mineral alite C3SThe hydration of alite can be
divided into 4 periods: a) Pre-induction period: Immediately after
contact with water, an intense, but short-lived hydration of C3S
gets under way. An intense liberation of heat may be observed in
this stage of hydration. The duration of this period is typically
no more than a few minutes. b) Induction (dormant) period: The
pre-induction period is followed by a period in which the rate of
reaction slows down significantly. At the same time the liberation
of heat is significantly reduced. This period lasts typically a few
hours. c) Acceleration (post-induction) period: After several hours
the rate of hydration accelerates suddenly and reaches a maximum
within about 5 to 10 hours. The beginning of the acceleration
period coincides roughly with the beginning of the second main heat
evolution peak. The Ca(OH)2 concentration in the liquid phase
attains a maximum at this time and begins to decline. Crystalline
calcium hydroxide (portlandite) starts to precipitate. The initial
set as determined by Vicat-needle is often just after the start of
this period and the final setting time just before the ending of
it. d) Deceleration period: After reaching a maximum the rate of
hydration starts to slow down gradually, however, a measurable
reaction may still persist even after months of curing. The reason
for this is that the hydration reaction becomes diffusion
controlled due to hydration products growing around the unhydrated
cement core in increasingly thickness.
Chapter 2: Background The overall alite hydration reaction may
ideally be written as 2 C 3S + 7 H C 3S 2 H 4 + 3 CH
8
The calcium hydroxide, CH, is crystalline, while the calcium
silicate hydrate is amorphous with a variable composition and
therefore often simply denoted CSH-gel.
2.2.3 Hydration and setting of ordinary Portland cementThe
overall hydration of ordinary Portland cement is basically a
combination of the description of the interstitial phase with
gypsum and alite as discussed in the preceding sections. Which of
the two dominates the setting is still a matter of discussion and
probably depends on the cement composition The hydration of
Portland cement can be associated with the liberation of hydration
heat. Figure 2.1 shows the heat evolution curve for a typical
Portland cement.
Rate of Heat Evolution
Dissolution Ettringite and CSH gel Formation
Formation of Monosulfate
Rapid Formation of CSH and CH
Induction Period Increase in Ca2+ and OH- Concentration
DiffusionControlled Reactions Final Set
Initial Set
Min
Hours
Days Time of Hydration
Figure 2.1 Hydration heat evolution of an ordinary Portland
cement. (Justnes, H., [6], p. 10)
In cements containing at least a fraction of the K+ in the form
of potassium sulphate, the hydration process may be marked by a
distinct initial endothermic peak immediately after mixing which is
due to the dissolution of this cement constituent in the mixing
water. A rather intense liberation of heat with a maximum within a
few
Chapter 2: Background
9
minutes is due to the initial rapid hydration of C3S and C3A.
Hydration of calcium sulphate hemihydrate to dehydrate may also
contribute to this exothermic peak. After a distinct minimum, due
to the existence of a dormant period in which the overall rate of
hydration is slowed down, a second, mean exothermic peak, with a
maximum after a few hours, becomes apparent. It is mainly due to
the hydration of C3S and the formation of the CSH phase and
portlandite. After that, the rate of heat release slows down
gradually and reaches very low values within a few days. In most
but not all cements, a shoulder or small peak may be observed at
the descending branch of the main peak, which is probably due to
renewed ettringite formation, there may even be a second shoulder
which is attributed to ettringite-monosulphate conversion.
(Hewlett, P., [7], p. 270-271)
2.3 Rheology2.3.1 General viscosityIn his Principa published in
1687, Isaac Newton formulated the following hypothesis about steady
simple shearing flow: The resistance which arises from the lack of
slipperiness of the parts of the liquid, other things being equal,
is proportional to the velocity with which the parts of the liquid
are separated from each other. This is shown in Figure 2.2.
Figure 2.2 Steady simple shearing flow. (Justnes, H., [6], p.
3)This lack of slipperiness is what we now call viscosity. It is
synonymous with internal friction and is a measure of resistance to
flow. The force per unit area required to produce the motion F/A is
denoted shear stress ( ) and is proportional to the velocity
gradient U/d (or shear rate, ). The constant of proportionality, ,
is called the shear viscosity (also called apparent viscosity):
=
Chapter 2: Background
10
The simplest rheological behaviour for liquids is the Newtonian
viscous flow and Hookes law for solid materials. Ideal viscous (or
Newtonian) flow behaviour is described using Newtons law
= Examples of ideal viscous materials are low molecular liquids
such as water, solvents, mineral oils, etc. and they are often
called Newtonian liquids. Hookes law states that the shear force
acting on a solid is proportional to the resulting deformation
= G where G is the rigidity modulus. Many materials especially
those of colloidal nature show a mechanic behaviour in between
these to border lines (Hookes an Newtons laws), i.e. they have both
plastic and elastic properties and are called viscoelastic. Samples
with a yield point only begin to flow when the external forces
acting on the material are larger than the internal structural
forces. Below the yield point, the material shows elastic
behaviour, i.e. it behaves like a rigid solid that under load
displays only a very small degree of deformation that does not
remain after removing the load. To describe the rheology of samples
showing a yield point the Bingham model is often used. The Bingham
model was extended by Herschel/Bulkley to include samples with
apparent yield point due to shear thinning or thickening:
= 0 + p pp = 1 for samples with Bingham behaviour (true yield
point) p < 1 for samples exhibiting shear thinning (apparent
yield point) p > 1 for samples with shear thickening
behaviour
Shear thinning is a reduction of viscosity with increasing shear
rate in steady flow. Samples with shear thinning behaviour can be
macromolecule solutions or melts where the individual molecules are
entangled. Under high shear load the macromolecules will stretch
out and may be disentangled, causing a reduction of the viscosity.
Furthermore, in dispersions or suspensions shearing can cause
particles to orient in the flow direction, agglomerates to
disintegrate or particles to change their
Chapter 2: Background
11
form. During this process the interaction forces between the
particles usually decrease and this also lowers the flow
resistance. Shear thickening is an increase of viscosity with
increasing shear rate. Shear thickening flow behaviour occurs in
concentrated chemically unlinked polymers due to mechanical
entanglements between the mostly branched molecule chains. The
higher the shear load the more the molecule chains prevent each
other from moving. If, during the shear process with highly
concentrated suspensions, the particles touch each other more and
more the consequences are similar: the resistance to flow
increases. Cement paste has shear thinning properties due to both
agglomerates of cement grains and growth of needle-shaped
ettringite in the fresh state. An extreme case of particles that
will change shape under shear load easily are entrained air
bubbles. There is often more air in concrete than in cement paste,
and this may make it difficult to correlate the concrete
rheological properties with those of the same paste using the
particle-matrix model. Note that concrete with 5 volume percentage
air corresponds to 15 20 volume percentage air in the matrix,
something that clearly will affect the matrix rheology.
2.3.2 Flow resistanceNumerous rheological models have been
proposed to describe cementitious materials. The Bingham model has
become very popular due to its simplicity and ability to describe
cementitious flow. The model describes the shear stress ( ) as a
function of yield stress ( 0 ), plastic viscosity ( p ) and shear
rate ( ) as = 0 + p
The concept of yield stress is sometimes a very good
approximation for practical purposes. It is however clear that the
Bingham model often only applies for limited parts of the flow
curve if the tested material has shear thinning or shear thickening
flow behaviour. The Bingham model is dependent on the shear rate
range for shear thickening materials. The shear thickening
behaviour results furthermore in negative yield stress values at
the high shear rate, which has no physical meaning (see Figure
2.3). There is a similar strong effect of the shear rate range on
the flow parameters of a shear thinning paste.
Chapter 2: Background
12
p
0
Figure 2.3 Shear thickening behaviour resulting in negative
yield stress values when using the Bingham model. The
Hershel/Buckley equation = 0 + p p can be used to fit flow curves
of pastes
showing shear thinning or shear thickening behaviour. However,
it may be difficult to compare viscosities ( p ) for different
mixes with different
p-factors. Negative yield stress values ( 0 ) with no physical
meaning can sometimes also be obtained using the Hershel/Buckley
equation. Therefore the area under the flow curve (Vikan, H. and
Justnes, H., [8]) was chosen as a measure of flow resistance
(Figure 2.4). This parameter, from here on referred to as flow
resistance, shall be used throughout to work to describe the flow
curve. The flow resistance will always be a positive value and not
depend on curve shape.
flow resistance
Figure 2.4 Flow resistance.
Chapter 2: Background
13
Furthermore, the choice between two parameters for correlation,
as for the Bingham model, can be omitted. It can be shown (Vikan,
H. and Justnes, H., [8]) that the area under the flow curve
represents something more physical than an apparent yield stress
from Bingham modeling. In a parallel plate set-up with shear area,
A [m2], and gap h [m] between the plates:
=
F A v h
[N/m2 or Pa] [m/s.m or s-1]
=
where F [N] is the force used to rotate the upper plate and v
[m/s] the velocity.
F v F v F v Area under the curve = = = = A h V A h where V [m3]
is the volume of the sample. The unit of the area under the curve
is then [N.m/m3.s or J/m3.s or W/m3]. It is in other words the
power required to make a unit volume of the paste flow with the
prescribed rate in the selected range. The power,
P [W], required to mix concrete for a certain time interval is
actually sometimes measured by simply monitoring voltage (U [V])
and current (I [A]) driving theelectrical motor of the mixer, since
P = U.I.
2.4 Plasticizers/retarders2.4.1. IntroductionWater-reducing
admixtures or plasticizers are all hydrophilic surfactants which,
when dissolved in water, deflocculate and disperse particles of
cement. By preventing the formation of conglomerates of cement
particles in suspension, less water is required to produce a paste
of a given consistency or concrete of particular workability.
Maintaining low water contents whilst achieving an acceptable level
of workability results in higher strengths for given cement content
as well as lower permeability and reduced shrinkage. An important
consequence of the reduction in the permeability is a major
enhancement of its durability. The permeability of concrete to
gases (oxygen, CO2), and water (carrying chlorides, sulfates, acids
and carbonates) is of major importance with respect to its
durability. Retarding admixtures, which extend the hydration
induction period and thereby lengthening the setting times, are
often treated together with plasticizing admixtures as the main
components used for retarding mixtures are also present in
water-reducing
Chapter 2: Background
14
admixtures. As a result, many retarders tend to reduce mixing
water and many water reducers tend to retard the setting of
concrete. A much greater reduction in the volume of mixing water
can be achieved using socalled superplasticizers or high-range
water-reducing admixtures in case of concretes of normal
workability. Normal water reducers are capable of reducing water
requirement by about 10-15%. Further reductions can be obtained at
higher dosages but this may result in undesirable effect on
setting, air content, bleeding, segregation and hardening
characteristics of concrete. Superplasticizers are capable of
reducing water contents by about 30%. (Ramachandran, V.S., [9], p.
211) Much of the following is based on Rheology of Cement based
Binders State-of-theArt by H. Justnes ([6]).
2.4.2. Common plasticizer typesThere are four generations of
plasticizers/water reducers in terms of time of discovery/use: 1.
Salts of hydrocarboxylic acids with strong retarding effects 2.
Calcium or sodium lignosulphonate (denoted CLS or NLS) as
by-products from pulping industry with medium retarding properties.
3. Synthetic compounds like naphtalene-sulphonate-formaldehyde
condensates (SNF) and sulphonated melamine-formaldehyde condensates
(SMF) with small retarding properties. 4. Synthetic polyacrylates
with grafted polyether side chains (PA) with small retarding
properties. The first generation plasticizers, the salts of organic
hydroxycarboxylic acids, are mostly used for their dominating
retarding behavior. As the name implies, the hydrocarboxylic acids
have several hydroxyl (OH) groups and either one or two terminal
carboxylic acids (COOH) groups attached to a relatively short
carbon chain. Figure 2.5 illustrates some typical hydroxycarboxylic
acids which can be used as water reducing or retarding admixtures.
Gluconic acid is perhaps the most widely used admixture. Citric,
tartaric, mucic, malic, salicylic, heptonic, saccharic and tannic
acid can also be used for the same purpose. Usually they are
synthetized chemically
Chapter 2: Background
15
Figure 2.5 Typical hydrocarboxylic acids used in water reducing
admixtures. (Ramachandran, V.S., [9], p.126)and have a very high
degree of purity as they are used as raw materials by
pharmaceutical and food industries. Some aliphatic hydrocarboxylic
acids, however, can also be produced from fermentation or oxidation
of carbohydrates and for this reason are also called sugar acids.
Hydrocarboxylic acids can be used alone as retarders or
water-reducing and retarding admixtures. For use as normal and
accelerating water reducers they must be mixed with an accelerator.
(Ramachandran, V.S., [9], p. 125) The second generation
plasticizers, the lignosulphonates, are still the most widely used
raw material in the production of water reducing admixtures.
Lignosulphonates are sulphonated macromolecules from partial
decomposition of lignin by calcium hydrogen sulphite. Under
sulphite pulping, lignin is sulphonated and rendered water soluble.
The spent sulphite liquor contains sulphonated lignin fragments of
different molecular sizes and sugar monomers after removing the
pulp. It can be further purified by fermentation to remove hexoses
and by ultrafiltration to enrich larger molecular fractions. In
addition to chemical modification of functional groups for special
applications, simple treatment by sodium sulphate will ion exchange
calcium
Chapter 2: Background
16
through formation of gypsum that is removed. A fragment of a
lignosulphonate is illustrated in Figure 2.6. Fractionation to
enrich larger molecular fractions increases the effectiveness of
lignosulphonate as a dispersant for cement in water and reduces the
retarding effect. Sodium lignosulphonates retard in general less
than calcium lignosulphonates.
Figure 2.6 Fragment of lignosulphonate. (Justnes, H., [6], p.
30)Due to the size of the molecule, it cannot be ruled out that
lignosulphonates disperse cement both through electrostatic
repulsion and steric hindrance. The average molecular weight of
common lignosulphonates used as plasticizers for cement may be
about 5,000-10,000. It is assumed that the structure of
lignosulphonates in solution consists of a mainly hydrophobic
hydrocarbon core with sulphonic groups positioned at the surface.
The bulk of the model is assumed to be made up of cross linked,
polyaromatic chains which are randomly coiled. The negatively
charged groups are positioned mainly on the surface or near the
surface of the particle, and a double layer
Chapter 2: Background
17
of counter ions is present in the solvent. The lignosulphonate
molecules behave as expanding polyelectrolytes as they expand at
low and contract at high salt concentrations. The third generation
plasticizers, the synthesized polymers with sulphonated groups, are
not covered here as they were not used in this work. The fourth
generation of plasticizers is based on a polyacrylate (PA) backbone
that is obtained by free radical polymerization of different vinyl
monomers. This backbone may vary widely in composition depending on
the choice of monomers as shown in Figure 2.7. The next step is to
graft on side chains of polyether (polyethylene oxide). Variations
in the nature and relative proportions of the different monomers in
the copolymer yield a group of products having broad ranges of
physico-chemical and functional properties. Since some of the
polyacrylates seem to enhance the segregation tendencies, they are
often combined with viscosifiers to counteract this effect.
Figure 2.7 Illustration of a generic group of polyacrylate
copolymers where R1 equals H or CH3, R2 is a poly-ether side chain
(e.g., polyethylene oxide) and X is a polar (e.g., CN) or ionic
(e.g., SO3) group. (Ramachandran, V.S. et al., [10], p.52) 2.4.3.
Mechanisms of dispersionThere are generally two main mechanisms
which explain how plasticizers disperse particles in a suspension:
electrostatic repulsion and steric hindrance. These two mechanisms
are sketched Figure 2.8 and Figure 2.9 respectively. Since its
ionic lattice is cut, any fractured mineral particle will have
domains of positive and negative charged sites. Negatively charged
polymers (common feature of most plasticizers) will absorb to the
positive charged sites and render the total particle surface
negatively charged. As negatively charged particles approach each
other there will be an electrostatic repulsion preventing them from
getting close and attach to form
Chapter 2: Background
18
Figure 2.8 Sketch of how negative charged polymers may adsorb to
both positively and negatively charged domains of particles. The
resulting overall negative charge of the particles will prevent
them to form agglomerates by electrostatic repulsion and they will
stay dispersed. The electrostatic repulsion effect increases with
increasing charge density of the adsorbed molecule. (Ramachandran,
V.S. et al, [10], p.200)
Figure 2.9 Sketch of branched macromolecules adsorbing on the
surface of grains that will create steric hindrance for them to get
close enough to form agglomerates. The size effect of steric
hindrance increases with increasing molecular weight (or actual
size) of the adsorbed molecule. (Ramachandran, V.S. et al, [10],
p.201)agglomerates. The latest generation of grafted polymers may
also have some negative charges on their backbone that can
co-ordinate on the positive sites but it should be noted that the
ester group of acrylates may co-ordinate strongly to calcium anyway
without any charge. The grafted polyether chains perpendicular to
the backbone may stretch out and hinder the particles to get close
enough to form agglomerates. This so-called steric hindrance is
based on the size of the adsorbed molecules perpendicular to the
particle surface. This is shown in Figure 2.10.
Chapter 2: Background
19
Figure 2.10 Idealized model on how a grafted polymer will lead
to steric hindrance by adsorbing the polymer backbone to the
surface and stretching the grafted side chains into the water
phase. (Justnes, H., [6], p. 26)
The model of the grafted polymer dispersing according to steric
hindrance in Figure 2.10 may be a simplification. It would then be
necessary for all the intermolecular bonds (van der Waals type
hydrogen bonds) to break and unwind the polyether chains to let
them stretch out into the water phase (even though the hydrophilic
nature of polyethers may aid in stabilizing such configuration).
Alternatively, the molecules may stay unwound as polymeric balls or
micelles that equally well will lead to steric hindrance (see
Figure 2.11). While the first three generations of plasticizers are
said to rely on electrostatic repulsion as mechanism for their
dispersion of cement agglomerates, the fourth generation is the
first to be designed to function through steric hindrance.
Macromolecular micelles
Cement surface Figure 2.11 Model of how macromolecules with
strong intramolecular forces still may disperse through steric
hindrance as polymer balls or micelles (after Justnes, H., [6], p.
26)Another effect that will prevent agglomerates formation is
called depletion as sketched in Figure 2.12. The mechanism of this
is that surplus polymer will not be adsorbed and will stay in the
water phase between the particles and for this reason prevents them
from getting close enough to form agglomerates.
Chapter 2: Background
20
cement particle s
cement particle s
polymer Figure 2.12 Surplus polymer in the water phase (not
adsorbed) may prevent the cement particles to get close enough to
form agglomerates. This depletion effect will not disperse by
itself, but rather help stabilize dispersions by preventing
flocculation. (after Justnes, H., [6], p. 27)
Rheology may also be improved by a tribology effect as sketched
in Figure 2.13. Tribology is the science of friction, abrasion and
lubrication. Low molecular weight compounds may reduce the friction
between particles and also reduce the surface tension of the water
face.
cement particle s
cement particle s
Low molecular weight compound Figure 2.13 Low molecular
compounds in the water phase may improve rheology of particle
suspensions by lubrication and by lowering the surface tension of
the water phase, which may be denoted as a tribology effect. (after
Justnes, H., [6], p. 27)Initial rheology of cement paste is also
governed by early hydration, unlike inert particles suspensions
(e.g. limestone). Thus, there are other mechanisms of how
plasticizers may improve rheology of cement pastes. One is
adsorption to active sites
Chapter 2: Background
21
and retardation of the formation of hydration products (see
Figure 2.14), another is changing the morphology of the hydration
products formed by reducing growth (see Figure 2.15) or by
intercalation in the hydration products (see Figure 2.16).
Figure 2.14 Rheology in cement pastes may improve due to less
hydration caused by adsorbed polymers co-ordinating to active sites
(). The effect increases with decreasing size of the molecules. LMW
= low molecular weight and HMW = high molecular weight.
(Ramachandran, V.S. et al, [10], p.201)
Figure 2.15 Schematic illustration of hydration nucleation and
growth inhibition by adsorbed molecules. Selective adsorption on
crystal planes can give morphology changes. (Ramachandran, V.S. et
al, [10], p.208)
Chapter 2: Background
22
Figure 2.16 Intercalation of plasticizer in hydration product
with structural alteration (e.g. lignosulphonates with hydration
products of C3A). (Ramachandran, V.S. et al, [10], p.209)
2.5 Calcium nitrateThis section is based on the paper Setting
Accelerator Calcium Nitrate,
Fundamentals, Performance and Applications by Justnes, H. and
Nygaard, E. ([11]). In the past a growing concern about the
chloride-induced corrosion of reinforcing barsembedded in Portland
cement concrete has led to the development of a number of
chloride-free set accelerating admixtures to replace the widely
used calcium chloride accelerator. In 1981, calcium nitrate,
Ca(NO3)2, was proposed as a basic component of a set accelerating
admixture. Calcium nitrate, denoted as CN, works as a pure set
accelerator (see Figure 2.17), and not as a strength development
accelerator. The pure set accelerating effect is beneficial in
preventing any increase in maximum temperature in massive
constructions due to the heat of hydration. In spite of this, an
increase in long term compressive strength is often observed,
probably due to binder morphology changes.
Hardening
Setting
Reference
Figure 2.17 Difference between set and hardening
accelerators.
Chapter 2: Background
23
The effectiveness of CN as a setting accelerator for cement is
dependent on the cement type. The set accelerating efficiency
appeared to be correlated with the belite, C2S, content, while no
correlation between set accelerating efficiency and C3A has been
found. In order to find the reason for the linear correlation
between accelerator efficiency and belite content, and possibly the
mechanism of CN as set accelerator for cement, Justnes and Nygaard
undertook a thorough analysis of the water in cement pastes from
mixing to paste setting for two different cement types (HS65 and
P30). For both cement pastes the most noticeable change when 1.55 %
CN by weight of the cement was added, was that the calcium
concentration increased and the sulphate concentration decreased.
Thus, the mechanism for accelerated setting is twofold: i) ii) an
increased calcium concentration leads to a faster super-saturation
of the fluid with respect to calcium hydroxide, Ca(OH)2, while a
lower sulphate concentration will lead to slower/less formation of
ettringite which will shorten the onset of aluminate, C3A,
hydration.
The difference between the two cements was that P30 contained
much more of the mineral aphthitalite, K3Na(SO4)2, which leads to a
high initial sulphate concentration in the fluid. When CN was
added, much of the calcium precipitated as sparingly soluble
gypsum. Even when 1.55 % CN was added to the P30 paste, the
sulphate concentration in the fluid was higher than in the water of
HS65 paste without CN. At the same time, the calcium concentration
in the fluid of P30 with CN was only slightly higher than for HS65
without CN. The Ca2+ concentration in the water of HS65 paste, on
the other hand, was increased with about 4 times when 1.55 % CN was
added. Thus, the reason why CN did not accelerate the setting of
P30 was that it contained a very soluble alkali sulphate
originating from the clinker process. The correlation between
belite content and set accelerating efficiency is understandable
since belite can incorporate a portion of the total alkalies in its
structure and consequently prevent them from taking part in the
early fluid chemistry since belite is a slow reacting mineral.
Hence, for a series of cements, with about equal total alkali
content and increasing belite content, it is expected that the set
accelerating efficiency of CN will increase. On the other hand, in
an investigation of calcium acetate, chloride and nitrate on belite
hydration, it has been found that after 1 day, the chemically bound
water was 6 times larger when 2 % CN was mixed in the water, while
2 % calcium acetate and 2 % calcium chloride only increased the 1
day chemically bound water by 30 % compared with the reference.
Therefore, a special influence of CN on -C2S can not be
excluded.
Chapter 3 Materials and apparatusThe purpose of this chapter is
to introduce and describe the materials and the apparatus that have
been used frequently throughout this work.
3.1 Materials3.1.1. Cements Two Portland cements have been used
in this thesis. Their physical characteristics are given in Table
3.1, chemical analysis according to producer and minerals by Bogue
estimation is given in Table 3.2 and the mineralogy of the cements
determined by multicomponent Rietveld analyses of XRD profiles,
specific surface determined by the Blaine method and content of
easily soluble alkalis determined by plasmaemissionspectrometry are
given in Table 3.3. Table 3.1 Physical characteristics of Portland
cements according to EN 196 Cement type Fineness: Grains + 90 m
Grains + 64 m Grains 24 m Grains 30 m Blaine (m2/kg) Water demand
Le Chatelier Initial set time c (MPa) at 1 day 2 days 7 days 28
days CEM I 52.5 R - LA 1.7% 4.1% 66.3% 75.6% 359 26.7% 0.5 mm 145
min. 17.1 27.5 42.5 58.6 CEM I 42.5 RR* 0.1% 0.5% 89.2% 94.8% 546
32.0% 0 mm 115 min. 32.7 39.9 49.3 58.9
24
Chapter 3: Materials and apparatus
25
Table 3.2 Chemical analysis (%) of the Portland cements
according to producer and minerals (%) by Bogue estimation. Cement
type Chemical analyses CaO SiO2 Al2O3 Fe2O3 SO3 MgO Free CaO K2O
Na2O Equiv. Na2O Cr6+ (ppm) Carbon Chloride LOI Fly Ash Minerals by
Bogue C3 S C2 S C3 A C4AF CS CEM I 52.5 R - LA CEM I 42.5 RR*
63.71 20.92 4.21 3.49 2.67 1.87 0.84 0.46 0.19 0.49 0.30 0.17
0.02 1.72 -
61.98 20.15 4.99 3.36 3.55 2.36 1.23 1.08 0.42 1.13 0.00 0.04
0.03 1.34 -
50.4 22.0 5.3 10.6 5.8
50.7 19.5 7.5 10.2 7.7
(* The RR term refers to the Norwegian standard NS 3086 (2003)
where RR means extra demands to 1 and 2 day strength compared to R.
42.5 RR should then have characteristic 1 day strength 20.0 MPa and
2 day strength 30.0 MPa.) It can be seen that the CEM I 42.5 RR
cement had a higher alkali and C3A content and a higher specific
surface than the CEM I 52.5 R LA cement and, as a consequence of
the latter two, had a higher water demand. CEM I 42.5 RR cement
pastes were therefore prepared with a w/c ratio of 0.50, whereas
CEM I 52.5 R LA cement pastes were prepared with a w/c ratio of
0.40 throughout this work.
Chapter 3: Materials and apparatus
26
Table 3.3 Mineral composition (%) and alkali content of Portland
cements obtained by QXRD and plasmaemissionspectrometry Cement type
Alite Belite Ferrite Cubic aluminate Orthorombic aluminate Lime
Periclase Gypsum Hemihydrate Anhydrite Calcite Portlandite Quartz
Arcanite Mullite Amporhous Blaine K (%) Na (%) Naeqv (%) CEM I 52.5
R - LA 65.0 12.9 9.6 0.5 3.0 0.6 0.3 1.4 1.5 0.4 4.0 0.3 0.4 0.0
364 0.32 0.74 0.26 CEM I 42.5 RR 64.7 14.8 7.5 5.9 1.1 1.0 1.6 0.0
1.8 0.6 0.5 0.3 0.0 0.3 546 0.92 0.22 0.76
3.1.2. Plasticizers/retarders Borregaard Lignotech, Sarpsborg,
Norway delivered two lignosulphonate powders denoted as Ultrazine
Na and Ultrazine Ca. Ultrazine Ca (CLS) was sugar reduced and large
molecular size enriched by ultra filtration of the basic calcium
lignosulphonate obtained in the sulfite process on spruce. In
Ultrazine Na (NLS) the calcium in Ultrazine Ca has been ion
exchanged with sodium. Solutions with 30% dry matter were prepared
before use. A polyether grafted polyacrylate water solution
containing 18% solids and a viscosifying agent has also been used
as a plasticizer. The molecular weight of the polyacrylate was
220,000. A number of substances were used as retarders. They were
all of analytical laboratory grade: - citric acid (C6H8O7 H2O ) -
sodium salt of gluconic acid (C6H11NaO7) - sodium salt of tartaric
acid (Na2C4H4O6 2H2O, right-turning form) - lead nitrate (Pb(NO3)2)
- zinc acetate (Zn(CH3OO)2 2H2O) - sucrose (C12H22O11)
Chapter 3: Materials and apparatus
27
The trisodiumphosphate (Na3PO4 12H2O) used in this work was from
technical quality. Household sugar was also used as a retarder.
3.1.3. Accelerator Technical calcium nitrate (CN) was used as an
accelerator. Its formula may be written as xNH4NO3 yCa(NO3)2 zH2O,
and named xyz CN according to short hand practice. The CN used in
the present work had x = 0.092, y = 0.500 and z = 0.826, or in
other words 19.00%Ca2+, 1.57% NH + , 64.68% NO3 and 14.10% H2O. The
CN was delivered in the form of 4
granules by Yara, Porsgrunn, Norway. Calcium nitrate was also
used in the form of a 50% aqueous solution of pure calcium nitrate
Ca(NO3)2, also obtained from Yara. The fluid is colourless, viscous
and can easily be blended into the mixing water.
3.2. Apparatus3.2.1. MixerThe cement pastes were blended in a
high shear mixer by Braun (MR5550CA) and by Tefal (Rondo 500) as
illustrated in Figure 3.1. The mixers had a rotational speed of
approximately 800 rpm. It will be notified which of the blenders
has been used in each chapter. The blending was performed by adding
cement to the water and mixing for minute, resting for 5 minutes
and blending again for 1 minute.
Figure 3.1 High shear blenders from Braun (left) and Tefal
(right)
Chapter 3: Materials and apparatus
28
3.2.2. RheometerRheological measurements have been performed
with a MCR 300 rheometer produced by Paar Physica (Figure 3.2). A
parallel-plate measuring system was used as illustrated in Figure
3.3. This measuring system consisted of two plates. The surfaces of
both the bob and the motionless plate were flat, but the upper
plate had a serrated surface of 150 m depth to avoid slippage.
Figure 3.2 MCR 300 rheometer by Paar Physica
Figure 3.3 The parallel plate measuring system (Mezger T., [12],
p. 177)
The geometry of the upper plate is determined by the plate
radius R being 2.5 cm. The distance H between the two parallel
plates must be much smaller than the radius R and has been
recommended to be at least 10 times larger than the largest of the
particles of the sample (Mezger T., [12], p. 177-179). The average
particle size of unhydrated cement being
Chapter 3: Materials and apparatus
29
approximately 10 m (Taylor, [13]), the gap between the plates
was set to 1 mm for all measurements. The temperature controlled
bottom plate was set to 20 C. The parallel plate measuring system
makes it possible to measure dispersions containing relatively
large particles as well as samples with three-dimensional
structures. The measuring system has however also a number of
disadvantages. There is no constant shear gradient in the
measurement gap because the shear rate (or shear deformation)
increases in value from zero at the center of the plate to the
maximum at the edge. Furthermore, several unwanted phenomena can
occur at the edge of the plate: inhomogeneities, emptying of the
gap, flowingoff and spreading of the sample, evaporation of water,
or skin formation (Mezger T., [12], p. 180-181). To reduce
evaporation both upper and lower plates were covered with a plastic
ring and a metallic lid while a water trap attached to the upper
plate was filled with water to ensure saturated water pressure. The
following measuring sequence was used to determine the flow
resistance (area under the (down) flow curve in the range from 2 to
50 1/s), the gel strength after 10 seconds of resting and the gel
strength after 10 minutes of resting: 1. 1 minute with constant
shear rate ( ) of 100 1/s to stir up the paste 2. 1 minute resting
3. Stress ( ) shear rate ( ) curve with linear sweep of from 2 up
to 200 1/s in 30 points lasting 6 s each (up curve) 4. Stress ( )
shear rate ( ) curve with linear sweep of from 200 down to 2 1/s in
30 points lasting 6 s each (down curve) 5. 10 s resting 6. Shear
rate ( ) stress ( ) curve with logarithmic sweep of from 1 to 100
Pa in 30 points lasting 6 s each to measure the gel strength after
10 s rest 7. 10 minutes resting 8. Shear rate ( ) stress ( ) curve
with logarithmic sweep of from 1 to 400 Pa in 70 points lasting 6 s
each to measure the gel strength after 10 minutes rest The
recording of the shear rate ( ) stress ( ) curves was stopped
whenever the shear rate ( ) exceeded 300 1/s to prevent the sample
from being lost from the measurement gap. A flow chart of the
mixing and measurement sequence is shown in Figure 3.4.
Chapter 3: Materials and apparatus
30
Shear ratemixing minute mixing 1 minute
gel strength up curve 1 minute at 100 1/s down curve
gel strength
transfer to rheometer
5 minutes rest
8 minutes
1 minute rest
10 seconds rest
10 minutes rest
Time
Figure 3.4 Flow chart of the mixing and measurement sequenceThe
reproducibility of the rheological measurements was investigated
for two different cement pastes. The cement pastes were made with
distilled water. The plasticizer was added to the water. Cement
paste 1 was prepared with CEM I 52.5 R LA cement and 0.30% sodium
lignosulphonate by weight and a w/c ratio of 0.40. Paste 2 was
prepared with CEM I 42.5 RR cement and 0.50% sodium lignosulphonate
by weight and a w/c ratio of 0.50. Total paste volume was
approximately 250 ml. Each of the two cement pastes was prepared 5
times. The rheological data has been transformed into flow
resistance (area under the flow curve in the range from 2 to 50
1/s), gel strength after 10 seconds of rest and gel strength after
10 minutes of rest. The results are shown in Table 3.3 for cement
paste 1 and Table 3.5 for paste 2. The data show that the
reproducibility of the flow resistance is reasonable. Measurements
of the gel strength show higher deviations, especially for the 10
minute gel strength of the CEM I 52.5 R LA cement pastes which had
a standard deviation of 27%.
Chapter 3: Materials and apparatus
31
Table 3.4 Reproducibility of rheological measurements for cement
paste 1 (w/c=0.40 CEM I 52.5 R LA 0.30% Ultrazine Na) PASTE 1 Flow
resistance [Pa/s] 391 383 394 419 384 394 15 4% Gel strength [Pa]
10 sec. 10 min. 2.4 14.2 2.4 13.0 2.8 9.2 2.8 10.0 2.8 7.1 2.7 10.7
0.2 3 9% 27%
Average Standard deviation % standard dev.
Table 3.5 Reproducibility of rheological measurements for cement
paste 2 (w/c=0.50 CEM I 42.5 RR 0.50% Ultrazine Na) PASTE 2 Flow
resistance [Pa/s] 2119 2375 2455 2343 2392 2337 128 5% Gel strength
[Pa] 10 sec. 10 min. 22.2 36.8 22.2 36.8 26.1 40.1 22.2 40.1 22.2
36.8 2.7 38.1 1.7 2 7% 5%
Average Standard deviation % standard dev.
3.2.3. CalorimeterAn eight-channel TAM Air Isothermal
Calorimeter from Thermometric AB, Sweden was used for the heat of
hydration measurements (Figure 3.5). The calorimeter was calibrated
at 20 C. The hydration heat was measured by weighing 6 to 7 grams
of cement paste into a glass ampoule after which the ampoule was
sealed and loaded into the calorimeter. The ampoules were wiped
with a paper tissue to make sure that they were perfectly clean and
dry when they were inserted into the calorimeter. When studying the
heat of hydration measurements it should be kept in mind that when
an ampoule is loaded into the calorimeter the temperature of the
calorimeter will be disturbed. If the temperature of the ampoule is
2 degrees higher than the thermostat temperature, an exothermic
heat flow, showing an exponential decay, of roughly 400 mW is
observed. This phenomenon explains the exponential decay in
specific heat which is observed in the first hour after mixing.
Chapter 3: Materials and apparatus
32
Figure 3.5 TAM Air Isothermal Calorimeter
3.2.4. Adsorption of plasticizersTo measure the consumed amount
of lignosulfonate on the cement a UV Spectrophotometer from Thermo
Spectronic was used as illustrated in Figure 3.6. The adsorption
measurements in this work utilized a wavelength of 285 nm. Pore
solutions were extracted from the cement pastes by filtering the
pastes through 0.45 m filter paper on a Bchner funnel using low
vacuum 15 minutes after water addition. They were then diluted 25,
50 or 100 times with a solution of artificial pore water (NaOH and
KOH with a K/Na molar ratio equal to 2 and pH = 13.2). The amount
of plasticizer in the water phase was read from calibration curves
which had been made with a dilution series of each of the two
lignosulfonates being used in this work. The difference between the
added and the measured content of plasticizer gave the bound
portion.
Figure 3.6 UV Spectrophotometer from Thermo Spectronic
Chapter 3: Materials and apparatus
33
The consumption of polyacrylate on cement was determined by
measuring Total Organic Carbon (TOC) left in the pore water with a
Shimadzu TOC Analyzer 5000A. The Shimadzu TOC 5000A works by
converting organic matter to carbon dioxide by combustion with a
catalyst that promotes the redox reaction with oxygen. The reaction
takes place at a temperature of 680 C. The amount of carbon dioxide
formed is measured to determine the carbon content. The amount of
plasticizer bound to the cement is given by the difference between
the added and the measured content of organic carbon.
Chapter 4 Counteracting plasticizer retardation
4.1 IntroductionPlasticizers are used to increase flow for
cementitious materials at equal water-to-cement ratio, but will
also to a variable extent retard cement setting as a side effect.
The objective was to find an accelerator that at least partially
would counteract this retardation without negatively affecting the
rheology too much. Earlier papers (Justnes, H., Petersen, B.G.,
[14] and [15]) focusing on this topic studied rheological
properties at high shear rate (i.e. relevant for mixing) for
relatively low dosages of plasticizer, whereas the study reported
in this chapter focused on the lower shear rate range (i.e.
relevant for pouring concrete) and higher dosages of plasticizer.
Three different plasticizers were tested in the present study, but
the accelerator was chosen to be calcium nitrate. The experimental
work is largely carried out on cement paste using a Physica MCR 300
rheometer to determine flow curves and gel strength and an
isothermal calorimeter for determination of heat of hydration
curves. Two promising admixture blends were also tried out in
mortar.
34
Chapter 4: Counteracting plasticizer retardation
35
4.2 Calorimetric and rheological measurements4.2.1. Experimental
The investigated cement pastes were made with distilled water.
Plasticizer and accelerator were added to the water before mixing,
except for one series of pastes marked with DA (delayed addition),
where the plasticizer was added 5 minutes after the start of
initial blending in a 30% aqueous solution. Both a CEM I 52.5 R LA
and a CEM I 42.5 RR Portland cement were used. Three different
plasticizers were studied: a sodium lignosulphonate (NLS), a
calcium lignosulphonate (CLS) and a polyether grafted polyacrylate
(PA). The setting accelerator calcium nitrate (CN), available in a
50% aqueous solution, was used to counteract the retardation. A
more detailed description of both plasticizers and accelerator can
be found in Chapter 3. Table 4.1 provides an overview of the
experimental program.
Table 4.1 Experimental program Plasticizer Accelerator Reference
(0%) 0.15% NLS* 0.15% NLS DA* 0.30% NLS 0.50% NLS 0.00% CN 0.30%
CLS 0.25% CN 0.50% CLS 0.50% CN 0.75% CN 0.10% PA 1.00% CN CEM I
42.5 RR Reference (0%) (w/c = 0.50) 0.50% NLS 1.00% NLS 0.50% CLS
1.00% CLS 0.10% PA (* The 1.00% CN dosage was not studied for these
series.) Cement type CEM I 52.5 R LA (w/c = 0.40)
In Chapter 3 it was pointed out that the CEM I 42.5 RR cement
had a higher alkali and C3A content and a higher specific surface
than the CEM I 52.5 R LA cement and, as a consequence of the latter
two, had a higher water demand. CEM I 42.5 RR cement pastes were
therefore prepared with a w/c ratio of 0.50, whereas CEM I 52.5 R
LA cement pastes were prepared with a w/c ratio of 0.40 throughout
this work. Total paste volume was approximately 250 ml. The
blending was performed in a high shear mixer of Braun (see 3.2.1)
by adding the cement to the water containing plasticizer and/or
accelerator and mixing for minute, resting for 5
Chapter 4: Counteracting plasticizer retardation
36
minutes and blending again for 1 minute. The cement pastes
containing 0.15% sodium lignosulphonate were mixed with a high
shear mixer by Tefal using the same blending sequence. The heat of
hydration versus time curves were measured by accurately weighing 6
to 7 grams of cement paste into a glass ampoule after which the
ampoule was sealed and loaded into the calorimeter. The rheological
properties were studied by performing the measurement sequence
discussed in section 3.2.2 on the cement pastes 15 minutes after
the start of the blending: To measure the consumed (adsorbed and
intercalated) amount of plasticizer by cement, pore solutions were
extracted from the cement pastes by filtering the pastes through
0.45 m filter paper on a Bchner funnel using low vacuum 15 minutes
after water addition. The consumed amount of lignosulphonate was
determined using a UV Spectrophotometer from Thermo Spectronic. The
adsorption measurements in this work utilized a wavelength of 285
nm. The pore solutions were diluted 25, 50 or 100 times with a
solution of artificial pore water (NaOH and KOH with a K/Na molar
ratio equal to 2 and pH = 13.5). The amount of plasticizer in the
water phase was read from calibration curves which had been made
with a dilution series of each of the two lignosulphonates being
used in this work. The calibration curves for NLS and CLS are given
in Figure 4.1 and Figure 4.2 respectively. The difference between
the added and the measured content of plasticizer gave the consumed
amount. The consumption of polyacrylate by cement was determined by
measuring Total Organic Carbon (TOC) left in the pore water with a
Shimadzu TOC Analyzer 5000A.
Chapter 4: Counteracting plasticizer retardation
37
Calibration curve, NLS0.9 0.8 0.7 Absorbance 0.6 0.5 0.4 0.3 0.2
0.1 0 0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 % Added y =
137.5844x R2 = 0.9992
Figure 4.1 Calibration curve for adsorbance of sodium
lignosulphonate (NLS).
Calibration curve, CLS0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0
0.001
Absorbance
y = 137.2718x R2 = 0.9997
0.002
0.003
0.004
0.005
0.006
0.007
% Added
Figure 4.2 Calibration curve for adsorbance of calcium
lignosulphonate (CLS). Prior to discussing the results, we shall
provide an overview of the way in which the read outs from the
rheometer were converted into flow resistance (area under the flow
curve in the range from 2 to 50 1/s, see also Chapter 2), gel
strength after 10 seconds of rest and gel strength after 10 minutes
of rest. The measurements on the cement paste made with CEM I 52.5
R LA cement without any admixtures shall be used to illustrate
this:
Chapter 4: Counteracting plasticizer retardation
38
1. The flow resistance is defined as the area under the down
flow curve in the range from 2 to 50 1/s. The down curve for the
paste made with CEM I 52.5 R LA cement is shown in Figure 4.3.
Table 4.2 shows the read outs from the rheometer. The area under
the curve was determined by calculating the average of the shear
stresses for every two consecutive measuring points in the range
from 2 to 50 1/s and multiplying this by the difference in shear
rate for these points. In this case a value of 2283 Pa/s was found
for the flow resistance.
Table 4.2 Rheometer read outs for the down curve. Meas. Pt. 1 2
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
28 29 30 Shear Rate [1/s] 200 193 186 180 173 166 159 152 145 139
132 125 118 111 104 97.6 90.8 83.9 77.1 70.3 63.4 56.6 49.8 43.0
36.1 29.3 22.5 15.7 8.83 2.01 Shear Stress [Pa] 98.2 96.9 95.7 94.6
93.4 92.9 91.1 89.9 88.8 87.5 86.4 85.1 83.8 82.5 81.1 79.5 77.7
75.7 73.6 71.4 69.4 67.3 64.7 61.4 58.0 53.3 47.3 40.1 30.8
22.2
Chapter 4: Counteracting plasticizer retardation
39
Down Curve120 100 Shear Stress [Pa] 80 60 40 20 0 0 50 100 Shear
Rate [1/s] 150 200
Figure 4.3 Down curve.
2. The 10 sec. gel strength can be derived from the shear rate (
) stress ( ) curve with logarithmic sweep of from 1 to 100 Pa in 30
points lasting 6 s each. The curve is plotted in Figure 4.4. The
rheometer read outs are given in Table 4.3. The 10 sec. gel
strength was calculated by taking the average of the shear stresses
of measuring points 19 and 20 (Table 4.3) as the breakthrough
happened somewhere in between. That way a value of 19 Pa was found
for the 10 sec. gel strength.
10 sec. gel strength180 160 140 Shear Rate [1/s] 120 100 80 60
40 20 0 0 20 40 60 80 100 Shear Stress [Pa]
gel strength
Figure 4.4 Shear rate stress curve to determine the 10 sec. gel
strength.
Chapter 4: Counteracting plasticizer retardation Table 4.3
Rheometer read outs to determine the 10 sec gel strength. Meas. Pt.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
26 27 28 29 30 Shear Rate [1/s] 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 3.44
9.32 12.9 16.0 21.3 26.1 35.0 48.7 73.6 110 155 Shear Stress [Pa]
1.00 1.17 1.37 1.61 1.89 2.21 2.59 3.04 3.56 4.18 4.89 5.74 6.72
7.88 9.24 10.8 12.7 14.9 17.4 20.4 24.0 28.1 32.9 38.6 45.2 53.0
62.1 72.8 85.3 100
40
3. The calculation of the 10 min. gel strength is completely
similar to that of the 10 sec. gel strength and shall therefore not
be treated.
Chapter 4: Counteracting plasticizer retardation 4.2.2. Results
and discussion for reference pastes
41
Figure 4.5 shows the flow resistances for both CEM I 52.5 R LA
and CEM I 42.5 RR reference cement pastes. The flow resistance of
the CEM I 42.5 RR cement paste (w/c = 0.50) is higher than the CEM
I 52.5 R LA paste (w/c = 0.40) in spite of the higher
water-to-cement ratio. This is due to the higher specific surface
and the content of cubic C3A. Addition of calcium nitrate appeared
to have no effect on the flow resistance of these pastes.
reference3500 CEM I 52.5 R LA CEM I 42.5 RR
Flow resistance (Pa/s)
3000 2500 2000 1500 1000 500 0 0.00
0.25
0.50
0.75
1.00
Calcium nitrate (%)
Figure 4.5 Flow resistance for CEM I 52.5 R LA (w/c = 0.40) and
CEM I 42.5 RR reference cement pastes (w/c = 0.50) for different
dosages of calcium nitrate.
The gel strengths after 10 seconds of rest are depicted in
Figure 4.6. In case of CEM I 52.5 R LA cement paste, an increasing
10 seconds gel strength was observed for increasing calcium nitrate
dosages up to 0.50%. Figure 4.7 shows the gel strengths after 10
minutes of rest. For both cement types an increasing (albeit less
pronounced in case of CEM I 42.5 RR cement) gel strength can be
seen for increasing calcium nitrate dosages.
Chapter 4: Counteracting plasticizer retardation
42
reference35
CEM I 52.5 R LA
10 sec. gel strength (Pa)
30 25 20 15 10 5 0 0.00
CEM I 42.5 RR
0.25
0.50
0.75
1.00
Calcium nitrate (%)
Figure 4.6 Gel strength after 10 seconds of rest for CEM I 52.5
R LA (w/c = 0.40) and CEM I 42.5 RR reference cement pastes (w/c =
0.50) for different dosages of calcium nitrate.
reference300
CEM I 52.5 R LA CEM I 42.5 RR
10 min. gel strength (Pa)
250 200 150 100 50 0 0.00 0.25 0.50 0.75 1.00
Calcium nitrate (%)
Figure 4.7 Gel strength after 10 minutes of rest for CEM I 52.5
R LA (w/c = 0.40) and CEM I 42.5 RR reference cement pastes (w/c =
0.50) for different dosages of calcium nitrate.
Chapter 4: Counteracting plasticizer retardation
43
The heat of hydration curves are shown in Figure 4.8 and Figure
4.9. It can be seen that calcium nitrate speeded up hydration with
approximately two hours for both cement types. The peak in the
hydration curve for the pastes without calcium nitrate was seen at
about 9 hours after water addition.CEM I 52.5 R LA - w/c = 0.40 -
reference2.5
1.00 % CN 2 0.75 % CN
Rate of hydration heat (mW/g)
0.50 % CN 1.5
0.25 % CN 1 0.00 % CN
0.5
0 1 3 5 7 9 11 13 15 17 19 21 23 25
Time (hours)
Figure 4.8 Heat of hydration curves for CEM I 52.5 R LA cement
pastes (w/c = 0.40) for different dosages of calcium nitrate.CEM I
42.5 RR - w/c = 0.50 - reference4
3.5
3
1.00 % CN 0.75 % CN
Rate of hydration heat (mW/g)
2.5
2
1.5 0.00 % CN 1 0.25 % CN 0.50 % CN 0.5
0 1 3 5 7 9 11 13 15 17 19
Time (hours)
Figure 4.9 Heat of hydration curves for CEM I 42.5 RR cement
pastes (w/c = 0.50) for different dosages of calcium nitrate.
Chapter 4: Counteracting plasticizer retardation 4.2.3. Results
and discussion for sodium lignosulphonate
44
Flow resistances, gel strengths after 10 seconds and 10 minutes
of rest measured on CEM I 52.5 R LA cement pastes (w/c = 0.40) are
listed in Table 4.4, Table 4.5 and Table 4.6, respectively. Those
measured on CEM I 42.5 RR cement pastes (w/c = 0.50) are listed in
Table 4.7, Table 4.8 and Table 4.9. Table 4.4 Flow resistance
(Pa/s) for CEM I 52.5 R LA cement paste (w/c=0.40). Flow resistance
[Pa/s] Reference 0.15% NLS 0.15% NLS DA 0.30% NLS 0.50% NLS Calcium
nitrate [%] 0.25 0.50 0.75 2253 2515 2418 1973 1815 2060 618 727
839 651 819 1030 287 528 671
0.00 2283 1552 683 353 147
1.00 2372
1201 881
Table 4.5 Gel strength after 10 seconds of rest (Pa) for CEM I
52.5 R LA cement paste (w/c=0.40). 10 sec. gel strength [Pa]
Reference 0.15% NLS 0.15% NLS DA 0.30% NLS 0.50% NLS 0.00 18.9 22.2
5.3 2.4