Get documents about "Early hydration of ordinary Portland cement with an
Document Sample
Advances in Cement Research, 2005, 17, No. 1, January, 1–8
Early hydration of ordinary Portland cement with
an alkaline shotcrete accelerator
Q. Xu* and J. Stark*
Bauhaus-University Weimar, Germany
The early hydration (, 24 h) of ordinary Portland cement (OPC) with an added alkaline shotcrete accelerator
was investigated by means of differential calorimetric analysis, an environmental scanning electron microscope,
inductively coupled plasma-optical emission spectroscopy and X-ray diffraction analysis. Compared with normal
cement paste, the presence of the accelerator first changes the hydration conditions, such as the ion
concentrations and the pH value of the liquid phase of the paste directly after mixing. It influences the reactions
of cement clinkers and the development of the hydrate phases. A foil-like hydration product composed of calcium
aluminate hydrate (CAH) phases, long fibre-like ettringite crystals and tabular monosulphates was detected in the
microstructure of the paste with the accelerator. The additional Al(OH)4 À from the accelerator takes part in the
hydration reactions to form CAH phases. The reactions of the accelerator are dependent on the concentration of
Ca2þ ions in the liquid phase of the paste in the early course of hydration.
and this rapid setting makes it difficult to prepare the
Introduction
sample. All of this restricts the theoretical study of the
With the wide application of shotcrete in tunnels, effect of accelerators on cement hydration.
bridges and other underground constructions, there is Recently, with the introduction of new technologies
increasing interest in improving its properties. Shotcrete in cement science, much progress has been made in the
has two main properties: rapid setting and high early study of early cement hydration. The use of the
strength. Both are closely related to the early hydration environmental scanning electron microscope (ESEM)
of cement. In order to attain rapid setting and high in cement, for example, makes it possible to investigate
early strength, a variety of set accelerators and hard- cement hydration in a moist atmosphere on the nano-
ening accelerators are used. There are many publica- metre scale. This could reveal the hydration phenomen-
4–7
tions describing the practical use of these accelerators on realistically. Stark and Moeser et al. have
1–3
in shotcrete. Most accelerators are selected from the systematically studied the hydration of synthetic clin-
standpoint of economy and technology. However, very kers, technical clinkers and technical cement with the
little has been reported about the mechanism of the help of ESEM equipped with a field emission gun
effect of accelerators on cement hydration, especially (FEG), and other techniques. As a result they have
very early hydration. proposed new statements and a new model of cement
It is difficult to study the early hydration of cement, hydration, especially in respect of hydration phase
because it is a very complex, multiphase system. There development and microstructure.
are still many unanswered questions on the kinetics and On the basis of the new findings, the reason for the
thermodynamics of cement hydration. Moreover, when rapid setting of cement paste caused by an alkaline set
chemical additives such as accelerators are added to the accelerator, and the effect of the accelerator on the early
cement, both the hydration conditions and the hydration strength of cement mortar, have been discussed in
8,9
process are changed. The accelerator also causes very previous publications. This work will focus on the
rapid hydration in the first few minutes after mixing, mechanism of the effects of the set accelerator on
cement hydration at the early stages. It is intended to
* F. A. Finger Institute for Building Materials Science, Director, investigate the hydration of cement in the presence of
Department of Civil Engineering, Bauhaus-University Weimar, the accelerator and the reactions of the accelerator in the
Coudraystrasse 11, 99421 Weimar, Germany.
cement system. The formation and growth of the hydra-
(ACR 3433) Paper received 6 February 2003; last revised 25 tion products and the dissolution of the set regulators in
November 2003; accepted 1 April 2004 the cement will also be discussed. The testing techni-
1
0951-7197 # 2005 Thomas Telford Ltd
Xu and Stark
ques include differential calorimetric analysis (DCA) for tion calorimeter. The formation and morphology of the
the hydration heat liberation, environmental scanning hydration products were investigated through ESEM-
electron microscope–energy-dispersive X-ray spectro- EDX after specified hydration times. The composition
metry (ESEM–EDX) for phase development in the of the liquid phase (pore solution) of the paste was
microstructure, X-ray diffraction analysis (XRD) for analysed by ICP-OES. It is noticeable that the accuracy
crystals of the solid phase of the cement paste, and of ICP-OES for aluminium concentration in the liquid
inductively coupled plasma-optical emission spectro- is 0.1 mg/l (’ 0.004 mmol/l). So the low concentration
scopy (ICP-OES) for the liquid phase of the paste. of the aluminate ions (Al(OH)4 À ) in the liquid phase of
the cement paste can be measured. The pH values were
tested by a pH meter with a Ross special pH electrode.
Experimental
Materials
A German OPC, CEM I 42.5 R, was used. The Results
shotcrete accelerator was a type of alkaline set accel-
erator. It is composed mainly of sodium aluminate, DCA
NaAl(OH)4 , and sodium hydroxide, NaOH. In addition, The DCA diagram (Fig. 2) displays different stages
synthetic C3 S and ß-hemihydrate (CaSO4 Á0.5H2 O) were of cement hydration: I, initial period; II, dormant
also used. period; III, acceleration period; and IV, deceleration
period. Compared with a normal paste, the presence of
Specimens the accelerator greatly enhances the process of cement
Cement pastes were prepared at a total w/c ratio of hydration. In the initial period the heat liberation rate
0.5 (including the water content of the accelerator). of the paste with 5% accelerator is 50 J/(g h) higher
The dosage of the accelerator was chosen as 5% of the than that of the normal paste (Fig. 2(a)). This suggests
weight of cement. All samples were stored after mixing that more reaction takes place in the first 30 min when
in sealed plastic bottles except the samples intended for accelerator is added. In the dormant period there is an
hydration heat liberation. The liquid phase of the fresh
paste was extracted by filter suction (pore size 0.8 ìm),
160
and the pore solution of the hardened paste was pressed
Rate of heat liberation: J/(gh)
by a compressive machine (Fig. 1); the maximum
pressure on the sample was 256 N/mm2 . 120
50 J/(gh) 5% accelerator
Testing methods
80
The hydration heat liberation of the cement paste
was measured by DCA and an isothermal heat conduc-
40
Press plate Normal paste
0
0 0·1 0·2 0·3 0·4 0·5
Plunger Hydration time: h
(a)
Base body 35
II III IV
Rate of heat liberation: J/(gh)
30
Bushing 25
10 J/(gh)
20 5% accelerator
15
10
Normal paste
Sample 5
0
5
0
5
0
5
0
5
·0
·5
·0
·5
·0
·5
·0
0·
2·
3·
5·
6·
8·
9·
11
12
14
15
17
18
20
Hydration time: h
Base plate (b)
Fig. 2. DCA diagram of hydration of normal cement pastes
Fig. 1. Schematic diagram of compressive machine for pore and paste with 5% accelerator: (a) I, initial; (b) II, dormant
solution expression from the hardened paste period; III, acceleration period; IV, deceleration period
2 Advances in Cement Research, 2005, 17, No. 1
Early hydration of ordinary Portland cement with an alkaline shotcrete accelerator
on the surfaces of cement particles and in the space between the
Growth of CSH, portlandite, ettringite crystals and monosulphate
Formation of a hydrating layer containing foil-like CAH phases
extra reaction peak occurring in the paste with accel-
Formation of short, needle-like CSH and tabular portlandite
erator (Fig. 2(b)). In the acceleration period the maxi-
mum heat liberation rate of the sample with 5%
accelerator is also ’ 10 J/(g h) larger than in normal
paste (Fig. 2(b)).
Formation and growth of tabular monosulphate
Formation of long, fibre-like ettringite crystals
Growth of long, needle-like ettringite crystals
ESEM-EDX
Table 1 describes the phase formation and growth in
the microstructure of both the normal paste and the
paste with accelerator, as observed by means of ESEM-
FEG. It is noticeable that the addition of 5% accel-
Reduction of CAH phases
erator varies the development of hydrate phases. In the
Growth of CAH phases
Paste with accelerator
initial period of cement hydration, instead of ettringite
crystals formed in the normal paste (Fig. 3), a hydrat-
ing layer containing a foil-like phase was detected as
the main hydration product surrounding the cement
particles
crystals
particles (Fig. 4). In the dormant period a large quantity
of long, fibre-like ettringite crystals was formed, and
they slowly replaced the early foil-like phases (Fig. 5).
Formation of small prismatic crystals of ettringite, max. length
In the acceleration period, a hydration product in a
tabular form was observed in the microstructure (Fig.
6). According to the EDX analysis, this tabular phase
was most probably monosulphate (Fig. 7). Neverthe-
Growth of CSH, portlandite and ettringite crystals
less, no syngenite crystals were found in the micro-
300 nm, on the surfaces of cement particles
Formation of short, needle-like CSH phases
structure of the paste with 5% accelerator.
Formation of tabular syngenite crystals
ICP–OES
More formation of syngenite crystals
It can be seen from Figs 8 and 9 that the addition of
Formation of tabular portlandite
the accelerator has strongly enhanced the concentra-
Growth of ettringite crystals
Growth of ettringite crystals
tions of Naþ ions and the pH value in the liquid phase
(pore solution) of the cement paste. The normal paste
has Naþ ion concentrations ranging from 20 to
Phase description
50 mmol/l, but the paste with 5% accelerator has from
Normal paste
600 to 700 mmol/l, an enhancement of about 15–30
times. The hugely increased content of alkali ions
results in a rise of the pH value to above 13.6, so high
Table 1. Formation and growth of hydrates as observed by ESEM-FEG
that the normal paste could never reach this value so
rapidly. Comparing Fig. 10 with Fig. 11, we can see
that the concentrations of SO4 2À and Al(OH)4 À ions
Paste with accelerator
are raised, but the concentration of Ca2þ ions is
reduced between 10 min and 1 h of hydration if accel-
Extra reactions
erator is added. For example, 10 min after the paste is
mixed with water the SO4 2À and Ca2þ ion concentra-
tions in the liquid phase of the normal paste are
approximately 120 mmol/l and 20 mmol/l respectively,
and the Al(OH)4 À ion concentration is close to the
Deceleration period
Acceleration period
accuracy of measurement (Æ 0.1 mg/l). Correspond-
Dormant period
ingly in the paste with 5% accelerator the SO4 2À ion
Normal paste
Initial period
concentration is much higher (about 160 mmol/l) but
the Ca2þ ion concentration is much lower—only
Period
0.58 mmol/l, or about 1/33 of that of the normal paste.
On the other hand, the Al(OH)4 À ion concentration is
greatly enhanced to 20 mmol/l because of the addition
of the aluminate-containing accelerator, although the
original contribution of the accelerator to the Al(OH)4 À
0.5 $ 2 h
8 $ 20 h
ion concentration is about 372 mmol/l. Afterwards the
2 $ 8 h
$ 0.5 h
Al(OH)4 À ion concentration decreases after 3 h hydra-
Time
tion in the paste with 5% accelerator and becomes
Advances in Cement Research, 2005, 17, No. 1 3
Xu and Stark
Acc.V Spot Magn Det WD Exp 2 µm Acc.V Spot Magn Det WD Exp 2 µm
25.0 kV 3·0 16000x GSE 10.1 0 9.7 Torr CEMI42.5R 10min/K 25.0 kV 3.0 16000x GSE 10.0 0 9.5 TorrCEMI42.5R %BE/6h
Fig. 3. Normal paste, 10 min hydration: short, prismatic Fig. 6. Paste with 5% accelerator, 6 h hydration: tabular
ettringite crystals hydrate phases
O
Al
Si S
Ca Ca
0·50 0·90 1·30 1·70 2·10 2·50 2·90 3·30 3·70 4·10 4·50
Acc.V Spot Magn Det WD Exp 2 µm Fig. 7. EDX analysis of tabular hydrate phases (see Fig. 6);
25.0 kV 3.0 16000x GSE 10.2 0 9.4 Torr CEMI42.5R 5%BE/10 min/K they have a similar Al/S ratio to that of monosulphates
Fig. 4. Paste with 5% accelerator, 10 min hydration: foil-like
hydrate phases around the cement particles
800
Na1 ion concentration: mmol/l
600
Normal paste
400
With 5% accelerator
200
0
10 min 1h 3h 6h 24 h
Hydration time
Fig. 8. Naþ ion concentration of the liquid phase
Acc.V Spot Magn Det WD Exp 2 µm comparable to that in the normal paste. Furthermore,
25.0 kV 3.0 8000x GSE 10.0 0 9.5 TorrCEMI42.5R %BE/3h the Ca2þ ion concentrations in the paste with accel-
erator are always much lower than in the normal paste.
Fig. 5. Paste with 5% accelerator, 3 h hydration: long, fibre- The changes of the SO4 2À, Ca2þ and Al(OH)4 À ion
like hydrate phases, assumed to be ettringite concentrations in the liquid phase (pore solution) cause
4 Advances in Cement Research, 2005, 17, No. 1
Early hydration of ordinary Portland cement with an alkaline shotcrete accelerator
14
Discussion
13·6 Initial period
Addition of this accelerator promotes the concentra-
pH values
13·2 tions of Naþ , Al(OH)4 À and SO4 2- ions and the pH
values in the pore solution of the cement paste. Thus
12·8
the cement particles are surrounded by a solution with
Normal paste high concentrations of Naþ , OH À , SO4 2À and
With 5% accelerator Al(OH)4 À ions upon mixing the cement with water.
12·4
10 min 1h 3h 6h 24 h The reactivity of C3 A seems to be largely enhanced by
Hydration time the rising pH values in the solution. A hydrate layer is
quickly formed around the cement particles. This
Fig. 9. pH values of the liquid phase hydrate layer consists of foil-like phases, which are
assumed to be CAH. As they are in the size range of
several nanometres, their identification by EDX is
approximate. But they are most probably products from
180
the hydration of C3 A and reaction of the added
SO4, normal paste
Ca21, normal paste aluminate because the amount of Al(OH)4 À ions in the
Concentrations of ions: mmol/l
150 Al31, normal paste liquid phase quickly decreases from 372 mmol/l to
20 mmol/l after 10 min hydration. The reactions are as
120
follows:
90
C3 A þ 21H ! C4 AH13 þ 2 AH8
C
(1)
60 ˜f G8 ¼ À415 kJ=mol
298
30 À À
ðOHÞ4 þ 6CH þ 9H C4 AH13 þ C2 AH8 þ 4OH
Al !
0 ˜f G298 ¼ À41 kJ=mol
8
10 min 1h 3h 6h 24 h
Hydration time (2)
Fig. 10. SO4 2À, Ca2þ and Al(OH)4 À ion concentrations in
the normal paste
The calculated Gibbs free energy indicates the possibi-
lity of reactions being displaced as shown. In addition
to the foil-like CAH phases, the initially formed
hydrating layer most probably contains gel-like silicate
and precipitated Ca(OH)2 . From the ESEM photos of a
synthetic C3 S with the same dosage of accelerator after
180 SO4, 5% accelerator 10 min hydration, a gel-like hydrate layer is also found
Ca21, 5% accelerator around the cement particles, whereas the sample with-
Concentrations of ions: mmol/l
150 Al31, 5% accelerator
out accelerator shows few changes to the surfaces of
120 the C3 S particles (Figs 12 and 13).
Furthermore it is very noticeable that the presence of
90
60
30
0
10 min 1h 3h 6h 24 h
Hydration time
Fig. 11. SO4 2À, Ca2þ and Al(OH)4 À ion concentrations in
the paste with 5% accelerator
the change of the proportions of these three ions,
especially during the first 3 h of cement hydration. This Acc.V Spot Magn Det WD Exp 2 µm
will dramatically influence the development of the 25.0 kV 3.0 8000x GSE 10.1 0 9.8 Torr2000 C3S/10 min RT-Hydr
hydrate phases and thus affect the performance of the
cement paste. Fig. 12. Synthetic C3 S, 10 min hydration
Advances in Cement Research, 2005, 17, No. 1 5
Xu and Stark
G G
500
G
400
P
300 S S S
cps
200 (a)
100
(b)
0
·6
·0
·7
·8
·3
·3
·4
11
18
20
21
23
24
25
Acc.V Spot Magn Det WD Exp 2 µm 2-Theta-Scale
25.0 kV 3.0 8000x GSE 9.9 0 8.9 Torr C3S 5%/10 min
Fig. 15. XRD analysis of the solid phase of â-CaSO4 Á0.5H2 O
in (a) water and (b) an alkaline solution after 1 h: G,
Fig. 13. Synthetic C3 S + 5% accelerator, 10 min hydration
gypsum; P, portlandite; S, Na3 K(SO4 )2 or K3 Na(SO4 )2 salts
the accelerator greatly increases the SO4 2À ion concen-
trations but decreases the Ca2þ ion concentrations in releasing 1 mole of Ca2þ ions and 1 mol of SO4 2À ions,
the liquid phase of the paste, especially in the first 1 h the molar Ca2þ /SO4 2À ratio is about 1. However, when
of hydration. As we know, after cement is mixed with â-CaSO4 Á0.5H2 O is present in the alkaline solution, the
water, alkaline sulphates from clinkers and calcium molar Ca2þ /SO4 2À ratio is drastically lowered (Fig. 14).
sulphates from the set regulator enter the liquid phase This result is similar to that for the paste with 5%
of the paste, contributing the content of sulphates in the accelerator. XRD analysis confirms the formation of
solution. Highly soluble salts, such as alkaline sul- portlandite and complex salts of alkali sulphates
phates, are rarely influenced by the liquid conditions, (Na3 K(SO4 )2 or K3 Na(SO4 )2 ) in the solid phase when
but the dissolution of calcium sulphates could be â-CaSO4 Á0.5H2 O dissolves in an alkaline solution (Fig
strongly varied by changes to the composition of the 15). Therefore it could be concluded that, when CaSO4
10
liquid phase. Locher and Richartz have reported that is present in the alkaline solution, it reacts with the
the solubility of gypsum decreases with increasing pH. alkaline hydroxide to form Ca(OH)2 and release SO4 2À
Therefore it could reasonably be concluded that both ions:
the SO4 2À and the Ca2þ ion concentrations will be
2OH À þ CaSO4 ! CaðOHÞ2 þSO2À
4
lowered by increasing pH. The results here do not (3)
completely support this conclusion, because the SO4 2À ˜f G8 ¼ À5:5 kJ=mol
298
ion concentrations are not lowered, but raised. Figures
14 and 15 show the analysis of the liquid phase The lower solubility of Ca(OH)2 compared with
composition and the solid phase of the hemihydrate CaSO4, especially in highly alkaline solution, leads to
dissolved in water and at pH 13.8 (similar to the the precipitation of Ca(OH)2 and the low Ca2þ ion
conditions of the cement paste with 5% accelerator). concentration in the liquid phase. On the other hand, it
When 1 mol of â-CaSO4 Á0.5H2 O dissolves in water, seems that the added NaAl(OH)4 reacts rapidly with
Ca(OH)2 , because much Al(OH)4 À is consumed from
the pore solution after 10 min of hydration (the
Al(OH)4 À concentration after 10 min was 20 mmol/l,
180 14
Ca ions whereas 5% accelerator contributed 372 mmol/l of
SO4 ions 12 Al(OH)4 À ions to the cement paste). This is another
Ions concentrations: mmol/l
150 pH value
10 reason for the low concentrations of Ca2þ ions in the
120
pore solution. Table 2 lists some possible chemical
pH value
8
90 equations relating to NaAl(OH)4 and their Gibbs free
6 energy.
60 It can been seen that all the reactions of NaAl(OH)4
4
30
have a negative Gibbs free energy, which means that
2
the equilibrium is displaced, as shown by the arrows.
0 0 Since no reaction in Table 2 has sufficiently more
In water In alkaline solution
negative free energy than the others, it is difficult to
Fig. 14. Ca2þ , SO4 2À ion concentrations and pH values of say which reaction is more favourable. Other testing
the liquid phase of â-CaSO4 Á0.5H2 O in water and in an methods are needed to confirm the reaction. In addi-
alkaline solution after 1 h tion, all the reactions consume Ca(OH)2 . So it could be
6 Advances in Cement Research, 2005, 17, No. 1
Early hydration of ordinary Portland cement with an alkaline shotcrete accelerator
11, 12
Table 2. Possible reactions of NaAl(OH)4 in cement paste and their Gibbs free energy
Reactions Free energy, ˜f G8 : kJ/mol
298
4Al(OH)4 À + 6CH + 9H ! C4 AH13 + C2 AH8 + 4OH À À41 (equation (2))
2Al(OH)4 À + 3CH + 3C SH2 +26H ! C3 AÁ3C SÁH32 + 2OH À À79
2Al(OH)4 À + 6CH + 3SO4 2À + 26H ! C3 AÁ3C SÁH32 + 8OH À À62 (equation (4))
4Al(OH)4 À + 6CH + C3 AÁ3C SÁH32 ! 3(C3 AÁC SÁH12 ) + 4OH À + 8H À44
2Al(OH)4 À + 3CH + C SH2 +4H ! C3 AÁC SÁH12 + 2OH À À41
2Al(OH)4 À + 4CH + SO4 2À + 6H ! C3 AÁC SÁH12 +4OH À À36
concluded that the reaction of NaAl(OH)4 is Ca(OH)2 - Meanwhile, as a metastable hydrate phase, the formed
controlled. CAH phases tend to transform to the more stable
Although there is a great deal of alkaline sulphate in phases such as ettringite crystals, so long as Ca(OH)2
the liquid phase, syngenite crystals (K2 SO4 Á CaSO4 Á is available in the pore solution:
H2 O) are not detected in the microstructure of the paste
C2 AH8 þ 4CH þ 3SO2À þ 23H ! C3 A Á 3CS Á H32 þ
4
with 5% accelerator: this could be ascribed to the lack
6OH À
of CaSO4, and further proves the low content of Ca2þ
˜f G298 ¼ À49 kJ=mol
8
in the pore solution.
(5)
Dormant period
Ettringite replaces the CAH phases and covers the
While the short prismatic ettringite crystals are surfaces of the cement particles.
growing in the normal cement paste, an extra reaction
occurs in the sample with the accelerator. ESEM has Acceleration period
observed that a long fibre-like phase is formed both on The CSH phases and portlandite crystals form and
the surfaces and between the particles in the micro- ettringite crystals grow, both in normal paste and in the
structure of the paste containing accelerator. The paste with accelerator. For the latter sample, as well as
formation of the phases is accompanied by the reduc- ettringite crystals, tabular monosulphate crystals are
tion of Al(OH)4 À and SO4 2À ion concentrations in the also observed. This is ascribed mainly to the sufficient
pore solution. XRD has also detected the ettringite content of aluminate and the deficient content of
peak after 1 h hydration in the paste with 5% accel- calcium sulphate available in the pore solution. On the
erator (Fig. 16). other hand, the ettringite crystals continue to replace
It can be thus concluded that ettringite crystals are the CAH phases until the latter completely vanish.
formed. The formation of the long fibre-like ettringite Owing to the low content of CaSO4, the residual C3 A
results from the reaction of the added NaAl(OH)4 : reacts with ettringite to form monosulphates:
À
2AlðOHÞ4 þ6CH þ 3SO2À þ 26H ! À Á
3 A Á 3CS Á H32 þ 2C3 A þ 4H ! 3 C3 A Á 3CS Á H12
4 C
C3 A Á 3CS Á H32 þ 8OH À
(4)
˜f G8 ¼ À417 kJ=mol
˜f G8 ¼ À62 kJ=mol
298
298
(6)
Meanwhile, the Al(OH)4 À ions could also react with
ettringite so long as Ca2þ ions are available:
P
C CÀ A Á 3CS Á H32 þÁ4AlðOHÞ4 À þ 6CH !
3
E 3 C3 A Á 3CS Á H12 þ 4OH À þ 8H
(7)
E ˜f G8 ¼ À44 kJ=mol
298
30 24 h
6h Deceleration period
20 3h
cps
In this period the hydrate phases, including the CSH
1h
10
phase, portlandite and ettringite crystals, are further
10 min
growing in both the normal sample and the sample
0 containing accelerator. For the paste with 5% accel-
9·07 15·7 18·0 22·8
erator, the CAH phases disappear. As metastable
2-Theta-Scale
phases, CAH in the paste transforms to ettringite and
Fig. 16. XRD analysis of paste with 5% accelerator after monosulphate phases. On the other hand, although the
hydration for 10 min, 1 h, 3 h, 6 h and 24 h hydration: E, DCA diagram showed a higher peak in this period for
ettringite; P, portlandite; C, calcite the sample with 5% accelerator, neither more CSH
Advances in Cement Research, 2005, 17, No. 1 7
Xu and Stark
phases nor more portlandite are observed. The rise of 3. Manns W. Special cement for shotcrete. Beton, 2001, 51, No.
the heat liberation here may be ascribed to the forma- 9, 482–486.
4. Eckart A. and Stark J. Observation of hydration products of
tion of monosulphates.
calcium aluminates and calcium ferrites in ESEM-FEG.
Proceedings of the 13th International Conference on Building
Materials (IBAUSIL), Weimar, 1997, 901–919.
Conclusions 5. Stark J., Moeser B. and Eckart A. A new approach to
(a) The early hydration of OPC is greatly changed by the cement hydration: Part I. Cement-Lime-Gypsum (ZKG), 2001,
54, No. 1, 52–60.
addition of the set accelerator containing sodium 6. Stark J., Moeser B. and Eckart A. A new approach to
aluminate. Far more hydration heat is liberated in the cement hydration: Part II. Cement-Lime-Gypsum (ZKG), 2001,
initial period, and an extra reaction is detected in the 54, No. 2, 114–119.
dormant period when 5% accelerator is present. 7. Moeser B. and Stark J. A new model to OPC hydration
derived by means of ESEM-FEG. Proceedings of the 5th
(b) Instead of the small, prismatic ettringite crystals
International Symposium on Cement and Concrete, Shanghai,
found in normal cement paste, a hydrate layer forms 2002, 56–70.
that consists of foil-like calcium aluminate hydrate 8. Xu Q. and Stark J. Effect of an alkaline shotcrete accelerators
phases, gel-like silicate hydrates and calcium hydro- on the setting of cement. Proceedings of the 5th International
xide. It surrounds the cement particles in the paste Symposium on Cement and Concrete, Shanghai, 2002, 1038–
with 5% accelerator. 1042.
9. Goepfert T., Xu Q. and Stark J. Effect of an alkaline
(c) The aluminate, Al(OH)4 À , introduced by the accel- shotcrete accelerator on the early strength of cement mortar.
erator containing sodium aluminate reacts to form Proceedings of the 5th International Symposium on Cement
calcium aluminate hydrate phases, ettringite and and Concrete, Shanghai, 2002, 1032–1037.
monosulphate phases. The reactions of the aluminate 10. Locher F. and Richartz W. Setting of cement, IV: Effect of
are dependent on the availability of Ca(OH)2 . the solution compositions. Cement-Lime-Gypsum (ZKG), 1983,
4, 224–231
(d ) The set regulator in the cement readily reacts with 11. Bellmann F. Chemical Reactivity of Fly Ash and its Effect on
alkaline hydroxide to release SO4 2À ions and cause Sulphate Resistance of Concrete. Diploma paper from F. A.
the precipitation of Ca(OH)2 if the pH value in the Finger Institute for Building Materials Science, Bauhaus
liquid phase is raised. University Weimar, Germany, 2002.
12. Taylor H. F. W. Cement Chemistry, 2nd edn. Thomas Telford,
London, 1997.
References
1. Kusterle W. Shotcrete technology. Proceedings of the 6th
International Conference on Shotcrete, Innsbruck, 1999.
2. Ruffert R., Brux G. and Badzong H. Shotcrete. Expert Discussion contributions on this paper should reach the editor by
Press, Renningen-Malmsheim, 1995. 1 July 2005
8 Advances in Cement Research, 2005, 17, No. 1
Related docs
