Anda di halaman 1dari 11

~

CementandConcreteResearch,Vol.24,No.6,pp.1033-1043,1994
Pergamon

Copyright 1994 Elsevier Science L~d


Printed in the USA. All fights reserved
0008-8846/94$6.00+.00

0008-8846(94) 00046-8
FACTORS AFFECTING THE STRENGTH OF ALKALI-ACTIVATED SLAG
Shao-Dong Wang, Karen L. Scrivener and P.L. Pratt
Department of Materials, Imperial College of Science, Technology and Medicine,
University of London, SW7 2BP, UK
(Communicated by A.J. Majuradar)
(Received October 1, 1993)

ABSTRACT
The effect of several factors on the strength of alkali activated slags has
been investigated. The most important factors were found to be: the type
of alkaline activator, the means of adding activator, the dosage of alkali,
the type and fineness of slag, SiO2/Na20 ratio (modulus, Ms) when using
waterglass solution, curing temperature, liquid/slag or water/slag ratio and
additive. Some of these factors are interdependent and the effect of
changing more than one is usually not additive. The optimum range for
each factor is suggested through reviewing previous work and our recent
results of a full factorial range strength study. The interaction of factors is
considered and discussed throughout the paper, hoping to gain a better
understanding of the processing of alkali-activated slag (AAS) cement and
concrete.
INTRODUCTION
In practice it is most important to understand the relationship between the processing of a
material and its properties. Compared to OPC, the quantitative or even qualitative relations
between strength and processing parameters in alkali-activated slag (AAS) cements are much less
well understood, in spite of much work on the influence of various parameters (1.14), and more
detailed studies reported by Andersson et al (15) and Tailing (16). Moreover, there is great
interaction (interdependence) of factors affecting the strength in AAS systems. Most of the
available data from the literature relates to strength measurements of concretes under different
conditions and some reported results are contradictory. The study reported here was conducted
in order to gain a more fundamental understanding of the factors affecting the strength of alkali
activated slag cements. Mortars with a fixed sand to binder ratio were used to avoid the
influence of factors related to the concrete mix design.
EXPERIMENTAL
Six slags were used in the present study ground to various finenesses: M and C1 as examples of
basic slags, W and KP as examples of neutral slags, and C2 and C3 as examples of acid slags.
The oxide compositions of these slags are shown in Table 1. The alkaline activators used are
detailed in Table 2. Other parameters investigated were: the type of alkaline activator, the means
1033

1034

S.-D. Wanget al.

Vol.24. No. 6

of adding activator, the dosage of alkali, SiO2/Na20 ratio (modulus, Ms) when using waterglass
solution, curing temperature, liquid/slag or water/slag ratio and additive. The mortars were
machine mixed for 3 minutes and cast into moulds 222 cm 3. After vibration, the samples were
cured in the moulds for one day before demoulding and then cured in air at 20+2C and RH >
95% to the required ages. Each strength result is an average of six specimens in which the
highest and the lowest values are discounted in the average. All samples have a sand to binder
ratio of 2.
Table 1 Chemical Composition of Slags
Notation Feature
SiO 2 AI~O3 CaO MgO !Fe~O3 MnO TiO~ Total Kq ~ Kb ~
.Vl
basic
35.5
12.6 40.3 9.0
0.6
0.5
0.7 99.2 1.69 1.02
C1
basic
33.6 11.3 41.8 5.8
2.9
0.3
1.7 97.4 1.65 1.06
W
hish Kq
32.8 16.1 35.1 11.0 0.75 N/A N/A 95.8 1.90 0.96
KP
~hos-sla~ 37.6
9.4
45.2
1.6
1.6
N/A N/A 98.8 1.37 1.00
C2
acid
41.8
11.7 32.9 3.8
2.7
N/A
1.1 94.0 1.13 0.69
C3
acid
38.6 10.8 33.3 4.3
2.4
N/A 2.4 91.8 1.18 0.76
* Kq (quality coefficient) = CaO+MgO+AI203/SiO2+TiO2+(P2Os+F+MnO)
** Kb (basicity coefficient) = CaO+MgO/SiO2+AI203
Table 2 Alkali Solutions (grams ~er 100 ml solution)
Notation
Na20
SiO 2
Total Solid
pH
2M Na2SO 4
12.4
/
28.5
-7
2M Na2CO 3
12.4
/
21.2
12.6
4M N a O I t
12.4
/
16.3
14.6
2M Na2SiO 3
12.4
12
24.4
13.3
2M W G ( I ) '~
12.4
12
24.4
13.4
2M WG(1.5)
12.4
18
30.4
13.3
1.6MWG(I.5)
9.9
14.4
24.3
13.2
* 2M WG(1) means waterglass solution with the concentration of 2 moles and the modulus of 1;
other waterglass solutions used (not listed in the Table) were prepared by mixing above
waterglass solutions at different proportions as required.
RESULTS AND DISCUSSION
1. Type of Alkaline Activator
The present work confirms previous findings that the anionic component of the activator has far
more effect than the cationic component on strength and other properties. In spite of the fact that
NaOH, NaSO4 and Na2CO3 are often used as activators of slag and may be found effective, it
was found that, in agreement with previous results (1,5,7,11,15-18), waterglass solutions with moduli
of 1-1.5 always gave the best strengths, regardless of curing conditions, type or fineness of slag
(Figure 1).
The effect of activator type on strength changes with other factors. For example, with normal
curing up to 28 days large differences in strength were observed (as seen in Figure 1), whereas
with accelerated curing the differences were smaller os,16). Nevertheless, the use of waterglass
solution is always the most favourable from the strength point of view. However from an

Vol.24, No. 6

STRENGTH,ALKALI-ACTIVATEDSLAG,WATERGLASS,TEMPERATURE

1035

economic viewpoint, the use of alkali-containing by-products rather than commercial waterglass
may be desirable especially under conditions of accelerated curing.
Although the pH value of NaOH is higher than that of waterglass and their activation in terms of
the amount of slag reacted (reactivity) are the same ~1,5~,the resulting strengths of AAS mortars
are markedly different (see Figure 1). This can only be explained by the effect of the anion.
Na:~CO3 due to its lower pH value often shows lower activation ability at early ages with respect
to both strength and reactivity of slag. However, at late ages, NaECO3 yields higher strengths
than NaOH. This is also attributed to the effect of its anion: CO3 2-. The use of Na:~SO4 on its
own seems to be useless in practice due to its very low early strength. High 28-day strengths
have been reported with Na2SO4 (up to 20-40 MPa) when finely ground basic slag is used ~5~,but
the products are subject to damage by frost or carbonation. Application ofNa2SO 4 plus a small
amount of OPC, lime or clinker results in a dramatic improvement in strength, providing finely
ground basic GBFS is used. Fine grinding of slags which are basic or with higher Al203/SiO:t
ratio, addition of small amounts of other alkaline additives and steam curing are more favourable
to the activation of slag in the case of using Na2SO4 as the main activator. By the correct
combination of the above conditions, a concrete with strength up to 70 MPa has been made ~.
The products made by this technique are more economic and save energy, however their long
term durability has not been studied thoroughly.
100

98

9O
8O
m

70

O.

6o

~ 5o
~-

40

30
2o
10
0
2MNa2SO4

4MNaOH

2MNa2CO3

1.75MWG(1)

2MWG(1)

2MWG(1.5)

Activators

Figure 1 Compressive strength of M slag activated by different activators.


(fineness = 4100 cm2/g; alkali solution/slag = 0.47 cm3/g~ sand/slag = 2.0)
2. Method of Adding Alkaline Component
There are three methods of adding alkaline component to slag: in solution; in the solid state alkali and slag ground together; and in the solid state - alkali and slag ground separately then
mixed. Many attempts have been made to add alkaline component to slag in the solid state so as
to supply AAS cement ready-to-use. However, the present work confirms that far higher
strengths are obtained when waterglass is added in solution (120 MPa) than when it is added in
the solid state by either method (strengths of only 10-80 MPa). The addition of alkali in the solid
state not only results in much lower strengths but also produces much greater strength
fluctuation. By adding anhydrous waterglass solids (free from chemically bound water) with Ms
< 1.2, strengths up to 80 MPa may be obtained, however the development of strength is very

1036

S.-D. Wang et al.

Vol. 24, No. 6

sensitive to the physical and chemical changes of slag and alkali, making its use impractical due to
the problems with quality control of AAS cement production. For example, the waterglass solids
with Ms < 1.2 may absorb moisture when ground with slag, resulting in a dramatic decrease in
strength after storage. Another interesting example is that using hydrous waterglass/sodium
metasilicate (containing chemically bound water) in the solid form produces very low or even
zero strength under normal curing conditions, however when the same alkali is added in solution
the strength obtained is as good as that with ordinary waterglass solution at the same dosage and
modulus, displaying the great influence of alkali form on strength.
Some results for Na2CO3 activated slag concrete led to the conclusion that grinding Na2CO3
together with slag gives the highest strength ~5~. However, the authors' experiments showed that
addition in solution also gives the highest strength in this case. When steam/autoclave curing is
used, the variation in strength with method of adding alkali is somewhat reduced. Due to its high
solubility in water, any means of adding NaOH produces a good activation effect with respect to
the hydration degree of slag, however the strength differences, though smaller, still exist. That is,
adding alkali by solution is better than when it is ground together with slag which is better than
when slag and alkali are ground separately and then mixed. Furthermore, grinding NaOH
together with slag is impractical, as NaOH is highly hygroscopic causing reaction of slag before
USe.

In brief, addition of high Ms waterglass or other alkalis with low solubility in solid form results in
low strength because it is difficult to dissolve during the course of hydration. Although
waterglass with low modulus and NaOH are easier to dissolve, their addition in solid form also
results in low strength or poor reproducibility of strength because it is hygroscopic. Moreover,
solid waterglass with Ms < 1.2 is rarely available commercially at a reasonable price, as its
production greatly reduces the lifetime of refractory furnace linings. Further efforts aimed at
trying to add alkaline component in solid state are not recommended from strength and economic
viewpoints, unless special considerations of other aspects are taken into account.
3. Dosage of Alkali Component
Several reports on this aspect have been made (1,3,9,10,12,15,16), but are inconsistent. For example, in
the case of using NaOH, within the range of 1-10% (by slag weight) reported by Isozaki et al (3)
and within the range of 2-5% reported by Parameswaran et al 02), the higher the dosage, the
better the strength, whereas other results (10,1~)showed that the variation of NaOH dosage within
the ranges of 3-11% and 1-8% has little effect on strength. In fact, as regards the optimum
dosage of alkali, it is not only a question of strength but also of cost and other properties.
Usually 2-8% Na20 by slag weight is necessary for stable activation of slag It seems true, within
the practical range: 2-8% Na20 by slag weight, that the more Na20 is used, the higher the
strength. However, when Na20% reaches a certain value (depending on slag, activator and
curing condition), there is no further significant increase in strength at higher dosages and
detrimental properties such as efflorescence and brittleness are increased because of the effect of
more free alkali in the product. In this case, trying to increase the strength by increasing the
alkali dosage is not recommended from both economic and property points of view. Therefore,
the optimum dosage taking all factors into account is most likely to be within the range of 3.05.5% Na20 by slag weight.
4. Modulus of Waterglass Solution
As the formation of silica gel makes a significant contribution to strength, there is an obvious
interaction between modulus and Na20 content. That is, if the Na20 content is kept constant,
the alkaline activation effect can be considered the same, the higher the modulus the more the

Vol. 24, No. 6

STRENGTH, ALKALI-ACTIVATED SLAG, WATERGLASS, TEMPERATURE

1037

contribution from silica gel and the higher the strength within a certain range. However, if the
waterglass solid content is kept constant, the higher the modulus the lower the Na20 content and
the smaller the alkaline activation effect but the greater is the amount of silica gel. These
competing effects result in a variable optimum modulus, depending on the slag, its fineness and
curing condition. When alkaline activation is insufficient and thus becomes the main factor
slowing down the hydration of slag, a lower modulus is preferred, otherwise a higher modulus is
preferred. Nevertheless, the highest strengths of AAS cement/concrete were obtained with
waterglass solutions with moduli of 1-1.5 ~:5,17). Our current results indicate that, with normal
curing and using finely ground GBFS with Blaine fineness above 3700cm2/g, the optimum moduli
of waterglass solution are 0.75-1.25 for acid slag, 0.9-1.3 for neutral slag and 1.0-1.5 for basic
slag when waterglass solid content or the specific gravity of the solution is kept constant (see
Figure 2).
120
100

~.
~

$0

~o

}
g
~,1

211
0

0.5

0.75

1.25

1.5

2.5

Moduli of Waterglass Solution

Figure 2 Moduli of waterglass solution vs. 28-day strength for different types of slag.
(20 C; fineness = 4500300 cm2/g; alkali solution/slag = 0.41; sand/slag = 2.0)
5. Types of Slag
The hydraulic activity of slag can be measured by chemical moduli, such as a basicity coefficient
(CaO+MgO/SiO2+A1203) and a quality coefficient (CaO+MgO+AI203/SiO2+TiO2), although it
is found that the correlation between these moduli and the strength of blended cement is not
always good. Numerous papers have been published, trying to find a method to estimate the
hydraulic activity of slag which correlates better with the strength of blended cement ~c.g.~9-22). It is
well known that the hydraulic activity of slag is affected by the chemical composition of the main
chemical constituents, minor chemical constituents, glass content and their combination. The
quality control of slag by the current standards is effective in avoiding noticeable quality
fluctuation in blended slag cement production. However, in AAS cement since slag is no longer
an additive but a main ingredient, the quality of AAS cement is much more sensitive to the quality
of slag and therefore the methods of appraisal of slag quality are more important to AAS cement.
Figure 3 shows the effect of the slag type on the 28 day strength for different activators. It can be
seen that the type of slag has a significant influence on strength but more so in the systems with
weak alkaline activator than in those with strong alkali such as NaOH. The specifications for
GBFS in current use relate essentially to use with OPC rather than with AAS cement. It is

1038

S.-D. Wang et al.

Vol. 24, No. 6

therefore necessary to investigate new criteria more appropriate to this application. From our
current results, a few points can be suggested in this regard:
(a) slag with high alumina content results in high early strength, a greater amount of slag
reacted and often quick setting;
(b) the reactivities of slag activated by different activators are not particularly relevant to the
strength of AAS cement; however with the same activator, the reactivity of the slag may be
used as a parameter of hydraulic activity which correlates well with the strength of AAS
cement within a certain range of chemical compositions of slag (19);
(c) minor constituents in slag, such as P, F, S, Mn and Ti, often have significant influence on
slag quality;
(d) it seems that the best way to estimate the quality of slag for use in AAS cement is to
standardise a method directly measuring its strength under an accelerated curing regime with
a consideration of the fineness of slag and proper choice of activator.

120
~

100

80

~.

- --

O)
t-.

.~

6o

U)

>"
m
~o

40
20
0
2MNa2OS4

2MNa2CO3

2MWG(1)

4MNaOH

Figure 3 Type of activator versus 28-day strength for different types of slag (20 C;
fineness = 4500+300 cm2/g; alkali solution/slag = 0.41; sand/slag = 2.0).
6. Fineness of Slag
It is known that increasing the fineness of GBFS improves the strength and decreases the setting
time of slag mixes, regardless of activator type (15~. Slags with fineness of 3500 and 5300 cm2/g
(Blaine) activated by 6% waterglass with modulus of 0.9 have compressive strengths of 29.7 and
56.6 MPa at 91 days respectively (l~. However, Shi and Li ~6~reported that increasing the GBFS
fineness above 4000 cm2/g had no effect on compressive strength at ages later than 28 days,
although 3 and 7-day strengths were markedly increased up to about 6000 cm2/g. Tailing (16,17)
reported that the optimum slag fineness was 4000 cm2/g, with higher fineness increasing the
water demand and causing false setting. Wang (1) reported the optimum range of slag fineness as
4000-5500 cm:Z/g. Gjorv ~3) reported that, in F-cement (activated by F-activator which is
composed of NaOH, Na2CO 3 and lignosulphonate as the main ingredients together with sodium
gluconate and tributyl phosphate as additives), increasing the fineness of slag from 4200 to 6400
cm2/g significantly improves the strength regardless of the age and curing temperature but more
so at early ages and low temperatures than at later ages and higher temperatures. Other reports
dealing with different cases indicated that the strength increases with the increase of the fineness
of slag up to 6000 cm2/g (3,12) and even to 10000 cm2/g (14). In fact, the optimum range of slag
fineness should be determined from a comprehensive understanding of properties and economics,

V.~l. 24, No. 6

STRENGTH, ALKALI-ACTIVATEDSLAG, WATERGLASS,TEMPERATURE

1039

not only strength. The optimum fineness also depends on slag, activators and curing conditions.
More generally, from our results (see Figure 4) and previous reports, the following points need to
be considered when choosing the optimum fineness of slag:
(a) it is true that the higher the fineness the higher the strength within an extended range of
fineness when keeping water/solid ratios the same without taking workability into account.
However, a certain workability is necessary for the shaping of products; if the workability
and alkali amount are kept constant, an optimum range of fineness exists (see Figure 4); this
is, 4500-6500 cmZ/g for acid and neutral slags, and 4000-5500 cm2/g for basic slags;
(b) in comparison with slag used in OPC, the slags used in AAS cement are generally required
to be finer. From a comprehensive consideration of economics, effectiveness of mill
grinding, strength and other properties, the suggested values of slag fineness in AAS cement
are 4000-5500 cm2/g, whereas those in OPC blends are probably 3500-4500 cm2/g;
(c) as will be discussed later, the effectiveness of increasing the fineness will be reduced by
using accelerated curing or highly reactive slag;
(d) the strength increment with increasing fineness is more pronounced at early ages (<14 days)
rather than at later ages;
(e) increasing the fineness of slag within its optimum range results in not only somewhat higher
strength but, more importantly, better reproducibility of strength.

,20!
.. ,00
80

~
~'

60
40

--O---acid slag

20 J/

- - 0 - - - basic slag I

3000

3500

4000

4500

5000

5500

6000

7000

Fineness cm2/g
Figure 4. Fineness vs. 28-day strength for different types of slag (waterglass activator,

Na20 = 5.5%; Ms = 1; sand/slag = 2; slag fineness variation = +300 cm2/g).


7. Curing Temperature
AAS cement shows an increase in compressive strength with the increase of temperature up to
150oC (23). For F-cement (13), it was found that increasing curing temperature from 20 to 60 C
and curing time up to 48 hours caused a continuous increase in compressive strength. In AAS
cement based on waterglass, increasing temperature up to 50 C and curing time up to 7 hours by
Tailing (16) and 80 C for 16 hours by Andersson et a105) yielded noticeable increases in strength.
In an AAS cement based on Na2CO3, increasing the curing time from 6 to 16 hours at 80 C has
been reported to have no definite effect on strength (15~. As regards to AAS cement using NaOH
as an activator, it was found that steam curing caused some decreases in both compressive and
flexural strengths compared with water and air curing (15), whereas others (5) reported that steam
curing of phosphorous slag plus NaOH at 95 C for 8 hours resulted in higher strengths. In brief,
although accelerated curing is more favourable to AAS cement than OPC, the quantitative

1040

S.-D. Wang et al.

Vol. 24, No. 6

relationship between optimum curing regime (temperature and time) and other factors such as the
types and fineness of slag, types and dosage of alkali has not been well understood. Due to the
fact that the apparent activation energies of most alkali-slag systems are higher than those of
OPC, the hydration of slag in AAS cement systems is more sensitive to curing temperature ~.
The hydrate products of AAS cement are generally amorphous, but higher temperature leads to
the formation of some crystalline hydrate products. In comparison with normal curing and OPC,
the following points regarding accelerated curing of AAS cement can be concluded from our
recent studies (see Figure 5).
(a) The effectiveness of accelerated curing is more pronounced with acid and neutral slags or
weaker alkaline activators (Figure 5 compared with Figure 3). For a further example, an
AAS cement based on phosphorous slag (neutral or acid in nature) and waterglass having a
strength of 30 MPa under normal curing can reach 62 MPa after steam curing and 71-76
MPa after autoclave curing, whereas an AAS cement based on GBFS (basic in nature) and
waterglass solution whose strength is 80 MPa under normal curing can only goes up to 8595 MPa after steam or autoclave curing;
(b) The effectiveness of accelerated curing is reduced by using finely ground slag or a strong
activator such as NaOH.
(c) Accelerated curing can greatly improve the durability, quality fluctuation and the occurrence
of efflorescence of the products.
(d) Therefore, using low quality slag and weak alkali or alkaline by-products or alkalicontaining waste becomes more feasible in practice by introducing accelerated curing.

120

117

~=

100
(~

~.
~:

80

~ ~o
g

4~

~o
o
2MNa2OS4

2MNa2CO3

2MWG(1)

4MNaOH

Figure 5 Type of activator vs. type of slag after curing at 80 C for 16 hours.
(Conditions: as Figure 3).
8. Water/Slag (w/s) and Liquid/Slag (Us) Ratios
When alkali is used in solid form the water/slag ratio must be considered. As in OPC, the higher
the w/s ratio, the lower the strength, providing it is above a certain value to ensure the fluidity
required for shaping. The water demand of AAS cement to give a standard consistency is about
17-22%, as compared to 25-29% for OPC. When alkali is used in solution it is more convenient
to consider the liquid/slag ratio rather than the dosage of alkali solid and water/slag ratio.
Increasing the l/s ratio results in an increase of alkaline activation due to the introduction of more
alkaline component and a simultaneous increase of pore volume in hardened concrete due to

Vol. 24, No. 6

STRENGTH, ALKALI-ACTIVATEDSLAG,WATERGLASS,TEMPERATURE

1041

introducing more water. Therefore, the effects of I/s ratio on strength are positive at lower I/s
ratio where alkaline activation is a dominant factor, nil within a certain range when positive and
negative effects are basically balanced and then negative at higher l/s ratio where the effect of
increasing pore volume is a key factor. For example, using 2M waterglass solution with modulus
of 1 and a sand/slag ratio of 2, the strength of AAS cement mortar increases with the increase of
l/s ratio up to 0.38, is fairly constant over the range 0.38-0.45, and then decreases with increasing
l/s ratio. When reducing the concentration or increasing the moduli of waterglass solution, similar
conclusions can be obtained and explained by the competing effects of l/s ratio on strength.
9. Additives
To improve some special properties of AAS cement, different additives have been
studied (1,5,9,12,17,18,24,25). Typical additives can contain any one or more of OPC clinker, lime/lime
slurry, fly ash (below 10% by slag weight), silica fume, CaF2, limestone, sodium lignosulphonate.
Douglas and Brandstetr (9) have carried out a comprehensive investigation using OPC, silica fume,
fly ash and lime. Partial substitution of fly ash for GBFS (below 10% by slag weight) may result
in an increase in strength (9,17), however substitution of fly ash for GBFS in higher proportions
reduces the strength noticeably (1,12). Fly ash may also improve the workability of a fresh mix of
AAS cement (17), however Parameswaran et al (12) reported that an Indian fly ash (characterised by
being less reactive and more angular in shape) did not improve the workability even at 40/'0 by
weight of total binder, implying that the characteristics of fly ash are important.
Lime or lime slurry has been used to improve the early strength (5,17,24) and to control the setting
time of AAS cement using sodium silicate with Ms=2.85 (18). Using a slag characterised by
relatively low CaO (31.7%), high MgO (12.4%) and a waterglass solution with Ms=l.5, Cheng et
al (24) reported that both the early and 28-day strengths increased dramatically by adding 1.9-3.4%
Ca(OH)2. It was also found that the addition of water reducer such as sodium lignosulphonate
(an ingredient of F-activator) and sulphonate naphthalene-based superplasticizer caused a
reduction of compressive strength without any improvement of workability (9).
OPC clinker is considered to have similar effects to lime/lime slurry (25). It is worth mentioning
that using OPC clinker as an additive is quite different from using OPC; the latter contains
gypsum, which neutralises the alkaline component and produces expansive non-binding mirabilite
(Na2SO4-10H20) and thus is harmful to strength.
Our current results regarding the effects of various additives on strength are summarised as
follows:
(a) lime, lime slurry and clinker can be used to improve the early (less than 7 day) strength,
however the later strength may be slightly reduced in many cases;
(b) when using Na2SO4 as the main activator, addition of 5-10% lime or clinker results in a
dramatic increase in strength;
(c) addition of fly ash (below 10%) has little effect on improving the workability of the mix, and
often causes slight reduction of strength under normal curing;
(d) addition of CaF:~ (0.5-3.0%) may slightly increase the early (less than 7 day) strength, but
has no effect on later strength in most cases;
(e) addition of CaCO3 (calcite fine powder, 1-3%) may also slightly increase the early (less than
3 day) strength, but has no or a slightly negative effect on later strength, regardless of the
type of activator such as Na2CO3;
(f) due to the surface-active effect of the alkaline cation, most water reducing agents used in
OPC are much less effective in improving the fluidity of AAS mixes;
(g) additions of other less reactive minerals such as high-lime fly ash, non-ferrous slag, steel

1042

S.-D. Wang et al.

Vol. 24, No. 6

slag, clay, phosphorous slag in a small quantity (below 10%) may have slightly positive
effect on strength; however most attempts to increase the amount of replacement materials
for GBFS have resulted in significant decreases of strength except in a case using quickly
cooled converter steel slag up to 30% by weight reported by WangO).
CONCLUSIONS
On the basis of the results obtained in this research and comparing with previous results, the
optimum range for each factor affecting the strength of AAS cement/concrete is summarised as
follows:
1. waterglass is the best activator in all cases;
2. adding alkaline component in solution form is the best method in all cases;
3. the best alkali dosage is within the range of 3.0-5.5% Na20 by slag weight;
4. the optimum ranges ofmoduli of waterglass solution are 0.75-1.25 for acid slag, 0.90-1.3 for
neutral slag and 1.0-1.5 for basic slag;
5. the optimum range of slag fineness is 4000-5500 cm2/g;
6. the optimum range of solution (volume) to slag (weight) ratio is within 0.38-0.45.
Acknowledgement

SDW is grateful to the Institution of Mining and Metallurgy (IMM), UK for the award of a
Stanley Elmore Fellowship, to the Committee of Vice-Chancellors and Principals (CVCP) and
Imperial College, UK for the award of an Overseas Research Studentship.
References

1. Wang S.D., Review of recent research on alkafi-activated concrete in China. Mag. Cone.
Res., 43(154), 29-35(1991).
2. Deja J. and Malolepszy J., Resistance of alkali-activated slag mortars to chloride solution.
3rd Int. Conf on Fly-ash, Silica Fume, Slag and Natural Pozzolans in Concrete, Trondheim,
Norway, SP114-75, Vol. 2, pp1547-1563 (1989).
3. Isozaki K., Iwamoto S. and Nakagawa K., Some properties of alkali-activated slag cement.
8th Int. Congr. Chem. Cem., Rio de Janeiro, Vol.6, pp395-399 (1986).
4. Kukko H. and Mannonen R., Chemical and mechanical properties of alkali-activated blast
furnace slag (F-concrete). Nordic Concrete Research, OSLO, 1982, No. 1, ppl6.1-16.16.
5. Pu X.C., Gan C.C., Wang S.D. and Yang C.H., Research reports of on alkali-activatedslag
cement and concrete, 6 vols in Chinese. Chongqing Institute of Architecture and Engineering,
China, (1988).
6. Shi C. and Li Y., Investigation on some factors affecting the characteristics of alkaliphosphorous slag cement. Cem Cone. Res., 19(4), 527-533(1989).
7. Glukhovsky V.D., Rostovskaja G.S. and Rumyna G.V., High strength slag-alkaline cements.
7th Int. Congr. Chem. Cem., Paris, 1980, Vol.3, ppV164oV168.
8. Voinovitch I., Raverdy M. and Dron R., Slag cement without clinker. 7th Int. Congr. Chem.
Cem., Paris, 1980, Vol.2, pplii122-III127.
9. Douglas E. and BrandstetrJ., Preliminary study on the alkali activation of ground
granulated blast-furnace slag. Cem. Cone. Res., 20(5), 746-756(1990).
10. Metso J. and Kajaus E., Activation of blast furnace slag by some inorganic materials. 1st
Int. Conf. on Fly-ash, Silica Fume, Slag and Natural Pozzolans in Concrete, Montbello,
Canada, 1983, Vol.2, pp1059-1073.

Vol.24, No. 6

STRENGTH,ALKALI-ACTIVATEDSLAG,WATERGLASS,TEMPERATURE

1043

11. Malolepszy J. and Petri M., High strength slag-alkaline binder. 8th Int. Congr. Chem. Cem.,
Rio de Janeiro, 1986, Vol.4, ppl08-111.
12. Parameswaran P.S. and Chatterjee A.K, Alkali activation of lndian blast furnace slags. 8th
Int. Congr. Chem. Cem., Rio de Janeiro, 1986, Vol.4, pp86-91.
13. Gjorv O.E., Alkali activation of a Norwegian granulated blast furnace slag. 3rd Inter. Conf.
on Fly-ash, Silica Fume, Slag and Natural Pozzolans in Concrete, Norway, 1989, Vol.2,
SP114-73, pp1501-1517.
14. Bijen J. and Waltje H., Alkali activated slag-fly ash cement. 3rd Inter. Conf. on Fly-ash,
Silica Fume, Slag and Natural Pozzolans in Concrete, Norway, 1989, Vol.2, SPl14-76,
pp1565-1578.
15. Andersson R., Gram H-E, Malolepszy J. and Deja J., Alkali-activated slag. Swedish Cement
and Concrete Research Institute, Stockholm, 1988, 104 ps.
16. Talling B., Effect of curing conditions on alkali-activated slags. 3rd Inter. Con on Fly-ash,
Silica Fume, Slag and Natural Pozzolans in Concrete, Norway, 1989, Vol.2, SPl14-72,
pp1485-1500.
17. Talling B. and Brandstetr J., Present state and future of alkali-activated slag concretes, 3rd
Inter. Conf. on Fly-ash, Silica Fume, Slag and Natural Pozzolans in Concrete, Norway, 1989,
Vol.2, SP114-74, pp1519-1546.
18. Douglas E., Bilodeau A., Brandstetr J. and Malhotra V.M., Alkali activated ground
granulated blast furnace slag concrete: preliminary investigation. Cem. Concr. Res., 21(1),
101-108(1991).
19. Wang S.D., A new method to estimate the hydraulic activity of slag: the reactivity method
Journal of The Chinese Ceramic Society, 20(2), 196-200(1992).
20. Takemoyo K., The variation of activity index of slag. 8th Int. Congr. Chem. Cem., Rio de
Janeiro, 1986, Vol.6, pp204-206.
21. Daugherty K.E., The glass content of slag and hydraulic activity. Silicates Industriels,
45(4/5), 107-110(1983).
22. Smolczyk H.G., Slag structure and identification of slags. 7th Int. Congr. Chem. Cem.,
Paris, 1980, Vol. 1, pplII1/3-III1/7.
23. Shi C., Wu X. and Tang M., Hydration of alkali-slag cements at 150 C. Cem. Concr. Res.,
21(~1), 91-100(1991).
24. Cheng Q.H., Tagnit-Hamou A. and Sarkar S.L., Strength and microstructural properties of
waterglass activated slag. Mat. Res. Soc. Symp. Proc., 245, 49-54(1992).
25. Glukhovsky V.D., Krivenko P.V., Rostovskaya G.S., Timkovich V.J. and Pankratov V.L.,
United States Patent No 4410365, 18 Oct 1983.

Anda mungkin juga menyukai