Cement and Concrete Research 30 (2000) 1031 ± 1036

Efficiency of GGBS in concrete

Abstract

The utilisation of supplementary cementitious materials is well accepted because of the several improvements possible in the concrete
composites and due to the overall economy. The present paper is an effort to quantify the 28-day cementitious efficiency of ground granulated
blast furnace slag (GGBS) in concrete at the various replacement levels. It was observed that this overall strength efficiency of GGBS
concretes can also be defined through a procedure adopted earlier for other cementitious materials like fly ash and silica fume. The overall
strength efficiency was found to be a combination of general efficiency factor, depending on the age and a percentage efficiency factor,
depending upon the percentage of replacement as was the case with a few other cementitious materials like fly ash and silica fume reported
earlier. This evaluation makes it possible to design GGBS concretes for a desired strength at any given percentage of replacement. D 2000
Elsevier Science Ltd. All rights reserved.

Keywords: Concrete; Mixture proportioning; Granulated blast furnace slag; Compressive strength; Efficiency

1. Introduction It was seen that among these, the reactive glass content and
fineness of GGBS alone will influence the cementitious/
Blast furnace slag cements are in use for a reasonably pozzolanic efficiency or its reactivity in concrete composites
long period due to the overall economy in their production significantly. Some of the earlier researchers tried to express
as well as their improved performance characteristics in this reactivity of GGBS in terms of slag activity index (SAI)
aggressive environments. Also, the use of pozzolans as or hydraulic index, considering its chemical composition.
additives to cement, and more recently to concrete, is well
accepted in practice. Ground granulated blast furnace slag
(GGBS) is one such pozzolanic material (termed by a few as 2. Slag activity index
a supplementary or complimentary cementitious material)
which can be used as a cementitious ingredient in either ASTM C989 defines SAI as the percentage ratio of the
cement or concrete composites. Research work to date average compressive strength of slag cement (50 ± 50%)
suggests that these supplementary cementitious materials mortar cubes to the average compressive strength of refer-
improve many of the performance characteristics of the ence cement mortar cubes at a designated age [2]. Based on
concrete, such as strength, workability, permeability, dur- this slag was classified into three grades Ð Grade 80, Grade
ability and corrosion resistance. To assess the effectiveness 100, and Grade 120, depending on the relative compressive
of GGBS in cementitious composites, some of the para- strength. Hooton and Emery [3] observed that the properties
meters like chemical composition, hydraulic reactivity, and of GGBS influencing its reactivity to be the glass content,
fineness have been carefully examined by many earlier. The chemical composition, mineralogical composition, fineness
details of the various chemical constituents and their effects of grinding and type of activation provided. Researchers
have been summarised by the authors in an earlier paper [1]. have suggested different compositional moduli to assess the
reactivity of GGBS. However, Mantel [4] came to conclu-
sion that hydraulic formulae for GGBS proposed in the
* Corresponding author. Tel.: +91-44-445-8623; fax: +91-44-235-0509
or 2545.
literature do not adequately predict the strength performance
E-mail addresses: kgbabu@acer.iitm.ernet.in, kgbabu99@hotmail.com of slag. He stated that there is no correlation between the
(K. Ganesh Babu). chemical composition of a cement or that of a slag and the

0008-8846/00/$ ± see front matter D 2000 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 0 8 - 8 8 4 6 ( 0 0 ) 0 0 2 7 1 - 4
1032 K. Ganesh Babu, V. Sree Rama Kumar / Cement and Concrete Research 30 (2000) 1031±1036

hydraulic activity of a blend made from that cement and there is a maximum percentage for obtaining an equivalent
slag. He also reported that the slag activity, tested as per strength (equivalent to the normal mortar at that age) and
ASTM, depends on the particle size distribution (fineness) also, that there is a specific percentage of GGBS at which
of slag and the cement used and showed that this ranges the maximum strength can be obtained at that age. From
from 62% to 115% at 28 days. He observed that cement this, it can be said that the compressive strength of GGBS
with high alkali content has not effected the hydraulicity of concretes depend both on percentage replacement level and
the slag. In contrast, Hogan and Rose [5] have said that high on the age. This observation is similar to that seen in the
alkali cement blends yield an appreciably greater SAI value case of fly ash and silica fume concretes studied earlier in
than the low alkali cement blends. It is to be noted that all this laboratory [8 ±10]. It is felt that the efficiency concept,
the above tests on SAI were conducted on mortar cubes proposed earlier for fly ash and silica fume, can also be used
only. Although it is well known that the behavior of mortar for understanding the behavior of GGBS in concrete.
is different from that of concrete and, in particular, the
reactivity of GGBS in mortar cannot directly be correlated
to its performance in concrete, concrete mix proportioning 3. Evaluation of efficiency
based on the reactivity of slag is not looked into by many.
The above discussion shows that there is a need to look at This paper attempts to assess the cementitious effi-
the possibility of proportioning mixes based on the reactiv- ciency of GGBS in concrete at the various replacement
ity of GGBS in concrete. percentages through the efficiency concept by establishing
There have been a few attempts of this nature reported in the variation of the strength to water-to-cementitious ma-
literature. Swamy and Bouikni [6] reported that by a proper terials ratio relations of the GGBS concretes from the
mix proportioning GGBS concretes can be produced with normal concretes at 28 days. In principle, this was done
strengths comparable to those with ordinary Portland ce- by using Dw concept, which attempts to bring the water-
ment from the 3rd day onwards. He also suggested that the to-cementitious material ratio [w/(c + g)] nearer to the
total cementitious material has to be increased by 10% for water-to-cement ratio of the control concrete (w/co) by
50% replacement of GGBS and by 20% for 65% replace- applying the cementitious efficiency factor k of GGBS at
ment to attain strengths comparable to normal concretes. For any particular strength. However, the first trails to bring
a general understanding of the reactivity of GGBS, the the water-to-cementitious materials ratio to strength rela-
compressive strength results of Hwang and Lin [7] on tions through a single efficiency value (general efficiency
GGBS mortars at different ages and at the various replace- factor ke), at all replacement percentages, did not lead to a
ment levels have been replotted (Fig. 1). This shows that good correlation. The remaining difference at this stage
was corrected through the ``percentage efficiency factor
(kp).'' The ``overall cementitious efficiency factor (k)'' is
the sum of the general efficiency factor ke and the
percentage efficiency factor kp.
A detailed presentation of this method for evolution of
efficiency of mineral admixtures in concrete was discussed
in detail earlier [8 ±10]. Thus, in this method, the water-to-
cement ratio (w/co) to strength relation of normal concrete
will also be valid for the GGBS concretes, by considering
the ``water-to-effective cementitious materials ratio.''
…w=co † ˆ ‰w=…c ‡ kg†Š ˆ ‰w=…c ‡ ke g ‡ kp g†Š
where k = ke + kp.
For this evaluation of the efficiency, the data available
from research efforts in the recent past [11 ± 21] were
collected and summarised in Table 1. It is to be noted that
this was to ensure that the results of these investigations are
representative of the cements and slags manufactured pre-
sently. It was made sure that these will form a fairly
representative group governing all the major parameters
that influence the behavior of GGBS in concrete and present
the complete information required for such an evaluation.
During the evaluation, it was seen that some of these mixes
do not form a part of normal concretes, due to variations
resulting from air entrainment, different curing conditions
Fig. 1. Effect of slag content on strength development in cement mortars [7]. and high fineness of slag, etc. and these were not considered
 K. Ganesh Babu, V. Sree Rama Kumar / Cement and Concrete Research 30 (2000) 1031±1036 1033

Table 1
Details of the concretes evaluated
Percentage Slump range 28-Day compressive Average efficiency
Slag no. replacement w/(c + g) Range (mm) strength range (MPa) @ 28 days Reference
1 0 0.23 ± 0.83 40 ± 170 19.7 ± 106 ± [6,12,14 ± 18,20,21]
2 10 0.26 ± 0.38 150 58.5 ± 105 1.29 [21]
3 30 0.26 ± 0.55 100 ± 150 49.1 ± 105 1.02 [14,16,21]
4 50 0.30 ± 0.80 35 ± 190 21.2 ± 89.3 0.84 [6,12,14 ± 16,18]
5 60 0.26 ± 0.50 150 43.5 ± 80 0.78 [19,21]
6 65 0.46 ± 0.75 100 23.0 ± 57.5 0.75 [6,17]
7 70 0.41 ± 0.61 45 ± 65 32.5 ± 62.5 0.73 [12,14,19]
8 80 0.5 ± 29.5 ± 32.5 0.70 [19]

for evaluation. Finally, out of the 175 mixes only about 70 of 15-cm size through accepted guidelines [22]. In most
concretes made with ordinary Portland cement confirming cases, the change was the variation from cylinder to cube
to ASTM type I cement (with fineness in the range of 283 ± strength, and was corrected by using a single factor of 0.9
391 m2/kg) and cured under normal conditions were eval- for concretes in the strength range of 55± 70 MPa. The
uated. The GGBS in these concretes confirm to the mini- water-to-cementitious materials ratio [w/(c + g)] to com-
mum characteristics specified by ASTM C989 [2] for use as pressive strength relations at different percentages of repla-
mineral admixture in concrete (with fineness ranging from cement were plotted for all the concretes at 28 days (Fig. 2).
350 to 465 m2/kg, SiO2 from 31.1% to 38.59% and CaO It can be seen from this that the 28-day compressive
from 32.8% to 43.9%). The replacement percentages range strengths of concretes containing GGBS up to 30% replace-
from 10% to 80%. Natural river sand was used as fine ment were all slightly above that of normal concretes and at
aggregate and the maximum size of the coarse aggregate all the other percentages the relationships were below that of
ranges from 10 to 20 mm. It is obvious that with replace- normal concretes. It was also observed that the variations
ment levels as high as 80%, some of these concretes had due to the different percentages of slag replacement were
superplasticizers for improving the workability, because of smaller than the corresponding variations in the case of fly
the high fines content in these concretes. In view of this, ash [9]. In order to bring the strength values at all replace-
only concretes up to a maximum superplasticizer percentage ment levels nearer to that of normal concrete, the water-to-
of 2% were considered for evaluation. Also, different cementitious material ratios were modified by applying the
researchers used specimens of different sizes and shapes ``general efficiency factor (ke),'' replacing the [w/(c + g)]
and these have been corrected to their equivalent for a cube with [w/(c + keg)]. After several trials with ke values varying

Fig. 2. Compressive strength variation with [w/(c + g)] at 28 days.

Fig. 3. Compressive strength variation with [w/(c + keg)] at 28 days.

from 0.85 to 1.0, the appropriate ke value was found to be efficiency, which can bring the GGBS concrete strength
0.9 for the 28-day strength of these concretes. As already values closer to that of normal concrete was found by
mentioned earlier, it was observed that this general effi- evaluating the remaining difference through a ``percentage
ciency factor ke could not bring the [w/(c + keg)] to strength efficiency factor (kp).'' This value was observed to be
relations very close to the water-to-cement ratio of normal varying between +0.39 and ÿ0.20 for replacement levels
concrete (w/co) at all percentage replacement levels (Fig. 3). between 10% and 80%. This results in an ``overall effi-
At this stage, the effect of percentage replacement on ciency factor (k),'' the sum of the ``general efficiency factor

Fig. 4. Compressive strength variation with [w/(c + kg)] at 28 days.

Fig. 5. Assessment of the reliability of the efficiencies evaluated.

(ke)'' and the ``percentage efficiency factor (kp),'' varying also found to be appropriate for the evaluation of GGBS.
between 1.29 and 0.70 for the percentage replacements This method recognises that the ``overall strength efficiency
varying from 10% to 80%. The typical variations of factor (k)'' of the pozzolan is a combination of the two
strengths with [w/(c + keg + kpg)] at 28 days were presented factorsÐthe ``general efficiency factor (ke)'' and the ``per-
(Fig. 4). This shows that by adopting the two efficiency centage efficiency factor (kp).''
factors ke and kp, the strength of GGBS concretes at (2) The evaluations have shown that at 28 days, the ``overall
different percentages could be brought close to that of strength efficiency factor (k)'' varied from 1.29 to 0.70 for
normal concrete. Fig. 5 shows best fit of the corrected percentage replacement levels varying from 10% to 80%.
water-to-cementitious materials ratio to strength relations (3) It was also seen that the ``over all strength efficiency
of the GGBS concretes in comparison to that of the normal factor (k)'' was an algebraic sum of a constant ``general
concretes. The regression coefficient for GGBS concretes as efficiency factor (ke),'' with a value of 0.9 at 28 days, and a
well as normal concretes was found to be 0.94 at 28 days. ``percentage efficiency factor (kp),'' varying from +0.39 to
The above evaluation also showed that the slag concretes ÿ0.20, for the cement replacement levels varying from 10%
based on this overall efficiency factor (k), will need an to 80% studied.
increase of 8.6% for 50% replacement and 19.5% for 65% (4) Overall, the prediction of the strength of concretes
replacement in the total cementitious materials for achieving varying from 20 to 100 MPa with GGBS levels varying
strength equivalent to that of normal concrete at 28 days. from 10% to 80% by this method was found to result in a
This agrees well with the observations of Swamy and regression coefficient of 0.94, which was also the same for
Bouikni [6] reported earlier. normal concretes.
(5) Finally, it was observed that for obtaining equal
strength in concretes at 28 days, by adopting the efficiencies
4. Conclusions evaluated in the present investigation, it will be required to
have an additional 8.5% and 19.5% increase in the total
This study was primarily concerned with the evaluation cementitious materials at 50% and 65% cement replacement
of the efficiency of GGBS in concretes containing normal levels, agreeing well with the values 10% and 20% addi-
Portland cements from the results of the investigations tional material reported earlier.
reported in recent years. The replacement levels in the
concrete studied varied from 10% to 80% and the strength
References
efficiencies at the 28 days were calculated. The primary
conclusions can be listed as follows. [1] K. Ganesh Babu, V. Sree Rama Kumar, Performance of GGBS in
(1) The earlier proposed method for evaluating the cementitious composites. Sixth NCB International Seminar on Ce-
efficiency of pozzolans like fly ash and silica fume was ment and Building Materials, New Delhi, India (1998) XIII-76.
1036 K. Ganesh Babu, V. Sree Rama Kumar / Cement and Concrete Research 30 (2000) 1031±1036

[2] ASTM C989-94a, Standard specification for ground granulated blast- [13] N. Nakamura, M. Sakai, K. Koibuchi, Y. Iijima, Properties of high
furnace slag for use in concrete and mortars, 04.02. strength concretes incorporating very finely ground granulated blast
[3] R.D. Hooton, J.J. Emery, Glass content determination and strength furnace slag, ACI SP 91, Detroit (1986) 1361 ± 1379.
development predictions for vitrified blast furnace slag, ACI SP 79, [14] J.J. Brooks, P.J. Wainwright, M. Boukendakji, Influence of slag type
Detroit (1983) 943 ± 962. and replacement level on strength, elasticity, shrinkage and creep of
[4] D.G. Mantel, Investigation into the hydraulic activity of five granu- concrete, ACI SP 132, Detroit (1992) 1325 ± 1342.
lated blast furnace slags with eight different Portland cements, ACI [15] N. Nakamura, M. Sakai, R.N. Swamy, Effect of slag fineness on the
Mater J 91 (1994) 471 ± 477. development of concrete strength and microstructure, ACI SP 132,
[5] F.J. Hogan, J.H. Rose, ASTM specification for ground iron blast Detroit (1992) 1343 ± 1365.
furnace slag: Its development use and future, ACI SP 91, Detroit [16] K. Fukudome, K. Miyano, H. Taniguchi, T. Kita, Resistance to freez-
(1986) 1551 ± 1576. ing and thawing and chloride diffusion of anti-washout underwater
[6] R.N. Swamy, A. Bouikni, Some engineering properties of slag con- concrete containing blast furnace slag, ACI SP 132, Detroit (1992)
crete as influenced by mix proportioning and curing, ACI Mater J 87 1565 ± 1582.
(1990) 210 ± 220. [17] R.N. Swamy, I.C. Laiw, Effectiveness of supplementary cementing
[7] C.L. Hwang, C.Y. Lin, Strength development of blended blast furnace materials in controlling chloride penetration into concrete, ACI SP
slag cement mortars, ACI SP 91, Detroit (1986) 1323 ± 1340. 153, Detroit (1995) 657 ± 674.
[8] K. Ganesh Babu, G.S.N. Rao, P.V.S. Prakash, Efficiency of pozzolans [18] R.M. James, G.A. Mark, Marine exposure of concrete under se-
in cement composites, Concrete 2000 Dundee (1993) 497 ± 509. lected South African conditions, ACI SP 163, Farmington Hills, pp.
[9] K. Ganesh Babu, G.S.N. Rao, Efficiency of fly ash in concrete, Cem 201 ± 214.
Concr Compos 15 (1993) 223 ± 229. [19] G.J. Osborne, Effectiveness of blast furnace and glassifies slags at
[10] K. Ganesh Babu, P.V. Surya Prakash, Efficiency of silica fume in reducing ingress of chloride ions into Portland cement concrete in
concretes, Cem Concr Res 25 (1995) 1273 ± 1283. marine environments, Concr Mar Struct (1986) 99 ± 119.
[11] J.W. Meusel, J.H. Rose, Production of granulated blast furnace [20] P.S. Mangat, B.T. Molloy, Chloride binding in concrete containing
slag at Sparrows point and the workability and strength potential PFA, GGBS or silica fume under sea water exposure, Mag Concr
of concrete incorporating the slag, ACI SP 79, Detroit (1983) Res 47 (171) (1995) 129 ± 141.
867 ± 890. [21] P.C. Domone, M.N. Soutsos, Properties of high strength concrete
[12] P.J. Wainwright, J.J.K. Tolloczko, Early and later age properties of mixes containing PFA and GGBS, Mag Concr Res 47 (173) (1995)
temperature cycled slag-OPC concretes, ACI SP 91, Detroit (1986) 355 ± 367.
1293 ± 1321. [22] A.M. Neville, Properties of Concrete, Longman, Singapore, 1988.