Professional Documents
Culture Documents
Models Relating Mixture Composition To The Density and Strength of Foam Concrete Using Response Surface Methodology - Nambiar & Ramamurthy 2006 PDF
Models Relating Mixture Composition To The Density and Strength of Foam Concrete Using Response Surface Methodology - Nambiar & Ramamurthy 2006 PDF
www.elsevier.com/locate/cemconcomp
Received 21 October 2005; received in revised form 11 June 2006; accepted 14 June 2006
Abstract
There have been several investigations in the past on the influence of mixture composition on the properties of foam concrete. Con-
ventionally strength is related to density alone and few models have been developed relating strength with density, porosity, gel–space
ratio, etc. Very little work has been reported in the literature in predicting the properties of foam concrete from the knowledge of its
mixture proportions. This paper discusses the development of empirical models for compressive strength and density of foam concrete
through statistically designed experiments. The response surface plots helps in visually analysing the influence of factors on the responses.
The relative influence of fly ash replacement on strength and density of foam concrete is studied by comparing it with mixes without fly
ash and brought out that replacement of fine aggregate with fly ash will help in increase in the strength of foam concrete at lower densities
allowing high strength to density ratio. Confirmatory tests have shown that the relation developed by statistical treatment of experimen-
tal results can act as a guideline in the mixture proportion of foam concrete.
Ó 2006 Elsevier Ltd. All rights reserved.
Keywords: Foam concrete; Response surface methodology (RSM); Fly ash replacement level; Filler–cement ratio; Strength–density ratio
0958-9465/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.cemconcomp.2006.06.001
E.K. Kunhanandan Nambiar, K. Ramamurthy / Cement & Concrete Composites 28 (2006) 752–760 753
Table 1 Table 2
Physical and chemical properties of cement and fly ash used Factors and factor levels adopted
Properties Cement Fly ash Notation Factor Low level High level
Physical F/C Filler–cement ratio 1 3
Blaine fineness 391 252 FA Fly ash as replacement 20 100
(m2/kg) for sand (%)
Specific gravity 3.13 2.09 FV Foam volume (% of total 10 50
28-day – volume of concrete)
compressive 54.7
strength (MPa)
% by IS 12269-1987 (17) % by ASTM C 618 class F fly ash conforming to ASTM C 618-89 [17] were
mass mass (Class F) (18) used. The properties of cement and fly ash used in this
Chemical study are presented in Table 1.
SiO2 19.3 – 63.6 – The mix proportions were arrived at as per the proce-
CaO 61 – 1.57 <10 dure given in ASTM C 796-97 [18]. As the standard deals
Al2O3 5.687 – 28.19 – with only cement slurry, the procedure was modified to
Fe2O3 6.036 – 2.99 –
include cement–sand–fly ash components. As the foam is
MgO 1.875 6 0.54 5 (max)
MnO – – 0.03 – added to the wet mix, the consistency of the wet mix is very
Na2O – – 0.05 – important to get the desired design density [15], which is
K2O – – 0.003 – expressed in terms of water–solids ratio required to pro-
SO3 1.67 2.5 0.26 – duce this consistency, and its range is narrow for a given
Loss on ignition 0.2963 4 0.85 –
mix. Based on several trails, the percent flow (consistency),
Soluble residue 1.489 2 – –
Al2O3/Fe2O3 0.94 0.66 – – measured in a standard flow table [19] (without raising/
SiO2 + Al2O3 – – 94.78 70 (min) dropping of the flow table as it may affect the foam bubbles
+ Fe2O3 entrained in the mix) was arrived at as 45 ± 5%. Earlier
studies by the authors have shown that within this range,
the water–solids ratio does not affect compressive strength.
the use of fly ash results in higher strength to density ratio As the water–solids requirement for obtaining this flow val-
and facilitates its high volume utilization [15]. This study ues varied for mixes with and without fly ash replacement
attempts to develop empirical models for predicting the (0.3–0.4 for mixes without fly ash and 0.35–0.6 for mixes
density and compressive strength of foam concrete from with fly ash at different replacement levels), it was decided
the mixture composition details like filler–cement ratio, to split experimental programme into two mixes, namely
addition of fly ash as partial/complete replacement for cement–sand mixes (M1) and cement–fly ash–sand mixes
sand and foam volume, through systematically designed (M2).
experiments.
2.2. Experimental design and validation
2. Experimental investigations
With an objective of developing relationships between
2.1. Materials and mixture composition the parameters and response (empirical models for com-
pressive strength and density), a statistical methods of
Foam was produced by aerating an organic based foam- experimental design based on response surface methodol-
ing agent (dilution ratio 1:5 by weight) using an indige- ogy (RSM) is adopted. For providing equal precision of
nously fabricated foam generator to a density of 40 kg/ estimates in all directions, a central composite design with
m3. 53 grade Ordinary Portland Cement [16], pulverized rotatability or equal precision is selected [20]. The factors
river sand finer than 300 lm (specific gravity = 2.52) and and factor levels are shown in Table 2. Mixture M1 has
Table 3
Observed and predicted responses for confirmation of models
Mix type Mixture composition Compressive strength, MPa Dry density, kg/m3
F/C FA, % FV, % 7-Day 28-Day 90-Day
Predicted Observed Predicted Observed Predicted Observed Predicted Observed
M1 1 – 20 9.0 7.8 11.6 10.7 13.2 12.3 1282 1318
2 – 35 2.41 2.14 2.67 2.30 2.91 3.01 998 1011
3 – 50 0.25 1.20 0.52 1.23 0.84 1.54 802 787
M2 1 70 10 13.8 12.3 18.9 17.8 24.7 23.0 1194 1224
2 65 30 3.66 2.93 5.08 4.84 9.23 8.03 858 854
3 50 40 1.03 1.87 1.10 1.96 3.83 3.01 664 650
754 E.K. Kunhanandan Nambiar, K. Ramamurthy / Cement & Concrete Composites 28 (2006) 752–760
a 50
7.8E2
1400 45
800 25 1.2E3
3
20 1.3E3
600
1.0 1.3E3
15
tio
1.5
ra
1.4E3
nt
2.0 10
me
10 15 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0
20 25
-ce
30 35 2.5
Foam Filler-cement ratio
ler
volum 40 45
3.0
Fil
e% 50
100
b 8.4E2
1050 90 8.6E2
950 60
50 9.2E2
900
40
9.8E2 9.4E2
850 30
20
Fly 3040 1.0 20 1E3 9.6E2
as 50 1.5 800 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0
hr 60
ep 2.0
lac 70 80 ratio Filler-cement ratio
em 2.5 ent
en 90 l e r - c m
e
t % 100 3.0 Fil
Fig. 1. Response surface and contour plot for dry density for (a) Mix M1; (b) Mix M2 (FV = 30%).
two variables (F/C and FA) and M2 has three (F/C, FA For cement–sand foam concrete mix (Mix M1):
and FV). The experimental design was done using Statisti- Dry density ¼ 1655:92 18:283 ðF=CÞ 17:753 ðFVÞ
cal Analysis Software (SAS) [21], resulting in 13 and 20 7-Day strength ¼ 25:328 5:903 ðF=CÞ 0:737 ðFVÞ
experimental runs for mixes M1 and M2, respectively. þ 0:00474 ðFVÞ2 þ 0:120 ðF=CÞ ðFVÞ
The fresh density of foam concrete was measured by fill- 28-Day strength ¼ 32:342 6:098 ðF=CÞ 1:004 ðFVÞ
ing a standard container. The compressive strength was
þ 0:00768 ðFVÞ2 þ 0:118 ðF=CÞ ðFVÞ
determined at the age of 7, 28, 90 days using six cube spec-
imens of size 50 mm after curing by immersing in water in a 90-Day strength ¼ 38:212 7:327 ðF=CÞ 1:225 ðFVÞ
curing tank. A sample size of six was arrived at as per þ 0:00929 ðFVÞ2 þ 0:155 ðF=CÞ ðFVÞ
ASTM E 122-72 [22]. The dry density was determined in
accordance with the section on Oven-Dry Weight of Test Mix M1 (without FA)
1500 Mix M2 with 20% FA
method ASTM C 495-91a [23]. Mix M2 with 60% FA
1400
The results were analysed using SAS to determine the Mix M2 with 100% FA
1300
Dry density kg/m3
For cement–sand–fly ash foam concrete mix (Mix M2): 28-Day strength ¼ 31:179 6:279 ðF=CÞ þ 0:117 ðFAÞ
Dry density ¼ 1438:536 þ 68:465 ðF=CÞ 2:629 ðFAÞ 1:053 ðFVÞ 0:00119 ðFAÞ2 þ 0:00799 ðFVÞ2
12:333 ðFVÞ 22:563 ðF=CÞ2 0:098 ðFVÞ2 þ 0:138 ðF=CÞ ðFVÞ
þ 0:038 ðFAÞ ðFVÞ 90-Day strength ¼ 32:002 5:348 ðF=CÞ þ 0:193 ðFAÞ
7-Day strength ¼ 23:770 3:669 ðF=CÞ 0:778 ðFVÞ 1:028 ðFVÞ 0:00140 ðFAÞ2 þ 0:00715 ðFVÞ2
þ 0:00628 ðFVÞ2 þ 0:0:082 ðF=CÞ ðFVÞ þ 0:095 ðF=CÞ ðFVÞ
Foam volume %
10 1.4
35
8
2.8
6 30
4.2
4 25
5.6
2 7.0
7.0
20
9.1 8.47.7
0
1.0 15 111111 9.8
9.8
tio
1.5 1312
13
ra
10
2.0
nt
30 35
Foam v 40 45
ler
olume 50 3.0
Fil
%
28-Day compressive strength, MPa
50
b 20
18 45
2.0
16 40 1.0
Foam volume %
14
12 35
10 30 4.0
8 6.0
6 25
8.0
4
20 10
2 12
0 15 14
1.0
16
1.5
o
ati
10
2.0
r
20 25
me
Foam v
olume 40 45
% 50 3.0
ler
Fil
c
90-Day compressive strength, MPa
50 1.2
20 45
1.2
40
Foam volume %
15 2.2
35
10 30 4.4
6.6
25
8.8
5
20 11
13
0 15 15
1.0 18
tio
1.5 20
10
ra
0 15 2.0 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0
nt
me
20 25 Filler-cement ratio
30 35 2.5
-ce
Foam v 40 45
ler
olume 50 3.0
%
Fil
Fig. 3. Response surface and contour plot for (a) 7-Day strength for Mix M1; (b) 28-Day strength for Mix M1; (c) 90-Day strength for Mix M1.
756 E.K. Kunhanandan Nambiar, K. Ramamurthy / Cement & Concrete Composites 28 (2006) 752–760
The models contain those terms that are statically signif- of the models was also checked by examining the residuals
icant (i.e. those terms that had a T-statistics greater than for trends by looking at the diagnostic plots like normal
that of chosen significance level, 0.05 in this study). The probability plots and residual plots. To confirm the accu-
adequacy checks on strength and density models (R- racy of the prediction relations, a few tests were carried
squared value, F-value and T-statistics) satisfy statistical out with some of the factor levels chosen from the experi-
requirements (details presented in Appendix A). Adequacy mental series. The predicted and observed responses for
a 100
1.4
14 90
ngth, MPa
8 60
5.6
6 50 7.0
4 40 8.4
9.8
2 30 11
100 13
0 90
10 80 20
15 70 % 10 15 20 25 30 35 40 45 50
20 nt
foa 25 30 50
60
eme Foam volume %
mv 35 lac
olu 40 ep
me 40 30 hr
% 45 as
50 20 Fly
b 100
ngth, MPa
20 90
2.0
18
Fly ash replacement %
80
16 4.0
28-day compressive stre
14 70
6.0
12 60 8.0
10
50 10
8
12
6 40
4 14
30 16
2 2.0
0 100 18
10 90 20
15 80
20 70 t% 10 15 20 25 30 35 40 45 50
25 60 en
Foa 30
35
50
c em Foam volume %
mv 40 la
olu 40 30 ep
me 45 20 shr
% 50 a
Fly
100
c 25 5.2
ngth, MPa
90
Fly ash replacement %
20 80 7.8
70
90-Day compressive stre
10
15
60
13
10 50
23
16
40
5
18
30
0 100
10 90 21 2.6
15 80 20
20 70 %
25 60 nt 10 15 20 25 30 35 40 45 50
Foa 30
35 40
50
e me
mv 40 lac Foam volume %
olu 45 30 ep
me
% 50 20 shr
a
Fly
Fig. 4. Response surface and contour plot for (a) 7-Day strength (Mix M2, F/C ratio: 1); (b) 28-Day strength (Mix M2, F/C ratio: 1); (c) 90-Day strength
(Mix M2, F/C ratio: 1).
E.K. Kunhanandan Nambiar, K. Ramamurthy / Cement & Concrete Composites 28 (2006) 752–760 757
both compressive strength and density were found to sity decreased linearly with an increase in foam volume.
match within satisfactory limits as given in Table 3. The The variation in density with filler–cement ratio was very
behaviour of foam concrete for different mixture composi- less compared to that with foam volume. For mixes with
tions are discussed in the next sections by plotting the fly ash replacement (Mix M2), the density varies nonlin-
response surface graphs developed from above equations. early with filler–cement ratio and shows a peak value as
indicated by the contour plot. For a given foam volume,
3. Response surface for dry density replacement of fine aggregate with fly ash, due to its rela-
tively lower specific gravity, reduces the density of foam
The response surface and contour plot for dry density of concrete (Fig. 2). Alternately, to achieve a particular den-
foam concrete mixes M1 and M2 are shown in Fig. 1(a) and sity of foam concrete, use of fly ash results in a reduction
(b), respectively. In cement–sand mixes (Mix M1), the den- in foam volume.
a b
28-Day compressive strength, MPa
Fig. 5. Foam volume–filler cement ratio interaction for Mix M2 at (a) 28-Day, (b) 90-Day.
a FA% :0
b
28-Day compressive strength, MPa
7-Day compressive strength, MPa
14 20 FA% :0
FA% :20
FA% :40 18 FA% :20
12
FA% :60 16 FA% :40
10 FA% :80 14 FA% :60
FA% :100 12 FA% :80
8
FA% :100
10
6 8
4 6
4
2
2
0 0
600 800 1000 1200 1400 600 800 1000 1200 1400
Dry density kg/m3 Dry density kg/m3
c FA% :0
90-Day compressive strength, MPa
25 FA% :20
FA% :40
20 FA% :60
FA% :80
15 FA% :100
10
4. Response surface for compressive strength Solid line-Strength-density ratio Dotted line-Strength
F/C ratio 1 F/C ratio 1
F/C ratio 2 F/C ratio 2
Fig. 3(a)–(c) shows the response surfaces and contour F/C ratio 3 F/C ratio 3
26 22
plots for compressive strength of cement–sand mix (M1)
at the age of 7, 28, 90 days. Compressive strength varies lin-
Table A.2b
Effect estimates of factors for cement–sand–fly ash mix M2
Responses 7-Day 28-Day 90-Day Dry density
Estimates T Prob > T T Prob > T T Prob > T T Prob > T
F/C 5.275 0.0001 7.587 0.0001 6.256 0.0001 3.765 0.0024
FA – – 3.611 0.0031 2.9406 0.0114 10.39 0.0001
FV 20.76 0.0001 20.98 0.0001 20.50 0.0001 55.07 0.0001
F/C * FA – – – – – – – –
F/C * FV 3.259 0.0053 4.4109 0.0007 2.1791 0.0483 2.40 0.0320
FA * FV – – – – – – – –
(F/C)2 – – – – – – – –
(FA)2 – – 4.103 0.0012 3.3604 0.0051 2.36 0.0342
(FV)2 6.754 0.0001 6.9168 0.0001 4.4130 0.0007 4.144 0.0011
density range, it is the foam volume which controls the ment over and above the pozzolanic and filler effects. The
strength of foam concrete, rather than the material proper- strength–density ratio is high for foam concrete mixes con-
ties. At 90 days, the contribution of fly ash at lower density taining fly ash as compared to sand.
is seemed to be better than other lower ages.
For a given strength and density, the percentage fly ash Appendix A
replacement can be appropriately selected from the
strength–density relation as shown in Fig. 6. Knowing See Tables A.1(a)–A.2(b).
the filler–cement ratio and fly ash replacement level the
foam volume requirement can be calculated using the equa- References
tion developed for dry density. The water requirement for
the foam concrete mixture with these parameters can be [1] Valore RC. Cellular concretes—Part I. Composition and methods of
determined based on the consistency (flow value of preparation. J Am Concr Inst 1954;50(48a):773–95.
[2] Valore RC. Cellular concretes—part II. Physical properties. J Am
45 ± 5 %). Thus the prediction relation developed using Concr Inst 1954;50(48b):817–36.
statistical treatment of experimental results can serve as [3] Short A, Kinniburgh W. Lightweight concretes. London: Applied
guidelines in the design of foam concrete mixes. Sciences Publishers; 1978.
Fig. 7 shows the effect of fly ash replacement on the 90- [4] Rudnai G. Lightweight concretes. Budapest: Akademiai Kiado;
day strength and strength–density ratio for foam concrete 1963.
[5] Fred C, Cormick Mc. Rational proportioning of preformed foam
with 10% foam volume and for different filler–cement cellular concrete. ACI Mater J 1967:104–9.
ratios. The strength–density ratio increases with finer filler [6] Richard TG. Low temperature behaviour of cellular concrete. ACI
like fly ash. It can be seen that even though there is a mar- Mater J 1977;74(4):173–8.
ginal reduction in strength after around 70% fly ash [7] Hoff CG. Porosity-strength considerations for cellular concrete. Cem
Concr Res 1972;2:91–100.
replacement level, while the strength to density ratio
[8] Durack JM, Weiqing L. The properties of foamed air cured fly ash
remains nearly same beyond this level. Hence a 100% based concrete for masonry production. In: Proceedings of fifth
replacement level will allow higher utilization of fly ash Australasian Masonry conference. Gladstone, Queensland, Australia,
with same strength to density ratio, with a small reduction June 1998. p. 129–38.
in strength. [9] Nehdi M, Dejebbar Y, Khan A. Neural network model for preformed
foam cellular concrete. ACI Mater J 2001;98(5):402–9.
[10] Tam CTT, Lim Y, Lee SL. Relation between strength and volumetric
6. Conclusions composition of moist-cured cellular concrete. Mag Concr Res 1987;
39:12–8.
[11] Kearsley EP, Wainwright PJ. The effect of high fly ash content on the
Conclusions pertain to the experimental investigations compressive strength of foamed concrete. Cem Concr Res 2001;31:
conducted and are valid for the factors and factor level 105–12.
considered are summarised below. [12] Kearsley EP, Wainwright PJ. Porosity and permeability of foamed
The prediction relation developed using statistical treat- concrete. Cem Concr Res 2001;31:805–12.
ment of experimental results can serve as guidelines in the [13] Kearsley EP, Wainwright PJ. Ash content for optimum strength of
foamed concrete. Cem Concr Res 2002;32:241–6.
design of foam concrete mixes. Fly ash when used as par- [14] Kearsley EP, Wainwright PJ. The effect of porosity on the strength of
tial/complete replacement for fine aggregate; cause an foamed concrete. Cem Concr Res 2002;32:233–9.
increase in compressive strength. The replacement level [15] Kunhanandan Nambiar EK, Ramamurthy K. Influence of filler type
for maximum strength depends on age of testing; it is on the properties of foam concrete. Cem Concr Compos 2006;28:
49% at 28 days and 71% at 90 days. Replacement of sand 475–80.
[16] Bureau of Indian standards. Specifications for 53 grade ordinary
with fly ash in the mixture generally reduces the density, Portland cement. New Delhi, BIS, 1987. IS 12269-1987.
resulting in reduced foam volume requirement for a given [17] American Society for testing and materials. Standard specification for
density and thus indirectly contributes to strength enhance- fly ash and raw or calcined natural pozzolana for use as a mineral
760 E.K. Kunhanandan Nambiar, K. Ramamurthy / Cement & Concrete Composites 28 (2006) 752–760
admixture in Portland cement concrete. Philadelphia, ASTM, 1989. [22] American Society for testing and materials. Standard recommended
ASTM C 618-89. practice for choice of sample size to estimate the average quality of a
[18] American Society for testing and materials. Standard test method for lot or process. Philadelphia, ASTM, 1972. ASTM E 122-72.
foaming agents for use in producing cellular concrete using preformed [23] American Society for testing and materials Standard test method for
foam. Philadelphia, ASTM, 1997. ASTM C 796-97. compressive strength of light weight insulating concrete. Philadelphia,
[19] American Society for testing and materials. Specification for flow ASTM, 1991. ASTM C 495-91a.
table for use in tests of hydraulic cement, Philadelphia, ASTM, 1998. [24] Johns MR, McCarthy MJ, McCarthy A. Moving fly ash utilization in
ASTM C 230-98. concrete forward: a UK perspective, International ash utilization
[20] Montgomery DC. Design and analysis of experiments. New York symposium. Centre for Applied Energy Research, University of
(NY): John Wiley & Sons; 2001. Kentucky; 2003.
[21] SAS Release 8.02, SAS Institute Inc., Cary, NC, USA.