APSEC-EACEF 2009, Universiti Teknologi Malaysia UTM Langkawi,

Malaysia, 4-6 August 2009

OPTIMUM MIX PROPORTIONING OF MIRHA FOAMED CONCRETE

USING TAGUCHI'S APPROACH

Muhd. Fadhil B NURUDDIN

Associate Professor, Civil Engineering Department - Universiti Teknologi PETRONAS Malaysia

Ridho BAYUAJI
Ph.D Student, Civil Engineering Department - Universiti Teknologi PETRONAS Malaysia
Lecturer, Institut Teknologi Sepuluh Nopember (ITS), Surabaya Indonesia
*Corresponding author Email: fadhilnuruddin@petronas.com.my

ABSTRACT:
Lightweight foamed concrete (LWFC) is essentially a concrete of low density that reduces the total self
weight of the structures. In this study, mix proportion parameters for LWFC are analyzed using the
Taguchi’s experiment design methodology for optimal design. For that purpose, the mixtures are designed
in L16 orthogonal array with five factors. The best possible mix proportions are determined via compressive
strength test and porosity values

Keywords: Lightweight Foamed Concrete (LWFC), Microwave Incinerated Rice Husk Ash (MIRHA),
Taguchi Method

1. INTRODUCTION The use of replacement materials in Portland cement

The primary feature differentiating of foamed
concrete with normal concrete is it can be designed to
the required light density by arranging air foam as an
“aggregate” in concrete [1-3]. Foamed concrete
possesses many advantages such as lightweight,
thermal and sound insulation, void filling, good shock
absorption and high impact resistance [4].
Nevertheless the criteria for structural lightweight
concrete are minimum 28 day cylinder compressive
strength of 17.5 N/mm2 and maximum dry density of
1850 kg/m3 [5].
In Europe, LWFC is considered load bearing if it
fulfils the requirement strength as structural
application at 28 day (typically greater than 20
N/mm2 compressive strength with actual plastic
density 1405 kg/m3) [6]. In South Africa, LWFC of
density 1500 kg/m3 is considered to be a load bearing
concrete if a minimum compressive strength of 17.5
N/mm2 at 28 day [7] is met. The low thermal
conductivity, high compressive strength and high
performance [8] of foamed concrete can be achieved
by changing the fine aggregate with high fly ash [6,
7], or silica fume and ultra-fine silica powder, fly ash
without using sand, decreasing w/c and other
measures.
Hamidah et al’s [9] showed that the lower
sand-cement ratio results in foamed concrete of
higher compressive strength and the mix proportion
need huge amount of cement. It showed that
cementitious material is an interesting factor to
explore in more detail.
has been gaining much attention in recent years.
Given the world situation with respect to cement
demand and the cost of energy, binders such as
ground granulated blast-furnace slag (GGBS) and
pulverized fuel ash (PFA), RHA can be used. In
blended cements, the replacement material may take
part in the hydration reactions and contribute to the
hydration products.
Two groups of cement replacements are commonly
recognized pozzolanas and are also latently hydraulic
materials. The first group reacts with the CH
produced during the cement hydration, while the
second has hydraulic characteristics itself, but its
hydration needs activation, for example by the
presence of hydration products (e.g., CH) or other
additives.
In Malaysia, large amount of rice husks are being
produced annually almost reaching 2.2 million tons
and the total annual world production is about 600
million tons. The utilization of rice husk ash as a
pozzolanic material in cement and concrete provides
several advantages, such as improved strength and
durability properties, reduced materials cost due to
cement savings and environmental benefits related to
the disposal of waste materials and to the reduction of
carbon dioxide emissions.
In this research [10], the investigation carried out is to
better understand the effect of the MIRHA, water
binder ratio, and sand cement ratio on the foamed
concrete. The experimental work (using the Taguchi
method) is designed to give the optimum working
conditions of the parameter that affects the physical
properties of concrete mixtures. One of the
advantages of Taguchi method over the conventional

694
experimental design, in addition to keeping the most robust with respect to noise factors. The Taguchi
experimental cost at the minimum level, is that it method utilizes orthogonal arrays from design of
minimizes the variability around the investigated experiments theory to study a large number of
parameters when bringing the performance value to variables with a small number of experiments. An
target value. Its other advantage is that the optimum orthogonal array significantly reduces the number of
working conditions determined from the laboratory experimental configurations to be studied.
work can also be reproduced in the real production Furthermore, the conclusions drawn from small-scale
environment. experiments are valid over the entire experimental
region spanned by the control factors and their
2. TAGUCHI METHODS AND RESEARCH settings.
MODEL
In this study, the following parameters are considered
in the mix proportions
Taguchi method is a statistical method developed by
- MIRHA contents;
Genichi Taguchi during the 1950s [11] as an
- sand cement ratio, (s/c);
optimization process technique. Taguchi’s approach
- water cementations ratio, (w/c);
to parameter design provides the design engineer with
- superplasticizer content;
a systematic and efficient method for determining
- foam content;
near optimum design parameters for performance and
There are ten steps in a systematic approach to the use
cost. The first concept of Taguchi that must be
of Taguchi's parameter design methodology [13].
discussed is the ‘‘noise factors”. Noise factors are
Figure 1 shows the detail procedure of Taguchi design
viewed as the cause of variability in performance,
methodology.
including why products fail. The signal–to-noise ratio
(S/N) is used in evaluating the quality of the product
[12]. The S/N measures the level of performance and 1. Problem recognition and formulation
the effect of noise factors on performance and is an
evaluation of the stability of performance of an output 2. Select quality characteristic
characteristic. Target values may be:
1. Smaller is better, choose when goal is to 3. Select design or process parameters
minimize the response. The S/N can be
calculated as given in Eq. (1) for smaller the
4. Classify design parameters
better
1 n 2
S/N =  10 * log 10   Yi 
 n i 1 
(1) 5. Determine levels

2. Larger is better: choose when goal is to maximize 6. Identify interactions

the response. The S/N is calculated as given in
Eq. (2) for larger the better
7. Choose appropriate orthogonal array
1 n 1 
S/N =  10 * log 10    (2)
n Y2 
 i 1 i  8. Conduct experiments
3. Nominal is better: choose when goal is to target
the response and it is required to base the S/N on 9. Perform statistical analysis
standard deviations only. The S/N is calculated as
given in Eq. (3) for smaller the better. 10 Perform a confirmatory experiment and
implement results
1
Yi  Y0 2 
n
S/N =  10 * log 10  
 n i 1 
(3)
Fig.1 Taguchi method algorithm

The variation levels for the considered parameters are

In Eqs. (1) – (3), Y shows the measured value of each
shown in Table 1.
response. When variability occurs, it is because the
physics active in the design and environment that
promotes change [12]. Noise factors can be classified Table 1. Parameters and their variation levels
into three groups: Parameter unit Level
- External noise factors: sources of variability that MIRHA (%) 0 5 10 15
come from outside the product, w/c ratio 0.3 0.35 0.4 0.45
- Unit-to-unit noise: due to the fact that no two s/c ratio 0.25 0.5 0.75 1
manufactured components or products are ever SP (%) 1 1.5 2 2.5
exactly alike, Foam (%) 20 25 30 35
- Internal noise: due to deterioration, aging, and
wear incurred in storage and use. Table 2 shows the details of the variables used in the
The objective is to select the best combination of the experiment. Note that the parameters are at four
control parameters so that the product or process is levels. Only 16 experiments are needed to study the

695
entire experimental parameters using L16 (45) Table 4. Mixture proportion of concrete
orthogonal arrays.
Cement Sand Water MIRHA Vol Foam SP
Table 2. Standard L16 orthogonal array Code
(kg/m3) (kg/m3) (kg/m3) (kg/m3) (liter/m3) (kg/m3)
Independent variables Perform.
Exp. LWFC-1 930 233 419 0 203 23
parameter
no Var. 1 Var. 2 Var. 3 Var. 4 Var. 5 LWFC-2 950 475 285 0 238 19
value
1 1 1 1 1 1 P1 LWFC-3 770 578 270 0 272 12
2 1 2 2 2 2 P2 LWFC-4 620 620 248 0 324 6
3 1 3 3 3 3 P3 LWFC-5 893 223 282 47 333 13
4 1 4 4 4 4 P4 LWFC-6 703 352 333 37 298 7
5 2 1 2 3 4 P5 LWFC-7 732 549 308 39 239 18
6 2 2 1 4 3 P6 LWFC-8 703 703 259 37 240 18
7 2 3 4 1 2 P7 LWFC-9 900 225 350 100 239 9
8 2 4 3 2 1 P8
LWFC-10 810 405 360 90 194 12
9 3 1 3 4 2 P9
10 3 2 4 3 1 P10 LWFC-11 572 429 286 64 345 11
11 3 3 1 2 4 P11 LWFC-12 648 648 216 72 305 16
12 3 4 2 1 3 P12 LWFC-13 748 187 352 132 284 15
13 4 1 4 2 3 P13 LWFC-14 663 332 273 117 342 17
14 4 2 3 1 4 P14 LWFC-15 774 580 273 137 206 8
15 4 3 2 4 1 P15 LWFC-16 565 565 299 100 267 8
16 4 4 1 3 2 P16

The mixes are prepared using a rotating planetary

3. METHODHOLOGY mixer and in accordance with BS 1305:1974. After
3.1. Material the mixing procedure is completed, tests are
The constituent materials used in the laboratory to conducted on the fresh concrete to determine the
produce foamed concrete comprised (i) Portland slump flow diameter and flow time. Segregation and
cement (Ordinary Portland Cement BSEN bleeding are visually checked during the slump flow
197-1-2000) (ii) natural sand, with 100 % passing 425 test. From each concrete mix, six 50 mm cube
mm sieved (BS EN 12620:2002), (iii) MIRHA, with samples are tested for 28 days compressive strength,
high reactive silica content, controlled combustion of while 25.4x25.4 mm cylinders were used for
ash (BS EN 450:2000) and (iv) free water (BS determining the porosity level.
3148:1980) (v) superplasticizer (BS EN 934-2:2001).
The surfactant used for the production of the 3.3. Curing regime
preformed foam by aerating palm oil based (again, The specimens are demolded 24 hour after casting
this is typical of industry practice) a ratio of 1:30 (by and placed in water tank at 20  2oC. The curing is
volume), aerated to a density of 110 kg/m3 (ASTM C done in accordance with BS EN 12390-2:2000
869-91 (reapproved 1999), ASTM C 796-97). The
chemical properties of MIRHA and OPC used are 3.4. Hardened concrete
shown in Table 3 For each concrete mix, the compressive strength test
is conducted in accordance BS EN 12390-3. The test
Table 3. Binder Properties results on hardened concrete can be seen in Table 5.
Oxide Weight % Table 5. Test results of hardened concrete
compositio MIRHA OPC
28-day
Na2O 0.02 0.02 Dry Density Compressive Porosity 28d
Code
MgO 0.63 1.43 (kg/m3) strength (%)
Al2O3 0.75 2.84 (N/mm2)
SiO2 90.75 20.44 LWFC-1 1856 55.14 29.61
LWFC-2 1899 68.89 22.42
P2O5 2.5 0.1
LWFC-3 1541 25.19 31.10
K2O 3.77 0.26 LWFC-4 1340 13.76 40.25
CaO 0.87 67.73 LWFC-5 1400 29.68 37.61
TiO2 0.02 0.17 LWFC-6 1504 26.57 33.68
LWFC-7 1918 54.81 23.93
Fe2O3 0.28 4.64 LWFC-8 1886 63.87 22.13
SO3 0.33 2.2 LWFC-9 1409 29.99 37.19
MnO 0.08 0.16 LWFC-10 1517 37.46 28.82
LWFC-11 1208 18.18 43.40
3.2. Test specimens LWFC-12 1668 40.14 25.51
The mix proportions of foam concrete are presented in LWFC-13 1249 27.84 36.34
Table 4. LWFC-14 1412 30.77 30.33
LWFC-15 1703 10.19 26.42
LWFC-16 1374 22.93 30.18

696
In Table 5, the dry density, the 28-day compressive strength and for the minimization of dry density and
strength and the porosity of lightweight-foamed porosity values using the Taguchi method.
concrete are in the range of 1,208 to 1,918 kg/m3, The performance statistics for ‘‘the larger the better”
10.19 to 68.89 N/mm2, 22.13 to 43.40%, respectively. situations are evaluated for maximization properties
The lowest dry density, the highest compressive and of LWFC and ‘‘the smaller the better” situations are
the lowest porosity captured are obtained from evaluated for minimization properties of LWFC.
LWFC-11, LWFC-2 and LWFC-8 mixes, respectively. Since all parameters had interaction between them in
LWFC mix, the best possible testing conditions of the
4. RESULT AND DISCUSSION LWFC properties can be determined from the main
The best possible levels of mix proportions are effect plot graphs from Figs. 2–4 for compressive
investigated for the maximization of compressive strength, porosity, and dry density, respectively.
47.5

42.5
28-day Compressive Strength (N/mm2)

37.5

32.5

27.5

22.5

17.5
0 5 10 15 0.3 0.35 0.4 0.45 0.25 0.5 0.75 1 1 1.5 2 2.5 20 25 30 35
MIRHA (%) w/c s/c SP (%) Foam (%)

Fig. 2. Main effect plot for 28-day compressive strength

38

36

34
Porosity (%)

32

30

28

26
0 5 10 15 0.3 0.35 0.4 0.45 0.25 0.5 0.75 1 1 1.5 2 2.5 20 25 30 35
MIRHA (%) w/c s/c SP (%) Foam (%)

Fig. 3. Main effect plot for Porosity

1725

1675

1625
Dry Density (Kg/m3)

1575

1525

1475

1425

1375

1325
0 5 10 15 0.3 0.35 0.4 0.45 0.25 0.5 0.75 1 1 1.5 2 2.5 20 25 30 35
MIRHA (%) w/c s/c SP (%) Foam (%)

Fig. 4. Main effect plot for dry density

697
A statistical analysis is performed to determine the
statistically significant factors and the data analysis is
presented in Table 6. Finally, degree of contribution
of each significant factor is obtained so as to
determine the level of its statistical importance in the
model. The contribution percentage in Table 6 gives
an idea about the degree of contribution of the factors
to the measured response. If the contribution
percentage is high, the contribution of the factors to
that particular response is more. Likewise, lower
contribution percentage will lower the contribution
factors on the measured response.
As it can be seen from Fig. 2, increasing MIRHA and
foam level decrease the compressive strength.
Superplasticizer is the most influencing factor on the
compressive strength of the LWFC with 39.8%
contribution. The second most influencing factor is
foam content with 25.7% contribution. The
contribution rank of the parameters on the
compressive strength can be seen in Table 6. An
optimal condition for maximum 28-day compressive
strength is obtained at 5% MIRHA, 0.3 w/c, 0.5 s/c,
2 % SP, and 25% Foam
The porosity of foamed concrete must be minimized
for durability purpose. In this research, the porosity is
a measure of the portion of the total volume of
concrete occupied by pores. In Figure2, increase of
w/c and foam content parameter increase the porosity
but increasing the content of superplasticizer
decreases the porosity. Foam content is the most
influencing factor on the porosity of the LWFC with
47.0% contribution. The second most influencing
factor is superplasticizer with 16.6 % contribution. An
optimal condition for minimum porosity is obtained
at 5% MIRHA, 0.3 w/c, 0.5 s/c, 2.5% SP, and 20%
Foam.
Dry density of concretes must be minimized for
reducing the self-weight of foamed concrete. The
decreasing density is affected by the increasing of
MIRHA, w/c and foam values (see Figure.3). Foam
content is the most influencing factor on the dry
density of the LWFC with 44.2% contribution. The
second most influencing factor is MIRHA with 24.1%
contribution. The optimum mix proportions for
minimization of dry density is 15% MIRHA, 0.45 w/c,
0.25 s/c, 1.5% SP, and 35% Foam.

Table 6. Analysis of variance of LWFC

28-day
statistical Dry
Parameter compr. Porosity
parameters Density
strength
MIRHA DFa 3 3 3
SSSb 204558.3 1068.8 40.5
ASSc 204558.3 1068.8 40.5
MSd 68186.1 356.3 13.5
Contribution (%) 24.1% 23.3% 6.6%
sand/cement DF 3 3 3
SSS 32772.6 391.2 97.7
ASS 32772.6 391.2 97.7
MS 10924.2 130.4 32.6
Contribution 3.9% 8.5% 15.8%
water/binder DF 3 3 3
SSS 79573.3 125.7 86.9
ASS 79573.3 125.7 86.9
MS 26524.4 41.9 29.0
Contribution (%) 9.4% 2.7% 14.1%
superplasticizer DF 3 3 3
SSS 156036.0 1829.7 102.2
ASS 156036.0 1829.7 102.2
MS 52012.0 609.9 34.1
Contribution (%) 18.4% 39.8% 16.6%
foamed DF 3 3 3
SSS 374984.9 1180.8 290.4
ASS 374984.9 1180.8 290.4
MS 124995.0 393.6 96.8
Contribution (%) 44.2% 25.7% 47.0%

a Degree of freedom
b Sequential sum of square
c Adjusted sum of square
d Mean square (variance)

698
According to Figs. 2–4, the best mix proportions of Using Neural Network. in National Post
the target properties are tabulated in Table 7
Graduated Conference UTP. 2008.
Table 7. Optimal mix-design proportions for
properties of foamed concrete
4. Rahman, A.M.A.Z.a.I.A., Investigation on
superpla
MIRHA Foam Foamed Concrete as A Novel Protective
Optimal mix Proportional w/c s/c sticizer
(%) (%)
(%) Structure. ICCE 2008, 2008.
28-day compressive strength 5 0.3 0.5 2 25
Porosity 5 0.3 0.5 2.5 20
Dry Density 15 0.45 0.25 1.5 35 5. ASTM-C330-69, Specification for
lightweight aggregates for structural
concrete 2001.
5. CONCLUSSION

A new method is established using Taguchi method

6. McCarthy, M.R.J.a., Preliminary views on
for determination of the optimum composition of
material proportion and the effect of MIRHA the potential of foamed concrete as a
properties and durability of foamed concrete. LWFC
structural material. Magazines of Concrete
consists of many components; it is critical to use a
systematic approach for identifying optimal mixes Research, 2005. 57, No1, February: p. 21-31.
and investigate the most effective factors under a set
of constraints. Due to this reason Taguchi method
with L16 (45) orthogonal array is used in this study to 7. Kearsley EP, P.W., The Effect of high fly ash
investigate ranking of the effective parameters and
content on the compressive strength of
best possible mix proportions of fresh and hardened
properties of LWFC. At the end of this research, it is foamed concrete. Cem. and Concr. Res, 2001.
seen that Taguchi method is a promising approach for
31: p. 105-112.
optimizing mix proportions of LWFC to meet
hardened concrete properties. Taguchi method can
simplify the test protocol required to optimize mix
8. PAN Zhihua1, F.H., Wee Tionghuan,
proportion of LWFC by reducing the number of trial
batches. This study has shown that it possible to Preparation of High Performance Foamed
design foamed concrete that satisfies the criteria of
Concrete from Cement, Sand and Mineral
high strength lightweight concrete. As it can be seen
in the results of fresh and hardened concrete Admixtures.DOI10.1007/s11595-005-2295-4
properties of produced concrete samples, they satisfy
.2006.
the expected properties of LWFC.

REFERENCES
9. MS Hamidah, I.A., MR A Ruslan, K Kartini,
1. Nuruddin MF, Bayuaji R, Artificial Neural NM Fadhil, Optimisatiom of Foamed
Network Prediction For Foamed Concrete Concrete Mix of Different Sand-Cement ratio
Compressive Strength. Journal of and Curing Conditions. Global
International Association for Concrete Construction: Ultimate concrete
Technology, 2007. Vol. 5 No.2 p. pp 11-23. opportunities, 2005.

2. Valore, Cellular Concrete. Journal of The 10. Bayuaji R, Nuruddin MF, ed. Optimum mix
American Concrete Institute, 1954. proportioning of Foamed Concrete Using
Taguchi Method. UTP Internal PG
3. Bayuaji R, Nuruddin MF, Prediction of Symposium. 2009: Universiti Teknologi
Foamed Concrete Compressive Strength PETRONAS.

699
11. Roy, R., A primer on the Taguchi method. 13. Jiju Antony, D.P., Chengbo Wang, Maneesh
1990. Kumar, An application of Taguchi method of
experimental design for new product design
12. Belavendram, N., Quality By Design Taguchi and development process. Assembly
Techniques for Industrial Experimentation. Automation, 2006: p. 18-24.
1995.

700