You are on page 1of 8

ACI MATERIALS JOURNAL TECHNICAL PAPER

Title no. 103-M06

Air-Void System of Foamed Concrete and its Effect on


Mechanical Properties
by Tiong-Huan Wee, Daneti Saradhi Babu, T. Tamilselvan, and Hwee-Sin Lim

A study has been undertaken to investigate the effect of air content Under some circumstances, however, the air voids may also
ranging from approximately 10 to 70% introduced by preformed coalesce during mixing and transporting and form larger air
foam of same quality on the air-void system and the effect of the voids; this is undesirable. A comprehensive review of
air-void system on the mechanical properties of foamed concrete literature dealing with both autoclaved and nonautoclaved
without sand. The linear traverse method (ASTM C 457) was used
aerated concrete properties suitable mostly for nonstructural
to determine air-void system, which refers to air content, average
air-void size, air-void frequency, and spacing factor. The cementi- applications has been reported recently.7 Legatski,1 ACI
tious material used in this study was blended cement with 50% of Committee 523,6 McCormick,8 and Tam et al.9 reported the
normal portland cement and 50% of ground-granulated blast- relationships of compressive strength and modulus of
furnace slag (GGBFS) by weight; the water-cementitious material elasticity with density mainly for low to medium strength
ratio (w/cm) of the foamed concrete was 0.3. For the range of air range (0.69 to 17.24 MPa,10 defined for a density range of
content adopted, the fresh densities of the foamed concrete ranged 800 to 1440 kg/m3) foamed concrete. In recent years, high-
from approximately 600 to 1900 kg/m3 with a corresponding strength foamed concrete had been developed with a low
compressive strength range of approximately 2 to 58 MPa were water-to-binder ratio (≤0.3) and using silica fume and ultra-
obtained. The results show that although the same quality of foam fine silica powder,11 and fly ash12,13 without using sand. A
was introduced, the air-void size and frequency and, hence, compressive strength of up to 60 MPa for the foamed
spacing factor in the foamed concrete, varied with different air concrete under special curing conditions has been reported
contents, which in turn affected the mechanical properties of the
concrete. In this study, it was found that the air-void system with a
together with the relationships between compressive
spacing factor, air-void size, and air content of 0.04 mm, 0.12 mm, strength and porosity and density.
and 42%, respectively, were to be optimal, and at which, a high Generally, air voids that govern the porosity of foamed
strength-to-weight ratio can be achieved. The results also indicated concrete are considered to have a significant effect on
that the inclusion of air voids in foamed concrete had a greater effect compressive strength of the concrete.14,15 As it is possible to
on compressive strength than the modulus of elasticity. A numerical have foamed concrete of different air-void sizes with the
analysis was carried out to verify the experimentally observed same porosity and in the same context—air voids of the same
effect of an air-void system on strength and modulus. The result of size with different porosity—Powers16 proposed a parameter
the numerical analysis concurred with the experimental observations.
known as a spacing factor to characterize the air-void system
in concrete. The relationships suggested by Powers for
Keywords: air void; compressive strength; modulus of elasticity.
estimating the spacing factor are
INTRODUCTION
Foamed concretes, produced by introducing preformed p - for p/A < 4.342
L = -------------- (1)
foam, are lightweight concretes consisting of a system of αA ⁄ V
macroscopic air voids of approximately 0.1 to 1 mm size,
uniformly distributed in either a matrix of aggregate and p 1⁄3
and L = --- 1.4  1 + ---
3
cement paste or cement paste alone. It can be produced –1 for p/A > 4.342 (2)
α  A
anywhere in any shape or building unit size using conventional
equipment and machines and be cost-effective and simple to
produce.1-3 The basic contribution of foamed concrete to the 4N
field of concrete technology is the ability to control its where α = ----------- (3)
A⁄V
density over a wide range. The fresh density of foamed
concrete typically ranges from approximately 320 to
1920 kg/m 3 (20 to 120 lb/ft3).1 The density control is is the specific surface area; L is the spacing factor that is
achieved by adding a calculated amount of foam to slurry of governed by air-void system such as air content, total
water and cement, with or without the addition of sand or number of voids, void size, and specific surface area; p is the
aggregate. However, the skin of the air voids must be tough paste content; A is the air content; V is the volume of the
and persistent to withstand the rigors of mixing and placing, specimen; and N is the average number of air-void sections
during which the air voids are separated, coated with cement encountered along the unit length of traverse line.
paste, and the concrete can be pumped or transported to the
casting location.1,4 ACI Materials Journal, V. 103, No. 1, January-February 2006.
MS No. 04-340 received October 19, 2004, and reviewed under Institute publication
The preformed foam process of making foamed concrete policies. Copyright © 2006, American Concrete Institute. All rights reserved, including
is the most economical and controllable pore-forming the making of copies unless permission is obtained from the copyright proprietors.
Pertinent discussion including authors’ closure, if any, will be published in the November-
process,5 consisting of stable, unconnected air voids.1,6 December 2006 ACI Materials Journal if the discussion is received by August 1, 2006.

ACI Materials Journal/January-February 2006 45


RESEARCH SIGNIFICANCE
ACI member Tiong-Huan Wee is an associate professor in the Department of Civil
Engineering, National University of Singapore, Singapore. His research interests Ensuring an optimal air-void system in foamed concrete is
include concrete technology, primarily in the development of lightweight high- essential to produce foamed concrete with high strength-to-
strength concrete with and without aggregate and other high-performance concrete to
match various applications such as defense, offshore and floating structures, and
weight ratios. For the same w/c, the density of foamed
structures exposed to coastal and aggressive underground conditions, control of concrete may be varied through the incorporation of
early-age thermal and other nonstructural cracking, and nondestructive testing and different amounts of foam, which may result in a different
quality assurance of concrete.
air-void system—that is, air content, air-void size, air-void
Daneti Saradhi Babu is a PhD scholar in the Department of Civil Engineering, frequency, and spacing factor. Therefore, any change in the
National University of Singapore. He received his MS from the Indian Institute of micro/macro structure of foamed concrete due to a variation
Technology Madras, India, in 2002. His research interests include development of
high-strength, structural, and nonstructural lightweight concretes; and durability, in the air-void system may affect the mechanical properties
creep, shrinkage, and cracking of concrete. significantly in relation to density. There are some reports on
foamed concrete that studied the effect of air-void size on
T. Tamilselvan is a research fellow in the Department of Civil Engineering, National
University of Singapore. He received his PhD in civil engineering from the National compressive strength, but none on the change of an air-void
University of Singapore in 1999. His research interests include lightweight concrete, system and its effect on mechanical properties. An under-
fracture mechanics, and thermal cracking of concrete. standing of the air-void system in foamed concrete would be
Hwee-Sin Lim is Director of DE Consultants, Singapore. His research interests essential to determine the optimal air-void system for the
include structural use of lightweight concrete, thermal analysis and cracking of concrete, production of concrete with a high strength-to-weight ratio.
and durability, creep, and shrinkage of high-performance lightweight concrete.

EXPERIMENTAL INVESTIGATION
The cementitious material of equal portions, by weight of
Thus far, the spacing factor has been the most widely used normal portland cement complying with BS 12 and ground-
parameter for measuring the air-void system in hardened granulated blast-furnace slag (GGBFS) complying with BS
normal concrete to determine the durability of air-entrained 6699, was blended and used to form the paste of the foamed
concretes in freezing-and-thawing environments. In previous concrete designed with a water-cementitious material ratio
reports on air-entrained concretes with typical air contents up to (w/cm) of 0.3. A protein-based foaming agent with a specific
10%, the influence of silica fume, high-range water-reducing gravity of 1.1 was used to produce preformed foam using a
admixtures, air-entraining agents, cement characteristics, foam generator. The mixture design method adopted by
type of mixing, retempering, and temperature on the stability Kearsley and Wainwright13 had been used to produce the
of the air-void system17-21 were considered. In these reports, foamed concrete for this study. The foam introduced was
the relationships between spacing factor and time of ascertained to be of the same quality by checking its density.
sampling, air content, and specific surface area were studied. Using this method, foamed concrete of consistent quality
The recent study on air-void stability in self-consolidating with a density of 1200 kg/m3 and a strength of 19 ± 2 MPa
concrete also reported on the relationships between spacing has also been successfully produced in large volumes, cast,
factor and specific surface area, air content, slump flow, and and cured on site just like normalweight concrete, using
rheological properties.22 conventional concrete batching facilities.26
The studies on autoclaved aerated concrete23 cited The molding of foamed concrete was achieved by tapping
insignificant change in the compressive strength for the the sides of the mold with a rubber hammer, as per ASTM
average air-void size ranging from 0.72 to 1.08 mm with C 796-97 recommendations. The ambient temperature during
corresponding bulk densities of 350 to 380 kg/m3. On the mixing and casting was 27 ± 3 °C. Immediately after casting
other hand, studies on foamed concrete24 made with a w/c of and finishing, the specimens were covered with plastic
0.58 and dry concrete densities ranging from approximately sheets to avoid direct exposure to wind and, hence, decrease
550 to 1230 kg/m3 indicated a significant increase of evaporation. The covered specimens were kept at the same
compressive strength when the average air-void size is temperature until demolding. After 24 hours, the specimens
decreased from approximately 0.35 to 0.1 mm. Few reports were demolded and stored in a fog room for curing. The
discussed the relation between the air-void size and compres- specimens were moist-cured in the fog room at 30 ± 2 °C and
sive strength of foamed concrete.24,25 In addition to the air- 100% relative humidity until the day of testing. The fresh
void size, the compressive strength of foamed concrete may density of foamed concrete was tested as per ASTM C 796.
also be affected by the void/paste ratio, number of air voids All tests to determine the properties of foamed concrete were
(frequency of air voids), and spacing of air voids (spacing carried out on three concrete specimens after 28 days of
factor). Existing reports did not emphasize the effect of the air- moist curing, and the average value was taken.
void system on the mechanical properties. Therefore, under- The specimen preparation for the measurement of the air-
standing the air-void system may be essential to produce void system was slightly different then from recommended
foamed concrete with a high strength-to-weight ratio. by ASTM C 457. ASTM C 457 specified the size and thickness
The main objective of this paper was to assess the optimal of the specimen and length of travel in the linear traverse
air-void system in foamed concrete for achieving high method (LTM), based on the size of aggregate. Mixtures
strength-to-weight ratios by understanding: 1) the effect of from this study, however, do not contain any aggregate
air content on average air-void size, air-void frequency, and (either coarse aggregate or sand) but consist of high amounts
spacing factor of foamed concrete that had been produced by of air (foam), and the resulting foamed concrete strength
adding preformed foam of identical quality into a paste made ranges from 2 to 58 MPa. To ensure the stability of the air-
with the same w/c; and 2) the effect of the air-void system on void walls during polishing, particularly in weaker specimens,
mechanical properties of foamed concrete by experimental all the specimens were vacuum-impregnated with slow-setting
studies and numerical analysis. Air content and the spacing epoxy. To ensure consistency in results, all the specimens
factor will be used as parameters of the air-void system to were prepared using similar techniques under the same
discuss its effect on mechanical properties of foamed concrete. environmental conditions, as follows.

46 ACI Materials Journal/January-February 2006


First, specimens of approximately 45 x 45 mm size with a given in Table 1 indicate that the specimens were relatively
minimum thickness of approximately 15 mm were cut from homogeneous.
the center of two randomly selected 100 mm cubes using a Besides this, the compressive strength and modulus of
diamond cutter. The face of the specimen was cut perpendicular elasticity of the concrete were also obtained. The compressive
to the casting direction. Sized specimens were saturated in strength test was carried out on a 100-mm cube in compression
acetone to stop further hydration reaction before drying at testing machine conforming to BS 1881: Part 116. The
105 °C. To ensure the stability of the air-void walls during modulus of elasticity of the concrete was obtained from a
polishing, the dried and cooled specimens were vacuum- 100 x 200 mm cylinder on a servo-controlled hydraulic
impregnated with slow-setting epoxy. The impregnated testing machine according to BS 1881: Part 121.
specimens were polished as per ASTM C 457. After
polishing and cleaning, the specimens were dried at room NUMERICAL ANALYSIS
temperature for 12 hours. Finally, an effective size 40 x 40 mm To study the effect of air voids on the strength and
was considered for air-void measurement, as shown in Fig. 1. modulus of foamed concrete, a numerical analysis using
The air voids were measured according to ASTM C 457 finite element method was carried out. To begin, single-sized
under a microscope with a magnification of 60× on two air voids arranged in a close hexagonal packing configuration,
specimens, prepared as per the procedure described previously, as shown in Fig. 2, was assumed for the foamed concrete.
for each concrete. The LTM was adopted and a total length This would provide for the most natural and compact
of 600 mm accumulated over a series of 15 parallel lines was arrangement of air voids as the spacing between adjacent
traversed on the polished face of the specimen. The air-void bubbles would be relatively uniform, although the adoption
system was then obtained from the equations provided in of a single-sized bubble may be an over-simplification. The
ASTM C 457. dimensions R and r in the figure refers to the maximum
To verify the homogeneity and, hence, the reliability of the possible and actual radius of the air voids. For the numerical
air-void system obtained, a statistical analysis was carried analysis, an arbitrary radius of 2r = 0.15 mm (being the
out. The air content and average air-void size were first overall weighted average void size of all the test specimens)
obtained for each of the 15 parallel lines traverse from which was first used and thereafter, the results were factored for the
the mean x , standard deviation s, and the coefficients of
variation (COVs) were then computed using the equations

∑ xi
x = i=1 -
------------ , (4)
n

∑ ( xi – x )
2

s = i----------------------------
=1 , (5)
n–1

s
and COV = -- × 100% (6)
x
Fig. 2—Single-size void arranged in hexagonal packing.
respectively, where xi is the air content or average air-void
size of each parallel line traversed, and n is the total number
of parallel lines traversed, for each specimen. The results Table 1—Mean, standard deviation, and coefficient
of variation of air content and air-void size
Mixture Air content, % Air-void size, mm
name Sample Mean SD COV Mean SD COV, %
1 72.15 3.29 4.569 0.226 0.0130 6.128
FC-6*
2 71.58 4.33 5.553 0.234 0.0247 10.559
1 62.28 5.69 9.147 0.181 0.0123 6.929
FC-8
2 63.70 3.44 5.631 0.177 0.0082 4.967
1 56.36 3.48 5.944 0.160 0.0116 6.839
FC-10
2 57.03 1.96 3.267 0.154 0.0092 5.338
1 42.32 5.05 11.94 0.121 0.0102 9.134
FC-13
2 41.62 1.53 3.668 0.118 0.0065 6.234
1 26.90 5.54 12.139 0.110 0.0091 8.857
FC-16
2 25.96 2.73 10.247 0.113 0.0073 7.528
1 11.47 3.00 12.881 0.099 0.0174 13.681
FC-19
Fig. 1—Typical prepared specimen used to measure air- 2 11.82 1.53 12.342 0.102 0.0135 11.862
*
void system. FC-6 = foamed concrete with designed density of 600 kg/m3.

ACI Materials Journal/January-February 2006 47


weighted average void size of each test specimen, accordingly. A total of eight configurations were analyzed for r/R ratios
The spacing factor was computed from Eq. (1) to (3) using ranging from 0 to 0.95. The model was meshed with 20-noded
the air content, paste content, and specific surface area quadratic brick elements, as shown in Fig. 3. The top and
obtained from the following equations, respectively bottom faces of the model were constrained for a symmetry
boundary condition while the six vertical faces were not
4 3 restrained. This would provide an upper-bound solution, as
--- πr in reality, the boundary conditions of the vertical faces
π r 3
A = ---------------- = ----------  ---
3
(7) would range from free to fully restrained.
4 2R
3
3 2  R The reduction in the strength of the foamed concrete due
to the air-void system may be assumed to be directly
p=1–A (8) governed by the resulting stress concentration. To determine
this stress concentration factor in relation to the applied
2 stress and the corresponding elastic modulus, a linear elastic
4πr 3
α = ----------- = --- (9) analysis would be adequate. A prescribed displacement
4 3 r
--- πr equivalent to 0.1% strain was first applied and, from the
3 analysis, the equivalent applied stress was determined. The
maximum Von Mises stress was compared with the applied
stress to deduce the stress concentration factor while the
overall modulus of elasticity of the model was determined
from the applied stress and prescribed strain of 0.1%. The
paste modulus of elasticity and Poisson’s ratio of 22 GPa and
0.2, respectively, had been used as the material properties in
the analysis. The results obtained from the numerical analysis
are provided in Table 2.

RESULTS AND DISCUSSION


Air-void system
The details of the air-void system such as air content,
weighted average air-void size, specific surface area, and
spacing factor obtained by LTM are given in Table 3. The air
Fig. 3—Model meshed with 20-node brick element. contents measured from the adopted technique slightly
under- and over-estimated the theoretical air contents at
higher and lower values, respectively. The resultant of high
measured air contents in low-air-content mixtures may be
due to the accidental trapping of air (entrapped air) while
mixing. COVs up to 13 and 14% were observed in the air
content and average air-void size measurements, respectively
(Table 1). No criteria, however, have been set so far for the
acceptance of air-void system measurements by LTM.27
The relationship between average air-void size and air
content is shown in Fig. 4. Although the properties of the
foam introduced into the foamed concrete and the paste
viscosity was uniform for all mixtures, it can be observed
that the average size of the voids varied with the air content
of the concrete. Concrete with higher air content tends to
result in larger air voids because of the proximity of the air
Fig. 4—Relationship between air content and average air- voids, which leads to higher incidence of void coalescing and
void size. forming larger air voids. This observation is more pronounced

Table 2—Numerical modeling results


Spacing factor L for different average air-void sizes, mm
Spacing
r/R ratio Air content, % factor, mm 0.230 0.179 0.157 0.120 0.112 0.101 σapplied /σmises* E/Ep†
0 0 — — — — — — — 1 1
0.25 1.157 0.389 0.597 0.464 0.407 0.311 0.291 0.262 0.717 0.979
0.5 9.256 0.157 0.241 0.187 0.164 0.126 0.117 0.106 0.590 0.84
0.75 31.239 0.055 0.084 0.066 0.058 0.044 0.041 0.037 0.374 0.529
0.85 45.475 0.030 0.046 0.036 0.031 0.024 0.022 0.020 0.252 0.370
0.9 53.981 0.021 0.033 0.025 0.022 0.017 0.016 0.014 0.187 0.288
0.93 59.561 0.017 0.026 0.020 0.018 0.014 0.013 0.011 0.151 0.241
0.95 63.487 0.014 0.022 0.017 0.015 0.012 0.011 0.010 0.131 0.213
*
σapplied = applied stress; σmises = von Mises stress.
†E = modulus of elasticity; E = modulus of elasticity of paste.
p

48 ACI Materials Journal/January-February 2006


in concrete with air content of more than 42%. It is apparent is reduced to 0%. The aforementioned phenomenon deduces
that when the paste content is less than 58%, the average air- that the spacing factor and, hence, the paste content, is critical
void size increases because there is less cement paste to to sustain the air-void size.
prevent the air voids from coalescing.
Figure 5 shows the relationship between the average air Mechanical properties
content and air-void frequency, which is simply the number The strength-to-weight ratio is one of the important
of air voids encountered per unit length traversed. The air- parameters that can be assumed as a measure of economy of
void frequency can be interpreted as a measure of the density the mixture while selecting suitable material for specific
of air void, that is, the number of air voids per unit volume application. A higher ratio would indicate a better returns as
of foamed concrete. It can be noted that the air-void less material had been used for the same strength or same
frequency initially increased with an increase in air content, amount of material now gives higher strength. In this study,
but at air contents higher than 42%, the air-void frequency the ratio of compressive strength over density and modulus
decreased with an increase in air content. In conjunction with of elasticity over density was considered for the discussion
the observation in Fig. 4, it seems that the air-void system is with respect to spacing factor, as shown in Fig. 7. An s-curve
affected disparately by air content and there exists a distinct relationship is observed for both curves. When the spacing
transitional air content that demarcates the trend clearly, factor increases up to 0.04 mm, the corresponding increase
which, in this study, is approximately 42%. This transitional in the strength and modulus ratio was significant. Likewise,
air content would depend on the viscosity of the cement
paste and the properties of the foam added before mixing.
Powers28 and Mielenz et al.29 reported that the coalescence
of air voids in air-entrained concrete may be due to the differ-
ence of surface tensions in different size of bubbles creating
difference of pressure. If the water surrounding a small bubble
should become saturated with respect to the pressure in the
small bubble, it will become supersaturated with respect to the
water surrounding a large bubble and, subsequently, one
should expect air to diffuse through the water from a smaller
to the larger bubble, diminishing the smaller and enlarging the
larger. This diffusion seems to be more dominant in lower-
density mixtures that contain lower paste content.
To understand the circumstances that lead to the coalescence
of air voids, the spacing factor was plotted against the
average void size, as shown in Fig. 6. The figure consists of
the most practical range of concrete densities γw ranging Fig. 5—Relationship between air content and air-void
between 600 to 1900 kg/m3. Interestingly, an asymptotic frequency.
tendency was observed for both the spacing factor and the
average air-void size with apparent limiting values of 0.02
and 0.1 mm, respectively. As spacing factor is a measure of
the mean paste thickness surrounding the air void, this lower
limit of 0.02 mm would be the threshold for the paste at
which the air void could be sustained without coalescing.
Therefore, at lower air content, the spacing factor was
sufficiently large to prevent the coalescence of air voids and
hence the change in average air void size was relatively
small. As the spacing factor approaches the threshold value,
however, a further increase in air content causes the air voids
to coalesce and stabilize at the threshold value of the spacing
factor. On the other hand, the average air-void size asymptotes
toward a threshold value of 0.1 mm, which would then be the
average air-void size of the foam introduced into the
concrete. This is also evident from Fig. 4, where the average Fig. 6—Relationship between average air-void size and
air-void size approaches a value of 0.1 mm when air content spacing factor.

Table 3—Mechanical properties and air-void system of foamed concrete


Mixture Wet density, 28-day fcu , Weighted average Air-void Spacing Specific surface
name kg/m3 MPa E, MPa Atheoretical, % Ameasured, % air-void size, mm frequency factor, mm area, 1/mm
FC-6 606 2.18 0.98 73.33 71.86 0.230 3.26 0.0209 18.13
FC-8 812 4.21 2.02 63.28 62.99 0.179 3.45 0.0260 21.92
FC-10 1033 7.98 4.42 53.23 56.69 0.157 3.61 0.0320 25.08
FC-13 1310 20.81 8.16 38.16 41.97 0.120 3.77 0.0420 36.15
FC-16 1620 36.82 14.03 23.09 26.43 0.112 2.64 0.0698 38.97
FC-19 1935 58.80 20.84 8.03 11.64 0.101 1.61 0.1438 48.78
Note: fcu = cube compressive strength; E = modulus of elasticity in compression; and A = air content.

ACI Materials Journal/January-February 2006 49


as shown in Fig. 6, when the spacing factor increases up to be different for foamed concrete made with different w/c and
0.04 mm, the average air-void size reduces significantly, binder contents. Further studies are required on suitable air-
which thereby contributes to the significant increase in the void system for different w/c and binder types.
strength and modulus ratio. As the spacing factor increases According to Hoff’s14 and Kearsley and Wainwright’s13
further from 0.04 to 0.14 mm, the strength and modulus ratio observations, the strength of cellular concrete varied with
did not increase significantly. This could be due to the small porosity, which is directly proportional to the density. The
change in the average air-void size when the spacing factor relationship between dry density versus compressive
increases from 0.04 to 0.14 mm, as shown in Fig 6. It is strength and spacing factor of the foamed concrete are
evident that air-void size in combination with the spacing shown in Fig. 8. The compressive strength increasing with
factor significantly governs the mechanical properties of the an increase in density is well-aligned with the trend reported
foamed concrete. A small air-void size in combination with for foamed concrete by many researchers.6,8,9,11-13 Figure 8
a larger spacing factor would lead to better mechanical prop- also shows that the spacing factor increases correspondingly
erties, and the optimal values of these factors would result in as the compressive strength with an increase in the density. It
optimal strength-to-weight ratio. Figure 7 shows the optimal can also be seen that the rate at which the compressive strength
spacing factor to be 0.04 mm, above which the strength or and the spacing factor increase with density are congruous,
modulus to density ratios did not increase significantly. suggesting that the spacing factor governs the compressive
Interestingly, this optimal value was at the transitional air strength for the foamed concrete made with same w/c.
content of 42%, which demarcated the trends of air-void size Balshin30 has shown that compressive strength of porous
(Fig. 4) and air-void frequency (Fig. 5) shown previously. materials can be closely related to their porosity by the equation
It is known that compressive strength of concrete is
controlled by w/c because it determines the porosity of σ = σ0(1 – P) y (10)
cement paste.4 Compressive strength is also controlled by
the size of the existing voids in the cement paste.15,24,25 The
relationship between strength or modulus to density ratios where σ is the compressive strength of the porous material;
and spacing factor, as presented in Fig. 7, indicates that the σ0 is the compressive strength of the matrix at zero porosity;
spacing factor also controls the mechanical properties of P is the porosity; and y is a constant. Foamed concrete would
foamed concrete produced with the same w/c. From these contain large air voids due to the introduction of foam and
observations, it can be concluded that the air-void system small pores inherent in the paste. Because in this study only
with a spacing factor, average air-void size, and air content the air content varies and is also relatively much larger that
of 0.04 mm, 0.12 mm, and 42%, respectively, seem to be the the pore structure in the paste, which in this case is assumed
optimal values at which high-strength foamed concrete can to be unchanged for all the mixtures, Eq. (10) may be used
be produced in relation to density. These optimal values may by replacing the porosity with air content. Figure 9 shows the
relationship between the compressive strength and the air
content of the foamed concrete, from which the relationship
can be obtained as

σ = 1.262σp(1 – A)2.962 (11)

where σ, σp, and A are the compressive strength of foamed


concrete, compressive strength of the cement paste, and air
content, respectively.
The exact form of the relation may be uncertain, as the
strength of foamed concrete may be influenced not only by
the entrained air from the preformed foam, but also by
entrapped air, capillary pores, gel pores (if any are present), and
the shape and size of the pores and its distribution,4 similar to
Fig. 7—Variation of compressive strength of modulus of normal concrete. Nevertheless, Eq. (11), which is analogous
elasticity to density ratio with spacing factor.

Fig. 8—Relationship between dry density versus compressive Fig. 9—Relationship between compressive strength and air
strength and spacing factor. content.

50 ACI Materials Journal/January-February 2006


to the Balshin equation, exhibited a better relationship between contents (4.3 to 9.3%),17-22 but an air content of 6% has
air content and compressive strength with a correlation been recommended.21
coefficient of 0.997.
The inclusion of air voids in foamed concrete also has a Numerical analysis
greater effect on strength than on modulus of elasticity, Some of the observations above may seem to be particular
which was observed in relation with strength (72 MPa) and only to the mixtures in this study. To generalize, a numerical
modulus (22 GPa) of paste. This observation was verified in analysis was carried out. Figure 10 and 11 shows the plots of
the following numerical analysis. the ratios of the maximum Von Mises and applied stress, and
It is also noteworthy that the foamed concrete in this study overall and paste modulus of elasticity versus the spacing
would have very good resistance against frost damage factor, respectively. The strength and modulus ratios increase
because the spacing factors are all within the limit of 0.2 mm with the increase in spacing factor, suggesting that the
recommended for good frost resistance.31 Alternatively, this spacing factor controls the mechanical properties. It can also
spacing factor can also be achieved with different air be observed in the numerical analysis that when the spacing
factor is larger than 0.04 mm, the strength and modulus ratio
decreased at slower rate, similar to the experimental
observations (Fig. 7), notwithstanding the fact that only
single-sized bubbles had been used in the numerical analysis.
The experimental results of compressive strength and
modulus of elasticity was also superimposed on the plots, as
shown in Fig. 10 and 11, respectively. The compressive
strength and modulus of elasticity plotted in Fig. 10 and 11,
respectively, had been normalized with the paste compressive
strength and modulus of elasticity for relative comparison
with the numerical analysis. The over- and under-prediction
by the numerical analysis in Fig. 10 and 11 at higher and
lower spacing factors, respectively, could be due to two
factors. First, in the numerical analysis, only a single air-void
size had been used, whereas in the actual test specimen, the
Fig. 10—Comparison of numerical analysis with experimental air voids are not single-sized. Second, only linear elastics
results for compressive strength. analysis was carried out to determine the magnitude of stress
concentration. This may not be adequate to emulate the size
effect and the actual failure mechanism accurately. Never-
theless, both numerical and experimental results collectively
show that the introduction of air voids into the concrete has
a greater effect on the compressive strength than the
modulus of elasticity and this is more prominent at lower air
content, which corresponds to a higher spacing factor.
Figure 12 shows the variation of the spacing factor with air
contents. The numerical plot was based on the presumed air-
void arrangement of single-sized bubble adopted in the
numerical analysis. The relative closeness of both plots
suggests that the presumed air-void arrangement may be
representative of the configurations of the air void in actual
samples. Furthermore, the variation in the two plots in Fig. 12
also explains the deviation of the numerical results from the
Fig. 11—Comparison of numerical analysis with experimental experimental results in Fig. 10 and 11.
results for modulus of elasticity.
CONCLUSIONS
Based on the experimental and numerical results presented
in this paper, the following conclusions may be drawn:
1. The effect of air content on the average air-void size and
air-void frequency changes at some transitional air content
for foamed concrete made with the same w/c. The transitional
air content in this study was found to be 42% beyond which
the coalescence of air voids was very significant;
2. The average air-void size and spacing factor increases
significantly when the air content was higher and lower than
the transitional value (42%), respectively. These approach the
threshold values 0.1 mm and 0.02 mm for average air-void
size and spacing factor, respectively. It can also be interpreted
that the initial air-void size introduced into the foamed
concrete is 0.1 mm, whereas the smallest spacing factor that
Fig. 12—Relationship between air content and spacing factor. could be sustained without air void coalescing is 0.02 mm;

ACI Materials Journal/January-February 2006 51


3. The air-void system with a spacing factor, air-void size, Concrete,” ACI JOURNAL, Proceedings V. 64, No. 2, Feb. 1967, pp. 104-110.
and air content of 0.04 mm, 0.12 mm, and 42%, respectively, 9. Tam, C. T.; Lim, T. Y.; Sri Ravindrarajah, R.; and Lee, S. L.,
“Relationship between Strength and Volumetric Composition of Moist-
were found to be optimal at which foamed concrete with a Cured Cellular Concrete,” Magazine of Concrete Research, V. 39, No. 138,
high strength-to-weight ratio can be achieved with the w/c 1987, pp. 12-18.
and the binder used in this study; 10. ACI Committee 213, “Guide for Structural Lightweight Aggregate
4. From both the experimental and the numerical studies, Concrete (ACI 213R-87),” American Concrete Institute, Farmington Hills,
Mich., 1987, 27 pp.
it was found that the inclusion of air bubbles in foamed 11. Fujiwara, H.; Sawada, E.; and Ishikawa, Y., “Manufacture of High-
concrete had a greater effect on compressive strength than Strength Aerated Concrete Containing Silica Fume,” Fly Ash, Silica Fume,
the modulus of elasticity; Slag, and Natural Pozzolans in Concrete, Proceedings of the Fifth International
5. The relationship between compressive strength σ and Conference, SP-153, V. M. Malhotra, ed., V. 2, American Concrete Institute,
air content A of foamed concrete for high air contents (11 to Farmington Hills, Mich., 1995, pp. 779-791.
12. Kearsley, E. P., “Just Foamed Concrete—An Overview,” Specialist
71%) was found to be as Techniques and Materials for Concrete Construction, Proceedings of the
International Conference, R. K. Dhir, ed., Scotland, 1999, pp. 228-237.
2.962 13. Kearsley, E. P., and Wainwright, P. J., “The Effect of High Fly Ash
σ = 1.262σ p ( 1 – A ) Content on the Compressive Strength of Foamed Concrete,” Cement and
Concrete Research, V. 31, No. 1, 2001, pp. 105-112.
14. Hoff, G. C., “Porosity-Strength Considerations for Cellular Concrete,”
6. From both the experimental and the numerical analysis, Cement and Concrete Research, V. 2, No. 1, 1972, pp. 91-100.
it can be seen that the combined effect of spacing factor and 15. Odler, I., and Robler, M., “Investigations on the Relationship
average air-void size controls the mechanical properties of between Porosity, Structure, and Strength of Hydrated Portland Cement
foamed concrete. Pastes II. Effect of Pore Structure and of Degree of Hydration,” Cement
and Concrete Research, V. 15, 1985, pp. 401-410.
16. Powers, T. C., “Void Spacing as a Basis for Producing Air-Entrained
ACKNOWLEDGMENTS Concrete,” ACI JOURNAL, Proceedings V. 50, 1954, pp. 741-760.
The present work is part of the more comprehensive R&D program titled 17. Pigeon, M.; Plante, P.; and Plante, M., “Air Void Stability, Part I:
“Development of High Strength Lightweight Concretes with and without Influence of Silica Fume and Other Parameters,” ACI Materials Journal,
Aggregates,” jointly funded by Building and Construction Authority of V. 86, No. 5, Sept.-Oct. 1989, pp. 482-490.
Singapore (BCA) and the National University of Singapore (NUS). The 18. Plante, P.; Pigeon, M.; and Saucier, F., “Air Void Stability, Part II:
financial support from BCA, NUS, and Taiheyo Cement is gratefully Influence of Superplasticizers and Cement,” ACI Materials Journal, V. 86,
acknowledged. The assistance and support from Tomita and Nakamura No. 6, Nov.-Dec. 1989, pp. 581-589.
from Taiheyo Cement are also gratefully appreciated. 19. Saucier, F.; Pigeon, M.; and Plante, P., “Air Void Stability, Part III:
Field Test of Superplasticized Concretes,” ACI Materials Journal, V. 87,
NOTATION No. 1, Jan.-Feb. 1990, pp. 3-11.
A = air content 20. Pigeon, M.; Saucier, F.; and Plante, P., “Air Void Stability, Part IV:
COV = coefficient of variation Retempering,” ACI Materials Journal, V. 87, No. 3, May-June 1990,
L = spacing factor pp. 252-259.
N = average number of air voids 21. Saucier, F.; Pigeon, M.; and Plante, P., “Air Void Stability, Part V:
n = total number of lines traversed Temperature, General Analysis, and Performance Index,” ACI Materials
P = porosity Journal, V. 88, No. 1, Jan.-Feb. 1991, pp. 25-36.
p = paste content 22. Khayat, K. H., and Assaad, J., “Air-Void Stability in Self-Consolidating
R = maximum possible radius of air void Concrete,” ACI Materials Journal, V. 99, No. 4, July-Aug. 2002, pp. 408-416.
r = actual radius of air void 23. Schober, G., “Effect of Size Distribution of Air Pores in AAC on
s = standard deviation Compressive Strength,” Advances in Autoclaved Aerated Concrete,
V = volume of specimen Proceedings of the 3rd RILEM International Symposium, F. H. Wittmann,
x = mean of air content or average air-void size ed., Switzerland, 1992, pp. 77-80.
xi = air content or average air-void size of each line length traverse 24. Kearsley, E. P.; and Visagie, M., “Micro-properties of Foamed
y = constant (strength-porosity) Concrete,” Specialist Techniques and Materials for Concrete Construction,
α = specific surface area Proceedings of the International Conference, R. K. Dhir, ed., Scotland,
1999, pp. 173-184.
σ = compressive strength of porous material
25. Toshio, Y.; Yoshinor, K.; Yasuhiko, Y.; and Itoshi, I., “High Strength
σ0 = compressive strength of matrix
Cellular Mortar and Lightweight Concrete,” Japanese Patent No. 112876,
σp = compressive strength of cement paste
May 1991, 10 pp.
26. Wee, T. H., “Recent Developments in High Strength Lightweight
REFERENCES Concrete with and without Aggregates,” Construction Materials: Perfor-
1. Legatski, L. M., “Cellular Concrete,” Special Technical Publication mances, Innovations and Structural Implications and Mindess Symposium,
169B, ASTM International, West Conshohocken, Pa., 1978, pp. 836-851. Proceedings of Third International Conference, N. Banthia, T. Uomoto,
2. Short, A., and Kinniburgh, W., Lightweight Concrete, Third Edition, A. Bentur, and S. P. Shah, eds., Vancouver, British Columbia, Canada,
Applied Science Publishers Ltd., London, 1978, 464 pp. 2005, 97 pp.
3. Widmann, H., and Enoekl, V., “Foam Concrete—Properties and 27. ASTM C 457, “Standard Test Method for Microscopical Determination
Production,” Concrete Precasting Plant and Technology, No. 6, 1991, of Parameters of the Air-Void System in Hardened Concrete,” ASTM
pp. 38-44. International, West Conshohocken, Pa., 1998, 14 pp.
4. Neville, A. M., Properties of Concrete, 4th Edition, Addison Wesley 28. Powers, T. C., The Properties of Fresh Concrete, John Wiley & Sons,
Longman, Essex, England, 1997, 844 pp. Inc, New York, 1967, pp. 380-385.
5. Valore, R. C., “Cellular Concrete—Part 1: Composition and Methods 29. Mielenz, R. C.; Wolkodoff, V. E.; Backstrom, J. E.; and Flack, H. L.,
of Preparation,” ACI JOURNAL, Proceedings V. 51, No. 9, 1954, pp. 50-66. “Origin, Evaluation, and Effects of the Air Void System in Concrete, Part I:
6. ACI Committee 523, “Guide for Cellular Concretes Above 50 pcf, Entrained Air in Hardened Concrete,” ACI JOURNAL, Proceedings V. 55,
and for Aggregate Concretes Above 50 pcf with Compressive Strengths 1958, pp. 95-121.
Less Than 2500 psi (ACI 523.3R-75),” ACI JOURNAL, Proceedings V. 72, 30. Balshin, M. Y., “Relation of Mechanical Properties of Powder Metals
No. 2, Feb. 1975, pp. 50-66. and their Porosity and Ultimate Properties of Porous Metal-Ceramic
7. Narayanan, N., and Ramamurthy, K., “Structure and Properties of Materials,” Dokl. Akad. Nauk. (SSSR), V. 67, No. 5, 1949, pp. 831-834.
Aerated Concrete: a Review,” Cement and Concrete Composites, V. 22, 31. Sommer, H., “The Precision of the Microscopical Determination of
No. 5, 2000, pp. 321-329. the Air Void System in Hardened Concrete,” Cement, Concrete and
8. McCormick, F. C., “Rational Proportioning of Preformed Foam Cellular Aggregates, V. 1, No. 2, 1979, pp. 49-55.

52 ACI Materials Journal/January-February 2006

You might also like