Construction and Building Materials 25 (2011) 658–662

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

Characteristics of lightweight concrete containing mineral admixtures

M.J. Shannag *,1
Department of Civil Engineering, King Saud University, P.O. Box 800, Riyadh 11421, Saudi Arabia

a r t i c l e i n f o a b s t r a c t

Article history: This research investigates the properties of fresh and hardened concretes containing locally available nat-
Received 4 April 2010 ural lightweight aggregates, and mineral admixtures. Test results indicated that replacing cement in the
Received in revised form 19 July 2010 structural lightweight concrete developed, with 5–15% silica fume on weight basis, caused up to 57% and
Accepted 28 July 2010
14% increase in compressive strength and modulus of elasticity, respectively, compared to mixes without
Available online 21 August 2010
silica fume. But, adding up to 10% fly ash, as partial cement replacement by weight, to the same mixes,
caused about 18% decrease in compressive strength, with no change in modulus of elasticity, compared to
Keywords:
mixes without fly ash. Adding 10% or more of silica fume, and 5% or more fly ash to lightweight concrete
Lightweight concrete
Mineral admixtures
mixes perform better, in terms of strength and stiffness, compared to individual mixes prepared using
Stress same contents of either silica fume or fly ash.
Strain Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction associated with the use of conventional LWA produced from clay,
slate and shale in concrete is that these porous aggregates absorb
Lightweight concrete (LWC) has been used for structural pur- a very large quantity of the mixing water [5]. The presence of a
poses for many years. The density of LWC typically ranges from shell structure on the LWA significantly influences the mechanical
1400 to 2000 kg/m3 compared with that of 2400 kg/m3 for normal properties of the lightweight concrete. When the lightweight
weight concrete (NWC). Some of the techniques used for producing aggregate concrete is constituted with stiff aggregates, stresses
LWC include using natural lightweight aggregates such as pumice, are transmitted between cement phases. Failure cracks will extend
diatomite, and volcanic cinders, or artificial by-products such as along the shells of the aggregates similar to that of the normal
perlite, expanded shale, clay, slate, and sintered pulverized fuel weight aggregates. For the soft aggregates or the aggregates with-
ash (PFA). Lightweight concrete has established itself as a suitable out the shell, failure mechanism can be totally different since
construction material whenever the conditions require strict sav- cracks will pass right through the aggregates [6].
ings in the dead-loads in structures and energy conservations The main objectives of this investigation include: (1) developing
and whenever there is an abundance of local lightweight aggre- lightweight concrete (LWC) mixes suitable for structural applica-
gates [1–5]. The demand for lightweight concrete in many applica- tions using locally available materials, (2) studying the properties
tions of modern construction is increasing, owing to the advantage of the LWC mixes developed, including workability, density, com-
that lower density results in a significant benefit in terms of load- pressive and tensile strength, and (3) studying the compressive
bearing elements of smaller cross sections and a corresponding stress–strain behavior of the LWC mixes developed.
reduction in the size of the foundation. However, despite the ef-
forts to improve the strength/weight ratio and versatility of struc-
tural lightweight concrete (SLWC), more research is needed for 2. Experimental program
exploring the potential application of this important building
The experimental program focused on investigating the properties of fresh and
material in structural design [6–12]. hardened concretes containing locally available natural lightweight aggregates, and
Lightweight aggregate (LWA), can be used for making masonry mineral admixtures. A total of ninety-nine 100 mm cubes, and sixty-six
blocks, wall panels, precast concrete elements, structural in situ 100  200 mm cylinders were cast to measure the density, compressive and split-
concrete, screeding and cladding. Its presence in concrete reduces ting tensile strengths, and stress–strain diagram in compression.
the dead weight of structure. The cellular structure of the aggre-
gate gives thermal insulation properties. One of the main problems
2.1. Materials

The materials used in this investigation include lightweight aggregates (LWA),

* Tel.: +966 1 467 6928; fax: +966 1 467 7008. cement, silica sand, and admixtures. The LWA’s were volcanic tuffs of scoria origin
E-mail addresses: mjshanag@ksu.edu.sa, mmshannag@just.edu.jo available in the outskirts of Al-Madina, Saudi Arabia. Fig. 1 shows enlarged photos
1
On leave from Jordan University of Science and Technology, Jordan. for the natural LWA used in this investigation. The physical properties of the aggre-

0950-0618/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.conbuildmat.2010.07.025
 M.J. Shannag / Construction and Building Materials 25 (2011) 658–662 659

gates were determined following ACI and ASTM standards [1,13], as shown in Table
1a. A brief summary of the properties of other materials used, is presented in Table
1b.

2.2. Mix proportions

The absolute volume method, ACI 211 [3], was used for designing the basic con-
crete mix. The final mix was optimized for workability, density and strength, using
the following ingredients: cement, silica sand, natural lightweight coarse and fine
aggregates, silica fume, fly ash, high range water reducers, and water. After casting
many trial mixes, and making necessary adjustments, the concrete mix that
achieved relatively a good degree of workability, minimum density and an accept-
able level of strength was considered as a basis for further investigation of the effect
of mineral admixtures on the behavior of SLWC. The concrete mixes designed in this
investigation were of similar workability and water to cementitious materials ratio.
They consisted of about 400 kg/m3 of Portland cement with the addition of 0%, 5%,
10% and 15% of silica fume and fly ash by weight of cement. The details of these
Fig. 1. Natural lightweight aggregates used in this investigation.
mixes are listed in Table 2.

2.3. Mixing and casting

Table 1a
Physical properties of lightweight aggregates used. The concrete mixes were prepared using a tilting drum mixer of 0.05 m3 capac-
ity. The interior of the drum was initially washed with water to prevent water
LWCA (Madina) LWFA (Madina)
absorption. The coarse and medium aggregate fractions were mixed first, followed
Density (kg/m3) by adding the amount of water absorbed by the aggregates and allowed to rest for
Loose 965 996 30 min to minimize the variation in the initial slump caused by the high water
Rodded 1071 1040 absorption of lightweight aggregates; then silica sand was added, followed by add-
Density (gm/cm3) ing cement, fly ash, silica fume, and the water containing about 75% of the superp-
Dry 2.04 2.10 lasticizer. One-fourth of the superplasticizer was always retained to be added
SSD 2.1 2.15 during the last 3 min of mixing period. The concrete mixes were poured in cubic
and cylindrical molds, and compacted using a vibration table at low speed. After
Water absorption (%) LWCA LWFA each mold was properly filled the vibration speed was increased to medium speed
Time to ensure sufficient compaction.
10 min 3.2 2.8
30 min 3.7 3.5
2.4. Curing and casting
1h 4.2 3.9
2h 4.7 4.3
After casting, the specimens were covered with wet burlap and stored in the
4h 5.1 5.6
laboratory at 23 °C and 65% relative humidity for 24 h and then demoulded and
24 h 6.9 6.3
placed under water. Each specimen was labeled as to the date of casting, mix used
Cylinder compressive strength of LWA 4.5 MPa and serial number. The specimens were then taken out of water a day before testing
and dried in air.

Table 3
Table 1b
Slump and density of LWC determined in this investigation.
Physical properties of cement, sand, and admixtures used.
Mix no. Slump (mm) Density (kg/m3)
Type of material Density Specific surface area (m2/kg)
(gm/ Fresh Air dry Oven dry
cm3)
1 160 2050 1950 1847
Portland cement (type I) – 3.15 300 2 150 2040 1971 1852
ASTM C 150 [14] 3 90 2025 1946 1854
Silica fume (powder form) 2.2 2000 4 110 2032 1995 1878
Fly ash (powder form) 2.3 500 5 130 2066 1968 1898
Superplasticizer ASTM 1.21 – 6 180 2050 1958 1896
494-type D [14] (liquid 7 180 2053 1947 1854
form) 8 155 2060 1970 1834
Silica sand (natural) 2.6 Fineness modulus: 1.65, water 9 150 2039 1947 1851
absorption: 0.5%, density: 1620 kg/ 10 160 2032 1954 1820
m3 11 160 2030 1935 1817

Table 2
Concrete mix proportions in kg/m3 using natural lightweight aggregates from Madina region.

Mix no. Cement Fly ash Silica fume Silica sand LWCA LWFA Water Superplasticizer (L)
1 400 0 0 200 550 350 250 1
2 380 0 20 199.3 548 348.7 249.1 3
3 360 0 40 198.6 546 347.4 248.2 3
4 340 0 60 200 550 350 250 4
5 380 20 0 199.2 548.2 348.9 249.2 2
6 360 40 0 198.7 546.6 347.8 248.4 1.5
7 360 20 20 198.6 546.3 347.6 248.3 3
8 340 40 20 198 544.6 346.5 247.5 2
9 340 20 40 197.9 544.3 346.4 247.4 4
10 320 40 40 197.3 542.6 345.3 246.6 4
11 320 20 60 199.2 548.2 348.9 249.2 3
660 M.J. Shannag / Construction and Building Materials 25 (2011) 658–662

3. Results and discussion lightweight aggregates, LWA, and mineral admixtures seems to
be feasible. The concrete produced possesses 28 days compressive
The performance of the LWC mixes developed in this investiga- strength of about 22.5–43 MPa with a corresponding air dry den-
tion was evaluated by determining the following properties: den- sity of about 1935–1995 kg/m3 which falls slightly above the ACI
sity, workability, compressive strength, splitting tensile strength, requirements of 1850 kg/m3.
and stress–strain diagram.
3.4. Splitting tensile strength
3.1. Workability
The 28 days splitting tensile strengths and the corresponding
The workability of the concrete mixes cast in this investigation compressive strengths at the same age, for the lightweight con-
was measured using the slump test. The slump test results listed in crete mixes cast in this investigation are listed in Table 4. Majority
Table 3 indicate that most of the LWC mixes showed a slump val- of the splitting tensile strength test results for air dried LWC shown
ues ranging from 90 mm to 180 mm immediately after mixing. The in the table were about 8–9% of the corresponding compressive
larger slump for LWC is desirable in order to account for the grad- strength. This is slightly below the standard range reported in
ual loss in workability, caused by the high water absorption of the the literature of 10% for normal weight structural concrete [3]. This
aggregates, which may occur 1–2 h after mixing, i.e. at the begin- could be due to the cellular structure of light weight concrete that
ning of pouring the concrete in the formwork. To be within the enhanced the initiation and growth of microcracks under tensile
scope of this investigation, the workability of all the LWC mixes loading, and thus resulted in larger decrease in tensile strength
cast, was kept almost the same by changing the dosage of superp- compared to normal weight concrete.
lasticizer whenever needed, in particular for the mixes containing
relatively high percentages of silica fume and fly ash. 3.5. Stress–strain diagrams in compression

3.2. Density The compressive behavior of the LWC mixes proposed in this
investigation can best be understood by plotting the complete
In this investigation, the densities of all the LWC mixes cast, stress–strain response. All specimens were tested under uniaxial
including fresh, air dry and oven dry were determined and pre- compression as shown in Fig. 2, by applying a vertical load gradu-
sented in Table 3. The fresh densities shown in Table 3 indicated ally until they reached complete failure. During the test, the dis-
that most of the LWC mixes made, showed a density varying from placement readings of the vertical LVDTs were recorded with the
2025 kg/m3 to 2066 kg/m3. Since the aggregates in the fresh state corresponding load. The readings of the vertical LVDTs attached
were completely saturated with water, therefore the fresh densi- to the specimen sides with a gage length of 120 mm were used
ties were considerably higher than the corresponding air dry and to record the axial deformation and axial strains at the surface of
oven dry densities as shown in Table 3. Most of the specification the specimen. The axial strains determined were also double
standards classify structural lightweight concrete based on air checked by pasting two electrical strain gages of 60 mm gage
dry density not exceeding 2000 kg/m3 [2]. The air dry density length on the sides of each specimen. The results of all tested spec-
shown in the table varied from 1935 to 1995 kg/m3. It can be no- imens were recorded and analyzed in terms of their axial stress–
ticed that the range of air dry density complies with the European strain curves as shown Figs. 3–6.
specifications for structural LWC of air dry density not exceeding The shape of the stress–strain curves of the LWC tested, can be
2000 kg/m3, but does not meet ACI requirements of air dry density characterised with a linear elastic response up to about 40–50% of
not exceeding 1850 kg/m3. It should be noted that the air dry den- its ultimate load carrying capacity; a curvilinear response up to the
sities can be reduced to meet ACI requirements by making some peak followed by a post peak curvilinear segment of decreasing
adjustments on the composition of the mixes without sacrificing slope. Close to the peak load, vertical hairline cracks started
the structural strength required at 28 days. It can be observed from appearing on the surface of the specimen. The number and width
Table 3 that the oven dry unit weights of the LWC mixes developed of these cracks kept on increasing with further increase in axial
varied from 1817 to 1898 kg/m3. load until they formed a major shear crack at an angle of 45° with
the longitudinal axis of the cylinder. Because of the porous nature
of LWA, the vertical cracks passed through the aggregates, and thus
3.3. Compressive strength

The test results listed in Table 4, indicate that producing struc-

tural lightweight concrete, SLWC, using locally available natural

Table 4
Compressive strength, splitting tensile strength, and modulus of elasticity, of LWC
mixes at 28 days.

Mix Compressive Splitting tensile Modulus of

no. strength (MPa) strength (MPa) elasticity (MPa)
1 29.3 2.75 19,788
2 28.8 3.15 19,343
3 38.0 3.47 20,413
4 43.2 3.18 22,477
5 27.7 2.68 20,795
6 22.5 2.76 19,751
7 32.2 2.91 18,696
8 32.4 2.39 17,457
9 39.0 3.44 20,213
10 33.7 2.86 18,694
Fig. 2. Test setup and instrumentation of the specimen used for determining the
11 36.7 2.64 18,587
complete stress–strain diagram.
 M.J. Shannag / Construction and Building Materials 25 (2011) 658–662 661

50
5% Silica Fume
45 10% Silica Fume
15% Silica Fume
40 0% Silica Fume
Compressive Stress (MPa)

35

30

25

20

15

10

0
0 2000 4000 6000 8000 10000

Strain mm/mm (10-6)

Fig. 3. Effect of silica fume on the compressive behavior of LWC. Fig. 6. Effect of silica fume, and fly ash on the compressive behavior of LWC.

14% in compressive strength and modulus of elasticity respectively

compared to mixes without silica fume. But, adding up to 10% fly
ash, to the same mixes, caused about 18% decrease in compressive
strength, with no change in modulus of elasticity, compared to
mixes without fly ash. The mixes containing 10% or more of silica
fume, and 5% or more fly ash, Figs. 5 and 6, exhibited a considerable
increase in compressive strength and modulus of elasticity com-
pared to individual mixes containing same content of either silica
fume or fly ash.
The increase in the strength of the LWC due to the addition of
silica fume and fly ash may be attributed to the improved aggre-
gate–matrix bond associated with the formation of a less porous
transition zone and a better interlock between the paste and the
aggregate [15,16]. The aggregate–matrix bond improvement in-
duced by these admixtures is probably the result of a combined fil-
ler and pozzolanic effect. The filler effect leads to reduction in
porosity of the transition zone and provides a dense microstructure
Fig. 4. Effect of fly ash on the compressive behavior of LWC. and thus increases the strength of the system. The pozzolanic effect
helps in the formation of bonds between the densely packed parti-
cles in the transition zone through the pozzolanic reaction with the
calcium hydroxide liberated during the hydration of Portland ce-
ment [15,16]. Therefore, it is recommended to use silica fume, fly
ash blends in producing structural LWC without sacrificing
strength and workability by incorporating the required dosage of
superplasticizer.

Fig. 5. Effect of silica fume, and fly ash on the compressive behavior of LWC.

forced longitudinal pieces of the cylinder to split apart. Typical fail-

ure modes of some of the LWC specimens tested are shown in
Fig. 7.
Fig. 3 illustrates that by adding up to 15% of silica fume, to the
mixes containing LWA, caused a significant increase of 57% and Fig. 7. Typical failure modes of LWC cylinders tested under axial compression.
662 M.J. Shannag / Construction and Building Materials 25 (2011) 658–662

4. Conclusions tories for their assistance during the execution of the experimental
program.
Based on the test results of this investigation, the following con-
clusions can be drawn: References

[1] ACI 213-8. Guide for structural lightweight aggregate concrete. ACI 213-87.
1. Lightweight concrete (LWC) mixes suitable for structural appli-
Manual of concrete practice, part 1. American Concrete Institute, Detroit,
cations were developed using locally available natural light- Michigan; 1995. p. 27.
weight coarse and fine aggregates. The mixes developed had a [2] Neville AM. Properties of concrete. 4th ed. New York: John Wiley and Sons Inc.;
compressive strength range of 22.5–43 MPa; an air dry density 1996. p. 844.
[3] Mindess S, Young F, Darwin D. Concrete. 2nd ed. New Jersey: Prentice-Hall;
of 1935–1995 kg/m3; and a high degree of workability. 1996. p. 644.
2. The stress–strain diagrams plotted for the structural LWC mixes [4] Zhang MH, Gjorv OE. Characteristics of lightweight aggregate for high strength
developed were comparable to typical stress–strain diagrams concrete. ACI Mater J 1991;88(2):150–8.
[5] Hossain KMA. Properties of volcanic pumice based cement and lightweight
for normal weight concrete with relatively larger strain capacity concrete. Cem Concr Res 2004;34:283–91.
at failure in case of LWC. [6] Chi JM, Huang R, Yang CC, Chang JJ. Effect of aggregate properties on the
3. Replacing cement with 5–15% silica fume on weight basis for strength and stiffness of lightweight concrete. Cem Concr Compos
2003;25(2):197–205.
LWC, caused up to 57% and 14% increase in compressive [7] Teo DC, Mannan MA, Kurian VJ, Ganapathy C. Lightweight concrete made from
strength and modulus of elasticity respectively compared to oil palm shell (OPS): structural bond and durability properties. Build Environ
mixes without silica fume. But, adding up to 10% fly ash, as par- 2007;42:2614–21.
[8] Babu KG, Babu DS. Behavior of lightweight expanded polystyrene concrete
tial cement replacement by weight, to the same mixes, caused containing silica fume. Cem Concr Res 2003;33:755–62.
about 18% decrease in compressive strength, with no change [9] Chen B, Liu J. Experimental application of mineral admixtures in lightweight
in modulus of elasticity, compared to mixes without fly ash. concrete with high strength and workability. Constr Build Mater
2008;22(4):655–9.
Adding 10% or more of silica fume, and 5% or more fly ash
[10] Waldron CJ. Demonstration of use of high-performance lightweight concrete
caused a considerable increase in compressive strength and in bridge superstructure in Virginia. J Perform Constr Facil 2005;19(2):146–54.
modulus of elasticity compared to individual mixes containing [11] Delsye CL, Abdul Mannan MD, Kurian JV. Flexural behavior of reinforced
same contents of either silica fume or fly ash. lightweight concrete beams made with oil palm shell (OPS). J Adv Concr
Technol 2006;4(3):1–10.
[12] Kayali O, Haque M, Zhu B. Some characteristics of high strength fiber
reinforced lightweight aggregate concrete. ACI Mater J 1990;81(2):199–204.
[13] ASTM C 33-97. Standard specification for concrete aggregates. Annual Book of
ASTM Standards, vol. 4.02. p. 10–6.
Acknowledgements [14] ASTM C 150-97a. Standard specification for Portland cement. Annual Book of
ASTM Standards, vol. 4.01. p. 134–7.
The author wishes to acknowledge the financial support pro- [15] Malhotra VM, Carette GG, Sivasundaram V. Role of silica fume in concrete. In:
Proceedings of the international conference on advances in concrete
vided by the Research Center at the College of Engineering of King technology, CANMET, Ottawa, Canada; 1992.
Saud University. He is also grateful to the engineers and the tech- [16] Popovics S. Portland cement-fly ash silica fume systems in concrete. Adv Cem
nicians at the research center, and concrete and structural labora- Based Mater J 1993;1:83–91.