Construction and Building Materials 66 (2014) 247–258

Contents lists available at ScienceDirect

Construction and Building Materials

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

Effect of substitution of wood shavings by barley straws

on the physico-mechanical properties of lightweight sand concrete
B. Belhadj a, M. Bederina a,⇑, N. Montrelay b, J. Houessou b, M. Quéneudec b
a
Structures Rehabilitation and Materials Laboratory, Civil Engineering Department, University A. Telidji of Laghouat, Algeria
b
Research Unit EPROAD (Eco-procédés, optimisation et aide à la décision), University of Picardie Jules Verne, Amiens, France

h i g h l i g h t s

 Resolution of problems posed by wood sand concrete.

 Substitution of wood shavings by barley straws in sand concrete.
 Existence of strong relations between the thermo-physical and mechanical properties.
 The presence of straws and wood together is better than the existence of each of them alone.
 Obtained results have been confirmed by microstructure study.

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

Article history: The paper mainly aims to study the effect of the incorporation of barley straws and wood shavings on the
Received 18 March 2014 physico-mechanical properties of sand concrete intended for the construction in arid zones. Basing on the
Received in revised form 25 May 2014 study at fresh state, the experimental work has been firstly devoted to searching for optimal composi-
Accepted 27 May 2014
tions of the composite. Then, a second study has focused on the physico-mechanical characterization
Available online 14 June 2014
of sand concrete without lignocellulosic materials. Finally, a special experimental program has been
devoted to the study of sand concrete with lignocellulosic materials. The obtained results show that it
Keywords:
is quite possible to develop a lightweight sand concrete by the incorporating, separately and in combina-
Sand–concrete
Lignocellulosic materials
tion, of barley straws and wood shavings. Moreover, it has been noted that the combined addition of
Eco-materials barley straws and wood shavings gives the best physico-mechanical results which were strongly con-
Quaternary binder firmed by a microstructure study. With combined addition, the problem of shrinkage was relatively
Wood shavings solved: decreases of about 11.76% and 39.02% were recorded in comparison with the cases where barley
Barley straws straws and wood shavings are taken separately. In addition, the thermal diffusivity has been reduced of
Mechanical properties about 35.47% with respect to the basic composition. Due to its higher deformability, the barley straw
Thermal properties improves the toughness and ductility of sand concrete and reduces its dimensional variations. Indeed,
Shrinkage
the studied composites, offer to the local constructions important technical and economic interests: good
thermal insulation and efficient energy saving and low cost of materials.
Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction
For climatic conditions in arid environments, new lightweight
concretes intended for the construction of external walls and roofs
are really needed to provide a better thermal resistance and to
compete with the traditional construction materials. It is probable
that, if it is easily maintained with highly reflective external finish,
⇑ Corresponding author. Address: Civil Engineering Department, University Amar
Telidji, BP 37G, Laghouat 03000, Algeria. Tel.: +213 773732230.
E-mail addresses: mdbederina@yahoo.com, m.bederina@mail.lagh-univ.dz
(M. Bederina).
a new lightweight concrete might fare better than current materi-
als (block, brick, etc.). Indeed, the rapid economic development in
several countries characterized by an arid climate, such as Algeria,
has led to the construction of a phenomenal number of new
buildings during the past 10 years. In response to the severe
summer and winter weather conditions, these new buildings are
universally heated or cooled by means of mechanical systems.
Indeed, although some thermal benefits are likely gained from
the contemporary use of current building envelope elements, the
energy burden imposed by the continued construction of thermally
inefficient building stock is of a great concern to these countries.
Energy burden is an important statistic for policy makers who

http://dx.doi.org/10.1016/j.conbuildmat.2014.05.090
0950-0618/Ó 2014 Elsevier Ltd. All rights reserved.
248 B. Belhadj et al. / Construction and Building Materials 66 (2014) 247–258
are considering the need for energy assistance. Moreover, the
environmental consideration requires the preservation of mineral
aggregates and the use of vegetal waste [1]. A large amount of lig-
nocellulosic waste coming from various sources is generated
worldwide which causes various environmental problems. The
use of such waste in the manufacture of cementitious building
materials can resolve these problems [2]. However, the formula-
tion of lightweight concrete based on local lignocellulosic materi-
als is a way of contribution for sustainable development in the
construction industry [1]. Therefore, the valorisation of local mate-
rials as well as agro-industrial waste could make an important con-
tribution where a consensus is necessary for improving the
performance and behavior of materials while minimizing costs as
well as energy consumption [3]. It’s in these two contexts that a
research program have been established to investigate the phys-
ico-mechanical characteristics of a new concrete consisting of local
materials and waste. It is noteworthy that, sand concrete based on
wood shavings has already shown an interesting lightness; how-
ever, its dimensional variations are very important. In order to
solve this problem, several studies have been performed using
different treatments of wood shavings [4] in order to reduce the
shrinkage to a value less than the target value generally recom-
mended (1 mm/m) [5]. That is the reason for which the assumption
to partially or totally substitute the wood shavings (coming from
woodwork activity waste) by another local lignocellulosic material,
characterized by a renewable and ecological character, is adopted.
By definition, lignocellulosic materials are composed of the combi-
nation of lignin, cellulose and hemicellulose in the structural cells
of woody plants [2]. Currently, the barley straws are excessively
produced compared to their use, i.e. it is a natural product, inex-
pensive and easy to obtain in the majority of countries. Thus, the
reuse of this material in the construction has significant interests
and leads to economical, ecological and sustainable development
[6]. Locally, around the city of Laghouat (south of Algeria), there
are two main types of straws namely barley and wheat. Because
it presents better thermal insulation than wheat straw, the barley
straw was selected to be used in this study [7]. On the other hand,
sand concrete, which is used as matrix in this case, is defined as a
concrete which contains only sand as aggregates (without any
gravels) and a cement content similar to that of ordinary concretes
[8,9]. Thus, the incorporation of lignocellulosic materials in cement
matrix aims to obtain a new lightweight concrete which is an
important topic of research because of the interesting thermo-
mechanical properties expected [10]. According to literature,
several works have studied the incorporation of lignocellulosic
materials in composites. Bouhicha et al. used barley straws in order
to reinforce soil composites [11]. Li et al. used hemp fibers in
cementitious composite [12]. Yang et al. used rice straws fibers
and wood particles at different ‘straw/wood’ ratios in wooden con-
struction materials [13]. Soroushian et al. used wheat straws
instead of wood in cementitious composites [14]. As for Ashour
et al., they used wheat and barley straws [7]. While Nozahic
et al. used two aggregates of plant, the first was of hemp fibers
and the second was the sunflower stem which has been selected
as an alternative for byproduct of the lignocellulose due to its
apparent similarities with hemp fibers [15]. Furthermore, Merta
and Tschegg used hemp fibers, miscanthus grass and wheat straws
[16]. Other authors used untreated and treated wood shavings
[4,17–19]. Concerning the present work, a comparative study
between the properties of four types of sand concretes: a sand con-
crete without lignocellulosic materials (SC-W-LM) which is consid-
ered as reference concrete, a sand concrete with wood shavings
(SC-WS) where content of wood is 60 kg/m3, a sand concrete with
barley straws (SC-BS) and finally a sand concrete with both barley
straws and wood shavings (SC-BS-WS) is provided. The only
respected condition during this work is that, at fresh state, the last
three concretes (sand concretes with lignocellulosic materials)
have an equal workability (an Abrams cone slump value of 6 cm).
In order to find the appropriate compositions, several preliminary
experiments were performed in order to determine the optimal
proportions of used lignocellulosic materials. Moreover, it should
be noted that a quaternary binder obtained by the partial substi-
tuting of cement by three mineral additives was used in order to
raise the economic level of the sought composites. It has been
reported that the present work is only a continuation of previous
studies and the effect of the quaternary binder is not the purpose
of this paper. Finally, the main objective of this work is to study
the contribution of two lignocellulosic materials, which are barley
straws and wood shavings, to the development of a lightweight
eco-sand–concrete based on river-dune sand and quaternary bin-
der. In fact, this work deals with search of optimal compositions
of insulating-bearer lightweight eco-sand–concrete intended for
the construction in arid environments.
2. Materials and experimental methods
2.1. Raw materials
The studied composites are composed of sand concrete as matrix and barley
straws or/and wood shavings as inclusions. The sand concrete consists of a mixture
of sand, cement, fillers and water. Besides these basic components, it typically
includes one or more admixtures [9]. The various used components introduced dur-
ing this work are characterized hereafter.
The used sand (RDS) is a local material which consists of a mixture of two other
local sands, river sand (RS) and dune sand (DS). Both of these sands are character-
ized by a continuous particle size distribution ranging from 0 to 0.63 mm and 0 to
5 mm respectively, contain a fraction of grains smaller than 80 lm below 5% and
present rounded shapes of grains (Fig. 1). The high values of the ‘‘Sand Equivalent’’
show that RS and DS are clean and the X-ray Diffraction analysis demonstrates their
essentially siliceous nature [20]. So the only difference between these two sands
lies in their fineness modulus values; RS presents more coarse grains while DS pre-
sents more fine grains. This is why the fine part of the RS grading curve was cor-
rected by the addition of DS according a mass ratio: (RS/DS) = 1.7 [20]. Table 1
lists the set of physical characteristics of all these sands. These results reveal that
RDS is the denser and the most compact sand. Its modulus of fineness is 2.81.
The fact that RS and SD are clean and of siliceous nature, RDS is therefore clean
and siliceous. The visualization of the general appearance of sand grains shows that
they are rounded with very low reasons of sustainable construction, the cement
used is a Portland limestone cement (PLC) of type CPJ CEM/II A-L 42.5 R. This type
of binder is a new class of cement that provides a similar performance to that of
conventional Portland cement but emits less CO2 (a reduction of up to 10%). Let
us note that the use of PLC in North America is a proactive action to help mitigate
climate change through energy savings and reduced emissions. The physical char-
acteristics of the used cement are as follow: apparent density: 1030 kg/m3, specific
density: 3030 kg/m3 and Blaine specific surface area: 444.9 m2/kg. Moreover, the
chemical analysis revealed the existence of the chemical products shown in Fig. 2
with the proportions presented in Table 2. On the other hand, the use of fillers in
sand concrete composition is essential [8]. Their use is intended to complete the
grading curve of sand in its finest part in order to obtain more compact concrete
and reduce the cement content and therefore the cost of concrete. In this work,
the used fillers have been obtained by sifting local crushing waste (to a sieve
opening of 80 lm). It enters also within the scope of waste recovery. The X-ray
100
90
DS
RS
Cumulative passing (%)
RDS
80
70
60
50
40
30
20
10
0
0,01 0,1 1 10
Grain diameter (µm)
Fig. 1. Particle size distribution of used sands.
 B. Belhadj et al. / Construction and Building Materials 66 (2014) 247–258 249

Table 1
Physical properties used sands (RS, DS and RDS).

Physical characteristics RS DS RDS

3
Apparent density (kg/m ) 1552 1501 1590
Specific density (kg/m3) 2425 2587 2485
Compactness (%) 0.64 0.58 0.66
Porosity (%) 0.36 0.42 0.34
Fineness modulus 2.89 1.98 2.81
Visual sand equivalent (%) 93.5 90.7 93.2
Sand equivalent (%) 82.3 81.4 80.3

Fig. 3. X-ray diffractogram analysis of used limestone fillers.

Table 3
Physical properties of used mineral additions.

Physical properties Limestone Pozzolan Lime

Blaine specific surface area (m2/kg) 280 388 650
Specific density (kg/m3) 2700 2850 2750
Apparent density (kg/m3) 1530 1005 666

Fig. 2. X-ray diffractogram analysis of used Portland limestone cement.

Diffraction analysis highlights their limestone nature (Fig. 3) [20] and reveals low
percentages of harmful components that can influence the cement hydration. The
physical characteristics of these fillers are shown in Table 3. In addition, the valori-
sation of local materials and waste generates a significant economic interest. To
give more importance to the economic aspect by developing a new concrete with
a lower price, a portion of cement was substituted with cheaper mineral additions.
The latters are limestone, natural pozzolan and lime. Their chemical analyses are
shown in Fig. 3 and their physical properties are summarized in Table 3 [21,22].
The limestone addition is an inert pounder which is the same product used as fillers
(Fig. 3, Table 3). The natural pozzolan is an active addition of volcanic origin
extracted from the deposit of the south region of the city of Beni Saf in Algeria.
The latter is especially an effusive volcanic rock of feldspathic composition (sili-
ceous). After grinding, using Los Angeles machine, the product is sieved through a
sieve of 80 lm. Its chemical analysis is shown in Fig. 4. The third used addition is Fig. 4. X-ray diffractogram analysis of used natural pozzolan.
an Algerian hydraulic lime. Generally, the lime is an active addition. Although it
is soluble in water, it was subjected to sifting to a sieve of 80 lm in order to follow
the same procedure used for the pozzolan and limestone. Its chemical analysis is
shown in Fig. 5. The results thus obtained show that the limestone and lime have
a high concentration of basic elements such as CaO and a low concentration of oxide
elements such as silicates oxides SiO2, A12O3 and aluminates. However, the pozzo-
lan presents a high concentration of SiO2. The wood shavings, which are also used in
this study, fall within the scope of recycling woodwork activity wastes. These shav-
ings derive from the fir trees. This choice is justified by the importance of their use
in the field of building construction. Moreover, the wood aggregates commercial-
ized in France are mostly obtained from softwood or hardwood [5]. In general, fir
shavings have several advantages such as the low density and the low thermal con-
ductivity. They have a particular structure which makes them brittle (Fig. 6). Their
particle size analysis is not significant; nonetheless it has been performed according
to standard (NF P 18-560). From these results, it has been noted that the used wood
shavings have a continuous particle size distribution ranging from 1 to 8 mm
(Fig. 7). Natural wood is a quite complex material. The fir wood used is composed
mainly by three natural polymeric structures, which are cellulose (38.8%), hemicel-
luloses (26.30%) and lignin (29.3%) [23]. The second lignocellulosic material, which
is intended to substitute the wood shavings, is in fact barley straws which are Fig. 5. X-ray diffractogram analysis of used lime.
harvested in 2012. Locally, this material is commercialized in the form of bale of
about 1.08  0.48  0.30 m of dimension. The used straws are characterized by a state, the straws present a high flexibility, i.e., after folding, the straws return to
water absorption coefficient of about 3.30 and a slow absorption kinetic, which is their initial form and length. Their diameter ranges from 1 to 4 mm. The chemical
in contrast with wood shavings whose water absorption kinetics is faster. At wet composition of the barley straw is mainly composed by cellulose (37.6%),

Table 2
Chemical analysis of used cement (%).

SiO2 CaO MgO Al2O3 Fe2O3 SO3 K2O Cl Na2O Free CaO Loss Ins.
16.93 62.23 1.03 5.26 2.82 2.89 0.65 0.02 0.04 1.782 7.83 1.61
250 B. Belhadj et al. / Construction and Building Materials 66 (2014) 247–258

(a) G=50 (b) Normal aspect

Fig. 6. General aspect of wood shavings.

100
90 as their, cover are also shown in Fig. 8. In order to have an idea about the rigidity of
the used barley straws, tensile tests were made on tubular straws whose diameter
Cumulative passing (%)

80
is about 2.5 mm using a Universal Testing Machine (Fig. 9a). The obtained
70 results are thus shown in Table 4. Fig. 9 likewise shows an example of a ‘‘force–
60 displacement barley straw’’ curve. These results demonstrate the good strength
and especially the good deformability of the material. The last used material, whose
50 use is essential in the case of sand concrete, is an adjuvant of type ‘‘MEDAPLAST SP
40 40’’, in conformity with the EN 934-2. It is a high range water reducer superplast-
icizer which allows the obtaining of a high quality concrete and mortar of high
30
quality. In addition to its main function of superplasticizer, it greatly reduces the
20 water content. This product is presented in liquid form that is easily diluted in
10 the mixing water. The used proportion is 2% (in mass) of the binder content.

0
0,1 1 10 100
2.2. Elaboration of composite
Grain diameter (mm)
The optimal composition of the matrix (sand concrete without lignocellulosic
Fig. 7. Particle size distribution of used wood shavings.
materials) (Table 5) is obtained in previous works, in particular those of Bederina
et al. [20]. Indeed, this basic composition (SC-W-LM) is taken as reference sand con-
hemicelluloses (34.9%) and lignin (15.8%) [24]. The chosen straw lengths are 3.5 cm crete. The percentage of mineral additions, added by mass substitution in the
for 7  7  28 cm prism and 2 cm for the 4  4  16 cm prism. Their shape, as well cement, is 10% (1/3 limestone, 1/3 natural pozzolan and 1/3 hydraulic lime).

Fig. 8. General aspect of used barley straws.

 B. Belhadj et al. / Construction and Building Materials 66 (2014) 247–258 251

(a) Tensile test led on of barley straw.

(b) Obtained load-displacement curve of barley straw.

Fig. 9. Mechanical characterization of barley straw.

1,4
Table 4
SC-WS
Obtained results of tensile test of barley straw.
Shrinkage at 90 days (mm/m)

1,2
Maximum stress (MPa) 115.38 ± 6.21
Displacement (mm) 0.863 ± 0.093 1,0
SC-BS
Elasticity modulus (MPa) 9917.19 ± 931
0,8 SC-W-LM SC-BS-WS

0,6

Among, the parameters taken into account in calculating the mixtures, it can be
0,4
cited: ‘‘S/LM’’ ratio (sand/lignocellulosic materials) and ‘‘WT’’ (total content of
water). The ‘‘S/LM’’ ratio is calculated by absolute volumes using Eq. (1):
0,2

V S0 ¼ V SN þ V LM ð1Þ
0,0
1 2 3 4
where VS0 is the volume of basic sand (Case without lignocellulosic materials), VNS
Sand concrete type
the new volume of sand (Case with addition of lignocellulosic materials) and VLM
the volume of lignocellulosic materials.
Fig. 10. Shrinkage of studied sand concretes measured at 90 days.
The real density of wood shavings can be assimilated to the apparent density of
massive wood [25,26]. The absolute density of the barley straws, which is also trea- W T ¼ W M þ W AbS ð2Þ
ted as fir wood, is of about 506.28 kg/m3. The proportions of straws considered in
this work vary from 0 to 20 kg/m3 with an interval of 5 kg. The total amount of where WT is total amount of water in the composition, WM amount of water in the
water in the composition is calculated by the Eq. (2): initial mixture (210 l/m3) and Wabs amount of water absorption of the straws.

Table 5
Compositions of studied sand concretes.

Sand concrete S (RDS) (kg/m3) C (kg/m3) MA (kg/m3) F (kg/m3) WT (l/m3) SPa (%) LM (kg/m3)
BS WS
SC-W-LM 1316 315 35 135 210 2 0 0
SC-WS 1030 315 35 135 288 2 0 60
SC-BS 1269 315 35 135 223 2 10 0
SC-BS-WS 1150 315 35 135 255 2 5 30
a
Mass percentage relative to the amount of cement.
252 B. Belhadj et al. / Construction and Building Materials 66 (2014) 247–258

0,9 Relating to the mixture ‘‘straw tube’’ and covers (broken straws) (Fig. 8), the
Y = 0,60082 + 0,00366.x proportions of each of them were determined basing on a series of preliminary
WS
R = 0,99985 tests. It has been found that the used quantity of straws is constituted from 70%
0,8 of tubular straws and 30% of covers. In order to better homogenize the mixture,
BW which contributes to a good quality of concrete and therefore a good control of
the different properties of the finished product, the mixture is performed by a mixer
0,7 BS
W/C ratio

BS of controls type according to the following procedure: a dry mixing of cement and
BS
BS mineral additions for 1 min at a low speed, thereafter the sand and the limestone
WLM fillers are added with a dry mixing for three other minutes at low speed. When
0,6
dry mixture is perfectly homogeneous, lignocellulosic materials are introduced in
the saturated state, while continuing mixing for 3 min at low speed. Finally the mix-
0,5 ing water is added gradually to the mixture without interruption in order to ensure
the homogenization of the mixture for 3 min at low speed [17,19]. After mixing, the
material is fed into molds of 4 (4  4  16 cm) for shrinkage tests and molds of 7
0,4 (7  7  28 cm) for the mechanical properties tests. The specimens are demoulded
0 10 20 30 40 50 60
after 24 h and the samples are kept until the day of the test, in the same room
Lignocellulosic material content (kg/m3) where the indoor climatic conditions of the laboratory room is closer to the actual
weather conditions, namely the temperature is (20 °C ± 5 °C), while the relative
Fig. 11. Evolution WT/C ratio according to lignocellulosic material content. humidity of (50% ± 10%). The compositions of the studied sand concretes are deter-
mined to have a constant workability. To control the workability, several experi-
mental tests inspired from the literature works [27,28] have been made. The
procedure is to mix and add water to the material repeatedly until obtaining the
workability value previously fixed. The reduction coefficient of water is around
1.8–2.2 for all the studied compositions, which means a reduction of 2 L from the
2500 initial mixing water for each 1 kg of saturated lignocellulosic material added, due
Y = 2055,59943 - 17,04313.x + 0,10455.x 2 to its favorable effect on workability. After several preliminary workability tests
R = 0,99629
of many compositions, four types of concretes have been selected as shown in
SC-W-LM
Table 5.
Density (kg/m )

SC-BS
3

2000
2.3. Experimental techniques
SC-BS-WS
The workability is tested using Abrams cone according to the standard (NF P
1500 SC-WS 18-451). The shrinkage tests are carried out according to NF P 433 standard. The
test specimens are 40  40  160 mm prisms. Three specimens of each composition
are also developed in order to obtain the average of three data. The shrinkage is
measured using a length comparator with digital. After demoulding at 24 h, an ini-
1000
tial comparator reading is taken for each composition. Then, the measurements
0 10 20 30 40 50 60 continue at different ages until 90 days. The mechanical properties are studied by
3 measuring compressive and flexural strength. The measurements of flexural
Lignocellulosic material content (kg/m )
strength are conducted on 7  7  28 cm prisms and compressive strength on
7  7 cm cubes using a universal press of type ‘‘Controls’’, in accordance with the
Fig. 12. Relationship between the density and the lignocellulosic material content.
standard (EN196-1). Three specimens by composition are also developed in order
to obtain the average of three data of measurement. The measurements of thermal
properties are performed using the Hot Disk TPS 500 based on the principle of the
TPS-technique. For each composition, all thermal parameters are identified in a sin-
gle operation. These parameters are: thermal conductivity, thermal diffusivity and
0,42
specific heat.
Y = 0,75948 - 2,67356E-4.x A Shimadzu IR Prestige-21 accessory equipped with a diamond crystal was used
0,40
R = 0,99911 for all of the spectra shown in this work. The test method was performed using
Porosity accessible to water

0,38 Attenuated total reflectance by absorbance. Attenuated total reflectance (ATR) is a

0,36 sampling technique used in conjunction with infrared spectroscopy which enables
0,34 samples to be examined directly in the solid or liquid state without further
preparation.
0,32
0,30
0,28 3. Results and discussion
0,26
0,24 3.1. Shrinkage
0,22
0,20 Fiber reinforcement improves tensile strength and some other
1300 1400 1500 1600 1700 1800 1900 2000 2100 properties. Among these properties, it can be cited: the shrinkage.
Dry density (kg/m3) The latter is satirically reduced by the fibers use. The quantity of
fibers is small, typically 1–5% by volume. Aspect ratio (length/
Fig. 13. Relationship between the dry density and the porosity accessible to water. diameter) and some other characteristics such as: shape and sur-
face texture, length and structure are significant [29]. Let us note
that in the present paper, the material which is considered as fiber
is the barley straws. The obtained results are shown in Fig. 10.
Table 6 Indeed, sand concrete based on wood shavings (SC-WS) has given
Thermal properties of the studied sand concretes.
shrinkage values of about 1.23 ± 0.04 mm/m which is greater than
Type sand Thermal Specific heat Thermal diffusivity the limit value generally recommended for concretes. This obser-
concrete conductivity (W/ (J/kg K) (m2/s (106)) vation has been already made in previous studies which led certain
m K)
authors to research different treatments for wood shavings before
SC-W-LM 1.40 1209.07 0.578 their use [4]. In this study, the problem is resolved otherwise. It is a
SC-BS 1.35 1209.63 0.592
substitution of wood shavings by barley straws. Indeed, the
SC-WS 0.80 1876.76 0.311
SC-BS-WS 1.00 1821.19 0.373 replacement of wood shavings by barley straws (SC-BS-WS and
SC-BS) has significantly solved this problem. A partial replacement
 B. Belhadj et al. / Construction and Building Materials 66 (2014) 247–258 253

0,65 0,0000007
Y = - 2,36257E-6 + 2,96941E-9.x - 7,47462E-13.x 2
Y = - 0,09529 + 0,49124.x R = 0,99091
0,60 SC-BS
R = 0,98769 0,0000006

Thermal diffusivity (m2/s)

SC-W-LM
Thermal diffusivity (m /s)
2

0,55

0,0000005
0,50

0,45
0,0000004 SC-BS-WS

0,40 SC-WS

0,0000003
0,35

0,30 0,0000002
0,7 0,8 0,9 1,0 1,1 1,2 1,3 1,4 1,5 1300 1400 1500 1600 1700 1800 1900 2000 2100
Thermal conductivity (W/m.K) Dry density (kg/m3)

Fig. 14. Thermal diffusivity versus thermal conductivity and dry density.

5 25
Y = 13,49122 - 0,01514.x + 5,196E-6.x 2 SC-W-LM Y = 80,51364 - 0,103.x + 3,59776E-5.x2
R = 0,97503 SC-W-LM
R = 0,982

Compressive strength (MPa)

Flexural strength (MPa)

4 20
SC-BS

SC-BS
3 15
SC-BS-WS
SC-WS

2 10 SC-BS-WS

SC-WS

1 5
1400 1500 1600 1700 1800 1900 2000 2100 1400 1500 1600 1700 1800 1900 2000 2100
Density (kg/m3) Density (kg/m3)

Fig. 15. Relationship between mechanical strength and density.

22 4,5
Y = 1,54841 + 0,13327.x
Y = 269,7697 - 675,57019.x + 433,06089.x 2
20 R = 0,99952
R = 0,99093
Compressive strength (MPa)

18 4,0
Flexural strength (MPa)

16
3,5
14

12
3,0
10

8 2,5
6

4 2,0
0,55 0,60 0,65 0,70 0,75 0,80 0,85 6 8 10 12 14 16 18 20 22
W/C ratio Compressive strength (MPa)

Fig. 16. Compressive strength versus WT/C ratio and flexural strength.

is even better than total replacement, of the fact that in the case of to SC-WS, the decrease is about 39.02%. Indeed the presence of the
SC-BS-WS, it has been recorded a shrinkage value of only two materials, wood shavings and barley straws, is beneficial for
0.75 ± 0.03 mm/m. This value is lower even than that obtained in the shrinkage, since the presence of straws without shavings has
the case of sand concrete with barley straw alone (SC-BS). It consti- given a higher shrinkage value. So, shavings and straw, all together,
tutes an increase of only 4.16% compared to the basic composition have a complementary positive effect. Or perhaps, the fibers con-
(SC-W-LM) and a decrease of 11.76% compared to SC-BS. Compared tent must not exceed a limit proportion [29]. Moreover, the chosen
254 B. Belhadj et al. / Construction and Building Materials 66 (2014) 247–258

0,40 3.2. Study of the materials in the fresh state

Y = 0,73996 - 2,61644E-4.x
0,38
R = 0,99615
0,36 At equal workability, the evolution of the relationship between
0,34 the content of lignocellulosic materials and the WT/C ratio follows a
purely linear relationship with good correlation (Fig. 11). Since WT
Rf/Rc ratio

0,32
represents the total water, a big part of this observation is due to
0,30
the amount of water absorbed by the lignocellulosic material
0,28
which linearly increases with the increase of the latter.
0,26
0,24
0,22
3.3. Study of the materials in the hardened state
0,20
1400 1500 1600 1700 1800 1900 2000 2100 3.3.1. Density
3
Density (kg/m ) At this state, the purpose generally sought by the addition of
lignocellulosic materials is, in fact, the degree of lightness and
Fig. 17. Relationship between density and Rf/Rc ratio. thermal insulating capacity. Regarding this work, the higher the
lignocellulosic materials content, the lower is the density of sand
concrete. The density decreases following a parabolic law with a
length of fiber, which is greater than the aggregate maximum size, good correlation as shown in Fig. 12. This findings agree with the
is enough sufficient to be higher than the limit length [29]. It research works already led on sand concrete based on wood
should be noted finally that the addition of wet wood reduces shavings [17,25]. Indeed, the addition of lignocellulosic materi-
the shrinkage about 27% compared to dry wood [25,26]. als increases the porosity of sand concrete. The study of the

(a1) SC-BS (After flexion) (b1) SC-BS-WS (After flexion)

(a2) SC-BS (After flexion) (b2) SC-BS-WS (After flexion)

(a3) SC-BS (After flexion) (b3) SC-BS-WS (After flexion)

(a4) SC-BS (After compression) (b4) SC-BS-WS (After compression)

Fig. 18. Aspect of cracks after mechanical tests.
 B. Belhadj et al. / Construction and Building Materials 66 (2014) 247–258 255

relationship between the dry density and the porosity accessible to

water, showed the existence of a negative linear relationship with
a very good correlation as shown in Fig. 13. This increase in poros-
ity is due to the porous structure of both lignocellulosic materials
and their highly hygroscopic nature which induces an additional
porosity in the matrix [17,25].

3.3.2. Thermal properties

The measurements of thermal properties led to the determina-
tion of the following parameters: thermal conductivity (k), thermal
diffusivity (a) and specific heat (cp). The obtained values of these
parameters are shown in Table 6. It is well known that the relation-
ship between the thermal conductivity and the specific heat can
lead to the determination of the thermal diffusivity.

a ¼ k=q  cp ðm2 =sÞ ð3Þ

where q is the density of the material. (a) SC-WS (G=2O)

Since the studied concrete is intended to the building shell con-
struction in arid environments, so, the interpretation of results
mainly focuses on thermal diffusivity value which connects three
other thermophysical properties of concrete (Eq. (3)). This thermal
property, that characterizes the concept of thermal inertia, has
given interesting values in the case of SC-BS-WS, with a reduction
of about 35.47% with respect to the basic composition, while the
reduction in the case of SC-WS is only 46.19%. It’s true that
SC-WS has given a value of thermal diffusivity slightly lower, but
SC-BS-WS is much better in shrinkage. Indeed, beside the good
shrinkage, SC-BS-WS is also good in terms of thermal diffusivity.
Moreover, SC-BS-WS have a higher density compared to SC-WS;
if its density is reduced and brought to a value close to that of
SC-WS, it may have the same thermal diffusivity or even better.
The study of the relationship between its dry density and its ther-
mal diffusivity is characterized by a polynomial regression with a
very good correlation as shown in Fig. 14. Similarly, the study of
the relationship between its thermal conductivity and its thermal (b) SC-BS (G=2O)
diffusivity is characterized by a linear evolution with good correla-
tion as shown in Fig. 14.

3.3.3. Mechanical strength and fracture behavior

The study of the mechanical properties, which are based on
compressive and flexural tests, showed that the lower the density
is, the lower are the mechanical properties. The variation of the
compressive strength as a function of density is similar to that of
flexural strength, as shown in Fig. 15. Regressions of polynomial
type, confirmed by very good correlation coefficients (1), are
obtained. Moreover, the highest mechanical strength is related to
a higher compactness and consequently to low W/C ratios; the var-
iation of the mechanical strength is inversely proportional to the
W/C ratio (Fig. 16). It nearly presents a hyperbole shape. This rela-
tionship is valid for any concrete regardless the type of aggregate
and age [29]. It is also clear that, the compressive strength
increases when the flexural strength increases. This evolves
according to linear function with very good correlation as shown (c) SC-BS-WS (G= 20)
in Fig. 16. That is exactly which has been observed in the study
of the relationship between the density and the ratio (Rf/Rc) Fig. 19. Visualization of the interface ‘‘lignocellulosic materials – cementitious
(Fig. 17). Indeed, barley straw has shown a positive effect on the matrix’’.
ductility improvement of sand concrete, either for the composition
of SC-BS-WS or the composition of SC-BS. An important improve-
ment, mostly in the flexural strength, was recorded (In the case adherence ‘straw-matrix’. Even after application of an additional
of SC-BS, it is higher of about 41% compared to SC-WS). force, the straw fibers continue to resist until a force value of about
In addition, after cracking, the specimen preserves its original 0.30 ± 0.011 kN, i.e. the separation of the two half-prisms can take
shape (no sudden separation) as shown in Fig. 18(a1, a2, b1, b2), place only after a large crack opening. (Fig. 18(a3 and b3)). It is also
i.e. the two half-prisms remain attached, one to the other, thanks possible to observe that, after rupture, most of the fibers of barley
to the straw fibers. This can be explained by the significant effect straws are debonded then pulled gradually from the matrix phase
of the tensile strength of the barley straw itself and the strong after the breaking (not cut). Moreover, it should be noted that the
256 B. Belhadj et al. / Construction and Building Materials 66 (2014) 247–258

0.15 EChant5(sab+cim+calc+pouz+chaux+bois)
EChant4(sab+cim+calc+pouz+chaux+paille+bois)
----- E1 ----- E2
EChant3(sab+cim+calc+pouz+chaux+paille)
EChant2(sab+cim+calc+pouz+chaux)
----- E3
EChant(sab-cim-cal) ----- E4
0.125
----- E5

0.1

0.075
Abs

0.05

0.025

4000 3750 3500 3250 3000 2750 2500 2250 2000 1750 1500 1250 1000 750
EChant5 (sab+cim+calc+pouz+chaux+bois) 1/cm

E1 : sand-cement-limestone
E2 : sand-cement- limestone-pozzolan-lime
E3 : sand-cement- limestone-pozzolan-lime-straw
E4 : sand-cement- limestone-pozzolan-lime-straw -wood
E5 : sand-cement- limestone-pozzolan-lime-wood

Fig. 20. Infrared spectroscopy of studied sand concretes.

used lignocellulosic materials did not undergo any chemical effect.
In addition, the behavior of the straw seems better than that of
wood shavings. The fibers of the barley straws have shown an
advantageous effect on the compressive strength. The latter is
clearly improved when the substitution of wood shavings by bar-
ley straws is increased and the specimens preserve better their ori-
ginal shapes after breaking as shown in Fig. 18(a4 and b4). The
improvement of the ductility and the toughness of sand concrete
by adding barley straws (cases of SC-BS and SC-BS-WS compared
to SC-WS), has clearly shown the beneficial effect of barley straws
compared to wood shavings. In fact, the straw of barley increases
the compressive strength of sand concrete thanks to the good
adherence ‘‘straw-matrix’’ which increases the pullout strength
(Fig. 19) and the high rigidity of the straws (Fig. 9). Fig. 19 also
shows that the cement matrix in the case of straw is more compact
than in the case of wood shavings, which increases the mechanical
strength of the barley straws composite. The additional porosity in
the matrix, which is due to the evaporation of water absorbed by
the lignocellulosic materials [17], is lower in the case of straws.
Perhaps this is because, in this case the evaporation is guided by
the tubular pore of straws, while, in the case of shavings, the evap-
oration is made in every direction. Moreover, and compared with
wood shavings, the straws solidify the skeleton of the sand con-
crete thanks to the effect of their length and stiffness. Even their
positive effect on the shrinkage is not without consequence; less
shrinkage leads to fewer micro-cracks that may influence the
porosity and consequently the mechanical strength.
3.3.4. Structure ‘‘matrix – lignocellulosic materials’’
Fig. 19 shows the tubular pores created in the matrix when the
straws are pulled out of the matrix during flexion test, which
proves the good straws strength (they are pulled out but not cut)
compared to wood shavings. Finally, it should be noted that
although both of the lignocellulosic materials adhere well to the
cement matrix, straws are more advantaged by their rigidity which
gives them the ability to help the sand concrete in compression
and especially in flexion. Another important observation has been
made: there is no penetration of cement paste inside the tubular
voids of straws, which has allowed to not affecting the gain in
lightness already acquired. Let us note finally that the orientation
of fibers has also its influence on all the properties of concrete
[30–33]. It is well known that the orientation of fiber relative to
the plane of a crack in concrete influences the reinforcing capacity
of fiber. The maximum benefit occurs when the fiber is unidirec-
tional and parallel to the applied tensile stress. On the contrary,
the fibers are of less benefit when they are randomly oriented to
three dimensions [29]. In this study, the orientation has not been
taken into consideration; the fibers have been randomly orientated
during the mixing and implementation operations. But despite
that, when the concrete is fresh enough, the fibers naturally put
themselves in the right direction into the molds due to the effect
of their density which is lower than that of the matrix (Fig. 18).
3.3.5. Infrared spectroscopy analysis (IR)
The infrared spectroscopy analysis (IR) has shown that in the E2
mixture, calcium hydroxide reacts with siliceous and aluminous
compounds of pozzolan and those of cement to form silicates
and aluminates of calcium which are called hydrates. The calcium
silicate hydrates (CASAH) are the main bond phases for several
concretes [34]. This product, coming from the predominant com-
ponent C3S, which is responsible for a large part of the mechanical
properties of the hardened material (Fig. 20).
This hydrate is indicated by the wide-band centered at
1005 cm1 due to the environment change of the chemical bond
SiAO (the hydration of C3S led to CASAH).
Indeed, the decrease in the height of the band related to sili-
cates after polymerization is due to the formation of CASAH [35].
A thin band centered on 3660 cm1, which characterizes the
elongation vibrations of OAH bonds, allows following the presence
of portlandite (CH).
 B. Belhadj et al. / Construction and Building Materials 66 (2014) 247–258
Table 7
Assignment of IR bands recorded on each material.
IR bands (cm1) Assignment
1005 SiAO: silicate, CASAH
3660 OAH: Portlandite Ca(OH)2
870 C@O: Calcite (CaCO3)
In the E3 mixture, the presence of the straw causes an increase
in the calcite (CaCO3) characterized by vibration of CAO bonds
observed at 870 cm1 reducing the polymerization of tetrahedral
ions ‘‘SiO4
4 ’’. The straw plays an inhibitory role in the mixture by
developing some hydration reactions allowing the formation of
hydrates. Unlike the mixture E3, E4 and E5 which contain wood,
the increase in polymerization of SiO4 4 leads to an increase of
hydrates, which consequently improves the mechanical properties
of the composites.
It should be also noted that the benefit is in favor of the mixture
E4 (SC-BS-WS) whose spectrum (in digital part) is almost superim-
posed with that of E2 (SC-W-LM).
So, the good results (mechanical, thermal and dimensional)
already obtained for the mixture E4, which is the concrete sand
based on wood shavings and barley straws, can be confirmed by
the analysis of its infrared spectroscopy studied structure.
Finally, Table 7 lists assignment of IR bands recorded on each
material. The obtained results are in agreement with the literature
results [34–37].
4. Conclusion
The approach to formulating eco-sand–concrete based on ligno-
cellulosic materials is an environmental and sustainable approach
because it helps meet the three pillars of sustainable development:
environment, economy and society. It is within this context that
this experimental work which studies the effect of lightening by
addition of lignocellulosic materials on the physico-mechanical
properties and the shrinkage of eco-sand–concrete studied at equal
workability was conducted. The obtained results reveal the exis-
tence of strong relations between the thermophysical and mechan-
ical properties of sand concretes which was confirmed by various
regressions obtained with very good correlation coefficients (close
to 1). The sand concrete based on both lignocellulosic materials,
wood shavings and barley straws (SC-BS-WS), is both insulating
and load-bearing concrete which can be used in the construction
of external walls or other elements in arid environments. However,
the sand concrete based on wood shavings (SC-WS) is an insulating
concrete which can be used for wall insulation and roofs, but the
only drawback is its high shrinkage. It should be noted that the
sand concrete based on both lignocellulosic materials, wood
shavings and barley straws (SC-BS-WS), has better respected the
compromise between the studied properties: density, thermal con-
ductivity, specific heat, thermal diffusivity, the mechanical proper-
ties and especially the shrinkage. It has presented the better
shrinkage and good thermal diffusivity. Its shrinkage increases of
only 4.16% compared to the basic composition, but decreases of
11.76% compared to SC-BS and 39.02% compared to SC-WS. More-
over, its thermal diffusivity has given interesting values in the case
of SC-BS-WS, with a reduction of about 35.47% with respect to the
basic composition. The barley straws improved the deformation
capacity of sand concrete, namely ductility and toughness. So,
the hypothesis of partial substitution of wood shavings by barley
straws has positively contributed to the improvement of the
shrinkage of sand concrete by reducing it to a value less than the
limit value without overly affecting the thermophysical properties,
hence the objective is achieved. On the other hand, the barley
straw adheres well to the cement matrix, increases its stiffness
257
and deformability and does not break after the fracture of the
matrix. Besides all this, it can be emphasized the beneficial effect
of the used additions substituted in the cement which makes the
economic interest of the studied composite more interesting.
Moreover, the obtained results have been confirmed by micro-
structure study. It should be noted that the good particle size
distribution of sand has also contributed in improving the homoge-
neity of the matrix which has led to the good obtained results.
Finally, the results are encouraging and promising within the con-
text of eco-materials intended for sustainable construction in arid
environments, such as the south of Algeria. It is possible to design
passive houses, more efficient in thermal insulation and energy
saving, especially with SC-BS-WS which has shown better results
compared to those of other studied sand concretes. In addition, it
should also be noted, the very low cost of the materials used within
the context of valorizing of local materials (dune sand available in
large quantities) and the re-use of waste (limestone fillers, wood
shavings and barley straws).
References
[1] Khazma M, Goullieux A, Dheilly RM, Quéneudec M. Coating of a lignocellulosic
aggregate with pectin/polyethylenimin mixtures: effect on flax shive and
cement-shive composite properties. Cement Concr Compos 2012;34:223–30.
[2] Karade SR. Cement-bonded composites from lignocellulosic wastes. Constr
Build Mater 2010;24:1323–30.
[3] Mathur VK. Composite materials from local resources. Constr Build Mater
2006;20:470–7.
[4] Bederina M, Gotteicha M, Belhadj B, Dheily RM, Khenfer MM, Queneudec M.
Drying shrinkage studies of wood sand concrete – effect of different wood
treatments. Constr Build Mater 2012;36:1066–75.
[5] Pimienta P, Chandellier J, Rubaud M, Dutruel F, Nicole H. Etude de faisabilité
des procédés de construction à base de béton de bois. Cahiers du CSTB, No.
2703; 1994 [in French].
[6] Ashour T, George H, Wu W. Performance of straw bale wall: a case of study.
Energy Build 2011;43:1960–7.
[7] Ashour T, Wieland H, Georg H, Bockisch FJ, Wu W. The influence of natural
reinforcement fibres on insulation values of earth plaster for straw bale
buildings. Mater Des 2010;31:4676–85.
[8] AFNOR, Concrete – Sand concrete. Projet P 18-500 [in French].
[9] Chauvin JJ, Grimaldi G. The sand concretes. Bulletin de Liaison des Laboratoires
des Ponts et Chaussées, No. 157; September–October 1988. p. 9–15 [in French].
[10] Khazma M, Goullieux A, Dheilly RM, Laidoudi B, Queneudec M. Impact of
aggregate coating with a PEC elastomer on properties of lightweight flax shive
concrete. Ind Crop Prod 2011;33:49–56.
[11] Bouhicha M, Aouissi F, Kenai S. Performance of composite soil reinforced with
barley straw. Cement Concr Compos 2005;27:617–21.
[12] Li Z, Wang X, Wang L. Properties of hemp fibre reinforced concrete composites.
Composites: Part A (Appl Sci Manuf) 2006;37:497–505.
[13] Yang HS, Kim DJ, Kim HJ. Rice straw–wood particle composite for sound
absorbing wooden construction materials. Bioresour Technol 2003;86:117–21.
[14] Soroushian P, Aouadi F, Chowdhury H, Nossoni A, Sarwar G. Cement-bonded
straw board subjected to accelerated processing. Cement Concr Compos
2004;26:797–802.
[15] Nozahic V, Amziane S, Torrent G, Saidi k, De Baynast H. Design of green
concrete made of plant-derived aggregates and a pumice-lime binder. Cem
Concr Compos 2012;34:231–41.
[16] Merta I, Tschegg EK. Fracture energy of natural fibre reinforced concrete.
Constr Build Mater 2013;40:991–7.
[17] Bederina M, Marmoret L, Mezreb K, Khenfer MM, Bali A, Quéneudec M. Effect
of the addition of wood shavings on thermal conductivity of sand concretes:
experimental study and modelling. Constr Build Mater 2007;21:662–8.
[18] Taoukil D, El bouardi A, Ajzoul T, Ezbakhe H. Effect of the incorporation of
wood wool on thermo physical properties of sand mortars. KSCE J Civil Eng
2012;16(6):1003–10.
[19] Bederina M, Laidoudi B, Goullieux A, Khenfer MM, Bali A, Quéneudec M. Effect
of the treatment of wood shavings on the physic-mechanical characteristics of
wood sand concretes. Constr Build Mater 2009;23:1311–5.
[20] Bederina M, Khenfer MM, Dheilly RM, Quéneudec M. Reuse of local sand: effect
of limestone filler proportion on the rheological and mechanical properties of
different sand concrete. Cem Concr Res 2005;35:1172–9.
[21] Belaidi ASE, Azzouz L, Kadri E, Kenai S. Effect of pozzolana and marbre powder
on the properties of self compacting concrete. Constr Build Mater
2012;31:251–7.
[22] Makhloufi Z, Kadri EH, Bouhicha M, Benaissa A. Resistance of limestone
mortars with quaternary binders to sulfuric acid solution. Constr Build Mater
2012;26:497–504.
[23] Hayoz Pascal, Peter Wolfgang, Rogez Daniel. A new innovative stabilization
method for the protection of natural wood. Prog Org Coat 2003;48:297–309.
258 B. Belhadj et al. / Construction and Building Materials 66 (2014) 247–258
[24] Suna JX, Xub F, Suna XF, Xiaoa B, Sunb RC. Physico-chemical and thermal
characterization of cellulose from barley straw. Polym Degrad Stab
2005;88:521–31.
[25] Ledhem A, Dheilly RM, Benmalek ML, Quéneudec M. Reuse of waste oils in the
treatment of wood aggregates. Constr Build Mater 2000;14:341–50.
[26] Rim K, Ledhem A, Roodenbeke M. Quéneudec-t’Kint, parameters study of
manufacturing a wood concrete based on clay matrix. Mater Struct
1996;29(8):514–8 [In French].
[27] Poon Chi-Sun, Chan Dixon. Effects of contaminants on the properties of
concrete paving blocks prepared with recycled concrete aggregates. Constr
Build Mater 2007;21:164–75.
[28] Kasemchaisiri R, Tangtermsirikul S. A method to determine water retainability
of porous fine aggregate for design and quality control of fresh concrete.
Constr Build Mater 2007;21:1322–34.
[29] Neville AM. Properties of concrete, Ed. EYROLLES; 2000 [in French].
[30] Vincent Thomas, Ozbakkaloglu Togay. Influence of fiber orientation and
specimen end condition on axial compressive behavior of FRP-confined
concrete. Constr Build Mater 2013;47:814–26.
[31] Deeb R, Karihaloo BL, Kulasegaram S. Reorientation of short steel fibres during
the flow of self-compacting concrete mix and determination of the fibre
orientation. Cem Concr Res 2014;56:112–20.
[32] Zerbino R, Tobes JM, Bossio ME, Giaccio G. On the orientation of fibers in
structural members fabricated with self-compacting fiber reinforced concrete.
Cement Concr Compos 2012;34:191–200.
[33] Sanal Irem, Ozyurt Zihnioglu Nilufer. To what extent does the fiber orientation
affect mechanical performance? Constr Build Mater 2013;44:671–81.
[34] Ebbert C, Grundmeier G, Buitkamp N, Kroger A, Messerschmidt F, Thissen P.
Toward a microscopic understanding of the calcium–silicate–hydrates/water
interface. Appl Surf Sci 2014;290:207–14.
[35] Ahmari S, Ren X, Toufigh V, Zhang L. Production of geopolymeric binder from
blended waste concrete powder and fly ash. Constr Build Mater
2012;35:718–29.
[36] Horgnies M, Willieme P, Gabet O. Influence of the surface properties of
concrete on the adhesion of coating: characterization of the interface by peel
test and FT-IR spectroscopy. Prog Org Coat 2011;72:360–79.
[37] Nozahic V, Amziane S. Influence of sunflower aggregates surface treatments on
physical properties and adhesion with a mineral binder. Composites: Part A
2012;43:1837–49.