Construction and Building Materials 27 (2012) 126129

Contents lists available at SciVerse ScienceDirect

Construction and Building Materials

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

Evaluation of using waste timber railway sleepers in woodcement

composite materials
Alireza Ashori a,, Taghi Tabarsa b, Fatemeh Amosi b
a
Department of Chemical Technologies, Iranian Research Organization for Science and Technology (IROST), P.O. Box 15815-3538, Tehran, Iran
b
Department of Wood and Paper Technology, Gorgan University of Agricultural Sciences and Natural Resources (GUASNR), Gorgan, Iran

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

Article history: In present study, the suitability of using waste timber railway sleepers (WTRSs) as ller in woodcement
Received 18 June 2011 composites was investigated. The effects of two variable factors namely press temperatures (25 and
Received in revised form 3 August 2011 60 C) and calcium chloride contents (3%, 5% and 7% w/w cement) on some physico-mechanical proper-
Accepted 4 August 2011
ties were also investigated. The following experimental parameters were constant: wood/cement ratio
Available online 30 August 2011
(40:60), board thickness (15 mm), press pressure (40 kg/cm2), and press time (5 min). Press temperature,
calcium chloride and the interaction of both variables had signicant effects (p < 0.01) on all the studied
Keywords:
properties. Test results showed that addition of calcium chloride tends to enhance both the physical and
Timber railway sleeper
Construction material
mechanical properties of boards. All properties of the boards were improved when the calcium chloride
Woodcement composite content was increased from 3% to 7%. The results also showed that as the press temperature was
Physico-mechanical properties increased from 25 to 60 C, signicant increased in water absorption and thickness swelling occurred.
Water absorption and thickness swelling (at 2 h and 24 h) compared favorably with values reported
for cement-bonded composites produced from virgin wood particles. In general, the strength properties
of the boards were found to be a maximum when press temperature and calcium chloride were 25 C and
7%, respectively. These properties can be exploited in many applications where lightweight concretes are
required. Therefore, WTRS is technically suitable for building construction such as paneling, ceiling and
partitioning.
2011 Elsevier Ltd. All rights reserved.

1. Introduction
Traditionally, wood has been an essential construction material.
However, with the invention of reinforced concrete, its application
has been reduced [1]. In recent years, there has been a growing
interest in utilizing wood for making low-cost building materials
[25]. The primary advantages of using woody materials as llers
in cement are the low density, low cost, nonabrasive nature, high
lling levels possible, low energy consumption, and wide variety
of wood species available throughout the world [6]. On the other
hand, increasing demand for forest resources in various applica-
tions has led to the shortages of wood supply. Thus, there is a need
to look for innovative ways of using non-traditional forest
resources to substitute wood raw materials for wood-based indus-
tries. Among the possible alternatives, using recycled wood is cur-
rently at the center of attention [1,79]. Every year, a large number
of old and deteriorated structures such as railways, buildings, fenc-
ing poles, furniture items and bridges are being demolished. Most
of the timbers used in construction are being treated with chemical
preservatives (such as chromated copper arsenate and creosote)
Corresponding author. Tel./fax: +98 228 2276629.
E-mail address: ashori@irost.org (A. Ashori).
during their service life to protect them from biological deteriora-
tion. These chemical materials contain toxic or polymeric sub-
stances which are not easily biodegradable. Presently these
wastes are either burnt or land lled. These approaches cause var-
ious environmental issues like air pollution, emission of green
house gases and occupation of useful land. The increasing charges
of landll are further aggravating the problem. Moreover, these
methods of disposal are certainly wastage of a primary natural re-
source. Therefore attempts have been made by researchers, both in
industry and academia, to reuse these wastes as the building mate-
rials (i.e. woodcement composites) [1,6].
Studies on the waste wood in the forms of bers, particles or
strands suggest that these materials have the potential for use as
reinforcing agent or ller in cement composites. Kasai et al. [10]
used wood particles from construction waste in Japan for making
wood-chip concrete. They made concrete with a density range of
0.921.25 g/cm3. They found the exural strength of the product
in the range of 47 MPa and compressive strength 58 MPa. The
ratio between exural strength and compressive strength was
0.50.9, greater than that for normal concrete. This indicates the
reinforcing effect of wood particles. They further reduced the den-
sity to about 0.78 g/cm3 by adding synthetic lightweight aggre-
gates. This resulted in comparatively lower bending and

0950-0618/$ - see front matter 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.conbuildmat.2011.08.016
 A. Ashori et al. / Construction and Building Materials 27 (2012) 126129 127

compressive strength values at 2.05 and 2.2 MPa, respectively. In Table 2

another study by Wolfe and Gjinolli [11] on the use of southern Mixing ratios of raw materials and their abbreviations.

pine wood particles derived from construction waste, it was found Code Press temp. (C) CaCl2 (%) Cement (%) WTRS (%) Replications
that the toughness index, which is used to compare the energy A1B1 25 3 60 40 3
absorption capacity during failure of a material, was greater than A1B2 25 5 60 40 3
whereas the average value of toughness index for other ber rein- A1B3 25 7 60 40 3
forced concrete products is 5. Therefore, they achieved a better A2B1 60 3 60 40 3
A2B2 60 5 60 40 3
toughness index. A2B3 60 7 60 40 3
A number of studies have reported the effects of virgin or waste C1 25 0 100 0 3
wood particles on the properties of cement composites [9,1215]. C2 60 0 100 0 3
However, no published reports are available that evaluate effects of
waste timber railway sleepers (WTRSs) on the strength and sorp-
tion properties of cement-bonded composites. In addition, reutili- 2.3.1. Physical properties
zation of WTRS has advantages for economy and environment. Physical properties in terms of apparent density (AD), water absorption (WA)
The main objective of present work is to evaluate the suitability and thickness swelling (TS) were evaluated. The specimens for WA and TS
(50  50 mm2) were completely submerged horizontally under distilled water
of using WTRS as reinforcing agent in the cement composites. Fur-
maintained at 25 C for 2 h and 24 h. After soaking, the samples were drained on
ther, the effects of two variable factors namely press temperatures paper towels for 10 min to remove excess water. The WA and TS were calculated
and calcium chloride contents on the selected physical and from the increase in weight and thickness of the specimen during submersion,
mechanical properties were investigated. respectively. At least three specimens of every board were tested to obtain a reliable
average and standard deviations.

2. Experimental procedures
2.1. Materials
The woody material used in this work was WTRS which was collected from a
local railway station, Karaj, Iran. They have been preserved by pressure treatment
with an EPA-registered pesticide containing creosote to protect them from insect
attack and decay. The WTRSs were initially chipped using a laboratory-scale drum
chipper (Pallmann PHT). The chips were then reduced into smaller pieces using a
laboratory-scale ring aker (Pallmann PZB). Dried particles with 5% moisture con-
tent were screened to eliminate oversized particles. The weight distributions of
wood particles are presented in Table 1.
The binding agent employed was commercial grade of Portland cement (type II).
Calcium chloride (CaCl2) was used as cement setting accelerator. It was an analyt-
ical grade from Merk Co, Germany.
2.3.2. Mechanical properties
Conditioned boards were sawn into test samples for modulus of rupture (MOR)
modulus of elasticity (MOE) and internal bonding (IB) strength. Three-point exural
testing was carried out using an Instron Universal Testing Machine, with a span of
180 mm and crosshead, bearer diameter of 25 mm and loading speed of 5 mm/min.
2.4. Statistical analysis
The experimental design consisted of two treatments press temperature and
percentage concentration of CaCl2 and three replications per treatment. Data for
each treatment were statistically studied by analysis of variance (ANOVA). When
the ANOVA indicated a signicant difference among factors and levels, a comparison
of the means was done employing Duncans multiple range test to identify the
groups that were signicantly different from others groups at 99% condence level.

2.2. Preparation of boards 3. Results and discussion

All the boards were made with 40:60 weight ratio for wood/cement, and a 1:0.6
weight ratio for cement/water (the amount of water in the wood particles was in-
cluded). The wood/cement ratio was selected based on commercially produced
woodcement composites. Two process variables considered in this study include
CaCl2 contents (3%, 5% and 7% w/w cement) and press temperature levels (25 and
60 C). Control samples contained neat cement and water. Board thickness
(15 mm), press pressure (40 kg/cm2) and press time (5 min) were kept constant
in all experiments. Table 2 presents the formulations and abbreviations of the
boards produced.
The wood particles for each board were rst sprayed evenly with the dilute
aqueous solution of CaCl2. Consequently, the cement and WTRS were uniformly
blended and then immediately transferred into a mould of 450 mm 
450 mm  15 mm. The mixture was evenly distributed and attened using a woo-
den block, the mould removed and a piece of plywood placed on top of the mat.
The resulting assemblage was pre-pressed to reduce its height while the mat for
the next board was mixed. This stack of mats was placed between two steel plates
and pressed. The pressed mats were kept under compression for 24 h by bolting
the two steel plates together using four 15 mm thick bolts. After 24 h, the boards
were declamped, stacked vertically and conditioned for 28 days at 35 1 C and
90 5% RH to allow the cement to cure and gain strength. The specimens were con-
ditioned in a controlled room for 28 days at 25 C and 65% RH.
2.3. Test procedure
Mechanical and physical properties were determined following ISO 8335:1987.
It is a procedure for boards of Portland or equivalent cement reinforced with wood
particles.
Table 1
Distribution of wood particles.
Size (mm) Percentage
<0.4 3.2
0.41 3.4
12 8.4
34, >4 82.1
The results of the physical and mechanical tests, with statistical
analysis, are shown in Tables 35 for all the fabricated materials.
3.1. Physical properties
The woodcement composites with low board density (61 g/
cm3) would be preferred for applications that require boards of
light weight such as ceiling and roong for easy installation. In
addition, the composites with low cement/wood ratio would be
more economical because a small quantity of cement is required
with a large quantity of particles used in this study (WTRS) during
CBB production. It must be noted that cement is more expensive to
procure than WTRS. According to Adhikary et al. [16] the water
absorption and swelling rate increased linearly with the decreasing
of the composite board density.
In general, the characteristics that make woodcement compos-
ites desirable to construction applications are low water absorp-
tion and good dimensional stability. On the other hand, water
absorption can affect on the mechanical properties. As Tables 3
and 4 present, a signicant difference in water absorption was ob-
served for the 8 types of boards after 2 h and 24 h of immersion.
A1B1, A1B2 and A1B3 boards showed intermediate values of water
absorption which increased with immersion time. A1B3 board
showed the lowest values of water absorption among the studied
boards. However, water absorption of woodcement composite
boards was signicantly higher than control samples (C1 and C2).
This could be attributed to their lower bulk density and hence
higher porosity already alluded to (Table 3). Asasutjarit et al. [17]
reported that the low density woodcement boards have more
void spaces than dense ones so that more water can be absorbed.
128 A. Ashori et al. / Construction and Building Materials 27 (2012) 126129

Table 3
Mean values of physical and mechanical properties of woodcement composites.

Code D (g/cm3) MOR (MPa) MOE (MPa) IB (MPa) WA (%) TS (%)

2h 24 h 2h 24 h
A1B1 0.98 (0.06) 8.3 (1.1) 2166 (102) 0.36 (0.06) 14.6 (1.8) 16.5 (1.2) 1.34 (0.8) 1.78 (0.93)
A1B2 1.00 (0.04) 9.5 (0.8) 2189 (87) 0.40 (0.09) 15.1 (1.6) 15.8 (2.5) 1.09 (0.12) 1.61 (0.47)
A1B3 1.02 (0.06) 10.1 (0.9) 2206 (152) 0.43 (0.05) 11.0 (1.8) 15.4 (1.2) 0.98 (0.91) 1.51 (0.82)
A2B1 0.94 (0.11) 1.9 (0.4) 1034 (92) 0.09 (0.02) 26.1 (2.3) 39.8 (2.7) 6.16 (1.14) 15.25 (1.35)
A2B2 0.96 (0.10) 2.1 (0.7) 1271 (120) 0.12 (0.04) 25.9 (2.7) 39.4 (4.1) 5.95 (2.01) 15.11 (2.06)
A2B3 0.98 (0.8) 2.8 (0.6) 1542 (84) 0.14 (0.07) 25.6 (2.4) 39.1 (2.8) 5.80 (1.97) 15.01 (1.69)
C1 1.13 (0.09) 15.4 (1.0) 2261 (186) 0.51 (0.10) 15.6 (1.9) 29.2 (2.1) 1.33 (1.01) 10.73 (2.03)
C2 1.17 (0.11) 14.6 (1.2) 2247 (191) 0.50 (0.08) 15.2 (1.4) 28.9 (1.4) 1.53 (0.98) 11.21 (1.27)

Note: Standard deviations are in parenthesis.

Table 4
Analysis of variances on the effects of press temperature (A) and CaCl2 (B) and their interaction (AB) on the physical properties.

Source of variations df WA 2-h WA 24-h TS 2-h TS 24-h

SS MS F SS MS F SS MS F SS MS F
A 1 791.092 791.092 518.81 2492.88 2492.88 7866.48 105.17 105.17 3312.31 815.54 815.54 62.97
B 2 12.62 6.31 4.14 2.416 1.208 3.81 0.405 0.202 6.37 0.284 0.142 3.09
AB 2 7.55 3.78 2.484ns 0.156 0.078 0.25ns 0.001 0.0006 0.02 ns 0.014 0.007 0.15ns
Error 12 18.29 1.53 3.169 0.317 0.317 0.0317 0.460 0.046
Total 17 829.56 2499.42 106.68 819.18

df = degree of freedom; MS = mean of squares; SS = sum of squares; F = F value; = signicant difference at the 1% level (p 6 0.01%); and ns = not signicant.

Table 5
Analysis of variances on the effects of press temperature (A) and CaCl2 (B) and their interaction (AB) on the density and mechanical properties.

Source of variations df MOR MOE IB AD

SS MS F SS MS F SS MS F SS MS F
A 1 222.18 222.18 877.52 3685612.5 3685612.5 5403.3 0.35 0.35 453.56 0.0845 0.0845 93089
B 2 6.1 3.05 13.05 225,424 112,712 165.2 0.0103 0.0051 6.71 0.0044 0.0022 24.63
AB 2 0.907 0.453 1.94ns 164,668 82,334 12.74ns 0.0002 0.0001 0.14 ns 0.00003 0.000016 0.19ns
Error 10 3.04 0.252 6821 682.1 1.07 8.94 5.93 9.94
Total 17 232.23 4084482.5 1.45 6.02

df = degree of freedom; MS = mean of squares; SS = sum of squares; F = F value; = signicant difference at the 1% level (p 6 0.01%); and ns = not signicant.

In other words, increasing board density decreases the void and
improves thickness swelling and water absorption characteristics.
It is interesting to mention that the increase in void content is also
due to the inadequate compaction of the unworkable mixes.
After addition of 7% CaCl2 to woodcement composites, a reduc-
tion ratio of 25% and 8% in water absorption after 2 h and 24 h of
water soaking, respectively, was observed. In addition, it was ob-
served that CaCl2 treated composites generally absorbed less water
at 2 h and 24 h, respectively. This is most probably due to changes in
the ber structure that rendered them stiffer and tougher by the cre-
ation of high cement-to-cement bonds and cementwood bonds.
These results are comparable with those obtained by Ashori et al.
[6] and Tabarsa and Ashori [18] who proposed the addition of CaCl2,
35%, could improved the water absorption and thickness swelling.
the WTRS with cement was improved considerably. The strength
properties of the boards were found to be a maximum when press
temperature and CaCl2 were 25 C and 7%, respectively.
As mentioned earlier, wood used in construction is typically
treated with chemical preservatives to make it resistant against
microbial degradation and biological deterioration. Among the
chemical preservatives, creosote or coal tar creosote is the most
widely used wood preservative, particularly for timber railway
sleeper in the world. The possible reason for strength reduction
with increasing press temperature, should be related to the diffu-
sion of creosote to the surface of wood particles. This could reduce
adhesion at the interface, which causes inferior interfacial bonding
strength as compared to the control (C1 and C2) samples.

3.2. Mechanical properties 4. Conclusions

The average values and statistical analysis of the mechanical
properties in terms of MOR, MOE and IB are presented in Tables
3 and 5, respectively. Mechanical properties generally decreased
with increase in press temperature (60 C) and increased with in-
crease in CaCl2 content. Further, ANOVA test showed that the
strength properties of the boards were signicantly inuenced by
the press temperature and CaCl2 contents and their interaction (Ta-
ble 5). All boards made with 25 C had the highest values among
the other types of specimens. In addition, all mechanical properties
of the boards were improved when the CaCl2 was increased from
3% to 7%. This may be due to the fact that the compatibility of
Based on the results of this work the following conclusions can
be drawn.
1. During the process of using recycled wood wastes for further
reutilization, some problems should be taken into consider-
ation such as chemical preservatives.
2. Statistical analysis of the test results conrmed that the
physical and mechanical properties of the woodcement
composites showed evidence of signicant difference
between the means of the two treatments (press tempera-
ture and CaCl2 contents) and their interaction tested.
 A. Ashori et al. / Construction and Building Materials 27 (2012) 126129
3. The results of the 2 h and 24 h water soaking test for both
water absorption and thickness swelling of the wood
cement composites indicated that as the press temperature
increases, these properties of the boards increases
signicantly.
4. Mechanical properties generally decreased with increase in
press temperature (60 C) and increased with increase in
CaCl2 content.
5. The strength properties of the boards were found to be a
maximum when press temperature and CaCl2 were 25 C
and 7%, respectively.
6. Mechanical and physical properties of the WTRS-cement
boards were found to be equivalent to those commercially
available and other results described in the literature.
References
[1] Karade SR. Cement-bonded composites from lignocellulosic wastes. Constr
Build Mater 2010;24(8):132330.
[2] Soroushian P, Shah Z, Won J-P, Hsu J-W. Durability and moisture sensitivity of
recycled wastepaper-ber-cement composites. Cem Concr Compos 1994;16(2):
11528.
[3] Jorge FC, Pereira C, Ferreira JMF. Woodcement composites: A review. Holz als
Roh und Werkstoff 2004;62(5):3707.
[4] Tonoli GHD, Rodrigues Filho UP, Savastano Jr H, Bras J, Belgacem MN, Rocco
Lahr FA. Cellulose modied bres in cement based composites. Composites:
Part A 2009;40(12):204653.
[5] Franek J, Kollr M, Makovny I. Microwave electromagnetic led and
temperature distribution in a multilayered woodcement board. J Electrical
Eng 2011;62(1):2530.
129
[6] Ashori A, Tabarsa T, Sepahvand S. Cement-bonded composite boards made from
poplar strands. Constr Build Mater, 2011. doi: 10.1016/j.conbuildmat.2011.
06.001.
[7] Basri HB, Mannan MA, Zain MFM. Concrete using waste oil palm shells as
aggregate. Cem Concr Res 1999;29(4):61922.
[8] Asokan P, Saxena M, Asolekar SR. Solid wastes generation in India and their
recycling potential in building materials. Build Environ 2007;42(6):231120.
[9] Ashori A, Tabarsa T, Azizi K, Mirzabeygi R. Woodwool cement board using
mixture of eucalypt and poplar. Ind Crop Prod 2011;34(1):11469.
[10] Kasai Y, Kawamura M, Zhou JD. Study on wood chip concrete with used timber.
In: Malhotra VM, editor. Recent advances in concrete technology: proceedings:
fourth CANMET/ACI/JCI international conference, SP 179, Tokushima, Japan;
1998, p. 90528.
[11] Wolfe RW, Gjinolli A. In: Proceedings of the conference on the use of recycled
wood and paper in building applications, Madison, Wisconsin, 1998, p. 84-6.
[12] Savastano Jr H, Warden PG, Coutts RSP. Potential of alternative ber cements
as buildings materials. Cem Concr Compos 2003;25(6):58592.
[13] Papadopoulos AN. An investigation of the suitability of some Greek wood
species in woodcement composites manufacture. Eur J Wood Prod
2007;65(3):2456.
[14] Olorunnisola AO. Effects of husk particle size and calcium chloride on strength
and sorption properties of coconut huskcement composites. Ind Crop Prod
2009;29(23):495501.
[15] Espinoza-Herrera R, Cloutier A. Physical and mechanical properties of gypsum
particleboard reinforced with Portland cement. Eur J Wood Prod
2011;69(2):24754.
[16] Adhikary KB, Pang S, Staiger MP. Long-term moisture absorption and thickness
swelling behaviour of recycled thermoplastics reinforced with Pinus radiata
sawdust. Chem Eng J 2008;142(2):1908.
[17] Asasutjarit C, Hirunlabh J, Khedari J, Charoenvai S, Zeghmati B, Shin UC.
Development of coconut coirbased lightweight cement board. Constr Build
Mater 2007;21(2):27788.
[18] Tabarsa T, Ashori A. Dimensional stability and water uptake of cement-bonded
wood composite. Polym Environ 2011;19(2):51821.