1. Introduction

Nowadays, one can note increasing the concern for recycling, reuse and rational exploitation of natural resources, since most of the raw materials are becoming scarce due to poor planning in relation to their use. Another issue is the negative environmental impact caused by many materials, such as asbestos and various types of polymers, which triggered the search for environmentally friendly inputs. [1].

Cement-wood composites are constituted of fibers/vegetal particles, mineral type binder, water and chemical additives, which are consolidated under pressure at room temperature [2]. These products have many advantages such as no emission of toxic during their manufacture and reduced cost due to cure without applying high temperatures [3, 4]. Moreover, panels made from these composites have high versatility in terms of finish, may be sawn, nailed, screwed, glued and towed; allowing wide application [5] as flat roofing, prefabricated structures, mobile homes, permanent formwork, cladding, sound barriers and paving, because they work as structural insulation panels [6].

Wood wastes have been potential materials for application in these composites, such as Eucalyptus grandis [7], wood preserved with pesticide containing creosote [8], tropical woods [9], Jack pine (Pinus banksiana) wood chips [10], among others. Karade [6] cites the possibility of using various lignocellulosic wastes as rice straw, coir, cork granules, hazelnut shell, oil palm, arhar stalks, bagasse; as well as residues from wood products - construction demolition wastes and furniture industry (waste MDF, timber, barks).

In this context, this study aimed to produce and characterize wood-cement composites reinforced with Pinus sp. wastes.

2. Material and Methods

Manufacture of composites and tests

Pinus sp. were obtained in sad sawmills in form of shavings, from Itapeva region, São Paulo State, Brazil. Distilled water was used for composites hydration and Portland cement II-32 as a binder. Wood wastes were converted into particles (Figure 1) in a knife mill Marconi model MA-680. Particles retained between sieves with openings of 10 and 20 mesh were selected, with dimensions closed to 1.7 mm.

Figure 1. Wastes from Pinus sp. wood (a) and particles obtained (b)

Treatment of particles aimed at removing extractives which could inhibit cement curing. Adapting procedure used by Beraldo e Carvalho [11], particles were immersed in distilled water at 80°C, using a thermal bath QUIMIS Q 250-E. Immersion was performed in three cycles (2 hours each), with subsequent washing of the particles in cold water. After, particles were dried in a drying oven at 105 (±) 1°C to remove moisture.

Proportion of cement/biomass/water used in the composites was 1/0.21/0.6. Experimental conditions were obtained according to type of biomass and treatment. Composites production consisted of mixture of the dry components - cement and particles, and after added to water. Six specimens were produced for each condition, cylindrical format and with dimensions of 50 x 100 mm (Figure 2). Specimens were removed from their molds after 48 hours, stored for 5 days in a plastic bag, to prevent cracking and 21 days outdoors, totalizing 28 days curing. Process of manufacturing and curing of specimens was adapted from Beraldo et al. [12], with some adjustments.

Figure 2. Specimens made with untreated wood (a) and with treated wood particles (b)

Properties of composites such as strength and modulus of elasticity in compression parallel to the grain were determined according to NBR 7215 [13] and NBR 8522 [14]. Density of composites was obtained by the ratio between mass and volume of specimens.

Statistical Analysis

To investigate the influence of treated particles on properties of density (ρ), modulus of elasticity (E) and strength (Sc) in parallel compression of wood-cement composites, analysis of variance (ANOVA) - Kruskal-Wallis (non parametric) and Student-Newman-Keuls - multiple comparisons test, by BioEstat ® version 5.0 software, were used. The significance level (α) set for ANOVA was 5%. The null hypothesis (H₀) was the equivalence between means and alternative hypothesis (H₁) was related to non-equivalence of averages. P-value from Kruskal-Wallis test below the significance level implies rejecting H0, accepting it otherwise.

3. Results and Discussions

Table 1 shows results of physical and mechanical properties for composite cement-wood, with and without fiber treatment.

Table 1. Results of physical and mechanical properties for composite cement-wood

Regarding the density, it was noted that values found are consistent with literature, compared to similar composites of 1.20 g/cm³ reinforced with 20% of Eucalyptus grandis particles [7]. After treatment of Pinus sp. particles, composites showed a lower density, possibly due to extractives removal.

Torkaman et al. [15] obtained compression strength to cementitious composites, containing 25% wood, approximately, 1.6 MPa in compression. It should be emphasized the values obtained in present study, with quantities of reinforcing fibers (≈ 20%), resulting in Sc around 5MPa for non-treated particles (UT) and approximately 10 MPa for treated particles (T). In addition to particles treatment, longer cure times can increase this property, according to a study involving composites of this type [16], which compared cure times of 28 and 90 days. In terms of elasticity (E), treatment of wood particles used as reinforcement resulted in increases of 126%.

Tables 2 to 4 present results of ANOVA Kruskal-Wallis test of properties investigated for two treatments evaluated. GL are degrees of freedom R1 and R2 median posts related to particles untreated and treated, respectively.

Table 2. Results of ANOVA Kruskal-Wallis for density

Table 3. Results of ANOVA Kruskal-Wallis for modulus of elasticity in parallel compression

Table 4. Results of ANOVA Kruskal-Wallis for strength in parallel compression

From Tables 2 to 4 it was noted that wood particles treatment was significant in the three properties (P-value <0.05). Particles treatment was responsible for reduction of 7% in composites density and increased values of strength and elasticity in parallel compression at 126% and 70%, respectively, as can be seen in Figure 3.

Figure 3. Graphs of main effects: density (a), modulus of elasticity in compression (b) and compression strength (c)

Studies of Lima [17] showed that thermal bath of wood particles can provide a positive effect on the compressive strength, and result in a lower loss value for the consistency (final-initial consistency) and, consequently, greater fluidity mortar.

4. Conclusions

Inclusion of treated wood particles in the manufacture of wood-cement composite was significant in all three properties investigated, by reducing the density values (7%) and increase values of modulus of elasticity (126%) and compressive strength (70%) compared to composites made from untreated wood. Large gains in these properties demonstrate the feasibility of wood-cement composites manufacture.