1. Introduction

Beams are structural elements present in most of buildings. Among the usual materials engineering highlights the wood, to be from natural and renewable source, low density and good mechanical performance. Timber structures when not treated properly can present problems due to the attack of biological degrading agents that contribute to the loss of their physical and mechanical properties, compromising the integrity of the structural components.

The study of repair and reinforcement in the structure of wood has been the focus of technical and scientific papers, aimed at developing viable solutions to be used in the recovery of the same[1-7].

Of the possible materials used as reinforcement and repair wooden structures stand out from the composites, because it is a material designed, in order to obtain a resultant mechanical properties superior to those of constituent phases[8].

The use of vegetable fibres such as sisal [9-14], coir, jute, bananaand bamboo as reinforcement in laminates composite are considered as a good solution, show good tensile strength[15-18], and are materials biodegradable and of low cost when compared to synthetic fibres[19].

With the purpose of developing alternatives as reinforcement in beams, this paper aims at the development and characterization of composite laminated polymer matrix reinforced with sisal fibres to be used as reinforcement in Eucalyptus grandis and Pinus elliottii timber beams. The wooden beams with and without the use of the composite laminate is tested in bending, by making use of the static three point bending tests, and comparing the maximum strengths condition to the faultless timber, and defective unreinforced and reinforced, and faulty, making it possible to evaluate the efficiency of the manufacture composite.

2. Material and Methods

The raw material used is a vegetable fiber and sisal as reinforcement and resin epoxy as matrix fase. The laminate composite was manufacture with a layer. The fiber used was obtained from the Sisal company (Brazil), with caution as the use of fibres from the same batch. The Pinus elliottii and Eucalyptus grandis timber used in the fabrication of the specimens was obtained in a local sawmill in São João del-Rei (MG-Brazil), having as a precautionarypre-screening of samples free defects.

Brackets have been manufactured of cast iron with 225 mm by 160 mm wide synthetic enamel coated. The sisal fibers were woven in a direction perpendicular to the length of the square, so as to stay closer to each other (Figure 1), and with tension on the nodes from the seams positioned on the metal rods, not allowing the presence of nodes in the structure of the laminate.

Figure 1. Woven of sisal fibres

To elaborate the composite material, the volume of the fibre should correspond to thirty precents of the total[8], and the remaining seventy percents should correspond to the volume of resin. From these data the total weight of resin to be applied in the composite was then calculated.

Figure 2. Dimensions of the specimens: (a) Flawless; (b) defective

Figure 3. (a) Laminate sisal; (b) Flawless timber; (c) Bonding area; (d) Composite fixed on timber

Eight specimens of timber, four of each species, were made by sawing them prismatic shape with square cross section with dimensions60×2.5×2.5cm (Figure 2). Four of these (two of each species) were damagein the centre of their bases, measuring 8×2.5×1cm (Figures 3b and 3c). Finally, two of the specimen defective (one for each species) have been reinforced with the laminate (Figure 3a) and its adhesion was performed by use of the resin while maintaining a bonding with 10cm²of area of each side groove (Figure 2c) and curing by seven days. For adhesion of the laminate to the timber (Figure 3d) was used in the same proportions resin used to manufacture the composite.

The mechanical bending tests were performed in an EMIC testing machine with loading speed of 1 mm/min. The modulus of elasticity (E_m) and strength flexural modulus (f_m) of the specimenswithout defect (no failure) was obtained according to the Brazilian standard NBR 7190[13], respectively expressed by Equations 1 and 2, F_10% and F_50% and 10% and 50% of maximum load (F_max), L is the length of the useful parts (distance between supports) and b and hthe width and height measures of the cross section respectively.

(1)

(2)

The dimensions of the specimens following the L≥21·h relation, neglecting the effect of shear forces in the calculus of the displacements[20-22].

3. Results andDiscussions

Os testes realizados com as madeiras íntegras geraram fraturas frágeis (Figura 4a) e também por propagação de trincas (Figuras 4b, 4c e 4d).

Figure 4. Fractures in timber due to the imposition of load on the bending test. (a) Fragile Fracture – Pinus elliiottii; (b) Fracture by crack propagation – Pinus elliiottii; (c) and (d) Fracture by crack propagation – Eucalyptus grandis

The tests performed with the wood cracked generated crack propagation precisely in points where there stress concentrators, as shown in Figure 5.

Figure 5. Fractures in the woods cracked by the imposition of load on the bending test. Fracture by crack propagation in a region of stress concentration: (a) Eucalyptus grandis and (b) Pinus elliiottii

Finally, tests carried out with the additional have differentfailure mechanisms. While in the Pinus was disruption of the composite material (Figure 6), there was a break in Eucalyptusspecies, which has subsequently damage the composite material (Figure 7).

Figure 6. Specimen of Pinus intact with rupture of the composite material in bending test (left top)

Figure 7. Specimen of Eucalyptus intact (top) and its disruption in the bending test (bottom)

Figure 8 illustrates the behaviour of the relationships between displacements and forces applied in the specimens obtained during the bending tests for the eight experimental conditions.

Figure 8. Results obtained in the bending test

From Figure 8 it is noted that the intact samples shows the maximum force (F_MAX) higher than the strengthened, which in turn was superior to those at the specimens with defects.

It is also noted that Pinus support a load lower than the Eucalyptus in all conditions tested, but can also be seen in Table 1.

Table 1. Maximum load achieved in bending tests. Experimental Conditions FMAX (N) Intact Pinus 1235.64 1873.07 Flawless Pinus 264.78 Reinforced Pinus 441.30 Intact Eucalyptus 1549.45 2530.12 Flawless Eucalyptus 529.56 Reinforced Eucalyptus 627.63

The cracks in Pinus provided an average drop in maximum load in relation to intact timber about 82.96%, while the reinforcement was able to increase the maximum load supported by 66.66% compared to flawless Pinus.Eucalyptus already cracked gave an average drop in maximum load in relation to intact timber about 74.04%, while the reinforcement was able to increase the maximum load supported by 18.52% compared to flawless Eucalyptus.Given the above, it appears that when the timber is not enhanced fracture during the imposition of charges, the strengthening of the composite material has become more efficient.

Table 2 shows the individual values of the modulus of elasticity in bending (E_m) and flexural strength modulus (f_m) obtained for the intact Pinus elliiottiiand Eucalyptus grandis timber.

In Table 2, the Eucalyptushad a higher modulus of elasticity and flexural strength modulus that Pinus timber.

Table 2. Mechanical properties of the timber obtained by the static three points bending test FMAX (N) fm (MPa) Em (MPa) Pinus 1235.64 6.17 8250 1873.07 9.35 13544 Eucalyptus 1549.45 7.73 10428 2530.12 12.63 17041

4. Conclusions

Currently, researches are being directed to the production of laminates for structural reinforcement and low cost. These factors are affected by material selection, environmental conditions of rolling, the characteristics of the tooling and manufacturing methodology.

After a few tests on the materials presented in this work, we can conclude that the flawless specimens had a considerable reduction of its resistance to bending in relation to the intact timber. The addition of natural fiber reinforcement allowed reasonable increase in the flexural strength modulus of the flawless timber.

In future studies, we intend to evaluate other bonding areas, new timber species and the variation in the number of layers used in the preparation of the composite.