Particleboard Manufactured with Bicomponent Polyurethane Resin Base on Castor Oil

This study aimed to investigate, with the aid of the Brazilian NBR 14810:2002 standard, the physical properties: bulk density and moisture and mechanical properties: strength flexural modulus and internal compliance of part icleboards manufacture with bicomponent polyurethane resins derived from castor oil and Cambará, Canelinha and Cedrinho fibres. The factors investigated were: nominal density (0.80 g/cm3), nominal thickness (10mm), moisture content of the particles (5%), percentage of resin (15%), t ime used in the pressing cycle (10min), hot pressing (100°C), pressing pressure (5MPa) and length of the particles (3 to 5mm), leading to a fu ll factorial design of 2 type, provid ing four different experimental conditions. The mechanical properties of both materials prepared showed higher values than the stipulated by the standard. Physical properties, only the density has overstepped the bounds set by the NBR 14810:2002, rating the particleboards produced as high density.


Introduction
Brazil and world wide researches have been developed focusing the particle boards, aiming to verify the feasibility of developing panels by some experimental factors and stipulated levels [1,2]. Silva et al.[3] evaluated the equilibriu m mo isture content of wood panels using cluster samp les and mult ilaminate plywood, hardboard, Oriented Strand Board (OSB), Mediu m Density Fiberboard (MDF), High Density Fiberboard (HDF), wood-cement and wood-plastic. A mong others, the authors conclude that Nelson's equation is efficient for estimating the equilibriu m mo isture content in all cases studied, no significant differences in hysteresis between the products studied, revealing that all have the same dimensional stability. Akgüla and Çamlibelb[4] evaluated the strength and stiffness of particle boards made fro m Rhododendron timber, founded in abundance in the Black Sea (Turkey), part icles having controlled humidity of 14% and held together by adhesive using urea-formaldehyde. The results indicate the use of Rhododendron timber in the manufacture of panels.
Belin i et al. [5] produced and evaluated MDF with three different conditions of chip shredding system, evaluated the anatomical features of wood cell co mponents. Among the results, the authors conclude that increasing the intensity refining of wood chips reduced the average fiber length and increased the percentage of broken fibers, and cause a decrease in the number of vessels and parenchyma cells.
Belin i et al. [6] evaluate the production of M DF using three samples of Eucalyptus grandis wood chips. The results of anatomical and morphological analyses were discussed and compared with information fro m specialized literature and correlated with the production process of MDF.
Abreu et al. [7] developed and evaluated waste wood panels intended to production of objects to compare their qualities, being used ten graphic designs reproduced by craftsmen in the industry. The authors concluded that the mixing o f d ifferent waste panels affected the evaluation of the attributes finishing, beauty and colour, no significant difference with respect to the attributes functionality, weight, proportion and symmetry. Saffian et al.[8] studied the feasibility of production of particle boards made with clones RRIM 2020 with four years of age, were evaluated the modulus of elasticity (M OE) and strength (MOR). The results indicate that is possible to manufacture panels with the rubber clones evaluated.
One aspect to consider due to the industrialization of particleboards is to use products that pollute the environment, mainly through the emission of gases. Accordingly it is necessary to develop new products, such as the proposed study by Bradi et al.[9], having considered the influence of a mixtu re of vegetable o il in the polyurethane matrix strength of wood fiber panels. The analyses possible to conclude that it is possible to use mixtures of vegetable oil in the polyurethane matrix in the ratio 35:65 (by weight) for the manufacture of wood fiber panels.
Silva and Lahr [10] evaluated the production of particleboard with homogeneous particles of wood in the Amazon for low and mediu m density (Erisma uncinatum, Nectranda lanceolata, Erisma sp). Particleboards were made with nominal thickness of 10 mm, no minal density of 0.75g/cm 3 , using 10% of bico mponent polyurethane resin derived fro m castor oil and part icles with a length fro m 0.02 mm to 6 mm, 90℃ of hot pressing, 40MPa of pressure and time of 10 minutes. In the assessments according to NBR14810:2002 standard, panels of wood particles Nectranda lanceolata showed higher strength values, which are above the limit set by the standard.
Dias et al. [11] evaluated the mechanical properties of plywood panels made of polyurethane resin based on castor oil. The results for the M OE did not reach the minimu m value of 18M Pa, being justified by poor distribution of the adhesive during the formation process of the panels.
Fiorelli et al. [12] developed particleboards with sugar cane bagasse and polyurethane resin derived fro m castor oil investigating the response variables: bulk density, swelling, absorption and modulus of elasticity and strength modulus. The results indicated as those made of h igh density, suitable for industrial use, showing the efficiency of the polyurethane resin.
Paes et al. [13] evaluated the effect of the combination of pressure (2.0, 3.0, 3.5 M Pa) and temperature (50, 60, 90℃) in particleboard with Pinus elliottii wood and polyurethane resin derived fro m castor oil on the response variables: density, swelling and water absorption (0-2h, 2-24, 0-24h), strength modulus, screw pull-out and internal adhesion, concluding that the combinations: 3.0 MPa and 90°C and 3.5 MPa and 60℃ showed the best results, proving to be a pressing temperature the most important variable for quality (workmanship) of the panels.
Sartori et al. [14] evaluated the mechanical performance of particleboards manufactured with reforestation wood, sugar cane bagasse and bicomponent polyurethane resin derived fro m castor oil as an alternative to the closure system side of the trunk collective management center for beef cattle. The physical and mechanical propert ies obtained confirmed the efficiency of the proposed structural model for use in management center.
According to the Brazilian Association of Wood Panels [15], Brazil is one of the most advanced countries of the world in the manufacture of part icle boards, whose annual production currently amounts to 612.000m 3 , accounting for worldwide production a very low percentage considering the wood potential of the country and the technology installed.
Considering the positive aspects of the current MDF production in Brazil and the need for studies that will permit the use of new adhesives, this study aimed to evaluate, with the aid of the Brazilian NBR 14810 [16] standard, the physical properties: bulk density and moisture and mechanical properties: strength flexural modulus (MOR) and internal adhesion (AI) of part icleboards produced with bicomponent polyurethane resin derived fro m castor oil and Cambará, Canelinha and Cedrinho wood fibres, making it possible to evaluate the potential of the materials developed.

Material and Methods
The factors and levels investigated in the preparation of the panels are nominal density (0.80 g/cm 3 ), nominal thickness (10mm), mo isture content of the particles (5%), resin (15%), cycle t ime used in pressing (10min), hot pressing (100℃), the pressing pressure (5MPa) and length of the particles (3 to 5 mm). With respect to procedures for producing the panels, it should be noted: -Addition of 15% o f bico mponent polyurethane resin: for each component was considered the addition of a part corresponding to 50% of the total content of the resin and according to the manufacturer's instructions, was initially added to and mixed with the substrate content corresponding to 1 part polyol, and then the content corresponding to a part of prepolymer. Th is procedure provided good homogenizat ion of resin to the substrate and prevented expansion reactions; -Pressing cycle: was to apply a constant pressure of 5MPa over 3 minutes, followed by the pressure relief of the press for a time of 30 seconds and subsequent use of the pressure of 5MPa for 7 minutes. This procedure was used for making up all the panels, there were no format ion of bubbles and cracks in the particleboards produced; -Temperature of pressing: used 90 o C, avoid ing bubble formation, leading to panels with good qualities; -Press pressing: the particleboards were p repared emp loying 5MPa, pressure used in the industrial production of mediu m density panels.
The particleboards were made fo llo wing the procedures as shown in Figure 1.
Were produced two panels for the investigated conditions, with dimensions 400×400×1mm.
Seven specimens were extracted fro m each panel produced to determine the strength flexu ral modulus (M OR), with no minal dimensions of 50×250×10mm.
For internal adhesion (AI) were extracted t wo samples of each specimen used to determine the strength flexural modulus, with nominal dimensions of 50×50×10mm.
With respect to density and moisture, each specimen used in the bending test were extracted two another specimens, with no minal dimensions of 50×250×10mm. Table 1 shows the mean and the standard deviation (S d ) values of the mechanical properties of the materials developed. According to NBR 14810 [16] standard, the min imu m suggested for strength flexu ral modulus (MOR) and internal adhesion (AI) should be 18MPa and 0.4 MPa respectively. The values of MOR and AI for both wood species were above the limits established by the standard. The lower values of MOR and AI were obtained for the panels made with Canelinha wood, still being 1.17 and 2.12 times higher than the standard.

Results and Discussions
The average values of bulk density (d ap ) and moisture (u m ) of the panels for both species are shown in Table 2. The bulk densities of the panels for both wood species had values above the threshold fro m 0.55 to 0.61 g/cm 3 established by the Brazilian standard, except made with Cedrinho, can be classified as high-density panels (HDP), and the moisture of the panels showed values within the limits 5-11% set by the NBR 14810 [16].

Conclusions
The analysis of the strength flexu ral modulus, the bulk density and mo isture allo wed in ferences about the manufacturing process of the panels and consequently on the experimental parameters used: -The mo isture of the particles (5%) did not result in expansion reaction of the bico mponent polyurethane resin which impair the homogeneity of the mixture; -The time pressing (10 minutes) was sufficient to cure the resin, however, the use of 100 o C caused bubbles to form in the surfaces of the particleboards; -The pressing pressure (5MPa) was suitable for compression of the panels that acquired density exceeding 0.8 g/cm 3 and a thickness of 10mm, as planned; -The use of the resin in the proportion 15% (dry weight basis) gave strong adhesion between the particles, because the values of sheet properties are better than the values suggested by NBR 14810 [16].
Even considering the significant results, mainly for the mechanical properties, it is necessary to adjust the variables of the process of manufacture the panels aiming the use of smaller amounts of resin, allo wing reduction of cost of the materials produced.