Abstract: Particleboard can be produced from a mixture of different lignocellulosic materials, which can be chosen depending on the density required for the panel and its applications. The aim of this study was to evaluate the mechanical properties of particleboard with bamboo, pine and mate for a new product of high density for the special applications currently served by HDF. Particles of bamboo (Phyllostachys aurea) finely chopped sticks of mate (Ilex paraguariensis) and commercial particles of southern pine (Pinus taeda) wood were used. These particles used 100 % by weight in the panel or in mixtures of 50 % each (three mixtures) or in a triple mixture of one third each, were glued in a drum-type rotary mixer with melamine-urea-formaldehyde (MUF) resin, and pressed in hydraulic press at 120 °C and 5,88 MPa for 10 minutes, up to 6 mm thickness. The panels were produced with 900 kg/m3 nominal density and, after pressing, were conditioned at 20 ºC and 65 % relative humidity. Statistical was performed by means the variance analysis and simplex centroid experimental design, with three replicates. It was found the use of pine particles contributed mainly to increase the panel's strength and stiffness, while the use of mate particles facilitated internal bond strength. The mixtures modeling showed that the water absorption, the strength and stiffness in bending and the internal bond strength are explained by the cubic model, while the thickness swelling and hardness are explained by the quadratic model. The best physical and mechanical properties results were found for the pine, bamboo and mate same ratio mixture.

Objetivou-se, neste trabalho, testar diferentes modelos matemáticos para o ajuste de curvas de índice de sítio e avaliar o modelo de Clutter para predição do crescimento e da produção em povoamentos de Pinus taeda na região norte do Estado de Santa Catarina. Os dados de altura dominante foram obtidos por meio da medição de 216 parcelas, com área de 400 m², em 2009. Para o modelo de Clutter, utilizaram-se dados de inventário contínuo, realizado entre os anos de 2004 e 2008, em talhões com idades entre 5 e 34 anos. As estatísticas coeficiente de determinação ajustado (R²aj), erro padrão da estimativa (Syx), erro padrão da estimativa em porcentagem (Syx%) e distribuição de resíduos em porcentagem foram utilizadas para selecionar o melhor modelo. Verificou-se que o modelo de sítio mais adequado foi o de Silva-Bailey e o modelo de Clutter foi considerado eficiente para estimar o crescimento e a produção em m³.ha-1, do ponto de vista estatístico e biológico.

The objective of this study was to test different models for fitting site index curves and evaluat the Clutter model for predicting Pinus taedagrowth and yield in the northern State of Santa Catarina. The dominant height data were obtained from 216 plots measured in 2009, with an area of 400 m². To fit the Clutter model we used data from permanent plots, conducted between 2004 and 2008 in areas aged between 5 and 34 years. The coefficient of determination (R²adj), standard error of estimate (Syx), standard error of estimate in percentage (Syx%) and residual percentage distribution were the statistics used to select the best model. The Silva-Bailey model was the most appropriated to represent site index curves, while the Clutter model performed well for estimating the growth and yield (m³.ha-1), from a statistical and biological point of view.