Este trabalho teve como objetivos construir, testar e avaliar o comportamento de vigas de madeira de eucalipto com emendas de topo ligadas por cobrejuntas inteiramente colados com adesivo resorcinólico e compará-las com o comportamento de vigas maciças feitas do mesmo material. Para isso, obteve-se um lote de madeira de Eucalyptus grandis seca em estufa. Realizou-se a caracterização completa da madeira de acordo com a norma NBR7190/ 97. Protótipos menores das vigas foram construídos a partir de tábuas retiradas ao acaso com as dimensões de 10 cm de largura, 3 cm de espessura e 170 cm de comprimento nas testemunhas e 85 cm nas duas partes da alma das vigas a serem coladas com cobrejuntas compostas por tábuas de 35 e 43 cm de comprimento e 1,5 cm de espessura. Essas vigas com 160 cm de vão foram ensaiadas até a ruptura. Concluiu-se que a melhor configuração foi o cobrejunta que ocupava 25% da superfície lateral das tábuas. As vigas de tamanho real foram construídas com tábuas de 20 cm de largura, 6 cm de espessura e 420 cm de comprimento nas testemunhas e duas tábuas com 210 cm de comprimento e mesma espessura e largura nas vigas com emendas. Estas apresentaram excelente desempenho em termos de resistência, em alguns casos, superando a resistência da viga-testemunha. Quanto à rigidez, o desempenho foi considerado adequado, já que os deslocamentos das vigas com cobrejuntas foram em média 20% inferiores aos da viga-testemunha submetida a uma mesma magnitude de carga, o que as proporciona melhor desempenho estrutural.
The objective of the present work was to construction, to test and to evaluate the behavior of wooden beams made of eucalyptus boards with mid-span end-grain joints. A resorcinol-formaldehyde glue was used as adhesive and the joints were covered by side members of the same species of wood. Strength and stiffness of these beams were compared with the results obtained for solid beams. Eucalyptus lumber was purchased for this purpose. Wood properties were determined for these two sub-samples, according to the recommendations of the Brazilian Association of Testing Standards, NBR 7190/97. Prototype beams were built to find the best size of the side members that symmetrically covered the joints. These beams had a length of 170 cm, 3 cm of width and 10 cm of deepness with cobrejuntas consist of boards 35 and, or 43 cm long and 1.5 cm thickness and were tested to failure using a center load on a 160-cm span. It was found that a 1.5 cm thick side member should cover at least 25% of the two sides of the beam to give the same results of strength and stiffness of solid 170 cm long beams. Larger beams were then built and tested. They had a length of 420 cm, 20 cm of deepness and 6 cm of width the controls were built of a single piece of lumber; the others were made of 210 cm long boards, joined at mid-section by 100 cm long, 3 cm thick side members. After the test, these glued beams presented strength similar to the control witness, but the displacements of beams with cobrejuntas were on average 20% lower than the control, submitted showing a better structural performance.
