Associate professor, Oregon State University, Corvallis, OR
Graduate student, Oregon State University, Corvallis, OR
(Received 29 May 1998; accepted 9 February 1999)
The five-point bending test is configured as a continuous two-span beam having point loads at each midspan. It was proposed by others as an approach to evaluating the shear capacity of structural size lumber and engineered wood products because earlier tests with wood-based structural panels yielded shear failure modes. In the tests with structural panels, the thickness/depth ratio was large relative to that in structural lumber, although the length/depth ratios were the same. In addition, the density of the structural panels was greater than that of structural lumber. At the load-bearing positions, rounded bearing surfaces were used for the tests of structural panels, but flat bearing plates have been the practice in tests of structural size lumber. In this study, finite-element analyses of the bearing plates and the resulting stress distributions on structural lumber were conducted, first on an orthotropic half-plane and then in the five-point test geometry for structural lumber. The analyses used linear elastic and nonlinear compressive stiffness perpendicular to the grain and nonlinear shear modulus. Our testing showed that elementary beam theory failed to accurately predict the stresses in the critical region of the specimens. In addition, it appears unlikely that sufficient bearing can be developed in smaller lumber sizes to make the five-point bending method reflective of the true shear capacity.
Paper ID: JTE12060J