Fracture development in wood resulting from bending stresses and detected using the acoustic emission phenomenon

A new approach to the problem of failure prediction of wood components and structures is that of fracture mechanics. In order to implement this approach knowledge of flaw growth (or crack propagation) in wood subjected to various stress systems and environmental conditions is required. The extension of flaws is accompanied by the release of energy in the form of acoustic or stress wave emissions, which can be detected and have been found to be reliable indicators of crack growth. Using acoustic emissions as a measure, flaw growth was investigated in stressed bending specimens of three species. The test pieces contained a variety of knot configurations, seasoning checks and resin pockets. Three wood moisture contents were used; specimen size was 2-by-4-inch (nominal) cross section and 50-inch length. A detailed description of the system used for detecting, measuring and recording acoustic emissions is given. Emission activity was measured using an electronic counter. Simultaneous count, load and deflection measurements were recorded and count-deflection and load-deflection curves plotted. The association between acoustic emissions and crack growth is discussed. Four types of integrated count- deflection patterns were found, which generally exhibited two common features. Firstly, a rate change point, where the count rate increased rapidly, and secondly a count increase just prior to failure. The rate change point was approximately coincident with the proportional limit in the load-deflection curve, suggesting an association between elastic behaviour and microfailure development. Several qualitative relationships between the count and load curves are presented. A correlation between stable crack growth and the entire load-deflection curve is postulated. The presence of checks and resin pockets affected the microfailure pattern, shown by increased acoustical activity, without appreciably influencing modulus of rupture. Green specimens for the most part produced lower counts than drier ones. Also clear or essentially clear boards exhibited lower emission counts. Species differences were not apparent. The estimation of modulus of rupture employing parameters from the count-deflection curves is discussed and compared to that using modulus of elasticity. Precision of the prediction was not improved using count parameters. A method for determining the direction of crack propagation in a fractured bending member is described.Forestry, Faculty ofGraduat