Estimation of tensile strength in timbers of ten eucalypt species as a function of anatomical parameters and apparent density

Abstract

Characterization of construction materials is a piece of essential knowledge to ensure the structural stability and strength of any building. The strength of materials is an effective prediction prescribed by technical stan-dards, which delimit the resistive behaviors in different conditions. Physical-mechanical properties of timber allow the rational application of this bioresource in structural projects. However, tests to determine mechanical properties require large equipment and a skilled workforce, found in few institutions and specific research centers. The microscopy and regression models may establish an alternative to understanding the properties of wood. The present paper aimed to demonstrate the feasibility of estimating the mechanical properties of Eucalyptus wood as a function of its anatomical characteristics and apparent density. Ten batches of different eucalypt species were selected to predict the tensile strength parallel to the grain, tensile strength normal to grain, and modulus of elasticity in tensile parallel to the grain, as well as three different anatomical parameters (fibers, vessels, and rays) of the wood samples according to two Brazilian standard documents. In practice, the proposed regression models were excellent, as the R2 values were equal to 90.55, 86.55, and 67.95% for estimating Et0, ft0, and ft90, respectively.