The Consequences of Wood Cellular Structure and Rolling-Shear in Crossbanded Veneer Composites

Simo Koponen The Consequences of Wood Cellular Structure and Rolling-Shear in Crossbanded Veneer Composites 28.04.2005 Helsinki University of Technology Laboratory of Structural Engineering and Building Physics Pages:23+ app. 119 ISBN: 951-22-7678-X ISSN: 1456-4297 A hierarchical modelling path of wood and wood composites’ properties is presented. First the calculation model (WOOD123) is developed for elastic and shrinkage behaviour based on the complex ultra-, microand macro-structure of wood, based on the properties of wood main constituents and their orientation. Then the results and the understanding obtained are utilised in the development of the calculation program for the analysis and design of transversely loaded layered wood composite plates (OptiPly). The model (WOOD123) predicts the elastic and shrinkage properties of wood at the cell wall level and at the macro-level, where wood substances consist of earlywood, latewood and ray cells. The modelling of the cell wall properties and behaviour is based on the properties of cellulose, hemicellulose and lignin. At the cellular level wood substances consist of individual softwood cell types (earlywood, latewood and ray cells). Finally at the highest level of the cellular modelling, earlywood, latewood and ray cells are combined together to predict anisotropic elastic properties of wood, the effects of moisture content on the elastic properties and the nonlinear shrinkage in different directions. WOOD123 model, developed by this author, has been the first model capable of building the full path from the ultraand micro-scale material properties, polymer orientation and arrangement to the orthotropic behaviour of wood. Earlier models are limited only to the prediction of cell wall properties or the layered structure of cell wall is neglected and the behaviour of cellular wood substance is modelled using homogenous cell walls. Due to the lack of suitable micro-scale testing devices and inadequate microscopy facilities and due to the large number of structural and material property parameters required in the model as input data, the development and verification of the model have been done mainly based on the material and structural parameters obtained from the literature. The elastic properties, the dependence of the elastic properties on the moisture content of wood and the complex shrinkage behaviour calculated by the model agreed well with the elastic constants and material behaviour presented in the literature. Next step was the modelling of the behaviour of layered wood composites. For the development and verification of OptiPly-program large experimental work has been carried out. Tests contain plywood and LVL manufactured using birch and spruce veneers, different veneer thicknesses and special lay-ups, shortand long-term tests were performed and test methods varied from standard tests to more complex beam-type and point-loaded plate tests. The tests showed that the shear deformations in cross veneers have a strong effect on the deflections and the strength of wood composite plates. However, the conventional theory based on Love-Kirchoff hypothesis does not take into account pronounced shear deformations of cross layers. The accuracy of the classical theory and the generalised Bernoulli’s hypothesis is compared. The calculation based on the exact solution of the generalised Bernoulli’s hypothesis by Heinisuo correlates well with the test results. The accuracy of prediction of the mid span deflection obtained using the generalised Bernoulli’s hypothesis is – 16%...+10%, and by the classical lamination theory is –26%...+13%, respectively. OptiPly-program was found to be reliable, slightly conservative and suitable for the practical design of plywood plates. At this moment Finnish plywood and LVL industry uses OptiPly-program in practice.