Development of an andvanced fe-numerical method for virtual grading of timber

Abstract

Structural heterogeneities in wood are changing the complexity of the numerical modelling of this material. Knots are the main strength governing parameters, which are changing the uniformity of the stress distributions through the boards. These weak locations may result in failure in the material. Comprehensive reconstruction of the boards, based on the knot data on the surfaces of the boards makes it possible to access the virtual model of the element and predict new material behaviors. The aim of the current study is to develop a more technologically advanced numerical method for grading of the wooden boards. Simulations are run using ABAQUS and PYTHON. By programming the structural model of the boards, covering the total quality range, the mechanical behavior is predicted under a uniform tension. Simulations are run for low-medium quality Douglas fir and mediumhigh quality spruce boards. Based on the extracted numerical parameters, the strength of the boards are predicted virtually. The results of the simulations are benchmarked to the currently available visual and machine grading methods. By getting a promising correlation between the tensile and the predicted strength results, a more advanced numerical method is provided for virtual strength grading of the boards. The method is later applied for the reconstruction of the glulam model based on the information of the single lamellas.