Fracture characteristics of acetylated birch - Experimental and numerical studies

Sammanfattning: Modication of wood is carried out to increase the durability and dimensional stability of wood, and one of the modication methods is acetylation. Acetylation of wood means that polymers in the cell wall react with acetic anhydride. Today, acetylated wood is rarely used in load-bearing structures and there are only a few studies on how the fracture characteristics are affected by acetylation. A study found in the literature indicates a 15-20 % reduction of the fracture energy in acetylated spruce compared to untreated spruce. Due to the small number of studies made, it is important to increase the knowledge of how acetylation affects fracture characteristics in order to determine whether acetylation is a useful modication method for structural applications. In this dissertation, an investigation is made on how birch is affected by acetylation, both through a numerical and an experimental study. A literature study is presented which includes a description of the micro and macro structures of wood, the acetylation process, basic fracture mechanics and a background of the nite element method. The experimental work, based on a Nordtest method, is performed to determine the fracture energy of the specimens. Values of the fracture energy are then used to calibrate the FE-models. 40 specimens were tested, half of which were untreated and half acetylated. The specimens had either rectangular or triangular shaped cross-section areas exposed to tension perpendicular to the grain, in order to investigate whether the geometry of the tested cross-section affected the outcome of stable or unstable response. The acetylated birch showed a 12 % higher mean value of the density than the untreated birch. The mean moisture content was 3 % in the acetylated wood and 11 % in the untreated. From the experimental work, a 55 % reduction of the mean value of fracture energy was obtained for the acetylated birch compared to the untreated birch. When comparing the results from the numerical models with the experimental results, both models predicted a value of the maximum load that was 15 % lower than what was obtained in the tests. The value of the fracture energy was 4.5 % less for the model of the acetylated birch and 3.7 % less for untreated birch. The triangular cross-section generated more stable responses compared to the rectangular.