Two-level self-organization of wood properties: a new paradigm for dimensional analysis and scaling
2010
Tekleyohannes, Anteneh T. | Avramidis, Stavros
Wood physical properties and chemical composition are described as self-organizing two-level systems (TLS) and the governing equations are derived. The product of the non-dimensional magnitudes of a physical property and corresponding chemical composition was able to generate power law distributions when the non-dimensional magnitudes were raised to the power of large numbers in the order of 2-26. The TLS analogy was able to reveal the mechanism that creates self-organizing criticality (SOC) from the interactions between two or more high and low dimensional components of wood. As a result of the TLS description, SOC is demonstrated to exist in the couplings of specific gravity and carbohydrates content, specific gravity and extractive substances, shrinkage and carbohydrate content, shrinkage and extractives content. The mechanism is also shown to exist within similar physical and chemical properties if the gross wood has undergone some changes like separation or extraction. Although previous studies failed to find a correlation between some wood properties such as specific gravity and celluloses content, this study revealed the existence of a scale invariant relationship. The scale invariance has an application in dimensional analysis and scaling of inputs to be used in models such as artificial neural networks. It can further have a direct application in developing power law-based models for wood properties. The feasibility of the latter is shown by establishing a scaling relation between shrinkage and dimensionally reduced variables (TLS of alpha-cellulose, lignin, pentosans and specific gravity) as predictors.
显示更多 [+] 显示较少 [-]