Modeling strength loss in wood by chemical composition. Part I. An individual component model for southern pine

Abstract

In this study, we develop models for predicting loss in bending strength of clear, straight-grained pine from changes in chemical composition. Although significant work needs to be done before truly universal predictive models are developed, a quantitative fundamental relationship between changes in chemical composition and strength loss for pine was demonstrated. In particular, this study explored a linear independent-component modeling approach. The models were evaluated across a range of environmental exposure conditions known to cause strength loss and with several chemical treatments capable of causing hydrolytic chemical degradation in wood. Simple linear models developed reasonably accurate predictions of strength loss of clear, straight-grained southern pine wood based on changes in its chemical composition. Side-chain sugars of hemicellulose were the most susceptible to acid hydrolysis. The extent of their degradation was a sensitive predictor of early strength loss. Those sugars associated with the hemicellulose backbone were the next most susceptible, but they were strongly correlated between themselves. This is known as collinearity and, as such, data from either mannose or xylose, or from Klason lignin or glucose, often precluded the need for the other in the models. A linear three-parameter model using changes in a side-chain hemicellulose (arabinose), a main-chain hemicellulose (mannose), and glucose as an indicator of the extent of cellulose degradation reasonably predicted bending strength loss. We believe that with further work, residual strength or serviceability models based on a linear accumulation of the changes in chemical composition of wood during microbiological attack, thermochemical treatments, or severe environmental exposures can be developed to provide sensitive predictors of post-treatment or in-service strength loss.