Nondestructive Evaluation of Wood Properties in Standing Trees Using Vibrational Spectra

Abstract

The objective of this study was to estimate the moisture content, wood density of corresponding moisture content and modulus of elasticity in standing trees using vibrational spectra. For the preliminary tests, fixed length and diameter logs were used for the following experiments. The logs were impregnated with water at normal pressure and the vibration tests were carried out at fixed intervals while the logs were left to dry under room conditions. The statistical models for the prediction of the wood properties were built using the vibration spectra generated from three excitation methods, namely, excitation of longitudinal vibration by hitting the transverse surface of logs, and excitation of lateral vibration or longitudinal vibration by hitting the log’s lateral surface. The Lasso (least absolute shrinkage and selection operator) and PLS (partial least squares) models could predict the moisture content and density independently of the excitation, with the exception of the modulus of elasticity. The vibrational spectra obtained by hitting the lateral surface of logs could give good prediction models, indicating that the techniques suggested in this study have high potential for application to standing trees.