Assessment of softwood lumber distortion by modelling three dimensional shrinkage variations

Thumbnail Image



Journal Title

Journal ISSN

Volume Title


University of New Brunswick


Distortion in lumber has been identified as one of the important problems for its utilization in building construction. A key objective of this study is to develop a methodology to study how lumber distortion is affected by tree growth characteristics and lumber conversion pattern. In this study, jack pine (Pinus banksiana Lamb.) and white spruce (Picea glauca (Moench) Voss) trees were harvested from a mixed-species spacing trial site. Within-stem tangential, radial and longitudinal shrinkage of wood were measured using the digital image correlation (DIC) technique. The variation patterns of wood shrinkage were modeled using a nonlinear mixed-effects modeling approach. These shrinkage models incorporate crown characteristics (crown length and crown ratio), distance away from crown base, cambial age, growth rate, and height within stem as explanatory variables and can provide input shrinkage properties for a finite element based lumber distortion model. Validation of lumber distortion model was conducted by comparing the measured and simulated distortion in lumber pieces. To provide required crown characteristics in the shrinkage models, tree height and crown base height were reconstructed based on growth ring analysis. Data from knot dissection technique on the same tree were used to calibrate the crown base height model from ring area data. The results showed that wood shrinkage exhibits different radial variation patterns in the three anatomical directions. Correlation analysis between shrinkage and wood density has confirmed the classical theory of the influence of earlywood and latewood interaction on difference between tangential and radial shrinkage values. Wood shrinkage value and variation pattern are strongly influenced by tree crown characteristics and growth rate. Tree growth conditions and stand density could have indirect influence on wood shrinkage. Overall, the lumber distortion model is able to predict the trend of the distortion, although there are some discrepancies between predicted and measured distortion values in the test lumber pieces. The shrinkage models and lumber distortion model developed can be used to study the influence of tree growth characteristics on distortion of resulting lumber but further validation using more extensive shrinkage dataset will be necessary before the results can be generalized for the studied species.