Predictive capacity of some wood properties by near-infrared spectroscopy

ABSTRACT

Near-infrared spectroscopy (NIRS) is a non-destructive method that has been used in wood property assessment. Preliminary studies have revealed its ability to predict wood density and mechanical properties, but with little attention given to the effect of surface quality and wood anisotropy. This study simulates sawmill conditions variable timber surface quality and cross-sectional annual ring orientation where NIR spectra were taken on rough and smooth surfaces of quarter-sawn and flat-sawn specimens. Two models were developed based on a mixed set which included both cross-sections, and another that was based on wood from these two cross-sections, separately. Promising predictive models were obtained for density ($R_p^2$= 0.66), stiffness ($R_p^2$= 0.78) and strength ($R_p^2$= 0.82). Rough surface and quarter-sawn specimens were mostly better than their counterparts. In general, density, stiffness, and strength of wood could swiftly and with relative accuracy be assessed using NIRS, especially with rough-surfaced timber.

KEYWORDS:

• Near-infrared spectroscopy
• timber
• density
• modulus of rupture (MOR)
• modulus of elasticity (MOE)
• partial least squares (PLS)
• modelling
• surface quality

Acknowledgements

This research was partially supported by a scholarship from the MasterCard Foundation Scholars Program for Africans. Many thanks go to Professor Shawn Mansfield’s for allowing the use of his NIR system in order to carry out the experiments.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research was partially supported by a scholarship from the MasterCard Foundation Scholars Program for Africans.