Non-destructive assessment of moisture content and modulus of rupture of sawn timber Hevea wood using near infrared spectroscopy technique

Abstract

The sawn timber Para rubber (Hevea brasiliensis) wood is an important wood product accounting for the highest export value of Thailand. The objective of this research was to build a prediction model of moisture content and modulus of rupture of sawn timber Hevea wood samples using desktop near infrared spectrometer. The timber samples were collected from the southern region and eastern region of Thailand and scanned using Fourier transform near infrared (FT-NIR) spectrometer in a range of 12489–3594 cm−1 (800-2700 nm) in diffuse reflectance mode. Then they were determined for moisture content and modulus of rupture (MOR) following ASTM. The predictive models were built by the partial least squares regression (PLSR). The result showed high performance in prediction of moisture content with correlation coefficient of prediction, Rp = 0.89 and root mean square error of prediction; RMSEP = 0.70%db. Regarding a predictive model of modulus of rupture, the results showed fair performance giving Rp = 0.78 and RMSEP = 17.11 MPa. Therefore, using near-infrared spectroscopy technique to predict the moisture content and strength based on the modulus of rupture of timber Hevea wood offered a rapid and non-destructive measurement as an alternative to the destructive checking the quality of sawn timber Hevea wood.

Timber moisture detection using wavelet packet decomposition and convolutional neural network

“…Moisture content (MC) is the key factor of the physical and mechanical properties of timber. Below the fibre saturation point, different MC could lead to both strength and stiffness changes [1,2]. Also, MC affects wood decay, fungal growth, weight gain, which could considerably reduce the safety and durability of timber structures [3][4][5].…”