Four-dimensional computed tomography and image processing to investigate moisture in wood

Sammanfattning: Kiln drying optimisation is a necessity to improve the resilience of sawmills against energy and environmental crises. X-ray Computed Tomography (CT) can aid to increase the efficiency of the process as it is a technique that can be applied to study moisture transport in wood during drying. The moisture content (MC) can be estimated at a sufficiently small scale, however many steps of image processing operations are required. This complexity has until now restricted the computation of MC distributions to one cross-sectional plane of the sawn timber volume, consequently limiting the potential of the method. The aim of this thesis was to develop a method to estimate the MC in each volume-element (voxel) of a CT-scanned volume of sawn timber during a drying process, and apply it to evaluate kiln drying schedules. The MC distribution was correlated with the sawn timber quality by proposing a new approach to derive moisture gradients.The developed image processing algorithms require limited pre-processing or parameter tuning, and with significantly improved computation time compared to algorithms previously suggested. The MC distributions were spatially averaged at a volume of 1 cm3 and compared with gravimetric measurements performed at the same scale. The accuracy of the method was quantified with an average R2 of X, the precision given by the average RMSE between both datasets was 4.8 percentage points (pp). The sources of error were identified as the image processing (3 pp) and the density measurement uncertainty of the CT scanner (1.8 pp). The method was applied to study the impact of aggressive drying schedules on sawn timber quality compared with a reference schedule used by a local sawmill. The evolution of moisture gradients and checking occurrence suggested that the drying simulators are too conservative, more adapted schedules can reduce the energy consumption without significantly degrading the sawn timber quality.