The determination of moisture profiles in Pinus radiata timber during kiln drying

Abstract

All timber extracted from the forest contains a large amount of water, most of which must be removed before the timber can give satisfactory service in use. The process of drying (seasoning) removes the surplus moisture as economically and with as little damage to the wood as possible. The moisture gradient is a critical parameter in the drying process because high moisture gradients cause drying defects, yet small gradients prolong drying. Kiln drying is not expensive when compared to the sawing process, but the defects that can occur during drying (such as splits, checks, cracks, warp, staining, casehardening and excessive shrinkage) can greatly affect the overall drying cost. Drying schedules are used to control kiln drying to minimise the occurrence of drying defects. These schedules are primarily time based because there is a lack of moisture content measurement methods that provide an accurate and continuous measurement during the entire drying cycle.

This research investigated the moisture gradient using both measurement and prediction techniques. Measurement of the moisture gradient involved using gravimetric methods to calibrate a newly developed dielectric moisture sensor (DMS) that inferred the moisture gradient from the measurement of the average and surface moisture contents. A combined constant drying-rate and diffusion model was used to predict the moisture content and moisture gradient of a single board during drying. Monte-Carlo simulation techniques were applied to the single board model to examine the moisture content and moisture gradient variation within a stack of timber based on the difference in the initial moisture content and morphological properties of the boards.

The experimental work conducted with the DMS (only an early prototype version was available in time for these trials) was affected by difficulties caused through moisture ingress into the electronic components inside the kiln. The results obtained suggested that as drying progresses there is a significant reduction in the drying rate as the surface moisture content approaches the fibre saturation point (25-30% dry weight basis moisture content). The decreasing slope is an indication of both the reduced drying rate and a reduction in the sensitivity of the instrument at low moisture contents. Additional work is required to determine the effect of kiln temperature, humidity and wood density on the sensor calibration.

The single board drying model was used to determine the constant drying-rate and diffusion coefficients by calibrating the model drying curves to fit measured gravimetric data over a wide range of temperatures, and using timber of various densities. The coefficients that were determined were then analysed to determine the effects on them of kiln temperature and board density for use in the stochastic simulation model.

The stochastic model showed that an accurate prediction of the drying times, moisture content and gradient could be made at several different kiln settings. The predicted drying curves obtained from the model showed that a large amount of variation exists between the boards during the middle stages of the drying cycle. This variation is reduced to within 5% and 10% for 25 and 50 mm timber respectively upon completion of drying. The moisture content differences between the average and surface of 2%, for 25 mm timber, and 6%, for 50 mm timber were predicted by the model at the completion of the simulation. The largest differences were observed during the middle stages in the drying cycle.