Dynamic sorption and hygroexpansion of wood wafers exposed to sinusoidally varying humidity

Abstract

Round wood wafers, 4 mm thick along the grain and 2 cm in cross-sectional diameter, of green basswood (Tilia americana L.), yellow birch (Betula alleghaniensis Britton), and black cherry (Prunus serotina Ehrh.), and initially dried to equilibrium in air of 77% relative humidity and 25°C, were exposed to sinusoidally varying relative humidity between 77 and 47% at 25°C for many cycles at each of four different cycling periods, 5.33, 10.67, 16.0 and 25.33 hours. Moisture changes and radial and tangential dimensional changes in response to the imposed humidities, measured during initial drying and subsequent cycling, gave the following results: 1. The moisture and dimensional changes were generally sinusoidal but lagged behind the imposed humidity. The phase lag decreased and the amplitude increased with increasing cycling period. Both responses and phase lags approached repetitive or “steady-state” values as cycling was prolonged. 2. A numerical solution for moisture diffusion, assuming a constant diffusion coefficient and sinusoidally varying boundary moisture conditions, was used to simulate the average moisture content in the wafer at any time. The resulting curves were qualitatively similar to those obtained experimentally, the differences attributed primarily to the effects of hysteresis and stress relaxation. 3. The mean moisture diffusion coefficients, calculated from steady-state phase lag data combined with an analytical solution of the diffusion equation, decreased with increasing cycling period. The values obtained increased with decreasing wood specific gravity as anticipated, but their magnitudes were somewhat lower than theoretical values. 4. The dynamic moisture expansion coefficient was relatively constant during successive cycles, with no consistent effect of cycling period. The dynamic values were generally higher than the static values. 5. The dynamic humidity expansion coefficient increased with increasing cycling period. It was only about half that obtained from static experiments, presumably because of hysteresis. 6. The dynamic moisture sorption coefficient, which is the effective slope of the dynamic sorption isotherm, decreased with increasing number of cycles and decreasing cycling period. It was less than half the calculated static sorption isotherm in the same humidity range, presumably due to hysteresis.