Dynamic Mechanical Studies of Wood, Paper and some Polymers Subjected to Humidity Changes

Doktorsavhandling, 1997

The dynamic mechanical response of several hygroscopic materials including wood (Scots pineveneer), paper (unbleached sulphate), cellophane, polyamide 6 (PA6) and poly(vinyl acetate)(PVAc) have been studied especially under conditions of a stepwise varying humidity in the atmosphere surrounding the samples. In all cases the in phase and out of phase components of the force necessary to achieve a given deformation were measured for various oscillation frequencies (from 0.005 Hz to 15 Hz). Of particular interest was the loss tangent tan.delta. obtainedas the ratio of the out of phase part of the measured force divided by the in phase part thereof. The relative humidity of the surrounding atmosphere was varied between 5% and 85% in a stepwise cyclical manner. When the relative humidity is raised the sample swells anisotropically and the elastic modulus decreases while the loss modulus also varies. In three point bending the force required to bend the sample a given amount often increases on humidification because the stiffness increase resulting from swelling the sample overcompensates the elastic modulus decrease. For uniaxial tension this is usually not the case. The loss tangent also varies with the moisture content of the sample. None of these changes is instantaneous, among other things because it takes time for the moisture to diffuse into the sample. This diffusion is typically non-Fickian. Here effective or best fit diffusion coefficients for absorption and desorption have been evaluated for the different materials. The time scale of this absorption and desorption is compatible with the time scale of the observed changes in the vibration force necessary to bend or stretch the sample after changes in the relative humidity of the surrounding atmosphere. Characteristic of the loss modulus or out of phase component of the vibration force is that there is a pronounced transient peak each time the humidity of the surrounding atmosphere is changes. This is reflected in observed peaks in the loss tangent tan.delta. The height of these peaks is stronglydependent on the vibration frequency, increasing with decreasing frequency in the range studiedhere. This indicates that the characteristic relaxation time involved with this process is of the order of seconds or minutes. The relaxation can be related to the "chemical" interaction between water molecules and binding sites in the polymer which also affects the strength of the material and is therefore interrogated by the mechanical vibration. Such a relaxation must be very complicated, typically involving more than one water molecule before coming to completion through a series of rearrangements and exchanges. Typical of such complicated phenomena is that they are relatively slow, as illustrated by the relatively long characteristic relaxation time noted above. Some experiments involving impregnation of wood and paper with a view to inhibiting moisture diffusion in the sample or blocking sites of water molecules were carried out to test the above model. In general the experimental results support the model in that for instance acetylation of wood and paper, which blocks water binding sites, reduces the strength of the loss tangent peaks. Treatment of wood with PEO, which inhibits diffusion significantly, increased the response time as observed in the dynamic mechanical measurements correspondingly. Similar results were obtained on treating the wood with glycerol.