Linear and nonlinear structural relaxation in glass

Enthalpy relaxation measurements were performed on several molecular liquid glasses as well as on a standard lead silicate glass designated NBS 711. Samples of these glasses were given various thermal histories in and below the glass transition and the resulting heat capacity peaks in the glass transition region were measured while heating on a differential scanning calorimeter (DSC). The results were compared and fitted to the Tool-Narayanaswamy structural relaxation model.;The results for the molecular liquid glasses were compared to linear relaxation data taken by frequency based measurements. The nonlinear DSC data agreed with the predictions of the linear data for the ortho-terphenyl solution but the two methods disagreed for glycerol and propylene glycol. The disagreement grew worse as the relaxation took place farther from equilibrium, indicating that the model may need revision at large departures from equilibrium.;The NBS 711 results indicated that good fits to the data could be obtained for rate cooled and moderate temperature annealing data. The fit quality deteriorated when low temperature annealing data were fit. The Adam-Gibbs relaxation time expression did a slightly better job than the Narayanaswamy expression, but no definitive superiority could be established. Several methods were tried to improve the fits including different relaxation functions, a spread of activation energies, challenging the assumption that the relaxation time depends on the average fictive temperature and a correction effective at large departures from equilibrium. None of these improved the fit quality enough to justify adding them to the model.;Hardness measurements were also performed on several of the NBS 711 glass samples. The hardness of a glass was shown to be a function of its thermal history, as the hardness increases as the fictive temperature decreases. An equation for viscous flow under an indenter using the Simmons non-Newtonian viscosity equation showed that flow under an indenter can be described by both the Douglas viscous flow model and the Marsh plastic flow model. The thermal history dependence of the hardness was then interpreted to mean that the flow stress under the indenter depends on thermal history.