Abstract
Literature mechanical relaxation data of two glasses, vitreous silica and poly(methyl methacrylate) (PMMA), an amorphous polymer, are interpreted in terms of the Gilroy-Phillips model. For freauencies up to 1 GHz, one finds a temperature-independent barrier density function f(V) at all temperatures in the glass phase. If one looks at still higher freauencies, with the picosecond optical techniaue or with Brillouin, Raman and neutron scattering, one can sample the same barriers at a higher temperature. As long as the high-freauency measurements are made at Iow temperatures, up to about one fifth of the glass temperature, one finds again the same barrier density function. For higher temperatures, the barrier density function of the high-frequency data exceeds that determined at lower freauencies. This holds for the silica neutron data from the literature as well as for new neutron data on PMMA reported here. The finding supports earlier Raman results from other glasses in the literature. The deviation from the simplest form of the Gilroy-Phillips model is not a strong effect (at least not as long as one stays in the glass phase), but clearly measurable. Possible reasons for the deviation are discussed.