Glass viscosity as a function of temperature and composition: A model based on Adam–Gibbs equation

Abstract

This paper shows that a generalized Adam–Gibbs relationship reasonably approximates the real behavior of glasses with four temperature-independent parameters of which two are linear functions of the composition vector. The equation is subjected to two constraints, one requiring that the viscosity–temperature relationship approaches the Arrhenius function at high temperatures with a composition-independent pre-exponential factor and the other that the viscosity value is independent of composition at the glass-transition temperature. Several sets of constant coefficients were obtained by fitting the generalized Adam–Gibbs equation to data of two glass families: float glass and Hanford waste glass. Other equations (the Vogel–Fulcher–Tammann equation, original and modified, the Avramov equation, and the Douglass–Doremus equation) were fitted to a float-glass data series and compared with the Adam–Gibbs equation, showing that Adam–Gibbs glass appears an excellent approximation of real glasses even as compared with other candidate constitutive relations.