Structural relaxation of a medieval cathedral glass

Abstract

Window glasses are obtained by rapidly cooling a liquid, generating a non-crystalline structure out of thermodynamic equilibrium. This similarity to liquids leads to confusion as, for example, an urban legend that the windowpanes of medieval cathedrals drained slowly over the centuries and grew thicker underneath. Approximate calculations have shown that the time required for typical window glass to flow at room temperature is absurdly high, with variations in thickness caused by the imprecise methods of manufacture. It is necessary to extrapolate the viscosity value of the material from the glass transition region to room temperature to perform these calculations. Recent works have shown that the experimental values of the relaxation times are close to those calculated using the equilibrium viscosity and that the temperature dependence of relaxation and viscous flow appears to be somewhat different. Still, the relaxation times are entirely different from those calculated using the non-equilibrium viscosity used in the latest works. Therefore, we intend to infer whether the temperature dependence of the experimental relaxation times is the same as those calculated via the Maxwell equation and to recalculate with greater precision the time required for a medieval cathedral glass to flow at some temperatures, including ambient. To this end, we plan to produce a glass with a medieval composition, characterize and measure properties at temperatures close to the glass transition, and then more accurately extrapolate the data to room temperature.