• Viscosity–temperature data of borosilicate melts were systematically analyzed. • AM and MYEGA models outperform VFT in viscosity extrapolation. • Low-temperature viscosity data provide more reliable extrapolations toward high-temperature regions than the opposite. • A new dimensionless parameter evaluates extrapolation reliability. • Including the calorimetric glass transition temperature improves the accuracy of high-temperature partial datasets. Borosilicate glasses and glass-ceramics are widely used in laboratory, optical, display, and nuclear technologies. Their viscosity–temperature relationship governs key processing steps such as melting, sintering, and crystallization. Here, viscosity–temperature data for various borosilicate compositions were compiled and analyzed using three models: Vogel–Fulcher–Tammann (VFT), Avramov–Milchev (AM), and Mauro–Yue–Ellison–Gupta–Allan (MYEGA). Statistical evaluation shows that AM and MYEGA outperform VFT in extrapolating partial datasets, with low-temperature data yielding superior predictions at higher temperatures. We also determine the infinite-temperature viscosity value for each model and identify compositions whose partial datasets successfully reproduce the full viscosity–temperature curve. A new dimensionless parameter ( φ ) is introduced to assess extrapolation reliability. Including the calorimetric glass transition temperature significantly improves the accuracy of extrapolated data. These findings provide a practical framework for optimizing viscosity experiments and predicting the complete viscosity–temperature behavior of borosilicate melts.
Rosante et al. (Wed,) studied this question.