A mathematical model will be develope d to determine the liquidus temperature of glass. This is an important physical parameter for the glass manufacturers. The research-work comprises of: a. experimental determination of liquidus temperatures and primary crystalline phase. b. creation of a theoretical and a semi-empirical mathematical model for liquidus computations. c. computing the parameters from the experimental results for the best- fit in a semi-empirical mathematical model.
Both empirical and thermodynamical modelling have been used to predict liquidus temperatures of simple soda lime silica melts. 50 complex glasses of this type with additions of potassium oxide, magnesium oxide, aluminium oxide, barium oxide, strontium oxide, boron oxide and lead oxide have also been made and their liquidus temperatures measured. Both sets of data have been used to develop as good a model as possible for predicting liquidus temperatures of glasses in the region of the complex 10 component system studied. It had been expected that using the thermodynamic modelling to predict the liquidus of the appropriate simple soda lime silica glass, then making corrections for the difference caused by adding each minor component, would prove most satisfactory. In fact one polynomial, determined by multiple regression and covering the whole set of data, proved best.
Investigations were carried out to develop models for the prediction of liquidus temperature of multicomponent silicate glasses, in this way helping to reduce processing costs of glass production.
The project had also to minimise costs of glass production processes by applying models predicting the liquidus temperature of silicate glasses, relating the glass chemical composition to viscosity and temperature through 2 alternatives, phenomenological model based on regression analysis of observed couplings of chemical composition subsequent liquidus temperature and viscosity, or a more accurate but costlier model, relating chemical and thermodynamic observations. An aggregate model, calibrated through regression analysis, linking glass composition and its liquidus temperature was developed. An alternative (and disaggregate) thermodynamic model was abandoned because of lack of sufficient data to calibrate its many thermochemical variables.
THE LIQUIDUS TEMPERATURE OF GLASS IS AN IMPORTANT DATA FOR THE GLASS MANUFACTURERS. THE RESEARCH WORK COMPRISES EXPERIMENTAL DETERMINATION OF LIQUIDUS TEMPERATURE, PRIMARY CRYSTALLINE PHASES AND HEAT CAPACITIES DETERMINATION. BESIDES EXPERIMENTAL WORK, A MATHEMATICAL MODEL WILL ALSO BE DEVELOPED TO IMPROVE THE EFFICIENCY OF LABORATORY EXPERIMENTS.
AN ECONOMICAL ASPECT OF THE WORK IS THAT IT ENABLES GLASSMAKERS TO OPTIMISE FORMULATIONS TO PRODUCT REQUIREMENTS WITH RESPECT TO RAW MATERIALS COST PRICE. FOR INSTANCE IF WE ARE ABLE TO REACH A CERTAIN VISCOSITY WITH SAVINGS IN RELATIVELY EXPENSIVE RAW MATERIALS, FINANCIAL SAVINGS AS HIGH AS 10% IN RAW MATERIALS COST PRICE ARE VERY COMMON, BUT OFTEN LACK OF KNOWLEDGE ONLY ENABLES US TO DO SO AFTER MANY EXPERIMENTS.
Funding SchemeCSC - Cost-sharing contracts
S1 4DU Sheffield