INFLUENCE OF BATCH MIXING TIME AND BATCH GRAIN SIZE ON HOMOGENEITY OF A SODA-LIME-SILICA GLASS*

Density-spread determinations were made on a series of melts of a soda-lime-silica glass employing batches compounded from raw materials varying in particle size from 20- to 60-mesh to minus 200-mesh, which had been mixed by tumbling for one, ten, or thirty minutes. The two melting techniques employed were (1) melting for sixteen hours in a stationary platinum crucible at 1400°C. under conditions of even temperature distribution and consequently a minimum of convection mixing and (2) melting for four hours at 1400°C. in a rotating tilted crucible; this condition was intended to simulate convection mixing Under the experimental conditions employed, it was found that (1) batch mixing time has relatively little effect on the homogenizing rate, (2) homogenization increases rapidly with decrease in grain size, and (3) in all cases the 4-hour melting treatment with moderate mixing gives better homogeneity than sixteen hours melting time in a stationary crucible, this effect increasing with decrease in grain size. No significant difference in glass homogeneity resulted from three different methods of mixing, namely, tumbling, ball milling, and mixing in a muller-type mixer, although the homogenizing influence imposed probably was sufficient to mask any differences in uniformity of mixing that might have existed.