Dry grinding kinetics of binary mixtures of ceramic raw materials by Bond milling

Ipek H., Ucbas Y., Yekeler M., Hosten C.

CERAMICS INTERNATIONAL, vol.31, no.8, pp.1065-1071, 2005 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 31 Issue: 8
  • Publication Date: 2005
  • Doi Number: 10.1016/j.ceramint.2004.11.006
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.1065-1071
  • Keywords: binary mixtures, breakage parameters, Bond milling, kaolin, K-feldspar, quartz
  • Middle East Technical University Affiliated: Yes


The kinetics of batch dry grinding of binary mixtures of ceramic raw materials, namely quartz-kaolin, quartz-potassium feldspar and kaolin-potassium feldspar, from the feed sizes of -3.350 + 2.360, -2.000 + 1.400, -0.850 + 0.600, -0.500 + 0.355 and -0.300 + 0.212 mm have been determined using a Bond mill with a mixture of ball sizes of 38.10, 31.75, 25.40, 19.05 and 12.70 mm diameter and total mass of 22.648 kg. The Bond mill used was a size of 30.5 cm diameter, 30.5 cm length, with a total volume of 22,272 cm(3). The fractional ball filling was 22% of mill volume and the mill speed was 70 rpm. The breakage parameters were obtained for those mineral mixtures to predict the product size distributions. As the feed sized given above, which were ground in the mill, increase, the specific rate of breakage (S-i) values also increase, which means faster breakage with higher Si value occurs in the order of quartz-kaolin, quartz-potassium feldspar and kaolin-potassium feldspar mixtures when comparing the characteristic alpha values (slope of S-i versus size relationship with higher alpha value). The cumulative breakage distribution function (B-ij) values obtained for these mineral mixtures were slightly different in terms of the fineness factor, gamma. This means that quartz-potassium feldspar mixture produced less fines with higher gamma value, while kaolin-potassium feldspar gave more fines with lower gamma value. The simulations of the product size distribution for these mixing were very close to the experimental data. Finally, slowing down effect, treated with false time concept, started earlier than the expected for these binary mixtures. There were some correlations found between the simulated time (theta) and experimental time (t). (c) 2004 Elsevier Ltd and Techna Group S.r.l. All rights reserved.