OXFORD has made 244 determinations of modulus of rupture, 1601 determinations of dry contraction 110 ºC, 1508 determinations of fired contraction 1220 ºC, 877 determinations of ignition loss, 1544 determinations of apparent porosity, 2288 determinations of dry linear dilation (dried at 110 ºC) and 1128 determinations of fired linear dilation (fired at 1220 ºC) on their kaolin ores. The averages, maximum, minimum and the statistical distribution of these values are shown on tables and graphics. Those determinations and field observations allow to the conclusion that kaolin deposits are of two different types, named Trigolândia model, originated from pegmatites, and Floresta model, derived from acid volcanic rocks. With the exception of the ore from Turvo Mine, the average values of the analyzed physical properties are always into the intervals accepted by Oxford industrial processes, that runs to produce tableware. From Turvo Mine, the green and the brown ores have the more discordant characteristics. For the other ores, from Turvo and the other mines, the average values of their physical characteristics almost always equal the central values of intervals that characterize good ores. If correlations between average values of different physical properties of ores are made, only Trigolândia shows some bad correlations. Taking into account all the analyzed values, and no more the averages, all the correlations are bad. The exception is the correlation between dry retraction and fired retraction, with r = 0,78. All other correlation coefficients are lower than 0,60.
kaolin ores
