ABSTRACT
The sintering behavior of 31 commercially available alumina powders has been investigated. Typical properties of the powders, such as particle size distribution, specific surface area, agglomerate factor and chemical composition were determined and their influence on the sintered density and the microstructure was investigated. Since none of the known sintering theories could describe the whole range of properties of the commercial powders, the 'neural network' method was used for the description of the relationship between the powder characteristics and the microstructure of the sintered samples. Following the analysis, a general rule for the sintering behavior of commercial alumina could be formulated which will be presented. Also the possible use of the 'neural network' theory for the characterisation of processes which do not have well defined parameters, like the sintering of commercial powders, will be shown.
