Dry ball mixing and deagglomeration of alumina and zirconia composite fine powders using a bimodal ball size distribution

Agglomerates lead to poor and unreproducible properties of ceramics. To ensure a reliable manufacturing process and final product properties, a batch needs to be well deagglomerated and also well mixed. This phenomenological study compares the application of a bimodal ball size distribution to monomodal ones of large or small balls in a tumbling mixer operated at 30 rpm. The ball sizes were adapted to the agglomerated particle sizes using Bond׳s ball sizing relation. Additionally, the diameters of the small and the large balls were adjusted according to Furnas׳ densest packing theory. The used filling ratio was 33 vol% and the batch of zirconia and alumina filled the interstices between the balls. The agglomerated batch had a d50=1.2μm and a d99=5.1μm. The small balls (d50=2.25mm) caused a good macroscopic mixing degree for conditioning 10–20 min as investigated by specific surface area and true density. After initial deagglomeration, reagglomeration occurred for times ≥10min. Conditioning with large balls (d50=15mm) led to a comparatively large d99=2.2μm but reagglomeration was suppressed for times ≥20min. The microscopic mixing degree was also good for times ≥20min as evaluated by scanning electron microscopy and X-ray diffraction. The bimodal ball size distribution led to a time-insensitive d99≈2.0μm between 10 and 40 min. Consequently, the robustness of the deagglomeration process increased according to the approach of Taguchi. Therefore, the simultaneous optimization of mixing and deagglomeration was simplified to the mixing optimization. An excellent mixing degree was achieved for times ≤10min. Therefore, the conditioning result by the bimodal ball size distribution was superior to applying monosized balls.