1987 年 103 巻 1195 号 p. 577-585
Agglomeration experiments were carried out using BaSO4 as model particles, sodium oleate (NaOl) as surface active reagent and kerosene as bridging liquid and agglomerate growth processes could be classified into three patterns as follows: Pattern 1) Agglomerates were formed immediately after the agglomeration experiment started and they gradually grew to certain sizes after which agglomerate growth slowed down (Figure 3). Pattern 2) Initially, micro-agglomerates were formed and after a certain aging time they grew rapidly to large agglomerates with intermediate size agglomerates also present (Figure 4 (a) and (b)). Pattern 3) Initially, micro-agglomerates were formed and after a certain aging time they grew to large agglomerates rapidly with almost no intermediate size agglomerates (Figure 5 (a) and (b)).
Simulations of these patterns were conducted using the population balance equations derived by the authors. The equations consist of a collision frequency function and a coalescence probability, formulated on the basis of Kolmogoroff's theory under isotropic turbulent flow conditions and the physical properties of agglomerates. Both the experimental and simulated results showed that in the case of pattern 1 agglomerate growth was restricted by a characteristic agglomerate diameter (CD), at which the coalescence probability between two agglomerates of the same diameter was zero. There was good correlation between the weight mean diameter of agglomerates (D50) and the CD. In the case of pattern 2 and 3, the rapid agglomerate growth after the certain aging time is interpreted in terms of rapid increases in the CD. It was.also shown that breakage of agglomerates became significant for agglomerates larger than a limiting diameter. From these results, it was concluded that the agglomerate growth were controlled by CDand the three patterns were mainly caused by the differences in CDchanges due to compaction during the experiments.