Incinerator bottom ash (IBA) is a product from incineration of municipal solid wastes. It contains heavy metals that may post environmental concerns. Research has shown that particle size reduction process may move the heavy metals in IBA to a certain size range. Further understanding of the attrition of IBA therefore is believed to lay the foundation of subsequent recycling heavy metal elements from it. In this study, glass beads were utilized to co-fluidized with IBA in a fluidized bed at constant gas velocity. Effects of glass bead size and glass bead composition were studied. The results reveal that glass beads can enhance the attrition rate of IBA, and increase the breakage of large particles. In terms of fines, the use of small glass beads has marginally higher rate of attrition compared to the use of large glass beads at the start. However, large glass beads eventually have higher degree of attrition given a sufficiently long attrition time. The results also indicate that glass beads have a stronger impact to the breakage of IBA in the corresponding size range.
Rosin-Rammler (RR) distribution has also been used to describe particle size distribution of IBA in the fluidization process. Five popular meta-modeling techniques were used to model the change in particle size distribution and the accuracy, efficiency and robustness of the models were evaluated. The results show that artificial neural network (ANN) and ANN-GA model which integrates ANN into the genetic algorithm achieve the best performance in terms of accuracy and robustness. Gaussian process regression with the pattern search as curve fitting technique (GPR-PS) is the most efficient model followed by support vector machine models which are integrated into the genetic algorithm (SVM-GA) and the pattern search (SVM-PS).