ABSTRACT Aiming at the problem that low-rank coal is unsuitable for making CWS (coal-water slurry) and lacks an accurate particle size gradation prediction model. This paper expounds on the effects of ball milling time and speed on the concentration and rheology of CWS under different grinding conditions. After size grading, the slurry’s bimodal parameters and particle size distribution are studied. Based on the BP neural network, a particle size distribution model more suitable for low-rank coal is established in this paper. The results show that the slurry concentration in the graded coal sample increases by 4% higher than that of the ungraded coal sample. The BP neural network model accurately predicts the relationship between bimodal parameters and slurry concentration based on bimodal parameters. In the context of grain size gradation, the optimized particle size cumulative distribution model, denoted as y = x n − x K n x 100 n − x K n + K { ( K + 10 % > x > K% } , demonstrates enhanced alignment with the attributes of low-rank coal. This underscores the paper’s contribution in providing a more fitting particle size distribution model for low-rank coal in the context of CWS production.
Read full abstract