Particle size distribution modeling of milled coals by dynamic image analysis and mechanical sieving

Abstract

ABSTRACT Particle size distribution (PSD) of lignite and hard coals ground by ball and Gy-Ro mills were determined and compared using mechanical sieving (MS; direct, width-classification) and dynamic image analysis (DIA; indirect, length-classification), expressed in terms of sieves opening diameter and equivalent circular area diameter. Ground coal PSD data (eight combinations) were visualized and studied using length-transformation on MS and DIA using log-normal distribution plots. Data were also analyzed and compared by developing PSD models, namely Gaudin–Schuhmann (GS) and Rosin–Rammler (RR). In general, in the fine particle range below 100 μm, all ground coal samples exhibited similar PSD, for both MS and DIA methods. On PSD modeling, RR model ( R 2 ≥ 0.81; AIC ≤ − 8.6) fitted well the observation over the entire range of particles sizes produced by both ball and Gy-Ro mills compared to GS model ( R 2 ≥ 0.69; AIC ≤ − 7.3). DIA can easily determine the particle size ranges below 100 μm, while MS have only limited sieves in this range. Besides, DIA has produced more accurate PSD results than MS especially 38 μm and below. Thus, DIA and RR model can be recommended for PSD analysis of fine particulate coals, minerals, and similar products.