Appraisal of Fine Distribution in a Pilot-scale Silo in Different Filling Conditions

During filling a silo, granular material containing a range of particle sizes, the fine material accumulates under the filling point. The inclined surface of stationary bed particle which is formed in silos during filling process acts similar to a sieve through which the smaller particle fall. This effect is called sifting. As a result of the mentioned effect, much finer particles form a vertical cylindrical zone of high concentration at the silo center. For optimal design in industrial process such as aeration of stored products in silos, filling silos, and wherever granular materials are handled, it is necessary to survey the distribution of the fine materials depending on product and process properties. The objectives of present study were: (a) To study fine change as affected by substantial parameters, (b) To model fine changes at different conditions in silos. In the present study, an experimental setup consist of a main container, elevator, trapezoidal container and experimental silo was designed and built. Fine content was defined by BCFM (broken corn and foreign material). By applying a new approach, sampling was performed in a radial and vertical direction. The position of each sampling point was determined with a scaled distance from center (R) and from bottom (Z). Local BCFM (BCL) was defined as the value of BCFM in each sampling point. Influential parameters namely, initial BCFM (BCI), volume flow rate (Q) and fill pipe diameter (DF) were considered as treatments. Non-linear regression technique was applied on the experimental data to predict the distribution pattern of fines into the pilot-scale silo. The most appropriate model in a try and error procedure was selected based on highest value of R2 and least value of χ2, RMSE and MRDM. According to the results of ANOVA, it was found that the effects of all parameters were significant at 5% probability. BCL decreased nonlinearly with a concave down decreasing trend along radial direction due to sifting effect. As a result, most amount of fines remained in the sections closer to the center of the silo. Fine distribution became more uniform with decreasing Z and increasing BCI and DF. Also, the distribution of fine became more uniform with increasing Q. BCL was a nonlinear function of R and a linear function of Z, BCI, Q and DF. Although including more and complex terms increased the model complexity but in the present study considering BCL as an exponential function of R and as an implicit function of Z and R (ZR) improved the quality of the model significantly. The values of 0.94, 1.14, 1.06, 11.39 for R2, χ2, RMSE and MRDM, respectively, gave the best model. The results showed, considerable accumulation of fines occurred at the center of the silo which increased with increase of level of Z. Also, low concentration of fine occurred at the periphery of the silo especially at higher levels of Z. It means that maximum non-uniformity of fine distribution occurred at higher levels of Z. The present study investigated distribution of fines during filling affected by main parameters namely, initial BCFM, volume flow rate and fill pipe diameter in a pilot scale silo. A new procedure was developed for measuring the fine material along radial and vertical directions. Distribution of fine was modeled using a developed equation considering the effects of main parameters. The results showed that distribution of fine becomes more uniform with decreasing height and increasing initial BCFM, volume flow rate and fill pipe diameter.