In powder handling industries, screw feeders are routinely used to control the flow of material from a hopper into the subsequent stage of a process. The accurate and consistent feeding of materials into reaction systems, mixers and other processing equipment is necessary to maintain optimal conditions, and to generate products of the required quality at the required rate.
However, predicting feed rates has historically relied on engineering estimation based on experience and the extrapolation of and pre-existing information on performance – either from pilot scale trials, or from data collected at earlier installations. This can lead to equipment that does not meet the target design requirements, and result in sub-standard operation or, in extreme cases, complete process failure.
In an attempt to more accurately predict the expected feed rate of powder in a given feeding process, one approach is to use statistical, multivariate models to evaluate the relationships between powder behaviour and process performance. This study demonstrates how to generate a design and operating model based around robust measurements of rheological properties of powders and standard Multiple Linear Regression (MLR) analysis.