In-line powder flow behaviour measured using electrostatic technology
Date
Authors
Advisors
Journal Title
Journal ISSN
ISSN
DOI
Volume Title
Publisher
Type
Peer reviewed
Abstract
Within solid-dose manufacturing processes, powder flow and powder triboelectrification are critical to the quality of the final product. Off-line testers do not simulate the shear and packing conditions that a powder would experience in-process and may be unreliable in predicting in-line flow and charging properties, which are key components to successful formulation and process design. In this work, a dual-electrode, electrostatic powder flow sensor (EPFS) was used to obtain electrostatic signals that were generated in response to the pattern of flow of pharmaceutical powders in two density modes: The first being powders in lean phase flow, generated by free-fall of the powder from the outlet of a screw-feeder. The second being dense phase flow, through either 19.1 mm Ii.Dd. stainless-steel pipe or at the outlet of a tablet-press hopper. Powders were selected from a range of low to high cohesivity so as to study the effect of powder cohesion on the flow pattern. Electrostatic signals were then analysed by three distinct signal processing methods (RMS signal averaging, cross correlation, and Fast-Fourier-Transform) with a view to determining certain characteristics of powder flow, i.e. mass flow rate; cohesivity; and triboelectrification. In the first application a calibration was attempted to establish the link between the root-mean-square (RMS) of the electrostatic signal and the mass flow, as determined by the accumulation of mass on a balance placed below the screw-feeder (in the case of lean phase application) and the 19.1 mm i.d. pipe (in the case of dense phase application). In both cases it proved unsuccessful, owing to the instability in the electrostatic signal (i.e. its dependence on factors other than mass flow, for example inherent and induced charge fluctuations and moisture content). An alternative method for determining mass flow rate was proposed based on the second signal processing method, which involved the cross-correlation of signal from both sensors to determine the free-fall velocity. This method might work in future applications if combined with a suitable technique for determining the powder density. In the second application, a Fast-Fourier-Transform (FFT) of the electrostatic signal to yield an FFT spectrum was used to establish whether this technique could determine aspects of powder cohesivity. A correlation in rank order of cohesivity was observed between the ratio of the summed or averaged amplitudes at the three principle frequencies to the summed or averaged of the baseline components respectively, and the cohesivity of the powders, as determined by off-line powder rheometry assessments of dynamic flow and bulk properties. In the third application, the RMS signal normalised to the powder mass flow rate was used to study the time-dependent powder charging behaviour, which is induced by the transportation of the powder within the screw feeder. Characteristic relative charging profiles were obtained for each powder, which in some cases were coupled to charge-induced adhesion of the powder to the equipment. In the last application, the RMS signal generated from the EPFS sensor located at the outlet of the hopper on a rotary tablet press was used to interrogate the dense-phase intermittent-flow resulting from the dosing of the tablet die. Those more cohesive powders gave a larger RMS signal at the lower electrode (relative to the upper electrode) whereas less cohesive powders had similar RMS signals at each electrode. While the exact explanation of this effect is currently unknown these results suggest that the technique might be useful in the determination of die filling as a function of the input material characteristics. In summary, this work has provided some insight into the potential applications of EPFS for in-line measurement of powder flow and charging characteristics. Future work should focus on (i) developing an integrated sensor with an independent measurement of density to yield the powder mass flow using an inferential approach, (ii) co-use of techniques (such as Faraday-cup and charge decay analysers) to validate the in-line charging behaviour, (iii) further exploration of the significance of the signal amplitude difference at the tablet press hopper outlet in on the characteristics of the tablet compact.