Investigation of the effect of particle size distribution on the flow properties of bulk solids Abstract This paper reports from a Defraud sponsored project which aims to develop a toolkit that can be used by formulators to assist in controlling the flow properties of blended or engineered powders. This paper presents the results of an investigation into the effect of the breadth of the particle size distribution on the flow function, bulk density and internal friction of four food grade fillers namely; lactose, dextrose, malnutrition and sodium chloride.
Physical property measurements, including particle size and article shape, are presented along with bulk flow properties which were measured using the Brookfield Powder Flow Tester (PET). The flow properties of the four food powders are compared and discussed with reference to the effect of the breadth of the particle size distribution. 1 Introduction The objective of this work is to investigate the effect of particle size distribution on the flow properties of four filler powders commonly used in the food manufacturing industry.
Due to the variety of lactose, we used six different lactose grades to compare different particle size distributions – from wide to very narrow. Dextrose was used because of its wide size distribution and by sieving analysis, sections from particle size distribution were removed to be tested in the shear tester. Sodium chloride and malnutrition have a narrower particle size distributions and the presence of very fine particles gave us the possibility to study the effect of particles below 45 mm.
Three samples were tested for each flow property measurement and the mean values of the flow functions, bulk densities and friction functions were plotted. These values showed a low standard deviation indicating that the data points tend to be very close to the mean, therefore, they are inappreciable in the graphs. 2 Methods 2. 1 Physical properties Particle size distribution for lactose was measured by laser diffraction using the Manlier Monasteries with dry powder feeder unit. Measured by sieving analysis using GRADED 2000 Particle Size Analyzer.
The bulk solid specimen is loaded vertically by a normal stress and then a shear deformation is applied on the specimen. Fig. 3 Brookfield PET While the majority of available shear testers [2, 3] undertake flow function measurements at a single consolidation stress level per powder sample, the standard est. algorithm for the Brookfield PAT undertakes failure property measurements over multiple increasing consolidation stress levels, significantly reducing the required testing time.
The results are presented graphically in a form familiar to those who use the Jennie silo design method . While this data is easy to interpret for experienced users, what is important for quality control purposes is deskilling the interpretation and using a single figure fallibility numbers to describe material.
Two single figure fallibility numbers that could be generated by the Brookfield PET, which are also concepts that an be easily understood by inexperienced operators are the critical arching dimension and critical rat-hole dimension. 3 Results and discussion The comparison of the mean values of the flow functions, friction functions and bulk densities of the four food powders tested over a range of different size fractions with the Brookfield PAT are shown in figures 4, 5 and 6 respectively.
In order to show clearly the effect of particle size distribution on the flow function, it was considered that the scales in vertical axis of the graphs of the flow functions should be different for each powder (see figure 4). By considering this figure, we can observe that the strength of the bulk solid increases when particle size decreases, even for free flow powders like sodium chloride. It is also noticeable that for all filler fallibility of the size fraction containing the DAD.
These effects can be observed in the friction functions of the powders as well. Fig. 4 Comparison of the mean values of the flow functions of the four food powders measured with the PET, a) lactose, b) dextrose, c) malnutrition, d) sodium chloride Inspection of the effective friction functions, fig 5 (a) shows that at low consolidation stresses, the fine particles Lactose 300 and Corolla 400 are agglomerating exulting in a lower friction than the coarser lactose 200.
By comparison with the other three fillers, the sodium chloride fig 5 (d) shows larger variations in internal friction angle as proportion of the measured unconfined failure strength. It is suspected that the reason for this behavior is that salt has cubic particles (see figure 2 where the radii of the particles are very much larger than the radii of the corners) which do not shear easily producing variability in the friction measurements.
The shape of the other filler powders is elongated (malnutrition) or approximately rounded (dextrose) as we can see in figure 2. Fig. Comparison of the mean values of the friction functions of the four food powders measured with the PET, a) lactose, b) dextrose, c) malnutrition, d) sodium chloride As we can see in figure 6, the bulk densities decrease when the particle size decrease due to the increase of the voiding, and so, there are more open packing structures in the powder.
The density curves corresponding to full size distributions are similar in shape to the size fraction containing the DAD whereas they show higher magnitude due to the presence of coarse particles. Fig. 6 Comparison of the mean values of the bulk densities of the four food powders our filler powders generated by the Brookfield PAT for conical mass flow hoppers. The values are approximate because wall friction tests have not been undertaken, so a critical flow factor (if=l . ) has been assumed to determine the intercept with the flow function and the critical outlet diameter. Inspection of fig 7 shows that the arching dimensions increase when the particle size decreases, especially for powders with low bulk densities (lactose and malnutrition) which show a significant increase of arching diameter at low particle size. It is also noticeable that again the behavior of the full size distribution is insistent with the arching diameter for the size fraction containing the DAD of the powders. Fig. Comparison of the approximate critical arching dimensions of the four food powders generated by the Brookfield PAT as a function of particle size 4 Conclusions The results of the study found that: For powders tested, with DAD around 100 mm and different particle size distributions, the behavior of the full size distribution is consistent with the size fraction containing the mommy because the proportional fines of the DAD dominates the behavior, even for powders like dextrose, with a wide particle size distribution.
Note that because of the number of materials tested, this conclusion is compromised and further work needs to be done with different materials. It has demonstrated the usefulness of Brookfield PAT as industrial tool to study the effect of changes in the particle size distribution to the flow behavior of the powder. Acknowledgements I would like to thank Defraud for sponsoring this project through the Advanced Food Manufacturing (Link) Programmer. References R. J. Berry and M. S. A. Bradley, Development of the Brookfield Powder Flow Tester. In: Bulk Solids Europe 2010 Conference and Niche Symposium in New Frontiers in Bulk