(473e) Stress Distribution and Relaxation in a Powder Column Under Compression

This work investigates the distribution of stresses transmitted by a compressed powder from a piston acting vertically on a powder column, onto the containing cylindrical vessel base and sidewalls. For this purpose, we used a pressure-sensitive Fujifilm Prescale film and a thin flexible SingleTact capacitive sensor to visualise pressure distributions and to quantify stresses over areas as small as 50 mm². The latter was the smallest and most sensitive flexible force sensor commercially available at the time this work was carried out, and it proved to have sufficient resolution to differentiate between the compaction properties of different powders. Of the materials tested, Whole Milk Powder and Corn Flour had the highest compressibility as measured using an FT4 Powder Rheometer, but in this study we measured significant differences in their stress distribution and retention behaviour in comparison to the less compressible Eskal (calcium carbonate) samples.

From the footprint left by the compressed powder on the pressure-sensitive film on the vessel base, we differentiated between two forms of stress distribution, according to the powder column height. For short columns the pressure transmitted through the powder onto the base was uneven. In fact, the shorter the distance between piston and base, the stronger and more localised we observed pressure to be transmitted by the powder on the base. In contrast, for powder columns taller than a critical value, the pressure on the base appeared evenly distributed (except around the outer edge of the base where the pressure is lower), albeit decreasing in magnitude the farther the base is from the piston.

In the region immediately below the piston we measured very high radial stresses on the vessel sidewall, in some cases even higher than the pressure applied by the piston. This result validates similar observations of a radial stress distribution along the compressed surface made through different methods (ball indentation, neutron diffraction), that the authors were inconclusive about at the time. We conjecture that the magnitude of the maximum radial stress in this area is powder dependent, as is the rate at which it decreases with depth from the piston. In Corn Flour, for example, the radial stress near the outer edge of the piston was particularly high and it decreased rapidly with depth. In this case there was a large radial stress gradient with depth and high shear between consecutive powder layers. We expect that powders that exhibit such high stress gradient are more susceptible to defects such as capping and lamination, so we propose that measurements of radial stresses in this narrow region may support a useful method for the prediction of tabletting defects. The SingleTact sensor proved to be sensitive enough and to have high enough resolution to differentiate between the stress gradients of the powders evaluated for this study, and future studies will focus on the differentiation of the compaction properties of pharmaceutical powders. As wider-range and more sensitive technology becomes available, we will also aim to reduce the dimensions of these experiments, to study powder behaviour at a scale more representative of real industrial processes.

The measurements obtained for this work were also highly differentiating in the powder stress retention properties. We measured that for all powders tested the fraction of the stress retained after compression increased linearly with depth. But the absolute fraction and its gradient of increase with depth are largely powder-dependent. We corroborated that powders that are less permeable and whose particles are more frictional and that tend to interlock (as measured from Permeability and Specific Energy measurements, respectively, using an FT4 Powder Rheometer), relax significantly less after compression. The results presented in this work may have a significant impact in the characterisation and design of formulations for optimal tabletting performance.