Abstract
Flow inconsistency of powders is a major problem in industrial processes, which often leads to product wastage, hence economical losses. Many powder flowability characterisation techniques exist, however, their applicability to prediction of in-process behaviour of a powder are limited by the data relevance to the process conditions. Most industrial processes operate under the intermediate flow regime, which is characterised by frictional and collisional interactions between particles. Therefore, prediction of flow performance in these processes require the use of dynamic flow characterisation techniques, which provide a more accurate reflection of the material performance at high strain rates. The FT4 powder rheometer is capable of dynamic flow characterisation, however, the complex shearing flows in the powder bed are not fully understood. There is also lack of understanding on the relationship between the prevailing shear stresses in the powder bed and the measured flow energy, and its dependence on material properties and blade operational conditions. In this work, the flow energy of model powders is experimentally measured using the FT4 powder rheometer and related to physical powder properties and blade tip speed. The Discrete Element Method (DEM) is used to simulate the flow energy test, which is validated against the experimental flow energy and velocity distributions in the powder bed measured by Positron Emission Particle Tracking (PEPT) and Particle Image Velocimetry (PIV). Initially, the sensitivity of the flow energy to sliding friction is evaluated, and it is demonstrated that a velocity-dependent sliding friction model provides better prediction of the experimental flow energy than the conventional velocity-invariant sliding friction model. The distribution of the prevailing stresses and strains and their dependence on material properties and blade operational conditions are analysed. The strain rate is highest in the region near to the front of the blade and decreases with increasing angular distance away from the blade. The shear stress is found to be approximately constant along the radial direction, but it dissipates in the angular direction away from the blade. The shear stress increases with increasing blade tip speed, interfacial energy and shape irregularity (decreasing aspect ratio). A predictive relationship for the average shear stress in the powder bed based on the torque acting on the blade and height of the moving bed is established. This estimated shear stress correlates well with the shear stress predicted by DEM for different material properties. Rheological relationships that describe the granular rheology at high strain rates are established in relation to physical and mechanical particle properties as well as process parameters. A non-dimensional shear stress correlates well with the inertial number providing data unification for a range of blade tip speeds for both downward and upward testing in the FT4 powder rheometer. The developed constitutive relationship can be used to predict rheological behaviour of granular materials under process-relevant conditions. The mixing and segregation behaviour of binary granular mixtures under the dynamic shear conditions is assessed experimentally using a newly designed protocol for the FT4 powder rheometer. The evolution of mixing index with time shows that the mixing rate increases with increasing ratios of size and density. The mixing and segregation mechanisms are determined by numerical analysis, with the percolation or sifting mechanism shown to be the dominant mechanism for mixtures consisting of components with different particle sizes. However, for cohesive mixtures, mixing occurs by shear-induced rupturing of agglomerates, which aids the convective mixing. A non-dimensional flow energy, normalised by the cohesive energy, correlates well with the mixing index providing data unification for a range of cohesive binary mixtures. This approach has a great potential to be used for prediction of the minimum energy input required to provide the optimum mixing performance under process-related conditions.