FRR - Final Report
Executive Summary
At moderate to high stresses shear cells are the preferred method of powder flow measurement, however at low stresses the determination of unconfined yield strength by this technique is often marred with inconsistencies in the measurement, or in comparison to observed behaviour. An alternative approach for measuring powder flow is ball indentation, which directly measures hardness; related to unconfined yield strength by the constraint factor. The ball indentation and shear cell methods are applied for a wide range of powders, and the constraint factor is found to increase if particle size is reduced or the particle size distribution is reduced. Constraint factor is found to be independent of stress for almost all tested powders, though varies from 1.8 – 5.5 for the different powders tested, and hence its determination is required in order to provide meaningful unconfined yield strength values from ball indentation. For many powders the variation in unconfined yield strength with major principal stress is shown to be much more pronounced at low stresses (< 1 kPa) when analysed by ball indentation. Shear cell measurements at this low stress range generally agree with this trend, though for some powders discrepancies exist between the values reported by each technique. The hardness measurement at low stress is highly reproducible, though relies on a constant constraint factor in order to determine unconfined yield strength. In contrast, the reproducibility of the yield locus generated in the shear cell weakens at lower stresses, with the minimum stress that provides reliable measurements being material dependent. For maize starch the shear cell measurement appears to be reliable even at a pre-shear normal stress of 60 Pa, and the determined unconfined yield strength agrees very well with that determined by ball inden
Executive Summary
Many particulate products are manufactured through powder compaction, in which die filling is a critical process stage as the die filling performance will determine the process efficiency and product quality. For aerated powders, the interaction between particle and the surrounding air could play an important role during die filling. Although die filling has attracted increasing attention over the last two decades, our understanding on die filling of aerated powders is still limited. In particular, how the system design will affect the die filling performance, how significant the presence of air will affect die filling behavior, and whether powders can segregate during die filling are still not well understood. To address these questions, both experimental and numerical investigation were carried out on this project, and this report summarise the key findings on these three aspects. This report contains 5 chapters: the first two chapters examine the effect of system design (die shape, size and orientation) with experimental work reported in Chapter 1, and numerical study in Chapter 2. The effect of air presence on die filling behavior was presented in Chapter 3 and Chapter 4, with chapter 3 focusing on theoretical modeling and Chapter 4 on measuring air pressure buildup in the die. Chapter 5 reports the size-induced segregation during die filling.
We would like to acknowledge IFPRI for financially supporting this project. We also would like to thank Michele Marigo and Tim Freeman for constructive discussions throughout this project.
Executive summary.
This reports presents the results of a series of experiments aimed at studying the factors affecting the amount of charge in dispersed and bulk powders, for how long the charge remains in the bulk powder before dissipating into the environment and the effect of the electric charge on the solid fraction of a bulk powder. The powders used in this study cover a range of particle sizes from 3 μm to roughly 1000 μm, although not all powders were used in all the experiments.
Experiment 1
In the first experiment presented in this report, particles are dispersed in a gas stream and charge due to collisions with a tribocharger. It is found that the main factor affecting the maximum amount of charge per particle is the particle size, since the charge is limited by the electric field on the surface of the particles and for the same charge on a particle, the electric field on its surface scales with the square of the diameter. In practical applications the particles may not experience enough number of collisions with solid surfaces to charge up to their maximum level, but the results presented in this report indicate than, on equal conditions, it is still the particle size the main parameter affecting the particle charge.
Experiment 2
In the second experiment we measure the charge distribution of the particles that come out of the tribocharger. We have found that the particle charge has a very wide distribution spanning both polarities. This finding may be explained if the charge transfer from the tribocharger to the dispersed particles causes a shift in a pre-existent charge distribution in the direction of the transferred charge. In consequence, the particles in a neutral bulk powder may carry electric charge, but on some of the particles the charge is positive and on the others is negative.
Experiment 3
In the third experiment we have measured the charge in a bulk powder formed by sedimentation of highly charged particles. We have found that while particle settles, the layer of bulk powder formed losses its charge. We propose a model that qualitatively describes the decay of the charge in the bulk powder based on the assumptions that the charge in the bulk powder has some mobility and that charge is dissipated on the surfaces of the bulk powder by neutralization with ions existing in the surrounding gas in order to keep the electric field on the surface of the powder at a value equal or below the breakdown field in the gas. The amount of charge in a bulk powder results from an equilibrium between charge dissipation into the surroundings and the accretion of new charge from the incoming particles and according to the model, depends on the charge on the particles that sediment, the mass flow rate at which they arrive and the effective electrical conductivity of the powder.
The effective electrical conductivity can be estimated from the typical time for charge dissipation, which is of the order of minutes, yielding a value of the effective electrical conductivity of the bulk powder of the order of nS/m.
Experiment 4
In the fourth set-up we measure directly the effective electrical conductivity of some powders as a function of consolidation and ambient humidity. The effective conductivity is found to be in the order of nS/m and it is highly dependent on humidity and to a lesser extend, on particle size. The strong dependence on humidity, specially for smaller particle sizes, may explain why the charge on bulk powders seem to be highly unpredictable in environments in which humidity is not controlled.
Experiment 5
In the fifth and last set-up, we measure the poured and tapped densities of charged and uncharged powders in order to determine if there is an effect of electric charge on the solid fraction, but within the accuracy of our experiment, we have found none.
Executive summary
Colloidal gels are used in industrial formulations to solve the ‘gravity problem’.
Particles are typically heavier than their suspending media, and will settle out over time. A strong enough short-range attraction will cause the formation of space-spanning networks that are strong enough to support their own weight. Such gel states are, however, metastable, and will, in time, evolve towards thermodynamic equilibrium. This is manifested in products as the collapse of the gel structure and the appearance of dense sediments. Our project is concerned with understanding such gravitational instabilities.
To do so, we set up a very well-defined experimental model system in which a short-range attraction between nearly-hard-sphere colloids was induced by added non-adsorbing polymers via the ‘depletion’ mechanism. Careful comparison between experimental observations and simulations allowed us to establish that gelation in our system was due to ‘arrested spinodal decomposition’, which gave rise to gels with bicontinuous texture.
Studying such gels using magnetic resonance and optical imaging and again comparing our findings with simulations, we have made a number of important, perhaps paradigm-shifting, discoveries. Two mechanisms operate in gel collapse:
- the accumulation of dense ‘debris’ (compact clusters) at the top, which then fall through the bulk, and
- the rise of solvent ‘bubbles’ from the bulk of the gel to the top.
In both cases, solvent back flow plays an essential role in the break up of spatial structures. Perhaps surprisingly, processes occurring right at the top of gels are vital in determining their fate. In particular, curved menisci at gel-air interfaces seem to generate copious ‘debris’, leading to continuous collapse without any latency (or delay) times, while filling a sample cuvette gives rise to gels that have finite gravitational stability before collapse.
Submicron crystals have the potential to enhance dissolution rates and absorption efficiency of active ingredients with low aqueous solubility used in pharmaceutical formulations and in a wide range of specialty chemicals. Precipitation from solution offers a direct and cost effective method to produce micron-sized and submicron particles. However, a precise knowledge of the particle formation mechanisms, involving the primary process of nucleation and crystal growth, is essential to control crystallization in this size range. Producing submicron crystals of organic compounds by antisolvent or reactive precipitation methods, as compared to inorganic crystals (ionic salts and metal oxides), is even more challenging due to their relatively slow nucleation rates at moderate supersaturation levels. Further, the presence of directional H-bonding in molecular crystals can lead to faster growth along certain facets resulting in crystals with higher aspect ratios.
In this research work, nucleation and growth phenomena of organic crystals from solution were investigated experimentally, with the aim of understanding the effects of solution conditions and process variables on submicron crystallization process. Initially, nucleation kinetics of the model compound, naproxen (a poorly water-soluble drug), was determined at various solute concentrations and in the presence of polymeric additives that are typically used to stabilize colloidal crystal dispersions. Nucleation rates were calculated from multiple induction time measurements at a constant supersaturation using a statistical approach. The results showed a reasonably good agreement between experimentally determined nucleation rates and that predicted using classical nucleation theory relationship. While the polymeric additive polyvinylpyrolidone (PVP) significantly promoted the nucleation kinetics in the entire range of supersaturation studied, the effect of hydroxypropyl methyl cellulose (HPMC) on the nucleation kinetics was supersaturation dependent. Thermodynamic and kinetic parameters for nucleation of naproxen crystals were derived from the experimental data and, in turn, linked to the mechanisms underpinning the effects of polymeric additives in producing submicron crystal dispersions.
The microstructure of gelled colloidal systems is a key determinant of their rheological response. In this project, we addressed three aspects of this critical problem.
Part One
In the first part, we identified high contact number, stress bearing configurations in colloidal gels that had been subjected to non-linear step strain, and determined that these stress bearing clusters are predictive of the non-linear elasticity of the gel. The idea, borne out by analysis of the experiments, is that the hydrodynamic interaction of a fluid of jammed, stress bearing clusters is the principal determinant of the post-yield rheology of colloidal gels. In fact, the abundance and size of these stress-bearing clusters could likely be identified from a measurement of the rheological response.
Part Two
In the second part, in collaboration with the Furst group at the University of Delaware, we developed a model colloidal system in which force, structure, and rheology could all be measured, so that the relationship between these three properties could be better understood, especially with respect to modeling the important rheological quantity of the yield stress. This model system needed to balance a number of constraints related to refractive index matching and density matching. We found a particular condition that satisfied all the constraints, and therefore opens this area to fundamental research.
Part Three
In a final project, we observed that we could produce colloids with controllable roughness. We studied the effect of this variable on a number of rheological properties of colloidal suspensions. We found that the shear thickening response was most affected by colloid roughness.
The rheology of colloidal gels arises from how the ramified network structure responds to an applied stress. Stress transmission is controlled by the topology of the microstructure and the nature of the interparticle “bonds” between colloids. Thus, to deform a gel, these bonds must stretch to accommodate extension and compression of the microstructure.
This project focused on developing and refining a model experimental system to directly and quantitatively measure gels over multiple lengthscales, including the “bond” mechanics, microstructure and bulk rheology. This work was performed in close collaboration with Professor Michael Solomon, University of Michigan. This report summarizes the development of the model system to achieve individual bond rupture measurements and the validation of bond rupture measurements by a dynamical model of bond rupture. We conclude with a comparison of the experimental gel modulus to the modulus predicted by current theory and an energy balance model.
Executive Summary
The aim of this research program was to develop the means to accurately measure the surface forces between mineral like surfaces in concentrated solutions of aqueous electrolytes.
At high salt concentrations, the electrostatic forces are highly screened and become very short in range. Therefore, the forces in concentrated salt solutions are governed by dispersion and hydration forces, the latter have a decay length of approximately 0.3 nm. This demands that the surfaces between which the measurements are performed have exceptionally low levels of roughness. We pursued Atomic Layer Deposition as a means to produce nearly ideal surfaces of the mineral oxides Alumina, Titania, Hafnia and Zirconia. These surfaces were characterised for roughness, composition, thickness, surface charge and stability and employed in surface force measurements for the first time. Also extensive theoretical calculations have been performed to evaluate the van der Waals forces in these layered systems. The adsorption of the surfactant CTAB and the forces between ALD surfaces in CTAB solutions was also studied.
Numerous surface force studies have been completed using Alumina and Titania surfaces covering the influence of pH, electrolyte concentration and a range of additives. A particularly surprising result is found for Titania surfaces at high pH, where the influence of the van der Waals force is not seen. The surface forces between Titania surfaces at very high salt concentrations has also been investigated extensively. For concentrated solutions we find that the forces are always attractive regardless of the type of salt or pH of the solution. The results obtained for the adhesion measurements between Titania surfaces in different electrolytes are not reproducible. This is attributed to extreme sensitivity to contamination as well as small differences in local geometry in the contact region. We conclude that the strength of the adhesion in systems of practical interest is likely to be governed by the nature and concentration of contaminants.
Our investigations in this area are ongoing. Hafnia and Zirconia surfaces have been produced and characterised and will shortly be used for surface force measurements. Additionally, we are developing a theoretical approach for including the effect of surface roughness in theoretical models of surface forces. This will allow experiments performed on non ideal surfaces to be compared to theory with the influence of roughness rigorously accounted.