No description available.

No description available.

No description available.

No description available.

No description available.

Predictions of the mechanical properties of a densely packed granular material so far have been quite inaccurate. Many of the phenomena observed by such a material leave a lot of unanswered questions. For example the stiffness of a sample under a strain path is poorly predicted by the most recent theoretical work. Part of the difficulty of a granular material is the interplay between discrete and continuous measures. For any sample of a practical scale, the size and numbers of grains involved makes consideration of each individual grain a near impossible process. So on the whole, continuum measures and theories have been used to try to describe the material behaviour without acknowledgement of the particulate constituents. However, that the mechanics of the material as a whole is dependent on the inter-particle forces is undeniable. These inter- particle forces work at the level of the individual particle where discrete measures such as force and displacements should be used. So how can practical, continuum measures be constructed from discrete constituents? This thesis provides the theoretical means to traverse from the discrete constituents to continuum measures. A key feature is the formalisation of a mesodomain, which is the unit of a granular material where the continuum and discrete regimes meet. Use is made of the formulation of a heterogeneous material to describe this mesodomain and it is shown how material properties can be scaled up from the mesoscale to the size of a sample of granular material. Work external to this project provides a method for describing the mechanics within the mesodomain. This is used to calculate the values of important tensors that represent the fabric of the mesodomain.

No description available.

In biaxial compression tests, the stress calculations based on boundary information underestimate the principal stresses leading to a significant overestimation of the shear strength. In direct shear tests, the shear strain becomes highly concentrated in the mid-plane of the sample during the test. Although the stress distribution within the specimen is heterogeneous, the evolution of the stress ratio inside the shear band is similar to that inferred from the boundary force calculations. It is also demonstrated that the dilatancy in the shear band significantly exceeds that implied from the boundary displacements. In simple shear tests, the stresses acting on the wall boundaries do not reflect the internal state of stress but merely provide information about the average mobilised wall friction. It is demonstrated that the results are sensitive to the initial stress state defined by K0 = sh/sv. For all cases, non-coaxiality of the principal stress and strain-rate directions is examined and the corresponding flow rule is identified. Periodic cell simulations have been used to examine biaxial compression for a wide range of initial packing densities. Both constant volume and constant mean stress tests have been simulated. The characteristic behaviour at both the macroscopic and microscopic scales is determined by whether or not the system percolates (enduring connectivity is established in all directions). The transition from non-percolating to percolating systems is characterised by transitional behaviour of internal variables and corresponds to an elastic percolation threshold, which correlates well with the establishment of a mechanical coordination number of ca. 3.0. Strong correlations are found between macroscopic and internal variables at the critical state.

This thesis reports the results of DEM (Discrete Element Method) simulations of rotating drums operated in a number of different flow regimes. DEM simulations of drum granulation have also been conducted. The aim was to demonstrate that a realistic simulation is possible, and further understanding of the particle motion and granulation processes in a rotating drum. The simulation model has shown good qualitative and quantitative agreement with other published experimental results. A two-dimensional bed of 5000 disc particles, with properties similar to glass has been simulated in the rolling mode (Froude number 0.0076) with a fractional drum fill of approximately 30%. Particle velocity fields in the cascading layer, bed cross-section, and at the drum wall have shown good agreement with experimental PEPT data. Particle avalanches in the cascading layer have been shown to be consistent with single layers of particles cascading down the free surface towards the drum wall. Particle slip at the drum wall has been shown to depend on angular position, and ranged from 20% at the toe and shoulder, to less than 1% at the mid-point. Three-dimensional DEM simulations of a moderately cascading bed of 50,000 spherical elastic particles (Froude number 0.83) with a fractional fill of approximately 30% have also been performed. The drum axis was inclined by 50 to the horizontal with periodic boundaries at the ends of the drum. The mean period of bed circulation was found to be 0.28s. A liquid binder was added to the system using a spray model based on the concept of a wet surface energy. Granule formation and breakage processes have been demonstrated in the system.

This project investigated how various minerals of different surface areas and morphologies can be used to adsorb isothiazoline biocides for controlled-release and antimicrobial purposes. The absorption of the biocides on the mineral powders was achieved by way of using a bench high shear mill (dry process), or combining them to hydrated minerals (wet process). The characterisation of the minerals was achieved by XRF (chemical composition), XRD (crystal composition), SEM (morphology), B.E.T nitrogen (surface area), and Light Scattering (particle size distribution). HPLC was used to determine the concentration of the biocide in solution, and the Flow Microcalorimeter used to measure the bond strength between the biocide molecules and the minerals. The minerals were added to an exterior paint made according to an Imerys in-house formulation. Various modifications of this initial coating formulation were made in order to compare the biocide 2-Octyl-4-isothiazolin-3-one (OIT) release profiles from impregnated and non-impregnated minerals. Montmorillonite clay was the best performing mineral in all experiments (adsorption and desorption both from the minerals and paints films, strength of bond analysis, and bioassay). All other minerals tested carried the biocide with varying degree of success. Optical and mechanical tests performed on paint films containing various minerals suggested there were no significant differences between the films. Rheology tests demonstrated that newly developed formulations were easy to apply to a surface.