Additive manufacturing (AM) is going through an exponential growth, due to its enormous potential for rapid manufacturing of complex shapes. One of the manufacturing methods is based on powder processing, but its major bottleneck is associated with powder spreading, as mechanical arching adversely affects both product quality and speed of production. Here we analyse transient jamming of gas-atomised metal powders during spreading. These particles are highly frictional, as they have asperities and multiple spheres and are prone to jamming in narrow gaps. Therefore their detailed characterisations of mechanical properties are critical to be able to reliably predict the jamming frequency as influenced by powder properties and process conditions. Special methods have been used to determine the physical and mechanical properties of gas-atomised stainless steel powders. These properties are then used in numerical simulations of powder spreading by the Discrete Element Method. Particle shape is reconstructed for the simulations as a function of particle size. The characteristic size D 90 by number (i.e. the particle size, based on the projected-area diameter, for which 90% of particles by number are smaller than this value) is used as the particle dimension accountable for jamming. Jamming is manifested by empty patches over the work surface. Its frequency and period have been characterised as a function of the spreader gap height, expressed as multiple of D 90 . The probability of formation of empty patches and their mean length, the latter indicating jamming duration, increase sharply with the decrease of the gap height. The collapse of the mechanical arches leads to particle bursts after the blade. The frequency of jamming for a given survival time decreases exponentially as the survival time increases.

Rigorous non-linear models of elasto-plastic contact deformation are time-consuming in numerical calculations for the distinct element method (DEM) and quite often unnecessary to represent the actual contact deformation of common particulate systems. In this work a simple linear elasto-plastic and adhesive contact model for spherical particles is proposed. Plastic deformation of contacts during loading and elastic unloading, accompanied by adhesion are considered, for which the pull-off force increases with plastic deformation. Considering the collision of a spherical cohesive body with a rigid flat target, the critical sticking velocity and coefficient of restitution in the proposed model are found to be very similar to those of Thornton and Ning's model. Sensitivity analyses of the model parameters such as plastic, elastic, plastic-adhesive stiffnesses and pulloff force on work of compaction are carried out. It is found that by increasing the ratio of elastic to plastic stiffness, the plastic component of the total work increases and the elastic component decreases. By increasing the interface energy, the plastic work increases, but the elastic work does not change. The model can be used to efficiently represent the forcedisplacement of a wide range of particles, thus enabling fast numerical simulations of particle assemblies by the DEM.

No abstract

In the seed processing industry, rotary batch seed coaters are widely used for providing a protective coating layer (consisting of various ingredients including fertilisers and crop protection chemicals) on the seeds. Seed motion and mixing are important in ensuring uniform coating. In the batch seed coater, the base of a cylindrical vessel rotates, whilst the cylindrical wall is stationary and two baffles turn the bed over for mixing. In the present study, the Discrete Element Method (DEM) is used to simulate the effect of particle shape on motion and mixing in this device. Corn seed is used as a model material and the effect of its shape on motion is analysed by considering two approaches: (1) manipulation of rolling friction to account for shape as it is commonly used in the field; (2) approximation of the actual shape by a number of overlapping spheres of various sizes. The geometry of corn seeds is captured using X-Ray micro tomography and then the ASG2013 software (Cogency, South Africa) is used to generate and optimise the arrangement of the overlapping spheres. A comparison is made of the predicted tangential and radial velocity distributions of the particles from DEM and those measured experimentally. It is concluded that for rapid shearing systems with short collisional contacts a small number of clumped spheres suffices to provide a reasonable agreement with experimental results. Equally well, manipulating the rolling friction coefficient can provide results that match experiments but its most suitable value is unknown a priori, hence the approach is empirical rather than predictive. 2

No abstract