Over the past decades, the prevalence of machine learning (ML) methods has made the development of ML-based constitutive models for granular materials undoubtedly a popular subject. Numerous studies have been made to feature the loading path or history-dependent stress-strain response of granular media using neural networks. In this work, a novel finite element method (FEM)–ML multiscale approach was developed by incorporating internal variables to improve the simulation accuracy of 3D history-dependent granular materials for the first time. To this end, a surrogate constitutive model based on the single-step-based multi-layer perceptron (MLP) neural network was used to replace representative volume element (RVE) simulations conducted by the discrete element method (DEM) in the multiscale FEM–DEM approach. Although the prediction principle of the MLP aligns with the FEM algorithm, artificially added internal variables are required to differentiate the loading history. To address this issue, history variables associated with the Frobenius norm are proposed to be fed into the MLP coupled with the strain tensor to extract the history-dependent behaviour of granular assemblies. The developed FEM–ML approach was demonstrated in 3D conventional triaxial compression (CTC) simulations. Compared to the multiscale FEM–DEM approach, the proposed FEM–ML method exhibits a significantly improved computational efficiency.

Silicon-nitride ceramic substrates exhibit high thermal conductivity, high electrical insulation, high mechanical strength, low expansion. The Cu-clad ceramic substrate is the key material for the encapsulation of insulated gate bipolar transistor modules. In this review, we divide the research of high thermal conductivity silicon nitride ceramics into four stages: 1) exploration of thermal conductivity with different sintering additives; 2) using the anisotropy of silicon nitride grains to improve the thermal conductivity, through the addition of silicon nitride grain seeds combined with tape casting technology, sintered to obtain a thermal conductivity of 155W/(m·K) silicon nitride, and with the development of the strong magnetic field alignment technology in the later stage, the thermal conductivity can be increased to 176W/(m·K); 3) using sintered of reaction-bonded silicon nitride technology to improve thermal conductivity, thermal conductivity can be increased to about 180W/(m·K); 4) the application of high thermal conductivity silicon nitride ceramic substrate. Finally, future preparation methods and applications of high thermal conductivity silicon nitride ceramics are envisioned.

This study explores an efficient approach to modelling the flow of particles in wet granular systems using the Discrete Element Method (DEM). Typically, when particles move in a viscous fluid, DEM is coupled with Computational Fluid Dynamics (CFD) or Smooth Particle Hydrodynamics (SPH) methods to capture both particle and fluid motion. However, the computational expense and time required for one- or two-way coupled simulations, such as DEM–CFD or DEM–SPH, can be significant. In this research, a lubrication approximation is introduced to address fluid viscosity within DEM, which is particularly suitable for dense granular systems where viscous forces play a dominant role. DEM simulations incorporating the lubrication approximation were compared with in situ data obtained from Positron Emission Particle Tracking (PEPT) experiments. These experiments involve a cylindrical setup of radius R=230mm and length L=200mm, filled with 10mm spherical glass beads at a 50% fill fraction, and various mixtures of water and glycerol (60%, 75%, and 90% by weight) as the fluid phase. Simulations are conducted within the LIGGGHTS–DEM framework, and detailed comparisons with PEPT data assess the suitability of the lubrication approximation for rotating drum systems. The analysis of tangential velocity profiles across different Stokes numbers reveals the applicability of the same constitutive equation for modelling both the PEPT and DEM data within a specified viscosity range, with the exception of the highest viscosity. This understanding is crucial for interpreting the flowing layer dynamics and optimising the simulation parameters for accurate predictions.

The Discrete Element Method (DEM) is a widely used numerical method for studying, analyzing, and optimizing particulate processes. Based on contact models, DEM can predict the interactions of particles and particles with wall surfaces. Various contact models have been developed that consider various deformation and adhesion behaviors, particle shapes, and surface morphologies. However, accurate prediction of the real behavior requires experimental estimation and calibration of the model parameters for the studied particles. The objective of this paper is to provide a comprehensive overview of recent advancements in particle-property measurement methods, with a particular focus on the parameters of contact models. The measurement methods are classified into different categories, including static and dynamic loading and single particle and particle bed tests, with and without consideration of the surrounding liquid. A range of measurement techniques for estimating elastic and plastic properties, as well as friction and restitution coefficients in both air and liquid environments, are described, including nano- and tribo-indentation, compression, and impact tests. The applicability of these techniques was demonstrated using our own elastic–plastic model. The impacts of various influencing factors on material parameters and contact interactions are discussed. This overview can help identify suitable experimental techniques for calibrating and validating DEM models.

Silos are widely used structures designed with hoppers or as flat bottoms to store different types of particulate materials (pellets, grains, powders, etc.). Although they have been used in industry for a long-time, further research is needed. This article provides design recommendations that prevent structural failures and ensure efficient discharge. The discrete element method (DEM) has been commonly used in recent decades in silo/bin research to simulate the behavior of stored materials. In this study, DEM was used to investigate the relationship between the slot width and particle size on the discharge behavior in a flat-bottom silo using particle shape configurations with different elongations and aspect ratios. Sixteen DEM models were developed using the multi-sphere approach to obtain the particle configurations, ranging from single spheres to 4×4×4 clusters of spheres arranged along orthogonal axes (x, y, and z). The simulations involved between 459 and 29,383 particles with a total mass of 40 kg in each model. The established comparisons include information on the bulk density, velocity profiles, residual masses, and discharge rates.

Aluminum powder is an attractive fuel, especially for micropropulsion, due to its wide availability, high volumetric and gravimetric energy densities, and environmental friendliness. For safety reasons, the powder is preferably stored and supplied to a combustion chamber separately from the oxidizer. However, powder supply systems for micropropulsion applications are technologically immature. To improve this situation, a powder supply system was developed, and the impacts of important design factors on the supply characteristics were studied. This included operating modes, carrier gas pressures, gas line conductances, and powder line lengths. The results showed that, when compared to continuous operation, pulsed operation achieved 1) significantly higher powder flow rates and 2) significantly higher ratios of powder to carrier gas mass flow rates, ϕ. Furthermore, higher carrier gas pressures can increase the powder mass flow rates, but this reduces ϕ. The gas line conductance had a lesser impact on the powder flow rates than the pressure, but increasing the former also increased the powder flow rates without significant impact on ϕ. The powder line length had a negative impact on the powder flow rate, but no significant impact on ϕ. These results will help the future development of micropropulsion systems that use powder fuels.

Dry powder coating, characterised by the blending of poorly flowing powders with finer coating powders to optimise flowability, represents a sophisticated and evolving approach to powder processing. The optimisation of this method involves precise formulation, carefully combining powders with different particle properties to achieve a desirable blend aimed at enhancing the flow characteristics during the application process. Over the last decade, this field has witnessed increasing activity, focusing on key mixing parameters, such as mixer type and mixing power, as well as understanding the influence of constituent powder characteristics, including size ratio, density, and cohesion. Various techniques have been used to assess the flowability improvement or quantify the degree of coating. This review aims to provide a comprehensive exploration of the literature on powder coating research, highlighting its significance in both academic research and industrial applications. This paper discusses current coating analysis techniques using state-of-the-art equipment and reviews recent findings, particularly the nascent attempts to establish regime maps for dry powder coating.

Particle motion affects microscopic momentum, heat/mass transfers, and macroscopic mixing/separation performances in engineering processes ranging from those involving granular flows to multiphase systems. The primary interests of researchers seeking to improve the performances of such processes are finding the best ways to understand particle motion. Several methods have been developed to understand particle motion, among which radiation-based techniques have shown particular advantages because they can non-invasively investigate dynamic particle motion. This article reviews four prominent radiation-based techniques for studying particle motion, including Radioactive Particle Tracking (RPT), Positron Emission Particle Tracking (PEPT), X-ray-based methods and Magnetic Resonance Imaging (MRI). The principles and characteristics of these techniques are explained, and recent advances and applications are reviewed.

In powder-based Additive Manufacturing (AM) the precise control of process parameters plays a significant role in the quality and efficiency of the printing process. Among these, the effect of temperature has received less attention in the literature, although it is a significant factor that influences the inter-particle forces and, consequently the powder flow and spreading behaviour of powders. In selective laser sintering (SLS) or selective laser melting (SLM), pre-heating the chamber and powder bed is a required step prior to sintering, hence, the temperature can significantly influence the layer adhesion and spread quality. In this context, the present study explores the effect of elevated temperature on the flow and spreading behaviours of AlSi10Mg powders. The flow properties of two different grades of aluminium alloy powders are characterised using the Carney and Hall flow tests, angle of repose and shear test techniques at different temperatures and correlated with the spreading behaviour at elevated temperatures, measured using the spreading rig with a heated bed developed at the University of Leeds. This study revealed that at elevated temperatures the spreadability of AlSi10Mg powders worsens because of changes in interparticle forces and particle surface interactions.