• Energy Research

Abstract In this paper, percolation of cubical particles through a packed bed is studied using discrete element method (DEM). The influences of some key variables e.g. restitution coefficient, sliding coefficient, rolling coefficient and diameter ratio on cubical particles percolation behaviour were comprehensively analyzed. The results show that, compared with the sphere particles, cubical particles also have a constant vertical velocity during percolating in the packed bed, but it has a larger percolation velocity and a less off-center of radial distance. Restitution coefficient affects the cube percolation behaviour. With increasing the restitution coefficient, the percolation velocity decreases, but the radial dispersion coefficient increases. Effects of sliding friction coefficient and rolling friction coefficient show the same tendency. The percolation velocity and radial dispersion decrease with the increase of friction coefficient. The ratio of cubical particle diameter to packing particle diameter (diameter ratio) is another dominant variable that significantly affect the percolation behaviour. The percolation velocity and radial dispersion coefficient increases with decreasing the diameter ratio. The simulations are useful for understanding percolation and segregation of multi-scale cubical materials and optimization in cubical particles handling and mixing.

Abstract In this paper, percolation of cubical particles through a packed bed is studied using discrete element method (DEM). The influences of some key variables e.g. restitution coefficient, sliding coefficient, rolling coefficient and diameter ratio on cubical particles percolation behaviour were comprehensively analyzed. The results show that, compared with the sphere particles, cubical particles also have a constant vertical velocity during percolating in the packed bed, but it has a larger percolation velocity and a less off-center of radial distance. Restitution coefficient affects the cube percolation behaviour. With increasing the restitution coefficient, the percolation velocity decreases, but the radial dispersion coefficient increases. Effects of sliding friction coefficient and rolling friction coefficient show the same tendency. The percolation velocity and radial dispersion decrease with the increase of friction coefficient. The ratio of cubical particle diameter to packing particle diameter (diameter ratio) is another dominant variable that significantly affect the percolation behaviour. The percolation velocity and radial dispersion coefficient increases with decreasing the diameter ratio. The simulations are useful for understanding percolation and segregation of multi-scale cubical materials and optimization in cubical particles handling and mixing.

Abstract In this work we report the Na incorporation from Na-doped Mo (Mo-Na) back contact for kesterite Cu2ZnSnS4 solar cells on flexible stainless steel substrates. It is demonstrated that Na can be effectively incorporated into CZTS by inserting Mo-Na layer at back contact. Direct contact of CZTS and Mo-Na layer leads to poor homogeneity and adhesion. The thickness of MoS2 formed at the back contact depends on the presence of Na and whether Mo contacts with CZTS directly. Back contact configuration with a Mo capping layer on Mo-Na layer is found to be helpful to maintain the advantages of Mo back contact and control the thickness of MoS2 interface. As a result, CZTS device fabricated on this configuration yields higher conversion efficiency of 6.2%. However, this efficiency is still far lower than that on traditional soda lime glass substrate which shows efficiency over 8%. The loss mechanism of device fabricated on stainless steel is investigated and analyzed according to the device performance and electrical parameters.

Abstract In this work we report the Na incorporation from Na-doped Mo (Mo-Na) back contact for kesterite Cu2ZnSnS4 solar cells on flexible stainless steel substrates. It is demonstrated that Na can be effectively incorporated into CZTS by inserting Mo-Na layer at back contact. Direct contact of CZTS and Mo-Na layer leads to poor homogeneity and adhesion. The thickness of MoS2 formed at the back contact depends on the presence of Na and whether Mo contacts with CZTS directly. Back contact configuration with a Mo capping layer on Mo-Na layer is found to be helpful to maintain the advantages of Mo back contact and control the thickness of MoS2 interface. As a result, CZTS device fabricated on this configuration yields higher conversion efficiency of 6.2%. However, this efficiency is still far lower than that on traditional soda lime glass substrate which shows efficiency over 8%. The loss mechanism of device fabricated on stainless steel is investigated and analyzed according to the device performance and electrical parameters.

Abstract The presence of kaolinite as mineral in coal may significantly influence coking properties of coal, therefore affects quality of coke that is produced from such coals. In this paper, the effect of kaolinite addition on pyrolysis behaviour of coal was examined by blending kaolinite with a medium rank, high vitrinite, low ash coal. Thermoplasticity of coal was characterised by using a Gieseler Plastometer. Char and tar samples were collected through pyrolysis at 450 °C in a Gray-King apparatus. The functional groups of the char samples were measured using FT-IR spectroscopy, while tar compositions were characterised using GC/MS. The results indicated that under the test conditions, addition of up to 5 wt% of kaolinite could increase the coal fluidity, attributed to the decrease in CH 3 /CH 2 ratio in the residual char. The proportion of C O/C C bonds affected the plastic range. Furthermore, the results showed that kaolinite addition suppresses the evolution of low molecular compounds in tar phase. This study provides useful information for optimising coal blending in cokemaking industry.

Abstract The presence of kaolinite as mineral in coal may significantly influence coking properties of coal, therefore affects quality of coke that is produced from such coals. In this paper, the effect of kaolinite addition on pyrolysis behaviour of coal was examined by blending kaolinite with a medium rank, high vitrinite, low ash coal. Thermoplasticity of coal was characterised by using a Gieseler Plastometer. Char and tar samples were collected through pyrolysis at 450 °C in a Gray-King apparatus. The functional groups of the char samples were measured using FT-IR spectroscopy, while tar compositions were characterised using GC/MS. The results indicated that under the test conditions, addition of up to 5 wt% of kaolinite could increase the coal fluidity, attributed to the decrease in CH 3 /CH 2 ratio in the residual char. The proportion of C O/C C bonds affected the plastic range. Furthermore, the results showed that kaolinite addition suppresses the evolution of low molecular compounds in tar phase. This study provides useful information for optimising coal blending in cokemaking industry.

AbstractIn order to understand the complicated phenomena of pulverized coal injection (PCI) process in blast furnace (BF), several mathematical models have been developed by the UNSW and BSR cooperation. These models are featuring from coal combustion in a pilot‐scale test rig, to coal combustion in a real BF, and then to coal/coke combustion in a real BF, respectively. This paper reviews these PCI models in aspects of model developments and model applicability. The model development is firstly discussed in terms of model formulation, their new features and geometry/regions considered. The model applicability is then discussed in terms of main findings followed by the model evaluation on their advantages and limitations. It is indicated that the three PCI models are all able to describe PCI operation qualitatively. The model of coal/coke combustion in a real BF is more reliable for simulating in‐furnace phenomena of PCI operation qualitatively and quantitatively. Such model gives a more reliable burnout prediction over the raceway surface, which could better represent the amount of unburnt char entering the coke bed. These models are useful for understanding the flow‐thermo‐chemical behaviours and then optimising the PCI operation in practice.

AbstractIn order to understand the complicated phenomena of pulverized coal injection (PCI) process in blast furnace (BF), several mathematical models have been developed by the UNSW and BSR cooperation. These models are featuring from coal combustion in a pilot‐scale test rig, to coal combustion in a real BF, and then to coal/coke combustion in a real BF, respectively. This paper reviews these PCI models in aspects of model developments and model applicability. The model development is firstly discussed in terms of model formulation, their new features and geometry/regions considered. The model applicability is then discussed in terms of main findings followed by the model evaluation on their advantages and limitations. It is indicated that the three PCI models are all able to describe PCI operation qualitatively. The model of coal/coke combustion in a real BF is more reliable for simulating in‐furnace phenomena of PCI operation qualitatively and quantitatively. Such model gives a more reliable burnout prediction over the raceway surface, which could better represent the amount of unburnt char entering the coke bed. These models are useful for understanding the flow‐thermo‐chemical behaviours and then optimising the PCI operation in practice.

Abstract Simulations based on Discrete Element Method3 (DEM) are carried out to investigate the effect of blade-supporting spokes of the impeller on overall mixing performance in a ribbon mixer. The influences of some key variables are studied, which include spoke number, particle cohesiveness and fill-level. The overall mixing performance is mainly assessed using the Lacey's mixing index based on the coordination number. The axial velocity variation, axial diffusion coefficient, and contact forces are also analysed to depict the effect of spokes on the mixing. It is found that for non-cohesive particles, increasing the number of spokes is favourable for achieving faster overall mixing when the fill level is large. But this comes at the expense of increased contact forces. It is also found that particle dispersion due to axial velocity fluctuations plays a major role in mixing of non-cohesive particles, but not the axial diffusion. For cohesive particles, no-spokes impeller shows the best overall mixing performance. The effect of spokes on the overall mixing performance is complex and depends on the fill level. At higher fill-levels, the mixing curves are relatively less affected by the number of spokes, while at lower fill-levels, mixing is delayed extensively for 2 and 6 spokes impellers, but the 4-spokes impeller shows some recovery of mixing performance towards that of no-spoke impeller. At low fill-levels, axial diffusion is a governing factor of the mixer performance. Conversely, at high fill-levels, particle dispersion due to the velocity fluctuations become the dominant mixing mechanism for all types of impellers considered. Generally, adding spokes to impellers causes only a slight increase in the contact forces between particles when mixing cohesive particles but a very significant increase in the contact forces when mixing non-cohesive particles. The results indicate that the attached parts in the impeller should be optimized under different operational conditions, which can be guided by DEM simulations.