Thesis

UNDERSTANDING PASSIVE CONTROL OF BULK
MATERIAL DUST EMISSIONS IN DISCRETE FEED

SITUATION
Amir A. Esmaili
Master of Engineering Management

Bachelor of Industrial Engineering
A thesis submitted in fulfilment of

the requirements for the degree of
Doctor of Philosophy
School of Engineering - Department of Mechanical Engineering

The University of Newcastle
Callaghan, New South Wales 2308
Australia
October 2016
DECLARATION
I hereby certify the work embodied in this thesis is the result of original research and has not
been submitted for a higher degree to any other University or Institute and, to the best of my
knowledge and belief, contains no material previously published or written by another person,
except where due reference has been made in the text.
(Signed)
Amir A. Esmaili
The University of Newcastle
October 2016
ii
Acknowledgements
ACKNOWLEDGEMENTS
The work embodied in this thesis took place in the Centre for Bulk Solids and Particulate
Technologies at the University of Newcastle Australia, and as such I wish to thank the
Centre directors Emeritus Professor Alan Roberts, and Professor Mark Jones for
providing the opportunity to study within the Centre.
I would also like to thank my principal supervisor, Associate Professor William McBride,
and co-supervisors Associate Professor Craig Wheeler and Dr Tim Donohue, for their
continued support and supervision of my research work.
Acknowledgement is also given to the staff at Tunra Bulk Solids as they have offered
tremendous technical support during my research, and assistance during the experimental
test work.
Further, I would like to acknowledge the Australian Research Council for the financial
support of this research project under the Discovery Projects Scheme.
The University of Newcastle Scholarship (UNRSE ARC) must also be acknowledged as

it provided additional financial support during my PhD study.
Above all, I wish to thank my family, in particular my wife and my parents, for all of the
encouragement and supports that they have given me during my studies. To them I am
deeply indebted.
iii
Table of Contents
Table of Contents
Chapter 1: Introduction ............................................................................................. 4

1.1 Thesis Structure .................................................................................................. 6
Chapter 2: Dust Definition and Related Environmental Standards ......................... 9

2.1 Dust - A Definition.............................................................................................. 9
2.2 Existing Dust Emission and Air Quality Standards ............................................ 12
2.3 Australian Standard (Australia Department of Sustainability, 2013) .................. 12
2.4 World Bank Guidelines (World Health Organization, 2005) .............................. 14
2.5 United States Standard (US Environmental Protection Agency (EPA), 1997) .... 15
2.6 Measurement Equipment ................................................................................... 16
2.7 Dust Control Methods ....................................................................................... 21
Chapter 3: Free-Falling Streams .............................................................................. 26

3.1 Background ....................................................................................................... 26
3.2 Experimental Work on Stream Diameter ........................................................... 29
3.3 Analysis of Velocity and Stream Diameter Method ........................................... 37
3.4 Theoretical Terminal Velocity Calculation of Gravity Falling Particles under Drag
Influence ................................................................................................................. 39
3.5 Results and Discussions .................................................................................... 41
3.6 Conclusion ........................................................................................................ 45
iv
Table of Contents
Chapter 4: Free-Falling Air Entrainment Measurement and Calculation ............. 47

4.1 Background ....................................................................................................... 47
4.2 Particle Motion under Gravity ........................................................................... 52
4.3 Development of an Entrained Air Calculation Method ...................................... 59
4.4 Experimental Set-up and Validation .................................................................. 63
4.5 Results and Discussions - Theory versus Experimental Data ............................. 66
4.6 Conclusion ........................................................................................................ 69
Chapter 5: Free-Falling – Air Entrainment Computer Simulation ........................ 70

5.1 Computational Fluid Dynamics (CFD) .............................................................. 71
5.2 Simulation Theory ............................................................................................. 73

5.2.1 Mass Conservation Equations ........................................................................ 73
5.2.2 Momentum Conservation Equations .............................................................. 74
5.2.3 Drag Model ................................................................................................... 75
5.2.4 Granular Shear Viscosity ............................................................................... 82
5.2.5 Turbulence Model ......................................................................................... 83
5.3 PIV Experimental Testing ................................................................................. 89
5.4 Simulation of Free-Falling................................................................................. 92
5.5 Results .............................................................................................................. 99
5.6 Conclusion ...................................................................................................... 106
Chapter 6: Review of Air Flow within Dump Hopper Geometry ......................... 108
6.1 Experimental Set-up ........................................................................................ 109
6.1.1 Test Procedure ............................................................................................. 115
v
Table of Contents
6.2 Model Simulation and CFD Simulation ........................................................... 122

6.2.1 CFD Multiphase Flow ................................................................................. 122
6.2.2 Mesh Sensitivity Analysis............................................................................ 126
6.3 Results ............................................................................................................ 127
6.4 Conclusions..................................................................................................... 136
Chapter 7: Application of the AVCF Theory in More Complex Mining Applications

................................................................................................................................. 138
7.1 Background ..................................................................................................... 138
7.2 Review of Particle Trajectory Theory .............................................................. 139
7.3 Material Stream Discharging from Haulage Truck - Air Entrainment Theory Work
............................................................................................................................. 141
7.4 Example Case ................................................................................................. 145
7.5 Comparison between the Introduced Theory Work and Simulation.................. 149
7.6 Conclusion ...................................................................................................... 151
Chapter 8: Conclusions and Recommendations .................................................... 153

8.1 Thesis Conclusion ........................................................................................... 154
8.2 Recommendations for Future Work ................................................................. 156

8.2.1 Further Testing on the AVCF Theory in Free- Falling case .......................... 156
8.2.2 Further Testing on the Introduced Dump Hopper Entrained Air Theory case study
............................................................................................................................. 156
8.2.3 Further Simulation with Prediction of the Flow of Dust Particles ................. 156
vi
List of Figures
List of Figures
Figure 1. Sizes of Common Particles- (Wark & Warner, 1981) .................................. 10
Figure 2. Respirable Dust as defined by AS 2985....................................................... 11
Figure 3. Frisbee Gauge .............................................................................................. 18
Figure 4. Deposit Gauge ............................................................................................. 19
Figure 5. Directional Dust Gauge ................................................................................ 20
Figure 6. Free-Falling Air Entrainment Schematic ...................................................... 27
Figure 7. Stream Diameter Test Rig - Concept 1 ......................................................... 29
Figure 8. Coal Measured Weight (gm) ........................................................................ 30
Figure 9. Coal Percentage in each Annulus ................................................................. 31
Figure 10. Stream 'S''ing Phenomenon ........................................................................ 34
Figure 11. Coloured Particles - Particle Trajectory Tracking ....................................... 35
Figure 12. Layout of Experimental Test Equipment .................................................... 36
Figure 13. An Example of Thresholding Used in Image Analysis (plastic pellets) ....... 37
Figure 14. Example Distribution of Stream Diameter .................................................. 38
Figure 15. Area Under Curve Given by Standard Deviations....................................... 39
Figure 16. Terminal Velocity of Particles with Spherical Shape .................................. 41
Figure 17. Velocity of Different Type and Size Particles versus Vertical Displacement
................................................................................................................................... 42
Figure 18. Stream Diameter versus Drop Height ......................................................... 43
Figure 19. Comparison of stream diameters for tests carried out with 26 and 52mm
hoppers ....................................................................................................................... 44
Figure 20. Measured Density of the Stream versus Vertical Distance ......................... 45
Figure 21. Schematic View of Free-falling Stream ...................................................... 60
Figure 22. Schematic View of Experimental Set-up .................................................... 63
Figure 23. Air Entrainment Test Rig ........................................................................... 65
Figure 24. Entrained Air versus Drop Height for Sand ................................................ 67
Figure 25. Entrained Air versus Drop Height for 3mm Glass Beads ............................ 67
Figure 26. Entrained Air versus Drop Height for 3 mm Plastic Pellets......................... 68
Figure 27. PIV Set-up Schematic ................................................................................ 90
Figure 28. CFD Geometry used for Benchmark Simulation (cylinder shape) ............... 93
i
List of Figures
Figure 29. Final CFD Geometry selected for Simulations (cone shape) ....................... 94
Figure 30. Air Entrainment Comparison; Cylindrical Domain Simulation versus Conical
Shape Simulation ........................................................................................................ 94
Figure 31. 3 mm Glass Bead Turbulence Model Comparison SST vs K Epsilon ......... 96
Figure 32. 3mm Plastic Pellets Turbulence Model Comparison SST vs K Epsilon ...... 97
Figure 33. 6mm Glass Beads Turbulence Model Comparison SST vs K Epsilon ......... 97
Figure 34. PIV Image – Velocity of 6 mm Glass Beads in Free-Falling ( ) ............ 100
Figure 35. PIV Image -Velocity of 3 mm Plastic Pellets in Free-Falling ( )........... 100
Figure 36. PIV Image - Velocity of 3 mm Glass Beads in Free-Falling ( ) ............ 101
Figure 37. PIV Velocity Histogram of 3mm Glass Beads ( ) ................................. 102
Figure 38. Cross-sectional Velocity Profile of 3 mm Glass Beads ............................. 103
Figure 39. Simulation Results for the Air Velocity Distribution at a Plane located 900 mm
below the Hopper Outlet ........................................................................................... 104
Figure 40. Air Entrainment Results from the CFD Simulations for 3mm Glass Beads 105
Figure 41. Air Entrainment Results from the CFD Simulations for 3mm Plastic Pellets
................................................................................................................................. 105
Figure 42. Air Entrainment Results from the CFD Simulations for 6mm Glass Beads 106
Figure 43. 3D Scale Model ROM Dump hopper and inclined feeder ......................... 109
Figure 44. Dump Hopper Test Facility - Air Velocity Measurement Points ............... 110
Figure 45. Key Components of the Tested Hopper – Base Model .............................. 111
Figure 46. Typical Elastic Curtain ............................................................................. 112
Figure 47. Kimo Air Velocity Measurement Unit with Hot Wire ........................ 112
Figure 48 Example of Hot Wire Set-up Mounted on a Pipe ....................................... 113
Figure 49. Iron Ore PSD ........................................................................................... 115
Figure 50. Base Geometry......................................................................................... 116
Figure 51. Bespoke Scale Model Dump Truck Tray .................................................. 117
Figure 52. Dump Hopper - Arrangement 1 ................................................................ 118
Figure 55. Schematic View of Mining Dump Truck .................................................. 122
Figure 56 CFD Air and Solid Velocity Results – Base Model ................................... 128
Figure 57. Experimental Hopper during Physical Testing - Base Model .................... 128
ii
List of Figures
Figure 58. Cross-Sectional Pressure Profile ̶ Base Model ........................................ 129

Figure 59. CFD Air and Solid Velocity Results - Arrangement 3 .............................. 131
Figure 60. Experimental Hopper during Physical Testing – Arrangement 3............... 131
Figure 61. Cross-Sectional Pressure Profile ̶ Arrangement 3...................................... 132
Figure 62. CFD Air and Solid Velocity Results – Arrangement 2 .............................. 134
Figure 63. Experimental Hopper during Physical Testing – Arrangement 2............... 134
Figure 64. Cross-Sectional Pressure Profile – Arrangement 2.................................... 135
Figure 65. Dump Truck Bulk Material Stream Trajectory in Discharge Condition..... 139
Figure 66. Dump Truck Schematic View .................................................................. 145
Figure 67. Stream Trajectory – Proposed Model ....................................................... 147
Figure 68. Stream Width CFD Simulation – Top View ............................................. 148
Figure 69. Dump Hopper Iron Ore Velocity .............................................................. 150
Figure 70. Simulation of Stream into a Box............................................................... 151
iii
List of Tables
List of Tables
Table 1. Australian Standards for Ambient Air Quality ............................................... 13
Table 2 Dust Concentration Limits for the State of NSW ............................................ 14
Table 3. World Bank Guidelines for Ambient Air Quality........................................... 15
Table 4. United States Standards for Ambient Air Quality .......................................... 16
Table 5. Summary of Materials ................................................................................... 35
Table 6. Properties of Materials Tested ....................................................................... 66
Table 7. Comparison of RNAS Turbulence Models (Ansys13.0, 2010; Bakker, 2002) 88
Table 8. Summary of Materials ................................................................................... 91
Table 9. Mesh Sensitivity Analysis ............................................................................. 95
Table 10. CFD Simulation Parameters – General ........................................................ 98
Table 11. CFD Simulation Parameters – Specific ........................................................ 99
Table 12. PIV, Simulation and Theoretical Velocity Comparison .............................. 102
Table 13. Dump Hopper Enclosure Volume .............................................................. 118
Table 14. CFD Model Description ............................................................................ 125
Table 15. Mesh Sensitivity Analysis ......................................................................... 126
Table 16. Velocity Comparison between Simulation and Experiments ...................... 136
i
Nomenclature
Nomenclature
A Cross-sectional area of bulk material stream [m2]

Ar Archimedes Number [-]
Ap Particle cross-sectional area [m2]
CD Drag coefficient [-]
Cp Entrainment constant [-]
Cij Convection term [-]
d Particle diameter [m]
d Free-falling stream diameter [m]
d Hopper outlet diameter [m]
d Average particle diameter [m]
F Drag force [N]
f Drag function [-]
DT,ij Turbulent diffusion [-]
DL,ij Molecular diffusion [-]
e Restitution coefficient [-]
g Gravitational constant, 9.81 [m/s2]
Gk Turbulent kinetic energy due to the mean velocity gradients [-]
Gb Turbulent kinetic energy due to buoyancy [-]
G1ε Constant - Equation 2-21 [-]
h Drop height [mm]
hb Total drop height extension (free falling) [m]
hs Hopper section height (free falling) [m]
I̿ Stress tensor [-]
I2D Second invariant of the deviatoric stress tensor [-]
Ksg Momentum exchange coefficient [-]
m mass flow rate of bulk material [kg/m3]
i
Nomenclature
Q Volumetric flow rate of entrained air [Lit/Sec]

Rep Particle Reynolds Number [-]
S User-defined source terms [-]
t Time [s]
t Time for a single particle free fall under the influence of drag [s]
Free stream velocity [m/s]
U Correction coefficient for random fluctuation velocity [-]
V Velocity (stream or particles) [m/s]
∞ Terminal velocity [m/s]
vo Initial velocity [m/s]
Y Turbulence dissipation [-]
β fluid-solid exchange coefficient [-]
ε Rate of dissipation [-]
εg Gas volume fraction [-]
εs Solids volume fraction [-]
Θ Granular temperature [m2/s2]
Turbulence kinetic energy [-]
λ Viscosity [Pa s]
λg Gas viscosity [Pa s]
λs Solids viscosity [Pa s]
μ Contact friction coefficient [-]
μe Equivalent friction [-]
μ Gas shear viscosity [kg/(m s)]
μ Solids shear viscosity [kg/(m s)]
μ , Collision viscosity [kg/(m s)]
μ , Kinetic viscosity [kg/(m s)]
μ , Friction viscosity [kg/(m s)]
Kinematic viscosity of fluid [m2/s]
ρf Fluid phase density [kg/m3]
ρp Particle density [kg/m3]
ii
Nomenclature
ρs Solid density [kg/m3]

Turbulent Prandtl for k [-]
Turbulent Prandtl for [-]
Particle relaxation time [μs]
τg Drag on the underlying surface [N/m2]
τ Gas stress tensor [N/m2]
τ Solid stress tensor [N/m2]
φ Specularity coefficient [-]
∇ Nabla operator [-]
iii
Abstract
Abstract
Fugitive emissions arising from the handling of particulate materials have occurred since
these materials were first handled. As the volume of materials handled has increased so
has the volume of fugitive materials; the fraction of the fugitive material focused on in
this thesis, we generally call dust. In modern operations, dust represents a lost income
stream, as the dust is often saleable product; increased cost of operations due to cleaning
costs; a public nuisance for those working and living adjacent to material-handling
operations; an explosion risk with some dust types; an environmental pollutant, and
significantly a multitude of health risks due to slips, trips and inhalation of dust
(Hamelmann & Schmidt).
Government regulation through various environmental pollution agencies is continually

reducing the so-termed acceptable limits for dust emissions associated with different
operations and in different environments. As these limits reduce, a greater level of
operational care and often cost is required by the operators to comply.
Active dust control is often seen as an easy method to reduce emissions. In the simplest
form, an air mover is used to pull dust laden air away from the process in question, before
air separation and returning the fine product back to the process line. Active measures
clearly carry an energy penalty to operate the air movers, but also have high costs
associated with duct and pipe wear, air-mover filtration system maintenance and
reliability issues. Further, there are many areas where active dust control is impractical.
Passive dust control in contrast, through an understanding of the mechanics of dust

generation, aims to reduce the total emissions through a range of mechanisms, without
external energy input. This can be achieved for example by embedding the smaller
particles fractions deeper within a flowing stream (Wheeler, Krull, Roberts, & Wiche,
2007), or careful positioning of baffles and stilling zones in equipment. Unfortunately
i
Abstract
with passive control systems, there is a much greater need to ‘understand the system’
before an appropriate passive solution can be implemented. This is unlike a high-cost
active system where a simple up-scaling of equipment can often overcome a poor
understanding of the system.
This thesis presents a volume of work expanding the overall understanding of the
underlying mechanics required for passive dust control, particularly related to transient
events such as dump hoppers.
Air entrained into a falling stream of bulk material under the influence of gravity is known
as the main source of dust generation (R. Ansart, A. d. Ryck, & J. A. Dodds, 2009b). For
calculating the amount of entrained air, several researches and formulas have been
published. However, the accuracy of different formulas varies under different conditions.
To understand the impact of different parameters on the behaviour of a material stream,

this research was undertaken in a particular sequence, to enable detailed analysis in a
controlled manner. In general, free falling largely contributes to the discrete bulk-
material-handling case activities. Therefore, research was initiated on free-falling streams
to better identify the impact of key influencing variables. This research provided an
insight into the free-falling and influencing parameters such as particle diameter, particle-
solid density, stream diameter and drop height. Resulting from these investigations, a new
approach to calculate the magnitude of air entrained into the free-falling stream of
particles (AVCF - Air-Entrainment using Volume Conservation Theory in Free-Falling)
has been developed (Esmaili, Donohue, Wheeler, McBride, & Roberts, 2013).
Computational Fluid Dynamics (CFD) simulation was successfully applied to different
free-falling conditions. These simulations were then verified with Particle Imaging
Velocimetry (PIV), high-speed videos and particle-tracking techniques. A close
correlation between the provided theory (AVCF), CFD and experimental data was
observed.
ii
Abstract
The knowledge obtained from the analysis of free falling was extended to transient bulk-
material-handling cases. Dump hoppers are known to be one the major dust generation
sources within the mining industry (Cecala, 2010), and was selected to extend the work
completed on free falling air entrainment research.
Scale model tests were conducted to provide air velocity readings, airflow patterns and
high-quality footage for simulation verification purposes. The two-phase CFD simulation
method is considered as a compromise solution, optimising accuracy, time consumption
and ease-of-use considerations. The potential to further improve performance in the scale
model tests were investigated using these CFD method. The influences of the drop height,
particle size, particle density and the design of the dump hopper geometry, in relation to
the air-flow pattern within the enclosure, and subsequently reducing dust emissions are
discussed. Moreover, the effects of ‘air cells’, ‘curtains’, ‘baffles’ and introduced new
technique (additional section) are investigated. The results reveal that the flow pattern of
the air within the dump hopper geometry can be improved and the flow slowed down if
air cells, baffles, curtains or any other technique is utilised appropriately.
In addition to the above, the introduced theory work for the air-entrainment-magnitude
calculation concept has been expanded to dump hoppers by utilising particle trajectory
analysis. It has been concluded that this formula provides a close correlation to simulation.
iii
Introduction
Chapter 1: Introduction
Industrial air pollution is recognized as a major environmental and health concern. Air
pollution occurs when the air contains unwanted gases, dust, fumes or odour in more
than allowable limits. In many industries across the world, the topic of dust control is a
major concern. Amongst different industries and applications, bulk materials handling
is known as one of the sources of fugitive dust generation (Liu, 2003). Quantifying and
controlling dust (generally particles under 100μm in diameter) pollution within many
industries is globally an increasing focal point as more stringent Occupational Health,
Safety and Environment (OHSE) regulations are specifying significantly lower
acceptable dust levels across all industries. Additionally, higher ambient dust levels in
work environments often necessitate equipment with higher dust-ingress protection,
requiring more advanced sealing and eventually, higher costs.
Dust control systems are classified into two main categories active and passive. Active
dust control systems are one of the most commonly employed means to minimise
fugitive dust from material handling systems. They use external energy (through air
movers) or surfactants such as water spray, foam and water-misting systems, dust-
filtering systems, enclosures and fan systems, to control fugitive dust. While these
systems are reasonably effective, their main drawback is the ongoing costs associated
with operation and maintenance. (Tooker, 1985), (Junemann & Holzhauer, 1992), (Weiss,
1992), (Herrmann and Evensen, 1994), (ACGIH, 1992).
In recent times, passive dust control has received much attention as a complementary
technology or in some cases a viable alternative to active dust control systems. Passive
dust control relies on containing, or re-entraining, the dust within the moving bulk
material stream and relies on a thorough understanding of the interaction between the
bulk material and the surrounding air. Hence, there are significant incentives to conduct
research on the fundamentals of passive dust control.
4
Introduction
Within bulk materials handling processes, the action of material free fall is generally
unavoidable. In this case, the air surrounding the free-falling stream is entrained into the
stream of bulk material as the particles naturally dilate, which results in generation of
dust upon subsequent compaction of the stream (Cooper & Arnold, 1995a). Obtaining
a better understanding of the way air is induced into the stream of bulk material, and the
quantity of the induced air, will assist in the design of more efficient dust control
systems.
The Office of Environment and Heritage (OEH) in NSW Australia, operates a

comprehensive air-quality monitoring network to provide the community with accurate
and up-to-date information about air quality. Dust is generally recognised as one
significant air pollution parameter, which can lead to health and environmental issues.
Bulk material handling and transportation are accountable for ninety percent of diffuse
dust emissions produced during the storage and transport of bulk material (Junemann &
Holzhauer, 1992). In conventional bulk material handling cases, the free falling of bulk
material either forms the complete task (i.e. dumping stations) or contributes to dust
generation in a more complicated applications (i.e. dump hoppers and transfer points).
The introduction of more stringent Environmental and Occupational Health and Safety
regulations demands significantly lower acceptable dust levels across all industries
(Australian Standard, 2012; British Standard, 1972). Similarly, much pressure is placed
on minimising spillage to improve workplace safety and reduce plant clean-up costs.
For example, it has been estimated that the annual cost to industry due to dust and
spillages from materials handling operations was US$300 million in 1987, in the UK
(Wilkinson, 1989). Furthermore, increased equipment costs, such as those associated
with the need for improved sealing efficiencies, are another strong stimulant for industry
to reduce dust emissions.
5
Introduction
1.1 Thesis Structure
The main purpose of the research is to better understand the behaviour of dust particles
in air and the effective parameters for the reduction of dust using a passive dust control
framework, which can then minimise any external energy or surfactant material
utilisation. The current work will generally focus on discrete bulk material handling
cases.
A variety of discrete bulk material handling equipment and scenarios are utilised daily
within industry. A considerable portion of material handling processes do contain a form
of free-falling, hence specific attention is given to the topic of free fall. As mentioned
earlier, within the free-falling process, the air surrounding a free-falling stream of bulk
material is entrained into the stream and upon compaction of the stream may be
exhausted into the atmosphere. The air being moved by the process of stream discharge
and exhausted air upon compaction has the potential to carry airborne (fine) particles in
the direction of air flow. Hence, a better understanding of the mechanics of air
entrainment into a stream of bulk material and the magnitude of the entrained air, will
assist in the design of a more efficient dust control system (Z. Q. Liu, P. Cooper, & P.
W. Wypych, 2007).
By utilising the knowledge obtained from work on particles free falling, an efficient,
flexible and cost-effective design concept is introduced to aid discrete bulk material
handling tools i.e. dump hoppers, with particular emphasis being placed on predicting
dust control performance. A theory is developed and the proposed method is verified by
experimental results and is applied to develop some rules to improve the efficiency of
the design of dump hoppers and any other type of enclosures, as far as dust-emission
reduction is concerned.
The following section highlights the line of thoughts followed during this work. The
high-level structure of the thesis is outlined as follows:
6
Introduction
Chapter 1 Reviews existing dust definitions and outlines dust emission associated
problems within industry. The shortcomings of different active dust control methods are
discussed and the significant advantages of passive dust control are introduced. Chapter
1 also outlines the scope and detailed structure of the thesis.
Chapter 2 Explores various existing standards on dust definition, dust emission and air
quality. High level dust control strategies and control methods are provided.
Chapter 3 Elaborates on free-falling stream diameter in detail. Existing approaches and

theories are reviewed. Several test procedures are discussed, highlighting their
advantages and disadvantages. An accurate and user-friendly method for free-falling
stream diameter measurement is developed and presented in this chapter.
Chapter 4 Reviews existing air-entrainment models for free-falling streams. Moreover

a new theoretical model for free-falling air entrainment is developed. A detailed
comparison of existing models versus the new model and experimental results is
discussed in detail. On the new proposed model, a detailed sensitivity analysis is
undertaken and reviewed in detail.
Chapter 5 Presents the detail works being undertaken in relation to modelling/validating

the introduced free-falling stream air entrainment model by utilising simulation and
experimental case studies. For CFD model verification purposes, a set of PIV (Particle
Imaging Velocimetry) experiments are conducted to guarantee that simulation results
are accurate.
Chapter 6 Focuses on liberated dust from discrete bulk material handling cases. The
dump hopper, as one of the more widely utilised discrete bulk material handling mining
equipment is selected and examined in detail. A detailed review of airflow pattern within
dump-hopper geometry is conducted using a CFD model and scale model experimental
testing. Key design parameters are identified and assessed.
7
Introduction
Chapter 7 Develops a new theoretical model for quantifying the magnitude of air
entrained into the stream of free-falling material, based on the introduced free-falling
air-entrainment model. A sample case is chosen and results obtained from the theory
works are compared with the CFD result.
Chapter 8 Presents a summary and evaluations of the contributions of this thesis and a
discussion on future research directions.
8
Dust Definition and Related Environmental Standards
Chapter 2: Dust Definition and Related

Environmental Standards
All industries are morally and legally obligated to comply with best practice and local
regulations and standards in relation to dust and particulate emission. Depending on the
resource availability and specific health and environmental requirements across various
countries, different standards have been developed and regulated to define the
compliance criteria.
Over the years, the impacts of fugitive dust have been understood better and as a result
dust control technology has been improved, a reduction in accepted emissions often
follows. This chapter intends to review the current standards relating to dust emissions
and its relation to the materials handling sector.
2.1 Dust - A Definition
There are numerous dust definitions existing within available standards and literatures
depending on their area of concern, however, the majority of them agrees that “dust” is
a generic term used to describe fine particles that are suspended in the atmosphere under
specific circumstances.
The United States Mine Safety and Health Administration (MSHA) describes dust as
“tiny solid particles that may become airborne from the original state without any
chemical or physical change other than fracture”. (Mine Safety and Health
Administration, 2001 Retrieved 7th February, 2012)
A more specific definition for dust is given by the ISO (International Organisation for
Standardisation) Standard defining dust as “small solid particles under 75μm in diameter
that settle under their own weight - which may remain suspended for some time” (ISO,
1994)
9
Furthermore, based on Australian Standard AS 60068.2.68-2003, The Environmental

Testing Standard, dust is defined as “particulate matter of unspecific origin or
composition whose size ranges from 1 micron to 150 microns” (Australian Standard,
2003).
According to the Glossary of Atmospheric Chemistry Terms (International Union of

Pure and Applied Chemistry, 1990), dust is defined as “small, dry, solid particles
projected into the air by natural forces, such as wind, volcanic eruption, and by
mechanical or manmade processes such as crushing, grinding, milling, drilling,
demolition, shovelling, conveying, screening, bagging, and sweeping”. Dust particles
are usually in the size ranging from about 1 to 100μm in diameter; they settle slowly
under the influence of gravity. Figure 1 presents the particle sizing of some common
phenomena.
Figure 1. Sizes of Common Particles- (Wark & Warner, 1981)
10
Airborne particles generally have irregular shapes, and their aerodynamic behaviour is
expressed by the diameter of an idealised spherical particle known as “aerodynamic
diameter”. Hence, particles having the same aerodynamic diameter may have different
dimensions and shapes.
Depending on the size of dust particles, they are generally categorised as respirable dust
and total inhalable dust. Total inhalable dust is the fraction of airborne material which
enters the nose and mouth during breathing which might deposit in respiratory system.
The particle sizes of total inhalable dust are up to 100μm. Respirable dust is the fraction
that penetrates to the deep lung where gas exchange takes place. The particle sizes of
respirable dust are up to 10μm.
However, there are different views on type and threshold of respirable and inhalable
dust. Based on Australian Standard 2985, respirable dust is defined as “the proportion
of airborne particulate matter that penetrates to the unciliated airways when inhaled”
(Australian Standard, 2009). Alternatively, respirable dust can be defined as “a
cumulative log-normal distribution with a median equivalent aerodynamic diameter of
4.25μm and a geometric standard deviation of 1.5μm” as AS2985, see Figure 2.
100
90
80
70
Respirability (%)
60
50
40
30
20
10
0
0 1 2 3 4 5 6 7 8 10 12 14 16 18
Equivalent Aerodynamic Diameter (μm)
Figure 2. Respirable Dust as defined by AS 2985
11
A moderate number of health-related researches have been conducted on the impact of

a variety of bulk material dusts on employees health. As an example, Mannino et al
(Mannino, Buist, Petty, Enright, & Redd, 2003) in one of his researches highlighted that
exposure to Respirable Crystalline Silica (RCS) may cause loss of lung function even at
the safe work occupational exposure limit of 0.1 mg/m3. He also concluded that Chronic
Obstructive Pulmonary Disease (COPD) could occur due to exposure to RCS. Both RCS
and COPD the obstructive and restrictive lung disease respectively are significant
predictors of premature death.
2.2 Existing Dust Emission and Air Quality Standards
Compliance with the local current dust-emission standards is required for every
industrial site across the world. There are solutions available to reduce the dust
emissions, though the majority of those are categorised as ‘active’ systems. Active
systems that are tolerant to poorly understood physical systems with the potential to
generate a business revenue have been developed and marketed strongly.
There are a number of ambient-air quality standards available across the world. In this
section dust particle thresholds have been extracted from some of the key ambient-air
quality standards in order to define the dust emission compliance.
2.3 Australian Standard (Australia Department of

Sustainability, 2013)
The Australian state and territory governments have developed the National Air Quality
Standards. The Australian air quality standard specifies the high-level guideline as per
Table 1, below.
12
Table 1. Australian Standards for Ambient Air Quality
(Australian Department of Sustainability 1998)
Pollutant Concentration and Averaging Period
Particles as PM ∗ 50 averaged over a 24-hour period

³
Particles as PM 2.5 25 over a one day period; 8 over a one year period
³ ³
*Particulate matter is the general term used for a mixture of solid particles and liquid
droplets in the air. It includes aerosols, smoke, fumes, dust, ash and pollen.
Based on the above guidelines, local regulatory bodies establish allowable dust levels
regionally. For example, the National Environment Protection Measure (Air NEPM)
provides air quality standards across Australia. In the state of New South Wales (NSW),
however, the air quality is regulated by Air NEPM as well as the NSW Department of
Environment and Conservation (NSW DEC) based on average particle concentrations
across a region. The NSW DEC regulations are consistent with Air NEPM, however,
they are more comprehensive and can be applied to more specific areas such as those in
vicinity of mine sites. Table 2 summarises the dust concentration thresholds.
13
Table 2 Dust Concentration Limits for the State of NSW
Concentration
Pollutant Averaging Period Set By
(µg/m3)
TSP* Annual 90 NSW DEC
NSW DEC
1 Day (24 hr) 50
NEPM (allows
PM 10 1 Day (24 hr) 50 exceedance 5
times/year)
Annual 30
NSW DEC
NEPM (advisory
1 Day (24 hr) 25 standard)
PM 2.5
Annual 8 NEPM (advisory
standard)
Dust Annual (total) 4g/m2/month NSW DEC

Deposition Annual (increase) 2
2g/ m /month NSW DEC
 TSP refers to the Total Suspended Particulate matter, which is defined as the
total amount of all particles suspended in the air.
 PM2.5 refers to all particles suspended in the air, which are smaller in size
than 2.5μm
 PM10 refers to all particles suspended in the air, which are smaller in size
than 10µm, but larger in size than 2.5µm.
2.4 World Bank Guidelines (World Health Organization,

2005)
The World Bank (as a part of the World Health Organisation) developed guidelines for
ambient air quality to assist the World Bank Group in making decisions regarding air
pollution matters. The air quality thresholds specified by the World Bank standard are
presented in Table 3.
14
Table 3. World Bank Guidelines for Ambient Air Quality
Parameter Annual Arithmetic Maximum 24 hrs.

Mean Average
Particles as PM 10 20 50
³ ³
Particles as PM 2.5 10 25
³ ³
2.5 United States Standard (US Environmental Protection

Agency (EPA), 1997)
In the USA, the Occupational Safety and Health Administration (OSHA) and Mine
Safety and Health Administration (MSHA) are accountable for the occupational health
and safety regulations. The United States Environmental Protection Agency (EPA)
under authority of the Clean Air Act have established two classes of air quality standard.
The primary standard is defined to protect public health with an adequate margin of
safety given to the health of sensitive populations such as asthmatics, children and
elderly people. The secondary standard sets limits to protect the public from any known
or anticipated adverse effects of air pollutants, including protection against decreased
visibility, damage to animals, crops, vegetation and buildings. The National Ambient
Air Quality Standards (NAAQS) under the Clean Air Act are referred to in Table 4
(Final rule cited on 14-Dec-2012).
15
Table 4. United States Standards for Ambient Air Quality
Averaging
Pollutant Type Standard Form
Time
Primary 12 Annual annual mean, averaged over 3 years

³
Not to be exceeded more than once

Secondary 15 ³
Annual
PM 2.5 per year on average over 3 years
Primary &
35 ³
24-hour Annual mean, averaged over 3 years
Secondary
Primary &
PM 10 150 µg/m³ 24-hour Annual mean, averaged over 3 years
Secondary
There are a number of regulatory standards existing in other countries to regulate safe
dust levels in the atmosphere. The areas surrounding mining sites typically have more
strict controls in place, with routine testing undertaken by the regulatory authorities.
This is to ensure the compliance with safe dust levels designated by the standards. In
the case of any non-compliance, major penalties are applied to the offenders.
2.6 Measurement Equipment
Due to the complex nature of mineral operations coupled with variable dispersion and
dilution characteristics of airborne dust, predicting the specific level of dust emissions is
extremely difficult. As such, it is often important to be able to differentiate dust from
multiple sources (therefore necessitating monitoring for direction, colour and possibly
mineralogy) and to correlate this with site-specific meteorological data.
As monitoring is conducted for both health and nuisance purposes, the sophistication of
different monitoring techniques differs greatly. The aim and purpose of dust monitoring
need to be clear if it is to be effective. Consideration should be given to the location and
number of monitoring stations, the monitoring objective (i.e. to assess the potential for
16
nuisance effects) and the duration and frequency of monitoring. Monitoring methods
are categorised into active and passive systems.
Typically, active dust monitoring systems are more suited to measuring over minutes
and hours whereas passive systems are best suited for measuring over days weeks and
months.
Active monitoring systems are primarily used for occupational health and safety
concerns. They are not commonly used for monitoring of nuisance dust, as the
equipment is complex and expensive. In comparison with active methods, passive
systems are generally low technology with limited development and are applied in
monitoring of nuisance dust. Nuisance is a subjective parameter due to the lack of
standard guidelines for nuisance dust emissions. Nuisance is generally related to the
visual effects observed by receptors (such as the soiling of surfaces over days and
weeks) or the short-term visual effect of dust clouds. Hence, the passive method is
capable of monitoring industrial or mine sites for deposited particulate matter, which
rapidly settles from the air. They are an inexpensive solution for background dust
monitoring where a real-time dust management system is not required (This equipment
has to comply with the Australian Standard AS/NZS 3580.10.1:2003). Passive
monitoring measurement devices can be found in two categories: non-directional and
directional.
There are a number of non-directional passive measurement devices available on the

market. Some of the most widely used ones are Glass Slides, Frisbee Gauge (Figure 3. )
and Deposit Gauge (Figure 4).
Glass Slide Deposit Gauges compare the 'gloss' on a glass microscope slide before and
after exposure. 'Gloss' or surface reflectance is measured using a calibrated gloss meter.
Representative sample is formed by exposing the equipment in the desired location for
a number of days (usually seven days).
17
Figure 3. Frisbee Gauge
The Frisbee gauge consist of a frisbee shape collecting bowl, rainwater collecting bottle
and a stand. The frisbee shape collecting bowl should be supported with the opening 1.7
m above ground level. A drain pipe will also connect the stem down to a rainwater
collecting bottle on the ground. The capacity of the rainwater collecting bottle can vary
depending on the environment of test, however they usually vary between 5 to 10 litters
for collection periods of one month. The collecting bowl should be lined with a 10-mm
thick, 240-mm diameter, disc of black, open-celled (10 pores per inch) polyester foam.
The gauge also should be protected by a bird-guard. At the end of the period (usually 1
month) liquid collected in the rain water collecting bottle will be filtered through a glass
filter and residual weighted accurately which forms the basis of mean rate of dust
deposition calculation.
The Dust Deposition Gauge is one of the most widely used tools. It consists of a tripod
stand, bottle assembly (including two 5L glass bottles), glass funnel, silicone bung and
a bird ring.
The process of measurement is to separate insoluble deposited material from the liquid
by gentle vacuum filtration. The sample is then dried and determined gravimetrically.
The results are expressed as mg/m per day. The main known drawback with this device
2
is sample losses in high winds.
18
Figure 4. Deposit Gauge
Automatic Methods use the same principles as manual devices, with an instrument to
trigger the process automatically and collect the time stamped readings.
The main principle of a directional gauge is to collect dust in flux as per Figure 5. This
method has in the past been commonly used and remains the preferred method of the
environment agencies for routine directional dust monitoring despite serious accepted
limitations.
19
Figure 5. Directional Dust Gauge
The directional dust gauge as illustrated in Figure 5 is essentially a grouping of four

vertical tubes with slots facing four separate directions. Dust deposition can be
determined from obscuration of a beam of light in a water-filled cell, or by gravimetric
determination as with other dust fall gauges. Results from the two methods are not
interchangeable.
Similar to passive measurement systems, active systems are grouped into directional
and non-directional types. Directional active monitoring devices generally draw a
volume of air over a designated time period to measure the amount (particle
concentration and mass) and types of dust (particle size fraction) suspended in the air.
Active systems rely on an air pump with constant measurable flow rate to determine the
volume of air sampled. The device samples ambient air for a fixed period with
particulate matter captured on a filter for subsequent analysis. Automatic methods use
the same principles along with an instrument to trigger the process automatically and
collect the time stamped reading.
20
2.7 Dust Control Methods
Dust control is a major issue in many bulk-material handling plants for several reasons.
Increasingly stringent Environmental and Occupational Health and Safety regulations
(Australian Standard, 2012) and (British Standard, 1972) demand significantly lower
acceptable dust levels are common across all industries. Furthermore, costs incurred
through spillage clean ups to improve workplace safety continue to increase as
acceptable levels reduce. For example, it has been estimated that the annual cost to
mining industry due to dust, mess and spillages from materials handling operations was
US$300 million in 1987 in the UK(Wilkinson, 1989). In addition, the increased
equipment costs associated with improving sealing efficiency e.g. enhancing the ingress
protection (IP rating) and hazardous area rating of the electrical equipment are another
strong stimulant for industry to effectively control dust emissions. By establishing an
appropriate dust control system, not only health, safety and the environment can benefit
from reduced dust emissions, but also equipment wear and maintenance costs can be
reduced. This means that effective dust control can result in cost savings, increased
profits and productivity and at the same time improved workers health, safety and
community relations.
Dust control techniques can be implemented in three different aspects: Prevention,

Control and Dilution (or Isolation) detailed as follows.
Prevention: Implemented within the design phase of the bulk material handling system.
Control: Relies on additional equipment or techniques to restrict the dust emission;

Equipment such as dust collection system, wet dust suppression system and airborne
dust capturing through water sprays are examples of dust control.
Dilution or Isolation: Achieves the control by diluting the contaminated air with
uncontaminated fresh air to reduce dust concentration, or isolating the workers from
harmful dust (Kissell, 2003).
21
In terms of dust management, the goal is to prevent rather than control or isolate. The
ramifications of retrofitting dust control measures after poor initial design and
fabrication is often difficult, impractical and costly. However, despite these drawbacks,
due to lack of sufficient study in design section, Control is often the main employed
method to reduce dust emissions.
Dust control systems could be further classified into two main categories: active and
passive. Active dust control systems are one of the most commonly employed
techniques to control and minimise fugitive dust from bulk material handling systems.
Active systems typically utilise two main techniques, namely suppression and filtration.
The suppression technique relies on a variety of water and chemical materials to

decrease the dust emissions, such as the application of chemical surfactants, water or
foam spray systems and water misting systems.
The filtration technique use mechanical devices, such as electrostatic precipitators, bag
houses and fan extraction, to reduce dust emissions. Although these techniques are
widely utilised, many limitations exist with these types of systems, which are detailed
in this chapter.
Dust suppression can be employed either in an open or closed bulk material transfer
system. The principle of these systems is based on wetting the bulk material with a
surfactant such as foam or water which results in increasing water meniscus forces.
With an increase in meniscus attraction, fine particles are less likely to be entrained by
local turbulence in the air boundary layer which prevents them from being further
suspended in the air. Although these dust suppression systems are effective, they have
major disadvantages.
A high level of water consumption is one of the significant drawbacks of this system.
Typical consumption rates have been estimated at about 10 to 50 litres per tonne
(Swinderman, Goldbeck, & Marti, 2002). These figures are confirmed by a coal mine
22
in Hunter Valley region in Australia which uses approximately 40 litres per tonne (X.
Chen, 2013). Given that the total production of raw black coal in Australia in 2011 was
468 million tonnes (Australian Coal Association, 2012), this equates to more than 18
billion litres of potable water per annum consumed in this industry alone.
The other issue is the sensitivity of these systems to cold temperatures, which requires
heating or insulation of surfactants to keep them unfrozen. The third issue is that these
systems often require high installation and maintenance costs.
Filtration techniques are generally effective only in a closed bulk material transfer
system. This technique uses an air mover (fan) to create a negative air pressure in a duct
which leads to the filtration system. In the filtration system, the air passes through a
filter, which removes the majority of the collected dust particles. However, filtration
systems are expensive to install and maintain. To remove the dislodged dust particles
from the filter, a shaker or other device is required, which contributes to the higher
manufacturing and maintenance cost. Moreover, the required high energy consumption
to create a region of negative air pressure in the ducts, is the other major disadvantage
of this method.
While active systems are reasonably effective, their main drawback is the ongoing cost
associated with operation and maintenance. (ACGIH, 1992), (Junemann & Holzhauer,
1992), (Tooker, 1985), (Weiss, 1992), (Herrmann and Evensen, 1994). Although active
dust control is the most widely used technique to manage and reduce fugitive dust, and
has the potential to resolve or significantly reduce many dust-related problems, there is
an ongoing need for an improved understanding of the interaction between bulk
materials and air in handling operations. This understanding provides additional refined
design inputs, which lead to a more effective and energy-efficient dust control system
design(Z. Liu, P. Cooper, & P. W. Wypych, 2007).
23
As an alternative solution, passive dust control is attracting more attention within

industry due to its advantages over active dust control systems. Passive dust control can
be used as a complementary technology, or in some cases, a viable alternative to active
dust control systems.
Passive dust control systems are designed to require minimal maintenance, no energy
consumption, no moving parts and no water, mist or other suppression agents. Passive
systems reduce the dust emission by preventing dust generation rather than reducing it
after it occurs. Passive dust control relies on containing or re-entraining the dust within
the moving bulk material stream which requires a thorough understanding of the
interaction between the bulk material and the surrounding air.
The aim of this research is to review the air entrainment process and identify major
influencing design parameters in discrete bulk material handling processes with the
objective of improving the dust emission liberation. Given the large number of variables,
such as bulk solid properties, throughput and equipment geometry, a rigorous scientific
approach must be undertaken to arrive at a suitable configuration and geometry for each
installation.
Air entrained into a falling stream of bulk material under the influence of gravity is
known as the main source of dust generation (Ansart, Ryck, et al., 2009b). Therefore,
this research was initiated on free-falling streams, to better identify the impact of key
influencing variables such as particle diameter, particle-solid density, particle shape,
stream diameter and drop height.
The findings from this free-falling analysis can be extended to the transient bulk material
handling cases. As dump hoppers are known to be a major source of the dust generation
within the mining industry (Cecala, 2010) this equipment is selected to further expand
on.
24
Fundamental techniques such as theoretical and CFD computer simulations models

verified by scale model testing, PIV (Particle Imaging Velocimetry), high speed footage
have been employed throughout this study.
25
Free-Falling Streams
Chapter 3: Free-Falling Streams
3.1 Background
Dust generation physics within the free-falling process of particles can be categorised
into the following two groups:
I. Liberation of dust from granular material during the free-falling process

II. Falling stream impacting a bed of material resulting in release of entrained air
and dust to the atmosphere (Ansart, Ryck, et al., 2009b)
A natural dilution takes place in the free-falling stream during the free fall process of
particulate materials. As the process of the free-falling initiates and continues, the
surrounding air entrains into the stream of bulk material, hence increasing the voidage
ratio.
The air-entrainment process is a complex phenomenon. From the physics of the free-
falling streams and air-entrainment perspective, a couple of scenarios could be looked
at as the reasons behind the air entrainment. One scenario is that the air entrainment
takes place due to particle dilution during the freefalling under gravity. A stream-wide
dilution gradually forms as the particles accelerate to their terminal velocity. Another
scenario would be that the dilution of the stream occurs due to the air entrainment. Once
the air starts to infiltrate, a higher surface drag on the particles at the periphery of the
falling stream forms a transverse speed profile.
Ansart (Ansart, Letourneau, de Ryck, & Dodds, 2011) reported that the vertical particle
velocity across the stream is Gaussian with the maximum being at the centre of the
stream. Given that finer particles have lower velocity (or terminal velocity), this, to some
extent, leads to the finer particles migrating to the outer skirt of the stream. The zone of
influence is small, but this might be some of the lateral dilution. The following figure
shows the schematic view of the free-falling stream and air entrainment principle.
26
Figure 6. Free-Falling Air Entrainment Schematic
Relatively little fundamental research has been carried out with regards to the fluid and
particle mechanics of “particle plumes” which have some similarities with other plume
and jet flows such as bubble plumes and water sprays (Ghosh, Hunt, & Phillips, 1993),
(R. Ansart, A. Ryck, & J. A. Dodds, 2009a).
In terms of air entrainment, several methods have been published to calculate the amount
of entrained air within a free-falling bulk material stream, however the accuracy of these
different methods varies considerably, and is heavily dependent on particle diameter,
solids density, stream diameter and drop height. Moreover, the majority of the methods
are for fine particles.
The first comprehensive work in this area was undertaken by Hemeon (Hemeon, 1955).
He analysed the induced air based on the work performed by a single particle. The
amount of air induced into the stream of free-falling granular material is given by:
27
0.66 g m (h A) (1)
Q =
d ρ
Where m is mass flow rate of bulk material, h is drop height of material, A is cross-
sectional area of falling stream of material, d is particle diameter, g is gravitational
constant and ρ is the solids density of the particle. The specific parameter of concern
in this study is the cross-sectional area of the falling stream (A) or the diameter of the
stream.
There is limited published research on the diameter of a falling stream. Ansart et al.
(Ansart, de Ryck, et al., 2009) investigated the diameter of a falling stream by employing
particle image velocimetry (PIV) and concluded that by decreasing the particle size, the
plume diameter increases. However, this work has only considered fine particles from
34 to 97 µm. In a more recent study, Ansart et al.(Ansart, de Ryck, et al., 2009) studied
the impact of drop height on the diameter and intensity of a fine particle stream by
utilising an image-analysis technique. This work concluded that by increasing the drop
height, the stream diameter increases, which resulted in the stream core intensity
decrease, with this finding being reproduced in the work of Esmaili et al. (Esmaili et al.,
2013). Moreover, a coarse-particle stream-diameter study was conducted by Esmaili et
al. in which a method for free-falling stream-diameter measurement using a high-speed
camera along with an image analysis technique in MATLAB software was proposed.
However, this work indicated that for a certain range of particles and sizes this shape is
closer to a cone.
‘Stream diameter’ is a terminology, which has been unclear until to date. To quantify
the stream diameter which was also referred to in Hemeon’s work, the author has trialled
three different approaches and suggested the most accurate and practical way. The
28
selected methodology will be used in the future chapters for air-entrainment volumetric
flow rate calculations.
3.2 Experimental Work on Stream Diameter
In order to determine the stream diameter in an efficient and reproducible way, a number
of experimental arrangements have been investigated, one of which is shown in Figure
7.
Figure 7. Stream Diameter Test Rig - Concept 1
As part of this experimental work, a test rig was designed and manufactured initially by
utilising a range of commonly available material as a proof-of-concept. The rig
comprised of four concentric close-ended pipes (just one covered end in each pipe) fitted
into each other from smallest to largest diameter. The uncovered ends of the pipes are
located under the stream, facing toward it, so that particles can enter into the annulus
formed in between the pipes. The thickness of the selected pipes were 3mm.The landing
location of particles within the annulus represents the free-falling stream diameter.
29
Additionally, by undertaking particle-size-distribution analysis (PSD), the granularity

of particles in each section of the stream was determined.
To conduct the test, the test rig was placed under the stream and in line with the centre
of the hopper outlet. For each round of experiments, the location of the test rig was
adjusted in two perpendicular directions, by employing a laser alignment tool to ensure
the hopper and test rig were concentric, and only then a free-falling stream of material
was released. Subsequent measurement of the weight and PSD (Particle Size
Distribution) of the particles landed in different pipes was conducted. The key advantage
of this methodology is that it could provide the cross-sectional particle size distribution
(PSD) of a free-falling stream that is shown in Figure 8.
Coal Sample
120.0
100.0
Particles Weight (gr)
80.0
60.0
40.0
20.0
0.0
40mm 80mm 100mm 150mm
5.6 to 4mm 111.0 56.0 1.5 1.0
4 to 2.8mm 84.7 53.3 2.1 0.8
2.8 to 2 mm 103.9 65.4 3.0 1.3
2 to 1 mm 83.2 59.9 4.4 2.1
smaller than 1mm 21.2 16.4 2.0 1.4
Figure 8. Coal Measured Weight (gm)
The above results indicate that the concentration of particles reduces from the centre
towards the outer layers of the stream. It also can be concluded that the population of
larger particles decreases as the distance from the centre increases.
30
Figure 9 represents the percentage of each particle-size range in each slot (the maximum
particle percentage in each slot is considered 100%). The presented results are the
average of two sets of tests.
Coal Sample
40.0%
35.0%
Particles Weight (%)
30.0%
25.0%
20.0%
15.0%
10.0%
5.0%
0.0%
40mm 80mm 100mm 150mm
5.6 to 4mm 27.5% 22.3% 11.5% 14.8%
4 to 2.8mm 21.0% 21.2% 16.2% 12.6%
2.8 to 2 mm 25.7% 26.1% 23.1% 20.1%
2 to 1 mm 20.6% 23.9% 33.8% 30.9%
smaller than 1mm 5.2% 6.5% 15.4% 21.5%
Figure 9. Coal Percentage in each Annulus
Furthermore, it can be seen that as the stream expands, the finer-particle population
increases towards the outer layers of the stream. Although weight of the particles could
be the best indicator of the particle size distribution, the quantity of the particles from
each size group (Figure 8.) in each annulus can play a role in this analysis as far as
probability of occurrence is concerned. The reason being is that the weight of one large
particle is equivalent to a large number of fine particles; therefore, if one large particle
lands in the outer, it could influence the results considerably. Hence, the quantity of
particles could be taken into account for further advanced analysis.
One of the main observations in design and performance of the concept model was the
wall thickness of the pipes used for annulus. By reviewing the high-speed camera
footage from the experiments, it became evident that some particles hit the top edge of
31
the pipes in annulus, bounced and eventually landed in a random slot. Therefore, the
wall thickness of the pipes needs to be as thin as possible. To minimise this issue in the
concept model, the edges of the utilised pipes were sharpened. However, a better
solution would be to utilise pipes with lower schedule (thinner walls). After making the
above modification, there was no material falling outside of the 150mm.
Moreover, the limited volume of the central pipe and tendency of the particles to stay at
the centre of the stream imposed a timing limitation on the experiment, which made it
harder to obtain the results from a fully developed stream.
In addition, after reviewing the stream behaviour using recorded footage from a high-
speed camera (up to 1000 fps – frame per second) as well as a conventional 30 fps
camera, a phenomenon was observed which will be henceforth referred to as the ‘S’
shape of the stream. This phenomenon will be discussed in more details in this chapter.
However, due to the stream not following a straight line and taking a ‘S’ shape, the
experiment requires a longer period in general, in order to collect better statistically
averaged data, and this was identified as one of the major limitations of this
methodology.
Although the test equipment developed for this experiment was a concept model, it
provided a good insight into particle-size distribution across the stream. According to
the obtained results, once the flow is fully developed, larger particles tend to stay at the
centre of the stream and finer particles are guided toward the outer layer of the stream.
Hence, the finer particles are more likely to become airborne.
It also can be concluded that the core of the stream has a relatively higher velocity due
to the higher concentration of heavier (and larger) particles and less drag effect which
is in agreement with Ogata’s finding.(K. Ogata & Tomita, 2004).
This experiment provided a good insight into the behaviour and physics of bulk material
free-falling, which could be used in developing a more advanced test rig, however, the
32
author decided to develop a more comprehensive methodology that will be discussed

later in this chapter (page37).
The second major set of experiments comprised of a high-speed camera along with a
background grid, as per Figure 10. The main advantage of this method in comparison
with the previous method (annulus experiment) was that the stream diameter could be
measured from a fully developed stream of material; however, this method cannot
provide the particle size distribution.
Although this method was easy to set up, the stream diameter reading was hard to
determine due to the nature of the stream and the number of photos needed to be
analysed. The ‘S’’ing phenomenon of the stream was observed, indicating that the free-
falling stream of granular material did not follow a straight line.The particles appeared
to take up a ‘S’ shape free flow with some clear horizontal velocity that varied along the
drop height such that when viewed the steam of particles always appeared as a long
drawn out S shape. This finding made the manual photo review much harder. Moreover,
this phenomenon has been witnessed at industrial site with various discharge
mechanisms. Some pictures from this observation are presented in Figure 10. Whilst the
observation is fascinating and research worthy in itself, this is not the focus of this thesis
and is cited for completeness only. However, this method formed the basis of the
technique, which will be expanded upon within the next section.
33
Figure 10. Stream 'S''ing Phenomenon
The materials employed for this study were uniformly shaped granular materials, most of which
were spherical to minimise the variables. A summary of the materials used can be seen in Table
5.
34
Table 5. Summary of Materials
Material Particle diameter (mm) Particle density (kg/m3)

Plastic Pellets 3.0 870
Sand 0.3 2600
Glass Beads 2.0 2450
For each of the above materials, a high-speed camera was used to record the falling
stream of particles over the range of drop heights studied. Some coloured particles were
also inserted into the material, to ease particle trajectory tracking as well as particle
velocity measurements, presented within
Figure 11. For particle tracking and velocity measurement purposes, a proprietary high-
speed camera software was employed.
Figure 11. Coloured Particles - Particle Trajectory Tracking
35
The variation in drop heights for each material was from 300 to 1500 mm, with
increments of 300 mm. The general layout of the experiment can be seen below in Figure
12.
Figure 12. Layout of Experimental Test Equipment
A high-performance light was used to illuminate the particles as they passed through the
field of view of the camera. In high-speed filming, having sufficient light is vital to
obtain appropriate results. Also shown above is a black background that was used to
assist in distinguishing the particles from the background (in the series of images). For
each material and each drop height, two 4 sec videos of the flow were captured once the
falling stream had fully developed. The frame rate used to capture the falling stream was
set at 2000 frames per second.
36
3.3 Analysis of Velocity and Stream Diameter Method
A detailed analysis of the digital footage was performed using MATLAB software
programming. The presented information in this section are reflecting 8 seconds of high-
speed camera captured footage, which corresponded to a total of 16000 individual fames
for every individual test.
For the image processing analysis, a MATLAB program was developed. The colour
thresholding function in MATLAB allows users to threshold colour images by
manipulating the colours and contrasts. Most of the thresholding procedures involved
setting of boundaries based on grey values or intensities of image pixels (Raof et al.,
2008).
For each image analysed from the high-speed camera, thresholding of the colour value
was necessary to differentiate between the particles and the. An example of this can be
seen in Figure 13 (left hand) which represents photo from actual experiment by high-
speed camera and Figure 13 (right hand) is the outcome of thresholding process.
Figure 13. Example of Thresholding Used in Image Analysis (plastic pellets)
37
Once this thresholding had taken place, a horizontal slice (30 mm in vertical height) was
taken across each image in order to collect data to determine the stream diameter. For
every pixel that existed in the horizontal slice, data was stored on whether or not each
pixel contained a particle at any given time. By performing this analysis over a series of
images, a distribution was built up which reflected the relative likelihood of each pixel
to contain a particle. From this distribution, the information could then be extrapolated
to infer a stream diameter of the falling particles. An example of the distribution can be
seen in Figure 14.
Figure 14. Example Distribution of Stream Diameter
Each distribution (typically as shown in Figure 14), most closely resembled a normal
distribution. Once the mean and standard deviation were found, the stream diameter of
the falling material could then be defined. For a typical normal distribution, the area
under the curve sums to 100% from the limits of the mean ± 4 standard deviations, see
Figure 15. However, for a more realistic approximation of the stream diameter, the mean
± 2 standard deviations were used (encompassing ~95% of the area). This assumption
was taken with a view to air induced by a free-falling stream of material, as it is the
38
opinion of the author that increasing the stream diameter to 3 or 4 stream diameters will
overestimate the diameter of the stream causing induced air.
Figure 15. Area Under Curve Given by Standard Deviations
3.4 Theoretical Terminal Velocity Calculation of Gravity

Falling Particles under Drag Influence
The motion of a free-falling particle without air resistance can be calculated by the
following equation:
(2)
= +2 ℎ
If drag is taken into account, without considering the buoyancy effects, the terminal
velocity is given by:
39
2 (3)
∞ =
In this equation, the key parameter of discussion is -drag coefficient. There are a
number of theories available for calculating the drag coefficient. The author is aware of
the fact that this topic is substantial, however a brief and basic explanation is provided,
as this information is required for the current and upcoming chapters.
Drag coefficient is defined by the following formula (Lapple & Shepherd, 1940):
d Particle diameter [m]

2 (4)
=
For spherical particles, Stokes has provided following theory, which is known as the
key theory in drag coefficient research topic:
= For streamline flow (Re < 1) (5)
. (6)
= .
For intermediate flow (2 < Re < 500)
= 0.44 For turbulent flow (Re > 500) (7)
40
Stokes comprehensive theory will be utilised within this chapter as a standard drag
coefficient model.
Behaviour of fine particles in a free-falling stream of bulk material is such that they
reach their terminal velocity considerably earlier than heavier ones, due to their lower
terminal velocity; hence the fine particle are more likely to become airborne (dust).
Figure 16 represents the terminal velocity of a spherical particle under drag influence.
Figure 16. Terminal Velocity of Particles with Spherical Shape
3.5 Results and Discussions
Figure 17 shows the theoretical versus experimental velocities of the tested material
along the vertical displacement. The experimental velocity measurement takes place at
300mm increments. The theoretical figures are inclusive of air drag force; particles are
considered single spherical with drag coefficient 0.44.
41
PP : Plastic Pellets
GB : Glass Beads
*experimental values shown as data points while the theoretical values are represented by solid/dashed lines
Figure 17. Velocity of Different Type and Size Particles versus Vertical Displacement
The results in Figure 17 show that the difference between the measured velocities and
the theoretical velocities are within a close range for all materials except sand. The
experimental velocities are reflective of the velocity of outer layer particle of the stream.
However, measuring the velocity at the core of the stream is unachievable with high-
speed camera.
Ogata et al (K. Ogata & Tomita, 2004) have measured the velocity of glass beads (fine
particles and dense flow) with mean diameter of 454 µm under the influence of gravity,
from a circular orifice. He observed that the stream travelled faster than a single particle.
In the case of sands, the experimental velocities are higher than the predicted velocities,
and the reason for this is that the effective air drag of the falling stream is lower than the
one of a single particle, due to the high concentration of particles.
42
The results for the stream diameters illustrated in Figure 18 show that for the materials
with smaller diameter particles (sand and the 2mm glass beads), the stream diameter is
relatively the same along the vertical displacement regardless of the drop height. This
translates directly to a lack of lateral dispersion as the material undergoes free-fall. In
contrast to this, the larger particles of glass beads show considerable dispersion. This
also could be attributed to the transient blockage of the outlet due to rise in the particle
size with constant hopper outlet diameter.
Figure 18. Stream Diameter versus Drop Height
To investigate the impact of hopper outlet(investigation of transient blockage of hopper

outlet) on stream diameter, the same tests as carried out above were repeated for the
6mm glass beads and the sand using a hopper with a 52mm outlet (the diameter of the
hopper outlet used in the previous tests was 26mm). As Figure 19 shows, the results for
the stream diameter using the 52mm outlet diameter hopper are very similar to the
results for the stream diameter using the 26mm outlet diameter hopper. For both
materials, the initial stream diameter for 52mm hopper outlet is slightly higher than
43
26mm outlet; however as the free-falling undergoes the results for the stream diameter
converge to the results for the 26mm outlet diameter hopper.
In addition to these results, a velocity analysis on the results reveals the velocities for
the testing carried out on the 52mm opening hopper(with identical hopper arrangement
and material) are similar to those presented in Figure 17, while the mass flow rates are
different meaning that the solids ratio is higher for the results of the 52mm hopper. This
increase in solids ratio is due to the fact that the stream diameter and velocity have
remained the same while the mass flow rate of the material out of the hopper has
increased (due to the larger opening size).
This result indicates that the lateral dispersion of the stream, or the stream diameter at
any point during freefall, is more likely to be a function of the properties of the material
(such as particle diameter, particle density and particle shape) rather than the geometry
of the hopper (for the range of experiments considered in this study).
Figure 19. Comparison of stream diameters for tests carried out with 26 and 52mm
hoppers
44
The other key finding observed in this research work was “Stream density”. Knowing
the mass flow rate from the hopper and the stream diameter (as calculated in the first
section of this chapter), the density of the stream can be calculated from the following
formula:
Q = ρAV (8)
The results of this analysis are shown in Figure 20, indicating that the finer particles
have denser stream. Moreover, the results show that the particles of the same size have
a similar stream density under the same test conditions regardless of their solid density.
900
800
Stream Density (Kg/m3)
700
600
500
400
300
200
100
0
0 200 400 600 800 1000 1200
Drop Height (mm)
3mm Glass Bead Sand 3mm Plastic Pellet
Figure 20. Measured Density of the Stream versus Vertical Distance
3.6 Conclusion
In this chapter, it has been demonstrated that an optical method with the aid of
MATLAB software has been developed and tested for the determination of stream
45
diameter. This method relies on an image-analysis method utilising a high-speed camera

in conjunction with MATLAB software.
The diameter of the falling stream of particles was seen to vary amongst a range of
materials of varying densities. Results for the velocity of the stream of particles were
also presented, along with a comparison to the theoretical results, which were
reasonably close. These experimental results for the diameter and velocities could then
be combined to calculate the solids ratio of the falling stream. The solids ratio revealed
that the stream density of the sand was much higher than other materials, which led to a
possible explanation for the discrepancy between theory and experiment for the sand.
A further study was conducted to investigate the effect of the hopper outlet diameter on
the stream diameter. For the two investigated materials sand and 6 mm glass beads, the
stream diameter was largely unaffected by the size of the hopper outlet. This indicates
that the stream diameter at any point along the free-fall is more dependent on the
properties of the material, such as diameter, shape and density of the particles rather
than the size of the feeding hopper outlet.
It has also been observed that a falling stream of larger particles will have a lower
particle concentration than an identical material with smaller particles. Additionally,
increasing the particle density will result in an increase in particle concentration at the
stream centre.
46
Free-Falling Air Entrainment Measurement and Calculation
Chapter 4: Free-Falling Air Entrainment Measurement and

Calculation
Many bulk-material handling operations are being employed in different industrial

applications on a daily basis, most containing a variation of free-falling (Liu, 2003). As
described in previous chapter, during the free-falling process the air surrounding a free-
falling stream of bulk material is entrained into the stream and upon subsequent
compaction of the stream, it will be exhausted into the atmosphere (along with dust
particles). Hence, a better understanding of the air-entrainment mechanics and
magnitude will assist in the design of a more efficient dust control system (Z. Q. Liu et
al., 2007). This chapter reports on a research study being undertaken to analyse passive
dust control technology for use within material handling systems, with a focus on free-
falling.
This chapter reviews the performance of the existing models for mass flow rate
calculation of the entrained air within a stream of free-falling granular material. The
predicted results are compared to the existing theories and other experimental data.
Several theories exist on predicting the air induced into a stream of free-falling granular
material, amongst which the work by Hemeon (1955) is the most comprehensive. Based
on the methodology developed for stream diameter measurement within Chapter 3:, the
accuracy of different methods is compared and discussed. In addition, a new theory is
developed and compared with the existing theories as well as experimental results.
4.1 Background
The most robust structured research in relation to the calculation of air-entrainment

flow rate into the stream of free-falling granular material was carried out by Hemeon
(Hemeon, 1955), in which, the mass flow rate is calculated as
47
(9)
0.66 g m (hA)
Q =
d ρ
Where m is mass flow rate of bulk material, h is drop height of material, A is cross-
sectional area of falling stream of material, d is particle diameter, g is gravitational
constant and ρ is the solids density of the particle.
Morrison (Morrison, 1971) modified Hemeon’s formula and developed an empirical

equation without an analytical derivation. He introduced the modified equation for
induced air as follows:
(10)
0.49 g m (hA)
Q =
d ρ
Where d is the average particle diameter. It was proven by field experiment that this
formula is more accurate than Hemeon’s formula.
The work by Marc et al. (Marc et al., 1995) is the only available paper in the literature
concentrating on dust generation rate in details. Based on the hypothesis of the
interaction between separation forces and binding forces that are critical to the dust-
generation process, he developed a practical method to predict dust generation as a
function of falling material flow rate in terms of parameters, including drop height, mass
flow rate, particle size and moisture content. He predicted the dust-generation
magnitude by material size distribution, material cohesion and impaction arising from
material handling. Impaction refers the rapid consolidation of a dilated stream of bulk
material often resultant from a rapid change in the velocity vector. He concluded that
48
size distribution was not as important as either cohesion or impaction. The resultant
relationship is:
. ± . . ± .
( ). ± . (11)
=
( ℎ ) . ± .
The exponent of the equation shows that cohesion has a stronger influence on dust
generation than impaction.
Further research in this area was conducted by Plinke (Plinke, Leith, Holstein, &
Boundy, 1991) who carried out experiments using fine particles, including cement and
sand. These experiments confirmed that the amount of the entrained air increases with
drop height, which is in general in agreement with Hemeon’s work.
While Hemeon’s model was the first robust structured published work in this area, it is
considered to grossly overestimate the amount of the entrained air (Ansart et al., 2009).
One of the main issues that Hemeon’s model fails to address is the difference between
the free-fall velocity of a stream compared to a single particle. Ogata et al. (2001)
measured the velocity of glass particles (mean diameter of 454µm) falling from a
circular orifice, and found that the velocity of the free-falling stream of particles was
higher than the velocity of a single free-falling particle.
Cooper and Arnold (1995) published their theory based on two extreme cases for fine
and coarse particle streams. In the first extreme case, the stream is made up of coarse
particles. The main concept of the formula is based on volume conservation theory.
Cooper and Arnold’s first theory corresponds to the “massive particle model”, which is
shown in Equation (12):
49
(12)
Q = ((V + 2gh) − V )A
Where V and A are the initial velocity and area of the stream respectively.
Gravitational acceleration is g and h represents the drop height of material. One of the
advantages of this formula is the absence of any empirical parameters. However, the
particular type of material can only influence the result via the initial velocity term. For
the second case where the stream is made up of fine particles, plume theory is considered
and is given by:
mg (13)
Q =C h
ρ
Where ρ is the particle density, m is mass flow rate of the particles, h is drop height
and C is entrainment constant, which is usually 0.1 for a miscible plume. However, this
constant is not known for different testing conditions and it could be considered as one
of the drawbacks of this theory. In addition, Cooper and Arnold (1995) reported that the
stream core cross-sectional area decreases with increasing drop height and the
contiguous outer layer cross-sectional area increases with increasing drop height. In
addition, increase in the drop height will result in increase in volume of entrained air.
Ogata et al. (2001) suggested a new theory for air entrainment calculation, given by:
0.147 m h . A (14)
Q =
.
ρ d
50
Where ρ and d are the particle density and particle diameter respectively, m is mass
flow rate of the particles, h is drop height and A is the cross-sectional area of the stream.
Equation (14) can be used to calculate the amount of air entrained into the stream of
material, for Reynolds numbers less than 500. Uchiyama (2004) compared simulation
results with both Equations (12) and (14) over a range of particle sizes from 300-500μm.
He concluded that the results agree well with that predicted by Equation (12).
Within the free-falling process, the adjacent air surrounding a free-falling stream of bulk
material is induced into the stream and upon compaction of the stream may be exhausted
into the atmosphere (along with dust particles). Hence, by analysing the way air is
induced into the stream of bulk material and the quantity of the entrained air, a more
efficient dust control system can be designed and developed. (Z. Q. Liu et al., 2007).
The air-entrainment pattern of a free-falling stream of bulk material can be influenced

by bulk material properties, such as particle size and density, as well as operating
conditions such as drop height, particle mass flow rate and the other environmental
factors (Cooper & Arnold, 1995b; Wypych, Cook, & Cooper, 2005). Because of these
observations, it has been acknowledged that there are other parameters, which can
influence the dust-generation pattern of bulk-material streams. These parameters are
categorised under the ‘passive’ dust control topic in which no external energy sources
are used. Passive dust control techniques have received substantial attention, as either a
complementary option to improve the efficiency of active systems, or to be used as a
viable alternative (X. L. Chen, Wheeler, McClean, Donohue, & Roberts, 2012)
As discussed in Chapter 3:, coarse-particle stream diameter needs to be understood for

further study in terms of air entrainment, as some of the existing theories are dependent
on this parameter. Furthermore, a detailed study was conducted by the author (Esmaili
et al., 2013) in which a method for free-falling stream diameter measurement using a
high-speed camera, along with a proposed image analysis technique in MATLAB.
51
This chapter has presented a range of studies conducted in terms of the free-falling of
coarse particles with a focus on velocity of air and solid particles as well as air
entrainment. The existing researches are limited to applications involving relatively fine
particles, with little research undertaken on particle diameters greater than 2mm. The
focus of this research is to investigate the amount of entrained air within a free-falling
stream of granular material consisting of particles of diameter greater than 2mm.
4.2 Particle Motion under Gravity
The motion of a free-falling particle is investigated under two conditions: with and
without air resistance. Assuming that the air resistance is negligible, the velocity of free-
falling material can be calculated by equation:
(15)
= +2 ℎ
If the air resistance is taken into account, the following equation presents the relation
between velocity of free-falling material and drop height:
1− (16)
∞ ∞ −
ℎ= ( − ∞)
1−
∞
The terminal velocity ∞ in Equation (16) has been investigated in details in the
literature, for spherical particles, however, to a lesser extent for non-spherical particles.
Without considering the buoyancy effects, the terminal velocity is given by:
52
2 (17)
∞ =
For spherical particles, when the buoyancy effects are taken into account, the terminal
velocity ∞ can be expressed as follows:
(18)
4 −
∞ =
3
Turton and Clark (Turton & Clark, 1987) introduced an explicit relationship where they
defined a dimensionless particle diameter ∗ and a dimensionless terminal velocity ∗
for spherical particles as follows:
(19)
−
∗ =
. (20)
1
∗ = . .
18 0.321
+
∗ ∗
Where, d is the diameter of sphere, is the density of fluid (air), is the density of
sphere, is the dynamic viscosity of the fluid, ∗ is the terminal velocity and ∗ is the
dimeter of the dimensionless sphere.
53
For the non-spherical particles, the particle sphericity factor ψ is introduced to take into
account the shape effect, shown by the following equation:
6 (21)
=
Where, is the projected area of the particle.
Based on the work of Pettyjohn and Christiansen (Pettyjohn & Christiansen, 1948),
Geldart (Geldart, 1990) proposed two distinct equations incorporating the particle
sphericity factor to determine the terminal velocity according to Reynolds Number:
(22)
∞ = 0.843 ∞ < 0.2
.
(23)
∞ = ( . )
1000 < ∞ < 3×10
.
Given that majority of the existing terminal velocity calculation models rely on drag
coefficient of particles, it is important to utilise an applicable and accurate drag
coefficient model. Therefore, a brief review of drag coefficient calculation models is
provided in this section.
There are a number of theories available for drag coefficient estimation, in which some
of the flow and solids properties are utilised, i.e. flow direction, object position, object
54
size, fluid density and fluid viscosity. In summary, drag coefficient is a function of
Reynolds Number (Re).
The first model for drag coefficient (C ) calculation was introduced by Schiller and
Naumann (Schiller & Naumann, 1935) which is presented in Equations (24) and (25) as
below:
24 . (24)
C = 1 + 0.15 Re Re < 1000
Re
C = 0.44 Re > 1000 (25)
In which Re represents Reynolds Number for particle. This model is mostly known as
a standard drag-coefficient calculation model and will be utilised for drag calculation
purposes within this chapter. Clift (Clift & Gauvin, 1970) refined this model based on
different ranges of particle Reynolds Number were recommended by:
24 3 (26)
Re < 0.01 C = +
Re 16
C Re (27)
0.01 < Re < 20 log −1
24
= −0.881 + 0.82w − 0.05
24 . . (28)
C = 1 + 0.315 Re
Re
55
C Re (29)
20 < Re < 260 log − 1 = −0.7133 + 0.6305w
24
24 . (30)
C = 1 + 0.1935 Re
Re
260 < Re < 1500 log C (31)

= 1.6435 − 1.1242w + 0.15585
1.5 x 10 < Re < 1.2 x 10 (32)
log C = −2.4571 + 2.5558w − 0.9295 + 0.1049
1.2 x 10 < Re < 4.4 x 10 (33)
log C = −1.9181 + 0.6370w − 0.0636
4.4 x 10 < Re < 3.38 x 10 (34)
log C = −4.3390 + 1.5809w − 0.1546
3.38 x 10 < Re < 4 x 10 C = 29.78 − 5.3w (35)
56
4 x 10 < Re < 10 C = 0.1w − 0.49 (36)
8 x 10 (37)
10 < Re C = 0.19 −
Re
In comparison with other drag models, Clift’s model was found to most closely match
the experimental data. However, that correlation consists of seven different equations in
the Reynolds Number region studies, which makes it inconvenient (Brown & Lawler,
2003)
Turton and Levenspiel (Turton & Levenspiel, 1986) developed an equation based on the
Clift et al work, to cover the entire range of Reynolds Number with a more accurate
estimate:
24 .
0.413 (38)
= (1 + 0.173 )+ .
1 + 16300
Haider and Levenspiel (Haider & Levenspiel, 1989) proposed two equations to
determine the drag coefficient for a spherical particle and a non-spherical particle
respectively. For a non-spherical particle, a particle sphericity factor is used to include
the effect of shape; therefore, no further adjustments are needed. Equations (39) and (40)
represent the drag coefficient calculations for spherical and non-spherical particles
respectively:
57
24 .
0.4251 (39)
= (1 + 0.1806 )+
1 + 6880.95
24 (40)
= (1 + )+
1+
Where:
A = exp(2.3288 - 6.4581ψ + 2.4486ψ2), (41)
B = 0.0964 + 0.5565ψ,
C = exp(4.905 - 13.8944ψ + 18.4222 ψ2 - 10.2599ψ3), and
D = exp(1.4681 + 12.2584ψ - 20.7322 ψ2 + 15.8855ψ3)
A new drag coefficient for irregularly shaped particles was developed by Tran-Cong et
al (Tran-Cong, Gay, & Michaelides, 2004) . In this method, rather than using a particle
sphericity factor, a particle circularity factor, c has been utilised. This method is limited
to the ranges Re < 1500, 0.8 < ds/dv < 1.5 and 0.4 < c < 1.0. The model is expressed
with the following equation:
(42)
24 0.15 . 0.42
= 1+ + .
√
√ 1 + 42500
58
Where, is the projected perimeter of the particle in the direction of the motion,
and are surface equivalent sphere and nominal diameter respectively, and c is particle
circularity.
As an iterative approach to calculate the drag coefficient, Karamanev (Karamanev, 1996)

developed a new method using Archimedes Number rather than Reynolds Number,
represented as follows:
432 0.517 (43)

= 1 + 0.0470 +
1 + 154
3
=
4
4.3 Development of an Entrained Air Calculation Method
This section will go through an approach that expands on (Cooper & Arnold, 1995b), to
calculate the amount of entrained air into a free-falling granular material based on
volume conservation theory. Similar to some of the existing theories, the following
method relies on accurate measurement of stream diameter. In Chapter 3:, a
methodology for stream-diameter measurement was introduced, which will be utilised
in conjunction with the following approach for free-falling air entrainment calculation
purposes.
The volume of the free-falling stream from a hopper with a circular outlet is estimated
to be the volume of a cone, shown pictorially in Figure 21. Initially it is assumed that
59
the velocity of the air moving along the cone is equal to the velocity of a single free-
falling particle with drag. It is also assumed that the pressure is constant within the
system and the stream diameter at the impact point is larger than the outlet diameter of
the hopper. Therefore, the fundamental basis of the new theory is reflected in Equation
(44):
V (44)
Q =
t
Figure 21. Schematic View of Free-falling Stream
Where Q is the volumetric flow rate of air induced into the stream of free-falling
particles, V is the volume of the cone and t is the time for a single particle to fall
through height “h” under the influence of drag. Assuming that the outlet shape is circular,
the volume of the cone is given by:
60
πh (45)
V = d + d d +d
12
The velocity of a single particle under the influence of air is given by:
(46)
2mg ρ gAc
v(t) = tanh t +V
ρ Ac 2m
Where m is mass of particle, ρ is the air density, c is the drag coefficient, A is particle
frontal area, t is the time, ρ and d are respectively particle density and diameter, and
g is the gravitational constant.
h
Given that, = v, the drop height is then given by:
(47)
2mg ρ gAc
h= tanh t . dt
ρ Ac 2m
After solving the above integral, the time of free-fall for a single particle under the
influence of drag is given by:
2m ρ (48)
t= cosh (e )
ρ gAc
61
By substituting Equation (45) and (48) into (44) Q is then given by:
π (49)
Q = = ρ
( )
ρ
π ρ π
Given m = and A = substituting into (49) gives:
π (50)
Q = = ρ
ρ ( )
ρ
( )
ρ
The Equation (50) can be applied and extended to any shape of hopper outlet by some
modifications on the calculated stream volume. For ease of referencing, this equation is
named as AVCF theory (Air-Entrainment using Volume Conservation Theory in Free-
Falling)
The stream diameter analysis outcome from Chapter 3: can be utilised in AVCF theory
(Equation (50)), in which, it was found that the prediction of the velocity of a stream of
particles based on the theory of a single particle was a close estimate (maximum error
of 5%) for particles greater than 2mm in diameter. However, for streams of fine
particles, this error increased to as high as 25% which matches matching the findings
of Ogata et al (Koichiro Ogata, Funatsu, & Tomita, 2001). Moreover, the ratio of
particles volumetric flow rate to air volumetric flow rate is well under 5%, to simplify
the AVCF theory this volume have been neglected form the calculations.
62
4.4 Experimental Set-up and Validation
Although scale model experimental studies are time consuming and expensive, the
significance of the outcomes of these works cannot be ignored. To investigate the
validity of AVCF theory (Equation (50)), an experimental program was undertaken to
measure the entrained air within a free-falling stream of granular material. The
experimental setup is shown diagrammatically in Figure 22 and consists of a Perspex
hopper with a 22-degree hopper half angle and 26mm diameter outlet that facilitates the
free-fall of granular material from different drop heights into a sealed enclosure.
Figure 22. Schematic View of Experimental Set-up
The enclosure has a central hole to allow the free-falling stream to enter, and a second
hole positioned in the corner of the enclosure to guide the entrained air into the
measurement venturi. The front of the enclosure is covered with a Perspex sheet paired
with a light source providing visibility of the test(located inside the enclosure). The
63
amount of air entrained into the stream of free-falling material is calculated by

measuring the pressure drop across the venturi. A second pressure gauge was used to
monitor the pressure inside the enclosure, with the objective of maintaining constant
pressure during each test. The complete test rig is presented in Figure 23. After
undertaking smoke test, it was observed that once the flow of the stream stops, there
will be some air escaping from the central hole of the enclosure. To overcome this issue,
a constant net inflow throughout the box was designed and maintained across the
experiments. A venturi arrangement shown in in Figure 22 has been utilised. Venturi is
designed and sized as per BS 5167 and the diameter of the constricted section was 44mm.
According to the BS standard the length of the downstream pipe (located at the end of
the diverged cone) is 6D which has been considered in the design. The venturi system
was calibrated by utilising a set of digital chock flow rate array proving known mass
flow rate of air.
64
Figure 23. Air Entrainment Test Rig
Three different materials were tested during the experimental work, with the properties
of each material detailed in Table 6. To ensure continuous discharge from the hopper,
the outlet diameter was selected to be significantly larger than the diameter of the largest
particle tested.
65
Table 6. Properties of Materials Tested
Hopper
Solid Bulk Average
mass
Material Density Density Particle
flow rate
(kg/ ) (kg/ ) Size(mm)
(kg/s)
3mm Plastic Pellets 870 522 4.0 0.09
3mm Glass Beads 2450 1453 3.0 0.34
Sand 2600 1566 0.22 0.38
By having the mass flow rate and bulk density of bulk material, the average initial
velocity of the material also can be calculated.
4.5 Results and Discussions - Theory versus Experimental

Data
Figure 24, Figure 25 and Figure 26 show the experimental data versus the theoretical
predictions of Hemeon – Equation (9), Cooper and Arnold – Equation (12) and the new
model given by AVCF theory (Equation (50)). Since Equations (9) and (50) depend on
the stream diameter of the free-falling stream, it is noted that the stream diameter used
in these equations was calculated from the work described in Chapter 3.
66
5
Entrained Air (liter/s)
4
0
0 200 400 600 800 1000 1200
Drop Height (mm)
Experiment Equation (9) Equation (12) Equation (50)
Figure 24. Entrained Air versus Drop Height for Sand
12
10
0
0 200 400 600 800 1000 1200
Drop Height (mm)
Figure 25. Entrained Air versus Drop Height for 3mm Glass Beads
67
9
8
7

6
5
4
3
2
1
0
0 200 400 600 800 1000 1200
Drop Height (mm)
Figure 26. Entrained Air versus Drop Height for 3 mm Plastic Pellets
Based on the data presented in Figure 24 to Figure 26, it is clear that Equation (9)
consistently over-predicts the experimental data for all three materials tested. Equation
(12) over-predicts the experimental data for sand; however, it provides a good estimate
for the coarser particles. AVCF theory (Equation (50)) provides a good estimate over
the full range of experimental data.
AVCF theory (Equation (50)) calculates the mass flow rate of entrained air in the
material stream, using the velocity of a single particle free-falling under the influence
of drag. Nevertheless, the velocity of the fluid is generally lower than the velocity of the
particles in a free-falling case (Uchiyama, 2004).
Conversely, the pressure within the system is assumed to be constant, which is not the
case in reality. However, these two estimates are acting in opposite directions in terms
of the entrained air quantity calculation and they can balance-up their effects to some
extent.
68
In terms of the sensitivity of AVCF theory (Equation (50)) to different parameters, it is

realised that the most effective parameter is the stream diameter and hence, it is
important to estimate the stream diameter accurately.
4.6 Conclusion
A comprehensive review of the existing air-entrainment formulas has been provided and
their strengths and weaknesses were compared in details, as far as their accuracies under
different environmental conditions were concerned. Furthermore, a comprehensive
review on particle motion has been presented and the impacts of all acting forces on a
particle in free-falling condition, under the influence of gravity have been discussed. A
new theoretical formula has been developed for air entrainment prediction and its
efficiency has been compared with the experimental validation results. The stream-
diameter measurement technique provided in Chapter 3 has been employed into
development of this formula.
AVCF theory (Equation (50) predicts the amount of the induced air comparatively
closely to the experimental results over the range of experiments undertaken. However,
in terms of the fine particles, stream velocity comparing to the velocity of single particle
can cause calculation error up to 25%. Therefore, in case of applying the formula to the
fine particles, it is expected that the accuracy of the formula decrease considerably,
although our experiment did not show it. AVCF theory (Equation (50) does not have
any empirical parameters and the results indicate that it provides a close estimation of
the real amount of air entrained into the stream of free-falling coarse particles, as
observed throughout the experiments (results are limited to drop height 1.2m).
In addition, it was observed that the developed theory is highly sensitive to the stream
diameter, whereas determining an exact stream diameter will increase the accuracy of
the results significantly.
69
Free-Falling – Air Entrainment Computer Simulation
Chapter 5: Free-Falling – Air Entrainment Computer

Simulation
Computational Fluid Dynamics (CFD) uses applied mathematics, physics and advanced
computer software to model and visualise the flow of fluids inside equipment or a
processing facility. Fluid flows are governed by partial differential equations, which
represent conservation laws for the mass, momentum and energy. CFD is based on
Navier Stocks equations, which describes how velocity, density, pressure and
temperature of a moving fluid are related. The fluid phase can be attributed to liquid,
gas or loose-particle form, or a combination of them. By employing computer
simulation, a wide range of operational variables and variations in physical design can
be tested and refined until optimum performance has been achieved. The broad
applications of CFD to new and existing plants in a variety of industrial fields, mainly
mineral processing, metal production and energy industries, have saved companies
many millions of dollars and has the potential to grow further.
In this chapter, the interaction between particles and air in the free-falling of coarse
particles is reviewed by employing both CFD computer simulations and experiments. A
series of physical experiments were carried out using the Particle Image Velocimetry
(PIV) technique to measure the fluid velocity in a free-falling stream of coarse material.
Each experiment is simulated using ANSYS CFX simulation software. The results
obtained from the experiments, along with the theoretical models are utilised for
simulation verification purposes. These results are used as the basis of air-velocity and
air-entrainment analysis.
Particle and air velocities as well as air entrainment readings from simulation are
compared with experimental measurements and theoretical formulas. The experimental
program is designed and conducted to classify the effect of particle diameter and particle
density independently.
70
Furthermore, the CFD methodologies as well as the governing formulas are explained
within this chapter. This chapter presents several distinct sections that detail the
experimental and simulation results together with a comparison of these two methods,
followed by a discussion of the existing theories.
5.1 Computational Fluid Dynamics (CFD)
CFD (Computational Fluid Dynamics) utilises algorithms and numerical analysis to

solve fluid-flow cases by aid of computers. This method is capable of simulating the
interactions between liquids and gases with surfaces (boundary conditions).. The
performance of the CFD software, in terms of accuracy and speed of simulation
scenarios, has been one of the key areas of improvement amongst researchers over the
last couple of decades and the author believes this will remain one of the key topics in
coming years.
Amongst a number of available CFD simulation models, the multiphase-flow simulation

model is the most applicable theory for the free-falling cases. In principal, multiphase
flow refers to the situation where more than one fluid is present. All fluids may share a
common flow field or each fluid may possess its own flow field. Unlike multi-
component flow, the fluids are not mixed in a microscopic scale in multiphase flow.
Rather, they are mixed in a macroscopic scale with a discernible interface between the
fluids.
ANSYS CFX simulation software includes a variety of multiphase models to enable the
simulation of multiple fluid streams, bubbles, droplets, solid particles, and free-surface
flows. Fundamentally, two distinct multiphase flow models are available in ANSYS
CFX for particle and air modelling: Eulerian-Lagrangian Particle Tracking multiphase
model and Eulerian-Eulerian multiphase model (Ansys13.0, 2010).
Eulerian-Lagrangian model solves equations of motion for each parcel individually. In

fluid dynamics, within the framework of continuum mechanics, a fluid parcel is a very
71
small amount of fluid, identifiable throughout its dynamic history while moving with
the fluid flow (Batchelor, 2000).
The fluid phase is treated as a continuum by solving the time-averaged Navier-Stokes

equations, while the dispersed phase is solved by tracking a large number of particles,
bubbles, or droplets through the calculated flow field (Fluent 6.3, 2006).
The dispersed phase can exchange momentum, mass, and energy with the fluid phase.
This method offers individual parcel tracking and therefore requires greater
computational resources. Moreover, this method is recommended for a lower-volume
fraction of solids and is limited to dilute flow. The particle or droplet trajectories are
computed individually at specified intervals during the fluid phase calculation. This
makes the model appropriate for the modelling of spray dryers, coal and liquid fuel
combustion, and some particle-laden flows. However, this model is not recommended
for modelling of liquid-liquid mixtures, fluidised beds, or any applications where the
volume fraction of the second phase cannot be neglected. For such applications, particle-
particle interactions can be included using the Discrete Element Model which is not
covered in this work.
The second model is the Eulerian-Eulerian model, which considers both phases as fluids.
This model can be used to simulate dense multiphase flows. Within the Eulerian-
Eulerian model, certain interphase transfer terms used in momentum, heat, and other
interphase transfer models, can be modelled using either, the Particle Model,
the Mixture Model or the Free Surface Model.
Multiphase modelling employs the notion of interpenetrating continua. Phases are

mixed at length scales much larger than molecular (Ansys 16 Help). Thus, in principle,
each phase is assumed to be present in every control volume, and is assigned to a volume
fraction equal to the fraction of the control volume occupied by that phase.
72
Hence, the first step to solve any multiphase simulation is to determine the best-suited
model to solve the case. Therefore, the model assumptions must be understood prior to
utilisation. Additionally, identifying the best method to estimate the key parameters (i.e.
through formulas or experiment) is the key to obtaining accurate results. Therefore in
this section, prior to providing the details on adopted models for bespoke simulation for
free-falling, a brief review on the governing equations is provided for the relative models.
Within the current work, the adopted models solve the governing equations of mass and
momentum by means of a multiphase Eulerian model incorporating the Kinetic Theory
of Granular Flow (KTGF) available in the CFX software. The energy conservation
equation is not utilised as the flow is isothermal. A brief summary of the model
equations are listed in the sequel.
5.2 Simulation Theory
5.2.1 Mass Conservation Equations
The conservation of mass for the gas and solid phases can be written as:
(51)
+ ∇∙ =0
(52)
( )+ ∇∙( )=0
Where ε and ρ are the volume fraction and density (for gas and solid), respectively, and
is the instantaneous velocity. The volume fraction constraint requires that + =
1.
73
5.2.2 Momentum Conservation Equations
Assuming no mass transfer takes place between phases and no lift and virtual mass
forces exist due to the large difference in their densities, only the drag force and gravity
are considered in this study. Therefore, the conservation of momentum for each phase
can be articulated as:
∂ (53)
ε ρ v + ∇∙ ε ρ v v
∂t
= ∇ ⋅ τ − ε ∇P + ε ρ g + K v −v
ερv + ∇∙ ε ρ v v = ∇ ⋅ τ − ∇P − ε ∇P + (54)
ε ρ g+K v −v
Where ̿ is the stress tensor, P is the pressure, g is gravity and is the momentum
exchange coefficient that represents the drag force between the gas and solid phases.
The stress tensor for each phase is given by:
2 (55)
τ =ε μ ∇v + ∇v +ε λ − μ ∇ ⋅ v I̿
3
2 (56)
τ = ε μ (∇v + (∇v ) ) + ε λ − μ ∇ ⋅ v I̿
3
Where, λ is the viscosity. The granular bulk viscosity represents the resistance of
granular particles to compression or expansion, and is modelled by Lun et al. (Lun,
Savage, Jeffrey, & Chepurniy, 1984) as:
74
(57)
4 Θ
λ = ε ρ d (1 + e )
5 π
Where, ds is the particle diameter, e is the particle-particle restitution coefficient and Θ

is the granular temperature.
The granular pressure Ps is derived from the kinetic theory of granular flow; it is
composed of a kinetic term and a term due to particle collisions. In the current work, the
model proposed by Lun et al. (Lun et al., 1984) given by the following equation is used:
P = ε ρ Θ + 2ρ (1 + e)ε g , Θ (58)
Where, g , is the radial distribution function that modifies the probability of collisions
between particles which is modelled as the following equation by Lun et al. (Lun et al.,
1984):
(59)
ε
g ,
= 1−
ε ,
5.2.3 Drag Model
A considerable number of research works have been undertaken in relation to drag

model studies due to the importance of the drag force in the terminal-velocity calculation
and fluidised bed. Generally, the drag force experienced by a single spherical particle in
75
a fluid-solid system can be represented by the slip velocity between the two phases and
the fluid-solid exchange coefficient β:
(60)
3 ε ρ v −v
f = C v −v =β v −v
4 d
Where is the drag coefficient. The fluid-solid exchange coefficient β can be written
in the form of:
ρ ε (61)
β=
τ
Where, f is the drag function defined differently for the various exchange-coefficient
models, and is the particle relaxation time defined as:
ρd (62)
τ =
18μ
An earlier drag model was proposed by Ergun (Ergun, 1952). This model is more
suitable for dense-phase systems since it was derived from the observations on a fixed
bed, where the particles have no relative motion. It is written as:
76
ε μ ε ρ |v − v | (63)
β = 150 + 1.75
α d d
Wen and Yu (Wen & Yu, 1966) proposed a correlation for high void fraction cases
( ≥ 0.8), which was an extension to the work of Ergun (Ergun, 1952). The model is
expressed as follows:
ε ε |v − v | (64)
3 .
β= C ε
4 d
Arastoopour et al. (Arastoopour, Pakdel, & Adewumi, 1990) developed a drag model to
describe gas-solid flow behaviour in dilute gas-solid systems. The model considered
particles of the same size and their boundary layers as a separate continuum phase; it is
expressed as:
ρ (65)
17.3 .
β= + 0.336 v −v ε ε
Re d
The particle Reynolds Number is given by:
d |v − v |ρ (66)
Re =
μ
77
To cover the entire range of void fraction cases, Gidaspow (Gidaspow, 1994) proposed
a model to combine the Ergun drag model and Wen-Yu drag model. It is expressed as:
ε μ ε ρ |v − v | (67)
β = 150 + 1.75 ε < 0.8
ε d d
(68)
3 ε ε |v − v | .
β= C ε ε ≥ 0.8,
4 d
As noted in Section 4.2, there are different models available to calculate (drag
coefficient). The first model for drag coefficient estimation, introduced by Schiller-
Naumann (Schiller & Naumann, 1935) is expressed as:
24 . (69)
C = 1 + 0.15 Re Re < 1000
Re
C = 0.44 Re > 1000 (70)
Syamlal and O’Brien (Syamlal & O'Brien, 1988) presented a drag model based on
investigation of the segregation in a liquid fluidised bed. The drag model was derived
from the terminal velocity of particles and it is expressed as:
78
ρ ε ε |v − v | (71)
3
β= C
4 f d
Where:
4.8 (72)
C = 0.63 +
Re ⁄f
Where the particle Reynolds Number Rep defines the same as Equation (79) and ft is the
terminal velocity correlation for the solid phase:
f = 0.5 M − 0.06 Re + (0.06 Re) + 0.12 Re(2N − M) + M (73)
With:
. (74)
M=ε
.
N=ε ε < 0.15,
.
N = 0.8ε ε ≥ 0.15.
79
Zhang and Reese (Zhang & Reese, 2003) extended the Gidaspow drag model to consider
the random fluctuation velocity of individual particles by introducing a correction
coefficient U . It is described as:
ε μ ε ρ U (75)
β = 150 + 1.75 α < 0.8
ε d d
(76)
3 ε ε |v − v | .
β= C ε α ≥ 0.8,
4 d
. (77)
8
U = v −v +
6 21 (78)
C = 0.28 + +
Where the particle Reynolds Number Rep is given by:
ρ Ud (79)
Re =
μ
Yang et al. (Yang, Wang, Ge, & Li, 2003) suggested a EMMS (energy-minimisation
multi-scale) model which was based on a new approach. This model is capable of
reflecting the real mechanisms of gas-solid interactions, which are highlighted as
follows:
80
ε μ ε ρ |v − v | (80)
β = 150 + 1.75 ε < 0.74
ε d d
(81)
3 ε ε v −v
β= C ε χ(ε ) ε ≥ 0.74
4 d
Where CD symbolises the standard drag coefficient for an individual particle as shown
in Equations (69) and (70), and χ is the drag-coefficient correction factor:
0.0214 (82)
χ = −0.5760 + (0.74 ≤ ε
4 ε − 0.7463 + 0.0044
≤ 0.82)
0.0038 (83)
χ = −0.0.0101 + (0.782 ≤ ε
4 ε − 0.7789 + 0.0040
≤ 0.97)
χ = −31.8295 + 32.8295ε (0.74 ≤ ε (84)
≤ 0.82)
The particle Reynolds Number Rep can be calculated through Equation (79).
81
5.2.4 Granular Shear Viscosity
The solids stress tensor contains shear and bulk viscosities arising from the particle
momentum exchange due to transition and collision (Fluent 6.3, 2006). In other words,
the solids shear viscosity consists of a collision term, a kinetic term and a friction term:
μ =μ +μ +μ (85)
, , ,
The collisional viscosity is taken from the kinetic theory of granular flow by Gidaspow
et al. (Gidaspow, Bezburuah, & Ding, 1992):
(86)
4 Θ
μ = ε ρ d g , (1 + e )
, 5 π
Gidaspow et al. (Gidaspow et al., 1992) described the kinetic viscosity as follows:
(87)
10ρ d πΘ 4
μ = 1 + ε g , (1 + e )
, 96(1 + e )g ,
5
Moreover Schaeffer (Schaeffer, 1987) defined the friction viscosity as follows:
P sin∅ (88)
μ ,
=
2 I
Where, P is the solid pressure, the constant ϕ = 30.0 is the default angle of internal
friction and I2D is the second invariant of the deviatoric stress tensor, which can be
written as:
82
1 (89)
I = (D −D ) + (D −D ) + (D −D ) +D +D
6
+D
1 ∂u , ∂u , (90)
D = +
2 ∂x ∂x
5.2.5 Turbulence Model
Turbulence is a transient phenomenon. Therefore, turbulence models are required for

closed forms of the time-independent Navier-Stokes equations, which involve unknown
Reynolds stress correlations. There have been many different concepts and attempts to
solve the turbulence closure problem in a general form over the past fifty years.
The first approach is to solve the Reynolds-averaged Navier-Stokes equations with

suitable models for turbulent quantities, named RANS. An alternative approach is to
compute them directly, named Direct Numerical Simulation (DNS). Fine grids and time
steps are required due to the minor fluctuations for small motions in turbulence flows.
Thus, a direct simulation becomes infeasible for high Reynolds Numbers. Furthermore,
due to intensive resource requirements and slower process involved in the calculation
phase, although this model can provide accurate results, it is less desirable for industrial
fluid simulation purposes.
Large Eddy Simulation (LES) is another model, which is in between RANS and DNS.
As highlighted earlier, the accuracy of this method is increased compared to RANS; also
less computational effort is required. However, due to high computational costs it is
effectively non-viable to currently solve industrial-sized problems by LES modelling.
Thus, RANS is currently the best option to study most industrial problems.
83
The RANS models can be classified into three main categories: linear eddy viscosity
models, non-linear eddy viscosity models and Reynolds Stress Models (RSM). More
specifically, linear eddy viscosity models include zero-equation models, one-equation
models and two-equation models. Among all RANS models, only three widely used
models including standard k-ε, standard k-ω and Reynolds Stress Model (RSM) will be
discussed; for the rest of the models and the details of the turbulence models, the reader
is referred to (Ansys13.0, 2010; Cebeci, 2004).
The k-ε model was proposed by Jones and Launder (Jones & Launder, 1972) based on
the eddy viscosity concept. It focuses on the mechanisms that affect the turbulence
kinetic energy. The turbulence kinetic energy k and its rate of dissipation ε are obtained
from the following transport equations:
(91)
( )+ ( )= + + + − − +
(92)
( )+ ( )=
+ + ( + )− +
Where, Gk represents the generation of turbulent kinetic energy due to the mean velocity
gradients. Gb is the generation of turbulent kinetic energy due to buoyancy.
YM represents the contribution of the fluctuating dilatation in compressible turbulence
to the overall dissipation rate. G , G and G are constants. σ and σ are the
turbulent Prandtl numbers for k and ε, respectively. Sk and Sε are user-defined source
terms.
84
Some improvements have been made to the k-ε model since both the advantages and
disadvantages became known. There are two variants available in ANSYS FLUENT:
the RNG k-ε model and the realisable k-ε model (Ansys13.0, 2010).
The k-ω model was proposed by Wilcox (Wilcox, 1998), based on model transport
equations for the turbulence kinetic energy (k) and the specific dissipation rate (ω),
which incorporates modifications for low-Reynolds-Number effects, compressibility,
and shear- flow spreading. For standard k-ω models, the turbulence kinetic energy k is
given by:
( )+ ( )= (93)
+ − +
The specific dissipation rate ω is given by:
(94)
( )+ ( )= + − +
Where, Gω represents the generation of ω, . represent the effective diffusivity of k

and ω, respectively. Y and Y represent the dissipation of k and ω due to turbulence
and S and S are user-defined source terms.
As the k-ω model has been in existence and modified over many years, some
improvements have been made to increase the accuracy of the model for predicting free-
shear flow. A variation of the standard k-ω model named the Shear Stress Transport
(SST) k-ω model.
85
The Reynolds Stress Model (RSM) originates from the work by Launder et al. (Launder,
Reece, & Rodi, 1975), which is the most elaborate type of turbulence model that is built
into ANSYS FLUENT. It involves calculation of the individual Reynolds stresses
′ ′
using different transport equations. The exact transport equations for the
transport of Reynolds stresses may be written as follows:
′ ′ ′ ′ (95)
+
′ ′ ′ ′ ′
=− + + +
′ ′ ′ ′ ′ ′
− +
′ ′
− +
′ ′ ′ ′
+ ′ + −2
′ ′ ′ ′
−2 Ω + +
Equation (95) can also be written as:
Local Time Derivate + Cij = DT,ij + DL,ij + Pij+ Gij + ij- ij + Fij + User (96)
Defined Source Term
Where, Cij is the convection-term, DT,ij equals the turbulent diffusion, DL,ij represents the
molecular diffusion, Pij is the term for stress production, Gij is the buoyancy production,
86
ij stands for the pressure strain, ij represents the dissipation and Fij is the system rotation.
Of these terms, Cij, DT,ij , Pij and Fij do not require any modelling, however, the rest of
the terms need to be modelled to solve the equations.
Different turbulence models have different performance, applicability and limitations.

Some popular and typically applied turbulence models including their strengths and
weaknesses are summarised in Table 7.
87
Table 7. Comparison of RNAS Turbulence Models (Ansys13.0, 2010; Bakker, 2002)
Model Strengths Weaknesses

Efficient and Economical; Model has not widely tested yet; lack
Spalart- satisfactory record of of sub models have been noticed (e.g.
Allmaras accomplishment for mildly complex combustion, buoyancy).
types of flow.
Robust, easy to implement and Average results for complex flows

economical, reasonably accurate; with severe pressure gradients, strong
converge relatively easily. streamline curvature, swirl and
rotation. Predicts that round jets
spread 15% faster than planar jets,
Stand k-ε whereas in reality they spread 15%
slower.
Only useable for fully turbulent flow
and non-separated flows.
Simplistic ε equation.
Recommended for moderately Subjected to limitations due to
complex flows like jet isotropic eddy viscosity assumption.
RNG k-ε impingement, separating flows, Same problem with round jets as
swirling flows, and secondary standard k-.
flows.
Offers largely the same benefits as Subject to limitations due to isotropic
Realisable
RNG but also resolves the round-jet eddy viscosity assumption.
k-ε
anomaly.
Good for wall-bounded boundary Separation is typically predicted to be
layer, shear and low Reynolds excessive and early. Requires mesh
Number flows. Suitable for resolution near the wall.
Stand k-ω
complex boundary layer flows
under adverse pressure gradient and
separation.
Provides more accurate and reliable Less suitable for free-shear flows due
results for a wider class of flows to dependency on wall distance, and
like adverse pressure gradient requires mesh resolution near the wall.
SST k-ω flows, airfoils, and transonic shock
Produces large turbulence levels in
waves.
regions with large normal strain, like
stagnation regions and regions with
strong acceleration.
Reynolds Physically most complete model Requires more CPU effort (2-3 times);
Stress (history, transport, and anisotropy tightly coupled momentum and
Model of turbulent stresses are all turbulence equations.
accounted for).
(RSM)
88
5.3 PIV Experimental Testing
To verify CFD results, some basic flow information needs to be measured with scale-
model- or full-size-model experiments and compared with the simulation results. The
experimental testing for this study utilised Particle Imaging Velocimetry (PIV), which
is an optical flow-visualisation method. The system consists of a high-performance
camera coupled with a laser-generation source.
PIV technology uses cross-correlation to find the distance that a particle pattern has
moved during the inter-image time and translates this into a velocity measurement. PIV
measures velocity fields by taking two images in quick succession and calculates the
distance that each individual particle travelled within this timeframe. Hence, the velocity
can be calculated based on the time between two captured images and the particle
displacement. The cross-correlated function is calculated on the small square areas
named “interrogation windows”. Depending on the desired accuracy, the size of the
interrogation windows and the resolution of the camera are adjusted.
An Oxford Firefly diode laser coupled with a high-performance camera was employed
for the experiments used in this research. The experimental arrangement is shown as per
Figure 27.
89
Figure 27. PIV Set-up Schematic
For the experiments, particles were loaded into an axisymmetric hopper with a 26mm
outlet and 22˚ half angle then discharged into an enclosure. The PIV system was set up
at 900mm below the hopper outlet. For each experiment, the free-falling stream of
particles falls into the enclosure that is prefilled with seeding particles, which are used
to visualise the airflow. To increase the uniformity of seeding particles, they were
injected into the enclosure through a filter paper to ensure the size of the seeding
particles was consistent across the entire experiments. The enclosure is a 1000mm
square shape cross sectional box with 1200mm height.(A.A. Esmaili, 2015)
The materials used in the experimental program were granular in nature (Table 8); their
sizes were selected to classify the effect of particle diameter and particle density,
independently.
90
Table 8. Summary of Materials
Solid Bulk Average 26mm Hopper

Material Density Density Particle Outlet mass flow
(kg/m³) (kg/m³) Size(mm) rate (kg/s)
3 mm Plastic Pellets 870 522 3.0 0.09
3 mm Glass Beads 2450 1445 3.0 0.34
6 mm Glass Beads 2450 1331 6.0 0.26
To obtain accurate results in the experiments conducted in this work, a number of

variables need to be considered. It has been observed that the size of seeding particles
and their distribution density within the enclosure can affect the results considerably.
Concerning the size of the seeding particles, the smaller sized particles become airborne
at lower velocities. Therefore, smaller particles are preferred seeding particles.
However, different PIV arrangements have different specifications and the minimum
seeding-particle diameter will be determined by the type of the laser. In this work,
Talcum powder with a size distribution between 2µm and 100µm and average diameter
of 15.5µm was used as the seeding agent.
The PIV system synchronises the camera, laser flashes and frame grabber to obtain a
pair of images separated by a short time delay (400μs). Pairs of images were obtained
at a frequency of 10Hz. The camera’s field of view was set to capture a region about
180×250mm. To minimise the random error, each test was run at least three times.
To validate the simulation results, a comparison between them and the results from
analytical methods (Chapter 4), PIV results and the experimental results from Chapter
4 is carried out.
91
5.4 Simulation of Free-Falling
There are several CFD software codes available for research and commercial simulation
purposes. ANSYS software has been used extensively for research purposes. It offers
two different solvers for CFD simulations: ANSYS CFX and ANSYS Fluent. In this
chapter, ANSYS CFX (v. 14.5) software has been employed for simulations. In the
multiphase modelling technique, both the particles and the fluid are modelled as a fluid;
the air is a continuous fluid phase while the particles are a dispersed solid phase. The
simulation was performed for 3mm and 6mm glass beads as well as 3mm plastic pellets.
The CFD models have been validated through experimental results and compared with
the theories. The majority of the existing theories of mass flow rate calculation for
entrained air in free-falling applications use the stream-diameter as a key parameter; the
stream-diameter variable influences the entrained air magnitude significantly. The
previous work by the author (Esmaili et al., 2013) measured the stream-diameter
experimentally, which results are used in the current Chapter. In addition to this, these
results can also be used to assist in the selection of the domain size for the CFD
simulations.
To land on a benchmark simulation a couple of domains are simulated in which the

computational domain was selected to be much larger than the region of interest in the
simulation. Two simulation were undertaken; one scenario with a 1000mm rectangular
shape cross sectional domain representing the experimental enclosure arrangement and
the other scenario with a 900mm radius cylindrical shape domain. Both geometries were
1200mm in height. After detail review of the outcomes, it was observed that the results
obtained from rectangular cross sectional domain were identical to cylindrical shape
domain. However cylindrical shape domain’s calculation timing was considerably
quicker due to smaller number of meshes. Hence the cylindrical shape domain has been
taken forward for further refinement.
92
Figure 28. CFD Geometry used for Benchmark Simulation (cylinder shape)
The computational domain in cylindrical geometry was set to be a 900mm diameter

cylinder as shown in Figure 28, which is much larger than the 26mm material inlet.
Following this simulation, a range of additional simulations was conducted in which the
computational domain was reduced in size further while keeping the simulation physics
the same. Comparisons between the benchmark simulation and the additional
simulations were made; and parameters such as the particle and air velocity, as well as
the entrained air were analysed. A conical shape with dimensions of 300mm (at top) and
500mm (at bottom) with a 1200mm drop height was selected as per Figure 29 as the
optimum geometry domain; it reduces the computational time (smaller mesh quantity)
while achieving the same results as the benchmark simulation.
93
Figure 29. Final CFD Geometry selected for Simulations (cone shape)
Figure 30 shows a comparison between the two studies geometries.
2.5
2.0
1.5
1.0
0.5
0.0
300 600 900 1200
Drop Height (mm)
Cylindrical Domain Simulation Conical Shape Domain Simulation
Figure 30. Air Entrainment Comparison; Cylindrical Domain Simulation versus Conical
Shape Simulation
94
In addition to the study of optimum domain size, mesh sensitivity analysis were
undertaken to ensure the most optimum mesh size is being utilised. A range of
simulations for the optimum mesh size was carried out. Four cases were simulated; the
cases had the same geometry and physics but with different mesh sizes, starting from a
4.36mm mesh size (coarse mesh) and decreasing to a 1.98mm mesh size (finer mesh).
The results shown in Table 9 indicate that the particle velocity was very close in all four
of the simulations. However, there were considerable differences for the air velocity and
entrained air volumetric flow rate across the different mesh sizes. The fine mesh showed
comparable results to the finest mesh, and hence to reduce the computational effort, the
fine mesh size was taken forward for the subsequent analysis.
Table 9. Mesh Sensitivity Analysis
Coarse Medium Fine Finer

Mesh Size (mm) 3.47 2.31 2.03 1.98
Number of Elements 32,185 77,315 103,985 109,559
Max 1.16 1.36 1.43 1.45
Air velocity at 1200 mm (m/s)
Ave 1.12 1.31 1.36 1.37
Max 3.96 4.08 4.106 4.137
Solid velocity at 1200 mm (m/s)
Ave 3.92 4.01 4.068 4.11
The selection of turbulence models was conducted by assessing the relative performance
of the two most applicable turbulence models for use in this study. Two simulation cases
were investigated for each material to assess the performance of the − turbulence
model and the Shear Stress Transport (SST) turbulence model. For evaluation of these
turbulence models, the simulation results were compared with the experimental results.
It should be noted that the − turbulence model is commonly used in CFD
simulations as it is known to be robust and reasonably accurate for a wide range of
applications; it is also considered as the industry standard model (Chowdhury, 2007).
The SST k-ω model was developed by (Menter, 1994) to effectively blend the robust
and accurate formulation of the k-ω model in the near-wall region with the free-stream
independence of the − model in the far field. This model is the next popular
95
turbulence model in industry and is recommended for a variety of cases including

complex geometries. The SST model uses more resources in comparison to the −
model, but it displays similar characteristics to the − model in terms of convergence.
Figure 31, 32 and 33 show a comparison of results between the experimental data,
AVCF theory (Equation (50) and the two turbulence model simulations for each of the
three materials.
3.5
3.0
2.5
Entrained Air (l/s)
2.0
1.5
1.0
0.5
0.0
0 200 400 600 800 1000 1200
Drop Height (mm)
Experiment Equation (50) Simulation K-Epsilon Model Simulation SST Model
Figure 31. 3 mm Glass Bead Turbulence Model Comparison SST vs K Epsilon
96
2.5
2.0
Entrained Air (l/s)
1.5
1.0
0.5
0.0
0 200 400 600 800 1000 1200
Drop Height (mm)
Figure 32. 3mm Plastic Pellets Turbulence Model Comparison SST vs K Epsilon
10
8
Entrained Air (l/s)
0
0 200 400 600 800 1000 1200
Drop Height (mm)
Figure 33. 6mm Glass Beads Turbulence Model Comparison SST vs K Epsilon
97
Analysis of the results of the turbulence model simulations indicates that while the
particle velocities were identical in both sets of simulations, the − model is more
sensitive and over-estimates the air entrainment as well as the air velocity in lower
particle densities. Therefore, the SST model was used exclusively for the simulation
work presented in this chapter.
Further to the CFD simulation parameters in the preceding discussion, Table 10 and
Table 11 tabulate various general and material specific parameters used in modelling.
Table 10. CFD Simulation Parameters – General
3mm Glass 3mm Plastic 6mm Glass Beads

Beads Pellets
Particle Density( ) 2449 870 2449
Gas Density ( ) 1.185 1.185 1.185
Mean Particle
3 3 6
Diameter(mm)
Initial Solid Packing 0.6 0.6 0.6
Inlet Boundary Normal Speed Normal Speed
Condition Type ( ) Normal Speed 0.37
0.47 0.35
Opening with Opening with
Opening with
Outlet Boundary Relative Relative Pressure
Relative Pressure 0
Pressure 0 0
To calculate the inlet velocity, the mass flow rate of each material through the selected
hopper was measured. For this purpose a set of physical experiment were conducted
with a specific amount of material (weighted). The discharge time of each lot through
the selected hopper was measured. Then by having the material weight, density,
discharge time and hopper outlet details, average mass flow rate as well as average inlet
velocity are calculated, reflected within the above table.
98
Table 11. CFD Simulation Parameters – Specific

Parameter Model
Drag Force Gidaspow
Air Continuous Fluid
Solid Particles Dispersed Solids with Particle Dimension
Drag function Gidaspow
Turbulence SST
Time Steady
Air wall boundary No slip
Inlet Normal Speed inlet
Stream Boundary Opening / Opening Pressure and Dirn
Outlet Opening / Opening Pressure and Dirn
Turbulence High Resolution
Advection Scheme High Resolution
5.5 Results
The PIV system come with a software named VIDPIV. This software is capable of
providing some basic analysis of the experimental data; however, to perform further
analysis and improve the data visualisation, MATLAB software was used. For this
purpose, the experimental data have been exported from VIDPIV software and imported
into Matlab software for further image analysis.
Figure 34, Figure 35 and Figure 36 show images of PIV measurements processed with
MATLAB. The arrows indicate the direction and magnitude of the air-velocity vectors,
with the magnitude of the velocity represented by the contours. The results presented
below show the flow velocity ( ) and the flow pattern (A.A. Esmaili, 2015).
99
Figure 34. PIV Image – Velocity of 6 mm Glass Beads in Free-Falling ( )
Figure 35. PIV Image -Velocity of 3 mm Plastic Pellets in Free-Falling ( )
100
Figure 36. PIV Image - Velocity of 3 mm Glass Beads in Free-Falling ( )
It was observed that the measured velocities out of the PIV results involved a mixture
of coarse particles and seeding particles. In an effort to separate the granular-particle
readings and the seeding-particle readings, a histogram analysis was used. As an
example of this, Figure 37 shows the histogram of PIV velocity measurements for the
3mm glass beads. By comparison, the theoretical velocity of a single free-falling 3mm
glass bead under the influence of drag at a 900mm drop height, is 4.1m/s.
It can be seen in the histogram that there are two main sections. In the region of the
higher velocities (approximately 4m/s), there are a large number of readings around the
theoretical velocity for 3mm particles. As this is the case, we can conclude that the
readings in this range can be attributed to the particles, meaning that the rest of the
readings can be considered as air-velocity readings. The same analysis conducted here
for the 3mm glass beads was also conducted for the two other size ranges. The results
of this analysis can be seen in Table 12, which shows a close match between the
experimental PIV air velocities and the CFD air velocities.
101
400
350
300
250
Frequency
200
150
100
50
Velocity
Figure 37. PIV Velocity Histogram of 3mm Glass Beads ( )
In addition to the above analysis, and by comparing Figure 34 to Figure 36, it can be
concluded that by increasing the particle diameter of an identical material, the particle
concentration in a falling stream will be decreased and the stream will be more dispersed.
Moreover, by comparing Figure 35 to Figure 36, it can be seen that by increasing the
density of the particles, the concentration of particles at the centre of the stream will
increase, if the particle size stays constant. These observations are reflected in Table 12.
Table 12. PIV, Simulation and Theoretical Velocity Comparison
PIV Theoretical Simulation

Particle Vel.
Material Name Ave. Air Ave Solid Ave Air Ave.
Drag Incl. (m/s)
Vel.(m/s) Vel.(m/s) Vel. (m/s) Solid
Vel.
(m/s)
3mm Glass Bead 1.65 4.09 4.10 1.77 3.97
3mm Plastic Pellet 1.32 3.96 3.93 1.38 3.91
6mm Glass Bead 1.07 4.35 4.15 0.87 4.30
102
Figure 38. Cross-sectional Velocity Profile of 3 mm Glass Beads
In relation to the air velocity, by comparing the PIV and simulation results of 3mm glass
beads and 3mm plastic pellets, it can be summarised that by increasing the particle
density the velocity of the surrounding air will increase. A cross-sectional velocity
profile of the 3mm glass beads from the simulation is presented in Figure 38. The effect
of particle size can be seen by analysing the results for 3mm and 6mm glass beads. The
results of both experiments and simulations indicate that by increasing the particle
diameter (of the same density), the velocity of air decreases. Figure 39 shows the air
velocity for the three tested materials at 900mm drop height.
103
2.0
1.8
1.6
1.4
Velocity (m/s)
1.2
1.0
0.8
0.6
0.4
0.2
0.0
-0.23 -0.18 -0.13 -0.08 -0.03 0.02 0.07 0.12 0.17 0.22
X Location across 900 mm drop height plane (m)
6mm GB.Velocity [ m s^-1 ] 3mm GB Fluid 1.Velocity [ m s^-1 ]

3PP Fluid 1.Velocity [ m s^-1 ]
Figure 39. Simulation Results for the Air Velocity Distribution at a Plane located 900 mm
below the Hopper Outlet
For the analysis of the air entrainment in the CFD simulations, the experimentally
measured stream-diameter value was used to define an iso-surface plane at each drop-
height location. Iso-surface is one the ANSYS software post processing options which
allows user to create a ‘cloud’ showing where a given value for a parameter exists. The
entrained air-flow rate is then calculated on the iso-plane by analysing the air-flow data
combined with the volume fraction data. The outcome has been compared with
experimental data and theoretical estimation using AVCF theory (Equation (50)), and is
presented in Figure 40, Figure 41 and Figure 42. It is assumed that all particles are
spherical and the drag coefficient is equal to 0.47. The results in these three graphs show
a close correlation between the simulation and experimental results. It can also be
observed that with the conditions reported in this study, the CFD simulations and
theoretical calculations produce more accurate predictions as the particle size increases.
104
3.0
2.5
Entrained Air (l/s)

2.0
1.5
1.0
0.5
0.0
0 200 400 600 800 1000 1200
Drop Height (mm)
Experiment Equation (32) Simulation
Figure 40. Air Entrainment Results from the CFD Simulations for 3mm Glass Beads
2
Entrained Air (l/s)
0
0 200 400 600 800 1000 1200
Drop Height (mm)
Figure 41. Air Entrainment Results from the CFD Simulations for 3mm Plastic Pellets
105
10
9
8
Entrained Air (l/s) 7
6
5
4
3
2
1
0
0 200 400 600 800 1000 1200
Drop Height (mm)

Figure 42. Air Entrainment Results from the CFD Simulations for 6mm Glass Beads
5.6 Conclusion
This chapter has compared the air velocity results for a free-falling stream of particles,
and has investigated both an experimental approach as well as a simulation approach,
to analyse the impact of particle density and particle diameter on the air velocity. The
PIV system was used to measure air velocities in a series of experimental tests.
Furthermore, CFD modelling was used to simulate the experimental cases; a systematic
set-up and verification of a CFD tool has been presented.
The proposed CFD model offered a close correlation on particle velocities between
simulations and theoretically calculated velocities for a single free-falling particle under
the influence of drag. This is probably due to the fact that the falling streams were quite
dispersed. An additional observation driven by both simulation and experiments is that,
by increasing the particle density for an identical particle size and shape, the velocity of
the surrounding air will increase. Furthermore, the effect of particle size has been
analysed both experimentally and by simulation, indicating that by increasing the
particle diameter (same density), the velocity of air decreases.
106
Another major section in this chapter focused on analysing the behaviour of turbulence
models applicable to this case. It was observed that the − model tended to
overestimate predicted air and solids velocities as well as air entrainment with identical
physics (under conditions defined in this study). Based on the modelling conditions used
in this study different simulations were conducted; results indicated that the SST model
results were closer to the experimental outcomes. In terms of the convergence of the
two turbulence models, the − model and the SST model showed no major
differences.
107
Review of Air Flow within Dump Hopper Geometry
Chapter 6: Review of Air Flow within Dump Hopper

Geometry
Dump hoppers are one of the most used equipment in many run of mine material-
handling facilities. Dump hoppers serve large-scale trucks and trains in offloading and
other material handling activities. The size of the trucks can be up to 10 x 15m in width
and length, respectively. The haulage capacity of dump trucks can be as high as 400
tonnes (Caterpillar 797). As mining dump hoppers are designed to service large scale
mining dump trucks and/or trains, they are appropriately large. Hence, dump hoppers
traditionally emit significant volumes of dust during their operation. For instance, Port
Waratah Coal Services, located at Newcastle, Australia, has three of road receival
hoppers with a 600 tonnes per hour capacity (Port Authority of NSW website).
Studies of air entrained into the stream of gravity falling material have been divided into
two principal directions of analysis. Significant attention has been placed on the first of
these free-falling material flow properties and influencing parameters, which were
discussed within the previous chapters. The second section of study predominantly
focuses on practical and more complicated material handling cases by employing the
knowledge obtained from work on free-falling material.
As discussed in detail in Chapter 3, during the free-falling process the air surrounding a
free-falling stream of bulk material is induced into the stream, and upon compaction of
the stream may be exhausted carrying dust particles with it. As the airflow slows down,
particles settle based on their size and velocity.
To this end, a better understanding of the way air is settled within the equipment
geometry will assist us to design more efficient equipment, as far as dust generation and
emission control are concerned.
108
6.1 Experimental Set-up
Based on an industrial installation, a test rig was developed to aid in the visualisation of
the air-flow patterns and turbulence under a range of conditions. The experimental set-
up comprised of a dump hopper with geometry as per a conventional industrial designs.
Figure 43 presents details of the manufactured model in 3D. A 1:10 scale model of an
existing industrial Run Of Mine (ROM) hopper was designed and built from Perspex by
the TUNRA bulk solids workshop to enable the material-flow visualisation. Flow
visualisation is one of the key tools for flow analysis and problem identification.
Figure 43. 3D Scale Model ROM Dump hopper and inclined feeder
109
A high-resolution video camera was employed to record footage for detailed analysis.
Additionally, two air-velocity measurement points (A and B) were fitted within the
geometry as indicated in Figure 44.
Figure 44. Dump Hopper Test Facility - Air Velocity Measurement Points
The air measurement points were selected to facilitate the analysis of air-flow trajectory
within the dump hopper enclosure. Two selected points enabled the measurement of the
air velocity at the lower end of the geometry (Point A) as well as at the top section of
the geometry (Point B), which is located on the main air (dust) outlet. Both points are
located on the central longitudinal datum of the enclosure and 80mm away from the
edge Q, as highlighted in Figure 45.
110
Edge Q
Figure 45. Key Components of the Tested Hopper – Base Model
Figure 45 highlights key components of the test rig. The model has two CFD outlets;
Outlet 1 is located at the far right side of the picture and is covered by elastic rubber
curtains. It is predominantly known as the material outlet with the minimum or
negligible air (dust) outlet. Elastic curtains (Figure 46) are flexible straps of rubber
commonly used in bulk material handling processes to limit dust emissions. Curtains
can segregate two hopper sub sections from each other allowing dust particles to settle
as the air velocity slows down. It is quite common practice to use several layers of elastic
rubber curtains to provide some overlapping and further isolation.
111
Figure 46. Typical Elastic Curtain
The second outlet (Outlet 2) is located at the top of the geometry (adjacent face to the
inlet). A conveyor belt is located at the lower end of the geometry to transfer the material
out of the enclosure. This arrangement will be referred to as the “base model”;
components and sizing of this model remain identical across the other analysed
arrangements within this chapter.
To measure the velocity in the hopper during

transient loading, a number of options were
available including pitot tubes and hot-wire
anemometry. It was suspected that, due to the
high fraction of dust during the experiments,
pitot tube blockage was highly likely. After
conducting a small number of experiments with
a pitot tube, this point became evident. Hence,
a hot-wire anemometer was used as the
measurement instrument for all the reported
experiments. This unit was a Kimo AMI 300
with hot wire. A picture of the unit can be seen
in Figure 47. Figure 47. Kimo Air Velocity
Measurement Unit with Hot Wire
112
The equipment was sent to the manufacturer to perform an ISO 9001 calibration just
prior to experiment to make sure that the measurements were accurate across the range
of experiments. The anemometers are known to be sensitive to the gas properties. It
should be acknowledged there is a small fraction of uncertainty introduced to the
measurements of this experiments due to the fact that the test environment of this
scenario is prone to low fraction of dust. However, as we are predominantly undertaking
comparative studies this point is of less significance.
A brief review of the basics of hot-wire anemometry principle is provided herein for
completeness. Two temperature sensors are used in these systems. The first temperature
sensor provides the gas/flow temperature. The second temperature sensor is kept at a
specified temperature. At zero flow, the difference between the two sensors is zero. With
an increase in the flow, the elevated temperature of the heated sensor is reduced due to
the heat transfer to the passing fluid. As an outcome, additional energy is required to
keep the flow sensor at the pre-set temperature. The increased current to maintain the
temperature of the flow- sensing element is then translatable to a fluid velocity. Figure
48 represents a sample hot-wire measurement system mounted on a pipe section.
Figure 48 Example of Hot Wire Set-up Mounted on a Pipe
113
There are a number of advantages for this method, which makes it more attractive for
this application. However, the key advantage is that it can provide results across a wide
operational range, from small to large flows. The key disadvantage is sensitivity to gas
properties. The diameter of the hot wire needs to be chosen carefully because it is known
as an inherent damper, so it acts as a filter in the flow measurements. Very thin wires
can record high-frequency changes in the flow, whereas thicker wires do not record this
nature of fluctuations.
The material tested was dry iron ore fines with a moisture content under 3%. The tested
material was selected in a way to facilitate a high dust level to allow better visualisation
of the air-flow patterns; the efficiency of different geometry arrangements aimed at
reducing the dust emission. Bulk and solid density of the iron ore (pre-test) measured at
2,166 and 4,356 respectively. The iron ore Particle Size Distribution (PSD)
results are presented in
Figure 49, illustrating that the particle size distribution of the test sample was 80% less
than 6mm diameter.
Particle Size Distribution

100.00
90.00
80.00
70.00
Percent Finer (%)
60.00
50.00
40.00
30.00
20.00
10.00
0.00
0.0 0.1 1.0 10.0 100.0
Sieve Opening (mm)
114
Figure 49. Iron Ore PSD
In addition to air velocity measurements, all experiments were recorded with a high-
resolution (HD) 35fps camera, to enable further air-trajectory review. To maximise the
image quality, the enclosure was washed and cleaned after each experiment. Scale
model testing provided a set of high-resolution videos and air-velocity measurements
for three different arrangements within the selected geometry; these were compared in
detail to highlight the advantages and disadvantages of each arrangement. General
model sizing information is as follows:
 Width max 1,000mm at top rear,

 Width min 420mm at top front
 Opening at the bottom 200mm
 Horizontal Length 1,600mm
 Height 1,990mm
 Drop Height 1,300mm
 Belt velocity was adjusted to be 0.4m/s to transfer the material out of the
geometry reasonably quickly to let analysis to be undertaken with minimum
residual material within the geometry.
6.1.1 Test Procedure
Three different arrangements, based on the “base model” were designed, tested and
analysed to assess the impact of chamber volumes and turbulence on dust generation.
Three arrangements are presented in
Figure 52, Figure 53 and Figure 54. Figure 50 highlights the key components of the
geometry marked as A, B, C, and D.
115
Figure 50. Base Geometry
As mentioned, the inlet is located at the lower left-hand sidewall of section A. The inlet
was supplied with material by a bespoke scale model dump truck tray, dumping the bulk
material into the enclosure as presented in Figure 51. The dumping station was a
component designed with the mechanics of a mine dump truck. This equipment was
attached to a forklift to simulate the process of dumping material from a dump truck.
This discharge tray was based on 1:10 scale of a Komatsu 830E truck body with Westec
trays; it was loaded with approximately 220kg of material. (Each experiment’s material
weight and test duration was measured exactly for average mass-flow calculation
purposes).
116
Figure 51. Bespoke Scale Model Dump Truck Tray
Sections B and C predominantly form the effective volume of the enclosure, which
allows the dust to settle. The right-hand sidewall of section D is the outlet of the model.
Sections C and D are separated by an elastic curtain. This curtain allows a negligible
portion of air to escape through to the outlet; this portion of air is neglected in this
analysis.
To enable an effective volume-analysis process, volumes of each section of the

geometry are provided within Table 13. The interesting point in relation to this design
is that 85.6% of the volume of the enclosure is located within two regions of the
geometry. Region A is mainly the inlet area and is partially part of the effective volume
of the geometry. Therefore, it is less desirable to make changes to this region. However,
region B and C are the focus points of this chapter; as it can be seen, arrangements 2
and 3 are targeting these two regions.
117
Table 13. Dump Hopper Enclosure Volume
Section of Enclosure based on Volume of Section Percent from Total Volume

Figure 50 ( ) of Enclosure
Section A 4.033 ×10 42.9%
Section B 4.0311 ×10 42.9%

Section C 8.243×10 8.8%
Section D 5.087 ×10 5.4%
Figure 52 represents the “base model” arrangement. Within the geometry of this
arrangement, the direction of airflow is mainly guided by the geometry walls and an
elastic curtain, which covers the outlet.
Figure 52. Dump Hopper - Arrangement 1
118
Figure 53 is the base model arrangement with a wall separating sections B and C of the
enclosure. This model was primarily designed to highlight the impact of the enclosure
volume on the air settlement and respectively liberated dust.
Figure 54 is the base model with an additional curved section, which separates sections
B and C of the enclosure. This arrangement is designed to utilise the stream kinetic
energy to generate managed and controlled airflow and turbulence. The key aim of the
controlled airflow is to keep the entrained air (dust) inside the geometry and
subsequently allow air (with dust particles) to slow down and settle within the enclosure.
Hence, this arrangement is primarily designed to test the effectiveness of managed and
controlled airflow and turbulence, which are generated by utilising the stream kinetic
energy.
119
To measure the flow rate of material, the weight of material was measured before the
start of experiments. Then the duration of each experiment was also measured. These
two figures provided an average mass flow rate of the experiments, which will be used
in CFD simulation.
In relation to measurement data collection, KIMO AMI 300 is capable of recording the
data every 5 seconds (frequency 0.2). Therefore, as the duration of the experiments took
around 50~55 sec, an average of 10 readings were collected for every individual test.
Each test was repeated twice to ensure consistency and accuracy of the results. The
average of the validated data from the two sets of experiments has been used for
reporting purposes.
In addition to air velocity measurement, all experiments were recorded with a high-
resolution (High Definition) camera with 35fps frame rate to enable the review of the
air trajectory afterward. To obtain the best footage quality, the hopper was washed and
cleaned after each individual experiment. By adopting this test procedure, the impacts
of different geometry components were measured and reviewed.
120
The radius of the new curved part (indicated in Figure 54), was adjusted to suit the tested
geometry. In this case, a range of radiuses have been tested and visually monitored,
filmed and reviewed. Furthermore, Smoke was used to track the air patterns within the
enclosure. From visual inspection of different experiments, the selected radius perceived
to provide the best outcome for this scenario. The overarching principal is that if the
radius is small, more air can escape from the lower end of the enclosure. However, by
increasing the radius, larger portion of the cross sectional area in the material flow
direction is covered by the curve, which might damage the curve section eventually.
During the experiment, it was noted that the mechanics of the dump truck did not work
in favour of dust suppression. Figure 55 represents a schematic view of a mining dump
truck, highlighting the location of the dumping compartment edge (1) and its associated
hinge (2). The position of the falling edge of the dumping compartment of the truck
moves horizontally during the discharge process, due to location of the hinge, which is
offset from the edge of the dumping component. This design predominantly helps to
reduce the force required from hydraulic jacks to move the dumping compartment,
which, in turn, reduces the production cost of the truck. Hence, as far as the mechanical
efficiency and associated design cost is concerned, this arrangement is an efficient
design, however, the design causes lateral movement in the stream of material during
the discharge process, which increases the air entrainment resulting in increased dust
emission mainly at higher drop heights.
121
Figure 55. Schematic View of Mining Dump Truck
6.2 Model Simulation and CFD Simulation
6.2.1 CFD Multiphase Flow
A CFD multiphase model was used to predict the performance of each dump hopper
configuration with respect to dust emissions.
ANSYS FLUENT and ANSYS CFX are two solvers developed independently, but they
have a number of similarities. Both are control-volume based for high accuracy and
utilise a pressure-based solution technique for a wide range of applications.
122
In the 1990s, it was reported that CFX was predict multiphase flow scenarios better than
Fluent(Soccol, Pandey, & Larroche, 2013). But nowadays both software have the
capability to deal with complex academic as well as industrial simulation scenarios. In
this chapter, the simulation software has altered to Fluent as a broader range of
solvers/methods were available in multiphase topic and also to undertake some cross
simulations output verification purposes, the academic version of commercial CFD
software, ‘FLUENT’ was used to analyse air and particle flow within the hopper
enclosure. The relative performance of each design was projected by analysing the air
and particle velocity.
6.2.1.1 Geometry Sizing Detail
To conduct CFD analysis, geometry is modelled in ANSYS as per the dimensions of the
scale model. A 3D model developed in Pro Engineer CREO was utilised. This model
was imported to ANSYS workbench and unnecessary details have been removed from
the model to minimise solution time.
As a number of simulations will be undertaken, it was important to keep the complexity

of the model at a certain level to keep the timing needed for the simulation cases at a
reasonable level. Therefore, the dumping component was modelled as an angled surface
on the geometry, and defined as the inlet of the model. This assumption enabled the
removal of a considerable computational load and complexity from the simulation, and
focus more on the model details and internal arrangement; the key focus of this work.
The inlet was defined as an angled wall at 15 degrees discharge angle. Based on the
experimental observations, and given that a sample of predominantly fine iron ore
particles were used for this experiment, material discharge occurs when the discharge
angle is between 10 and 15 degrees. The width was modelled pro rata to the sizing of
the dumping component.
123
It was not easy to measure or estimate the thickness of the stream at the discharge or
inlet. However, based on the video footage obtained from the experiments and further
image processing analysis, it has been observed that the thickness of the stream changes
in different discharge angles. This could be a topic for future study and analysis.
However, in relation to stream thickness, observations from the experimental footage as
well as the simulation have been used for measurement purposes. Based on above
evidences, stream thickness of 40mm was considered across the entire the simulation
cases.
The belt conveyor located at the lower end of the geometry was defined as a moving
wall. The velocity was set according to the experimental velocity, which was 0.4 m/s.
Furthermore, inlet was set as velocity inlet; velocity calculated according to the average
mass flow rate measured from similar experimental conditions. Other experiemntal
arrangements are represented in Figure 52 to Figure 54.
6.2.1.2 Model Details
Due to the high-volume fraction of particles, a two-phase Euler-Euler approach was

adopted to model the granular stream and airflow in the dump hopper simulation. An
Eulerian treatment was used for both the fluid and solid phase. A common pressure is
experienced by the air and particles and the continuity and momentum equations were
solved separately for each phase. The models and boundary conditions used in the
simulation are summarised in Table 14.
124
Table 14. CFD Model Description
Description Model Description Model
Bulk viscosity Lun et al. Solver Segregated
Granular viscosity Gidaspow Time Unsteady

Fractional viscosity Shaeffer Volume fraction
Discretisation 2nd Order
Granular temperature Algebraic wall boundary No slip or free
slip
Drag function Gidaspow Inlet Uniform velocity
inlet
Fractional pressure Based-ktgf Opening Pressure inlet
Solid pressure Lun et al. Outlet Pressure outlet
Radial distribution Lun et al.
Multiphase Model Eulerian/Implicit Viscous Model k-epsilon

Turbulence Multiphase Dispersed
Model
The air is considered incompressible with a constant density, while the particles are
assumed to be spherical and have a uniform density. A particle diameter of 4mm was
selected for the simulations. Solid density of iron ore was measured as 4,356 kg/m3.
Solid average velocity at inlet was set at 0.15 m/s.
In relation to turbulence model, previously within the free falling simulations, it was
concluded that SST model provides more accurate results across the analysed case.
Given that this case was more hardware intensive, the author ran a comparison
simulation between two k-epsilon and SST turbulence model (with identical physics,
geometry and mesh size). It was concluded that both turbulence models readings were
identical. This could be translated as the measurement points are beyond the boundary
limits (which refers to the key difference of the two reviewed models). Therefore due to
less memory requirement of k-epsilon model and quicker calculation timing, k-epsilon
is chosen over SST model and utilised across all of the simulations in this chapter.
125
With the above settings, the average computational time for each dump hopper
configuration was restricted to less than 10 days. Belt velocity was set to 0.4 m/s as per
experimental conditions. All simulations were run to simulate 3 minutes of actual
experiments, which represented a fully developed flow. The restitution coefficient
between particles was assumed to be 0.9.
Three cases were simulated in line with the tested scale models, as per Figure 52, Figure
53 and Figure 54.
6.2.2 Mesh Sensitivity Analysis
The arrangement 2 configuration was selected for mesh sensitivity analysis. Three
different mesh settings were tested as per Table 15 to ensure that the optimum mesh size
was used during the bulk of simulations. Each of the mesh sizes listed in Table 15 were
simulated with the same physics. The velocity at point B for arrangement 2, was
experimental measured at 1.16 m/s.
Evidently, the coarsest mesh size did not provide any satisfactory results as it was
substantially different from the experimental measurement. However, medium and fine
mesh results were very close to the experimental benchmark with an acceptable
variance. To reduce the simulation time, medium mesh was selected. Detailed results
are highlighted in Table 15.
Table 15. Mesh Sensitivity Analysis
Experimental Measured
Air Velocity
Velocity
Mesh Size (m) at Point B (m/s)
at Point B (m/s)
Fine - 0.060 1.189 1.16

Medium - 0.075 1.143
Coarse - 0.090 0.586
126
6.3 Results
The first simulated model was the base model. This arrangement was simulated with
Fluent software and a comparison of results with experimental data have been discussed
in detail. Within the experimental testing procedure, air velocity was measured in two
pre-located critical points within the geometry as per Figure 44.
Experimental air velocity at Point A (lower section measurement point) was measured
at 0.63 m/s. By comparing experimental readings versus CFD simulation (which shows
0.69 m/s), it can be concluded that there is reasonable correlation between the
experimental results and simulations. Therefore, adopted models in the simulation were
selected accurately, and they provided a reasonable estimate. Moreover, air velocity
within the geometry is categorised as moderate amongst the other two arrangements.
For this arrangement, the readings from the nominated top section measurement point
indicated that the experimental air velocity was 0.27 m/s versus 0.29 m/s from the
simulation; this reflects an acceptable agreement between simulation and experimental
results. The measured velocity at this point also has a direct relation with the air
exhausted out of the enclosure from the top section. Based on these observations, this
arrangement is ranked as the second preferred design across the three tested
arrangements.
Figure 56 and Figure 57 represent the simulation and experimental results, respectively.
Experimental results (hi-res video) show a moderate level of dust emission being
liberated to the environment through the top opening of the geometry. Additionally,
experiment video footage confirmed that the level of dust escaping through the curtains
can be neglected which is in-line with the key assumptions of this chapter.
127
Figure 56 CFD Air and Solid Velocity Results – Base Model
Figure 57. Experimental Hopper during Physical Testing - Base Model
128
An additional important factor influencing the design efficiency is the pressure

distribution within the hopper geometry, which can be analysed from the cross-sectional
pressure profile. High-pressure areas within the system could influence the air
circulation pattern. Figure 58 illustrates the pressure profile from CFD simulation. This
profile is quite consistent within the effective volume of the geometry. A slight pressure
increase can be seen towards the area where the curtain is located. In the simulation, a
“solid wall” has been utilised to model the curtain. However, in the real model, a curtain
generates a minimal pressure increase within the geometry, which can be neglected.
Figure 58. Cross-Sectional Pressure Profile ̶ Base Model
The second tested arrangement is the base model with an additional curved section.
Figure 59 and Figure 60 represent the results obtained by adding a curved section to the
base geometry. The key design criteria for the curved section is to allow the maximum
material input, therefore the curved-shape design will guide the particles which may
collide with this section, towards the internal sections of the enclosure. The chambers B
129
and C also need to be as isolated as possible, to retain the air within the enclosure until
it (and the fine particles) settles.
The turbulence, which is being generated by the newly added curved section will assist
to re-entrain the air into the stream. This will help to keep the air within the effective
volume of the geometry and reduce dust emission being liberated to the environment.
The curved section converts stream energy to a form of turbulence, which promotes
holding the air within the geometry and helps air to settle within the enclosure.
A comparison between experimental measurements and simulation readings will reveal

the fact that this curved section is reducing the velocity at the upper designated
measuring point (showing as 0 m/s), which means less air is leaving the geometry). This
velocity was negligible.
Furthermore, the lower-section velocity measured at the lower designated reading point,
is lower than in the other two reviewed cases; this allows a larger fraction of particles
to settle within the geometry (depending on their size). The experimental reading at the
lower designated point was measured as 0.5 m/s versus 0.36 m/s in the simulation result.
These readings show a reasonable correlation between simulation and experimental
measurement.
130
Figure 59. CFD Air and Solid Velocity Results - Arrangement 3
Figure 60. Experimental Hopper during Physical Testing – Arrangement 3
131
The cross-sectional pressure profile from the simulation was reviewed. No

abnormalities in relation to pressure build-ups were observed, as per Figure 61. This
profile is quite consistent within the effective volume of the geometry. Similar to the
previous arrangement, a slight pressure increase can be seen towards the curtain
location. The increased pressure is slightly higher than the base model, which is largely
due to the added curved part. As discussed, the pressure increase in the geometry
towards the curtain can be neglected.
Figure 61. Cross-Sectional Pressure Profile ̶ Arrangement 3
The last tested arrangement was based on the base model with a section separating
section B from C. This specific arrangement was designed to test the influence of the
hopper effective volume on air velocity, and subsequently dust generation. Figure 62
and Figure 63 represent computer simulation and experimental results, respectively. The
hi-resolution video obtained from the experiment indicates that the dust level at the top
outlet is the highest across all tested cases.
132
The velocity measurements were taken only at the upper designated measurement point,
as the air velocity at the lower end was very low. The readings were 1.16 m/s for
experiment versus 1.14 m/s for simulation.
By comparing the results from this arrangement with the two previous tested
arrangements, it can be concluded that the air velocity at the top outlet is higher due to
the smaller volume inside the enclosure (as the sections B and C are separated), and
ineffective design in relation to air pattern within the geometry. Hence, it can be
concluded that effective volume of enclosure plays an important role in dust-emission
reduction as long as the enclosure is designed properly. The observed high velocity is
largely due to lack of enough volume within the geometry for the entrained air to settle.
As far as dust-emission liberation to the environment is concerned, this model has the
highest air velocity and consequently the least efficient design of the three reviewed
models.
133
Figure 62. CFD Air and Solid Velocity Results – Arrangement 2
Figure 63. Experimental Hopper during Physical Testing – Arrangement 2
134
Again, the cross-sectional pressure profile was reviewed, based on the obtained
simulation results. The pressure profile is quite different to the other two tested cases,
as chamber C is not considered predominantly as a part of the effective volume of the
geometry.
In this arrangement, no major pressure increases were observed. The air can easily
escape from the geometry due to no barriers existing in its way; hence, the air velocity
is the highest amongst all three tested arrangements. This arrangement has the least
efficient design of the three tested models.
Figure 64. Cross-Sectional Pressure Profile – Arrangement 2
135
Table 16. Velocity Comparison between Simulation and Experiments
Simulation Experiments
Simulation results Top Bottom Top Bottom
m/s m/s m/s m/s

Curved Part 0.36 0.5
Base Geometry 0.43 0.63 0.29 0.63
Middle Separator - Old

1.23 1.16
Arrangement
6.4 Conclusions
A detailed computer simulation was carried out for three different physical
arrangements for which experimental data was collected for comparison. Simulation
results were presented and compared with the experimental results as a verification test,
which provided a reasonable correlation between experimental results and simulation.
In addition, the design of dump hopper models was critically reviewed and analysed.
Details of the geometry and the impact of turbulence, as well as the volume of the
geometry was analysed and presented. It was concluded that the volume of the enclosure
plays a significant role in providing sufficient time for dust particles to settle; this results
in dust-emission reduction within the work environment. The airflow pattern within the
geometry was reviewed and discussed in detail; pressure variation through the hopper
enclosures was used as a parameter to present the effectiveness of the design.
Additionally, a new arrangement was designed and tested which, proved to reduce the
air velocity and thus the dust emission by generating an effective turbulence. This, then
allow airborne particles to either re-entrant back into the stream or settle within the
effective volume of enclosure. The turbulence is generated by the stream energy and
136
this finding is therefore valuable from a sustainable energy design perspective. Hence
by utilising the stream energy more effective dust control systems can be designed.
It was also observed that the mechanics of the dump trucks do not work in favour of
dust-emission reduction due to the location of the tipping component hinges, which
causes a lateral movement in the free-falling stream of material. However, this is
considered as a mechanically efficient design as it requires a smaller force to dump the
material.
137
Application of the AVCF Theory in More Complex Mining Applications
Chapter 7: Application of the AVCF Theory in More

Complex Mining Applications
7.1 Background
Dump hoppers are one of the widely used pieces of equipment across a range of mining
applications. They are primarily designed to assist in material discharging processes and
transfer of material to desired locations within industrial sites. This process is known to
be one of the major dust-generation sources across mine sites.
To expand more on theory aspect of dust generation in transient applications, the author
selected dump hopper due to popularity in industry and the significant role they play in
material handling process. To achieve this objective, a scale model of an existing design
was employed and studied.
The author believes that dust emission can be reduced, or even prevented, if dump
hoppers are designed appropriately. To reduce dust emissions, the designer must
understand the mechanism by which the dust is re-entrained into the particle stream
within the hopper enclosure. By reviewing the available standards and designs, it can be
concluded that the design procedure for such systems is still quite empirical. Moreover,
the basic flow properties of the bulk material and the mechanism by which the dust
particles interact with the air, are being ignored. Dump hoppers are usually designed
with minimum attention to dust emission and after installation, the dust emission topic
is dealt with as a trial-and-error process, which is very high in cost and limited in terms
of the performance achieved.
Because of the many variables involved in the dump-hopper design process, a scientific
approach must be undertaken to determine the optimum solution. The solution not only
has to guarantee a reasonable material flow but aim to reduce the dust emission, which
is the key incentive of this project.
138
One key knowledge area of importance to improve the operation performance of dump
hoppers is understanding the possible particle trajectory from the delivery device, in this
situation a haul truck.
In this chapter, the AVCF theory (developed by the author on the free-falling air-
entrainment) is extended to dump hopper applications; a new theoretical model is
developed to calculate the magnitude of the air entrained into a stream of material
dumped out of a dump truck. The principles of this example can be applied to any other
applications with a similar process.
7.2 Review of Particle Trajectory Theory
There are a number of trajectory models available in the literature, which determines
the discharge angle of a material stream. The majority of these theories were developed
for high-speed situations such as conveyor-belt cases.
Figure 65. Dump Truck Bulk Material Stream Trajectory in Discharge Condition
139
and are the initial velocities of the underside and topside geometries and ∝ is the
discharge angle of the bulk material.
Assuming that the motion of the object trajectory is being measured from a free-falling
frame which starts at (x,y)=(0,0) at t=0 (drop height H) , the equation of the particle
trajectory is as per Equation (97):
= tan(∝) (97)
In addition, it is known that:
(98)
=−
2
Given that, the vertical component of the velocity is given by = Cos(∝) then:
=− ( ) (99)
2
The motion of a particle with an initial velocity of Vo, neglecting the drag force, can be
defined as (K. Golka, Bolliger, & Vasili, 2006):
= tan - (100)
140
7.3 Material Stream Discharging from Haulage Truck - Air

Entrainment Theory Work
In Chapter 4:, a new theory was introduced and verified to for measurement of free-
falling air-entrainment. In this section, the same principle is modified and applied to the
dump hoppers. By adopting the volume-conservation theory, the magnitude of air
entrainment into the stream of discharging material from a dump hopper can be
calculated.
Key parameters of the interest of the previously developed theory in relation to air
entrainment of free-falling stream were stream volume and stream velocity. It was also
discussed that the stream diameter had a significant impact on the results. Hence, a better
understanding of the possible particle trajectory is critical to be able to calculate the
volume of air entrained into the stream of discharging bulk material from a haul truck.
There is limited research available in relation to the discharge trajectory of bulk material
from a haul truck. However a number of studies have been undertaken regarding the
discharge trajectory from a belt conveyor under different conditions; these works can be
tailored to suit the dump truck discharge trajectory condition (Hastie & Wypych, 2007).
For granular bulk material discharge (as per schematic view represented in Figure 66)
without cohesion and adhesion in a vacuum, the parabolic trajectory of particles is
described as follows (K Golka, 1992) :
=v cos (101)
= tan -
( ) (102)
141
=v cos + (103)
= tan -
( )
+ (104)
Where represents divergence coefficient for the underside and topside limits of the
trajectory paths. Assuming and are the initial velocities, and H and H- are the drop
heights of the underside and topside geometries as shown in Figure 66 we have:
= + sin >0 > 0 t=T (105)
− = − sin >0 <0 t=T (106)
To calculate the stream volume , firstly, the areas under curves y and y , which
are represented by A and B, respectively, are calculated and deducted from each
other; then the subtracted value is multiplied by the nominal width of the stream W
(which is assumed constant along the falling trajectories). The calculation details
based on Equations (102) to (104) are as follows:
= . = ( tan −
( )
). (107)
A= tan −
( )
, = (108)
142
The falling time for the particles travelling trajectory y can be calculated using
Equation (105) with = :
( ) (109)
=
Now by substituting in Equation (101), we have as the final horizontal

displacement of trajectory y as follows:
+ (110)
= (− sin ( sin ) +2 )
−
The Equations (101) and (102) represent the trajectory y ; by applying integral to
trajectory y , the area B (under trajectory y ) is calculated as shown in Equations (107)
and (108), by substituting = .
Now the same approach will be applied to y using Equation (104), as follows:
(1 + ) (111)
= . = ( tan − − ).
2
B= tan −
( )
− . , = (112)
Where, is the final horizontal displacement of trajectory y which can be

calculated by finding (falling time for trajectory y ) from Equation (106) with
= :
143
+ (113)
− sin ( sin ) −2 ( − )
= −
By substituting in Equation (103), we have as the final horizontal

displacement of trajectory y as follows:
. + (114)
= (− sin ( sin ) −2 ( − ) +
−
The Equations (103) and (104) represent the trajectory y ; by applying integral to
trajectory y , the area A (under trajectory y ) is calculated as shown in Equations (111)
(112), by substituting = .
By deducting A (Equation (108)) from B (Equation (112)), the area of stream can be
calculated. V (Volume of the Stream) can be calculated using W (nominal width of the
stream):
(B-A) x W = V (115)
the volume of solids is now deducted from to calculate the volume of air as
per Equation (116). The entrained-air flow rate Q can be calculated using as the
volume of air that is displaced over the entire free-falling time as shown in Equation
(117). The free-falling time is considered as t = max( T , T ).
144
V (Vol. of material) – V (Vol. of Solids) = (Vol. of the air) (116)
V (117)
Q =
t
Figure 66. Dump Truck Schematic View
7.4 Example Case
To test the accuracy of the proposed formula, an example case has been defined and
calculated in line with the Iron ore sample case in Chapter 6: with bulk and solid density
of 2,166 and 4,356 , respectively.
145
By employing Equations (101), (102), (103) and (104), four key parameters ( ,
and ) are plotted in an Excel spreadsheet based on the given time-step increments,
up to the time in which and become equal to the desired drop height ( = =
). It is clear that the two limit curves of the stream ( and ) go along the two
different trajectories and hence the timing to reach the ground is different. The
parameters considered in this calculation are as follows:
Parameters Value Units

0.05 m/s
10 degree
10 -
0.25 m/s
15 degree
1e-6 -
ῤ 2,100 kg/m3
0 m
0.04 m
H (Drop Height) 1.2 m
Stream Width 0.8 m
The shape of the stream can be modified to model any type of material stream from a
dump hopper by using and factors. Then, as per the steps provided within the
introduced theory work, Equations (108) and (112) are utilised to calculate the area
under the curve for both and at the designated drop height (H), which in this case
is 1.2m, and deducted from each other. The shape of the stream resulting from case
study is shown in Figure 67.
146
Dump Hopper Trajectory

(m)
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
0.11
0.20000
0.00000
-0.20000
-0.40000
(m)
-0.60000
-0.80000
-1.00000
-1.20000
Y1 Y2
Figure 67. Stream Trajectory – Proposed Model
The thickness of the stream at inlet is considered as 40mm (as discussed in Chapter 6:),
and drop height is set at 1200mm. In relation to stream thickness, we have measured the
stream diameter at the lower end of the stream from two sources: simulation and footage
of experiments. Figure 68 represents results from the simulation, in which, the width of
the stream at the lower end varies between 80mm and 150mm. After reviewing the
simulation results utilising Iso-surface, based on particle volume fraction, it was
concluded that the core of the stream can be considered to be 100mm thick
(encompassing ~90% of the stream), which was utilised to adjust the theoretical
147
calculation (Iso-surface Volume Fraction of Particles > 0.0035). Therefore, in analytical

model calculation, the and factors are adjusted in a way that the same stream
thickness is obtained. Iso-surface is one of the ANSYS software post processing options,
which allows user to create a “cloud” showing where a given value for a parameter
exists.
Figure 68. Stream Width CFD Simulation – Top View
By adopting the introduced theory work based on AVCF and utilising the case study
parameters and criteria (which were selected as per one of the simulations carried out
and verified in Chapter 6:), the entrained air-flow rate is calculated at 83.7 lit/sec.
148
In relation to the sensitivity analysis of the introduced theory, it has been observed that
and have the maximum impact on the entrained-air readings. These factors have
been adjusted based on the properties of the stream to provide the measured stream
width at the impact point (Visual Comparison between Figure 67 and Figure 68).
The next key influencing parameter is the Initial Velocity. Therefore, it is important to
measure and input the initial velocity accurately. A combination of Initial Velocity,
Stream Thickness and and can model any types of stream trajectory.
7.5 Comparison between the Introduced Theory Work and

Simulation
As stated earlier in this chapter, to measure the performance of the provided theory, the
verified simulation physics in Chapter 6: was used. The enclosure is converted to a box
with all sidewalls defined as “No Slip Walls”. Moreover, the top face is defined as the
opening and the bottom face is defined as the outlet. The inlet remained in the same
location as for the previous model.
As the mesh-sensitivity analysis has been undertaken within the dump hopper analysis,
the mesh sizing and other simulation physics were kept identical to obtain consistent
results.
Furthermore, as simulation was verified by experimental results within Chapter 6:, the
readings from this simulation are used as verification benchmark for the proposed theory
works in this Chapter.
As per the information obtained from simulation during 1.7s (equivalent of real time) of
modelling, the particle velocity within lateral parts of the stream is higher than in the
dump hopper case. The angled sidewalls of the dump hopper are effective in reducing
the stream-velocity, due to the impact of particles with walls and frictional velocity
149
reduction (see Figure 69 and Figure 70); hence, this will be working in favour of dust
suppression.
Figure 69. Dump Hopper Iron Ore Velocity
The actual simulation time took six consecutive days using a system equipped with an
i7 CPU and 24GB of RAM. Moreover, a Graphic card was utilised (GPU) to share the
load between available resources in the machine. This process provided results
equivalent to 1.7sec of real-time experiments.
The volume of air entrained into the stream of bulk material at 1200mm drop height was
measured at 88.3 lit/sec on defined Iso-surface as per the simulation. As discussed in
detail in the Example Case, by employing the proposed theoretical model, this value
was calculated at 83.7 lit/sec. This represents a reasonable correlation between the
verified simulation and the proposed theoretical model.
150
Figure 70. Simulation of Stream into a Box
7.6 Conclusion
A new theoretical work has been proposed which can model the boundary layers of a
free- falling stream from a dump truck. This model can be used for any other free-falling
case with initial velocity from an angled inlet. This model was then extended to provide
a theoretical model for calculating the volume of air entrained into a stream of material
falling from a dump truck, which is done for the first time, to the author’s best
knowledge.
A verified simulation physics discussed in the previous chapter has been utilised to
verify the proposed theoretical model. In terms of the performance of the model, a close
correlation between verified simulation results and the proposed theory has been
observed.
151
Furthermore, it has been observed that the particle velocity in the dump hopper was
lower than that of free-falling particles, due to frictional velocity reduction of particles
within the dump hopper sidewalls. This is mainly due to collision of particles with walls
and inter-particle collisions.
The proposed theoretical model is an extension to the theoretical free-falling model

proposed in Chapter 4:, which enables analysis on more complicated bulk material
cases. This model is a mathematical model, which can be extended to any other bulk-
material handling case with simular principles.
The findings of this chapter, can assist designers to understand the magnitude of the
entrained air in to the bulk-material stream (free-falling with initial velocity) to be able
to employ appropriate mitigation methods as necessary.
152
Conclusions and Recommendations
Chapter 8: Conclusions and Recommendations

The work presented expands the understanding of the underlying mechanics associated
with passive dust control. Through a considered program of study, an extension of the
understanding and knowledge around free-falling of granular bulk material has been
achieved using a combination of experimental and numerical studies. In addition, the
presented works reviewed the mechanics of dust generation in dump hopper and
undertook physical testing of several designs to reduce the apparent liberated dust. All
of this has resulted in a better understanding of how dust emission from transient flow
(dump hoppers) can be better controlled. From the work carried out, this author believes
that these design concepts can make a substantial impact on the amount of liberated dust
in these situations.
In this work, a comprehensive review across different dust-generation influencing

parameters, with a focus on air entrainment as a principal driver, were provided.
As discussed in detail, during free-fall the air surrounding falling stream of bulk material
is entrained into that stream and upon compaction of the stream will be exhausted,
carrying dust particles with it. Since entrained air has been identified as one of the key
sources of dust generation, a major focus was placed on quantifying the degree of
entrained air, which will assist in the design of more efficient and environmentally
friendly equipment.
Studies of air entrained into a stream of gravity-falling material have been divided into
two sections. Substantial attention has been placed on the first of these, free-falling
material flow properties and influencing parameters, which have been analysed in detail.
The second section of study predominantly focuses on practical and more complicated
material-handling cases, transient flow situations an example of which is a dump hopper,
by employing the knowledge obtained from the work on free-falling.
153
Based on the work presented in this thesis, the following conclusions are made and
recommendations are offered as to where the author believes further research may be
directed.
8.1 Thesis Conclusion
Handling of bulk materials is a necessary aspect of any process utilising granular

products. Liberation of dust and fine particles represent both an environmental cost and
a cost from loss of product and additional cleaning needs.
In general terms, free-falling is one of the inseparable fragments of bulk material

handling process. Therefore, the research was initiated by studying the flow behaviour
of granular materials in free-falling condition. In this thesis, a new approach for
measuring the stream diameter of free-falling material using an image analysis
technique (by employing high-speed camera and MATLAB image processing
technique) was discussed and provided which proved to be significantly accurate. As
part of this research work, a new theory was proposed to quantify the volumetric flow
rate of entrained air into a stream of free-falling granular material by utilising the
proposed technique for stream diameter measurement. A comprehensive test rig (Figure
22) was developed and a range of tests were undertaken to measure the volumetric flow
rate of air entrained into a stream of free-falling particles. The results presented in
Section 4.3 proved to be more accurate than the existing models across the performed
tests.
Computation Fluid Dynamic simulations were carried on the proposed free-falling air
entrainment analytical model; the results were compared to the experimental results. In
addition, Particle Imagining Velocimetry technique was also adopted to provide
accurate experimental measurements of velocity and trajectory of air. The accuracy of
the simulation models was verified based on particle and air velocity readings (obtained
from Particle Imagining Velocimetry). Then, the air entrainment results from CFD were
154
compared with the experimental results and the AVCF theory (proposed analytical
mode) results, which indicated a close correlation between them. Based on the
performance of the models and comparison with experimental results, a set of models
were determined as the best performing models for this nature of simulations; it can
assist passive dust control system designers in utilising CFD software for dust emission
purposes.
The knowledge gained from studies on free-falling was extended to transient bulk
material handling cases. The dump hopper was chosen as an example of a discrete bulk
material handling case. The performance of a dump hopper has a significant impact, not
only on the efficiency of the bulk material handling systems, but also on the level of
fugitive dust emitted across the bulk material handling plant. The proposed AVCF
theory (free-falling air-entrainment theory) was readjusted to the dump hopper case.
This theory was compared with simulation and experimental results, and it was
concluded that this model provides a good degree of accuracy.
The key findings from this research work, which may help to improve the design of dust
suppression equipment as far as dust emission mitigation is concerned are:
- A quantification of drop height and dust generation potential. This work has
demonstrated a significant reduction of air entrainment by limiting free-fall.
- Effective use of turbulences in complex geometries as demonstrated in dump
hopper studies
- Efficient use of enclosure volume to allow dust particles to settle in enclosures
as demonstrated in dump hopper studies
- A theoretical formula can be used for both passive and active system designs.
This model can provide the magnitude of the entrained air, which will assist the
passive and active systems designer to provision for appropriate dust mitigation
technique; this work will assist in selecting the appropriate dust control method
resulting in lower implementation and operational cost.
155
8.2 Recommendations for Future Work
The author believes there are four main areas that could be studied further.
8.2.1 Further Testing on the AVCF Theory in Free- Falling case
As discussed comprehensively within this work, the air entrained into a stream of bulk
material is known to be the main source of dust generation. Accurate measurement of
entrained air allows a designer to understand the air-flow magnitude, which is the key
to design the equipment more efficiently as far as the dust generation is concerned.
Although significant improvements in predicting and understanding of air entrainment
has been presented, an extension of this work to consider greater drop heights could be
undertaken to further verify the theory presented.
8.2.2 Further Testing on the Introduced Dump Hopper Entrained Air

Theory case study
Following successful correlation between results from free-falling air entrainment

theory and experiments, the theory was extended to transient bulk material handling
cases and hence dump hopper was chosen as a case study equipment. The theory was
tested and a good level of correlation with experimental and Computation Fluid
Dynamic model results were achieved. However, an interesting research topic is how to
extend this work other transient bulk material handling equipment cases with a similar
nature.
8.2.3 Further Simulation with Prediction of the Flow of Dust Particles
Within the current work, the capability of predicting the dump hopper performance (as
far as dust suppression is concerned), was verified according to the air velocity.
156
However, it would be beneficial if the dust particles could be directly modelled. To the
author’s knowledge, three strategies could possibly provide the solution to this problem.
As discussed within the presented work the Eulerian-Lagrangian model has the
capability to track the particles, hence a combination of the Eulerian-Lagrangian model
and the Eulerian-Eulerian model would be one of the viable solutions. The dust particles
can be tracked using the Eulerian-Lagrangian model, while the mainstream flows,
including both the particle and airflows, are modelled using the Eulerian-Eulerian model.
This method can only give an approximation of the dust particle distributions in a dump
hopper. It is worthwhile mentioning that this method is not suitable to predict the dust
flux, due to the high number of particles and the resulting computational cost.
Employing the Dense Discrete Phase Model (DDPM) could be the second option, which
accounts for the effect of disperse-phase volume fraction on the continuous phase in the
Discrete Phase Model (DPM). This method requires less computational resources
compared to the three-phase Eulerian-Eulerian model, as there is no need to define
classes to handle particle size distributions.
Two-way coupled DEM/CFD should soon be at a point of refinement to enable the

granular flow in dump hoppers (inclusive of the dust phase) simulation to occur. To the
author’s best knowledge, the open-source DEM software LIGGGHTS, together with the
open-source CFD-DEM coupling framework CFDEM, has the capability to run two way
coupled CFD/DEM to a certain level of complexity. However, given the complication
of this case, both software and hardware were incapable of solving the case.
157
Reference
ACGIH. (1992). Industrial Ventilation - A Manual of Recommended Practice (21st

Edition). Paper presented at the American Conference of Governmental Industrial
Hygien, Ohio USA.
Ansart, R., de Ryck, A., Dodds, J. A., Roudet, M., Fabre, D., & Charru, F. (2009). Dust
emission by powder handling: Comparison between numerical analysis and
experimental results. Powder Technology, 190(1-2), 274-281.
Ansart, R., Letourneau, J.-J., de Ryck, A., & Dodds, J. A. (2011). Dust emission by
powder handling: Influence of the hopper outlet on the dust plume. Powder
Technology, 212(3), 418-424.
Ansart, R., Ryck, A., & Dodds, J. A. (2009a). Dust emission in powder handling: Free
falling particle plume characterisation. Chemical Engineering Journal, 152(2-3),
415-420. doi:10.1016/j.cej.2009.04.070
Ansart, R., Ryck, A. d., & Dodds, J. A. (2009b). Dust emission in powder handling:
Free falling particle plume characterisation. Chemical Engineering Journal, 152(2-
3), 415-420.
Ansys13.0. (2010). Ansys Help System: Fluent.
Arastoopour, H., Pakdel, P., & Adewumi, M. (1990). Hydrodynamic analysis of dilute
gas—solids flow in a vertical pipe. Powder Technology, 62(2), 163-170.
doi:10.1016/0032-5910(90)80080-i
Australia Department of Sustainability. (2013). Air Quality Standards. Australia:

Environment.
Australian Coal Association. (2012). Black Coal Production.
Australian Standard. (2003). Environmental testing AS 60068.2.68: Standards Australia.
Australian Standard. (2009). Workplace atmospheres - Method for sampling and

gravimetric determination of respirable dust AS 2985. Australia: Standards Australia.
Australian Standard. (2012). Respiratory protective devices 1716: Standards Australia.
158
Bakker, A. (2002). Lecture 10 - Turbulence Models Applied Computational Fluid

Dynamics. Retrieved from PowerPoint.:
Batchelor, G. K. (2000). An Introduction to Fluid Dynamics: Cambridge University

Press.
British Standard. (1972). British Standard 1747 Methods for the Measurement of Air
Pollution, Part 2, Determination of Concentration of Suspended Matter 1747-5:
British Standard.
Brown, P. P., & Lawler, D. F. (2003). Sphere Drag and Settling Velocity Revisited.
Journal of Environment Engineering. doi:10.1061/~ASCE!0733-
9372~2003!129:3~222!
Cebeci, T. (2004). Turbulence Models and Their Application:Efficient Numerical

Methods With Computer Programs: Springer.
Cecala, A. B. (2010). Dust Control in Mineral Processing Operations from Office of

Mine Safety and Health Research
Chen, X. (2013). Experimental and Numerical Investigation of Dust Emission From

Transfer Chutes. (Doctor of Philosophy), The University of Newcastle.
Chen, X. L., Wheeler, C. A., McClean, R., Donohue, T. J., & Roberts, A. W. (2012).
Evaluation of transfer chute geometry on dust emission using experimental and CFD
simulation. International Journal of Mineral Processing, 10.
Chowdhury, S. J. (2007). A modified k-ε model of turbulence. . Journal of Mechanical

Engineering, 38, 4.
Clift, R., & Gauvin, W. H. (1970). The Motion of Particles in Turbulent Gas Streams.
Proc. Chemeca, 70(1), 14.
Cooper, P., & Arnold, P. C. (1995a). Air entrainment and dust generation from falling
stream of bulk material. Kona Powder and Particle, 13, 10.
Cooper, P., & Arnold, P. C. (1995b). Air Entrainment Processes and Dust Generation
from a Falling Stream of Bulk Material. Kona Powder and Particle(13), 125-134.
Ergun, S. (1952). Fluid flow through packed columns. Chemical Engineering Progress,
48(2), 89-94.
159
Esmaili, A. A., Donohue, T. J., Wheeler, C. A., McBride, W. M., & Roberts, A. W.
(2013). A New Approach for Calculating the Mass Flow Rate of Entrained Air in a
Freefalling Material Stream. Particulate Science and Technology, 31(3), 8.
doi:10.1080/02726351.2012.715617
Esmaili, A.A., Donohue, T. J., Wheeler, C. A., McBride, W. M., & Roberts, A. W.
(2015). On the analysis of a coarse particle free falling material stream. International
Journal of Mineral Processing, 142, 9.
Fluent 6.3. (2006). Ansys Fluent User Guide.
Geldart, D. (1990). Estimation of basic particle properties for use in fluid--particle

process calculations. Powder Technology, 60(1), 1-13. doi:10.1016/0032-
5910(90)80099-k
Ghosh, S., Hunt, J. C. R., & Phillips, J. C. (1993). Dynamic Turbulent Air-Flow in Drop
Driven Sprays. Applied Scientific Research, 51, 8.
Gidaspow, D. (1994). Multiphase Flow and Fluidization: Continuum and Kinetic

Theory Descriptions: Academic Press, Boston.
Gidaspow, D., Bezburuah, R., & Ding, J. (1992). Hydrodynamics of circulating

fluidized beds: Kinetic theory approach. Paper presented at the Conference: 7th
international conference on fluidization, , Gold Coast, Australia.
Golka, K. (1992). Discharge Trajectories of Bulk Solids. Paper presented at the 4th
International Conference on Bulk Materials, Storage, Handling and Transportation:
7th International Symposium on Freight Pipelines.
Golka, K., Bolliger, G., & Vasili, C. (2006). Belt Conveyors: Principles for Calculation
and Design: Vasili, Costa.
Haider, A., & Levenspiel, O. (1989). Drag coefficient and terminal velocity of spherical
and nonspherical particles. Powder Technology, 58(1), 63-70. doi:10.1016/0032-
5910(89)80008-7
Hamelmann, F., & Schmidt, E. (2003). Methods of Estimating the Dustiness of

Industrial Powders – A Review(21), 7-18. Retrieved from
Hastie, D. B., & Wypych, P. W. (2007). Conveyor trajectory discharge angles. Paper
presented at the nternational Conference for Bulk Materials Storage, Handling and
Transportation, Newcastle, Australia.
160
Hemeon, W. C. L. (1955). Plant and process ventilation (First Edition ed.). New York
The Industrial Press.
Herrmann and Evensen. (1994). Newly - Developed Combination Product Supression

in the Storage of Bulk Materia. Bulk Solid Handling, 14(1), 4.
International Union of Pure and Applied Chemistry. (1990). Glossary of atmospheric

chemistry terms. Pure and Applied Chemistry, 62(11), 2167-2219.
ISO. (1994). Air Quality - Particle Size Fraction Definitions for Health-related Sampling
(Vol. ISO Standard 7708). International Standardization Organization: International
Organization for Standardization (ISO), Geneva.
Jones, W. P., & Launder, B. E. (1972). The Prediction of Laminarization with a Two-
Equation Model of Turbulence. International Journal of Heat and Mass Transfer,
15, 301-314.
Junemann, R., & Holzhauer, P. (1992). The Reduction of Dust Emissions in Bulk
Handling Installations. Bulk Solids Handling, 12(2), 10.
Karamanev, D. G. (1996). EQUATIONS FOR CALCULATION OF THE TERMINAL

VELOCITY AND DRAG COEFFICIENT OF SOLID SPHERES AND GAS
BUBBLES. Chemical Engineering Communications, 147(1), 75 - 84.
Kissell, F. N. (2003). Handbook for Dust Control in Mining U.S. Department of Health
and Human Services, CDC/NIOSH Office of Mine Safety and Health Research
Lapple, C. E., & Shepherd, C. B. (1940). CALCULATION OF PARTICLE

TRAJECTORIES. Industrial & Engineering Chemistry, 32(5), 605-617.
doi:10.1021/ie50365a007
Launder, B. E., Reece, G. J., & Rodi, W. (1975). Progress in the Development of a
Reynolds-Stress Turbulent Closure. Jounral of Fluid Mechanics, 68(3), 537-566.
Liu, Z., Cooper, P., & Wypych, P. W. (2007). Experimental Investigation of Air
Entrainment in Free-Falling Particle Plumes. Particulate Science and Technology:
An International Journal, 25(4), 357-373.
Liu, Z. Q. (2003). Air entrainment in free falling bulk material. (Doctor of Philosophy),
University of Wollongong, University of Wollongong Thesis Collection.
161
Liu, Z. Q., Cooper, P., & Wypych, P. W. (2007). Experimental Investigation of Air
Entrainment in Free-Falling Particle Plumes. Particulate Science and Technology:
An International Journal, 25(4), 357 - 373.
Lun, C. K., Savage, S. B., Jeffrey, D. J., & Chepurniy, N. (1984). Kinetic theories for
granular flow: inelastic particles in Couette flow and slightly inelastic particles in a
general flowfield. Journal of Fluid Mechanics, 140, 223-256.
Mannino, D., Buist, A., Petty, T., Enright, P., & Redd, S. (2003). Lung function and
mortality in the United States: data from the First National Health and Nutrition
Examination Survey follow up study, Thorax 58, 388-393. Thorax, 58, 6.
Marc, A. E., Plinke, D., Holstein, L., Maryanne, D., Boundy, G., & Loffler, F. (1995).
Dust Generation from Handling Powders in Industry. American Industrial Hygiene
Association Journal, 56, 251-257.
Menter, F. R. (1994). Two-equation eddy-viscosity turbulence models for engineering

applications. AIAA Journal, 32(8), 1598-1605. doi:10.2514/3.12149
Mine Safety and Health Administration. (2001 Retrieved 7th February, 2012).
Retrieved from http://www.cdc.gov/msha/topics/ABLES/ables-description.html
Morrison, J. N. (1971). Controlling dust emissions at belt conveyor transfer points.

Trans.Soc. Mining Engineers,AIME, 47-53.
Ogata, K., Funatsu, K., & Tomita, Y. (2001). Experimental investigation of a free falling
powder jet and the air entrainment. Powder Technology, 115(1), 90-95.
Ogata, K., & Tomita, Y. (2004). Experimental investigation of velocity profiles of free
falling powder Jet discharged from hopper. Paper presented at the The proceeding
og the 8th International Conference on Bulk Materials Handling and Transportation.
Pettyjohn, E. S., & Christiansen, E. B. (1948). Effect of Particle Shape on Free-Settling

Rates of Isometric Particles. Chemical Engineering Progress, 44(2), 157-173.
Plinke, M. A. E., Leith, D., Holstein, D. B., & Boundy, M. G. (1991).

EXPERIMENTAL EXAMINATION OF FACTORS THAT AFFECT DUST
GENERATION. American Industrial Hygiene Association Journal, 52(12), 521-
528.
Raof, R. A. A., Salleh, Z., Sahidan, S. I., Mashor, M. Y., Noor, S. S. M., Idris, F. M., &
Hasan, H. (2008, 12-14 Nov. 2008). Color thresholding method for image
162
segmentation algorithm of Ziehl-Neelsen sputum slide images. Paper presented at

the Electrical Engineering, Computing Science and Automatic Control, 2008. CCE
2008. 5th International Conference on.
Schaeffer, D. G. (1987). Instability in the evolution equations describing incompressible

granular flow. Journal of Differential Equations, 66(1), 19-50. doi:10.1016/0022-
0396(87)90038-6
Schiller, L., & Naumann, A. (1935). A drag coefficient correlation. Vdi Zeitung, 77(318),
51.
Soccol, C. R., Pandey, A., & Larroche, C. (2013). Fermentation Processes Engineering
in the Food Industry: CRC Press.
Swinderman, T., Goldbeck, L., & Marti, A. (2002). Foundations 3 - The Practical
Resource for Total Dust and Material Control.
Syamlal, M., & O'Brien, T. J. (1988). Simulation of granular layer inversion in liquid
fluidized beds. International Journal of Multiphase Flow, 14(4), 473-481.
doi:10.1016/0301-9322(88)90023-7
Tooker, G. E. (1985). "Establishing Design criteria for Fugitive Dust Collectio. Bulk
Solids Handling, 5, 5.
Tran-Cong, S., Gay, M., & Michaelides, E. E. (2004). Drag coefficients of irregularly
shaped particles. Powder Technology, 139(1), 21-32.
doi:10.1016/j.powtec.2003.10.002
Turton, R., & Clark, N. N. (1987). An explicit relationship to predict spherical particle
terminal velocity. Powder Technology, 53(2), 127-129. doi:10.1016/0032-
5910(87)85007-6
Turton, R., & Levenspiel, O. (1986). A short note on the drag correlation for spheres.
Powder Technology, 47(1), 83-86. doi:10.1016/0032-5910(86)80012-2
Uchiyama, T. (2004). Numerical analysis of particulate jet generated by free falling

particles. Powder Technology, 145(2), 123-130.
US Environmental Protection Agency (EPA). (1997). National Ambient Air Quality

Standards for Particulate Matter: USA, 40 CFR Part 50.
163
Wark, K., & Warner, C. F. (1981). Air Pollution - Its Design and Control. Collins New
York.
Weiss, M. (1992). A Practical Approach to Dust Control for Special Applications. Bulk
Solid Handling, 12, 4.
Wen, C. Y., & Yu, Y. H. (1966). Mechanics of fluidization. Chemical Engineering

Progress Symposium Series, 62(62), 100-111.
Wheeler, C., Krull, T., Roberts, A. W., & Wiche, S. (2007). Design of ship loading
chutes to reduce dust emissions. Process Safety Progress, 26(3), 229-234.
Wilcox, D. C. (1998). Turbulence Modeling for CFD: DCW Industries, Inc., La

Canada,California,.
Wilkinson, H. N. (1989). The Cost to U.K. Industry of Dust, Mess and Spillage in Bulk
Materials Handling Plants. Bulk Solid Handling, 9(1).
World Health Organization. (2005). WHO air quality guidelines global update 2005.
Retrieved from Bonn,Germany:
Wypych, P., Cook, D., & Cooper, P. (2005). Controlling dust emissions and explosion
hazards in powder handling plants. Chemical Engineering and Processing, 44(2),
323-326.
Yang, N., Wang, W., Ge, W., & Li, J. (2003). CFD simulation of concurrent-up gas–
solid flow in circulating fluidized beds with structure-dependent drag coefficient.
Chemical Engineering Journal, 96(1–3), 71-80. doi:10.1016/j.cej.2003.08.006
Zhang, Y., & Reese, J. M. (2003). The drag force in two-fluid models of gas–solid flows.
Chemical Engineering Science, 58(8), 1641-1644. doi:10.1016/s0009-
2509(02)00659-0
164

Thesis

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Thesis

Uploaded by

Copyright:

Available Formats

UNDERSTANDING PASSIVE CONTROL OF BULK

MATERIAL DUST EMISSIONS IN DISCRETE FEED

Master of Engineering Management

A thesis submitted in fulfilment of

School of Engineering - Department of Mechanical Engineering

The University of Newcastle

The University of Newcastle Scholarship (UNRSE ARC) must also be acknowledged as

Chapter 1: Introduction ............................................................................................. 4

Chapter 2: Dust Definition and Related Environmental Standards ......................... 9

2.2 Existing Dust Emission and Air Quality Standards ............................................ 12

2.3 Australian Standard (Australia Department of Sustainability, 2013) .................. 12

2.4 World Bank Guidelines (World Health Organization, 2005) .............................. 14

2.6 Measurement Equipment ................................................................................... 16

2.7 Dust Control Methods ....................................................................................... 21

Chapter 3: Free-Falling Streams .............................................................................. 26

3.2 Experimental Work on Stream Diameter ........................................................... 29

3.3 Analysis of Velocity and Stream Diameter Method ........................................... 37

3.5 Results and Discussions .................................................................................... 41

3.6 Conclusion ........................................................................................................ 45

Chapter 4: Free-Falling Air Entrainment Measurement and Calculation ............. 47

4.2 Particle Motion under Gravity ........................................................................... 52

4.3 Development of an Entrained Air Calculation Method ...................................... 59

4.4 Experimental Set-up and Validation .................................................................. 63

4.5 Results and Discussions - Theory versus Experimental Data ............................. 66

4.6 Conclusion ........................................................................................................ 69

Chapter 5: Free-Falling – Air Entrainment Computer Simulation ........................ 70

5.2 Simulation Theory ............................................................................................. 73

5.3 PIV Experimental Testing ................................................................................. 89

5.4 Simulation of Free-Falling................................................................................. 92

5.5 Results .............................................................................................................. 99

5.6 Conclusion ...................................................................................................... 106

6.2 Model Simulation and CFD Simulation ........................................................... 122

6.3 Results ............................................................................................................ 127

6.4 Conclusions..................................................................................................... 136

Chapter 7: Application of the AVCF Theory in More Complex Mining Applications

7.2 Review of Particle Trajectory Theory .............................................................. 139

7.4 Example Case ................................................................................................. 145

7.6 Conclusion ...................................................................................................... 151

Chapter 8: Conclusions and Recommendations .................................................... 153

8.2 Recommendations for Future Work ................................................................. 156

Figure 58. Cross-Sectional Pressure Profile ̶ Base Model ........................................ 129

A Cross-sectional area of bulk material stream [m2]

Q Volumetric flow rate of entrained air [Lit/Sec]

ρs Solid density [kg/m3]

Government regulation through various environmental pollution agencies is continually

Passive dust control in contrast, through an understanding of the mechanics of dust

To understand the impact of different parameters on the behaviour of a material stream,

The Office of Environment and Heritage (OEH) in NSW Australia, operates a

1.1 Thesis Structure

Chapter 3 Elaborates on free-falling stream diameter in detail. Existing approaches and

Chapter 4 Reviews existing air-entrainment models for free-falling streams. Moreover

Chapter 5 Presents the detail works being undertaken in relation to modelling/validating

Chapter 2: Dust Definition and Related

2.1 Dust - A Definition

Furthermore, based on Australian Standard AS 60068.2.68-2003, The Environmental

According to the Glossary of Atmospheric Chemistry Terms (International Union of

Figure 1. Sizes of Common Particles- (Wark & Warner, 1981)

Figure 2. Respirable Dust as defined by AS 2985

A moderate number of health-related researches have been conducted on the impact of

2.2 Existing Dust Emission and Air Quality Standards

2.3 Australian Standard (Australia Department of

Table 1. Australian Standards for Ambient Air Quality

(Australian Department of Sustainability 1998)