Parapari2021 - Effects of Loading Mechanisms and Texture On Ore Breakage A Multidimensional Study

L ICE N T IAT E T H E S I S
Parisa Semsari Parapari Effects of loading mechanisms and texture on ore breakage – A multidimensional study
Department of Civil, Environmental and Natural Resources Engineering
Division of Minerals and Metallurgical Engineering
ISSN 1402-1757 Effects of loading mechanisms

ISBN 978-91-7790-554-7 (print)
ISBN 978-91-7790-555-4 (pdf)
and texture on ore breakage –
Luleå University of Technology 2020
A multidimensional study
Parisa Semsari Parapari
Mineral Processing
Effects of loading mechanisms and
texture on ore breakage –
A multidimensional study
Luleå University of Technology

Department of Civil, Environmental and Natural Resources Engineering
Division of Minerals and Metallurgical Engineering
Printed by Luleå University of Technology, Graphic Production 2020
ISSN 1402-1757
ISBN 978-91-7790-554-7 (print)
ISBN 978-91-7790-555-4 (pdf)
Luleå 2020
www.ltu.se
Effects of loading mechanisms and texture on
ore breakage – A multidimensional study
Licentiate thesis in Mineral Processing

Division of Minerals and Metallurgical Engineering Luleå
Luleå University of Technology
SE-971 87 LULEÅ
Sweden
May 2020
Abstract
In comminution machines, the product properties (particle size distribution, mineral
liberation characteristics) and process consumables (energy for size reduction, wear) are
affected by various parameters. On the one hand, understanding and optimizing these
parameters can provide an energy efficient process and a specified product. On the other
hand, a fundamental understanding of the breakage process can even be used for designing
new or improved comminution machines.
In this thesis, breakage fundamentals are analyzed and set against the principles of various
comminution machines. The study of the breakage fundamentals is crucial for a better
understanding of the effect of different comminution environments on ore types and their
textures in order to achieve a desired product size and liberation. This work defines three
main areas of breakage processes with breakage fundamentals, namely “loading
mechanism”, “breakage mechanism” and “breakage mode”. The “loading mechanism” is
defined as the physical action that is applied to a particle or several particles in order to
introduce mechanical stress. The resulting pattern of the particle failure is named “breakage
mechanism”. Finally, the “breakage mode” defines the particle breakage in terms of being
random or non‐random. Non‐random breakage depends on the ore texture, which can be
categorized as preferential breakage and phase boundary breakage.
Promoting the breakage mode to the phase boundary breakage could help to increase the
liberation degree. Various studies have assessed the effect of ore texture and operational
parameters on mineral liberation. While ore texture is related to the particle inherent
characteristics, operational conditions such as loading mechanism are related to the
comminution environment. In all these investigations, little attempt has been made to
explore the combined effects of loading mechanism and quantitative ore texture features
on breakage mode and mineral liberation. In addition, a lack of fundamental understanding
of the breakage process and mineral liberation can be seen.
Accordingly, a more fundamental study of the causes behind the effects of loading
mechanism and texture is required in order to optimize the comminution process in terms
of mineral liberation. The objective of this work is, therefore, to investigate the effects of
different loading mechanisms on particle breakage and breakage mode. In order to achieve
this goal, work has started with using two methods including three‐dimensional
deformation and two‐dimensional crack quantification. The former method involved X‐ray
iii
computed micro‐tomography (XCT) imaging and Digital Volume Correlation (DVC)
measurements which determiners the breakage mode in terms of being random or non-
random. Whereas the latter was done using an image processing code in MATLAB to
quantify cracks in terms of random and non-random breakage (preferential or phase
boundary) from Scanning Electron Microscopy (SEM) images. In addition, XCT 3D
imaging was used in order to track the propagated cracks in the third dimension. Moreover,
phase boundary breakage in magnetite grains was studied qualitatively based on optical
microscopy images in order to identify and characterize the propagated cracks.
iv
Acknowledgments
First, I wish to thank my supervisors for their support and valuable contribution throughout
this research work. Professor Jan Rosenkranz for keeping this work on track through
constant reviews, and Dr. Mehdi Parian for the constructive discussions and constant
support.
The financial support from the Centre of Advanced Mining and Metallurgy (CAMM) at
Luleå University of Technology is gratefully appreciated.
Thanks to Henrik Lycksam and Fredrik Forsberg for helping with X-ray computed micro-
tomography scans; Thomas Aiglsperger for advising and guiding me on Scanning Electron
Microscopy analysis; Mattias Gustafsson from LKAB for sampling.
I would like to thank all my colleagues and friends from MiMeR division. I am also
thankful for the fruitful discussions that I had with Pratama Istiadi Guntoro and Dr. Saeed
Chehreh Chelgani.
I wish to thank my family; my amazing parents and sisters who offer invaluable support.
And, most of all for my supportive and loving husband, Ehsan.
Above all, I thank and praise God.
v
vi
List of papers
The thesis is based on the following papers:
1. Semsari Parapari, P., Parian, M., Rosenkranz, J., Breakage process of mineral
processing comminution machines – an approach to liberation, Manuscript
submitted to Advanced Powder Technology.
2. Semsari Parapari, P., Parian, M., Forsberg, F., Rosenkranz, J., Characterization of
ore texture crack formation and liberation by quantitative analyses of spatial
deformation, Manuscript submitted to Minerals Engineering.
The list of papers not included in this thesis:
1. Chelgani, S.C., Parian, M., Semsari Parapari, P., Ghorbani, Y. and Rosenkranz, J.,
2019. A comparative study on the effects of dry and wet grinding on mineral
flotation separation – A review. Journal of Materials Research and Technology,
Volume 8, Issue 5, September–October 2019, Pages 5004-5011.
2. Koch, P.H., Tiu, G., Semsari Parapari, P., Lishchuk, V., Parian, M., Ghorbani, Y.,
Rosenkranz, J., 2019. Textures in geometallurgy: from characterization to
applications. Manuscript submitted to Minerals Engineering Journal.
3. Semsari Parapari, P., Parian, M., Forsberg, F., Rosenkranz, J., 2020. Spatial
deformation of ore texture by quantitative characterization. Conference in minerals
engineering, Luleå, Sweden.
vii
viii
Contents
Part I ..................................................................................................................................... 1
Chapter 1 – Introduction ...................................................................................................... 3
1.1 Background ........................................................................................................... 3
1.2 Objectives .............................................................................................................. 4
Chapter 2 - Fundamentals of ore breakage .......................................................................... 7
2.1 Loading mechanism .............................................................................................. 7
2.2 Ore texture breakage behavior ............................................................................ 10
2.3 Breakage mechanism........................................................................................... 13
2.4 Breakage mode .................................................................................................... 16
Chapter 3 - Analysis of crack formation ............................................................................ 21
3.1 Sample characterization ...................................................................................... 21
3.2 Deformation measurements................................................................................. 23
3.3 Modified Brazilian test ........................................................................................ 25
3.4 In-situ loading along with XCT .......................................................................... 26
3.5 Crack quantification ............................................................................................ 28
Chapter 4 - Results and discussion .................................................................................... 29
4.1 Maximum principal strain variations .................................................................. 29
4.2 Crack quantification and characterization ........................................................... 32
4.2.1 Crack characterization ................................................................................. 32
4.2.2 Crack quantification ..................................................................................... 36
Chapter 5 - Summary and future work .............................................................................. 39
References .......................................................................................................................... 43
Part II – Papers ................................................................................................................... 55
ix
Part I
1
2
Chapter 1 – Introduction
1.1 Background
Nowadays, the mining industry is exploring and extracting more low-grade ores due to the
depletion of high-grade and coarse-grained ores. This trend has two consequences for the
minerals industry; first to process larger tonnages of mineral raw material and second to
grind the raw materials to finer sizes (Fuerstenau, 1995) which in turn consumes higher
energy. Comminution is an energy-intensive process; mineral raw materials and cement
clinker comminution processes consume 3% of the world-generated electricity (Deniz,
2013). The strategies to reduce energy consumption for mineral comminution are first to
avoid comminution where possible and second to use more efficient and suitable
comminution technology (Reichert et al., 2015). Consequently, an appropriate selection of
the comminution machine for a certain ore type leads to a specified product with the least
energy consumption. This selection is generally based on the feed material characterization
(Yokoyama and Inoue, 2007), final product specification (Chelgani et al., 2019), available
operational system (Mohd et al., 2017) and energy efficiency of the machines (Chen et al.,
2014). While having a proper selection is important, King (1993) pointed out that having a
fundamental understanding of the breakage process should even be considered in the design
of new comminution machines, which ultimately improves industrial comminution
processes. He proposed that recognizing the link between the micro-processes and the
particulate system dynamics is critical for understanding the breakage process (King, 1994,
1993).
Moreover, Bradshaw (2014) suggested ‘process mineralogy’ as a bridge for the ore
processing, geology, and metallurgy. She highlighted that the textural study of the ore
would help to design the process flowsheet. Indeed, it is not possible to consider all the
geological textural features; however, the dominant ore textures can be used as an indicator
for the ore. A reasonable approach to tackle this issue could be to study the ore texture in a
quantitative manner (Lamberg, 2011). Therefore, the two strategies for optimizing size
reduction units that should be further investigated in the future are: 1) optimizing the
available comminution technology with respect to the ore and its properties, i.e. ore texture,
and 2) introducing the ore texture into the design procedure for new comminution
machines. These strategies require a better understanding of comminution by analyzing
breakage fundamentals.
3
In comminution, loading mechanism, breakage mechanism, and breakage mode are the
parameters that control particle breakage. Loading mechanisms initiate the particle
breakage in the comminution process. Breakage mechanisms are influenced by the loading
mechanism along with the comminution environment and material properties (Hogg and
Cho, 2000; Little et al., 2016). Governing comminution environment parameters are
temperature, wet or dry processes, grinding media, etc. The main material properties are
mechanical strength, toughness, and brittleness. All of these factors can affect the product
with regard to particle shape (Celik and Oner, 2006; Little et al., 2016; Vizcarra et al.,
2011), particle size (Kotake et al., 2011; Ozcan and Benzer, 2013; Tavares, 2005), and
mineral liberation (Cao et al., 2019; Clarke and Wills, 1989; Hamid et al., 2018; Xiao et
al., 2012). While loading and breakage mechanisms are related to the fracture, the breakage
mode defines the minerals breakage in terms of being random or non-random. Random
breakage, i.e. breakage independent of ore texture, has a slight contribution to mineral
liberation at least in coarse size fractions. On the other hand, non-random breakage (i.e.
preferential phase breakage and phase boundary breakage) significantly increases the
liberation of the fractured phase (King and Schneider, 1998). Comparatively, the
contribution of phase boundary (or grain boundary) breakage to liberation is higher than
preferential phase breakage. Therefore, identifying and understanding the breakage mode
in comminution systems and its relation to ore texture and operating conditions is the major
step towards improving liberation.
While King and Schneider (1998) suggested that the breakage mode is affected by the ore
texture, other studies highlighted the influence of the loading and breakage mechanisms on
breakage mode (Celik and Oner, 2006; Garcia et al., 2009; Roufail, 2011; Xu et al., 2013).
It is of interest to note that, previous research findings of the effect of operational
parameters and ore texture on breakage mode and mineral liberation have been inconsistent
and contradictory (Garcia et al., 2009; Ozcan and Benzer, 2013; Vizcarra et al., 2010).
While it is well recognized that breakage mode and liberation can be affected by ore
characteristics (such as ore texture) or by operational conditions (such as loading rate),
surprisingly a lack of fundamental understanding of these factors is existing.
1.2 Objectives
This work focuses on designing and optimizing comminution processes with respect to
mineral liberation based on ore texture and process-related factors. Therefore, the effect of
loading displacement rate (i.e. a comminution process related factor) on the breakage mode
4
and liberation of iron ore textures is investigated. To measure the minerals breakage mode,
Parian et al. (2018) introduced a method based on the fragmented particles. According to
their work, the mineral liberation can be predicted based on its ore texture. In the current
work, ore texture cracks and deformations are used as indicators for fragmented particles
in comminution process.
Accordingly, the objectives of this licentiate study have been:
1. Review the literatures of ore texture breakage fundamentals.

2. Development of a method to analyze the breakage mode based on deformation and
crack characterization.
3. Investigate the breakage mode and liberation utilizing the loading rate and ore
texture.
To achieve these objectives, the following steps were undertaken in this study:
1. Screening the literature on breakage fundamentals, which ranges across the

spectrum of time and researchers to define (Paper I):
o Classification of areas of comminution in mineral processing where
breakage fundamentals have been used for process comprehension.
o Analysis of process-related factors and ore textural breakage behavior on
the product properties received in the comminution process.
2. Investigating the effects of process-related factors namely loading mechanisms on
ore texture breakage mode (Paper II):
o The rate of loading mechanism and ore textural parameters as controlling
parameter for breakage mode.
o Application of multidimensional deformation characterization and two-
dimensional crack quantification and characterization.
5
6
Chapter 2 - Fundamentals of ore breakage
In this chapter, a detailed explanation of breakage fundamentals is introduced and is mainly
divided into three main subcategories including loading mechanism, breakage mechanism,
and breakage mode. In addition, for having a better perspective of the breakage process,
the breakage behavior of materials under loading and the theories behind the breakage
process are discussed in brief. Lastly, the breakage mode is introduced in order to
understand the mechanisms for improving minerals liberation.
2.1 Loading mechanism

The loading mechanism is defined as the action of physical quantities, either force or
momentum, which is applied to a particle or several particles in order to introduce
mechanical stress (Andreatidis, 1995; Mütze and Husemann, 2008; Rumpf, 1965). These
physical quantities have a relative direction (Andreatidis, 1995) and rate (Mütze and
Husemann, 2008) compared to particle motion. When particles start to undergo loading,
potential energy is transformed into elastic energy (Evertsson, 2000). In brittle materials,
the elastic energy is stored in the crystal lattice and, as it reaches a critical value, a micro-
crack will be formed in the interatomic bonds (Anderson, 2012). In the cases where the
material has natural micro-cracks, the stress is accumulated in crack tips and fracture occurs
due to the coalescence of many micro-cracks (Kranz, 1983). A range of direction and rates
define the type of loading mechanism. Researchers use a variety of terms to define the
loading mechanism. The most common terms used for the loading mechanism in literature
are compression, impact, shear, attrition, and abrasion. While a variety of definitions for
the term loading mechanism have been suggested, this paper uses the definition with
respect to the loading type, loading rate, and the effective physical quantity. Figure 1
illustrates the type of loading mechanism in a comminution process for a single particle,
which can also be valid for multiple particles or a bed of particles. Generally based on the
loading direction, there are three types of loading including compression, impact, and shear
in comminution machines. Tensile stress affects the breakage process indirectly as an effect
of compression or impact loading (Wills and Atkinson, 1993).
7
Figure 1. Loading mechanisms in a comminution environment
In terms of loading rate, qualitative information suggested that a fast loading rate has an
instantaneous collision, while a slow rate breaks particle at slower rates. Therefore, the
impact is faster compared to compression and shear can be either slow or fast. In a recent
study, Saeidi et al. (2017) categorized the strain rate in three groups including static, quasi-
static, and dynamic. Figure 2 depicts the different loadings strain rates for various
comminution environments.
Figure 2. Strain rates associated with different types of loading – modified after Saeidi et
al. (2017)
8
A compression loading mechanism occurs if the particle is pressed between two media, two
tools or media and the comminution machine. For instance, Evertsson (2000) introduced
cone crushers as a compressing crusher in which the rock is broken by squeezing between
two surfaces. He calculated the working velocity in a specific type of cone crusher at choke-
level from 0.5-1 m/s.
In general, impact and strike loading mechanisms are caused by fast loading. The impact
loading mechanism according to the definition provided by Kelly and Spottiswood (1982),
is the rapid version of the loading mechanism. There is a degree of uncertainty around the
rate of impact and strike loading mechanism and the effective physical phenomena behind
it. However, with the help of simulation and modeling, it is now revealed that different
effective physical quantities in the comminution environment can occur at various impact
rates based on machine geometry and function (Bracey et al., 2016; Sinnott and Cleary,
2015; Wang et al., 2012). In comminution machines like tumbling mills, the loading rate is
controlled by the falling height of the grinding media and gravity (Wang et al., 2012). Two
mechanisms are taking place under the effect of the gravity force. The first mechanism
occurs as the particle falls and hits the comminution machine’s internal surface (impact
loading mechanism). As the second mechanism, another particle or media, for instance, the
grinding balls in a tumbling mill, falls onto the particle (strike loading mechanism). In
another case of impact loading mechanism, the blades of the comminution machine or
accelerated media, e.g. hammer mill, hit the particle. This collision imparts a momentum
to the particle and consequently increases the speed of the particle to up to 160 m/s (Bracey
et al., 2016). In this loading mechanism, each particle can hit other particles or the liner of
the comminution machine (Barley et al., 2004; Sinnott and Cleary, 2015).
Another effective loading type in a communion environment is shear. Shear loading results
in wear and shear stress, which generates the attrition loading mechanism. This loading
mechanism includes two states; (i) the particle is nipped between two media or the surfaces
of other particles, (ii) side collision of two particles as they pass by each other. The
consequence of shear loading is the attrition breakage mechanism. As it was mentioned,
the other introduced types of loading mechanisms in the literature are attrition and abrasion.
Attrition is the act of wearing or grinding down by physical quantities e.g. friction (Cleary
and Morrison, 2016; Inoue and Okaya, 1996) and abrasion is a consequence of attrition
(Cleary and Morrison, 2016; Frances et al., 2001; Inoue and Okaya, 1996; Little, 2016;
Rumpf, 1965). Therefore, based on this definition, attrition is the act (loading mechanism)
9
that causes abrasion (breakage mechanism). However, scientists have used abrasion as a
condition for the loading mechanism (Evertsson, 2000; Pourghahramani, 2012; Roufail and
Klein, 2010). Moreover, in some cases attrition has been introduced for particles hit
between two other particles or media and abrasion is as the particles are rubbing each other
(Lynch, 2015). In general, the definition of attrition and abrasion is closely related to the
shear loading mechanism.
Overall, variations of loading mechanisms occur in each comminution machine; however,

each machine has one or several dominant loading mechanisms (Evertsson, 2000; Frances
et al., 2001; Hogg and Cho, 2000; Schubert and Tomas, 2007). Moreover, some loading
mechanisms can have a higher probability of occurrence in crushing rather than grinding
and vice versa (Frances et al., 2001; Mütze and Husemann, 2008).
2.2 Ore texture breakage behavior

For mining valuable minerals, geological materials namely ores are the ones, which
undergo loading. Generally, for all materials, the particle breakage starts with crack
initiations and propagations, which in case of ore particles having a texture consequently
leads to mineral liberation. Overall, for all material types failure occurs when the applied
stress and work are sufficient to break the bonds that hold atoms together. The bond
strength is due to the attractive forces between atoms. The potential energy is at a minimum
where the equilibrium spacing occurs. In order to break the atomic bond, a tensile force is
required for increasing the separation distance from the equilibrium value; this force must
exceed the cohesive force to sever the bond completely (Figure 3).
Figure 3. Force as a function of atomic separation (x0 is equilibrium spacing) (Anderson,

2012)
10
Cohesive stress (σc) can be described by the following equation:
𝐸𝛾
𝜎𝑐 = √ (1)
𝑥0
where γ is the material surface energy per unit area i.e. material constant (J/m2) and E is
the Young modulus. However, the experimental work done on brittle material revealed that
the fracture strengths are usually three to four times or even more below this value. This is
due to inherent flaws (micro-defects or micro-cracks) in these materials (Anderson, 2012).
Based on the definition provided by Simmons and Richter (1976), micro-cracks in rocks
are the openings where one or two of their dimensions are significantly smaller than the
third. They added that in 2D micro-cracks, one dimension is significantly smaller and the
crack aspect ratio, i.e. the width to length ratio, must be less than 10-2 (usually 10-3 - 10-5).
Furthermore, the length of the micro-crack is typically less than 100 µm.
Rocks as geological materials mostly show brittle breakage behavior and commonly
contain micro-cracks (Kranz, 1983; Simmons and Richter, 1976). Micro-cracks are formed
when the localized stress goes beyond the local strength (Hamdi et al., 2015). They are
categorized into two groups i.e. natural and stress-induced micro-cracks (Nur and
Simmons, 1970). Natural micro-cracks are formed under pressure and temperature during
geological events. For instance, Nadan and Engelder (2009) classified quartz natural micro-
cracks in three categories: (1) healed micro-cracks often associated with fluid inclusion
planes that formed during isobaric cooling, (2) filled micro-cracks containing foreign
minerals, and (3) open micro-cracks that are produced during the isothermal decompression
process and tectonic activities. On the other side, stress-induced micro-cracks are formed
due to the man-made mechanical operation (Nur and Simmons, 1970). Natural and stress-
induced micro-cracks are the source of localized stress as the material undergoes
mechanical stresses. The critical stress level (σf) that a solid body with an induced crack
can sustain is:
2𝛾𝐸
𝜎𝑓 = √ ∗ (2)
𝜋𝑐
At constant c*, when σ (induced stress) exceeds σf, the specimen fails. Moreover, at
constant stress when c ˃ c* (c: crack length), the specimen fails by instantaneous nucleation.
By increasing the stress level, meso-cracks are formed due to the activating and growing
11
of micro-cracks and, finally, the development of fast macro-crack propagation is occuring,
which causes the sample failure (Figure 4).
Figure 4. Fracture process of brittle material under stress – modified after Dyskin and
Sahouryeh (1997)
In terms of rock and ore texture breakage, micro-cracks are not the only factor affecting
breakage behavior. All of the effective factors can be categorized in nano-scale (Prasher,
1987) and micro-scale (Hamdi et al., 2015) and are here referred to as ore texture
heterogeneity (Figure 5).
Figure 5. Ore texture heterogeneity

12
Ore texture heterogeneity at micro scale is studied due to its effect on rock strength and
stress-induced fracture patterns. For instance, it was shown that the compressive strength
of coarse-grained granite is roughly 10-20% less than for fine-grained granite (Martin and
Stimpson, 1994). Some definitions of ore textural heterogeneity are fully recognized in the
field of mineral processing and process mineralogy. Weak grain boundaries e.g. mineral
cleavage planes can be studied as a part of mineral heterogeneity. For instance, it was
shown that in rock breakage behavior, the existence of biotite and feldspar changes the
crack formation and rock breakage (Mahabadi et al., 2014; Nadan and Engelder, 2009). In
addition, the physical properties of mineral types combined with mineral association show
the evidence of natural micro-cracking (Nur and Simmons, 1970) (Figure 6).
Figure 6. Ore texture and its effect on heterogeneity
2.3 Breakage mechanism

Depending on the loading mechanisms in the comminution machines, particles face energy
transmission (Nematollahi, 2002; Wills and Atkinson, 1993). Griffith, as quoted by
Ozkahraman (2010) postulated that if the effective stress exceeds the local strength of the
material, the particle starts to break. The way a particle fails is referred to as the ‘breakage
mechanism’. Understanding of particle behavior under loading will help to recognize the
mechanism behind the breakage.
The breakage mechanism is often described based on particle breakage pattern, product
particle size distribution or particle shape (Table 1). Shattering, cleavage, chipping, and
13
abrasion are the most frequent terms that have been used for the type of breakage
mechanisms.
Table 1. Breakage mechanisms in a comminution environment
Feed particle Breakage pattern Product particles Breakage mechanism
Shattering
Cleavage
Chipping
Abrasion
Shattering
Shattering occurs when the loading is perpendicular to the particle and the applied energy
is a multiple of that required for the breakage. Therefore, the particle experiences a load
higher than the particle strength and than what is needed for single breakage (Kelly and
Spottiswood, 1982). King (2012) expressed that shattering involves multiple breakage
processes, in fact, a series of steps that the generated product particles undergo in an
immediate sequential of breakage. This process continues until all the energy is dissipated.
Berthiaux and Dodds (1999) have separated the shattering mechanism into two stages of a
sequential process. The first stage is the primary cleavage, where the particle breaks with
large fragmentation. While in the second step of the sequential breakage, the secondary
cleavage occurs if enough energy is still available to break the particle. Otherwise, if the
energy is not high enough for another destructive breakage event, fine particles will be
produced (Berthiaux and Dodds, 1999). In a similar study, Varinot et al. (1997) call the
shattering process “fracture” because of its intense stress, which leads to relatively large
particle sizes. Other definitions used for the shattering are destructive fracture and impact
fragmentation. Frequently, shattering is described as a process, which is caused by impact
14
and strike loading mechanism (Hennart et al., 2009; Leon, 2018; Little et al., 2017, 2016;
Palaniandy et al., 2008; Prasher, 1987; Varinot et al., 1997).
Cleavage
Cleavage is the breakage along preferred material surfaces i.e. the cracks move along the
cleavage planes in crystals (Wills and Atkinson, 1993). In mineralogy, cleavage is the
tendency of a mineral for being split along crystallographic planes. This tendency comes
from the structural locations of atoms and ions in a crystal, which create planes of relative
weakness (Haldar and Tišljar, 2014).
King (2012) describes cleavage as a type of breakage mechanism that generates several
large particles along with fine particles that originate from the points of applied stress. The
definition is similar to that found by Kelly and Spottiswood (1982) who outlined that
cleavage happens as the applied energy is sufficient for breaking a few regions of the
particle. However, Kelly and Spottiswood (1982) did not consider fine particles in their
description as King (2012) did. On the other hand, Little et al. (2017) stated that cleavage
is not a breakage mechanism but is related to material properties. It is of interest to note
that, a wide range of investigations categorized cleavage as a result of the compression
loading mechanism (Leon, 2018; Mariano, 2016; Ye et al., 2010a) and solely called it
compression. On the other hand, King (2012) stated that, similar to shattering, cleavage
takes place with the rapid imposition of compression stress. In general, some agreements
exist that the cleavage breakage mechanism occurs if the kinetic energy is enough to break
the relative weak planes, e.g. cleavage planes of the particle (Hennart et al., 2009; Kelly
and Spottiswood, 1982; King, 2012; Ye et al., 2010a).
Chipping
Particles with topographical properties or angular surfaces experience chipping breakage

(Little et al., 2017; Unland, 2007; Vizcarra et al., 2011). If the applied load on an irregular
particle is angular, rough corners can break off (Austin et al., 1986; Wills and Napier-
Munn, 2005). Contrary to the opinion of Austin et al. (1986) regarding the angular loading,
Fuerstenau and Han (2018) stated that the chipping breakage mechanism breaks off the
sharp edges and corners without a specific direction of applying load. This definition is
similar to those reported by Kelly and Spottiswood (1982) expressing that chipping is due
to the breaking of edges or corners of a particle. They considered chipping as a special case
15
of cleavage. The particle size distribution in this breakage mechanism is bimodal and the
size difference between the two peaks of size distribution is relatively small (Unland, 2007).
Abrasion
The attrition loading mechanism is one of the reasons that can cause abrasion. These terms
have been widely and interchangeably used in comminution studies. Abrasion is caused by
either the tangential load or insufficient energy of the applying load and the acting forces
are parallel to the surface (Wills and Napier-Munn, 2005). This definition is somehow used
in some other studies (Mariano, 2016; Roufail, 2011; Ye et al., 2010b) that defined abrasion
as an effect of scrubbing or trimming of a material. Furthermore, Kelly and Spottiswood
(1982) believed that abrasion takes place if the applied energy is not sufficient for causing
significant particle breakage. Palaniandy et al. (2008) stated that abrasion is characterized
by the formation of fines and no noticeable change in the feed particle mean diameter. The
latter definition was used by Hogg (1999) for modeling breakage and Little et al. (2017)
for investigating the particle shape after breakage.
2.4 Breakage mode

One of the main purposes of breakage in mineral comminution is to liberate the mineral
phases from each other. King (2012) attempted to explain mineral liberation based on the
breakage mode. In his work, the breakage mode has been expressed as random and non-
random breakage. The breakage mode is random if it is independent of mineral texture.
Oppositely, it is non-random breakage if it depends on the mineral texture. King (2012)
classified non-random breakage as selective breakage, differential breakage, preferential
breakage, phase boundary breakage, liberation by detachment, and boundary region
fracture. Recently, Little et al. (2016) suggested that the first three breakage modes should
be categorized as preferential breakage and the other three modes are phase boundary
breakage (Figure 7). Preferential breakage is used to refer to situations in which breakage
more frequently occurs in one mineral. Phase boundary breakage is a type of breakage in
which breakage is happening preferentially along boundaries rather than across the phases
(King and Schneider, 1998).
16
Intact ore Random Preferential Phase boundary
Figure 7. Intact ore and different types of breakage mode.
Another definition of breakage mode is based on the initiated and propagated cracks (Baud
et al., 2014; Brotóns et al., 2013; Eberhardt et al., 1999; Shea and Kronenberg, 1993).
Therefore, identifying the cracks and quantifying them would help to characterize the
cracks and ultimately breakage mode can be analyzed. On the other hand, based on micro-
crack size and detecting methods resolution, measuring the deformation can analyze the
final breakage mode in terms of being random or non-random.
Since breakage starts with the application of the loading mechanism, it is essential to
investigate the process related factors on breakage mode and mineral liberation. Promoting
the breakage mode to the phase boundary breakage could help to increase the liberation
degree. Moreover, since the ore texture has formed during geological deformations and has
been extracted through mining, having a broader perspective of texture combined with
quantitative descriptions would help to optimize the mineral liberation with respect to
process and ore texture.
Previous findings on the effect of loading mechanism type on mineral liberation have been
inconsistent and contradictory. For instance, some investigations have shown that the
loading mechanism has an effect on mineral liberation improvement (Garcia et al., 2009;
Ozcan and Benzer, 2013; Roufail, 2011; Xu et al., 2013). In contrast to that, other studies
have reported that the sample liberation degree does not depend on the comminution
machine type with its dominant loading mechanism (Andreatidis, 1995; Bérubé and
Marchand, 1984; Manlapig et al., 1985; Mariano and Evans, 2018; Vizcarra et al., 2010).
Bérubé and Marchand (1984) found that liberation depends on the textural characteristics
of minerals and can vary within the same ore body. However, they investigated only one
ore texture and did not consider other ore textures and their effect on liberation. Similarly,
Andreatidis (1995) showed that liberation is independent of the loading mechanism and
added that liberation is affected by the ore texture.
17
Table 2 summarized the results of the loading mechanisms on liberation improvement.
Based on this table, the compression loading mechanism is the most effective mechanism
for mineral liberation improvement.
Table 2. Summary of loading mechanism effects on mineral liberation
Impact (Roufail, 2011) (Apling and Bwalya, 1997) (Apling and Raissi,
2000)
Positive effect
Strike (Apling and Bwalya, 1997) (Han et al., 2012)

Compression (Celik and Oner, 2006) (Ozcan and Benzer, 2013) (Garcia et al.,
2009) (Xu et al., 2013) (Han et al., 2012)
Attrition
No effect (Vizcarra et al., 2010) (Mariano and Evans, 2018)
Attrition (Roufail, 2011)
Negative effect
Compression
Strike (Apling and Bwalya, 1997)
Impact
Based on the findings of the sources summarized in Table 2, there is a need for proper
studies of breakage mode and liberation:
1. Quantitative textural features have not been used in order to characterize the ore
texture. The influencing features of ore texture are mineralogy, grain size, and
minerals association (Parian et al., 2018).
2. The loading mechanisms should be investigated more systematically in terms of
energy intensity and strain rate. Each loading mechanism should be studied
separately through specific machines (Mwanga et al., 2015).
3. Little attempt has been made so far to explore the combined effects of loading
mechanism and ore texture.
4. Comminution of single, multiple (monolayer) or bed of (multi-layer) particles
with loading mechanism and textural effects should be investigated.
5. A more fundamental understanding of loading mechanism and texture effects is
required. This approach would help to understand the type of breakage mode (e.g.
phase boundary breakage) and its effect on mineral liberation.
18
With regard to ore texture, the effects can be studied following Figure 6. In this regard, for
instance, a combination of mineral type and mineral association shows the evidence of
natural micro-cracking. The main role of micro-cracks is to act as a stress concentrating
mechanism and are the source of the local stress concentration which ultimately causes
rock failure (Dyskin, 1999). The other effect of mineral type can be the minerals with weak
grain boundaries. For instance, Mahabadi et al. (2014) showed the effect of biotite due to
its cleavage planes on macro-crack formation and breakage. These ore textural features are
at the micro scale. However, as it was shown in Figure 5, ore textural heterogeneity
continues down to crystalline size. Therefore, further investigations of the crystalline
structure of ore texture should illuminate the effect of elemental association and distribution
in the breakage behavior of minerals.
19
20
Chapter 3 - Analysis of crack formation
In this chapter, a systematic analysis of the crack formation and propagation with respect
to phase boundary, preferential and random breakage mode is investigated. The approach
is having a better analysis of breakage mode in ore textures, which leads to optimizing the
loading mechanism with respect to mineral liberation. In order to achieve this goal, a
quantitative analysis of single ore texture particles was used. The identified ore textures
were tested with compression loading with different displacement rates followed by
quantitative analysis including three‐dimensional deformation and two‐dimensional crack
quantification (Figure 8).
Figure 8. The schematic representation of workflow
3.1 Sample characterization

Magnetite ore samples from Malmberget ore deposit located in Northern Sweden have been
used for this study. The samples were collected from the conveyor belts feeding to the
beneficiation plant. The collected samples were characterized at a macro- and micro-scale.
Figure 9 illustrates the characterization procedure. In the first step, the collected samples
were classified and separated based on similarity in terms of grain size and geological
features i.e. mineralization, alteration and veining. The separated samples were then studied
through reflected optical microscopy for quantifying the average grain size. In order to
prepare the samples for reflected optical microscopy studies, intact samples were molded
and polished. The magnetite grain size distribution was obtained by measuring a minimum
of 200 grains. In the next step, samples were discriminated into fine, medium and coarse
grain size according to their d80 (Table 3).
21
Figure 9. Feed ore characterization flowchart
Table 3. Range of d80s for defining fine, medium and coarse grains
Grain size d80 (µm)

Fine 250-350
Medium 500-750
Coarse 950-1050
At the final stage of characterization, two different textures were selected for XCT studies.
The selection criteria were the mineral proportions and the grain size of the minerals.
Samples contained both magnetite and gangue minerals with roughly equal proportions and
scattered distribution of minerals. Moreover, since the ore type may have hematite (Lund,
2013), only those samples without hematite were selected. As the density and attenuation
number of hematite and magnetite are roughly the same range, the two minerals would
otherwise be detected with the same gray level in XCT images even though having different
physical properties. On the other hand, a proper selection of the grain size had to be done.
Too fine grain sizes can lead to two problems. First, grains may not be distinguished at the
selected resolution. Second, the sample will have a high particle strength, which may not
deform evenly at high loading intensities (Hatzor and Palchik, 1997). A too coarse grain
size, on the other hand, may cause the following problems. First, in the selected sample
size for XCT studies, there will not be a sufficient number of grains for XCT studies.
Second, it has been shown that when grains get coarser, samples will break only along the
grain boundaries (Eberhardt et al., 1999). Hence, samples of two textures were selected,
named as (A) and (B). In order to identify the minerals present in the selected textures,
SEM-EDS analysis was done (Table 4).
22
Table 4. Selected samples for XCT
Sample Visual logging d80 Mineralogy
700 Magnetite, apatite, actinolite,

(A)
µm chlorite, quartz, biotite
950 Magnetite, apatite, actinolite,

(B)
µm chlorite, calcite
3.2 Deformation measurements

Based on its size, a micro-crack will be visible or invisible due to the resolution limit of the
measuring techniques, e.g. SEM, used for image analysis. In the cases where micro-cracks
are not visible, measuring the deformation around micro-cracks can be a way to investigate
the material failure. The deformation of a plane can be studied by Digital Image Correlation
(DIC) method (Pan et al., 2009). This method is an optical experimental method, which
measures the displacement and strain field in two dimensions using digital image
processing and numerical computing (Larsson et al., 2018; Pan et al., 2009). While optical
imaging gives information of two-dimensional space, there are other methods that give the
internal (volume) structure of a specimen (Buljac et al., 2018). X-ray computed micro-
tomography (XCT) is one of the techniques for scanning the internal features of an object
in 3D. X-rays passing through the material are attenuated according to the differences in
density and atomic composition of the internal phases. A detector then measures the
attenuation of the X-ray signals. The attenuated signals are reconstructed in virtual slices
with gray levels of brightness (Kyle and Ketcham, 2015) and by stacking multiple
sequential slices. By that, a 3D image is reconstructed. In this method, the denser phases
will attenuate more and in the gray brightness of slices of the reconstructed images, the
denser material will occur brighter.
3D measuring techniques can be used in order to measure the internal structural

deformation of the material (Maire and Withers, 2014). One of the methods to quantify
internal deformation and strain in a 3D volume is Digital Volume Correlation (DVC) which
23
is a non-destructive method based on a theory developed for DIC (Bay et al., 1999). In this
technique, it is possible to follow the non-destructive evolution or deformation of a material
by time-lapse 3D imaging at different stress levels (Yu et al., 2016). One of the advantages
of this method is that the 3D deformation field can be superimposed on top of a
microstructure that is the ore texture here. This is used for relating the heterogeneities in
strain to the underlying structure (Maire and Withers, 2014). The second advantage of DVC
is that it can reveal those deformations, which are not visible by internal three-dimensional
imaging methods even under high resolutions (Lenoir et al., 2007).
The measurements of 3D displacement and strain fields in DVC are based on the two steps
of volume images of a test sample before and after loading (Wang et al., 2018). The feature
for comparing the volumetric images is the grayscale variation. For having a proper DVC
measurement, the variation of gray level should be continuous in space (Maire and Withers,
2014). The measurements within DVC are based on two approaches, namely the so-called
local approach (using sub-volume or sub-regions or integration window) and the global
approach (nodal spacing) (Dall’Ara et al., 2017). In both approaches, the spatial resolution
is defined as the length scale of the displacement and strain measurements. In the local
approach, the size of the correlation window is used while the global approach is based on
twice the element size. Figure 10 illustrates the local approach for the image correlation
procedure. In this figure, f and g are the gray levels of the initial and deformed volumes,
respectively, and u(X) is the sought displacement field. X and X* refer to the coordinates
(in voxels) of the same point in the reference and the deformed state. The displacement
vector connects the center of the sub-volume and the point of the highest correlation.
Figure 10. The local approach for deformation measurements.
24
3.3 Modified Brazilian test
After the samples were characterized and the textures were classified and selected, samples
were prepared for a modified Brazilian test. The Brazilian test setup was used for the
following reasons:
 Due to the small surface sharing of the particle and the platen, the friction force will
be less effective. Friction will cause the apparent compressive strength to vary inside
the sample (Obert et al., 1946).
 There will be less irregularity on the cylindrical specimen surface. In this case, the
applied stress will not accumulate on a specific part (Prasher, 1987).
 It is a method with indirect measurements of tensile strength and is used widely for
fracture behavior of brittle materials (Yuan et al., 2018).
Another advantage of the cylindrical shape samples is the full scan field for XCT studies
of circular fields of view (Ketcham and Carlson, 2001). Accordingly, a small-scale
Brazilian test procedure and apparatus fitting the XCT chamber were designed in this study.
The modified Brazilian test along with XCT can be used for studying the structural
deformation (in combination with DVC measurements) of minerals within each texture. To
perform the Brazilian test, a cylindrically shaped sample was drilled out of the
characterized samples. The sample size was selected with the same height and width of 7
mm due to the restrictions for XCT with respect to sample size and resolution. The sample
was handled carefully to reduce the stress-induced micro-cracks during the preparation
procedure. The samples were then polished with a polishing wheel to have parallel planes
at the end faces of the specimen with a smooth edge. The moisture content of the sample
was as received.
For conducting the in-situ Brazilian test, an apparatus with a suitable sample holder was
needed. Criteria for selecting the proper sample holder material were to have a (i) light
density with (ii) low plastic behavior. For preventing the plastic deformation inside the
apparatus, it was recommended to use steel (Bieniawski and Hawkes, 1978). However, the
X-ray signal attenuation of steel is high. Considering the above criteria, two materials were
selected: Polyoxymethylene (POM) with an average density of 1.41 g/cc (Pious and
Thomas, 2016) and aluminum with a density of 2.7 g/cc. POM is usually used for the
production of engineering components due to its rigidity under loading (Pious and Thomas,
2016). Aluminum has a good strength to weight ratio (Claisse, 2016). Finally, an aluminum
25
apparatus was chosen due to difficulties in POM machining. Preliminary experiments were
done on samples with and without the aluminum apparatus (A-A) (Table 5). The results
showed that the amount of aluminum extension is small enough after 60 minutes to allow
neglecting the stress relaxation.
Table 5. Sample extension under quasi-static loading
Loading (N) With A-A (mm) Without A-A (mm) Differentiate (mm)
500 0.085 0.080 0.005
1000 0.125 0.111 0.014
Another criterion for the apparatus design was its proper size setting. Two apparatus were
manufactured with two different size settings. One apparatus was designed according to the
procedure recommended in Bieniawski and Hawkes (1978) where the jar radius was 1.5
times larger than the sample. The second apparatus was prepared with a radius twice the
sample size. Preliminary studies showed that the sample started rotating inside the second
apparatus as the loading was applied and consequently the stress could not be properly
transferred into the sample. However, for the sample tested in the first apparatus, the stress
was well distributed and the cracks were mostly following the loading direction. Therefore,
an apparatus with aluminum and a radius of 1.5 times more than the sample diameter was
used for the in-situ loading (Figure 11).
Figure 11. Brazilian test apparatus
3.4 In-situ loading along with XCT

The 3D scanning was carried out using a Zeiss Xradia 510 Versa 3D microscope system.
The scanning voltage was 140 kV, the power was 10 W, and the scanning current was 71.5
µA. A CCD (charge-coupled-device) detector was used which included an image
intensifier. For the detector, an exposure time of 1 sec with a binning number of 2 was
26
utilized. The spatial resolution (i.e. the voxel size in the reconstructed slice) was 20 × 20 ×
20 µm3. The field of view (i.e. the beam section size) was 20 mm × 20 mm to cover the
specimen and the Brazilian apparatus with an objective of 4X. For having comparable scan
results, the scanning parameters were consistent in each scan.
For carrying out the DVC measurements, an in-situ loading device (CT5000, Deben Ltd,
UK) was used along with the XCT and the Brazilian apparatus. The quasi-static
compression loading mechanism was applied at two different displacement rates (0.1 and
1 mm/min) at room temperature. The temperature was controlled during the process to
prevent a temperature effect on particle deformation. Three loading steps were used for all
textural types and loading displacement rates, including normal loads of 30 N, 500 N and
the load at the breakage point. In each step, the ore texture sample was loaded to the before-
mentioned force and scanned while the force remained constant (Figure 12).
Figure 12. Schematic drawing of applied loading while XCT scanning and the DVC
measurements superimposed to the particle
DVC measurements were performed on the obtained XCT datasets, using LaVision DVC
software (Davis, 8.4, LaVision GmbH, Göttingen, Germany). The local correlation
algorithm in LaVision is a Fast Fourier Transform cross-correlation, which correlates the
sub-volumes independently as a discrete function of gray levels. Each dataset was
correlated against the as-received condition (loading cycle 0) with two different settings
(Table 6).
27
Table 6. The input setting for correlation
Setting Loading Overlap Peak

Sub-volume Binning Passes
Number cycle (%) search
(0) 128 50 16 8×8×8 2
(1) 96 50 8 4×4×4 2
(1)
(2) 64 50 4 2×2×2 2
(3) 64 75 2 None 2
(0) 128 50 16 8×8×8 2
(1) 64 50 8 4×4×4 2
(2)
(2) 32 50 4 2×2×2 2
(3) 32 75 2 None 2
3.5 Crack quantification

In this study, crack quantification based on 2D SEM images was done in MATLAB. For
this purpose, high-resolution 2D images were acquired by SEM coupled with energy-
dispersive X-ray spectroscopy (EDS) from a carbon-coated polished section. The SEM
analysis was carried out with a Zeiss Sigma 300 VP with Bruker Instrument EDS detector
and Mineralogic Mining software with a resolution of 20 μm per pixel and a field of view
of about 7 × 7 mm per image. A summary of the image preparation and analysis
methodology in MATLAB is given below:
1. Gray level SEM images were binarized for discriminating cracks.

2. Morphological processing was applied to the crack binary images. Thinning
processing was used in order to “skeletonize” the cracks and was iterated until there
was no more modification. Topological properties (number of holes and
connections) were conserved in this process (Lam et al., 1992; Pratt, 2001).
3. By combination of the cracks’ binary image and SEM backscattered electron (BSE)
image, the crack width and surrounding phases were calculated and recorded.
An example for SEM image processing and the resulting network of cracks obtained by the
steps described above is presented in Figure 13.
BSE Binarized cracks Morphed cracks
Figure 13. Steps for 2D crack quantification.

28
Chapter 4 - Results and discussion
In this study, the maximum principal strain was used since rock is a brittle material and the
maximum principal stress is an appropriate measure for assessing damage initiation
(Meisels et al., 2015; Toifl et al., 2015). The maximum principal strain indicates maximum
principal stress, which is the quantity of normal stress calculated at an angle when shear
stress is considered as zero. While shear stress is the reason for ductile material failure, in
brittle material failure is due to the normal stress.
In order to see the effect of mineralogy on the characteristic of cracks, SEM images, and
optical microscopy images were used. Figure 14 shows an example of the selected slices
for texture (B) - 0.1 mm/min sample; (a), (b), and (c) show subsets obtained by cuts in the
y-z plane through the specimen at several x-positions, namely 1.4 mm, 2.12 mm, and 5.67
mm. The criterion for subsets selection was based on the selected slices of 2D images for
optical microscopy and SEM imaging. Figure 14 was generated using Dragonfly software,
Version 4.1 for Windows (Object Research Systems (ORS) Inc, Montreal, Canada, 2018;
software available at http://www.theobjects.com/dragonfly).
(1) (2) (3)

Figure 14. Selected sections for analyzing the maximum principal strain in 3D
Cracks were characterized qualitatively and quantitatively. With the help of three-
dimensional images and selected slices for two-dimensional images, the reported
observations can be tracked in the whole specimen. In order to study cracks quantitatively,
2D images of SEM analysis were utilized. For investigating the magnetite breakage mode,
optical microscopy images were used.
4.1 Maximum principal strain variations

For measuring the strains and deformations, the assumption is micro-cracks are equally
distributed without any specific orientation. Figure 15 shows the variations of the maximum
principal strain for 500 N in the samples. According to this image, in both textural types
and loading displacement rates, at 500 N, both tensile and compressive strain are observed.
29
The strain distribution in brittle material before its failure is consistent with measurements
obtained in previous studies (Melenka and Carey, 2015; Wu et al., 2019). It seems possible
that this type of strain distribution before material failure is due to the possibility of the
crack closure stage of rock breakage (Peng et al., 2015). As the applying force is increased
up to the breakage point of the particle, the maximum principal strain is mostly tensile as
it was expected (Mahabadi et al., 2014; Wills and Atkinson, 1993) (Figure 16).
Sample DR* Scale

Strain mapping
code (mm/min) bar
Slice (1) Slice (2) Slice (3)
0.1
(A)
0.1
(B)
Figure 15. Strain mapping and cross-correlation section of 3D. *DR stands for
displacement rate. In the scale bar, negative values indicate compressive strain, 0 to 0.001
does have roughly no deformation. Positive values indicate tensile strain.
In order to compare the effect of loading displacement rate and textural type on strain type
and its distribution, sections of the 3D images of the core were used (Figure 15). By
increasing the loading displacement rate, the magnitude of tensile strain drastically
increased in texture (A). This drastic strain difference gives a possible expectation of higher
deformation of inherent micro-cracks under the same time circumstances. In addition, it
was shown that the failure of a brittle material under compression involves a large number
of micro-cracks. However, in the tensile case, the failure occurs with only a few micro-
cracks (Kaplan, 1963). Thus, the probability of particle failure will increase in locations
30
where the tensile strain is accumulated. In this study, the micro-cracks, which are involved
in particle failure under tensile stress, are named “effective micro-cracks”. It is possible
that all the micro-cracks and micro-defects are not visible due to the resolution limit of the
device. As a result of higher tensile strain, the deformation may occur faster. This finding
is consistent with that of Koch (2017) who pointed out that this theory that if the speed of
crack propagation is slow enough, the tip of the crack can adapt accordingly with the
upcoming grains. However, if the crack propagation occurs rapidly, the adaption will not
occur, resulting in crack propagation independent of the medium ahead of the tip. It is
possible, therefore, that the incremental increase of effective micro-cracks combined with
the theory of the less micro-crack adaption and selectivity with a higher crack inducing rate
cause random breakage mode. The probability for random breakage mode is higher than
for the non-random breakage mode (i.e. phase boundary breakage and preferential
breakage). This finding is similar to a study that was set up to determine the effect of
displacement rate on phase boundary breakage and a potential increase of phase boundary
breakage and liberation to be achieved at slower displacement rates (Garcia et al., 2009).
While in the case of the texture (A), the loading rate has an effect on breakage mode and
liberation, the texture (B) shows no significant difference in terms of deformation, which
is high in both displacement rates (Figure 15). Therefore, in texture (B), the inducing rate
of “effective micro-cracks” is high and the crack propagation would be random. A possible
explanation is that the crystalline structure of the texture (B) has more natural micro-cracks,
which have a potential of meso-crack and ultimately macro-crack formation and
propagation. These findings support that the ore texture nature and loading mechanism have
an effect on minerals breakage mode and liberation.
Sample code (A) (B)

DR (mm/min) 0.1 1 0.1 1
Material
failure
Figure 16. Ore texture failure at the failure point- strain ranging from -0.16 (compression)
to +0.16 (tensile).
31
4.2 Crack quantification and characterization
4.2.1 Crack characterization
From the SEM analysis of texture (A), it was observed that some quartz grains were covered
by biotite (Figure 17). This phenomenon is due to the ionic thermal reaction when the rock
is forming according to the reaction of quartz with water and elements i.e. potassium, iron,
magnesium, aluminum (Yardley, 1977). From this reaction, new biotite grains in the matrix
can be formed or enlarge the existing grains. More examples of the SEM and optical
microscopy images of the failed texture (A) with 0.1 mm/min displacement rate are given
in Figure 17. What is clearly seen from these images is the crack diversion to the biotite
grain rather than passing through the quartz (Figure 17 (h)). This might be due to the
cleavage planes of biotite, which deviate the cracks. This type of crack formation is
attributed to the splitting along cleavage planes and correspond with kinked structures in
deformed biotite which is a sign of weak planes heterogeneity (Wilson and Bell, 1979). On
the other hand, in the cases where quartz did not have biotite around (Figure 17 (l) and (g)),
cracks passed through the quartz.
In terms of breakage mechanism and pattern of breakage, there are four main categories
including shattering, cleavage, chipping, and abrasion. The shattering breakage mechanism
is expected to be observed where the applied energy is a multiple of that required for the
breakage. Closer inspection of the texture (A)-0.1 mm/min case shows that crack formation
patterns in magnetite lead to the shattering breakage mechanism (Figure 17 (p)). Micro-
indentation tests done on magnetite and quartz show that they have roughly similar
toughness (Middlemiss and King, 1996; Whitney et al., 2007). However, Figure 17 (j)
shows that there is no shattering breakage mechanism where quartz, apatite, and actinolite
are accumulated. This observation may be explained by the fact that the magnetite phase
has comparatively lower toughness due to its natural micro-cracks. The natural micro-
cracks in magnetite might be due to the recrystallization process of magnetite under the
metamorphism process and random orientation of magnetite with anisotropic crystalline
grains which causes intergranular stresses (geometrical and micro-crack heterogeneity)
(Bergman et al., 2001; Nur and Simmons, 1970). In addition to the metamorphism process
and its pressure and temperature variation, due to the higher values of bulk moduli and
thermal expansion coefficient of quartz in the magnetite matrix, natural micro-cracks can
be formed in magnetite (micro-crack heterogeneity) (Nadan and Engelder, 2009; Nur and
32
Simmons, 1970). Therefore, it can be concluded that, based on the material type in the
compression zone, a shattering breakage mechanism may occur or not.
The observations of magnetite failure in optical microscopy images (Figure 17 - Figure 21)
shows that the cracks were propagating into magnetite grains rather than along the grain
boundaries. By comparing the grains images from optical microscopy with DVC
measurements, a stronger tensile strain can be seen (Figure 17). This may partly be
explained by the natural micro-cracks inside the magnetite grain and crystals, which can be
deformed under the stress. These micro-cracks cause preferential breakage in the magnetite
grains. Overall, these results indicate that micro-cracks, which are not visible with the
current resolution, contributed to the failure of the sample. In the case of Figure 17 (o), it
can be seen that the micro-crack has propagated between apatite and quartz grains. This
type of micro-crack is referred to as “en passant” and can be due to the interaction of the
stress fields around two approaching crack tips and locally applied stress (Kranz, 1979).
Figure 17. Strain, grain boundary, and BSE mapping illustration of texture (A)-0.1
mm/min. Yellow color: grain boundaries. Red color: Explained cases in the text. Strain
code color code as Figure 15.
In addition, similar results in terms of biotite covering quartz grains have been observed
also in texture B (Figure 19 (q), (s), and (t)) at 1 mm/min loading displacement rate.
However, in certain areas, the quartz grain boundary selectivity decreased and cracks
propagated within the quartz (Figure 19 (k), (v), and (y)). This can be due to the decreasing
selectivity of micro-crack propagation at higher displacement rates (Koch, 2017). Similar
to lower loading displacement rate or the conditions where quartz is alone without biotite
coverage, the cracks penetrated into the quartz (Figure 19 (r), (u), and (w)).
33
Dimension BSE Optical
2D
XCT-Prior slice XCT-Later slice
3D
Figure 18. Crack formation and distribution in texture (A) with displacement rate of 0.1
mm/min. Yellow color: grain boundaries. Red color: Explained cases in the text.
2D
3D
Figure 19. Crack formation and distribution in texture (A) with displacement rate of 1
The observations of crack propagation in magnetite grains for texture (B) showed that the
micro-cracks were propagating in magnetite grains (Figure 20 and Figure 21). Therefore,
at both displacement rates, the magnetite breakage behavior is similar to that of texture (A).
For (B) (0.1 mm/min), micro-cracks propagated along with apatite and magnetite grain
boundaries ((1) and (2)) and more into apatite grains ((2), (3), (4), and (6)). For texture (B)
(1 mm/min), cracks propagated with a similar behavior as texture (B) (0.1 mm/min).
34
Compared to the texture (A), the loading displacement rate has less effect on the crack
formation and distribution of the texture (B). It seems possible that these results are due to
higher quantities of natural micro-defects inside the texture (B), which act more as
“effective micro-cracks” and the crack tips propagate faster. The crack formations support
evidence from previous observations of strain measurements.
2D
3D
Figure 20. Crack formation and distribution in texture (B) with displacement rate of 0.1
2D
3D
Figure 21. Crack formation and distribution in texture (B) with displacement rate of 1
35
4.2.2 Crack quantification
In total four minerals were selected according to the pixel value of the mineral including
quartz, apatite, actinolite, and magnetite. Quantification of minerals occurrence on both
sides of the crack was done along the entire crack. The crack length inside each mineral
and between two minerals quantified and was normalized by the total length of the cracks
(Figure 22). Due to the resolution restrictions, it was not possible to distinguish rim shaped
biotite around the quartz grains. In terms of loading rate, the remarkable change is seen in
the cracks passing inside magnetite as well as apatite. In this regard, while the relative crack
length inside apatite in both textures increases with loading rate, the occurrence of the
cracks inside magnetite decreases. It is known that magnetite and apatite are often highly
associated together in this type of ore (Parian et al., 2018), therefore the cracks occurring
in between magnetite and apatite or solely inside magnetite contribute most to the liberation
of magnetite. Besides, the higher number of cracks in between actinolite and magnetite in
slower loading rate contribute even further to higher magnetite liberation. The other notable
differences are relative crack length in quartz and actinolite in texture A compared to
texture B. This could be explained by considering relatively large abundance of these
phases in texture A (Table 7).
Figure 22. Phase boundary and preferential cracking quantification for each mineral along
entire crack length.
36
In overall, the cracks dominantly occurred in magnetite phase. It seems that the inherent
micro-cracks and the large abundance of magnetite are the main reasons. In terms of
breakage mode, the slower loading rate gives higher preferential breakage for magnetite
while this is opposite for apatite. Phase boundary breakage between minerals e.g. apatite
and magnetite or quartz and magnetite are not significantly different with tested loading
rates. Moreover, it is observed that the slower loading rate is more favorable for the
liberation of magnetite.
Table 7: Minerals volumetric grade in each texture
Volumetric grade (%)

Mineral Texture A Texture B
Quartz 21.24 0.94
Actinolite 8.64 2.16
Apatite 3.81 6.45
Magnetite 66.31 90.44
37
38
Chapter 5 - Summary and future work
The strategies for reducing energy consumption in comminution are (i) to design flowsheets
and select equipment based on particle breakage behavior and (ii) to design new types of
comminution machines. In order to implement these strategies, a fundamental
understanding of the micro-processes is needed which leads to optimized breakage
processes with respect to dominant ore texture and process parameters. In this work the
fundamentals behind the particle breakage in comminution machines have been analyzed.
In general, the particle breakage process has been classified according to loading
mechanism, breakage mechanism, and breakage mode. The loading mechanism is defined
as the action of physical quantities, which are applied to a particle or several particles in
order to introduce mechanical stress. The consequence of particle failure is the breakage
mechanism, which is influenced by loading mechanism and material properties. The
breakage mode defines the particle breakage in terms of being random or non-random.
Based on King and Schneider (1998), the breakage mode is affected by the ore texture. In
contrast to that, other studies highlighted the influence of the loading mechanisms on
breakage mode. Hence, a fundamental investigation of the microscopic spatial deformation
of ore texture under loading mechanism test is essential for optimizing the process with
respect to loading mechanism and ore texture.
Ore texture type and loading displacement rate are the two parameters, which have been
identified to affect the ore crack formation and consequently breakage mode and mineral
liberation. In this study, a new aspect of ore texture was introduced as ore texture
heterogeneity, which is a complex parameter comprising mineral heterogeneity,
geometrical heterogeneity, weak grain boundaries, and micro-cracks. The combined effect
of ore texture heterogeneity and loading displacement rate was investigated through two
methods including Digital Volume Correlation and crack quantification. Digital Volume
Correlation illustrates the textural deformations and strains received from in-situ loading
of the specimen while conducting X-ray computed micro-tomography scanning. Crack
quantification was done through a combination of the cracks’ binary image and scanning
electron microscopy backscattered electron image. Moreover, a detailed characterization
of cracks was done based on optical and characteristic scanning electron microscopy
images. The following conclusions can be drawn:
39
 Digital Volume Correlation measurements showed that the deformations of
invisible micro-cracks help to analyze the breakage mode in terms of being random
or non-random within the ore texture.
 The findings suggest that both ore texture and loading displacement rate may affect
the breakage mode and liberation, although the current study is based on small size
samples.
 The density of crack formation in a specific mineral indicates the natural micro-
cracks within the mineral, which cause a preferential breakage mode.
 Geological heterogeneity deformations within an ore texture affect the minerals’

natural micro-crack and alter breakage mechanism.
 Biotite weak planes and cleavages cause a preferential breakage mode through its
boundaries with minerals.
 Geological deformation e.g. metamorphism causes geometrical and micro-crack

heterogeneity.
 Crack quantification gives a powerful tool for detecting the type of random and non-
random breakage mode in terms of being phase boundary and preferential breakage.
 The evaluation of the comminution process in terms of mineral liberation should be

based on the combinations of ore texture and process-related parameters.
This licentiate thesis aimed at establishing such a comprehensive method and, by using a
case study from the Malmberget iron ore, Northern Sweden, at demonstrating how and why
both ore texture and process related factors should be considered in the investigation of ore
breakage and mineral liberation.
Overall, in terms of liberation studies, this PhD project aims to integrate process-related
parameters to minerals breakage mode and liberation, and investigates the fundamentals
behind the ore texture effecting breakage model (Figure 23). The appropriate technique to
analyze minerals breakage mode and liberation without the energy-intensive fragmentation
process will also be developed.
In continuation of the achievements of licentiate objectives, the following tasks are planned
for the future studies:
40
1. The loading mechanisms will be further investigated in terms of other energy
intensities and strain rates on multi and multiple particle level.
2. Breakage mechanism effect on ore texture particle fragmentation and breakage rate
will be investigated through a small scale batch grindability test (Mwanga et al.,
2017).
3. The quantification of cracks in third dimension will pave the way to validate
breakage mode analysis in whole particle volume.
Figure 23. Workflow for the PhD thesis. Blue boxes are the features that were selected to
study in this licentiate thesis.
41
42
References
Anderson, T.L., 2012. Fracture Mechanics: Fundamentals and Applications, Fourth Edi.
ed, Taylor & Francis Group. https://doi.org/10.1016/j.jmps.2010.02.008
Andreatidis, J.P., 1995. Breakage Mechanisms and Resulting Mineral Liberation in a Bead
Mill. Dissertation/Thesis, University of Queensland.
Apling, A., Bwalya, M., 1997. Evaluating high pressure milling for liberation enhancement
and energy saving. Miner. Eng. 10, 1013–1022. https://doi.org/10.1016/S0892-
6875(97)00080-0
Apling, A.C., Raissi, A., 2000. Evaluation of enhanced liberation by comminution under
high pressure. Miner. Process. Extr. Metall. 109, 117–120.
https://doi.org/10.1179/mpm.2000.109.3.117
Austin, L.G., Barahona, C.A., Menacho, J.M., 1986. Fast and slow chipping fracture and
abrasion in autogenous grinding. Powder Technol. 46, 81–87.
https://doi.org/10.1016/0032-5910(86)80102-4
Barley, R.W., Conway-Baker, J., Pascoe, R.D., Kostuch, J., McLoughlin, B., Parker, D.J.,
2004. Measurement of the motion of grinding media in a vertically stirred mill using
positron emission particle tracking (PEPT) Part II. Miner. Eng. 17, 1179–1187.
https://doi.org/10.1016/j.mineng.2004.06.034
Baud, P., Wong, T. fong, Zhu, W., 2014. Effects of porosity and crack density on the
compressive strength of rocks. Int. J. Rock Mech. Min. Sci. 67, 202–211.
https://doi.org/10.1016/j.ijrmms.2013.08.031
Bay, B.K., Smith, T.S., Fyhrie, D.P., Saad, M., 1999. Digital volume correlation: Three-
dimensional strain mapping using x-ray tomography. Exp. Mech. 39, 217–226.
https://doi.org/10.1007/BF02323555
Bergman, S., Kübler, L., Martinsson, O., 2001. Description of regional geological and
geophysical maps of northern Norrbotten county (east of the Caledonian orogen),
SGU Geolog. ed.
Berthiaux, H., Dodds, J., 1999. Modelling fine grinding in a fluidized bed opposed jet mill.
Part I: Batch grinding kinetics. Powder Technol. 106, 78–87.
https://doi.org/10.1016/S0032-5910(99)00049-2
43
Bérubé, M.A., Marchand, J.C., 1984. Evolution of the mineral liberation characteristics of
an iron ore undergoing grinding. Int. J. Miner. Process. 13, 223–237.
https://doi.org/10.1016/0301-7516(84)90005-X
Bieniawski, Z.T., Hawkes, I., 1978. Suggested methods for determining tensile strength of
rock materials. Int. J. Rock Mech. Min. Sci. Geomech. Abstr. 15, 99–103.
https://doi.org/10.1016/0148-9062(78)90003-7
Bracey, R.J., Weerasekara, N.S., Powell, M.S., 2016. Performance evaluation of the novel
multi-shaft mill using DEM modelling. Miner. Eng. 98, 251–260.
Bradshaw, D., 2014. The role of process mineralogy in improving the process performance
of complex sulphide ores, in: XXVII International Mineral Processing Congress
(IMPC) Santiago Chile: The Role of Process Mineralogy in Improving the Process
Performance of Complex Sulphide Ores. pp. 1–23. https://doi.org/10.1007/978-3-658-
07529-3
Brotóns, V., Tomás, R., Ivorra, S., Alarcón, J.C., 2013. Temperature influence on the
physical and mechanical properties of a porous rock: San Julian’s calcarenite. Eng.
Geol. 167, 117–127. https://doi.org/10.1016/j.enggeo.2013.10.012
Buljac, A., Jailin, C., Mendoza, A., Neggers, J., Taillandier-Thomas, T., Bouterf, A.,
Smaniotto, B., Hild, F., Roux, S., 2018. Digital Volume Correlation: Review of
Progress and Challenges. Exp. Mech. 58, 661–708. https://doi.org/10.1007/s11340-
018-0390-7
Cao, J., Liu, L., Han, Y., Feng, A., 2019. Comminution behavior and mineral liberation
characteristics of low-grade hematite ore in high pressure grinding roll. Physicochem.
Probl. Miner. Process. 55, 575–585. https://doi.org/10.5277/ppmp18169
Celik, I.B., Oner, M., 2006. The influence of grinding mechanism on the liberation
characteristics of clinker minerals. Cem. Concr. Res. 36, 422–427.
https://doi.org/10.1016/j.cemconres.2005.09.011
Celik, I.B., Oner, M., Can, N.M., 2007. The influence of grinding technique on the
liberation of clinker minerals and cement properties. Cem. Concr. Res. 37, 1334–
1340. https://doi.org/10.1016/j.cemconres.2007.06.004
44
Chelgani, S.C., Parian, M., Parapari, P.S., Ghorbani, Y., Rosenkranz, J., 2019. A
comparative study on the effects of dry and wet grinding on mineral flotation
separation–a review. J. Mater. Res. Technol. 1–8.
https://doi.org/10.1016/j.jmrt.2019.07.053
Chen, X., Peng, Y., Bradshaw, D., 2014. The effect of particle breakage mechanisms during
regrinding on the subsequent cleaner flotation. Miner. Eng. 66, 157–164.
Claisse, P.A., 2016. Alloys and nonferrous metals, in: Civil Engineering Materials.
Elsevier, pp. 361–368. https://doi.org/10.1016/B978-0-08-100275-9.00032-2
Clarke, A.J., Wills, B.A., 1989. Technical note Enhancement of cassiterite liberation by
high pressure roller comminution. Miner. Eng. 2, 259–262.
https://doi.org/10.1016/0892-6875(89)90046-0
Cleary, P.W., Morrison, R.D., 2016. Comminution mechanisms, particle shape evolution
and collision energy partitioning in tumbling mills. Miner. Eng. 86, 75–95.
Dall’Ara, E., Peña-Fernández, M., Palanca, M., Giorgi, M., Cristofolini, L., Tozzi, G.,
2017. Precision of digital volume correlation approaches for strain analysis in bone
imaged with micro-computed tomography at different dimensional levels. Front.
Mater. 4. https://doi.org/10.3389/fmats.2017.00031
Deniz, V., 2013. Comparisons of dry grinding kinetics of lignite, bituminous coal, and
petroleum coke. Energy Sources, Part A Recover. Util. Environ. Eff. 35, 913–920.
https://doi.org/10.1080/15567036.2010.514591
Dyskin, A. V., 1999. On the role of stress fluctuations in brittle fracture. Int. J. Fract. 100,
29–53. https://doi.org/10.1023/A:1018664101433
Dyskin, A. V., Sahouryeh, E., 1997. Mechanism of dilatancy and fracture of brittle
materials in uniaxial compression. Int. J. Fract. 86.
Eberhardt, E., Stimpson, B., Stead, D., 1999. Effects of grain size on the initiation and
propagation thresholds of stress-induced brittle fractures. Rock Mech. Rock Eng. 32,
81–99. https://doi.org/10.1007/s006030050026
Evertsson, C.M., 2000. Cone crusher performance. Dissertation/Thesis, Chalmers
45
University of Technology. https://doi.org/10.13140/RG.2.1.2212.7526
Frances, C., Le Bolay, N., Belaroui, K., Pons, M.N., 2001. Particle morphology of ground
gibbsite in different grinding environments. Int. J. Miner. Process. 61, 41–56.
https://doi.org/10.1016/S0301-7516(00)00025-9
Fuerstenau, D.W., 1995. Grinding Aids. KONA Powder Part. J. 13, 5–18.
https://doi.org/10.14356/kona.1995006
Fuerstenau, M.C., Han, K.N. (Eds.), 2018. Principles of Mineral Processing, Society for
Mining, Metallurgy, and Exploration, Inc. https://doi.org/10.1016/B978-0-12-
814022-2.00013-7
Garcia, D., Lin, C.L., Miller, J.D., 2009. Quantitative analysis of grain boundary fracture
in the breakage of single multiphase particles using X-ray microtomography
procedures. Miner. Eng. 22, 236–243. https://doi.org/10.1016/j.mineng.2008.07.005
Haldar, S.K., Tišljar, J., 2014. Introduction to Mineralogy and Petrology. Elsevier.
https://doi.org/10.1016/C2012-0-03337-6
Hamdi, P., Stead, D., Elmo, D., 2015. Characterizing the influence of stress-induced
microcracks on the laboratory strength and fracture development in brittle rocks using
a finite-discrete element method-micro discrete fracture network FDEM-μDFN
approach. J. Rock Mech. Geotech. Eng. 7, 609–625.
https://doi.org/10.1016/j.jrmge.2015.07.005
Hamid, S., Alfonso, P., Anticoi, H., Guasch, E., Oliva, J., Dosbaba, M., Garcia-Valles, M.,
Chugunova, M., 2018. Quantitative Mineralogical Comparison between HPGR and
Ball Mill Products of a Sn-Ta Ore. Minerals 8, 151.
https://doi.org/10.3390/min8040151
Han, Y., Liu, L., Yuan, Z., Wang, Z., Zhang, P., 2012. Comparison of low-grade hematite
product characteristics in a high- pressure grinding roller and jaw crusher. Miner.
Metall. Process. 29, 6. https://doi.org/3059792
Hatzor, Y.H., Palchik, V., 1997. The influence of grain size and porosity on crack initiation
stress and critical flaw length in dolomites. Int. J. Rock Mech. Min. Sci. 34, 805–816.
https://doi.org/10.1016/S1365-1609(96)00066-6
Hennart, S.L.A., Wildeboer, W.J., van Hee, P., Meesters, G.M.H., 2009. Identification of
46
the grinding mechanisms and their origin in a stirred ball mill using population
balances. Chem. Eng. Sci. 64, 4123–4130. https://doi.org/10.1016/j.ces.2009.06.031
Hogg, R., 1999. Breakage mechanisms and mill performance in ultrafine grinding. Powder
Technol. 105, 135–140. https://doi.org/10.1016/S0032-5910(99)00128-X
Hogg, R., Cho, H., 2000. A review of breakage behavior in fine grinding by stirred-media
milling. KONA Powder Part. J. 18, 9–19. https://doi.org/10.14356/kona.2000007
Inoue, T., Okaya, K., 1996. Grinding mechanism of centrifugal mills - A simulation study
based on the discrete element method. Int. J. Miner. Process. 44–45, 425–435.
https://doi.org/10.1016/0301-7516(95)00049-6
Kaplan, M.F., 1963. Strains and Stresses of Concrete at Initiation of Cracking and Near
Failure. J. Proc. 60, 853–880. https://doi.org/10.14359/7882
Kelly, E.G., Spottiswood, D.J., 1982. Introduction to mineral processing, John Wiley. ed.
The Australian Mineral Foundation.
Ketcham, R.A., Carlson, W.D., 2001. Acquisition, optimization and interpretation of x-ray
computed tomographic imagery: Applications to the geosciences. Comput. Geosci.
27, 381–400. https://doi.org/10.1016/S0098-3004(00)00116-3
King, R.P., 2012. Modeling and Simulation of Mineral Processing Systems, 2nd ed,
Modeling and Simulation of Mineral Processing Systems. Society for Mining,
Metallurgy & Exploration. https://doi.org/10.1016/C2009-0-26303-3
King, R.P., 1994. Comminution and liberation of minerals. Miner. Eng. 7, 129–140.
https://doi.org/10.1016/0892-6875(94)90059-0
King, R.P., 1993. A success story that has not yet ended, in: Proceedings XVIII,
International Mineral Processing Congress, the Australasian IMM. pp. 39–45.
King, R.P., Schneider, C.L., 1998. Mineral liberation and the batch comminution equation.
Miner. Eng. 11, 1143–1160. https://doi.org/10.1016/S0892-6875(98)00102-2
Koch, P.-H., 2017. Particle generation for geometallurgical process modeling. Licent.
thesis / Luleå Univ. Technol., Licentiate thesis / Luleå University of Technology.
https://doi.org/10.1007/978-3-319-06590-8
Kotake, N., Kuboki, M., Kiya, S., Kanda, Y., 2011. Influence of dry and wet grinding
47
conditions on fineness and shape of particle size distribution of product in a ball mill.
Adv. Powder Technol. 22, 86–92. https://doi.org/10.1016/j.apt.2010.03.015
Kranz, R.L., 1983. Microcracks in rocks: A review. Tectonophysics 100, 449–480.

https://doi.org/10.1016/0040-1951(83)90198-1
Kranz, R.L., 1979. Crack-crack and crack-pore interactions in stressed granite. Int. J. Rock
Mech. Min. Sci. Geomech. Abstr. 16, 51. https://doi.org/10.1016/0148-
9062(79)90512-6
Kyle, J.R., Ketcham, R.A., 2015. Application of high resolution X-ray computed
tomography to mineral deposit origin, evaluation, and processing. Ore Geol. Rev. 65,
821–839. https://doi.org/10.1016/j.oregeorev.2014.09.034
Lam, L., Lee, S.-W., Suen, C.Y., 1992. Thinning methodologies-a comprehensive survey.
IEEE Trans. Pattern Anal. Mach. Intell. 14, 869–885.
https://doi.org/10.1109/34.161346
Lamberg, P., 2011. Particles-the bridge between geology and metallurgy, in: Proceedings
Conference in Minerals Engineering, Luleå, Sweden.
Larsson, S., Gustafsson, G., Häggblad, H.Å., Jonsén, P., 2018. Experimental and numerical
study of potassium chloride flow using smoothed particle hydrodynamics. Miner. Eng.
116, 88–100. https://doi.org/10.1016/j.mineng.2017.11.003
Lenoir, N., Bornert, M., Desrues, J., Bésuelle, P., Viggiani, G., 2007. Volumetric Digital
Image Correlation Applied to X-ray Microtomography Images from Triaxial
Compression Tests on Argillaceous Rock. Strain 43, 193–205.
https://doi.org/10.1111/j.1475-1305.2007.00348.x
Leon, L.G., 2018. Analysis and Modelling of Mineral and Element Composition in
Compression Breakage. Dissertation/Thesis, Chalmers University of Technology.
Little, L., 2016. The development and demonstration of a practical methodology for fine
particle shape characterisation in minerals processing. Dissertation/Thesis, University
of Cape Town.
Little, L., Mainza, A.N., Becker, M., Wiese, J., 2017. Fine grinding: How mill type affects
particle shape characteristics and mineral liberation. Miner. Eng. 111, 148–157.
48
Little, L., Mainza, A.N., Becker, M., Wiese, J.G., 2016. Using mineralogical and particle
shape analysis to investigate enhanced mineral liberation through phase boundary
fracture. Powder Technol. 301, 794–804.
https://doi.org/10.1016/j.powtec.2016.06.052
Lund, C., 2013. Mineralogical, chemical and textural characterisation of the Malmberget
iron ore deposit for a geometallurgical model. Dissertation/Thesis, Luleå University
of Technology.
Lynch, A.J. (Ed.), 2015. Comminution handbook. Australasian Institute of Mining and
Metallurgy.
Mahabadi, O.K., Tatone, B.S.A., Grasselli, G., 2014. Influence of microscale heterogeneity
and microstructure on the tensile behavior of crystalline rocks. J. Geophys. Res. Solid
Earth 119, 5324–5341. https://doi.org/10.1002/2014JB011064
Maire, E., Withers, P.J., 2014. Quantitative X-ray tomography. Int. Mater. Rev. 59, 1–43.
https://doi.org/10.1179/1743280413Y.0000000023
Manlapig, E.V., Drinkwater, D.J., Munro, P.D., Johnson, N.W., Watsford, R.M.S., 1985.
Optimisation of Grinding Circuits at the Lead/Zinc Concentrator, Mount Isa Mines
Limited. IFAC Proc. Vol. 18, 265–274. https://doi.org/10.1016/S1474-
6670(17)60520-2
Mariano, R., 2016. Measurement and modelling of the liberation and distribution of
minerals in comminuted ores. Dissertation/Thesis, The University of Queensland.
https://doi.org/10.14264/uql.2016.1081
Mariano, R.A., Evans, C.L., 2018. The effect of breakage energies on the mineral liberation
properties of ores. Miner. Eng. 126, 184–193.
Martin, C.D., Stimpson, B., 1994. The effect of sample disturbance on laboratory properties
of Lac du Bonnet granite. Can. Geotech. J. 31, 692–702. https://doi.org/10.1139/t94-
081
Meisels, R., Toifl, M., Hartlieb, P., Kuchar, F., Antretter, T., 2015. Microwave propagation
and absorption and its thermo-mechanical consequences in heterogeneous rocks. Int.
J. Miner. Process. 135, 40–51. https://doi.org/10.1016/j.minpro.2015.01.003
49
Melenka, G.W., Carey, J.P., 2015. Evaluation of fiber reinforced cement using digital
image correlation. PLoS One 10, 1–18. https://doi.org/10.1371/journal.pone.0128644
Middlemiss, S., King, R.P., 1996. Microscale fracture measurements with application to
comminution. Int. J. Miner. Process. 44–45, 43–58. https://doi.org/10.1016/0301-
7516(95)00016-X
Mirzaei, Z.S., Khalesi, M.R., 2019. Development of a Simulator for Random and Non-
Random Breakage of Particles and Liberation of Grains Based on Voronoi
Tessellation. Minerals 9, 341. https://doi.org/10.3390/min9060341
Mohd, S., Brahim, J., Latiffi, A.A., Fathi, M.S., Harun, A.N., 2017. Developing building
information modelling (BIM) implementation model for project design team,
Malaysian Construction Research Journal. https://doi.org/10.1002/ejoc.201200111
Mütze, T., Husemann, K., 2008. Compressive stress: Effect of stress velocity on confined
particle bed comminution. Chem. Eng. Res. Des. 86, 379–383.
https://doi.org/10.1016/j.cherd.2007.11.007
Mwanga, A., Rosenkranz, J., Lamberg, P., 2017. Development and experimental validation
of the Geometallurgical Comminution Test (GCT). Miner. Eng. 108, 109–114.
Mwanga, A., Rosenkranz, J., Lamberg, P., 2015. Testing of Ore Comminution Behavior in
the Geometallurgical Context—A Review. Minerals 5, 276–297.
https://doi.org/10.3390/min5020276
Nadan, B.J., Engelder, T., 2009. Microcracks in New England granitoids: A record of
thermoelastic relaxation during exhumation of intracontinental crust. Bull. Geol. Soc.
Am. 121, 80–99. https://doi.org/10.1130/B26202.1
Nematollahi, H., 2002. Mineral processing handbook. University of Tehran Press.
Nur, A., Simmons, G., 1970. The origin of small cracks in igneous rocks. Int. J. Rock Mech.
Min. Sci. 7, 307–314. https://doi.org/10.1016/0148-9062(70)90044-6
Obert, L., Winder, S.L., Derwall, W.I., 1946. Standardized test for determinig the physical
properties of mine rock. U. S. Bur. Mines Rep. Investig. 3891, 67 pages :
https://doi.org/10.1016/0278-4343(88)90009-X
50
Ozcan, O., Benzer, H., 2013. Comparison of different breakage mechanisms in terms of
product particle size distribution and mineral liberation. Miner. Eng. 49, 103–108.
Ozkahraman, H.T., 2010. Breakage mechanisms and an encouraging correlation between

the Bond parameters and the friability value. J. South. African Inst. Min. Metall. 110,
153–159.
Palaniandy, S., Azizli, K.A.M., Hussin, H., Hashim, S.F.S., 2008. Effect of operational
parameters on the breakage mechanism of silica in a jet mill. Miner. Eng. 21, 380–
388. https://doi.org/10.1016/j.mineng.2007.10.011
Pan, B., Qian, K., Xie, H., Asundi, A., 2009. Two-dimensional digital image correlation
for in-plane displacement and strain measurement: A review. Meas. Sci. Technol. 20.
https://doi.org/10.1088/0957-0233/20/6/062001
Parian, M., Mwanga, A., Lamberg, P., Rosenkranz, J., 2018. Ore texture breakage
characterization and fragmentation into multiphase particles. Powder Technol. 327,
57–69. https://doi.org/10.1016/j.powtec.2017.12.043
Peng, J., Rong, G., Cai, M., Zhou, C.B., 2015. A model for characterizing crack closure
effect of rocks. Eng. Geol. 189, 48–57. https://doi.org/10.1016/j.enggeo.2015.02.004
Pious, C.V., Thomas, S., 2016. Polymeric Materials—Structure, Properties, and

Applications, in: Printing on Polymers. Elsevier, pp. 21–39.
https://doi.org/10.1016/B978-0-323-37468-2.00002-6
Pourghahramani, P., 2012. Effects of ore characteristics on product shape properties and
breakage mechanisms in industrial SAG mills. Miner. Eng. 32, 30–37.
Prasher, C., 1987. Crushing and grinding process handbook. Chichester: Wiley.
Pratt, W.K., 2001. Digital Image Processing. John Wiley & Sons, Inc., New York, USA.
https://doi.org/10.1002/0471221325
Reichert, M., Gerold, C., Fredriksson, A., Adolfsson, G., Lieberwirth, H., 2015. Research
of iron ore grinding in a vertical-roller-mill. Miner. Eng. 73, 109–115.
51
Roufail, R., Klein, B., 2010. Mineral Liberation and Particle Breakage in Stirred Mills.
Can. Metall. Q. 49, 419–428. https://doi.org/10.1179/cmq.2010.49.4.419
Roufail, R.A., 2011. The effect of stirred mill operation on particles breakage mechanism
and their morphologyical features. Dissertation/Thesis, University Of British
Columbia.
Rumpf, H., 1965. Die Einzelkornzerkleinerung als Grundlage einer technischen

Zerkleinerungswissenschaft. Chemie Ing. Tech. 37, 187–202.
https://doi.org/10.1002/cite.330370303
Saeidi, F., Yahyaei, M., Powell, M., Tavares, L.M., 2017. Investigating the effect of applied
strain rate in a single breakage event. Miner. Eng. 100, 211–222.
Schubert, W., Tomas, J., 2007. Chapter 24 Liberation of Valuables Embedded in Particle
Compounds and Solid Waste, in: Handbook of Powder Technology. pp. 989–1018.
https://doi.org/10.1016/S0167-3785(07)12027-3
Shea, W.T., Kronenberg, A.K., 1993. Strength and anisotropy of foliated rocks with varied
mica contents. J. Struct. Geol. 15, 1097–1121. https://doi.org/10.1016/0191-
8141(93)90158-7
Simmons, G., Richter, D., 1976. Microcracks in rocks. Phys. Chem. Miner. rocks 105–137.
Sinnott, M.D., Cleary, P.W., 2015. Simulation of particle flows and breakage in crushers
using DEM: Part 2 - Impact crushers. Miner. Eng. 74, 163–177.
Tavares, L.M., 2005. Particle weakening in high-pressure roll grinding. Miner. Eng. 18,
651–657. https://doi.org/10.1016/j.mineng.2004.10.012
Toifl, M., Meisels, R., Hartlieb, P., Kuchar, F., Antretter, T., 2015. 3D numerical study on
microwave induced stresses in inhomogeneous hard rocks. Miner. Eng. 90, 29–42.
Unland, G., 2007. Chapter 4 The Principles of Single-Particle Crushing, in: Handbook of
Powder Technology. pp. 117–225. https://doi.org/10.1016/S0167-3785(07)12007-8
Varinot, C., Hiltgun, S., Pons, M.N., Dodds, J., 1997. Identification of the fragmentation
52
mechanisms in wet-phase fine grinding in a stirred bead mill. Chem. Eng. Sci. 52,
3605–3612. https://doi.org/10.1016/S0009-2509(97)89693-5
Vizcarra, T.G., Wightman, E.M., Johnson, N.W., Manlapig, E. V., 2011. The effect of
breakage method on the shape properties of an iron-oxide hosted copper-gold ore.
Miner. Eng. 24, 1454–1458. https://doi.org/10.1016/j.mineng.2011.07.007
breakage mechanism on the mineral liberation properties of sulphide ores. Miner. Eng.
Wang, B., Pan, B., Lubineau, G., 2018. Morphological evolution and internal strain
mapping of pomelo peel using X-ray computed tomography and digital volume
correlation. Mater. Des. 137, 305–315. https://doi.org/10.1016/j.matdes.2017.10.038
Wang, M.H., Yang, R.Y., Yu, A.B., 2012. DEM investigation of energy distribution and
particle breakage in tumbling ball mills. Powder Technol. 223, 83–91.
Whitney, D.L., Broz, M., Cook, R.F., 2007. Hardness, toughness, and modulus of some
common metamorphic minerals. Am. Mineral. 92, 281–288.
https://doi.org/10.2138/am.2007.2212
Wills, B.A., Atkinson, K., 1993. Some observations on the fracture and liberation of
mineral assemblies. Miner. Eng. 6, 697–706. https://doi.org/10.1016/0892-
6875(93)90001-4
Wills, B.A., Napier-Munn, T., 2005. Mineral Processing Technology An Introduction to

the Practical Aspects of Ore Treatment and Mineral, Wills’ Mineral Processing
Technology. https://doi.org/10.1016/B978-075064450-1/50008-4
Wilson, C.J.L., Bell, I.A., 1979. Deformation of biotite and muscovite: Optical
microstructure. Tectonophysics 58, 179–200. https://doi.org/10.1016/0040-
1951(79)90328-7
Wu, Q., Chen, L., Shen, B., Dlamini, B., Li, S., Zhu, Y., 2019. Experimental Investigation
on Rockbolt Performance Under the Tension Load. Rock Mech. Rock Eng. 52, 4605–
4618. https://doi.org/10.1007/s00603-019-01845-1
Xiao, X., Zhang, G., Feng, Q., Xiao, S., Huang, L., Zhao, X., Li, Z., 2012. The liberation
53
effect of magnetite fine ground by vertical stirred mill and ball mill. Miner. Eng. 34,
63–69. https://doi.org/10.1016/j.mineng.2012.04.004
Xu, W., Dhawan, N., Lin, C.-L., Miller, J.D., 2013. Further study of grain boundary fracture
procedures. Miner. Eng. 46–47, 89–94. https://doi.org/10.1016/j.mineng.2013.03.016
Yardley, B.W.D., 1977. The nature and significance of the mechanism of sillimanite
growth in the Connemara schists, Ireland. Contrib. to Mineral. Petrol. 65, 53–58.
https://doi.org/10.1007/BF00373570
Ye, X., Gredelj, S., Skinner, W., Grano, S.R., 2010a. Regrinding sulphide minerals -
Breakage mechanisms in milling and their influence on surface properties and
flotation behaviour. Powder Technol. 203, 133–147.
Ye, X., Gredelj, S., Skinner, W., Grano, S.R., 2010b. Evidence for surface cleaning of
sulphide minerals by attritioning in stirred mills. Miner. Eng. 23, 937–944.
Yokoyama, T., Inoue, Y., 2007. Chapter 10 Selection of Fine Grinding Mills, in: Handbook
of Powder Technology. pp. 487–508. https://doi.org/10.1016/S0167-3785(07)12013-
3
Yu, B., Bradley, R.S., Soutis, C., Withers, P.J., 2016. A comparison of different approaches
for imaging cracks in composites by X-ray microtomography. Philos. Trans. R. Soc.
A Math. Phys. Eng. Sci. 374. https://doi.org/10.1098/rsta.2016.0037
Yuan, Z.N., Chen, H., Li, J.M., Dai, B., Zhang, W. Bin, 2018. In-situ X-ray tomography
observation of structure evolution in 1,3,5-Triamino-2,4,6-Trinitrobenzene based
polymer bonded explosive (TATB-PBX) under thermo-mechanical loading. Materials
(Basel). 11, 1–14. https://doi.org/10.3390/ma11050732
54
Part II – Papers
55
Paper 1
1
Breakage process of mineral processing comminution machines –
an approach to liberation
Parisa Semsari Paraparia*, Mehdi Parianb, Jan Rosenkranzc

*
Corresponding author
a
Minerals and Metallurgical Engineering, Dept. of Civil, Environmental and Natural Resources
Engineering, Luleå University of Technology, SE-971 87, Luleå, Sweden. Email: parisa.semsari@ltu.se,
Tel: +46(0)920492115
b
Engineering, Luleå University of Technology, SE-971 87, Luleå, Sweden. Email: mehdi.parian@ltu.se
c
Engineering, Luleå University of Technology, SE-971 87, Luleå, Sweden. Email: jan.rosenkranz@ltu.se
Abstract
In comminution machines, the product properties are affected by various parameters.
Optimizing these parameters can provide an energy-efficient process and a specified
product. In this study, breakage fundamentals are analyzed and set against various
comminution machines. The study of the breakage fundamentals is essential for a better
understanding of the different comminution environments effect on ore types in order to
achieve a desired product size and liberation. While numerous investigations have been
conducted to study the effect of breakage fundamentals on the comminution process, the
definitions of breakage fundamentals are often used interchangeably and without caution.
Thus, this review defines three main areas of breakage processes with breakage
fundamental, namely “Loading mechanism”, “Breakage mechanism” and “Breakage
mode”. The focus is on investigating the effects of breakage fundamentals on the
comminution process and their effects on downstream processes of mineral processing
plants. Despite the advances on the design of the comminution machines and relation to
various ore types, still, the combined effect of breakage fundamentals and ore properties
such as ore texture are not fully understood. Addressing these gaps will pave the way for
choosing the proper comminution environment concerning a certain ore type.
Keywords: Breakage mechanism; Loading mechanism; Ore texture; Liberation;

Comminution
2
1 Introduction
Nowadays, with the increasing depletion of high grade and coarse-grained ores, the trend
is towards the extraction of low-grade ores [1]. This can have two consequences for the
minerals industry; first to process larger tonnages of mineral raw material and second to
grind the raw materials to finer sizes [2], which in turn leads to higher energy consumption.
It is well known that the comminution process is an energy-intensive process. Comminution
of mineral raw materials and cement clinker consumes 3% of the world-generated
electricity [3]. The strategies to reduce energy consumption for mineral comminution are
first to avoid comminution where possible and second to use more efficient and suitable
comminution technology [4]. Therefore, a proper selection of the comminution machine
for a certain mineral system leads to a specified product with the least energy consumption.
The comminution machine generates a repeated sequence of events that lead to breakage.
King pointed out that recognizing the link between the micro-processes and the particulate
system dynamics is critical to understand the breakage process [5]. He suggested that a
fundamental understanding of the breakage process should even be used in the design of
new comminution machines, which then should lead to improved industrial comminution
processes [5,6]. The selection of comminution machines is generally based on the feed
material characterization [7], final product specification [8], available operational system
[9], and energy efficiency of the machines [10].
Moreover, Bradshaw suggested ‘process mineralogy’ as a bridge for the ore processing,
geology, and metallurgy. She highlighted that the textural study of the ore would help to
design the process flowsheet [11]. Naturally, one should keep in mind that it is not possible
to consider all the geological textures; however, the key ore textures can be used. A
reasonable approach to tackle this issue could be to study the ore texture in a quantitative
manner [12]. Therefore, the two strategies, which should be taken into account for
optimizing size reduction units, are: 1) optimizing the available comminution technology
with respect to the ore and its properties, i.e. ore texture, and 2) introducing the ore texture
into the designing procedure for new comminution machines. These strategies need to be
understood through analyzing breakage fundamentals and ore textural features. In this
section, breakage fundamentals are briefly discussed. A detailed explanation of breakage
fundamentals is presented later.
3
In comminution, loading mechanism, breakage mechanism and breakage mode are the
control parameters of particle breakage. Particle breakage in the comminution process
commences with loading mechanisms. The loading mechanism along with the
comminution environment and material properties are the factors that affect the breakage
mechanism [13, 16]. Governing comminution environment parameters are temperature,
wet or dry processing, grinding media, etc. In addition, the main material properties are
mechanical strength, toughness and brittleness. All of these factors can affect the product
with regard to particle shape [15–17], size [18–20], and liberation [21–24]. While loading
and breakage mechanisms are related to the ore breakage, the breakage mode defines its
breakage in terms of being random or non-random. King and Schneider suggested that if the
breakage is affected by the mineral texture non-random breakage mode happens; otherwise, it is a
random process [25]. On the other hand, other studies highlighted that the breakage mode can be
influenced by the loading and breakage mechanisms [17,26–28].
This article aims at bringing together the literature, which ranges across the spectrum of
time and researchers to define the breakage fundamentals and their effects on the product
properties received in the comminution process. The purpose of this comprehensive review
of breakage fundamentals is threefold: (i) to present the most common definitions for
breakage fundamentals that are used based on the presented information, (ii) to highlight
and classify the areas of comminution in mineral processing where breakage fundamentals
have been used for process comprehension, and (iii) to identify the areas that require further
research and better understanding for an effective process.
2 Loading mechanism
The loading mechanism is defined as the action of physical quantities, either force or
momentum, which is applied to a particle or several particles in order to introduce
mechanical stress [29, 31, 32]. These physical quantities have a relative direction [31] and
rate [32] compared to particle motion. A range of direction and rates defines the type of
loading mechanism. Researchers use a variety of terms to define the loading mechanism.
The most common terms used for the loading mechanism in the literature are compression,
impact, shear, attrition, and abrasion (compare Table 1), which are discussed in the
following.
When particles start to undergo compression between two surfaces, potential energy is
transformed into elastic energy [33]. In brittle materials, the elastic energy is stored in the
4
crystal lattice and as it reaches a critical value, a micro-crack will be formed in the
interatomic bonds [34]. In the cases where the material has natural micro-cracks, the stress
is accumulated in crack tips and fracture occurs due to the coalescence of many micro-
cracks [35]. It is challenging to define the compression loading mechanism in terms of rate
in the field of solid or rock mechanics. In this regard, for conducting a compression test in
the aforementioned fields, the loading rate should be less than 0.2 mm/min to avoid any
dynamic effects [36]. However, in mineral processing, it is not common to reach this low
level of rate. For instance, Evertsson introduced cone crushers as a compressing crusher in
which the rock is broken by squeezing between two surfaces [33]. He calculated the
working velocity in a specific type of cone crusher at choke-level from 0.5-1 m/s.
The impact loading mechanism, as a definition provided by Kelly and Spottiswood, is the
rapid version of the loading mechanism [14]. There is a degree of uncertainty around the
rate of impact loading mechanism and the effective physical phenomena behind it.
However, with the help of simulation and modelling, it is now revealed that different
effective physical quantities in the comminution environment can occur at various impact
rates based on machine geometry and function [37–39].
Shear is challenging to define because while some references introduce it as force [40–42]
others define it as stress [14,29,43,44]. Shear force is the principal force which acts parallel
to the surface of the object and the shear stress is the shear force divided by the surface of
the object [45]. Although different terms for this loading mechanism still exist, it appears
to be some common understanding that shear refers to the physical quantities that act in
parallel with the particle and tool motion.
The other introduced types of loading mechanisms in the literature are attrition and
abrasion. Attrition is the act of wearing or grinding down by physical quantities e.g. friction
[46,47] and abrasion is a consequence of attrition [29,43,46–48]. Therefore, based on this
definition, attrition is the act (loading mechanism) that causes abrasion (breakage
mechanism). However, scientists have used abrasion as a condition for the loading
mechanism [33,49,50]. Moreover, in some cases attrition has been introduced for particles
hit between two other particles or media and abrasion is as the particles are rubbing each
other [51]. In general, the definition of attrition and abrasion is closely related to the shear
loading mechanism.
5
Table 1: Loading and breakage mechanisms and their terminology based on different researchers and their investigations based on these mechanisms
Term used for
Term used for loading Type of loading Reported breakage
References breakage Investigation
mechanism mechanism mechanism
mechanism
Impact, double impact,
[29] Stressing mode compression, attrition, - - -
shear
Slow compression, rapid
Fracture Abrasion, cleavage,
[14] Stress compression (impact), -
mechanism shatter
shear, attrition
Fast fracture, slow
Breakage Mathematical simulations for breakage rate of
[52] - Massive impact fracture, chipping,
mechanism autogenous mill
abrasion
Rupture Abrasion, cleavage, Mathematical simulations for breakage rate and
[53] Stress Compression, impact
mechanism fracture breakage function of stirred media mill
Breakage Massive fracture,
[54] Stress - Mathematical simulation of breakage mechanism
mechanism attrition
Fragmentation Cleavage, destructive Mathematical simulations for breakage rate and
[55] - -
mechanism breakage breakage function of jet mill
Comminution Compression, impact, Mathematical simulations for breakage rate and
[33] - -
principle abrasion and attrition breakage function of cone crusher
Compression, shear,
Mode of Abrasion, chipping Product particle morphology studies of ball mill,
[43] Stress attrition, impact, and
fragmentation and massive breakage jet mill, attrition mill, stirred bead mill
internal forces
Mode of Massive fracture, Product particle shape of hammer mill,
[56] - -
breakage attrition disk mill, ring and puck mill, planetary ball mill
Simulated condition of single particle breakage
Particle
[20] - - - and roller crusher, hammer mill and high
weakening
pressure grinding roll (HPGR) on particle size
Compressive, tensile, Crushing, impact,
[57] Mode of fracture Breakage -
shear attrition
Weakening, cracking,
Compression, percussion,
Comminution breaking, crumbling,
[30] Loading case bending, friction, shearing, -
effects chipping, splitting,
cutting, splitting, impact
disintegrating
6
Destructive breakage, Operational parameters effects such as feed rate,
Breakage abrasion, primary classifier rotational speed and grinding pressure
[58] Breakage mechanism Impact, shear
mechanism cleavage, secondary on particle size and distribution, and particle
cleavage shape
Mode of
[40] Force Shear and normal forces Abrasion, chipping -
breakage
The fundamental understanding of high-shear
Shear, compression, Breakage Abrasion, cleavage,
[44] Stress stirred ball mill breakage mechanism through
impact mechanism fracture
particle shape and population balance model
Destructive breakage,
Breakage Effect of regrinding with tumbling mill and
[59] - - abrasion,
mechanism stirred media mill on flotation
compression, attrition
Impact, compression, Liberation and morphological studies of particles
[49] Breakage mode - -
abrasion or attrition in stirred media mill
Impact, particle bed Minerals liberation in hammer mill and particle-
[17, 60] Breakage mechanism - -
compression bed piston–die compression
Particles morphological features and breakage
Impact, compressive, Breakage Massive fracture,
[27] Force mechanisms in stirred media mill through
abrasion mechanism abrasion, attrition
experimental and discrete element modelling
Fracture Shattering, cleavage,
[61] Stress - -
pattern abrasion and attrition
Attrition, compression,
Breakage stress Abrasion, chipping Particles flotation behavior using ball mill and
[62] shear, impact and internal Breakage mode
mechanisms and impact breakage stirred media mill
forces
[50] Breakage mechanism Impact, abrasion, attrition - - Product particle shape in semi autogenous mill
Impact, abrasion, shear, Minerals liberation using standard drop weight
[18] Breakage mechanism - -
compression tests and particle-bed piston–die compression
Impact, abrasion, Crack formation in HPGR and cone crusher and
[63] Breakage mechanism - -
compression its effect in leaching
Massive fracture,
Compression, impact, Grinding Effect of ball mill and stirred media mill on
[16] Grinding mechanism abrasion, attrition,
shear outcome particle shape and mineral liberation
chipping
Compression, impact, Impact, attrition,
Modes of Mineral liberation and crack formation using JK
[64] - scrubbing and trimming or shear and
breakage Rotary Breakage Tester
cleaving action compression breakage
7
The effect of regrinding using stirred media mill
[65] Breakage mechanism Impact fracture, attrition - -
on of rougher concentrate
Body breakage,
Impact breakage, body
incremental damage,
breakage, Sliding at Breakage Particle breakage mechanism in comminution
[47] - attrition or abrasion,
corners, Small scale body mechanism machines using discrete element modelling
rounding and
breakage
chipping
Shear, impact (rapid
Contact events (force Breakage Random fracture, Product particle shape and size using ball mill
[41] compression),
application) outcome cleavage, abrasion and stirred media mill
compression
Particle breakage in ball mill using discrete
Comminution Abrasion, crushing,
[66] - - element modelling, ball mill and drop ball
process impacting
experiments
Simulated condition for single particle breakage
[67] Breakage mechanism Impact, compression - - using short impact load cell and compression
tester
Impact breakage, Fundamental understanding of ore breakage after
Impact, compression, Breakage
[42] Force attrition, compressive compressive crushing in terms of minerals and
shear force principles
breakage elemental distribution
Tension-dominated, shear-
Finite and discrete element method (FDEM) and
[68] Fracture mechanism dominated and mix-mode Fragmentation Impact fragmentation
cohesive crack model of brittle sphere
fracture
The effect of strain rate on single particle
[69] Breakage mechanism Impact, compression - -
breakage and applied energy
8
While a variety of definitions for the term loading mechanism have been suggested, this
paper uses the definition with respect to the loading type, loading rate, and the effective
physical quantity. Figure 1 illustrates the type of loading mechanism in a comminution
process for a single particle, which can also be valid for multiple particles or a bed of
particles. Generally based on the loading direction, there are three types of loading
including compression, impact, and shear in comminution machines. Tensile stress affects
the breakage process indirectly as an effect of compression or impact loading [70].
Figure 1. Loading mechanisms in a comminution environment
According to Table 1 and in terms of loading rate, qualitative information suggested that a
fast loading rate has an instantaneous collision, while a slow rate breaks particle at slower
rates. Therefore, the impact is faster compared to compression and shear can be either slow
or fast. In a recent study, Saeidi et al. categorized the strain rate in three groups including
static, quasi-static, and dynamic [67]. Figure 2 depicts the different loadings strain rates for
various comminution environments.
9
Figure 2. Strain rates associated with different types of loading – modified after Saeidi et
al. [67]
A compression loading mechanism occurs if the particle is pressed between two media, two
tools or media and the comminution machine. Impact and strike loading mechanisms are
caused by fast loading. In comminution machines like tumbling mills, the loading rate is
controlled by the falling height of the grinding media and gravity [39]. Two mechanisms
are taking place under the effect of the gravity force. The first mechanism occurs as the
particle falls and hits the comminution machine’s internal surface (impact loading
mechanism). As the second mechanism, another particle or media, for instance, the
grinding balls in a tumbling mill, falls onto the particle (strike loading mechanism). In
another case of impact loading mechanism, the blades of the comminution machine or
accelerated media, e.g. hammer mill, hit the particle. This collision imparts a momentum
to the particle and consequently increases the speed of the particle to up to 160 m/s [38]. In
this loading mechanism, each particle can hit other particles or the liner of the comminution
machine [37,71]. Another effective loading type in a communion environment is shear.
Shear loading results in wear and shear stress, which generates the attrition loading
mechanism. This loading mechanism includes two states; (i) the particle is nipped between
two media or the surfaces of other particles, (ii) side collision of two particles as they pass
by each other. Overall, variations of loading mechanisms occur in each comminution
10
machine; however, each machine has one or several dominant loading mechanisms
[13,33,43,72]. Moreover, some loading mechanisms can have a higher probability of
occurrence in crushing rather than grinding and vice versa [32, 43].
3 Breakage mechanism
3.1 Geological materials breakage behavior
To understand the breakage process, it is necessary to understand the behavior of certain

materials under stress. Two types of materials can be distinguished based on their breakage
behavior: ductile and brittle. Brittle materials subject to stress will break without significant
plastic deformation. They have relatively large Young’s modulus, high ultimate strength
with no yield point. Therefore, the ultimate strength and breaking strength are the same.
Typical brittle materials like glass do not show any plastic deformation but fail while the
deformation is elastic. In general, almost all geological materials are brittle [73].
Material failure occurs when the applied stress and work are sufficient to break the bonds
that hold atoms together. The bond strength is due to the attractive forces between atoms.
The potential energy is at a minimum where the equilibrium spacing occurs. In order to
break the atomic bond a tensile force is required for increasing the separation distance from
the equilibrium value (x0), this force must exceed the cohesive force to sever the bond
completely. Cohesive stress (σc) can be described by the following equation:
𝜎𝑐 = √𝐸𝛾/𝑥0 (1)
where γ is the material surface energy per unit area i.e. material constant (J/m2) and E is
the Young modulus. However, the experimental work done on brittle material revealed that
the fracture strengths are usually three to four times or even more below this value. This is
due to inherent flaws (micro-defects or micro-cracks) in these materials [34]. Based on the
definition provided by Simmons and Richter, micro-cracks in rocks are the openings where
one or two of their dimensions are significantly smaller than the third. They added that in
2D micro-cracks, one dimension is significantly smaller and the crack aspect ratio (i.e. the
width to length ratio) must be less than 10-2 (usually 10-3 - 10-5). Furthermore, the length of
the micro-crack is typically less than 100 µm [74]. Geological material breakage is the
result of initiating and propagating of the induced micro-cracks. Creating new (crack)
surfaces consumes energies because surfaces carry higher energy than the body. Based on
11
Griffith’s criterion for brittle crack propagation, a crack grows when it reaches its critical
point. The total energy for the system is the sum of surface energy and elastic energy.
Change in system energy = ΔU = Increase in surface energy + Reduction in elastic energy
Increase in surface energy = 4γc
Reduction in elastic energy = πc2𝜎 2 /𝐸
Where:
γ: free surface energy per unit area (material constant) (J/m2)
c: crack length (m)
σ: Stress intensity (N/m2)
E: Young modules
Two conditions are essential for the crack propagation under stress: 1) sufficiently high
energy and 2) critical crack length. Based on the Griffith criterion for brittle crack
propagation, a crack grows when it reaches its critical length [75]. The crack would remain
stable below the critical crack length. The critical crack length (c*) is determined by the
following equation:
𝑐 ∗ = 2𝛾𝐸/𝜋𝜎 2 (2)
At constant stress when c ˃ c , the specimen fails by instantaneous nucleation. Moreover,
*
the critical stress level that a cracked body can sustain is:
𝜎𝑐 = √ 2𝛾𝐸/𝜋𝑐 ∗ (3)
At constant c when σ exceeds σc, specimen fails. According to equation (1), the critical
*
crack length for the failure will decrease if the stress increases. This results in two important
implications. Firstly, the critical stress level for a given crack length varies with materials,
viz. materials with high surface energy are tougher compared to others. Secondly, the
critical stress level decreases with crack length, i.e. the larger the crack, the easier it will
become unstable [34]. An illustration of equation (2) is shown in Figure 3, which exhibits
a maximum of energy level at the critical crack length.
12
Figure 3. Energetics of Griffith crack in uniform tension: linear elastic. c*: critical crack
length, c0: maximum crack length before failure [34].
In terms of rock and ore texture breakage, micro-cracks are not the only factor that affects
breakage behavior. All of the effective factors can be categorized in nano-scale [76] and
micro-scale [77] and are named as ore texture heterogeneity (Figure 4).
Figure 4. Ore texture heterogeneity
13
3.2 Definition of breakage mechanism
Depending on the loading mechanisms in the comminution machines, particles start to

experience energy transmission [70,78]. Griffith, as quoted by Ozkahraman postulated that
if the effective stress exceeds the local strength of the material, particle starts to break [79].
The consequence of the particle failure is referred to as the ‘breakage mechanism’.
Understanding of particle behavior under loading will help to recognize the mechanism
behind the breakage. Kelly and Spottiswood highlighted that there is little consistency for
terms defining the breakage mechanism [14]. Mode of breakage, breakage mechanism,
fracture mechanism, and fragmentation mechanism are the most commonly used
definitions, to avoid confusion, the term ‘breakage mechanism’ is used in this paper.
3.3 Types of breakage mechanisms
The breakage mechanism is often identified based on particle breakage pattern, product
particle size distribution or particle shape (Table 2). Shattering, cleavage, chipping, and
abrasion are the most frequent terms that have been used for the type of breakage
mechanisms (Table 1) that need to be discussed in more detail.
Table 2: Breakage mechanisms in a comminution environment
Feed particle Breakage pattern Product particles Breakage mechanism
Shattering
Cleavage
Chipping
Abrasion
14
3.3.1 Shattering
Shattering occurs when the loading is perpendicular to the particle and the applied energy
is a multiple of that required for the breakage. Therefore, the particle experiences a load
higher than the particle strength and what is needed for single breakage [14]. King
expressed that shattering involves multiple breakage processes, in fact, a series of steps
which the generated product particles undergo in an immediate sequential of breakage [61].
This process continues until all the energy is dissipated.
Berthiaux and Dodds have separated the shattering mechanism into two stages of a
sequential process [55]. The first stage is the primary cleavage, where the particle breaks
with large fragmentation. While in the second step of the sequential breakage, the
secondary cleavage occurs if enough energy is still available to break the particle.
Otherwise, if the energy is not high enough for another destructive breakage event, fine
particles will be produced [55]. In a similar study, Varinot et al. call the shattering process
“fracture” because of its intense stress, which leads to relatively large particle sizes [53].
Other definitions used for the shattering are destructive fracture and impact fragmentation.
Frequently, shattering is described as a process, which is caused by impact and strike
loading mechanism [16, 41, 42, 44, 53, 58, 76].
3.3.2 Cleavage
Cleavage is the breakage along preferred material surfaces i.e. the cracks move along the
cleavage planes in crystals [70]. In mineralogy, cleavage is the tendency of a mineral for
being split along crystallographic planes. This tendency comes from the structural locations
of atoms and ions in a crystal, which creates planes of relative weakness [81].
King describes cleavage as a type of breakage mechanism that generates several large
particles along with fine particles that originate from the points of applied stress [61]. The
definition is similar to that found by Kelly and Spottiswood who outlined cleavage happens
as the applied energy is sufficient for breaking a few regions of the particle [14]. However,
Kelly and Spottiswood [14] did not consider fine particles in their description as King did
[61]. On the other hand, Little et al. stated that cleavage is not a breakage mechanism and
is related to material properties [41]. It is of interest to note that, a wide range of
investigations categorized cleavage as a result of the compression loading mechanism
[42,59,64] and solely called it compression. On the other hand, King stated that similar to
15
shattering, cleavage takes place with the rapid imposition of compression stress [61]. In
general, some agreements exist that cleavage breakage mechanism occurs if the kinetic
energy is enough to break the relative weak planes, e.g. cleavage planes of the particle
[14,44,59,61].
3.3.3 Chipping
Particles with topographical properties or angular surfaces experience chipping breakage

[15,30,41]. If the applied load on an irregular particle is angular, rough corners can break
off [52,57]. On the contrary to the opinion of Austin et al. regarding the angular loading
[52], Fuerstenau and Han stated that the chipping breakage mechanism breaks off the sharp
edges and corners without a specific direction of applying load [82]. This definition is
similar to those reported by Kelly and Spottiswood expressing that chipping is due to the
breaking of edges or corners of a particle [14]. They considered chipping as a special case
of cleavage. The particle size distribution in this breakage mechanism is bimodal and the
size difference between the two peaks of size distribution is relatively small [30].
3.3.4 Abrasion
The attrition loading mechanism is one of the reasons that can cause abrasion. These terms
have been widely used interchangeably in comminution studies. Abrasion is caused by
either the tangential load or insufficient energy of the applying load and the acting forces
are parallel to the surface [57]. This definition is somehow used in some other studies
[27,64,83] which defined abrasion as an effect of scrubbing or trimming of a material.
Furthermore, Kelly and Spottiswood believed that abrasion takes place if the applied energy
is not sufficient for causing significant particle breakage [14]. Palaniandy et al. stated that
abrasion is characterized by the formation of fines and no noticeable change in the feed
particle mean diameter [58]. The latter definition was used by Hogg for modeling breakage
[54] and Little et al. for investigating the particle shape after breakage [41].
3.3.5 Concluding remarks
Based on the above-mentioned phenomena, it can be concluded that the difference between
shattering and cleavage is the amount of energy conveyed to the particle. A combination of
shattering and cleavage breakage mechanism is often called massive fracture
[27,43,48,54,56]. In the abrasion breakage mechanism, the applied energy is either local or
angular. The breakage process for chipping and abrasion is similar. These occur when the
16
particle experiences a tangential loading mechanism. When the energy is enough to chip
off coarser parts from the surface, the mechanism is chipping. However, if the energy is
only sufficient to take away fine particles off the surface the breakage mechanism is
abrasion.
3.4 The effect of breakage mechanism on particle size and shape
3.4.1 Particle size
One of the greatest challenges in comminution is to achieve the desired particle size. The
particle size is of interest because it is essential for achieving liberation, product size, and
applicable size for the downstream process [84–86]. For instance in flotation, coarse
particles (>100 µm) are too heavy to be floated while too fine particles (<25 µm) cause
selectivity problem [87,88]. Understanding the breakage and loading mechanism and their
effects on particle breakage would help to predict the particle size distribution. Therefore,
product particle size distribution is a way to distinguish the type of breakage mechanism in
a comminution environment.
Numerous studies linked the dominant breakage mechanism in the comminution

environment to the product particle size distribution [44,62,65,89]. Foundation of these is
the pioneer works by King [90] and Kelly and Spottiswood [14]. Later Hennart et al. added
quantitative information to these two works [44]. Table 3 demonstrates the particle size
distribution in frequency form for each breakage mechanism. Since the product particle
size distribution strongly depends on the size of feed particle, for comparing the effect of
loading and breakage mechanism type on product particle size distribution, the size should
be normalized (Andreatidis [31]).
17
Table 3: Size distribution in frequency form for each breakage mechanism
Reference Product particle size distribution
[90]
[14]
[44]
While various studies tried to relate the breakage mechanism to the particle size
distribution, some others have attempted to explain the effect of loading mechanism
(impact, compression, and attrition) on particle size by direct particle size measurements
[17,59,67,91–95], and analysis such as specific surface areas [17,59]. These studies are
18
generally focused on the loading mechanism type and suffer from overlooking the effects
of feed particle properties (e.g. shape) or comminution ambient conditions, e.g. the
temperature.
3.4.2 Particle shape
The shape is an important aspect of a particle that affects abrasiveness bulk density and
slurry rheology [92, 93]. In mineral processing, particle shape has a significant role on the
downstream process [98] such as collector adsorption through flotation separation [56, 95–
97]. In this regard, various descriptions have been proposed to characterize the shape by
different shape descriptors [102].
Kelly and Spottiswood noted that depending on the breakage mechanism, shape patterns
may vary (Table 2) [14]. Shattering produces highly irregular particles with sharp edges
[56] and cubical shape [103]. Chipping removes coarse uneven edges and the feed particle
has a uniform shape and the broken parts are irregular [104]. If abrasion occurs as an effect
of scrubbing, the particle shape will have the same condition as chipping but with finer
proportions of removed particles [56] and becomes spherical. If the abrasion is caused by
insufficient applied energy, the feed particle roughly conserves the particle shape and
produces irregular fine particles [104].
Particles undergoing a mechanical loading can be either single, multiple (monolayer) or a

bed of (multi-layer) particles. The form of particle layers can alter the shape of product
particles [103,105]. Holt argued that particles experiencing the direct breaking at metal
surfaces have likely cubical shape than the broken particles inside the bed [103]. Similarly,
Durney and Meloy indicated that with the choke flow particles have fewer asperities and
are blocky [105].
The relationship between breakage mechanisms and product particle size and shape is fully
recognized. A key aspect of identifying the breakage mechanism type in the comminution
is analyzing particle size and shape [106].
4 Effective factors in breakage mechanism

The breakage mechanism is known to be dependent on particle and environment properties
[54,65,107]. Examples of particle properties include particle size, shape, and ore texture
(i.e. mineral type, grain size, mineral association, and porosity, inherent cracks).
19
Comminution environment properties involve comminution temperature, loading
mechanism intensity, wet or dry comminution. Besides, particle breakage mechanism can
be influenced by external factors such as pre-treatment, for instance, chemical pre-
treatment [2,108,109], thermal, electrical or ultrasonic weakening [110]. Strategies, in
order to reduce rock strength, would alter the breakage mechanism (Figure 5). In the
following, the different particle properties and comminution environment parameters are
described.
Figure 5. Effective factors in the breakage mechanism
4.1 Material related factors
4.1.1 Particle size
The breakage behavior of a particle under loading is influenced by its size. This is due to
different internal stress patterns, which are related to the flaw size and abundance, stored
elastic energy and plastic deformation effects. As the particles become finer, the flaw size
reduces as well as the probability of occurring flaws. Thus, the stress level for finer particle
breakage increases accordingly [34]. The stored elastic energy in a particle, in the moment
of crack release, decreases proportionally with particle volume. However, the required
energy for crack propagation diminishes in cross-sections. The amount of energy required
for breaking coarse particles is higher than the finer ones [111]. Therefore, depending on
20
particle size, the same amount of applied energy to particles can lead to different breakage
mechanism. Gaudin found that coarse particles tend to be more rounded as they break with
abrasion and chipping [93]. In addition, with the same amount of energy finer particles tend
to break with shattering breakage mechanism as product particle shape is angular [27,93].
Lastly, the probability of plastic deformation increases with particle size reduction or below
the material grinding limit [27,112]. For example, it was suggested that in quartz for
particles finer than five microns the deformation has plastic behavior without any crack
propagation [113]. Later, Kendall calculated the critical diameter of a crack at which the
particle changes its breakage behavior from brittle to plastic. For instance, the size of
calcium carbonate particles at which the fracture behavior changes from elastic to plastic
is one micron [76]. In his investigation, he did not consider the particle shape, temperature,
and rate effects [114].
4.1.2 Particle shape
The shape of a particle is known to be a factor that affects the breakage process [27,115].
Investigations regarding the effect of flakiness on breakage showed that the flaky shape
particles that are loaded in their longer dimensions are weaker compared to the ones that
are loaded horizontally [116,117]. In these studies, flakes were classified as flat and
elongated shapes and non-flakes as blocky, spherical shapes. Compared to flaky particles,
the particle strength of non-flake shape lies somewhere between horizontal and vertical
directions. Moreover, particles with rounder shape require higher energy for breakage
compare to irregular particles. This is due to the accumulation of energy in the tip of the
irregular parts (i.e. the sharp edges or corners) [27]. These irregular shapes on the particle
are the source of breakage and can transfer energy sufficiently to the particle. This implies
that while for the irregular shape particles with the same amount of energy, the breakage
mechanism is shattering or cleavage, for regular particles the breakage mechanism is
abrasion.
4.1.3 Particle texture
Ore texture can be classified according to its description in geology or process mineralogy.
In geology, ore texture description focuses on qualitative aspects of the texture such as ore
formation and genesis [123, 124]. While in mineral processing and process mineralogy,
texture has a quantitative description with textural parameters such as mineral type and
21
grain size [120–122]. Figure 6 illustrates the effective factors of ore texture in the breakage
mechanism.
Figure 6. Ore texture relation to the breakage mechanism
Mineral type and association: Based on the mineral type of a particle, breakage mechanisms
can vary under the same loading mechanism. This is because of the toughness variation of
each mineral due to its specific strength [27,123]. In addition, investigations on the shape
of the product particles after breakage revealed that the shape of particles has been changed
depending on the material type [124–126]. For instance, glass tends to have slimmer shape
while limestone shows more convexity [127] and minerals with distinctive cleavage were
found to have a cleavage characteristic shape [56,103]. Cleavage breakage can occur due
to the weak planes of mineral cleavage, which is named as weak grain boundary
heterogeneity. Griffith introduced micro-cracks as a reason for the macroscopic failure of
a particle under stress [75]. The new experimental and acoustic emission studies revealed
that the failure process involves a multiplicity of micro-cracks [128–130]. The number of
micro-cracks in the structure of rock increases the probability of its breakage [131].
Therefore, as the number of micro-cracks increases in the rock structure, rock tends to break
easier under a certain loading mechanism intensity. Investigations showed that the number
of inherent micro-cracks variates based on the dominant mineral (matrix) and inclusions.
Nur and Simmons investigated the effect of quartz as a matrix and inclusion. They found
that during the geological formation due to the higher amounts of bulk moduli and thermal
22
expansion of quartz has a significant effect on the number of natural micro-cracks [132].
Therefore, based on minerals type and their association, the breakage behavior can be
changed which ultimately influences the type of breakage mechanism.
Grain size: Minerals grain size is categorized as one of the important aspects that determine
the rock strength. Several systematic approaches have been taken in order to obtain an
empirical correlation between grain size and rock strength [133–137]. Hatzor and Palchik
reported that fracture initiation stress is inversely related to the grain size of dolomite [134].
This feature of ore texture is called geometrical heterogeneity.
Porosity: Porosity is a micromechanical factor that needs to be taken into account when
analyzing the initiation and propagation of stress-induced cracks [138–140]. Drawing on
an extensive range of sources, Li and Aubertin reviewed the effect of porosity on the
strength of materials. They indicate that there is a relationship between the porosity of the
rock and its strength [141]. It was also reported that the rock strength reduces as the
porosity increases [139]. In recent studies, the effect of porosity on various rock types’
strength has been investigated [140,142,143]. Similarly, the effect of apparent porosity may
affect particle strength indirectly by its water content and temperature. Kawatra et al.
reported that the rock strength is reduced as the water freezes in the pores and cracks of a
rock [144].
4.2 Process related parameters
4.2.1 The loading mechanism intensity
The most important condition in the failure process of a particle is to provide the applied
stress and consequently sufficient energy higher than the strength of the particle. However,
the energy increment after a certain amount may cause a shattering breakage mechanism
[61]. Comminution machines employing high energy inputs, produce particles containing
a high proportion of newly created surfaces with irregular shapes indicating the shattering
and cleavage breakage mechanisms [56, 58, 142]. On the other hand, lowering loading
mechanism intensity leads to abrasion and chipping breakage mechanisms
[55, 58, 142, 143].
The motion of the grinding media and its material type affects the loading mechanism
intensity in tumbling and stirred media mill. If the grinding media is too small, particles
cannot be nipped [147,148]. The other effect of the grinding media is its specific gravity.
23
In this regard, the breakage mechanism of particles ground with a high specific gravity
media is shattering rather than abrasion [149]. In a study on stirred media mill, Khonthu
showed that bead size, density, and distribution are the effective parameters, which
determine the breakage mechanism [62].
The loading mechanism rate may also affect the breakage mechanism [150]. For instance,
in the attrition loading mechanism, the rate of induced velocity determines the breakage
mechanism. At low velocities, the particles tend to show the abrasion breakage mechanism
whereas, at high velocities, particles tend to show chipping [107]. Accordingly, the
rotational speed in the stirred media mill is low where particles break by the abrasion
breakage mechanism.
4.2.2 Temperature
The friction of the media and particles inside the comminution machine elevates the
temperature of the comminution environment. The temperature can rise to more than 100°C
in tumbling mills [150]. It was shown that the particle strength of quartz reduced after
heating above 573°C and quenching in aqueous media [151]. Delle Piane et al. showed that
increasing the rock temperature results in wider crack openings. Moreover, raising rock
temperature will reduce its compression strength through the diminution of the frictional
coefficient [152]. They also showed that the space between minerals grain boundaries in
marble samples increases drastically when the temperature is increased from 300°C to
600°C. In this case, the breakage mechanism may change to cleavage, as the grain
boundaries are weak. However, in conventional comminution environments, the
temperature does not increase to the aforementioned temperature. Kawatra et al.
investigated rock breakage in the cryogenic condition. It was stated that freezing the rock
increases the brittleness and creates more cracks during blasting. However, the breakage
test results in temperatures between -25 to 25°C showed that rock brittleness does not vary
noticeably [144].
4.2.3 Wet or dry
The effect of wet and dry processes are studied through viscosity and energy consumption.
Increasing the viscosity of the slurry decreases the suspensions turbulence [107] and
particle movement [153] which results in the de-escalation of grinding. Therefore,
24
insufficient energy for destructive breakage takes away small portions of material and
breakage mechanism becomes abrasion.
Dry grinding consumes more energy compared to wet grinding and excessive energy usage
appears in the form of defects [85]. The dry grinding samples have a relatively higher
amount of microstructural defects [154] which can be due to the higher amount of
shattering or cleavage breakage mechanism [56]. On the other hand, the wet grinding
samples have a smoother and cleaner surface [154] which can be due to abrasion and
chipping breakage mechanisms. By contrast, Frances et al. found that a wet or dry condition
in various grinding environments does not affect the breakage mechanism [43].
4.3 Pre-treatments
The third major parameter for changing the particle breakage mechanism is applying pre-
treatments. Pre-treatments are used to decrease the strength of the material and can be
classified according to the applied method including energy applied and grinding aids
[155]. Energy applied methods include microwave exposure, thermal breakage, thermal
shock, high-voltage pulse breakage, and ultrasonic breakage [110]. Moreover, it was shown
that reagent addition before comminution can reduce particle strength [2,108,109].
5 Breakage mode
One of the main purposes of breakage in mineral comminution is to liberate the mineral
phases from each other. King attempted to explain mineral liberation based on the breakage
mode [61]. In his work, the breakage mode has been expressed as random and non-random
breakage. The breakage mode is random when it is independent of mineral texture.
Oppositely, it is non-random breakage when it depends on the mineral texture. King
classified non-random breakage as selective breakage, differential breakage, preferential
breakage, phase boundary breakage, liberation by detachment, and boundary region
fracture [61]. Recently, Little et al. suggested that the first three breakage modes should be
categorized as preferential breakage and the other three modes are phase boundary
breakage (Figure 7) [16]. Preferential breakage is used to refer to situations in which
breakage more frequently occurs in one mineral. Phase boundary breakage is a type of
breakage in which breakage is happening preferentially along boundaries rather than across
the phases [25].
25
Another definition of breakage mode is based on the initiated and propagated cracks
[137,140,156,157]. In this regard, two types of cracks are recognized including grain
boundary and transgranular cracks. The former is related to cracks along the grain
boundaries while the latter is the cracks crossing the grains. The grain boundary cracks lead
to phase boundary breakage mode and the transgranular cracks lead to preferential or
random breakage mode. The grain boundary cracks give a better liberation degree
[49,70,158,159]. Overall, findings indicate that the type of breakage mode can affect the
liberation type and its degree, which consequently affects the downstream separation
process.
6 Loading and breakage mechanisms in comminution

machines
In comminution machines, there is a variety of loading and breakage mechanism. Different
methods have been proposed to determine the breakage and loading mechanism types in
comminution machines. These methods either come from experimental data including
product particle size [43,54,65,89,148], shape [43,65], computational methods including
discrete element method (DEM) [47,160–162], particle fragmentation (breakage function)
[53,163,164], and rate of breakage (selection function) studies [52,54,89,148,165]. These
investigations have attempted to explain the loading mechanism based on the represented
breakage mechanism mostly as shattering for impact and strike, compression for cleavage,
and attrition with abrasion.
In most of the investigations, the possibility of various breakage mechanisms within a

comminution process has been emphasized [53]. However, one or two breakage
mechanisms can have a dominant effect [89]. Table 4 summarizes the most commonly used
comminution machines in mineral processing with their dominant loading mechanism.
26
Table 4: Comminution machines in mineral processing with their dominant loading mechanism
Loading mechanism
Comminution
Sub category Impact/ Reference
machine type Compression Attrition
Strike
 [166]
  [167]
Ball mill   [50]
 [65]
  [89]
   [52]
Tumbling mill   [168]
Autogenous mill  [166]
 [169]
  [170]
  [65,171]
All
 [10]
Vertical  [65]
Horizontal  [10]
Stirred media
 [172]
mill Without sub
 [57]
category
 [31]
[63,86,180,181,167,1
HPGR - 
73–179]
Jaw crusher  [30,171,182]
Without sub
 [30,171,182]
category
Crusher Without sub
 [30,33,171,182]
category
Without sub
 [30,171,182]
category
All  [30,33,37,101,183]
Vertical shaft
 [37,183]
Impact/ impact crusher
hammer
Horizontal shaft
crusher/mill  [37,184]
impact crusher
Multi-shaft
  [38]
impact mill
Jet mill -   [58]
7 Effects of loading and breakage mechanisms on

minerals liberation and breakage mode
Wills and Atkinson pointed out that more investigation is needed to understand the effects
of breakage and loading mechanisms on the breakage mode. They suggested that shifting
the breakage mode to the phase boundary breakage could help to increase the liberation
degree [70]. The effect of the loading mechanism on mineral liberation has been a subject
27
of interest for several investigations. The breakage and loading mechanisms have been
studied either with the comminution machine with its dominant loading and breakage
mechanism or by designing a simulated test with similar characteristics.
Preliminary studies on synthesized samples have shown that the loading mechanism type
and consequently, the breakage mechanism might improve the liberation [185]. The
liberation improvement can be directly studied either through the produced particles
[17,26,28,31,64,186,187] or indirectly through the type of the initiated cracks [188] of
different breakage mechanisms and loadings.
Previous findings on the effect of loading mechanism type on mineral liberation have been
inconsistent and contradictory (Table 5). For instance, some investigations have shown that
the loading mechanism has an effect on mineral liberation improvement [18,26–28]. In
contrast, other studies have reported that the sample liberation degree does not depend on
the comminution machine type with its dominant loading mechanism [31,60,189–191].
Bérubé and Marchand found that liberation depends on the textural characteristics of
minerals and can vary within the same ore body. However, they investigated only one ore
texture and did not consider other ore textures on liberation [190]. Similarly, Andreatidis
showed that liberation is independent of the loading mechanism and added that liberation
is affected by the ore texture [31]. Table 6 summarized the results of the loading
mechanisms on liberation improvement. Based on this table, the compression loading
mechanism is the most effective mechanism for mineral liberation improvement.
28
Table 5: The investigated liberation improvements in terms of loading mechanisms
Liberation Loading
Textural Comminution
type/ Breakage Ore type Target minerals mechanism Concluding remarks Reference
feature environment
mode feature
Sphalerite has Vibratory rod as A higher recovery in flotation was
Liberation finer grain Vibratory rod impact achieved in HPGR for both sphalerite
Lead-Zinc ore Galena and sphalerite [187]
degree compared to mill and HPGR HPGR as and galena. Sphalerite is more affected
galena compression than galena.
Vibratory rod as
HPGR gives a better liberation degree
Liberation Vibratory rod impact
Lead-Zinc ore Galena and sphalerite - than rod mill. Sphalerite is more [177]
degree mill and HPGR HPGR as
affected than galena.
compression
Breakage
HPGR causes breakage along grain
mode (phase Clinker minerals (alite HPGR and ball HPGR as
- - boundaries in both minerals compared [17]
boundary and belite) mill compression
to ball mill
breakage)
Breakage
Slow A range of
mode (phase Copper The slower the compression rate the
Chalcocite - compression applying force [26]
boundary sulphide ore better liberation in grain boundary
test rate
breakage)
Frist ore: a low First ore: quartz,
grade iron – feldspar, magnetite, Piston-die press
Mineralogical
Liberation oxide chalcopyrite Piston-die press as compression No difference in minerals liberation
differences of [60]
degree Second ore: a Second ore: Hammer mill Hammer mill as degree
two ores
high-grade zinc sphalerite, quartz, impact
ore illite, siderite
Breakage
Impact: liberation in grain boundary
mode (phase
Lead-Zinc Ore Galena, quartz - Stirred mill - Attrition: liberation through grain [27]
boundary
boundaries
breakage)
While jaw crusher product has more
Breakage HPGR and jaw HPGR as random micro-cracks, HPGR samples
Iron ore Hematite - [188]
mode crusher compression has more preferential and phase
boundary micro-cracks
29
Breakage
Slow A range of
mode (phase Copper The slower the compression rate the
Chalcocite - compression applying force [28]
boundary sulphide ore better liberation in grain boundary
test rate
breakage)
For both ore types, the liberation
Mineralogical
A range of degree in coarse size fraction is better
difference in Piston-die press
Liberation Two copper applying force in compression.
Chalcopyite terms of Drop weight [18]
degree ores rate with a For both ore types, the liberation
gangue test
specific energy degree in fine size fraction is same in
minerals
compression and impact.
Two ores Mineralogical
Three energy
including: For gold-bearing differences of
JK Rotary levels:
Liberation A gold-bearing pyrite ore: pyrite ores No difference in liberation degree of
Breakage 0.1 kWh/t, 1.0 [191]
degree pyrite ore and a For copper sulphide Grain size various applied energy
Tester kWh/t and 2.5
copper sulphide ore: chalcopyrite distribution of
kWh/t
ore target minerals
Table 6 Summary of loading mechanism effects on mineral liberation
Impact [27] [187] [177]
Strike [187] [188]

Positive effect
Compression [17] [18] [26] [28] [188]
Attrition
No effect All loading mechanism [60] [191]
Attrition [60] [191]
Compression [27]
Negative effect
Strike
Impact [187]
30
Based on the analysis of Table 5, some investigations on single and multi-particle level has
been done. However, the following conclusions can be drawn for a proper breakage mode
and liberation studies:
1. Quantitative textural features have not been used in order to characterize the ore
texture. The influencing features of ore texture are mineralogy, grain size, and
minerals association [122].
2. The loading mechanisms should be investigated more systematically in terms of
energy intensity and strain rate. Each loading mechanism should be studied
separately through specific machines e.g. drop weight tester [192].
3. Little attempt has been made so far to explore the combined effects of loading
mechanism and ore texture.
4. Comminution of single, multiple (monolayer) or bed of (multi-layer) particles with
loading mechanism and textural effects should be investigated.
5. A more fundamental understanding of loading mechanism and texture effects are
required. This approach would help to understand the type of breakage mode (e.g.
phase boundary breakage) and its effect on mineral liberation.
The ore texture effect can be studied through the features of Figure 6. In this regard, for
instance, a combination of mineral type and mineral association shows the evidence of
natural micro-cracking. The main role of micro-cracks is to act as a stress concentrating
mechanism and are the source of the local stress concentration which ultimately causes
rock failure [193]. The other effect of mineral type can be the minerals with weak grain
boundaries. For instance, Mahabadi et al. showed the effect of biotite due to its cleavage
planes on macro-crack formation and breakage [194]. These ore textural features are in the
scale of micro. However, as it was shown in Figure 4, ore textural heterogeneity goes
further to crystalline size. Therefore, further investigations in the crystalline structure of
ore texture shows the effect of elemental association and distribution in the breakage
behavior of minerals.
8 Conclusions
Nowadays, with the depletion of high grade and coarse-grained ores, the low-grade and
fine-grained ores are gaining more attention. The processing of low-grade ores has two
consequences for the minerals industry; first to comminute larger tonnages of mineral raw
31
material and second to fine grind the raw materials, which leads to higher energy
consumption. Therefore, a proper selection of the comminution machine leads to having
the proper product with the least energy consumption. The strategy should, therefore, be (i)
to design of flowsheets and select equipment based on particle breakage behavior and (ii)
to design new types of comminution machines. In order to tackle this issue, a fundamental
understating of micro-processes is needed that leads to a better understanding of the
breakage process.
This review has analyzed the fundamentals behind the particle breakage in comminution
machines. As a result, this review has classified particle breakage with breakage
fundamentals including loading mechanism, breakage mechanism, and breakage mode.
The loading mechanism is defined as the action of physical quantities, which are applied to
a particle or several particles in order to introduce mechanical stress. The consequence of
particle failure is the breakage mechanism. The breakage mode defines the particle
breakage in terms of being random or non-random. It was shown that breakage
fundamentals might have or not have an effect on the downstream processes including
minerals liberation. Much uncertainty still exists about the relationship between textural
effect and breakage fundamentals.
After reviewing the literature published, it is clear that further research is needed to
understand the interaction between loading and breakage mechanisms with respect to
particle size reduction and mineral liberation as the basis for designing new comminution
machines. The lines of study that are critical for a better understanding of the ore texture
breakage phenomena in comminution machines have been highlighted. This will prove
useful in expanding our understanding of ore texture in breakage and its effect on designing
new comminution machines or comminution flowsheets.
Acknowledgments
The financial support of the Centre for Advanced Mining and Metallurgy (CAMM), a
strategic research environment established at Luleå University of Technology funded by
the Swedish government, is gratefully acknowledged.
References
[1] B.J. Skinner, A Second Iron Age Ahead ?, in: Sigma Xi, The Scientific Research
32
Society: New Haven, CT, United States, 1984: pp. 258–269.
[2] D.W. Fuerstenau, Grinding Aids, KONA Powder Part. J. 13 (1995) 5–18.
https://doi.org/10.14356/kona.1995006.
[3] V. Deniz, Comparisons of dry grinding kinetics of lignite, bituminous coal, and
petroleum coke, Energy Sources, Part A Recover. Util. Environ. Eff. 35 (2013) 913–
920. https://doi.org/10.1080/15567036.2010.514591.
[4] M. Reichert, C. Gerold, A. Fredriksson, G. Adolfsson, H. Lieberwirth, Research of

iron ore grinding in a vertical-roller-mill, Miner. Eng. 73 (2015) 109–115.
https://doi.org/10.1016/j.mineng.2014.07.021.
[5] R.P. King, A success story that has not yet ended, in: Proc. XVIII, Int. Miner.
Process. Congr. Australas. IMM, 1993: pp. 39–45.
[6] R.P. King, Comminution and liberation of minerals, Miner. Eng. 7 (1994) 129–140.
https://doi.org/10.1016/0892-6875(94)90059-0.
[7] T. Yokoyama, Y. Inoue, Selection of Fine Grinding Mills, in: Handb. Powder
Technol., 2007: pp. 487–508. https://doi.org/10.1016/S0167-3785(07)12013-3.
[8] S.C. Chelgani, M. Parian, P.S. Parapari, Y. Ghorbani, J. Rosenkranz, A comparative

study on the effects of dry and wet grinding on mineral flotation separation–a review,
J. Mater. Res. Technol. (2019) 1–8. https://doi.org/10.1016/j.jmrt.2019.07.053.
[9] S. Mohd, J. Brahim, A.A. Latiffi, M.S. Fathi, A.N. Harun, Developing building
information modelling (BIM) implementation model for project design team, 2017.
https://doi.org/10.1002/ejoc.201200111.
[10] X. Chen, Y. Peng, D. Bradshaw, The effect of particle breakage mechanisms during
regrinding on the subsequent cleaner flotation, Miner. Eng. 66 (2014) 157–164.
[11] D. Bradshaw, The role of process mineralogy in improving the process performance
of complex sulphide ores, in: XXVII Int. Miner. Process. Congr. Santiago Chile Role
Process Mineral. Improv. Process Perform. Complex Sulphide Ores, 2014: pp. 1–
23. https://doi.org/10.1007/978-3-658-07529-3.
[12] P. Lamberg, Particles-the bridge between geology and metallurgy, in: Proc. Conf.
33
Miner. Eng. Luleå, Sweden, 2011.
[13] R. Hogg, H. Cho, A review of breakage behavior in fine grinding by stirred-media

milling, KONA Powder Part. J. 18 (2000) 9–19.
https://doi.org/10.14356/kona.2000007.
[14] E.G. Kelly, D.J. Spottiswood, Introduction to mineral processing, John Wiley, The
Australian Mineral Foundation, 1982.
[15] T.G. Vizcarra, E.M. Wightman, N.W. Johnson, E. V. Manlapig, The effect of
breakage method on the shape properties of an iron-oxide hosted copper-gold ore,
Miner. Eng. 24 (2011) 1454–1458. https://doi.org/10.1016/j.mineng.2011.07.007.
[16] L. Little, A.N. Mainza, M. Becker, J.G. Wiese, Using mineralogical and particle
shape analysis to investigate enhanced mineral liberation through phase boundary
fracture, Powder Technol. 301 (2016) 794–804.
https://doi.org/10.1016/j.powtec.2016.06.052.
[17] I.B. Celik, M. Oner, The influence of grinding mechanism on the liberation
characteristics of clinker minerals, Cem. Concr. Res. 36 (2006) 422–427.
https://doi.org/10.1016/j.cemconres.2005.09.011.
[18] O. Ozcan, H. Benzer, Comparison of different breakage mechanisms in terms of

product particle size distribution and mineral liberation, Miner. Eng. 49 (2013) 103–
108. https://doi.org/10.1016/j.mineng.2013.05.006.
[19] N. Kotake, M. Kuboki, S. Kiya, Y. Kanda, Influence of dry and wet grinding
conditions on fineness and shape of particle size distribution of product in a ball mill,
Adv. Powder Technol. 22 (2011) 86–92. https://doi.org/10.1016/j.apt.2010.03.015.
[20] L.M. Tavares, Particle weakening in high-pressure roll grinding, Miner. Eng. 18
(2005) 651–657. https://doi.org/10.1016/j.mineng.2004.10.012.
[21] A.J. Clarke, B.A. Wills, Technical note Enhancement of cassiterite liberation by high
pressure roller comminution, Miner. Eng. 2 (1989) 259–262.
https://doi.org/10.1016/0892-6875(89)90046-0.
[22] X. Xiao, G. Zhang, Q. Feng, S. Xiao, L. Huang, X. Zhao, Z. Li, The liberation effect
of magnetite fine ground by vertical stirred mill and ball mill, Miner. Eng. 34 (2012)
63–69. https://doi.org/10.1016/j.mineng.2012.04.004.
34
[23] S. Hamid, P. Alfonso, H. Anticoi, E. Guasch, J. Oliva, M. Dosbaba, M. Garcia-
Valles, M. Chugunova, Quantitative Mineralogical Comparison between HPGR and
Ball Mill Products of a Sn-Ta Ore, Minerals. 8 (2018) 151.
https://doi.org/10.3390/min8040151.
[24] J. Cao, L. Liu, Y. Han, A. Feng, Comminution behavior and mineral liberation
characteristics of low-grade hematite ore in high pressure grinding roll,
Physicochem. Probl. Miner. Process. 55 (2019) 575–585.
https://doi.org/10.5277/ppmp18169.
[25] R.P. King, C.L. Schneider, Mineral liberation and the batch comminution equation,
Miner. Eng. 11 (1998) 1143–1160. https://doi.org/10.1016/S0892-6875(98)00102-
2.
[26] D. Garcia, C.L. Lin, J.D. Miller, Quantitative analysis of grain boundary fracture in
the breakage of single multiphase particles using X-ray microtomography
procedures, Miner. Eng. 22 (2009) 236–243.
[27] R.A. Roufail, The effect of stirred mill operation on particles breakage mechanism
and their morphologyical features, Dissertation/Thesis, University Of British
Columbia, 2011.
[28] W. Xu, N. Dhawan, C.-L. Lin, J.D. Miller, Further study of grain boundary fracture
procedures, Miner. Eng. 46–47 (2013) 89–94.
[29] H. Rumpf, Die Einzelkornzerkleinerung als Grundlage einer technischen

Zerkleinerungswissenschaft, Chemie Ing. Tech. 37 (1965) 187–202.
https://doi.org/10.1002/cite.330370303.
[30] G. Unland, The Principles of Single-Particle Crushing, in: Handb. Powder Technol.,
2007: pp. 117–225. https://doi.org/10.1016/S0167-3785(07)12007-8.
[31] J.P. Andreatidis, Breakage Mechanisms and Resulting Mineral Liberation in a Bead
Mill, Dissertation/Thesis, University of Queensland, 1995.
[32] T. Mütze, K. Husemann, Compressive stress: Effect of stress velocity on confined
35
particle bed comminution, Chem. Eng. Res. Des. 86 (2008) 379–383.
https://doi.org/10.1016/j.cherd.2007.11.007.
[33] C.M. Evertsson, Cone crusher performance, Dissertation/Thesis, Chalmers

University of Technology, 2000. https://doi.org/10.13140/RG.2.1.2212.7526.
[34] T.L. Anderson, Fracture Mechanics: Fundamentals and Applications, Fourth Edi,
2012. https://doi.org/10.1016/j.jmps.2010.02.008.
[35] R.L. Kranz, Microcracks in rocks: A review, Tectonophysics. 100 (1983) 449–480.
https://doi.org/10.1016/0040-1951(83)90198-1.
[36] R. Ulusay, The ISRM Suggested Methods for Rock Characterization, Testing and
Monitoring: 2007-2014, 2015. https://doi.org/10.1007/978-3-319-07713-0.
[37] M.D. Sinnott, P.W. Cleary, Simulation of particle flows and breakage in crushers
using DEM: Part 2 - Impact crushers, Miner. Eng. 74 (2015) 163–177.
[38] R.J. Bracey, N.S. Weerasekara, M.S. Powell, Performance evaluation of the novel
multi-shaft mill using DEM modelling, Miner. Eng. 98 (2016) 251–260.
[39] M.H. Wang, R.Y. Yang, A.B. Yu, DEM investigation of energy distribution and
particle breakage in tumbling ball mills, Powder Technol. 223 (2012) 83–91.
[40] D.W. Fuerstenau, K.N. Han, Particle Strength and Breakage Energy Requirement.,
in: Princ. Miner. Process., 2009: pp. 63–71.
[41] L. Little, A.N. Mainza, M. Becker, J. Wiese, Fine grinding: How mill type affects
particle shape characteristics and mineral liberation, Miner. Eng. 111 (2017) 148–
[42] L.G. Leon, Analysis and Modelling of Mineral and Element Composition in
Compression Breakage, Dissertation/Thesis, Chalmers University of Technology,
2018.
[43] C. Frances, N. Le Bolay, K. Belaroui, M.N. Pons, Particle morphology of ground

gibbsite in different grinding environments, Int. J. Miner. Process. 61 (2001) 41–56.
36
https://doi.org/10.1016/S0301-7516(00)00025-9.
[44] S.L.A. Hennart, W.J. Wildeboer, P. van Hee, G.M.H. Meesters, Identification of the
grinding mechanisms and their origin in a stirred ball mill using population balances,
Chem. Eng. Sci. 64 (2009) 4123–4130. https://doi.org/10.1016/j.ces.2009.06.031.
[45] W.A. Nash, Schaum’s outline of theory and problems of strength of materials,
McGraw-Hill Professional, 1988.
[46] T. Inoue, K. Okaya, Grinding mechanism of centrifugal mills - A simulation study

based on the discrete element method, Int. J. Miner. Process. 44–45 (1996) 425–435.
https://doi.org/10.1016/0301-7516(95)00049-6.
[47] P.W. Cleary, R.D. Morrison, Comminution mechanisms, particle shape evolution
and collision energy partitioning in tumbling mills, Miner. Eng. 86 (2016) 75–95.
[48] L. Little, The development and demonstration of a practical methodology for fine
particle shape characterisation in minerals processing, Dissertation/Thesis,
University of Cape Town, 2016.
[49] R. Roufail, B. Klein, Mineral Liberation and Particle Breakage in Stirred Mills, Can.
Metall. Q. 49 (2010) 419–428. https://doi.org/10.1179/cmq.2010.49.4.419.
[50] P. Pourghahramani, Effects of ore characteristics on product shape properties and

breakage mechanisms in industrial SAG mills, Miner. Eng. 32 (2012) 30–37.
[51] A.J. Lynch, ed., Comminution handbook, Australasian Institute of Mining and
Metallurgy, 2015.
[52] L.G. Austin, C.A. Barahona, J.M. Menacho, Fast and slow chipping fracture and
abrasion in autogenous grinding, Powder Technol. 46 (1986) 81–87.
https://doi.org/10.1016/0032-5910(86)80102-4.
[53] C. Varinot, S. Hiltgun, M.N. Pons, J. Dodds, Identification of the fragmentation

mechanisms in wet-phase fine grinding in a stirred bead mill, Chem. Eng. Sci. 52
(1997) 3605–3612. https://doi.org/10.1016/S0009-2509(97)89693-5.
[54] R. Hogg, Breakage mechanisms and mill performance in ultrafine grinding, Powder
37
Technol. 105 (1999) 135–140. https://doi.org/10.1016/S0032-5910(99)00128-X.
[55] H. Berthiaux, J. Dodds, Modelling fine grinding in a fluidized bed opposed jet mill.
Part I: Batch grinding kinetics, Powder Technol. 106 (1999) 78–87.
https://doi.org/10.1016/S0032-5910(99)00049-2.
[56] E. Kaya, R. Hogg, S. Kumar, Particle Shape Modification in Comminution, KONA

Powder Part. J. 20 (2002) 188–195. https://doi.org/10.14356/kona.2002021.
[57] B.A. Wills, T. Napier-Munn, Mineral Processing Technology An Introduction to the

Practical Aspects of Ore Treatment and Mineral, 7th Editio, Elsevier Science &
Technology Books, 2005. https://doi.org/10.1016/B978-075064450-1/50008-4.
[58] S. Palaniandy, K.A.M. Azizli, H. Hussin, S.F.S. Hashim, Effect of operational

parameters on the breakage mechanism of silica in a jet mill, Miner. Eng. 21 (2008)
[59] X. Ye, S. Gredelj, W. Skinner, S.R. Grano, Regrinding sulphide minerals - Breakage
mechanisms in milling and their influence on surface properties and flotation
behaviour, Powder Technol. 203 (2010) 133–147.
[60] T.G. Vizcarra, E.M. Wightman, N.W. Johnson, E. V. Manlapig, The effect of
breakage mechanism on the mineral liberation properties of sulphide ores, Miner.
Eng. 23 (2010) 374–382. https://doi.org/10.1016/j.mineng.2009.11.012.
[61] R.P. King, Modeling and Simulation of Mineral Processing Systems, 2nd ed.,
Society for Mining, Metallurgy & Exploration, 2012.
https://doi.org/10.1016/C2009-0-26303-3.
[62] T. Khonthu, Investigation of the flotation behaviour of ball mill and IsaMill
products, Dissertation/Thesis, University of Cape Town, 2012.
[63] Y. Ghorbani, A.N. Mainza, J. Petersen, M. Becker, J.P. Franzidis, J.T. Kalala,
Investigation of particles with high crack density produced by HPGR and its effect
on the redistribution of the particle size fraction in heaps, Miner. Eng. 43–44 (2013)
[64] R. Mariano, Measurement and modelling of the liberation and distribution of

minerals in comminuted ores, Dissertation/Thesis, The University of Queensland,
38
2016. https://doi.org/10.14264/uql.2016.1081.
[65] V. Resabal, Investigation of the Breakage Mechanism in High Speed Disk-type

Impeller Stirred Mill and its Influence on the Mineral Liberation,
Dissertation/Thesis, University of Queensland, 2016.
[66] Z. Yin, Y. Peng, Z. Zhu, Z. Yu, T. Li, Impact load behavior between different charge
and lifter in a laboratory-scale mill, Materials (Basel). 10 (2017).
https://doi.org/10.3390/ma10080882.
[67] F. Saeidi, M. Yahyaei, M. Powell, L.M. Tavares, Investigating the effect of applied
strain rate in a single breakage event, Miner. Eng. 100 (2017) 211–222.
[68] G. Ma, Y. Zhang, W. Zhou, T.T. Ng, Q. Wang, X. Chen, The effect of different
fracture mechanisms on impact fragmentation of brittle heterogeneous solid, Int. J.
Impact Eng. 113 (2018) 132–143. https://doi.org/10.1016/j.ijimpeng.2017.11.016.
[69] D. Gong, S. Nadolski, C. Sun, B. Klein, J. Kou, The effect of strain rate on particle
breakage characteristics, Powder Technol. 339 (2018) 595–605.
[70] B.A. Wills, K. Atkinson, Some observations on the fracture and liberation of mineral
assemblies, Miner. Eng. 6 (1993) 697–706. https://doi.org/10.1016/0892-
6875(93)90001-4.
[71] R.W. Barley, J. Conway-Baker, R.D. Pascoe, J. Kostuch, B. McLoughlin, D.J.

Parker, Measurement of the motion of grinding media in a vertically stirred mill
using positron emission particle tracking (PEPT) Part II, Miner. Eng. 17 (2004)
[72] W. Schubert, J. Tomas, Liberation of Valuables Embedded in Particle Compounds

and Solid Waste, in: Handb. Powder Technol., 2007: pp. 989–1018.
https://doi.org/10.1016/S0167-3785(07)12027-3.
[73] E. Hoek, C.D. Martin, Fracture initiation and propagation in intact rock - A review,
J. Rock Mech. Geotech. Eng. 6 (2014) 287–300.
https://doi.org/10.1016/j.jrmge.2014.06.001.
[74] G. Simmons, D. Richter, Microcracks in rocks, Phys. Chem. Miner. Rocks. (1976)
39
105–137.
[75] A.. Griffith, The Phenomena of Rupture and Flow in Solids, Philos. Trans. 221
(1920) 163–198.
[76] C. Prasher, Crushing and grinding process handbook, Chichester: Wiley, 1987.
[77] P. Hamdi, D. Stead, D. Elmo, Characterizing the influence of stress-induced

microcracks on the laboratory strength and fracture development in brittle rocks
using a finite-discrete element method-micro discrete fracture network FDEM-
μDFN approach, J. Rock Mech. Geotech. Eng. 7 (2015) 609–625.
https://doi.org/10.1016/j.jrmge.2015.07.005.
[78] H. Nematollahi, Mineral processing handbook, University of Tehran Press, 2002.
[79] H.T. Ozkahraman, Breakage mechanisms and an encouraging correlation between

the Bond parameters and the friability value, J. South. African Inst. Min. Metall. 110
(2010) 153–159.
[80] P. yun Xu, C. Hu, M. Gan, J. Li, X. Pan, H. qi Ye, Analyses on uniformity of particles
under HPGR finished grinding system, J. Cent. South Univ. 25 (2018) 1003–1012.
https://doi.org/10.1007/s11771-018-3800-1.
[81] S.K. Haldar, J. Tišljar, Introduction to Mineralogy and Petrology, Elsevier, 2014.
https://doi.org/10.1016/C2012-0-03337-6.
[82] M.C. Fuerstenau, K.N. Han, eds., Principles of Mineral Processing, 2018.
https://doi.org/10.1016/B978-0-12-814022-2.00013-7.
[83] X. Ye, S. Gredelj, W. Skinner, S.R. Grano, Evidence for surface cleaning of sulphide
minerals by attritioning in stirred mills, Miner. Eng. 23 (2010) 937–944.
[84] T.P. Meloy, N. Clark, Liberation of target material from locked particles: A
theoretical analysis, Part. Sci. Technol. 3 (1985) 15–26.
https://doi.org/10.1080/02726358508906424.
[85] D. Feng, C. Aldrich, A comparison of the flotation of ore from the Merensky Reef
after wet and dry grinding, Int. J. Miner. Process. 60 (2000) 115–129.
https://doi.org/10.1016/S0301-7516(00)00010-7.
40
[86] M. Athayde, M.C. Bagatini, Iron Ore Concentrate Particle Size Controlling Through
Application of Microwave at the HPGR Feed, Mining, Metall. Explor. 36 (2019)
353–362. https://doi.org/10.1007/s42461-018-0013-y.
[87] T. Miettinen, J. Ralston, D. Fornasiero, The limits of fine particle flotation, Miner.
Eng. 23 (2010) 420–437. https://doi.org/10.1016/j.mineng.2009.12.006.
[88] H. Soto, G. Barbery, Flotation of coarse particles in a counter-current column cell,

Miner. Metall. Process. 8 (1991) 16–21. https://doi.org/10.1007/bf03402925.
[89] M. Gao, E. Forssberg, Prediction of product size distributions for a stirred ball mill,
Powder Technol. 84 (1995) 101–106. https://doi.org/10.1016/0032-5910(95)02990-
J.
[90] R.P. King, Modeling and simulation of mineral processing systems, 1nd ed., Society
for Mining, Metallurgy & Exploration, 2001.
[91] H. Cimilli, An investigation of the principles of laboratory-scale particle-bed

comminution, Dissertation/Thesis, Middle East Technical University, 2008.
[92] N.A. Aydoǧan, L. Ergün, H. Benzer, High pressure grinding rolls (HPGR)
applications in the cement industry, Miner. Eng. 19 (2006) 130–139.
[93] A.M. Gaudin, An investigation of crushing phenomena, Trans. AIME. 73 (1926) 23.
[94] L.. Tavares, R.. King, Single-particle fracture under impact loading, Int. J. Miner.
Process. 54 (1998) 1–28. https://doi.org/10.1016/S0301-7516(98)00005-2.
[95] W. Zuo, F. Shi, Ore impact breakage characterisation using mixed particles in wide
size range, Miner. Eng. 86 (2016) 96–103.
[96] B. Ndlovu, M. Becker, E. Forbes, D. Deglon, J.P. Franzidis, The influence of

phyllosilicate mineralogy on the rheology of mineral slurries, Miner. Eng. 24 (2011)
[97] C.I. Walker, M. Hambe, Influence of particle shape on slurry wear of white iron,
Wear. 332–333 (2015) 1021–1027. https://doi.org/10.1016/j.wear.2014.12.029.
[98] L. Little, M. Becker, J. Wiese, A.N. Mainza, Auto-SEM particle shape
41
characterisation: Investigating fine grinding of UG2 ore, Miner. Eng. 82 (2015) 92–
[99] K. Kashiwaya, T. Noumachi, N. Hiroyoshi, M. Ito, M. Tsunekawa, Effect of particle

shape on hydrocyclone classification, Powder Technol. 226 (2012) 147–156.
[100] P.T.L. Koh, F.P. Hao, L.K. Smith, T.T. Chau, W.J. Bruckard, The effect of particle
shape and hydrophobicity in flotation, Int. J. Miner. Process. 93 (2009) 128–134.
https://doi.org/10.1016/j.minpro.2009.07.007.
[101] T.G. Vizcarra, S.L. Harmer, E.M. Wightman, N.W. Johnson, E. V. Manlapig, The
influence of particle shape properties and associated surface chemistry on the
flotation kinetics of chalcopyrite, Miner. Eng. 24 (2011) 807–816.
[102] P. Pourghahramani, E. Forssberg, Review of applied particle shape descriptors and

produced particle shapes in grinding environments. Part I: Particle shape descriptors,
Miner. Process. Extr. Metall. Rev. 26 (2005) 145–166.
https://doi.org/10.1080/08827500590912095.
[103] C.B. Holt, The shape of particles produced by comminution. A review, Powder
Technol. 28 (1981) 59–63. https://doi.org/10.1016/0032-5910(81)87010-6.
[104] G.M.H. Meesters, Particle Strength in an Industrial Environment, in: Handb. Powder
Technol., 2007: pp. 915–939. https://doi.org/10.1016/S0167-3785(07)12024-8.
[105] T.E. Durney, T.P. Meloy, Particle shape effects due to crushing method and size,
Int. J. Miner. Process. 16 (1986) 109–123. https://doi.org/10.1016/0301-
7516(86)90078-5.
[106] P. Pourghahramani, E. Forssberg, Review of applied particle shape descriptors and

produced particle shapes in grinding environments. Part II: The influence of
comminution on the particle shape, Miner. Process. Extr. Metall. Rev. 26 (2005)
167–186. https://doi.org/10.1080/08827500590912103.
[107] C.R. Bemrose, J. Bridgwater, A review of attrition and attrition test methods, Powder
Technol. 49 (1987) 97–126. https://doi.org/10.1016/0032-5910(87)80054-2.
[108] P.B. Rajendran Nair, R. Paramasivam, Analysis of the influence of grinding aids on
42
the breakage process of calcite in media mills, Adv. Powder Technol. 10 (1999) 223–
243. https://doi.org/10.1163/156855299X00316.
[109] M. Camalan, The influence of chemical pre-treatment on particle fracture pattern

and mineral liberation, Dissertation/Thesis, Middle East Technical University, 2018.
[110] A. Somani, T.K. Nandi, S.K. Pal, A.K. Majumder, Pre-treatment of rocks prior to
comminution – A critical review of present practices, Int. J. Min. Sci. Technol. 27
(2017) 339–348. https://doi.org/10.1016/j.ijmst.2017.01.013.
[111] L.A. Panek, T.A. Fannon, Size and shape effects in point load tests of irregular rock
fragments, Rock Mech. Rock Eng. 25 (1992) 109–140.
https://doi.org/10.1007/BF01040515.
[112] D. Tromans, J.A. Meech, Fracture toughness and surface energies of covalent
minerals: Theoretical estimates, Miner. Eng. 17 (2004) 1–15.
[113] K. Steier, K. Schönert, Verformung und Bruchphänomene unter

Druckbeanspruchung von sehr kleinen Körnen aus Kalkstein, Quarz und Polystyro,
Dechema Monogr. 69 (1972) 167–192.
[114] K. Kendall, Complexities of Compression Failure., Proc R Soc London Ser A. 361
(1978) 245–263. https://doi.org/10.1098/rspa.1978.0101.
[115] D. Golchert, R. Moreno, M. Ghadiri, J. Litster, Effect of granule morphology on

breakage behaviour during compression, Powder Technol. 143–144 (2004) 84–96.
[116] R. Chandramohan, P. Holtham, M. Powell, The influence of Particle Shape in Rock

Fracture, in: XXV Int. Miner. Process. Congr., 2010: pp. 3163–3171.
[117] T. Kojovic, Prediction and Control of Aggregate Shape Resulting From Crushing,
in: Inst. Quarr. Conf. Sydney, 1994.
[118] J.R. Craig, Ore-mineral textures and the tales they tell, Can. Mineral. 39 (2001) 937–
956. https://doi.org/10.2113/gscanmin.39.4.937.
[119] J.O. Nystroem, F. Henriquez, Magmatic features of iron ores of the Kiruna type in
Chile and Sweden; ore textures and magnetite geochemistry, Econ. Geol. 90 (1995)
43
473–475. https://doi.org/10.2113/gsecongeo.90.2.473.
[120] L. Vink, Textures of the Hilton North Deposit, Queensland, Australia, and their
relationship to liberation, Dissertation/Thesis, University of Queensland, 1997.
https://doi.org/DOI 10.1111/j.1398-9995.2009.02166.x.
[121] N. Bonnici, The mineralogical and textural characteristics of copper-gold deposits

related to mineral processing attributes, Dissertation/Thesis, University of
Tasmania, 2012.
[122] M. Parian, A. Mwanga, P. Lamberg, J. Rosenkranz, Ore texture breakage

characterization and fragmentation into multiphase particles, Powder Technol. 327
(2018) 57–69. https://doi.org/10.1016/j.powtec.2017.12.043.
[123] O. Lecoq, P. Guigon, M.N. Pons, A grindability test to study the influence of
material processing on impact behaviour, Powder Technol. 105 (1999) 21–29.
https://doi.org/10.1016/S0032-5910(99)00114-X.
[124] F.C. Bond, Contol Particle Shape and Size, Chem. Eng. (1954) 195–198.
[125] H. Heywood, Powder Production by Fine Milling, Soc. Chem. Ind. Powders in
(1962) 25–26.
[126] H.E. Rose, Particle Shape, Size and Surface Area, Soc. Chem. Ind. Powders in
(1961) 130–149.
[127] J. Tsubaki, G. Jimbo, The identification of particles using diagrams and distributions
of shape indices, Powder Technol. 22 (1979) 171–178.
https://doi.org/10.1016/0032-5910(79)80023-6.
[128] S.A.F. Murrell, P.J. Digby, The Theory of Brittle Fracture Initiation under Triaxial
Stress Conditions—I, Geophys. J. R. Astron. Soc. 19 (1970) 309–334.
https://doi.org/10.1111/j.1365-246X.1970.tb06050.x.
[129] T.F. Wong, Micromechanics of faulting in westerly granite, Int. J. Rock Mech. Min.
Sci. 19 (1982) 49–64. https://doi.org/10.1016/0148-9062(82)91631-X.
[130] B. Menéndez, W. Zhu, T.F. Wong, Micromechanics of brittle faulting and cataclastic
flow in Berea sandstone, J. Struct. Geol. 18 (1996) 1–16.
https://doi.org/10.1016/0191-8141(95)00076-P.
44
[131] L. Griffiths, M.J. Heap, P. Baud, J. Schmittbuhl, Quantification of microcrack
characteristics and implications for stiffness and strength of granite, Int. J. Rock
Mech. Min. Sci. 100 (2017) 138–150. https://doi.org/10.1016/j.ijrmms.2017.10.013.
[132] A. Nur, G. Simmons, The origin of small cracks in igneous rocks, Int. J. Rock Mech.
Min. Sci. 7 (1970) 307–314. https://doi.org/10.1016/0148-9062(70)90044-6.
[133] W.A. Olsson, Grain size dependence of yield stress in marble, J. Geophys. Res. 79
(1974) 4859–4862. https://doi.org/10.1029/JB079i032p04859.
[134] Y.H. Hatzor, V. Palchik, The influence of grain size and porosity on crack initiation
stress and critical flaw length in dolomites, Int. J. Rock Mech. Min. Sci. 34 (1997)
805–816. https://doi.org/10.1016/S1365-1609(96)00066-6.
[135] J.T. Fredrich, B. Evans, T.-F. Wong, Effect of grain size on brittle and semibrittle
strength: Implications for micromechanical modelling of failure in compression, J.
Geophys. Res. 95 (1990) 10907. https://doi.org/10.1029/JB095iB07p10907.
[136] R.H.C. Wong, K.T. Chau, P. Wang, Microcracking and grain size effect in Yuen
Long marbles, Int. J. Rock Mech. Min. Sci. Geomech. 33 (1996) 479–485.
https://doi.org/10.1016/0148-9062(96)00007-1.
[137] E. Eberhardt, B. Stimpson, D. Stead, Effects of grain size on the initiation and
propagation thresholds of stress-induced brittle fractures, Rock Mech. Rock Eng. 32
(1999) 81–99. https://doi.org/10.1007/s006030050026.
[138] C. Chang, M.D. Zoback, A. Khaksar, Empirical relations between rock strength and
physical properties in sedimentary rocks, J. Pet. Sci. Eng. 51 (2006) 223–237.
https://doi.org/10.1016/j.petrol.2006.01.003.
[139] L. Vernik, M. Bruno, C. Bovberg, Empirical relations between compressive strength

and porosity of siliciclastic rocks, Int. J. Rock Mech. Min. Sci. 30 (1993) 677–680.
https://doi.org/10.1016/0148-9062(93)90004-W.
[140] P. Baud, T. fong Wong, W. Zhu, Effects of porosity and crack density on the
compressive strength of rocks, Int. J. Rock Mech. Min. Sci. 67 (2014) 202–211.
https://doi.org/10.1016/j.ijrmms.2013.08.031.
[141] L. Li, M. Aubertin, A general relationship between porosity and uniaxial strength of
engineering materials, Can. J. Civ. Eng. 658 (2003) 644–658.
45
https://doi.org/10.1139/L03-012.
[142] A. Basu, D.A. Mishra, A method for estimating crack-initiation stress of rock
materials by porosity, J. Geol. Soc. India. 84 (2014) 397–405.
https://doi.org/10.1007/s12594-014-0145-8.
[143] I.N. Sevostyanova, Y.T. Sablina, N.L. Savchenko, S.N. Kulkov, Deformation
behaviour of zirconia with a different pore structure, IOP Conf. Ser. Mater. Sci. Eng.
447 (2018). https://doi.org/10.1088/1757-899X/447/1/012022.
[144] S.K. Kawatra, S.A. Moffat, T.C. Eisele, K.A. DeLa’O, Effects of freezing conditions
on rock breakage, Miner. Metall. Process. 11 (1994) 178–184.
[145] S.M. Tasirin, D. Geldart, Experimental investigation on fluidized bed jet grinding,
Powder Technol. 105 (1999) 337–341. https://doi.org/10.1016/S0032-
5910(99)00156-4.
[146] M. Mebtoul, J.F. Large, P. Guigon, High velocity impact of particles on a target -
An experimental study, Int. J. Miner. Process. 44–45 (1996) 77–91.
https://doi.org/10.1016/0301-7516(95)00020-8.
[147] M.J. Mankosa, G.T. Adel, R.H. Yoon, Effect of media size in stirred ball mill
grinding of coal, Powder Technol. 49 (1986) 75–82. https://doi.org/10.1016/0032-
5910(86)85008-2.
[148] K. Shinohara, B. Golman, T. Uchiyama, M. Otani, Fine-grinding characteristics of

hard materials by attrition mill, Powder Technol. 103 (1999) 292–296.
https://doi.org/10.1016/S0032-5910(99)00042-X.
[149] J.M. Parry, Ultrafine grinding for improved mineral liberation in flotation
concentrates, Univ. Br. Columbia. (2006) 1–119.
https://circle.ubc.ca/handle/2429/18118.
[150] G.C. Lowrison, Crushing and Grinding: The Size Reduction of Solid Materials,
reprint, Butterworths, 1974.
[151] J. Pocock, T.J. Veasey, L.M. Tavares, R.P. King, The effect of heating and
quenching on grinding characteristics of quartzite, Powder Technol. 95 (1998) 137–
142. https://doi.org/10.1016/S0032-5910(97)03333-0.
46
[152] M. Masri, M. Sibai, J.F. Shao, M. Mainguy, Experimental investigation of the effect
of temperature on the mechanical behavior of Tournemire shale, Int. J. Rock Mech.
Min. Sci. 70 (2014) 185–191. https://doi.org/10.1016/j.ijrmms.2014.05.007.
[153] M.J. Mankosa, G.T. Adel, R.H. Yoon, Effect of operating parameters in stirred ball
mill grinding of coal, Powder Technol. 59 (1989) 255–260.
https://doi.org/10.1016/0032-5910(89)80084-1.
[154] M.M. Ahmed, Effect of comminution on particle shape and surface roughness and
their relation to flotation process, Int. J. Miner. Process. 94 (2010) 180–191.
https://doi.org/10.1016/j.minpro.2010.02.007.
[155] V. Singh, P. Dixit, R. Venugopal, K.B. Venkatesh, Ore Pretreatment Methods for
Grinding: Journey and Prospects, Miner. Process. Extr. Metall. Rev. 40 (2019) 1–
15. https://doi.org/10.1080/08827508.2018.1479697.
[156] W.T. Shea, A.K. Kronenberg, Strength and anisotropy of foliated rocks with varied
mica contents, J. Struct. Geol. 15 (1993) 1097–1121. https://doi.org/10.1016/0191-
8141(93)90158-7.
[157] V. Brotóns, R. Tomás, S. Ivorra, J.C. Alarcón, Temperature influence on the physical
and mechanical properties of a porous rock: San Julian’s calcarenite, Eng. Geol. 167
(2013) 117–127. https://doi.org/10.1016/j.enggeo.2013.10.012.
[158] I.B. Celik, M. Oner, N.M. Can, The influence of grinding technique on the liberation
of clinker minerals and cement properties, Cem. Concr. Res. 37 (2007) 1334–1340.
https://doi.org/10.1016/j.cemconres.2007.06.004.
[159] Z.S. Mirzaei, M.R. Khalesi, Development of a Simulator for Random and Non-
Random Breakage of Particles and Liberation of Grains Based on Voronoi
Tessellation, Minerals. 9 (2019) 341. https://doi.org/10.3390/min9060341.
[160] P.W. Cleary, Industrial particle flow modelling using discrete element method, Eng.
Comput. (Swansea, Wales). 26 (2009) 698–743.
https://doi.org/10.1108/02644400910975487.
[161] M. Sinnott, P.W. Cleary, R. Morrison, Analysis of stirred mill performance using
DEM simulation: Part 1- Media motion, energy consumption and collisional
environment, Miner. Eng. 19 (2006) 1537–1550.
47
[162] N.S. Weerasekara, M.S. Powell, P.W. Cleary, L.M. Tavares, M. Evertsson, R.D.
Morrison, J. Quist, R.M. Carvalho, The contribution of DEM to the science of
comminution, Powder Technol. 248 (2013) 3–4.
[163] J.M. Menacho, Some solutions for the kinetics of combined fracture and abrasion
breakage, Powder Technol. 49 (1986) 87–95. https://doi.org/10.1016/0032-
5910(86)85010-0.
[164] E.G. Kelly, D.J. Spottiswood, The breakage function; What is it really?, Miner. Eng.
3 (1990) 405–414. https://doi.org/10.1016/0892-6875(90)90034-9.
[165] J. Yue, Rheological effects on ultra-fine grinding in stirred mills,

Dissertation/Thesis, University Of British Columbia, 2003.
[166] E. Forssberg, S. Sundberg, Z. Hongxin, Influence of different grinding methods on

floatability, Int. J. Miner. Process. 22 (1988) 183–192. https://doi.org/10.1016/0301-
7516(88)90063-4.
[167] N. Solomon, M. Becker, A. Mainza, J. Petersen, J.P. Franzidis, Understanding the

influence of HPGR on PGM flotation behavior using mineralogy, Miner. Eng. 24
(2011) 1370–1377. https://doi.org/10.1016/j.mineng.2011.07.015.
[168] G.G. Stanley, Mechanisms in the Autogenous Mill and Their Mathematical
Representation., J. South African Inst. Min. Metall. 75 (1974) 77–98.
[169] B.K. Loveday, The use of fag and sag batch tests for measurement of abrasion rates
of full-size rocks, Miner. Eng. 17 (2004) 1093–1098.
[170] G.W. Delaney, P.W. Cleary, R.D. Morrison, S. Cummins, B. Loveday, Predicting
breakage and the evolution of rock size and shape distributions in Ag and SAG mills
using DEM, in: Miner. Eng., 2013: pp. 132–139.
[171] Y. Ghorbani, M. Becker, J. Petersen, S.H. Morar, A. Mainza, J.P. Franzidis, Use of
X-ray computed tomography to investigate crack distribution and mineral
dissemination in sphalerite ore particles, Miner. Eng. 24 (2011) 1249–1257.
48
[172] T.J. Napier-Munn, S. Morrell, R.D. Morrison, T. Kojovic, Mineral comminution

circuits: their operation and optimisation, Julius Kruttschnitt Mineral Research
Centre, University of Queensland, 1996.
[173] K. Schönert, A first survey of grinding with high-compression roller mills, Int. J.
Miner. Process. 22 (1988) 401–412. https://doi.org/10.1016/0301-7516(88)90075-
0.
[174] D.W. Fuerstenau, A. Shukla, P.C. Kapur, Energy consumption and product size
distributions in choke-fed, high-compression roll mills, Int. J. Miner. Process. 32
(1991) 59–79. https://doi.org/10.1016/0301-7516(91)90019-F.
[175] A.M. Kellerwessel, High Pressure Particle Bed Comminution.pdf, (1996).
[176] W.I.L. Lim, J.J. Campbell, L.A. Tondo, The effect of rolls speed and rolls surface
on high pressure grinding rolls performance, Miner. Eng. 10 (1997) 401–419.
https://doi.org/10.1016/S0892-6875(97)00017-4.
[177] A.C. Apling, A. Raissi, Evaluation of enhanced liberation by comminution under

high pressure, Miner. Process. Extr. Metall. 109 (2000) 117–120.
https://doi.org/10.1179/mpm.2000.109.3.117.
[178] C.L. Schneider, V.K. Alves, L.G. Austin, Modeling the contribution of specific
grinding pressure for the calculation of HPGR product size distribution, Miner. Eng.
22 (2009) 642–649. https://doi.org/10.1016/j.mineng.2009.03.006.
[179] C.L. Lin, J.D. Miller, C.H. Hsieh, Particle damage during HPGR breakage as
described by specific surface area distribution of cracks in the crushed products, in:
Xxvi Int. Miner. Process. Congr., 2012: pp. 3397–3410.
[180] F. Shi, A review of the applications of the JK size-dependent breakage model: Part
1: Ore and coal breakage characterisation, Int. J. Miner. Process. 155 (2016) 118–
129. https://doi.org/10.1016/j.minpro.2016.08.012.
[181] M.J. Daniel, Energy efficient mineral liberation using HPGR technology,
Dissertation/Thesis, University of Queensland, 2007.
[182] P.W. Cleary, M.D. Sinnott, Simulation of particle flows and breakage in crushers
49
using DEM: Part 1 - Compression crushers, Miner. Eng. 74 (2015) 178–197.
[183] M. Bengtsson, Quality-driven production of aggregates in crushing plants,

Dissertation/Thesis, Chalmers University of Technology, 2009.
[184] E. Hulthén, Real-time optimization of cone crushers, Dissertation/Thesis, Chalmers

University of Technology, 2010.
[185] L. Kiss, K. Schonert, Liberation of two component material by single particle

compression and impact crushing, Aufbereit. Tech. 21 (1980) 223–230.
https://doi.org/10.1002/anie.200600263.
[186] Ç. Hoşten, C. Özbay, A comparison of particle bed breakage and rod mill grinding
with regard to mineral liberation and particle shape effects, Miner. Eng. 11 (1998)
871–874. https://doi.org/10.1016/S0892-6875(98)00074-0.
[187] A. Apling, M. Bwalya, Evaluating high pressure milling for liberation enhancement
and energy saving, Miner. Eng. 10 (1997) 1013–1022.
https://doi.org/10.1016/S0892-6875(97)00080-0.
[188] Y. Han, L. Liu, Z. Yuan, Z. Wang, P. Zhang, Comparison of low-grade hematite

product characteristics in a high- pressure grinding roller and jaw crusher, Miner.
Metall. Process. 29 (2012) 6. https://doi.org/3059792.
[189] E.V. Manlapig, D.J. Drinkwater, P.D. Munro, N.W. Johnson, R.M.S. Watsford,
Optimisation of Grinding Circuits at the Lead/Zinc Concentrator, Mount Isa Mines
Limited, IFAC Proc. Vol. 18 (1985) 265–274. https://doi.org/10.1016/S1474-
6670(17)60520-2.
[190] M.A. Bérubé, J.C. Marchand, Evolution of the mineral liberation characteristics of
an iron ore undergoing grinding, Int. J. Miner. Process. 13 (1984) 223–237.
https://doi.org/10.1016/0301-7516(84)90005-X.
[191] R.A. Mariano, C.L. Evans, The effect of breakage energies on the mineral liberation
properties of ores, Miner. Eng. 126 (2018) 184–193.
[192] A. Mwanga, J. Rosenkranz, P. Lamberg, Testing of Ore Comminution Behavior in

the Geometallurgical Context—A Review, Minerals. 5 (2015) 276–297.
50
https://doi.org/10.3390/min5020276.
[193] A. V. Dyskin, On the role of stress fluctuations in brittle fracture, Int. J. Fract. 100
(1999) 29–53. https://doi.org/10.1023/A:1018664101433.
[194] O.K. Mahabadi, B.S.A. Tatone, G. Grasselli, Influence of microscale heterogeneity

and microstructure on the tensile behavior of crystalline rocks, J. Geophys. Res.
Solid Earth. 119 (2014) 5324–5341. https://doi.org/10.1002/2014JB011064.
51
Paper 2
1
Characterization of ore texture crack formation and liberation by
quantitative analyses of spatial deformation
a*
Parisa Semsari Parapari , Mehdi Parian b , Fredrik Forsberg c , Jan
d
Rosenkranz
*
Corresponding author
a
Engineering, Luleå University of Technology, SE-971 87, Luleå, Sweden. Email: parisa.semsari@ltu.se
b
Engineering, Luleå University of Technology, SE-971 87, Luleå, Sweden. Email: mehdi.parian@ltu.se
c
Fluid and Experimental Mechanics, Department of Engineering Sciences and Mathematics, Luleå
University of Technology, 971 87, Luleå, Sweden. Email: Fredrik.Forsberg@ltu.se
d
Engineering, Luleå University of Technology, SE-971 87, Luleå, Sweden. Email: jan.rosenkranz@ltu.se
ABSTRACT
In comminution, particle breakage starts with crack induction and propagation. The path of
cracks defines the breakage mode, e.g. preferential breakage or phase boundary breakage.
For investigating crack formation behavior, the description by displacement fields can be
applied. The displacement fields of the mineral phases can then be used to understand
breakage mode and liberation. Ore texture and operational conditions such as loading
mechanisms will affect the system. One of the ore texture aspects is the ore texture
heterogeneity, which is a complex quantity comprising mineral heterogeneity, geometrical
heterogeneity, weak grain boundaries, and micro-cracks. This study aims at investigating
the effects of ore texture and loading displacement rate on breakage mode and liberation.
The approach is to describe the spatial displacement fields in different ore textures. In order
to obtain these, in-situ compression loading tests with different displacement rates were
conducted, followed by X-ray computed micro-tomography (XCT) and Digital Volume
Correlation (DVC). In addition, the resulting cracks from ore breakage were analyzed and
quantified in order to analyze the breakage mode. Moreover, XCT imaging was used for
tracking the propagated cracks in the third dimension. For identifying mineral phases,
automated scanning electron microscopy (SEM) complemented by energy dispersive
2
spectroscopy was applied. The outcomes showed that both ore texture and loading
mechanism should be considered for describing crack formation and mineral liberation.
Keywords: Crack quantification; Spatial deformation, X-ray micro-tomography; Ore
texture; Breakage mode
1 Introduction
The main purpose of comminution in ore processing is to liberate valuable minerals. In
order to understand the contribution of particle breakage in comminution to mineral
liberation, the breakage mode is defined. In general, the breakage mode is expressed as
random and non-random breakage (e.g. King (2001)) reflecting the dependency to ore
texture (Figure 1). Random breakage (i.e. breakage independent of ore texture) has a slight
contribution to mineral liberation at least in coarse size fractions, whereas non-random
breakage (i.e. preferential phase breakage and phase boundary breakage) significantly
increases the liberation of the fractured phase. Comparatively, phase boundary breakage
contribution to liberation is higher than preferential phase breakage. Therefore, identifying
and understanding the breakage mode in comminution systems and its relation to ore
texture and operating conditions is the major step towards improving liberation.

Previous research findings on the effect of operational parameters and ore texture on
breakage mode and mineral liberation have been inconsistent and contradictory. In an
investigation into the loading rate on mineral liberation, Vizcarra et al. (2010) used piston-
die compression representing slow loading rate comminution machines and a hammer mill
representing fast rate comminution machines. They found that the liberation degree did not
change with the loading rate and suggested that liberation is related to the ore texture
characteristics only. In a similar study, Ozcan and Benzer (2013) found that the slow rate
loading has an effect on coarse particle liberation degree, however, as the particle size
3
became finer, the loading rate did not affect liberation. In contrast to these investigations,
Garcia et al. (2009) conducted a three-dimensional breakage mode study and showed that
liberation improves due to phase boundary breakage when the loading displacement rate is
slow. In another study, a vibratory rod mill was used for fast rate and high pressure grinding
roll (HPGR) for slow compression rates. The results showed that the liberation of galena is
more affected than for sphalerite in slow compression rates (Apling and Raissi, 2000).
Therefore, it is well recognized that breakage mode and liberation can be affected by ore
characteristics such as ore texture as well as or by operational conditions such as the loading
rate. However, the different results show that there is a lack of fundamental understanding
of the microscopic spatial deformation of ore texture under loading rate and the cracks
initiation and propagation.
In order to tackle this problem, a systematic analysis of the crack formation and propagation
with respect to phase boundary, preferential, and random breakage will lead to better
predicting the breakage mode and liberation. The prediction of breakage mode will pave
the way to optimize the comminution process in terms of mineral liberation.
1.1 Ore texture heterogeneity

Ore texture heterogeneity is an inherent component of geological material, which is at the
nano scale and micro scale. Nano-scale heterogeneities are due to the lattice defects of
single crystals/grains including (1) Interstitial impurity atom, (2) Edge dislocation, (3) Self
interstitial atom vacancy, (4) Precipitate of impurity atoms, (5) Vacancy type dislocation
loop (6) Interstitial type dislocation loop, and (7) Substitutional impurity atom. Micro-scale
heterogeneity includes four types: (1) mineral heterogeneity due to different grain
mineralogy, (2) geometrical heterogeneity associated with grain sizes, (3) heterogeneity
associated with anisotropy and weak grain boundaries and (4) heterogeneity due to
existence of natural micro-cracks (Hamdi et al., 2015). Ore texture heterogeneity is studied
due to its effect on rock strength and stress-induced fracture patterns. For instance, it was
shown that the compressive strength of coarse grain granite is roughly 10-20% less than
fine-grain granite (Martin and Stimpson, 1994). In the field of mineral processing and
process mineralogy, mineral heterogeneity and geometrical heterogeneity are two aspects,
which are mostly defined as ore texture characteristics. Weak grain boundaries (e.g. mineral
cleavage planes) can be studied as a part of mineral heterogeneity. For instance, it was
shown that in rock breakage behavior, the existence of biotite and feldspar changes the
crack formation and rock breakage (Mahabadi et al., 2014; Nadan and Engelder, 2009).
4
Micro-crack heterogeneity can be influenced by ore texture mineralogy and its formation.
This type of ore texture heterogeneity is defined in more detail in the following.
1.2 Crack formation and propagation

Particle breakage and consequently mineral liberation starts with crack initiations and
propagations. Material failure is a consequence of sufficient applied stress and work that
break the bonds that hold atoms together. The bond strength is due to the attractive forces
between atoms. In addition, where the equilibrium spacing occurs the potential energy is
minimum. A tensile force can increase the separation distance from the equilibrium value,
which breaks the atomic bond. In order to sever the bond completely, the tensile force must
exceed the cohesive force (Figure 2).
Figure 2 Force as a function of atomic separation (x0 is equilibrium spacing) (Anderson,

2012)
Cohesive stress (σc) can be described by the following equation:
𝐸𝛾
𝜎𝑐 = √ (1)
𝑥0
5
where, γ is the material surface energy per unit area (i.e. material constant (J/m2)) and E is
the Young modulus of the material. However, the fracture strengths of brittle materials are
usually three to four times or even more below this value. This was revealed in the
experimental studies conducted on brittle material. This is due to inherent flaws (micro-
defects or micro-cracks) in these materials (Anderson, 2012). Simmons and Richter (1976)
claimed that micro-cracks in rocks are the openings where one or two of their dimensions
are considerably smaller than the third. They also defined that in 2D micro-cracks, one
dimension is significantly smaller and the crack aspect ratio (i.e. the width to length ratio)
must be less than 10-2 (usually 10-3 - 10-5). Furthermore, the length of the micro-crack is
typically less than 100 µm.
Rocks as geological materials mostly have brittle breakage behavior and commonly contain
micro-cracks (Kranz, 1983; Simmons and Richter, 1976). Micro-cracks are formed when
the localized stress goes beyond the local strength (Hamdi et al., 2015). They are
categorized into two groups, natural and stress-induced micro-cracks (Nur and Simmons,
1970). Natural micro-cracks are formed under pressure and temperature during geological
events. For instance, Nadan and Engelder (2009) classified quartz natural micro-cracks in
three categories: (1) healed micro-cracks often associated with fluid inclusion planes that
formed during isobaric cooling, (2) filled micro-cracks containing foreign minerals, and (3)
open micro-cracks are produced during the isothermal decompression process and tectonic
activities. On the other side, stress-induced micro-cracks formed due to the human
mechanical operation (Nur and Simmons, 1970). Natural and stress-induced micro-cracks
are the source of localized stress as the material undergoes mechanical stresses. The critical
stress level (σf) that a solid body with an induced crack can sustain is:
2𝛾𝐸
𝜎𝑓 = √ ∗ (2)
𝜋𝑐
At constant c*, when σ (induced stress) exceeds σf, the specimen fails. Moreover, at
constant stress when c ˃ c* (c: crack length), the specimen fails by instantaneous nucleation.
By increasing stress level, meso-cracks are formed due to the activating and growing of
micro-cracks and finally, the development of fast macro-crack propagation, which causes
the sample failure (Figure 3).
6
Figure 3. Fracture process of brittle material under stress – modified after Dyskin and
Sahouryeh (1997)
Cracks can be quantified in terms of density and their geometrical characteristic

(microstructural features) (Arena et al., 2014; Devasto et al., 2012; Griffiths et al., 2017;
Mamtani et al., 2012; Roux et al., 2013). In the case of ore breakage for mineral liberation,
crack quantification will help to identify random or non-random breakage mode.
1.3 Deformation measurements

Based on its size, a micro-crack can be visible or invisible due to the resolution limit of the
measuring techniques (e.g. SEM used for image analysis). In the cases where micro-cracks
are not visible, measuring the deformation around micro-cracks can be a way to investigate
a material failure. The deformation of a plane can be studied by Digital Image Correlation
(DIC) method (Pan et al., 2009). This method is an optical experimental method, which
measures the displacement and strain field in two dimensions using digital image
processing and numerical computing (Larsson et al., 2018; Pan et al., 2009). While optical
imaging gives information of two-dimensional space, there are other methods that give the
internal (volume) structure of a specimen (Buljac et al., 2018). X-ray computed micro-
tomography (XCT) is one of the techniques for scanning the internal features of an object
in 3D. X-rays passing through the material are attenuated according to the differences in
density and atomic composition of the internal phases. A detector then measures the
attenuation of the X-ray signals. The attenuated signals are reconstructed in virtual slices
with gray levels of brightness (Kyle and Ketcham, 2015) and by stacking multiple
sequential slices. By that, a 3D image is reconstructed. In this method, the denser phases
7
will attenuate more and in the gray brightness of slices of the reconstructed images, the
denser material will occur brighter.
3D measuring techniques can be used in order to measure the internal structural

deformation of the material (Maire and Withers, 2014). One of the methods to quantify
internal deformation and strain in a 3D volume is Digital Volume Correlation (DVC) which
is a non-destructive method based on a theory developed for DIC (Bay et al., 1999). In this
technique, it is possible to follow the non-destructive evolution or deformation of a material
by time-lapse 3D imaging at different stress levels (Yu et al., 2016). One of the advantages
of this method is that the 3D deformation field can be superimposed on top of a
microstructure that is the ore texture here. This is used for relating the heterogeneities in
strain to the underlying structure (Maire and Withers, 2014). The second advantage of DVC
is that it can reveal those deformations, which are not visible by internal three-dimensional
imaging methods even under high resolutions (Lenoir et al., 2007).
The measurements of 3D displacement and strain fields in DVC are based on the two steps
of volume images of a test sample before and after loading (Wang et al., 2018). The feature
for comparing the volumetric images is the grayscale variation. For having a proper DVC
measurement, the variation of gray level should be continuous in space (Maire and Withers,
2014). The measurements within DVC are based on two approaches, namely the so-called
local approach (using sub-volume or sub-regions or integration window) and the global
approach (nodal spacing) (Dall’Ara et al., 2017). In both approaches, the spatial resolution
is defined as the length scale of the displacement and strain measurements. In the local
approach, the size of the correlation window is used while the global approach is based on
twice the element size. Figure 4 illustrates the local approach for the image correlation
procedure. In this figure, f and g are the gray levels of the initial and deformed volumes,
respectively, and u(X) is the sought displacement field. X and X* refer to the coordinates
(in voxels) of the same point in the reference and the deformed state. The displacement
vector connects the center of the sub-volume and the point of the highest correlation.
8
Figure 4. The local approach for deformation measurements.
1.4 Objectives
The objective of this study has been to investigate the effect of ore texture and loading
mechanism on the spatial deformation and micro-crack propagation within an intact ore
sample. The first step is to utilize the DVC method along with in-situ loading to measure
the spatial deformation of ore textures from XCT images. The second step is to quantify
the propagated cracks in 2D to investigate breakage mode in terms of being random or
preferential breakage. To assess this, an image processing code in MATLAB was used to
quantify cracks in order to identify random or preferential breakage from SEM images.
Additionally, XCT 3D imaging was used for tracking the propagated cracks in the third
dimension. Moreover, phase boundary breakage and breakage mechanism were studied
qualitatively based on optical microscopy images in order to identify and characterize the
propagated cracks.
2 Materials and Methods

2.1 Materials and characterization
Magnetite ore samples from Malmberget ore deposit located in Northern Sweden have been
used for this study. The samples were collected from the conveyor belts feeding to the
beneficiation plant. The collected samples were characterized at a macro- and micro-scale.
Figure 5 illustrates the characterization procedure. In the first step, the collected samples
9
were classified and separated based on similarity in terms of grain size and geological
features i.e. mineralization, alteration and veining. The separated samples were then studied
through reflected optical microscopy for quantifying the average grain size. In order to
prepare the samples for reflected optical microscopy studies, intact samples were molded
and polished. The magnetite grain size distribution was obtained by measuring a minimum
of 200 grains. In the next step, samples were discriminated into fine, medium and coarse
grain size according to their d80 (Table 1).
Figure 5. Feed ore characterization flowchart
Table 1: Range of d80s for defining fine, medium and coarse grains
Grain size d80 (µm)

Fine 250-350
Medium 500-750
Coarse 950-1050
At the final stage of characterization, two different textures were selected for XCT studies.
The selection criteria were the mineral proportions and the grain size of the minerals.
Samples contained both magnetite and gangue minerals with roughly equal proportions and
scattered distribution of minerals. Moreover, since the ore type may have hematite (Lund,
2013), only those samples without hematite were selected. As the density and attenuation
number of hematite and magnetite are roughly the same range, the two minerals would
otherwise be detected with the same gray level in XCT images even though having different
physical properties. On the other hand, a proper selection of the grain size had to be done.
Too fine grain sizes can lead to two problems. First, grains may not be distinguished at the
selected resolution. Second, the sample will have a high particle strength, which may not
deform evenly at high loading intensities (Hatzor and Palchik, 1997). A too coarse grain
10
size, on the other hand, may cause the following problems. First, in the selected sample
size for XCT studies, there will not be a sufficient number of grains for XCT studies.
Second, it has been shown that when grains get coarser, samples will break only along the
grain boundaries (Eberhardt et al., 1999). Hence, samples of two textures were selected,
named as (A) and (B). In order to identify the minerals present in the selected textures,
SEM-EDS analysis was done (Table 2).
Table 2: Selected samples for XCT
Sample Visual logging d80 Mineralogy
Magnetite, apatite, actinolite, chlorite,

(A) 700 µm
quartz, biotite
Magnetite, apatite, actinolite, chlorite,

(B) 950 µm
calcite
2.2 Modified Brazilian test
After the samples were characterized and the textures were classified and selected, samples
were prepared for a modified Brazilian test. The Brazilian test setup was used for the
following reasons:
 Due to the small surface sharing of the particle and the platen, the friction force will
be less effective. Friction will cause the apparent compressive strength to vary inside
the sample (Obert et al., 1946).
 There will be less irregularity on the cylindrical specimen surface. In this case, the
applied stress will not accumulate on a specific part (Prasher, 1987).
 It is a method with indirect measurements of tensile strength and is used widely for
fracture behavior of brittle materials (Yuan et al., 2018).
11
Another advantage of the cylindrical shape samples is the full scan field for XCT studies
of circular fields of view (Ketcham and Carlson, 2001). Accordingly, a small-scale
Brazilian test procedure and apparatus fitting the XCT chamber were designed in this study.
The modified Brazilian test along with XCT can be used for studying the structural
deformation (in combination with DVC measurements) of minerals within each texture. To
perform the Brazilian test, a cylindrically shaped sample was drilled out of the
characterized samples. The sample size was selected with the same height and width of 7
mm due to the restrictions for XCT with respect to sample size and resolution. The sample
was handled carefully to reduce the stress-induced micro-cracks during the preparation
procedure. The samples were then polished with a polishing wheel to have parallel planes
at the end faces of the specimen with a smooth edge. The moisture content of the sample
was as received.
For conducting the in-situ Brazilian test, an apparatus with a suitable sample holder was
needed. Criteria for selecting the proper sample holder material were to have a (i) light
density with (ii) low plastic behavior. For preventing the plastic deformation inside the
apparatus, it was recommended to use steel (Bieniawski and Hawkes, 1978). However, the
X-ray signal attenuation of steel is high. Considering the above criteria, two materials were
selected: Polyoxymethylene (POM) with an average density of 1.41 g/cc (Pious and
Thomas, 2016) and aluminum with a density of 2.7 g/cc. POM is usually used for the
production of engineering components due to its rigidity under loading (Pious and Thomas,
2016). Aluminum has a good strength to weight ratio (Claisse, 2016). Finally, an aluminum
apparatus was chosen due to difficulties in POM machining. Preliminary experiments were
done on samples with and without the aluminum apparatus (A-A) (Table 3). The results
showed that the amount of aluminum extension is small enough after 60 minutes to allow
neglecting the stress relaxation.
Table 3: Sample extension under quasi-static loading
Loading (N) With A-A (mm) Without A-A (mm) Differentiate (mm)
500 0.085 0.080 0.005
1000 0.125 0.111 0.014
Another criterion for the apparatus design was its proper size setting. Two apparatus were
manufactured with two different size settings. One apparatus was designed according to the
procedure recommended in Bieniawski and Hawkes (1978) where the jar radius was 1.5
times larger than the sample. The second apparatus was prepared with a radius twice the
12
sample size. Preliminary studies showed that the sample started rotating inside the second
apparatus as the loading was applied and consequently the stress could not be properly
transferred into the sample. However, for the sample tested in the first apparatus, the stress
was well distributed and the cracks were mostly following the loading direction. Therefore,
an apparatus with aluminum and a radius of 1.5 times more than the sample diameter was
used for the in-situ loading (Figure 6).
Figure 6. Brazilian test apparatus
2.3 In-situ loading along with XCT

The 3D scanning was carried out using a Zeiss Xradia 510 Versa 3D microscope system.
The scanning voltage was 140 kV, the power was 10 W, and the scanning current was 71.5
µA. A CCD (charge-coupled-device) detector was used which included an image
intensifier. For the detector, an exposure time of 1 sec with a binning number of 2 was
utilized. The spatial resolution (i.e. the voxel size in the reconstructed slice) was 20 × 20 ×
20 µm3. The field of view (i.e. the beam section size) was 20 mm × 20 mm to cover the
specimen and the Brazilian apparatus with an objective of 4X. For having comparable scan
results, the scanning parameters were consistent in each scan.
For carrying out the DVC measurements, an in-situ loading device (CT5000, Deben Ltd,
UK) was used along with the XCT and the Brazilian apparatus. The quasi-static
compression loading mechanism was applied at two different displacement rates (0.1 and
1 mm/min) at room temperature. The temperature was controlled during the process to
prevent a temperature effect on particle deformation. Three loading steps were used for all
textural types and loading displacement rates, including normal loads of 30 N, 500 N and
13
the load at the breakage point. In each step, the ore texture sample was loaded to the before-
mentioned force and scanned while the force remained constant (Figure 7).
Figure 7. Schematic drawing of applied loading while XCT scanning and the DVC
measurements superimposed to the particle
DVC measurements were performed on the obtained XCT datasets, using LaVision DVC
software (Davis, 8.4, LaVision GmbH, Göttingen, Germany). The local correlation
algorithm in LaVision is a Fast Fourier Transform cross-correlation, which correlates the
sub-volumes independently as a discrete function of gray levels. Each dataset was
correlated against the as-received condition (loading cycle 0) with two different settings
(Table 4).
Table 4: The input setting for correlation
Overlap Peak
Setting Number Loading cycle Sub-volume Binning Passes
(%) search
(0) 128 50 16 8×8×8 2
(1) 96 50 8 4×4×4 2
(1)
(2) 64 50 4 2×2×2 2
(3) 64 75 2 None 2
(0) 128 50 16 8×8×8 2
(1) 64 50 8 4×4×4 2
(2)
(2) 32 50 4 2×2×2 2
(3) 32 75 2 None 2
14
The reliability of the displacement measurements was assessed in terms of the quality of
the cross-correlation coefficient. These settings were utilized for finding the desired
resolution with the least decorrelation value. A normalized cross-correlation coefficient,
rDVC, based on gray level gaps was used in this context (Equation 3):
∑𝑋∈𝑉𝑂𝐼 𝑓(𝑋)𝑔(𝑋 ∗ )
𝑟𝐷𝑉𝐶 = (3)
√∑𝑋∈𝑉𝑂𝐼 𝑓(𝑋)2 ∑𝑋 ∗∈𝑉𝑂𝐼 𝑔(𝑋 ∗ )2
A correlation coefficient value rDVC close to 1 indicates a perfect match. The results of the
correlation analysis show that in both input settings there is a high correlation value (Figure
8). However, the results of the second setting have a higher resolution and therefore this
setting has been used. From the displacements at the center of the sub-volumes, all the
components of the Green–Lagrange strain tensor were calculated using a centered finite
differences scheme (Germaneau et al., 2008).
: sub-volume 32 voxels, : sub-volume 64 voxels

Figure 8. Cross-correlation values for the tested samples
2.4 Crack quantification

In this study, crack quantification based on 2D SEM images was done in MATLAB. For
this purpose, high-resolution 2D images were acquired by SEM coupled with energy-
dispersive X-ray spectroscopy (EDS) from a carbon-coated polished section. The SEM
analysis was carried out with a Zeiss Sigma 300 VP with Bruker Instrument EDS detector
and Mineralogic Mining software with a resolution of 20 μm per pixel and a field of view
15
of about 7 × 7 mm per image. A summary of the image preparation and analysis
methodology in MATLAB is given below:
1. Gray level SEM images were binarized for discriminating cracks.

2. Morphological processing was applied to the crack binary images. Thinning
processing was used in order to “skeletonize” the cracks and was iterated until there
was no more modification. Topological properties (number of holes and
connections) were conserved in this process (Lam et al., 1992; Pratt, 2001).
3. By combination of the cracks’ binary image and SEM backscattered electron (BSE)
image, the crack width and surrounding phases were calculated and recorded.
An example for SEM image processing and the resulting network of cracks obtained by the
steps described above is presented in Figure 9.
BSE Binarized cracks Morphed cracks
Figure 9. Steps for 2D crack quantification.
3 Results and discussion

In this study, the maximum principal strain was used since rock is a brittle material and the
maximum principal stress is an appropriate measure for assessing damage initiation
(Meisels et al., 2015; Toifl et al., 2015). Maximum principal strain indicates maximum
principal stress, which is quantity of normal stress calculated at an angle when shear stress
is considered as zero. While shear stress is the reason of ductile material failure, in brittle
material failure is due to the normal stress.
In order to see the effect of mineralogy on the characteristic of cracks, SEM images and
optical microscopy images were used. Figure 10 shows an example of the selected slices
for texture (B) - 0.1 mm/min sample; (a), (b), and (c) show subsets obtained by cuts in the
y-z plane through the specimen at several x-positions, namely 1.4 mm, 2.12 mm, and 5.67
mm. The criterion for subsets selection was based on the selected slices of 2D images for
16
optical microscopy and SEM imaging. Figure 10 was generated using Dragonfly software,
Version 4.1 for Windows (Object Research Systems (ORS) Inc., Montreal, Canada, 2018;
software available at http://www.theobjects.com/dragonfly).
(1) (2) (3)

Figure 10. Selected sections for analyzing the maximum principal strain in 3D
Cracks were characterized qualitatively and quantitatively. With the help of three-
dimensional images and selected slices for two-dimensional images, the reported
observations can be tracked in the whole specimen. In order to study cracks quantitatively,
2D images of SEM analysis were utilized. For investigating the magnetite breakage mode,
optical microscopy images were used.
3.1 Maximum principal strain variations

For measuring the strains and deformations, the assumption is micro-cracks are equally
distributed without any specific orientation. Figure 11 shows the variations of the maximum
principal strain for 500 N in the samples. According to this image, in both textural types
and loading displacement rates, at 500 N, both tensile and compressive strain are observed.
The strain distribution in brittle material before its failure is consistent with measurements
obtained in previous studies (Melenka and Carey, 2015; Wu et al., 2019). It seems possible
that this type of strain distribution before material failure is due to the possibility of crack
closure stage of rock breakage (Peng et al., 2015). As the applying force is increased up to
the breakage point of the particle, the maximum principal strain is mostly tensile as it was
expected (Mahabadi et al., 2014; Wills and Atkinson, 1993) (Figure 12).
17
Sample DR* Scale
Strain mapping
code (mm/min) bar
Slice (1) Slice (2) Slice (3)
0.1
(A)
0.1
(B)
Figure 11. Strain mapping and cross-correlation section of 3D. *DR stands for
displacement rate. Dark blue (below zero strain) shows the compressive strain and lighter
blue (0 to 0.001) does have roughly no deformation. The other colors of the scale bar
show the tensile strain.
In order to compare the effect of loading displacement rate and textural type on strain type
and its distribution, sections of the 3D images of the core were used (Figure 11). By
increasing the loading displacement rate, the magnitude of tensile strain drastically
increased in texture (A). This drastic strain difference gives a possible expectation of higher
deformation of inherent micro-cracks under the same time circumstances. In addition, it
was shown that the failure of a brittle material under compression involves a large number
of micro-cracks. However, in the tensile case, the failure occurs with only a few micro-
cracks (Kaplan, 1963). Thus, the probability of particle failure will increase in locations
where the tensile strain is accumulated. In this study, the micro-cracks, which are involved
18
in particle failure under tensile stress, are named “effective micro-cracks”. It is possible
that not all the micro-cracks and micro-defects are visible due to the resolution limit of the
device. As a result of higher tensile strain, the deformation may occur faster. This finding
is consistent with that of Koch (2017) who pointed out that this theory that if the speed of
crack propagation is slow enough, the tip of the crack can adapt accordingly with the
upcoming grains. However, if the crack propagation occurs rapidly, the adaption will not
occur, resulting in crack propagation independent of the medium ahead of the tip. It is
possible, therefore, that the incremental increase of effective micro-cracks combined with
the theory of the less micro-crack adaption and selectivity with a higher crack inducing rate
cause random breakage mode. The probability for random breakage mode is higher than
for the non-random breakage mode (i.e. phase boundary breakage and preferential
breakage). This finding is similar to a study that was set up to determine the effect of
displacement rate on phase boundary breakage and liberation to be achieved at slower
displacement rates (Garcia et al., 2009).
While in the case of the texture (A), the loading rate has an effect on breakage mode and
liberation, the texture (B) shows no significant difference in terms of deformation, which
is high in both displacement rates (Figure 11). Therefore, in texture (B), the inducing rate
of “effective micro-cracks” is high and the crack propagation would be random. A possible
explanation is that the crystalline structure of the texture (B) has more natural micro-cracks,
which have a potential of meso-crack and ultimately macro-crack formation and
propagation. These findings support that the ore texture nature and loading mechanism have
an effect on minerals breakage mode and liberation.
Sample
(A) (B)
code
DR
0.1 1 0.1 1
(mm/min)
Material
failure
Figure 12. Ore texture failure at the failure point- strain ranging from -0.16 (compression)
to +0.16 (tensile).
19
3.2 Crack quantification and characterization
3.2.1 Crack characterization
From the SEM analysis of texture (A), it was observed that some quartz grains were covered
by biotite (Figure 13). This phenomenon is due to the ionic thermal reaction when the rock
is forming according to the reaction of quartz with water and elements i.e. potassium, iron,
magnesium, aluminum (Yardley, 1977). From this reaction, new biotite grains in the matrix
can be formed or enlarge the existing grains.
More examples of the SEM and optical microscopy images of the failed texture (A) with
0.1 mm/min displacement rate are given in Figure 13. What is clearly seen from these
images is the crack diversion to the biotite grain rather than passing through the quartz
(Figure 13 (h)). This might be due to the cleavage planes of biotite, which deviate the
cracks. This type of crack formation is attributed to the splitting along cleavage planes and
correspond with kinked structures in deformed biotite which is a sign of weak planes
heterogeneity (Wilson and Bell, 1979). On the other hand, in the cases where quartz did
not have biotite around (Figure 13 (l) and (g)), cracks passed through the quartz.
In terms of breakage mechanism and pattern of breakage, there are four main categories
including shattering, cleavage, chipping, and abrasion. The shattering breakage mechanism
observed when the applied energy is a multiple of that required for the breakage. Closer
inspection of the texture (A)-0.1 mm/min case shows that crack formation patterns in
magnetite lead to result of the shattering breakage mechanism (Figure 13 (p)). Micro-
indentation tests done on magnetite and quartz show that they have roughly similar
toughness (Middlemiss and King, 1996; Whitney et al., 2007). However, Figure 13 (j)
shows that there is no shattering breakage mechanism where quartz, apatite, and actinolite
are accumulated. This observation may be explained by the fact that the magnetite phase
has comparatively lower toughness due to its natural micro-cracks. The natural micro-
cracks in magnetite might be due to the recrystallization process of magnetite under the
metamorphism process and random orientation of magnetite with anisotropic crystalline
grains which causes intergranular stresses (geometrical and micro-crack heterogeneity)
(Bergman et al., 2001; Nur and Simmons, 1970). In addition to the metamorphism process
and its pressure and temperature variation, due to the higher values of bulk moduli and
thermal expansion coefficient of quartz in the magnetite matrix, natural micro-cracks can
be formed in magnetite (micro-crack heterogeneity) (Nadan and Engelder, 2009; Nur and
20
Simmons, 1970). Therefore, it can be concluded that, based on the material type in the
compression zone, a shattering breakage mechanism may occur or not.
The observations of magnetite failure in optical microscopy images (Figure 13 - Figure 17)
shows that the cracks were propagating into magnetite grains rather than along the grain
boundaries. By comparing the grains images from optical microscopy with DVC
measurements, a stronger tensile strain can be seen (Figure 13). This may partly be
explained by the natural micro-cracks inside the magnetite grain and crystals, which can be
deformed under the stress. These micro-cracks cause preferential breakage in the magnetite
grains. Overall, these results indicate that micro-cracks, which are not visible with the
current resolution, contributed to the failure of the sample. In the case of Figure 13 (o), it
can be seen that the micro-crack has propagated between apatite and quartz grains. This
type of micro-crack is referred to as “en passant” and can be due to the interaction of the
stress fields around two approaching crack tips and locally applied stress (Kranz, 1979).
Figure 13. Strain, grain boundary, and BSE mapping illustration of texture (A)-0.1
mm/min. Yellow color: grain boundaries. Red color: Explained cases in the text. Strain
code color code as Figure 11.
In addition, similar results in terms of biotite covering quartz grains have been observed
also in texture B (Figure 15 (q), (s), and (t)) at 1 mm/min loading displacement rate.
However, in certain areas, the quartz phase boundary selectivity decreased and cracks
propagated within the quartz (Figure 15 (k), (v), and (y)). This can be due to the decreasing
selectivity of micro-crack propagation at higher displacement rates (Koch, 2017). Similar
to lower loading displacement rate or the conditions where quartz is alone without biotite
coverage, the cracks penetrated into the quartz (Figure 15 (r), (u), and (w)).
21
2D
3D
Figure 14. Crack formation and distribution in texture (A) with displacement rate of 0.1
2D
3D
Figure 15. Crack formation and distribution in texture (A) with displacement rate of 1
The observations of crack propagation in magnetite grains for texture (B) showed that the
micro-cracks were propagating in magnetite grains (Figure 16 and Figure 17). Therefore,
at both displacement rates, the magnetite breakage behavior is similar to that of texture (A).
For (B) (0.1 mm/min), micro-cracks propagated along with apatite and magnetite grain
boundaries ((1) and (2)) and more into apatite grains ((2), (3), (4), and (6)). For texture (B)
(1 mm/min), cracks propagated with a similar behavior as texture (B) (0.1 mm/min).
22
Compared to the texture (A), the loading displacement rate has less effect on the crack
formation and distribution of the texture (B). It seems possible that these results are due to
higher quantities of natural micro-defects inside the texture (B), which act more as
“effective micro-cracks” and the crack tips propagate faster. The crack formations support
evidence from previous observations of strain measurements.
2D
3D
Figure 16. Crack formation and distribution in texture (B) with displacement rate of 0.1
2D
3D
Figure 17. Crack formation and distribution in texture (B) with displacement rate of 1
23
3.2.2 Crack quantification
In total four minerals were selected according to the pixel value of the mineral including
quartz, apatite, actinolite, and magnetite. Quantification of mineral occurrence on both
sides of the crack was done along the entire crack length. The crack length inside each
mineral and between two minerals was quantified and normalized by the total length of the
cracks (Figure 18). Due to the resolution restrictions, it was not possible to distinguish rim
shaped biotite around the quartz grains. In terms of the loading rate, the remarkable change
is seen in the cracks passing inside magnetite as well as apatite. In this regard, while the
relative crack length inside apatite in both textures increases with loading rate, the
occurrence of the cracks inside magnetite decreases. It is known that magnetite and apatite
are often highly associated together in this type of ore (Parian et al., 2018); therefore, the
cracks occurring in between magnetite and apatite or solely inside magnetite contribute
most to the liberation of magnetite. Besides, the higher number of cracks in between
actinolite and magnetite at a slower loading rate contribute even further to higher magnetite
liberation. The other notable differences are relative crack length in quartz and actinolite in
texture A compared to texture B. This can be explained by considering relatively large
abundance of these phases in texture A (Table 5).
Figure 18. Phase boundary and preferential cracking quantification for each mineral along
entire crack length
In overall, the cracks dominantly occurred in the magnetite phase. It seems that the inherent
micro-cracks and the large abundance of magnetite are the main reasons for that. In terms
of breakage mode, the slower loading rate gives higher preferential breakage for magnetite
24
while this is opposite for apatite. Phase boundary breakage between minerals e.g. apatite
and magnetite or quartz and magnetite are not significantly different for the tested loading
rates. Moreover, it is observed that the slower loading rate is more favorable for the
liberation of magnetite.
Table 5: Minerals volumetric grade in each texture.
Volumetric grade (%)

Mineral Texture A Texture B
Quartz 21.24 0.94
Actinolite 8.64 2.16
Apatite 3.81 6.45
Magnetite 66.31 90.44
4 Conclusions
Ore texture type and loading displacement rate are the two parameters, which affect the ore
crack formation and consequently breakage mode and mineral liberation. In this study, a
new aspect of ore texture was introduced as ore texture heterogeneity, which is a complex
property comprising mineral heterogeneity, geometrical heterogeneity, weak grain
boundaries, and micro-cracks. The combined effect of ore texture heterogeneity and
loading displacement rate was investigated through two methods including Digital Volume
Correlation and crack quantification. Digital Volume Correlation illustrates the textural
deformations and strains through in-situ loading of specimen X-ray computed micro-
tomography scanning. Crack quantification was done through combination of the cracks’
binary image and scanning electron microscopy backscattered electron image. Moreover,
a detailed characterization of cracks was done with optical and characteristic scanning
electron microscopy images. The following conclusions can be drawn from this work:
 Digital Volume Correlation measurements showed that the deformations from
invisible micro-cracks help to predict the breakage mode in terms of being random
or non-random within the ore texture.
 The findings suggest that both ore texture and loading displacement rate may affect
the breakage mode and liberation, although the current study is based on small size
samples.
 The density of crack formation in a specific mineral indicates the natural micro-
cracks within the mineral, which cause a preferential breakage mode.
25
 Geological heterogeneity deformations within an ore texture affect minerals’
natural micro-crack and alter breakage mechanism.
 Biotite weak planes and cleavages cause a preferential breakage mode through its
boundaries with minerals.
 Geological deformation e.g. metamorphism causes geometrical and micro-crack
heterogeneity.
 Crack quantification gives a powerful tool for detecting the type of random and non-
random breakage mode in terms of being phase boundary and preferential breakage.
 The evaluation of the comminution process in terms of mineral liberation should be
based on the combinations of ore texture and process-related parameters.
Acknowledgements
The financial support of the Centre for Advanced Mining and Metallurgy (CAMM), a
strategic research environment established at Luleå University of Technology funded by
the Swedish government, is gratefully acknowledged.
References
Anderson, T.L., 2012. Fracture Mechanics: Fundamentals and Applications, Fourth Edi.
ed, Taylor & Francis Group. https://doi.org/10.1016/j.jmps.2010.02.008
Apling, A.C., Raissi, A., 2000. Evaluation of enhanced liberation by comminution under
high pressure. Miner. Process. Extr. Metall. 109, 117–120.
https://doi.org/10.1179/mpm.2000.109.3.117
Arena, A., Delle Piane, C., Sarout, J., 2014. A new computational approach to cracks
quantification from 2D image analysis: Application to micro-cracks description in
rocks. Comput. Geosci. 66, 106–120. https://doi.org/10.1016/j.cageo.2014.01.007
Bay, B.K., Smith, T.S., Fyhrie, D.P., Saad, M., 1999. Digital volume correlation: Three-
dimensional strain mapping using x-ray tomography. Exp. Mech. 39, 217–226.
https://doi.org/10.1007/BF02323555
Bergman, S., Kübler, L., Martinsson, O., 2001. Description of regional geological and
geophysical maps of northern Norrbotten county (east of the Caledonian orogen),
SGU Geolog. ed.
Bieniawski, Z.T., Hawkes, I., 1978. Suggested methods for determining tensile strength of
26
rock materials. Int. J. Rock Mech. Min. Sci. Geomech. Abstr. 15, 99–103.
https://doi.org/10.1016/0148-9062(78)90003-7
Buljac, A., Jailin, C., Mendoza, A., Neggers, J., Taillandier-Thomas, T., Bouterf, A.,
Smaniotto, B., Hild, F., Roux, S., 2018. Digital Volume Correlation: Review of
Progress and Challenges. Exp. Mech. 58, 661–708. https://doi.org/10.1007/s11340-
018-0390-7
Claisse, P.A., 2016. Alloys and nonferrous metals, in: Civil Engineering Materials.
Elsevier, pp. 361–368. https://doi.org/10.1016/B978-0-08-100275-9.00032-2
Dall’Ara, E., Peña-Fernández, M., Palanca, M., Giorgi, M., Cristofolini, L., Tozzi, G.,
2017. Precision of digital volume correlation approaches for strain analysis in bone
imaged with micro-computed tomography at different dimensional levels. Front.
Mater. 4. https://doi.org/10.3389/fmats.2017.00031
Devasto, M.A., Czeck, D.M., Bhattacharyya, P., 2012. Computers & Geosciences Using
image analysis and ArcGIS s to improve automatic grain boundary detection and
quantify geological images. Comput. Geosci. 49, 38–45.
https://doi.org/10.1016/j.cageo.2012.06.005
Dyskin, A. V., Sahouryeh, E., 1997. Mechanism of dilatancy and fracture of brittle
materials in uniaxial compression. Int. J. Fract. 86.
Eberhardt, E., Stimpson, B., Stead, D., 1999. Effects of grain size on the initiation and
propagation thresholds of stress-induced brittle fractures. Rock Mech. Rock Eng. 32,
81–99. https://doi.org/10.1007/s006030050026
Garcia, D., Lin, C.L., Miller, J.D., 2009. Quantitative analysis of grain boundary fracture
procedures. Miner. Eng. 22, 236–243. https://doi.org/10.1016/j.mineng.2008.07.005
Germaneau, A., Doumalin, P., Dupré, J.C., 2008. Comparison between X-ray micro-
computed tomography and optical scanning tomography for full 3D strain
measurement by digital volume correlation. NDT E Int. 41, 407–415.
https://doi.org/10.1016/j.ndteint.2008.04.001
Griffiths, L., Heap, M.J., Baud, P., Schmittbuhl, J., 2017. Quantification of microcrack
characteristics and implications for stiffness and strength of granite. Int. J. Rock Mech.
27
Min. Sci. 100, 138–150. https://doi.org/10.1016/j.ijrmms.2017.10.013
Hamdi, P., Stead, D., Elmo, D., 2015. Characterizing the influence of stress-induced
microcracks on the laboratory strength and fracture development in brittle rocks using
a finite-discrete element method-micro discrete fracture network FDEM-μDFN
approach. J. Rock Mech. Geotech. Eng. 7, 609–625.
https://doi.org/10.1016/j.jrmge.2015.07.005
Hatzor, Y.H., Palchik, V., 1997. The influence of grain size and porosity on crack initiation
stress and critical flaw length in dolomites. Int. J. Rock Mech. Min. Sci. 34, 805–816.
https://doi.org/10.1016/S1365-1609(96)00066-6
Kaplan, M.F., 1963. Strains and Stresses of Concrete at Initiation of Cracking and Near
Failure. J. Proc. 60, 853–880. https://doi.org/10.14359/7882
Ketcham, R.A., Carlson, W.D., 2001. Acquisition, optimization and interpretation of x-ray
computed tomographic imagery: Applications to the geosciences. Comput. Geosci.
27, 381–400. https://doi.org/10.1016/S0098-3004(00)00116-3
Koch, P.-H., 2017. Particle generation for geometallurgical process modeling. Licent.
thesis / Luleå Univ. Technol., Licentiate thesis / Luleå University of Technology.
https://doi.org/10.1007/978-3-319-06590-8
Kranz, R.L., 1983. Microcracks in rocks: A review. Tectonophysics 100, 449–480.

https://doi.org/10.1016/0040-1951(83)90198-1
Kranz, R.L., 1979. Crack-crack and crack-pore interactions in stressed granite. Int. J. Rock
Mech. Min. Sci. Geomech. Abstr. 16, 51. https://doi.org/10.1016/0148-
9062(79)90512-6
Kyle, J.R., Ketcham, R.A., 2015. Application of high resolution X-ray computed
tomography to mineral deposit origin, evaluation, and processing. Ore Geol. Rev. 65,
821–839. https://doi.org/10.1016/j.oregeorev.2014.09.034
Lam, L., Lee, S.-W., Suen, C.Y., 1992. Thinning methodologies-a comprehensive survey.
IEEE Trans. Pattern Anal. Mach. Intell. 14, 869–885.
https://doi.org/10.1109/34.161346
Larsson, S., Gustafsson, G., Häggblad, H.Å., Jonsén, P., 2018. Experimental and numerical
study of potassium chloride flow using smoothed particle hydrodynamics. Miner. Eng.
28
Lenoir, N., Bornert, M., Desrues, J., Bésuelle, P., Viggiani, G., 2007. Volumetric Digital
Image Correlation Applied to X-ray Microtomography Images from Triaxial
Compression Tests on Argillaceous Rock. Strain 43, 193–205.
https://doi.org/10.1111/j.1475-1305.2007.00348.x
Lund, C., 2013. Mineralogical, chemical and textural characterisation of the Malmberget
iron ore deposit for a geometallurgical model. Dissertation/Thesis, Luleå University
of Technology.
Mahabadi, O.K., Tatone, B.S.A., Grasselli, G., 2014. Influence of microscale heterogeneity
and microstructure on the tensile behavior of crystalline rocks. J. Geophys. Res. Solid
Earth 119, 5324–5341. https://doi.org/10.1002/2014JB011064
Maire, E., Withers, P.J., 2014. Quantitative X-ray tomography. Int. Mater. Rev. 59, 1–43.
https://doi.org/10.1179/1743280413Y.0000000023
Mamtani, M.A., Vishnu, C.S., Basu, A., 2012. Quantification of microcrack anisotropy in
quartzite --- a comparison between experimentally undeformed and deformed
samples. J. Geol. Soc. India 80, 153–166. https://doi.org/10.1007/s12594-012-0128-6
Martin, C.D., Stimpson, B., 1994. The effect of sample disturbance on laboratory properties
of Lac du Bonnet granite. Can. Geotech. J. 31, 692–702. https://doi.org/10.1139/t94-
081
Meisels, R., Toifl, M., Hartlieb, P., Kuchar, F., Antretter, T., 2015. Microwave propagation
and absorption and its thermo-mechanical consequences in heterogeneous rocks. Int.
J. Miner. Process. 135, 40–51. https://doi.org/10.1016/j.minpro.2015.01.003
Melenka, G.W., Carey, J.P., 2015. Evaluation of fiber reinforced cement using digital
image correlation. PLoS One 10, 1–18. https://doi.org/10.1371/journal.pone.0128644
Middlemiss, S., King, R.P., 1996. Microscale fracture measurements with application to
comminution. Int. J. Miner. Process. 44–45, 43–58. https://doi.org/10.1016/0301-
7516(95)00016-X
Nadan, B.J., Engelder, T., 2009. Microcracks in New England granitoids: A record of
thermoelastic relaxation during exhumation of intracontinental crust. Bull. Geol. Soc.
Am. 121, 80–99. https://doi.org/10.1130/B26202.1
29
Nur, A., Simmons, G., 1970. The origin of small cracks in igneous rocks. Int. J. Rock Mech.
Min. Sci. 7, 307–314. https://doi.org/10.1016/0148-9062(70)90044-6
Obert, L., Winder, S.L., Derwall, W.I., 1946. Standardized test for determinig the physical
properties of mine rock. U. S. Bur. Mines Rep. Investig. 3891, 67 pages :
https://doi.org/10.1016/0278-4343(88)90009-X
Ozcan, O., Benzer, H., 2013. Comparison of different breakage mechanisms in terms of
product particle size distribution and mineral liberation. Miner. Eng. 49, 103–108.
Pan, B., Qian, K., Xie, H., Asundi, A., 2009. Two-dimensional digital image correlation
for in-plane displacement and strain measurement: A review. Meas. Sci. Technol. 20.
https://doi.org/10.1088/0957-0233/20/6/062001
Parian, M., Mwanga, A., Lamberg, P., Rosenkranz, J., 2018. Ore texture breakage
characterization and fragmentation into multiphase particles. Powder Technol. 327,
57–69. https://doi.org/10.1016/j.powtec.2017.12.043
Peng, J., Rong, G., Cai, M., Zhou, C.B., 2015. A model for characterizing crack closure
effect of rocks. Eng. Geol. 189, 48–57. https://doi.org/10.1016/j.enggeo.2015.02.004
Pious, C.V., Thomas, S., 2016. Polymeric Materials—Structure, Properties, and

Applications, in: Printing on Polymers. Elsevier, pp. 21–39.
https://doi.org/10.1016/B978-0-323-37468-2.00002-6
Prasher, C., 1987. Crushing and grinding process handbook. Chichester: Wiley.
Pratt, W.K., 2001. Digital Image Processing. John Wiley & Sons, Inc., New York, USA.
https://doi.org/10.1002/0471221325
Roux, S. Le, Medjedoub, F., Dour, G., Rézaï-aria, F., 2013. Image analysis of microscopic
crack patterns applied to thermal fatigue heat-checking of high temperature tool steels
44, 347–358.
Simmons, G., Richter, D., 1976. Microcracks in rocks. Phys. Chem. Miner. rocks 105–137.
Toifl, M., Meisels, R., Hartlieb, P., Kuchar, F., Antretter, T., 2015. 3D numerical study on
microwave induced stresses in inhomogeneous hard rocks. Miner. Eng. 90, 29–42.
30
breakage mechanism on the mineral liberation properties of sulphide ores. Miner. Eng.
Wang, B., Pan, B., Lubineau, G., 2018. Morphological evolution and internal strain
mapping of pomelo peel using X-ray computed tomography and digital volume
correlation. Mater. Des. 137, 305–315. https://doi.org/10.1016/j.matdes.2017.10.038
Whitney, D.L., Broz, M., Cook, R.F., 2007. Hardness, toughness, and modulus of some
common metamorphic minerals. Am. Mineral. 92, 281–288.
https://doi.org/10.2138/am.2007.2212
Wills, B.A., Atkinson, K., 1993. Some observations on the fracture and liberation of
mineral assemblies. Miner. Eng. 6, 697–706. https://doi.org/10.1016/0892-
6875(93)90001-4
Wilson, C.J.L., Bell, I.A., 1979. Deformation of biotite and muscovite: Optical
microstructure. Tectonophysics 58, 179–200. https://doi.org/10.1016/0040-
1951(79)90328-7
Wu, Q., Chen, L., Shen, B., Dlamini, B., Li, S., Zhu, Y., 2019. Experimental Investigation
on Rockbolt Performance Under the Tension Load. Rock Mech. Rock Eng. 52, 4605–
4618. https://doi.org/10.1007/s00603-019-01845-1
Yardley, B.W.D., 1977. The nature and significance of the mechanism of sillimanite
growth in the Connemara schists, Ireland. Contrib. to Mineral. Petrol. 65, 53–58.
https://doi.org/10.1007/BF00373570
Yu, B., Bradley, R.S., Soutis, C., Withers, P.J., 2016. A comparison of different approaches
for imaging cracks in composites by X-ray microtomography. Philos. Trans. R. Soc.
A Math. Phys. Eng. Sci. 374. https://doi.org/10.1098/rsta.2016.0037
Yuan, Z.N., Chen, H., Li, J.M., Dai, B., Zhang, W. Bin, 2018. In-situ X-ray tomography
observation of structure evolution in 1,3,5-Triamino-2,4,6-Trinitrobenzene based
polymer bonded explosive (TATB-PBX) under thermo-mechanical loading. Materials
(Basel). 11, 1–14. https://doi.org/10.3390/ma11050732
31

Parapari2021 - Effects of Loading Mechanisms and Texture On Ore Breakage A Multidimensional Study

Uploaded by

Copyright:

Available Formats

Parapari2021 - Effects of Loading Mechanisms and Texture On Ore Breakage A Multidimensional Study

Uploaded by

Document Information

Original Title

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

Parapari2021 - Effects of Loading Mechanisms and Texture On Ore Breakage A Multidimensional Study

Uploaded by

Copyright:

Available Formats

L ICE N T IAT E T H E S I S

ISSN 1402-1757 Effects of loading mechanisms

Parisa Semsari Parapari

Parisa Semsari Parapari

Luleå University of Technology

Parisa Semsari Parapari

Licentiate thesis in Mineral Processing

Above all, I thank and praise God.

The list of papers not included in this thesis:

Chapter 1 – Introduction ...................................................................................................... 3

1.1 Background ........................................................................................................... 3

1.2 Objectives .............................................................................................................. 4

Chapter 2 - Fundamentals of ore breakage .......................................................................... 7

2.1 Loading mechanism .............................................................................................. 7

2.2 Ore texture breakage behavior ............................................................................ 10

2.3 Breakage mechanism........................................................................................... 13

2.4 Breakage mode .................................................................................................... 16

Chapter 3 - Analysis of crack formation ............................................................................ 21

3.1 Sample characterization ...................................................................................... 21

3.2 Deformation measurements................................................................................. 23

3.3 Modified Brazilian test ........................................................................................ 25

3.4 In-situ loading along with XCT .......................................................................... 26

3.5 Crack quantification ............................................................................................ 28

Chapter 4 - Results and discussion .................................................................................... 29

4.1 Maximum principal strain variations .................................................................. 29

4.2 Crack quantification and characterization ........................................................... 32

4.2.1 Crack characterization ................................................................................. 32

4.2.2 Crack quantification ..................................................................................... 36

Chapter 5 - Summary and future work .............................................................................. 39

Part II – Papers ................................................................................................................... 55

Accordingly, the objectives of this licentiate study have been:

1. Review the literatures of ore texture breakage fundamentals.

1. Screening the literature on breakage fundamentals, which ranges across the

2.1 Loading mechanism

Overall, variations of loading mechanisms occur in each comminution machine; however,

2.2 Ore texture breakage behavior

Figure 3. Force as a function of atomic separation (x0 is equilibrium spacing) (Anderson,

Figure 5. Ore texture heterogeneity

Figure 6. Ore texture and its effect on heterogeneity

2.3 Breakage mechanism

Table 1. Breakage mechanisms in a comminution environment

Feed particle Breakage pattern Product particles Breakage mechanism

Particles with topographical properties or angular surfaces experience chipping breakage

2.4 Breakage mode

Table 2. Summary of loading mechanism effects on mineral liberation

Strike (Apling and Bwalya, 1997) (Han et al., 2012)

Figure 8. The schematic representation of workflow

3.1 Sample characterization

Grain size d80 (µm)

Sample Visual logging d80 Mineralogy

700 Magnetite, apatite, actinolite,

950 Magnetite, apatite, actinolite,

3.2 Deformation measurements

3D measuring techniques can be used in order to measure the internal structural

Figure 10. The local approach for deformation measurements.

Table 5. Sample extension under quasi-static loading

Figure 11. Brazilian test apparatus

3.4 In-situ loading along with XCT

Setting Loading Overlap Peak