See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/379477826

Some Aspects on Grindability of Feldspar Ore Using Ball Mill

Article in Particulate Science And Technology · April 2024

DOI: 10.1080/02726351.2023.2293168

CITATIONS READS

0 82

4 authors, including:

Khaled Yassin El-Sayed R. E. Hassan

Central Metallurgical Research and Development Institute Central Metallurgical Research and Development Institute
12 PUBLICATIONS 32 CITATIONS 26 PUBLICATIONS 287 CITATIONS

SEE PROFILE SEE PROFILE

Abdalla Elbendari
Central Metallurgical Research and Development Institute
15 PUBLICATIONS 60 CITATIONS

SEE PROFILE

All content following this page was uploaded by El-Sayed R. E. Hassan on 01 April 2024.

The user has requested enhancement of the downloaded file.

 Particulate Science and Technology
An International Journal

ISSN: (Print) (Online) Journal homepage: www.tandfonline.com/journals/upst20

Some aspects on grindability of feldspar ore using

ball mill

Khaled E. Yassin, El-Sayed R. E. Hassan, N. A. Abdel Khalek & A. M. Elbendari

To cite this article: Khaled E. Yassin, El-Sayed R. E. Hassan, N. A. Abdel Khalek & A. M. Elbendari
(01 Apr 2024): Some aspects on grindability of feldspar ore using ball mill, Particulate Science
and Technology, DOI: 10.1080/02726351.2023.2293168

To link to this article: https://doi.org/10.1080/02726351.2023.2293168

Published online: 01 Apr 2024.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=upst20
PARTICULATE SCIENCE AND TECHNOLOGY
https://doi.org/10.1080/02726351.2023.2293168

Some aspects on grindability of feldspar ore using ball mill

Khaled E. Yassin, El-Sayed R. E. Hassan, N. A. Abdel Khalek, and A. M. Elbendari
Minerals Beneficiation and Agglomeration Department, Central Metallurgical R& D Institute (CMRDI), Helwan, Egypt Cairo

ABSTRACT KEYWORDS
Egyptian feldspar ores are usually of low grade and need beneficiation to improve their grade for Feldspar; ball milling;
industrial applications. Therefore, feldspar should be ground to a reasonable liberation degree grinding rate; slurry filling
prior to beneficiation processes. Mineralogical studies showed that a good degree of liberation is ratio; optimal ball size
below 0.25 mm. This work aims to optimize the grinding parameters for an Egyptian feldspar ore
that ensures maximum production of size −250 þ 45 mm and minimum % − 45 mm. Grinding was
performed using a laboratory ball mill. The results displayed that the grinding performance
was increased with decreasing the solid content with U between 0.5 and 1. A high milling rate
was achieved at ball filling 20–30%. The ideal ball-filling ratio was determined to be 25%, with
less production of fines. Hence, it was decided to crush the feed to less than 2 mm before grind­
ing with 36.6 mm balls to reduce fines.

1. Introduction for establishing the best condition in wet milling. The solids
content in slurry is an essential factor in the wet grinding
Feldspar minerals (KAlSi3O8, NaAlSi3O8, CaAl2Si2O8) are the
process due to its influence on the motion of the balls inside
most abundant aluminosilicate rock-forming minerals, com­
the mill and the ground product size. The presence of fine
prising 60% of the earth’s crust. They are widely applied in
particles during milling increases pulp viscosity, which lowers
producing glass ceramics and as fillers and extenders (Abdel-
grinding efficiency (El-Mofty et al. 2020). The balls must be
Khalek, Yehia, and Ibrahim 1994). The flotation technique is coated with an ore layer; pulp dilution raises metal-to-metal
considered as a primary separation method for beneficiating contact, resulting in an increase in steel consumption and a
feldspar (Abdel-Khalek, Yehia, and Ibrahim 1994). An urgent decrease in efficiency (Makokha et al. 2012).
demand should not only optimize feldspar flotation but also In general, increasing the powder amount led to the fill­
other associated operations, including grinding. ing of collision voids between balls, thus improving grinding
Feldspar pretreatment includes crushing, grinding, de- action. Filling up the adequate space, ball collision space,
sliming, and classifications. A jaw crusher is applied for with powder yielded the highest values of breakage rates.
coarse crushing, while roll and hammer crushers are applied Additional powder increases mill holdup but has no effect
for fine crushing (Zhang et al. 2018). Grinding in mineral on breakage as the colliding area has become saturated, and
processing technology is an essential process in producing additional powder simply enters a powder reservoir (Deniz
the required particle size for the liberation of valuable miner­ 2011a).
als from gangue for further concentration process (Lorig and A plateau of steady breakage rate is obtained. Eventually,
Clarence 2017; Hassan et al. 2020). Grinding is the largest overfilling minimizes collisions by powder cushioning, caus­
energy-consuming operation in mineral processing. About ing the ball–powder bed to expand, resulting in weak ball–
50% of the energy in a mining operation is consumed in a ball powder-nipping collisions and reducing the breakage
comminution operation (Abouzeid et al. 2022). Therefore, rates (Umucu and Deniz 2014; Deniz 2011a, 2011b).
the intensification of the grinding processes is a huge duty. Shoji et al. (1982) examined the effects of powder filling
Determination of the mineral liberation size is the first and on the milling of quartz in a laboratory ball mill for dry and
most crucial step to overcome this challenge, which elimi­ wet grinding. The optimal filling ratio for maximizing the
nates over-grinding and, as a result, lowers the cost of grind­ rates of breakage at all ball loads was determined at U from
ing (Elbendari et al. 2020). Grinding performance belongs to 0.6 to 1.1. Ball filling is one of the most important factors
the grinding media movement in a milling drum. It is influencing mill efficiency, as the tumbling action and the
affected by operating variables including grinding time, rota­ rate of breakage is determined by how much the balls fill
tional speed, mill volume, and sample charge (Cho et al. the mill volume. It refers to the volume occupied by the
2013; Deniz 2012; Santos, Machado, and Nascimento 2022). bulk ball charge at rest in a mill, which includes about 40%
Solid concentration and slurry viscosity are essential variables of the void space between balls (Deniz 2016; Milan,

CONTACT El-Sayed R. E. Hassan [email protected] Minerals Beneficiation and Agglomeration Department, Central Metallurgical R& D Institute
(CMRDI), P.O. Box 87 Helwan, 11722 Cairo, Egypt.
� 2024 Taylor & Francis Group, LLC
2 K. E. YASSIN ET AL.

Magdalinovi�c, and Trumi�c 2007; Austin, Klimpel, and intensities (I/Io) were obtained. The diffraction charts and
Luckie 1984). relative intensities were obtained.
The ball charge increases the energy input to a mill, and Thin and polished sections were conducted to specify the
its maximum value is achieved at a charge volume of about different types of minerals. The specimen is ground as the
50% (Cayirli 2018). The grinding conditions optimization surfaces are parallel to each other. One side of the ground
has been studied from the experimental effort and empirical surfaces is then carefully polished on a glass plate, and glued
technique (Petrakis and Komnitsas 2022). on a slide glass with Canada balsam as the polished surface
The previous studies on Egyptian feldspar (Wadi Zirib is faced to the glass plate. The glued specimen is polished
locality) proved that the suitable size for subsequent separ­ until it becomes 0.03 mm thick, and then, it is covered with
ation processes is less than 250 microns (Ahmed et al. 2016; a slide glass.
Yassin et al. 2023). Therefore, this paper is focused on opti­
mizing the grinding conditions of feldspar ore that ensure 2.2. Sample preparation
maximum production of feed size (−250 mm) with minimum
slimes (fine particles less than 45 microns). The feldspar sample was, first, primary crushed in a “5 � 6
Denver” Jaw crusher followed by secondary crushing using a
“Wedag” roller to produce different product of sizes less
2. Materials and methods than 8 mm, 4 mm, and 2 mm. The crushed products were
wet ground using a laboratory ball mill with dimensions of
A low-grade Egyptian feldspar sample from Wadi Zirib
DxL ¼ 150 � 200 mm and screened in 250 and 45 um sieves
locality, in the Eastern desert, was used for this study.
to determine the percentage of þ250 um, −250 þ 45 um
and −45 um size fractions. The experiments were performed
at a constant speed of n ¼ 76 rpm which is 70% of its critical
2.1. Sample characterization
speed. The mill charge consisted of forged steel balls with
The feldspar ore sample was crushed and subjected to dry density qb ¼ 7.8 g/cm3. The parameters that describe the
size analysis, using ASTM standard sieves of 2.0, 1.2, 0.84, mass of grinding media J and material fc, respectively, are
0.6, 0.4, 0.297, 0.25, 0.18, 0.15, 0.106, 0.075, 0.063, and calculated from Equations (1) and (2). J is the ball filling
0.045 mm using a laboratory RetschTM Vibratory Sieve ratio (which is the mill zone filled with the media bed) while
Shaker. fc is the filling ratio of the material fed in the mill (Deniz
Chemical analysis for feldspar sample was carried out 2012; Deniz 2011b; Umucu and Deniz 2014; Cayirli 2018).
using X-ray fluorescence (XRF). Quantitative determination � �
Volume of solid balls 1
of the oxides content was carried out using X-ray fluores­ J¼ : (1)
mill volume 1-;
cence analysis (XRF), Phillips, PW 2404, Netherlands.
Feldspar sample was fired at 1000 � C for 2 hours for calcu­ � �
Volume of solid material 1
lating its ignition loss (LOI). A portion of the ignited sample fc ¼ : (2)
mill volume 1-;
was mixed and pressed with 2 gm of wax as a binder in alu­
minum cup, and then exposed to X-ray as a disk. � �
fc
Qualitative X-ray diffraction (XRD) analysis was con­ U¼ (3)
ducted for identifying the minerals phase structure. It was 0:4:J
performed using PAN analytical X‫׳‬Pert PRO with Secondary Where U is the bed porosity of balls and material
Monochromator, Cu-radiation (k ¼ 1.542 Å) at 45 K.V., (assumed to be 40%). The space fraction between balls at
35 M.A. and scanning speed 0.04 o/sec. were used. The dif­ rest, U, that is filled with material bed (interstitial filling)
fraction peaks between 2h ¼ 2� and 70� and relative can be calculated from Equation (3). The grinding rate (the

Table 1. Wet ball mill grinding test conditions.

Item Description First series Second series Third series
Mill Inner diameter (D), mm 150
Length, mm 200
Volume, cm3 3532
Mill speed Critical, Nc, [rpm] 109
Operational speed, Uc (0.7Nc), [rpm] 76
Media (Balls) Material Forged steel
Specific gravity 7.8
Mass of balls, g 3320 4150 4980
Assumed porosity 40 %
Ball size 25.4 19.1-25.4-36.6 25.4
Ball filling volume fraction J, [%] 20 25 30
Material Feldspar ore
Specific gravity, g/cm3 2.6
Initial feed size 2mm 2mm-4mm-8mm 2mm
Interstitial filling U, [%] 1 0.5-0.75-1- 1.5-2 1
% S in slurry by weight 65% 50% − 65%- 80 % 65%
 PARTICULATE SCIENCE AND TECHNOLOGY 3

decrease rate of coarse particle) of various tests has been K2O, Fe2O3) are more or less evenly distributed among the
observed through the milling rate constant k in the law different size fractions.
equation of grinding kinetics as a first order grinding R¼R0
e-kt, where R is the coarse fraction remained un-ground of
3.2. Factors affecting on wet grinding of feldspar using
the sample after grinding time (t), and R0 is the sample
ball mill
mass for grinding (Milan, Magdalinovi�c, and Trumi�c 2007;
Austin, Klimpel, and Luckie 1984). The wet ball mill grind­ 3.2.1. Effect of % solid content in slurry
ing test presented in Table 1. The effect of increasing slurry solid content on the ground
products and grinding rate has been studied at constant
grinding time (15 min), U ¼ 0.5 and constant ball mill speed
3. Results and discussion (76 rpm). Figure 3 shows the results of grinding at different
3.1. Mineralogy and chemical composition of the solid contents at constant (15 min) grinding time. Grinding
sample rate decreased sharply at 80% dense slurry, and the percent­
age of fraction less than 45 um is high in comparison to
XRF analysis of the sample shows that it contains: 75.85% that conducted at 65% and 50% solids. These results align
SiO2, 12.48% Al2O3, 1.4% Fe2O3, 3.22% Na2O, 5.36% K2O, with the research conducted by Mulenga and Moys (2014),
and 0.93% CaO. Table 2. who suggested that the net production of fines increased
Figure 1 shows XRD of the original feldspar sample. It steadily, as the slurry pool was increased. Increasing slurry
contains albite and microcline as the main feldspar minerals density will result in a decrease in breakage rate, and caus­
while quartz is the main gangue in minor amount of hema­ ing problems in slurry flow due to increased viscosity
tite impurities. These results were confirmed by mineral­ (Jankovic 2003). When the largest degree in size reduction,
ogical studies where illustrated that the rock is composed with producing the lowest fines percentage occurred, there
mainly of alkali feldspar (microcline, quartz and plagioclase was no significant difference between 50% and 65% solids
(albite) together with rare amounts of mica (biotite and content. For the next experiments, 65% solids content was
muscovite). Secondary minerals are represented by clay min­ chosen as the optimum value from an economic standpoint,
erals, sericite, carbonates, chlorite and iron oxides as shown whereas too dilute slurry may cause metal-to-metal contact,
in Figure 2. resulting in high steel ball consumption (Tangsathitkulchai
In the meantime, Table 3 depicts the chemical analysis 2003).
(using XRF) of different size fractions of the secondary
crushed sample. The different oxides (such as Al2O3, Na2O, 3.2.2. Influence of slurry filling on the grinding
performance
Table 2. XRF of the feldspar sample.
The concept of slurry filling in the mills was first introduced
Oxide SiO2 Al2O3 Na2O K2O Fe2O3 CaO MgO Cl
by Austin, Klimpel, and Luckie (1984). The amount of pow­
% 75.87 12.48 3.22 5.36 1.4 0.93 0.074 0.04
der in a mill in relation to the percentage of media is

Figure 1. XRD of the feldspar sample.

4 K. E. YASSIN ET AL.

Figure 2. Photomicrographs showing all minerals in the feldspar sample through the thin section (25X, C.N.). Photomicrographs showing all minerals in the Wadi
zeriba feldspar sample through the thin section; (a) Perthite and granophyric texture in a head sample (thin section 25X, C.N.), (b) Plagioclase, Quartz and
Microcline (thin section 25X, C.N.), (c) Plagioclase & Quartz as major, minerals (thin section 25X, C.N.), (d) Micro veinlet of Quartz, perthite, and microcline (thin sec­
tion 25X, C.N.), (e) Biotite and Muscovite (thin section 25X, C.N.).

typically expressed as ‘‘the fractional powder filling is the gaps, so the amount of solid in the active breakage zones
fraction of the mill bed gaps (at rest in the case of a packed will depend on the slurry concentration and U (interstitial
bed) filled with powder, determined by the powder’s bulk filling) (Austin, Klimpel, and Luckie 1984). The specific
density and the assumption that a bed of balls has a porosity throughput has been calculated at different grinding time.
of 40%” The above description is simple to estimate in dry Table 4 gives the values of specific throughput of mill (Qs)
grinding, but in wet grinding, it is the slurry that fills the and Constants of milling rate (k) for grinding of feldspar at
 PARTICULATE SCIENCE AND TECHNOLOGY 5

Table 3. Chemical analysis of different size fractions. different slurry filling and different time. The amount of
Wt. SiO2 Al2O3 Na2O K2O Fe2O3 CaO Cl slurry is increased till the slurry volume becomes 2 times of
Size, mm % % % % % % % %
the balls bed voidage (the balls volume disregarding their
−4.0 þ 2.0 1.79 77.23 11.73 2.95 6.26 0.85 0.87 0.05
−2.0 þ 1.63 3.57 75.17 13.09 3.15 6.44 1.24 1.07 0.05
voids). The effect of increasing the U value from 0.5 to 2 at
−1.63 þ 1.16 10.57 77.08 12.26 3.02 5.77 0.79 0.82 0.048 a constant grinding time of 15 minutes is presented in
−1.16 þ 1.0 12.26 77.73 12.41 2.68 5.57 0.71 0.81 0.04 Figure 4. It has been demonstrated that decreasing the U
−1.0 þ 0.84 12.75 77.73 12.41 2.69 5.58 1.71 0.81 0.03
−0.84 þ 0.58 18.16 77.23 12.47 3.82 5.03 1.45 0.69 0.04 value significantly increases breakage rate. The tests associ­
−0.58 þ 0.4 13.74 76.96 12.54 3.52 5.19 1.83 0.82 0.042 ated to Figure 4 are presented in Figure 5 for different val­
−0.4 þ 0.295 9.11 76.81 12.68 3.40 5.19 1.45 0.93 0.043 ues of the grinding time. The values of grinding time were
−0.295 þ 0.25 3.10 76.84 12.47 3.00 5.42 1.11 0.97 0.051
−0.25 þ 0.125 8.76 76.59 12.46 3.42 4.85 1.34 1.11 0.046 chosen to be 15 min, 18 min, 24 min, 36 min, and 48 min.
−0.125 þ 0.105 1.74 75.85 12.39 2.83 5.19 1.96 1.36 0.045 while using a feed quantity of each 300 gm, 450 gm, 600 gm,
−0.105 þ 0.075 0.80 75.03 12.54 2.80 5.12 2.41 1.75 0.057 900 gm, and 1200 gm, respectively. According to Figure 5
−0.075 þ 0.053 2.18 73.39 13.20 3.47 5.06 2.33 1.69 0.06
−0.053 þ 0.045 0.75 68.26 12.29 2.64 5.56 0.153 4.66 0.093 and Table 4 the best grinding rate occurs at the slurry filling
−0.045 0.72 69.59 12.87 2.76 5.64 0.85 4.08 0.08 (U) between 0.5 and 1, corresponding to feed quantity
Assay% 100.00 76.87 12.48 3.22 5.36 1.35 0.93 0.04
origin 75.87 12.48 3.22 5.36 1.40 0.93 0.04
between 300 to 600 gm. Also, if the grinding time is similar
(at 15 min) and increasing of U, the size distribution of the

Figure 3. (a, b). Effect of % solids content in slurry at 15 min grinding time.
6 K. E. YASSIN ET AL.

Table 4. Grinding rate constant (k) and specific throughput at different slurry filling and different grinding time.
Size fraction, mm
−250 þ 45 −45
U Feed quantity, g Grinding time, min Mass Fr., % Qs, kg/hr. k k(t) Mass Fr., % Qs kg/hr. k k(t)
0.5 300 7.5 32 0.768 0.0826 0.7335 14.2 0.34 0.0204 0.153
15 60 0.72 0.137 2.055 27.2 0.326 0.0211 0.316
0.75 450 11.25 31.4 0.753 0.0575 0.647 16.28 0.39 0.0157 0.177
22.5 58.13 0.698 0.1052 2.367 32.5 0.39 0.0174 0.393
1 600 15 32.5 0.78 0.0409 0.6135 13.4 0.32 0.0096 0.1438
30 60.83 0.73 0.0675 2.025 26 0.312 0.0100 0.3011
1.5 900 22.5 28.37 0.68 0.0238 0.5355 13.1 0.313 0.0062 0.1404
30 37.12 0.668 0.0263 0.789 17.5 0.315 0.0064 0.1923
45 54.37 0.652 0.0362 1.629 26 0.3117 0.0066 0.3011
2 1200 30 24.5 0.58 0.015 0.45 12 0.284 0.0042 0.1278
60 47 0.56 0.0205 1.23 23.76 0.283 0.0045 0.2713

Figure 4. Effect of increasing feed quantity at 15 min grinding time.

Figure 5. Effect of increasing feed quantity at different grinding time.

products becomes coarser and coarser. On the contrary, if there is enough material between the grinding balls to avoid
the grinding time values are not similar, initially with the them from impacting each other excessively. In that case,
increase of U, the grinding rate is improved in which its the mechanism of grinding might involve abrasion between
maximum occurs at U ¼ 1, after that, the grinding rate is the particles as well as the impact from the balls. However,
reduced again (U 1.5 and 2). These findings can be when the slurry filling is reduced to a low U value, in the
explained by the fact that when the slurry filling is high, range of 0.5 to 1, the particles are ground by direct contact
 PARTICULATE SCIENCE AND TECHNOLOGY 7

Figure 6. The effect of ball filling at 65% solid content in slurry, 10% powder filling, and 30 min grinding time.

Figure 7. Particle size distribution of milling at 65% solid content in slurry, 10% powder filling, different ball filling and different grinding time.

between the balls, resulting in a strong impact (Austin, € uzo�glu and Uçurum (2016) investi­
grindability. Also, Oks€
Klimpel, and Luckie 1984; Gao and Forssberg 1989). gated that a value of fc ¼ 0.05 filling ratio is optimal for
Furthermore, at U ranging from 0.5 to 1 it seems that the gypsum. According to Serkan (2018) study, which investi­
influence of slurry filling on product size distribution can be gated the impact of powder filling at values ranging from fc
compensated by grinding time. Product size distributions ¼ 0.1–0.2, the ideal fc value was 0.125, which was consistent
obtained with a low slurry filling (U ¼ 0.5) and a short with the general rule (U ¼ 0.6 to 1.1).
grinding time (15 min) are very similar to those obtained Based on the data presented in Table 4, U ¼ 1 was
with a medium and high slurry filling (U ¼ 0.75, U ¼ 1) and chosen as the optimum slurry filling value, resulting in the
the corresponding grinding times (22.5 and 30 min), highest throughput of the desired class (−250 þ 45mm) and
respectively. the lowest fines observed.
Previous work of Deniz and Onur (2002) revealed that,
an interstitial filling of U ¼ 0.4 is indicative of an optimal
powder-ball loading ratio for pumice. Furthermore, Deniz’s 3.2.3. Effect of ball filling (J, %)
2011 research revealed that a high value of U ¼ 1.5–2.0 is an Ball mill filling has a great influence on power draw and
effective ratio for gypsum. The differences in these findings energy consumption. When the mass inside the mill
are most likely due to differences in ore types, densities, and increases, more energy is required. More power is needed
8 K. E. YASSIN ET AL.

Table 5. Grinding rate constant (k) at different ball filling, U ¼ 1.

Size fraction, mm
−250 þ 45 −45
Grinding time, min Ball filling (%) Mass fraction, % k Mass fraction, % k
15 20 22.92 0.02774 11.12 0.00785
25 32.58 0.04107 13.42 0.0096
30 39.57 0.05715 18 0.0132
30 20 40.8 0.03387 23 0.0087
25 60.83 0.0675 26 0.0100
30 62.76 0.1199 34.5 0.0141

when there is an increase in mass within the mill (Austin, (whereby small diameter balls are used for feed material of
Klimpel, and Luckie 1984; Cayirli 2018). small sizes while larger diameter balls are used for larger
Figures 6, 7 and Table 5 show the effect of ball filling at sized feed material) (Magdalinovic, Trumic, and Andric
65% solid concentration and 10% powder filling (600 gm) 2012). The experiment was carried out to ascertain the effect
during grinding for 30 minutes and the dependence of mill­ of ball diameter sizes on milling (at constant 25% ball filling
ing rate constant (k) upon the percentage ball filling. and 15 min grinding time). Balls of sizes 36 mm, 25 mm,
The milling rate was shown to be significantly higher and 19 mm are used. A mill run with an equal mixture of
when the ball filling was increased from 20% to 30%, which the three ball diameter sizes was also conducted.
provides the highest rate of grinding, in terms of the milling The results depicted in Figure 8. It has been shown the
rate constant (k). When the ball filling was increased from lowest grinding rate was observed during grinding with
20% to 30%, the percentage weight of fraction less than smaller ball diameters (19 mm). The 19 mm ball size was
250 mm increased from 63.8% to 97.26%, and fines −45 mm unable to break up the large particle, making it the least
increased from 23.5% to 34.5%. effective where a minimum particle breakage rate was
These findings can be explained by the fact that collision observed. Whether using 25 mm or an equal mixture of
spaces between the balls were filled, resulting in higher three ball sizes, the grinding rate is approximately the same,
breakage rates at a filling ratio of 0.30. A decrease in milling and it results in a roughly uniform product size distribution.
performance was seen as a result of underfilling (J ¼ 0.20), The highest grinding rate was achieved when grinding with
which caused the ball-powder bed to expand, resulting in larger ball diameters (36.6 mm). The proportion of coarse
poor ball-ball powder nipping collisions. A high throughput fractions þ250 microns in the product is lower when grind­
of the desired flotation class (−250 þ 45mm) is achieved with ing with a 36.6 mm ball compared to 25.4 and 19.1 mm,
a 25% filling, while fewer fines are produced (26%) com­ whereas the percentage of fractions less than 250 microns is
pared to 34.5% in the case of a 30% filling. higher when grinding with 36.6 mm ball compared to 25.4
Previous work of Gupta and Yan (2006) suggested that a and 19.1 mm. It means that larger balls are needed for quick
ball charge of 35–45% of the mill volume would be the ideal grinding of larger coarse fractions, whereas a greater propor­
conditions for grinding. Also, Deniz (2012) determined that tion of small balls (25.4 or 19.1 mm) are needed for finer
an effective filling ratio of 0.35 was appropriate for barite product sizes.
grinding based on his research on the effects of media filling These findings are consistent with previous research by
on barite breakage parameters. Furthermore, Mulenga and Khumalo et al. (2006), who claimed that, in general, larger
Moys (2014) revealed that a high ball filling ensures more grinding media would break larger particles faster, but
optimal milling. Then, Serkan Cayirli’s (2018) examined the smaller balls would produce a finer product. Smaller grind­
impact of ball filling ratios ranging from 0.2 to 0.45, and it ing media, on the other hand, are thought to support abra­
was determined that a ratio of Jb ¼ 0.30 was the optimal sion and attrition, which are energy-ineffective breakage
choice. mechanisms (Katubilwa et al. 2011; Danha, Hlabangana, and
One could draw the conclusion that among the most sig­ Muzenda 2018; Chimwani, Mulenga, and Hildebrandt 2015).
nificant aspects directly affecting the product fineness is ball Also, Oksuzoglu and Ucurum’s studied the dry grinding
or media filling. The grinding rate significantly increased of gypsum using different ball size and distributions. They
with increased ball filling. Considering the minimum % of found that 20% (2 cm), 60% (3 cm) and 20% (4 cm) is the
fines (−45 mm) and the maximum production of the desired optimal for maximum grinding. They also revealed that the
size (−250 þ 45 mm), a ratio of 25% ball filling was deter­ impact of other variables such as ball size distribution on
mined to be the optimal for the next tests. product size was minor compared with other studied param­
eters such as mill speed, ball and powder filling ratio, and
grinding time
3.2.4. Effect of ball size diameter According to Serkan (2018) in wet calcite grinding, the
The diameter of the balls used in the ball mill plays a sig­ best results were obtained at (10% (40 mm) −10% (32 mm)
nificant role in the improvement and optimization of the − 40% (20 mm) − 40% (12 mm)).
efficiency of the mill (Lameck 2005). The size of the steel It is concluded that, using an appropriate ball size distri­
balls used during the milling process is related to the size of bution could improve the grinding efficiency and reduce the
the mill charge material and the desired size reduction energy consumption required for the process. The choice of
 PARTICULATE SCIENCE AND TECHNOLOGY 9

Figure 8. Particle size distribution at constant 25% ball filling, 15 min grinding time and different ball size diameter and distribution.

Figure 9. Effect of initial feed size milled at different ball size at 15 min grinding time.

ball sizes can influence the grinding kinetics, including the studies about the effect of grinding media size and initial
rate of breakage and the grinding time required to achieve a feed size on the grinding performance of different minerals.
specific particle size. This can affect production rates and Figure 9 illustrates how the initial feed particle size (−8mm,
process optimization. −4mm, and −2mm) and grinding media size (25.4 and
36.6 mm) affect milling performance. Throughout this study,
the grinding time, slurry and ball filling ratio remained con­
3.2.5. Effect of initial feed size stant. In terms of grinding rate, larger (36.6 mm) balls usu­
A number of researchers (Deniz 2012; Bwalya et al. 2014; ally perform better than smaller (25.4 mm) balls for all
Petrakis, Stamboliadis, and Komnitsas 2017) carried out examined feed sizes. The lowest grinding rate was observed
10 K. E. YASSIN ET AL.

Table 6. Different grinding conditions for producing the desired fraction (−250 þ 45 mm).
Ball filling Feed Product, %
Grinding Feed Water, Ball size, Fraction of
time, min quantity, g ml mm critical speed (Uc) kg J, % 100 % d50 −250 þ 45 mm −45 mm
15 600g (U ¼ 1) 324 36.6 0.70 4.15 25 −2 mm 1.15 mm 35.8 12.9
30 600g (U ¼ 1) 324 25.4 0.70 4.15 25 −2 mm 1.15 mm 60.83 26
30 600g (U ¼ 1) 324 36.6 0.70 4.15 25 −2 mm 1.15 mm 66.8 26.5
32 600g (U ¼ 1) 324 36.6 0.70 4.15 25 −2 mm 1.15 mm 70 27.3
30 600g (U ¼ 1) 324 25.4 0.70 4.98 30 −2 mm 1.15 mm 62.76 34.5
30 600g (U ¼ 1) 324 36.6 0.70 4.98 30 −2 mm 1.15 mm 63.5 36.5

larger ball diameters (36.6 mm). Based on the results, it was

decided to crush the feed to less than 2 mm then grinding
using 36.6 mm balls to reduce the amount of produced fines
Table 6.
Figure 10(a, b) displays the relationship between breakage
rate constant and different variables, % solid in slurry, %
ball filling, initial feed size, ball size diameters and distribu­
tion, for size fractions 250 and 45 micron, respectively. It is
shown that the maximum grinding rate for the size fraction
less than 250 micron was achieved when grinding a feed of
initial size 4 mm using 25.4 mm balls at maximum ball fill­
ing and minimum solid % in slurry. The minimum breakage
rate of size fraction less than 45 micron was achieved when
grinding a feed of initial size 2 mm using 36.6 mm balls at
minimum ball filling and at 65% solid in slurry. Considering
the maximum production of the desired size range
(−250 þ 45 mm) and minimum % of fines, it is decided to
crush the feldspar feed to less than 2 mm. The crushed feld­
spar should be grinded using 36.6 mm balls with 25% ball
filling and 65% solid in slurry.

4. Conclusion
Results displayed in this work showed that the grinding rate
decreased sharply at a high dense slurry of 80%, and at the
same time, the percentage of fine fraction (less than 45 um) is
high.
Also, increasing slurry density will result in a decrease in
mill rate and cause problems in slurry flow due to increased
viscosity. A solids content (of 65% solids) was chosen as the
optimum value from an economic standpoint.
The breakage rate increased as the ore quantity decreased.
The best grinding occurs at the slurry filling (U) between 0.5
and 1, corresponding to feed quantity between 300 to 600 gm.
Ball filling is one of the most critical factors influencing
milling, as, the tumbling action and the rate of breakage is
determined by how much the mill volume is filled by the
balls. Mill filling has a major influence on power draw and
energy consumption. The milling rate was shown to be sig­
nificantly higher when the ball filling was increased from
20% to 30%, which provides the highest rate of grinding in
terms of the milling rate constant k. When the ball filling
Figure 10. Relationship between breakage rate constant and different studied was increased from 20% to 30%, the percentage weight of
grinding conditions for size fractions: (a) −250 micron and (b) −45 micron at
U ¼ 0.5, and grinding time 15 minute. fractions less than 250 mm increased from 63.8% to 97.26%,
and fines −45 mm increased from 23.5% to 34.5%.
Balls of small diameter (e.g., 19 mm) were unable to
when grinding the feed of size less than 8 mm. The highest break up the large particle, making it the least effective
grinding rate with less produced fines (−45mm) was where a minimum grinding rate was observed. The highest
achieved when grinding the feed of size less than 2 mm with grinding rate was achieved when grinding with larger ball
 PARTICULATE SCIENCE AND TECHNOLOGY 11

diameters (36.6 mm). For fast grinding of a more significant Deniz, V. 2012. The effects of ball filling and ball diameter on kinetic
coarse fraction, the larger the ball size needed, the finer the breakage parameters of barite powder. Advanced Powder Technology
23 (5):640–6. doi: 10.1016/j.apt.2011.07.006.
product size, and the more significant the proportion of Deniz, V. 2016. An investigation on the effects of the ball filling on the
small balls (25.4 or 19.1 mm) required. breakage parameters of natural amorphous silica. Advanced Powder
Considering the maximum production of the desired size Technology 27 (4):1272–9. doi: 10.1016/j.apt.2016.04.017.
range (−250 þ 45 mm) and minimum % of fines, it was Deniz, V., and T. Onur. 2002. Investigation of the breakage kinetics of
pumice samples as dependent on powder filling in a ball mill.
decided to crush the feed to less than 2 mm, then grinding
International Journal of Mineral Processing 67 (1–4):71–8. doi: 10.
32 minutes using 36.6 mm balls with 25% ball filling, 65% 1016/S0301-7516(02)00041-8.
solid in slurry and slurry filling value U ¼ 1, at these condi­ Elbendari, A., A. Aleksandr, N. Nadezhda, and A. Anastasiya. 2020.
tions, the percentage of −250 þ 45 mm was 70% with fewer Selective flotation of phosphorus-bearing ores. E3S Web of
fines −45 mm ¼ 27.3%. Conferences 192:02021. doi: 10.1051/e3sconf/202019202021.
El-Mofty, S. E., A. M. Elbendari, K. A. Abuhasel, and A. A. El-Midany.
2020. Ultrafine dry grinding of talc by planetary mill: Effects of
operating conditions. Obogashchenie Rud 6:21–5. doi: 10.17580/or.
Disclosure statement 2020.06.04.
No potential conflict of interest was reported by the author(s). Gao, M. W., and E. Forssberg. 1989. The effect of powder filling on
selection and breakage functions in batch grinding. Powder
Technology 59 (4):275–83. doi: 10.1016/0032-5910(89)80086-5.
Funding Gupta, A., and D. S. Yan. 2006. Mineral processing design and
operation- An introduction. Amsterdam: Elsevier B.V.
The authors would like to thank the Science, Technology & Innovation Hassan, E.-S R. E., F. Mutelet, N. A. Abdel-Khalek, M. A. Youssef,
Funding Authority (STDF), Ministry of Higher Education and M. M. Abdallah, and A. H. El-Menshawy. 2020. Beneficiation and
Scientific Research, Egypt, for supporting this research work under separation of Egyptian tantalite ore. Key Engineering Materials 835:
grant number 45915. 208–13. doi: 10.4028/www.scientific.net/KEM.835.208.
Jankovic, A. 2003. Variables affecting the fine grinding of minerals
using stirred mills. Minerals Engineering 16 (4):337–45. doi: 10.1016/
References S0892-6875(03)00007-4.
Katubilwa, F. M., M. H. Moys, D. Glasser, and D. Hildebrandt. 2011.
Abdel-Khalek, N. A., A. Yehia, and S. S. Ibrahim. 1994. Technical note An attainable region analysis of the effect of ball size on milling.
beneficiation of Egyptian feldspar for application in the glass and Powder Technology. 210 (1):36–46. doi: 10.1016/j.powtec.2011.02.009.
ceramics industries. Minerals Engineering 7 (9):1193–201. doi: 10. Khumalo, N., D. Glasser, D. Hildebrandt, B. Hausberger, and S.
1016/0892-6875(94)90006-X. Kauchali. 2006. The application of the attainable region analysis to
Abouzeid, A., A. Seifelnassr, A. Ghorashi, Y. Mustafa, and A. F. M. comminution. Chemical Engineering Sciences. 61 (18):5969–80. doi:
Ibrahim. 2022. Comparative Studies on the grindability of some ore 10.1016/j.ces.2006.05.012.
minerals. Journal of Petroleum and Mining Engineering 24 (1):66–71. Lameck, N. N. 2005. Effects of grinding media shapes on ball mill per­
doi: 10.21608/jpme.2022.127156.1120. formances. Johannesburg, Gauteng, South Africa: University of the
Ahmed, M. M., G. A. Ibrahim, A. M. E. Rizk, and N. A. Mahmoud. Witwatersrand http://hdl.handle.net/10539/1506.
Lorig, H. G., and H. Clarence. 2017. Mineral processing. Encyclopedia
2016. Reduce the iron content in Egyptian feldspar ore of Wadi
Britannica, November 17, 2006. Accessed September 9, 2023. https://
Zirib for industrial applications. International Journal of Mining
www.britannica.com/technology/mineral-processing.
Engineering and Mineral Processing 5:25–34. Magdalinovic, N., M. Trumic, and L. Andric. 2012. The optimal ball
Austin, L. G., R. R. Klimpel, and P. T. Luckie. 1984. The process engin­ diameter in a mill. Physicochemical Problems of Mineral Processing.
eering of size reduction: Ball milling, 99–110. New York: SMEAIME. 48 (2):329–39.
Bwalya, M. M., M. H. Moys, G. J. Finnie, and F. K. Mulenga. 2014. Makokha, B., H. M. Moys, M. A. Muumbo, and J. R. Kiprono. 2012.
Exploring ball size distribution in coal grinding mills. Powder Optimization of in-mill ball loading and slurry solids concentration
Technology. 257:68–73. doi: 10.1016/j.powtec.2014.02.044. in grinding of UG-2 ores: A statistical experimental design
Cayirli, S. 2018. Influences of operating parameters on dry ball mill approach. Minerals Engineering 39:149–55. doi: 10.1016/j.mineng.
performance. Physicochemical Problems of Mineral Processing 54 (3): 2012.05.020.
751–62. Milan, T., N. Magdalinovi�c, and G. Trumi�c. 2007. The model for opti­
Chimwani, N., F. K. Mulenga, and D. Hildebrandt. 2015. Ball size dis­ mal charge in the ball mill. Journal of Mining and Metallurgy A:
tribution for the maximum production of a narrowly-sized mill Mining 43:19–31.
product. Powder Technology. 284:12–8. doi: 10.1016/j.powtec.2015. Mulenga, F. K., and M. H. Moys. 2014. Effects of slurry pool volume
06.037. on milling efficiency. Powder Technology 256:428–35. doi: 10.1016/j.
Cho, H., J. Kwon, K. Kim, and M. Mun. 2013. Optimum choice of the powtec.2014.02.013.
€ uzo�glu, B., and M. Uçurum. 2016. An experimental study on the
Oks€
make-up ball sizes for maximum throughput in tumbling ball mills.
Powder Technology 246:625–34. doi: 10.1016/j.powtec.2013.06.026. ultra-fine grinding of gypsum ore in a dry ball mill. Powder
Danha, G., N. Hlabangana, and E. Muzenda. 2018. Effect of ball and Technology. 291:186–92. doi: 10.1016/j.powtec.2015.12.027.
Petrakis, E., and K. Komnitsas. 2022. Effect of grinding media size on ferro­
feed particle size distribution on the milling efficiency of a ball mill:
nickel slag ball milling efficiency and energy requirements using kinetics
An attainable region approach. South African Journal of Chemical
and attainable region approaches. Minerals 12 (2):184. https://www.
Engineering 25 (1):79–84. doi: 10.1016/j.sajce.2018.02.001. researchgate.net/publication/358237016. doi: 10.3390/min12020184.
Deniz, V. 2011a. Comparison with some porous materials and the Petrakis, E., E. Stamboliadis, and K. Komnitsas. 2017. Evaluation of the
effects of powder filling on breakage parameters of diatomite in dry relationship between energy input and particle size distribution in
ball milling. Particulate Science and Technology 29 (5):428–40. doi: comminution with the use of piecewise regression analysis.
10.1080/02726351.2010.509855. Particulate Science and Technology 35 (4):479–89. doi: 10.1080/
Deniz, V. 2011b. Influence of interstitial filling on breakage kinetics of 02726351.2016.1168894.
gypsum in ball mill. Advanced Powder Technology 22 (4):512–7. doi: Santos, A. A., D. L. Machado, and S. M. Nascimento. 2022.
10.1016/j.apt.2010.07.004. Experimental study of clinker grinding in a ball mill and the
12 K. E. YASSIN ET AL.

behavior of the grinding media inside the equipment using DEM. Umucu, Y., and V. Deniz. 2014. The evaluation of grinding behaviors
Brazilian Journal of Chemical Engineering 40 (2):469–79. doi: 10. of quartz and feldspar. Tojsat 4:60–7.
1007/s43153-022-00261-3. Yassin, K. E., M. Ahmed, M. G.-E. Khalifa, and A. A. Hagrass. 2023.
Shoji, L. G., F. Austin, K. Smaila, P. Brame, and T. Luckie. 1982. Removing iron impurities from feldspar ore using dry magnetic
Further studies of ball and powder filling effects in ball milling. separation (part one). Archives of Mining Sciences 68 (1):19–33.
Powder Technology 31 (1):121–6. doi: 10.1016/0032-5910(82)80013-2. Zhang, Y., Y. Hu, N. Sun, R. Liu, Z. Wang, L. Wang, and W. Sun.
Tangsathitkulchai, C. 2003. Effects of slurry concentration and powder 2018. Systematic review of feldspar beneficiation and its comprehen­
filling on the net mill power of a laboratory ball mill. Powder sive application. Minerals Engineering 128:141–52. doi: 10.1016/j.
Technology 137 (3):131–8. doi: 10.1016/j.powtec.2003.08.048. mineng.2018.08.043.

View publication stats