PTEC-12803; No of Pages 6

Powder Technology xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

Powder Technology

journal homepage: www.elsevier.com/locate/powtec

The effect of dry granulation on flow behaviour of pharmaceutical

powders during die filling
Serena Schiano a, Lan Chen b,⁎, Chuan-Yu Wu a,⁎
a
Deprtment of Chemical & Process Engineering, University of Surrey, Guildford GU2 7XH, UK
b
School of Medical Instrument and Food Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China

a r t i c l e i n f o a b s t r a c t

Article history: Flowability that quantifies the flow behaviour of powders is an important material attribute for such
Received 26 May 2017 applications as packing, hopper flow and powder transport. It is also one of the critical material attributes
Accepted 31 August 2017 of pharmaceutical formulations for solid dosage forms. It is anticipated that size enlargement via dry/wet
Available online xxxx
granulation will improve the flowability of feed powders, but it is still unclear how significant the
flowability can be enhanced. Therefore, in this study, an experimental investigation was performed to
Keywords:
Flowability
explore how dry granulation affects the flowability of pharmaceutical powders, such as microcrystalline
Dry granulation cellulose (MCCs), mannitol and lactose. Both as-received powders and binary mixtures were considered.
Roll compaction Granules of various sizes were produced using roll compaction followed by ribbon milling, and the
Critical filling speed flowability of as-received powders and produced granules was characterised using two methods: 1) the
Die filling critical filling speed measured using a model die filling system and 2) the flow index measured using a
Flodex tester. It was shown that the flowability increases as the size of the granules increases for all materials
considered. Furthermore, it was found that there is a strong correlation between the critical filling speed and
the flow index: the critical filling speed is proportional to the flow index to a power of −5/2.
© 2017 Elsevier B.V. All rights reserved.

1. Introduction
In the pharmaceutical industry, it is well recognized that composi-
tion variation and the quality of tablets are determined by material
properties and process conditions. One of the greatest challenges in
pharmaceutical development is to identify i) the causal relationship
between material properties, process variables and final product
properties, and ii) the critical material attributes dominating the
product properties, which is of practical importance to obtain high qual-
ity products. Pharmaceutical tablets are generally manufactured by
compressing dry powders or granules in a die, i.e., the die compaction,
which is the so-called tabletting process. The tabletting process consists
of three primary stages: die filling, compaction and ejection [1]. Die
filling is a process in which powders are deposited into a die under
the effect of gravity or suction. It is a critical process step during
tabletting, as the flow behaviour during die filling determines tablet
properties (e.g. weight variation, content uniformity), and dictates
the segregation tendency of powder blends during the tabletting
process [1–3].
During die filling, flow behaviour of powders depends on particle
density (i.e. true density) and bulk density. For example, Xie and Puri
[4] investigated the die filling process using a pressure deposition tester
⁎ Corresponding authors.
E-mail addresses: lanchen@usst.edu.cn (L. Chen), C.Y.Wu@surrey.ac.uk (C.-Y. Wu).
with alumina powders of different bulk densities. The tester generated
real time pressure distributions of powders at the base of the die during
die filling. It was shown that materials of low bulk density led to irreg-
ular and low reproducible pressure profiles, implying that powders of
lower bulk density had a higher tendency to produce non-uniform
packing during die filling, which led to non-homogeneous tablet density
during tabletting. The influence of particle density during die filling was
also examined by Guo et al. [5], who explored segregation behaviour of
a binary mixture consisting of particles having the same size but differ-
ent densities using a coupled discrete element method with computa-
tional fluid dynamic (DEM-CFD). It was shown that the difference in
densities caused segregation during die filling, in particular, light parti-
cles tend to settle on the top of the packed powder bed while dense par-
ticles at the bottom. It was also found that this tendency was enhanced
in the presence of air.
Particle size also affects die filling behaviour.For instance, Mills
and Sinka [6] explored the effect of particle size on gravity and
suction filling with different grades of microcrystalline cellulose
powders and found that fine particles showed intermittent flow
behaviour due to strong cohesion, while smooth mass flow was
observed for large particles. Wu et al. [7] investigated powder flow
behaviour during die filling using the positron emission particle
tracking technique (PEPT) that measured the velocities of individual
particles. Two grades of spherical microcrystalline cellulose powders
with different particle sizes were examined using a model shoe

https://doi.org/10.1016/j.powtec.2017.08.064
0032-5910/© 2017 Elsevier B.V. All rights reserved.

Please cite this article as: S. Schiano, et al., The effect of dry granulation on flow behaviour of pharmaceutical powders during die filling, Powder
Technol. (2017), https://doi.org/10.1016/j.powtec.2017.08.064
2 S. Schiano et al. / Powder Technology xxx (2017) xxx–xxx
system [1]. It was shown that coarse powders resulted in a higher
critical filling speed, indicating that they possess a better flowability
and a higher die filling efficiency can be achieved, comparing to the
fine powder. This was attributed to the presence of air in the die,
which can significantly inhibit the flow of small particles. These ob-
servations were consistent with the numerical results of Guo et al.
[8], who modelled the die filling process in vacuum (i.e. absence of
air) and in the presence of air using DEM-CFD. They found that, dur-
ing die filling in presence of air, the air inhibited the flow of powders
of small particles and of low density, as no significant difference was
observed for die filling with particles of different sizes and densities
in vacuum.
Many attempts were also made to understand the correlation
between powder flowability and die filling behaviour. For example,
Xie and Puri [4] argued that fine particles had poor flowability due to
the small size and the increased surface area, which led to an increase
of cohesive forces. Consequently, lower die filling efficiency and poorer
content uniformity were obtained. However, other studies showed that
there was not a strong correlation between powder flowability and
filling behaviour. For example, Felton et al. [9] investigated capsule
filling with mixtures composed of MCC and silicified microcrystalline
cellulose (SMCC) using a tamping-tape encapsulation machine, aiming
to understand the influence of powder flowability on the encapsulation
filling. They showed that the fill weight was higher and more reproduc-
ible with SMCC that has better flowability. However, similar results
were obtained for the MCC powder even though their flowability was
poorer. They hence suggested that powder flowability might not be a
critical parameter for encapsulation filling. Wu et al. [10] evaluated
the correlation between flow behaviours of powders during die filling
and flow properties characterised using various methods, include the
shear cell tester, flowmeter, angle of repose, Hausner index and Carr
index, and found that the flowability testing methods mimicking the
powder flow in an actual application gave the better indication of the
ability of powder to flow in that specific application than other powder
flowability tests.
Although die filling has attracted increasing attention in the past two
decades, previous studies primarily focused on feed powders but the
influence of the granule properties on die filling behaviour was not
investigated. Hence, in this study, the flow behaviour of granules
produced using dry granulation with roll compaction was investigated.
The influence of granule size was examined and the correlation
between flow behaviour of granules and flowability measurements
using the Flodex tester was explored.
2. Materials and methods
Three commonly used pharmaceutical excipients, microcrystalline
cellulose (MCC) of three different grades: Avicel PH 101, Avicel PH 102
and DG (FMC, Biopolymer, USA), lactose monohydrate (Granulac 140,
Meggle GmbH, Germany) and mannitol (Pearlitol 200 SD, Roquette,
UK) were considered. All the powders are of pure component apart
from MCC DG that is a formulated microcrystalline-based excipient
composed of 75% of MCC and 25% of anhydrous calcium phosphate
[11]. Three binary mixtures (see Table 1) of MCC Avicel PH 102 and
lactose were also considered. These samples were produced by mixing
the powders in a mixer (TURBULA T2F, Wab, UK) for 15 min at a
constant speed of 34 min−1 and named as mixture 1, 2 and 3 based on
their compositions.
Table 1
Mixtures composition.
Mixture Lactose (%) MCC Avicel PH102 (%)
1 25 75
2 50 50
3 75 25
Please cite this article as: S. Schiano, et al., The effect of dry granulation on flow behaviour of pharmaceutical powders during die filling, Powder
Technol. (2017), https://doi.org/10.1016/j.powtec.2017.08.064
The same custom-made gravity fed roll compactor as reported in Wu
et al. [12] was used to produce ribbons with a roll gap of 1.2 mm at a roll
speed of 1 rpm. The ribbons were then milled into granules using
a cutting mill (SM 100, Retsch, Germany) equipped with a 4 mm
mesh size screen at a constant speed of 1500 rpm. The produced
granules were then sieved into different granule size cuts (1–90,
90–250, 250–500, 500–1000, 1000–1400, 1400–2360 μm).
The flowability of the granules, as-received powders and powder mix-
tures was characterised using a Flodex tester (Flodex™, Gradco, UK) that
assesses the ability of powder to flow freely through an orifice in a funnel.
The diameter of the smallest orifice through which the sample can pass
three times was considered as the flow index. For each test, around 50 g
of sample were poured in the funnel, and the lever device was triggered
to open the orifice quickly to initiate the powder flow. All tests reported
here were repeated three times. From the Flodex tests, the higher the
flow index (the larger the orifice diameter through which the sample
can flow), the poorer the powder flowability.
Die filling experiments were then performed with the produced
granules, the as-received powders, and powder mixtures, using a
model die filling system (see also [1]), which consists of a shoe driven
by a pneumatic driving unit, a positioning controller unit and a displace-
ment transducer. Shoe speeds in the range of 10 to 400 mm/s were
employed. In each test, the mass deposited in the die was weighed
and the fill ratio was calculated by
mx
δ¼ ð1Þ
mT
where mx is the mass deposited into the die at a certain shoe speed,
and mT is the mass in a completely filled die that can be directly mea-
sured or calculated using the bulk density of a powder and the vol-
ume of the die. In this study, mT was directly measured when the
die is full of a powder.
From the definition of the fill ratio adopted by Wu and Cock [13],
the critical filling speed Vc and the index, n, were determined using
 n
Vc
δ¼ ð2Þ
Vs
where Vc is the critical filling speed, Vs is the shoe speed and n is a
parameter having a value of 0.8–1.6 for most powders [2,13].
3. Results
3.1. Flowability of granules made of pure powders
The flow indices (ψ) measured using the Flodex tester for granules
made of pure powders are presented in Fig. 1, in which the flow indices
of the as-received powders are also superimposed. The flow index is
Fig. 1. Flow index (ψ) as a function of granule size for granules made of four
different materials.
 S. Schiano et al. / Powder Technology xxx (2017) xxx–xxx 3

referred to as the orifice size through which the material can flow
continuously. Therefore, a lower flow index means that a smaller orifice
size is needed for the material to flow, implying a better flowability. For
the as-received powders, MCC powders have much larger flow index
values compared to the mannitol powders, implying that the mannitol
powder has the best flowability among four powders considered. Where-
as the MCC powders have a similar flow index value, close examination
reveals that the flowability of MCC PH102 is slightly better than MCC
PH101 and MCC DG. It is also interesting to note that when the granule
is sufficiently large (N 1000 μm) the flow index does not vary much
with the granule size, indicating that it becomes insensitive to the granule
size.
For the granules produced, it can be seen that the flow index decreases
with the increasing granule size for all the materials investigated. This in-
dicates that coarse granules possess better flowability than fine ones. This
is consistent with the results obtained by Xie and Puri [4], who suggested
that smaller particles have a larger surface area that increases the cohe-
sive forces between particles, thus a poor flowability. It is interesting to
notice that, for MCC, the granules generally have lower flow index values
than the as-received powders, indicating that improved flowability is
achieved using dry granulation. However, the mannitol granules general-
ly have higher flow index values than the as-received powder, indicating
that the flowability of granules is poorer than the as-received powder,
even though the granules are generally larger in size than the as-
received powders, which is somehow counter-intuitive. This is believed
to be induced by the change in particle shape when the particles are
converted to granules, as the mannitol particles are almost round while
the granules are angular, as demonstrated by Perez-Gandarillas et al.
[14]. This is in broad agreement with the observation of Mellmann et al.
[15], who also found that spherical particles have better flowability than
needled-shaped and rough particles.
Die filling experiments were also performed using the same
materials considered in Fig. 1. For each material, the fill ratios at
various filling speeds were determined using Eq. (1). Fig. 2 presents
the fill ratio as a function of the filling speed for granules of three
different size cuts: small (0–90 μm), medium (250–500 μm) and large
(1.400–2.360 μm), as well as for the as-received powders. It is clear
that the fill ratio decreases with the increase of filling speed for all the
materials investigated. From the definition of the fill ratio, a higher fill
ratio implies a higher mass deposited into the die. Hence, Fig. 2 shows
that, as the filling speed increases, less powder is generally deposited
into the die. To achieve a given fill ratio, lower filling speeds are gener-
ally required for finer granules. On the other hand, at the same filling
speed, a higher fill ratio indicates a better flowability, as more powder
flows into the die. In order to obtain the same fill ratio, die filling with
fine granules has to be run at a lower shoe speed than coarse granules.
Therefore, it can be seen in Fig. 2 that coarse granules have better
flowability than fine ones.
The critical filling speed (Vc) and the index n were determined
through multi-variate fitting of the fill ratio data as a function of the
filling speed, as shown in Fig. 2 for all materials and granule size cuts
considered. In Fig. 2, the lines are fitted curves using Eq. (2) for δb1.
The critical filling speed Vc as a function of the granule size is then plot-
ted in Fig. 3. It can be seen that, for all materials considered, as the gran-
ule size increases, the critical filling speed generally increases until it
reaches a plateau when the granule size is large (say N1000 μm). It is
also noticed that with such a large granule size (N1000 μm), the critical
filling speeds for MCC PH102 and MCC DG granules are similar and
slightly higher than the other two materials considered. Since a high
critical filling speed indicates a better flowability (as the die can be filled Fig. 2. Fill ratio as a function of filling speed for three different granule size cuts made of
(a) MCC PH 102, (b) MCC PH 101, (c) MCC DG and (d) mannitol.
with such a powder at a higher filling speed), Fig. 3 shows that the
flowability generally increases as the granule size increases. However,
when the granule size is very large (N 1000 μm), the flowability becomes
independent of the granule size. The corresponding n values are plotted interesting to observe that the n values generally vary between 0.8 and
against the granule size in Fig. 4, which shows that the n value for large 1.4, which is in broad agreement with the experimental data of Wu and
granules (N1000 μm) is slightly higher than that for small granules. It is Cocks [13] and Schneider et al. [2].

Please cite this article as: S. Schiano, et al., The effect of dry granulation on flow behaviour of pharmaceutical powders during die filling, Powder
Technol. (2017), https://doi.org/10.1016/j.powtec.2017.08.064
4 S. Schiano et al. / Powder Technology xxx (2017) xxx–xxx

Fig. 3. Critical filling speed as a function of granule size for various materials Fig. 5. Flow index (ψ) as a function of granule size for three binary mixtures.
considered.

3.2. Flowability of granules made of binary mixtures

Flow index values for the three binary powder mixtures and their
granules of different size cuts are presented in Fig. 5. It is shown that,
for the binary powder mixtures, mixture 1 (i.e. 75% MCC and 25%
lactose) has the highest flow index value, while mixture 3 has the
lowest one. This implies that for powder mixtures with a higher MCC
percentage, a larger orifice size is needed to achieve a steady powder
flow; hence it has a poorer flowability. For all the granules considered,
a similar behaviour to that shown in Fig. 1 was observed, i.e. when the
granule is small (b1000 μm), the flow index decreases with increased
granule size for all the mixtures; when the granule is sufficiently large
(N 1000 μm), a similar flow index value is obtained for all three mix-
tures. This further confirms that increasing granule size generally im-
proves the flowability, but further increase in the granule size has
a limited effect on the flow index when the granule size is large
(say N1000 μm).
Fig. 6 shows the fill ratio as a function of the filling speed for
granules made of the binary mixtures in three different size cuts:
small (0–90 μm), medium (250–500 μm) and large (1400–2360 μm).
The same features as shown for pure powders (Fig. 2) are also
observed for binary mixtures: the fill ratio decreases with the increase
of the filling speed; at the same filling speed, the fill ratio is higher for
larger granules, indicating an improved flowability. The critical filling
speed (Vc) and the index n for various sized granules made of three bi-
nary mixtures are shown in Figs. 7 & 8, respectively. Again, it is shown
that the critical filling speed increases as the size increases, but becomes
insensitive to the granule size when the granules become too large (say
N1000 μm). A close examination of Fig. 7 also reveals that a higher crit-
ical filling speed is generally obtained for the granules made of higher

Fig. 6. Fill ratio as a function of filling speed for granules of three different size cuts: small
(0–90 μm), medium (250–500 μm) and large (1400–2360 μm), made of (a) mixture 1,
Fig. 4. The index n as a function of granule size for various materials considered. (b) mixture 2 and (c) mixture 3.

Please cite this article as: S. Schiano, et al., The effect of dry granulation on flow behaviour of pharmaceutical powders during die filling, Powder
Technol. (2017), https://doi.org/10.1016/j.powtec.2017.08.064
 S. Schiano et al. / Powder Technology xxx (2017) xxx–xxx
Fig. 7. Critical filling speed as a function of granule size for granules made of different
binary mixtures.
fraction of MCC. The index n (see Fig. 8) increases as the granule size in-
creases and it varies in the range of 0.8–1.6, which is in broad agreement
with the value obtained previously.
4. Discussions
4.1. The effect of particle/granule size on flowability
It is recognized that powder flowability can be improved through
size enlargement, because the large particles generally have a smaller
specific surface area (i.e. surface area per unit mass) and lower inter-
particle cohesive forces, comparing to the small particles [4]. It is
hence expected that the granules produced from dry granulation pro-
cesses will have improved flowability compared to the feed powders.
However, limited evidence was provided in the literature to substanti-
ate this argument. In the present study, using 4 as-received powders
and 3 binary mixtures, granules were produced with roll compaction
and ribbon milling, and the effect of granule size was evaluated by divid-
ing the granules into various size cuts. The flowability of these granules
was evaluated using two different techniques: 1) the Flodex tester,
which determines the critical orifice size (flow index) through which
the materials can flow; and 2) the critical filling speed, i.e. the highest
speed at which the die can be completely filled in one passage. The
Flodex results (Figs. 1 & 5) showed that a smaller flow index is obtained
with larger granules, confirming that the flowability was generally
improved by granulating the fine powder into large granules. Similar
conclusions can also be drawn from the analysis of the critical filling
speed (Figs. 3 & 7), as the critical filling speed increases as the granule
size increases.
Nevertheless, two interesting phenomena were also observed in the
present study: 1) when the granule is very large (say N1000 μm),
Fig. 8. Parameter n as a function of granule size for granules made of different
binary mixtures.
Please cite this article as: S. Schiano, et al., The effect of dry granulation on flow behaviour of pharmaceutical powders during die filling, Powder
Technol. (2017), https://doi.org/10.1016/j.powtec.2017.08.064
5
further increase in the granule size offers limited advantage in
improving the flowability, as both the flow index (Figs. 1 & 5) and
critical filling speed (Figs. 3 & 7) remain unchanged as the granule
size increases further. This is believed to be due to the diminished
effect of inter-particle forces when the granules are sufficient large.
2) Particle shape can also play an important role in powder flow
behaviour, as demonstrated with the mannitol powder. It clearly
showed that, for mannitol, although the granules are generally larger
than the as-received feed powder the feed powder can be discharged
through a smaller orifice size than the granules. This is primarily due
to the change in particle/granule shape in dry granulation. It should
also be noted that the effect of particle shape appears dependent
on the actual flow conditions. As shown in Fig. 2d, during die filling,
the as-received mannitol powders only had a better flowability than
the granules of small size (b 500 μm), as a higher critical filling speed
than that of as-receive mannitol powders was obtained when the
granules are large. Hence a more comprehensive understanding of
particle shape on powder flowability deserves further study.
4.2. Correlation between flow index and critical filling speed
As argued by Wu et al. [1], powder flow behaviour during the die
filling process, which was used to determine the critical filling speed
in the present study, is to some extent similar to hopper flow as
happened in the Flodex measurement. For all the materials used and
the orifice sizes considered in the Flodex measurement, the mass flow
rate during the discharge can be approximated using the Beverloo
equation [16]:
M ¼ Cρb g 1=2 ðD − kdÞ
5=2
ð3Þ
where M is the mass flow rate, C is the Beverloo constant, D is the orifice
size, d is the particle/granule size, k is a parameter related to particle
shape and g is the gravitational acceleration. As the Flodex tester
determines the smallest discharge orifice size ψ, the lowest mass flow
rate of the powder during hopper discharge can be given as
5=2
Mmin ¼ Cρb g 1=2 ðψ − kdÞ ð4Þ
The flowability of the powder, Φ, can then be expressed as

a a
Φ ¼ f Mmin ¼ ¼ ð5Þ
Mmin Cρb g 1=2 ðψ − kdÞ5=2
where a is an empirical parameter. Eq. (5) indicates that a smaller value
of the flow index ψ represents a high value of Φ, i.e. better flowability.
The flowability of the powder, Φ, can be interpreted as the shortest
time to discharge a powder or pack the powder into a container/die.
For the die filling process at the critical filling speed Vc, the effective
discharge/filling time τ can be approximated as
L
τ¼ ð6Þ
Vc
where L is the effective length for die filling, which is related to the
length of the shoe and the die. To quantify the flowability of powder
Φ in this process, the flowability Φ can be related to the effective
discharge time as
b bV c
Φ ¼ f ðτ Þ ¼ ¼ ð7Þ
τ L
where b is a constant. Eq. (7) indicates that the shorter the effective
discharge time required to fully fill the die, the better the powder
flowability is. Moreover, it shows that the higher the critical filling
speed, the better the flowability.
6 S. Schiano et al. / Powder Technology xxx (2017) xxx–xxx
Fig. 9. Critical filling speed Vc as a function of flow index (ψ) for granules made of pure
powders and binary mixtures considered.
Combining Eqs. (5) and (7) leads to
aL
Vc ¼ 5=2
Cbρb g 1=2 ðψ − kdÞ
For d ≪ ψ, Eq. (8a) can be simplified to
aL
Vc ¼
Cbρb g 1=2 ψ5=2
Eq. (8b) indicates that the critical filling speed is proportional
to ψ−5/2, i.e.
V c ¼ αψ−5=2
Using the data reported in Figs. 1, 4, 5 and 7, a plot of the critical
filling speed against the flow index ψ is shown in Fig. 9. It can be seen
that the data using both the as-received powders and binary mixtures
can be well approximated using Eq. (9) with α=4.87×104 (mm7/2/s),
indicating that there is a strong correlation between the critical filling
speed Vc and the flow index ψ. On the other hand, Fig. 9 demonstrates
that Eq. (9) can well describe the interdependence of Vc with the flow
index ψ.
5. Conclusions
In order to explore whether dry granulation can improve the
flowability of feed materials, flowability of feeds powders, their
mixtures, and the granules of various sizes, were evaluated using a
Flodex tester and a model die filling system, from which the flow
index and the critical filling speed were determined, respectively. It
was found that coarse granules generally have a better flowability
Please cite this article as: S. Schiano, et al., The effect of dry granulation on flow behaviour of pharmaceutical powders during die filling, Powder
Technol. (2017), https://doi.org/10.1016/j.powtec.2017.08.064
than fine ones, and the flowability can generally be improved when
the granules are sufficiently large (say N500 μm). For small granules
(b500 μm), the effect of particle shape can be dominant as the round
feed particles appears to have a better flowability than the angular
granules, even though the latter may have a larger size than the
former. It was also observed that when the granule size is very
large (N1000 μm), the flowability became insensitive to the granule
size. Furthermore, it was found that there is a strong correlation
between the critical filling speed and the flow index. Both experimental
and theoretical analyses showed that the critical filling speed is
proportional to the flow index to a power of − 5/2.
Acknowledgements
This work was supported by the IPROCOM Marie Curie initial
training network, funded through the People Programme (Marie Curie
Actions) of the European Union's Seventh Framework Programme
FP7/2007–2013/ under REA grant agreement no. 316555. The authors
are grateful to FMC Chemicals prl, Brussels, Belgium, for the gift of
MCC powders used in this study.
ð8aÞ
References
[1] C.Y. Wu, L. Dihoru, A.C.F. Cocks, The flow of powder into simple and stepped dies,
Powder Technol. 134 (2003) 24–39.
ð8bÞ [2] L.C.R. Schneider, I.C. Sinka, A.C.F. Cocks, Characterization of the flow behaviour of
pharmaceutical powders using a model die-shoe filling system, Powder Technol.
173 (2007) 59–71.
[3] C.Y. Wu, DEM simulations of die filling during pharmaceutical tabletting,
Particuology 6 (2008) 412–418.
[4] X. Xie, V.M. Puri, Powder deposition in three parallel-oriented dies of cylindrical and
E shapes, Adv. Powder Technol. 23 (2012) 1–7.
ð9Þ [5] Y. Guo, C.-Y. Wu, K.D. Kafui, C. Thornton, 3D DEM/CFD analysis of size-induced
segregation during die filling, Powder Technol. 206 (2011) 177–188.
[6] L.A. Mills, I.C. Sinka, Effect of particle size and density on the die fill of powders, Eur.
J. Pharm. Biopharm. 84 (2013) 642–652.
[7] C.-Y. Wu, X.F. Fan, F. Motazedian, J.P.K. Seville, D.J. Parker, A.C.F. Cocks, Quantitative
investigation of powder flow during die filling using positron emission particle
tracking, Proc. Inst. Mech. Eng. E 224 (E3) (2010) 169–175.
[8] Y. Guo, K.D. Kafui, C.Y. Wu, C. Thornton, J.P.K. Seville, A coupled DEM/CFD analysis of
the effect of air on powder flow during die filling, AICHE J. 55 (1) (2009) 49–62.
[9] L.A. Felton, C.J. Wiley, A.L. Friar, Enteric coating of gelatin and cellulosic capsules
using an aqueous-based acrylic polymer, AAPS Pharm. Sci. 4 (2002) T3320.
[10] C.Y. Wu, B. Armstron, N. Vlachos, Characterisation of powder flowability for die
filling, Part. Sci. Technol. 30 (4) (2012) 378–389.
[11] FMC BioPolymer, In: Application Bulletin: Avicel® DG Specialty Binder for Dry
Granulation, 2014.
[12] C.-Y. Wu, W.-L. Hung, A.M. Miguélez-Morán, B. Gururajan, J.P.K. Seville, Roller
compaction of moist pharmaceutical powders, Int. J. Pharm. 391 (2010) 90–97.
[13] C.Y. Wu, A.C.F. Cocks, Flow behavior of powders during die filling, Powder Metall. 47
(2004) 127–136.
[14] L. Perez-Gandarillas, A. Perez-Gago, A. Mazor, P. Kleinebudde, O. Lecoq, A. Michrafy,
Effect of roll-compaction and milling conditions on granules and tablet properties,
Eur. J. Pharm. Biopharm. 106 (2016) 38–49.
[15] J. Mellmann, T. Hoffmann, C. Furll, Flow properties of crushed grains as a function of
the particle shape, Powder Technol. 249 (2013) 269–273.
[16] J.P.K. Seville, C.Y. Wu, Particle Technology and Engineering, Elsevier, London, 2016.