Journal of Electrostatics xxx (2017) 1e5

Contents lists available at ScienceDirect

Journal of Electrostatics
journal homepage: www.elsevier.com/locate/elstat

Study of charged particles trajectories in free-fall electrostatic

separators
Hakima Labair a, Seddik Touhami b, Amar Tilmatine a, *, Samir Hadjeri c, Karim Medles b, d,
Lucian Dascalescu d
a
APELEC Laboratory, Djillali Liabes University of Sidi Bel-Abbes, Algeria
b
IRECOM Laboratory, Djillali Liabes University of Sidi Bel-Abbes, Algeria
c
ICEPS Laboratory, Djillali Liabes University of Sidi Bel-Abbes, Algeria
d
PPRIME Institute, CNRS-Poitiers University-ENSMA, F86962 Futuroscope, Poitiers, France

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

Article history: The objective of this paper is to develop a mathematical model for computing the millimeter-sized
Received 14 October 2016 particles trajectories in the free-fall electrostatic separator. The simulation involves the numerical so-
Received in revised form lution of motion equations of the particles subjected to electric and gravitational forces. The air-drag
7 January 2017
force and the impact of the particles with the electrodes were also considered. The resolution algo-
Accepted 10 January 2017
rithm is implemented as a MATLAB program that uses the results of electric field computation performed
Available online xxx
with COMSOL software. The model obtained is used to study the factors influencing the quality of the
products recovered at the outlet of the separator.
Keywords:
Electrostatic separation
© 2017 Elsevier B.V. All rights reserved.
Electric field
Electric charges
Particles trajectory

1. Introduction leads to an increase of the energy consumed for milling. Therefore,

The free-fall electrostatic separators are widely used in sorting
and purification of sub-millimeter size ores [1e4]. Currently, they
are also successfully applied to the sorting of granular plastics from
waste electrical and electronic equipment [5e8]. In these facilities,
granular material sorting is produced by the electrostatic force
acting on particles pre-charged in devices that use the triboelectric
effect. The quality of products recovered at the outlet of the elec-
trostatic separator is strongly related to the charge/mass (Q/m)
ratios of the particles in the mixture [9,10].
Several triboelectric charging devices have been developed in
recent years [10e13]. Currently, the fluidized bed is considered to
be the best solution [14,15], as characterized by moderate energy
consumption, as well as low operating and maintenance costs.
Improving the tribocharging conditions by reducing the size of the
treated particles is a way to increase the efficiency of the electro-
static separations.
However, particle size reduction is sometimes unnecessary and
* Corresponding author.
E-mail addresses: amar.tilmatine@univ-sba.dz (A. Tilmatine), lucian.dascalescu@
univ-poitiers.fr (L. Dascalescu).
attempts were made to optimize the electrostatic separation of
mm-size particles, by using numerical modelling techniques. Thus,
Calin et al. [16] were the first to perform numerical modelling of
particle trajectories in a free-fall electrostatic separator. However,
the analytical method they employed for the electric field compu-
tation was far from being accurate, especially in the proximity of
electrode edges. Recently, Younes et al. [17] compared the outcome
of free-fall electrostatic separation experiments with the results of
numerical simulations of particle trajectories, but the computa-
tional conditions were not very clearly described.
This paper aims at validating an accurate numerical model of
particle trajectory in a free-fall electrostatic separator. The model
uses information on the geometrical and electrical parameters of
the separation system and of the physical characteristics of the
particles of the product under study. Thus, it relies on measures of
Q/m ratio performed for the constituents of a mixture (ABS/PVC)
recovered at the outlet of a laboratory separator (Fig. 1).
The results obtained may be used to define the optimal values of
the control variables of the separator that would give maximum
gap between the collection points of two products, having different
mass and charge distributions.

http://dx.doi.org/10.1016/j.elstat.2017.01.010
0304-3886/© 2017 Elsevier B.V. All rights reserved.

Please cite this article in press as: H. Labair, et al., Study of charged particles trajectories in free-fall electrostatic separators, Journal of
Electrostatics (2017), http://dx.doi.org/10.1016/j.elstat.2017.01.010
2 H. Labair et al. / Journal of Electrostatics xxx (2017) 1e5

 
! 0
Fg¼ (4)
mg

where m represents the mass of the particle and g ¼ - 9.81 (m/s2) is

the gravitational acceleration.
The electrostatic force applied to a charged particle in the inter-
electrode region can be assessed by:
 
! ! E ðx; yÞ
F el ¼ Q E ¼ Q x (5)
Ey ðx; yÞ

where Q is the electrical charge of the particle and Ex (x, y); Ey (x, y)
components of the electric field vector.
The frictional force with the air applied to a particle is:
  
! 1 vx
Fa¼ Cf rprp2 v (6)
2 vy

where vx and vy are the x and y component of particle velocity ! v ; rp :

radius of the spherical particle of equal volume to the actual par-
ticle, Cf:friction coefficient between the particle and the air; r: mass
density of the particle.
Substituting (4e6) into (1) gives the equation of motion:
Fig. 1. Dimensions used in the free-fall laboratory separator.
 
1 pD2
max ðx; yÞ ¼ QEx ðx; yÞ  Cf r vvx ðx; yÞ (7.1)
2. Numerical simulation 2 4

The separator used in this study consists in two PMMA plates
suspended in a rectangular tower of same material (Fig. 1). Each
plate carries an aluminum electrode connected to a DC high-
voltage supply, to generate the electric field zone in which the
separation takes place. The inclination of the plates with respect to
the vertical plane is adjustable.
The charged particles are introduced into the electric field zone
through a funnel. The movement of the particles in this zone can be
characterized by the equation:
d!v X!
m ¼ F
dt
where m and ! v represent respectively the mass and velocity of the
P!
particle and F is the sum of the forces acting on the moving
particle.
At every moment, the charged particles are subjected to an
electrostatic force due to the action of the electric field in the inter-
electrode zone. The distribution of the electric potential U in this
area is given by the Laplace equation:
v2 U v2 U v2 U
þ 2 þ 2 ¼0
vx2 vy vz
The intensity E of the electrostatic field in the separation zone
can then be calculated from:
E ¼ VU
In this study, the solution of (2) is obtained using software
specialized in solving partial differential equations by the finite
element method (COMSOL). The results of calculations made by
this method are shown in Fig. 2.
During their movements in the electrostatic separation zone,
the charged particles are subjected to the force of gravity, electro-
static force and the friction force with the air.
The gravity force applied to a particle of mass m is:
 
1 pD2
may ðx; yÞ ¼ QEy ðx; yÞ  Cf r vvy ðx; yÞ þ mg (7.2)
2 4
The equation of motion described by the equation system is
solved using a finite difference method:
vnþ1 ¼ vn þ an $Dt
(8)
pnþ1 ¼ pn þ vnþ1 $Dt
where an, vn and pn respectively represent the vectors acceleration,
velocity and position at the time n·Dt.
Fig. 3 shows typical trajectories plotted by the simulation pro-
(1) gram. In a first series of simulations, the trajectory of a particle is
calculated for fixed values of the charge/mass ratio (Fig. 3a): 10 nC/
g; 7.5 nC/g; 5 nC/g; 2.5 nC/g. The obtained results show that
increasing the ratio Q/m causes an increase in the horizontal
deflection of the particle with respect to the symmetry axis of the
separator. The particle having a charge per mass ratio 10 nC/g
collides with the electrode of the installation at the point (x, y) ¼ (-
0.142 m, 0.742 m). The horizontal deflection of the particles
equals 0.206 m for 7.5 nC/g; 0.139 m for 5 nC/g and 0.069 m for
2.5 nC/g. In this simulation the maximum value of the horizontal
deflection is obtained with charge/mass ratios of the order of
(2) 7.5 nC/g. To avoid particle-electrode collisions, the charge should be
less than this value that can be predicted by numerical simulation.
The results of the calculation of particles trajectories for
different values of potential difference DU applied to the electrodes
(20 kV; 40 kV and 60 kV), are given in Fig. 3b. They show that the
increase of the potential difference DU causes an increase of the
(3)
particles deviation with respect to the symmetry axis of the
installation. The deviation varies from 0.046 m for DU ¼ 20 kV,
to 0.137 m for DU ¼ 60 kV.
An increase of the potential difference DU is desired to obtain a
good separation. However, an excessive increase of DU in the
presence of particles characterized by a high Q/m ratio may cause
collisions between particles and electrode that affect the quality of
the product recovered at separator outlet. It should be noted also
that excessive increase of DU may cause air breakdown in the inter

Please cite this article in press as: H. Labair, et al., Study of charged particles trajectories in free-fall electrostatic separators, Journal of
Electrostatics (2017), http://dx.doi.org/10.1016/j.elstat.2017.01.010
 H. Labair et al. / Journal of Electrostatics xxx (2017) 1e5 3

Fig. 2. Electrostatic field map and equipotential lines in (a): the median vertical plan of the separator; (b): horizontal plane at the upper end of the electrodes; (c): horizontal plane
at the lower end of the electrodes (electrode inclination angle: a ¼ 7 ).

Fig. 3. Trajectories of particles for different: (a) Q/m ratio, (b) potential difference DU, and (c): friction coefficient between the particles and air Cf (a ¼ 7 ; d1 ¼ 0.1 m; Q/m ¼ 5 nC/g;
rp ¼ 2 mm; r ¼ 1100 kg/m3; (x0,y0) ¼ (0,0) m; (vy0, vx0) ¼ (0,0) m/s; Cf ¼ 0; DU ¼ 60 kV).

electrode space.
Fig. 3c represents the results of trajectories calculations for
different values of the friction coefficient between the particles and
air: Cf ¼ 1/50, Cf ¼ 1/100 and Cf ¼ 0. Simulation results show that
the increase in the friction coefficient causes an increase in the
deflection of the particles.
3. Experimental study
In all experiments carried out in this work, the mixture to be
separated is composed of 50 g of ABS and 50 g of PVC. The distri-
butions of the particle mass of the two products are shown in Fig. 4,
obtained from a database of 100 measured values. Each value,
which represents an estimate of the mass of a single particle, was
obtained by measuring the total mass of ten particles randomly
selected divided by ten. The database values are then classified in
0.02 g intervals, the first class including the particles mass between
0.01 g and 0.03 g.
The tribocharging of the particles is performed in a worm-type
device (Fig. 5). The particles are continuously fed to the device and
get charged due to the impacts they have with the metallic worm.
The tribocharging process is controlled by adjusting the rotation
speed of the worm and the inclination angle of the tube with
respect to the horizontal. The charged particles are then introduced
into the free-fall separator. After separation, the charge per mass Q/
m was measured and the average particle radius was estimated for
each collector cell.
The specific charge Q/m of the particles is measured by using a

Please cite this article in press as: H. Labair, et al., Study of charged particles trajectories in free-fall electrostatic separators, Journal of
Electrostatics (2017), http://dx.doi.org/10.1016/j.elstat.2017.01.010
4 H. Labair et al. / Journal of Electrostatics xxx (2017) 1e5

Faraday pail connected to an electrometer and a high resolution

(0.001 g) scale. The results for a potential difference DU ¼ 60 kV and
an inclination of the electrodes with respect to the vertical a ¼ 7
are summarized in Table 1.
In this experience the majority of particles are recovered in the
extreme cells of the collector (Fig. 6). The gravity centers of posi-
tively and negatively charged products are located respectively at
xp ¼ 14.11 cm and xn ¼ 12.82 cm (positions measured relative to
the symmetry axis of the installation).
The values of charge/mass ratio of the product recovered in each
cell of the collector are displayed in Fig. 5b. The product recovered
in the vicinity of the central symmetry plane of the separator has a
very low charge/mass ratio as compared with that in the extreme
cells of the collector. On the other hand, the charge/mass ratio of
the positively charged product is higher than that of the negatively
charged product.
To facilitate the interpretation of the experimental results, an
image processing program was written that calculates the radius
of the circle having the area equal to the orthogonal projection
of the particle shape on a plane. The program was used to
analyze the size of the particles recovered in each cell of the
collector. Fig. 7 shows the distribution of the particle size in cells
{2; 6; 11}. It point out that the electrostatic separation of plastic
mixtures in a free-fall separator is based on the nature of the
materials of the mixture (ABS dominance in the cell 2; PVC
dominance in the cell 11, a balanced mixture in the cells 6) and
the particle size of each product. The particles recovered in the
Fig. 4. Distribution of particle mass for (a): ABS; (b) PVC.

Fig. 5. Worm-type tribocharging device.

Table 1
Charge Mass and charge/Mass of particles recovered in cells product collector
{1e12}; DU ¼ 60 kV a ¼ 7 ; «-» empty cell.

Cell N Xc (m) Q (109C) M (103 kg) Q/m (106C/kg)

1 e e e e
2 0.18 106.27 20.98 5.13
3 0.14 43.96 9.33 4.59
4 0.10 32.85 6.78 5.10
5 0.06 11.63 3.53 3.03
6 0.02 3.63 2.38 1.55
7 0.02 3.18 2.55 0.78
8 0.06 9.62 3.65 2.66
9 0.10 21.78 8.98 2.28
10 0.14 27.46 8.45 3.22
11 0.18 51.00 14.58 3.36
Fig. 6. (a): Total weight of the product in the collector cells and (b): Charge/Mass ratio
12 e e e e
measured in each cell.

Please cite this article in press as: H. Labair, et al., Study of charged particles trajectories in free-fall electrostatic separators, Journal of
Electrostatics (2017), http://dx.doi.org/10.1016/j.elstat.2017.01.010
 H. Labair et al. / Journal of Electrostatics xxx (2017) 1e5
Fig. 7. Radius of the particles recovered in the cells 2, 6 and 11.
Fig. 8. Comparison between simulation and experimental results.
extreme cells of the collector have small radii compared to those
in the central cells.
4. Validation of the numerical simulation results
In order to test the validity of the numerical modelling, the
computations were performed for 1000 particles having the pa-
rameters randomly selected in the following intervals: initial po-
sition of the particle, 1 to 1 cm; initial speed of the particle,
0.5e1.5 m/s; direction of the initial speed, 2 to 2 ; radius of the
particles, 1e3 mm; mass of particles, 0.1e4 mg; charge/mass
ratio, 5.5e5.5 nC/g.
After the analysis of the simulation results, the particles were
grouped according to their respective positions in the compart-
ments of collector. Then, the charge/mass ratio of the product
recovered in each box was estimated (Fig. 8). The error of the model
is lower than 10% in the four cells situated at the center of the
collector, i.e. from 8 cm to þ8 cm around the symmetry axis.
5. Conclusions
The numerical analysis of particle trajectories enables the
evaluation of the effects of different factors affecting the operation
of the free-fall electrostatic separator, including the voltage applied
to the electrodes and the force of friction with the air. The simu-
lation results are in good agreement with the results obtained
experimentally on a laboratory separator.
The obtained results show that a free-fall tribo-electrostatic
separator characterized by a height of 1.5 m makes possible the
separation of PVC/ABS granules of 4 mm average size. The particles
with charge/mass ratios varying from 3 to 5 nC/g are collected at a
Please cite this article in press as: H. Labair, et al., Study of charged particles trajectories in free-fall electrostatic separators, Journal of
Electrostatics (2017), http://dx.doi.org/10.1016/j.elstat.2017.01.010
5
distance ranging between 6 and 18 cm relative to the symmetry
axis of the installation (depending on the nature and the size of the
particle).
The simulation results also show that the deviation of weakly
charged particles can be improved by increasing either the elec-
trostatic field intensity or the height of the separator.
References
[1] R. Ciccu, R. Peretti, A. Serci, M. Tamanini, A. Zucca, Experimental study on
triboelectric charging of mineral particles, J. Electrostat 23 (1989) 157e168.
[2] R. Peretti, A. Serci, A. Zucca, Electrostatic K-Feldspar/Na-Feldspar and feldspar/
quartz separation : influence of feldspar composition, Mineral Process. & Extr.
Metall. Miner 33 (2012) 220e231.
[3] I.I. Inculet, Electrostatic Mineral Separation, RSP Ltd., John Wiley, 1984.
[4] K. Haga, Applications of the electrostatic separation technique, in: J.S. Chang,
A.J. Kelly, J.M. Crowley (Eds.), Handbook of Electrostatic Processes, Dekker,
New York, 1995, pp. 365e386.
[5] L. Dascalescu, Electrostatic separation of plastics from industrial wastes,
J. Electrostat 25 (2001) 282e288.
[6] I.I. Inculet, G.S.P. Castle, J.D. Brown, Electrostatic separation of plastics for
recycling, Part. Sci. Technol. 16 (1998) 91e100.
[7] S. Bendimerad, A. Tilmatine, M. Ziane, L. Dascalescu, Plastic wastes recovery
using free-fall triboelectric separator, Int. J. Environ. Stud. 66 (2009) 529e538.
[8] A. Tilmatine, K. Medles, S. Bendimered, F. Boukhoulda, L. Dascalescu, Elec-
trostatic separators of particles. Application to plastic/metal, metal/metal and
plastic/plastic mixtures, Waste Manag. 29 (2009) 228e232.
[9] M. Zelmat, A. Tilmatine, M. Rizouga, R. Gouri, K. Medles, L. Dascalescu,
Experimental analysis of a cyclone tribocharging device for free-fall tribo-
electric separation of plastic particles, IEEE Trans. Diel. Elect. Insul. 20 (2013)
1584e1589.
[10] A. Iuga, A. Samuila, R. Morar, M. Bilici, L. Dascalescu, Tribocharging techniques
for the electrostatic separation of granular plastics from waste electric and
electronic equipment, Part. Sci. Technol. 34 (2016) 45e54.
[11] M. Blajan, R. Beleca, A. Iuga, L. Dascalescu, Triboelectrification of granular
plastic wastes in vibrated zigzag shaped square pipes in view of electrostatic
separation, IEEE Trans. Ind. Appl. 46 (2010) 1558e1563.
[12] G. Buda, M. Bilici, A. Samuila, L. Dascalescu, Experimental study of the tri-
bochargingproces of plastic granular materials on a vibratory feeder device,
IEEE Trans. Diel. Elect. Insul. 20 (2013) 1489e1496.
[13] R. Kohnlechner, Triboelectric charging and electrostatic separation of diverse,
non-conductive mixed waste, especially plastic, employs enclosed vibro-
conveyor followed by in-flight separation influenced by non-linear electro-
static field. Patent DEI 9901 743, 2000-07-20.
[14] A. Iuga, L. Calin, V. Neamtu, A. Mihalcioiu, L. Dascalescu, Tribocharging of
plastics granulates in a fluidized bed device, J. Electrostat 63 (2005) 937e942.
[15] L. Calin, A. Mihalcioiu, A. Iuga, L. Dascalescu, Fluidized bed device for plastic
granules triboelectrification, Part. Sci. Technol. 25 (2007) 205e211.
[16] L. Calin, A. Mihalcioiu, S. Das, V. Neamtu, C. Dragan, L. Dascalescu, A. Iuga,
Controlling particle trajectory in free-fall electrostatic separators, IEEE Trans.
Ind. Appl. 44 (2008) 1038e1044.
[17] A. Younes, M. Younes, H. Sayah, A. Samuila, L. Dascalescu, Experimental and
numerical modeling of a new tribo-electrostatic separation process for
granular plastics mixtures, Part. Sci. Technol. 33 (2014) 189e196.