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

net/publication/222359159

Shape Separation of Particles

Article  in  Powder Technology · November 1992

DOI: 10.1016/0032-5910(92)87001-Q

CITATIONS READS
31 1,412

2 authors, including:

Masami Furuuchi
Kanazawa University
160 PUBLICATIONS   1,065 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Development of High Volume Air Sampling Tool for Aerosol Nanoparticles and Its Application to Dynamic Behavior Analysis and Risk Assessment View project

East Asia Nanoparticle Monitoring Network (EA-Nanonet) View project

All content following this page was uploaded by Masami Furuuchi on 25 October 2017.

The user has requested enhancement of the downloaded file.

Powder Technology, 73 (1992) l-9 1

Review Paper

Shape separation of particles

Masami Furuuchi” and Keishi Gotoh

Department of Energy Engineering Toyohashi University of Technology, Tempaku-cho, Toyohashi 441 (Japan)

Abstract

The separation of particles according to shape is necessary for improving the quality of powder products. The
shape of a particle significantly affects its bulk properties. Various types of shape separators reported so far are
reviewed in this paper to provide useful information when choosing the most effective one, and the separation
mechanism and features are compared. Problems in the further development of the technique are also explained.

Introduction etc., and a combination thereof, particle separation

based on a particular single property is difficult in
Bulk mean properties of particles depend on the general. Hence, in the shape separation of particles a
properties of individual particles, such as size, density, dynamic characteristic predominantly affected by the
shape, cohesive force, etc. Hence, the particle quality, particle shape is utilized, while keeping the other factors
based not only on the size but also on the particular unchanged.
property, should be controlled. In order to adjust the The shape separation methods proposed so far can
particle size, many separation techniques have been be classified into four groups which make use of the
developed and extensively used so far in powder in- difference in 1) the particle velocity on a tilted solid
dustries [l]. Research into particle separation based wall, 2) the time the particles take to pass through a
on shape, however, is at an early stage, in spite of its mesh aperture, 3) their cohesive force to a solid wall,
significance in improving the quality of powder products and 4) their settling velocity in liquid. In the following,
PI* the separation mechanism and features of various meth-
In the present article, the shape separation techniques ods will be explained.
reported so far are reviewed in order to compare their
separation mechanisms and features. Problems and
future subjects in developing the techniques are also Shape separation by tilted plate
discussed.
This method is the most basic and simple and many
shape separation methods proposed so far belong to
Shape separation methods thiS category. As pointed out by Yamamoto et al. [3],
Nakagawa et al. [4] and Furuuchi et al. [5], in this
method particles tend to be separated according to the
The difference in the dynamic behavior of particles
flatness of the side view of the moving particle. The
can be utilized in the separation on the basis not only
shape separation appears applicable particularly for
of shape but also of the other criteria. Since the dynamic
round particles which roll on the plate but not for flat
behavior of particles depends on shape, size, density,
particles which slide on it [5]. As reported by Shinohara
*Current address: Dept. of Civil Engineering, Faculty of En- [6], the jumping of particles on the plate decreases the
gineering, Kanazawa University, Kodatsuno 2-40-20, Kanazawa difference in. the particle velocity due to the particle
920 (Japan). shape. Hence, the plate is usually tilted by a few degrees

0032-5910/92/$5.00 0 1992 - Elsevier Sequoia. All riehts reserved

2

to prevent particles from sliding or jumping. The tilted [9] blew off the particles with the air jet. It is hard to
rotating disk, the tilted rotating cylinder and the tilted separate and collect particles at the same time in the
vibrating trough have been devised for practical ap- lower part of the disk. Contrarily, the conical disk can
plication. The lower limit to the particle size in these separate particles everywhere on the disk so that the
shape separators may be a few hundred pm, because separation capacity becomes large and separated par-
small particles become liable to agglomerate owing to ticles can be collected at any position along the disk
the influence of the humidity and the cohesive force edge [15, 16, 19, 201. Klar [9] continuously separated
between them. atomized copper particles of #140 _ #200 into 95% of
spherical particles and the rest non-spherical ones with
Tilted rotating disk a capacity of 14 kg/day. Since non-spherical particles
The trajectory of particles fed onto the center of a are apt to move randomly on the tilted disk and hence
tilted or conical disk rotating at a low speed (< 10 are trapped into the flow of spherical particles, the
rpm) changes according to the particle shape. Spherical separation procedure should be repeated to attain high
particles roll down faster while non-spherical ones stay accuracy [14]. In order to separate finer particles, the
longer on the disk [7-211. Figure 1 shows a schematic rotating disk is vibrated to agitate the particles [21].
diagram of the shape separator devised by Yamamoto A similar type of shape separator, a tilted belt conveyor,
et al. [12], which consists of a tilted glass disk, hoppers has recently been devised [22].
and vibratory feeders. The binary mixture of spherical
and non-spherical particles is continuously fed onto the Tilted rotating cylinder
point f near the disk center and scraped by a logarithmic- The residence time of particles in a tilted, rotating
spiral curved scraper placed on the disk. Spherical and cylinder may be related to the particle shape [23, 241.
non-spherical particles are collected respectively in Figure 2 depicts the shape separator devised by Waldie
reservoirs P, and Pi. The disk inclination usually ranges [25]. A sample mixture of particles is loaded into a
from 3 to 7” and various types of particle scrapers are rotating cylinder A. The cylinder is gradually tilted to
devised: Viswanathan et al. [17] and Aravamudhan et the position B and spherical particles flow out. Before
al. [18] used a plate scraper and Riley [8] and Klar non-spherical particles reach the lower end, the cylinder
is returned to the original position A. This procedure
is repeated using a cam mechanism. A baffle shown
in Fig. 3 is set in the cylinder to agitate the particle
flow and hence to eliminate the effect of interaction
among particles.
Figure 4 shows a shape separator consisting of a
tilted cylinder with a bottom cover and a row of tilted
blades attached to the inside wall [4, 261. When the
cylinder, tilted by 2 N 4”, rotates, particles fed into the
bottom are scraped up by the blades and move upward
on the tilted wall, and then flow down owing to gravity.
Particles having positive net velocity toward the upper
end flow out of the cylinder, leaving the others in the
bottom. For each particle, there exists a critical ro-
tational frequency of the cylinder above which the

1 Yapper for spherical particles

2 Hopper for non-spherical particles
3,4 Electric vibratory feeder
5 Collecting vessel
6 Electric vibratory conveyor
7 inclined rotating disk
8 Stationary spiral scraper
Ps Spherical particle products
Pi N,on-spherical particle products
Fig. 2. The tilted rotating cylinder driven by a cam mechanism.
Fig. 1. The tilted rotating disk with a logarithmic-spiral scraper. (A) Charge and return position, (B) Separating position.
 3

114,
T
Tilted vibrating trough
w
Abe et al. [29] devised a shape separator utilizing
I I
I the difference in the particle motion on a tilted vibrating
Tube-l ,
I I trough shown in Fig. 5. A brass open channel, 6 cm
I in width and 23 cm in length, is mounted on a vibrator
I I
with inclination angle (Y.The trough is vibrated in the
Baffle -

5:
direction of the angle p with a frequency of 60 Hz.
The amplitude is regulated at 132 pm to prevent particles
from jumping. A mixture of spherical and non-spherical
Liz
particles is fed continuously onto the vibrating trough
through a slit placed at -5 mm above the bottom
plate. The particle size is in the range 210 to 1190
pm. According to the coefficient of static friction, the
!i!
._I - 4 particle velocity on the trough becomes positive or
Fig. 3. A baffle inserted in the cylinder shown in Fig. 2. negative as shown in Fig. 6. Solid curves in Fig. 6
denote the theoretical prediction by Taniguchi et al.
[30]. Since the coefficient of static friction depends on
the particle shape, particles can be separated on the
ho\der
basis of shape by adjusting the operational conditions
of the trough. According to the numerical analysis by
Shinohara et al. [31] for the particle motion on a tilted
vibrating plate, the velocity and the moving direction
of the particle depend on its size and aspect ratio and

\
roller

S = particle trajectory

Fig. 4. The tilted rotating cylinder with a raw of blades.

particle flows out. Particles can be separated according

to the difference in the critical rotational frequency
N,, which is sensitive to the particle shape [4, 51. Fig. 5. The tilted vibrating trough.
Particles can be continuously separated by feeding them
from the middle of the cylinder with no bottom cover
[27]. The merits of this shape separator are 1) that
high separation efficiency can be attained in a short
time owing to the multi-staged separation by the row
of blades and 2) that particles can be classified according
to shape merely by adjusting the rotational frequency
of the cylinder [4, 261. Keunen [28] classified particles
according to their residence time on a tilted rolling
cylindrical trough. This may belong to the type of the
tilted rotating cylinder.
In the tilted rotating cylinder, particles flow on the
cylinder bottom so that the effective separation area
is small and the interaction among particles may be
larger than that on the conical rotating disk. It can be 2 ---- Glass beads
L -60 ‘\\I
I-
improved, however, by agitating the particle flow with t k

blades or baffles. Since the difference between the -so- 0 0.1 0.2 0.3 0.4 0.5 0.6
velocities of spherical and cylindrical particles is neg- Amplitude (mm)

ligible in the rotational direction of the cylinder but Fig. 6. The relation between the transporting velocity of particles
is very large in the axial direction, the mixture of these and -the amplitude of the vibration. Solid curves denote the
particles can be easily separated. theoretical results by Taniguchi et al. [30].
the wall friction as well as the vibratory condition. Tilted chute
Particles having an aspect ratio less than about 0.7 can In the shape separators mentioned above, the solid
be separated, which cannot be done by the separation wall is tilted only by a few degrees so as to prevent
methods mentioned previously. Figure 7 depicts the particles from sliding or jumping. In contrast, Glezen
tilted vibrating trough used by Ridgway et al. [32, 331 et al. [36] utilized the jumping motion positively. As
and Lenn et al. [34], which was originally designed for shown in Fig. 9, the shape separator consists of a feed
sorting industrial diamond crystallines (Jeffrey-Galion hopper, chute A with gate, deflector, one collector vial,
Ltd., Johannesburg). It consists of a triangular matt- chute B with gate, deflector and two collector vials.
surfaced deck with a small slope to the horizontal and Each chute is made of six long narrow passages shown
a larger slope to the vertical. Particles are fed by a in Fig. 10. The material of the chute floor and walls
vibratory feeder into one corner of the deck and move is aluminum sand-blasted to the texture of 600-grit
across it by gravity and the deck vibration, following emery paper. The slopes of chute A and B are re-
different trajectories according to shape. Spherical par- spectively 26.0 and 18.5”, which are much larger than
ticles leave the deck at the lower point of the right those of other shape separators. Particles fed from the
edge, while irregular ones leave at the higher point. hopper through the gate bounce up and down and
The particles are collected in 13 pots placed next to collide with the wall repeatedly while flowing downwards
the deck in order to classify them into various shapes. owing to gravity. Flatter particles are unlikely to bounce
Iwata et al. [35] made use of the shape difference in and experience higher friction on the wall so that the
the particle trajectory on a horizontal circularly vibrating descending velocity becomes lower. Particles which trav-
disk with peripheral wall (Fig. 8). Non-spherical par- erse chute A in more than 10.3 s are classified as
ticles, fed between the wall and the circular sink at tabular; particles which traverse chute A in less than
the center, move toward the wall then along the wall 10.3 s and traverse chute B in less than 12.5 s are
to the direction inverse to the vibration, while spherical classified as spherical; and the others as intermediate.
ones with higher outward velocity rebound from the This procedure is automated by changing the flow
wall to the sink, where they are captured. The circular direction of particles by the deflectors. It is said that
vibration may be an effective way of dispersing the the device is applicable to any particle larger than
particles. 0.35 mm.

Feed
Sample

h
_+

~,&iiG?2L
Vibrator
Spherical
particles

Fig. 7. The Jeffrey-Galion shape-sorting table. Intermediate Spherical

Fig. 9. Shape separation by tilted chutes.

Vibrating direction

Ceiling

Partition

Wall

Floor

Fig. 8. The horizontal circularly vibrating disk. Fig. 10. Chute battery.
Shape separation by sieves (lj: Sph&d

Shape dependency on the time required for a particle

to pass through a mesh aperture has been reported
[37-43]. As the particle elongation increases, for in- 4 0.10

stance, the passage time increases because the elongated Z- 0.08

(2)
particle takes a long time to change its orientation and 0.06

pass through the mesh aperture. The relation between 0.04

0.02
the particle shape and the passage time has been (1)
0.001 k
investigated by Ludwick et al. [38], Roberts et al. [39], 0 4 8 12 16 20
Distance Iid
Endoh et al. [41, 421 and Nakayama et al. [43]. Endoh
et al. [41, 421 concluded that the rate for a cylindrical Fig. 12. The relation between the distance from the feed point
and the average shape of particles collected at the corresponding
particle to pass through the rectangular mesh screen distance. VAR denotes the secondary moment of the radius of
is inversely proportional to the cubic power of the the particle profile about its centroid.
particle elongation (length)/(diameter). The passage
rate, however, is also affected by the operational con-
ditions of sieves such as the vibratory wave form and
the shape of the sieve opening. Hence, the relation
between the particle shape and the passage rate is
determined experimentally. Although this separation
method can be applied to a wide range of particle
sizes, the lowest limit may exist because of the choking
of the particles on the screen. The particle size should
be made more uniform than that in the shape separators
of the tilted plate type. A mixture of spheres and
elongated or flat particles as well as a mixture of
elongated particles may be easily separated by this
method.
I
h-tap f’
Tilted vibrating screen base plate
Figure 11 depicts the shape separator reported by Fig. 13. The sieve-cascadograph.
Hsyung et al. [44] and Monts et al. 1451 which is composed
of an inclined screen, particle collectors aligned under
and becomes zero for circles. As the distance increases,
the screen, an oscillator, an amplifier and a vibrator.
the particle elongation becomes larger. The peak of
Particles fed onto the upper end of the screen flow
the curve in Fig. 12 may be caused by the mismatch
down by gravity and the vibration. Spherical particles
of the operational conditions and the choking of the
liable to pass through the screen are collected at the
particles on the screen.
upstream while elongated ones flow toward the lower
end and are collected. The relation between the distance
fibrating stacked screens
from the feed point and the average shape of particles
collected at the corresponding distance is shown in Fig. Meloy et al. [47, 481 and Clark et al. [49] proposed
12, where VAR denotes the secondary moment of a new device to measure the shape distribution of
radius of the particle profile about its centroid [46] particles. The device, called a sieve-cascadograph, con-
sists of a stack of identical screens placed in a sieve
shaker (see Fig. 13 [49]). The stack of 20 sieves is used
[47-49]. Previously sieved sample particles are poured
into the top sieve and the shape distribution is de-
termined by measuring the flow rate of the particles
at the bottom sieve. As shown in Fig. 14 [49], the
residence time of spherical particles in the stack is
much less than that of elongated ones, so that the
mixture of these particles can be separated. Even if
E: Particle collector I the shape difference is small, increasing the number
F: Shelf
of sieves gives rise to an increase in the separation
Fig. 11. The tilted vibrating screen. efficiency. As can be seen from its separation mechanism,
6

particles on the Screen

Time (~1

Fig. 14. The difference between normalized cumulative plots for

angular stone and spherical red wheat.

only the batch operation is applicable. Orr [50] in- Fig. 15. The shape separator consisting of a rotary drum with
vestigated the particle flow through a vibrating vertical small circular holes to hold particles on by air suction.
column loosely packed with glass beads to determine
the influence of the particle size and shape on the rate
of passage. The residence time of particles may be
related to the distance traveled by the particles through
the column as well as the particle size and shape and
the void size. The method may also be used for the
shape separation of particles.

Rotating cylindrical sieve

A rotating cylindrical sieve is used for separating
elongated zinc particles on the basis of the aspect ratio
[51, 521. The particles fed into the horizontal drum
are separated according to residence time. Longer
particles are collected at a later period of time. It may
be a merit of the sieve drum that during the operation, : I I I I I
6 0 20 40 60 60 100
lodged particles can easily be removed. Continuous 4:

separation can be conducted by tilting the sieve, where Relative humidity [%I (ZO’C)

the particles continuously fed from the upper end of Fig. 16. Adhesion forces of glass powders having different ir-
the drum are separated according to the aspect ratio regularities measured at 50% remaining particles on the cen-
trifuged cylinder in relation to the relative humidity.
[511.

pressure above 20 mmH,O, the mixture of spherical

Particle holding or adhesion methods glass beads and non-spherical crushed ones of #40 _ #45
in size was efficiently separated. The surface roughness
Sano et al. [53, 541 recently developed a new shape of particles as well as their elongation or flatness may
separator. As shown in Fig. 15, the separator consists affect the performance of the shape separator.
of a drum with circular holes of 0.3 mm in diameter, The adhesion force of particles to a solid wall increases
whose inside is divided into the inhaling section A and with increasing particle sphericity at a constant humidity,
the exhaling section D. Particles fed from the hopper as reported by Sano et al. [55] (see Fig. 16), according
are spread on the drum and held on the holes by the to whom the adhesion force of irregular particles onto
air suction as shown in the magnified figure. Spherical the solid wall is unstable owing to the small effective
particles are held more tightly than non-spherical ones contact area. Sano et al. [56] devised a shape separator
because of the narrower gap between the particle and shown in Fig. 17 which utilizes the shape dependency
the hole. Making use of the difference in the particle of the particle adhesion ‘force. A glass cylinder of
holding force, particles can be separated according to 0150~ 120 mm is rotated at 4 rpm and vibrated by
shape. The non-spherical particles are brushed off by an electric magnet with an amplitude of 0.35 mm and
brush B and collected into reservoir C while the spherical a frequency of 50 Hz. The mixture of spherical and
ones are carried to section D, blown off and collected non-spherical glass beads fed from a hopper is spread
into reservoir E. The separation accuracy is determined on the cylinder. The non-spherical particles fall down
by the suction pressure in section A. By keeping the into reservoir A while the spherical ones carried on
 7

sample
1 /hopper 1

10
3

0” 1

0.1
10-1 100 10’ 102 103 194 105 10s 10’
II I I. .I I brush I / 0, \ I 1
Rs (-1

Fig. 19. The drag coefficient against Reynolds’ number for sphere,
disk and cylinder.

Fig. 17. The shape separator making use of the adhesion force
of the particle onto the solid wall.

-
‘OOr-

Fig. 20. Dynamic shape factors averaged over all orientations as

a function of axial ratios of the spheroid.

x
is Stokes’ condition (Re -=zl), Davies [58] theoretically
g 20
a zoo-2704
obtained the relation between the aspect ratio of the
: /
L
1 1 spheroid and the dynamic shape factor K as depicted
01 ’ ’ ’ “’ ’ “If ’ ’ ’ J in Fig. 20 where K = VJV, R = (major axis)/(minor axis),
0 20 40 60 80 100
V, is the terminal velocity of a sphere of the same
relative humidity C%I (20 “C)
volume as the spheroid and V is the terminal velocity
Fig. 18. The recovery efficiencies of spherical and non-spherical of the spheroid. As can be seen in Fig. 19, the terminal
particles. velocity becomes smaller for the flatter or elongated
particle with the same volume. Hence, the particles
the cylinder are brushed off and collected into reservoir can be separated according to the difference in the
B. Since the humidity and the temperature affect the settling velocity. It should be noted that a spherical
adhesion force, they must be adjusted so as to make particle does not always settle faster than a non-spherical
it large. The recovery efficiency of the spherical particles one even if it is the same size because the settling
increases with increasing humidity, as shown in Fig. velocity of the particle depends not only on the drag
18. It should be noted that the shape separator can coefficient but also on the mass and projected area in
efficiently separate particles smaller than 100 pm in the settling direction. In the Stokes region, the settling
diameter. The separation performance is sensible to velocity becomes smaller for particles with irregular
the humidity and temperature of the atmosphere and surfaces [58]. The effect of the particle orientation
the contamination of the cylinder wall so that the caused by the velocity gradient cannot be ignored in
operation is very hard in practice, though this is a the flowing fluid. The smallest size separable is limited
general problem encountered in any process dealing by the formation of colloids or agglomerates. In prin-
with fine particles. ciple, particles of a few pm in size can be separated.
Figure 21 shows the shape separator developed by
Murali et al. [59]. It consists of a rectangular channel
Settling velocity methods B, a pump J, a sump H and a fine wire mesh D to
rectify the flow direction. The water which has over-
Particles settling in a fluid experience the drag force. flowed from a weir F is returned to the sump. Sample
The drag coefficient C, depends on the particle shape particles funneled into the liquid at a short distance
as well as the particle Reynolds number Re as shown from the mesh rectifier settle along some trajectory to
in Fig. 19 [57]. For a spheroidal particle settling under the bottom of the channel, where they are collected
8

A Conclusions

Various shape separators proposed so far are sum-

marized and compared. The separation performance
of the shape separators has been explained by using
a single shape descriptor measured from one direction
despite the fact that the particles are separated on the
basis of the three-dimensional form and various physical
properties. Hence, the difference in the performance
Fig. 21. Basic construction of the continuous settler: A=feed between shape separators remains unknown so that it
funnel; B=channel; C=inlet for liquid; D =rectifier mesh; E= is impossible to choose the best method even if the
collection funnels; F = V-notch weir; G = liquid outlets; H = sump;
shape characteristics of particles are given. Although
J = circulation pump.
the authors tried to elucidate the features of each
shape separator, most descriptions may be just a pre-
TABLE 1. Results of shape separation of binary mixtures of
spherical and non-spherical particles using the continuous settler
sumption. For practical applications, the features and
merits of each separation method should be specified
Flow rate Collection (g)
more clearly.
(cm’ s-l) Using the shape separators proposed so far, a wide
First funnel Second funnel range of particles can be separated. However, new
separation techniques based on a different principle
Spherical Non-spherical Spherical Non-spherical
should be developed in order to solve extensive problems
285.2 5 17 12 8
in the shape separation of particles. The high speed
281.5 3.5 20.5 9.5 2.5 shape separation of particles with sizes less than 100
277.7 6 21 14 8 pm may be important for practical applications.
253.6 4 17 11 5
240.4 11 23.5 8 3
223.2 5.5 21 7 4
216.6 8 30 6.5 2 References
203.9 8 22 6.5 1.5
185.2 8 24 6.5 1.5 1 J. K. Beddow, Chem. Eng., 10 (1981) 70.
138.8 10 23 4 0.3 2 J. K. Beddow, Particle characterization in Technology, vol. 2,
127.8 7 23 2.5 0.6 CRC Press, Boca Raton, FL, 1984.
106.4 7 19 4 0 3 K. Yamamoto and M. Sugimoto, .I. Sot. Powder Technol. Jpn.,
22 (1985) 3.
Feed rate =2.0 g s -‘; feed amount=50 g; feed ratio 1:l; feed
4 M. Nakagawa, M. Furuuchi, M. Yamahata, K. Gotoh and J.
at water surface.
K. Beddow, Powder Technol., 44 (1985) 195.
5 M. Furuuchi, E. Ohno and K. Gotoh, A&. Powder TechnoZ.,
in funnels E and removed in a batch or continuous I (1990) 89.
6 K. Shinohara, Powder Technol., 48 (1986) 151.
mode. A binary mixture of round (mustard seeds) and
7 F. G. Carpenter and V. R. Deitz, J. Res. Nat. Eur. Stand.,
non-spherical (silica gel) particles, both of which are 47 (1951) 139.
sieved to the average size of 1.5 mm, is separated. 8 G. S. Riley, Powder Technol., 2 (1968/69) 315.
Typical results are listed in Table 1. Most non-spherical 9 E. Klar, Powder Technol., 3 (1969/70) 313.
particles are collected in the first funnel, and spherical 10 M. Sugimoto and K. Yamamoto, .I. Res. Ass. Powder Technol.
Jon., 8 (1971) 35.
ones in the downstream funnels, mainly in the second.
11 M. Sugimoto; K. Yamamoto and J. C. Williams, J. Chem.
The recovery efficiency of each particle and the overall Eng. Jpn., 10 (1977) 137.
separation efficiency are dependent on the flow rate 12 K. Yamamoto and M. Sugimoto, .I. Sot. Powder Technol. Jpn.,
of water and the height of the feed funnel. The overall 16 (1979) 521.
efficiency is comparable with that of the tilted rotating 13 S. Endoh, Y. Kuga, J. Koga and K. Yamaguchi, Kagaku
Kogaku Ronbunshu, 8 (1982) 476.
disk separator devised by Aravamudhan et al. [18] who
14 S. Endoh and K. Yamaguchi, J. Chem. Eng. Jpn., I7 (1984)
used the same particles. The apparatus could be im- 21.
proved to achieve more accurate separation of the 15 K. Yamamoto and M. Sugimoto, J. Sot. Powder Technol. Jpn.,
particle shape by setting a number of narrower funnels 22 (1985) 626.
on the channel floor. 16 K. Yamamoto and M. Sugimoto, .I. Sot. Powder Technol. Jpn.,
The hydroclone was recently used for the shape 22 (1985) 813.
17 K. Viswanathan, S. Aravamudhan and B. P. Mani, Powder
separation of fine particles with diameter lo- 100 pm Technol., 39 (1984) 83.
and blocky mica particles were separated from flaky 18 S. Aravamudhan, N. Premkumar, S. S. Yerrapragada, B. P.
ones [60]. Mani and K. Viswanathan, Powder Technot., 39 (1984) 93.
 9

19 M. Sugimoto, Funsai, 31 (1987) 40. 40 Y. Nakajima, W. J. Whiten and M. E. White, Trans. Inst.
20 K. Yamamoto, T. Tsukada and M. Sugimoto, J. Sot. Powder Min. MetalL Sect. C, 87 (1978) 194.
Technol. Jpn., 26 (1989) 12. 41 S. Endoh, J. Koga and K. Yamaguchi, .I. Sot. Powder Technol.
21 K. Yamamoto and M. Sugimoto, Proc. Spring Meet. Sot. Jpn., 20 (1983) 8.
Powder Technol. Jpn., Kyoto, May 30-31, 1989, pp. 18. 42 S. Endoh, J. Koga and K. Yamaguchi, Kona, 2 (1984) 7.
22 H. Ohya, S. Endoh, M. Yamamoto, H. Iwata and H. Nakamura, 43 S. Nakayama, T. Shibata and K. Yamaguchi, Proc. 25th Forum
Proc. Spring. Meet. Sot. Powder Technol. Jpn., Kyoto, 1991, on Powder, Sapporo, 24-26 Sept., 1987, pp. 188.
pp. 183. 44 N. B. Hsyung, J. K. Beddow and A. F. Vetter, Proc. 8th
23 M. Furuuchi and K. Gotoh, Proc. 15th Fine Particle Sot. Meet., Powder and Bulk Solids Con$, Rosemont, IL, 1983.
Orlando, FL, July 28-31, 1984, pp. 74. 45 K. W. Monts, J. K. Beddow and A. F. Vetter, Proc. 9th
24 M. Furuuchi, E. Ohno and K. Gotoh, Adv. Powder Technol., Powder and Bulk Soli& Con$, Rosemont, IL, 1984.
1 (1990) 101. 46 A. F. Vetter and P. A. Swanson, Part. Sci. Technol., 1 (1983)
25 B. Waldie, Powder Technol., 7 (1973) 244. 127.
26 M. Furuuchi, M. Nakagawa, M. Suzuki, M. Tsuyumine and 47 T. P. Meloy, N. N. Clark, T. E. Dumey and B. Pitchmani,
K. Gotoh, Powder Technol., 50 (1987) 137. Part. Sci. Technol., 2 (1984) 259.
27 M. Furuuchi and IL Gotoh, Powder Technol., 54 (1988) 31. 48 T. P. Meloy, N. N. Clark, T. E. Dumey and B. Pitchmani,
28 Ph. H. Kuenen, Developments in Sedimentology, vol. 1, Elsevier, Chem. Eng. Sci., 40 (1985) 1077.
Amsterdam, 1964, pp. 207. 49 N. N. Clark and T. P. Meloy, Powder Technol., 54 (1988)
29 E. Abe and H. Hirosue, J. Chem. Eng. Jpn., 15 (1982) 323. 271.
30 0. Taniguchi, M. Sakata, Y. Suzuki and Y. Osanai, Trans. 50 C. Orr, Powder Technol., 50 (1987) 217.
Sot. Mech. Eng. Jpn., 28 (1962) 485. 51 M. Furuuchi, C. Yamada and K. Gotoh, Proc. 2nd World
31 K. Shinohara, T. Maenami and T. Tanaka, Part. Sci. Technol., Cong. Part. Technol., Kyoto, Sept. 19-22, 1990, p. 151.
4 (1986) 161. 52 M. Furuuchi, C. Yamada and K. Gotoh, submitted to Powder
32 K. Ridgway and R. Rupp, J. Pharm. Pharmac., 21, suppl. Technol.
(1969)30. 53 S. Sano and S. Yashima, Proc. Spring Meet. Sot. Powder
33 K. Ridgway and J. B. Scotton, Sedimentology 20 (1973) 263. Technol. Jpn., Osaka, May 26-27, 1987, p. 1.
34 C. P. Lenn and C. B. Halt, in N. G. Stanley-Wood and T. 54 S. Sano and M. Nikaidoh, Proc. 2nd World Cong. Part, Technol.,
Allen (eds.), Particle Size Analysti 1981, Wiley Heydon, 1982, Kyoto, Sept. 19-22, 1990, p. 214.
pp. 233. 55 S. Sano, F. Saito and S. Yashima, Kagaku Kogaku Ronbunshu,
35 H. Iwata, H. Ohya, K. Masuda, M. Yamamoto and S. Endoh, 10 (1984) 17.
Proc. Spring Meet. Sot. Powder Technol. Jpn., Kyoto, May 56 S. Sano and S. Yashima, Proc. Spring Meet. Sot. Powder
28-29, 1991, pp. 178. Technol. Jpn., Osaka, May 26-27, 1987, p. 5.
36 W. H. Glezen and J. C. Ludwick, J. Sediment. Petrol., 33 57 H. W. King, C. 0. Wisler and J. G. Woodburn, Hydraulics,
(1963) 23. 5th edn., Krieger, Malabar, FL, 1980, p. 306.
37 G. Rittenhouse, Znd. Eng. Chem., 15 (1943) 153. 58 C. N. Davies, J. Aerosol Sci., 10 (1979) 477.
38 J. C. Ludwick and P. Henderson, Sedimentology, I1 (1968) 59 C. Murali, B. Pitchmani and N. N. Clark, Znt. J. Min. Process.,
197. 18 (1986) 237.
39 T. A. Roberts and J. K. Beddow, Powder Technol., 2 (19681 60 S. Endoh and H. Iwata, Proc. 24th Autumn Meet. Sot. Chem.
69) 121. Eng. Jpn., Nagoya, Oct. 16-18, 1991, p. 404.

View publication stats