Powder Technology 196 (2009) 163–168

Contents lists available at ScienceDirect

Powder Technology
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / p ow t e c

New measurement of particle size distribution by a buoyancy weighing-bar method

Eiji Obata, Yuichi Ohira ⁎, Mitsuhiro Ohta
Division of Applied Sciences, Graduate School of Engineering, Muroran Institute of Technology, Muroran, Hokkaido 050-8585, Japan

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

Article history: The rates which particles from JIS Test Powders 1, Class 2 (silica sand), Class 3 (silica sand), Class 16 (calcium
Received 14 January 2009 carbonate, heavy), and JIS Test Powder 2, Class GBM-20 (barium titanate glass) settled out of homogenous
Received in revised form 7 July 2009 suspensions were determined using the principles of the buoyancy weighing-bar method as well as a
Accepted 22 July 2009
sedimentation balance. Samples were standardized by Japanese Industrial Standard, and the dispersing agent
Available online 29 July 2009
was NaPP or NaHMP, while the viscosity improver was a starch syrup solution. Although constructing a
handmade sedimentation balance is difficult, developing a handmade weighing tool capable of measuring
Keywords:
Particle size distribution
particle size distribution is easy. Herein three weighing-bars, which were composed of aluminum, stainless
PSD steel, and copper, were used to measure the particle size distribution. However, an aluminum slit-cylinder
Buoyancy occasionally replaced the weighing-bars. The weighing data obtained via an analytical balance were con-
Buoyant mass trolled by connecting the balance to a personal computer with an RS-232C interface, and the determined
Sedimentation balance particle size agreed well with the data obtained by Andreasen analyses and the sedimentation balance.
© 2009 Elsevier B.V. All rights reserved.

1. Introduction The proposed method is accurate just as the Andreasen pipette method
and economical. Thus, it can be used in developing countries.
Various approaches have been used to measure particle size dis-
tribution [1]. For solid–liquid systems, it is important to measure par- 2. Theory-analogy between the fluidization curve and
ticle size distribution (PSD) by the Stokes diameter. However, the sedimentation curve
Andreasen pipette method [2], sedimentation balance method [3], cen-
trifugal sedimentation method [4], etc. can measure PSD in suspensions BWM is a sedimentation method capable of measuring particle
because these methods measure the migration velocities of particles size [17]. PSD can be classified as either a cumulative or frequency
in solution and then use Stokes formula to calculate the particle size. function [1]. The distribution obtained in this method is the same
The disadvantages of these methods are that they are time consuming cumulative function used in the Odén balance method [16] and
and require special skills. On the other hand, PSD can be analyzed using manometric method. Although sedimentation curves are not really
a different principle through micrographs [5], laser diffraction/scatter- relevant to fluidization curves, they are analogous to plots of mass vs.
ing methods [6], and Coulter counter method [7]; the drawbacks to this settling time and plots of pressure drop vs. superficial velocity. Fig. 1
type of method is that numerous samples must be used for accurate plots the pressure drop ΔP throughout the entire bed against the
PSD. Morever, although laser diffraction/scattering and Coulter counter superficial velocity u [8,9,15]. If the superficial velocity u is u3, then:
methods produce highly accurate results within a shorter time, the
equipment is extremely expensive. Therefore, a simple and cost-


ðρP −ρÞM0 g DðxÞ dΔP
effective new method to determine PSD is in high demand. ΔP = +u ≡0Y;
ρP A 100 du
Obata et al. have reported five measurements for PSD by fluid- ð1Þ
ization in the laminar to turbulent flow region [8–15], and one of


ðρP −ρÞM0 g DðxÞ dΔP
them can be obtained by a graphical measurement using a fluidization ≡0X; u ≡XY;
ρP A 100 du
curve, i.e. the superficial velocity vs. pressure drop is plotted [8,9].
The fluidization curve is constructed using the same algorithm as
the sedimentation (Odén) balance method [16]. We experimentally where M0, A, and D(x) are the total particle mass, cross-sectional area
investigate the applicability of the buoyancy weighing-bar method of the bed, and cumulative mass percentage undersize of particle size
(BWM) to measure PSD. Although it is difficult to construct a hand- x, respectively. Fig. 1 graphically depicts the PSD using the fluidization
made Odén balance, assembling a handmade weighing tool is easy. curve.
The volume of the submerged weighing tool is VB = Ah, where A
is the cross-sectional area of the weighing tool and h is the length
⁎ Corresponding author. Tel.: +81 143 46 5768; fax: +81 143 46 5701. of the submerged weighing tool in the suspension. The densities of
E-mail address: ohira@mmm.muroran-it.ac.jp (Y. Ohira). dispersion liquid containing the dispersant and particles are denoted

0032-5910/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.powtec.2009.07.015
164 E. Obata et al. / Powder Technology 196 (2009) 163–168

Fig. 1. Construction to determine particle size distribution by plot of pressure drop ΔP vs. superficial velocity u in fluidization.

as ρL, and ρP, respectively. The initial solid concentration of the sus- Stokes diameter. Eq. (11) describes the mass balance of particles in a
pension is C0 [kg-solid/m3-suspension]. suspension [1].
At the settling time t = 0, the initial density of the suspension ρS0
is:
x x vðxÞt
WBt = WB0 ∫xmax f ðxÞdx + WB0 ∫xmin f ðxÞdx; ð11Þ
C h
ρS0 = ρL + 0 ðρP −ρL Þ: ð2Þ
ρP
where v(x) is the sedimentation velocity of particle x and f(x) is the
Because the initial buoyant mass of the submerged weighing tool WB0 mass frequency of particle size x. The first term is the total mass of
depends on the particles in the suspension from the suspension particles from particle size x to xmax, while the second term is the
surface to depth h, WB0 can be defined as: mass of particles from particle size xmin to x. Differentiating Eq. (11)
with respect to the time t yields:
WB0 = VB ρS0 : ð3Þ

Then the apparent mass of the weighing tool in the initial suspension dWBt x vðxÞ
= WB0 ∫xmin f ðxÞdx ð12Þ
is: dt h

GB0 = VB ρB −WB0 = VB ðρB −ρS0 Þ: ð4Þ From Eqs. (11) and (12),

 
where ρB is the density of the weighting tool. dWBt
WBt = WRt + t; ð13Þ
At t = t, the density of the suspension, buoyant mass and apparent dt
mass of the submerged weighing tool in the suspension are:
where WRt is the mass of particles greater than particle size x,
C x
ρS = ρL + ðρ −ρL Þ: ð5Þ WB0 ∫xmax f ðxÞdx. The apparent mass of the submerged weighing tool
ρP P GBt is given by Eq. (7). Thus, the apparent mass gradually increases
from GB0 to GB∞, while the volume and the density of the submerged
WBt = VB ρS ; ð6Þ
weighing tool remain constant. Differentiating Eq. (7) with respect to
time t gives:
GBt = VB ρB −WBt = VB ðρB −ρS Þ: ð7Þ

At t = ∞, the solid concentration of the suspension is zero. The final dGBt dWBt
=− : ð14Þ
density of the suspension, the final buoyant mass, and the final dt dt
apparent mass of the submerged weighing tool in the suspension are
Combining Eqs. (7), (13), and (14) produces:
respectively:
   
ρS∞ = ρL ; ð8Þ dGBt dGBt
GBt = VB ρB −WRt + t = GRt + t; ð15Þ
dt dt
WB∞ = VB ρL ; ð9Þ
where GRt = VBρB−WRt. The cumulative mass percentage oversize is:
GB∞ = VB ρB −WB∞ = VB ðρB −ρL Þ: ð10Þ

In the sedimentation balance (Odén balance) theory [16], the mass x GRt −GB0
RðxÞ = 100∫xmax f ðxÞdx = × 100 = 100−DðxÞ: ð16Þ
frequency function of a given sample particle is f(x) where x is the GB∞ −GB0
 E. Obata et al. / Powder Technology 196 (2009) 163–168 165

Particle size x can be expressed by the following equation using

the Stokes formula:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
18μL vðxÞ
x= ; ð17Þ
gðρP −ρL Þ

where g is the gravitational acceleration and µL is the viscosity of

the dispersion liquid containing the dispersant. The sedimentation
velocity of the particles v(x) can be calculated by Eq. (18):

h
vðxÞ = ; ð18Þ
t

where h is the submerged length of the weighing tool and t is the

settling time. Particle size x given by Eq. (17) is the Stokes diameter.
Thus, the theory of the BWM is the similar to that for sedimentation
balance analysis [16].
Fig. 2 shows the PSD calculation method. The upper right graph
depicts the change in the apparent mass of weighing tool GB with
time, while the lower right graph shows the relationship between
time and the reciprocal of the particle size. From Eqs. (17) and (18),
time is proportional to the second power of the reciprocal of particle
size. Hence, in this method, particle size x can be calculated at time
t while GRt can simultaneously be calculated from tangent line based Fig. 3. Schematic diagram of experimental apparatus.
on Eq. (15). Mass percentage undersize D(x) can be calculated by
Eq. (16). Thus, the upper left graph illustrates the PSD from the
calculated particle size x and D(x).
submerged volume: 9.74 × 104 mm3), and the weighing-bars (φ9.0 ×
3. Experiments 210.0, submerged length: 200.0, submerged volume: 1.27 × 104 mm3),
which were composed of aluminum (density: 2700 kg/m3), SUS304
3.1. Experimental apparatus (density: 7900 kg/m3), and copper (density: 8960 kg/m3). The ana-
lytical balance (GR-300: A&D Co., Ltd., minimum readout mass 0.1 mg)
Fig. 3 schematically illustrates this experiment. Fig. 4 shows had a below-balance-weighing hook for hanging measurement, and
photographs of the weighing and stirring tools. The weighing tools was equipped with an RS-232C interface. The measuring glass cylinder
were the aluminum slit-cylinder (O.D.: 39.8, I.D.: 30.0, cylinder length: was 1 l. A personal computer was connected to an analytical balance
210.0, slit depth: 5.0, slit length: 200.0, submerged length: 200.0, with RS-232C, and the data was collected in 5-second intervals.

Fig. 2. Method to determine the particle size distribution using the relationship between the settling time and the apparent mass of the weighing tool.
166 E. Obata et al. / Powder Technology 196 (2009) 163–168

4. Results and discussion

4.1. Spherical glass beads (JIS Test Powders 2, Class GBM-20)

Fig. 5 shows the change with time in the apparent mass of the slit-
cylinder GB when JIS Test Powder 2, Class GBM-20 was used. The
apparent mass of the slit-cylinder increased until all the particles
settled below the lower end of the slit-cylinder, and then the apparent
mass of the slit-cylinder became constant. The change in the apparent
mass was due to the change in the buoyant mass against the slit-
cylinder as well as particle settling.
Fig. 6 shows the PSDs of the JIS Test Powder 2, Class GBM-20.
Additionally, the figure shows PSDs from commercial measuring
instruments, which included the Andreasen pipette method, Coulter
counter method, laser diffraction/scattering method, and micrograph
method. The PSD of the glass beads measured by BWM agreed well
with those of the Andreasen pipette method and micrograph method.
Fig. 4. Weighing-bars, slit-cylinder and stirring tool. Likewise, PSDs measured by the Coulter counter and laser diffraction/
scattering methods gave results similar to that of BWM. Hence, BWM
can measure the PSD of a spherical particle.
3.2. Sample particles

Table 1 shows the properties of the samples. The sample materials

were JIS Test Powders 1, Class 2 (silica sand), Class 3 (silica sand), and
Class 16 (calcium carbonate, heavy) and JIS Test Powder 2 (barium
titanate glass and Class GBM-20), which were standardized by Japanese
Industrial Standards. The PSDs were measured using the Andreasen
pipette method, laser diffraction/scattering method (MT3300EX, Nikkiso
Co., Ltd.), and micrograph method (SZX7, Olympus). Additionally, the
PSDs from the sedimentation balance method (JIS Test Powders 1, Class
2, Class 3, and Class 16) and the Coulter counter method (JIS Test Powder
2, Class GBM-20) were obtained by Japan Industrial Standards.

3.3. Experimental conditions and methods

All suspensions had a solid concentration of 10 kg/m3 (ca. 1 wt.%).

NaPP (sodium pyrophosphate decahydrate, Wako Pure Chemical In-
dustries, Ltd.) for silica sand and NaHMP (sodium hexametaphosphate,
Kanto Chemical Co., Inc.) for calcium carbonate were used as dis-
persants. A solution of starch syrup was used as a viscosity improver
to measure large sized particle. Both room temperature and liquid
Fig. 5. Time course of apparent mass of weighing cylinder in the barium titanate-glass
temperature were approximately 298 K. Although each suspension was
beads (JIS Test Powders 2, Class GBM-20) suspension.
initially stirred for 10 min, the measuring time was 2 h. In this work, all
of the particles were in the laminar flow region.
To prepare a suspension, 1000 ml liquid and the particles to be
tested were mixed in a glass cylinder. Using a hanging wire, which did
not extend due to the weight of the weighing tool, a weighing tool was
hung from the analytical balance. After thoroughly stirring the sus-
pension, the weighing tool was set in suspension, and this was re-
corded as t = 0. The measuring data, which consisted of time t and the
corresponding apparent mass of the submerged weighing tool GB,
were recorded on a personal computer. After the measurements, we
calculated the PSD of the tested particles based on the aforementioned
theory.

Table 1
Properties of sample particles (granted by Japan Industrial Standard).

Sample particles Median size Particle density

[m] [kg/m3]

Barium titanate glass beads 21–23 × 10− 6 4.0–4.2 × 103

(JIS Test Powders 2, GBM-20)
Silica sand (JIS Test Powders 1, Class 2) 27–31 × 10− 6 2.6–2.7 × 103
Silica sand (JIS Test Powders 1, Class 3) 6.6–8.6 × 10− 6 2.6–2.7 × 103
Calcium carbonate, heavy 3.6–4.6 × 10− 6 2.7–2.8 × 103
Fig. 6. Particle size distributions of the barium titanate-glass beads, JIS Test Powders 2,
(JIS Test Powders 1, Class 16)
Class GBM-20.
 E. Obata et al. / Powder Technology 196 (2009) 163–168
4.2. Optimum concentration of dispersants
Figs. 7 and 8 show the PSDs of the JIS Test Powder 1, Class 3 (silica
sand) and JIS Test Powder 1, Class 16 (calcium carbonate, heavy),
where the filled dots denote measurements by Japan Industrial
Standards using the sedimentation balance method and the solid line
indicates the PSD of the laser diffraction/scattering method. The ex-
perimental results agreed with the calculation. Calculations using the
Stokes formula indicate that particle sizes up to 5 µm can be measured
in 2 h. In Fig. 7, the optimum concentration of the dispersant (NaPP) in
the JIS Test Powder 1, Class 3 (silica sand) suspension was 0.1 wt.%. On
the other hand, in Fig. 8, the optimum concentration of the dispersant
(NaHMP) in the JIS Test Powder 1, Class 16 (calcium carbonate, heavy)
suspension was 0.3 wt.%. The PSD of the JIS Test Powder 1, Class 16
(calcium carbonate, heavy) with the optimum concentration of the
dispersants measured by the BWM agreed with those from both the
sedimentation balance method and laser diffraction/scattering meth-
od. Hence, BWM can measure PSDs of non-spherical particles.
4.3. Shape and density of the weighing tool
The buoyant mass exerted on the slit-cylinder was about eight-
fold the weight exerted on the weighing-bar because the submerged
volumes of the slit-cylinder and the aluminum bar were 9.74 ×
104 mm3 and 1.27 × 104 mm3, respectively. As shown in Fig. 8, the
PSD with the weighing-bar was similar to that with the slit-cylinder.
Fig. 7. Particle size distributions of silica sand, JIS Test Powders 1, Class 3.
Fig. 8. Particle size distributions of calcium carbonate, heavy, JIS Test Powders 1, Class 16.
167
Additionally, the shape of the weighing tool did not affect the results.
From Eqs. (4) and (10), the difference of the apparent mass of the
weighing tool can be calculated by the following formula:
VB C0
GB∞ −GB0 = ðρP −ρL Þ ð19Þ
ρP
The minimum readout mass of the analytical balance was constant.
When the value of the apparent mass difference GB∞−GB0 increased,
the accuracy of BWM improved. Hence, the accuracy of BWM should
remain unchanged if C0 is reduced to one-eighth the original value
with an eight-fold increase in the submerged volume VB.
The effect of weighing-bar density on PSD was investigated ex-
perimentally. Fig. 9 shows the PSDs using a bar composed of alu-
minum, copper, and SUS304. The material of the weighing-bar did not
influence the PSD measurement.
4.4. Viscosity improver solutions for a large sized particle
BWM has a theoretical upper limit for particle size. Measuring
particles greater than 50 µm in water is difficult because the Reynolds
number of a large particle is above the laminar flow region. Thus, to
measure large particles, the sedimentation velocity based on Stokes
formula must be reduced. Hence, a starch syrup, which is a Newtonian
fluid, was used to measure large particles. Fig. 10 shows the PSDs of
Fig. 9. Effect of the weighing-bar density on particle size distributions of calcium
carbonate, heavy, JIS Test Powders 1, Class 16.
Fig. 10. Particle size distributions of silica sand, JIS Test Powders 1, Class 2.
168 E. Obata et al. / Powder Technology 196 (2009) 163–168

the JIS Test Powder 1, Class 2 (silica sand) in starch syrup solutions Acknowledgements
where the filled dots denote measurements by Japan Industrial
Standards using the sedimentation balance method, while the solid The authors would like to thank our students for their experi-
line indicates the PSD of the laser diffraction/scattering method. Upon mental assistance: T. Kashiwagi, D. Iwamoto, M. Shigihara, H. Ohira,
the addition of the starch solution, the viscosity of the dispersing K. Tsuchiya and H. Mori.
liquid was approximately 4 mPa s. The PSD of the large sized particles
measured by the BWM was consistent with those of the sedimenta- References
tion balance method and laser diffraction/scattering method. Thus,
[1] T. Allen, Particle Size Measurement, Fourth edition, Chapman and Hall, London,
BWM can measure the PSD of large particles. Because the liquid 1990, pp. 345–355.
viscosity was ca. four-fold, the sedimentation velocities of particles [2] Society of Chemical Engineering of Japan, Chemical Engineering Handbook, 5th
were ca. one-half, and particles of this size could be measured in 2 h. edition, Maruzen, Tokyo, Japan, 1988, pp. 224–231.
[3] K. Fukui, H. Yoshida, M. Shiba, Y. Tokunaga, Investigation about data reduction and
sedimentation distance of sedimentation balance method, Journal of Chemical
5. Conclusion Engineering of Japan 33 (2000) 393–399.
[4] M. Arakawa, G. Shimomura, A. Imamura, N. Yazawa, T. Yokoyama, N. Kaya, A New
apparatus for measuring particle size distribution based on centrifugal sedimen-
We experimentally investigated the applicability of BWM to mea- tation, Journal of the Society of Materials Science of Japan 33 (1984) 1141–1145.
sure PSD. This method is easy, useful, and economical. Regardless of [5] M. Kuriyama, H. Tokanai, E. Harada, Maximum stable drop size of pseudoplastic
the shape of the weighing tool, this method has accuracy equivalent dispersed-phase in agitation dispersion, Kagaku Kogaku Ronbunshu 26 (2000)
745–748.
to that of the sedimentation balance method or Andreasen's pipette
[6] H. Minoshima, K. Matsushima, K. Shinohara, Experimental study on size
method. Hence, a PSD up to 5 µm can be measured in 2 h. The PSD of distribution of granules prepared by spray drying: the case of a dispersed slurry
large sized particles can be measured in a viscous liquid using BWM. containing binder, Kagaku Kogaku Ronbunshu 31 (2005) 102–107.
[7] Y. Ohira, H. Takahashi, M. Takahashi, K. Ando, Wall heat transfer in a double-tube
coal-slurry bubble column, Kagaku Kogaku Ronbunshu 30 (2004) 360–367.
Nomenclature   [8] E. Obata, H. Watanabe, N. Endo, Measurement of size and size distribution of particles
α constant = hðρ18μ L
P −ρL Þg
, m2·s by fluidization, Journal of Chemical Engineering of Japan 15 (1982) 23–28.
A cross-sectional area of weighing tool or fluidization bed, m2 [9] E. Obata, H. Watanabe, Measurement of particle sizes by fluidization, Encyclopedia
of Fluid Mechanics, vol. 4, Gulf Publishing, Houston, 1986, pp. 221–236.
C solid concentration of suspension, kg/m3 [10] E. Obata, H. Watanabe, K. Mukaida, M. Akiyoshi, K. Ando, Measurement of particle size
D(x) mass percentage undersize of particle size x, % distribution by numerical values of fluidization curve, Kagaku Kougaku Ronbunshu 12
f(x) mass frequency of the particle size x, − (1986) 619–621.
[11] E. Obata, H. Takahashi, M. Akiyoshi, K. Ando, H. Watanabe, Measurement of
g gravitational acceleration, m/s2 particle size distribution by superficial velocity and pressure drops across liquid–
GBt mass of weighing tool at t = t, kg fluidized beds, Kagaku Kougaku Ronbunshu 14 (1988) 103–106.
GRt VBρB−WRt, kg [12] E. Obata, H. Takahashi, M. Akiyoshi, K. Ando, The measurement of the size
distribution of irregular particles by liquid–solid fluidization in a high Reynolds
h submerged length of weighing tool, m
number, Journal of the Society of Powder Technology 27 (1990) 301–307.
R(x) mass percentage oversize of particle size x, % [13] N.G. Stanly-Wood, E. Obata, H. Takahashi, K. Ando, Liquid fluidisation curves,
t time, s Powder Technology 60 (1990) 61–70.
[14] E. Obata, M. Maruyama, Y. Ohira, M. Akiyoshi, K. Ando, The simultaneous mea-
u superficial velocity, m/s
surement of crystal sizes and mass in a fluidized crystallizer, Journal of the Society of
v(x) sedimentation velocity of particle size x, m/s Powder Technology 33 (1996) 456–461.
VB submerged volume of weighing tool, m3 [15] E. Obata, K. Ando, Particle size measurements by fluidization: From laminar flow
WB buoyant mass of the submerged weighing tool in the region to the turbulent flow region, Encyclopedia of Fluid Mechanics, Supplement
2, Gulf Publishing, Houston, 1993, pp. 169–189.
suspension, kg [16] S. Odén, The size distribution of particles in soils and the experimental methods of
WRt WB0 ∫xxmax f ðxÞdx, kg obtaining them, Soil Science 19 (1925) 1–35.
x particle size, m [17] E. Obata, Y. Ohira, M. Ohta, New measurement of particle size distribution by
buoyancy weighing method, Proceedings of the 5th World Congress on Particle
µL liquid viscosity, Pa · s Technology, Yuichi Ohira, Florida, USA, CD-ROM, 2006.
ρL liquid density, kg/m3
ρB density of weighing tool, kg/m3
ρP particle density, kg/m3
ρS density of suspension, kg/m3

Subscripts
max maximum
min minimum
0 initial t = 0
∞ infinity t = ∞