Engineering MECHANICS, Vol. 20, 2013, No. 3/4, p.

309–318 309

SIZE ANALYSIS OF SOLID PARTICLES USING LASER

DIFFRACTION AND SIEVE ANALYSIS

Michaela Hrnčı́řová*, Jiřı́ Pospı́šil*, Michal Špiláček*

The article compares two methods of determination of the particle size distribution of
fine ash. The respective methods compared are the sieve analysis and laser diffraction.
The former is one of the oldest methods in use, applicable to size separation of large
grain size particles. The latter is a modern advanced method that employs laser
diffraction to determine the size distribution. For the measurement itself, the sieve
analyser Analysette 3 PRO and the universal laser diffraction device Analysette 22
MicroTec plus, both from the Fritsch company, were applied. The comparison was
made by analysing fine ash particles from biomass burners. The particles were in the
size range of 0.08 μm to 2000 μm.
Keywords : solid particles, size distribution, laser diffraction, sieve analysis

1. Introduction
The particle size determination is a parameter of great importance in materials science,
medical science, biology, energetics and other science branches. By the size of a particle, its
linear dimension, i.e. length, is understood. The size can be unequivocally determined only
for a spherical particle by measuring the particle length that corresponds to its diameter.
For non-spherical particles in order for them to be comparable with the spherical ones,
there is a need to transform them into spherical ones of a corresponding diameter. The
equivalent diameters are the most common transformations. These define the diameters
of equivalent spheres that characterise the original non-spherical particles correctly from
a physical point of view, depending on further utilization of the determined diameters (the
maximum particle dimension, equivalent aerodynamic diameter etc.) The sieve analysis
compares the diameter of a circle circumscribed about a non-spherical particle with the
equivalent sieve diameter, Dsieve , which corresponds to a maximum diameter of a sphere
that passes through a particular size of a sieve mesh, the sieve mesh being either square
or round. The equivalent diameter, DL , determined by laser diffraction is a diameter of
a sphere that would result in the same electronic response to an optical signal in a detector
– diffraction pattern – as if it were the case with the non-spherical particle and the same
detector. Assuming the Fraunhofer approximation, the diameter should correspond to an
equivalent plane diameter of a particle. The equivalent plane diameter, DP , is a diameter
of a sphere or circle of the same projected area as the non-spherical particle; the diameter
as such depends on the orientation of the particle especially when anisometric.

* Ing. M. Hrnčı́řová, doc. Ing. J. Pospı́šil, Ph.D., Ing. M. Špiláček, Brno University of Technology, Faculty of
Mechanical Engineering, Energy Institute, Technická 2896/2, Brno
310 Hrnčı́řová M. et al.: Size Analysis of Solid Particles Using Laser Diffraction ...

2. Methods of particle size determination

Apart from laser diffraction and sieve analysis, there are other methods for the size
distribution assessment of solid particles such as sedimentation method, microscopic image
analysis etc.
The sedimentation method is one of the traditional methods of particle size analysis
applicable to particles 0.02–200 μm. The method determines the particle size indirectly
from its sedimentation rate when exposed to gravity or centrifugal force in either liquid or
gaseous environment.
The microscopic image analysis of picture analysis is to transform the amount of visual
information contained in the picture into simple interpretable quantitative information in
form of simple charts or even only few values. With optical microscopy only sizes bigger
than few μm can be reliably measured due to the diffraction occurring at the edges of small
particles.
The sieve analysis is based on utilization of a set of sieves of a known mesh size. It is
mainly used for separating the coarse particles from the fine ones. One of the advantages of
this method is the known size range of particles captured on a particular sieve. The main
disadvantages are its time demanding character and its damaging effect on fragile particles
that affects granularity. During the sieving process, these particles tend to reduce their size
dramatically due to abrasion and collisions which leads to inaccurate results that do not
reflect the real size distribution of particles in the measured sample.
Laser diffraction is the present most widespread method used for the particle size mea-
surement. The physical principle has been known since the beginning of the 20th century,
but it was the invention of suitable laser devices and computers that made it advanced
enough. The method supersedes the others in its flexibility and speed nowadays.

3. Description of the experimental apparatus

The sieve analysis uses a set of sieves placed on top of each other, each with a different
mesh size that decreases in the direction of the gravitational gradient – sieve stack. For the
analysis itself, standardized round sieves with a rectangular mesh made of metal fibres are
used. These sieves are usually used for analysing powder samples with particle size ranging
from 40 μm to 4 mm. In our case, the dry method of sieving employed ANALYSETTE 3 PRO
(Fig. 1) measuring device. It is a standard piece of equipment in which the sieves vibrate
noticeably to the naked eye. The device makes use of a cradle to which a stack of sieves is
mounted. The cradle is connected to an actuator that controls the amplitude of vibrations
and sieving time.
When the sieving process ends, each sieve retains a fraction of the original sample. Each
fraction contains particles with the size range that matches the mesh size of the current and
previous sieve. These fractions are then weighted, resulting in mass fractions of particular
size ranges, which are considered a great advantage of the method over the other ones.
ANALYSETTE 22 MicroTec plus (Fig. 2) is a universal analysing device for particle
measurement of suspensions, emulsions and solid matters that utilizes laser diffraction. The
device is composed of a central measuring unit and a dispersion unit. There are two semi-
conductor lasers, each with the output of 7 mW and wavelength of 532 nm and 940 nm
respectively, located in the central measuring unit. The measuring range is from 0.08 μm
Engineering MECHANICS 311

Fig.1: Sieving analyzer Fig.2: Laser analyser

Fig.3: Concept of the laser beam shielding

up to 2000 μm. An ultrasonic water bath comprises the wet dispersion unit and can deliver
vibrations at the power output of 50 W of and frequency of 40 kHz to the liquid; organic
liquids or inorganic saturated salt solutions may replace water for short term measurements
without any risk of damage to the device. The entire device is governed by MaScontrol
software.
The laser diffraction method makes use of a laser light which is scattered by a suspension
particle, when passing through. Depending on the size and optical properties of the particles,
the light scattered at particular spatial angle. A Fourier lens concentrates the scattered light
into the focal plane of a detector array which measures the light intensity distribution due
to the light scattering. Based on the light distribution obtained, the size distribution of
particles is calculated.
The concentration of particles in the sample inserted into the device must be sufficiently
small to prevent too much scattering resulting in opacity. On the other hand, too few
particles in a sample cause to be too weak the optical signal. The optimum optical thick-
ness lies between 10 and 15 % for the wet dispersion unit used. The unit commences the
measurement process when the sample concentration is large enough. Fig. 3 illustrates the
shielding effect of a sample on a laser beam. There is no sample inserted into the device
in the upper figure thus 100 % light intensity is captured by the photo diode. The lower
figure depicts what happens to the laser beam when a sample is inserted into the device.
For optimum evaluation of the particle size distribution, both Mie solution to Maxwell’s
equations (small particles) and Fraunhofer approximation (particles of a size much larger
than the light wavelength) are applied.
312 Hrnčı́řová M. et al.: Size Analysis of Solid Particles Using Laser Diffraction ...

4. Experiments and result

Two types of ash were examined both produced by biomass furnaces. The first one (P1)
came from Dakon Damat Pyro 20G furnace, which was fuelled with spruce and pine logs.
The ash was taken from the furnace ash pan. The second one (P2) resulted from combustion
of walnut wood chips in Slokov Variant SL 22D furnace. Further experiments were conducted
on the ash collected from the bed of a fluidized-bed boiler after the burn out and cooling
down. These samples chiefly consisted of an inert material – sand.

4.1. Assessment of ash samples

Sieving analyser
First, both samples were sieved in order to remove coarse particles. Multiple sieving
courses were performed with gradually increased amplitude of vibrations and sieving times.
The ash sample P1 underwent the sieving procedure three times (subsamples 1, 2 and 3)
unlike the ash sample P2, which was processed twice (subsamples 4, 5). Nine sieves with
different mesh sizes were used for the analysis. Meshes with the size of 2 mm and 1 mm were
used to remove the coarsest particles. After this, the samples went through the rest of the
sieve stack with sieve sizes of 500 μm, 200 μm, 160 μm, 125 μm, 90 μm, 63 μm and 45 μm.
The masses retained on sieves below 500 μm (included) were then weighted and used for
a detailed analysis employing the laser analyser.
Tables 1 and 2 indicate the mass fractions as captured on a particular sieve. The charts
in figures 4 and 5 show the ash size distributions for samples P1 and P2 which correspond
to the data stated in tables 1 and 2.
It is apparent from the charts that repeated sieving at gradually increased amplitude,
interval and overall sieving time cumulatively increases the weight for sieves with smaller
mesh. The presumption is that the particles undergone abrasion which is causing breakup
into bigger amount of smaller particles. Input values for each measurement are listed in
table 3.
x [μm] 500 200 160 125 90 63 45
P1, sample 1 m [g] 7.9844 15.2033 17.7154 18.3719 18.6402 18.7095 18.7440
P1, sample 2 m [g] 2.1791 6.1464 8.1324 10.3205 13.0092 16.0908 18.0090
P1, sample 3 m [g] 1.6208 7.3865 9.0455 12.7262 15.2944 17.1447 18.1217
Tab.1: Cumulative mass of sample P1 (sieve analyser)

x [μm] 500 200 160 125 90 63 45

P2, sample 4 m [g] 4.3596 8.2407 9.7828 12.6219 14.0381 14.7743 15.0855
P2, sample 5 m [g] 3.5827 7.4638 10.2013 14.6475 16.0726 16.9372 17.3228
Tab.2: Cumulative mass of sample P2 (sieve analyser)

unit P1 sam. 1 P1 sam. 2 P1 sam. 3 P2 sam. 4 P2 sam. 5

Initial weight [g] 20.002 20.009 20.001 30.005 30.003
Period time [min] 1 3 5 5 5
Total time [min] 1 3 5 1 5
Amplitude [–] 0.2 0.6 0.8 0.5 0.5
Tab.3: Input values for the sieve analysis
Engineering MECHANICS 313

Fig.4: Size distribution of ash sample P1 Fig.5: Size distribution of ash sample P2

Laser analyser
The laser analyser used displays the measurement results in the form of data sheets
containing relative frequencies of the particular size range of particles. All the data acquired
are illustrated diagrammatically in figures 6 and 7.

Fig.6: Ash P1, sample 2, 45 μm Fig.7: Ash P1, sample 1, 200 μm

At figures 6 and 7 are the results of laser diffraction analysis. Specifically the figure 6
contains information about specimen that was retrieved from a sieve with 45 μm mesh.
According to the setup of the sieve analyser this specimen should contain particles the same
or bigger than 45 μm up to 63 μm that is mesh size placed above the 45 μm mesh sieve. From
the graph it is clear that the specimen contains a big amount of particles smaller than 45 μm.
Logical assumption is that the small particles agglomerate and then remains on sieves with
bigger mesh. This leads to distortion and inaccurate results. As was mentioned above
with more intensive sieving leads into a better particle separation. However, a mechanical
abrasion occurs too and the particles break up leading into distortion as well. With the sieve
analysis, each specimen must be approached individually. Unlikely, when wet dispersion is
used the agglomerates in the specimen are separated by ultrasound and all and each ash
particles are detected. The resulting graph shows that the specimen contains a small amount
of particles bigger than 63 μm. These particles get in the specimen due to their shape that
allowed them to fall through the mesh.
Similar result is shown in Fig. 7 which represents fraction retained on the sieve of the
mesh size of 200 μm. The only difference is there were no particles greater than 200 μm left
314 Hrnčı́řová M. et al.: Size Analysis of Solid Particles Using Laser Diffraction ...

on the sieve found by the laser analyser. All the particles as identified by the laser analyser
were smaller than the sieve size. Similar results were obtained for most of the ash samples
analysed.
In table 4 are listed percentage results from the measurement from laser diffraction using
values when after each executed analysis existed particles bigger than the respective mesh
(see figure 6). Value A represents number portion of particles detected as smaller than
respective mesh from which the sample was taken. Opposite is the value B shows the
number portion of bigger particles. Both values were acquired from graph belonging to each
sample. All acquired data were transformed in weight portions (C, D [%]) with premise
that all particles have spherical shape with density of 1200 kg/m3 . At figures 8 and 9 are
graphs that show dependence of number (B [%]) and weight portion (D [%]) to respective
mesh size.

Tab.4: Non-zero relative frequencies and mass fractions of samples assessed

Fig.8: Relative frequency of particles greater than Fig.9: Mass fractions of particles greater than
the mesh size of a corresponding sieve the mesh size of a corresponding sieve

The ash samples analysed show there are large amounts of fine particles in the individual
samples. By converting the frequencies into mass fractions, the dominant effect of relatively
small number of large particles on the results obtained by sieving becomes apparent. The ash
samples analysed exhibit significant cohesion of ash particles which leads to over estimation
of the ash mass fractions as obtained from the sieves of large mesh sizes. Further test proved
inefficient in removing the over estimation drawback despite repetitive screening and/or more
intense shaking of the sieve stack.
With respect to the ash particles cohesion, it seems that a high-quality dispersion of
the sample in the liquid bath, which prevents particles from agglomerating although not
Engineering MECHANICS 315

from disintegration of the fragile ones, is the key factor for laser diffraction when compared
with the sieve analysis. The accuracy of laser diffraction as well as the device was validated
against comparative measurement performed with the use of calibrated particles supplied
by Fritsch.

Fig.10: Ash P1, sample 2, 200 μm Fig.11: Ash P2, sample 5, 200 μm

Fig.12: Ash P2, sample 5, 125 μm Fig.13: Ash P2, sample 5, 125 μm

Fig. 10 shows remnants of a sample on the sieve of the mesh size of 200 μm after sieving.
A naked eye inspection identifies two distinct particle sizes there. The larger ones are
supposed to have their size greater or equal to the sieve size of 200 μm. The much smaller
ones just did not drop through the sieve. There is an ash sample of the size of 200 μm
in Fig. 11 as seen through a microscope. The microscopic view revealed the presence of
small particles. It displays them the same way as Fig. 13 does, i.e. sprinkle over a grid of
lines 0.05 mm thick and 1 mm apart. Particle agglomerates are also noticeable in Fig. 12
which captures finer fraction of the ash sample. Because of the particles cohesion, the dry
dispersion is not perfect which influences results of the sieve analysis. Fig. 13 shows particle
agglomerates at a magnification of 100×.

4.2. Inert material experiment

Based on the above mentioned results, it is clear the greatest trouble of ash particles
analysis by means of sieving lies in poor dispersion, which causes the fine particles to stick
the large ones and/or to agglomerate. This chapter brings up a comparison with an analysis
of fine particles which are assumed not to cluster or disintegrate due to mechanical abrasion.
The sample particles originate from a fluidized- bed boiler collected after the fuel had burnt
out and cooled down. The major part of the sample resembles an inert material – sand.
316 Hrnčı́řová M. et al.: Size Analysis of Solid Particles Using Laser Diffraction ...

Table 5 states the frequencies and mass fractions as obtained from the laser analysis of
a ‘sand’ sample, which experienced the same sieving procedure with the use of the same
sieve stack as described in the above chapters. All experiments including this sample were
carried out under the same conditions as was the ash sample.
The assessment of the particle size distribution of the sample with the use of laser diffrac-
tion identified some amount of particles greater than the corresponding mesh size in all cases.
Furthermore, the frequency analysis of particles greater than the mesh size indicated there
was 37.8 % of a large particle on average in the sample. If the sample is evaluated based on
the mass of individual fractions, we find out the particles greater than the mesh size carry
97.8 % of the sample mass on average. It can be said that the particles greater than the
mesh size that were captured represent the major mass of the sample; the mass associated
with the smaller particles is below 5 % of the sample.

Tab.5: Frequencies and mass fractions of sand sample

Fig.14: Sand, 500 μm Fig.15: Sand, 45 μm

The laser analysis results of sand are illustrated by Fig. 14 and 15. The former represents
the fraction retained on the sieve size of 500 μm. Notice that almost 70 % of the particles in
the sample were greater than the sieve size. Although the finer particles are more frequent
in Fig. 15, the major mass (95 %) is still carried by the particles greater than 45 μm.
A complete overview of the results is given in table 6 (all zero and non-zero frequencies
and mass fractions) for comparison. The data were assessed using the same procedure as
it was the case with ash sample P2, subsample 5 (see chapter 4.1). When looking at the
data in table 6, it is clear the laser diffraction identified fraction of particles greater than
the respective mesh size is both in frequencies and mass fractions significantly smaller for
the ash sample P2, subsample 5, than for a sand sample. Moreover, particles greater than
the mesh size were not identified in many cases at all.
Engineering MECHANICS 317

Tab.6: Frequencies and mass fractions of P2 samples, subsample 5

Fig.16: Mesh size effect on agglomeration of ash Fig.17: Mesh size effect on agglomeration of ash
P2, subsample 5, and ‘sand’, particle P2, subsample 5, and ‘sand’, particle
size distribution against frequency size distribution against mass fraction

Fig. 16 and 17 express graphically the obtained frequencies and mass fractions of solid
particles which size exceeds the mesh size of a respective sieve as identified by laser diffraction
for both the ash and sand samples. By comparing the data, it is evident the mass fractions
of particles acquired for a sand sample are greater than the respective sieve size is virtually
independent of the sieve size. The frequencies of such particles tend to decrease along with
the decrease of the mesh size. The ash sample shows an entirely irregular course both in
frequency and mass fraction data, containing even zero frequencies and fractions; therefore
it is impossible to comment on the data correctly.

5. Conclusion
The sieve analysis is a commonly used method of size distribution assessment of solid
residues of combustion processes both for its low financial and temporal cost. Its correct
application is closely related to the origin of solid residues. The fine fraction of combustion
products comprise incombustible fuel residues (ash) and small fractions of inert substances
especially used in fluidized-bed boilers to create a stable fluidized-bed (sand), and/or ap-
plied to absorb selected pollutants (limestone). Hence it creates a complex set of particles,
containing all types of particles of specific proportions. When using the sieve analysis it is
good to know the origin and compositions of the tested sample.
The experiments performed rendered the sieve analysis to be of high accuracy (97 %)
when assessing gravimetrically the size distribution of the inert mineral material (sand) of
318 Hrnčı́řová M. et al.: Size Analysis of Solid Particles Using Laser Diffraction ...

granularity ranging from 45 μm to 500 μm. On the contrary, the method proved unsuitable
for fine fractions of ash coming from organic fuel combustion (biomass). Such fine ash
particles tend to agglomerate, stick together and break up significantly while being sieved.
For that reason, the results of the sieve analysis are significantly inaccurate as particular
values depend on specific ash properties. For the ash samples the inaccuracy of measurement
using sieve analysis nears 100 %.
Usage of modern method for determination of size distribution of solid particles based
on laser diffraction seems like a more appropriate for analysis of ash particles than the sieve
analysis. Nevertheless, also with this method mechanical degradation of sample occurs when
the particles are being dispersed by the ultrasound in water bath. Fine ash particles partially
break up and this process is significantly supported by the water environment in which the
particles must by dispersed. The most appropriate approach is to use a combination of both
methods.

Acknowledgement
The present work has been supported by European Regional Development Fund in the
framework of the research project NETME Centre under the Operational Programme Re-
search and Development for Innovation.

References
[1] Pabst W., Gregorová E.: Charakterizace částic a částicových soustav, Praha: Vysoká škola
chemicko technologicka, 2007, s. 1–22
[2] Firemnı́ podklady firmy Fritsch k přı́stroji ‘ANALYSETTE 22’ MicroTec plus
[3] Wanogho S., Gettinby G., Caddy B.: Particle size distribution analysis of soils using laser
diffraction, Forensic Science International, Vol. 33, 1987, pp. 117–128.
[4] Tiehm A., Herwig V., Neis U.: Particle size analysis for improved sedimentation and filtration
in waste water treatment, Water Science and Technology, Vol. 39, 1999, pp. 99–106.
[5] Jandačka P., Študentová S., Hlavač Libor M., Kvı́čala M., Mádr V., Hredzák S.: Velikost
povrchu partikulárnı́ch látek, Chemické Listy, Vol. 105, 2011, s. 146–155
[6] Nascimento C.A.O., Guardani R., Giulietii M.: Use of neural networks in the analysis of
particle size distributions by laser diffraction, Powder Technology, Vol. 90, 1997, pp. 89—94

Received in editor’s office : August 31, 2012

Approved for publishing : September 18, 2013