Adv. Radio Sci., 9, 27–30, 2011
www.adv-radio-sci.net/9/27/2011/
doi:10.5194/ars-9-27-2011
© Author(s) 2011. CC Attribution 3.0 License.
Advances in
Radio Science
Evaluation of a concept for density measurement of solid particle
flows in pneumatic conveying systems with microwaves (8–12 GHz)
C. Baer1 , T. Musch1 , M. Gerding1 , and M. Vogt2
1 Ruhr-Universität
2 Ruhr-Universität
Bochum, Institute for Electronic Circuits, 44780 Bochum, Germany
Bochum, High Frequency Engineering Research Group, 44780 Bochum, Germany
Abstract. This paper presents a novel density measuring
concept for gas/coal particle compositions in pneumatic conveying systems. The proposed monitoring system uses horn
antennas to perform complex electromagnetic transmittance
measurements through the cross section of the conveying
tube. The phase of the complex transmittance gives information about the effective permittivity, which is related to the
mean volume fraction of the coal. Electromagnetic field simulations have been performed for the evaluation of the concept and the performance of the designed setup. A test stand
for measurements on coal dust under reproducible conditions
and with well-defined particle concentrations has been developed and implemented. The test tube has a diameter of
200 mm and a length of 400 mm. Coal particles with a diameter between 20 µm and 100 µm have been dispersed by
injecting nitrogen gas inside the test tube. Complex transmission measurements are performed in the implemented setup
with a calibrated vector network analyzer within a frequency
range of 8–12 GHz. Results of the conceptual evaluation by
measurements with different concentrations of coal particles
are presented and discussed.
of the volume concentration are based on cut-off frequency
measurements (Conrads, 1996), determination of backscatter coefficients (Happel, 2006), different resonator properties
(Penirschke and Jakoby, 2010), and various optical, acoustical, and electro-statical concepts. In this contribution, concepts for the measurement of the volume concentration by
means of microwaves are discussed and evaluated. The
proposed concept is based on transmission measurements
through the cross-section of a conveying tube. The complex transmittance between the antennas is measured in order to determine electrical material properties of the gas/dust
composition, which are representative for the mean volume
fraction of the pulverized coal. The applied wide frequency
range (8–12 GHz) permits a robust and precise analysis.
2
Monitoring sytem
2.1
Concept
The most important parameter in the described application is
the mass flow ṁ of the coal particles, which is defined as:
∂m
= ρ · V̇
(1)
∂t
In Eq. (1), ρ is the mass density of the coal. The volume flow
V̇ of the coal dust is given as follows:
ṁ =
1
Introduction
Pneumatic conveying is of interest for the handling of particulate bulk materials like grain, pellets, etc., and pulverized
fuel like coal dust. Mass or volume flow rate are parameters to quantify the mass flow (Miller et al., 2000). Common
methods for the measurement of the particles velocity are
correlation and Doppler techniques. Besides the flow velocity, the mass density and volume concentration of the solid
particles inside the conveying tube are additional parameters to be assessed. Available techniques for the monitoring
Correspondence to: C. Baer
([email protected])
∂V
= v ·A·ζ
(2)
∂t
In Eq. (2), v is the mean particle flow velocity over the crosssection of the tube, A the cross-section of the tube and ζ
the mean volume fraction of the coal dust through the crosssection. By means of the subsequently described measurement concept, the mean volume fraction ζ is measured. Figure 1 schematically shows the setup of the implemented monitoring system. Two horn antennas are arranged opposite to
each other with the line-of-sight being oriented perpendicularly to the tube axis. By measuring the complex transmittance S21 (ω), a projection along the tube’s cross-section of
V̇ =
Published by Copernicus Publications on behalf of the URSI Landesausschuss in der Bundesrepublik Deutschland e.V.
tance S21 (ω), a projection along the tube’s cross-section of
the particles is obtained. With the high propagation speed of
the electromagnetic waves in relation to the particle flow velocity,
the particle density within the measuring tube can be
28
assumed as static.
horn antenna
port 1
coal particles
horn antenna
port 2
acrylic glass tube
Fig. 1: Schematic of setup with
with conveying
conveying tube:
tube: Measuremeasurement tube between two rectangular horn antennas.
In order to obtain a robust model, various material properthe particles
is obtained.
Withtothebehigh
propagation
speed of
ties
of the coal
particles have
taken
into consideration.
the electromagnetic
waves
in relation
to the
the particle
veBecause
the maximum
diameter
K of
particlesflow
is less
locity,
theµm,
particle
density within the measuring tube can be
than
100
it follows:
assumed as static.
(3)
K In
<<order
λminto obtain a robust model, various material properties
theiscoal
have to be taken
into consideraIn
(3), of
λmin
the particles
smallest wavelength
in the used
frequency
tion.
Because
the
maximum
diameter
K
of
thedescribed
particles as
is
band i.e. at 12 GHz. In addition, coal can be
less than 100 µm, it follows:
a dielectric medium with a relative permittivity ε r,coal in a
range
1984):
(3)
K <<(Ruhrkohlenhandbuch,
λmin
2.4λ ... is
2.8the smallest wavelength in the used fre(4)
εInr,coal
Eq. =(3),
min
quency
band i.e. σatof12
GHz.
In addition,
canapplicabe deThe
conductivity
coal
is negligibly
smallcoal
in this
scribed
as
a
dielectric
medium
with
a
relative
permittivity
tion in relation to the product of angular frequency and perεr,coal in aConsequently,
range (Ruhrkohlenhandbuch,
1984): can be demittivity.
the coal/gas composition
scribed
a quasi-homogeneous
dielectric medium described
εr,coal =as2.4
... 2.8
(4)
by a single effective permittivity ε r,eff which is directly reζ ofsmall
the coal.
on
lated
to the meanσvolume
The conductivity
of coal fraction
is negligibly
in thisBased
applica(Sihvola,
2000) to
thethe
effective
given as follows:
tion in relation
productpermittivity
of angular is
frequency
and permittivity. Consequently, the coal/gas composition can be de3 · ζ · εr,gas · (εr,coal − εr,gas )
asr,gas
a quasi-homogeneous
dielectric medium described
εscribed
+
(5)
r,eff = ε
εr,coalpermittivity
+ 2 · εr,gas −εζr,eff
· (εr,coal
)
by a single effective
which−isεr,gas
directly
related
to the mean
volume fraction
ζ of εthe
coal.
Based
on
By
measuring
the effective
permittivity
,
the
mean
volr,eff
Sihvola
(2000)
thethe
effective
given(5)asby
follows:
ζ of
coal canpermittivity
be obtainedisfrom
taking
ume
fraction
the known permittivities into consideration. Formula (5) is
3 · ζ · εr,gas · εr,coal − εr,gas
known
the+Maxwell-Garnett mixing formula forspherical
εr,eff = as
εr,gas
(5)
εr,coal
+ 2 · which
εr,gas −isζ a· good
εr,coalapproximation
− εr,gas
particles (Sihvola,
2000),
for
the
of the
behavior
of the coalε particles.
By description
measuring the
effective
permittivity
, the mean volr,eff
ume fraction ζ of the coal can be obtained from Eq. (5)
2.2 Measurement approach
by taking the known permittivities into consideration. Formula (5)
known as the
mixing
formula
Under
theisassumption,
thatMaxwell-Garnett
standing waves and
multiple
refor
spherical
particles
(Sihvola,
2000),
which
is
a
good
apflections in the system can be neglected, the phase φ(ω)
proximation for the description of the behavior of the coal
particles.
2.2
see Fig. 1, as follows:
ω
ds + φ0 (ω)
φ(ω) =
C. Baer et
al.:
Density measurement
of solid particle flows
s vph (s)
ω
= transmission
·
εr (s)path
ds +
φ0 (ω) the two ports, see Fig.(6)
along the
between
1,
c0 s
as follows:
ω √
= Z · εr,eff · D + φ0 (ω)
c0 ω
φ (ω) =
ds + φ0 (ω)
By integrating
s vph (s)the effective permittivity over the cross secZ q
tion a closed
for the phase relation can be derived.
ω expression
0
=
·
(6)
0 (ω)measuring tube, ω the anIn (6), Dc is the εdiameter
ofφthe
r (s) ds +
0
s
gular frequency
ω √ and c 0 the speed of light. The frequency= phase
· εr,eff
D +denotes
φ0 (ω) the wave propagation inside
dependent
φ 0·(ω)
c0
the antennas and feeding networks. By means of a reference
measurement
any particles
insideover
the the
measurement
By
integratingwithout
the effective
permittivity
cross secgiven
in (6)
can becan
determined:
tube,a the
phase
term φ 0 (ω)
tion
closed
expression
for the
phase
relation
be derived.
In Eq. (6), D is the diameter of the measuring tube, ω the
angular frequency
ω √ and c0 the speed of light. The frequencyφref (ω) = phaseεr,gas
· Ddenotes
+ φ0 (ω)the wave propagation inside
(7)
dependent
φ0 (ω)
c0
the antennas and feeding networks. By means of a reference
(ω) obtained
measurement
The measuredwithout
phase φany
meas
measurement
particles
insideinthea measurement
can
be
eliminated
for
φ
(ω)
as
follows,
by
use
of
(7):
tube, the phase term φ00 (ω) given in Eq. (6) can(6)beand
determined:
(8)
Δφ(ω) = ω
φmeas
√ (ω) − φref (ω)
φ0 (ω) (7)
φref (ω) = ω ε√
r,gas · D +√
= c0 · εr,eff − εr,gas · D
c0
The measured phase φmeas (ω) obtained in a measurement
√
Δφ(ω)
is a linear function of εEqs., and
canThe
be phase
eliminated
forinφ(8)
(6)
0 (ω) as follows, by use of r,eff
the mean
and
(7): volume fraction ζ can be assessed by means of (5).
Based on a linear regression fit of the function Δφ(ω), the
(8)
1φ
φref (ω)
(ω) = φmeas
measurement
is (ω)
very−robust.
ω √
√
=
εr,eff − εr,gas · D
· system
2.3 Prototype
c0
√
The
1φmeasurements
is coal
a linear
of εr,eff ,
(ω) in Eq. (8) on
A testphase
tube for
dustfunction
under reproducible
conditions
andvolume
well-defined
particle
hasmeans
been
and
the mean
fraction
ζ canconcentrations
be assessed by
developed
implemented.
particlesfitare
inof
Eq. (5). and
Based
on a linearCoal
regression
of dispersed
the function
side(ω),
the test
tube by injecting
nitrogen
gas.
1φ
the measurement
is very
robust.
Figure 2 shows a photo of the test stand consisting of an
2.3
Prototype
system
acrylic
glass tube
(I), two rectangular horn antennas (II), a
nitrogen gas inlet pipe (III), and a filtering exhaust (IV). The
A
testparticles
tube forare
measurements
on coal
under of
reproducible
coal
dispersed with
a gasdust
pressure
5 bar. The
conditions
and iswell-defined
concentrations
been
exhausted gas
cleaned in aparticle
gas scrubbing
systemhas
in order
developed
and implemented.
Coal particles
arehas
dispersed
to avoid contamination.
The whole
test stand
a heightinof
side
theand
testatube
by injecting
nitrogen
gas.
80 cm
diameter
of 20 cm.
The complex
transmittance
2 shows
photo of
test stand
of an
is Figure
measured
with aa Rohde
& the
Schwarz
ZVK consisting
two-port vector
acrylic
tube (I), two rectangular horn antennas (II), a
networkglass
analyzer.
nitrogen gas inlet pipe (III), and a filtering exhaust (IV). The
coal particles are dispersed with a gas pressure of 5 bar. The
3 Simulations
exemplary
exhausted
gas is and
cleaned
in a gas measurements
scrubbing system in order
to avoid contamination. The whole test stand has a height of
For the verification of the measurement concept, various sim80 cm and a diameter of 20 cm. The complex transmittance
ulations with a 3D-field simulator were performed.
is measured with a Rohde & Schwarz ZVK two-port vector
network analyzer.
Measurement approach
3
Under the assumption, that standing waves and multiple reflections in the system can be neglected, the phase φ (ω) of
the complex transmittance S21 (ω) is given as the line-integral
Adv. Radio Sci., 9, 27–30, 2011
Simulations and exemplary measurements
For the verification of the measurement concept, various simulations with a 3-D-field simulator were performed.
www.adv-radio-sci.net/9/27/2011/
C.
Baer
et
al.:
Density
measurement
of
C.
Baer
et
al.:
Density
Measurement
of solid
Solidparticle
Particleflows
Flows
C.
Solid
Particle
Flows
C. Baer
Baer et
et al.:
al.: Density
Density Measurement
Measurement of
29
333
0
0
II
IV
IV
dB
dB
-10
-10
-20
-20
-30
-30
-40
-40
(a) 3D simulation, magnitude of the E-field at a fremagnitude
of the
at aa frefre3D simulation,
(a) 3-D
simulation,magnitude
magnitudeof
the E-field
E-field at
quency of 10 GHz with a homogeneous relative perof 10
10 GHz
GHz with
with aa homogeneous
homogeneousrelative
relativeperperquency of
mittivity of εr,eff =1.
mittivity of εεr,eff
= 1.
r,eff
r,eff =1.
0
0
III
III
II
II
Fig.
2:
Photo
of
the
implemented
test
stand. I: Acrylic glass
Fig. 2:
2: Photo
Photo of
of the
the implemented
implemented test
test stand.
Fig.
I: Acrylic glass
cylinder,
II:
Rectangular
horn
antennas,
III: Nitrogen gas incylinder,II:
II: Rectangular
Rectangular horn
horn antennas,
antennas, III:
cylinder,
Nitrogen gas inlet
pipe,
IV:
Filtering
exhaust.
let pipe,
pipe, IV:
IV: Filtering
Filtering exhaust.
exhaust.
let
3.1
Simulations
3.1
Simulations
3.1 Simulations
Simulations
3.1
Fig.
shows
Figure
showssimulation
simulationresults
resultsfor
for two
two different
different setups
setups of
Fig. 3333shows
shows
simulation
results
for
two
different
setups
of
Fig.
simulation
results
for
two
different
setups
of
the
measuring
tube.
the
measuring
the
measuring
tube.
In
both
configurations,
the
measuring
tube.
In
both
configurations,
the
measuring
the measuring tube. In both configurations, the measuring
tube
is
placed
in
center,
antennas
are
artube
is
placed
in
center,
while
the
two
horn
tube is
is placed
placed in
in center,
center, while
while the
the two
two horn
horn antennas
antennas are
are arartube
ranged
side
wise,
permittivity
ranged
side
wise,
opposite
to
each
other.
The
ranged
side
wise,
opposite
to
each
other.
The
permittivity
ranged
side wise,
opposite to each other. The
permittivity
of
the
inner
volume
varied.
A
tranof
the
inner
volume
of
the
tube
has
been
of the
the inner
inner volume
volume of
of the
the tube
tube has
has been
been varied.
varied. A
A trantranof
sient
solver
was
used
While
sient
solver
was
used
for
the
following
simulations.
sient
solver
was
used
for
the
following
simulations.
While
sient
solver
was used for the was
following
simulations.
While
the
field
monitor’s
set
to
10
all
cases,
the
field
monitor’s
frequency
GHz
in
the field
field monitor’s
monitor’s frequency
frequency was
was set
set to
to 10
10 GHz
GHz in
in all
all cases,
cases,
the
=
1
and
=
4
have
relative
permittivities
relative
permittivities
ε
=
1
and
ε
=
4
have
been
used
r,eff
r,eff
r,eff
r,eff
=
1
and
ε
=
4
have
been
used
relative
permittivities
ε
r,eff
r,eff
εr,effrespectively.
= 4 have beencomused
relative
permittivities ε3a
r,eff = 1 and 3b,
for
the
3a
and
Fig.
3b,
respectively.
A
for
the
results
in
Fig.
for
the
results
in
Fig.
3a
and
Fig.
3b,
respectively.
A
comfor
the results
in Fig. 3a and Fig. 3b,
respectively.
A comparison
shows
in
Fig.
tighter
than
parison
shows
that
the
phase
fronts
3b
are
parisonshows
shows that
that the
the phase
phase fronts
fronts in
in Fig.
Fig. 3b
3b are
are tighter
tighter than
than
parison
the
phase
3a.
This
is
in
the
the
phase
fronts
in
Fig.
3a.
This
is
in
agreement
with
the
phase
fronts
in
Fig.
3a.
This
is
in
agreement
with
the
the
phase fronts
in
Fig. 3a. This is in agreement with
the
assumption
that
the
on
the
assumption
that
the
slope
of
the
phase
line
depends
assumption
that
the
slope
of
the
phase
line
depends
on
the
assumption
that
the
slope
of the
phaseshow
line that
depends on the
permittivity.
Fig.
3a
and
Fig.
3b
also
re3a and
permittivity.
Figures
permittivity. Fig.
Fig.
3a and
and
Fig.3b
3balso
also show
show that
that multiple
multiple rerepermittivity.
3a
Fig.
3b
also
show
that
multiple
flections,
effected
with
a
relative
flections,
effected
by
the
acrylic
glass
cylinder
flections,
effected
by
the
acrylic
glass
cylinder
with
a
relative
flections,
effected by=the
acrylic
glass cylinder
with a relative
3,
are
negligible
as
they
than
permittivity
permittivity
r,acryl==3,
3,are
arenegligible
negligible as
as they
they are
are more
more than
than
permittivity εεεεr,acryl
r,acryl = 3,
are
negligible
as
they
are
more
permittivity
r,acryl
40
dB
smaller
compared
to
the
maximum
40
dB
smaller
compared
to
the
maximum
field
strength.
40
dB
smaller
compared
to
the
maximum
field
strength.
40 Fig.
dB smaller
compared
to the maximum
field strength.
shows
Figure
showsthe
thesimulated
simulated phase
phase Δφ
1φ versus
frequency
Fig. 4444shows
shows
the
simulated
phase
Δφ
versus frequency
frequency
Fig.
the
simulated
phase
Δφ
versus
=
1
ε
=
1.8
of
for
increasing
ε
r
r
f,f,
for
increasing
permittivities
ε
=
1
to
ε
=
1.8
rr = 1 to εrr = 1.8 of
of the
the ininf,
for
increasing
permittivities
ε
the
inf,ner
forvolume.
increasing
permittivities ε r = 1 to εr = 1.8 of
Comparing
from
the
ner
volume.
Comparing
the
effective
permittivities
ner
volume.
Comparing
the
effective
permittivities
from
the
ner
volume.
Comparing the effective
permittivities from the
slopes
of
in
Fig.
slopes
of
the
linear
functions
to
the
permittivities
slopes of
of the
the linear
linear functions
functions in
in Fig.
Fig. 444 to
to the
the permittivities
permittivities
slopes
used
for
an
evaluation
error
of
about
%
is
used
for
the
simulation,
evaluation
error
of
about111%
used
for
the
simulation,
an
evaluation
error
of
about
% is
is
used
for theThis
simulation, an evaluation error of the
about
1%
obtained.
proposed
obtained.
This
confirms
the
applicability
of
obtained.
This
confirms
the
applicability
of
the
proposed
obtained. This confirms the applicability of the proposed
www.adv-radio-sci.net/9/27/2011/
dB
dB
-10
-10
-20
-20
-30
-30
-40
-40
(b) 3D
3-Dsimulation,
simulation,magnitude
magnitude
E-field
of of
thethe
E-field
at a at
fre-a
(b) 3D simulation, magnitude
of
magnitude
of the
the E-field
E-field at
at aa frefrefrequency
of GHz
10 GHz
a homogeneous
relative
quency of 10
withwith
a homogeneous
relative
perquency of 10 GHz with
with aa homogeneous
homogeneous relative
relativeperperpermittivity
εr,eff
=4.=4.
mittivity εr,eff
mittivity εεr,eff
r,eff =4.
Fig. 3: Screenshots of transient 3-D field simulations at constant frequency for changing permittivity, CST Microwave
Fig.
3: 2010.
Screenshots of
transient 3D
3D field simulations
simulations at conconFig.
Studio
Fig. 3:
3: Screenshots
Screenshots of
of transient
transient 3D field
field simulations at
at constant
frequency
for
changing
permittivity,
CST
Microwave
stant
stant frequency
frequency for
for changing
changing permittivity,
permittivity, CST
CST Microwave
Microwave
Studio
2010.
Studio
Studio 2010.
2010.
concept and model. Due to the linear run of the phases in
Fig. 4 it can be assumed, that standing waves are also negliconcept
and
model.
Due
to the
the linear
linear
runruns.
of the
the phases
phases in
in
gible
as they
would
cause
nonlinear
phase
concept
and
model.
Due
to
run
concept
and
model.
Duethat
to the
linear
run of
ofare
thealso
phases
in
Fig.
4
it
can
be
assumed,
standing
waves
negliFig.
44 it
can
be
assumed,
that
standing
waves
are
also
negliFig.
it
can
be
assumed,
that
standing
waves
are
also
negligible
as
they would
cause nonlinear
nonlinear phase runs.
runs.
3.2
gible
as
gible Measurements
as they
they would
would cause
cause nonlinear phase
phase runs.
3.2
Measurements
In
a next
step, measurements with different homogeneous
3.2
3.2 Measurements
Measurements
materials with known permittivities have been performed.
In
a phantoms
next step,
measurements
with
different homogeneous
homogeneous
Two
have
been made with
of polyethylene
(PE, relative
In
step,
measurements
different
In aa next
next with
step,
measurements
with
different
homogeneous
materials
known
permittivities
have
been
performed.
permittivity
ε
=
1.4)
and
polyoxymethylene
(POM,
r,eff,PE
materials
with
known
permittivities
have
been
performed.
materials
with have
known
permittivities
have
been(PE,
performed.
Two
phantoms
been
made
of
polyethylene
relative
5
shows
the
relative
permittivity
ε
=
3.8).
Figure
r,eff,POM
Two
have
been
made
of
polyethylene
(PE,
relative
Two phantoms
phantoms
have
been
made
ofpolyoxymethylene
polyethylene
(PE,
relative
=
1.4)
and
(POM,
permittivity
ε
measured
phase
1φ
versus
frequency
f
for
both
phantoms.
r,eff,PE
1.4)
and
polyoxymethylene (POM,
permittivity
εεr,eff,PE =
=the
1.4)slope
and
permittivity
r,eff,PE
relative
permittivity
εε r,eff,POM
= 3.8).
3.8).
Fig.
shows
the(POM,
meaLike
in the
simulation,
ofpolyoxymethylene
theFig.
phase
increases
with
relative
permittivity
=
555 shows
the
r,eff,POM
relative
permittivity
ε
=
3.8).
Fig.
shows
themeamear,eff,POM
sured
phase
Δφ
versus
frequency
f
for
both
phantoms.
Like
increasing
permittivity.
The
permittivites
calculated
from
the
sured
phase
Δφ
versus
frequency
ff for
phantoms.
Like
sured
phase
Δφ
versus
frequency
for both
both
phantoms.
Like
in
the
simulation,
the
slope
of
the
phase
increases
with
inslopes
show
a
relative
measurement
error
of
9.2%
for
PE
and
in
simulation,
the
slope
of
increases
with
inin the
the
simulation,
the The
slopepermittivites
of the
the phase
phasecalculated
increasesfrom
withthe
increasing
permittivity.
0.7
% for
POM, respectively.
creasing
permittivity.
The
permittivites
calculated
from
the
creasing
permittivity.
The
permittivites
calculated
from
the
slopes
aa relative
error of
of 9.2
% for
forconPE
Both,show
simulation
andmeasurement
phantom measurement
results,
slopes
show
relative
measurement
slopes
show
aPOM,
relative
measurement error
error of 9.2
9.2 %
% for PE
PE
and
0.7
%
for
respectively.
firm
the
validity
of
the
proposed
method.
and
and 0.7
0.7 %
% for
for POM,
POM, respectively.
respectively.
Adv. Radio Sci., 9, 27–30, 2011
30
44 4
Δφ / rad
Δφ
Δφ // rad
rad
55 5
00 0
-5
-5
-5
-10
-10
-10
-15
-15
-15
-20
-20
-20
-25
-25
-25
-30
-30
-30
-35
-35
-35
-40
-40
-40
-45
-45 8.0 8.5
-45
9.0
8.0 8.5
8.5
9.0
8.0
9.0
C.
Baer
al.:
Density
measurement
solid
particle
flows
Baer
etet
al.:
Density
Measurement
ofof
Solid
Particle
Flows
C.C.
Baer
al.:
Density
Measurement
Solid
Particle
Flows
C.
Baer
etetal.:
Density
Measurement
ofofSolid
Particle
Flows
epsr,eff=1
epsr,eff=1
=1
eps
r,eff
eps
=1,2
r,eff
eps
=1,2
epsr,eff
r,eff=1,2
epsr,eff=1,4
epsr,eff=1,4
=1,4
eps
r,eff
eps
=1,6
r,eff
epsr,eff=1,6
=1,6
eps
r,eff
eps
=1,8
r,eff
eps =1,8
=1,8
eps
measuredcoal
coal /g
measured
measured
coal /g/g
300
300
Gcoal,meas
300
coal,meas
GGcoal,meas
Gcoal,actual
coal,actual
GGcoal,actual
250
250
250
200
200
200
r,eff
r,eff
150
150
150
100
100
100
50
50
50
9.5
10.0 10.5 11.0 11.5 12.0
10.0 10.5 11.0
9.5
12.0
11.0 11.5
11.5 12.0
f /GHz
f /GHz
Fig.4:4: Simulated
Simulated phase differencesΔφ
Δφ fordifferent
different hoFig.
Fig.
4: Simulated
Simulated phase
phase differences
for
hoFig.
4:
differences 1φ for
for different
differenthohomogenous
permittivities.
mogenous permittivities.
permittivities.
mogenous
permittivities.
mogenous
Δφ/ rad
/ rad
Δφ
Δφ
/ rad4
10
4
0104
10
00
reference
reference
reference
polyethylene
-0.5
polyethylene
polyethylene
-0.5
-0.5
polyoxymethylene
polyoxymethylene
polyoxymethylene
-1
-1
-1
-1.5
-1.5
-1.5
-2-2
-2
-2.5
-2.5
-2.5
-3-3
-3
-3.5
-3.5
-3.5
-4-4 8.0 8.5
10.0 10.5
10.5 11.0 11.5
11.5 12.0
-4
8.0 8.5
8.5
9.09.0 9.5
9.59.5 10.0
10.0
11.0 11.5 12.0
8.0
9.0
10.5 11.0
12.0
f
/GHz
ff /GHz
/GHz
Fig.
Measuredphase:
phase:Polyethylene
Polyethyleneand
andpolyoxymethypolyoxymethyFig.
5:
polyethylene
and
Fig.
5:5:Measured
Measured
Fig.
5:
Measured
phase: Polyethylene
and polyoxymethypolyoxymethylene
phantom.
lene
phantom.
lene
phantom.
lene phantom.
Measurements on a coal/gas composition have been perBoth,simulation
simulationand
andphantom
phantommeasurement
measurement
results,
conBoth,
results,
conformed
by
inserting and
different
amounts
of coal particles
with
Both,the
simulation
phantom
measurement
results, confirm
validity
of
the
proposed
method.
firm the validity of the proposed method.
well
weight
intoproposed
the tube.method.
Since the total volume of
firm known
the validity
of the
Measurementsonona acoal/gas
coal/gascomposition
compositionhave
have
been
perMeasurements
been
pertheMeasurements
measurement tube
and
the density
of thehave
coal been
is known,
on
a
coal/gas
composition
performed
by
inserting
different
amounts
of
coal
particles
with
formed by inserting different amounts of coal particles with
measurements
resultsdifferent
can be compared
tocoal
the known
volume
formed
by inserting
amounts
of
particles
withof
wellknown
known
weightinto
intothethetube.
tube.Since
Since
total
volume
well
weight
thethe
total
volume
of
fractions.
Figure
6
shows
the
measured
total
coal
mass
verwell
known
weight
into
the
tube.
Since
the
total
volume
of
themeasurement
measurementtube
tubeand
andthethedensity
densityofof
coal
known,
the
thethe
coal
is is
known,
sus
the
actually inserted
and
dispersed
coal
mass.
The
relathemeasurements
measurement
tube
and
the
density
of
the
coal
is
known,
resultscan
canbebe
compared
known
volume
measurements results
compared
toto
thethe
known
volume
measurements
results
can
be
compared
to
the coal
known
volume
tive
measurement
error
isthe
approximately
13%.
fractions.
Fig.
shows
themeasured
measured
total
mass
versus
fractions.
Fig.
6 6shows
total
coal mass
versus
fractions.
Fig.
6 shows
the
measured
total
coal The
mass
versus
theactually
actually
inserted
and
dispersed
coal
mass.
The
relative
the
inserted
and
dispersed
coal
mass.
relative
the
actually inserted
and
dispersed coal
The relative
measurement
errorisis
approximately
%.
measurement
error
approximately
1313
%.mass.
4measurement
Conclusions
error is approximately 13 %.
In4 4this
paper, a novel density measuring method for pulConclusions
Conclusions
4 Conclusions
verized
fuels in pneumatic conveying systems is proposed.
InInthis
method
The
concept
is abased
ondensity
electromagnetic
transmission
meathispaper,
paper,
anovel
novel
densitymeasuring
measuring
methodforforpulpulIn
this paper,
novel
density
measuring
method
for
pulverized
fuels
inainpneumatic
conveying
is isproposed.
surements
along
the
cross-section
of asystems
conveying
tube
converized
fuels
pneumatic
conveying
systems
proposed.
verized of
fuels
in pneumatic
conveying
systems is setup,
proposed.
sisting
acrylic
glass. In
the implemented
two
Adv. Radio Sci., 9, 27–30, 2011
70
70
70
140
210
140
210
140
210
actually inserted coal /g
actuallyinserted
insertedcoal
coal/g/g
actually
280
280
280
Fig.
Comparison
between
mass
of
inside
Fig.
6:6:Comparison
Comparison
between
mass
coal
measured
inside
Fig.
6:6:
between
mass
ofofof
coal
measured
inside
Fig.
Comparison
between
mass
coal
measured
inside
the
measurement
tube
and
actually
inserted
coal
mass
the
measurement
tube
and
actually
inserted
coal
mass.
the
themeasurement
measurementtube
tubeand
andactually
actuallyinserted
insertedcoal
coalmass
mass
horn-antennas are arranged opposite to each other with the
The
concept based
is based
on electromagnetic
transmission
meaThe
concept
on
transmission
meaThe
conceptisisbeing
basedoriented
onelectromagnetic
electromagnetic
transmission
mealine-of-sight
perpendicularly
to the tube
axis.
surements
along
the
cross-section
of
a
conveying
tube
consurements
along
the
ofofa aconveying
tube
consurements
along
thecross-section
cross-section
conveying
tubeis
conThe complex
transmittance
between
both
antennas
measistingof ofacrylic
acrylicglass.
glass. In
the
implemented
setup, two
sisting
the
implemented
setup,
sisting
oforder
acrylic
glass. InInthe
themean
implemented
setup,two
two
sured in
to determine
volume fraction
of
the
horn-antennas
are
arranged
opposite
to
each
other
with
the
horn-antennas
are
opposite
totoeach
other
the
horn-antennas
arearranged
arranged
opposite
each
otherwith
with
the
gas/coal particle
flow.
The
phase of
the
complex
transmitline-of-sight
being
oriented perpendicularly
to the
tube axis.
line-of-sight
being
oriented
totothe
tube
line-of-sight
being
orientedperpendicularly
perpendicularly
the
tubeaxis.
axis.of
tance
contains
information
about
the effective
permittivity
The
complex
transmittance
between
both
antennas
is meaThe
complex
between
both
antennas
isismeaThe
complextransmittance
transmittance
between
both
antennas
mea-of
the
quasi-continuous
medium.
The
phase
contributions
sured
order to
determine
the mean
volume
fraction the
of the
sured
ininin
order
the
fraction
sured
orderto
todetermine
determine
themean
meanvolume
volume
fractionofof
the
antennas
and
feeding
networks
are
compensated
bythe
takgas/coal
particle
flow.
The
phase
of
the
complex
transmitgas/coal
particle
ofofthe
transmitgas/coal
particleflow.
flow. The
Thephase
phase
thecomplex
complex
transmiting
a
reference
measurement
with
the
empty
tube
into
tance
contains
information
about the
effective
permittivityacof
tance
contains
information
about
permittivity
tance
contains
information
aboutthe
theeffective
effective
permittivityof
of
count.
Furthermore,
electromagnetic
field
simulations
have
the
quasi-continuous
medium.
The
phase
contributions
of
the
quasi-continuous
medium.
contributions
ofof
the
quasi-continuous
medium.
Thephase
phase
contributions
been
performed
tofeeding
evaluate
theThe
concept
and
the performance
the
antennas
and
networks
are
compensated
takthe
antennas
and
feeding
networks
are
compensated
bybyby
takthe
antennas
and
feeding
networks
are
compensated
takofing
thea method.
Concerning
thewith
permittivity
measurements,
reference
measurement
the
empty
tube
into
acing
a
reference
measurement
with
the
empty
tube
into
acing
aevaluation
reference error
measurement
empty
tube
acthe
iselectromagnetic
smallerwith
thanthe
1%.
The
errorinto
of meacount.
Furthermore,
field
simulations
have
count.
Furthermore,
electromagnetic
field
simulations
have
count.
Furthermore,
electromagnetic
field simulations
have
surement
with two
homogenous
phantoms
with
different
perbeen
performed
evaluate
the
concept
and
the
performance
been
performed
tototo
evaluate
the
concept
and
the
performance
been
performed
evaluate
the
concept
and
the
performance
mittivities
are 0.7%
and 9.7%,the
respectively.
Different
tests on
of
the
method.
Concerning
permittivity
measurements,
of
the
Concerning
the
measurements,
of
the method.
method.
Concerning
thepermittivity
permittivity
measurements,
dispersed
coalerror
particles
revealed
an
of
13%.
prothe
evaluation
error
is smaller
than
1error
%.
The
errorThe
of meathe
evaluation
is
smaller
than
1
%.
The
error
the
evaluation
error
is
smaller
than
1
%.
The
errorof
ofmeameaposed
monitoring
concept
has
been
verified
based
on
various
surement
with
two
homogenous
phantoms
with
different
persurement
with
homogenous
phantoms
with
persurement
withtwo
twosimulations
homogenous
phantoms
withdifferent
different
per-of
electromagnetic
and
measurements.
The tests
goal
mittivities
are
0.7
%
and
9.7
%,
respectively.
Different
tests
mittivities
are
0.7
%
and
9.7
%,
respectively.
Different
mittivities
are
0.7
%
and
9.7
%,
respectively.
Different
tests
future
developments
is
to
improve
the
measurement
accuracy
on
dispersed
coal
particles
revealed
an
error
of
13
%.
The
on
dispersed
coal
particles
revealed
ananerror
1313%.%.The
on
dispersed
coal
particles
revealed
errorofof
The
by
employing
more
sophisticated
calibration
techniques.
proposed
monitoring
concept
been
verified
based
proposed
monitoring
concept
hashas
been
verified
based
on on
var-varproposed monitoring concept has been verified based on various
electromagnetic
simulations
measurements.TheThe
ious
electromagnetic
simulations
andand
measurements.
ious electromagnetic simulations and measurements. The
goal
future
developments
is to
improve
measurement
goal
of of
future
developments
is to
improve
thethe
measurement
References
goal
of future developments is to improve the measurement
accuracy
employing
more
sophisticated
calibration
techaccuracy
byby
employing
more
sophisticated
calibration
techaccuracy by employing more sophisticated calibration techniques.
niques.
Conrads, H. G.: Method and device for a contact-free measurement
niques.
of the mass flow rate in a two phase pneumativ transport using
microwaves, Eur ER Patent 07176269 A2, 1996.
References
References
Happel,
J.: Method and device for measuring a mass flow, US Patent
References
7102133, 2006.
Conrads,
Method
device
a contact-free
measurement
Conrads,
H. H.
G.:G.:
Method
andand
device
forfor
a contact-free
measurement
Miller, D.,
Baimbridge,
P., and
Eyre,
D.:
Technology
status of PF
Conrads,
H.mass
G.:
Method
device
for pneumativ
a contact-free
measurement
flow
a two
phase
pneumativ
transport
using
of of
thethe
mass
flow
raterate
inand
aintwo
phase
transport
using
flow
measurement
andincontrol
methods
for utilitytransport
boilers, Report
of
the
mass
flow
rate
a
two
phase
pneumativ
using
microwaves,
Patent
07176269
microwaves,
EurEur
ERER
Patent
07176269
A2.A2.
No. COAL R201
DTI/PUB
00/1445, Crown
2000.
microwaves,
Eur ER
Patent
07176269
A2. a Copyright
Happel,
Method
device
measuring
a mass
, US
Happel,
J.: J.:Method
andand
device
forfor
measuring
mass
flowflow
, US
Penirschke,
A. and Jakoby,
R.: Design
of a Moisture
Independent
Happel,
J.:
Method
and
device
for
measuring
a
mass
flow
, US
Patent
7102133.
Patent
7102133.
Microwave
Mass Flow
Detector
forTechnology
Particulate
Solids,
Patent
7102133.
Miller,
D.,D.,
Baimbridge,
P., P.,
Eyre,
D.:D.:
Technology
status
of PF
flowflow
Miller,
Baimbridge,
Eyre,
status
of GEMIC
PF
2010.
Miller,
D.,
Baimbridge,
P.,
Eyre,
D.:
Technology
status
of
PFNo.
flow
measurement
andand
control
methods
forfor
utility
boilers,
Report
measurement
control
methods
utility
boilers,
Report
No.
Ruhrkohlenhandbuch:
Glueckaufverlag
GmbH,
Essen,
1984.
measurement
and
control
methods
for
utility
boilers,
Report
COAL
R201
DTI/PUB
00/1445,
Crown
Copyright
2000.
COAL
R201
DTI/PUB
00/1445,
Crown
Copyright
2000. No.
Sihvola,
Mixing
Rules
with ofcomplex
Dielectric
Coefficients,
COAL A.:
R201
DTI/PUB
00/1445,
Crown
Copyright
2000.
Penirschke,
A.,A.,
Jakoby,
R: R:
Design
aofMoisture
Independent
Mi-MiPenirschke,
Jakoby,
Design
a Moisture
Independent
Subsurface
Sensing
Technologies
Applications
Vol.
1,MiNo.
Penirschke,
A.,
Jakoby,
R:Detector
Design
aand
Moisture
Independent
crowave
Mass
Flow
Detector
foroffor
Particulate
Solids,
GEMIC
crowave
Mass
Flow
Particulate
Solids,
GEMIC
4, 2000. Mass Flow Detector for Particulate Solids, GEMIC
crowave
www.adv-radio-sci.net/9/27/2011/