ArticlePDF Available

Influence of Peanut Orientation on Microwave Sensing of Moisture Content in Cleaned Unshelled Peanuts

Authors:

Abstract

An investigation of the influence of unshelled peanut orientation on microwave moisture sensing from measurement of dielectric properties is presented for the first time in this article. A free-space dielectric measurement technique was used for measurement on cleaned unshelled peanuts at 9.6 GHz with a pair of low-cost printed Yagi-Uda antennas. The peanut sample was placed between the transmitting and receiving antennas in a cubical polyester sample container (76 mm ×76\times76 mm ×76\times76 mm). Influence of orientations of the peanut pods, were studied with 2, 3, 4 and 6-direction measurements of the dielectric properties of the sample. The different measurement directions were obtained by rotation of the cubical sample container. The dielectric properties were determined from measurements of the attenuation and phase-shift caused by the peanut sample when placed between the antennas. The dielectric properties of the peanut sample were found to be dependent on the orientation distribution of pods. The average value of dielectric properties determined from 6-direction measurement was more representative of the dielectric properties of the bulk peanut sample. A temperature-compensated function for the dielectric properties and subsequent moisture content calibration equation were established and used during the in-field testing. Results of the in-field microwave moisture measurements are compared with values obtained by using the official moisture determination method (capacitance-type meter).
1530-437X (c) 2021 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2022.3168664, IEEE Sensors
Journal
IEEE SENSORS JOURNAL, VOL. XX , NO. XX, MONTH X, XXXX 1
XXXX-XXXX © XXXX IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
AbstractAn investigation of the influence of unshelled
peanut orientation on microwave moisture sensing from
measurement of dielectric properties is presented for the first
time in this article. A free-space dielectric measurement
technique was used for measurement on cleaned unshelled
peanuts at 9.6 GHz with a pair of low-cost printed Yagi-Uda
antennas. The peanut sample was placed between the
transmitting and receiving antennas in a cubical polyester
sample container (76 mm x 76 mm x 76 mm). Influence of
orientations of the peanut pods, were studied with 2, 3, 4 and 6-
direction measurements of the dielectric properties of the
sample. The different measurement directions were obtained by
rotation of the cubical sample container. The dielectric properties were determined from measurements of the
attenuation and phase-shift caused by the peanut sample when placed between the antennas. The dielectric properties
of the peanut sample were found to be dependent on the orientation distribution of pods. The average value of
dielectric properties determined from 6-direction measurement was more representative of the dielectric properties of
the bulk peanut sample. A temperature-compensated function for the dielectric properties and subsequent moisture
content calibration equation were established and used during the in-field testing. Results of the in-field microwave
moisture measurements are compared with values obtained by using the official moisture determination method
(capacitance-type meter).
Index TermsBulk properties, cleaned unshelled peanuts, complex permittivity, free-space measurement, microwave
sensing, moisture content.
I. Introduction
OISTURE sensing by microwave measurements is of
interest for quality assurance in the peanut production and
processing industry [1]-[6]. Because of their polar nature, water
molecules have a strong interaction with the electric field of the
electromagnetic wave at microwave frequencies. The ability of
the water dipoles to be polarised by the electric field is defined
by the relative complex permittivity (εr=ε'-jε''), where dielectric
constant (ε') represents the energy storage and dielectric loss
factor (ε'') represents the energy loss. Therefore, the relative
complex permittivity is a unique parameter for each condition
of cleaned unshelled peanut. Typically, the water dipole
polarisation plays an important role in relative complex
permittivity measurements for hygroscopic materials at
microwave frequencies above 2.5 GHz, while the dielectric
behavior for materials at frequencies lower than 2.5 GHz is
dominated by ionic conductivity [7]-[9]. It is well known that
the cost and complexity of microwave components increase
with frequency. Therefore, a tradeoff between cost and
This paragraph of the first footnote will contain the date on which you
submitted your paper for review. It will also contain support information,
including sponsor and financial support acknowledgment. For example,
“This work was supported in part by the U.S. Department of Commerce
under Grant BS123456.”
The next few paragraphs should contain the authors’ current
affiliations, including current address and e-mail. For example, F. A.
measurement frequency has been a concern in microwave
moisture meter development.
Cleaned unshelled peanuts have a unique shape and the size
of peanut pods depends on the type and variety. Accounting for
about 80% of USA peanut production are runner type peanut,
because it is high yield and it is commonly used for making
peanut butter. Typically, the shape of runner type peanut is
nearly cylindrical and they contain two kernels inside. For the
single kernel peanut pods, they are nearly of spherical shape. It
was reported by researchers from University of Georgia [10]
that for the runner type peanut the pod density increased with
peanut pod maturity while the moisture content decreased
significantly with pod maturity. The average pod densities for
immature kernels, mid-mature kernels and mature kernels were
540 kg m-3, 620 kg m-3 and 690 kg m-3, respectively. Therefore,
the pod density implies the edible kernels in the peanut sample.
The peanut maturity determination was predicted by observing
the color of peanut pods [11]. The peanut sample was taken
from four to five peanut plants (180 - 200 pods) which were
Author is with the National Institute of Standards and Technology,
Boulder, CO 80305 USA (e-mail: author@ boulder.nist.gov).
S. Julrat and S. Trabelsi are with USDA Southeast Poultry Research
Laboratory, 934 College Station Road Athens, GA 30605 USA (e-mail:
sakol.julrat@gmail.com).
Influence of peanut orientation on microwave
sensing of moisture content in cleaned
unshelled peanuts
Sakol Julrat and Samir Trabelsi, Fellow, IEEE
M
Authorized licensed use limited to: Prince of Songkla University provided by UniNet. Downloaded on May 30,2022 at 03:09:07 UTC from IEEE Xplore. Restrictions apply.
1530-437X (c) 2021 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2022.3168664, IEEE Sensors
Journal
8 IEEE SENSORS JOURNAL, VOL. XX , NO. XX, MONTH X, XXXX
randomly dug in the field. The maturity level for harvest was
70% or more of peanut pods from the brown and black color
classes. Later, [12] proposed a peanut bulk density estimator
technique for in-field peanut measurements. A peanut sample
container was mounted on a vibrator to minimize the space
between the peanut pods or some other abnormal peanuts such
as damage peanut pods. The average bulk density of uncleaned
peanut runner type, Spanish type and Virginia type were 290 kg
m-3, 307 kg m-3 and 256 kg m-3, respectively, obtained from the
2002, 2003 and 2004 Uniform Peanut Performance Test [12],
the dielectric properties of bulk peanut samples are those of
nonhomogeneous materials and different from those of
biological anisotropic materials such as chicken meat [13] and
woods [14]-[15].
The dielectric properties of cleaned unshelled peanuts were
first measured with a microwave free-space measurement
system involving an automatic vector network analyser and
horn-lens antennas system [1]. This technique is
nondestructive, does not require contact with the sample, and
can be flexible in terms of sample size which can be achieved
by changing the sample thickness. The measured dielectric
properties of cleaned unshelled peanuts were not only related to
moisture content but also to bulk density which is characteristic
of particulate materials [16]. Later [1]-[4], a density-dependent
calibration function expressed in terms of dielectric properties
was established for cleaned unshelled peanuts moisture content
determination. It was also implemented in the first portable
unshelled peanut microwave moisture meter [17] for in-field
cleaned unshelled peanuts measurements. Furthermore, a
nondestructive and noncontact microwave measurement
system for foreign materials sensing in peanut samples was
developed [18]. The measured dielectric properties were
significantly different with the foreign material contents and
can be used for the development of algorithms for determining
foreign material content.
To the authors’ knowledge, the influence of peanut pods
orientation has not been reported with respect to dielectric
properties measurement and moisture content determination.
Studying the influence of peanut orientation and measurements
for different positions of the peanut sample will provide insight
on the influence of peanut orientation and is useful in
developing accurate and robust algorithms for nondestructive
moisture content determination.
II. MATERIALS AND METHODS
A. Microwave circuit design and measurement
technique
A free-space measurement technique for a small unshelled
peanut sample (76 mm x 76 mm x 76 mm), was designed and
developed based on our previous measurement system for
foreign-materials sensing in cleaned unshelled peanuts. A pair
of printed Yagi-antennas with vertical electromagnetic
polarisation [18] was used. Figure 1 shows the microwave
components of the proposed microwave free-space
measurement system. The peanut sample in cubical sample
container (76 mm x 76 mm x 76 mm) was placed in the middle
between transmitter and receiver antennas with the interfaces
31mm away from the antennas. The microwave source (Analog
Devices: 110227-HMC734LP5), was a low-cost voltage-
controlled oscillator with 5 Volt reference operating at 9.6
GHz. An 18 dB isolator was used to prevent any reflected signal
from reaching the source. A power divider (3 dB) was used to
provide the microwave reference signal to the local oscillator
(LO) port of the IQ-demodulator (Polyphase microwave:
AD90120B) and a transmitted signal to the transmitting
antenna. The electromagnetic energy radiates into the material
sample, which was 76 mm thick. The microwave phase was
shifted, and the microwave power was attenuated due to the
microwave-material interaction mechanisms. The receiving
antenna collects the transmitted electromagnetic wave from the
peanut sample and it was amplified by a 20 dB amplifier (Mini-
Circuits: ZX60-183A-S+) before sending the signal to the radio
frequency (RF) port of the IQ-demodulator.
The data acquisition unit (Measurement computing: usb-
1608fs) was connected to the IQ-demodulator which converted
the analog signals into digital signals which were sent to the
personal computer via universal port (USB). The measured
transmission parameters, attenuation and phase shift, are related
to the voltages at the IQ-demodulator output as follows:
𝑀𝑎𝑔𝑛𝑖𝑡𝑢𝑑𝑒 = 𝑉+𝑉 (1)
𝑃ℎ𝑎𝑠𝑒 = tan 󰇣
󰇤 (2)
where VI and VQ are the voltage readings at the IQ-demodulator.
The attenuation (A) and phase shift (
) due to microwave-
material interaction can be expressed as follows:
𝐴 = 𝑀𝑎𝑔𝑛𝑖𝑡𝑢𝑑𝑒, 𝑆 (𝑑𝐵) 𝑀𝑎𝑔𝑛𝑖𝑡𝑢𝑑𝑒, 0 (𝑑𝐵) (3)
= 𝑃ℎ𝑎𝑠𝑒, 𝑆 (𝑟𝑎𝑑) 𝑃ℎ𝑎𝑠𝑒, 0 (𝑟𝑎𝑑) (4)
where Magnitude, S and Magnitude,0 are the magnitude with
and without sample, respectively. Phase, S and Phase,0 are the
phase with and without sample, respectively. The dielectric
properties can be determined from measured attenuation and
phase shift as follows [19]-[20];
Fig. 1.
Layout of microwave components for a peanut moisture
measurement system
Authorized licensed use limited to: Prince of Songkla University provided by UniNet. Downloaded on May 30,2022 at 03:09:07 UTC from IEEE Xplore. Restrictions apply.
1530-437X (c) 2021 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2022.3168664, IEEE Sensors
Journal
8 IEEE SENSORS JOURNAL, VOL. XX , NO. XX, MONTH X, XXXX
𝜀󰆒= 󰇣1 + ∅
󰇤 (5)
𝜀󰆒󰆒 =
.
𝜀󰆒 (6)
where 𝜀󰆒is the dielectric constant and 𝜀󰆒󰆒 is the dielectric loss
factor, c = 3*108 m s-1 is the speed of light, d is the sample
thickness in meters (width of sample container), and f is the
frequency in hertz.
A cubical sample container (76 mm x 76 mm x 76 mm) was
designed with a clear color nGen co-polyester 3D printing
filament material [21]. The dielectric properties can be
measured from all 6 faces of the cubical sample container. Face
number 5 is the lid of the cubical container and can be opened
for loading the sample. The lid (with a small pin at each corner)
was designed to fit snugly so that the cubical sample container
is tightly closed. Unshelled peanuts were needed to completely
fill the cubical container. Figure 2 shows, the system integration
in the control box (160 mm x 390 mm x 140 mm). The
transmitter and receiver antennas were mounted on the top
compartment. All microwave components and power supply
unit were placed at the bottom compartment under the absorber
sheet (Laird: ECCOSORB®AN79). A Styrofoam piece was
designed to correctly position the cubical sample container
which rested on the absorber sheet. To avoid the unwanted
reflection signals, the absorber sheets were applied to all sides
of the top compartment. Figure 3 shows the complete peanut
moisture measurement system. The Styrofoam top (160 mm x
390 mm x 20 mm) had a square opening (76 mm x 76 mm) for
inserting the cubical sample container in the proper position.
Six different orientations of the peanut sample were measured
by inserting the cubical sample container (face 1 to 6) into the
square slot in the Styrofoam top, with the face being measured
pointing to the transmitter antennas. From each set of
measurement, the average (𝜀) and standard deviation (𝜀)
of dielectric constant or dielectric loss factor were calculated as
follows:
𝜀 =
𝜀
 (7)
𝜀 =
(𝜀𝜀)
 (8)
where i is face number (i = 1, 2, 3, ... ,6), n is number of
face measurement, εi is the dielectric constant or dielectric loss
factor value at face ith.
B. Peanut Moisture Conditioning
Peanut samples (runner type) were collected from a buying
point located in Georgia, USA. Mass distribution of the peanut
samples (6.9% moisture content) was determined and plotted in
Fig. 4. Each bag of peanuts weighed about 2 kilograms. The
initial moisture content of these peanut samples was on average
about 6.9% (wet basis) [22]. To obtain samples with higher
moisture levels, peanuts were sprayed with a fine mist of
deionized water and sealed in plastic bags. For each target
moisture level, the amount of water to be added was calculated.
After the water was added, the samples were kept in a cold room
at 4°C for three days to equilibrate. Each day the samples were
stirred one by one to allow an even distribution of the water
within the sample. A total of 10 peanut samples were prepared
with moisture ranging from 6.9% to 14.9%. Peanut pod samples
were arbitrarily loaded in a container of size 0.210 m x 0.121 m
x 0.205 m, the total weight was 1.825 kg (bulk density is 337.1
kg m-3) with a total of 1499 peanut pods. The two-kernels
peanut pods were the main population with an average mass
between 1 to 1.75 grams and corresponding average length and
diameter of 25 mm and 6 mm, respectively. The average mass
of single kernel pods was about 0.5 grams, they are nearly round
with average diameter of 6 mm.
Fig. 2. System integration in the top compartment of the control box.
Fig. 3. Peanut moisture measurement system operating at 9.6 GHz.
Fig. 4. Mass distribution of cleaned unshelled peanut samples (6.9%
moisture content) -Single kernels and -Two-kernels.
Authorized licensed use limited to: Prince of Songkla University provided by UniNet. Downloaded on May 30,2022 at 03:09:07 UTC from IEEE Xplore. Restrictions apply.
1530-437X (c) 2021 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2022.3168664, IEEE Sensors
Journal
8 IEEE SENSORS JOURNAL, VOL. XX , NO. XX, MONTH X, XXXX
C. Oriented Peanut Loading
The cleaned unshelled peanut pods were carefully oriented and
loaded into the cubical sample container, layer by layer, as
illustrated in Fig. 5 (a). With the cover lid tightly in place,
measurements were taken for each face of the container
resulting in six independent measurements for each sample. In
all, 180 peanut pods (two-kernels = 150 pods with average
length = 25 mm and diameter 6 mm, single kernels = 30 pods
with average diameter 6 mm) were used to fill the cubical
sample container. The bulk density of oriented peanut sample
was 375.8 kg m-3. It should be noted that there was a small gap
between the lid and peanut sample which allowed the lid to
close tightly the cubical sample container. The lid was held by
the friction force between the pins and container.
D. Arbitrary Peanut Loading
The cleaned unshelled peanut samples were arbitrarily loaded
by hand in the cubical sample container (no peanut pods size
selection). Empty pockets near the corners and edges were
carefully avoided by spreading the cleaned unshelled peanut
samples as illustrated in Fig. 5 (b). In all, 147 peanut pods (two-
kernels = 132 pods with average length = 25 mm and diameter
6 mm, single kernels = 15 pods with average diameter 6 mm)
were used to fill the cubical sample container. The bulk density
of arbitrarily loaded peanut sample was 318.9 kg m-3.
E. Oven Moisture Content Determination
Oven moisture content [22] was used as the standard
technique for determining the moisture content in peanut
samples. The oven moisture content was determined for 100
grams of peanut pod sample. The peanut pod sample was placed
in an aluminum pan and kept in the oven at 130°C for 6 hours.
Once out of the oven, the samples were placed in a desiccator
over anhydrous CaSO4 and allowed to cool to room
temperature. Moisture content of the samples was calculated on
wet basis as follows:
MC (%)=
100 (9)
where MC (%) is moisture content, 𝑚 is the mass of the water
in the sample, and 𝑚 is the mass of the dry matter in the
sample.
III. RESULTS AND DISCUSSION
A. Oriented and Arbitrary loadings
Measurements of oriented and arbitrary loadings of cleaned
unshelled peanut samples were investigated in this section as an
example to show the influence of pod orientations in these two
extreme situations. To avoid moisture loss during the peanut
loading process, peanut pod samples of low moisture content
(6.9% wet basis [22]), from a buying point in Georgia were
used. The sample was stored in a plastic bag and kept at room
temperature (23°C) for 12 Hours before the measurements. An
empty cubical sample container was used for system calibration
by rotating and measurement at all 6 orientations.
Measurement results of oriented cleaned unshelled peanut
loading are shown in Fig. 6 (green bars), the dielectric
properties measurements from faces 1 and 3 show the highest
dielectric properties values compared to faces 2, 4, 5 and 6. This
was not only related to the orientation of pods only but also to
the number of kernels interacting with the electric field. As
shown in Fig.5 (a), a higher number of peanut kernels was
presented in the measurement direction for faces 1 to 3 (6
kernels) than for faces 2 or 4, (5 kernels).
However, sometimes 5 kernels were presented in direction
faces 1 to 3 because of the longer length of peanut pods.
Generally, more air gaps were presented for the 5-kernel
direction than for the 6-kernel direction. It should be noted that
more air gaps could be present at the top because some space
Fig. 5.
Cleaned unshelled peanut in the cubical container with two
always with their long axis in the direction of faces 1,3 and the 1-
kernel
pods were used to fill up the gap when the 2-
kernel pods cannot fill the
gaps) and (b) arbitrary
(a)
(b)
Fig. 6.
Measured dielectric properties of a cleaned unshelled peanut
sample (6.9% moisture content) at 9.6 GHz and room temperature
(23°C) with oriented and arbitrary peanut loadings for 1-direction, 2-
direction, 4-direction and 6-direction measurements, (a) diele
ctric
constant and (b) dielectric loss factor. -oriented and -arbitrary.
Authorized licensed use limited to: Prince of Songkla University provided by UniNet. Downloaded on May 30,2022 at 03:09:07 UTC from IEEE Xplore. Restrictions apply.
1530-437X (c) 2021 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2022.3168664, IEEE Sensors
Journal
8 IEEE SENSORS JOURNAL, VOL. XX , NO. XX, MONTH X, XXXX
was required between peanut pods and lid for the proper closing
of the lid.
Measurement results of arbitrary cleaned unshelled peanut
loading are shown in Fig. 6 (orange bars). Variations of
dielectric properties measured for arbitrary peanuts loading
were smaller than those of the oriented peanuts. The measured
dielectric properties were close to those of the oriented pods
peanut measurement in the 5-peanut kernel direction rather than
those of the 6-peanut kernel direction. This implies the
influence of peanut orientation.
Small variation in the 2-direction measurements ((faces1,3),
(faces2,4) and (faces5,6)) were shown for the oriented and
arbitrary peanut loadings in Fig.6. This implies a reciprocal
property of the peanut sample (very low-loss material).
Measurements with arbitrary orientation of peanut pods show
much less variation with 4-direction (Face1,2,3,4) and 6-
direction measurements than for those obtained for oriented
pods. Therefore, arbitrary peanut loading was preferable to the
oriented peanut loading.
B. System Calibration for Moisture Content
Determination at 23
°
C
Ten samples of various moisture contents were removed from
the 4°C chamber and held at room temperature (23°C) for 12
hours before the measurements were taken. The samples were
arbitrarily loaded into the cubical sample container layer by
layer to avoid empty pockets near the corners. For each sample,
measurements in 6 directions were performed. Figure 7 shows
the dielectric properties measurements versus oven moisture
content (pods) [22] at indicated number of measurements, 6-
direction = faces 1 to 6, 4-direction = faces 1 to 4, 3-direction =
faces 1, 2 and 5, 2-direction = faces 1 and 2. As shown, the
dielectric properties measurements of 2, 3 and 4-directions
agreed closely with the 6-direction measurements. This means
that it was possible to reduce the number of directions for
measurements.
Moisture content determination algorithms were developed
to evaluate the performance of the different directional
measurements. From the literature [2], [18], [23], it is known
that the dielectric constant and loss factor are related to moisture
content (MC) and bulk density (ρ). Figure 8 show 𝜀󰆒
 and
𝜀󰆒󰆒 from 6-direction measurement as a function of pod
moisture content. High coefficients of determination were
obtained for both relationships. Therefore, moisture content can
be determined from the following relationships:
𝜀󰆒
 = 0.01437𝑀𝐶 + 0.9261 R² = 0.930, SE = 0.0349 (10)
𝜀󰆒󰆒 = 0.0408𝑀𝐶 0.0285 R² = 0.988, SE = 0.0131 (11)
The standard error of calibration (SEC) can be used for
evaluating the performance of equation (10) and (11), the
standard error of calibration was calculated as follows:
SEC =
 (∆𝑀)
 (12)
where N is number of samples, P is number of independent
variables in the regression equation with which the calibration
is performed (P = 2 for equation (10) and P = 1 for equation
(11)), ∆Mi is the difference between predicted value of moisture
content and that determined by a standard oven method for the
ith sample. It should be noted that the unit of SEC is the
prediction error in percentage moisture. The calculated SEC for
6 directions measurement for moisture content prediction were
0.624% and 0.305% for equation (10) and (11), respectively.
Cross validation was applied to the other directional
measurements. The standard error of performance (SEP) for
moisture content was calculated for the cross validation (data
set that were not used in the calibration) as follows:
SEP =
 (∆𝑀𝑀)
 (13)
where N is number of samples, ∆Mi is the difference between
predicted value of moisture content and that determined by a
standard oven method for the ith sample, and
𝑀 = (∆𝑀
 )/
𝑁. The unit of SEP and SEC is the same. The SEPs for 2, 3, and
(a)
(b)
Fig. 7. Measured dielectric properties of cleaned unshelled peanuts
(arbitrarily loaded) as a function of moisture content at 9.6 GHz and room
temperature (23°C) at indicated directions of measurement, 6-direction = all
faces, 4-directions = faces 1 to 4, 3-directions = faces 1, 2 and 5, 2-direction
= faces 1 and 2, (a) dielectric constant and (b) dielectric loss factor. - 6-
direction,
- 4-directions, - 3-directions and - 2-direction.
Authorized licensed use limited to: Prince of Songkla University provided by UniNet. Downloaded on May 30,2022 at 03:09:07 UTC from IEEE Xplore. Restrictions apply.
1530-437X (c) 2021 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2022.3168664, IEEE Sensors
Journal
8 IEEE SENSORS JOURNAL, VOL. XX , NO. XX, MONTH X, XXXX
4-directions were listed in Table I. The SEPs value for equation
(11) were lower than for equation (10) because of the higher
correlation for equation (11). For equation (10), validation of
SEP value for 3-direction measurement compared well with the
SEC for the 6-direction measurement because of the reciprocal
property. However, the SEPs for 4- and 2-direction
measurements were higher than the 6-direction measurement,
respectively. These values show the influence of orientation of
peanut pods in the cubical sample container.
C. System Calibration for Moisture Content
Determination with Temperature Compensation
In some instances, measurements were carried out in
environments where temperature fluctuates. Because dielectric
properties are temperature dependent [19] – [20], there is a need
for compensation for temperature effect when determining
moisture from these properties. Therefore, equations (10) and
(11) need to be established with temperature as a variable.
Figure 9 shows the results 𝜀󰆒
 and 𝜀󰆒󰆒 as functions of
moisture content and temperature for 6-direction measurement.
Regression analyses provided the following relationships:
𝜀󰆒
 = 𝑎 𝑀𝐶 + 𝑏 𝑇 + 𝑦 (14)
𝜀󰆒󰆒 = 𝑔 𝑀𝐶 + 𝑇 𝑦 (15)
where a, b, g, h, y0 and y1 are the fitting parameters for the
temperature-dependent moisture calibration functions. The
fitting parameters, regression coefficient and the SEC are listed
in Table II. High coefficients of determination were noted for
all measurements. The SECs for equation (15) were lower than
those for equation (14) because of the higher correlation
obtained for equation (15). The variation of the SEC from all
measurements demonstrates the influence of peanut orientation
in the cubical container on the measurements. The SECs for the
6-direction measurements by using equation (14) and (15) were
higher than equation (10) and (11), respectively, indicating
temperature-related errors. The SECs calculations for 6-, 4-, 3-
, and 2-direction measurements of equation (15) show a trend
similar to that observed in the results obtained in equation (11)
shown in Table I.
D. In-field validation measurements
The peanut moisture measurement system was taken to a
buying point in Georgia for in-field measurements. Several
peanut samples were measured. The relationship between the
oven pod moisture content and kernel moisture content was
reported previously [17]. The peanut samples with foreign
materials (about 4000 grams) were sampled from the trailer. A
splitting machine was used to divide the sample into 2 bags
(about 2000 grams each). Bag number 1 was supplied to the
grading room for foreign material and moisture content
determinations. Another bag (bag 2) was reserved, in case
mistakes occurred with bag 1. Therefore, the sample from bag
2 was used for testing the peanut moisture measurement system.
The sample was arbitrarily loaded into the cubical sample
Fig. 8. 𝜀󰆒
and 𝜀󰆒󰆒 for 6-
direction measurement (arbitrarily loaded)
as a function of moisture content at 9.6 GHz and room temperature
(23°C). -
𝜀
󰆒
 and -
𝜀
󰆒󰆒.
TABLE I
SEC. AND SEP. FOR MOISTURE CONTENT PREDICTION AT ROOM
TEMPERATURE (23°C) AND 9.6 GHZ
Directional
measurement
𝜺
󰆒
𝝆
𝟏𝟎𝟎𝟎
Eq. (10)
𝜺
󰆒󰆒
Eq. (11)
6-diection = all
faces
SEC 0.624% 0.305%
4-direction =
faces 1 to 4
SEP 0.622% 0.357%
3-direction =
faces 1,2 and 5
SEP 0.644% 0.304%
2-direction =
faces 1 and 2
SEP 0.617% 0.375%
(a)
(b)
Fig. 9. Calibration functions dependence on moisture content and
temperature at 9.6 GHz from 6-
direction measurement (arbitrarily loaded) of,
(a)
𝜀
󰆒
 and (b)
𝜀
󰆒󰆒.
Authorized licensed use limited to: Prince of Songkla University provided by UniNet. Downloaded on May 30,2022 at 03:09:07 UTC from IEEE Xplore. Restrictions apply.
1530-437X (c) 2021 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2022.3168664, IEEE Sensors
Journal
8 IEEE SENSORS JOURNAL, VOL. XX , NO. XX, MONTH X, XXXX
TABLE II
REGRESSION COEFFICIENTS, COEFFICIENTS OF DETERMINATION
AND SEC FOR CALIBRATION EQUATIONS (14) AND (15) AT 9.6
GHZ.
Directions
Regression
coefficients
Coefficients of
determination
Prediction error,
% MC.
A b y0 R
2
SE SEC
Equations (12)
6 0.0414 0.0047 0.8567 0.933 0.0328 0.782%
4 0.0421 0.0049 0.8552 0.939 0.0319 0.748%
3 0.0415 0.0047 0.8551 0.929 0.0339 0.806%
2 0.0423 0.0049 0.8514 0.935 0.0331 0.772%
G h y1 R
2
SE SEC
Equations (13)
6 0.0372 0.0028 -0.0547 0.963 0.0203 0.541%
4 0.0375 0.0028 -0.0603 0.958 0.0220 0.578%
3 0.0368 0.0028 -0.0498 0.962 0.0204 0.548%
2 0.0373 0.0029 -0.0600 0.954 0.0229 0.606%
TABLE III
SEP OF IN-FIELD MOISTURE CONTENT PREDICTION AT 9.6 GHZ
Direction measurements
𝜺
󰆒
𝝆
𝟏𝟎𝟎𝟎
Eq. (14)
(
𝜺
󰆒󰆒
)
Eq. (15)
4-direction = faces 1, 2, 3 and 4 SEP 1.166% 0.841%
2-direction = faces 1 and 3 SEP 1.493% 0.967%
2-direction = faces 2 and 4 SEP 1.149% 0.909%
Authorized licensed use limited to: Prince of Songkla University provided by UniNet. Downloaded on May 30,2022 at 03:09:07 UTC from IEEE Xplore. Restrictions apply.
1530-437X (c) 2021 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2022.3168664, IEEE Sensors
Journal
8 IEEE SENSORS JOURNAL, VOL. XX , NO. XX, MONTH X, XXXX
container layer by layer to avoid empty pockets near the
corners.
The cleaned unshelled peanuts were collected from bag 2 for
the measurement. In total, 74 peanut samples were measured.
The temperature of the peanut sample was measured with
an infrared thermometer (Fluke® 561) at different locations in
the cubical sample container before the microwave
measurement. The measured temperature of the sample ranged
from 15 °C to 32 °C. For the in-field measurement, 4-direction
measurement instead of 6-direction was performed to avoid
temperature drift during the measurements.
Figure 10 shows the in-field kernel moisture content
prediction (bag 2) by using equations (12) and (13) versus the
official meter kernel moisture prediction (peanut sample from
bag 1). It should be noted that the official moisture meter was a
capacitance-type meter utilizing measurements at 2 MHz. As
we can see the predicted kernel moisture content of the
proposed system follows the same trend with the official meter.
However, the official moisture meter tests were time consuming
because the peanuts must first be shelled. It is to be noted that
the microwave moisture meter was calibrated in the laboratory
against the moisture determined with oven-drying technique
and not the official meter.
To evaluate the influence of the peanut orientation, the SEP
was calculated with the moisture from the official meter taken
as reference. Values of the SEP as listed in Table III. The best
SEP was obtained from equation (15) with 4-direction
measurement. The SEP is slightly higher for 2-direction
measurements for both equations (14) and (15). This may imply
that there was influence of peanut orientation in the dielectric
properties measurements. The box plot of standard deviation of
the in-field dielectric properties measurements is plotted in Fig.
11. As can be seen the variation between 2-direction and 4-
direction measurements was very obvious. The variations in
Fig.11 agreed well with results shown in Fig. 6. This was
another confirmation of the influence of peanut pod orientation
in the in-field measurements.
IV. CONCLUSION
Influence of peanut sample orientation on the dielectric
properties and moisture content determination has been
investigated. A free-space dielectric measurement system
operating at 9.6 GHz was designed for measuring cleaned
unshelled peanuts in a cubical sample container (6 faces).
Oriented and arbitrary loadings of peanut pods were
investigated for purpose of comparison. The dielectric
properties measurement was not related to the orientation of
(a)
(b)
Fig. 10. In-field kernel moisture content prediction by using 4-direction
measurement (arbitrarily loaded) at 9.6 GHz plotted against kernel moisture
content obtained with an official moisture meter, (a) equation (14) and (b)
equation (15). -predicted and -ideal line.
(a)
(b)
Fig. 11. Box plot of standard deviation of the in-field measurement of, (a) 𝜀󰆒
and (b)
𝜀
󰆒󰆒.
Authorized licensed use limited to: Prince of Songkla University provided by UniNet. Downloaded on May 30,2022 at 03:09:07 UTC from IEEE Xplore. Restrictions apply.
1530-437X (c) 2021 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2022.3168664, IEEE Sensors
Journal
8 IEEE SENSORS JOURNAL, VOL. XX , NO. XX, MONTH X, XXXX
pods only but also to the number of kernels interacting with the
electric field. Prediction algorithms of moisture content based
on dielectric constant 𝜀󰆒
 and dielectric loss factor 𝜀󰆒󰆒)
were separately developed for arbitrary loading (practical). The
measurement results showed that moisture content had a high
correlation with 𝜀󰆒󰆒 rather than 𝜀󰆒
 which is close
agreement with [18]. For moisture content predicted with 𝜀󰆒󰆒,
the average dielectric measurement for the 6 directions were
more representative of the peanut sample and therefore
provided better standard error of calibration for moisture
content prediction compared to 4- and 2-direction
measurements. The standard error of performance (compared to
the official meter) was about 0.84% and 1.17% for in-field
moisture content prediction by using the 𝜀󰆒󰆒 and 𝜀󰆒
,
respectively. By using the 𝜀󰆒󰆒, the SEP for the 2-direction
measurement was higher than that for the 4-direction
measurements. The cost of the proposed sensor operating at 9.6
GHz is less than $7000 which is about 7 to 10 times less than
the cost of a vector network analyzer operating at the same
frequency. By using the average value of multi-direction
measurements of dielectric properties one can reduce the
influence of peanut orientation which is useful in developing
low-cost, accurate, and robust microwave instruments for
peanut moisture sensing.
ACKNOWLEDGMENT
This work was supported in part by the Georgia Federal-
State Inspection Service/Ga Institute of Technology under
Grant 58-6040-7-018 and in part by the U.S. Department of
Agriculture, Agricultural Research Service under contract for
service with the Oak Ridge Institute for Science and Education,
managed by Oak Ridge Associate University under the ARS
Research Participation The authors would like to thank Dr.
Stuart O. Nelson for his insightful comments and suggestions.
The authors would like to thank Dr. Micah A. Lewis, an
agriculture engineer at USDA-ARS in Athens, Georgia and Mr.
Ron Dozier, buying point manager of Peanut Producers, LLC
at Bartow, Georgia, for supporting the in-field measurements.
REFERENCES
[1] S. Trabelsi, S. O. Nelson, “Microwave dielectric properties of shelled and
unshelled peanuts,” Transactions of the ASAE, vol. 47, pp. 1215, 2004.
[2] S. Trabelsi, S. O. Nelson, “Microwave sensing technique for
nondestructive determination of bulk density and moisture content in
unshelled and shelled peanuts,” Transactions of the ASABE, vol. 49,
pp.1563-1568, 2006.
[3] S. Trabelsi, S. O. Nelson, “Nondestructive sensing of bulk density and
moisture content in shelled peanuts from microwave permittivity
measurements,” Food Control, vol,17, pp.304-311, 2006.
[4] S. Trabelsi, M. A. Lewis, S. O. Nelson, “Microwave meter for rapid,
nondestructive determination of in-shell peanut kernel moisture content
from dielectric measurements on cleaned and uncleaned pod samples,”
ASABE Paper No. 121338109, American Society of Agricultural and
Biological Engineers. St. Joseph, MI. 2012.
[5] M. A. Lewis, S. Trabelsi, S. O. Nelson, “Real-time monitoring of peanut
drying parameters in semitrailers,” Drying Technology, vol.35, pp.747-
753, 2017.
[6] M. A. Lewis, S. Trabelsi, S. O. Nelson, “Estimating energy costs of
nonbeneficial dryer operation by using a peanut drying monitoring
system,” Applied Engineering in Agriculture, vol.34, pp. 491-496, 2018.
[7] A. M. Paz, S. Trabelsi, S. O. Nelson, “Measurement of the dielectric
properties of sawdust between 0.5 and 15 GHz,” IEEE Trans. Instrum.
Meas., vol.60, no.10, pp.3384-3390, 2011.
[8] F. Jafari, K. Khalid, Y. J. Hassan, A. Zulkifly, N. S. M. Salim, “Variation
of microwave dielectric properties in the glucose biosensor system,”
International Journal of Food Properties, vol.18, no.7, pp.1428-1433,
2015.
[9] S. Zhang, L. Zhou, B. Ling, S. Wang, “Dielectric properties of peanut
kernels associated with microwave and radio frequency drying,"
Biosystems Engineering, vol.145, pp.108-117, 2016.
[10] K.S. Rucker, C. K. Kvien, K. Calhoun, R. J. Henning, P. E. Koehler, S. R.
Ghate C. C. Holbrook, “Sorting peanuts by pod density to improve quality
and kernel maturity distribution and to reduce aflatoxin,” Peanut Science,
vol.21, no.2, pp.147-152, 1994.
[11] E. Carter, P. Troy, D. Rowland, B. Tillman, K. Wynne, K. Pierre, M.
Mulvaney, “Methods to Evaluate Peanut Maturity for Optimal Seed
Quality and Yield,” EDIS, vol.2016, no.7. 2016
[12] C. L. Butts, C. V. K. Kandala, R. B. Sorensen, M.C. Lamb, “In-shell Bulk
Density as an Estimator of Farmers Stock Grade Factors,” Peanut Science,
vol.34, no.2, pp.135-141., 2007.
[13] J. Roelvink, S. Trabelsi, S. O. Nelson, “Measuring the complex
permittivity tensor of uniaxial biological materials with coplanar
waveguide transmission line,” IEEE Microwave and Wireless Components
Letters, vol.24, no.11, pp.811-813, 2014.
[14] T.C. Mai, S. Razafindratsima, Z. M. Sbartaï, F. Demontoux, F. Bos,“Non-
destructive evaluation of moisture content of wood material at GPR
frequency,” Construction and Building Materials, vol.77, pp.213-217,
2015.
[15] R. Martínez-Sala, I. Rodríguez-Abad, R. D. Barra, R. Capuz-Lladró,
“Assessment of the dielectric anisotropy in timber using the nondestructive
GPR technique,” Construction and Building Materials, vol.38, pp.903-
911, 2013.
[16] Nelson S. O., Measurement of microwave dielectric properties of
particulate materials,” Journal of Food Engineering, vol.21, pp.365-384,
1994.
[17] S. Trabelsi, M. A. Lewis, S. O. Nelson, “Microwave moisture meter for in-
shell peanut kernels,” Food Control, vol.66, pp.283-290, 2016.
[18] S. Julrat, S. Trabelsi, “Measuring dielectric properties for sensing foreign
material in peanuts. IEEE Sensors Journal, vol.19, pp.1756-1766, 2019.
[19] S. Trabelsi, A. Kraszewski, S. O. Nelson S. O., “A microwave method for
on-line determination of bulk density and moisture content of particulate
materials,” IEEE Trans. Instrum. Meas., vol.47, pp.127-132.1998
[20] S. Trabelsi, A. Kraszewski, S. O. Nelson S. O., “New density-independent
calibration function for microwave sensing of moisture content in
particulate materials,” IEEE Trans. Instrum. Meas., vol.47, pp.613-622,
1998.
[21] NGEN. Available from: https://colorfabb.com/filaments/co-polyesters-
filaments/ngen, [assessed March 10 2020], 2014.
[22] ASAE, ASAE S401.1 Moisture measurement - Peanuts, ASAE Standards
2002. American Society of Agricultural Engineers, St. Joseph, Michigan,
pp.600 – 601.2002.
[23] S. Trabelsi, “New calibration algorithms for dielectric-based microwave
moisture sensors," IEEE Sensors Letters, vol.1, pp.1-4, 2017.
Authorized licensed use limited to: Prince of Songkla University provided by UniNet. Downloaded on May 30,2022 at 03:09:07 UTC from IEEE Xplore. Restrictions apply.
Article
Full-text available
Within the framework of the recent agri-food technological advancement, the design and validation of a methodology for the water content estimation in the Carasau bread manufacturing process is herein presented. Following a thorough evaluation of the dough dielectric properties, a suitable antenna layout has been selected, pointing out the advantages in the choice of a contactless narrow-band antenna in comparison to wide-band and dual-band ones. The presented simulated results are then validated using a prototype sensor and an ad hoc measurement system to confirm the antenna ability to discriminate among doughs with different water content. In addition, an accurate analysis of possible sources of misinterpretation of the results is presented. INDEX TERMS dielectric properties, moisture sensor, Carasau bread, patch antenna, food engineering
Article
The moisture content (MC) is a critical factor influencing the quality of wheat. Presently, existing MC sensing methods do not provide a swift and precise determination of the wheat’s MC. This study develops a non-contact microwave sensing approach for the accurate prediction of wheat’s MC. To be more specific, the system employs microwave signals for transmitting wheat samples, enabling the measurement and calculation of the dielectric constant ε′ and dielectric loss ε″. Additionally, it takes into account the natural packing density of the sample, thereby enhancing detection accuracy. Meanwhile, the system receives the signal attenuation difference ΔA and phase shift value Δφ corresponding to each frequency component, MC levels, and bulk density. Experimental results demonstrate that the wheat’s dielectric characteristic parameters monotonically increase with the MC level in the selected frequency range. According to the data feature, this study proposes an Adaptive Boosting for Support Vector Regression (AdaBoost-SVR) moisture prediction model to consider dielectric characteristic parameters and bulk density comprehensively. These results are also compared with Support Vector Regression (SVR) regression and the linear regression predictions using only one dielectric parameter. The comparison denotes that the best prediction performance is obtained using the three feature inputs of dielectric characteristic indicators and the bulk density (ε′ & ε″,ρ), achieving an root mean square error ( RMSE ) of 0.2201 and an coefficient of determination ( R 2 ) of 0.991. This study strongly suggests that a combination of multiple indicators is inevitable when predicting MC in stockpiled wheat or future silos.
Article
Full-text available
A new method for simultaneous and independent on-line determination of bulk density and moisture content in particulate materials by measurements of the relative complex permittivity is proposed. The bulk density is determined, based on a representation in the complex plane of relative complex permittivity normalized to bulk density. For moisture content determination, a new density-independent function, exclusively dependent on the dielectric properties, is used. Results obtained from measurements on wheat over broad ranges of microwave frequencies, temperatures, densities, and moisture contents are presented.
Article
Full-text available
Free-space dielectric properties measurements at 10 GHz with four pairs of low-cost printed Yagi–Uda antennas are proposed for detecting foreign material in unshelled peanuts, including sticks, peanut shells, peanut raisins, and stones. The four-element array design of the Yagi–Uda antenna allowed better averaging over the entire sample volume and improved the measurement resolution. The antenna gain is about 9.7 dBi. The beam width of each antenna is about 18° and 38° for the E-plane and H-plane at 3 dB, respectively. The attenuation and phase shift were measured for each pair of antennas by controlling automatically the microwave switches. Measurement repeatability of the system was confirmed by using different thicknesses of polyethylene dielectric slab samples. The standard deviation was less than 0.5 dB for the attenuation and less than 5.0° for phase measurements. Sensitivity of each antenna pair for unshelled peanuts and different foreign materials, each loaded in a small volume (4.4 cm ×4.4\times4.4 cm ×12.1\times12.1 cm), was tested by using a partitioned sample container. It was shown that a difference in attenuation was more than 1.5 dB for different types of foreign materials and also for unshelled peanut samples with a 1% moisture content difference. Prediction of moisture content in cleaned unshelled peanut samples agreed well with the standard oven-drying method (standard error of calibration = 0.51%). The dielectric properties of foreign materials mixed with cleaned unshelled peanut samples were measured through the attenuation and phase shift. The measurement repeatability of each antenna and all four pairs of antennas is presented.
Article
Full-text available
Knowledge of kernel moisture content during peanut drying is important to ensure that the bed of peanuts is dried appropriately. However, the lack of a commercially available, industry-accepted solution for real-time kernel moisture content determination during peanut drying makes its detection cumbersome and laborious. Samples of unshelled peanuts are extracted from the semitrailer by an operator periodically, and the samples have to be cleaned and shelled to determine kernel moisture content with the official meter. A peanut drying monitoring system that includes a microwave kernel moisture sensor, developed within the USDA ARS, provides a means for monitoring in-shell kernel moisture content in real-time. The system determines kernel moisture content with a standard error of prediction (SEP) of 0.55% moisture content when compared to the reference oven-drying method. During recent peanut harvest seasons, peanut drying monitoring systems were placed in 13.7-m (45-ft) drying semitrailers, one 3 m (10 ft) from the front of the trailer and the other 3 m (10 ft) from the back of the trailer. As the peanuts dried, pod and kernel moisture content, temperature of the drying peanuts, temperature and relative humidity of the air exhausted from the peanuts, and temperature and relative humidity of the air being blown into the peanuts were measured every 12 seconds. The continuous data, provided by the monitoring systems, were useful in observing the loss of moisture by the peanuts throughout drying. The data also revealed periods of at least 3 hours during which dryer operation did not result in loss of moisture from the peanuts; thus, identifying nonbeneficial dryer operation. Such periods cause a peanut buying point to accumulate unnecessary expenses for propane and/or electric energy which can total up to $3,250 annually for an average-size buying point. © 2018 American Society of Agricultural and Biological Engineers.
Article
Full-text available
New calibration algorithms for determining moisture content in granular and particulate materials from measurement of the dielectric properties at a single microwave frequency are proposed. The algorithms are based on identifying empirically correlations between the dielectric properties and the partial densities of water and dry matter of a given material. Both situations where the bulk density is known or unknown were considered. Validation of the performance of the new algorithms are presented for grain and seed samples for which the dielectric properties were measured by free-space transmission at 10 GHz and 24 oC.
Article
Full-text available
Knowledge of peanut drying parameters, such as temperature and relative humidity of the ambient air, temperature and relative humidity of the air being blown into the peanuts and kernel moisture content, is essential in managing the dryer for optimal drying rate. The optimal drying rate is required to preserve quality and desired flavor. In the current peanut drying process, such parameters are elusive in real-time and are either not measured or only measured periodically by an operator. A peanut drying monitoring system, controlled by an embedded microcontroller and consisting of relative humidity and temperature sensors and a microwave peanut moisture sensor, was developed to monitor drying parameters in real-time. It was deployed during the 2014 peanut harvest season at a peanut buying point in central Georgia, USA. It was placed in 45-ft (13.7-m) drying semitrailers to monitor in-shell kernel moisture content, temperature of the drying peanuts, temperature and relative humidity of the exhaust air from the peanuts and relative humidity of the air being blown into the peanuts in real-time. In-shell kernel moisture content was determined with a standard error of performance of 0.55% moisture content when compared to the reference oven-drying method. Data from drying parameters were time-stamped and stored on a CompactFlash card every 12 seconds and were used to assess the efficiency of dryer control settings. Ambient air conditions were measured by an on-site weather station. Results of the study support the value of such a monitoring system and show that implementation of the system for dryer control has the potential for saving a buying station, in the current economical context, as much as $22,000 annually in costs of electric energy and propane.
Article
To develop advanced drying methods based on microwave (MW) and radio frequency (RF) heating, knowledge of dielectric properties is essential for understanding the interaction between electromagnetic fields and peanuts. In this study, dielectric properties of peanut kernels were measured between 10 and 4500 MHz using an open-ended coaxial-line probe and network analyser at temperatures between 25 and 85 °C and moisture contents between 10% and 30% on a wet basis (w.b.). The results showed that both dielectric constant and loss factor of peanut kernels decreased sharply with increasing frequency over RF range (10-300 MHz), but gradually over the MW range (300-4500 MHz). Both dielectric constant and loss factor increased with increasing moisture content and temperature. The rate of increase was greater at higher temperature and moisture levels than at their lower levels. Penetration depth decreased with increasing frequency, moisture constant, and temperature. The measured dielectric properties were finally applied to determine the temperature profiles of RF heated samples under three moisture levels using experiment and simulation. This study on dielectric properties may provide useful guidelines in developing effective dielectric drying methods with a suitable drying thickness for peanuts.
Article
Glucose biosensor is generally based on the reaction between glucose and glucose oxidase, which produces gluconic acid and hydrogen peroxide. The gluconic acid is a conductive material, while hydrogen peroxide has polar molecules. This article examines the changes of dielectric properties due to the conductive loss below 2 GHz and dipole orientation of above that frequency of this reaction. The difference between the dielectric loss of glucose solution and the dielectric loss of glucose-enzyme reaction can be related to the glucose concentration in the samples, such as orange juice, black grape juice, sugarcane juice, and sapodilla juice. A good sensitivity to these differences due to the effect of ionic conductivity and dipole orientation was found at 1 and 16.44 GHz, respectively. The minimum detection limit of glucose concentration in the proposed technique was about 0.01 M (0.20 g/100 ml) with an optimum reaction ratio of about 1:1 between the enzyme solution and the glucose solution. This technique could benefit the future development of microwave biosensor by which both ionic conductivity and dipole effects can be considered simultaneously.
Article
A simple and accurate technique is described for measuring the uniaxial permittivity tensor of biological materials with a coplanar waveguide transmission-line configuration. Permittivity tensor results are presented for several chicken and beef fresh meat samples at 2.45 GHz.