An ASABE Meeting Presentation Paper Number: 076215
Phase angle and Impedance Measurements for Nondestructive Moisture Content Determination of InShell Peanuts Using a Cylindrical Sample Holder. Chari V. Kandala National Peanut Research Laboratory, ARS, USDA, Dawson, Georgia.
[email protected]
Christopher L. Butts National Peanut Research Laboratory, ARS, USDA, Dawson, Georgia.
[email protected] Written for presentation at the 2007 ASABE Annual International Meeting Sponsored by ASABE Minneapolis Convention Center Minneapolis, Minnesota 17 - 20 June 2007 Abstract. Two parallel-plate electrodes were mounted inside a cylinder, made of a non-conducting material. The space between the plates was filled with peanut pods and the capacitance and phase angle of this system was measured with a prototype low-cost impedance meter, designed for this purpose. Measurements were made at frequencies of 1 and 5 MHz on peanut pod (in-shell) samples of known moisture content between 7 and 17% (wet basis). Using these measured values an empirical equation was developed from which the average moisture content of a peanut pod sample could be calculated from its impedance values. The calculated moisture content values were compared with their air-oven values and were found to be within 1% of the air-oven values for over 95% of the samples tested. The size of the sample used was about 100 g and no shelling or cleaning of the sample was involved. This method is rapid and nondestructive. Keywords: Parallel-plate, Impedance, Peanut pods, Phase angle, Capacitance. The authors are solely responsible for the content of this technical presentation. The technical presentation does not necessarily reflect the official position of the American Society of Agricultural and Biological Engineers (ASABE), and its printing and distribution does not constitute an endorsement of views which may be expressed. Technical presentations are not subject to the formal peer review process by ASABE editorial committees; therefore, they are not to be presented as refereed publications. Citation of this work should state that it is from an ASABE meeting paper. EXAMPLE: Author's Last Name, Initials. 2007. Title of Presentation. ASABE Paper No. 07xxxx. St. Joseph, Mich.: ASABE. For information about securing permission to reprint or reproduce a technical presentation, please contact ASABE at
[email protected] or 269-429-0300 (2950 Niles Road, St. Joseph, MI 49085-9659 USA).
Introduction Several methods are in vogue for determining the moisture content of food materials such as nuts and grain. These include chemical, analytical and electrical methods. While the chemical methods are generally destructive in nature, the later two methods are nondestructive for a considerable variety of food products. The conductivity and capacitance methods are used extensively in the development of several commercially available moisture meters. Moisture content (mc) measurement and control are very important aspects in the peanut industry. Peanuts (Arachis hypogaea L.) after harvesting have to be dried to less than 10.5% mc for grading and sale (USDA, Update 2000). They can also be stored at this mc level when provided with proper ventilation. A device with which the mc of peanuts could be measured, rapidly and nondestructively, is useful at several stages of their drying, storage and processing. Presently available commercial meters for measuring the mc of peanuts are mostly of the capacitance type. However, peanuts have to be shelled and cleaned before their mc could be measured by these instruments. This process is time consuming and needs additional labor and once the peanuts are shelled the samples are usually discarded resulting in the loss of considerable quantities of edible peanuts. It would be very useful if the peanut kernel mc could be estimated from physical measurements made on the in-shell peanut (pod) itself. This process would not only eliminate the need for shelling and cleaning of peanut samples being tested, resulting in considerable savings in time and labor, but also prevents the destruction of the test samples. Measurements of capacitance, phase angle and/or dissipation factor of a parallel-plate capacitor with a few grain kernels between the plates has shown promise for nondestructively and rapidly measuring the moisture content of single or small samples of corn, wheat or peanut kernels (S.O. Nelson et al., 1992). Earlier experiments (Butts et al., 2004) showed a good correlation between the peanut pod and the kernel moistures as measured by the standard oven method. Thus if the peanut pod moisture can be measured using the RF Impedance method then the kernel mc may be estimated to an acceptable accuracy. In this article an electrical instrument that measures the phase angle and impedance of a parallel-plate system, fitted inside a nonconducting cylindrical tube, for obtaining mc of in-shell peanut kernels is described. The measured moisture content values (in the range of 7 to 17%) were compared with the mc values obtained by the standard air-oven method. This method can be used for similar mc measurements for grain such as corn and wheat.
Materials and Methods THEORETICAL CONSIDERATIONS: The dependence of dielectric constant on moisture of yellow–dent field corn in the frequency range of 1 MHz to 11 GHz was earlier investigated (Nelson, 1978). The variation of the dielectric constant was more pronounced at 1 and 5 MHz, and it was found earlier that these variations could be used as an useful parameter in estimating the mc of single corn kernels (Kandala et al., 1989) and single peanut kernels (Kandala and Nelson, 1990). The capacitance of a parallel-plate capacitor with plate area A and plate separation d, filled with a dielectric material, at a frequency f1 is given by:
2
C1 = εr1ε0A/d
[1]
And at a frequency f2 is given by: C2 = εr2ε0A/d
[2]
where εr1 and εr2 are the dielectric constants of the filled material at the two frequencies and ε0 is the permittivity of the free space (8.854x10-12 farad/m). Using these two equations we can write the difference in the dielectric constants as: εr1 - εr2 = (C1 - C2) d/(ε0A)
[3]
It was found earlier that (C1 - C2) was a good estimate of the variation in the dielectric constant at these frequencies and could be used to estimate the mc of the peanut sample filling the space between the plates (Kandala, 2004). In the case of a lossy dielectric sample such as a moist peanut sample held between two parallel-plates forming a capacitor, the capacitance and dissipation factor of such a capacitor are used in the medium frequency range to determine the complex permittivity of the sample. Any lossy dielectric sample can be modeled as an ideal capacitor in series with a resistor. The dissipation factor D, is related to the phase angle θ as tan θ = (1/D)
[4]
The impedance Z of this system is a complex quantity and consists of the real (R) and imaginary (X) parts. The dissipation factor is related to the complex impedance and can be written as D = R/X
[5]
From equations (3), (4) and (5) it can be seen that the complex permittivity of a lossy material is a function of the capacitance for a capacitor of fixed plate area and separation and the dissipation factor is a function of phase angle and complex impedance. Thus from measured values of capacitance, phase angle and complex impedance of a nut or grain sample placed between the plates of a parallel-plate capacitor it should be possible to estimate the mc of the sample. IMPEDANCE MEASURING CIRCUIT An electronic circuit that measures the two required parameters, impedance and phase angle at 1 and 5 MHz, of a parallel-plate electrode system is described below. Incorporating this circuit a prototype instrument was built and an attempt was made to calibrate it to estimate the mc of peanut samples filled between a set of parallel-plate electrodes fitted inside a non-conducting cylindrical tube. The two frequencies 1 and 5 MHz are generated by two crystal oscillators as shown in the block diagram (Fig. 1). These signals are applied to the parallel-plate electrode system alternately by switching through a multiplexer. Initially at 1.0 MHz the current, flowing through this system with an impedance Z, is fed into an op-amp. The same current would flow through the feed-back resistor Rr. The output voltage of the op-amp and the original 1 MHz signal from the oscillator are rectified and measured as em1 and er1 respectively. The current through Z is calculated as em1/Rr and the magnitude of the impedance of the parallel-plate system with the peanut kernel between them is obtained as |Z1| = Rr (er1/em1).
3
The phase angle at 1MHz is determined by comparing the signal emerging out of the op-amp with that of the original signal, using a phase detector. However, the phase detector can compare signals of the same amplitude only. To keep the amplitude of the output signal from the op-amp constant and equal to the amplitude of the original signal, a comparator is used. 1 MHZ
O S C IL L A T O R
F IL T E R
ATTENUATO R
CO M PARATO R
R E C T IF IE R
e r1
PHASE DETECTO R
ep1
F IL T E R
Rr
Z
MULTIPLEXER
OP AMP
R E C T IF I E R
em 1
R E C T IF I E R
e r2
GROUND
Rr
OP AMP
CO M PARATO R
F IL T E R
GROUND
ATTENUATO R
O S C IL L A T O R
F IL T E R
PHASE DETECTO R
ep2
R E C T IF IE R
em 2
5 MHZ
Fig. 1. Block diagram of the electronic circuit to measure magnitude and phase angle at 1 and 5MHz. The comparator would output a square wave, and a filter is used to convert it to a sine wave. The original signal from the oscillator is attenuated to the same amplitude as this signal and the two signals are fed into the phase detector. The phase detector compares the two signals and gives an output voltage ep1 proportional to the phase angle θ1 between the two. The computer then switches the multiplexer to allow the 5 MHz to pass through the parallelplate system. The signals are processed through a circuit similar to the 1 MHz circuit but with a range resistor of a different value. The impedance magnitude |Z2| and the phase angle θ2 are determined at this frequency as was done for 1 MHz. From the values of Z and θ the real and imaginary parts of the impedance R and X, at each frequency is calculated as R = |Z| Cos θ and X = |Z| Sin θ. The values of capacitance C, of the parallel-plate system with the peanut sample between them are given as C = -1/2πfX
[6]
The power supply consists of two 12V rechargeable lead-acid batteries from which the voltages, required to operate the circuits, were derived. A Fujitsu P Series Life Book computer, S6210 Model (Fig. 3) was used to register data from the system, compute the calibration constants and calculate the moisture content.
4
MEASURING SYSTEM
1
2
3
Fig. 2. Measuring system: 1. Computer, 2. Impedance meter, 3. Cylindrical tube with sensors. A cylindrical tube (Fig. 2) made of polyethylene, fitted with two sets of parallel-plate electrodes, served as the sample holder. The tube is 305-mm- long, 62- mm-diameter and has a wall thickness of 5-mm. The electrode assembly consists of two pairs of rectangular aluminum plates, 88-mm-long and 38-mm-wide. The electrode pairs were connected together to form two parallel capacitors and they are glued into the inner walls of the cylinder, as shown at in Fig.2, at a distance of 38-mm from the ends. This cylinder sits on the top of a rectangular polyethylene box and in a circular hole, centered about 35-mm from the front side of the box. This box is provided with a polyethylene drawer that can slide in and out of the box when pulled, holding the non-conducting handle that is provided. The upper side of this drawer was covered up to 75mm from the font end, using a polyethylene plate. When the drawer is pushed all the way into the box, this plate would cover the hole in which the cylinder is sitting, and prevents any peanuts placed in the cylinder from dropping into the drawer. Except for the electrodes, no metal parts were used in the assembly of the electrode system or the sample collecting system to prevent any interaction with the RF signal used in the measurements. With the drawer pushed all the way in, the cylinder was filled with the peanut samples and the impedance measurements were taken. After the completion of the measurements, the drawer is pulled out, till the plate moves from under the circular hole enabling the peanuts to fall into the drawer. The drawer was emptied before another sample was placed in the cylinder for measurement. With the peanut pods occupying the space between the electrodes, the analyzer measured the capacitance, dissipation factor and phase angle of this electrode system at 1 and 5 MHz, and a computer controlled and collected the data. PEANUT SAMPLES Peanuts of the Georgia Green cultivar harvested in 2005 and stored at 4º C at the National Peanut Laboratory were used for these studies. The initial moisture content (mc) of these peanuts was about 6% as measured with a Dickey-john Grain Analysis Computer 2100 (Dickeyjohn, Inc., Auburn, Illinois). From these, four sub-lots, called the calibration lots, were placed in quart jars. Leaving one jar of peanuts at the original moisture level, appropriate quantities of water were added to samples in the other jars to raise their moisture levels to obtain four moisture levels ranging from 6% to 19%. Similarly from the original lot another four sub-lots called the validation lots were separated and placed in four jars. The peanuts in these jars were
5
conditioned to different moisture levels but in the same moisture range (6% to 19%) by adding appropriate amounts of water to samples in each jar. Thus four calibration levels with nominal mc values of 6%, 10%, 14% and 19% and four validation levels with nominal mc values of 8.5%, 10.5%, 12.5% and 16.5% were generated. All the jars were sealed and held at 4 0C to equilibrate. The jars were removed from cold storage and the kernels were allowed to reach room temperature in the jars before the measurements were made. MEASUEREMENTS From each of the sub-lots three samples each weighing about 100 g were placed in small aluminum containers and their mc was determined by the standard air-oven method (ASAE 2000). The bulk values of the moisture contents of the eight sub-levels were 6.66%, 8.72%, 10.48%, 10.64%, 12.43%, 14.15%, 16.63%, 18.39%. These were separated into two groups, one called the calibration group and the other validation group. The calibration group consisted of samples with bulk mc levels of 6.66%, 10.48%, 14.15%, 18.39% and the validation group had bulk mc levels of 8.72%, 10.64%, 12.43% and 16.63%. Measurements were made on 30 samples from each moisture level. Peanut samples were transferred from the jars into the cylindrical electrode system, till the space between the two plates of the cylinder is completely filled. The cylinder accommodated about 80 to 100 gm of peanut pods. The room temperature during the measurements varied from 21ºC to 23 ºC. Peanut pods from the jar with nominal mc level of 6% were transferred into the cylinder with the drawer sitting fully inside the box, till the space between the two lower plates of the cylinder is filled. In this position measurements of impedance (z) and phase angle (θ) were taken on the moisture meter at 1 and 5 MHz. The sample was then collected in the drawer by gently pulling it out and tapping on the cylinder for the peanuts to drop down. The drawer was emptied and reset in the box. The procedure was repeated on rest of the 29 samples in this mc level and for all other mc levels in the calibration and validation groups.
Results and Discussion From the measured values of impedance and phase angle the capacitance value for each sample was obtained using Eq. 6. Using the capacitance value, measured values of impedance and phase angle, and the oven determined mc value of the samples in the calibration lots, a semi-empirical equation was developed . This equation had the following form: mc = A0 + A1 (∆θ) + A2 (∆Z) + A3 (∆C) + A4 (∆θ)2 + A5 (∆Z)2 + A6 (∆C)2
[7]
where ∆θ, ∆Z and ∆C are the differences between the phase angle, impedance and capacitance at 1 and 5 MHz. The values of the constants in Eq. (7) determined with the help of SAS procedures (SAS, 2001) for regression analysis were: A0 = -84.183, A1 = -5.873, A2 = 69.227, A3 = 9.256, A4 = -0.064, A5= - 85.309, A6= 0.480 The coefficient of determination was 0.97. Using these values in Eq. 7, the mc of each pod sample in the four calibration lots was calculated, averaged over the 30 samples in each mc level, and the results are shown in table 1 along with the bulk mc values obtained by the air-
6
oven method. Also shown are the differences between the calculated and the oven values and the standard deviations of the calculated values. It can be seen that the mean calculated values Table 1.Comparison of oven and calculated mc values for 4 Calibration lots (Average of 30 pod samples in each level) Oven bulk
Moisture Content (%)
% mc value 6.66
Difference
From Eq.(7)
(%)
6.93 ± 0.37*
-0.27
*
0.40
10.48
10.08 ± 0.43
14.15
14.66 ± 0.30*
-0.51
18.39
18.08 ± 1.02*
0.31
*
Standard deviation
agree very closely with the bulk oven values. An R2 value of 0.97, relatively small differences in the calculated and the bulk oven values, and no visible outliers suggested the suitability of the model for mc predictions. The mc values of each of the pod samples in the four validation lots were similarly calculated averaged over 30 pod samples and the results are shown in Table 2. The mc of a 200g sample determined by the standard oven method is estimated to yield precisions (error/true mean) at the 95% confidence level of ± 6% for in-shell peanuts in the 20% moisture range (ASAE, 2002). On a similar note the accuracies of the values calculated using Eq. (7) were expected to fall within 1% of the average values obtained by the standard oven method. The predictability as shown in the last column is the percentage of pod samples for which mc was predicted with in Table 2. Comparison of Oven and calculated mc values for 4 validation lots (Average of 30 pod samples in each level) Nominal
Bulk oven
Moisture Content (%)
Difference
Predictability
% mc value
% mc value
From Eq. (7)
(%)
(%)
8.5
8.72
9.08 ± 0.40
-0.36
100
10.5
10.64
10.45 ± 0.42
0.19
100
12.5
12.43
12.21 ± 0.34
0.22
100
16.5
16.63
16.54 ± 0.80
0.09
83
1% of the air-oven value in each moisture level. The predictability was at least 80 % at any level and averaged over 95 % over all the moisture levels in the validation group. A bar graph comparing the mc values determined by the oven and the impedance methods at the four moisture levels in the validation group is shown in Fig. 3.
7
Calculated %M
20 Oven %M Pred. %M
15 10 5 0 8.5
10.5
12.5
16.5
Nominal %mc
Fig. 3. Comparison of mc values determined by the air-oven and impedance measurement methods It could be seen from Fig. (4) that the average values predicted by the impedance method agree well with the standard air-oven values. The samples used were not of the same weight or volume and consisted of peanut pods of different sizes. This did not seem to have affected the predictability.
Conclusions By measuring capacitance and phase angle of a parallel-plate system fitted inside a nonconducting cylindrical tube the average moisture content of about 100g of in-shell peanuts could be predicted rapidly and nondestructively. The moisture range of the peanuts tested was between 6% and 18% and the predicted mc values were within 1% of the air-oven bulk values for over 95% of the samples tested from the 2005 harvest. Ability to estimate the mc content without the need to shell and clean the peanuts would save considerable amount of time and labor during their drying process and would save the destruction of large quantities of edible peanuts. The RF Impedance measurement method provides a basis for the development of a practical instrument that can measure mc of in-shell peanuts rapidly and nondestructively.
References ASAE.Standards, 49th edition. 2002. S410.1: Moisture Measurement – Peanuts. St. Joseph, Mich.: ASABE. Butts, C.L., J.I. Davidson, Jr., M.C. Lamb, C.V. Kandala, and J.M. Troeger. 2004. “ Estimating drying time for a stock peanut curing decision support system”. Trans. ASAE, 47(3): 925-932. Kandala, C.V.K., S.O. Nelson and K.C. Lawrence. 1989. Nondestructive electrical measurement of moisture content in single kernels of corn. J. Agric. Eng. Res. 44:125-132 Kandala, C.V.K., and S.O. Nelson. 1990. Measurement of moisture content in single kernels of peanuts: A nondestructive electrical method. Trans. ASAE, 33(2):567-572. Kandala, C.V.K. 2004. “Moisture determination in single peanut pods by complex RF Impedance measurement,” IEEE Trans. Instrum. Meas. 53(6): 1493-1496.
8
Nelson, S.O., C.V.K. Kandala and K.C. Lawrence. 1992.“Moisture determination in single grain kernels and nuts by RF Impedance measurements”. IEEE Trans. Instrum. Meas. 41(6): 1027-1031. Nelson, S.O. 1978. “Frequency and moisture dependence of the dielectric properties of highmoisture corn. J. Microwave Power, 13(2): 213-218. SAS. 2001. SAS User’s guide. Version 8. 2001. Carey, NC. SAS Institute, Inc. USDA. 2000. AMS farmers stock peanuts inspection instructions. Updated 2000, Washington, D.C., USA.
9
