Progress 09/01/13 to 08/31/18
Outputs
Target Audience:1- Readers of IEEE Sensors Journal 2- Scientists an Engineers at International Conferences 3- Students at the School of Engineering, the University of Georgia Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?The project funded in part a PhD graduate student and a full-time Visiting Scholar How have the results been disseminated to communities of interest?1- Conferences and journal papers 2- PhD thesis 3- Seminars and lectures at the University of Georgia What do you plan to do during the next reporting period to accomplish the goals?
Nothing Reported
Impacts
What was accomplished under these goals?
Moisture content is a key parameter for quality assessment of pelleted biomass and other biomass products. It is often needed in pricing, binding, combustion, and storage. With highly automated processes, knowledge of moisture content in real time is a must. Dielectric-based sensors operating at microwave frequencies provide a means for in-line moisture determination. I. Dielectric data collection A fundamental step in developing indirect methods and sensors for determining a physical parameter of interest is calibrating the measurement method against the conventional standard. In this instance, this means calibrating a dielectric-based method and sensor for moisture content determination in biomass products against the standard oven-drying technique. Also, of prime importance are the selection of the measurement technique that provides the most accurate dielectric properties and the identification of an algorithm correlating these properties to the target parameter (moisture content). Among microwave measurement techniques for dielectric properties measurement, free space (reflection and transmission) offers the most advantages. It is nondestructive, no sample preparation is needed, often no contact with sample is required, and can be used for measurements at multiple frequencies over a broad microwave frequency range. With access to such data, performance of indirect method at different frequencies can be assessed and the best frequency (ies) selected. In addition, it can be applied for instantaneous measurements on static (benchtop) and dynamic (flowing/moving) situations. This is of particular interest for in-process monitoring and control of biomass during handling and transformation processes. II. Transmission-type measurement of dielectric properties of biomass pellets 1. Static situation Dielectric properties of water-containing materials are dependent on moisture content, temperature, and for granular and particulate materials on bulk density. Dielectric properties of peanut-hull, pine, and hardwood pellets were measured in free space over a broad frequency range (5 GHz to 15 GHz) for temperature between 10 oC and 50 oC, moisture content between 4.9% and 16%, and bulk density ranging from loosely packed to compacted. For dielectric properties measurements, each sample was placed in a Styrofoam sample holder between two horn lens antennas which were connect to a vector network analyzer through high quality coaxial cables. At a given microwave frequency, the dielectric properties were found to increase with temperature, bulk density, and moisture. Therefore, use of the dielectric properties for moisture content determination requires compensation for temperature and bulk density. 2. Dynamic situation Dielectric properties of flowing pelleted biomass derived from peanut hulls and pine sawdust were collected at 5.8 GHz and 23 oC. Moisture content (wet basis) of pine sawdust pellets ranged from 5.4% to 9.9% and for peanut-hull pellets the moisture ranged from 8.9% to 14.5%.The flowing-material measurement system consisted of a conical 14.7-gallon stainless steel hopper into which pellets are poured. The material then flows through PVC piping into a polycarbonate (Lexan) rectangular-cross-section chute with a thickness of 12.1 cm. On either side of the Lexan chute, two inexpensive 19 dBi high-gain microstrip patch antennas are placed for free-space transmission measurement of the dielectric properties. The patch antennas were connected to a vector network analyzer. Samples then flow through a stainless-steel rectangular-to-circular transition and an adjustable iris which allows control of mass-flow rate. All components are supported by a wooden stand and the system achieves flow through gravity. III. Reflection-type measurement of dielectric properties of sawdust For reflection measurements novel and original low-cost sensors were developed and tested on static and moving sawdust samples. 1. Static situation a. Open transverse-slot substrate integrated waveguide sensor A novel open transverse-slot substrate-integrated waveguide sensor was designed and fabricated for measurements of the dielectric properties of sawdust at 8 GHz. The sensor is connected to a vector network analyzer through a high quality coaxial cable and the dielectric properties are determined from measurement of the modulus and phase of the reflection coefficient. Measurements were performed on sawdust samples at room temperature (23 oC), different bulk densities, and moisture content ranging from 6.8% to 39% (wet basis). Numerical modeling and fitting showed that there exists a linear correlation between the dielectric properties and moisture content which can be implemented in microwave sensors for real-time determination of moisture content in biomass samples. b. Open-ended half-mode substrate-integrated waveguide This is another novel inexpensive microwave sensor for moisture determination in biomass Samples. The measurement principle is similar to that used with open transverse-slot substrate-integrated waveguide sensor. The sensor operating at 5 GHz was used to measure the dielectric properties of sawdust samples at room temperature (23 oC), different bulk densities, and moisture content ranging from 6.8% to 22.8% (wet basis). The dielectric properties were found to increase linearly with moisture content with the effect of bulk density more pronounced for the dielectric constant. 2. Dynamic situation For measurement on moving sawdust, a conveyor belt was used. Four open-ended rectangular substrate-integrated waveguide (ORSIW) sensors were connected through three radiofrequency switches to in-house made six-port reflectometer. The four ORSIWs measured the reflection coefficient of sawdust while it was moving at different speeds ranging from 0.15 m/s to 0.35 m/s. Dielectric properties computed from measurement corresponding to all four ORSIW sensors were found to increase linearly with moisture content. IV. Development of algorithms for moisture determination from measurement of dielectric properties Direct correlations between the dielectric properties and moisture content were established. These calibration equations can be used to predict moisture content from instantaneous measurement of the dielectric properties at single microwave frequency. In this instance, there is a need to compensate for bulk density and temperature. This is means additional devices for the measurement of temperature and bulk density which is not practical, in particular for flowing situations. A density-independent and material-independent algorithm which was originally developed grains was tested successfully for biomass both in static and dynamic situations. The algorithm allowed moisture prediction from dielectric properties without knowledge of bulk density. In addition, a single calibration can be established for measurements on different types of biomass materials. V. Prototype sensors Several cost-effective sensor prototypes were assembled, calibrated and tested for moisture determination in biomass materials. The prototypes were built with inexpensive off-the-shelf components and operated at microwave frequencies. The sensing elements of these sensors were made in house by using numerical simulation software COMSOL and PROTOMAT. Some of the designs were made for static measurements with benchtop-type instrumentation and some of the designs were made for implementation for in-process monitoring and control operations. For reflection-type portable sensors, the six- port technology was used. The six port technology is an inexpensive replacement of the vector network analyzers and provides the same level of accuracy. Some of the portable prototypes allowed moisture determination with a standard error of performance of less than 1%.
Publications
- Type:
Theses/Dissertations
Status:
Published
Year Published:
2017
Citation:
Murat Sean McKeown, Microwave sensor development for real-time permittivity determination and moisture measurement in biomass materials , PhD dissertation, The University of Georgia, May 2017.
- Type:
Journal Articles
Status:
Published
Year Published:
2018
Citation:
Sakol Julrat, Samir Trabelsi, and Stuart O. Nelson, Open-ended half mode substrate integrated waveguide sensor for complex permittivity measurement , IEEE Sensors Journal, 18 (7): 2759-2767. April 1, 2018.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2018
Citation:
Sakol Julrat and Samir Trabelsi, Open-ended coplanar waveguide sensor for dielectric permittivity measurement , IEEE I2MTC Conference, May 14-18, 2018 Houston, TX.
- Type:
Journal Articles
Status:
Published
Year Published:
2017
Citation:
Sakol Julrat, Samir Trabelsi, Density-independent algorithm for sensing moisture content of sawdust based on reflection measurements , Biosystems engineering 158 (2017) 102-109.
- Type:
Journal Articles
Status:
Published
Year Published:
2017
Citation:
Murat Sean McKeown, Samir Trabelsi, Stuart Nelson, Ernest William Tollner, Microwave sensing of moisture in flowing biomass , Biosystems engineering 155 (2017) 152-160.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2015
Citation:
M.S. McKeown, S. Trabelsi, E.W. Tollner, "Microwave moisture measurements of flowing biomass." ASABE Annual International meeting, New Orleans, Louisiana, July 26-29 2015.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2015
Citation:
M.S. McKeown, S. Trabelsi, E.W. Tollner, "Effects of temperature and material on dielectric properties of pelleted biomass" IEEE Southeast Conference, Fort Lauderdale, Florida, April 9-12, 2015.
|
Progress 09/01/16 to 08/31/17
Outputs
Target Audience:1- Readers of Biomass and Sensors International Journals Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?A Postdoctoral Research Associate and a graduate student learned how to use COMSOL software to design a variety of inexpensive microwave sensors. Also, they fabricated in house these sensors by using Protomat and assembled microwave circuits, made with off-the-shelf components, to test the performance of these antennas in predicting moisture content of biomass materials. How have the results been disseminated to communities of interest?1- Conference and Journal papers 2- seminars at the University of Georgia What do you plan to do during the next reporting period to accomplish the goals?Investigate ways to improve accuracy in prediction of moisture content of biomass metrials.
Impacts
What was accomplished under these goals?
I-Portable six-port reflectometer for determining moisture content in biomass material A portable microwave reflection measurement system operating at 5.15 GHz and based on six-port reflectometer (SPR) technology was designed, tested, and calibrated for moisture determination in biomass material. It has the advantage of being low-cost and simple compared to laboratory-grade instrumentation such as the automatic network analyzer (ANA). The proposed sensor is made with off-the-shelf components and consists of a radio-frequency oscillator, a six-port reflectometer, four radio-frequency power detectors, a 16-bit data acquisition unit, a microcontroller, and open-ended half-mode substrate integrated waveguide (HMSIW) sensor. The latter was optimized using by COMSOL software and made in- house with Protomat LPFK. The moisture sensor principle consists of measurement of the dielectric properties at a single microwave frequency and the use of a dielectric-based algorithm for moisture prediction. The dielectric properties of biomass measured with the SPR sensor compared very well with those determined with the ANA. Linear correlations between the dielectric properties divided by bulk density and moisture content with a high coefficient of determination (> 0.96) were identified. These equations allowed moisture prediction with a standard error of performance of 0.967% in biomass samples with moisture content ranging from 6% to 36%, wet basis. The proposed sensor is a novel cost-effective system that can be used for instantaneous and nondestructive determination of moisture content in biomass materials. II - Inline microwave reflection measurement technique for determining moisture content of biomass material An in-line microwave reflection measurement technique for determining moisture content in biomass material (sawdust) moving past the sensor on a conveyer was designed, calibrated, and tested for material speeds of 0.15, 0.23, and 0.35 m/s. The proposed system consists of a 3.0 GHz six-port reflectometer connected to four open-ended rectangular substrate-integrated waveguide (ORSIW) sensors which were connected through three RF-switches. The ORSIW sensors were mounted flush on sidewalls of the conveyor and were used for measuring the dielectric properties of the sawdust as it moved on the conveyor belt. The proposed system consists of four ORSIW sensors, three RF switches, a six-port reflectometer (SPR), an RF oscillator, four RF power detectors, a 16-bit data acquisition (DAQ) system and a personal computer (PC) running the data acquisition software. The ORSIW sensors were optimized by using COMSOL software and fabricated in-house using Protomat LFPK. The measurement principle of the system was based on identifying correlations between dielectric properties measured with the four ORSIW sensors and the moisture content of sawdust samples. For the purpose of verification of the accuracy of the measured dielectric properties, measurements obtained with the ORSIW were compared with those determined with an open-ended coaxial commercial probe and they were in good agreement for moisture contents between 8.5% and 37.7%, wet basis. In the next step, a dielectric-based algorithm was developed and used for moisture prediction from measurement of the dielectric properties at a single microwave frequency of 3.0 GHz. This algorithm is based on the existence of linear relationships between the dielectric properties each divided by bulk density and moisture content. For all four ORSIW sensors and for biomass moving at a speed between 0.15 m/s and 0.35 m/s a high coefficient of determination (r2 > 0.94) was found for correlations between the dielectric properties and moisture content. For sawdust samples of moisture content between 8.5% and 37.7%, wet basis, the standard errors of performance for all ORSIW were between 2% and 3%. Further investigations are needed to improve the level of accuracy. The proposed system is suitable for in-line determination of moisture content of moving biomass materials.
Publications
- Type:
Journal Articles
Status:
Published
Year Published:
2017
Citation:
1- Sakol Julrat and Samir Trabelsi, "Portable six-port reflectometer for determining moisture content of biomass material, IEEE Sensors Journal 17(15): 4814-4819, August 1, 2017.
2-Sakol Julrat and Samir Trabelsi, Density-independent algorithm for sensing moisture content of sawdust based on reflection measurements, 158: 102-109, 2017.
|
Progress 09/01/15 to 08/31/16
Outputs
Target Audience:1- Scientists and engineers at international conferences. 2- Undergraduate students in the "Introduction to Enginering" course at the University of Georgia, Fall semester 2016. Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?A graduate student and a post-doctoral research associate learnedhow touse Protomat for the inhouse fabrication of low-cost substrate integrated waveguide sensors. They also investigated radiation patterns of different configurations of these sensors through simulation using COMSOL. How have the results been disseminated to communities of interest?1- Conferences and journal papers 2- Seminars and lectures at the University of Georgia What do you plan to do during the next reporting period to accomplish the goals?During the next reporting period, the focus will be on implementing these novel sensors for moisture measurement in moving biomass.
Impacts
What was accomplished under these goals?
Sensors Design and Testing Several substrate integrated waguide sensors working in reflection mode were designed, tested and calibrated for measurements the dielectric properties of sawdust at microwave frequencies. Moisture content and bulk density of sawdust were determined from measurement of the dielectric properties at a single microwave frequency Substrate waveguide based sensors were selected because they are inexpensive , easy to fabricate, and can be easily implemented for real-time, in-line measurements without disturbing the process. The choice of reflection measurement is attractive because the sensor is placed on one side of the material. Two type of substrate wavegude-based sensors were fabricated in house with Protomat and their radiation pattern investigated with advanced numerical software COMSOL. The first is the Transverse Slot Substrate Waveguide and the second is a Half Mode Substrate Integrated Waveguide. Transverse slot substrate waveguide This design configuration has the sensing area on the traverse or top face of the substrate waveguide. The design was accomplished by removing a portion of the metal layer of a double-sided copper clad FR4 circuit board to make a structure similar to a slotted waveguide antenna. In order to apply well-known waveguide techniques, an empirical relationship between geometries of substrate waveguides and those of classical waveguides was used to dimension the waveguide. A microstrip-to-Substrate waveguide transition was used to couple the sensor to a coaxial cable for connection to power sources. The width and length of the substrate waveguide sensor determine the radiation pattern of the sensor. Specifically, the sensing depth and the field width can be controlled by these parameters. The final configuration of the sensor had a width of 20 mm, a length of 68mm (omitting the microstrip-to-substrate waveguide transition region) , and a thickness of 1.6mm. The sensing region on the face of the substrate waveguide had a width of 19mm and length of 10mm and is offset from the end wall of the substrate waveguide by 1mm. Sawdust samples were obtained from a local saw mill consisting of a mixture of oak and hickory. The sawdust was seperated into bags that were conditioned to moisture contents between 6.8% and 39%, (wet basis). For each sample, moisture content was determined with oven-dryin method which consited of drying three 16-g samples at 103 °C ± 1°C for 16 hours.A Linderberg/ Blue M mechanical convection oven was used for this perpuse. Samples were then removed and allowed to cool in a dessicator and weighed to the nearest 0.0001 g. The samples were then placed back into the oven to dry an additional 2 h and weighed again. If the total weight change between weighings was less than 0.2% moisture content, the result was accepted as the final moisture content. For measurement of the dielectric properties, sawdust was loaded into Styrofoam sample holder filling a volume of 12.1 x 12.0 x 14.7 cm. All measurements were performed at room temperature, 22-23°C. Once the sawdust was loaded on the sensor, measurements of phase and magnitude of the reflection coefficient were obtained at frequencies between 4 and 20 GHz using an Agilent E5071C Network Analyzer. The same samples were also measured using an existing free-space transmission system that uses a Hewlett-Packard 8510C Network Analyzer to obtain dielectric properties of the material. Results showed that both magnitude and phase have linear relationships to moisture content for measurements. Of the range of frequencies investigated 8.00 GHz, 9.36 GHz, and 10.21 GHz show promise for prediction of dielectric properties. These measurements constitute the foundation to obtain an algorithm to extract dielectric properties from measurements of magnitude and phase from which moisture content and bulk density could be predicted. Half-mode Substrate integrated waveguide (HMSIW) A half-mode substrate integrated waveguide fed by a microstrip line, was used as a open-ended waveguide sensor with a ground flange for measuring the dielectric properties of sawdust. For accurate determination of the dielectric properties from measurement of the reflection coefficient, the sensor was calibrated using the three-material calibration technique. For liquid samples, aplastic container was mounted on the top of the ground flange. A high quality coaxial cable connected the open-ended HMSIW sensor to the automatic network analyzer (ANA, ENA series E5071C). The one-port reflection coefficient measurement was performed with the ANA. Three materials, i.e, air, water and 25% ethanol aqueous solution were used to calibrate the open-ended HMSIW sensor. A 50% ethanol aqueous solution was used to validate the sensor calibration. The predicted results agreed well with those measured with an open-ended Agilent Technologies coaxial probe for frequencies between 4.5 GHz to 5.25 GHz and corresponding to the dominant mode of the HMSIW. Dielectric properties of eight sawdust samples with different moisture contents were measured with the open-ended HMSIW sensor. Each sawdust sample was loaded into the container (70 mm x 120 mm x 50 mm), where the sawdust is resting on the waveguide sensor. The sawdust sample was 50 mm thick. Effect of bulk density on dielectric measurement was investigated by varying the compaction of sawdust. For the low density, the sawdust sample was carefully poured in the container without any pressure on the sawdust. To obtain higher density, constant loads of 0.5 kg, 1.0 kg and 5 kg were applied for packing the sawdust sample while adding more material to obtain the 50 mm thickness. A density-independent algorithm for determining moisture content of sawdust based on a microwave reflection technique has been developed and tested. The technique proposed allows one to determine the moisture content of sawdust from measurement of the dielectric properties determined from measurement of the reflection coefficient at a single frequency. For the definition of the density-independent function, both cases of complex-plane representation of the dielectric properties divided by either moist or dry bulk density were considered. Calculation of the standard errors of performance for moisture prediction in sawdust samples by using either function were very close, ie. 0.955% and 0.957%. The reflection-based algorithm, coupled with the low cost of the sensing device, shows potential for use of this method for in-process moisture determination in sawdust during the densification process where moisture content ranges from 10% to 20%.
Publications
- Type:
Journal Articles
Status:
Under Review
Year Published:
2016
Citation:
Microwave Sensing of Moisture in Flowing Biomass Pellets, Biosystems Engineering, Elsevier, 2016.
- Type:
Journal Articles
Status:
Under Review
Year Published:
2016
Citation:
Density-independent algorithm for sensing moisture content of sawdust based on reflection measurements
|
Progress 09/01/14 to 08/31/15
Outputs
Target Audience:1- Scientists and engineers at international conferences. 2- Undergraduate students in the "Introduction to Engineering" course at the university of Georgia, Fall 2015. Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?Findings from this project will be presented at International Confrences and guest lecture will be provided for undergraduate students at the University of Georgia. How have the results been disseminated to communities of interest?1- Conference and journal papers. 2- Seminars and lectures at the University of Georgia What do you plan to do during the next reporting period to accomplish the goals?During the next reporting period, the focus will be on testing in the laboratory subdtrate integrated sensors and similar designsto predict moisture in solid biomass from measurement of their dielectric properties a single microwave frequency.
Impacts
What was accomplished under these goals?
Measurements on Flowing Solid Biomass Measurements on pine pellets and peanut-hull pellets were performed with a flow through system. The system consisted of a conical 14.7-gallon stainless steel hopper into which pellets are poured. The material then flows through PVC piping into a polycarbonate (Lexan) rectangular-cross-section chute with a thickness of 12.1 cm. The Lexan chute permits microwave transmission with negligible energy loss. On either side of the Lexan chute, two 19 dBi high-gain microstrip patch antennas are placed for free-space transmission measurement. Samples then flow through a stainless-steel rectangular-to-circular transition and an adjustable iris which allows control of mass-flow rate. All components are supported by a wooden stand and the system achieves flow through gravity. The patch antennas were connected to a Hewlett-Packard 8510C Network analyzer for determination of attenuation and phase shift at 5.8 GHz and 23 °C. Acomputer program automates measurements and calculates ?' (dielectric constant) and ?" (loss factor) from the attenuation, A, and phase shift, φ. The system was calibrated in transmission mode by using a response-type calibration. After calibration, samples were poured into the hopper of the flowing material system. The sampling rate was 6 measurements per second. Three different mass flow rates were used, low, medium, and high, with each mass flow rate run in triplicate. With a known mass of material flowing through the system and monitoring of flow start and stop time, the mass flow rate of each run could be determined. Prior to each flowing measurement a static measurement of dielectric properties was taken as well. Reference moisture content determination of pelleted biomass was performed by a standard oven-drying reference technique [13]. Triplicate 16-g samples were loaded into aluminum cans and dried at 103 °C ±1 °C in a Lindberg/Blue M mechanical convection oven for 18 hrs. After removal, warm samples were placed in a desiccator with anhydrous calcium sulfate to equilibrate to room temperature and reweighed.. The moisture content was calculated in terms of the wet basis according to the following equation. where M% is the moisture content, mw is the mass of water in the sample and md is the mass of dry matter. An initial calibration for correlating dielectric properties to moisture content was performed by loading pelleted biomass into the flowing-material measurements system. Enough material was loaded to completely fill the polycarbonate chamber with known volume. With the mass of the material added known, bulk density of the material could be calculated. Static microwave measurements of pelleted biomass of different moisture contents were used to generate a moisture content prediction equation. A density-independent moisture calibration function was used for moisture prediction from measured dielectric properties. A unified calibration was obtained for both pine and peanut-hull pellets. To evaluate performance of this calibration in predicting moisture content in flowing solid biomass, the standard error of calibration (SEC) was calculated. In general, moisture predicted for flowing biomass are in good agreement with those predicted for static biomasssamples. For pine pellets with moisture ranging from 5.4% to 9.9%, the SEC comparing flowing versus static ranged from 0.15% to 0.54%. For peanut-hull pellets with moisture ranging from 8.9% to 14.46%, the SEC comparing flowing versus static ranged from 0.17% to 1.39%. The average mass flow rateranged from 0.2 kg/s to 2.89 kg/s for pine pellets and 0.25 kg/s to 2.98 kg/s. Substrate Integrated Waveguide Sensor Researchhas been initiated to identify sensors better suited for measurements in industrial environments. A subdtrate integrated (SIW) sensor was designed by removing a portion of the metal layer of a double-sided copper-clad circuit board to make a structure similar to a slotted waveguide antenna. Preliminary work has been performed to determine the feasibility of the design, and initial experimental resultslook promising.
Publications
- Type:
Journal Articles
Status:
Submitted
Year Published:
2015
Citation:
Effects of Temperature and Material on Dielectric Properties of Pelleted Wood-based Biomass, Biomass and Bioenergy, Elsevier, September, 2015
|
Progress 09/01/13 to 08/31/14
Outputs
Target Audience: 1- Graduate students, College of Engineering, the University of Georgia: Guest lecture "Microwave Sensing for Quality Determination of Pelleted Biomass' Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided? Findings from this project will be presented at two conferences: 1- IEEE Southeast Conference, Fort Lauderdale, Florida, April 9-12, 2015. 2- ASABE Annual International meeting, New Orleans, Louisiana, July 26-29 2015. Graduate stute was invited to give a Guest Lecture at the College of Engineering, The University of Georgia: "Microwave sensing for quality determination of pelleted biomass." How have the results been disseminated to communities of interest? Findings from this research were presented at the College of Engineering, The University of Georgia. What do you plan to do during the next reporting period to accomplish the goals? During the next reporting period we will focus on further refining the measurement protocol and assembling and testing different sensing configurations suitable for in-line implementation in industrial environment.
Impacts
What was accomplished under these goals?
The production of biomass is a significant emerging industry in the United States, with biomass energy consumption growing more than 60% between 2002 to 2013. A major part of this industry is the production of pelleted biomass which increases the specific density of biomass. This allows for the biomass to be handled in a more efficient and cheaper manner through use of systems and equipment designed for handling grain. A critical quality factor of pelleted biomass is moisture content, which is important in pricing, storage, combustion optimization, and binding. There is a growing need for sensors for in-line determination of moisture content of solid biomass. Recently, our research group proposed several algorithms and sensing methods for nondestructive instantaneous determination of bulk density and moisture content of grain and seed from measurement of their dielectric properties at microwave frequencies. Building on that success, our goal is to develop microwave sensors for in-line determination of moisture content in solid biomass. In this phase of the project, we focused on building a database of dielectric properties of solid biomass and identifying correlations between these properties and moisture content. It is well known that the dielectric properties of water-containing materials are dependent on frequency, temperature, and moisture content. For granular and particulate materials, they are also dependent on bulk density. Therefore, if moisture content is the target parameter, at a given frequency there is a need to compensate for the influence of temperature and bulk density. Methods for eliminating the effect of density were proposed by our research group and others. Therefore, at a given frequency only the effect of temperature needs compensation. Analytical two-dimensional and three-dimensional models provided moisture calibration equations at different microwave frequencies with overall standard errors of calibration (SEC) for moisture prediction from measurement of the dielectric properties less than 1%. The data collected and correlations established between the dielectric properties and moisture content will be valuable for the scientist and engineer looking at developing solutions for rapid, in-process sensing of moisture content in solid biomass. Database Dielectric properties of peanut-hull pellets were measured over wide ranges of frequency (between 5 Gigahertz and 15 Gigahertz); temperature (between 0 °C and 50 °C in increments of 10 °C); moisture content (between 6% and 16%); and three level of bulk density (low, medium, and high) for each moisture content. The measurements were performed with a free-space- transmission technique where the sample was place between two antennas and the attenuation (loss of energy) and phase shift (change in the wave velocity) were measured. The dielectric properties were determined from attenuation and phase. A total of two thousand data points was collected. Graphical representations of the dielectric properties as a function of the different variables revealed that the dielectric properties remained constant or decreased slightly with frequency, and they increased with moisture content, temperature, and bulk density. This database was used to develop models and algorithms for determining moisture content in solid biomass from measurements of their dielectric properties at a single microwave frequency. It will also constitute the basis for understanding the interaction between microwaves and solid biomass. Algorithms for moisture determination in solid biomass One of the main objectives of this investigation is to develop analytical models and algorithms that can be used for nondestructive and instantaneous of determination moisture content in solid biomass from measurement of their dielectric properties at microwave frequencies. The core of any algorithm for moisture determination in solid biomass is based on identifying correlations between the dielectric properties and moisture content. In this phase of the project, two- dimensional and three-dimensional data analysis allowed the establishment of calibration equations for moisture determination independent of bulk density, and with temperature compensation, from measurement of the dielectric properties at a single microwave frequency. Statistical analysis provided correlations between moisture content and the dielectric properties with high coefficients of determination (between 0.94 and 0.98). These equations were used in a user-friendly computer program which was written to automate the dielectric measurements in laboratory settings on static and flowing solid biomass. For measurements on static samples of peanut-hull pellets use of these calibration equations provided moisture content with a standard error of calibration (SEC) between 0.5% and 0.98% moisture for frequencies between 5 and 15 GHz. A flow-through system was designed and tested for measurements on flowing biomass. The system was designed with in-line implementation in mind, and the antennas used for sensing moisture in the flowing biomass were inexpensive off-the-shelf patch antennas operating at 5.8 GHz. In this system, sample material flows through a stainless steel hopper which flows into a lexan rectangular cross section chute of a thickness of 12.1 cm. The Lexan chute allows for transmission microwave measurements on the exposed pelleted biomass. Sample then flows through a stainless steel rectangle-to-circular transition eventually exiting through an adjustable iris which allows for control of mass flow rate. Flow is achieved through gravity. Preliminary measurements on flowing peanut-hull pellets were performed at a single moisture content. Here, a moisture calibration function was developed by using static measurements on solid biomass. This calibration function was then applied to flowing biomass. Measurements were performed at three different mass flow rates that were achieved by varying the diameter of the iris. Results of this study demonstrated that static and flowing predicted moisture content were in good agreement. Further work will be performed to validate these results across different moisture contents and different pellet materials.
Publications
- Type:
Conference Papers and Presentations
Status:
Accepted
Year Published:
2015
Citation:
1- M.S. McKeown, S. Trabelsi, E.W. Tollner, "Microwave moisture measurements of flowing biomass." ASABE Annual International meeting, New Orleans, Louisiana, July 26-29 2015.
2- M.S. McKeown, S. Trabelsi, E.W. Tollner, "Effects of temperature and material on dielectric properties of pelleted biomass" IEEE Southeast Conference, Fort Lauderdale, Florida, April 9-12, 2015.
|
|