Improved antenna design for grain bin electromagnetic imaging
Loading...
Date
2022-06-21
Authors
Mohamadi, Taghi
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Grain such as wheat and canola is usually stored in silos or bins, often for long periods of time. Grain can and regularly does spoil in such bins, thus monitoring the grain quality and quantity inside the bin is crucial. It is possible to monitor grain within metallic grain bins using three-dimensional ElectroMagnetic Imaging (EMI). To acquire data in EMI, multiple antennas must transmit and receive ElectroMagnetic (EM) energy within the imaging bin. These antennas subjected to significant mechanical forces, they must have a low profile. These antennas should also be easy to model in computational EM software (using a point source). Moreover, since the EMI algorithm uses linear polarization of EM fields, it is common to choose antennas that can detect a linear polarization and reject others, such as antennas that measure magnetic fields normal to the antenna’s loop and reject electric fields. This can be done through using Shielded Loop Antennas. Given their desirable physical and EM characteristics, previous grain bin imaging systems have been designed with Shielded Half Loop Antennas (SHLA). While capable of detecting primarily the tangential magnetic field at the bin wall, existing SHLA suffer from a large reflection coefficient at the antenna terminal, meaning they can be drastically improved by reducing this reflection coefficient, thus reducing noise and receiving more desired signal. Herein, we consider two methods of improving upon this design: matching circuits and improved antennas. For matching techniques, we consider passive and active circuits. Through simulation, we show how these methods can improve antenna performance. For an improved antenna design, we show that loading a SHLA with ferrite material can enhance its performance. To model the behavior of a ferrite-loaded shielded loop antenna, we develop an approximate lumped circuit model which can predict antenna resonance frequency, its reflection coefficient, and its efficiency. This model aids in the design of larger sized ferrite-backed antennas. An antenna at resonance frequency of 260 MHz is fabricated and tested in a small (labratorysized) grain bin. Its sensitivity to magnetic field and rejecting electric fields is tested in Gigahertz Transverse EM (GTEM) cell at the Electromagnetic Imaging Laboratory (EIL) of the University of Manitoba, and show that the ferrite-backed antenna increases signal strength while maintaining the rejection of normal-E fields. The small ferrite loaded shielded half-loop antenna provided a 6 18 dB improvement in signal level over existing half-loop antennas over the imaging frequencies of interest (approx 200600 MHz). Furthermore, experimental laboratory imaging tests with hard-red winter wheat indicate that the proposed antennas can be modeled more accurately using simple polarized point sources. A full 3D image of a wet-grain target set within a dry-grain background inside the enclosure shows that the improved antennas enhance the quality and accuracy of the reconstructed images, mainly where the antenna performance improvements are prominent. Based on these promising results, we designed and constructed a scaled up ferrite-backed halfloop antenna suitable for a larger industrial grain bin and at a lower frequency (a resonance of 120 MHz). Preliminary measurements of these antennas show that these larger versions not only can detect magnetic field and reject electric fields, but also increase the signal strength in the grain bin.
Description
Keywords
Inversion algorithms, Electromagnetic fields, Ferrite, Electromagnetic imaging, Shielded loop antenna, Non-Foster matching technique