Dipole antenna measurement on vegetable oils at 1 – 4 GHZ by using FEM, MoM, and FIT methods

A comparative study on the performance of computational electromagnetic methods (CEMs), namely finite integration technique (FIT) and finite element method (FEM) via their corresponding electromagnetic software and moment method (MoM) using RWG basis function via MATLAB, has been conducted for dipole antenna measurement of vegetable oils (palm oil, olive oil, and canola oil). These methods are widely used in the field of CEMs with several advantages and disadvantages in finding the same goal. There are many interests in understanding vegetable oil, one of it is to identify types of vegetable oil in order to avoid adulteration to generate profits. However, past research in identifying types of vegetable oil are too expensive, requiring laboratory assistant, or limited to its cavity size such as cavity resonators. This study proposed a dipole antenna to measure the reflection coefficient (S11) of vegetable oils. For this purpose, the antenna was fabricated using copper wire as a prototype to conduct measurements on air, water, and vegetable oils using Anritsu VNAMaster and simulation using CEMs. Next, the simulations were compared with the measurements with respect to the S11 at a frequency range of 1–4 GHz for air, water, and vegetable oils. The permittivity values were gathered for vegetable oils indicated distinctive values of dielectric constant at 1.4–2 GHz for palm oil, olive oil, and canola oil. The theoretical values for FEM, MoM, and FIT methods were analysed and compared to choose the best setting for vegetable oil simulation. The effect on S11 for different meshes, dipole heights, and dipole lengths was presented in this study to determine the best method, this study also considered accuracy, memory usage, versatility, and execution time. The frequency ranges of 1.6–1.8 GHz and 2.5–3.5 GHz were chosen to further analyse vegetable oils. At these frequency ranges, vegetable oil can be distinguished from one another from the measurement of S11. It is further validated by ANOVA analysis and suggested that frequency range 3.3 – 3.5GHz provide the best range to identify types of vegetable oil. Feature selective validation (FSV) was employed to test the relationship of accuracy between simulated and measurement results. Overall assessments were analysed for a placed on the IEEE's interpretation scale. It was found that FEM showed a fairly good relationship between measured and simulated results of air, palm oil, canola oil, and olive oil with GDM values of 0.5791, 0.2986, 0.5446, and 0.2820, respectively, and mean relative errors of 0.0551, 0.1630, 0.0381, and 0.1050, respectively, for the frequency range of 1.6–1.8 GHz. FEM also showed a good relationship between measured and simulated results of air, palm oil, canola oil, and olive oil with GDM values of 0.2258, 0.2755, 0.3986, and 0.3169, respectively, and least mean relative errors of 0.0397, 0.0800, 0.0581, and 0.0833, respectively, for the frequency range of 2.5–3.5 GHz. Overall, this study determined that FEM offers the best fitting between measured and calculated results. However, the FIT method followed closely with FEM results, which provides better execution time and memory usage.

Text 118235.pdf Download (1MB) Official URL or Download Paper: http://ethesis.upm.edu.my/id/eprint/18364

Actions (login required)

View Item