A New Approach to Predicting the Complex Permittivity of Rice

The initial attempts in relation to the present study, i.e., to develop a moisture calibration from the intrinsic dielectric properties of rice and wheat kernels, were unsuccessful. It was observed that the behavior of the loss factor with the variation of moisture content of the samples was less than predictable. The foregoing statement corroborates earlier results. Earlier researchers introduced the term "moisture density" (product of decimal moisture content and bulk density) and showed that variation in dielectric constant attributable to bulk density variation can be reduced when data are presented in terms of moisture density. The data, however, for the loss factor were not presented. In an attempt to find the moisture dependence of the relative permittivity and loss factor of different cultivars and kinds of rice at 2.45 GHz to produce quadratic and cubic models for the variation of the relative permittivity and loss factor with moisture content and density, a new term, "moisture-specific volume" (ratio of decimal moisture content to the bulk density), is introduced in the present work. Data are presented in light of this new term and in light of seven equations for the dielectric properties of random media. Further, two equations out of the seven are found to be equivalent. The data for the moisture-specific volume and hence for the moisture content of different samples and the relationships provide the estimates for the relative permittivity and loss factor of rice kernels over almost the entire range of moisture content, without the aid of reference data points. A comparative study of the two proposed models with those of previous researchers is also reported. Results for the two dielectric properties for rough, brown, and white or milled rice kernels of 'Lebonnet' and 'Pecos' cultivars, in terms of moisture-specific volume presented in the form of quadratic and cubic models, are found satisfactory.