Abstrato

An analytical approach to predicting moisture contents of spring barley kernels through microwave permittivity data

Zeeshan Ahmad, Ashutosh Prasad, Kamlesh Prasad

In the present paper, an attempt has been made at finding the moisture dependence of the relative permittivity and dielectric loss factor of Spring Barley kernel samples at 2.45 GHz and 240C to give quadratic and cubic models for the variation of the two dielectric properties with (i) decimal moisture content, m(ii) moisture density (product of decimalmoisture content and bulk density of the sample, md) and (iii) moisture specific volume (ratio of decimalmoisture content to bulk density of the samplemv). The last parameter has been introduced by the authors in order to remove or minimize the irregularities in the trend of variation of dielectric loss factor with moisture content. Resulting data and the plots of such variations, derived from eight equations for the effective dielectric function of random media are presented. On their bases, it is shown that the moisture content of a given sample could instantly be estimated for a given set of values for the two dielectric properties and the technique may be applied over the entire acceptable range of moisture contents without the aid of any reference data points.Abetter performance of the present models compared with those of Nelson and Kraszewski is reported. Average percentage errors of prediction of about 1.4 and 1.5 from the present quadratic and cubic models for relative permittivity versus decimal moisture content have been achieved as compared to 3.61 and 3.5 from the corresponding NelsonÂ’s models. The average percentage errors in loss factor from the present models are about 0.4 and 0.5 as compared to 24.8 fromthe NelsonÂ’s solitarymodel.