Composition-based prediction of dielectric properties of foods.

Prediction of accurate dielectric property data from fundamental principles for systems as complex as foods has not been possible. Simple prediction models based on easily measurable composition data can serve many useful purposes. Literature dielectric data on foods and their composition were statistically correlated. Dielectric data on salt solutions were measured to explain some of the results. When composition data were not available, standard handbook compositions were used. Inclusion of all types of foods (meats, fruits, and vegetables) inhibited any useful correlation with composition. Based on a smaller data set of meats, both dielectric constant and loss increased with water and salt content. Dielectric constant generally decreased with temperature whereas dielectric loss decreased with temperature at lower salt concentrations and increased with temperature at higher salt concentrations.