An Inverse Numerical Analysis to Predict the Temperature History during Freezing of Foods

Abstract

Most of the published methods to predict temperature history during food freezing require knowledge of the thermal properties of foods. However, it is difficult to estimate the precise thermal properties of food which show a significant change during freezing. In this paper, an inverse numerical procedure was developed using the modified Marquardt iterative method; a procedure which collects time/temperature data during the early stages of cooling, analyzes these data to determine the thermal property coefficient ratio associated with the heat conduction equation and predicts the time/temperature profile during the remainder of the freezing phase using the system parameters. A numerical experiment using the literature data of lean beef validated that this procedure was able to predict the temperature history even when the measured temperature included a random error of ±0.2°C.