Optimization of drying food products: application to fruits

Air drying is one of the most used unit operation in food processing. Simulation or designing the air-drying operation requires the mathematical description of food moisture evolution during the process, known as drying kinetics. The structure of the food materials goes through deformations due to the simultaneous effect of heat and mass transfer during the drying process. These deformations lead to changes in many quality attributes and directly affect physical properties (i.e. the mass diffusion coefficient). Changes in volume, “shrinkage”, negatively impact the quality perception of dried products by consumer. The knowledge of the changes in the properties of foods that occur during drying is hence needed for better designing the process. The aim of this study is to understand the effect of drying on quality of fruit which typically has high moisture content (i.e. pear and grape) and to develop a mathematical model which quantify the drying of this food matrix. To this purpose the drying kinetics, shrinkage and structure modifications (through SEM analysis) were experimentally evaluated. During drying, the macroscopic transport of water through the cellular tissue constituting the fruit is largely controlled by the microscopic distribution of water and air on a cellular and subcellular distance scale and by membranes permeability. 1H pulsed low resolution NMR allows to obtain quantitative information on water distribution and diffusion by detecting the proton signal predominantly due to H2O contained in vegetable tissue. Therefore, portable-NMR was used to determine the drying moisture profile and thickness reduction of pears. Portable-NMR also allowed to investigate water mobility in fresh and dried pears by measuring the longitudinal and transverse relaxation times, and the self-diffusion coefficient. For fruit, such as grape, drying is a slow and very energy intensive process because the waxy peel has low permeability to moisture. In order to enhance the drying rate, pretreatments are used. In this thesis, both chemical (by dipping in an ethyl oleate solution) and physical (by abrasion) pretreatments were analysed. It was found that ethyl oleate and abrasion pretreatments have the same effect in reducing the drying time, but the second one is to be preferred because it avoids the use of chemical additives and permits safer raisin to be produced. Moreover, the samples pretreated by peel abrasion showed less shrinkage and no cracks on the peel surface with respect to those pretreated with ethyl oleate solution. To obtain a detailed prediction of moisture distribution during drying, in the thesis, a diffusion model with Fickian moisture transfer was coupled with an empirical law which considers the effect of shrinkage. A numerical solution technique, based on the method of finite elements, is used with an adaptive mesh. The results match well with the experimental results. In particular, a good agreement between experimental and theoretical data of moisture ratio during drying was found for both fruits. The model also well predicted the moisture profile along the pear thickness obtained by NMR.