| Pressurized-steam segmented-flow aseptic processing (PSAP) is a novel technology being developed to process particulate foods, e.g. mushrooms, soybeans, beans, peas, apple slices, corn, etc. For the commercial application of this system, it is essential to model the heat and mass transfer, steam distribution and penetration in the heating of those foods. Models will be helpful in adjusting the segmented-flow aseptic processing system in order to process different particulate foods in this system. Such capability can bring great benefits to the food and agricultural industry of the Commonwealth of Pennsylvania and beyond.;For modeling purposes, each segmented unit in PSAP was simplified to be a packed bed. Glass beads with 2, 3, and 5 mm diameter were used to simulate the particulate foods. In preliminary experiments, a capillary fringe (CF) was observed, which was a region with 100% water saturation existing at the bottom of packed beds after vertical drainage. CF prevented steam from penetrating the whole bed and slowed down the heating, which is a negative effect.;Four major heat transfer mechanisms were important in this research: conduction, convection, condensation, and steam penetration. Although many researchers have modeled these processes, there was no applicable, integrated model available for predicting the temperature changes in the packed beds with the same scenario as in PSAP. The overall objective of this research was to develop a numerical model to simulate the heat and mass transfer in packed beds during isobaric steam heating so that the maximum bed depth can be determined, with which the temperature distribution from top to bottom of packed bed is still uniform.;Firstly, the characteristics of packed beds were studied. Although different sizes of glass beads have different particle densities and bulk densities, the porosity turned out to be the same, 0.46, for each size of glass beads. The sphericities were 0.989, 0.991, and 0.992 for 2, 3, and 5 mm glass beads, respectively, suggesting ideal spheres of the glass beads. The measurement of static holdup showed that at 1 cm above upper boundary of capillary fringe in packed beds, static holdup decreased rapidly from 0.46 to 0.024, while above that elevation it remained almost constant at 0.24. Under low water flux, the water flow velocity (parent velocity) in packed bed had positive polynomial relationship with dynamic holdup. Under high water flux, the relationship between dynamic holdup and water flow velocity was assumed to be linear, based on the literature. The capillary fringe thickness were 2.4, 1.4, 0.6 cm for 2, 3, and 5 mm glass beads, respectively, under atmospheric pressure. Less thickness of capillary fringe was expected for packed beds at high temperature due to the decreased surface tension of water.;A PSAP simulator was designed and built, and steam heating experiments were performed to investigate the temperature histories in packed beds with different bed depths and particle diameters. By modifying the experimental setup, temperature and weight changes in packed beds in steam heating were simultaneously measured. The general trend of temperature change of packed bed in steam heating was that the elevations from top to bottom were heated in order, and the bottom portion, CF, was heated much slower than other elevations. Further studies showed that with the decrease of particle diameter and increase of bed height, it took longer time for CF to reach a target accomplished temperature fraction (ATF).;The overall scheme for developing the numerical model was to divide the unsaturated flow zone (UFZ) and CF into N and M finite layers, in which the energy and mass balances were developed. Finite difference method was used to develop this numerical model, which was programmed and solved in an Excel spreadsheet. Model validation showed that after calibration of two parameters the one-dimensional numerical model performed well in predicting temperature and mass change, local porosity and liquid holdup and lethality, etc., in each layer.;A comparison of heating time versus bed depth showed that the predicted and observed data had the same trend and were close to each other, which doubly confirmed that the model works well. |
