Numerical simulation and optimization of commercial sterilization of liquid enteral nutrition using batch microwave heating

Abstract

This thesis contains 3 parts. The first part was to investigate the effect of ingredients on product’s stability and to find optimized oral-feeding enteral nutrition formulas. Since oral-feeding enteral nutrition is high in caloric density, it contains a high amount of calorie sources. Thus, it is likely to have either too high viscosity and/or unstable emulsion. Consequently, 9 factors which are caloric content, oil type, lecithin concentration, fructose syrup to acid ratio, percentage of hydrolyzed whey protein to total protein, percentage of calorie from fat and from fructose syrup to total calorie, complex whey protein concentration, and percentage of whey protein isolate to total complex whey protein were optimized. The desired responses were the complex viscosity at 50 Hz and 25 ºC, and the percentage of emulsion separation. By using the Iconographic Correlation method (IC), the number of experiment was reduced from 524 treatments for classical response surface methodology (RSM) with Doehlert matrix (DM) to 17 treatments. IC proposed models described of significant logical interactions between factors and responses with excellent correlation (R2adj = 0.99, and 0.93 for complex viscosity, and emulsion separation, respectively). The second part aimed to investigate the effect of microwave heating on the enteral nutrition and finding optimal conditions for heating tube-feeding formula. DM with 5 heating times (35 s to 60 s) and 3 specific powers (3 W/mL to 7 W/mL) was assigned to investigate the effect of microwave heating for a pouch of 150-mL product contained 1 kcal/mL and had the caloric distribution from carbohydrate: fat: protein at 50: 30: 20. The surface temperature measured by an infrared camera, relative tryptophan loss and FAST index by fluorometric spectroscopy were determined as responses. The FAST index and the change in average surface temperature were correlated with heating time and specific power, respectively. Comparing the models proposed by RSM and IC, RSM gave more predictive models than those of IC did for most responses. The last part involved the development of numerical simulation for microwave heating process. Proximate composition, physical, thermal, rheological, and dielectric properties of 1 kcal/mL, 2.5 kcal/mL, and 3.78 kcal/g liquid enteral nutrition products were determined. Products in a retortable pouch were heated at 450 W or 850 W by continuous heating or intermittent heating at different heating times by a 2,450-MHz microwave oven. Higher caloric density had higher heating rate and heterogeneity in temperature distribution while intermittent heating enhanced the homogeneity of heating. In addition, heating did not increase the FAST index of the samples. A numerical model was developed by using the Finite Element Method to solve a convective heat transfer in fluid coupling with electromagnetic propagation. The model proposed the conditions reaching the commercial sterilization for the products. Moreover, it can predict the average surface temperature for continuous heating and the FAST index, but not the temperature distribution and the temperature profile from intermittent heating.