14A-45

M. CRISTIANINI, Department of Food Technology, University of Campinas, Faculty of Food Engineering, CP 6121, Campinas-SP, 13083-970, Brazil and P. R. Massaguer.

The effect of heat sterilization on microbial destruction and quality retention of food have always been a concern of food processors along the years. The use of retortable pouches has become significant recently due to their rapid heating characteristics. Accurate prediction of transient temperature distribution during heat processing of food products is important for process design and optimization. Accurate prediction of temperature history within the product is very dependent on the mathematical model and it can be used to evaluate nutrient loss and process lethality. Variation on processing temperatures or product thermal properties can strongly influence microbial destruction during heat sterilization of foods. The finite element method is a useful tool to model heat transfer on food products since it is applicable to objects having any shape or size. In this study a mathematical model (3-dimensional) using the finite element technique was built considering an actual retortable pouch shape (19mmx190mmx180mm). A factorial experimental design (2⁶) was used to evaluate the influence of retort temperature (120.1°C, 121.9°C), convective heat transfer coefficient (255.8 W/m²K, 342.6W/m²K), food initial temperature (23.5°C, 26.5°C), thermal conductivity (0.437W/m°C, 0.509W/m°C), specific heat (3314.3KJ/Kg°C, 3545.3KJ/Kg°C) and density (1003.2Kg/m³, 1045.8Kg/m³) on product critical point and the mass average sterilizing values (FP and FM, respectively). Statistical analysis showed that all variables were significant (95%). Retort temperature, thermal conductivity and convective heat transfer coefficient were the variables that most affected FP and FM. A 1.8°C change on retort temperature caused a variation of 2,77min (39,7%) on FP and 3,0min (37%) on FM. The two empirical models obtained to predict FP and FM had R2 values higher than 0.999. The results showed that a process originally designed to have a FP of 7min ended up with values ranging from 3.7 to 11.5min and FM from 4.6 to 12.8min.