Mathematical modeling of thermal processes: Effects on food safety and starch

Guibing Chen, Purdue University


Unavoidable process temperature deviations in retort operations can compromise product safety and quality as well as production efficiency if improperly handled. Two methods, the Fixed Point (FP) and the Worst Case (WC) methods, aimed to automatically correct the adverse effect of those process temperature deviations during thermal processes in the operations of continuous retorts were developed. These methods have the potential to quickly analyze and accurately estimate the effect of process deviations on the thermal lethality of canned foods processed in continuous retorts. The FP algorithm consists of the selection of a point that for convenience is located close to the retort steam chamber exit, whereas the WC algorithm consists of searching for the worst case container, defined as the one requiring the lowest conveyor speed. Simulations showed that the computation time for estimating conveyor speed adjustments using both algorithms was less than 1 second in a personal desktop Pentium IV computer. Accurate heat transfer calculations are the basis for developing suitable control algorithms. The Apparent Position Numerical Solution (APNS) method, which was used in the control algorithms developed in this study, did not provide a good temperature prediction in the initial stage of the process cooling phase. A mathematical method was developed to fix this problem. Heat transfer calculations using numerical methods rely on accurate heat penetration parameters. For thermal processes of starch based food systems, which exhibit a broken-line heating behavior, the heating parameters are strongly associated to the viscosity changes during the heating stage. Major factors affecting the viscosity of swollen crosslinked corn starch (CCS) suspensions were identified and a mathematical model to estimate changes of viscosity was developed in this study. A method to predict the momentary granule sizes, which are closely related to viscosity of a starch suspension, under non-isothermal conditions was also developed. Accurate evaluation of thermal lethality is essential for thermal process design and control. The Weibull model provides a better estimation of process lethality than conventional first-order kinetics. A numerical method suitable for on-line estimation of thermal lethality under non-isothermal conditions was also developed in this study.




Campanella, Purdue University.

Subject Area

Agricultural engineering|Food science

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server