A new method for designing optimum package size and processing conditions for re-tortable flexible or microwaveable packages has been specified. The method employed the random centroid optimization technique. The method also included a computer simulation
model, used to calculated process lethality and nutrient degradation, and an objective function used to calculate total process cost from the above values as well as from the processing parameters.
The objective function was original in that it was economicaly based, rather than concentrating solely on achieving processes with high product quality. Typical processing costs such as the cost of energy or materials were included, as well as more qualitative costs such as those associated with decreases in product quality. Most of these costs were inferred from conventional can processing costs, as information regarding processing costs for flexible or microwaveable retortable packages was often unavailable.
For the simulation model, a recently developed one dimensional finite difference technique
of high accuracy (the exponential finite difference method) was extended to three dimensions and used to solve the partial differential heat flow equation. The simulation model also comprised an original algorithm to model headspace and a new numerical integration technique to calculate lethality and nutrient degradation from the time temperature
profile.
The headspace model calculated headspace volume as a function of retort temperature and pressure and then used steady state heat transfer theory to model heat flow across the package headspace. The numerical integration technique fitted an exponential curve to the time temperature profile and then integrated this curve analytically.
The accuracies of the exponential finite difference method with and without headspace included and of the numerical integration technique were tested by comparison with both analytical solutions and other well established numerical methods. In addition, the uncertainty associated with the assumption of uniform paramaters was evaluated using a Monte Carlo technique.
After the validity of the simulation model was established, it was used in conjunction with the objective function and the random centroid optimization method to search for a global cost minimum. One trial optimization using a simple objective function was made, followed by three optimization runs using the more complex objective function desinged to calculate process cost. Each optimization used a different set of decision variables, which varied in number from three to five. The results were evaluated as to whether a minimum was found, and if so, whether it constituted a global minimum. / Science, Faculty of / Botany, Department of / Zoology, Department of / Graduate
| Identifer | oai:union.ndltd.org:UBC/oai:circle.library.ubc.ca:2429/29604 |
| Date | January 1990 |
| Creators | Greaves, Karen F. |
| Publisher | University of British Columbia |
| Source Sets | University of British Columbia |
| Language | English |
| Detected Language | English |
| Type | Text, Thesis/Dissertation |
| Rights | For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use. |
Page generated in 0.0021 seconds