Steel is commonly used for food packaging. It is usually protected by means of various coatings and lacquers. However, corrosion resulting from foodstuff/metal interactions, and ultimately can degradation and failure is of concern in the packaging industry. In this study, an investigation into the effects of different food products on the corrosion of steel for food cans has been carried out. Both real and simulated food products have been used to gain an understanding of critical constituents of food products in relation to corrosion of steel used in food cans. I-dimensional artificial pits were used to study localised corrosion processes related to the corrosion of the steel used for making food cans. Experimental conditions allowed for controlled electrochemical tests to be carried out, while simulating the foodstuff/steel interaction within a can. Test conditions similar to a can environment, namely deaeration and pH did not have a significant effect on the corrosion of the artificial pits studied. Data obtained for characteristics of iron artificial pits fitted well with theory. Known "problem" food products were investigated, and some individual constituents were shown to have a significant effect on the corrosion of steel. During observation of the effects of fish products and fruit juices on the corrosion of iron artificial pits, acetic acid and nitrate were shown to play significant roles in the corrosion behaviour of the artificial pits. Significant levels of these constituents were found to be detrimental to steel corrosion in relation to food cans. Other constituents, for example sugars, were found to suppress corrosion of the steel.

No description available.

No description available.

No description available.

No description available.

This thesis concerns the controllability of fermentation processes. Fermentation processes are often described by unstructured process models. A control system can be used to reduce the effect of the uncertainties and disturbances. A process is called controllable if a control system satisfying suitably defined control objectives can be found. Controllability measures based on linear process models are identified. The idealised control objective for perfect control allows fast evaluation of the controllability measures. These measures are applied to compare different designs of a continuous fermentation process by identifying the controllability properties of the process design. The operational mode of fed batch fermentations is inherently dynamic. General control system design methods are not readily applicable to such systems. This work presents an approach for the design of robust controllers suitable for these processes. The control objective is to satisfy a set of robustness constraints for a given set of model uncertainties and disturbances. The optimal operation and design problems are combined into a single optimal control problem. The controller design is integrated into the process design problem formulation. In this way the control system and the process are designed simultaneously. Different problem formulations are investigated. The proposed approach is demonstrated on complex fermentation models. The resulting operating strategies are controllable with respect to the aims of control.

The whole of the thesis is motivated by a particular problem from the food-sorting industry in which grains of food, typically rice, flow down chutes. As they fall down the chute, the grains form a rapidly moving mono-layer. This project starts with a discrete model treating individual grains based on partcle dynamics without air resistance. Single grains and then many grains are addressed, and the methodology used includes large computations describing the particle paths, velocities and other key features accompanied by analysis. Much of the thesis thereafter is concerned with the development of a continuum model for the chute flow of grains, inspired by the clusters and voids seen in the above computations and based on analogies with the Lighthill-Whitham model of traffic flow. The crucial difference here is that the flow is not uni-directional and so a multi-valued flow-density relation is required. The introduction of such a law yields complex and rich flow behaviours. The mathematical interest is in solving hyperbolic and parabolic partial differential equations, incorporating shocks and fans into analytical and numerical solutions of the governing equation, in asymptotic methods used in analysis of particular limits, especially those relating to clashing or separation of grains, and in seeking steady state solutions for the density profile. Finally, air effects are studied. In particular, local viscous or inviscid effects in the gaps between grains are investigated, followed by the subsequent interaction effects on many grains.

No description available.

No description available.

No description available.