Modelling microwave cooking; theory and experiment

Microwave heating is an essential technique for cooking a wide variety of foodstuffs in both domestic and commercial applications. However, the simulation of this process remains problematic because of the complexity of both calculating a varying electromagnetic field and also determining how this interacts with the foodstuff (which is typically a complex multi-phase material). In one-dimension models have been employed to investigate moisture movement and moisture dependent parameters with ??, ??, [?] examining the use of the Lambert law approximation to the electric field. In two-dimensions models have been employed to investigate thermal runaway through phase changes and temperature dependent parameters in Nylon ??. and a paper by Zeng and Fahgri ?? extends Lambert’s law to two dimensions to investigate heating effects inside cylindrical samples of food . In three dimensions complex numerical simulations have been employed to investigate standing wave patterns set up in the microwave cavity. For example Metaxas [] has investigated the changing field patterns in turntable microwave ovens and the research team at the University of Greenwich has coupled full three dimensional calculations of the electromagnetic field with CFD calculations for the heat and moisture transport with a phase change from solid to liquid [],[]. This simulation also considers the changing dielectric properties of the food arising from heating and phase changes. Such complex calculations necessarily take a long time (typically hours or longer), even on fast computers, making fully three dimensional based computations of the microwave cooking process cumbersome to use for design calculations. A significant conclusion of the three dimensional field simulations is that they reveal that the fields in a typical microwave cavity (such as a domestic cooker) are very sensitive to small system variations. Thus, not only is a field calculation very expensive, if may not be particularly representative for any particular geometry. Indeed a simpler, averaged, model may give more representative solution to the problem of simultating the microwave heating of a foodstuff.