Time-dependent modelling of in-package pasteurization.

Brandon, H., Gardner, R., Huling, J. and Staack, G.

Abstract
The purpose of tunnel pasteurisation is to increase shelf life of beer by destroying spoilage bacteria. Heat is transferred from hot water, sprayed over each container forming a film of water, to the contents by forced convection. Pasteurisation is a time dependent process as beer temperature varies with time spent in the pasteuriser. Flow patterns of beer wtihin container during pasteurisation are complex. There are three main regions: a) mixing region at the top of the container, b) the main core, c) the thin boundary along the wall of the container. As the boundary layer moves up the wall of the container it passes inwards just below the surface of the liquid and mixes with the cooler fluid in the main core. In the main core temperature changes are mainly in the axial direction. This flow field has been studied via a theoretical model and flow patterns have been further examined by use of thermocouples. For the theoretical model the product was split into many small elements (360 for a 12 oz can and 620 for a 12 oz bottle). Equations of mass, momentum and energy conservation were solved for each element. From the data time-dependent temperature and fluid velocity distributions were expressed graphically. External heat input into the container was defined as the container's surface temperature distribution. Surface temperature data were measured by use of a thermocouple. during the heating cycle the can surface had a large axial temperature gradient which disappeared after 10 minutes. For the first five minutes of heating with the can upright the contents heat up faster at the bottom than when the can was inverted, after which temperature distributions are the same. The opposite effect was noted in the main core. These effects are associated with the 'V' at the base of the can. Graphs of ulow velocities and temperature distributions are presented for contents of a 12 oz can (inverted) and a 12oz bottle one and two minutes after heating from 35 degrees F. After two minutes regular temperature distributions were achieved. Conduction during pasteurisation also heats the base of the can; thus warm liquid at the bottom will mix with that passing down the main core. Movement of liquid from the bottom of a bottle was examined by a flow visualisation technique. Photographs showed there was a lag phase of 61 seconds when non uniform heating occurred across the base. Two seconds later vortex motion at the centre carries liquid up the side of the bottle, followed after another two seconds by a mushroom flow pattern. At 3, 6.5 and 9 minutes, thermals (regions of buoyant liquid suddenly released upwards) were noted. Thermals were spaced regularly across the heated surface. The output of the mathematical model gave a flow pattern consistent with experimental data. Although costs of computers mean that laboratory work will continue as the main source of information about pasteurisation, the computer model can give extra data to improve understanding of the process.
Keywords: beer flow glass bottle model simulation pasteurisation