Different numerical and physical models for glass tanks were compared by a simple round robin. Most models produced qualitatively similar velocity and temperature distributions. The minimum residence time of tracers in numerical models gave similar values too, but differed to the physical scale model. An explanation for this is the different kind of batch modelling in both methods of simulation. Velocity vectors at some fixed test points showed larger deviations. In regions with steep gradients, small variations in the local flow distribution can lead to large changes at fixed test points. These differences therefore should not be rated very high.
Origin
Schott Glass (Germany) & Tno Inst Appl Physics (Netherlands)
Journal Title
Glastech Ber Glass Sci Tech 71 6 1998 153-156
Sector
General
Class
G 1868