Abstract
Abstract
The prediction of cleaning in pipe-lines is important for equipment manufacturers, who wish to optimize designs with respect to hygienic performance. Degree of cleaning correlates with the level of fluctuations in the signal recorded in discrete points during wall shear stress measurements using an electrochemical method. For optimization of process equipment with respect to cleaning, the levels of local fluctuations across entire surfaces are needed. Trends of fluctuations in the geometries used can be predicted using computational fluid dynamics (CFD). Two sensitivity studies were carried out to investigate the robustness of the CFD predictions: one with the zone of interest located two 90° bends and a straight pipe with a length to diameter ratio of 14:8 downstream of the inlet (full geometry) and one with only a straight pipe with a length to diameter ratio of 1:7 between the inlet and the zone of interest (shortened geometry). Both studies having a fully developed turbulent velocity inlet profile with changing turbulence parameters: turbulence intensity 0.01-30 per cent and turbulence length scale 7.5-30 per cent of inlet pipe diameter. For the full geometry no sensitivity in the fluctuation profile trend estimated by use of CFD was found, whereas for the shortened geometry a great sensitivity in the estimated fluctuation profile trend when changing inlet conditions was seen. In the process Star-CD and Fluent (used in the present study) were compared and showed comparable predictions of turbulence intensity.
Get full access to this article
View all access options for this article.