Simulation of Solid -Liquid Two Phase Flow in Holding Tubes

碩士 === 國立海洋大學 === 食品科學系 === 89 === Abstract The study used computational fluid dynamics (CFD) software FLUENT to calculate solid/liquid flow in holding tube, and further discussed its simulation capability.The result of FLUENT simulation agreed with analytical and experimental...

Full description

Bibliographic Details
Main Authors: Yue_Roo Huang, 黃毓儒
Other Authors: Wen-Rong Fu
Format: Others
Language:zh-TW
Published: 2001
Online Access:http://ndltd.ncl.edu.tw/handle/73010711798247550075
Description
Summary:碩士 === 國立海洋大學 === 食品科學系 === 89 === Abstract The study used computational fluid dynamics (CFD) software FLUENT to calculate solid/liquid flow in holding tube, and further discussed its simulation capability.The result of FLUENT simulation agreed with analytical and experimental results from literature for fluid phase in laminar flow. Particle trajectories in straight tube flow showed similar tendency when comparing with results from literature. Difficulties arose when particle flowing near tube wall region and/or colliding with tube wall in U-bend section. Particles would travel along different trajectories even when they were released at the same radial position at the entrance. Two Layer Zonal and Standard k-ε models worked reasonably well for straight tube flow and provided comparable results to experimental data. The mean and standard deviation of a single particle residence time distribution decreased with increasing Reynolds number for both horizontal tube inclined 1.194o upward and vertical tube. The mean and standard deviation of dimensionless residence time distribution increased with solid/liquid density ratio for horizontal tube flow. The opposite was observed for vertical tube flow. Dimensionless standard deviation in vertical tube flow was always less than that found in horizontal tube flow, and the increase of solid/liquid density ratio augmented the difference. FLUENT is not capable of outputting results for individual particle. Users have to write their own post-processing program, if such is desired. Flow in U-bend was calculated by Two Layer Zonal and RNG k-εturbulent models. The constant axial velocity mapping at the cross-section of U-bend mid-way showed a positive pressure gradient (adverse pressure gradient) was present. A flow separation was likely to be present near the inside of the U-bend. But the point where the minimum axial velocity located did not shift with Reynolds numbers studied. Incorporation of particle colliding effects, replacing trajectory integration method and particle-fluid coupling are necessary to improve application of FLUENT in aseptic processing simulation.