CFD-ASSISTED THIN CHANNEL MEMBRANE CHARACTERIZATION MODULE DESIGN

Citation
K. Darcovich et al., CFD-ASSISTED THIN CHANNEL MEMBRANE CHARACTERIZATION MODULE DESIGN, Journal of membrane science, 124(2), 1997, pp. 181-193
Citations number
13
Categorie Soggetti
Engineering, Chemical","Polymer Sciences
Journal title
ISSN journal
03767388
Volume
124
Issue
2
Year of publication
1997
Pages
181 - 193
Database
ISI
SICI code
0376-7388(1997)124:2<181:CTCMCM>2.0.ZU;2-J
Abstract
This project involved the design of a thin channel cross-flow module f or the characterization of flat ceramic membranes. A primary objective of this work was to ensure that the flow characteristics over the per meating area were uniform. To house these membranes, a thin channel mo dule with a long rectangular base was envisioned. The module feed is s upplied by a multi-inlet tube-type plenum meant to provide a uniform f low distribution through pressure equilibration attained in its volume . The design criteria for the module were minimization of both the flo w non-uniformity and the pressure drop across the permeating area whic h was a central rectangular portion of a larger slab-style cell, The f low non-uniformity was taken as the normalized standard deviation of t he velocity field above the permeating area, The pressure drops consid ered were those across the inlet plenum and across the permeating area normalized with respect to the outlet pressure. The computational flu id dynamics (CFD) scheme which calculated the above module characteris tics was a k-epsilon based turbulent transport model which used the fi nite difference method. Design variables considered were: the plenum d iameter, module width, height and length, and the diameters, distribut ion and number of the inlets on the plenum. The distribution of the in let diameters was determined by two variables: either a linear or para bolic profile of variable slope and model coefficient. The operating v ariables were the cross-flow velocity and the plenum inlet pressure. A two-level factorial design was used to screen the design variables. A refined three-level factorial design was used with a reduced set of d esign variables to optimize the module and study the response surface. The final module design parameters were chosen such that the design c riteria of flow uniformity and low pressure drop were met under a pres et range of operating conditions. The local gradients of the response surface were used to verify that the design criteria were not overly s ensitive to the selected module design parameters.