The effects of oscillatory flow and bulk flow components on residence timedistribution in baffled tube reactors

Citation
P. Stonestreet et Pmj. Van Der Veeken, The effects of oscillatory flow and bulk flow components on residence timedistribution in baffled tube reactors, CHEM ENG R, 77(A8), 1999, pp. 671-684
Citations number
30
Categorie Soggetti
Chemical Engineering
Journal title
CHEMICAL ENGINEERING RESEARCH & DESIGN
ISSN journal
02638762 → ACNP
Volume
77
Issue
A8
Year of publication
1999
Pages
671 - 684
Database
ISI
SICI code
0263-8762(199911)77:A8<671:TEOOFA>2.0.ZU;2-D
Abstract
A characteristic of oscillatory flow mixing in a baffled tube is that the r esidence time distribution performance can be affected independently of net flow conditions. The effects of both the oscillatory velocity and the thro ughput velocity on the residence time distribution performance has been inv estigated in a 24 mm diameter, 2.8 m long, baffled tube oscillatory flow re actor. The experiments were performed by applying the tanks-in-series model to 'perfect pulse' tracer experiments over a wide range of oscillatory con ditions and flow rates. An optimum set of oscillatory and net flow conditio ns was found which resulted in near plug flow behaviour, usually 50 tanks-i n-series (N) or greater. For a throughput velocity, N was found to be sensi tive to both the amplitude and frequency of oscillation, and this dependenc e could be expressed by means of the oscillatory flow Reynolds number (Re-0 ), which combines both parameters. A unique value for Re-0 for each value o f the net flow Reynolds number (Re-n) gave the closest approach to a plug f low RTD. To relate the oscillatory and net flows, a dimensionless velocity ratio, psi, was used, defined as Re-0/Re-n. In order to relate the RTD perf ormance for different flow rates, the tanks-in-series model was non-dimensi onalized by the use of a stage-wise efficiency term, eta, defined as the ra tio of N to the theoretical number of stages. Over the range of oscillatory and net flow conditions studied, it was found that the range 2 less than o r equal to psi less than or equal to 4 corresponded to the optimum RTD cond itions, where efficiencies close to 1 were achieved. It was concluded that these dimensionless parameters were sufficient to select, a priori, the osc illatory parameters necessary to obtain the optimum RTD in an oscillatory f low reactor based on a desired throughput specification.