Pearson 1959

  • Upload
    anmar

  • View
    222

  • Download
    0

Embed Size (px)

Citation preview

  • 7/25/2019 Pearson 1959

    1/4

    Chemical Engineering S me. 1952, Vol. 10, pp. 281 to 284. Pergsmon Press Ltd. London

    Printed in Great Britain

    A note on the ( Danckwerts boundary conditions for continuous flow

    reactors

    J. R. A.

    PEARSON

    Imperial Chemical Industries Ltd., Akers Research Laboratories, The Frythe, Welwyn, Herts.

    Abstract-A mathematical justification is given for the use of the Danckwerts boundary

    conditions for continuous flow reactors. It is shown that the apparent indeterminacy, which

    DANCE~ERTSesolves intuitively, is caused by the use of a discontinuous coefficient of diffusion.

    By treating this as the limit of a cont~uous fusion and imposing eont~uity of the reactant

    concentration as the physically relevant boundary condition, the Danekwerts solution is obtained

    in the limit.

    RBsum&--Lauteur donne une justification mathematiques des conditions aux limites utilisees

    par DANCKWERTSans le cas dun reaeteur a Bcoulementcontinu.

    I1 montre que lindttermination

    apparente resolue ~t~tivement par D~~cxwmzrs est inherente it un coefficient de fusion

    discontinu.

    I1 con&d&e ce coefficient comme la limite dune fonction continue et il impose une

    continuite a la concentration du reactant comme &ant la condition limite physiquement correcte :

    la solution de DANCKWERTSest alors obtenue a la limite,

    Zusammenfassnng-Die G~~be~n~ngen nach DANCKWERTSfur den kon~uierlich durch-

    str6mten Reaktor werden mathematisch gerechtfertigt. Die anscheinende Unbestimmtheit, die

    DANCKWERTSntuitiv aufltist, ist durch die Verwendung eines diskontinuierlichen Diffusions-

    koeffizienten verursacht. Bebandelt man diesen als Grenzfall einer kontinuierlichen Funktion

    und setzt die Stetigkeit der Konzentration des Reaktanten als die physikalisch entscheidende

    Grenzbedingung fest, so erhalt man die DANCKWERTS-L~SIJNCm Grenzfall.

    1. INTRODUCTION

    IN an

    oft-quoted paper (Chem. Etgng. Sci.

    I953

    2

    I ,

    ANCKWERTS has considered the steady state

    ffow of r&&ant through a packed tubular reaction

    vessel in terms of a second order ordinary differen-

    tial equation (equation 30 in Eoc.

    cit. .

    This

    equation for the steady state concentration, c,

    of reactant in a first order reaction supposes that

    c is a function of one space variable only, y, the

    distance down the tube. The streaming velocity,

    21, the rate constant,

    k

    and the (constant) coeffi-

    cient of ~sion, 23, enter as parameters into

    the diffusion equation, which may thus be written

    = 0-t.

    (1)

    The boundary condition at the entry to the tube,

    y = 0, where the diffusion coefficient discon-

    tinuously changes from zero, is obtained by a

    consideration of mass balance, and is

    UC* = w: -

    D .

    dc

    qj

    Y =

    0,

    (2)

    where c* is the concentration in the entering

    stream.

    A similar relation is obtained for c at the exit

    from the tube, y = L, tihere D again changes

    discontinuously, but is replaced by DAMXWERTS,

    on intuitive grounds, by the stronger condition

    de

    - = 0

    dY

    y = L.

    Conditions (2) and (3) lead to a unique solution

    for c

    ;

    however this solution, as presented in

    t D~~c~vvnnrs notation is retained for ease of comparison.

    231

  • 7/25/2019 Pearson 1959

    2/4

    DANCKWERTS paper, appears to rely for its

    uniqueness on the acceptance of an intuitive

    boundary condition. If only because this boun-

    dary condition has not been universally accepted,

    it seems desirable to investigate a little more

    closely the formal. mathematical implications of

    the idealization represented by equation

    1)

    and

    the boundary conditions (2) and (3).

    First of all, we observe that in the general

    solution to l), using boundary condition (2) at

    y = 0 and a similar condition at y =

    L, a discon-

    tinuity in c, at either y = 0 or y = L, or both, is

    necessarily consequent upon the imposed discon-

    tinuities in

    D. Within the reactor, i.e. where

    0 < y