Premium
Calculation of soret‐shifted dew points by continuous mixture thermodynamics
Author(s) -
Rosner Daniel E.,
AriasZugasti Manuel,
LaMantia Barbara
Publication year - 2005
Publication title -
aiche journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.958
H-Index - 167
eISSN - 1547-5905
pISSN - 0001-1541
DOI - 10.1002/aic.10515
Subject(s) - thermodynamics , dew point , chemistry , isobaric process , dew , binary number , thermophoresis , combining rules , gaussian , statistical physics , physics , mathematics , computational chemistry , condensation , heat transfer , nanofluid , arithmetic
Soret‐induced local concentration changes, very large numbers of condensable species, and condensate nonideality combine to make the calculation of isobaric dew‐point surface temperatures, T dp , and corresponding incipient condensate compositions (x i , i = 1, 2, …, N) nontrivial, even with advanced computational methods. As usual in the canonical isobaric “dew‐point ” situation, the mainstream vapor mixture state [composition y i (i = 1, 2, …, N), T, and p] is fully specified. We consider here, for illustrative purposes, the tractable but frequently encountered limiting case of a very large number (perhaps dozens, if not hundreds) of condensable, nonreacting species that are members of a homologous series (such as n‐alkanes in a fuel “blend”), dilute in a “noncondensable ” carrier gas (such as N 2 ). Pseudo‐binary Ludwig–Soret coefficients for the vapor phase are estimated from the corresponding Schmidt numbers and gas‐kinetic theory. As in our earlier studies (which emphasized binary “regular ” solutions of alkali sulfates), of particular current interest is the extent of systematic Soret‐induced shifts in the predicted dew‐point (wall) temperature and associated condensate composition/properties (such as viscosity). This often‐overlooked feature removes the present class of univariate problems from the province of classical continuous thermodynamics. Three efficient continuous mixture theory (CMT)–based computational methods are proposed/illustrated/compared, two moment‐based, and all exploiting Gaussian quadratures. We find that accurate results can be obtained while avoiding prescribing the shape of the distribution function in the condensate, using as few as three to four rationally selected “pseudo‐components.” © 2005 American Institute of Chemical Engineers AIChE J, 2005