See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/234095414

Derivation and application of the Stefan-Maxwell equations

Article in Revista mexicana de ingeniería química · December 2009

CITATIONS READS

40 5,657

1 author:

Stephen Whitaker
University of California, Davis
237 PUBLICATIONS 16,353 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Diffusion of Charged Species in Liquids View project

Interpore Interview with Prof. Stephen Whitaker View project

All content following this page was uploaded by Stephen Whitaker on 26 May 2014.

The user has requested enhancement of the downloaded file.

 Revista Mexicana de Ingeniería Química
Vol. 8, No. 3 (2009) 213-243

DERIVATION AND APPLICATION OF THE STEFAN-MAXWELL EQUATIONS

DESARROLLO Y APLICACIÓN DE LAS ECUACIONES DE STEFAN-MAXWELL

Stephen Whitaker*

Department of Chemical Engineering & Materials Science

University of California at Davis

Received 5 of June 2009; Accepted 9 of November 2009

Abstract

The Stefan-Maxwell equations represent a special form of the species momentum equations that are used to
determine species velocities. These species velocities appear in the species continuity equations that are used to
predict species concentrations. These concentrations are required, in conjunction with concepts from
thermodynamics and chemical kinetics, to calculate rates of adsorption/desorption, rates of interfacial mass
transfer, and rates of chemical reaction. These processes are central issues in the discipline of chemical engineering.
In this paper we first outline a derivation of the species momentum equations and indicate how they simplify
to the Stefan-Maxwell equations. We then examine three important forms of the species continuity equation in
terms of three different diffusive fluxes that are obtained from the Stefan-Maxwell equations. Next we examine the
structure of the species continuity equations for binary systems and then we examine some special forms associated
with N-component systems. Finally the general N-component system is analyzed using the mixed-mode diffusive
flux and matrix methods.

Keywords: continuum mechanics, kinetic theory, multicomponent diffusion.

Resumen

Las ecuaciones de Stefan-Maxwell representan una forma especial de las ecuaciones de cantidad de movimiento de
especies que son usadas para determinar las velocidades de especies. Estas velocidades de especies aparecen en las
ecuaciones de continuidad de especies que son usadas para predecir las concentraciones de especies. Estas
concentraciones son requeridas, en conjunción con los conceptos de termodinámica y cinética química, para calcular las
velocidades de adsorción/desorción, las velocidades de transferencia de masa interfacial, y las velocidades de reacción
química. Estos procesos son elementos centrales en la disciplina de la ingeniería química.
En este artículo presentamos primeramente un desarrollo de las ecuaciones de cantidad de movimiento de
especies e indicamos como se simplifican a las ecuaciones de Stefan-Maxwell. Posteriormente examinamos tres formas
importantes de la ecuación de continuidad de especies en términos de tres diferentes fluxes difusivos que se obtienen de
las ecuaciones de Stefan-Maxwell. Más adelante examinamos la estructura de las ecuaciones de continuidad de especies
para sistema binarios y examinamos algunas formas especiales asociados con sistemas de N-componentes. Finalmente se
analiza el sistema general de N-componentes usando métodos matriciales y de flux difusivo de modo mixto.

Palabras clave: mecánica del continuo, teoría cinética, difusión multicomponente.

* Corresponding author. E-mail: whitaker@mcn.org 213

Publicado por la Academia Mexicana de Investigación y Docencia en Ingeniería Química A.C.

 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

Contents

1. Introduction 2

1.1 Conservation of mass 3

1.2 Laws of mechanics 5

2. Mass continuity equation 15

3. Molar continuity equation 16

4. Mixed-mode continuity equation 19

5. Binary systems 20

5.1 Mass diffusive flux 20

5.2 Molar diffusive flux 23

5.3 Mixed-mode diffusive flux 25

6. Special forms for N-component systems 26

6.1 Dilute solution diffusion equation 26

6.2 Dilute solution convective-diffusion equation using J*A 28

6.3 Dilute solution convective-diffusion equation using J A 30

6.4 Diffusion through stagnant species 31

7. General form for N-component systems: Constant total molar concentration 33

8. General form for N-component systems: Constant total mass density 37

9. Conclusions 41

Nomenclature 41

Acknowledgment 41

References 44

Appendix A: Chemical reaction and linear momentum 46

Appendix B: Thermodynamic pressure 50

Appendix C: Algebraic relations 55

Appendix D: Assumptions, Restrictions and Constraints 58

Appendix E: Restrictions for constant total concentration and density 62

214 www.amidiq.org
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

1. Introduction
Our derivation of multi-component transport
equations is based on the concept of a species body.
In Part I of Fig. 1 we have illustrated a system
containing both species A and species B and these
are illustrated as discrete particles. We have also
illustrated a region from which we have cut out both
a species A body and a species B body. In Part II of
Fig. 1 we have indicated that the species A body will
be treated as a continuum while the discrete
character of species B is retained for contrast. As
time evolves the two species bodies separate because
their velocities are different. This separation is
illustrated in Part III of Fig. 1 where we have also
indicated that the species B body will be treated as a
continuum. The continuum velocities of species A
and species B are designated as vA and vB. In
general, the continuum hypothesis should be
satisfactory when the distance between molecules is
very small compared to a characteristic length for
the system.

1.1 Conservation of mass

In terms of the species A body illustrated in Fig. 1,
we state the two axioms for the mass of multi- Fig.1. Motion of species A and species B bodies
component systems as
In order to extract a governing differential equation
Axiom I:
from Eq. (1), we make use of the general transport
d equation (Whitaker, 1981, Sec. 3.4, with w  v A )
dt VA(t )
 A dV   rA dV , A  1, 2, ...., N (1)
VA (t ) d  A
dt V A( t )
 A dV   dV
A N VA ( t )
t
(5)
Axiom II: r A 0 (2)
A 1  
AA ( t )
 A v A  n dA , A  1, 2, ...., N
Here  A represents the mass density of species A
and rA represents the net mass rate of production per and the divergence theorem (Stein and Barcellos,
unit volume of species A owing to chemical reaction. 1992, Sec. 17.2)
In Eqs. (1) and (2) we have used a mixed-mode
nomenclature making use of both letters and   A v A  n dA       A v A  dV ,
numbers to identify individual species. For example, AA ( t ) VA (t ) (6)
Axiom II could be expressed in terms of alphabetic A  1, 2, ...., N
subscripts as
in order to express Eq. (1) in the form
Axiom II: rA  rB  rC  rD  .....  rN  0 (3)
  A 
or we could use numerical subscripts to represent   t      A v A   rA  dV  0 , (7)
VA (t )
this axiom as
A  1, 2, ...., N
Axiom II: r1  r2  r3  r4  .....  rN  0 (4)
Since V A (t ) illustrated in Fig. 1 is arbitrary, and
This latter result can obviously be compacted to since it is plausible to assume that the integrand in
produce Eq. (2); however, the use of alphabetic Eq. (7) is continuous, the integrand in Eq. (7) must
subscripts to represent molecular species is prevalent be zero and the governing differential equation
in the chemical engineering literature. Because of associated with Eq. (1) is given by
this we will use alphabetic subscripts to identify
A
distinct molecular species, and we will use the    (  A v A )  rA , A  1, 2, ..., N (8)
nomenclature contained in Eq. (2) to represent the t
various sums that appear in this paper.

www.amidiq.org 215
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

If we sum Eq. (8) over all species and impose the d

dt V A( t ) 
Axiom II we obtain  A v A dV   A b A dV
VA ( t )


   ( v ) = 0 (9) BN
t Axiom I:  
AA ( t )
t A( n ) dA   P
V A ( t ) B 1
AB dV (15)
in which the total density and the total mass flux are
determined by  
VA (t )
rA vA dV , A  1, 2 , ..., N
A N A N
  
A1
A , v  
A1
A vA (10)
With an appropriate interpretation of the
nomenclature, one finds that this result is identical to
The mass average velocity, v, can be expressed in the second of Eqs. 5.10 of Truesdell (1969, page 85)
terms of the mass fraction, A, and the species provided that one interprets Truesdell’s growth of
velocity, vA, according to linear momentum as the last two terms in Eq. (15). In
A N terms of the forces acting on species A, we note that
v 
A 1
A v A , A   A  , A  1, 2,..., N (11)  A b A represents the body force, t A(n ) represents the
surface force, and PAB represents the diffusive force
The typical treatment of Eq. (8) involves the solution
exerted by species B on species A. This diffusive
of N  1 species continuity equations along with a
force is constrained by
solution of Eq. (9). This suggests a decomposition of
the species velocity into the mass average velocity, PAA  0 , A  1, 2, 3,..., N (16)
v, and the mass diffusion velocity, uA
The last term in Eq. (15) represents the increase or
vA  v  uA , A  1, 2, ..., N (12) decrease of species A momentum resulting from the
so that the species continuity equations take the form increase or decrease of species A caused by chemical
reaction, and this term is discussed in Appendix A.
A
   (  A v )     (  Au A )  rA , The angular momentum principle for the
t    species A body is given by
 convective diffusive chemical (13)
accumulation transport transport reaction
d
dt V A( t )
A  1, 2, ...., N  1 r   A v AdV   r   Ab AdV
VA (t )
Here we note that only N  1 of the diffusive
transport terms are independent since Eqs. (10) and BN

(12) require the constraint Axiom II:  

AA ( t )
r  t A( n ) dA    rP
V A ( t ) B 1
AB dV (17)

A N

 A1
A uA  0 (14)   r  rA v A dV , A  1, 2, ..., N
VA ( t )

In order to solve Eqs. (9) and (13) we need in which r represents the position vector relative to a
governing differential equations for the mass fixed point in an inertial frame. Truesdell (1969,
diffusion velocity, uA, and the mass average velocity, page 84) presents a more general version of Axiom II
v. These are determined by the axioms for the in which a growth of rotational momentum is
mechanics of multi-component systems. included, and Aris (1962, Sec. 5.13) considers an
analogous effect for polar fluids. The analysis of
1.2 Laws of mechanics Eq. (17) is rather long; however, the final result is
simply the symmetry of the species stress tensor as
Our approach to the laws of mechanics for multi-
indicated by
component systems follows the work of Euler and
Cauchy (Truesdell, 1968), the seminal works of TA  TA , A  1, 2, ..., N
T
(18)
Chapman & Cowling (1939) and Hirschfelder,
Curtiss & Bird (1954), along with the recent work of The constraint on PAB is given by Truesdell (1962,
Curtiss & Bird (1996, 1999). In terms of the species Eq. (22) as
body illustrated in Fig. 1 the linear momentum A N B  N
principle for species A is given by Axiom III:  P
A1 B 1
AB  0 (19)

and a little thought will indicate that this is satisfied

by
PAB   PBA (20)

216 www.amidiq.org
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

One can think of this as the continuum version of This result is identical to Eq. 4.20 of Bearman and
Newton’s third law of action and reaction Kirkwood (1958) for the case in which rA  0
(Whitaker, 2009a). provided that one takes into account the different
Hirschfelder et al. (1954, page 497) point out nomenclature indicated by (with the subscript
that “even in a collision which produces a chemical   A)
reaction, mass, momentum and energy are
conserved” and the continuum version of this idea Bearman and Kirkwood:
for linear momentum gives rise to the constraint:
 Ab A  c A X A ,    TA   Au Au A      σ A ,
A N (27)
Axiom IV: r A v 
A  0 (21) BN
A1
P
B 1
AB  c AFA(1)
This result, along with Eq. (19), is contained in the
second of Eqs. 5.12 of Truesdell (1969). Bearman and Kirkwood refer to σ A as the partial
Returning to the linear momentum principle, stress tensor and note that it consists of a “molecular
we note that the analysis associated with Cauchy’s force contribution” represented by  TA and a
fundamental theorem (Truesdell, 1968) can be “kinetic contribution” represented by  A u Au A .
applied to Eq. (15) in order to express the species
stress vector in terms of the species stress tensor Equation (23) can be represented in more
according to compact form using the species continuity equation.
We begin by multiplying Eq. (8) by the species
t A( n )  n  TA (22) velocity to obtain
This representation can be used in Eq. (15), along   
with the divergence theorem and the general v A  A    (  A v A )   rA v A , A  1, 2, ...., N (28)
  t 
transport theorem, to extract the governing
differential equation for the linear momentum of
Subtraction of this equation from Eq. (23) leads to
species A given by
  vA 
 A   v A  v A    Ab A    TA
  Av A   

   A v A v A    Ab A  
     TA   t 
t
  body surface (29)
convective
local acceleration force force
BN
 rA  v A  v A  ,
acceleration

BN
(23)  P
B 1
AB A  1, 2, ..., N
 P AB  rA vA
 , A  1, 2, ..., N
Bird (1995) has pointed out that Chapman and



B 1 source of momentum
diffusive owing to reaction Cowling (1939) first obtained this result 1 for dilute
force
gases by means of kinetic theory provided that
Equation (23) is identical to Eq. A2 of Curtiss and rA  0 . From the continuum point of view, Eq. (29)
Bird (1996) for the case in which rA  0 provided is given by Truesdell and Toupin (1960, Eq. 215.2),
that one takes into account the different Truesdell (1962, Eq. 22), and Curtiss and Bird (1996,
nomenclature indicated by Eqs. 7b and A7) all with rA  0 . The correspondence
with Truesdell (1962) is based on the nomenclature
 Ab A  G A ,   TA     σ A ,
Curtiss & Bird: (24)  Ab A   Af A ,   TA  div t A ,
BN Truesdell: (30)
P AB  FA BN
B 1
P
B 1
AB   pˆ A
One can make use of the identity
In its present form, Eq. (29) represents a governing
 A v A v A   A  v A v  vv A  vv    Au Au A (25) equation for the species velocity, vA, and we want to
use this result to derive a governing equation for the
in order to express Eq. (23) in the from mass diffusion velocity, uA. To carry out this
derivation, we need the total momentum equation

  A v A       A  v A v  v v A  v v  that is developed in the following paragraphs.
t
  Ab A     TA   Au Au A  (26)
BN
 P
B 1
AB  rA vA , A  1, 2, ..., N 1
See the species momentum equation following Eq. 6 on
page 135.

www.amidiq.org 217
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

 v   A N 
1.2.1 Total momentum equation   v v    b       TA   Au A u A   (38)
  t   A 1 
The traditional analysis of momentum transport in
multi-component systems makes use of the sum of In order to use this result to predict the mass average
Eqs. (23) over all N species to obtain the total velocity, we need a constitutive equation for the sum
momentum equation that is used to determine the of the species stress tensors. This problem is
mass average velocity, v. The remaining N  1 considered in the following paragraphs.
independent species momentum equations can then
used to determine the individual species velocities, 1.2.2 Governing equation for the mass diffusion
v A , v B , … v N 1 . We begin by taking into account velocity
Axioms III and IV so that the sum of Eq. (23) leads Our objective here is to develop the governing
to differential equation for the mass diffusion velocity,
uA. We begin by multiplying Eq. (38) by the mass
 A N A N
fraction A
t

A1
A v A      Av Av A
A1
(31)  v 
A N A N
A   v  v    Ab
 t 
 
A1
A b A     TA
A1
(39)
 A N

The first and third terms in this result can be   A     TA   Au Au A  
 A1 
simplified by the definitions
A N A N and subtracting this result from Eq. (29) to obtain the

A1
Av A   v , 
A1
Ab A   b (32) desired governing differential equation given by

 u A 
and Eqs. (10) and (12) can be used to obtain A   v A  u A  u A   v    A ( b A  b )
  t 
A N A N


A1
A v A v A   vv  
A1
A v Au A (33)  A N 
  TA   A     TA   Au Au A   (40)
 A1 
Application of Eq. (14) allows us to simplify the
BN
convective momentum transport to the form  P AB  rA  v A  v A  , A  1, 2,..., N  1
A N A N B 1


A1
A v A v A   vv  
A1
A u Au A (34) Here it is important to note that this result is based
only on the two axioms for mass given by Eqs. (1)
and substitution of Eqs. (32) and (34) in Eq. (31) and (2), and the four axioms for the mechanics of
provides multi-component systems given by Eqs. (15), (17),
(19) and (21). In addition, we have made use of
  A N 
  v       vv    b       TA   Au Au A  (35) classical continuum mechanics to obtain the result
t  A 1  given by Eq. (22).
At this point we need to be specific about the
Concerning the last term in this result, we note that
species stress tensor, TA , and to guide our thinking
Truesdell and Toupin (1960, Sec. 215) refer to
 Au Au A as the “apparent stresses arising from and constrain the subsequent development, we
propose that:
diffusion” and we note that this term also appears in
the analysis of Curtiss and Bird (1996, Eq. A7). In The analysis is restricted to mixtures that
that case one needs to make use of the second of behave as Newtonian fluids (Serrin, 1959,
Eqs. (24) along with Sec. 59; Aris, 1962, Sec. 5.21).
A N A N
Given this restriction for the mixture, we follow
π   π A     TA   Au Au A 
A1 A1
(36) Slattery (1999, Sec. 5.3) and write
A N

to complete the correspondence. At this point we can

T  T
A1
A   pI   (41a)
use Eq. (9) to obtain
in which p is the thermodynamic pressure and  is
  the extra stress tensor given by (Serrin, 1959,
v    ( v )  =0 (37)
 t  Eq. 61.1; Slattery, 1999, Eq. 5.3.4-3; Bird et al.,
2002, page 843)
and this allows us to express Eq. (35) in the form

218 www.amidiq.org
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

    v  v T    (  v ) I (41b) In Appendix A we show that difference between vA

and v A should be on the order of the diffusion
Given these results, Eq. (38) provides the Navier- velocity
Stokes equations containing an additional term
associated with the sum of the diffusive stresses. v 
A  v A   O( u A ) (47)
Here we need to point out that Eqs. (41) can
be obtained by following a classic continuum Arguments are given elsewhere (Whitaker, 1986,
mechanics analysis; or this result can be obtained 2009b) indicating that several of the terms in
from kinetic theory (Hirschfelder, Curtiss & Bird, Eq. (46) are generally negligible. This leads to the
1954, Eqs. 7.2-45 and 7.6-29). The advantage of this simplifications given by
latter approach is that a method of calculating the uA
coefficients  and  is created within the framework A  p A (48a)
of the theory. The disadvantage is that the t
calculations associated with the determination of 
for a dense gas or a liquid may be much more  A  v A  u A  u A  v   p A (48b)
difficult than the associated experiment.
A N
Given that the behavior of the mixtures under  A    Au Au A  p A (48c)
consideration is described by Eqs. (41), we propose A1

that the species stress tensor can be represented by

  A       A   p A (48d)
Proposal: TA   p AI   A , A  1, 2,..., N (42)
rA  v A  v A   p A (48e)
in which pA is the partial pressure defined by
(Truesdell, 1969, page 97) as The first of these indicates that the governing
equation for u A is quasi-steady; the second indicates
pA   2
A   A   A T ,  B , C ,.......
(43)
that diffusive inertial effects are negligible, the third
indicates that the diffusive stresses are negligible, the
Here  A is the Helmholtz free energy of species A
fourth indicates that viscous effects are negligible,
per unit mass of species A. In general it is more and the final inequality indicates that the effects of
convenient to work with the internal energy and homogeneous chemical reactions are negligible.
define the partial pressure by (Whitaker, 1989,
Chapter 10) When the restrictions given by Eqs. (48) are
imposed, the governing equation for the mass
p A   A2   eA   A  s ,  (44) diffusion velocity takes the form
B , C ,.......

 p A   A p   A ( b A  b )
in which eA is the internal energy of species A per (49)
unit mass of species A. A detailed discussion of the BN
partial pressure and the total pressure is given in  P
B 1
AB , A  1, 2,..., N  1
Appendix B. At this point we define the total
pressure and the total viscous stress tensor by
Truesdell (1962, Eq. 7) represents the left hand side
A N A N of this result by p d A and cites Hirschfelder, Curtiss
p  p
A1
A ,   
A1
A (45)
& Bird (1954) as the source. Curtiss & Bird (1999,
Eq. 7.6) represent the left hand side of Eq. (49) by
and we use these definitions along with Eq. (42) in cRT d A and refer to it as the generalized driving
order to express Eq. (40) as force for diffusion. At this point we make use of the
identity
 uA 
A   v A   u A  u A  v 
 t
1 p
   pA p   p A  A2 p (50)
p p
 A N 
  A     A u A u A      A       A  in order to express Eq. (49) in the form
 A1  (46)
 p A   Ap   A ( b A  b)   p A p   p 1  p A p    A  p
(51)
BN BN
 P AB  rA  v A  v A   p 1 A ( b A  b)  p 1  PAB , A  1, 2,..., N  1
B 1 B 1

www.amidiq.org 219
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

In order to see how this result is related to the work
of Hirschfelder, Curtiss & Bird (1954), we make use
of their Eqs. 7.4-48 and 7.3-27 (in terms of the
nomenclature used in this work) to obtain 2
Hirschfelder et al.
than suggested by Eq. (54), and Rutten (1992),
among many others, has documented these
complexities for ternary systems. Putting aside the
seminal problem associated with PAB , we make use
of Eq. (54) in Eq. (51) to obtain

x A  p 1  x A   A  p   p A p   p 1   p A p    A  p
 
BN pressure diffusion
x A xB
 p 1  A ( b A  b )   DAB
 vB  v A  (52) BN
x A xB
B 1
 p 1 A ( b A  b) 
   DAB
 vB  v A  (55)
BN
x A xB  D
T
D  T
B 1
 
B 1

B

DAB   B
  ln T , A  1, 2,..., N  1
A 
A forced diffusion

BN
x A xB  DBT DT 
The right hand side of this result is approximate in
  
B 1 DAB   B
 A  ln T ,
A 
A  1, 2,.., N  1
that (1) it is based on dilute gas kinetic theory, and   
thermal diffusion
(2) the binary diffusivities, DAB , have been used in
place of the coefficient of diffusion, DAB (see Here we have explicitly identified the terms
Hirschfelder, Curtiss & Bird, 1954, page 485). The associated with pressure diffusion, forced diffusion,
left hand side of Eq. (52) is identical to the left hand and thermal diffusion. This form of the species
momentum equation is restricted by the following:
side of Eq. (51) provided that p A is replaced by
x A p , and this is consistent with the idea that I. The basic assumptions associated with
Eq. (52) was developed for ideal gases. In terms of continuum mechanics.
the work of Chapman and Cowling (1970), we note II. The constitutive equation given by Eq. (42)
that their Eqs. 18.2,6 and 18.3,13 lead to Eq. (52)
with the last term in Eq. (52) expressed as III. The simplifications indicated by Eqs. (48).
BN
x A xB  DBT DAT  IV. The form of the terms that appear on the
kTA ln T  
B 1

DAB   B
   ln T
A 
(53) right hand side of Eq. (55).
One should remember that Eq. (55) is the governing
When dealing with ideal gases, one can proceed with equation for the diffusion velocity, and this becomes
Eq. (52); however, for more general cases that are more apparent if we replace v B  v A with u B  u A .
consistent with Eq. (42), one should make use of
Eq. (51) and this means dealing with the force, PAB . In general, thermal diffusion creates very
small fluxes that are difficult to measure (Whitaker
and Pigford, 1958) and in this study we will neglect
1.2.3 Non-ideal mixtures
this term to obtain
The simplest approach for non-ideal mixtures is to
use the form associated with dilute gas kinetic theory   p A p   p 1   p A p    A  p
 
in order to represent the right hand side of Eq. (51) pressure diffusion
as
BN (56)
x A xB
Proposal:  p 1  A ( b A  b ) 
  
B 1 DAB
 v B  v A ,
x x x x  DT DT  forced diffusion
1
p PAB  A B  v B  v A   A B  B  A   ln T (54) A  1, 2,..., N  1
DAB DAB   B A 
Chapman & Cowling (1970, page 257) discuss the
Here the diffusion coefficients are to be determined impact of pressure diffusion on the distribution of
experimentally with the idea that this form for PAB is chemical species in the atmosphere, and both Deen
an acceptable approximation, and that Eq. (20) (1998, page 452) and Bird et al. (2002, page 772)
would be utilized as a solution to Axiom III. provide an example of this effect in terms of a
Truesdell (1962, Sec. 6) refers to this approximation separation process using an ultracentrifuge. The
as the special case of binary drags. However, process of forced diffusion of electrically charged
multicomponent diffusion in liquids is more complex particles is analyzed by Chapman & Cowling (1970,
Chap. 19) among others.
2
Here one should note that the logarithm of a quantity Estimates (Whitaker, 2009b, Sec. 5.6) of the
having units is not defined, thus the appearance of lnT in terms on the left hand side of Eq. (56) indicate that
Eq. (52) represents a dangerous situation. However, the
these terms are generally quite small leading to the
gradient of ln T leads to T 1T and some might argue relatively simple relation given by
that the danger has been avoided.

220 www.amidiq.org
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

BN These equations can (in principle) be used to

x A xB
0     pA p   
B 1 DAB
vB  v A  , determine all the species mass densities,  A ,  B ,…,
(57)
 N in the same way that the momentum equations
A  1, 2,..., N  1 represented by Eqs. (59) and (60) can be used to
Here one should remember that the first term in this determine all the species velocities, v A , v B , …, v N .
result is based on the use of Eq. (42) and that the The mass diffusive flux,  Au A , is often represented
second term represents a less than robust model for as (Bird et al, 2002, page 537)
non-ideal mixtures in which the binary diffusivities,
DAB , must be determined experimentally. jA   Au A (63)

1.2.4 Ideal mixtures so that Eq. (62) takes the form

At this point we are ready to make the final Species Mass:

simplification given by
 A
   (  A v )    j A  rA , A  1, 2, ..., N  1 (64)
p A  xA p , ideal mixture (58) t

in order to obtain the classic Stefan-Maxwell Here we note that the mass diffusive fluxes are
equations that will be examined in the remainder of constrained by
this paper A N

Species Momentum: j
A1
A  0 (65)

BN
x A xB ( v B  v A ) and we need to determine N  1 of these diffusive
0   x A  
B 1 DAB
, A  1, 2,..., N  1 (59)
fluxes in order to develop a solution for Eq. (64).
In many liquid-phase diffusion processes, the
To complete our formulation of the mechanical governing equation for the total density given by
problem, we recall Eq. (38) in the form of the Eq. (61) is replaced by the assumption
Navier-Stokes equations
Assumption:   constant (66)
Total Momentum:
and we need only solve the N  1 species continuity
v 
  v  v    b  p   2 v (60) equations given by Eqs. (64).
 t 
3. Molar continuity equation
in which the diffusive stresses have obviously been Chemical engineers are primarily interested in
neglected. The determination of v A , v B , …, v N chemical reactions, interfacial mass transfer, and
using Eqs. (59) and (60) is a very complex problem adsorption/desorption phenomena, thus molar
and the chemical engineering literature contains concentrations and mole fractions are more useful
many simplified treatments of this problem. than mass densities and mass fractions. Because of
However, the domain of validity of these simplified this, the molar form of the species continuity
treatments is not always clear, and in the following equation is often preferred. This form is obtained
sections we attempt to clarify the basis for some of from Eqs. (8) by the use of the relations
the special forms of the Stefan-Maxwell equations.
 A  c A M A , rA  RA M A , A  1, 2, ..., N (67)
2. Mass continuity equation
We begin this study with the total mass continuity This leads to the species molar continuity equation
equation [see Eq. (9)] given by

  cA
   ( v )  0    ( c A v A )  RA , A  1, 2, ..., N (68)
Total Mass:
t
(61) t

while the constraint on the mass rate of reaction

along with N  1 species mass continuity equations
given by Eq. (2) provides
[see Eqs. (13)]
A N
Species Mass: R
A1
A MA  0 (69)
A
   (  A v )     (  Au A )  rA ,
t (62)
A  1, 2, ..., N  1

www.amidiq.org 221
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

The total molar continuity equation is analogous to the mass average velocity in Eq. (64) can be
Eq. (61) and it is developed by constructing the sum determined by the application of Eq. (60). In order to
of Eqs. (68) over all species to obtain eliminate the molar average velocity from Eq. (77)
we return to Eq. (73), multiply by  A , and sum over
c BN
Total Molar:
t
   (c v )  R
B 1
B (70) all species to obtain
BN BN BN

Here the total molar concentration and the molar  v

B 1
B B   v  
B 1
B


B 1
B uB (78)
average velocity are defined by
A N A N
On the basis of the second of Eqs. (10) this takes the
c   cA ,
A1
c v   cA v A
A1
(71) form
BN
v  v   B uB (79)
The development in Sec. 2 indicates that Eq. (70) B 1
should be solved along with the N  1 species
continuity equations given by and we are now confronted with the mixed-mode
term B uB that involves a mass fraction and a molar
Species Molar:
diffusion velocity. We would like to express B uB in
 cA terms of molar diffusive fluxes, and to do so we
   (c A v A )  RA , A  1, 2, ..., N  1 (72)
t manipulate this term as follows

This allows for the determination of all the species  B uB

B uB 
molar concentrations, c A , cB ,…, cN .  A   B   C  .....   N
The form of Eqs. (70) through (72) suggests M B cB uB
(but does not require) a decomposition of the species  (80)
M A c A  M B cB  .......  M N cN
velocity given by
M B J B
 
vA  v  u , *
A A  1, 2, ..., N (73) c  x A M A  xB M B  ...... x N M N 

in which u*A is the molar diffusion velocity. A little If we define the mean molecular mass as
thought will indicate that the molar diffusion
velocities are constrained by M  x A M A  xB M B  ...  x N M N (81)

A N we can express Eq. (80) in compact form according

c
A1
A u*A  0 (74) to

M B J B
When Eq. (73) is used in Eq. (72) the transport of B uB  (82)
species A can be represented in terms of a convective cM
part, c A v , and a diffusive part, c A u*A , leading to
At this point we return to Eq. (79) to develop the
 cA following relation between the molar average
   ( c A v )     ( c Au*A )  RA , velocity and the mass average velocity:
t (75)
A  1, 2, ...., N  1 1 BN M B 
v  v   JB
c B 1 M
(83)
The molar diffusive flux, c A uA , is often identified as
(Bird et al, 2002, page 537) Substitution of this expression for the molar average
velocity into Eq. (77) allows us to express that form
J A  c AuA (76) of the species continuity equation as

so that Eq. (75) takes the form Species Molar:

 cA  BN
M 
 cA
   ( c A v  )     J A  RA ,    ( c A v )     J A  x A  B J B   RA ,
t (77) t  B 1 M  (84)
A  1, 2, ..., N  1 A  1, 2, ...., N  1
This result is similar in form to Eq. (64) for the in which the molar diffusive fluxes are constrained
species mass density; however, there is no governing by
equation for the molar average velocity, v  , whereas

222 www.amidiq.org
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243
A N
J
A1
*
A  0 (85)
Here we can see that this convection-diffusion
problem is inherently nonlinear in terms of the
diffusive flux; however, if the mole fraction of
species A is sufficiently small it is possible that the
term involving the sum of the diffusive fluxes in
Eq. (84) can be neglected. By “sufficiently small” we
mean that the following inequality
BN
MB 
xA  J B  J A (86)
B 1 M as a hybrid flux and identified as (Bird et al, 2002,
is satisfied and Eq. (84) becomes linear in the molar
diffusive flux, J A .
To complete our formulation of the molar
forms of the species continuity equation, we make
use of Eq. (83) in Eq. (70) to obtain
Total Molar:
c BN
M BN
   ( c v )     B J B   RB (87)
t B 1 M B 1
This total molar transport equation should be
compared with Eq. (61) in order to appreciate the
complexity associated with the molar form of the
species transport equations. In many gas phase mass
transfer processes, Eq. (87) can be replaced by the
assumption
Assumption: c  constant (88)
and we need only solve the N  1 species
continuity equations given by Eqs. (84).
4. Mixed-mode continuity equation
The motivation for a mixed-mode or hybrid species
continuity equation is based on the applications that
are dominant in the area of chemical engineering,
and on the mechanical problem under consideration.
To be explicit, we note two facts:
(1) Chemical reactions and interfacial
mass transfer are usually represented
in terms of the molar concentration,
cA, or the mole fraction, xA, thus we
are motivated to use the molar form of
the continuity equation given by
Eq. (68) as opposed to the mass form
given by Eq. (62).
(2) The species continuity equation
involves velocities that must be
determined by the laws of mechanics,
thus we are motivated to use the mass
decomposition of the species velocity
given by Eq. (12) as opposed to the
molar decomposition given by
Eq. (73).
www.amidiq.org
In order to obtain a mixed-mode or hybrid continuity
equation, we begin with the species mass continuity
equation given by Eq. (62) and divide by the
molecular mass of species A to obtain
 cA
   ( c A v )     ( c A u A )  RA ,
t (89)
A  1, 2, ...., N  1
Here the diffusive flux is represented in terms of a
molar concentration and a mass diffusion velocity.
This mixed-mode diffusive flux is often referred to
page 537)
J A  cA u A (90)
Use of this representation in Eq. (89) leads to
Species Molar:
 cA
   ( c A v )     J A  RA ,
t (91)
A  1, 2, ..., N  1
The constraint on this diffusive flux is more complex
than that for either the mass diffusive flux or the
molar diffusive flux and is given by
A N
M
A1
A JA  0 (92)
This hybrid diffusive flux, JA, lacks popularity;
however, the transport equation given by Eq. (91)
has the advantage that it is linear in the diffusive
flux. In terms of the mixed-mode diffusive flux, the
total molar continuity equation takes the form
Total Molar:
c BN BN
   ( cv )      J B   RB (93)
t B 1 B 1
and we are still confronted with a complex form of
the total molar transport equation. This complexity
often serves to generate the assumption that the total
molar concentration is constant as indicated by
Assumption: c  constant (94)
Often gas phase diffusion problems lead to the use of
a molar form of the species continuity equation
because Eq. (94) provides a reasonable
simplification. On the other hand, liquid phase
diffusion problems suggest the use of the mass form
of the species continuity equation because Eq. (66)
provides a reasonable simplification. The author is
unaware of any solution to a diffusion problem that
does not make use of either Eq. (66) or Eq. (94), and
removing these assumptions remains as a significant
challenge.
223
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

5. Binary systems Before attacking the binary result given by

Binary systems are often used to introduce the
phenomena of diffusion, and we will follow that
approach in order to explore the nature of the mass,
molar, and mixed-mode forms of the species
continuity equation.
5.1 Mass diffusive flux
For a binary system, Eq. (59) reduces to
x x (v  v A )
0   x A  A B B
DAB
Eq. (102) it is convenient to list some results for N-
component systems. We begin with the definitions
for the mass fraction,  A , the mole fraction, xA , and
the mean molecular mass, M . These are given by
A   A  , xA  cA c ,
(103)
M  x A M A  xB M B  .....  xN M N
in which MA represents the molecular mass of
species A. In addition to these results, we make use
(95)
of

and we think of this as the governing differential cA   A M A , c   M , A  1, 2,...N (104)

equation for vA. The value of vB is available from a
solution for v and vA which can be used in the second to obtain the following relations between the mole
of Eqs. (10) to obtain fractions and the mass fractions

1 cA c M A MAM M
vB   v  Av A  (96) xA   A   A ,
B c   MA (105)
A  1, 2,... N
For a binary system, the two mass continuity
equations are given by
At this point we direct our attention to binary
A systems and make use of the following relations
   (  A v )     (  Au A )  rA (97)
t x A   xB , B  1   A ,
 1   (106)
   (  v)  0 (98)  A  B
t M MA MB

and we need to determine uA and v in order to solve along with several algebraic steps (see Appendix C)
these equations. Given the form of Eq. (97) it will be to arrive at
convenient to express Eq. (95) in terms of mass
M2
diffusion velocities, and the use of Eq. (12) leads to x A   A (107)
M AM B
x A xB (u B  u A )
0  x A  (99) Substitution of this expression for the gradient of the
DAB
mole fraction of species A into Eq. (102) leads to
For a binary system, Eq. (14) provides
M2 1 x A xB
0    A  ( u ) (108)
 A u A  B u B  0 (100) M AM B  DAB  AB A A

and this can be used in Eq. (99) to obtain From Eqs. (105) we see that
1 x A xB M2
0   x A  ( Au A ) (101) x A xB
 (109)
DAB  AB AB M AM B

Multiplying and dividing the second term by the total and Eq. (108) simplifies to the classic form of Fick’s
density allows us to express this result as Law given by

0   x A 
1 x A xB
(  Au A ) (102) Fick’s Law: jA   Au A    DAB  A (110)
 DAB  AB
Returning to Eq. (97), we make use of this form of
Here we have a mixed-mode representation in which Fick’s Law to obtain the following governing
the mass diffusive flux,  Au A , is expressed in terms equation for the species density,  A
of the gradient of the mole fraction, x A , along with
A
the mixed-mode term, xA xB  AB .    (  A v )      DAB    A     rA (111)
t

224 www.amidiq.org
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

For liquid systems this result can often be simplified by J*A . We begin by using Eq. (73) to express the
on the basis of the assumption single Stefan-Maxwell equation as
Assumption:   constant (112)
x A xB ( u*B  u*A )
0   x A  (118)
which leads to DAB

A and employ the form of Eq. (76) for both species to
   (  Av)
t obtain
(113)
   constant
    DAB  A   rA ,  0   x A 
x A J *B  xB J *A
(119)
 binary system c DAB
Here we have an attractive, linear transport equation Application of the binary version of Eq. (85)
for the species density,  A .
When confronted with chemical reactions and J *A  J *B  0 (120)
interfacial transport, we generally prefer to work
allows us to express Eq. (119) in the classic form of
with the molar form of the species continuity
Fick’s Law given by
equation. This form can be extracted from Eq. (113)
by the use of Fick’s Law: J A  c A u*A   c DAB x A (121)
 A  cA M A , rA  RA M A (114)
This is the molar analogy of Eq. (110), and
which leads to substitution of this result into Eq. (116) leads to the
molar analogy of Eq. (111).
 cA
   (cA v) c A
t    (c A v)     M B M  cDAB   c A c    RA (122)
(115) t
   constant
    DAB c A   RA , 
 binary system If we ignore variations in the total molar
concentration on the basis of the assumption
This is an attractive form to use with liquids where
the assumption of a constant density is likely to be a Assumption: c  constant (123)
valid approximation. When the total density is not
constant, one must solve Eq. (111) simultaneously we see that Eq. (122) takes the form
with Eq. (98).
 cA
   ( c A v )     M B M  DAB c A 
5.2 Molar diffusive flux t
(124)
 c  constant
Because of the prevalence of molar concentrations  RA , 
and mole fractions in chemical engineering analysis,  binary system
the species molar continuity equation is generally
preferred. This form can be extracted from Eq. (84) in which the presence of M leads to the non-
according to linearity associated with
1
Species Molar: M B M   x A  M A M B  1  1 (125)
 cA  B 2
M 
   ( c A v )      J A  x A  B J B   RA (116) In order to obtain the so-called dilute solution form
t  B 1 M  of Eq. (124), we impose
while the total molar continuity equation given Restriction: x A  M A M B  1  1 (126)
earlier by Eq. (87) takes the form
Total Molar: and Eq. (124) simplifies to the classic convective-
diffusion equation given by
c B 2
M
   ( c v )     B J B   RA  RB  c A
   ( c A v )     DAB c A 
(117)
t B 1 M
t
Ignoring for the moment the difficulties associated  c  constant (127)

with the total molar continuity equation, we direct  RA ,  x A  M A M B  1  1
our attention to the molar diffusive flux represented 
 binary system

www.amidiq.org 225
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

Here it is very important to note that this result is

identical to Eq. (115). However, Eq. (127) is not Fick’s Law:
based on the constraint that the density is constant.
Instead, Eq. (127) is based on the assumption of J A  c A u A    M B M  c DAB x A (135)
“constant total molar concentration” indicated by
Eq. (123), and the assumption of a “dilute solution” Substitution of this result into Eq. (131) provides the
indicated by Eq. (126). For binary systems we have following governing equation for the species A molar
(see Eqs. (103) and (104)) concentration
 cA
  c  x A M A  xB M B  (128)    (c A v)     M B M  c DAB   c A c    RA (136)
t
which can be arranged in the form This result is identical to Eq. (122) indicating that
  c x A  M A M B   1  1  M B
both the molar representation given by Eq. (116) and
(129) the mixed-mode representation given by Eq. (131)
lead to the same result for a binary system.
When the two restrictions associated with Eq. (127)
are imposed, the total density is essentially constant It is of some interest to note that the mixed-
and Eq. (127) is consistent with Eq. (115). mode diffusive flux can be expressed as
Returning to Eq. (127), we note that the 1
maximum value of the mole fraction for species A
cA u A  A uA (137)
MA
will usually be known a priori, and this allows us to
express the constraint associated with Eq. (127) as and on the basis of Eq. (110) this takes the form
Constraint: ( x A ) max  M A M B  1  1 (130) 1
cA u A    DAB  A (138)
MA
One should remember that there is a restriction
associated with every assumption and when one Use of this result in Eq. (131) yields what appears to
imposes the restriction one always assumes that be an unattractive form given by
small causes give rise to small effects (Birkhoff,
1960). In addition, one should remember that behind c A  1 
every restriction there is a constraint (see Appendix    (cA v)      DAB  A   RA (139)
t MA 
D); however, constraints can often be very difficult
to develop.
However, if we impose the condition
5.3 Mixed-mode diffusive flux Assumption:   constant (140)
In this case we return to the mixed-mode species and make use of the first of Eqs. (114) we find
continuity equation [see Eq. (89)]
c A
c A    ( c A v )     DAB c A 
   ( c A v )     ( c A u A )  RA (131) t
t (141)
   constant
 RA , 
and direct our attention to the single Stefan-Maxwell  binary system
equation given by Eq. (101) and repeated here as
which was given earlier by Eq. (115).
1 x A xB
0   x A  ( Au A ) (132)
DAB  AB 6. Special forms for N-component systems
Given the complexity of the binary forms described
It is convenient to rearrange this result in the form in the previous sections, we should expect additional
1 xB complexities for N-component systems. This
0   x A  ( c Au A ) (133) naturally leads to the search for simplifications, and
c DAB B
we will examine some of these simplifications in this
section.
in order to obtain the mixed-mode diffusive flux
given by 6.1 Dilute solution diffusion
B There are mass transfer processes in which all the
cA u A   c DAB x A (134)
xB molar fluxes are the same order of magnitude and the
dominant diffusing species is dilute. In this special
At this point we can use Eq. (105) to obtain the case, it is convenient to represent the Stefan-
mixed-mode form of Fick’s Law given by

226 www.amidiq.org
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

Maxwell equations in terms of the molar flux defined (148a) applies to these three components. Use of
by Eq. (147) allows us to express the dilute forms of
N A  cA v A , A  1, 2,..., N (142) Eqs. (146) as
N A  c DAm x A , A  1, 2, ..., G  N (149)
which allows us to express Eqs. (59) as
BN
At this point we recognize that Eqs. (72) can be
x A N B  xB N A
0   x A  
B 1 c DAB
, A  1, 2, ..., N  1 (143) expressed in terms of N A to obtain
B A  cA
   N A  RA , A  1, 2, ..., N  1 (150)
At this point we separate the second term to obtain t

BN
NB BN
xB and that Eq. (149) can be used to obtain a dilute
0   c x A  x A D
B 1
 NA D
B 1
, solution diffusion equation given by
B A
AB
B A
AB (144)
c A
A  1, 2, ..., N  1     c DAm x A   RA , A  1, 2, ..., G  N (151)
t
and we define the mixture diffusivity by
We are still confronted with the complexity of the
1 BN
xB transport equation for the total molar concentration
DAm
 D
B 1
, A  1, 2, ..., N  1 (145) given by Eq. (87), and this difficulty is classically
B A
AB avoided by assuming that the total molar
concentration is a constant in order to obtain
so that the Stefan-Maxwell equations can be
c A
expressed as     DAm c A   RA ,
t
BN
DAm (152)
0   c DAm x A  x A D
B 1
NB  N A ,  GN
A  1, 2, ..., G , 
AB (146)
B A  other conditions
A  1, 2, ..., N  1
in which the other conditions associated with this
For some processes, such as diffusion in porous result are given by
media (Whitaker, 1999) in which the flux of all the
species is driven by heterogeneous reaction or by Assumption: c  constant (153a)
adsorption/desorption, we can impose the
simplification Restriction: N B  O  N A  , B  1, 2, ..., N (153b)

BN
DAm Constraint: ( x A ) max  1 , A  1, 2, ..., G  N (153c)
xA D
B 1
N B  N A ,
B A
AB (147) The constraint identified by Eq. (153c) is generally
A  1, 2, ..., G , G  N available in terms of the problem statement, and
when this constraint is satisfied it is probable that the
when the following two conditions are satisfied: assumption given by Eq. (153a) and the restriction
given by Eq. (153b) are also valid.
Constraint: ( xA ) max  1 , A  1, 2, ..., G  N (148a)
6.2 Dilute solution convective-diffusion equation
Restriction: N B  O  N A  , B  1, 2, ..., N (148b)
using J*A
The first of Eqs. (148) is identified as a constraint In order to develop the convective-diffusion version
since the maximum values of the mole fractions are of Eqs. (152), we begin with the generally valid form
generally known a priori, while the second given by Eqs. (84) and repeated here as
inequality is identified as a restriction since it is not
expressed in terms of quantities that are known a c A  BN
M 
priori. Equation (148b) should be interpreted to    ( c A v )      J A  x A  B J B 
t  B 1 M 
mean that N B is not significantly larger than N A (154)
and if species B is stagnant, N B would be zero.  RA , A  1, 2, ...., N
In Eqs. (147) and (148a) we have indicated
that our N-component system contains G The Stefan-Maxwell equations can be expressed as
components that are dilute. For example, if we may BN
x A J B  xB J A
have a five-component mixture in which three 0   x A   , A  1, 2, .., N  1 (155)
components have mole fractions that are small B 1 cDAB
B A
compared to one, we have G  3 and Eqs. (147) and

www.amidiq.org 227
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

and the summation can be separated leading to c A

   ( c A v )     c DAmx A   RA ,
t (163)
BN
J B BN
x
0   cx A  x A   J A  B , A  1, 2, ..., G  N
B 1 DAB B 1 DAB (156)
B A B A
In addition to the inequalities given by Eqs. (160)
A  1, 2, ..., N  1 and (162), we assume that the total molar
concentration is constant in order to obtain the
The definition of the mixture diffusivity given by classic linear convective-diffusion equation for
Eq. (145) can be used to express this result in the species A.
form
c A
BN DAm    ( c A v )     DAmc A 
0   c DAm x A  x A  J B  J A , t
DAB
B 1
B A
(157)  GN (164)

A  1, 2, ..., N  1  RA , A  1, 2, ..., G ,  c  constant
 x  1
In making judgments about this result, we need to  A
remember that the diffusive fluxes are constrained by This special form of the species continuity equation
BN is ubiquitous in the chemical engineering literature;
J
B 1
*
B  0 (158) however, the simplifications associated with this
result are generally not made clear. In addition to the
dominant restrictions listed in Eq. (164), one should
indicating that the diffusive fluxes tend to be the keep in mind the restriction given by Eq. (162) that
same order of magnitude. This means that the would appear to be automatically satisfied by
following inequality
Eqs. (158) and (160) unless there is a serious
Restriction: disparity in the molecular masses.
BN DAm 6.3 Dilute solution convective-diffusion equation
xA  J B  J A , A  1, 2, ..., G  N (159)
B 1 DAB using J A
B A
In this case we begin with Eq. (91)
has considerable appeal when the mole fraction of
species A is small compared to one as indicated by c A
   (c A v)    J A  RA , A  1, 2, ...., N  1 (165)
t
Restriction: x A  1 , A  1, 2, ..., G  N (160)
and note that the Stefan-Maxwell equations can be
Use of the inequality given by Eq. (159) in the expressed as
Stefan-Maxwell equations given by
Eqs. (157) leads to the multi-component form of BN
x A J B  xB J A
Fick’s Law 0  x A  
B 1 c DAB
, A  1, 2, ..., N  1 (166)
B A
“Fick’s Law”
Separating the terms in the sum leads to
 GN
J A  c DAm x A , A  1, 2, ..., G ,  (161)
 x A  1
BN BN
JB xB
0   c x A  x A 
B 1 D
 J A 
B 1 D
,
which is analogous to the result for binary systems B A
AB AB
B A
(167)
given by Eq. (121). A  1, 2, ..., N  1
We now turn our attention to the species
continuity equation given by Eq. (154). Use of the and use of the definition of the mixture diffusivity
dilute solution condition indicated by Eq. (160) and given by Eq. (145) provides
the constraint on the diffusive fluxes given by BN
DAm
Eq. (158) leads to the restriction 0   c DAm x A  x A D
B 1
JB  J A ,
B A
AB (168)
BN
MB 
Restriction: xA  J B  J A (162) A  1, 2, ..., N  1
B 1 M
In making judgments about this result we need to
Use of this inequality along with the multi- remember that the diffusive fluxes are constrained by
component form of Fick’s Law given by A N
Eq. (161) in Eq. (154) leads to the following form of
the convective-diffusion equation
M
A1
A JA  0 (169)

228 www.amidiq.org
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

thus if the mole fraction of species A is small Note that this result is not restricted to a dilute
compared to one, we can make use of the restriction solution; however, we have imposed the condition on
given by the velocities given by
Restriction: Assumption: v B  v C  ...  v N  0 (177)
BN
DAm This assumption could be replaced with the
xA D
B 1
J B  J A , A  1, 2, ..., G  N (170)
restriction
AB
B A
Restriction: v B  v A , v C  v A , ..., v N  v A (178)
Under these circumstances, the Stefan-Maxwell
equation for species A takes the form in which the use of the absolute values of the
velocities is understood. Here one should remember
“Fick’s Law”:
that we are repeatedly relying on Birkhoff’s (1960)
J A   cDAm x A , A  1, 2, ..., G  N (171) plausible intuitive hypothesis that small causes give
rise to small effects. Use of Eq. (176) in Eq. (150)
Use of this result in Eq. (165) leads to the following leads to
form of the convective-diffusion equation c A
    cDAm x A   RA (179)
c A t
   ( c A v )     cDAmx A   RA (172)
t and we can assume that the total molar concentration
This result, based on the single restriction given by is constant to obtain
Eq. (170), is identical to that given earlier by c A  c  constant
Eq. (163). To complete the analysis of the mixed-     DAm c A   RA ,  (180)
t  other conditions
mode diffusive flux, we assume that the total molar
concentration is constant so that Eq. (172) takes the where the other conditions are those indicated by
form Eqs. (178). This result is identical to Eq. (152)
except for the fact that there is only a single
c A
   ( c A v )     DAmc A   RA , component that could satisfy this equation. As a
t reminder of the difference between Eq. (180) and
 GN (173) Eq. (152) we summarize the conditions upon which
 it is based
A  1, 2, ..., G ,  c  constant
 x  1 Restriction: v B  v A , v C  v A , ..., v N  v A (181)
 A
Restriction: x A c  c A (182)
Certainly the route to Eq. (173) is simpler than that
followed in the development of Eq. (164); however, Comparing these two restrictions with Eqs. (153)
the preferred approach might still be considered to be indicates that Eqs. (152) and (180) describe rather
a matter of choice. different physical phenomena even though the two
equations are identical. In reality, it seems unlikely
6.4 Diffusion through stagnant species that a process restricted by Eq. (181) could involve
significant homogeneous reaction, thus a more
The case of binary transport of species A through a realistic version of Eq. (180) would require that we
stagnant species B has been treated in terms of the set RA equal to zero. Nevertheless, the fact that
classic Stefan diffusion tube (Whitaker, 2009b, Sec. Eq. (152) and Eq. (180) are identical in form
2.7). Moving beyond the binary system, we consider suggests that we must be very careful to understand
the case in which species A is diffusing and all other the precise meaning of the special forms of Eq. (68).
species are stagnant. Under these circumstances, the
Stefan-Maxwell equation for species A reduces to 7. General solution for N-component systems:
BN
Constant total molar concentration
x A xB v A
0   x A  
B 1 DAB
(174) From the analysis in previous sections, it seems clear
B A that the most efficient route to the determination of
the molar concentration is via the mixed-mode
and this can be arranged in the form continuity equation described in Sec. 4. This is
BN especially true for the case in which we develop an
xB
0   cx A  D NA (175) exact solution of the Stefan-Maxwell equations. In
B 1 AB this section we consider the case of constant total
B A
molar concentration and in the next section we
Use of the definition of the mixture diffusivity given examine the case of constant total mass density. The
by Eq. (145) immediately leads to completely general case for which neither  nor c is
N A   c DAm x A (176) constant remains as a challenge.

www.amidiq.org 229
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

In this treatment we make use of Eq. (91) repeated We begin our analysis of the diffusive flux
here as with the Stefan-Maxwell equations given by
Eq. (166), and we make use of the mixture
c A
   ( c A v )     J A  RA , diffusivity defined by Eq. (145) to obtain
t (183)
BN
DAm
A  1, 2, ..., N  1 J A   c DAm x A  x A  JB ,
B 1 DAB (187)
B A
along with the constraint on the mixed-mode
diffusive flux given by A  1, 2, ...., N  1
A N
We want to use Eq. (184) to eliminate J N and it will
M
A1
A JA  0 (184)
be convenient to express that constraint on the
mixed-mode diffusive fluxes in the alternate form
For N-component systems, it is convenient to work given by
in terms of matrices, thus we define the following A N
column matrices that will be used in subsequent
paragraphs.
 J M
A1
A A MN   0 (188)

 cA   c A  At this point we extract J N from the sum in

 c   c  Eq. (187) in order to obtain
 B   B 
 c   cC 
c    ...C  , c    ...  , J A   c DAm x A  x A
B  N 1


DAm D
J B  x A Am J N ,
    B 1 DAB DAN (189)
 ...   ...  B A
    A  1, 2, ..., N  1
 c N 1   cN 1 
and from Eq. (188) we have the following
 x A   JA  representation for J N
 x   J 
 B   B  B  N 1
 xC   J 
x    ...  ,  J    ...C  , JN   
B 1
JB  M B M N  (190)
   
 ...   ... 
    In order to use this result with Eq. (189), we need to
x N 1   J N 1  condition the sum with the constraint indicated by
B  A and this leads to
 RA  B  N 1
 R 
 B 
JN   
B 1
JB M B M N   J A  M A M N  (191)
 R  B A
 R    ...C 
  Use of this result in Eq. (189) provides the following
 ...  form of the Stefan-Maxell equations
  (185)
 RN 1 
 M D  B  N 1
M D D 
J A  1  x A A Am   x A   B Am  Am  J B
Use of the first, fourth and fifth of these matrices  M D
N AN  B 1  M D
N AN DAB  (192)
B A
allows us to express Eq. (183) as
  c DAm x A , A  1, 2, ...., N  1
 [c ]
   [c ] v       J    R  (186)
t This can be expressed in compact form according to

and our single objective at this point is to develop a  JA   D Am x A 

J   D x 
useful representation for  J  . A similar approach  B   Bm B 
using v and J A with A  1, 2,..., N  1 is given by  JC   D Cm xC 
 H      c  ...  (193)
Bird et al. (2002, Sec. 22.9). In addition, Quintard et  ...   
al. (2006) have studied the formulation and the  ...   ... 
numerical solution for this problem using both the    
 J N 1   D N 1m x N 1 
molar forms, v  and J A , and the mass forms, v and
jA . in which  H  is an ( N  1)  ( N  1) square matrix

230 www.amidiq.org
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

 H AA H AB . . . H AN 1   JA 
 H DAm ....
0 0 0   x A 
H BN 1  J 
 0   x B 
H BB . . .
 BA   B  0 D 0 ....
 H . . . . .   JC   Bm  
   cH   0
1
[ H ]   CA . (194)  0 D .... Cm 0   xC 
 . . . . . .   ...    
 .   ...   . . . .... .  .. 
. . . . .
     0 0 0 .... DN 1m  x N 1 

 H N 1 A . . . . H N 1N 1   J N 1 
(198)
having the elements defined by
The diffusivity matrix is now defined by
M D
H AA  1  x A A Am ,
M N DAN DAm 0 0 .... 0 
(195a)  0 DBm 0 .... 0 
A  1, 2,..., N  1  
[ D ]  [ H ] 1  0 0 DCm.... 0  (199)

M D D   . . . .... . 
H AB  x A  B Am  Am  ,  0 0 0 .... DN 1m 
 N AN DAB 
M D 
(195b)
A, B  1, 2,.., N  1 , A  B and this allows us to express Eq. (198) as

We assume that the inverse of [ H ] exists in order to  JA   x A 

J   x 
express the column matrix of the mixed-mode  B   B 

diffusive flux vectors in the form  JC   xC 

    c [ D]   (200)
 DAm x A   ...   ... 
 JA   ...   ... 
J   D x 
 Bm B     
 B   J N 1   xN 1 
 JC   DCm xC 
   cH  
1
  (196)
 ...   ...  with the compact form given by
 ...   ... 
    J   c [ D ]  x  (201)
 J N 1   DN 1m x N 1 
This represents the N-component analog of Fick’s
The column matrix on the right hand side of this Law given by Eq. (135) that we recall here as
result can be expressed as
Fick’s Law: J A   c  M B M  DAB  x A (202)
 DAm x A 
 D x 
 Bm B  Use of Eq. (201) in Eq. (186) leads to
 DCm xC 
   [c ]
...    [c ]v      c [ D ]x    R  (203)
  t
DN 1m xN 1 
 
(197) Once again we may be faced with the difficult task
DAm 0.... 0 0   x A  of determining the total molar concentration on the
 0 D 0 .... 0   x B 
basis of Eq. (93), and to avoid this problem we
 Bm   restrict Eq. (203) to the case of constant total molar
  0 0 D .... Cm 0   xC  concentration. This leads to
  
 . . . .... .  ..   [c ]
 0 0 0 .... DN 1m  xN 1     [c ]v     [ D ] c    R 
 t
(204)
so that the matrix representation for the mixed-mode  c  constant
diffusive flux becomes 
 N  component system

Here it is important to remember that [ D] depends

explicitly on the mole fractions, as indicated by the
definitions given in Eq. (195) and implicitly as
indicated by the definition of the mixture diffusivity
given by Eq. (145). This means that a trial-and-error
numerical solution will be necessary in which the

www.amidiq.org 231
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

assumed values used for the mole fractions are  N     A  B  C  .....   N 1  (211)
upgraded after each iteration. The solution for [c]
This allows us to eliminate  N from Eq. (210) and
will provide values of c A , cB , ...., cN 1 and the
express that result in the form
concentration cN can be determined by the first of
 B  N 1
 M M  
Eqs. (71). Similarly, the solution for  R will x A 
M
 A   A     B  , (212)
MA  B 1  M B MN  
provide values of RA , RB , ...., RN 1 and the reaction
A  1, 2,... N  1
rate RN can be determined by Eq. (69). In the case
of complex kinetics, the column matrix of reaction Here we need to condition the sum with the
rates will need to be expressed as constraint indicated by B  A and this leads to

 R  F (cA , cA , ..., cN ) (205) M    M M 

x A  1   A      A
MA    M N M A 
and the trial-and-error procedure will be more
complex.  (213)
B  N 1
 M M  
 A     B  , A  1, 2,... N  1
8. General solution for N-component systems: B 1  MN MB  
B A 
Constant total mass density
In addition to the N-component form of the species which can be expressed as a matrix equation given
continuity equation based on the assumption of a by
constant total molar concentration, it would be  xA   WAA WAB WAC . . WAN 1    A 
useful to develop the analogous form for constant  x   W WBB . . . WBN 1   B 
 B   BA
total density. Our starting point for this analysis is  xC   WCA . . . . .   C 
 
.   . 
Eq. (203) and the analysis requires that we express
x  in terms of the gradient of the mass fractions,  .   . . . . .
 .   . . . . . .  . 
 A , B , etc. We begin the analysis with     
Eq. (105) repeated here as xN 1  WN 1 A WN 1B . . . WN 1N 1  N 1 
M (214)
xA  A , A  1, 2,... N (206)
MA
Here the elements of this ( N  1)  ( N  1) square
in which the mean molecular mass can be expressed matrix are defined as
as in terms of the mass fractions in order to obtain
M M  M M 
(see Eq. C11 in the Appendix C) WAA   A   ,
MA MA  MN MA  (215a)
1    
 A  B  C  ...  N (207) A  1, 2,..., N  1
M M A M B MC MN
M  M M 
We can use Eq. (206) to express the gradient of the WAB   A   ,
mole fraction as MA  MN MA  (215b)
M M A, B  1, 2,..., N  1 , A B
x A  A   A , A  1, 2,... N (208)
MA MA At this point we recall Eq. (200) and make use of
Eq. (214) to obtain
while the gradient of the mean molecular mass is
given by  JA    A 
J    
BN
B  B   B 
M   M 2  (209)  JC   C 
MB
   c [ D ][W ] 
B 1
  (216)
Use of Eq. (209) in Eq. (208) leads to  ...   . 
 ...   . 
 B N
    
M M  J N 1  N 1 
x A     A   A  M  B  ,
MA  B 1  (210)
in which the square matrix W  is defined explicitly
B

A  1, 2,... N
by
At this point we can make use of the fact that the
sum of the mass fractions is equal to one so that the
gradients are related by

232 www.amidiq.org
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

 WAA WAB WAC . . WAN 1  J   []c  (222)

W WBB . . . WBN 1 
 BA  in which the new diffusivity matrix is given by
W . . . . . 
W    CA
. 
(217) M A M 0 0 . . 0 
 . . . . .  
 . . . . . .   0 MB M 0 . . 0 
   0 0 MC M . . . 
WN 1 A WN 1B . . . WN 1N 1     [ D]W   
 . . . . . . 
Use of the third of Eqs. (104) leads to the total mass  . . . . . . 
density as a multiplier and Eq. (216) takes the form  
 0 0 . . . M N 1 M 
 JA    A 
J     (223)
 B   B 
Use of Eq. (222) in Eq. (186) yields
 JC    C 
    M [ D ]W  
1
  (218)  [c ]
 ...   .     [c ]v     [][c ]   R 
 ...   .  t
    (224)
   constant
 J N 1  N 1  
 N  component system
We are now in a position to impose the condition
that the total mass density is a constant in order to In the trial-and-error solution of this transport
express the mixed-mode fluxes in the form equation, values of the mole fractions will be
required as in the solution of Eq. (204); however, in
 JA    A  this case it is the total mass density, , that is a
J    
 B   B  specified constant and not the total molar
 JC     C 
concentration, c. This requires that we first determine
   M [ D ] W  
1
 ,  N and then cN according to
 ...   .  (219)
 ...   .  A N 1
 
 J N 1 
 
 N 1 
N    
A1
c A M A , cN   N M N (225)

  constant The mole fractions required for the evaluation of

At this point we make use of the first of Eqs. (104) to  D  would then be determined by
express the column matrix of the gradients of the
BN
species densities as xA  cA c
B 1
B (226)
  A   M A 0 0 . . 0   c A 
    0 MB 0 . . 0   cB  while the mass fractions required for the evaluation
 B  
  C   0 0 MC . . .   cC  of W  would be calculated according to
 
.   . 
(220)
 .   . . . . . B N

 .   . . . . . .  .  A  cA M A c B MB (227)
     B 1

 N 1   0 0 . . . M N 1  cN 1 
The result for constant total density given by
Substitution of this result into Eq. (219) leads Eq. (224), along with that for constant total molar
to concentration given by Eq. (204), should prove to be
 JA  M A 0 0 . 0   c A  useful for a wide range of mass transfer problems,
J   0 M 0 . 0   c  provided that the Stefan-Maxwell equations are an
 B   B  B  acceptable representation for the diffusive fluxes. A
 JC  [ D ] W   0 0 MC . .   cC  discussion of the conditions for which the total molar
     concentration and total mass density may be treated
 ...  M  . . . . .  . 
 ...   . as constants is given in Appendix E.
. . . .  . 
    
 J N 1   0 0 . . M N 1  cN 1  9. Conclusions
In this study we have examined the derivation of the
(221)
Stefan-Maxwell equations and we have explored the
in which it is understood that the total mass density structure of these equations in terms of the mass
is assumed to be constant. We can represent this diffusive flux, the molar diffusive flux, and the
result in compact form mixed-mode diffusive flux. Several classic special

www.amidiq.org 233
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

cases have been examined and the assumptions, R gas constant, J/mol K
restrictions and constraints have been identified t time, s
whenever possible. A general method of solution of t A(n ) stress vector for species A, N/m2
the Stefan-Maxwell equations has been presented in
TA stress tensor for species A, N/m2
terms of the mixed-mode diffusive flux.
U total internal energy in a volume V, J
Nomenclature uA v A  v , mass diffusion velocity, m/s
AA (t ) surface area of a species A material uA v A  v , molar diffusion velocity, m/s
volume, m2 vA velocity of species A, m/s
bA body force per unit mass exerted on A N

species A, N/kg
v 
A 1
A v A , mass average velocity, m/s
b body force per unit mass exerted on the A N
mixture, N/kg v x
A 1
A v A , molar average velocity, m/s
cA molar concentration of species A,
moles/m3 vA velocity associated with the net rate of
c total molar concentration, moles/m3 production of species A momentum
dA driving force for diffusion of species A in owing to chemical reaction, m/s
V A (t ) volume of a species A body, m3
an ideal solution, m
DAB DBA , binary diffusion coefficient for xA c A / c , mole fraction of species A
species A and B, m2/s
DAm mixture diffusivity for species A, m2/s Greek Letters
A mass density of species A, kg/m3
[ D] diffusivity matrix used with constant total
 total mass density, kg/m3
molar concentration, m2/s
[] diffusivity matrix used with constant total
 viscosity, N/m2s
mass density, m2/s  viscous stress tensor, N/m2
A viscous stress tensor for species A, N/m2
DAT thermal diffusion coefficient for species
A, kg m 2 s A  A /  , mass fraction of species A
G number of molecular species that are
dilute Acknowledgment
g gravitational body force per unit mass, This paper grew out of a presentation at the Second
N/kg International Seminar on Trends in Chemical
jA  Au A , mass diffusive flux of species A, Engineering, the XXI Century, Mexico City, January
kg/ m2s 28 – 29, 2008. The encouragement of students from
J A c A uA , molar diffusive flux of species A, Puebla to prepare a more complete discussion of the
moles/ m2s Stefan-Maxwell equations is greatly appreciated. In
JA c A u A , mixed-mode diffusive flux of addition, the thoughtful comments of Francois
Mathieu-Potvin helped to clarify some of the issues
species A, moles/ m2s treated in this work. Finally, the comments of
MA molecular mass of species A, g/mole Professor R.B. Bird have clarified my understanding
M mean molecular mass of a mixture, of the complex process of multicomponent mass
g/mole transfer.
N total number of molecular species
References
NA cA v A , molar flux of species A, mole/m2s
n unit normal vector Aris, R. (1962). Vectors, Tensors, and the Basic
PAB force per unit volume exerted by species Equations of Fluid Mechanics, Prentice-Hall,
Englewood Cliffs, New Jersey.
B on species A, N/m3 Bearman, R.J. and Kirkwood, J.G. (1958). Statistical
A N
p p
A 1
A , total pressure, N/m2 mechanics of transport Processes. XI
Equations of transport in Multicomponent
pA partial pressure of species A, N/m2 systems. Journal of Chemical Physics 28,
136-145.
rA net mass rate of production of species A
Bird, R.B., Stewart, W.E. and Lightfoot, E.N.
owing to homogeneous reactions, kg/m3s (1960). Transport Phenomena, First Edition.
RA net molar rate of production of species A John Wiley and Sons, Inc., New York.
owing to homogeneous reactions, Bird, R.B. (1995) personal communication.
moles/m3s

234 www.amidiq.org
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243
Bird, R.B., Stewart, W.E. and Lightfoot, E.N.
(2002). Transport Phenomena, Second
Edition. John Wiley and Sons, Inc., New
York.
Bird, R.B. (2009). Notes for the 2nd edition of
Transport Phenomena,
http://www.engr.wisc.edu/che/faculty/bird_by
ron.html.
Birkhoff, G. (1960). Hydrodynamics, A Study in
Logic, Fact, and Similitude, Princeton
University Press, Princeton, New Jersey.
Chapman, S. and Cowling, T.G. (1939). The
Mathematical Theory of Nonuniform Gases,
First Edition, Cambridge University Press.
Chapman, S. and Cowling, T.G. (1970). The
Mathematical Theory of Nonuniform Gases,
Third Edition, Cambridge University Press.
Curtiss, C.F. and Bird, R.B. (1996). Multicomponent
diffusion in polymeric liquids. Proceedings of
the National Academy of Sciences USA 93,
7440-7445.
Curtiss, C.F. and Bird, R.B. (1999). Multicomponent
diffusion. Industrial and Engineering
Chemistry Research 38, 2515-2522.
Deen, W. M. (1998). Analysis of Transport
Phenomena. Oxford University Press, New
York.
Gibbs, J.W. (1928). The Collected Works of J.
Willard Gibbs, Volume I: Thermodynamics,
Longmans, Green and Co., New York.
Hirschfelder, J.O., Curtiss, C.F. and Bird, R.B.
(1954). Molecular Theory of Gases and
Liquids, John Wiley & Sons, Inc., New York.
Quintard, M., Bletzacker, L., Chenu, D. and
Whitaker, S. (2006). Nonlinear, multi-
component mass transfer in porous media,
Chemical Engineering Science 61, 2643-
2669.
Rutten, Ph.W.M. (1992). Diffusion in Liquids (PhD
thesis), Delft University Press, The
Netherlands.
Serrin, J. (1959). Mathematical Principles of
Classical Fluid Mechanics, in Handbuch der
Physik, Vol. VIII, Part 1, edited by S. Flugge
and C. Truesdell, Springer Verlag, New York.
Slattery, J.C. (1999). Advanced Transport
Phenomena, Cambridge University Press,
Cambridge.
Stein, S.K. and Barcellos, A. (1992). Calculus and
Analytic Geometry, McGraw-Hill, Inc., New
York.
Truesdell, C. and Toupin, R. (1960). The Classical
Field Theories, in Handbuch der Physik, Vol.
III, Part 1, edited by S. Flugge, Springer
Verlag, New York.
Truesdell, C. (1962). Mechanical basis of diffusion,
Journal of Chemical Physics 37, 2336-2344
Truesdell, C. (1968). Essays in the History of
Mechanics, Springer-Verlag, New York.
Truesdell, C. (1969). Rational Thermodynamics,
McGraw-Hill Book Company, New York.
www.amidiq.org
Truesdell, C. (1971). The Tragicomedy of Classical
Thermodynamics, Springer-Verlag, New
York.
Whitaker, S. and Pigford, R.L. (1958). Thermal
diffusion in liquids. Measurements and a
molecular model. Industrial and Engineering
Chemistry 50, 1026-1032.
Whitaker, S. (1981). Introduction to Fluid
Mechanics, R.E. Krieger Pub. Co., Malabar,
Florida.
Whitaker, S. (1986). Transport processes with
heterogeneous reaction, pages 1 to 94 in
Concepts and Design of Chemical Reactors,
edited by S. Whitaker and A.E. Cassano,
Gordon and Breach Publishers, New York.
Whitaker, S. (1988). Levels of simplification: The
use of assumptions, restrictions, and
constraints in engineering analysis. Chemical
Engineering Education 22, 104-108.
Whitaker, S. (1989). Heat transfer in catalytic packed
bed reactors, in Handbook of Heat and Mass
Transfer, Vol. 3, Chapter 10, Catalysis,
Kinetics & Reactor Engineering, edited by
N.P. Cheremisinoff, Gulf Publishers,
Matawan, New Jersey.
Whitaker, S. (1999). The Method of Volume
Averaging, Kluwer Academic Publishers,
Dordrecht.
Whitaker, S. (2009a). Newton’s laws, Euler’s laws,
and the speed of light. Chemical Engineering
Education, Spring.
Whitaker, S. (2009b). Chemical engineering
education: Making connections at interfaces,
Revista Mexicana de Ingeniería Química 8, 1-
32.
Appendix A: Chemical Reaction and Linear
Momentum
The rate of change of linear momentum of species A
owing to chemical reaction, rA vA , can be caused
either by the increase of species A (production) or by
the decrease of species A (consumption). If species A
is consumed by chemical reaction, it seems plausible
that the rate of change of linear momentum is given
by rA v A . Here we need to note that the molecular
velocity (Hirschfelder et al., 1954, page 453) of
species A is much larger than the continuum velocity,
v A ; however, the average velocity associated with
the consumption of species A should be adequately
represented by v A . If species A is produced by a
chemical reaction, the rate of change of linear
momentum depends on the velocities of the species
that react to form species A.
The simple reaction illustrated in Fig. A1 can
be described as 2B  A , and we assume that the
loss of momentum by species B is equal to the gain
of momentum of species A. We express this idea as
[see Eq. (21)]
235
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

r vB  rA vA  0 (A1)  rate of change of 

B   
loss gain
linear momentum   rA v A  rA  u B  u A  (A8)
 of species A 
and note that conservation of mass [see Eq. (2)]  
requires
This leads to the estimate
rB  
rA  0 (A2)
loss gain  rate of change of 
 
linear momentum   rA v A  O  rAu A  (A9)
On the basis of the argument given above, we  of species A 
assume that  
v B  v B (A3) suggested by Whitaker (1986, Eqs. 1-19 and 1-55).
If we consider the slightly more complex
reaction illustrated in Fig. A2, the concepts
illustrated in Eqs. (A2) and (A4) take the form
rB  rC  
rA  
rD  0 (A10)
 
loss gain gain

rB v B  rC v C  rA v A  rD v D  0 (A11)
   
loss gain gain

In this case constructing a value for vA is not as

Fig. A1. Reaction of species B to form species A simple as the result illustrated by Eq. (A5). In terms
of molar rates of reaction, we have
rB  M B RB , rB  M B RB ,
(A12)
rB  M B RB , rB  M B RB
in which M B represents the molecular mass of
species B and RB represents the molar rate of
reaction for species B. In terms of molecular masses
and molar rates of reaction, we can express
Eq. (A11) as
M B RB v B  M C RC v C  M A RA v A
   
loss gain
Fig. A2. Reaction of B and C to produce A and D 
(A13)
 M D RD v  0
 D
and Eq. (A1) takes the form gain

r v B  rA vA  0 (A4) and Eq. (A10) can be replaced by

B 
loss gain RB  RC , RC   RA , RA  RD
When Eq. (A2) is used with this result we find that Use of this constraint on the molar rates of reaction
vA is given by in Eq. (A13) leads to
v A  v B (A5) M B v B  M C v C  M A vA  M D v D (A14)
and the rate of change of linear momentum of We now express the species velocities in terms of the
species A can be expressed as mass average velocity and the diffusion velocities in
order to obtain
 rate of change of 
  vA  v  uA , vB  v  uB ,
linear momentum   rA v A  rA v A  rA  v B  v A  (A6)

(A15)
 of species A  vC  v  uC
 
Use of these relations in Eq. (A14) provides
The species velocities can be expressed in terms of
the mass average velocity and the diffusion velocity M B u B  M C uC   M B  M C  v  M A  v A  v A 
to obtain (A16)
vA  v  uA , vB  v  uB (A7)  M D  v D  v D   M A u A  M D u D   M A  M D  v
and these results can be used in Eq. (A6) so that the
that can be simplified to
rate of change of momentum of species A takes the
form

236 www.amidiq.org
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

M A  vA  v A   M D  vD  v D  Here  A is a partial mass quantity such as the

(A17) species internal energy represented in Eq. (B1),
  M B u B  M C uC    M A u A  M D u D  while  is a total mass quantity defined by
A N
Provided that the molecular masses are all the same
order of magnitude, this result suggests that the
   
A1
A A (B7)

difference, vA  v A , is on the order of the diffusion In Eq. (B6) we have used  to represent some
velocities. Given the general constraint on the thermodynamic state variable such as the
diffusion velocities [see Eq. (14)], the result given by temperature, the total mass density, etc.
Eq. (A17) suggests that
We begin this proof with some variable 
v A  v A  O  u A  (A18) that can be represented as

which is equivalent to Eq. (A9).    A  A   B  B  ...   N  N (B8)

The two cases represented in Figs. A1 and A2 or in a manner identical to Eq. (B7)
are especially simple; however, most chemical    A  A  B  B  ...  N  N (B9)
reactions are likely to be binary in nature, thus
Eq. (A18) represents a plausible estimate of the Here the mass fractions are defined by the second of
velocity vA . Eqs. (11) and they are constrained by
 A  B  ...  N  1 (B10)
Appendix B: Thermodynamic pressure
Because of this constraint all the mass fractions are
The decomposition given by Eq. (42) indicates that not independent and the functional representation for
TA is represented in terms of the partial pressure,  is given by
p A , and the viscous stress tensor,  A . The partial      , T ,  A , B , ...., N 1  (B11)
pressure of species A can be defined by (Whitaker,
1989, Chapter 10) If we differentiate  with respect to  A we can
hold all the mass fractions constant except one. For
p A   A2   eA   A  s ,  (B1) convenience we choose this one to be  N and write
B , C ,.......

in which eA is the internal energy of species A per (Slattery, 1999, page 447)
unit mass of species A, and  A is the mass density of     A  , T ,  B ( B  A, N )
  A  N (B12)
species A. We defined the total pressure in terms of
the partial pressures according to This allows us to express  A as
A N
 A      A  , T ,   N (B13)
p  p
A1
A (B2) B ( B  A, N )

and Eq. (B8) can be used to obtain

However, the total pressure, p, can also be expressed
A N A N
  
 
as
  A      N
p   2  e   s ,  ,  ,...,     A   , T , B ( B  A, N )
A A
(B3) A1 A1
B C N

in which e is the total internal energy defined by (B14)

A N Subsequently we will use this result in the form
e   A eA (B4)
A1 A N
  
        N (B15)
   A   , T , B ( B  A , N )
A
In this appendix we wish to show that there is no A1
conflict between Eqs. (B1), (B2) and (B3), and this
requires that we demonstrate the following: At this point we consider the special case in which
  A
A N
  eA   e    , A  , A  1, 2,..., N (B16)
  A2    2   (B5)  
A1    A s,  B ,  C ,..., N    s ,  B , C ,...,  N
Use of this result in Eq. (B15) yields
In order to illustrate how the thermodynamic A N
       N
definition of the partial pressure is related to the  A
 A
 



(B17)
thermodynamic definition of the total pressure, we A1
 , T , B ( B  A , N )
need the following theorem:
Here we write Eq. (B13) for the variables  A and 
A N
 A 
Theorem:  A

 

(B6) to obtain
A1
 A      A  , T ,   N (B18)
B ( B  A, N )

www.amidiq.org 237
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

Use of this result in the left hand side of the theorem Directing our attention to the derivative on the left
we wish to prove leads to hand side of Eq. (B24) we note that it can be
expressed as (Stein and Barcellos, 1992, page 149)
A N
 A A N       N 
 A

  A    
     A   , T , B ( B  A, N )  
  eA   e 
  A
  A 
A1 A1    
   s ,  B , ...  N  A  s ,  B ,...  N     B , ...  N

(B19)
Error! Reference source not found.
Changing the order of differentiation in the first term
and carrying out the summation with the second term Since the mass density for species A can be
provides expressed as
 A  A  (B30)
A N
 A A N
      N
 A

  A
 A


we have
A1 A1
 , T , B ( B  A , N )   A   ,  , ....    A (B31)
B C N

(B20) and Eq. Error! Reference source not found. takes

Substitution of this result in Eq. (B17) provides the the form
desired proof given by
 eA  s ,    A  eA  A  s ,  (B32)
A N
 A  B ,  C , ....  N B, C , ....  N
Theorem:  A
A1 
 

(B21)
Use of this relation in Eq. (B25) leads to
At this point we want to verify the relations A N
 eA   e 
contained in Eq. (B5), and we begin with the  2
   2   (B33)
   s ,  B , C , ....  N    s ,  B , C ,...,  N
A
A1
following representation of the partial pressure
p A   A2   eA   A  s ,  (B22) and on the basis of the definition of the partial
B , C ,.......
pressure, this takes the form
which can be summed over all species to obtain A N
 e 
A N A N p  2   (B34)
p    e   A s,     s ,  B , C ,...,  N
A
 2
(B23) A1
B ,  C ,..., N
A A A
A1 A1

Our objective now is to represent the right hand side We now define the total pressure according to [see
of this result in terms of the total thermal energy. We Eq. (B2)]
A N
begin with Eq. (B21)in the form p  p A (B35)
A N
 e   e  A1

 A  A     (B24) which leads to

A1    s ,  B , C ,...,  N    s ,  B , C ,...,  N
 e 
p  2   (B36)
and multiply by the total density to obtain    s ,  B , C ,...,  N
A N
 e A   e 
     2   (B25) At this point we have proved Eq. (B5).
   s ,  B , C ,...,  N    s ,  B , C ,...,  N
A
A1 To complete this discussion we need to
indicate how this representation of the total pressure
The functional dependence of eA can be represented is related to the classic description for equilibrium
in terms of the mass fractions or the species densities systems. If we represent the volume per unit mass as
as indicated by
v  1 (B37)
e A  eA   , s,  A , B , ...., N 1  (B26a)
we see that Eq. (B36) leads to the following
expression for the total pressure
e A  eA  s,  A ,  B , ....,  N  (B26b)
 e 
In addition, the density of species A, for example, p    (B38)
 v  s ,  B , C ,...,  N
can be expressed as
 A      B   C  ....   N  (B27)
In terms of thermo-statics (Truesdell, 1971), we
consider a system at equilibrium having a mass m
or in the functional form given by with the volume and internal energy given by
V  mv, U  me
 A   A   ,  B , .... ,  N  (B28)
(B39)
Under these circumstances the equilibrium pressure
On the basis of this representation for  A we can takes the classic form (Gibbs, 1928, page 33) given
express Eq. (B26b) as a composite function given by by
p    U V  S (B40)
e A  eA  s,  A   ,  B , ... ,  N  ,  B , ....,  N  (B29)

238 www.amidiq.org
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

For use in the Stefan-Maxwell equations we need the

product form of this result that is given by
Appendix C. Useful algebraic relations
M2
x A xB   AB , A, B  1, 2,..., N (C14)
We begin by noting that the total mass density and M AM B
total molar concentration for a N-component system
are given by In order to develop a relation between the gradient of
   A   B  C  ...   N (C1a) the mole fraction, x A , and the gradient of the mass
fraction,  A , for a binary system we begin with
c  c A  cB  cC  ...  cN (C1b)
Eq. (C13) and take the gradient to obtain
The mass fractions and mole fractions take the form
M M
A cA x A  A   A , A  1, 2,..., N (C15)
A  , xA  , MA MA
 c (C2)
In terms of binary systems, the gradient of the mean
A  1, 2,..., N molecular mass takes the form
and the constraints on these quantities are given by M   x A  M A    x B  M B
BN BN (C16)

B 1
A 1, x
B 1
A 1 (C3)   M A  M B  x A

The mean molecular mass is defined by and use of this result in the binary form of Eq. (C15)
provides
M  x A M A  xB M B  xC M C  ...  xN M N (C4)
A M
and multiplication by the total molar concentration x A   M A  M B  x A   A ,
MA MA (C17)
gives
A  1, 2
c M  c A M A  cB M B  cC M C  ...  cN M N (C5)
Collecting terms leads to
The species molar concentration and the species
mass density are related by    M
x A 1  A  M A  M B     A ,
 A  cA M A , cA   A M A (C6)  MA  MA (C18)
and the use of the first of these in Eq. (C5) provides A  1, 2
c M   A   B   C  ...   N (C7) which can be simplified to the form
Use of Eq. (C1a) allows us to express this result as x A B M A   A M B   M  A ,
(C19)
cM   (C8) A  1, 2

We can use Eq. (C1b) and the second of Eqs. (C6) to At this point we use Eq. (C13) to obtain
obtain x M M x M M 
x A  B B A  A A B   M  A ,
A B C N  M M  (C20)
c     ...  (C9)
MA MB MC MN A  1, 2
Dividing both sides by the total mass density which can be simplified to (Bird, 2009)
provides the following result
M2
A B C N x A   A , A  1, 2 (C21)
c      ...  (C10) M AM B
MA MB MC MN
This result is Eq. (107) in the section on binary
and on the basis of Eq. (C8) we have systems.
1    
 A  B  C  ...  N (C11) Appendix D: Assumptions, restrictions and
M M A M B MC MN
constraints
For a N-component system, the mole fraction of
species A is given by Throughout this paper we have imposed various
assumptions associated with the analysis. The most
cA cA M   A M A  M M frequent of these concerned the total mass density
xA     A ,
c   MA (C12) and the total molar concentration, and an example
A  1, 2,..., N concerning the total mass density is given in
Eq. (63). In engineering analysis there is a logical
and in compact form we express this result as sequence of events that begins with a simplifying
M assumption, or an idea, and leads to a theory with an
xA  A , A  1, 2,..., N (C13)
identifiable domain of validity. In this section we
MA
wish to illustrate this sequence of events with an

www.amidiq.org 239
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243
example from fluid mechanics where the path from
an assumption to a constraint is well known
(Whitaker, 1988).
A large class of fluid mechanical problems
can be described by the continuity equation for
incompressible flow
v  0 (D1)
and the Navier-Stokes equations
 v 
  v  v    p   g   2 v (D2)
 t  is the magnitude defined by
The development of these two equations requires
assumptions that may be supported by restrictions or
reinforced by constraints; however, we will simply
accept Eq. (D1) and Eq. (D2) without inquiring into
their limitations.
As an illustration of the development of
assumptions, restrictions and constraints, we
consider Eq. (D2) and assume that the convective
inertial effects are negligible in order to obtain
v
   p   g  2 v (D3)
t
This linear form can be easily solved for a wide
variety of initial and boundary conditions, whereas
the general form given by Eq. (D2) represents a
difficult problem. It is important to clearly identify
the assumption that leads one from Eq. (D2) to
Eq. (D3) , and one way to express this is
Assumption:  v  v  0 (D4)
Equation (D4) indicates exactly what is being done
in a mathematical sense, but it is not necessarily a
precise description of the physics of any particular
fluid mechanical process. Strictly speaking, Eq. (D4)
can only be true when the velocity vector is a
constant and this is not likely to occur in any real
flow.
From a physical point of view, the
simplification of Eq. (D2) to Eq. (D3) is based on the
idea that the convective inertial term,  v v , is
negligible. This immediately raises the question:
“Negligible relative to what?” and one answer is that
the convective inertial term is negligible relative to
the viscous term. This represents the second level in
our process of simplification, and in this case we
express our simplification as a restriction.
Restriction:  v  v   2 v (D5)
In writing inequalities of this type, it is understood
that the comparison is being made between the
absolute values of the vectors under consideration. If
we apply the idea represented by Eq. (D5) to the
Navier-Stokes equations, we again obtain Eq. (D3)
provided we are willing to assume that small causes
give rise to small effects (Birkhoff, 1960). Equation
(D5) has a definite advantage over Eq. (D4) since it
indicates what is required in a physical sense;
however, neither Eq. (D4) nor Eq. (D5) indicate
240 www.amidiq.org
when Eq. (D3) will be valid in terms of parameters
that are known a priori. In order to determine
precisely under what circumstances Eq. (D3) will be
valid, one must be able to estimate the magnitude of
the terms in Eq. (D2).
We begin our analysis of the inertial term by
expressing the velocity in terms of a unit vector and
the magnitude according to
v  v (D6)
Here  is a unit tangent vector to a streamline and v
v  vv (D7)
The representation given by Eq. (D6) allows us to
express the inertial term as
 v  v  v  v (D8)
in which   is known as the directional derivative
(Stein and Barcellos, 1992, Sec. 14.7). The
directional derivative can be expressed as
d
   (D9)
ds
where s represents the arc length measured along a
streamline. Use of Eq. (D9) in Eq. (D8) provides the
following exact representation of the inertial term
dv
 v v   v (D10)
ds
While this form is not often used in the development
of solutions of the Navier-Stokes equations, it is
extremely useful in the development of an estimate
of the magnitude of the inertial term. To do so, we
need only think about how the velocity vector
changes as we proceed along a streamline and this
suggests that we define an inertial length, L , by the
estimate
dv
 O  v L  (D11)
ds
One should think of the inertial length as being the
distance, along a streamline, over which significant
changes in the velocity take place. A little thought
will indicate that the estimate of L requires an
intuitive knowledge of the flow field, and this
intuitive knowledge is based primarily on a
knowledge of the no-slip condition. Use of Eq. (D11)
allows us to estimate the inertial term as
dv
 v v   v  O   v 2 L  (D12)
ds
and we need only develop an estimate of the viscous
term in Eq. (D5) to complete our analysis.
We begin developing an estimate of the
magnitude of the viscous term by expanding the
Laplacian in rectangular, Cartesian coordinates to
obtain
2v 2v 2v
v   
2
(D13)
 x2  y2  z2
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

In terms of order of magnitude estimates, we express density. In general the assumption of constant total
this result as molar concentration is associated with gas-phase
 v   v y   v  diffusion processes, and the assumption of constant
 v  O 2 x   O 2   O 2 z 
2
(D14) total mass density is associated with liquid-phase
 L     
 x   Ly   Lz  diffusion processes. In this appendix we will treat
Here v x represents the change of v that takes place only the first of these two cases with the thought that
the second case can be explored on the basis of our
over the distance Lx , and the meaning of v y and analysis of the first.
v z is analogous for the y and z-directions. We now The assumption that the total molar
concentration is constant can be expressed as
represent the largest of the three terms on the right
hand side of Eq. (D14) as v L  so that our
2 Assumption: c  constant (E1)
estimate of the viscous term becomes however, nothing is constant and what is meant by
Eq. (E1) is that the variations of the total molar
 v  O  v L 
2 2
(D15) concentration are small enough so that they can be
neglected. The general application of the Stefan-
We refer to L as the viscous length and note that in
Maxwell equations requires that we replace cx A
general it is quite different than the inertial length.
with c A and this leads to the restriction given by
Once again we note that knowledge of the no-slip
condition is crucial for the determination of a reliable Restriction: x Ac  cx A (E2)
estimate of the viscous length. For a large class of
problems, v in Eq. (D15) is on the order of the If small causes give rise to small effects (Birkhoff,
velocity itself because of the no-slip condition, i.e., 1960), the condition represented by Eq. (E2) will
lead to the multi-component transport equation given
v  v , because of the no-slip condition (D16) by Eq. (204) and the binary form given by Eq. (124).
and this allows us to estimate the viscous term in In order to identify the conditions under
Eq. (D5) as which Eq. (E2) is valid, we need to express this
 v  O   v L2 
2
(D17) inequality in terms of parameters that are known a
priori and to achieve this we follow the approach
Use of this result, along with the estimate of the
outlined in Appendix D. While that approach has led
inertial term, in the restriction given by Eq. (D5)
to an established success, the problem under
leads to the inequality
consideration in this appendix is more difficult and
 v2 v further study is in order.
 2
(D18)
L L As an example we consider the process
illustrated in Fig. E1 in which the -phase represents
It is traditional to define the Reynolds number in
a flowing fluid. The  interface might be an
terms of the viscous length
interface at which adsorption or desorption occurs, or
 vL an interface at which a catalytic reaction occurs, or
Re  (D19)
 an interface at which mass transfer between the -
and this allows us to express Eq. (D18) as a phase and the -phase occurs. This could occur
constraint that takes the form because the -phase is a porous catalyst phase or
Re  L L   1
because of a difference in the chemical potential of
Constraint: (D20)
species A between the -phase and the -phase. The
We refer to this as a constraint with the thought that direction of mean flow is indicated by the unit vector
the Reynolds number, the viscous length, and the λ , and the direction orthogonal to the mean flow is
inertial length will all be known, at least in an indicated by the unit vector n .
approximate sense. This means that the domain of Directing our attention to Eq. (E2) we express
validity of Eq. (D3) is established by Eq. (D20), and that result in the form
this is something that is not done by either the
assumption given by Eq. (D4) or the restriction Restriction: c 1 c  x A1 x A (E3)
given by Eq. (D5). And note that we need to consider the gradients in
the direction of flow (the λ -direction indicated in
Appendix E: Constraints for constant total molar
Fig. 1E) and the direction orthogonal to the direction
concentration
of flow (the n -direction indicated in Fig. 1E). We
Throughout this paper we have imposed the represent the two inequalities associated with
condition of constant total molar concentration and Eq. (E3) as
constant total mass density in order to obtain
Restriction: c 1 λ  c  x A1 λ  x A (E4a)
transport equations that could be used to determine
the species molar concentration or the species mass

www.amidiq.org 241
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

Restriction: c 1 n  c  x A1 n  x A (E4b) The magnitude of the velocity in the direction of the
mean flow is given by
and note that the second of these is likely to be the
most important restriction. We limit our analysis to v  v λ (E11)
ideal gases so that the equation of state is given by and this allows us to express Eq. (E10) in the form
1   v 1
p 1 λ   p  O    O   v  v 
p  t  p
(E12)
 O   g  λ   O  2 v 
1 1
p p
Following the development given in Appendix D, we
estimate the inertial and viscous terms according to
 v  v  O   v 2 L  ,   v  O   v L  (E13)
2 2

and we estimate the local acceleration as

v  v
  O   (E14)
t  t 
in which t  is a characteristic process time. Use of
Eqs. (E13) and (E14) in Eq. (E12) leads to the
Fig. E1. Mass transfer in a two-phase system following estimate for the pressure gradient in the
direction of the mean flow:
pV  nRT , ideal gas (E5)
O   v t    O   v 2 L 
1 1
and the total molar concentration can be expressed as p 1 λ   p 
p p
(E15)
c  p RT (E6)
 O   g  λ   O   v L 
1 1 2

For an ideal gas, the left hand side of Eq. (E3) takes p p
the form Here we have assumed that the no-slip condition is
c 1c  p 1 p  T 1 T (E7) valid at the  interface, thus v can be replaced
and this yields two restrictions associated with with v .
Eq. (E3) that are given by The magnitude of the pressure can be
Restrictions: estimated as
p  O C2  (E16)
p 1 p  x A1 x A , T 1 T  x A1 x A (E8)
In this appendix we will consider only the first of where C is the speed of sound (Whitaker, 1981, Sec.
these restrictions and that requires an analysis of the 10.3). Use of this result in Eq. (E15) leads to
Navier-Stokes equations. The second restriction  M  M2 
p 1 λ   p  O     O 
requires an analysis of the thermal energy equation,
Ct   L 
and in many cases it would be appropriate to include   
(E17)
the rate of chemical reaction and the heat of reaction.  M2 1  M2 1 
 O  O
We begin our analysis of the first of Eqs. (E8)  Fr L   Re L 
     
by considering the direction of the mean flow
illustrated in Fig. E1. This leads to a restriction given in which the following dimensionless quantities have
by been used:
Restriction: p 1 λ  p  x A1 λ  x A (E9) M  Mach number  v C
Fr  Froude number  v 2 g  λ L (E18)
and we can make use of Eq. (D2) to obtain
Re  Reynolds number   vL 
1   (v  λ ) 
1
λ  p  O  
 t 
p Directing our attention to the right hand side of
p 
Eq. (E9) we estimate the gradient as

1 1
O  v  ( v  λ )   O   g  λ  (E10) λ xA  O  x A Lc  (E19)
p  p in which Lc represents the convective length scale
1
 O   2 ( v  λ )  for the transport of species A. Use of this result along
p with Eq. (E17) in Eq. (E9) leads to the constraint
given by

242 www.amidiq.org
 S. Whitaker / Revista Mexicana de Ingeniería Química Vol. 8, No. 3 (2009) 213-243

 M   M2  M2 1   M 
O    O   O p 1 n  p  O     O  M 2  
Ct   L   Fr L  Ct 
     
(E27)
(E20)
 M2 1   x A 1  M2 1  M2 1 
 O 
 O
 Re L 
 O    Fr L   Re L 
    x A Lc       

Since x A xA will be less than or on the order of In this case the Froude number is defined by
one, a conservative representation of this constraint Fr  Froude number  v 2 g  n L (E28)
is given by
Directing our attention to the right hand side of
 M Lc   M 2 Lc   M 2 Lc 
 O   O Eq. (E22) we estimate the gradient as
 Fr L 
O  
 Ct   L 
     
(E21) n x A  O  x A LD  (E29)
 M Lc 
2
O   1
 Re L 
in which LD represents the diffusive length scale for
   the transport of species A. Use of this result along
For Mach numbers small compared to one, it will be with Eq. (E27) in Eq. (E22) leads to the constraint
difficult to violate this constraint; however, one must
 M   M2 1 
O     O  M 2    O 
 Fr L 
remember that this constraint is based on Eq. (E9)
and we also need to consider the restriction given by Ct    
(E30)
Restriction: p 1 n p  x A1 n x A (E22)  M2 1   x 1 
   O  A
To explore this restriction, we express Eq. (E10) in  Re L  
the form     x A LD 
1   (v  n )  Since x A x A will be less than or on the order of
p 1 n p  O  
p   t  one, a conservative representation of this constraint
is given by
1
 O   v ( v  n )  (E23)
p   M LD   M 2 LD 
 O  M 2 LD    O 
 Fr L 
O  
1 1  Ct    
 O   g  n   O   2 ( v  n )  (E31)
p p  M 2 LD 
   1
For the special case in which n is replaced by the  Re L 
  
unit normal vector to a streamline, the inertial term
takes the form (Whitaker, Sec.7.4, 1968) Once again, it will be difficult to violate this
constraint whenever the Mach number is small
 v ( v  n )   v v  n   v 2 (E24) compared to one.
in which  is the curvature (Stein & Barcellos, Sec. At this point we have developed a constraint
13.4, 1992). Here it is important to keep in mind that associated with the restriction given by the first of
n is a constant unit vector as indicated in Fig. E1 Eqs. (E8); however, the second restriction given by
while the unit normal to a streamline will be a Restriction: T 1 T  x A1x A (E32)
function of position. Equation (E24) helps us to
estimate the inertial term in Eq. (E23) as still needs to be explored. This will require an
1 1 analysis of the thermal energy equation and for most
O   v  ( v  n )    v 2  (E25) realistic systems the thermal energy equation will be
p  p coupled to a mass transfer and reaction process. In
in which  represents some appropriate mean addition, the constraints associated with non-ideal
curvature associated with the system illustrated in gases need to be developed along with the
Fig. E1. About the other terms in Eq. (E23) we can constraints associated with the assumption for liquid
only say that they will be smaller than the analogous phase mass transfer processes given by
terms in Eq. (E10). This means that we can over Assumption:   constant (E33)
estimate p 1 n p as
The constraints associated with this assumption will
1  (v  λ )  1
p 1 n p  O    O   v  
2 be much more difficult to obtain than those given by
p  t  p Eqs. (E21) and (E31); however, the developments
1 1 presented in this appendix should provide guidance
 O   g  n   O   2 ( v  λ )  for the attack on Eq. (E33).
p p
(E26)
and follow our earlier development to obtain

www.amidiq.org 243

View publication stats