1

Module 4:
Multicomponent Transport

Lecture 36:
Solving the Multi-component Flux
Equations

NPTEL,IITKharagpur,Prof.SaikatChakraborty,DepartmentofChemicalEngineering

Given the continuity equation

Ci
+ .( vCi ) = .Fi
t

(3.7.35)

Where the multicomponent flux

Fi = Dij C j

(3.7.36)

j =1

Reformulate this problem in a form resembling a binary diffusion problem.

--------------------Eqs. (3.7.35), (3.7.36) can be written in matrix form

C
+ .( vC) = .F
t

F = D C

(3.7.37)

(3.7.38)

Where C is the vector of species concentrations and D is the multicomponent diffusion matrix
These are subject to

NPTEL,IITKharagpur,Prof.SaikatChakraborty,DepartmentofChemicalEngineering

Initial Conditions :

C ( x, y , z , t = 0) = C0

(3.7.39)

C ( B , t ) = 0

(3.7.40)

C
(b , t ) = 0
Z

(3.7.41)

Boundary Conditions :

Where B and b represent boundaries of the system,

C is a concentration difference that generally

varies with position and time.

Note: The boundary condition on all concentrations must have the same functional form for this analysis
to be applicable. This may be a serious restriction where chemical reactions on boundaries take place.
Assume there exists a non-singular matrix t which can diagonalize D , i.e

1 0 0
0
0
2
1
t . D. t = =
0 0 3

.... .... ....

....
....
....

....

(3.7.42)

1
Here t is the inverse of t , and is the diagonal matrix of the eigenvalues of matrix D .

NPTEL,IITKharagpur,Prof.SaikatChakraborty,DepartmentofChemicalEngineering

For a ternary mixture

t11
t =
t21

D12

D22 1

22

2
1

t12 D12

t22
det( t )

Note: find the components of for a ternary mixture

Now, use matrix t

C = t .

to define a new concentration ( ) variable, such that

(3.7.43)

Combining (3.7.37), (3.7.38) and (3.7.43)

+ .v = 2
t

(3.7.44)

Which represents a set of scalar equations

i
+ .v i = i 2 i
t

(3.7.45)

NPTEL,IITKharagpur,Prof.SaikatChakraborty,DepartmentofChemicalEngineering

Note: We have assumed that D (hence both t and ) are not functions of composition.
Eqs. (3.7.39) (3.7.41) can also be written in terms of .

( x, y, z , t = 0) = 0 = t 1. C0

(3.7.46)

( B, t ) = 0

(3.7.47)

(b, t ) = 0

(3.7.48)

This way, a set of coupled PDEs has been reduced to a set of uncoupled PDEs in terms of .

Eqs. (3.7.45) (3.7.48) are of the same form as the associated binary diffusion problem.

C1
+ .v C1 = D 2C1
t
If this binary problem has the solution

C1 = G ( D) C10
Then eqs. (3.7.45) (3.7.48) have the solution

i = G ( i ) i 0
NPTEL,IITKharagpur,Prof.SaikatChakraborty,DepartmentofChemicalEngineering

Where the eigenvalue i is substituted everywhere that the diffusivity occurring in the binary solution. In
terms of the actual concentration,

C = t .G ( ). t 1. C0

(3.7.49)

NPTEL,IITKharagpur,Prof.SaikatChakraborty,DepartmentofChemicalEngineering