Transport Phenomena in Metallurgical

Processes (MMC 401)

Dr. Manas Kumar Mondal

Associate Professor,
Department of Metallurgical and Materials Engineering
National Institute of Technology, Durgapur
 Disclaimer
The study materials/presentations are solely meant for academic purposes and they can
be reused, reproduced, modified, and distributed by others for academic purposes only
with proper acknowledgements.

Lecture 18
References:
 D.R. Poirier and G.H. Geiger, “Transport Phenomena in Materials Processing”, (Springer International Publishers. Switzerland,2016 ).
 Julian Szekely and N.J. Thermelis, “Rate Phenomena in Process Metallurgy” ,(John Wiley & Sons Inc (1 November 1971).
 R. Byron Bird, Warren E. Stewart, and Edwin N. Lightfoot, “TRANSPORT PHEOMENA”, (OHN WILEY & SONS, Inc., 2002).
 D. R. Gaskell, “An Introduction to Transport Phenomena in Materials Engineering”, (MOMENTUM PRESS, LLC, NEW JERSEY,
2013).
 S. P. Sukhatme, “A Textbook on Heat Transfer”, Fourth Edition (University Press India Ltd., 2005).

 J. P. Holman, “Heat Transfer”, (The McGraw-Hill Companies, Inc., )

MASS TRANSPORT
 Problem: Calculate the values of i, Ci, Xi and Ni for oxygen and nitrogen in air at NTP assuming these gases to behave
ideally. Molecular weights of O2 = 32 and that of N2 = 28, R = 8.314  103 kg m2s2kg-mol1K1 and 1 atm = 1.0133  105
kg m1s2.

Solution:
Air is known to contain 79% N2 and 21% O2 by volume. Therefore, one cubic meter of air at NTP will contain 0.79 m3 and
0.21 m3 of N2 and O2 respectively.

According to the Ideal gas law PV  nRT  C  n  p i

i i

V RT
Pi = volumetric fraction of the component i  total pressure = PVi

The volumetric fraction of N2, VN2 is 0.79 and of O2, VO2 is 0.21.

P  VN 1.0133  105  0.79

CN  2
  3.53  10 2 kg-mole. m3
2
RT 8.314  10  273
3

and mass concentration N2 is calculated from the value of CN2 and its molecular weights MN2 as

N2 = MN2 CN2 = 28  3.53  102 = 0.99 kg m3

P  VO 1.0133  105  0.21

CO  2
  9.4  10 3 kg-mole. m3
2
RT 8.314  10  273
3
 O2 = MO2 CO2 = 32  9.4  103 = 0.3 kg m3
N 0.99 X O'  1  X N'  1  0.77  0.23
Mass fraction are calculated as X 
' 2
  0.77 and
 N  O 0.99  0.3
N2 2 2

2 2

CN 3.53  10 2
Mole fraction are calculated as X   2
 0.79
C N  CO 3.53  10 2  9.4  10 3
N2

2 2

and X O  1  X N  1  0.79  0.21

2 2

FACTORS AFFECTING DIFFUSION

Diffusing Species:

 The magnitude of the diffusion coefficient D is indicative of the rate at which atoms diffuse. The diffusion species as well
as the host material influence the Diffusion coefficient.
 For example, there is a significant difference in magnitude between self and carbon interdiffusion in an iron at 500C, the
D values being greater for the carbon interdiffusion (3.01021 vs 2.41012 m2/sec). This comparison also provides a
contrast between rates of diffusion via vacancy and interstitial modes.
 Self diffusion occurs by a vacancy mechanism, where as carbon diffusion in iron is interstitial.
 Temperature:
 By experiments it is observed that diffusion speeds up with increasing temperature.
 Temperature has a most profound influence on the coefficients and diffusion rates.
 For example, for the self diffusion of Fe in -Fe, the diffusion coefficient is increase approximately six orders of
magnitude (from 3.01021 to 1.81015 m2/sec) is rising temperature from 500 to 900C.

 The temperature dependence of the diffusion coefficient is

 Q 
D  D exp  ......................(24)
d

 RT 
0

where D0 = a temperature independent pre-exponentional (m2/s)

Qd = the activation energy for diffusion (J/mol)

R = the gas constant, 8.314 J/mol-K

T = absolute temperature (K)

 The activation energy may be thought of as that energy requires to produce the diffusion motion of one mole of atoms. A
large activation energy result in a relatively small diffusion coefficient.
 Take natural logarithm on both side of equation (24),
 Q 
ln D  ln D exp   d

 RT 
0

In terms of logarithm to the base 10

Qd  1 
log D  log D0   ...................( 25)
2.303 R  T 

Since D0, Qd & R are all constant. Equation (25) takes the forms of an equation of a straight line y=b+mx.

Where y & x are analogous, respectively to the variables log D and 1/T.

Thus, if log D is plotted versus the reciprocal of the absolute temperature, a

straight line should result which have the slope and intercept of  Qd/2.303R
and log D0 respective.

Figure 3: Variation of diffusion coefficient

with reciprocal of temperature
The Type of diffusion Mechanism:

 Whether the diffusion interstitial a substitutional will affect the diffusion process. Small atoms diffusion is called
interstitial and approximately same size atoms diffuse substitutional in the FCC or BCC.

The type of crystal Structure:

 The atomic packing of BCC crystal structure is 0.68 while in FCC is 0.74, means more inter-atomic spaces available in
BCC crystal structure resulting more carbon atom can diffuse in BCC crystal structure of iron in comparison to FCC
iron.
 Experimentally it is found that the diffusivity in BCC iron is 2.4 1012 m2/s at 500C and 5  1015 m2/s in case of FCC.

Fick’s Second law of Diffusion

 Fick’s second law is concerned with diffusion under unsteady condition

i.e. the flux is a function of both space and time. The flux J varies from
section to section and varies with time at a given section.
 In this case the concentration profile will be function of space and time.
This kind of situation normally exists in practical system and so Fick’s
second law is applicable to real system.
 Consider section of thickness Δx with unit area of cross-section in which there is diffusion of atoms. Let the flux at x
be Jx and (x+Δx) be Jx+Δx.

J
J J  x......................(26)
x
x  x x

Difference in flux between two surfaces is

J J
J J J J  x  
x
x.....................(27)
x

x x
x x  x x x

The difference in flux gives the rate of accumulation of the diffusing atom within the element, which can be
expressed in terms of time dependence of the concentration.
 C
Rate of a accumulation of depletion of atoms in the section of volume (Δx  1) cm3 = Cx   x
t t
C J C C 2

x   x  D x
as Jx =  D C/x …………………(28)
t x t x 2

It is assumed that the diffusion coefficient is independent of concentration and x. This is (equation (28)) called Fick’s
second law of diffusion. The solution of the Fick’s second law is
 x 
C  x, t   A  Berf  .....................( 29)
 2 Dt 
Where erf( ) is special called error function and A & B are constant. The error function erf(Z) is defined as

2
erf Z   d
Z
e
 2

 0

where,  is a variable that gets eliminated when the limits of integration is substituted.
The A & B are determined by fixing the initial condition and boundary conditions. The most practical application the
following conditions can be applied.

Consider a long rod in which diffusion of atoms take place. One end of the rod is exposed to the diffusing atoms (called
solute) maintain at a particular concentration, say CS.

Initially the concentration of the solute in the material is uniform equal to C0 throughout the material. The bar is
considered to be infinitely long or is long enough that the concentration remains C0 at the other end of the bar during
the time over which diffusion take place.

Taking the end of the bar exposed to the diffusing atoms as the origin the initial concentration are:

C(x,0) = C0 i.e. initial concentration of the solute in the bar is uniform throughout length of the bar.

C(0,t) = CS i.e. the concentration at the end which is exposed to the solute is maintain constant throughout the
diffusion process.
Using above condition we get

C(x,0) = C0 = A B (as erf() = 1)

C(0,t) = CS = A (as erf(0) = 0)

Therefore, B = (CS  C0)

Then the solution of equation (29) is

 x   x 
C  x, t   C S  CS  C0 erf    C S  C0   C0  CS  C0 erf  
 2 Dt   2 Dt 

  x 
 C  x, t   C  C  C 1  erf  
 2 Dt 
0 S 0

C  x, t   C  x 
  1  erf 
0
................................(31)
C  C 
S 0  2 Dt 

Equation (31) is a very useful equation for determining the concentration of the solute atoms in a solid at different depth,
when rod is exposed to the solute for a known interval of time.
 Just maintaining a concentration gradient in a solid is not sufficient to cause diffusion. For diffusion to take place
under concentration gradient, the atoms need a minimum energy to overcome the barrier due to the potential which
binds them to crystal lattice. Thermal energy can provide the required minimum energy. So the diffusion coefficient
D is a function of temperature. Normally, diffusion process is carried out at a sufficiently high temperature.

 The diffusion coefficient D is the material constant that describe the kinetic rate of diffusion. The diffusion coefficient
depends on the temperature because as the temperature increases the energy with which atoms vibrate about their
equilibrium position within the crystal lattice increases, occasionally an atomic vibration is a sufficient energy to allow
an atom to move new lattice position. Due to increase in temperature atomic vibration increases in frequency and
amplitude. Therefore, we can say that temperature increase the base energy level of the material, making the energy
barrier easier to overcome. Thus, diffusion is a thermally active process.

 The relationship between these variables obeys the Arrhenius relationship

 E 
D  D exp  ....................(32)
 RT 
0

where D0 = frequency factor

Qd = the activation energy for diffusion (J/mol)
R = the gas constant, (8.314 J/mol-K)
T = absolute temperature (K)
 Problem: A piece of steel having 0.10 % carbon is carburized at 900C for 600 mins (10 hrs). Draw the graph showing
the carbon concentration in the case after 10 hrs carburizing operation. Carbon is diffusing in FCC iron.
Take; D0 = 0.21cm2/s, E = 142 KJ/mole, R = 8.314 J/mol-K, CS =1.3 %, erf(0.083) = 0.1, erf(0.415) = 0.45, erf(0.83) =
0.77, erf(1.66) = 0.98, and erf(2.49) = 1

Solution: Diffusion coefficient

 E   142000 
D  D exp    0.21exp    9.97  10 cm / s
8 2
0
 RT   8 .314  900  273 

For the calculation of the carbon concentration at various depth following relationship may be used

 x 
C  x, t   C  C  C erf  
 2 Dt 
S 0 S

 x 
 C  x, 10  3600   1.3  0.10  1.3erf  
 2 9.97  10  3600  10 
8

 C  x   1.3  1.2erf 8.3 x ...................(1)

Now if we put different value of x in equation (1), then we will get value of C(x) for corresponding value of x.
 x C(x)
0 1.3
0.01 1.18
0.05 0.76
0.1 0.38
0.2 0.12
0.3 0.1

Problem: If boron is diffused into a slice of pure silicon at 1100C for 2 hrs, what is depth below the surface at which
the concentration 1017 atoms/cm2, if surface concentration is 1018 atom/cm3. Diffusion coefficient of boron into
silicon is 4  1013 cm2/s. erf(1.2) =0.9
Solution: The solution of the Fick’s second law of diffusion is C  x, t   A  Berf 
x 

 2 Dt 
C (0,t) = 1018 atoms/cm3 at the surface.
As the boron is diffused into a slice of pure silicon, then C(x,0) = AB = 0
Therefore, A = B = 1018 atoms/cm3
C(x, 2hr) = 1017 atoms/cm3
  x  A  C  x, t  10  10
18 17

erf   B
  0.9 erf (1.2)  0.9
 2 Dt  10 18

x
 1.2  x  1.2  2  4  10  3600  2  1.29  10 cm
13 4

2 4  10  3600  2
13

Carburized case Depth

Carburized case depth can be defined as the depth corresponding to a carbon concentration that is the average of the
initial and saturation carbon concentration.
C C
Thus the carbon concentration at case depth is C x   S 0

2
p

 x 
The solution of the Fick’s second law of diffusion is C  x, t   C  C  C erf  
 2 Dt 
S 0 S

C C  x 
As erf (0.5) =0.5 therefore,  S
 C  C  C erf 
0

2  2 Dt 
S 0 S

x p
 0.5  x  Dt 1C C  x 
2 Dt
p
  erf 
0
 S

2C C 0  2 Dt  S

So case depth is (Xp) = Dt  x  1

 erf  
 2 Dt  2
Problem: Decomposition of poly (vinyl chloride)
Poly (vinyl chloride) or PVC, like other polymers, is subject to degradation reactions at high temperature. In the case of
PVC, HCl is evolved when a monomer in the chain undergoes the reaction
 C H Cl   C H    HCl
2 3 2 2

This take palace throughout the polymer body at a constant rate G which is independent of HCl concentration and we’ll
assume the back-reaction occurs at a negligible rate even at maximum HCl concentrations.

A PVC rod of radius R is held at high temperature for a long time, so that the concentration of HCl reaches a steady
state. On the surfaces, the rod reaches equilibrium with its surrounding at a fixed HCl concentration CHCl,s. Since the
properties of PVC are affected by the presence of dissolved HCl, we want to know the concentration profile of HCl in
the rod.

(a) Write down the differential equation for steady state diffusion and homogeneous chemical reaction at constant
reaction rate G in a cylindrical rod. What is the general form of the solution to this equation?

(b) For the boundary conditions at the centre and on the surface of the rod given above, what is the solution? What is the
maximum concentration of HCl?
2
GR
(c) Consider the dimensionless quantity DC . What does the value of the quantity tell you about the uniformity of
Hcl , S
concentration of HCl through the rod?
 Solution:

(a) Consider a cylinder of radius R and length L. Moreover, consider a shell which is extended from r to Δr.

Heat balance over this cylinder shell extending from r to r + Δr under steady state condition yields

0   J 2rL    J 2rL 
r r r r  r
 G.2rLr

d
Again we divide by 2LΔr and let Δr go to zero   J   Gr  0
dr
r

d  dC 
Substituting Fick’s first law gives us the differential equation:  rD   Gr  0
dr  dr 
dC Gr 2

This integrates to rD  A
dr 2
Substituting Fick’s first law gives us the differential equation:

d  dC 
 rD   Gr  0
dr  dr 

dC Gr 2

This integrates to rD  A
dr 2

dC rG A
For a constant A. Now divide by r and rearrange to give  
dr 2 D Dr

rG 2
A
 A ln r  B where once again A 
'
and integrate both sides for the general solution C   '

4D D

(b) Because the rod is symmetric about its axis, the concentration everywhere around the axis will be the same, so there
dC
will be no flux through the centre. This means at the centre (r = 0)  0.
dr

dC rG A '

Therefore,   0
dr 2D r

Thus, A  0. If the concentration is finite at the centre then A must be zero.

' '
 At the surface (r = R) C  C Hcl HCl , S

Then C RG 2
RG 2

  B  B C 
4D 4D
Hcl , S Hcl , S

rG 2
RG
2
G R  r
2 2

The solution is C   C  C 
4D 4D 4D
Hcl , S Hcl , S

GR
2

The maximum concentration is the concentration at r = 0 C C 

4D
max Hcl , S

C GR 2

(c) If we divide the maximum concentration by the surface concentration, we get: max
 1
C
Hcl , S
4 DC Hcl , S

GR 2
C
This means that if is small, max
will be very close to one, so the HCl concentration will be uniform. If it is
DC Hcl , S C Hcl , S

large, the maximum concentration will be much larger than that at the surface of the rod.