BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2A.

1 Page 1 of 2

Problem 2A.1
Thickness of a falling film. Water at 20°C is flowing down a vertical wall with Re = 10.
Calculate (a) the flow rate, in gallons per hour per foot of wall width, and (b) the film thickness
in inches.

Answer: (a) 0.727 gal/hr · ft; (b) 0.00361 in

Solution

The Reynolds number is related to the film thickness δ and the average velocity hvz i by Eq.
2.2-25 on page 47.
4δhvz iρ
Re = (2.2-25)
µ
We have
4δhvz iρ
= 10.
µ
Solve this for δ.
5 µ
δ= (1)
2 ρhvz i
Part (a)

The mass flow rate w of a falling film is given by Eq. 2.2-21 on page 46.

w = ρW δhvz i (2.2-21)

Divide both sides by W to obtain the mass flow rate per unit of wall width.
w
= ρδhvz i
W
Use equation (1) here.
5 µ
=ρ· hvz i
2 ρhvz i
5µ
=
2
µ, the viscosity of water, at 20°C can be found in Table 1.1-2 on page 14: µ = 1.0019 mPa · s
(centipoise). Use the conversion factor in Table F.3-4 on page 870 to convert this to lbm /ft · hr.
lbm
w 5 2.4191 ft·hr lbm
= · 1.0019   ×  ≈ 6.0592

centipoise
W 2 1
centipoise
   ft · hr
We have to divide this result by the density of water to get units of volume in the numerator.
Before we do, find the density in units of lbm /gal, starting with the known value 1000 kg/m3 and
using the conversion factors on page 868.
 3 3
kg 2.2046 lbm 1Zm 0.13368 ft
Z Z lbm
1000 3 × × × ≈ 8.345
Z
mZ
Z 1 kg 3.28 @
ft 1 gal gal

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2A.1 Page 2 of 2

Therefore, we have for the flow rate

w lbm 1 gal gal
≈ 6.0592 × ≈ 0.726 .

ρW ft · hr 8.345 
lbm hr · ft

Part (b)

The film thickness is given by equation (1) above. The average velocity hvz i of the film is given by
Eq. 2.2-20 on page 45.
ρgδ 2 cos β
hvz i = (2.2-20)
3µ
β represents the angle from the vertical that the film is flowing down. Since the wall is vertical,
β = 0 and cos β = 1. Substitute this expression for hvz i into equation (1).
5µ 3µ
δ= ·
2ρ ρgδ 2
15µ2
δ3 = 2
2ρ g
s
15µ2
δ= 3
2ρ2 g
s
15(1.0019 × 10−3 Pa)2
≈ 3
2(1000 kg/m3 )2 (9.81 m/s2 )
3.28 @
ft 12 in
≈ 9.16 × 10−5 
×
m ×
1m
 1@
ft
Therefore,
δ ≈ 0.00360 in.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2A.2 Page 1 of 1

Problem 2A.2
Determination of capillary radius by flow measurement. One method for determining the
radius of a capillary tube is by measuring the rate of flow of a Newtonian liquid through the tube.
Find the radius of a capillary from the following flow data:

Length of capillary tube 50.02 cm

Kinematic viscosity of liquid 4.03 × 10−5 m2 /s
Density of liquid 0.9552 × 103 kg/m3
Pressure drop in the horizontal tube 4.829 × 105 Pa
Mass rate of flow through tube 2.997 × 10−3 kg/s

What difficulties may be encountered in this method? Suggest some other methods for
determining the radii of capillary tubes.

Solution

The mass rate of flow is given by the Hagen-Poiseuille equation,

π(P0 − PL )R4 ρ
w= .
8µL
Solve this equation for the radius R.
8µLw
R4 =
π(P0 − PL )ρ
Take the fourth root of both sides.
s
8µLw
R= 4
π(P0 − PL )ρ

We want to write this in terms of kinematic viscosity ν, one of the given quantities. Use Eq.
1.1-3, ν = µ/ρ, on page 13.
s
8νLw
= 4
π(P0 − PL )

Now plug in the numbers.

s
4 8(4.03 × 10−5 m2 /s)(0.5002 m)(2.997 × 10−3 kg/s)
=
π(4.829 × 105 Pa)

Therefore,
R ≈ 7.51 × 10−4 m or R ≈ 0.751 mm.
The problem with this method is that there are many assumptions that have to hold in order for
the Hagen-Poiseuille equation to be valid. The assumptions are summarized on page 52: flow is
laminar, density is constant, flow is steady, fluid is Newtonian, capillary length must be much
larger than its radius, fluid behaves as a continuum, and no slip at capillary wall. To determine
the radius of a capillary tube, measure it directly with a micrometer. If one is not available, find
out what volume of liquid it takes to fill up the capillary tube and use V = πR2 L to find R.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2A.3 Page 1 of 1

Problem 2A.3
Volume flow rate through an annulus. A horizontal annulus, 27 ft in length, has an inner
radius of 0.495 in. and an outer radius of 1.1 in. A 60% aqueous solution of sucrose (C12 H22 O11 )
is to be pumped through the annulus at 20°C. At this temperature the solution density is
80.3 lbm /ft3 and the viscosity is 136.8 lbm /ft · hr. What is the volume flow rate when the
impressed pressure difference is 5.39 psi?

Answer: 0.110 ft3 /s

Solution

We obtain the volume flow rate from the mass flow rate w.
dm d(ρV ) dV
w= = =ρ
dt dt dt
Dividing both sides by the density ρ, we obtain the volume flow rate.
dV w
= .
dt ρ
The mass flow rate in an annulus is given by Eq. 2.4-17 on page 55.
π(P0 − PL )R4 ρ (1 − κ2 )2
 
4
w= (1 − κ ) − (2.4-17)
8µL ln(1/κ)
Thus,
π(P0 − PL )R4 (1 − κ2 )2
 
dV 4
= (1 − κ ) − .
dt 8µL ln(1/κ)
Our aim now is to convert all the given quantities to SI units before we plug them into this
formula. The conversion factors are in Table F.3-2 and Table F.3-4 on page 869 and page 870,
respectively.
lbf 6.8947 × 103 Pa
P0 − PL = 5.39 2 × lb
≈ 3.72 × 104 Pa
in f
1 in2

4.1338 × 10−4 Pa · s
· hr × ≈ 0.05655 Pa · s

µ = 136.8 
lb
m /ft

· hr

1
lb
m /ft

1m
L = 27 ft × ≈ 8.2 m
3.28 ft
1@
ft 1m
R = 1.1 in × × ≈ 0.028 m
12 in 3.28 @ ft
The dimensionless quantity κ is the ratio of the inner radius to the outer radius.
0.495 in
κ= = 0.45
1.1 in
Now substitute all these numbers into the formula. Therefore,
π(3.72 × 104 Pa)(0.028 m)4 (1 − 0.452 )2
 
dV 4
≈ (1 − 0.45 ) −
dt 8(0.05655 Pa · s)(8.2 m) ln(1/0.45)
3
  3 3
m 3.28 ft ft
≈ 0.00310 × ≈ 0.110

.
s 1m
 s

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2A.4 Page 1 of 2

Problem 2A.4
Loss of catalyst particles in stack gas.

(a) Estimate the maximum diameter of microspherical catalyst particles that could be lost in
the stack gas of a fluid cracking unit under the following conditions:

Gas velocity at axis of stack = 1.0 ft/s (vertically upward)

Gas viscosity = 0.026 cp
Gas density = 0.045 lbm /ft3
Density of a catalyst particle = 1.2 g/cm3

Express the result in microns (1 micron = 10−6 m = 1µm).

(b) Is it permissible to use Stokes’ law in (a)?

Answers: (a) 110 µm; Re = 0.93

Solution

Convert all the given quantities to SI units and assign variables to them. The conversion factors
can be found on page 868 and 870.

ft 1m m
Gas velocity at axis of stack v∞ = 1.0 × ≈ 0.3049
s 3.28 ft s
10−3 Pa · s
Gas viscosity µ = 0.026 ×
cp = 2.6 × 10−5 Pa · s
1cp

 3
lbm 1 kg 3.28 @
ft kg
Gas density ρg = 0.045 3 × × ≈ 0.7203 3

ZftZ 2.2046 
lbm 1m m
 3
g 1 kg cm
100 H kg
Density of a catalyst particle ρc = 1.2 H
3 × × = 1200 3
H
cmH 1000 g
1m m

Part (a)

Gas particles in a stack flow upward. Those at the center of the stack travel at the highest
velocity since they are furthest from the walls. Consequently, to find the maximum diameter of a
catalyst particle, we consider one at the center of the stack. Catalysts that are lost do not rise up
with the rest of the gas; rather, they fall at terminal velocity in the stack and have an acceleration
of zero. The sum of the forces acting on one in the y-direction must be equal to zero. The
gravitational force is pulling the catalysts down, and the buoyant and kinetic (drag) forces are
pushing them up.

X
Fy = F − Fg = 0
4 3 4 3
= πR ρg g + 2πµRv∞ + 4πµRv∞ − πR ρc g =0
|3 {z } | {z } | {z }
form drag friction drag
|3 {z }
buoyant force weight of catalyst
4
= πR3 g(ρg − ρc ) + 6πµRv∞ = 0
3

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2A.4 Page 2 of 2

Figure 1: Free body diagram of a catalyst particle at the center of a stack. Define upward forces
to be positive.

Solve this equation for the radius R.

s
6µv∞ 3
R= ·
g(ρc − ρg ) 4
s
9µv∞
=
2g(ρc − ρg )
s
9(2.6 × 10−5 Pa · s)(0.3049 m/s)
≈
2(9.81 m/s2 )(1200 − 0.7203) kg/m3
≈ 5.5 × 10−5 m

The diameter is twice the radius.

D = 2R ≈ 2(5.5 × 10−5 m)
1 micron
≈ 1.1 × 10−4 
×
m = 110 microns
10−6 
m

Part (b)

Calculate the Reynolds number.

Dv∞ ρg (1.1 × 10−4 m)(0.3049 m/s)(0.7203 kg/m3 )

Re = ≈
µ (2.6 × 10−5 Pa · s)
≈ 0.93

Since the Reynolds number is less than 1, Stokes’ law is valid.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.1 Page 1 of 4

Problem 2B.1
Different choice of coordinates for the falling film problem. Rederive the velocity profile
and the average velocity in §2.2, by replacing x by a coordinate x measured away from the wall;
that is, x = 0 is the wall surface, and x = δ is the liquid–gas interface. Show that the velocity
distribution is then given by
 
2 1 2
vz = (ρgδ /µ) (x/δ) − (x/δ) cos β (2B.1-1)
2
and then use this to get the average velocity. Show how one can get Eq. 2B.1-1 from Eq. 2.2-18
by making a change of variable.

Solution

Figure 1: Schematic of the falling film with x measuring the distance from the wall.

We assume that the fluid flows in the z-direction and that its velocity varies in the x-direction.

vz = vz (x)

As a result, only φx̄z (the z-momentum perpendicular to the x-direction) and φzz (the
z-momentum perpendicular to the z-direction) contribute to the momentum balance. We also
assume that the pressure only varies with depth.

p = p(x̄)

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.1 Page 2 of 4

Figure 2: This is the shell over which the momentum balance is made for the falling film. As a
result of the assumption vz = vz (x̄), the dashed-underlined terms are equal to zero.

Rate of z-momentum into the shell at z = 0: (W ∆x̄)φzz |z=0

Rate of z-momentum out of the shell at z = L: (W ∆x̄)φzz |z=L
Rate of z-momentum into the shell at x̄: (W L)φx̄z |x̄
Rate of z-momentum out of the shell at x̄ + ∆x̄: (W L)φx̄z |x̄+∆x̄
Component of gravitational force on the shell in z-direction: (W L∆x̄)ρg cos β

If we assume steady flow, then the momentum balance is

Rate of momentum in − Rate of momentum out + Force of gravity = 0.

Considering only the z-component, we have

(W ∆x̄)φzz |z=0 − (W ∆x̄)φzz |z=L + (W L)φx̄z |x̄ − (W L)φx̄z |x̄+∆x̄ + (W L∆x̄)ρg cos β = 0.

Factor the left side.

W ∆x̄( φzz |z=0 − φzz |z=L ) + W L( φx̄z |x̄ − φx̄z |x̄+∆x̄ ) + W L∆x̄ρg cos β = 0

Divide both sides by the volume of the shell W L∆x̄.

φzz |z=0 − φzz |z=L φx̄z |x̄+∆x̄ − φx̄z |x̄

− + ρg cos β = 0
L ∆x̄
Since we assumed that the pressure and velocity only vary as a function of x̄, φzz is the same at
z = 0 as it is at z = L.
φx̄z |x̄+∆x̄ − φx̄z |x̄
− + ρg cos β = 0
∆x̄
Now take the limit as ∆x̄ → 0.
φx̄z |x̄+∆x̄ − φx̄z |x̄
lim − + ρg cos β = 0
∆x̄→0 ∆x̄

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.1 Page 3 of 4

This is how the first derivative is defined.

dφx̄z
− + ρg cos β = 0
dx̄
Since φx̄z = τx̄z + 
ρv
xv
 z , we get
dτx̄z
= ρg cos β.
dx̄
There are two boundary conditions associated with this differential equation—one at the surface
of the film x = δ and one at the wall x = 0.

B.C. 1 (free surface): τx̄z = 0 when x̄ = δ

B.C. 2 (no-slip): vz = 0 when x̄ = 0

Integrate both sides of it with respect to x̄.

τx̄z (x̄) = ρg(cos β)x̄ + C1

Apply the first boundary condition.

τx̄z (δ) = ρg(cos β)δ + C1 = 0 → C1 = −ρg(cos β)δ

So we have
τx̄z (x̄) = ρg(cos β)x̄ − ρg(cos β)δ.
From Newton’s law of viscosity we know that τx̄z = −µ(dvz /dx̄), so

dvz
−µ = ρg(cos β)x̄ − ρg(cos β)δ.
dx̄
Divide both sides by −µ.
dvz ρg ρg
= − (cos β)x̄ + (cos β)δ
dx̄ µ µ
Integrate both sides with respect to x̄ once more.
ρg ρg
vz (x̄) = − (cos β)x̄2 + (cos β)δ x̄ + C2
2µ µ
Apply the second boundary condition.

vz (0) = C2 = 0

So we have
ρg ρg
vz (x̄) = − (cos β)x̄2 + (cos β)δ x̄.
2µ µ
Factor (ρgδ 2 /µ) cos β.
ρgδ 2 x̄2
 
x̄
vz (x̄) = − 2+ cos β
µ 2δ δ
Therefore, the velocity distribution is

ρgδ 2  x̄  1  x̄ 2
 
vz (x̄) = − cos β.
µ δ 2 δ

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.1 Page 4 of 4

The average velocity is obtained by integrating the velocity distribution over the cross-sectional
area and then dividing by that area.
ˆ
1
hvz i = vz dA
A
ˆ δ
1
= vz (W dx̄)
Wδ 0
ˆ
1 δ
= vz dx̄
δ 0
ˆ δ  
ρgδ 2 1  x̄ 2

x̄
= cos β − dx̄
µδ 0 δ 2 δ
 2 δ
ρgδ x̄ 1 x̄3
= cos β −
µ 2δ 2 3δ 2 0
 2
1 δ3

ρgδ δ
= cos β −
µ 2δ 2 3δ 2
 
ρgδ δ
= cos β
µ 3

Therefore, the average velocity is

ρgδ 2
hvz i = cos β,
3µ
which is the same as Eq. 2.2-20. From Eq. 2.2-18,

ρgδ 2 cos β
  x 2 
vz = 1− , (2.2-18)
2µ δ

one can get Eq. 2B.1-1 by making the substitution x = δ − x̄.

"  #
ρgδ 2 cos β δ − x̄ 2

vz = 1−
2µ δ
2 x̄ 2
 
ρgδ cos β 
= 1− 1−
2µ δ
2 x̄ x̄2
  
ρgδ cos β
= 1− 1−2 + 2
2µ δ δ
2 2
 
ρgδ cos β x̄ x̄
= 2 − 2
2µ δ δ
ρgδ cos β x̄ 1 x̄2
2
 
= −
µ δ 2 δ2
ρgδ 2  x̄  1  x̄ 2
 
= − cos β
µ δ 2 δ

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.2 Page 1 of 2

Problem 2B.2
Alternate procedure for solving flow problems. In this chapter we have used the following
procedure: (i) derive an equation for the momentum flux, (ii) integrate this equation, (iii) insert
Newton’s law to get a first-order differential equation for the velocity, (iv) integrate the latter to
get the velocity distribution. Another method is: (i) derive an equation for the momentum flux,
(ii) insert Newton’s law to get a second-order differential equation for the velocity profile, (iii)
integrate the latter to get the velocity distribution. Apply this second method to the falling film
problem by substituting Eq. 2.2-14 into Eq. 2.2-10 and continuing as directed until the velocity
distribution has been obtained and the integration constants evaluated.

Solution

The two equations from the text are as follows.

dτxz
= ρg cos β (2.2-10)
dx
dvz
τxz = −µ (2.2-14)
dx
Substituting Eq. 2.2-14 into Eq. 2.2-10, we obtain
 
d dvz
−µ = ρg cos β
dx dx
d2 vz
−µ 2 = ρg cos β.
dx
This is a second-order differential equation for the velocity profile. Divide both sides by −µ.

d2 vz ρg cos β
=−
dx2 µ
There are two boundary conditions—one at the surface x = 0 and one at the wall x = δ.
Substitute Eq. 2.2-14 into B.C. 1 to get it in terms of vz .

dvz dvz
B.C. 1 (free surface): τxz = 0 when x = 0 → −µ =0 → =0
dx x=0 dx x=0

B.C. 2 (no-slip): vz = 0 when x = δ

Integrate both sides of the differential equation with respect to x.

dvz ρg cos β
=− x + C1
dx µ
Apply B.C. 1 here.
dvz
= C1 = 0
dx x=0
Integrate both sides of the differential equation with respect to x once more.
ρg cos β 2
vz (x) = − x + C2
2µ

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.2 Page 2 of 2

Apply B.C. 2 here.

ρg cos β 2 ρg cos β 2
vz (δ) = − δ + C2 = 0 → C2 = δ
2µ 2µ
So we have
ρg cos β 2 ρg cos β 2
vz (x) = − x + δ
2µ 2µ
ρg cos β 2
= (δ − x2 ).
2µ
Therefore,
ρgδ 2 cos β
  x 2 
vz (x) = 1− .
2µ δ

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.3 Page 1 of 6

Problem 2B.3
Laminar flow in a narrow slit (see Fig. 2B.3).

Figure 1: Fig. 2B.3 in the text. Flow through a slit, with B  W  L.

(a) A Newtonian fluid is in laminar flow in a narrow slit formed by two parallel walls a distance
2B apart. It is understood that B  W , so that “edge effects” are unimportant. Make a
differential momentum balance, and obtain the following expressions for the momentum-flux
and velocity distributions:
P0 − PL
 
τxz = x (2B.3-1)
L
(P0 − PL )B 2
  x 2 
vz = 1− (2B.3-2)
2µL B
In these expressions P = p + ρgh = p − ρgz.
(b) What is the ratio of the average velocity to the maximum velocity for this flow?
(c) Obtain the slit analog of the Hagen–Poiseuille equation.
(d) Draw a meaningful sketch to show why the above analysis is inapplicable if B = W .
(e) How can the result in (b) be obtained from the results of §2.5?

2
Answers: (b) hvz i/vz,max =
3
2 (P0 − PL )B 3 W ρ
(c) w =
3 µL

Solution

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.3 Page 2 of 6

Part (a)

We assume that the fluid flows in the z-direction and that its velocity varies as a function of x.

vz = vz (x)

As a result, only φxz (the z-momentum in the positive x-direction) and φzz (the z-momentum in
the positive z-direction) contribute to the momentum balance. We also assume that the pressure
varies with height.
p = p(z)

Figure 2: This is the shell over which the momentum balance is made for the flow in a slit.

Rate of z-momentum into the shell at z = 0: (W ∆x)φzz |z=0

Rate of z-momentum out of the shell at z = L: (W ∆x)φzz |z=L
Rate of z-momentum into the shell at x: (W L)φxz |x
Rate of z-momentum out of the shell at x + ∆x: (W L)φxz |x+∆x
Component of gravitational force on the shell in z-direction: (W L∆x)ρg

If we assume steady flow, then the momentum balance is

Rate of momentum in − Rate of momentum out + Force of gravity = 0.

Considering only the z-component, we have

(W ∆x)φzz |z=0 − (W ∆x)φzz |z=L + (W L)φxz |x − (W L)φxz |x+∆x + (W L∆x)ρg = 0.

Factor the left side.

W ∆x( φzz |z=0 − φzz |z=L ) − W L( φxz |x+∆x − φxz |x ) + (W L∆x)ρg = 0

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.3 Page 3 of 6

Divide both sides by the volume of the shell W L∆x.

φzz |z=0 − φzz |z=L φxz |x+∆x − φxz |x
− + ρg = 0
L ∆x
Take the limit as ∆x → 0.
φzz |z=0 − φzz |z=L φxz |x+∆x − φxz |x
− lim + ρg = 0
L ∆x→0 ∆x
The second term is the definition of the first derivative.
φzz |z=0 − φzz |z=L dφxz
− + ρg = 0
L dx
Now substitute the expressions for φxz and φzz .

φxz = τxz + 
ρv
xv
 z = τxz
2
φzz = pδzz + H
τzz
H + ρvz vz = p(z) + ρvz

Since vz does not depend on z, the ρvz2 terms cancel and we get
2  2 
p(0) +  z z=0 − p(L) − 
ρv ρv
z z=L dτxz
− + ρg = 0.
L dx
Make it so ρg is part of the fraction.
p(0) − p(L) + ρgL dτxz
− =0
L dx
Subtract ρg0 from the numerator.
p(0) − ρg0 − [p(L) − ρgL] dτxz
− =0
L dx
Substitute Pz = p(z) − ρgz.
P0 − PL dτxz
− =0
L dx
So we have
dτxz P0 − PL
= .
dx L
From Newton’s law of viscosity we know that τxz = −µ(dvz /dx), so
P0 − PL
 
d dvz
−µ = .
dx dx L
Bring −µ in front of the derivative and then divide both sides by it.
d2 vz P0 − PL
2
=−
dx µL
We assume the fluid has zero velocity at the walls (x = ±B), i.e. the no-slip boundary condition,
and that the maximum velocity occurs furthest from the walls (x = 0). That is,
dvz
B.C. 1: = 0 when x = 0
dx
B.C. 2: vz = 0 when x = ±B.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.3 Page 4 of 6

Integrate both sides of the differential equation with respect to x.

dvz P0 − PL
=− x + C1
dx µL
Apply the first boundary condition.
dvz
= C1 = 0
dx x=0
Integrate both sides of the differential equation with respect to x once more.
P0 − PL 2
vz (x) = − x + C2
2µL
Apply the second boundary condition.
P0 − PL 2 P0 − PL 2
vz (±B) = − B + C2 = 0 → C2 = B
2µL 2µL
With the constants determined, we know the velocity profile.
P0 − PL 2 P0 − PL 2
vz (x) = − x + B
2µL 2µL
P0 − PL 2
= (B − x2 )
2µL
Therefore,
P0 − PL 2
  x 2 
vz (x) = B 1−
2µL B
and
dvz P0 − PL
τxz = −µ =− x.
dx L
Part (b)

As mentioned before, the maximum velocity occurs when x = 0.

P0 − PL 2
vz (0) = vz,max = B
2µL
Now we will find the average velocity. It is obtained by integrating vz (x) over the area the fluid
flows perpendicular to and then dividing by that area.
ˆ
1
hvz i = vz dA
A
ˆ B
P0 − PL 2
  x 2 
1
= B 1− (W dx)
W (2B) −B 2µL B
ˆ 
B P0 − PL B  x 2 
= 1− dx
4µ L −B B
B
B P0 − PL x3

= x−
4µ L 3B 2 −B
B P0 − PL 4
 
= B
4µ L 3
P0 − PL 2
= B
3µL

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.3 Page 5 of 6

Therefore, the ratio of the average velocity to the maximum velocity is

P0 −PL 2 1
hvz i 3µL B 3 2
= P0 −PL 2
= 1 = .
vz,max 2µL B 2
3

Part (c)

Here we will find the rate of mass flow w in the slit. Assume that fluid density ρ is constant.

dm d(ρV ) dV
w= = =ρ
dt dt dt
The volumetric flow rate dV /dt is equal to average velocity times cross-sectional area.

= ρhvz i(2BW )
P0 − PL 2
=ρ B (2BW )
3µL
Therefore,
2 (P0 − PL )B 3 W ρ
w= .
3 µL
Part (d)

Figure 3: This figure illustrates what happens when B = W .

When W = B hardly any part of the flow is not signficantly affected by the edges. The
assumption that the velocity only flows in the z-direction and varies in the x-direction is no
longer reasonable.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.3 Page 6 of 6

Part (e)

§2.5 analyzes flow in a slit lying horizontally for two immiscible fluids with different viscosities (µI
and µII ), so gravity does not appear in the equations there as it does here. The main results from
that section for fluid I are the velocity profile,

(p0 − pL )b2 2µI

  I
µ − µII  x   x 2
  
I
vz = + − , (2.5-18)
2µI L µI + µII µI + µII b b

and the average velocity,

(p0 − pL )b2 7µI + µII
 
hvzI i = . (2.5-20)
12µI L µI + µII
Set p0 − pL = P0 − P1 and set µI = µII = µ to obtain the corresponding equations for one fluid
moving vertically downward in a slit as a result of a pressure difference and gravity. Also, in §2.5
b is used instead of B, so set b = B.

(P0 − PL )B 2
  x 2 
I
vz = 1−
2µL B
(P0 − PL )B 2
hvzI i =
3µL
Set x = 0 to obtain the maximum velocity.

I (P0 − PL )B 2
vz,max = vzI (x = 0) =
2µL

The ratio of the average velocity to the maximum velocity is the same as that obtained in part (b).
(P0 −PL )B 2 1
hvzI i 3µL 3 2
I
= (P0 −PL )B 2
= 1 =
vz,max 2
3
2µL

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.4 Page 1 of 5

Problem 2B.4
Laminar slit flow with a moving wall (“plane Couette flow”). Extend Problem 2B.3 by
allowing the wall at x = B to move in the positive z direction at a steady speed v0 . Obtain (a)
the shear-stress distribution and (b) the velocity distribution. Draw carefully labeled sketches of
these functions.
P0 − PL (P0 − PL )B 2
    x 2  v 
µv0 0 x
Answers: τxz = x− ; vz = 1− + 1+
L 2B 2µL B 2 B

Solution

The analysis in this problem is the same as in Problem 2B.3 except that the boundary conditions
are different. Here the velocity at x = B is equal to v0 , and the velocity at x = −B is equal to 0.

Figure 1: This is Fig. 2B.3 with the wall at x = B moving in the positive z-direction with speed
v0 . Fluid is flowing through the slit with B  W  L.

We assume that the fluid flows in the z-direction and that its velocity varies as a function of x.

vz = vz (x)

As a result, only φxz (the z-momentum in the positive x-direction) and φzz (the z-momentum in
the positive z-direction) contribute to the momentum balance. Also, the boundary conditions are
as follows.

B.C. 1: vz = 0 when x = −B
B.C. 2: vz = v0 when x = B.

We also assume the pressure varies with height.

p = p(z)

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.4 Page 2 of 5

Figure 2: This is the shell over which the momentum balance is made for the flow in a slit.

Rate of z-momentum into the shell at z = 0: (W ∆x)φzz |z=0

Rate of z-momentum out of the shell at z = L: (W ∆x)φzz |z=L
Rate of z-momentum into the shell at x: (W L)φxz |x
Rate of z-momentum out of the shell at x + ∆x: (W L)φxz |x+∆x
Component of gravitational force on the shell in z-direction: (W L∆x)ρg
If we assume steady flow, then the momentum balance is
Rate of momentum in − Rate of momentum out + Force of gravity = 0.
Considering only the z-component, we have
(W ∆x)φzz |z=0 − (W ∆x)φzz |z=L + (W L)φxz |x − (W L)φxz |x+∆x + (W L∆x)ρg = 0.
Factor the left side.
W ∆x( φzz |z=0 − φzz |z=L ) − W L( φxz |x+∆x − φxz |x ) + (W L∆x)ρg = 0
Divide both sides by the volume of the shell W L∆x.
φzz |z=0 − φzz |z=L φxz |x+∆x − φxz |x
− + ρg = 0
L ∆x
Take the limit as ∆x → 0.
φzz |z=0 − φzz |z=L φxz |x+∆x − φxz |x
− lim + ρg = 0
L ∆x→0 ∆x
The second term is the definition of the first derivative.
φzz |z=0 − φzz |z=L dφxz
− + ρg = 0
L dx

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.4 Page 3 of 5

Now substitute the expressions for φxz and φzz .

φxz = τxz + 
ρv
xv
 z = τxz
2
φzz = pδzz + H
τzz
H + ρvz vz = p(z) + ρvz

Since vz does not depend on z, the ρvz2 terms cancel and we get
2  2 
p(0) +  z z=0 − p(L) − 
ρv ρv
z z=L dτxz
− + ρg = 0.
L dx
Make it so ρg is part of the fraction.

p(0) − p(L) + ρgL dτxz

− =0
L dx
Subtract ρg0 from the numerator.

p(0) − ρg0 − [p(L) − ρgL] dτxz

− =0
L dx
Substitute Pz = p(z) − ρgz.
P0 − PL dτxz
− =0
L dx
So we have
dτxz P0 − PL
= .
dx L
From Newton’s law of viscosity we know that τxz = −µ(dvz /dx), so

P0 − PL
 
d dvz
−µ = .
dx dx L

Bring −µ in front of the derivative and then divide both sides by it.

d2 vz P0 − PL
2
=−
dx µL
Integrate both sides of the differential equation with respect to x.
dvz P0 − PL
=− x + C1
dx µL
Integrate both sides of the differential equation with respect to x once more.
P0 − PL 2
vz (x) = − x + C1 x + C2
2µL
Apply the boundary conditions now to determine C1 and C2 .
P0 − PL 2
vz (−B) = − B − C1 B + C2 = 0
2µL
P0 − PL 2
vz (B) = − B + C1 B + C2 = v0
2µL

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.4 Page 4 of 5

Solving this system of equations for C1 and C2 gives

v0
C1 =
2B
v0 P0 − PL 2
C2 = + B ,
2 2µL
so we get
P0 − PL 2 v0 v0 P0 − PL 2
vz (x) = − x + x+ + B .
2µL 2B 2 2µL
Factor this result.
P0 − PL 2 v0  x
vz (x) = (B − x2 ) + 1+
2µL 2 B
P0 − PL 2 2
 
x v0  x
= B 1− 2 + 1+
2µL B 2 B

Therefore,
(P0 − PL )B 2
  x 2  v 
0 x
vz = 1− + 1+ .
2µL B 2 B

Since τxz = −µ(dvz /dx), we have

P0 − PL
 
v0
τxz = −µ − x+ .
µL 2B

Therefore,
P0 − PL µv0
τxz = x− .
L 2B
The maximum velocity can be obtained by taking the derivative of vz (x) and setting it equal to
zero.
dvz P0 − PL v0
=− x+ =0
dx µL 2B
Solving for x yields the x-coordinate where the maximum occurs.
µLv0
x=
2B(P0 − PL )

Plugging this into vz (x) and simplifying, we get

[2B 2 (P0 − PL ) + µLv0 ]2

 
µLv0
vz =
2B(P0 − PL ) 8B 2 µL(P0 − PL )

for the maximum velocity.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.4 Page 5 of 5

Figure 3: This is a carefully labeled sketch of the velocity distribution vz (x) and the shear-stress
(momentum-flux) distribution τxz (x) for flow in a slit with one wall moving at speed v0 .

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.5 Page 1 of 2

Problem 2B.5
Interrelation of slit and annulus formulas. When an annulus is very thin, it may, to a good
approximation, be considered as a thin slit. Then the results of Problem 2B.3 can be taken over
with suitable modifications. For example, the mass rate of flow in an annulus with outer wall of
radius R and inner wall of radius (1 − ε)R, where ε is small, may be obtained from Problem 2B.3
by replacing 2B by εR, and W by 2π(1 − 12 ε)R. In this way we get for the mass rate of flow:

π(P0 − PL )R4 ε3 ρ
 
1
w= 1− ε (2B.5-1)
6µL 2

Show that this same result may be obtained from Eq. 2.4-17 by setting κ equal to 1 − ε
everywhere in the formula and then expanding the expression for w in powers of ε. This requires
using the Taylor series (see §C.2)
1 1 1
ln (1 − ε) = −ε − ε2 − ε3 − ε4 − · · · (2B.5-2)
2 3 4
and then performing a long division. The first term in the resulting series will be Eq. 2B.5-1.
Caution: In the derivation it is necessary to use the first four terms of the Taylor series in Eq.
2B.5-2.

Solution

Eq. 2.4-17 in the text,

π(P0 − PL )R4 ρ (1 − κ2 )2
 
4
w= (1 − κ ) − , (2.4-17)
8µL ln(1/κ)

gives the mass flow rate w in an annulus. Set κ = 1 − ε as instructed.

π(P0 − PL )R4 ρ [1 − (1 − ε)2 ]2

 
4
w= [1 − (1 − ε) ] −
8µL ln[1/(1 − ε)]

Expand the first term in curly braces and the numerator of the fraction. Also, use the minus sign
in front of the fraction to invert the logarithm’s argument.

π(P0 − PL )R4 ρ 4ε2 − 4ε3 + ε4

 
2 3 4
w= 4ε − 6ε + 4ε − ε +
8µL ln(1 − ε)

Substitute the Taylor series expansion for ln(1 − ε), making sure to use the first four terms.
!
π(P0 − PL )R4 ρ 2 3 4 4ε2 − 4ε3 + ε4
w= 4ε − 6ε + 4ε − ε +
8µL −ε − 21 ε2 − 13 ε3 − 14 ε4 − · · ·

Factor out the minus sign in the denominator.

!
π(P0 − PL )R4 ρ 4ε2 − 4ε3 + ε4
w= 4ε − 6ε2 + 4ε3 − ε4 −
8µL ε + 12 ε2 + 13 ε3 + 14 ε4 + · · ·

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.5 Page 2 of 2

Proceed with the long division.

8 1
4ε − 6ε2 + ε3 − ε4 + · · ·
 3 3
1 2 1 3 1 4
ε + ε + ε + ε + ··· 4ε2 − 4ε3 + ε4 + 0ε5
2 3 4
4
(−) 4ε2 + 2ε3 + ε4 + ε5 + · · ·
3
−6ε3 − 13 ε4 − ε5 − · · ·
(−) − 6ε3 − 3ε4 − 2ε5 − · · ·
8 4
3ε + ε5 + · · ·
8 4 4 5
(−) ε + ε + ···
3 3
1 5
−3ε − · · ·

The first four terms of the quotient will suffice.

π(P0 − PL )R4 ρ
  
2 3 4 2 8 3 1 4
w= 4ε − 6ε + 4ε − ε − 4ε − 6ε + ε − ε + · · ·
8µL 3 3
π(P0 − PL )R ρ4
 
2 3 4 2 8 3 1 4
= 4ε
 − 6ε
Z + 4ε − ε − 4ε

 + 6ε
Z − ε + ε − · · ·
8µL 3 3
Z Z

π(P0 − PL )R ρ 4 3 2 4
4
 
= ε − ε − ···
8µL 3 3
π(P0 − PL )R ρ 4 3
4
 
1
= ε 1 − ε − ···
8µL 3 2
π(P0 − PL )R ε ρ
4 3
 
1
= 1 − ε − ···
6µL 2

Keeping only the first term in the series, we therefore have Eq. 2B.5-1.

π(P0 − PL )R4 ε3 ρ
 
1
w= 1− ε
6µL 2

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.6 Page 1 of 6

Problem 2B.6
Flow of a film on the outside of a circular tube (see Fig. 2B.6). In a gas absorption
experiment a viscous fluid flows upward through a small circular tube and then downward in
laminar flow on the outside. Set up a momentum balance over a shell of thickness ∆r in the film,
as shown in Fig. 2B.6. Note that the “momentum in” and “momentum out” arrows are always
taken in the positive coordinate direction, even though in this problem the momentum is flowing
through the cylindrical surfaces in the negative r direction.

Figure 1: Fig. 2B.6 in the text. Velocity distribution and z-momentum balance for the flow of a
falling film on the outside of a circular tube.

(a) Show that the velocity distribution in the falling film (neglecting end effects) is

ρgR2
  r 2  r 
vz = 1− + 2a2 ln (2B.6-1)
4µ R R

(b) Obtain an expression for the mass rate of flow in the film.

(c) Show that the result in (b) simplifies to Eq. 2.2-21 if the film thickness is very small.

Solution

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.6 Page 2 of 6

Part (a)

We assume that the fluid flows only in the z-direction and that its velocity varies with radius r.

vz = vz (r)

As a result, only φrz (the z-momentum in the positive r-direction) and φzz (the z-momentum in
the positive z-direction) contribute to the momentum balance. The fluid is assumed to fall
because of gravity and not because of a pressure difference.

Figure 2: This is the shell over which the momentum balance is made for flow down the exterior
of a cylindrical tube.

Rate of z-momentum into the shell at z: (2πr∆r)φzz |z

Rate of z-momentum out of the shell at z + L: (2πr∆r)φzz |z+L
Rate of z-momentum into the shell at r: (2πrL)φrz |r
Rate of z-momentum out of the shell at r + ∆r: [2π(r + ∆r)L]φrz |r+∆r
Component of gravitational force on the shell in z-direction: (2πr∆rL)ρg

If we assume steady flow, then the momentum balance is

Rate of momentum in − Rate of momentum out + Force of gravity = 0.

Considering only the z-component, we have

(2πr∆r)φzz |z − (2πr∆r)φzz |z+L + (2πrL)φrz |r − [2π(r + ∆r)L]φrz |r+∆r + (2πr∆rL)ρg = 0.

Factor the left side.

−2πr∆r( φzz |z+L − φzz |z ) − 2πL[ (r + ∆r)φrz |r+∆r − rφrz |r ] + 2πr∆rLρg = 0

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.6 Page 3 of 6

Divide both sides by 2π∆rL.

φzz |z+L − φzz |z (r + ∆r)φrz |r+∆r − rφrz |r

−r − + ρgr = 0
L ∆r
Take the limit as ∆r → 0.
φzz |z+L − φzz |z (r + ∆r)φrz |r+∆r − rφrz |r
−r − lim + ρgr = 0
L ∆r→0 ∆r
The second term is the definition of the first derivative of rφrz .

φzz |z+L − φzz |z d

−r − (rφrz ) + ρgr = 0
L dr
Now substitute the expressions for φrz and φzz .

φrz = τrz + 
ρv
rv
 z = τrz

φzz =   τzz + ρvz vz = ρv 2

pδ
zz + HH z

Since vz does not depend on z, the ρvz2 terms cancel.

2  2
z z+L − 
ρv ρv
z z d
−r −

(rτrz ) + ρgr = 0
L dr
So we have
d
(rτrz ) = ρgr.
dr
From Newton’s law of viscosity we know that τrz = −µ(dvz /dr), so
 
d dvz
−µr = ρgr.
dr dr

At the wall r = R, we assume the no-slip boundary condition, and at the gas-liquid interface
r = aR, we assume the free-surface boundary condition.
dvz
B.C. 1: = 0 when r = aR
dr
B.C. 2: vz = 0 when r = R

Integrate both sides of the differential equation with respect to r.

dvz ρg 2
−µr = r + C1
dr 2
Apply the first boundary condition now to determine C1 .

ρg ρga2 R2
0= (aR)2 + C1 → C1 = −
2 2
Divide both sides by −µr to solve for dvz /dr.

dvz ρg C1
=− r−
dr 2µ µr

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.6 Page 4 of 6

Integrate both sides of the differential equation with respect to r once more.
ρg 2 C1
vz (r) = − r − ln r + C2
4µ µ
Apply the second boundary condition now to determine C2 .
ρg 2 C1 ρg 2 C1
vz (R) = − R − ln R + C2 = 0 → C2 = R + ln R
4µ µ 4µ µ
ρg 2 ρga2 R2
= R − ln R
4µ 2µ
With the constants of integration in hand, the velocity distribution is known.
ρg 2 ρga2 R2 ρg 2 ρga2 R2
vz (r) = − r + ln r + R − ln R
4µ 2µ 4µ 2µ
ρgR2 r2
 
2 2
= − 2 + 2a ln r + 1 − 2a ln R
4µ R
2
 
ρgR  r 2
= 1− + 2a2 (ln r − ln R)
4µ R
Therefore,
ρgR2
  r 2  r 
2
vz = 1− + 2a ln .
4µ R R
Part (b)

The mass flow rate w is, assuming constant density ρ,

dm d(ρV ) dV
w= = =ρ .
dt dt dt
The volumetric flow rate dV /dt is average velocity times cross-sectional area.

w = ρhvz iA
The average velocity is obtained by integrating the velocity over the area the fluid is flowing
through and then dividing by that area.
 ˆ 
1

= ρ
vz dA A

A
ˆ aR
=ρ vz (2πr dr)
R
ˆ aR
= 2πρ rvz dr
R
ˆ aR
ρgR2
  r 2  r 
2
= 2πρ r 1− + 2a ln dr
R 4µ R R
Make a substitution to solve the integral.
r
u= → r = Ru
R
dr
du = → R du = dr
R

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.6 Page 5 of 6

The mass flow rate becomes

ˆ a
ρgR2
w = 2πρ (Ru)(1 − u2 + 2a2 ln u)(R du)
1 4µ
ˆ
πρ2 gR4 a
= (u − u3 + 2a2 u ln u) du
2µ 1
 a
πρ2 gR4 u2 u4
 2
u2

2 u
= − + 2a − + ln u
2µ 2 4 4 2
2 4
 2 4
 2 2
 12
14
 2
12

πρ gR a a a a 1 1
= − + 2a2 − + ln a − + − 2a2 − + ln 1
2µ 2 4 4 2 2 4 4 2
2 4
 2 4 4 2

πρ gR a a a 1 a
= − − + a4 ln a − +
2µ 2 4 2 4 2
2 4
 
πρ gR 3 1
= a4 ln a − a4 + a2 −
2µ 4 4

Therefore,
πρ2 gR4
w= (4a4 ln a − 3a4 + 4a2 − 1).
8µ
Part (c)

If the film thickness is very small, then we can say that a is only slightly above 1. That is, we can
say a = 1 + ε, where 0 < ε  1. Substitute this into the final answer for w.

πρ2 gR4
w= [4(1 + ε)4 ln(1 + ε) − 3(1 + ε)4 + 4(1 + ε)2 − 1]
8µ

The Taylor series expansion for ln(1 + ε) is as follows.

ε2 ε3 ε4
ln(1 + ε) = ε − + − + ···
2 3 4
Substitute this into the equation for w.

πρ2 gR4 ε 2 ε3 ε4
   
4 4 2
w= 4(1 + ε) ε − + − + · · · − 3(1 + ε) + 4(1 + ε) − 1
8µ 2 3 4

Expand the other terms.

πρ2 gR4 ε 2 ε3 ε4
   
2 3 4 2 3 4
= (4 + 16ε + 24ε + 16ε + 4ε ) ε − + − + · · · − 4ε − 14ε − 12ε − 3ε
8µ 2 3 4

Multiply the two terms in parentheses together. Only powers up to ε4 in the resulting series are
needed.
πρ2 gR4
       
4 2 16 4 3 24 16
= 4ε + 16 − ε + 24 − + ε + 16 − + − 1 ε4 + · · · − 4ε − 14ε2 − 12ε3 − 3ε4
8µ 2 2 3 2 3
2 4
 
πρ gR 2 52 3 25 4 2 3 4
= 4ε
 + 14ε
H+ ε + ε + ··· −  − 14ε
4ε H − 12ε − 3ε
H H
 H H
8µ 3 3

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.6 Page 6 of 6

We’re left with

πρ2 gR4 16 3 16 4
 
w= ε + ε + ···
8µ 3 3
2 4
πρ gR 16 3
= ε (1 + ε + · · · )
8µ 3
2πρ2 gε3 R4
= (1 + ε + · · · )
3µ
2πρ2 gε3 R4 2πρ2 gε3 R4
= +ε + ··· .
3µ 3µ
Keep only the first term of this series for the flow rate.

2πρ2 gε3 R4
w≈
3µ
Comparing this with Eq. 2.2-21 on page 46,

ρ2 gW δ 3 cos β
w= , (2.2-21)
3µ
we see that they are equivalent if W = 2πR and δ = εR. Note that since the fluid here is flowing
straight down, β = 0 and cos β = 1.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.7 Page 1 of 7

Problem 2B.7
Annular flow with inner cylinder moving axially (see Fig. 2B.7). A cylindrical rod of
radius κR moves axially with velocity vz = v0 along the axis of a cylindrical cavity of radius R as
seen in the figure. The pressure at both ends of the cavity is the same, so that the fluid moves
through the annular region solely because of the rod motion.

(a) Find the velocity distribution in the narrow annular region.

(b) Find the mass rate of flow through the annular region.
(c) Obtain the viscous force acting on the rod over the length L.
(d) Show that the result in (c) can be written as a “plane slit” formula multiplied by a
“curvature correction.” Problems of this kind arise in studying the performance of
wire-coating dies.1

vz ln(r/R)
Answers: (a) =
v0 ln κ
πR2 v0 ρ (1 − κ2 )
 
2
(b) w = − 2κ
2 ln(1/κ)
(c) Fz = −2πLµv0 / ln(1/κ)
 
−2πLµv0 1 1
(d) Fz = 1 − ε − ε2 + · · · where ε = 1 − κ (see Problem 2B.5)
ε 2 12

Solution

Unlike the previous problem, the fluid here is flowing horizontally, so gravity will not influence its
velocity. If we assume no-slip boundary conditions, then the fluid velocity at the cylindrical rod
r = κR is v0 and the fluid velocity at the outer wall r = R is 0.

Part (a)

Choose a cylindrical coordinate system with the positive z-axis pointing to the right, the direction
the inner cylinder is moving in. Then the fluid flows in the z-direction and varies as a function of
radius from the cylinder’s axis.
vz = vz (r)
1
J. B. Paton, P. H. Squires, W. H. Darnell, F. M. Cash, and J. F. Carley, Processing of Thermoplastic Materials,
E. C. Bernhardt (ed.), Reinhold, New York (1959), Chapter 4.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.7 Page 2 of 7

As a result, the boundary conditions are written as follows.

B.C. 1: vz = v0 when r = κR
B.C. 2: vz = 0 when r = R

Gravity is pointing down, so we can say that the pressure does not depend on z.

p 6= p(z)

Because vz = vz (r), only φrz (the z-momentum in the positive r-direction) and φzz (the
z-momentum in the positive z-direction) contribute to the momentum balance.

Figure 1: This is the shell over which the momentum balance is made for flow in a horizontally
oriented annulus.

Rate of z-momentum into the shell at z: (2πr∆r)φzz |z

Rate of z-momentum out of the shell at z + L: (2πr∆r)φzz |z+L
Rate of z-momentum into the shell at r: (2πrL)φrz |r
Rate of z-momentum out of the shell at r + ∆r: [2π(r + ∆r)L]φrz |r+∆r
Component of gravitational force on the shell in z-direction: 0

If we assume steady flow, then the momentum balance is

Rate of momentum in − Rate of momentum out + Force of gravity = 0.

Considering only the z-component, we have

(2πr∆r)φzz |z − (2πr∆r)φzz |z+L + (2πrL)φrz |r − [2π(r + ∆r)L]φrz |r+∆r = 0.

Factor the left side.

−2πr∆r( φzz |z+L − φzz |z ) − 2πL[ (r + ∆r)φrz |r+∆r − rφrz |r ] = 0

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.7 Page 3 of 7

Divide both sides by 2π∆rL.

φzz |z+L − φzz |z (r + ∆r)φrz |r+∆r − rφrz |r

−r − =0
L ∆r
Take the limit as ∆r → 0.
φzz |z+L − φzz |z (r + ∆r)φrz |r+∆r − rφrz |r
−r − lim =0
L ∆r→0 ∆r
The second term is the definition of the first derivative of rφrz .

φzz |z+L − φzz |z d

−r − (rφrz ) = 0
L dr
Now substitute the expressions for φrz and φzz .

φrz = τrz + 
ρv
rv
 z = τrz
2
φzz = pδzz + H
τzz
H + ρvz vz = p + ρvz

Since p and vz do not depend on z, the p and ρvz2 terms cancel.

2 H2
p|
z+L + ρvz X X− z − ρvH
p|
 XX
z+L zHz d
−r  − (rτrz ) = 0
L dr
So we have
d
(rτrz ) = 0.
dr
From Newton’s law of viscosity we know that τrz = −µ(dvz /dr), so
 
d dvz
−µr = 0.
dr dr

Integrate both sides of the differential equation with respect to r.

dvz
−µr = C1
dr
Divide both sides by −µr.
dvz C1
=−
dr µr
Integrate both sides of the differential equation with respect to r once more.
C1
vz (r) = − ln r + C2
µ
Apply the boundary conditions now to determine C1 and C2 .
C1
vz (κR) = − ln(κR) + C2 = v0
µ
C1
vz (R) = − ln R + C2 = 0
µ

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.7 Page 4 of 7

Solving the system of equations, we get

µ
C1 = − v0
ln κ
ln R
C2 = − v0 .
ln κ
Armed with the constants of integration, the velocity distribution is known.
1 ln R
vz (r) = v0 ln r − v0
ln κ ln κ
v0
= (ln r − ln R)
ln κ
v0 r
= ln
ln κ R
Therefore,
vz ln(r/R)
= .
v0 ln κ
Part (b)

The mass flow rate w is, assuming constant density ρ,

dm d(ρV ) dV
w= = =ρ .
dt dt dt
The volumetric flow rate dV /dt is average velocity times cross-sectional area.
w = ρhvz iA
The average velocity is obtained by integrating the velocity over the area the fluid is flowing
through and then dividing by that area.
 ˆ 
1
= ρ

vz dA A

A
ˆ R
=ρ vz (2πr dr)
κR
ˆ R
= 2πρ rvz dr
κR
ˆ R
v0 r
= 2πρ r ln dr
κR ln κ R
Make a substitution to solve the integral.
r
u= → r = Ru
R
dr
du = → R du = dr
R
The mass flow rate becomes
ˆ 1
v0
w = 2πρ (Ru) ln u(R du)
κ ln κ
ˆ
2πρv0 R2 1
= u ln u du
ln κ κ
1
2πρv0 R2
 2
u2

u
= − + ln u
ln κ 4 2 κ

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.7 Page 5 of 7

2πρv0 R2
 2
12   κ2 κ2

1
w= − + ln 1 + + ln κ
ln κ 4 2 4 2
2πρv0 R2 κ2 − 1 κ2
 
= + ln κ
ln κ 4 2
2
 2 
2πρv0 R κ −1 2
= + 2κ
4 ln κ
πρv0 R2 1 − κ2
 
2
= + 2κ
2 − ln κ

Therefore,
πR2 v0 ρ (1 − κ2 )
 
2
w= − 2κ .
2 ln(1/κ)
Part (c)

The viscous stress τrz physically represents the force in the z-direction on a unit area
perpendicular to the r-direction. By evaluating τrz at r = κR and multiplying it by the surface
area of the inner cylinder, we obtain the viscous force acting on it over its length. The final point
to note is that because the fluid is acting from a larger radius r on the inner cylinder, which has a
smaller radius κR, we place a minus sign in front of τrz .

Fz = −τrz |r=κR · 2π(κR)L

From Newton’s law of viscosity, we have

dvz
τrz = −µ
dr
d  v0 r
= −µ ln
dr ln κ R
v0 R 1
= −µ ·
ln κ r R
µv0 1
=− · .
ln κ r
Now we can find the force.
 
µv0 1
Fz = − − · · 2π(κR)L
ln κ r r=κR
2πκRLµv
 0 1
= · 
ln κ κR

2πLµv0
=
ln κ
2πLµv0
=−
− ln κ
Therefore,
2πLµv0
Fz = − .
ln(1/κ)
The minus sign here makes sense because the viscous (frictional) force of the fluid opposes the
motion of the cylinder, which is moving in the positive z-direction.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.7 Page 6 of 7

Part (d)

To make the annulus a plane slit, we let the radius of the inner cylinder tend towards the radius
of the outer cylinder; that is, κ is very slightly less than 1.

κ = 1 − ε, where 0 < ε  1

Substitute this into the formula for Fz .

2πLµv0
Fz =
ln κ
2πLµv0
=
ln(1 − ε)

The Taylor series expansion for ln(1 − ε) is as follows.

ε2 ε3
ln(1 − ε) = −ε − − − ···
2 3
Substitute this formula into Fz .
2πLµv0
Fz = 2 3
−ε − ε2 − ε3 − · · ·
2πLµv0 1
=− 2
ε 1 + 2 + ε3 + · · ·
ε

Use long division to obtain a series for the fraction.

ε ε2
1− − − ···
2 12
ε ε2

1+ + + ··· 1 + 0ε + 0ε2
2 3
ε ε2
(−) 1 + + + ···
2 3
ε2
− 2ε − 3 − ···
ε ε2
(−) − − − 2ε5 − · · ·
2 4
ε2
− 12 − ···

Therefore,
ε2
 
−2πLµv0 ε
Fz = 1− − − ··· .
ε 2 12
The viscous stress τxz for laminar slit flow with a wall moving at speed v0 was obtained in
Problem 2B.4.
P0 − PL
 
µv0
τxz = x−
L 2B

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.7 Page 7 of 7

If we orient the slit horizontally so that pressure and gravity do not influence the fluid flow as is
the case in this problem, then τxz simplifies.

Figure 2: This is Couette flow in a horizontal slit with one moving wall at x = B.

µv0
τxz = −
2B
The viscous force acting on the moving wall is Fz = +τxz |x=B · W L. There is no minus sign here
because the fluid acting on the wall has a lower x-coordinate than the moving wall at x = B.
 µv 
0
Fz = − · WL
2B x=B
W Lµv0
=−
2B
Comparing this formula with the boxed result, we see that they are equivalent if ε = 2B, the slit
width, and W = 2π. W should be a distance, but it’s only an angle; thus, the fraction in the
boxed result is not truly a plane slit formula. The remaining series in parentheses then must be a
curvature correction.
ε2
 
2πLµv0 ε
Fz = − 1− − − ···
| {zε } | 2 12
{z }
“plane slit” formula curvature correction

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.8 Page 1 of 5

Problem 2B.8
Analysis of a capillary flowmeter (see Fig. 2B.8).

Determine the rate of flow (in lbm /hr) through the capillary flow meter shown in the figure. The
fluid flowing in the inclined tube is water at 20°C, and the manometer fluid is carbon
tetrachloride (CCl4 ) with density 1.594 g/cm3 . The capillary diameter is 0.010 in. Note:
Measurements of H and L are sufficient to calculate the flow rate; θ need not be measured. Why?

Solution

For the capillary flow meter, choose a cylindrical coordinate system with the positive z-direction
pointing in the direction of the flow. The fluid velocity is assumed to vary as a function of radius
r.
vz = vz (r)
As a result, only φrz (the z-momentum in the positive r-direction) and φzz (the z-momentum in
the positive z-direction) contribute to the momentum balance. Figure 1 on the next page shows
the shell the momentum balance is made over.

Rate of z-momentum into the shell at z = 0: (2πr∆r)φzz |z=0

Rate of z-momentum out of the shell at z = L: (2πr∆r)φzz |z=L
Rate of z-momentum into the shell at r: (2πrL)φrz |r
Rate of z-momentum out of the shell at r + ∆r: [2π(r + ∆r)L]φrz |r+∆r
Component of gravitational force on the shell in z-direction: (2πr∆rL)ρg sin θ
If we assume steady flow, then the momentum balance is
Rate of momentum in − Rate of momentum out + Force of gravity = 0.
Considering only the z-component, we have
(2πr∆r)φzz |z=0 − (2πr∆r)φzz |z=L + (2πrL)φrz |r − [2π(r + ∆r)L]φrz |r+∆r + (2πr∆rL)ρg sin θ = 0.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.8 Page 2 of 5

Figure 1: This is the shell over which the momentum balance is made for fluid going through the
capillary flow meter.

Factor the left side.

−2πr∆r( φzz |z=L − φzz |z=0 ) − 2πL[ (r + ∆r)φrz |r+∆r − rφrz |r ] + 2πr∆rLρg sin θ = 0

Divide both sides by 2π∆rL.

φzz |z=L − φzz |z=0 (r + ∆r)φrz |r+∆r − rφrz |r

−r − + ρgr sin θ = 0
L ∆r
Take the limit as ∆r → 0.
φzz |z=L − φzz |z=0 (r + ∆r)φrz |r+∆r − rφrz |r
−r − lim + ρgr sin θ = 0
L ∆r→0 ∆r
The second term is the definition of the first derivative of rφrz .

φzz |z=L − φzz |z=0 d

−r − (rφrz ) + ρgr sin θ = 0
L dr
Now substitute the expressions for φrz and φzz .

φrz = τrz + 
ρv
rv
 z = τrz
2
φzz = pδzz + H
τzz
H + ρvz vz = p + ρvz

Since vz does not depend on z, the ρvz2 terms cancel.

2  2 
p|z=L +  z z=L − p|z=0 − 
ρv ρv
z z=0 d
−r − (rτrz ) + ρgr sin θ = 0
L dr
Make it so that ρgr sin θ is in the fraction.

p|z=L − p|z=0 − ρgL sin θ d

−r − (rτrz ) = 0
L dr

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.8 Page 3 of 5

From the schematic in Fig. 2B.8, we see that h = L sin θ. It is here where θ disappears from the
equation; hence, it does not need to be measured.

p|z=L − p|z=0 − ρgh d

−r − (rτrz ) = 0
L dr
So we have
d ρgh + ( p|z=0 − p|z=L )
(rτrz ) = r.
dr L
It is thanks to the manometer underneath the capillary that the principles of fluid statics can be
applied to compute the quantity in the numerator. Let

ρ = the density of water

ρC = the density of CCl4
H 0 = the distance between points B and C in Fig. 2B.8.

D and E are at the same height in the CCl4 ; thus, the pressures at these levels must be equal.

p| + ρgH 0 + ρC gH = p|z=0 + ρgh + ρgH 0 + ρgH

| z=L {z } | {z }
pressure at D pressure at E

The ρgH 0 terms cancel, so the distance between points B and C is not needed. Solve this
equation for the quantity in the numerator

ρgh + p|z=0 − p|z=L = (ρC − ρ)gH

and substitute it into the differential equation.

d (ρC − ρ)gH
(rτrz ) = r
dr L
From Newton’s law of viscosity we know that τrz = −µ(dvz /dr), so
 
d dvz (ρC − ρ)gH
−µr = r.
dr dr L

We thus have a differential equation for the velocity distribution in the capillary. The boundary
conditions for it are obtained from the assumptions that the velocity is maximum furthest from
the wall (at r = 0) and that no slipping occurs between the fluid and the wall (at r = R).

dvz
B.C. 1 : = 0 when r = 0
dr
B.C. 2 : vz = 0 when r = R

Integrate both sides of the differential equation with respect to r.

dvz (ρC − ρ)gH 2

−µr = r + C1
dr 2L
Apply the first boundary condition now to determine C1 .

dvz (ρC − ρ)gH 2

−µ(0) = (0) + C1 → 0 = C1
dr r=0 2L

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.8 Page 4 of 5

Divide both sides by −µr.

dvz (ρC − ρ)gH
=− r
dr 2µL
Integrate both sides of the differential equation with respect to r once more.

(ρC − ρ)gH 2
vz (r) = − r + C2
4µL
Apply the second boundary condition now to determine C2 .

(ρC − ρ)gH 2 (ρC − ρ)gH 2

vz (R) = − R + C2 = 0 → C2 = R
4µL 4µL
With the constants of integration in hand, the velocity distribution is known.

(ρC − ρ)gH 2 (ρC − ρ)gH 2

vz (r) = − r + R
4µL 4µL
(ρC − ρ)gH 2
= (R − r2 )
4µL
This result can be used to obtain the mass rate of flow w.
dm d(ρV ) dV
w= = =ρ
dt dt dt
The volumetric flow dV /dt is equal to the average velocity in the capillary times its
cross-sectional area.
w = ρhvz i · πR2
The average velocity is found by integrating vz over the area of the cross-section and then
dividing by that area.
 ˆ 
1 2
w=ρ 2
vz dA ·  πR 

πR
ˆ R


=ρ vz (2πr dr)
0
ˆ R
= 2πρ rvz dr
0
ˆ R
(ρC − ρ)gH 2
= 2πρ r (R − r2 ) dr
0 4µL
ˆ
πρ(ρC − ρ)gH R
= (rR2 − r3 ) dr
2µL 0
R
πρ(ρC − ρ)gH r2 R2 r4

= −
2µL 2 4 0
 4 4

πρ(ρC − ρ)gH R R
= −
2µL 2 4
πρ(ρC − ρ)gHR 4
=
8µL

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.8 Page 5 of 5

Use Eq. 1.1-3, ν = µ/ρ, to write the density of water ρ in terms of the kinematic viscosity ν and
viscosity µ. Also, write the radius in terms of the diameter.
µ
D 4


π ρC − ν gH 2
w=
8νL
π (ρC ν − µ) gHD4
=
128ν 2 L
Before we plug in the numbers, convert the units so that the desired units of lbm /hr are obtained.
µ and ν for water at 20°C are given on page 14 in Table 1.1-2. Other conversion factors are on
page 868 and 870.
 3
g
2.54 H
cm 1@kg
@ × 2.2046 lbm ≈ 0.0575863 lbm
ρC = 1.594 H 3 × ×
H
cmH 1 in 1000 g
1@kg
@ in3
2
cm2 in2
 
 1 in 3600 As
ν = 0.010037 × × ≈ 5.60066
As 2.54 
cm 1 hr hr
lbm
2.4191 1@
ft lbm
· s × ft·hr
µ = 1.0019 
mPa × ≈ 0.201975
 A
1 ·s
mPa
 
12 in in · hr
 2
m 3.28 @
ft 12 in 3600 As in
g = 9.81 2 × × × ≈ 5.00414 × 109

s@
@ 1
m
 1@
ft 1 hr hr2
H = 1.0 in
D = 0.010 in
L = 120 in

Therefore,
lbm ·in5
π(0.0576 · 5.6 − 0.202)(5 × 109 )(1.0)(0.010)4 hr3
w≈
in5
128(5.6)2 (120) hr2
lbm
≈ 3.9 × 10−5 .
hr

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.9 Page 1 of 6

Problem 2B.9
Low-density phenomena in compressible tube flow2,3 (see Fig. 2B.9). As the pressure is
decreased in the system studied in Example 2.3-2, deviations from Eqs. 2.3-28 and 2.3-29 arise.
The gas behaves as if it slips at the tube wall. It is conventional2 to replace the customary
“no-slip” boundary condition that vz = 0 at the tube wall by
dvz
vz = −ζ , at r = R (2B.9-1)
dr
in which ζ is the slip coefficient. Repeat the derivation in Example 2.3-2 using Eq. 2B.9-1 as the
boundary condition. Also make use of the experimental fact that the slip coefficient varies
inversely with the pressure ζ = ζ0 /p, in which ζ0 is a constant. Show that the mass rate of flow is

π(p0 − pL )R4 ρavg

 
4ζ0
w= 1+ (2B.9-2)
8µL Rpavg

in which pavg = 12 (p0 + pL ) and ρavg is the average density calculated at pavg .
When the pressure is decreased further, a flow regime is reached in which the mean free path
of the gas molecules is large with respect to the tube radius (Knudsen flow ). In that regime3
r   
2m 4 3 p0 − pL
w= πR (2B.9-3)
πkT 3 L

in which m is the molecular mass and k is the Boltzmann constant. In the derivation of this
result it is assumed that all collisions of the molecules with the solid surfaces are diffuse and not
specular. The results in Eqs. 2.3-29, 2B.9-2, and 2B.9-3 are summarized in Fig. 2B.9.

Solution

2
E. H. Kennard, Kinetic Theory of Gases, McGraw-Hill, New York (1938), pp. 292-295, 300-306.
3
M. Knudsen, The Kinetic Theory of Gases, Methuen, London, 3rd edition (1950). See also R. J. Silbey and R.
A. Alberty, Physical Chemistry, Wiley, New York, 3rd edition (2001), §17.6.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.9 Page 2 of 6

This problem wants us to consider a different boundary condition than the typical no-slip one in
§2.3. Up until the boundary conditions are applied, the analysis will be the same. We assume
that the fluid flows only in the z-direction and that its velocity varies with radius r.

vz = vz (r)

As a result, only φrz (the z-momentum in the positive r-direction) and φzz (the z-momentum in
the positive z-direction) contribute to the momentum balance. The pressure is assumed to vary
with height z.
p = p(z)

Figure 1: This is the shell over which the momentum balance is made for flow through a cylindrical
tube oriented vertically.

Rate of z-momentum into the shell at z = 0: (2πr∆r)φzz |z=0

Rate of z-momentum out of the shell at z = L: (2πr∆r)φzz |z=L
Rate of z-momentum into the shell at r: (2πrL)φrz |r
Rate of z-momentum out of the shell at r + ∆r: [2π(r + ∆r)L]φrz |r+∆r
Component of gravitational force on the shell in z-direction: (2πr∆rL)ρg

If we assume steady flow, then the momentum balance is

Rate of momentum in − Rate of momentum out + Force of gravity = 0.

Considering only the z-component, we have

(2πr∆r)φzz |z=0 − (2πr∆r)φzz |z=L + (2πrL)φrz |r − [2π(r + ∆r)L]φrz |r+∆r + (2πr∆rL)ρg = 0.

Factor the left side.

−2πr∆r( φzz |z=L − φzz |z=0 ) − 2πL[ (r + ∆r)φrz |r+∆r − rφrz |r ] + 2πr∆rLρg = 0

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.9 Page 3 of 6

Divide both sides by 2π∆rL.

φzz |z=L − φzz |z=0 (r + ∆r)φrz |r+∆r − rφrz |r

−r − + ρgr = 0
L ∆r
Take the limit as ∆r → 0.
φzz |z=L − φzz |z=0 (r + ∆r)φrz |r+∆r − rφrz |r
−r − lim + ρgr = 0
L ∆r→0 ∆r
The second term is the definition of the first derivative of rφrz .
φzz |z=L − φzz |z=0 d
−r − (rφrz ) + ρgr = 0
L dr
Now substitute the expressions for φrz and φzz .

φrz = τrz + 
ρv
rv
 z = τrz
2
φzz = pδzz + H
τzz
H + ρvz vz = p(z) + ρvz

Since vz does not depend on z, the ρvz2 terms cancel.

2 X2
p|z=L + X
ρvX
zX X − p|z=0 − ρvX
z=L zXz=0 d
−r − (rτrz ) + ρgr = 0
X
L dr
Make it so that ρgr is in the fraction.
p|z=L − p|z=0 − ρgL d
−r − (rτrz ) = 0
L dr
Place ρg0 in the numerator.
( p|z=L − ρgL) − ( p|z=0 − ρg0) d
−r − (rτrz ) = 0
L dr
The point of doing this is that now we can use the modified pressure Pz = p(z) − ρgz.
PL − P0 d
−r − (rτrz ) = 0
L dr
So we have
d P0 − PL
(rτrz ) = r.
dr L
From Newton’s law of viscosity we know that τrz = −µ(dvz /dr), so

P0 − PL
 
d dvz
−µr = r.
dr dr L
One boundary condition is obtained from the assumption that the velocity is maximum furthest
from the wall (at r = 0). The second boundary condition is provided in the problem statement at
the wall r = R.
dvz
B.C. 1 : = 0, at r = 0
dr
dvz
B.C. 2 : vz = −ζ , at r = R
dr

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.9 Page 4 of 6

Integrate both sides of the differential equation with respect to r.

dvz P0 − PL 2
−µr = r + C1
dr 2L
Apply the first boundary condition here to determine C1 .
dvz P0 − PL 2
−µ(0) = (0) + C1 → 0 = C1
dr r=0 2L

Divide both sides by −µr to solve for dvz /dr.

dvz P0 − PL
=− r
dr 2µL
Integrate both sides of the differential equation with respect to r once more.
P0 − PL 2
vz (r) = − r + C2
4µL
Apply the second boundary condition here to determine C2 .
dvz
vz (R) = −ζ
dr r=R
P0 − PL 2 P0 − PL P0 − PL
− R + C2 = ζ R → C2 = R(R + 2ζ)
4µL 2µL 4µL
With the constants of integration in hand, the velocity distribution is known.
P0 − PL 2 P0 − PL
vz (r) = − r + R(R + 2ζ)
4µL 4µL
P0 − PL
= [R(R + 2ζ) − r2 ]
4µL
The mass flow rate w is, assuming constant density ρ,
dm d(ρV ) dV
w= = =ρ .
dt dt dt
The volumetric flow rate dV /dt is average velocity times cross-sectional area.

w = ρhvz i · πR2

The average velocity is obtained by integrating the velocity over the area the fluid is flowing
through and then dividing by that area.
 ˆ 
1 2
=ρ 2
vz dA · 
πR

πR
ˆ R
=ρ vz (2πr dr)
0
ˆ R
= 2πρ rvz dr
0
ˆ R
P0 − PL
= 2πρ r [R(R + 2ζ) − r2 ] dr
0 4µL

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.9 Page 5 of 6

ˆ
π(P0 − PL )ρ R
w= [R(R + 2ζ)r − r3 ] dr
2µL 0
R
π(P0 − PL )ρ r2 r4

= R(R + 2ζ) −
2µL 2 4 0
π(P0 − PL )ρ 2 R4
 
R
= R(R + 2ζ) −
2µL 2 4
π(P0 − PL )ρ 4
 
R
= R3 ζ +
2µL 4
So we have for the mass flow rate,
PL − P0
 
πρ
w= − (R4 + 4R3 ζ).
8µ L
In the case of compressible flow (flow of a gas), the weight can be assumed to be negligible.
 
πρ pL − p0
w= − (R4 + 4R3 ζ)
8µ L
Take the limit as the length of the cylinder goes to 0.
 
πρ pL − p0
w= − lim (R4 + 4R3 ζ)
8µ L→0 L
The term in parentheses is the first derivative of p with respect to z.
 
πρ dp
w= − (R4 + 4R3 ζ)
8µ dz
The density ρ can be written in terms of the pressure p using the ideal gas law,

pV = nRg T,

where Rg represents the gas constant.

m
pV = Rg T,
N
where m is the mass and N is the molar mass (a constant). Divide both sides by V .
ρ
p= Rg T
N
Since the flow is assumed to be isothermal, the temperature T is also constant. This equation
holds for every value of z, so
ρ
At z : p= Rg T
N
ρ0
At z = 0 : p0 = Rg T.
N
Divide the first equation by the second one to get a relationship between the pressure and density
at z = 0 and the pressure and density at z. This can be solved for ρ in terms of p.
p ρ ρ0
= → ρ= p
p0 ρ0 p0

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.9 Page 6 of 6

Substitute this into the equation for w.

 
πρ0 dp
w= −p (R4 + 4R3 ζ)
8µp0 dz

Make use of the experimental fact that the slip coefficient varies inversely with pressure ζ = ζ0 /p,
where ζ0 is a constant.   
πρ0 dp ζ0
w= −p R4 + 4R3
8µp0 dz p
Solve this differential equation by separation of variables.
 
πρ0 4 3 ζ0
w dz = (−p dp) R + 4R
8µp0 p

Integrate both sides. ˆ ˆ

L pL  
πρ0 4 3 ζ0
w dz = (−p dp) R + 4R
0 p0 8µp0 p
The mass flow rate is assumed to be constant at every height z.
ˆ L ˆ pL
πρ0
w dz = − (R4 p + 4R3 ζ0 ) dp
0 p0 8µp0

Proceed with the integration.

 2
 pL
πρ0 4p 3
wL = − R + 4R ζ0 p
8µp0 2 p0

Divide both sides by L to solve for w.

p2 p2
 
πρ0
w=− R4 L + 4R3 ζ0 pL − R4 0 − 4R3 ζ0 p0
8µLp0 2 2
 4 
πρ0 R 2 2 3
=− (p − p0 ) + 4R ζ0 (pL − p0 )
8µLp0 2 L
 4 
πρ0 R 3
=− (pL − p0 )(pL + p0 ) + 4R ζ0 (pL − p0 )
8µLp0 2
πρ0  4
R (pL − p0 )pavg + 4R3 ζ0 (pL − p0 )

=−
8µLp0
π(pL − p0 )pavg ρ0 R4
 
4ζ0
=− 1+
8µLp0 Rpavg

The equation relating the pressure to density also holds for the averages.
p ρ ρ0 ρ0
= → ρ= p → ρavg = pavg
p0 ρ0 p0 p0
Therefore,
π(p0 − pL )ρavg R4
 
4ζ0
w= 1+ ,
8µL Rpavg
where pavg = 12 (p0 + pL ) and ρavg is the average density calculated at pavg .

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.10 Page 1 of 3

Problem 2B.10
Incompressible flow in a slightly tapered tube. An incompressible fluid flows through a
tube of circular cross section, for which the tube radius changes linearly from R0 at the tube
entrance to a slightly smaller value RL at the tube exit. Assume that the Hagen-Poiseuille
equation is approximately valid over a differential length, dz, of the tube so that the mass flow
rate is
π[R(z)]4 ρ
 
dP
w= − (2B.10-1)
8µ dz
This is a differential equation for P as a function of z, but, when the explicit expression for R(z)
is inserted, it is not easily solved.
(a) Write down the expression for R as a function of z.
(b) Change the independent variable in Eq. 2B.10-1 to R, so that it becomes
πR4 ρ
  
dP RL − R0
w= − (2B.10-2)
8µ dR L

(c) Integrate the equation, and then show that the solution can be rearranged to give
π(P0 − PL )R04 ρ 1 + (RL /R0 ) + (RL /R0 )2 − 3(RL /R0 )3
 
w= 1− (2B.10-3)
8µL 1 + (RL /R0 ) + (RL /R0 )2
Interpret the result. The approximation used here that a flow between nonparallel surfaces
can be regarded locally as flow between parallel surfaces is sometimes referred to as the
lubrication approximation and is widely used in the theory of lubrication. By making a
careful order-of-magnitude analysis, it can be shown that, for this problem, the lubrication
approximation is valid as long as4
 !
RL 2

RL
1− 1 (2B.10-4)
R0 R0

Solution

Part (a)

R needs to be a linear function of z that is equal to R0 at z = 0 and RL at z = L. It has the

general form,
R(z) = mz + b.
Apply the conditions to determine the slope m and slope-intercept b.

At z = 0 : R(0) = m · 0 + b = R0
At z = L : R(L) = m · L + b = RL

Solve this system of equations for m and b.

RL − R0
m= and b = R0
L
4
R. B. Bird, R. C. Armstrong, and O. Hassager, Dynamics of Polymeric Liquids, Vol. 1, Wiley-Interscience, New
York, 2nd edition (1987), pp. 16-18.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.10 Page 2 of 3

Therefore,
RL − R0
R(z) = z + R0 .
L
Part (b)

Now that R(z) is determined, the chain rule can be used to write the derivative of P with respect
to R.
dP dP dR
=
dz dR dz
Take the first derivative of R(z) and plug it in for dR/dz.
 
dP dP RL − R0
=
dz dR L
Substitute this result into Eq. 2B.10-1.

π[R(z)]4 ρ
 
dP
w= −
8µ dz
4
  
πR ρ dP RL − R0
= −
8µ dR L
Therefore,
πR4 ρ
  
dP RL − R0
w= − .
8µ dR L
Part (c)

Solve the differential equation with the method of separation of variables.

 
dR πρ RL − R0
w = (−dP)
R4 8µ L
Integrate both sides.
ˆ RL ˆ PL  
dR πρ RL − R0
w= (−dP)
R0 R4 P0 8µ L
Proceed with the integration.
RL  
w 1 πρ RL − R0
− = − (PL − P0 )
3 R3
R0 8µ L
   
w 1 1 πρ RL − R0
− 3 = (P0 − PL )
3 R03 RL 8µ L
 3 3   
w RL − R0 πρ RL − R0
= (P0 − PL )
3 R03 RL
3 8µ L
  2 2   
L − R0 )(RL + R0 RL + R0 ) RL−

w  (R  πρ R
0
3 3 = (P0 − PL ) 
3 R0 RL 8µ L

Multiply both sides by R04 .

2 + R R + R2 ) 
π(P0 − PL )R04 ρ

w R0 (RL 0 L 0
3 =
3 RL 8µL

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.10 Page 3 of 3

Expand the term in square brackets.

R02 R03 π(P0 − PL )R04 ρ

 
w R0
+ 2 + 3 =
3 RL RL RL 8µL

Solve this equation for w now.

 
π(P0 − PL )R04 ρ 3
w= 
R02 R03

8µL R0
+ +
RL 2
RL 3
RL
 
π(P0 − PL )R04 ρ 3
=  2

R03

8µL RL RL
3
RL R02
+ R0 +1
R3
 
π(P0 − PL )R04 ρ  3 RL3
0
= RL2

8µL RL
+ R0 + 1
R02
 2
RL 2
RL R3

RL RL
π(P0 − PL )R04 ρ  1 + R0 + R02 − 1 − R0 − R02
+ 3 RL3
0
= R2

8µL 1 + RL + L R0 R02
 2
RL 2
RL R3

RL RL
π(P0 − PL )R04 ρ 1+ R0 + R02
−1 − R0 − R02
+ 3 RL3
0
=  2
RL
+ 2
RL

8µL 1+ RL
+ 1+ RL
+
R0 R02 R0 R02
R2 R3
 
π(P0 − PL )R04 ρ  1+ R
R0 + R02 − 3 R03
L L L

= 1− 2
RL

8µL 1+ RL
+ R0 R02

Therefore,

π(P0 − PL )R04 ρ 1 + (RL /R0 ) + (RL /R0 )2 − 3(RL /R0 )3

 
w= 1− .
8µL 1 + (RL /R0 ) + (RL /R0 )2

In the event that R0 = RL , there is no taper in the tube, and the formula for w simplifies to the
Hagen-Poiseuille equation. If R0  RL , then the mass flow rate tends to 0. This doesn’t make
sense, though, because as long as RL is big enough, something should flow through the tube exit
regardless of how big R0 is. In the problem statement the assumption was made that RL was only
slightly less than R0 , so R0  RL is not a case we need to consider.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.11 Page 1 of 5

Problem 2B.11
The cone-and-plate viscometer (see Fig. 2B.11). A cone-and-plate viscometer consists of a
flat plate and an inverted cone, whose apex just contacts the plate. The liquid whose viscosity is
to be measured is placed in the gap between the cone and plate. The cone is rotated at a known
angular velocity Ω, and the torque Tz required to turn the cone is measured. Find an expression
for the viscosity of the fluid in terms of Ω, Tz , and the angle ψ0 between the cone and the plate.
For commercial instruments ψ0 is about 1 degree.

(a) Assume that locally the velocity distribution in the gap can be very closely approximated
by that for flow between parallel plates, the upper one moving with a constant speed. Verify
that this leads to the approximate velocity distribution (in spherical coordinates)
 
vφ (π/2) − θ
=Ω (2B.11-1)
r ψ0
This approximation should be rather good, because ψ0 is so small.
(b) From the velocity distribution in Eq. 2B.11-1 and Appendix B.1, show that a reasonable
expression for the shear stress is
τθφ = µ(Ω/ψ0 ) (2B.11-2)
This result shows that the shear stress is uniform throughout the gap. It is this fact that
makes the cone-and-plate viscometer quite attractive. The instrument is widely used,
particularly in the polymer industry.
(c) Show that the torque required to turn the cone is given by
2
Tz = πµΩR3 /ψ0 (2B.11-3)
3
This is the standard formula for calculating the viscosity from measurements of the torque
and angular velocity for a cone-plate assembly with known R and ψ0 .
(d) For a cone-and-plate instrument with radius 10 cm and angle ψ0 equal to 0.5 degree, what
torque (in dyn · cm) is required to turn the cone at an angular velocity of 10 radians per
minute if the fluid viscosity is 100 cp?

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.11 Page 2 of 5

Answer: (d) 40, 000 dyn · cm

Solution

Part (a)

Because the linear velocity of the cone’s top varies with r (V = Ωr), we can treat each slice of the
circle with circumference 2πr and thickness ∆r as a linear element with length L = 2πr and
thickness ∆r moving at speed V . The problem essentially boils down to Couette flow, where a

fluid flows in a slit and one wall moves while the other remains stationary. This is illustrated in
part (c) of Fig. 2B.11. Thus, a rectangular shell will be considered to determine the velocity
distribution. We assume that the fluid flows in the x-direction and that its velocity varies in the
y-direction.
vx = vx (y)
Consequently, only φyx (the x-momentum in the positive y-direction) and φxx (the x-momentum
in the positive x-direction) contribute to the momentum balance. We also assume that the flow is
only due to the spinning top and is not a result of gravity or a pressure difference.

Figure 1: This is the shell over which the momentum balance is made for the flow in a slit.

Rate of x-momentum into the shell at x = 0: (∆r∆y)φxx |x=0

Rate of x-momentum into the shell at x = L: (∆r∆y)φxx |x=L
Rate of x-momentum into the shell at y: (L∆r)φyx |y
Rate of x-momentum into the shell at y + ∆y: (L∆r)φyx |y+∆y
Component of gravitational force on the shell in x-direction: 0

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.11 Page 3 of 5

If we assume steady flow, then the momentum balance is

Rate of momentum in − Rate of momentum out + Force of gravity = 0.

Considering only the x-component, we have

(∆r∆y)φxx |x=0 − (∆r∆y)φxx |x=L + (L∆r)φyx |y − (L∆r)φyx |y+∆y = 0.

Factor the left side.

−∆r∆y( φxx |x=L − φxx |x=0 ) − L∆r( φyx |y+∆y − φyx |y ) = 0

Divide both sides by L∆r∆y.

φxx |x=L − φxx |x=0 φyx |y+∆y − φyx |y

− − =0
L ∆y
Take the limit as ∆y → 0.

φxx |x=L − φxx |x=0 φyx |y+∆y − φyx |y

− − lim =0
L ∆y→0 ∆y
The second term is the first derivative of φyx with respect to y.

φxx |x=L − φxx |x=0 dφyx

− − =0
L dy
Now substitute the expressions for φyx and φxx .

φyx = τyx + 
ρv
y v
x = τyx

φxx =   τxx + ρvx vx = ρv 2

pδ
xx + HH x

Since vx does not depend on x, the ρvx2 terms cancel and we get
| − φxx
φxx | dτyx
− x=L  x=0 − =0
L dy
or
dτyx
= 0.
dy
From Newton’s law of viscosity, τyx = −µ(dvx /dy), so we have
 
d dvx
−µ = 0.
dy dy
Viscosity is assumed to be constant, so we can pull it and the minus sign in front of the
derivative. Divide both sides by −µ.
d2 vx
=0
dy 2
Using the coordinate system in part (c) of Fig. 2B.11, the boundary conditions are as follows.

B.C. 1: vx = 0 when y = 0
B.C. 2: vx = V when y = b

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.11 Page 4 of 5

Integrate the differential equation with respect to y.

dvx
= C1
dy
Integrate the differential equation with respect to y once more.

vx (y) = C1 y + C2

Apply the two boundary conditions now.

vx (0) = C1 (0) + C2 = 0
vx (b) = C1 (b) + C2 = V

Solving this system of equations, we get C1 = V /b and C2 = 0. Hence, we have for the velocity
distribution
V
vx (y) = y.
b
Now we can change back to rotational flow in spherical coordinates; that is, the fluid actually
flows in the φ direction and its velocity varies with r [vφ = vφ (r)]. We have V = Ωr,
b = r sin ψ0 ≈ rψ0 , and y ≈ r(π/2 − θ).

Ωr
 π 
vφ = r −θ
rψ

0 2
Therefore,  
vφ (π/2) − θ
=Ω .
r ψ0
Part (b)

The shear stress on a surface element of the cone by the fluid is given by τθφ (the θ-direction is
perpendicular to the surface element, and the shear is in the φ-direction). According to Appendix
B.1 on page 844,  
sin θ ∂  vφ  1 ∂vθ
τθφ = −µ + .
r ∂θ sin θ r sin θ ∂φ
Since the velocity is in the φ-direction only, the second term in the square brackets is zero. Also,
θ is a constant on the cone’s surface (θ = θ0 ), so the formula for τθφ simplifies considerably.
 
sin θ0 ∂ vφ
τθφ = −µ
r ∂θ sin θ0
1 ∂vφ
= −µ
r ∂θ 
π
2 −θ

µ ∂
=− Ωr
r ∂θ ψ0
  
µ 1
=− Ωr −
r ψ0

Therefore,
Ω
τθφ = µ .
ψ0

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.11 Page 5 of 5

Part (c)

Since the shearing force is in the φ-direction, it is perpendicular to the radial direction. The
torque then is just the product of this force with the moment arm.
Tz = rF
To get the shearing force, we multiply (integrate) the shear stress by the area of the cone that the
fluid is in contact with. ˆ
Tz = r · τθφ |θ= π −ψ0 dA
2
| {z }
force
Since the fluid is at a lower value of θ than the cone’s surface at θ = π2 − ψ0 , no minus sign is
needed in front of τθφ . Looking at part (b) in Fig. 2B.11, we see that the area differential is
dA = r dr dφ.
ˆ 2π ˆ R
Tz = r · τθφ |θ= π −ψ0 r dr dφ
0 0 2
ˆ 2π ˆ R  
Ω
= r2 µ dr dφ
0 0 ψ0
ˆ 2π  ˆ R 
Ω 2
=µ dφ r dr
ψ0 0 0
 3
Ω R
= µ (2π)
ψ0 3
Therefore,
2 R3
Tz = πµΩ .
3 ψ0
Part (d)

We have the following values for the variables. The conversion factor for Pa to dyn/cm2 is found
in Table F.3-2 on page 869.
R = 10 cm
π rad π
ψ0 = 0.5° × = rad
180° 360
rad 1min
 1 rad
Ω = 10  × =
min
 60 s 6 s
−3 dyn
10 PaH H· s 10 cm 2 dyn
µ = 100 
cp
 × × =1 2
·s
1cp
 1 Pa
HH cm
Plugging the numbers into the formula for torque Tz , we get
2 R3
Tz = πµΩ
3 ψ0
H (10 cm)3
  H 
2 dyn 1 rad
= π 1 · s π H
3 cm2
6
s 360 rad
H
dyn
= 40, 000 · cm3 = 40, 000 dyn · cm.
cm2

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.12 Page 1 of 3

Problem 2B.12
Flow of a fluid in a network of tubes (see Fig. 2B.12). A fluid is flowing in laminar flow
from A to B through a network of tubes, as depicted in the figure. Obtain an expression for the
mass flow rate w of the fluid entering at A (or leaving at B) as a function of the modified
pressure drop PA − PB . Neglect the disturbances at the various tube junctions.

3π(PA − PB )R4 ρ
Answer: w =
20µL

Solution

The main idea with this problem is the law of conservation of mass; that is, mass is neither
created nor destroyed. Each of the corners in this network can be referred to as a node. The rate
of mass flowing into a node must be equal to the rate of mass flowing out of it, assuming no mass
accumulates there. Let wA be the rate of mass flowing into node A and let wB be the rate of
mass flowing out of node B. In addition, let wij be the rate of mass flowing from node i to node
j. As each tube is a cylinder with radius R and length L, the Hagen-Poiseuille equation can be
used for the mass flow rate,
π(Pi − Pj )R4 ρ
wij = ,
8µL
where Pi is the modified pressure at node i and Pj is the modified pressure at node j.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.12 Page 2 of 3

Figure 1: The network with rates of mass flow wij labeled for each tube. The assumed direction
that the mass is flowing in each tube is shown as well.

Applying the law of conservation of mass to each node, we obtain the following eight equations.

At node A : wA = wA1 + wA2 + wA3

At node 1 : wA1 = w14 + w15
At node 2 : wA2 = w24 + w26
At node 3 : wA3 = w35 + w36
At node 4 : w14 + w24 = w4B
At node 5 : w15 + w35 = w5B
At node 6 : w26 + w36 = w6B
At node B : w4B + w5B + w6B = wB

Substitute the Hagen-Poiseuille equation into the equations from nodes 1-6.

π(PA − P1 )R4 ρ π(P1 − P4 )R4 ρ π(P1 − P5 )R4 ρ

wA1 = w14 + w15 → = +
8µL 8µL 8µL
π(PA − P2 )R ρ
4 π(P2 − P4 )R ρ π(P2 − P6 )R4 ρ
4
wA2 = w24 + w26 → = +
8µL 8µL 8µL
π(PA − P3 )R ρ
4 π(P3 − P5 )R ρ π(P3 − P6 )R4 ρ
4
wA3 = w35 + w36 → = +
8µL 8µL 8µL
π(P1 − P4 )R ρ π(P2 − P4 )R ρ
4 4 π(P4 − PB )R4 ρ
w14 + w24 = w4B → + =
8µL 8µL 8µL
π(P1 − P5 )R ρ π(P3 − P5 )R ρ
4 4 π(P5 − PB )R4 ρ
w15 + w35 = w5B → + =
8µL 8µL 8µL
π(P2 − P6 )R ρ π(P3 − P6 )R ρ
4 4 π(P6 − PB )R4 ρ
w26 + w36 = w6B → + =
8µL 8µL 8µL

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2B.12 Page 3 of 3

Multiply both sides of these equations by 8µL/πR4 ρ.

(PA − P1 ) = (P1 − P4 ) + (P1 − P5 )

(PA − P2 ) = (P2 − P4 ) + (P2 − P6 )
(PA − P3 ) = (P3 − P5 ) + (P3 − P6 )
(P1 − P4 ) + (P2 − P4 ) = (P4 − PB )
(P1 − P5 ) + (P3 − P5 ) = (P5 − PB )
(P2 − P6 ) + (P3 − P6 ) = (P6 − PB )

Solve this system of equations for the modified pressure at each numerical node in terms of the
modified pressures at A and B.
1
P1 = P2 = P3 = (3PA + 2PB )
5
1
P4 = P5 = P6 = (2PA + 3PB )
5
Now that these are known, the mass flow rate entering at A can be determined. Use the equation
from node A to do this.

wA = wA1 + wA2 + wA3

π(PA − P1 )R4 ρ π(PA − P2 )R4 ρ π(PA − P3 )R4 ρ
= + +
8µL 8µL 8µL
4
πR ρ
= (3PA − P1 − P2 − P3 )
8µL
πR4 ρ
 
1 1 1
= 3PA − (3PA + 2PB ) − (3PA + 2PB ) − (3PA + 2PB )
8µL 5 5 5
4
 
πR ρ 3
= 3PA − (3PA + 2PB )
8µL 5
4
 
πR ρ 9 6
= 3PA − PA − PB
8µL 5 5
4
 
πR ρ 6 6
= PA − PB
8µL 5 5
4
πR ρ 6
= · (PA − PB )
8µL 5
Therefore, the mass flow rate of fluid entering at A is

3π(PA − PB )R4 ρ
w= .
20µL
The same result is obtained for wB , the rate of mass flowing out of B.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.1 Page 1 of 7

Problem 2C.1
Performance of an electric dust collector (see Fig. 2C.1)5 .

Figure 1: This is Fig. 2C.1 in the text. A possible particle trajectory. The particle that begins at
z = 0, x = B and ends at x = −B may not necessarily travel the longest distance in the z direction.

(a) A dust precipitator consists of a pair of oppositely charged plates between which dust-laden
gases flow. It is desired to establish a criterion for the minimum length of the precipitator in
terms of the charge on the particle e, the electric field strength E , the pressure difference
(p0 − pL ), the particle mass m, and the gas viscosity µ. That is, for what length L will the
smallest particle present (mass m) reach the bottom just before it has a chance to be swept
out of the channel? Assume that the flow between the plates is laminar so that the velocity
distribution is described by Eq. 2B.3-2. Assume also that the particle velocity in the z
direction is the same as the fluid velocity in the z direction. Assume further that the Stokes
drag on the sphere as well as the gravity force acting on the sphere as it is accelerated in the
negative x direction can be neglected.

(b) Rework the problem neglecting acceleration in the x direction, but including the Stokes
drag.

(c) Compare the usefulness of the solutions in (a) and (b), considering that stable aerosol
particles have effective diameters of about 1-10 microns and densities of about 1 g/cm3 .

Answer: (a) Lmin = [12(p0 − pL )2 B 5 m/25µ2 eE ]1/4

Solution

5
The answer given in the first edition of this book was incorrect, as pointed out to us in 1970 by Nau Gab Lee of
Seoul National University.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.1 Page 2 of 7

Part (a)

Consider a particle with mass m and charge e in the slit. If there’s an electric field in the
x-direction, then the particle will accelerate in that direction until it hits the lower wall. Our aim
is to find out how long it takes for this to happen. Once this time is known, we will use the fluid
velocity in the z-direction to determine how far it travels in the z-direction. Essentially this is a
kinematics problem. A charge e in an electric field E experiences a force eE . Since the charge is
moving from positive x to negative x, the electric field is negative. Apply Newton’s second law in
the x-direction to determine the particle’s acceleration in this direction.
X
Fx = e(−E ) = max

Acceleration is the second derivative of position with respect to time.

d2 x
m = −eE
dt2
Divide both sides by m.
d2 x eE
=−
dt2 m
Initially the particle is assumed to be at rest, so one of the initial conditions is x0 (0) = 0. The
second one is trickier because as mentioned in the problem statement, a particle that starts at
z = 0, x = B won’t necessarily travel the furthest distance in the z-direction. What we will do
then is assume an arbitrary initial position x(0) = x0 and then in the end choose the value of x0
that maximizes L. Integrate the differential equation with respect to t.
dx eE
= − t + C1
dt m
Apply the first initial condition to determine C1 .
eE
x0 (0) = − (0) + C1 = 0 → C1 = 0
m
Integrate the differential equation with respect to t once more.
eE 2
x(t) = − t + C2
2m
Apply the second initial condition to determine C2 .
eE
x(0) = − (0)2 + C2 = x0 → C2 = x 0
2m
So we have
eE 2
x(t) = x0 −t . (1)
2m
Plug in x = −B and solve for t to find how long it takes for the particle to go from its initial
position to the lower wall at x = −B.
r
eE 2 eE 2 2m
−B = x0 − t → t = x0 + B → t = (x0 + B)
2m 2m eE

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.1 Page 3 of 7

Now that we know how long the particle is moving for, we can determine how far it moves in the
z-direction. For a fluid in laminar flow, the formula for the velocity in a vertical slit is given by
Eq. 2B.3-2 on page 63.

(P0 − PL )B 2
  x 2 
vz (x) = 1− (2B.3-2)
2µL B
P0 − PL 2
= (B − x2 )
2µL
Since the slit we’re dealing with here is extended horizontally, gravity does not affect the motion
of the fluid, so the modified pressure P reduces to the pressure p.
p0 − pL 2
vz (x) = (B − x2 )
2µL
Substitute equation (1) for x to get vz as a function of t.
"  2 #
p0 − pL eE
vz (t) = B 2 − x0 − t2
2µL 2m

Expand the terms inside the square brackets.

e2 E 2 4
 
p0 − pL 2 2 eE 2
vz (t) = B − x0 + x0 t − t
2µL m 4m2
To get the distance traveled in the z-direction, integrate the velocity with respect to time. No
constant of integration is necessary because the particle starts from z = 0.
ˆ
z(t) = vz dt

By choosing the lower limit of integration to be when the particle starts moving (at t = 0) and
the upper limit of integration to be when the particle hits the wall at x = −B and stops moving,
we obtain L, the distance in the z-direction that the particle travels.
ˆ
q
2m
eE
(x0 +B)
L= vz dt
0
ˆ
q
2m
(x0 +B)
e2 E 2 4
 
eE p0 − pL 2 2 eE 2
= B − x0 + x0 t − t dt
0 2µL m 4m2
ˆ 2m (x0 +B) ˆ 2m (x0 +B) 2E 2 ˆ
 q q q 
2m
(x0 +B)
p0 − pL  2 eE eE eE e eE
= (B − x20 ) dt + x0 t2 dt − 2
t4 dt
2µL 0 m 0 4m 0
 #5 
" # 3 "r
e E
r r
p0 − pL  2m eE 2m 2 2 2m 
= (B 2 − x20 ) (x0 + B) + x0 (x0 + B) − (x 0 + B)
2µL  eE 3m eE 20m2 eE 
( )
2 r
e2 E 2 2m
  
p0 − pL 2 2 eE 2m 2m
= (B − x0 ) + x0 (x0 + B) − 2
(x0 + B) (x0 + B)
2µL 3m eE 20m eE eE
 r
p0 − pL 2 2 2 1 2 2m
= (B − x0 ) + x0 (x0 + B) − (x0 + B) (x0 + B)
2µL 3 5 eE

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.1 Page 4 of 7

 r
p0 − pL 4 2 4 8 2m
L= B + Bx0 − x20 (x0 + B)
2µL 5 15 15 eE
r
p0 − pL 4 2m
= · (3B − 2x0 )(x0 + B) (x0 + B)
2µL 15 eE
r
p0 − pL 2 2m
= · (3B − 2x0 )(x0 + B)3/2
µL 15 eE
Multiply both sides of the equation by L.
r
2 p0 − pL 2 2m
L = · (3B − 2x0 )(x0 + B)3/2 (2)
µ 15 eE
Now we’re ready to determine the initial position x0 in the slit that maximizes L. Take the
derivative of L2 with respect to x0 and set the result equal to 0.
r
d 2 p0 − pL 2 2m 5 p
(L ) = · · (B − 2x0 ) B + x0 = 0
dx0 µ 15 eE 2
This is only satisfied if
B
B − 2x0 = 0 → x0 =
2
B + x0 = 0 → x0 = −B.

The maximum value of L is obtained if the particle’s initial position is x0 = B/2. Now plug this
into equation (2) to determine the maximum value of L2 .
r   3/2
2 p0 − pL 2 2m B B
L = · 3B − 2 +B
µ 15 eE 2 2
3B 3/2
r  
p0 − pL 2 2m
= · (2B)
µ 15 eE 2
p0 − pL m 12 √ 5
r r
= B
µ eE 25
Take the square root of both sides to solve for L.
s
p0 − pL m 12 √ 5
r r
L= B
µ eE 25

Use the fourth root to simplify the answer.

s
p0 − pL 2 m 12

4
L= · · · B5
µ eE 25

Therefore, s
4 12(p0 − pL )2 B 5 m
Lmin = .
25µ2 eE
The subscript “min” refers to the fact that this is the length for the smallest mass.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.1 Page 5 of 7

Part (b)

Here the acceleration in the x-direction will be neglected (i.e. ax set equal to 0) and the Stokes
drag (Fk = 6πµRv∞ ) will be included in the sum of the forces. Since ax = 0, the particle will be
moving at terminal velocity from its initial position x = x0 to the lower wall x = −B. The Stokes
drag acts in the direction opposing the particle’s motion, so it will be in the positive x-direction.
As in part (a), the electric field is responsible for pushing the particle down. Consider Newton’s
second law in the x-direction.
X
Fx = e(−E ) + 6πµRv∞ = max = 0

Solve for v∞ , the velocity that the fluid is approaching the particle from below.
eE
v∞ =
6πµR
The terminal velocity is the negative of this, as the particle is moving down.
eE
vt = −
6πµR
Velocity is the derivative of position with respect to time.
dx eE
=−
dt 6πµR
As mentioned in the problem statement, a particle that starts at z = 0, x = B won’t necessarily
travel the furthest distance in the z-direction. What we will do then is assume an arbitrary initial
position x(0) = x0 and then in the end choose the value of x0 that maximizes L. Integrate the
differential equation with respect to t.
eE
x(t) = − t + C3
6πµR
Apply the initial condition to determine C3 .
eE
x(0) = − (0) + C3 = x0 → C3 = x 0
6πµR
So we have
eE
x(t) = x0 − t. (3)
6πµR
Plug in x = −B and solve for t to find how long it takes for the particle to go from its initial
position to the lower wall at x = −B.
eE eE 6πµR
−B = x0 − t → t = x0 + B → t= (x0 + B)
6πµR 6πµR eE
Now that we know how long the particle is moving for, we can determine how far it moves in the
z-direction. As explained in part (a), the velocity for the flow in a horizontally extended slit is
p0 − pL 2
vz (x) = (B − x2 ).
2µL

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.1 Page 6 of 7

Substitute equation (3) for x to get vz as a function of t.

"  2 #
p0 − pL eE
vz (t) = B 2 − x0 − t
2µL 6πµR

Expand the terms inside the square brackets.

e2 E 2
 
p0 − pL 2 2 eE 2
vz (t) = B − x0 + x0 t − t
2µL 3πµR 36π 2 µ2 R2
To get the distance traveled in the z-direction, integrate the velocity with respect to time. No
constant of integration is necessary because the particle starts from z = 0.
ˆ
z(t) = vz dt

By choosing the lower limit of integration to be when the particle starts moving (at t = 0) and
the upper limit of integration to be when the particle hits the wall at x = −B and stops moving,
we obtain L, the distance in the z-direction that the particle travels.
ˆ 6πµR (x0 +B)
eE
L= vz dt
0
ˆ 6πµR
(x0 +B)
e2 E 2
 
eE p0 − pL 2 2 eE 2
= B − x0 + x0 t − t dt
0 2µL 3πµR 36π 2 µ2 R2
" ˆ 6πµR (x0 +B) ˆ 6πµR (x0 +B) 2E 2 ˆ 6πµR (x0 +B) #
p0 − pL eE eE eE e eE
= (B 2 − x20 ) dt + x0 t dt − t2 dt
2µL 0 3πµR 0 36π 2 µ2 R2 0
( 2 3 )
2E 2
   
p0 − pL 6πµR eE 6πµR e 6πµR
= (B 2 − x20 ) (x0 + B) + x0 (x0 + B) − (x0 + B)
2µL eE 6πµR eE 108π 2 µ2 R2 eE
 
p0 − pL 6πµR 2 2 6πµR 2 2πµR 3
= (B − x0 )(x0 + B) + x0 (x0 + B) − (x0 + B)
2µL eE eE eE
 
p0 − pL 6πµR 2 2 1 2
= · (B − x0 ) + x0 (x0 + B) − (x0 + B) (x0 + B)
2µL eE 3
p0 − pL 3πR 1
= · · (2B − x0 )(x0 + B)(x0 + B)
L eE 3
p0 − pL πR
= · (2B − x0 )(x0 + B)2
L eE
Multiply both sides by L.
π(p0 − pL )R
L2 = (2B − x0 )(x0 + B)2 (4)
eE
Now we’re ready to determine the initial position x0 in the slit that maximizes L. Take the
derivative of L2 with respect to x0 and set the result equal to 0.
d π(p0 − pL )R
(L2 ) = · 3(B − x0 )(B + x0 ) = 0
dx0 eE
This is only satisfied if
B − x0 = 0 → x0 = B
B + x0 = 0 → x0 = −B.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.1 Page 7 of 7

The maximum value of L is obtained if the particle’s initial position is x0 = B. Now plug this
into equation (4) to determine the maximum value of L2 .

π(p0 − pL )R
L2 = (2B − B)(B + B)2
eE
π(p0 − pL )R 3
= 4B
eE
4π(p0 − pL )B 3 R
=
eE
Therefore, r
4π(p0 − pL )B 3 R
Lmin = .
eE
Interestingly Lmin is independent of the particle’s mass and the fluid’s viscosity. Also, releasing
the particle at x = B does lead to the maximum value of L in this case because the acceleration
in the x-direction was neglected.

Part (c)

With the diameter and density of an aerosol particle, one can compute its mass by

m=ρ·V
4
= ρ · πR3
3
In SI units the mass is exceedingly small: a 10 micron diameter particle has a mass of about
5 × 10−13 kg. By making the assumption in part (a) that the Stokes drag can be neglected, one is
effectively saying that it is negligible compared to the inertial force max in Newton’s second law.
X
Fx = e(−E ) + 
6πµRv ∞ = max



Given how small the mass is, this is not justified, and I would deem the result in part (a) to be
useless. On the other hand, neglecting the acceleration in the x-direction is justified because of
the small mass, which is what we do in part (b).
X
Fx = e(−E ) + 6πµRv∞ =  ma
x

Because of the really small diameter of aerosol particles (10−5 m to 10−6 m), the Reynolds
number,
Dv∞ ρ
Re = ,
µ
will be less than 0.1 for reasonable fluids and velocities. Stokes’s law will be valid as a result, and
the answer in part (b) should give an accurate value for Lmin . Also, on page 61 it says Stokes’s law
is used in the study of the motion of aerosol particles. That tells us something about its utility.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.2 Page 1 of 3

Problem 2C.2
Residence time distribution in tube flow. Define the residence time function F (t) to be
that fraction of the fluid flowing in a conduit which flows completely through the conduit in a
time interval t. Also define the mean residence time tm by the relation
ˆ 1
tm = tdF (2C.2-1)
0

(a) An incompressible Newtonian liquid is flowing in a circular tube of length L and radius R,
and the average flow velocity is hvz i. Show that

F (t) = 0 for t ≤ (L/2hvz i) (2C.2-2)

2
F (t) = 1 − (L/2hvz it) for t ≥ (L/2hvz i) (2C.2-3)

(b) Show that tm = (L/hvz i).

Solution

Part (a)

Consider the cross-section of fluid in a circular cylinder at the entrance z = 0. The fluid at the
center flows fastest since it is furthest from the walls, so it will reach the cylinder’s exit z = L
first. F (t) will only start to increase from 0 at this point. The amount of time it takes for the
fluid at the entrance’s center to reach the exit’s center is
distance L
t= = .
speed vz,max
At the moment we can say that
(
L
0 t≤ vz,max
F (t) = L
.
To be determined t≥ vz,max

Consider now the fluid at a radius r (0 < r < R) at the entrance. By the time it arrives at the
exit, all the fluid with a smaller radius will have exited, whereas the fluid with larger radius will
still be inside the cylinder. We can get the fraction of fluid that has exited by dividing the
volumetric flow rate through a circle of radius r by the total volumetric flow rate through the
cylinder.
dV
dt r
F (r) = dV
´dtr R
vz (2πr dr)
= ´ 0R
vz (2πr dr)
´0r
rvz dr
= ´ 0R
0 rvz dr

For the flow in a cylindrical tube, the velocity is

  r 2 
vz (r) = vz,max 1 − .
R

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.2 Page 2 of 3

Figure 1: This is an illustration of fluid that has exited the cylinder. Here t > L/vz,max and
F (t) > 0.

Substitute this result for vz .

´r h
r 2

i
0 rvz,max 1 − R dr
F (r) = ´ h
i
R r 2
0 rv z,max 1− R dr

Cancel vz,max , multiply the numerator and denominator by R2 , and expand the integrands.
´r
(rR2 − r3 ) dr
= ´ 0R
(rR2 − r3 ) dr
0 2 2 r
r R r4
2 − 4 0
=  R
r 2 R2 r4
2 − 4 0
r 2 R2 r4
2 − 4
= R4
4
 r 2  r 4
=2 −
R R
Here we have F as a function of r. Now we will get it in terms of t. The amount of time it takes
for fluid to go from the entrance to the exit at a radius of r is
L
t=
vz (r)
L
t= h i
r 2


vz,max 1 − R

Solve this equation for (r/R)2 .

 r 2 L  r 2 L
1− = → =1−
R vz,max t R vz,max t

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.2 Page 3 of 3

Substitute this result into F (r) to get F (t).

   2
L L
F (t) = 2 1 − − 1−
vz,max t vz,max t
 2
L L L
=2−2  −1+2  −
vz,max t v
z,max t vz,max t

 2 
L
=1−
vz,max t
The final point to note is that for flow in a circular cylinder, vz,max = 2hvz i. Therefore,

0 t ≤ 2hvLz i
F (t) = 
L
2
L
.
1 − t ≥
2hvz it 2hvz i

Part (b)

Since F is a function of time, the chain rule can be used to rewrite the definition of tm .
ˆ 1 ˆ ∞
dF
tm = t dF = t dt
0 0 dt
When F = 0, no fluid has flown completely through the conduit, so t = 0 is the new lower limit.
Though, we could use any value for t up to (L/2hvz i), since F remains 0 up to this point. When
F = 1, all fluid has flown completely through the conduit (this happens after a really long time),
so t = ∞ is the new upper limit. The function F is defined for two separate time intervals, so the
integral will have to be split into two. First find dF/dt from F (t).

for t ≤ 2hvLz i
(
0 dF 0 for t ≤ 2hvLz i
F (t) =  2 → = L2
1 − L
2hv it for t ≥ 2hvL i dt 2hvz i2 t3
for t ≥ 2hvLz i
z z

So we have
ˆ ∞
dF
tm = t dt
0 dt
ˆ L ˆ ∞
2hvz i dF dF
= t dt + t dt
0 dt L dt
2hvz i
ˆ L ˆ ∞ 
L2

2hvz i
= t(0) dt + t dt
0 L 2hvz i2 t3
2hvz i
ˆ ∞
L2
= dt
L2hvz i2 t2
2hvz i
 ∞
L2 1
= 2
−
2hvz i t L
2hvz i

L2
 
2hvz i
= 2
.
2hvz i L
Therefore,
L
tm = .
hvz i

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.3 Page 1 of 4

Problem 2C.3
Velocity distribution in a tube. You have received a manuscript to referee for a technical
journal. The paper deals with heat transfer in tube flow. The authors state that, because they
are concerned with nonisothermal flow, they must have a “general” expression for the velocity
distribution, one that can be used even when the viscosity of the fluid is a function of
temperature (and hence position). The authors state that a “general expression for the velocity
distribution for flow in a tube” is ˆ 1
(ȳ/µ)dȳ
vz y
=ˆ 1 (2C.3-1)
hvz i 3
(ȳ /µ)dȳ
0
in which y = r/R. The authors give no derivation, nor do they give a literature citation. As the
referee you feel obliged to derive the formula and list any restrictions implied.

Solution

We assume that the fluid flows only in the z-direction and that its velocity varies with radius r.

vz = vz (r)

As a result, only φrz (the z-momentum in the positive r-direction) and φzz (the z-momentum in
the positive z-direction) contribute to the momentum balance. The pressure is assumed to vary
with height z.
p = p(z)

Figure 1: This is the shell over which the momentum balance is made for flow through a cylindrical
tube oriented vertically.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.3 Page 2 of 4

Rate of z-momentum into the shell at z = 0: (2πr∆r)φzz |z=0

Rate of z-momentum out of the shell at z = L: (2πr∆r)φzz |z=L
Rate of z-momentum into the shell at r: (2πrL)φrz |r
Rate of z-momentum out of the shell at r + ∆r: [2π(r + ∆r)L]φrz |r+∆r
Component of gravitational force on the shell in z-direction: (2πr∆rL)ρg

If we assume steady flow, then the momentum balance is

Rate of momentum in − Rate of momentum out + Force of gravity = 0.

Considering only the z-component, we have

(2πr∆r)φzz |z=0 − (2πr∆r)φzz |z=L + (2πrL)φrz |r − [2π(r + ∆r)L]φrz |r+∆r + (2πr∆rL)ρg = 0.

Factor the left side.

−2πr∆r( φzz |z=L − φzz |z=0 ) − 2πL[ (r + ∆r)φrz |r+∆r − rφrz |r ] + 2πr∆rLρg = 0

Divide both sides by 2π∆rL.

φzz |z=L − φzz |z=0 (r + ∆r)φrz |r+∆r − rφrz |r

−r − + ρgr = 0
L ∆r
Take the limit as ∆r → 0.
φzz |z=L − φzz |z=0 (r + ∆r)φrz |r+∆r − rφrz |r
−r − lim + ρgr = 0
L ∆r→0 ∆r
The second term is the definition of the first derivative of rφrz .
φzz |z=L − φzz |z=0 d
−r − (rφrz ) + ρgr = 0
L dr
Now substitute the expressions for φrz and φzz .

φrz = τrz + 
ρv
rv
 z = τrz
2
φzz = pδzz + H
τzz
H + ρvz vz = p(z) + ρvz

Since vz does not depend on z, the ρvz2 terms cancel.

2 X 2X
p|z=L + X
ρvX
zX X − p|z=0 − ρvX
z=L z z=0 d
−r − (rτrz ) + ρgr = 0
X
L dr
Make it so that ρgr is in the fraction.
p|z=L − p|z=0 − ρgL d
−r − (rτrz ) = 0
L dr
Place ρg0 in the numerator.
( p|z=L − ρgL) − ( p|z=0 − ρg0) d
−r − (rτrz ) = 0
L dr

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.3 Page 3 of 4

The point of doing this is that now we can use the modified pressure Pz = p(z) − ρgz.
PL − P0 d
−r − (rτrz ) = 0
L dr
So we have
d P0 − PL
(rτrz ) = r.
dr L
From Newton’s law of viscosity we know that τrz = −µ(dvz /dr), so
P0 − PL
 
d dvz
−µr = r.
dr dr L
Here we assume that the viscosity varies radially as a function of r, not axially as a function of z.

µ = µ(r)

One boundary condition is obtained from the assumption that the velocity is maximum furthest
from the wall (at r = 0). The second boundary condition is obtained from the assumption that
the fluid does not slip at the wall r = R.
dvz
B.C. 1 : = 0, at r = 0
dr
B.C. 2 : vz = 0, at r = R

Integrate both sides of the differential equation with respect to r.

dvz P0 − PL 2
−µ(r)r = r + C1
dr 2L
Apply the first boundary condition here to determine C1 .
dvz P0 − PL 2
−µ(r)(0) = (0) + C1 → 0 = C1
dr r=0 2L
Divide both sides by −µ(r)r to solve for dvz /dr.
dvz P0 − PL r
=−
dr 2L µ(r)
Integrate both sides of the differential equation with respect to r once more.
ˆ
P0 − PL r s
vz (r) = − ds + C2
2L µ(s)
The second boundary condition can be satisfied by choosing the lower limit of integration to be R
and setting C2 = 0. Use the minus sign to switch the limits of integration.
ˆ
P0 − PL R s
vz (r) = ds
2L r µ(s)

Make the following substitution in the integral.

s
ȳ = → s = Rȳ
R
ds
dȳ = → ds = R dȳ
R

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.3 Page 4 of 4

We have ˆ
P0 − PL 1
Rȳ
vz (r) = (R dȳ).
2L r/R µ(Rȳ)
Let y = r/R.
ˆ
(P0 − PL )R2 1
ȳ
vz (y) = dȳ
2L y µ
Our aim now is to determine the average velocity. We do this by integrating the velocity over the
cross-section that the fluid is flowing through and then dividing by that area.
ˆ
1
hvz i = vz dA
A
ˆ R
1
= vz (2πr dr)
πR2 0
ˆ R
2
= 2 rvz dr
R 0

Let y = r/R.
ˆ 1  
2 dr
= 2 (Ry)vz dy
R 0 dy
ˆ 1
2
= 2 (Ry)vz (R dy)
R 0
ˆ 1
=2 yvz dy
0
ˆ 1  ˆ
(P0 − PL )R2 1 ȳ

=2 y dȳ dy
0 2L y µ(ȳ)
ˆ ˆ
(P0 − PL )R2 1 1 ȳ
= y dȳ dy
L 0 y µ(ȳ)
ˆ ˆ
(P0 − PL )R2 1 ȳ ȳ
= y dy dȳ
L 0 0 µ(ȳ)
ˆ 1  ˆ ȳ
(P0 − PL )R2

ȳ
= dȳ y dy
L 0 µ(ȳ) 0
ˆ 1
(P0 − PL )R2
 2
ȳ ȳ
= dȳ
L 0 µ(ȳ) 2

The average velocity is thus ˆ

(P0 − PL )R2 1
ȳ 3
hvz i = dȳ.
2L 0 µ
Therefore,
(P0 −P )R
2 ˆ 1 ȳ ˆ 1
L

 2L dȳ (ȳ/µ)dȳ
vz y µ y
2 ˆ 1 ȳ 3
= =ˆ 1 .
hvz i (P0 −P )R
L


3

 2L

dȳ (ȳ /µ)dȳ
0 µ 0

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.4 Page 1 of 9

Problem 2C.4
Falling-cylinder viscometer (see Fig. 2C.4).6 A falling-cylinder viscometer consists of a long
vertical cylindrical container (radius R) capped at both ends, with a solid cylindrical slug (radius
κR). The slug is equipped with fins so that its axis is coincident with that of the tube.
One can observe the rate of descent of the slug in the cylindrical container when the latter is
filled with fluid. Find an equation that gives the viscosity of the fluid in terms of the terminal
velocity v0 of the slug and the various geometrical quantities shown in the figure.

(a) Show that the velocity distribution in the annular slit is given by

vz (1 − ξ 2 ) − (1 + κ2 ) ln(1/ξ)
=− (2C.4-1)
v0 (1 − κ2 ) − (1 + κ2 ) ln(1/κ)

in which ξ = r/R is a dimensionless radial coordinate.

(b) Make a force balance on the cylindrical slug and obtain

(ρ0 − ρ)g(κR)2 1 − κ2
   
1
µ= ln − (2C.4-2)
2v0 κ 1 + κ2

in which ρ and ρ0 are the densities of the fluid and the slug, respectively.

(c) Show that, for small slit widths, the result in (b) may be expanded in powers of ε = 1 − κ to
give
(ρ0 − ρ)gR2 ε3
 
1 13 2
µ= 1 − ε − ε + ··· (2C.4-3)
6v0 2 20
See §C.2 for information on expansions in Taylor series.

Solution
6
J. Lohrenz, G. W. Swift, and F. Kurata, AIChE Journal, 6, 547-550 (1960) and 7, 6S (1961); E. Ashare, R. B.
Bird, and J. A. Lescarboura, AIChE Journal, 11, 910-916 (1965) F. J. Eichstadt and G. W. Swift, AIChE Journal,
12, 1179-1183 (1966); M. C. S. Chen, J. A. Lescarboura, AIChE Journal, 14, 123-127 (1968).

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.4 Page 2 of 9

Part (a)

For this problem we choose a cylindrical coordinate system with the origin at the bottom of the
slug’s center. We assume that as the slug falls, the fluid in the annular slit flows in the z-direction
and varies as a function of radius r.
vz = vz (r)
As a result, only φrz (the z-momentum in the positive r-direction) and φzz (the z-momentum in
the positive z-direction) contribute to the momentum balance. The pressure is assumed to vary
with height z.
p = p(z)

Figure 1: This is the shell over which the momentum balance is made for flow through an annular
slit.

Rate of z-momentum into the shell at z = 0: (2πr∆r)φzz |z=0

Rate of z-momentum out of the shell at z = H: (2πr∆r)φzz |z=H
Rate of z-momentum into the shell at r: (2πrH)φrz |r
Rate of z-momentum out of the shell at r + ∆r: [2π(r + ∆r)H]φrz |r+∆r
Component of gravitational force on the shell in z-direction: − (2πr∆rH)ρg

If we assume steady flow, then the momentum balance is

Rate of momentum in − Rate of momentum out + Force of gravity = 0.

Considering only the z-component, we have

(2πr∆r)φzz |z=0 − (2πr∆r)φzz |z=H + (2πrH)φrz |r − [2π(r + ∆r)H]φrz |r+∆r − (2πr∆rH)ρg = 0.

Factor the left side.

−2πr∆r( φzz |z=H − φzz |z=0 ) − 2πH[ (r + ∆r)φrz |r+∆r − rφrz |r ] − 2πr∆rHρg = 0

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.4 Page 3 of 9

Divide both sides by −2π∆rH.

φzz |z=H − φzz |z=0 (r + ∆r)φrz |r+∆r − rφrz |r

r + + ρgr = 0
H ∆r
Take the limit as ∆r → 0.
φzz |z=H − φzz |z=0 (r + ∆r)φrz |r+∆r − rφrz |r
r + lim + ρgr = 0
H ∆r→0 ∆r
The second term is the definition of the first derivative of rφrz .

φzz |z=H − φzz |z=0 d

r + (rφrz ) + ρgr = 0
H dr
Now substitute the expressions for φrz and φzz .

φrz = τrz + 
ρv
rv
 z = τrz
2
τzz
φzz = pδzz + HH + ρvz vz = p(z) + ρvz

Since vz does not depend on z, the ρvz2 terms cancel.

2  2 
p|z=H +  z z=H − p|z=0 − 
ρv ρv

z z=0 d
r + (rτrz ) + ρgr = 0
H dr
Make it so that ρgr is in the fraction.

p|z=H + ρgH − p|z=0 d

r + (rτrz ) = 0
H dr
Place ρg0 in the numerator.

p|z=H + ρgH − p|z=0 − ρg0 d

r + (rτrz ) = 0
H dr
The point of doing this is that now we can use the modified pressure Pz = p(z) + ρgz.

PH − P0 d
r + (rτrz ) = 0
H dr
So we have
d PH − P0
(rτrz ) = − r.
dr H
From Newton’s law of viscosity we know that τrz = −µ(dvz /dr), so

PH − P0
 
d dvz
−µr =− r.
dr dr H

The boundary conditions for this differential equation are obtained from assuming that no slip
occurs between the fluid and the walls of the slug and container. That is, at r = κR the fluid
travels with the slug (vz = −v0 ) and at r = R the fluid is stationary (vz = 0).

B.C. 1 : vz = −v0 , at r = κR
B.C. 2 : vz = 0, at r = R

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.4 Page 4 of 9

Integrate both sides of the differential equation with respect to r.

dvz PH − P0 2
−µr =− r + C1 .
dr 2H
Divide both sides by −µr.
dvz PH − P0 C1
= r−
dr 2µH µr
Integrate both sides of the differential equation with respect to r once more.
P H − P 0 2 C1
vz (r) = r − ln r + C2
4µH µ
Apply the boundary conditions here to determine C1 and C2 .
PH − P0 C1
vz (κR) = (κR)2 − ln(κR) + C2 = −v0
4µH µ
P H − P 0 2 C1
vz (R) = R − ln R + C2 = 0
4µH µ
Solving this system of equations, we get
(PH − P0 )R2
 
µ 2
C1 = (1 − κ ) − v0
ln(1/κ) 4µH
(PH − P0 )R2
 
2 ln R ln R
C2 = (1 − κ ) −1 − v0 .
4µH ln(1/κ) ln(1/κ)
So we have for the velocity distribution,

PH − P0 2 (PH − P0 )R2
 
ln r 2
vz (r) = r − (1 − κ ) − v0
4µH ln(1/κ) 4µH
(PH − P0 )R2
 
2 ln R ln R
+ (1 − κ ) −1 − v0 .
4µH ln(1/κ) ln(1/κ)
Factor the right side.
PH − P0 2 (PH − P0 )R2 1 − κ2 v0
vz (r) = (r − R2 ) + (ln R − ln r) + (ln r − ln R)
4µH 4µH ln(1/κ) ln(1/κ)
Thus,
(PH − P0 )R2  r 2 1 − κ2
 
R v0 R
vz (r) = −1+ ln − ln .
4µH R ln(1/κ) r ln(1/κ) r
Here we introduce the dimensionless radial coordinate ξ = r/R.
(PH − P0 )R2 2 1 − κ2
 
v0
vz (ξ) = ξ −1+ ln(1/ξ) − ln(1/ξ) (1)
4µH ln(1/κ) ln(1/κ)
Our aim now is to eliminate the coefficient of the square brackets because it’s not in the desired
answer. As the slug descends in the container, it displaces a certain volume of fluid per unit time.
That same volume per unit time must be what flows up the side of the slug in the annular slit.
The following relation can be written from this.
dV dV
=
dt displaced from bottom of slug dt up side of slug in annular slit

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.4 Page 5 of 9

The volumetric flow rate is velocity times area, so on the left side it’s just v0 · π(κR)2 . Since the
velocity varies radially in the slit, we will have to integrate the velocity over the area on the right
side.
ˆ
v0 · πκ2 R2 = vz dA
ˆ R
2 2
v0 · πκ R = vz (2πr dr)
κR
ˆ R
2 2
v0 · πκ R = 2π rvz dr
κR

ˆ
v0 κ2 R2 R
(PH − P0 )R2  r 2 1 − κ2
   
R v0 R
= r −1+ ln − ln dr
2 κR 4µH R ln(1/κ) r ln(1/κ) r
Make the substitution,
r
ξ= → Rξ = r
R
dr
dξ = → R dξ = dr,
R
to get
ˆ 1
v0 κ2 R2 (PH − P0 )R2 2 1 − κ2
   
v0
= (Rξ) ξ −1+ ln(1/ξ) − ln(1/ξ) (R dξ)
2 κ 4µH ln(1/κ) ln(1/κ)
2 ˆ 1 ˆ 1
v0 κ2 R2 2 (PH − P0 )R 1 − κ2
   
3 v0
=R ξ −ξ+ ξ ln(1/ξ) dξ − ξ ln(1/ξ) dξ
2 4µH κ ln(1/κ) ln(1/κ) κ
ˆ  ˆ 1
v0 κ2 (PH − P0 )R2 1 3 1 − κ2

v0
= ξ −ξ− ξ ln ξ dξ + ξ ln ξ dξ
2 4µH κ ln(1/κ) ln(1/κ) κ
Bring the last term on the right side over to the left.
ˆ 1 ˆ 
v0 κ2 (PH − P0 )R2 1 3 1 − κ2

v0
− ξ ln ξ dξ = ξ −ξ− ξ ln ξ dξ
2 ln(1/κ) κ 4µH κ ln(1/κ)
1  1
v0 κ2 (PH − P0 )R2 ξ 4 ξ 2
 2
ξ2 1 − κ2
 2
ξ2

v0 ξ ξ
− − + ln ξ = − − − + ln ξ
2 ln(1/κ) 4 2 κ 4µH 4 2 ln(1/κ) 4 2 κ
2 (PH − P0 )R 2 2 4 2
 
v0 κ v0 1 κ κ 1−κ
− (−1 + κ2 −  2κ 2 
ln κ) = − + − − (−1 + κ2 − 2κ2 ln κ)
2 4 ln(1/κ) 4µH 4 2 4 4 ln(1/κ)
(PH − P0 )R 2 4 (1 − κ2 )2
 
v0 2 1 κ
(1 − κ ) = − + +
4 ln(1/κ) 4µH 4 4 4 ln(1/κ)
(PH − P0 )R 2 (1 − κ2 )2
 
v0 2 4
(1 − κ ) = κ −1+
ln(1/κ) 4µH ln(1/κ)
(PH − P0 )R 2 (κ2 − 1)2
 
v0 2 2 2
− (κ − 1) = (κ + 1)(κ − 1) +
ln(1/κ) 4µH ln(1/κ)
(PH − P0 )R 2 2
 
v0 κ −1
− = (κ2 + 1) +
ln(1/κ) 4µH ln(1/κ)
(PH − P0 )R (κ + 1) ln(1/κ) + κ2 − 1
2
 2 
v0
− =
ln(1/κ) 4µH ln(1/κ)

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.4 Page 6 of 9

Consequently,
(PH − P0 )R2 v0
=− 2 . (2)
4µH (κ + 1) ln(1/κ) + κ2 − 1
Substitute this result into equation (1).

1 − κ2
 
v0 2 v0
vz (ξ) = − ξ − 1 + ln(1/ξ) − ln(1/ξ)
(κ2 + 1) ln(1/κ) + κ2 − 1 ln(1/κ) ln(1/κ)

Divide both sides by v0 and factor the minus signs.

 
ξ 2 − 1 + 1−κ2 ln(1/ξ)
vz ln(1/κ) ln(1/ξ) 
= − 2 2
+
v0 (κ + 1) ln(1/κ) + (κ − 1) ln(1/κ)
 2
(ξ − 1) ln(1/κ) + (1 − κ2 ) ln(1/ξ) (κ2 + 1) ln(1/κ) ln(1/ξ) + (κ2 − 1) ln(1/ξ)

=− +
ln(1/κ)[(κ2 + 1) ln(1/κ) + (κ2 − 1)] ln(1/κ)[(κ2 + 1) ln(1/κ) + (κ2 − 1)]
 2 2)( ( (((
(ξ − 1) ln(1/κ) + ( (1(−( κ( ((
ln(1/ξ) + (κ2 + 1) ln(1/κ) ln(1/ξ) + ( 2−
(κ( (1)
(( ln(1/ξ)
=−
ln(1/κ)[(κ2 + 1) ln(1/κ) + (κ2 − 1)]
 2
XXX + (κ2 + 1)X

(ξ − 1)X ln(1/κ) ln(1/κ)
XXX ln(1/ξ)
=− XXX 2 + 1) ln(1/κ) + (κ2 − 1)]
ln(1/κ)[(κ
X

(1 − ξ 2 ) − (κ2 + 1) ln(1/ξ)
 
=− − 2
(κ + 1) ln(1/κ) + (κ2 − 1)
(1 − ξ 2 ) − (κ2 + 1) ln(1/ξ)
 
=−
−(κ2 + 1) ln(1/κ) − (κ2 − 1)

Therefore,
vz (1 − ξ 2 ) − (1 + κ2 ) ln(1/ξ)
=− ,
v0 (1 − κ2 ) − (1 + κ2 ) ln(1/κ)
where ξ = r/R.

Figure 2: This is a plot of the velocity distribution when κ = 0.4 and v0 = 2 for κ ≤ ξ ≤ 1.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.4 Page 7 of 9

Part (b)

To determine the viscosity, the sum of the forces in the z-direction will be considered. There are
five forces that need to be taken into account in the free body diagram of the slug: (1) the
gravitational force, (2) the buoyant force, (3) the viscous force acting on the side from the flow of
fluid in the annulus, (4) the weight of the water (pressure) acting over the top of the slug’s
surface, and (5) the weight of the water (pressure) acting over the bottom of the slug’s surface.
The forces due to pressure are normal to the slug’s surface, and the viscous force is parallel to the
slug’s surface in the direction opposing the slug’s motion.

Figure 3: This is the free body diagram of the slug.

Gravitational force = Density of slug × Volume of slug × Acceleration = ρ0 × π(κR)2 H × g

Buoyant force = Density of fluid × Volume of slug × Acceleration = ρ × π(κR)2 H × g
Viscous force = Shearing stress × Surface area = ( −τrz |r=κR ) × 2πκRH
Force at bottom = Pressure at bottom × Surface area = P0 × π(κR)2
Force at top = Pressure at top × Surface area = PH × π(κR)2

The minus sign in the shearing stress is due to the fact that the fluid in the annulus is at a higher
radius acting on the slug’s surface, which has a lower radius. Use Newton’s law of viscosity,
τrz = −µ(dvz /dr), to evaluate it.

Viscous force = ( −τrz |r=κR ) × 2πκRH

 
dvz
= − −µ × 2πκRH
dr r=κR
dvz
= × 2πκRHµ
dr r=κR
v0 1 − κ2
=− × 2πκRHµ
κR 1 − κ2 − (1 + κ2 ) ln(1/κ)
1 − κ2
=− × 2πHµv0
1 − κ2 − (1 + κ2 ) ln(1/κ)

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.4 Page 8 of 9

Because the slug is falling at terminal velocity, the acceleration is zero, so the sum of the forces in
the z-direction is equal to 0. X
Fz = maz = 0
Plug in the forces on the left side. The positive z-direction is chosen to be upward, so forces
pointing up are positive and those pointing down are negative.

1 − κ2
− ρ0 π(κR)2 Hg + ρπ(κR)2 Hg − × 2πHµv0
1 − κ2 − (1 + κ2 ) ln(1/κ)
+ P0 × π(κR)2 − PH × π(κR)2 = 0

Divide both sides by π(κR)2 H.

1 − κ2 2µv0 P0 − PH
(ρ − ρ0 )g − 2 2
× 2 2+ =0
1 − κ − (1 + κ ) ln(1/κ) κ R H
Solve equation (2) for (P0 − PH )/H and substitute the expression here.
1 − κ2 2µv0 v0 4µ
(ρ − ρ0 )g − × + × =0
1− κ2 − (1 + κ2 ) ln(1/κ) κ2 R2 (κ2 + 1) ln(1/κ) + κ2 − 1 R2
All that’s left to do is to solve for µ.
2µv0 (1 − κ2 ) 4µv0 κ2
κ2 R2 (ρ − ρ0 )g − − =0
1 − κ2 − (1 + κ2 ) ln(1/κ) 1 − κ2 − (1 + κ2 ) ln(1/κ)
2µv0 (1 + κ2 )
κ2 R2 (ρ − ρ0 )g − =0
1 − κ2 − (1 + κ2 ) ln(1/κ)
κ2 R2 (ρ − ρ0 )g(1 − κ2 ) − κ2 R2 (ρ − ρ0 )g(1 + κ2 ) ln(1/κ) − 2µv0 (1 + κ2 ) = 0
κ2 R2 (ρ − ρ0 )g[(1 − κ2 ) − (1 + κ2 ) ln(1/κ)] = 2µv0 (1 + κ2 )
κ2 R2 (ρ − ρ0 )g (1 − κ2 )
 
− ln(1/κ) =µ
2v0 1 + κ2
Therefore,
(ρ0 − ρ)g(κR)2 1 − κ2
   
1
µ= ln − .
2v0 κ 1 + κ2
Part (c)

For small slit widths, κ is just barely less than one: κ = 1 − ε, where 0 < ε  1. Substitute this
into the result of part (b).
(ρ0 − ρ)g(κR)2 1 − κ2
   
1
µ= ln −
2v0 κ 1 + κ2
(ρ0 − ρ)g(1 − ε)2 R2 1 − (1 − ε)2
 
1
= ln −
2v0 1 − ε 1 + (1 − ε)2
(ρ0 − ρ)gR2 2ε − ε2
 
2
= (1 − ε) − ln(1 − ε) −
2v0 2 − 2ε + ε2
The Taylor series expansion for ln(1 − ε) is
1 1 1 1
ln(1 − ε) = −ε − ε2 − ε3 − ε4 − ε5 − · · · .
2 3 4 5

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.4 Page 9 of 9

Plug this in to the formula for µ.

(ρ0 − ρ)gR2 2ε − ε2
   
2 1 2 1 3 1 4 1 5
µ= (1 − ε) − −ε − ε − ε − ε − ε − · · · −
2v0 2 3 4 5 2 − 2ε + ε2
2 2ε − ε2
 
(ρ0 − ρ)gR 2 1 2 1 3 1 4 1 5
= (1 − 2ε + ε ) ε + ε + ε + ε + ε + · · · −
2v0 2 3 4 5 2 − 2ε + ε2
(ρ0 − ρ)gR2 2ε − ε2
  
1 2 1 3 1 4 1 5 2 2
= ε + ε + ε + ε + ε + · · · (1 − 2ε + ε ) − (1 − 2ε + ε )
2v0 2 3 4 5 2 − 2ε + ε2
Use long division to obtain a series for the quotient.
1 1 1
ε + ε2 − ε4 − ε5 + · · ·
2 4 4
2 − 2ε + ε2 2 3


2ε − ε + 0ε
(−) 2ε − 2ε2 + ε3
ε2 − ε3
1
(−) ε2 − ε3 + ε4
2
− 12 ε4
1 1 1
(−) − ε4 + ε 5 − ε6
2 2 4
1 5 1 6
−2ε + 4ε

Having terms up to ε5 will suffice.

(ρ0 − ρ)gR2
 
1 2 1 3 1 4 1 5
µ= ε + ε + ε + ε + ε + · · · (1 − 2ε + ε2 )
2v0 2 3 4 5
  
1 2 1 4 1 5 2
− ε + ε − ε − ε + · · · (1 − 2ε + ε ) .
2 4 4
Multiply the series together.

(ρ0 − ρ)gR2
        
2 1 3 1 4 1 2 1 5 1 1 1
= ε + ε −2 + +ε 1−1+ +ε − + +ε − + + ···
2v0 2 3 2 3 4 3 2 5
     
2 1 3 4 1 1 5 1 1
−ε − ε −2 + − ε (1 − 1) − ε − −ε −
2 2 4 2 4
(ρ0 − ρ)gR2 1 3
 
1 1 1 1
= ε + ε4 + ε5 − ε4 − ε5 + · · ·
2v0 3 12 30 4 4
Combine like-terms.
(ρ0 − ρ)gR2 1 3 1 4 13 5
 
= ε − ε − ε + ···
2v0 3 6 60

Factor ε3 /3 to obtain the final result. Therefore,

(ρ0 − ρ)gR2 ε3
 
1 13
µ= 1 − ε − ε2 + · · · .
6v0 2 20

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.5 Page 1 of 4

Problem 2C.5
Falling film on a conical surface (see Fig. 2C.5).7 A fluid flows upward through a circular
tube and then downward on a conical surface. Find the film thickness as a function of the
distance s down the cone.

(a) Assume that the results of §2.2 apply approximately over any small region of the cone
surface. Show that a mass balance on a ring of liquid contained between s and s + ∆s gives:
d d
(sδhvi) = 0 or (sδ 3 ) = 0 (2C.5-1)
ds ds

(b) Integrate this equation and evaluate the constant of integration by equating the mass rate
of flow w up the central tube to that flowing down the conical surface at s = L. Obtain the
following expression for the film thickness:
s  
3 3µw L
δ= 2
(2C.5-2)
πρ gL sin 2β s

Solution
7
R. B. Bird, in Selected Topics in Transport Phenomena, CEP Symposium Series #58, 61, 1-15 (1965).

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.5 Page 2 of 4

Part (a)

Figure 1: It’s a bit difficult to visualize, but the shell under consideration here is exterior to the
conical surface, above the conical frustum at s + ∆s, and below the conical frustum at s.

The law of conservation of mass states that matter is neither created nor destroyed. The same
amount of fluid that enters a shell per unit time must leave at that same rate; otherwise, fluid will
build up or accumulate within the shell. The mathematical expression for this idea, a mass
balance, is as follows.

rate of mass in − rate of mass out = rate of mass accumulation

Assuming that no fluid accumulates on the conical surface, the right-hand side is zero.

rate of mass in − rate of mass out = 0

Mass flows down the conical surface, so mass flows into the shell at s and out of the shell at
s + ∆s.
dm dm
− =0
dt s dt s+∆s
Mass is density ρ times volume V .

d(ρV ) d(ρV )
− =0
dt s dt s+∆s

Assuming density is constant, it can be pulled out and cancelled from both sides.

dV dV
− =0
dt s dt s+∆s

The volumetric flow rate is average velocity times the cross-sectional area that the fluid flows
through.
(hviA)|s − (hviA)|s+∆s = 0

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.5 Page 3 of 4

To obtain the cross-sectional area we multiply the circumference of the circle at s or s + ∆s by

the film thickness δ(s) or δ(s + ∆s), respectively. For the average velocity we use the result for a
falling film on an inclined surface in §2.2.

ρgδ 2 cos β
hvi =
3µ
The mass balance becomes
ρg[δ(s)]2 cos β ρg[δ(s + ∆s)]2 cos β
· 2πrs δ(s) − · 2πrs+∆s δ(s + ∆s) = 0.
3µ 3µ

Divide both sides by ρg cos β · 2π/3µ.

rs [δ(s)]3 − rs+∆s [δ(s + ∆s)]3 = 0

The relationship between r and s is r = s sin β, so

rs = s sin β
rs+∆s = (s + ∆s) sin β.

Substitute these results into the mass balance.

s sin β[δ(s)]3 − (s + ∆s) sin β[δ(s + ∆s)]3 = 0

Divide both sides by −(sin β)∆s.

(s + ∆s)[δ(s + ∆s)]3 − s[δ(s)]3

=0
∆s
Take the limit as ∆s → 0.
(s + ∆s)[δ(s + ∆s)]3 − s[δ(s)]3
lim =0
∆s→0 ∆s

This is how the first derivative of s[δ(s)]3 is defined. Therefore,

d
(sδ 3 ) = 0.
ds
Part (b)

Integrate both sides of this differential equation with respect to s.

sδ 3 = C1

Solve this for δ 3 .

C1
δ3 =
s
At s = L, we have
C1
δ3 = . (1)
L

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.5 Page 4 of 4

In §2.2 the mass flow rate w of a film falling down an incline is given in Eq. 2.2-21.

ρ2 gW δ 3 cos β
w= (2.2-21)
3µ
To adapt this result to a film falling down a conical surface, set the width of the film W equal to
2πr, or 2πs sin β.

ρ2 g(2πs sin β)δ 3 cos β

w=
3µ
Use the trigonometric identity 2 sin β cos β = sin 2β.

πρ2 gs(sin 2β)δ 3

w=
3µ
To determine C1 we use the condition that the mass flow rate up the circular tube w is equal to
the mass flow rate down the conical surface at s = L. That is,

dm
w=
dt
 2s=L
πρ gs(sin 2β)δ 3

=
3µ s=L
πρ2 gL(sin 2β)δ 3
= .
3µ

Solve this equation for δ 3 .

3µw
δ3 = (2)
πρ2 gL sin 2β
Set equations (1) and (2) equal to each other to solve for C1 .

C1 3µw 3µw
= 2
→ C1 =
L πρ gL sin 2β πρ2 g sin 2β
Now that C1 is determined, the film thickness is known.
3µw 1
δ3 =
πρ2 g sin 2β s
 
3µw L
=
πρ2 gL sin 2β s

Therefore, s  
3 3µw L
δ= 2
.
πρ gL sin 2β s

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.6 Page 1 of 4

Problem 2C.6
Rotating cone pump (see Fig. 2C.6). Find the mass rate of flow through this pump as a
function of the gravitational acceleration, the impressed pressure difference, the angular velocity
of the cone, the fluid viscosity and density, the cone angle, and other geometrical quantities
labeled in the figure.

Figure 1: This is Fig. 2C.6 in the text. A rotating cone pump. The variable r is the distance from
the axis of rotation out to the center of the slit.

(a) Begin by analyzing the system without the rotation of the cone. Assume that it is possible
to apply the results of Problem 2B.3 locally. That is, adapt the solution for the mass flow
rate from that problem by making the following replacements:

replace (P0 − PL )/L by − dP/dz

replace W by 2πr = 2πz sin β

thereby obtaining
B 3 ρ · 2πz sin β
 
2 dP
w= − (2C.6-1)
3 dz µ
The mass flow rate w is a constant over the range of z. Hence this equation can be
integrated to give
3 µw L2
(P1 − P2 ) = ln (2C.6-2)
4π B 3 ρ sin β L1

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.6 Page 2 of 4

(b) Next, modify the above result to account for the fact that the cone is rotating with angular
velocity Ω. The mean centrifugal force per unit volume acting on the fluid in the slit will
have a z-component approximately given by

(Fcentrif )z = KρΩ2 z sin2 β (2C.6-3)

What is the value of K? Incorporate this as an additional force tending to drive the fluid
through the channel. Show that this leads to the following expression for the mass rate of
flow: " #
4πB 3 ρ sin β (P1 − P2 ) + 21 KρΩ2 sin2 β (L22 − L21 )


w= (2C.6-4)
3µ ln(L2 /L1 )
Here Pi = pi + ρgLi cos β.

Solution

Part (a)

Without rotation the flow through the pump is essentially flow through a slit at an angle β. The
mass flow rate through a slit from Problem 2B.3 is

2 (P0 − PL )B 3 W ρ
w= ,
3 µL
where P = p − ρgz and z measures vertical distance. Make the indicated replacements.

dP B 3 ρ · 2πz sin β
 
2
w= −
3 dz µ

Because the positive z-direction goes from the bottom to the top of the pump, P here is
p + ρgL cos β. Assuming w is constant, the differential equation can be solved by separation of
variables.
3µw dz
3
= −dP
4B ρπ sin β z
Integrate both sides.
ˆ L2 ˆ P2
3µw dz
3
= (−dP)
L1 4B ρπ sin β z P1

Proceed with the integration.

3µw
(ln L2 − ln L1 ) = (P1 − P2 )
4B 3 ρπ sin β
Therefore,
3 µw L2
(P1 − P2 ) = 3
ln .
4π B ρ sin β L1

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.6 Page 3 of 4

Part (b)

The mean centrifugal force can be found by considering the centrifugal force of fluid at the center
of the slit.
v2
Fcentrifugal = macentrifugal = m
r
Since Ω is perpendicular to the radial direction, it is related to v by v = Ωr. For fluid at the
center of the slit a factor of 1/2 is included because it is halfway between a wall that is stationary
(the container) and a wall that is moving (the cone). It is assumed that the fluid does not slip on
these surfaces.
 2
m 1
Fcentrifugal = Ωr
r 2
m Ω2 r 2
=
r 4
1
= mΩ2 r
4

Figure 2: A relationship between r and z can be obtained from this triangle: r = z sin β.

1
Fcentrifugal = mΩ2 z sin β
4
This centrifugal force is in the radial direction. To get the component of this force in the
z-direction, see Figure 3.
1
(Fcentrifugal )z = mΩ2 z sin2 β
4
Finally, to get the centrifugal force in the z-direction per unit volume, divide both sides by V .
Replace m/V with the density ρ.

(Fcentrifugal )z 1
(Fcentrif )z = = ρΩ2 z sin2 β
V 4
Therefore, K = 1/4. Include this force in the equation for the mass flow rate to account for the
rotation of the cone.
dP B 3 ρ · 2πz sin β
    3
2 2 dP 1 2 2 B ρ · 2πz sin β
w= − → w= − + ρΩ z sin β
3 dz µ 3 dz 4 µ
| {z }
no rotation

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.6 Page 4 of 4

Figure 3: A relationship between
a radial force Fr and a force in the z-direction Fz can be obtained
from this triangle: Fr cos π2 − β = Fz , or Fr sin β = Fz .

Once again, the differential equation can be solved by separation of variables, assuming the mass
flow rate w is constant.
3µw 1 dP 1
3
=− + ρΩ2 z sin2 β
4πB ρ sin β z dz 4
dP 3µw 1 1 2
=− 3
+ ρΩ z sin2 β
dz 4πB ρ sin β z 4
 
3µw 1 1 2 2
dP = − + ρΩ z sin β dz
4πB 3 ρ sin β z 4
Integrate both sides.
ˆ P2 ˆ L2  
3µw 1 1 2
dP = − + ρΩ z sin2 β dz
P1 L1 4πB 3 ρ sin β z 4
L2
1 2 z2
 
3µw
(P2 − P1 ) = − ln z + ρΩ sin2 β
4πB 3 ρ sin β 4 2 L1

3µw 1 2 − L21
2 L2
(P2 − P1 ) = − 3
(ln L2 − ln L1 ) + ρΩ sin2 β
4πB ρ sin β 4 2
Combine the logarithms and solve for w.
3µw L2 1
ln = (P1 − P2 ) + ρΩ2 (L22 − L21 ) sin2 β
4πB 3 ρ sin β L1 8
Therefore, " #
(P1 − P2 ) + 1 2 2
(L22 − L21 )


4πB 3 ρ sin β 2 KρΩ sin β
w= ,
3µ ln(L2 /L1 )
where K = 1/4 and Pi = pi + ρgLi cos β.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.7 Page 1 of 8

Problem 2C.7
A simple rate-of-climb indicator (see Fig. 2C.7). Under the proper circumstances the simple
apparatus shown in the figure can be used to measure the rate of climb of an airplane. The gauge
pressure inside the Bourdon element is taken as proportional to the rate of climb. For the
purposes of this problem the apparatus may be assumed to have the following properties: (i) the
capillary tube (of radius R and length L, with R  L) is of negligible volume but appreciable
flow resistance; (ii) the Bourdon element has a constant volume V and offers negligible resistance
to the flow; and (iii) flow in the capillary is laminar and incompressible, and the volumetric flow
rate depends only on the conditions at the ends of the capillary.

(a) Develop an expression for the change of pressure with altitude, neglecting temperature
changes, and considering air to be an ideal gas of constant composition. (Hint: Write a shell
balance in which the weight of gas is balanced against the static pressure.)

(b) By making a mass balance over the gauge, develop an approximate relation between gauge
pressure pi − po and rate of climb vz for a long continued constant-rate climb. Neglect
change of air viscosity, and assume changes in air density to be small.

(c) Develop an approximate expression for the “relaxation time” trel of the indicator—that is,
the time required for the gauge pressure to drop to 1/e of its initial value when the external
pressure is suddenly changed from zero (relative to the interior of the gauge) to some
different constant value, and maintained indefinitely at this new value.

(d) Discuss the usefulness of this type of indicator for small aircraft.

(e) Justify the plus and minus signs in the figure.

Answers: (a) dp/dz = −ρg = −(pM/RT )g

(b) pi − po ≈ vz (8µL/πR4 )(M gV /Rg T ), where Rg is the gas constant
and M is the molecular weight.
1
(c) trel = (128/π)(µV L/D4 p̄), where p̄ = (pi + po )
2

Solution

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.7 Page 2 of 8

Part (a)

Following the hint, a rectangular shell will be considered where forces due to gravity and pressure
act on it.

Figure 1: This is the shell over which the force balance is made to determine the change in pressure
with altitude.

The positive z-direction is chosen to point upward, and the forces due to pressure act normal to
the shell’s surface on the top (at z + ∆z) and bottom (at z).

Force due to pressure at top: − p|z+∆z · HW

Force due to pressure at bottom: p|z · HW
Gravitational force on shell: − ρg · HW ∆z

Assuming the shell is in equilibrium, the sum of the forces acting on it is equal to 0.
X
Fz = 0

Place the forces above on the left side.

− p|z+∆z · HW + p|z · HW − ρg · HW ∆z = 0

Divide both sides by HW ∆z and factor a minus sign from the first two terms.

p|z+∆z − p|z
− − ρg = 0
∆z
p|z+∆z − p|z
= −ρg
∆z
Now take the limit as ∆z → 0.
p|z+∆z − p|z
lim = −ρg
∆z→0 ∆z
The term on the left is how the first derivative of p with respect to z is defined.
dp
= −ρg
dz

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.7 Page 3 of 8

If the air outside the aircraft is assumed to be an ideal gas, then the equation of state,

pv = nRg T,

can be used to write the density ρ in terms of the pressure p. Start by writing the number of
moles n as the mass m divided by the molar mass M .
m pM m
pv = Rg T → = =ρ
M Rg T v

Therefore,
dp pM
=− g.
dz Rg T
Part (b)

The differential equation above is for the pressure outside the aircraft.
dpo po M
=− g
dz Rg T

The boundary condition that goes with it is po (0) = ps , namely at z = 0 (sea level) the pressure is
known to be ps . Solve the differential equation with separation of variables.
dpo Mg
=− dz
po Rg T

Integrate both sides. ˆ ˆ

dpo Mg
= − dz
po Rg T
Mg
ln po = − z+C
Rg T
Exponentiate both sides.  
Mg
po (z) = exp − z+C
Rg T
Introduce a new constant of integration.
 
Mg
po (z) = A exp − z
Rg T

Apply the boundary condition here: po (0) = A = ps .

 
Mg
po (z) = ps exp − z (1)
Rg T

Mass flows through the capillary tube as a consequence of the pressure difference between the
inside of the Bourdon element and the outside, pi − po .

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.7 Page 4 of 8

Figure 2: Schematic of the capillary tube. It is assumed that R  L.

Assuming no mass accumulates inside the tube, the following mass balance can be made.

Rate of mass in − Rate of mass out = 0

wi − wo = 0
Bring wo to the right side.

wi = wo = w
dm dm
=
dt i dt o
d(ρv) d(ρv)
=
dt i dt o

Density is assumed to be constant.

dv dv
ρ =ρ
dt i dt o

Divide both sides by ρ.

dv dv dv
= = .
dt i dt o dt
From these equations we conclude that the rate of mass flow at one end of the tube is equal to
that at the other end (use w to represent it). The same can be said for the volumetric flow (use
dv/dt to represent it). The fact that a capillary tube connects the Bourdon element with the
outside means that the Hagen-Poiseuille equation can be used for the mass flow rate through it.

π(pi − po )R4 ρ
w=
8µL
Solve this equation for the pressure difference. For ρ an average density for air at both ends of the
tube should be used. Then it cancels out.
8µL
pi − po = w
πR4 ρ
8µL dv
= ρ
πR4 ρ dt
8µL dv
= (2)
πR4 dt
Our aim now is to obtain an expression for dv/dt. Use the equation of state to solve for v.
nRg T
pv = nRg T → v=
p

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.7 Page 5 of 8

So we have
 
dv d nRg T
=
dt dt po
 
d 1
= nRg T
dt po
 
1 dpo
= nRg T − 2
p dt
 o
1 dpo dz
= nRg T − 2 ,
po dz dt
where dz/dt is the rate of climb of the aircraft. This is assumed to be equal to vz , a constant.
Substitute the expression for po in equation (1) here.
        
dv 1 Mg Mg Mg
= nRg T − 2 exp 2 z ps − exp − z vz
dt ps Rg T Rg T Rg T
 
nRg T M g Mg
= exp z vz
ps Rg T Rg T
 
Mg Mg
=V exp z vz
Rg T Rg T
Plug this expression for dv/dt back into equation (2) for the pressure difference.
 
8µL M gV Mg
p i − p o = vz exp z
πR4 Rg T Rg T
Use the Taylor series expansion for the exponential function.
"    2 #
8µL M gV Mg Mg
pi − po = vz 1+ z + z + ···
πR4 Rg T Rg T Rg T

Therefore,
8µL M gV
p i − p o ≈ vz .
πR4 Rg T
Part (c)

Suppose initially that the outside pressure and the inside pressure are the same at pi and that all
of a sudden the outside pressure lowers to po and stays there. As a result of the pressure
difference, mass will flow through the capillary tube to the outside. The pressure inside the
Bourdon element will thus lower as a function of time and eventually come to equilibrium with
the outside pressure. According to the Hagen-Poiseuille equation, the mass flow rate through the
tube is
π[P (t) − po ]R4 ρ̄
w= ,
8µL
where P (t) represents the pressure on the inside of the Bourdon element as a function of time and
ρ represents the average density of air in the tube ρ = (ρi + ρo )/2. The average density is used as a
simplification so that P (t) is the only unknown here. Replace w with −dm/dt. The minus sign is
included because the Bourdon element loses mass as time passes.
dm π[P (t) − po ]R4 ρ̄
− =
dt 8µL

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.7 Page 6 of 8

Replace m with the molar mass M times the number of moles n in the Bourdon element (a
function of time as well).
d(M n) π[P (t) − po ]R4 ρ̄
− =
dt 8µL
M is assumed to be constant, so it can be pulled in front of the derivative.

dn π[P (t) − po ]R4 ρ̄

−M =
dt 8µL
n can be written in terms of P using the equation of state,
PV
P V = nRg T → n= .
Rg T
Substitute this into the Hagen-Poiseuille equation.

π[P (t) − po ]R4 ρ̄

 
d PV
−M =
dt Rg T 8µL
V , the volume of the Bourdon element, is assumed to be constant, so it can be pulled in front of
the derivative. Rg and T are also assumed to be constant.

M V dP π[P (t) − po ]R4 ρ̄

− =
Rg T dt 8µL
This is an inhomogeneous first-order differential equation that can be solved with an integrating
factor. Put it into standard form first.
dP Rg T πR4 ρ̄ Rg T πR4 ρ̄
+ P (t) = po
dt M V 8µL M V 8µL
The integrating factor I has the form,
ˆ t
Rg T πR4 ρ̄ Rg T πR4 ρ̄
  
I = exp ds = exp t .
M V 8µL M V 8µL
Multiply both sides of the differential equation by I.
Rg T πR4 ρ̄ dP Rg T πR4 ρ̄ Rg T πR4 ρ̄ Rg T πR4 ρ̄ Rg T πR4 ρ̄
     
exp t + exp t P (t) = po exp t
M V 8µL dt M V 8µL M V 8µL M V 8µL M V 8µL
The left side is just d/dt(IP ) as a result of the product rule.

Rg T πR4 ρ̄ Rg T πR4 ρ̄ Rg T πR4 ρ̄

     
d
exp t P (t) = po exp t
dt M V 8µL M V 8µL M V 8µL
Integrate both sides of the equation with respect to t.
ˆ t
Rg T πR4 ρ̄ Rg T πR4 ρ̄ Rg T πR4 ρ̄
   
exp t P (t) = po exp s ds + C1
M V 8µL M V 8µL M V 8µL
Evaluate the integral on the right side.
Rg T πR4 ρ̄ Rg T πR4 ρ̄
   
exp t P (t) = po exp t + C1
M V 8µL M V 8µL

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.7 Page 7 of 8

Divide both sides by I to solve for P (t).

Rg T πR4 ρ̄
 
P (t) = po + C1 exp − t
M V 8µL
The constant of integration is determined from the initial condition P (0) = pi .

P (0) = po + C1 = pi → C1 = pi − po

Thus, we have
Rg T πR4 ρ̄
 
P (t) = po + (pi − po ) exp − t
M V 8µL
for the pressure inside the Bourdon element as a function of time.

Figure 3: This is a plot of the pressure inside the Bourdon element as a function of time. Note
that initially the pressure is pi and that as t goes to infinity, it comes to equilibrium at po .

trel , the quantity we’re interested in, is the time for the pressure to drop to 1/e of its initial value.
This is found by setting the exponent of the exponential function equal to −1 and then solving for
the time.
1 Rg T πR4 ρ̄
P (trel ) = po + (pi − po ) ⇒ − trel = −1
e M V 8µL
Solve this equation for trel .
8µL M V
trel =
πR4 ρ̄ Rg T
8µL M V
=
4
π D ρ̄ Rg T
2
128 µV L M
=
π D4 ρ̄Rg T

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2C.7 Page 8 of 8

The average density ρ̄ can be related to the average pressure p̄ using the equation of state.
m m Rg T Rg T 1 M
p̄V = nRg T = Rg T → p̄ = = ρ̄ → =
M V M M p̄ ρ̄Rg T

Therefore,
128 µV L
trel = ,
π D4 p̄
where p̄ = (pi + po )/2.

Part (d)

The approximations that were made (neglecting changes in air temperature, viscosity, density,
and composition) are reasonable ones, provided that the plane does not fly too high. A small
aircraft is not only expected to fly at low altitudes, but also to have a constant rate of climb. This
sort of indicator would therefore be useful for one.

Part (e)

As the aircraft ascends (positive rate of climb), the pressure outside decreases and the Bourdon
element uncoils, making the needle point to the right. Conversely, as the aircraft descends
(negative rate of climb), the pressure outside increases and the Bourdon element coils up, making
the needle point to the left.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2D.2 Page 1 of 6

Problem 2D.2
Drainage of liquids9 (see Fig. 2D.2). How much liquid clings to the inside surface of a large
vessel when it is drained? As shown in the figure there is a thin film of liquid left behind on the
wall as the liquid level in the vessel falls. The local film thickness is a function of both z (the
distance down from the initial liquid level) and t (the elapsed time).

(a) Make an unsteady-state mass balance on a portion of the film between z and z + ∆z to get

∂ ∂δ
hvz iδ = − (2D.2-1)
∂z ∂t

(b) Use Eq. 2.2-18 and a quasi-steady-assumption to obtain the following first-order partial
differential equation for δ(z, t):
∂δ ρg 2 ∂δ
+ δ =0 (2D.2-2)
∂t µ ∂z

(c) Solve this equation to get r

µ z
δ(z, t) = (2D.2-3)
ρg t
What restrictions have to be placed on this result?

Solution
9
J. J. van Rossum, Appl. Sci. Research, 47, 121-144 (1958); see also V. G. Levich, Physicochemical Hydrodynamics,
Prentice-Hall, Englewood Cliffs, N.J. (1962), Chapter 12.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2D.2 Page 2 of 6

Part (a)

The law of conservation of mass states that matter is neither created nor destroyed. The same
amount of fluid that enters a shell per unit time must leave at that same rate; otherwise, fluid will
build up or accumulate within the shell. The mathematical expression for this idea, a mass
balance, is as follows.

rate of mass in − rate of mass out = rate of mass accumulation

Because of viscosity, the fluid in this problem doesn’t flow freely. Some of it gets stuck to the

Figure 1: This is the shell over which the mass balance is made for liquid draining.

wall, so the rate of accumulation is not zero. Mass flows in at z and out at z + ∆z, so with the
shell in Figure 1 the mass balance becomes

dm dm dm
− = .
dt z dt z+∆z dt shell

Mass is density times volume.

d(ρV ) d(ρV ) d(ρV )

− =
dt z dt z+∆z dt shell

As the density is assumed to be constant, it can be pulled out of each derivative and cancelled.

dV dV dV
− =
dt z dt z+∆z dt shell

For the left-hand side, the volumetric flow rate is the average velocity times the cross-sectional
area that the fluid flows through.

dV
(hviA)|z − (hviA)|z+∆z =
dt shell

The cross-sectional areas at z and z + ∆z are W δ(z, t) and W δ(z + ∆z, t), respectively.

dV
hvi|z [W δ(z, t)] − hvi|z+∆z [W δ(z + ∆z, t)] =
dt shell

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2D.2 Page 3 of 6

Factor −W from the left side.

dV
−W [ hvi|z+∆z δ(z + ∆z, t) − hvi|z δ(z, t)] =
dt shell

For the right-hand side, an expression for the volume of the shell from z to ∆z can be obtained by
integrating the cross-sectional area over that interval.
ˆ z+∆z
V |shell = A(s) ds
z

We get ˆ z+∆z
d
−W [ hvi|z+∆z δ(z + ∆z, t) − hvi|z δ(z, t)] = [W δ(s, t)] ds.
dt z
Now take the limit as ∆z → 0. It is assumed that δ is continuous and differentiable.
ˆ
d z+∆z
lim −W [ hvi|z+∆z δ(z + ∆z, t) − hvi|z δ(z, t)] = lim [W δ(s, t)] ds
∆z→0 ∆z→0 dt z

Bring the limit inside the derivative.

ˆ z+∆z
d
−W lim [ hvi|z+∆z δ(z + ∆z, t) − hvi|z δ(z, t)] = lim [W δ(s, t)] ds
∆z→0 dt ∆z→0 z

The interval of integration becomes so small that the integrand is essentially constant and can be
pulled in front of the integral.
ˆ z+∆z
d
−W lim [ hvi|z+∆z δ(z + ∆z, t) − hvi|z δ(z, t)] = lim [W δ(z, t)] ds
∆z→0 dt ∆z→0 z

Evaluate the integral.

d
−W lim [ hvi|z+∆z δ(z + ∆z, t) − hvi|z δ(z, t)] = lim [W δ(z, t)]∆z
∆z→0 dt ∆z→0
W and ∆z are constants with respect to time.
 
∂δ
−W lim [ hvi|z+∆z δ(z + ∆z, t) − hvi|z δ(z, t)] = W lim ∆z
∆z→0 ∂t ∆z→0

Divide both sides by W ∆z.

hvi|z+∆z δ(z + ∆z, t) − hvi|z δ(z, t) ∂δ

− lim =
∆z→0 ∆z ∂t
The limit on the left is how the first derivative of hviδ with respect to z is defined.

∂ ∂δ
− hvz iδ =
∂z ∂t
Therefore,
∂ ∂δ
hvz iδ = − .
∂z ∂t

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2D.2 Page 4 of 6

Part (b)

Eq. 2.2-18 gives the velocity distribution for a fluid flowing down an inclined plane.

ρgδ 2 cos β
  x 2 
vz = 1− (2.2-18)
2µ δ
ρg cos β 2
= (δ − x2 )
2µ
To adapt the formula to this problem, set β = 0 because the wall is vertical. To obtain the
average velocity hvi, integrate this velocity over the cross-sectional area that the fluid flows
through and then divide by that area.
ˆ
1
hvi = vz dA
A
ˆ δ(z,t)
1
= vz (W dx)
W δ(z, t) 0
ˆ δ(z,t)
1
= vz dx
δ(z, t) 0
ˆ
1 δ ρg 2
= (δ − x2 ) dx
δ 0 2µ
δ
x3

ρg 2
= δ x−
2µδ 3 0
 3
ρg 2δ
=
2µδ 3
ρgδ 2
=
3µ

Substitute this formula for hvi into the mass balance.

∂ ρgδ 2
 
∂δ
δ=−
∂z 3µ ∂t
3
 
∂ ρgδ ∂δ
=−
∂z 3µ ∂t
2
ρgδ ∂δ ∂δ
=−
µ ∂z ∂t

Therefore, we obtain a nonlinear first-order partial differential equation (PDE) for the film
thickness.
∂δ ρg 2 ∂δ
+ δ =0
∂t µ ∂z

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2D.2 Page 5 of 6

Part (c)

For a function of two variables δ = δ(z, t), its differential is defined as

∂δ ∂δ
dδ = dz + dt.
∂z ∂t
Divide both sides by dt.
dδ ∂δ dz ∂δ
= +
dt ∂z dt ∂t
This equation is marvelous because it tells us the relationship between an ordinary derivative and
the partial derivatives of a function. Comparing the PDE we have with this equation, we see that
along the curves in the tz-plane (known as characteristics) defined by

dz ρg 2
= δ , (1)
dt µ

the PDE simplifies to an ordinary differential equation (ODE),

dδ
= 0. (2)
dt
Integrate both sides of equation (2) with respect to t to solve for δ.

δ = f (ξ),

where f is an arbitrary function and ξ is a coordinate along the characteristics. ξ can be written
in terms of z and t by solving equation (1). Equation (2) tells us that δ is a constant with respect
to time along the characteristics, so equation (1) can be solved by integrating both sides with
respect to t.
ρg 2
z= δ t+ξ
µ
Solve for ξ.
ρg 2
ξ=z− δ t
µ
Thus, the general solution for δ is
 
ρg 2
δ(z, t) = f z − δ t ,
µ

though this is of little interest to us. Since the origin of the coordinate system is chosen so that
the film thickness starts to develop at z = 0, the solution on the ξ = 0 characteristic is the one we
care about.
ρg
0 = z − δ2t
µ
Therefore, r
µ z
δ(z, t) = .
ρg t
The solution only applies above the liquid level: z > 0, t > 0, and z < st. This can be written
compactly as 0 < z < st.

[Link]
BSL Transport Phenomena 2e Revised: Chapter 2 - Problem 2D.2 Page 6 of 6

Figure 2: This is a plot of the film thickness δ vs. z at various times with µ/ρg set equal to 1. The
curves in red, orange, yellow, green, blue, and purple correspond to t equal to 0.1, 0.2, 0.4, 1, 3,
and 20, respectively.

[Link]