491-582_cengel_ch10.

qxd 9/2/09 10:08 PM Page 572

572
APPROXIMATE SOLUTIONS OF NS EQS

1019 For each case, calculate an appropriate Reynolds

number and indicate whether the flow can be approximated
by the creeping flow equations. (a) A microorganism of
diameter 5.0 mm swims in room temperature water at a speed
of 0.25 mm/s. (b) Engine oil at 140C flows in the small gap
of a lubricated automobile bearing. The gap is 0.0012 mm
thick, and the characteristic velocity is 15 m/s. (c) A fog
droplet of diameter 10 mm falls through 30C air at a speed
of 2.5 mm/s.
1020 Estimate the speed and Reynolds number of the
sperm shown in Fig. 1010. Is this microorganism swimming
under creeping flow conditions? Assume it is swimming in
room-temperature water.
1021 A good swimmer can swim 100 m in about a
minute. If a swimmers body is 1.8 m long, how many body
lengths does he swim per second? Repeat the calculation for
the sperm of Fig. 1010. In other words, how many body
lengths does the sperm swim per second? Use the sperms
whole body length, not just that of his head, for the calculation. Compare the two results and discuss.
1022 A drop of water in a rain cloud has diameter D
 42.5 mm (Fig. P1022). The air temperature is 25C, and
its pressure is standard atmospheric pressure. How fast does
the air have to move vertically so that the drop will remain
suspended in the air? Answer: 0.0531 m/s

r, m

FIGURE P1022
1023
Run FlowLab with template Creep_
Reynolds. In this exercise, the Reynolds number is varied
from 0.1 to 100 for flow over a 2  1 ellipsoid. (a) Plot CD
as a function of Re and discuss. (b) Generate a plot of velocity profiles along the y axis (x  0) for the two extreme
Reynolds numbers (0.1 and 100). Compare and discuss.
(Hint: Change the vertical axis scale limits to 0 and 0.05 m
so that the profile shapes are more clearly visible.)
1024 A slipper-pad bearing (Fig. P1024) is often
encountered in lubrication problems. Oil flows between two
blocks; the upper one is stationary, and the lower one is moving in this case. The drawing is not to scale; in actuality, h
 L. The thin gap between the blocks converges with
increasing x. Specifically, gap height h decreases linearly
from h0 at x  0 to hL at x  L. Typically, the gap height
length scale h0 is much smaller than the axial length scale L.
This problem is more complicated than simple Couette flow
between parallel plates because of the changing gap height.

In particular, axial velocity component u is a function of both

x and y, and pressure P varies nonlinearly from P  P0 at x
 0 to P  PL at x  L. (P/x is not constant). Gravity
forces are negligible in this flow field, which we approximate
as two-dimensional, steady, and laminar. In fact, since h is so
small and oil is so viscous, the creeping flow approximations
are used in the analysis of such lubrication problems. Let the
characteristic length scale associated with x be L, and let that
associated with y be h0 (x  L and y  h0). Let u  V. Assuming creeping flow, generate a characteristic scale for pressure
difference P  P  P0 in terms of L, h0, m, and V. Answer:
mVL/h02
L

h0

u(x, y)
h(x)
x

hL

FIGURE P1024
1025 Consider the slipper-pad bearing of Prob. 1024. (a)
Generate a characteristic scale for v, the y-component of
velocity. (b) Perform an order-of-magnitude analysis to compare the inertial terms to the pressure and viscous terms in
the x-momentum equation. Show that when the gap is small
(h0  L) and the Reynolds number is small (Re  rVh0/m
 1), the creeping flow approximation is appropriate. (c)
Show that when h0  L, the creeping flow equations may
still be appropriate even if the Reynolds number (Re
 rVh0/m) is not less than 1. Explain. Answer: (a) Vh0/L
1026 Consider again the slipper-pad bearing of Prob.
1024. Perform an order-of-magnitude analysis on the
y-momentum equation, and write the final form of the
y-momentum equation. (Hint: You will need the results of
Probs. 1024 and 1025.) What can you say about pressure
gradient P/y?
1027 Consider again the slipper-pad bearing of Prob.
1024. (a) List appropriate boundary conditions on u. (b)
Solve the creeping flow approximation of the x-momentum
equation to obtain an expression for u as a function of y (and
indirectly as a function of x through h and dP/dx, which are
functions of x). You may assume that P is not a function of y.
Your final expression should be written as u(x, y)  f(y, h,
dP/dx, V, and m). Name the two distinct components of the
velocity profile in your result. (c) Nondimensionalize your
expression for u using these appropriate scales: x*  x/L, y*
 y/h0, h*  h/h0, u*  u/V, and P*  (P  P0)h02/mVL.
1028 Consider the slipper-pad bearing of Fig. P1028.
The drawing is not to scale; in actuality, h  L. This case
differs from that of Prob. 1024 in that h(x) is not linear;

491-582_cengel_ch10.qxd 9/2/09 10:08 PM Page 576

576
APPROXIMATE SOLUTIONS OF NS EQS

#
c
1 c V L
1 f

and u u 
   0. In this chapter,
r u 2pr
r u
r
we started with the equation for uu to generate expressions
for the velocity potential function and the stream function for
the line source. Repeat the analysis, except start with the
equation for ur, showing all your work.

y or z

uu

Consider an irrotational line vortex of strength  in

f
the xy- or ru-plane. The velocity components are u r 
r
c
1 c

1 f

 0 and u u 
 
. Generate expresr u
r u
r 2pr
sions for the velocity potential function and the stream function for the line vortex, showing all your work.

ur

1067

u
Rotational
symmetry
f

Axisymmetric
body

FIGURE P1063
1064

Show that the incompressible continuity equation for

axisymmetric flow in spherical polar coordinates,

1  2
(r u r)
r r

1 
(u sin u)  0, is identically satisfied by a stream
sin u u u
1 c
1 c
function defined as u r   2
and u u 
, so
r sin u r
r sin u u

long as c is a smooth function of r and u.

1065 Consider a uniform stream of magnitude V inclined
at angle a (Fig. P1065). Assuming incompressible planar
irrotational flow, find the velocity potential function and the
stream function. Show all your work. Answers: f  Vx cos a
 Vy sin a, c  Vy cos a  Vx sin a

a
x

1068 Water at atmospheric pressure and temperature (r 

998.2 kg/m3, and m  1.003  103 kg/m  s) at free stream
velocity V  0.100481 m/s flows over a two-dimensional circular cylinder of diameter d  1.00 m. Approximate the flow
as potential flow. (a) Calculate the Reynolds number, based
on cylinder diameter. Is Re large enough that potential flow
should be a reasonable approximation? (b) Estimate the minimum and maximum speeds |V|min and |V|max (speed is the
magnitude of velocity) and the maximum and minimum pressure difference P  P in the flow, along with their respective locations.
1069
Consider the cylinder flow of Prob.
1068. Run FlowLab with template Cylinder_2D_viscosity,
for which the flow speed and other parameters are set to the
same as those of Prob. 1068. Run both caseswith viscosity
turned on and with viscosity turned offand record the pressure that is calculated by the CFD program at the shoulder
(top) of the cylinder. (a) For each case, print the contour plot
of velocity magnitude (this is the default contour plot that
appears). Compare the two cases and discuss. (b) Compare the
pressure at the shoulder calculated by CFD to that calculated
using the potential flow approximation (Prob. 1068). (c) Discuss the difference between potential flow and flow in which
the viscosity of the fluid is set to zero. Why are these two
approximations not identical?
1070 The stream function for steady, incompressible, twodimensional flow over a circular cylinder of radius a and
free-stream velocity V is c  V sin u(r  a2/r) for the case
in which the flow field is approximated as irrotational (Fig.
y
V

FIGURE P1065

r
a

u
x

.
1066 Consider an irrotational line source of strength V/L
f
in the xy- or ru-plane. The velocity components are u r 
r

FIGURE P1070

491-582_cengel_ch10.qxd 9/2/09 10:08 PM Page 577

577
CHAPTER 10

P1070). Generate an expression for the velocity potential

function f for this flow as a function of r and u, and parameters V and a.
1071 Superpose a uniform
stream of velocity V and a
#
line source of strength V /L at the origin. This generates
potential flow over a two-dimensional half-body called the
Rankine half-body (Fig. P1071). One unique streamline is
the dividing streamline that forms a dividing line between
free-stream fluid coming from the left and fluid coming from
the source. (a) Generate an equation# for the dividing stream
function cdividing as a function of V /L. (Hint: The dividing
streamline intersects the stagnation point at the nose of the
body.) (b) Generate# an expression for half-height b as a function of V and V /L. (Hint: Consider the flow far downstream.) (c) Generate an equation for the dividing # stream
function in the form of r as a function of u, V, and V /L. (d)
Generate an expression# for stagnation point distance a as a
function of V and V /L. (e) Generate an expression for
(V/V)2 (the squared nondimensional velocity magnitude)
anywhere in the flow as a function of a, r, and u.

nice looking streamline plot; 200 intervals and a range of 1

to 1 is recommended for good color contrast.)

Boundary Layers
1073C In this chapter, we make a statement that the
boundary layer approximation bridges the gap between the
Euler equation and the NavierStokes equation. Explain.
1074C For each statement, choose whether the statement
is true or false and discuss your answer briefly. These statements concern a laminar boundary layer on a flat plate (Fig.
P1074C).
(a) At a given x-location, if the Reynolds number were to
increase, the boundary layer thickness would also increase.
(b) As outer flow velocity increases, so does the boundary
layer thickness.
(c) As the fluid viscosity increases, so does the boundary
layer thickness.
(d) As the fluid density increases, so does the boundary layer
thickness.
y

y
V

U(x) = V

Outer flow
x

FIGURE P1071

d(x)

Boundary layer

FIGURE P1074C
1072
Superpose #a uniform stream of velocity
V and a line source of strength V /L at the origin. (a) Generate
#
an equation in the form of r as a function of u, V, and V /L
that will enable you to plot a stream
function of arbitrary
#
value c. (b) For V  5 m/s and V /L  10 m2/s, plot several
streamlines outside of the dividing streamline. (c) Run
FlowLab with template Pitot_static_viscosity with viscosity
turned off. This template calculates flow over a body shaped
somewhat like a Rankine half-body. Plot streamlines and
compare qualitatively with those plotted analytically. (Postcontour-Deactivate-streamlines-Activate-Modify. In the new
window, Edit. Adjust the Color Map properties until you get a

1075C We usually think of boundary layers as occurring

along solid walls. However, there are other flow situations in
which the boundary layer approximation is also appropriate.
Name three such flows, and explain why the boundary layer
approximation is appropriate.
1076C A laminar boundary layer growing along a flat
plate is sketched in Fig. P1076C. Several velocity profiles
and the boundary layer thickness d(x) are also shown. Sketch
several streamlines in this flow field. Is the curve representing d(x) a streamline?

U(x) = V

d(x)
x
Boundary layer

FIGURE P1076C

491-582_cengel_ch10.qxd 9/2/09 10:08 PM Page 580

580
APPROXIMATE SOLUTIONS OF NS EQS

and momentum thickness as functions of d, based on this linear approximation. Compare the approximate values of d*/d
and u/d to the values of d*/d and u/d obtained from the Blasius solution. Answers: 0.500, 0.167

10108 Calculate the value of shape factor H for the limiting case of a boundary layer that is infinitesimally thin (Fig.
P10108). This value of H is the minimum possible value.

10103 For the linear approximation of Prob. 10102, use

the definition of local skin friction coefficient and the Krmn integral equation to generate an expression for d/x. Compare your result to the Blasius expression for d/x. (Note: You
will need the results of Prob. 10102 to do this problem.)
The Blasius boundary layer profile is an exact
solution of the boundary layer equations for
flow over a flat plate. However, the results are somewhat
cumbersome to use, since the data appear in tabular form (the
solution is numerical). Thus, a simple sine wave approximation (Fig. P10104) is often used in place of the Blasius solupy
tion, namely, u(y)  U sin a b for y  d, and u  U for
2d
y  d, where d is the boundary layer thickness. Plot the Blasius profile and the sine wave approximation on the same
plot, in nondimensional form (u/U versus y/d), and compare.
Is the sine wave profile a reasonable approximation?

U(x)

10104

U(x) = V
V
d(x)
x

FIGURE P10104
10105 The streamwise velocity component of a steady,
incompressible, laminar, flat plate boundary layer of boundary layer thickness d is approximated by the sine wave profile of Prob. 10104. Generate expressions for displacement
thickness and momentum thickness as functions of d, based
on this sine wave approximation. Compare the approximate
values of d*/d and u/d to the values of d*/d and u/d obtained
from the Blasius solution.
10106 For the sine wave approximation of Prob. 10104,
use the definition of local skin friction coefficient and the
Krmn integral equation to generate an expression for d/x.
Compare your result to the Blasius expression for d/x. (Note:
You will also need the results of Prob. 10105 to do this problem.)
10107 Compare shape factor H (defined in Eq. 1095) for
a laminar versus a turbulent boundary layer on a flat plate,
assuming that the turbulent boundary layer is turbulent from
the beginning of the plate. Discuss. Specifically, why do you
suppose H is called a shape factor? Answers: 2.59, 1.25
to 1.30

FIGURE P10108
10109 Integrate Eq. 5 to obtain Eq. 6 of Example 1014,
showing all your work.
10110 Consider a turbulent boundary layer on a flat plate.
Suppose only two things are known: Cf, x 0.059 (Rex)1/5
and u 0.097d. Use the Krmn integral equation to generate
an expression for d/x, and compare your result to column (b)
of Table 104.
10111
Boundary layers are generated in other
flows besides forced flows over bodies. For example, boundary layers form along bodies in natural convection flows. Run
FlowLab with template Convection_natural_fluid. In this
exercise, natural convection over a circular cylinder is modeled for two fluids (air and water) with otherwise identical
conditions. (a) Generate temperature contour plots for the
two cases and compare and discuss (File-Print Graphics,
select File and the type of image file desired, type in a file
name, and Accept). (b) Generate plots of velocity and temperature profiles at y  0 along the side of the cylinder (use
the Plot Velocity & Temperature profiles button). Compare
and discuss.

Review Problems
10112C For each statement, choose whether the statement
is true or false, and discuss your answer briefly.
(a) The velocity potential function can be defined for threedimensional flows.
(b) The vorticity must be zero in order for the stream function
to be defined.
(c) The vorticity must be zero in order for the velocity potential function to be defined.
(d) The stream function can be defined only for two-dimensional flow fields.
10113 In this chapter, we discuss solid body rotation (Fig.
P10113) as an example of an inviscid flow that is also rotational. The velocity components are ur  0, uu  vr, and
uz  0. Compute the viscous term of the u-component of the
NavierStokes equation, and discuss. Verify that this velocity
field is indeed rotational by computing the z-component of
vorticity. Answer: zz  2v

491-582_cengel_ch10.qxd 9/2/09 10:08 PM Page 581

581
CHAPTER 10
uu

uu

L
uu = 2pr

uu = vr

FIGURE P10113

FIGURE P10116

10114 Calculate the nine components of the viscous stress

tensor in cylindrical coordinates (see Chap. 9) for the velocity field of Prob. 10113. Discuss your results.

10117 Calculate the nine components of the viscous stress

tensor in cylindrical coordinates (see Chap. 9) for the velocity field of Prob. 10116. Discuss.

10115
As discussed in Chapter 8, the Pitot
formula is derived by neglecting viscous effectsthe flow
everywhere except very close to the walls of the Pitot-static
probe is considered to be an inviscid region of flow. With
CFD, we can compare flows with and without viscosity.
Run FlowLab with template Pitot_static_viscosity, where the
flow of air (r  1.225 kg/m 3) over a two-dimensional Pitotstatic probe is modeled. Run with viscosity on and with viscosity off to compare. Be sure to store the results in the table.
(a) Generate velocity contour plots for the two cases and
compare and discuss. (Post-contour-Activate-Modify: In
the new window, Edit where it says that it is a pressure contour plot, the default. Change DOF to velocity-magnitude
and Apply. Drag the middle- and right-mouse buttons to
move and zoom, respectively, so that you can reposition near
the probe. Save a graphics file of the contour plot by
File-Print Graphics, select File and the type of image file
desired, type in a file name, and Accept.) (b) Calculate the
inferred velocity from the Pitot formula (using the values of
P1 and P2 calculated by FlowLab, and compare to the actual
inlet velocity. Compare and discuss. (c) Generate strea
mline plots for the two cases and compare and discuss.
Specifically, what is different about the streamlines close
to the body? Explain. (Post-contour-Deactivate-streamlinesActivate-Modify: In the new window, Edit. Adjust the Color
Map properties until you get a nice looking streamline plot;
200 intervals and a range of 1 to 1 is recommended for
good color contrast.)

10118
As discussed in Chapter 8, the Pitot
formula is derived by neglecting viscous effectsthe flow
everywhere except very close to the walls of the Pitot-static
probe is considered to be an inviscid region of flow. This
approximation should be reasonable for high Reynolds numbers but is not reasonable for low Reynolds numbers. Run
FlowLab with template Pitot_static_Reynolds where the flow
of air (r  1.225 kg/m3 ) over an axisymmetric Pitot-static
probe is modeled. Run at Re  1 and at Re  5000 to compare. Be sure to store the results in the table. (a) Generate
velocity contour plots for the two cases and compare and discuss. (Post-contour-Activate-Modify: In the new window,
Edit where it says that it is a pressure contour plot, the
default. Change DOF to velocity-magnitude and Apply. Drag
the middle- and right-mouse buttons to move and zoom,
respectively, so that you can reposition near the probe. Save a
graphics file of the contour plot by File-Print Graphics, select
File and the type of image file desired, type in a file name,
and Accept.) (b) Calculate the inferred velocity from the Pitot
formula (using the values of P1 and P2 calculated by
FlowLab, and compare to the actual inlet velocity. Compare
and discuss your results.

10116 In this chapter, we discuss the line vortex (Fig.

P10116) as an example of an irrotational flow field. The
velocity components are ur  0, uu  /(2pr), and uz  0.
Compute the viscous term of the u-component of the
NavierStokes equation, and discuss. Verify that this velocity
field is indeed irrotational by computing the z-component of
vorticity.

10119
The extent of influence of a body moving through a fluid is greater at very low values of Reynolds
number, since viscous effects tend to extend far into the surrounding fluid. We test this with CFD. Run FlowLab with
template Cylinder_2D_Reynolds where the flow of water
(r  998.2 kg/m 3) is simulated over a 2-D circular cylinder.
Run at Re  1 and at Re  100 to study the extent of influence. Generate velocity contour plots for the two cases. In
this template, velocity magnitude contour plot is the default
graphical image displayed after the flow is solved. Save a
graphics file of the contour plot by File-Print Graphics, select
File and the type of image file desired, type in a file name,
and Accept. Compare the two cases and discuss your results.

491-582_cengel_ch10.qxd 9/2/09 10:08 PM Page 582

582
APPROXIMATE SOLUTIONS OF NS EQS

10120
Repeat Prob. 10119 but use FlowLab
template Cylinder_axi_Reynolds, which is the axisymmetric
version of the 2-D cylinder case, i.e., a sphere. (a) Compare
and discuss velocity contours at Re  1 and 100. (b) Compare and discuss velocity contours for the cylinder (Prob.
10119) and the sphere.
10121 Water falls down a vertical pipe by gravity alone.
The flow between vertical locations z1 and z2 is fully developed, and velocity profiles at these two locations are sketched
in Fig. P10121. Since there is no forced pressure gradient,
pressure P is constant everywhere in the flow (P  Patm).
Calculate the modified pressure at locations z1 and z2. Sketch
profiles of modified pressure at locations z1 and z2. Discuss.

z = z2

z = z1

10122 Suppose the vertical pipe of Prob. 10121 is now

horizontal instead. In order to achieve the same volume flow
rate as that of Prob. 10121, we must supply a forced pressure gradient. Calculate the required pressure drop between
two axial locations in the pipe that are the same distance
apart as z2 and z1 of Fig. P10121. How does modified pressure P change between the vertical and horizontal cases?
10123
In this problem, incompressible turbulent flow of water through a converging-diverging duct is
modeled using CFD. Run FlowLab with template
Duct_boundary_layers with inlet velocity set to 5.0 m/s. Plot
boundary layers at four streamwise locations: x  0.0450 m
(near the inlet in the converging portion), x  0.0675 m (near
the throat), x  0.1215 m (in the diverging portion), and x 
0.2025 m (in the diverging portion near the duct exit). To
generate the plots, Compute Results-XY Plots-XY Plot,
select the desired x location, and Plot. To write the data to a
file, File-Export data, type a file name, Export, and OK.
Compare boundary layer thickness and shape and the velocity outside the boundary layer for the four cases. Discuss,
considering whether the pressure gradient at each location is
favorable or adverse. Note: The boundary layers are upsidedown since the wall is on the top of the domain.

Design and Essay Problem

z

FIGURE P10121

10124 Explain why there is a significant velocity overshoot for the midrange values of the Reynolds number in the
velocity profiles of Fig. 10136, but not for the very small
values of Re or for the very large values of Re.