flow. Different aspects of flow are described in this
chapter, with two models introduced for dynamic
fluids: the ideal dynamic fluid and the Newtonian fluid.
Both are idealized, as we assume the fluid is incompress-
ible and turbulence free. The flow under these conditions
is called laminar flow,
In the ideal dynamic fluid, molecular interactions are
limited to elastic collisions. Ths yields frictionless motion
of the fluid at stationary walls. Two laws determine the
properties of the resulting flow: the equation of continuity
is an expression of the conservation of fluid mass, and
Bernoulli's law represents the conservation of energy.
‘These laws predict that the flow through a tapering tube
accelerates, and that the pressure in the fluid decreases
with increasing speed
In a Newtonian fluid, the fluid molecules interact
inelastically with each other and with the container walls,
“The equation of continuity still applies, but Bernoulli’ law
‘sno longer sufficient to describe the flow of the fluid. We
develop the concept of viscosity to take into account flow
resistance, In a Newtonian fluid, to forces are present in
the direction ofthe flow: a forward-acting force based on
1a pressure difference along the tube, and a resistance force
that depends on the viscosity of the fluid. In steady state,
a parabolic velocity distribution results in a cylindrical
tube, and the volume flow rate is proportional to the fourth
power ofthe radius of the tube (Poiseuille’s law)
When the properties of a flowing fluid violate the
assumptions made for a Newtonian fluid, non-Newtonian
bbchaviour is observed, most notably as deviations from
Poiseuille’s law. Examples include turbulent flow at high
fluid speeds and velocity-dependent interactions within
the fluid when it isin a mixed phase
‘Two physiological systems in the human body fea-
ture fluid flow: the respiratory system as air streams back
A= that is not in mechanical equilibrium will
CHAPTER 12
Fluid Flow
‘and forth into the lungs, and the cardiovascular system as
blood flows through the pulmonary and systemic citcula-
tions in sequence. We were able to discuss key properties
Of the respiratory system when developing the gas laws,
for the cardiovascular system we need the fluid model
introduced in the previous chapter
12.1: Basic Issues in Blood Flow
‘Our survey of the most important features of the eardio-
vascular system starts with Fig. 12.1, in which we com-
bined an anatomical overview with quantifying its prime
physiological function:
‘= In the systemic circulation, blood is pumped out of|
the left vericle ofthe heart to the capillaries in the
‘organs throughout the body, where it delivers oxygen
and nutrients. Loaded with carbon dioxide, a by-
product of cellular respiration, the blood then returns
{othe heart. From the right atrium it proceeds into the
right ventricle, entering
‘= the pulmonary circulation. From the right ventricle
‘blood is pumped to the capillary bed in the lungs,
‘where carbon dioxide is exchanged for oxygen, From
the lungs, oxygenated blood then returns to the left
atrium to complete a full eyee.
In the systemic circulation, organs are arranged in
parallel to allow the body to prioritize oxygen supply
based on the vital relevance of the organ and the current
metabolic demand. For the major organs, Fig. 12.1 shows
the fraction of the received blood flow, Hy. = AVIAT
is the volume flow rate through an organ, and J, is the
total volume flow rate of the systemic circulation. The
term volume flow rate refers to a fluid volume that flows
through a vessel per unit time. /, can be determined from
CHAPTER 12 | FLUIO FLOW
ausare not confined to particular containers. For example, we
‘want to apply these laws tothe low of air past the wing
of a bird, orto a steam of water falling from a water
tap toward the kitchen sink. Deriving laws in this chapter
is easier when refering toa fuid-confining container or
tube, however, the existence ofthe container or its shape
‘may not matter to the laws" applicability
Inthe easiest ease, the ideal dynamic uid isan ideal
gas. In this case, the third assumption is equivalent to
the respective assumption in the kinetic gas theory we
introduced in Chapter 8, Elastic collisions were studied
in Chapter 7, including results for a collision with
‘wall. This interaction is shown in Fig. 12.6: the com-
ponent of the velocity of the fluid particle parallel to
the wall (|-direction in Fig. 126) remains unchanged,
while the velocity component perpendicular to the
wall (L-rection in Fig. 12.6) changes its sign. In the
kinetic gas theory, a non-zero component parallel 19
the wall allows for collisions with variable angles, for
a flowing ideal gas, the component parallel to the wall
also contain the velocity of the collective flow. Since
this velocity component does not change as result of
an elastic collision, no additional effect occurs due to
the partcle-wall interaction that is not already cov-
cred in the discussion of the ideal gas. This means wall
ieractions are frictionless and can be neglected in
the discussion of flow of an ideal gas. Unfortunately,
though, we saw in the previous chapter thatthe ideal
{gs isnot always a good example for a uid because it
is compressible
If the ideal dynamic fluid is a liquid, neglecting
inelastic collisions is problematic in a dense liquid, the
fluid particles interact extensively with each other and
with the walls. We exclude inelastic collisions in the
Figure 126 asic colison of a fuid particle witha stationary wall
Its vlostycompanent paral to the wal () remains unchanged: its
lacy component perpen tothe wal (1) changes sgn but not
ts magne.
current section regardless, but will be foreed to abandon
this assumption when we introduce Newtonian fluids
12.2.2: Equation of Continuity
We now establish the laws governing flow of an ideal
dynamic fluid. We start with Fig. 12.7, which shows a
fluid flowing from left to right through a tube of varying
The cross-sectional area of the tube
changes from 4, to 4,, with 4, > A,, From experience, we
know that fluid is neither created nor lost along the tube,
‘Thus, during any given time interval, the same amount of |
‘mass of fluid that enters through cross-section 4, (grey
area in upper sketch) must leave through cross-section
(rey area in lower sketch):
Am,
a2y
a
Since the fui is incompressible,
Using p for the density of the fluid, Eq. 12.1] is written
in the form:
‘The volume of a fluid segment in Fig, 12.7 can be
written as the product of the eross-sectional area of the
tube and the segment’ length:
Al,
A
Figure 12.7 Model ofa tapering tube that we ute to denve te emiation
of contin, Fi ow occurs rom ett ight. Top) The hd segment
nay epreseted by te grey area atte lt wth cross-sectional area
‘A; and fd spe v, oton) Late, th fd segment is postoned in
the grey area athe it wth cross-sectional area A, and fow speed v
320__ PART TWO | ENERGY, THERMODYNAMICS, AND TRANSPORT PHENOMENA seWAL is now interpreted as a displacement of the fluid
along the tube per time interval At, This is equivalent to
the speed of the fluid in the tube. With » = |F] = AUAn
we find
Anema
Because Fg, [122] applies between any two points | and 2
along the tube, we ean wnite it in a generalized form:
(12.2)
23)
fate that the rea-
soning that led to Eq. [12.3] also establishes a second
usefl equation for the volume flow rate:
oa
which applies anywhere along the tube, including at the
cross-seetional areas with indexes 1 and 2 in Fig. 127.
Let's check this law carefully. We still need to estab
lish whether it applies to laminar flow and/or to an ideal
dynamic fluid. The two differ in that the former does
not include an assumption about the interactions within
the fluid and with the container wall, If you read once
more through the derivation of the equation of conti-
nuity, you note that the type of interaction plays no role
the mass of the fluid is conserved regardless.
CONCEPT QUESTION 12.1
‘A tube widens from a cross-sectional area A, to
‘cross-sectional area A, = 34,. AS a result, the speed
Cf an ideal dynamic fuid in the tube changes trom v,
10 vp, where
We study the steady flow of water from a water tap, for
‘example, in your kitchen sink. The jet of water (2) broadens
ast fas, (b)narows asit fal, c} does not change ts cross
‘sectional shape, or (slows before hitting the bottom ofthe
‘ink Hint: Neglect effects that could lead to the break-up of
‘a continuous flow into droplet formation,
‘Answer (t). Even though the flow does not occur in a tube,
the equation of continuity applies as long as the flow is
‘continuous. From faucet o sink, the water accelerates due to
‘graviy. Wit the spoed ofthe fd increasing and the volume
flow rte constant, the cross-section ofthe uid must minis,
EXAMPLE 12.1
(6) Ifwe assume that the
blood passes through al systemic capilaris in our body
in series, how fast would it have to flow through each cap-
itary? Use data from Fig. 12.3. Would this result make
‘sense? (6) What is the speed of blood ina capilary # we
Instead assume thatthe blood flows in parallel through the
‘systemic capilaries inthe human body?
Solution to prt (The amount of ood Nowing through
the aorta per minute comeepond toa volume flow rate ct
ae
ro
Soliton opto The ameter of te ara even
nF. 125 and 26 cy, which eae oan outer cee”
sectional area A = m(d/2)' = 6.3 om? (also shown in the
figure). The inner diameter defines the lumen, which is the
ee ee te ee le ee
aeqelerot he Reed weal ote aa fea
teen Fs eccor The Reson cl cial srt oi
ted the Dood vase wal fos betwoon 189% and 20%:
thus, typical ner ametr of the ata
Ts bate tthe Gee -ocktel oa oto are
‘The speed of the blood inthe aorta fs obtained from.
Its inner cross-sectional area and the volume passing per
‘second, Using Ea, [12.4], we obtain:
CHAPTER 12 | FLUID FLOW3ez
Tis isa frequently used result: blood fows through the
‘ora at an average speed of about 20 ors.
Solution to part (C: Let's assume blood passes
‘through each single systemic capilay atthe ate found in
par (@) For the outer clamatr ofa capilary we use 9m
from Fig. 123. This value leads to an inner ctamete of
Alapiay = 7 mE (eapitary wal thickness fs about 1 pi).
‘Te coss-sectonal area ofthe capilary is:
Coher
Aspe,
‘We use the equation of continuity to derive the speed
for blood inthe capil, Iie:
A weal sal
laptan|
FRegtnl =
Even if the capltaries could sustain blood rushing through
at such speed, it would no longer be possible to exchange
‘oxygen and nutrients with the surrounding tissue; that is,
‘the physiological purpose of the cardiovascular system
would be lost.
Solution to part (€) A slow flow of blood in the sys
‘temic capilares is achieved by arranging them parallel to
each other with a combined cross-section that is larger
than the cross-section of the aorta. Fig. 12.3 suggests
that this isthe case with 3500 em forthe capiliries, com-
pared to 5.3 cm? for the aorta. The equation of continuity
allows us to determine the actual speed of blood in the
capillaries once we have corrected the cross-sectional
‘areas from Fig. 12:3 to represent the lumen. A correction
‘actor kis defined asthe rato ofthe lumen cross-sectional
‘area to the outer erose-sectional area, and ie quantiied
with the inner and outer diameter for atypical capillary:
‘Again, thie i a frequently used value: blood flows very
slowly through the capilares, at less than 1 mvs,
ae illustrated
‘experimentally with an instrument ‘called Venturi meter,
‘which is shown in Fig. 12.8. The instrument consists of a
tube with a constriction zone at its centre. A W-shaped tube
‘connects a the left and right tothe wide sections of the main
‘ube; in the middle itis open to the constriction zone. The
\W-shaped tube is partially filled with aliquid to indicate the
pressure in the section of the horizontal tube above each
‘column Initially, while the fluid inthe main horizontal tube
is ates, the liquid in the W-shaped tube is atthe sume level
inall three columns, incieating thatthe pressure isthe same.
‘When the fluid flows through the main tube, a pressure di
ference is observed: the liquid in the middle column rises
highest, indicating that the pressure in the constriction zone
tis named after Giovanni
Battista Venturi (1746-1822), an Ktalian physicist.
Daniel Bemoulli (1700-1782) quantified this obser
vation starting from Fig, 12.9. Shown is a horizontal tube
that tapers from cross-section 4, to 43, We assign pressure
1p; to the wider section of the tube and p, to the narrower
section. With these definitions, the grey fluid segment
Figure 128 The Versi meer san istrunent to measure the speed
‘of a fluid in foal ibe. Top tincicates the same pressure in every
‘secon ofthe ube wh he di a est. tom) When the fd ws
2 indcaed by the atoms ine tube, pressure variatons become ev
‘dent the pressure higher where the speed afte fi Is tower.
PART TWO | ENERGY, THERMODYNAMICS, AND TRANSPORT PHENOMENA et.Fluid ow
>
LA 3
~ fui] ®
Figure 12.9. A sketch defining the parameters needed to derive
Berul’s law. We study a fuid segment (rey) tat inal occupies
the volume A, op and lr the volume A. (bottom). Fi ow
Inthe sketch occurs fom left o ight trough tapering ube. The fd
‘rsaur vais tom p to atthe constcbn, The change in the speed
of he fui causes a change in he kntic energy tat Is accounted for
‘yy awork tem associated wth the rans of id note constriction.
is studied. We want to determine its kinetic energy at
4n initial and a final instant, and then relate the change
in the kinetic energy to the work needed to move it into
the constricted section. We specifically choose the initial
time (index 1) when the fluid segment occupies the grey
volume shown in the top part of Fig. 12.9. At that instant
the volume of the fuid segment is AV’= 4, Ax. The final
instant is shown in the bottom part of Fig. 12.9 (index 2),
Now its volume is AV'= 4,A.x,
‘The cross-sectional area and the speed of the fluid are
related by the equation of continuity. A force F, has to be
applied to accelerate the fluid segment through the tube.
The change in the kinetic energy of the fuid segment is
Mig = Ame} ~ 3m,
in which Am is the mass ofthe fuid segment. With p the
density ofthe fluid, we rewrite the mass as Am = pAV:
1
My = 503 — DAV,
Since the tube becomes narrower, the speed must
imerease, and thus the kinetic energy of the fluid seg-
ment increases. To achieve this increase in kinetic energy,
work must be dane on the uid segment. This work is
required to transfer the uid segment from its initial to its
final position. In a gedanken experiment, we can split the
Work into two contributions: removing the uid segment
in the part of the tube with pressure p,, and adding the
fluid segment to the part of the tube with pressure
Quantitatvely, this means thatthe volume ofthe segment
is changed from AV’to 0 in the top part of Fig. 12.9, and,
concurrently, its volume is changed from 0 to AV in the
bottom part. "The work is:
W= ~pAV~ p\(-a)
= -(r- p)av
The conservation of energy for the fluid segment
requies that AU = IV, because no heat exchange takes
place, The only form of energy that changes in Fig. 12.9
‘isthe kinetic energy, thus AU = AF;
1 a yaar
(2 — PAV = pr} — VAY 125)
We separate all the terms related to positions 1 and 2 in
the equation and divide by AV"
aie r eR (26)
Eq. [12.6] applies at any position along the tube:
p+ bps? = cmt «27
‘This is Bernoulli's law
In the above derivation, we assumed thatthe fuid
flows in a horizontal tube where there is no change in
height. A more general form of Bernoulli's law, which
can accommodate changes in height, canbe obtained by
modifying the above derivation. Specifically, when the
height varies from =, 10 the change in energy should
include the change im potential energy. Tha is, term
pele) ~ =)AM should be added to the right-hand side
or (125)
~ (ps ~ ppAV'= dole} — »DAP + pels —
(28)
Then, Eq, [12.6] becomes
Aedpi en l=Dpe mene) «29
Finally, Eq, [12.7] becomes
CHAPTER 12 | FLUID FLOW
(12.10)
323324
‘This is the complete form of Bemoulli’s law: However,
asa common approximation, when the variation in height
is within a relatively small range, the term pgz is often
ignored. Such an approximation is particularly useful
when the details of the variation in height are irclevant,
unknown, or complicated, as ilustrated in Case Study 12.2
and Example
eee on limiting the interactions of the
fiuid particles withthe container wall to elastic collisions
is necessary to apply the conservation of energy in the
form used to derive Bemoulli’s law. When we revise this
condition in the next section, we must revisit the calcula
tion of the pressure in the flowing fluid
eae Sn eee
‘A blood vessel of radius r spits into two smal vessels,
‘each with racius 1. the speed of the blood in the lage
YOSSI ng: Whats the speed ofthe blood in each ofthe
‘smaller vesSa Vang)? Tet blood as an ideal dynamic fd.
9) Yat ee
©) erat * Vinge
(0 rat = Vege
(0 Wat = Meryl
(© Maat ™ Mer!
GEE
(@) In person with advanced arteriosclerosis (artery con-
‘Sriction due to accumulated plaque onthe inner wal, a8
shown in Fig. 12.10, Bernoulls effect produces a symptom.
called vascular flute. To maintain a constant volume flow
rate inthis situation the blood must travel faster than normal
through the constriction. At a suficiently high blood speed,
the artery collapses and Immediately reopens, leading to
{8 repeated temporary interruption ofthe blood flow that
‘can be heard with stethoscope. Why does vascular fer
‘occur? (b) An aneurysm is a weakened spot of an artery
where the artey walls balloon outward; Fig. 12.11 shows,
‘an aneurysm ofthe aorta. Blood flows more slowly through
this region, resulting in an increas in pressure atthe aneu-
rysm relative to the pressure in adjacent sections of the
‘artery This condition is dangerous because the increased
Figure 12.10. Advanced arteriosclerosis is an artery constcton due
to accumlates ptacue onthe ner vessel walls. Stown sa coronary
artery cross-section vith atharosiertc plaque (yell) nthe lumen.
Figure 12:11 An ast’ rendering ofan aneurysm ofthe aorta The
Inner tye ofthe aorta wall nk has ruptured. Blood (ark ed) has
led inthe fissure of the wal producing the vile bulge. Surgical
repair ofthe sort i requires.
pressure can cause the artery to rupture (see Chapter 13 for
detailed discussion of aneurysms) What siows blood flow
ian anouysmn?
‘Solution to part (a) The artory collapses since the high
‘9p2ed ofthe blood Inside the vessal lowers the pressure in
the bloodstream relative to the pressure in the stationary
‘extracellular fd, This is due to Bemoull's law: a high
value forthe speed, v, leads to a low value of the pres-
‘sure, p. Once the pressure diference is large enough to
‘lose the artery, the blood flow stops momentariy. When
‘this happens, the blood upstream from the clogged vessel
‘causes a pressure increase that is sufficient to reopen the
artery. The closing and reopening ofthe artery then con-
tinues in a cycte manner.
‘Solution to part (b) The cross-sectional area of a blood
‘vessel and the speed of blood in the vessel ar related
by the equation of continuity. Thus, the speed of blood
flow decreases in a blood vessel when its cross-section
increases in an aneurysm.
PART TWO | ENERGY. THERMODYNAMICS, AND TRANSPORT PHENOMENA me‘EXAMPLE 12.2
Blood flows smocthly through the aorta as its cross-
‘section tapers to 75% of its intial value, similar to the
‘case ilustrated in Fig. 12.9. What is the pressure ctf
lence Ap between the wide and the narrow sections? Hint:
Data we used before and required for this calculation are
‘the volume flow rate inthe aorta AV7As = 83 em?/s, the
ensity of blood p = 1.06 g/em®, and the lumen cross~
Section ofthe aorta yyy = 3.8 om?
‘Solution
‘We start with Bernoul's law, Let index 1 in Ea [12.6] refer
1 the wide section and index 2 to the narrow section of
the aorta. From the equation of continuity we know that
the flow Is faster in the narrow section; that is vz >
Inserting tis inequality in Eq, [12.6] leads to p> p. Thus,
We predict thatthe pressure drop from saction 1 to 260.
tion 2. This must be taken into account when writing the
‘pressure diferance in the form Ap = p — pn this form,
‘4p > 0. Note that we could have chosen to define Ap as
‘P:~ Py whlch case Ap would be a negative value.
Eq, [12.4] allows us to quantiy the two speeds, »,
and r, sine® we know how the two cross-sectional areas
‘are related: A, = 075.4, Thus:
ene eazy
— (avian?
we quae
a
which leads to
lavas?
wna
‘Toe given values are substituted next:
12.3: Flow of a Newtonian Fluid
Up to this point, fluid low was discussed for the ideal
dynamic fluid. We established two laws: the equation of
continuity and Bemoulli’s law. We now test how closely
predictions based on these laws correlate with expe
‘mental observations, The experiment we use is illustrated
in Fig, 12.12: a liquid flows through a horizontal tube
from left to right (direction of arrow), Three smaller, ver~
tical columns are placed at different positions along the
tube. The height to which the liquid rises in each of these
‘columns depends on the pressure in the flowing liquid
below If that liquid ean be modelled as an ideal dynamic
‘uid, al three columns have to be equally high, as shown
in Fig. 12.12(a). This prediction results from the laws
‘of the previous section: the equation of continuity states
that the luid speed does not change for a constant cross-
section of the tube, then Bernoulli's law states that the
pressure in the fluid does not change either.
«
cy
Figure 12.12 Comparison of (a) an ideal dynamic id and () a
Newtoran fi owing rough a brzonal ibe. Flow resistance leads
‘pressure crop song te tube as indexed by he lower columa heh
ofthe Md above the ube a eign pat)
The actual experimental result for a real liquid is
shown in Fig. 12.12(b) The farther the liquid progresses
along the tube, the shorter the vertical columns. Thus, the
‘experimental esult differs fundamentally from the predic=
tion in Fig, 12.12(a: the liquid speed andor the pressure
along the tube must change; that is, the equation of conti-
nuity and/or Bernoullis law must be modified. The equa-
tion of continuity applies as long as fluid low is laminar
because fluid cannot accumulate in or vanish from the
tube. Therefore, an approach beyond Bemoulli’s law is
needed to quantify the uid pressure
12.3.1: Newtonian Fluid Model
‘The Newtonian fluid model is developed to correctly
describe the observation in Fig. 12.12(b) Ibis derived from
the ideal dynamic fluid model by removing the assump-
tion that was identified as too restrictive: a dense fluid
such as a liquid cannot travel past a solid wall without
‘extensive interactions, We noted these interactions already
forthe ideal stationary fluid in the last chapter, when we
CHAPTER 12 | FLUID FLOW
325326
attributed the capillarity effect to a significant interface
energy term,
How do we formulate a condition to replace the elastic
collision restriction we used in the previous section? In
the natural sciences we usually proceed by conducting.
experiments, In this particular ease, the macroscopic
observation of viscosity provides a promising approach
even though it is not primarily a phenomenon describing,
the interaction between a fluid and a eontainer wal
Res isis illustrated in Fig. 12.13,
which two parallel fluid layers of area 4 are highlighted.
Let’s assume the lower layer is at rest. This could be due,
for example, to close proximity to the resting walls of
the tube. The upper layer, a distance Ay away from the
lower layer, moves with velocity A¥ toward the right
The moving layer encounters a resistance force R that
tries to slow it down. To maintain a constant velocity
Newton’s first law requires the presence of a second
force, Fe, with which the upper layer of fluid is pushed
forward. Thus, if a fluid encounters flow resistance, an
external force must be applied to push the fluid through
the tube, The magnitude of this external force is found
empirically by submerging two parallel test plates into a
resting fluid and moving one plate relative to the other,
12.13 illustrates
a2
Figure 1213 Expense to menue vscasy costes. Two
paral pts ar marod in fda dtanoe yam each tha
The overpass and te uper pate pusod by an extra
force Fa, The estas fre of the i beances the extemal toe,
len conan vest Av fhe upper te ova he it.
PART TWO | ENERGY, THERMODYNAMICS, AND TRANSPORT PHENOMENA
Since the required external force varies further from
fluid to fluid, with smaller forces typically needed in
“Table 12.1 lists vis-
costy coefficients for several fluids,
TABLE 12.
Viscosity coefficients of various uids
Gases
Ne 178x108 25
°, 2.08 x 10-° 25
ie wins °
He 90x10 5
H aaxto* °
#0 98x 108 25
Liquids
4,0 479x109 °
Ho 101x103 [a=
4,0 2ax104 100
(Whole) blood) 23-27 10 7
Blood senm | 16-22 10% 20
Ethanol 419x109 [2
Giycetine 6 20
Note that the third column in Table 12.1 gives the
‘temperature at which the reported values apply. implying
‘that these values change with temperature. Viscosity is
‘one of the transport phenomena like diffusion and heat
‘conduction, which we discussed in Chapter 10.A micro-
scopic model is needed in addition to the phenomeno-
logical law of Eq. [12.11] to describe the temperature
dependence of the viscosity coefficient
Eg, [12.11] establishes that viscosity is a dynamic
effec tha requites velocity gradient perpendicular tothe
(GED In a Newtonian fluid, the inelastic interaction
with the fluid-confining walls causes velocity gradients.
Viscosity replaces the assumption of elastic collisions
required between the ideal dynamic fluid and its con-
fining walls. In Newtonian fuids, the flow is laminar andANO-FR-thick coating of is placed between
fro mitocipe she of SES
‘each. Find the force required to move the microscope
‘slides at a constant speed of/0.10 mis relative to each
‘other. The viscosity coefficient of alycerine is found in
Table 12.1.
Solution
‘This problem is solved with Eq, [12.11], Each ofthe terms:
‘on the right-hand side of the equation is given in the
‘example text. The area is:
[Note that isn't twice this value forthe two faces ofa side
‘bocause only the cross-sectional area entors Eq, [12.11]
‘and the coating
Using 1 =
= 021N
‘This isa notable force, given the rather wide 1 mm sopara-
tion between the sides. Imagine you reduce ther separa-
‘ton to 1m. Now a force of F= 210 Nis required; thats, a
‘mass of more than 20 kg has to be suspended from one of
the slides to achieve the stated motion. This phenomenon
‘was used inthe development of adhesive tape.
12. luid Velocity Profile
in a Cylindrical Vessel
In a Newtonian fluid the inelastic interaction of fuid
molecules with a stationary wall causes velocity gradi-
‘ents perpendicular to the flow lines. This leads to a non-
‘uniform velocity profile across the fluid in the direction
Perpendicular to the stationary wall. The actual velocity
profile depends on the shape of the stationary wall; we
confine the discussion in this section to cylindrical tubes,
‘which are physiologically important because they include
blood vessels,
Eq, [12.11] is used to quantify the inclusion of
inelastic collisions of the fluid particles with the stationary
wall. Eq, [12.11] can be used only when the fluid flow has
reached a steady state; thus, steady state is an additional
‘assumption for quantitative predictions forthe Newtonian
fluid. This assumption is reasonable for blood flow in
the cardiovascular system. A more general approach
would require us to replace Eq. [12.11] with a formula
that includes transient fuid accelerations. Recall that we
excluded transient behaviour in transport phenomena,
before, for example, when using Fick’s law for diffusion,
In steady-state flow, two forces act on the fluid in the
tube. In the direction of the motion of the fluid, there is
a force due to a pressure difference along the tube. This
foree pushes the fluid through the tube. The viseosity of
the fluid causes a resistance force acting inthe direction
‘opposite to the direction of motion ofthe fluid. This force
ties to slow the moving fluid. The mechanical equilib-
rium between these two forces varies with the position
in the tube: near the stationary wall, viscosity dominates
and the fluid flows slowly (with the speed vanishing
directly at the wall), whereas toward the centre of the
tube, the force pushing the fluid dominates and the fluid
moves comparably fast. The velocity profile is written
{quantitatively in the form:
(212
in which rane isthe radius ofthe tube, Ap is the pressure
difference along a segment of the tube of length /, and 7
is the viscosity coefficient of the fluid. The term Apt is
‘ constant pressure gradient along the tube, Eq. [12.12]
rovides a parabolic velocity distribution, as shown in
Fig. 12.14. This velocity profile can be demonstrated with
a slow-flowing fluid such as honey. If you open a jar of
hhoney at room temperature and turn it upside down, the
honey travels fastest at the centre of the jar, while honey
near the glass surface won't flow out
Figure 12:14 The steady-state velocity profile of 2 Newtonian fd
owing trough a cynical tbe. The highest sped is reached atthe
Cente ofthe tube Gongest rd arrows, wil thet ayr oct in
cana wah he wal does not nave. Nate that te fgure shows a physical
seth fhe system, whch incu the wal ofthe tbe, nd a diagram
wie veloc axis pointing downward andthe position axis poning
tothe ight,
CHAPTER 12 | FLUID FLOW
327In hospitals and in the fod industry, workers are required
to waoh their hands trequenty and extensively. Why are
elaborate hand-washing procedures required in these
envrorments?
Anower Ea, [1212] shows that uid flow near stationary
surface varshes (7 Of0rF— Fug) Thus, lowing water
to flow past your hands doesnot wash away pathogens and
toxins that ae attached othe skin, where the water fs
stow if at al. Oty extensive rubbing with soap may loosen
these. Tis aways applies; however, hospitals are more
drecty concemed because of he lager numberof patho-
gens and toxins ther employees come in contact wth
Jean Leonard Poiseuille used Eq, [1212] to determine the
fe motivate
his result with a simplified ary
that an average velocity of th
a Me
CV oad
with the proportionality factor
‘7/8 that applies specifically to a cylindrical tube.
Thus, a narrower tube reduces the flow severely, For
‘example, when the diameter of a tube is reduced by a
factor of 2, the flow through the tube is diminished by a
factor of 161
EXAMPLE 12.4
Consider the flow of a viscous fd through a cylindrical
‘ube: the variabies that may affect the volume flow rate
AVTAr have boen identified as the pressure gradient Ap,
the viscosity n, and the inner radius r ofthe tube. Use the
‘method of dimensional analysis introduced in Chapter 1 to
‘derive Poisoulle's law.
We express the volume flow rate in terms ofthe relevant
‘variables using a general equation with exponents to be
determined:
Eyer
“where kis a dimensionless constant. I is straightforward
{0 determine the dimension of each of the quantities in
the equation:
(#]- ur
Fl f-varen
tne [aga] - a=
i= {4}
= (eyo
‘Then, the dimensional equation is:
Wir aera or ve
= [ae PU) -8 aE
quate the exponents on both sides ofthe equation:
o=a+p
3a -2a-p+y
I= 20-8.
‘Solve forthe unknown exponents:
‘The value of the dimensionless constant k can be
determined experimentally. For a cylindrical tube, i
turns out
“Then, we have the exact equation of Poisuil’s law, that
is, Eq, [12.13]
PART TWO | ENERGY, THERMODYNAMICS, AND TRANSPORT PHENOMENA weEq. [12.13] can be generalized for arbitrarily shaped
containers in the form of Ohm's law
ap RE.
Where R is the flow resistance, with unit Pa sim
a2)
(EEE Ohm's law states thatthe volume flow rate of a
Newtonian fluid is proportional to the pressure difference
‘along the tube, and thatthe proportionality constant isthe
flow resistance.
This relates to the everyday use of the word reise
tance since, if he resistance is high, a large pressure dif-
ference leads to only a small volume flow rate. In our
discussion of electric currents in Chapter 18 we will com-
pare F, [12.14] to one of the laws of electricity, which
is also called Ohm's law. We will see then that both laws
ate conceptually the same, except that we study viscous
flow of fluids in the caren chapter and then the flow of
charges in a conductor
For a cylindrical tube, the flow resistance is defined
by Eq, [12.13] I is directly proportional tothe viscosity
coefficient» ofthe fui:
R=— 1. (2.3)
Its important to note that Poseuile’s law cannot be
extrapolated tothe case 9 = 0. In particular, Eqs, {12.12}
and [1213] donot predict an infinite velocity or an ininte
volume flow rate in this ease. Ths interpretation would be
inconsistent with many other laws of physics. Why is that
so? Essentially, the answer is that 9 > 0 isan assumption
inthe derivation of Fg [12.12]- we used mechanical equi-
librium between a fore pushing the fluid forward and a
force holding it back. If] = 0, this equilibrium requires
thatthe pressure gradient along the tbe is also zero. Thus,
substituting the conditions for an ideal dynamic fuid into
qs. [12.12] and [12.13] leads to a division of zero by
zero, which i mathematically undefined
‘A\Nowtonian fd is forced through tube to obtain a cer.
tain volume flow rate (experiment 1). the same fd is
‘thon forced through atube of the same cross-sectional area
‘wth double the ent experiment 2, ow nas the pressure
diference Ap, along the tube changed from the previous
value 3p, if we observe the same volume flowrate?
@) 4p, = 40,
(0 Ap,= 44,
(@ dp, = 840,
(© Ap)= 1649,
What isthe pressure gradient the drop of pressure per
length uni) in the aorta? Assume that blood flows as a
Newtonian fui. Te viscosity coefficient of blood is rom
Table 1211 Not this example
dirs rom Example 12.2 because no change inthe aoe
crose-eecton ie ascumed.
Solution
‘The volume Now ate the aortas AVZN= 83x10"
See ee en meerererin
trove MELEE py Plea’ to
2
12.3.5: Newtonian Fluid Flow with Variable
Tube Size
When ideal dynamic fluids flow through a tube of vari-
able size, we see that the combination of the equation
of continuity and Bernoulli's law is sufficient to deter-
‘mine all luid parameters at any point along the tube. The
fluid pressure in particular varies because of variations in
the speed of the fluid. The associated changes in kinetic
energy are accounted for with work done against the pres-
sure in the fluid
‘Once we move fiom the ideal dynamic fluid to the
Newtonian fluid, that is, once we include interactions of |
the luid molecules with each other and the stationary tube
walls, an additional effect on the pressure inthe fluid has
to be included: to overcome viscous flow resistance to
sustain steady-state flow, a pressure gradient is required
along the tube that leads to a decreasing pressure toward
downstream.
In the next step we return with the Newtonian uid
to the case of variable tube size. More specifically, we
‘identify which role the equation of continuity, Bernoull's
law, and Poiseuille’s law play inthis ease
= Fig. 12.15(b) illustrates why the equation of continue
ity must also apply to Newtonian fluids in tubes with
Variable tube size- with or without flow resistance, no
place exists for either excess fluid to collect or fluid
1 disappear along the tube. Thus, all equations from
the beginning of this chapter up to Eq. 12.4] apply
to Newtonian fluids
CHAPTER 12 | FLUID FLOW
229334
[A detailed view ofthe glomeruius mn Bowman's cap-
‘ule is shown in Fig, 12.18, which is a sketch that shows,
from left to right, two different magnifications. In part a,
the supplying arteriole fe visible atthe top left. It leads
to tangled loops of capilaries—resembling a skein of
‘wool—that are embedded in a capsule and fnally leave
‘8 a blood vessel at the top right. The renal tubule, which
Collects the ftrate, is shown at the bottom where it leaves
‘Bowman's capsule, which therefore serves as the primary
{uid collection container.
Figure 12.18 Detailed sketch of Bowman's capel. (2) lad
‘supply othe capul is shown tthe op andthe renal tbe or >
trate removal a he botiom.() A cross-secton ofthe wall inside
Bowman's capsule is highlighted (green do wih the capilary mem
ano (), te tasement membrane 2 andthe podooytas (nhs
seth, lod is atthe tp and urn is atthe boom.
Fig. 12.180) shows a cross-section through the capi-
lary inside Bowman's capsule. The capilary membrane (1)
's porous, but enclosed by the basement membrane (2).
‘The basement membrane is embraced by podocytes (3)—
cals with arm-ike extensions —leaving slits open fr fig
flow. The pores in the capilary membrane on the blood
‘se are typically 20 nm in diameter, The basement mem=
‘brane at the centre is §0 nm to 80 nm thick and contains
‘pores of 12 nm diameter. Thus, the pores inthe basement
smembcane determine the volume flow rate.
‘The actual value for the pressure drop across the
basement membrane varies between zero and 1.3 kPa to
low the body to regulate the flow using variations in the
blood plasma pressure.
We user... = 6 nm forthe radius ofthe pores in the base-
‘ment membrane; 4p = 1.8 kPa is the maximum pressure
difference across the membrane; 50 nm for the length
(of the pore, which is equivalent to the thickness of the
basement membrane; and 9 = 1.4 x 10-9 N sim? for
the viscosty coefficient. With these values we find from
Poiseul's law the volume flow rate of a single pore:
AV (6 10-%m)4(1.3 x 108A)
Jo nandie he daty titration of 180 L, the two Keaneys
rust have more than 1V = 2 x 10 pores. This number is
‘obtained from the result inthe last equation:
(Bon (i),
22x 10! = 2 x 10!
‘The number of pores per nephron, the unit shown
in Fig. 12.17, s obtained by dividing N by the number of
nephrons in our kidneys, which is about 24 milion neph=
rons. Thus, we need roughly 9 x 10? pores per nephron,
“which ig a number close to 100 milion. We see that the
physiological performance of the kidneys. on a macro-
‘scopic scale is based on the physical properties ofa tre-
‘mendous numberof microscopic functional units,
12.4.1: Beyond Laminar Flow: Turbulence
and Convection
ea ‘Turbulent flow is a superposition
‘of laminar low (discussed above) and vortex formation!
Vortex motion. The different flow patterns for laminar
flow and turbulent flow around a solid cylinder are shown
in Fig. 12.19. Inthe turbulent case, flow lines are not con-
tinuous but terminate or start in vortices
id flow
>
Fluid flow
>
== =
==
B= BE
|
Figure 12:19 Two flow pattems: laminar fw (and tubulen fow
(ght around a soit cinder immersed fi. Noe the vertex forma
tin for turbulent fw.
me ‘This is illustrated in Fig. 12.20, which shows
the volume flow rate as a function of pressure difference
along a given tube, At low pressure differences Ohm's law
applies that is, the volume flow rate is proportional to the
pressure difference, The vertical dashed line indicates the
PART TWO | ENERGY, THERMODYNAMICS, AND TRANSPORT PHENOMENA wepressure difference at which the flow undergoes the tran-
sition from laminar to turbulent flow. AC larger pressure
differences Ohm’s law no longer applies, and increasing
the pressure difference to obtain a volume flow rate
increase beyond the transition is ineffective. Thus, once
the flow has become turbulent, no significant increase of
the volume flow rate ean be achieved
av
a
w
Figure 12.20. Tenston ofthe volume flow rte asa function of res
‘ue ference along a tube (6 curve) fom laminar to turbulent flow
(dashed threshold li). Turtuent fow is characterized by minimum or
‘ogi volume flow inerements tor Ieressing pressure stones.
A semi-empirical threshold numbe foduced
to serine whe fo nin ren tis
called the’ hhonour of Osborne
‘ho proposed it in 1883. The present discus-
geometry, he REMRIGS BBE given as
cylindrical tube,
The Reynolds number is useful when we discuss
systems in which turbulence control is required. For
‘example, turbulence suppression is a design erterion|
for birds’ wings in addition to providing lift and thrust
to overcome air drag, Wings are shaped primarily such
tht air above the wing travels faster than that below. Due
to Bernoullis law, a lower air pressure results above the
wing. The net effect isa lift force sufficient to compensate
the effect of gravity on the bied. The faster the bird flies
through the air the stronger the Lift force, and flapping of
nly one reason why large birds, such as most
birds of prey, use a flap-and-plide ight pattern. Avoiding
flapping or flapping slowly further addresses turbulence:
air is a Newtonian fluid in which a transition from lam-
{nar to turbulent flow occurs at high relative speeds of ait
‘and wing surface, that is, when large velocity gradients
are involved, Ceasing wing motion therefore minimizes
the occurrence of turbulences. Smaller birds, particularly
finches and woodpeckers, use another approach to mini
‘mize flow resistance related to slowing of thet light they
rise on one or two wing beats, then fold their wings to the
body and dart through the air, eliminating turbulent air
‘motion past their bodies at high speed. These birds ean be
identified by their undulating flight pattern because they
need 10 re-establish lift through another few wing beats
after several meires to avoid crashing to the ground
Birds show a good sense for turbulence in many
‘ways, even when they are not airborne, When frigid winds
blow along the seashore in winter, you can see seagulls on
the beach all facing in the same direction. As illustrated in
Fig. 12.21, the birds align their streamlined bodies such
that they offer the least resistance to the oncoming breeze.
'
5
Hy
Figure 12.21 When cold winds bow across he shor, seaguls alin the todas wih the wind oallow alana ail, Minimizing turbulence i
‘esse or these animals asthe a vrtces would ut some of ha faters and allow he body se heat excess,
CHAPTER 12 | FLUIDFLOW 335,336
‘This leads to a laminar flow of air around their bodies,
avoiding the ruling of their feathers due to turbulences
that would expose their body tothe low temperatures and
possibly cause hypothermia
exanptei29
‘aon rowan noe
trachea, and (e) blood flow inthe aorta.
Solution
‘Solution to
‘Solution to part (b): We assume 15 inhalations per
‘minute of 0.50 Leach, For the volume flow rate of ai
2(15 min-)01)
‘An additional factor of 2 ie introduced since each
Inhalation is followed by an exhalation, doubling the
volume flow through the trachea per breath. Using the
diameter ofthe trachea as d = 1.0 om, the average speed
(of air is determined from the equation of continuity:
Tap ee
A Ar 26x 10m
Using the density of ar as p= 12 kg/m?, and the vis-
cosity coaficient as 1 = 2 x 10 N sim? (Table 12.1), we
find Re = 2000, that is, @ value near the threshold to tur
bulent flow. The actual ow is turbulent because the inner
trachea surface Is not smooth. Turbulent low is desired
‘because the inhaled ar must be moistened in the trachea;
‘moistening occurs when dry ai is brought into contact
with the moist trachea wal. This contact fs more efficient
for turbulent flow. Fig. 12.20 then shows why airflow in the
trachea has a Reynolds number close to the leminar-to-
turbulent transition. Once low is turbulent, tle gain in the
volume flow rate is achieved by increasing pressure gradi-
tents along the tube. Thus, operating far into the turbulent
regime would unnecessarily increase the physical work
required for breathing.
‘Solution to part (6: The average speed of blood in the
‘ortais 0.20 mis (see Example 12.1) Using d = 2.2 om for
the inner diameter of the aorta, n= 2.5 x 10-2 N sim for
the viscosity coeticient of blood, and p=1.08 x 10° kg/m?
for the density of blood, we find the Reynolds number
{s Re = 1900, that is, again a value close to the transi-
tion laminar to turbulent. you Keep in mind that blood
‘low into the aorta is pulsatile, wth peak velocities in the
112 mvs range, turbulent flow seems to be favoured.
However, turbulent ow is particularly undesirable in blood
‘vessels since it oreetly diminishes the volume flow rate
for a given pressure graciont. Nature again maximizes the
‘efficiency ofthe aortic blood flow based on Fig. 12.20; it
pushes the volume flow rate tothe greatest possible value
for laminar flow, then develops a way to compensate for
the peak velocities at which blood flow would have to be
‘turbulent: immediately beyond the heart the aorta arches
+180", which allow it to butfer the rushing blood withthe
Windkessel effect. We discuss this effect in Chapter 13
because the elastic response of the aortic wall plays 2
key role in it.
Pathological vasoconstriction (vessels becoming
rarrower due to illness) may cause turbulent blood flow.
‘Based on the equation of continuity, the blood speed (+)
increases inthis case because the heart stil pumps the
‘same amount of blood through the aorta. This leads to @
potentially dangerous increas in the Reynolds number
‘even though the diameter of the blood vessel, is
reduced. To llustrate the net effect, we use the equation
‘of continuity to determine the dependence of the blood
‘speed on the vessel diameter:
=
nd}
Feces aoe)
reduced in the case of pathological vasoconstriction, the
‘overall effect onthe Reynolds number is an increase
na
7
ci
= ied,
const fixe
Re
Lets focus specitialy onthe aorta. Peak blood veloc
‘es inthe aorta can also Increase due to a defective aortic
valve. Fig. 12.22 shows the peak vetoty of biood ejected
‘rom the heart as a function of time fora particular patent.
vim)
‘Time (eae)
Figure 12.22. Peok velocity of tho ejected rom the heat as &
function of time fora particular patient, as measured by Doppler
utrasound. The patents peak bond low voc increased steaiy
uring a 3-year observation period, approaching atreshol at Sm.
‘Once tis threshold is reached, open-heart surgery Is required to
sss a detective heat ve.
PART TWO | ENERGY, THERMODYNAMICS, AND TRANSPORT PHENOMENA we.“The speed of blood is measured by Doppler ultrasound,
{8 diagnostic too! we discuss in Chapter 23. This patients
‘peak blood flow velocty increased steadily during a S-year
‘observation period, approaching a threshold at 5 m/s,
which is considered clincaly the maximum tolerable vale.
‘This value exceeds the value we used to calculate the
Reynolds number for this system by a factor of 25, lus
trating the added tolerance due othe elasticity ofthe blood
‘vessel walls. Data such a8 those shown in Fig. 12.22 allow
the medical team to plan a heart operation and peepare the
patient forthe subsequent treatment over a considerable
time span.
12.4.2: Lepr Newtonian Fluids:
We have so far treated blood as a Newtonian fluid
However, its leads to novel
properties that we cannot explain with the model devel-
‘oped for the Newtonian fluid, We illustrate this point for
the viscosity coefficient of blood 1, as listed in Table 12.1,
and the flow resistance R, which we defined in Eq, [12.15]
AAs that equation showed, both parameters are closely
related in the form R % n, Earlier in this chapter we
defined the viscosity coefficient as a material constant,
which depends only on the temperature. As no depen-
dence on other macroscopic parameters was identified,
blood viscosity inthe cardiovascular system of an endo-
thermic species should be constant. The low resistance R
in turn depends on geometric factors, such as tube length
and tube radius, and the viscosity coefficient 7. Thus, f
blood flow in a particular blood vessel of an endotherm,
flow resistance should be constant as well. Physiological
“observations illustrate, however, that the viscosity of
blood and its flow resistance in a given blood vessel
depend strongly on two additional Factors:
ient of blood (rela-
tive tothe viscosity coeficient of water) asa function
of the hematocrit value
The flow velocity of the blood:
‘This is illustrated in Fig, 12.24, which shows the vol-
‘ume flow rate of a Newtonian fluid (1) and the volume
flow rate of blood (2) as a function of the pressure
difference along a vessel, The viscosity coefficient
affects the flow resistance; itis constant in the case
of a Newtonian fluid (blue curve in Fig, 12.24), but
varies as a function of pressure difference for biood
(ced curve).
> a 30 af foo
Hemannese “t 6
Fur 1223 Te vet otic fr wil td ie othe
erst cot wat act ef De hens we ih
Se promt msorg te vobme faci of Ded es wo
Bind Teseoge act va rns an fem
Tha ero gy ol fi ct coed
be cae res
ae
Figure 1224 Purpingiequted ofc Newtonian and non-eutonan
‘is or example, water and bod) at varios rates of flow trough 2
straight tbe As tho intended volume Now ate inereases, te pressure the
‘ump must produce increases as wel. Note that for Newtonian fui (1)
‘he ope of thetine ean fw rat and pessre is constant. Howeve or
‘ond 2, te flow resitance is very ih at low ow rates but approaches
the ae for Newtonian isa higher flow aes.
consequence of Fi
J. as a dangerous
feedback-loop effect may oceur. We illustrate this feedback
Joop in the context of an anaphylactic shock. An anaphy
lactic shock is an allergic reaction of the body's immune
system in response to a second contact with an antigen
to which the body has become sensitized. An example is
a severe peanut allergy. The initial reaction of the body
is a histamine release that leads to a peripheral vasodila-
i that is, blood vessels: aa Tiere
CHAPTER 12 | FLUID FLOW238
But why does the viscosity change with the flow
velocity? The answer lies in the fact that blood is hetero-
‘geneous. Normally, blood cells are well-immersed in the
blood plasma due to the fact that they are nucleus free,
contain a low viscosity cytoplasm, and have a highly flex-
ible cell membrane. This allows blood to behave like a
‘The red blood cells form a structure resembling a
stack of eoins. Nummulation ereates a highly viscous sus-
pension (mixture of solid in liquid)
‘= Laminar flow: flow tubes (around flow lies) are nt ere
ated inthe flow; they do not intersect or vanish inthe flow
‘# [deal dynamic fd: uid whose flow is lamina The uid
molecules interact only through elastic collisions with con-
fining walls (retonless motion).
‘= Newtonian fluid: a uid whose flow is laminar The fuid
meds mre errcaly wh uh oer tnd wh
Seco caaega oly gal ol
1 VS ao tow i id
tte sa crea
Moos AV patina sates deme aya
toad neta
Where 4 isthe area ofthe plates,» isthe viscosity coef
ficient, and |AVV/Ay is a velocity gradient,
‘= Reynolds number for the transition from laminar to turbu-
lent flow ina eylindrical tube (Re < 2000 is laminar, Re >
2000 is turbulent)
vi
7
re (ste erg peed of te i dis he an
Croft be 9 ven eneloop te
Sv ft fad
Re=
UNITS.
= Volume flow rate AV/AE mls
= Mass flow rate Amis: ka/s
PART TWO | ENERGY, THERMODYNAMICS, AND TRANSPORT PHENOMENA
= Viscosity coeficint nN st
‘= Flow resistance R: Pain?
Laws,
‘= Equation of continuity (id mass conservation) for laminar
fow:
ay
nar const,
where AVAis the volume flow rate isthe cross-sectional
area ofthe tube, and vis the speed ofthe Dud,
‘= Bernoulli's law for an ideal dynamic fluid:
1s
D+ 3p¥? + pa: = const,
‘where p is the pressure is the height, and v isthe speed
of the Mui,
‘= Poisculle’s law for a Newtonian fluid in a cylindrical tube:
4 we
aw,
ar an
here Fag ithe ais othe tube and Api the pressure
Aiference along the length ofthe ube
‘= Ohms law fora Newtonian fluid
av
ap RE,
Where B is the low resistance. The flow resistance in a
«eylindica tube is
MULTIPLE-CHOICE QUESTIONS
MO-12.1. The diameter of tapering tbe decreases from dy t0