- III.

FLUID FLOW -

The Fluid

Viscosity.- There are basically three states of matter - solid, liquid, and gas. H 2O is
commonly called "ice" in the solid state, "water" in the liquid state, and "water vapor" in the
gaseous state. Assume one has a piece of ice and side forces are applied to it (called shearing
forces). Very large forces are needed to deform or break it. The solid has a very high internal
friction or resistance to shearing. The word for internal friction is viscosity and for a solid its
value is generally very large.
Liquids and gases are considered to be fluids since they behave differently from a solid.
Imagine two layers of water or air. If shear forces are applied to these layers, one discovers a
substantial and sustained relative motion of the layers with the air layers sliding faster over one
another than the water layers. However, the fact that a shear force must be applied to deform the
fluids indicates that they also possess internal friction.
Water, under normal temperatures, is about fifty times more viscous than air. Ice is 5 x 10 16
times more viscous than air. One concludes that, in general, solids have extremely high
viscosities whereas fluids have low viscosities. Under the category of fluids, liquids generally
possess higher viscosities than gases. Air, of primary interest in aerodynamics, has a relatively
small viscosity, and in some theories, it is described as a perfect fluid-one that has zero viscosity
or is "inviscid." But it will be shown that even this small viscosity of air (or internal friction) has
important effects on an airplane in terms of lift and drag.
Compressibility.- All fluids are compressible (that is, density increases under increasing
pressure) to some extent, but liquids are generally highly incompressible Compared with gases.
Even gases may be treated as incompressible provided the flow speeds involved are not great.
For subsonic flow over an airplane below about 150 m/sec, air may be treated as incompressible
(that is, no change in density throughout the flow). At higher speeds the effects of
compressibility must be taken into account

The Flow

Pathlines and streamlines.- A fluid flow may be described in two different - the Lagrangian
approach and the Eulerian approach. From the Lagrangian standpoint, one particle is chosen and
it is followed as it moves through space with time. The line traced out by that one particle is
called a particle pathline. An example is a transmitting ocean buoy shown in figure 20(a). Its
position has been marked at 6-hour intervals over a period of several days. The path observed is
the particle pathline.
In order to obtain a clearer idea of the flow- field at a particular instant, a Eulerian approach
is adopted. One is looking at a "photograph" of the flow. Figure 20(b) shows the surface ocean
currents at a particular fixed time. The entire flow field is easily visualized. The lines comprising
this flow field are called streamlines.

26
 (a) Particle airline.
(b) Streamlines.

Figure 20. Particle pathline and streamlines.

It is important to note the differences between a particle pathline and a streamline. A

pathline refers to the trace of a single particle in time and space whereas a streamline presents the
line of motion of many particles at a fixed time. The question of whether particle pathlines and
streamlines are ever the same is considered next.
Steady flow compared with unsteady flow.- Of basic importance in understanding fluid
movements about an object is the concept of a "steady flow." On a windy day a person calls the

27
wind steady if where he stands it blows constantly from the same direction at a constant speed.
If, however, the speed or direction changes, the wind is "gusty" or unsteady. In a similar manner
the flow of a fluid about an object is steady if its velocity (speed and direction) at each point in
the flow remains constant - this does not necessarily require that the velocity be the same at all
points in the fluid.
To consider this further, figure 21(a) presents the fluid flow (of air) about a house on a
windy day at one instant of time and figure 21(b) shows the flow an instant of time later. One
sees that this flow is unsteady. There are many areas where the flow pattern is different; the
streamlines are changing their position and shape with time. Particle pathlines and streamlines
for this flow are not equivalent.

(a) Streamlines at time t0.

(b) Streamlines at time t1.

Figure 21. Unsteady flow of air about a house.

Figure 22 shows a nicely "streamlined'' body (as opposed to the bluff-shaped house) in a
wind tunnel. At time to the tunnel is not running and no air is flowing. At time t 1 the tunnel is
started and air begins flowing about the body; the flow develops further at time t 2 and finally
reaches a constant pattern at time t 3. The flow appears unchanged at time t 4 and time t5. When the
flow starts. it passes through an unsteady transient state; that is. particle pathlines and streamlines
are not the same. From time t3 onwards a steady flow is established. Streamlines appear fixed in
position with respect to the body. A particle P shown on a streamline at time t 3 moves

28
downstream along that streamline as shown at times t 4 and t5. The particle pathline coincides
with the streamline.

Figure 22. Unsteady and steady flout

Summarizing, this means that for a steady flow a particle pathline and streamline are
equivalent and the Lagrangian point of view is the same as the Eulerian approach for flow
visualization.
Rotational and irrotational flow.- Fluid flow can be rotational or irrotational. Ii the elements
of fluid at each point in the flow have no net angular (spin) velocity about the points, the fluid
flow is said to be irrotational. One can imagine a small paddle wheel immersed in a moving fluid
as in figure 23(a). If the wheel translates without rotating, the motion is irrotational. If the wheel
rotates in a flow, as illustrated in figure 23(b), the flow is rotational.

29
 (a) Irrotational flow.

(b) Rotational flow.

(c) Inviscid, irrotational flow about an airfoil.

Figure 23. Rotational and irrotational flow.

According to a theorem of Helmholtz, assuming zero viscosity, if a fluid flow is initially

irrotational, it remains irrotational. In figure 23(c), an observer is fixed to the airfoil section
shown. The flow far ahead of the airfoil section is uniform and of [30] constant velocity. It is
irrotational. As the airflow passes about the airfoil section, it remains irrotational if zero viscosity
is assumed. In real life, viscosity effects are limited to a small region near the surface of the
airfoil and in its wake. Most of the flow may still be treated as irrotational.
One-dimensional flow.- A simplifying argument often employed to aid in understanding
basic ideas is that of a one-dimensional flow. Figure 24(a) shows a bundle of streamlines of a
simple flow. Each streamline can be thought of as a stream tube since fluid flows along it as if in
a tube. In the case of steady flow, the stream tube is permanent. Taken together, the bundle of
stream tubes comprise an even larger stream tube. Fluid flows through it as, for example, water
flows through a pipe or channel. The velocity varies across the tube, in general, according to the
individual streamline velocity variation, as shown in figure 24(b). One can easily imagine an
"average" uniform value of velocity at the cross section to represent the actual varying value as
indicated in figure 24(c). The velocity then is considered "one dimensional" since it varies only
with the particular distance along the tube where observations are made. In addition to velocity,

30
pressure, density, temperature, and other flow properties must also be uniform at each cross
section for the flow to be one dimensional.

(a) Stream tubes.

(b) Real velocity flow profile.
(c) One-dimensional flow profile.

Figure 24. Stream tubes and one-dimensional flow.

31
 In order to understand how aerodynamic forces arise, two basic principles must be
considered. They are the laws of conservation of mass and conservation of energy. Simply
stated, they convey the facts that mass and energy can neither be created nor destroyed.
For introductory purposes, simplifying assumptions are made. The fluid is considered to be
inviscid and incompressible (and hence, "perfect"). The flow is considered steady and one
dimensional.

Ideal Fluid Flow

The continuity equation.- The continuity equation is a statement of the conservation of mass
in a system. Consider a pipe which is uniform in diameter at both ends, but has a constriction
between the ends as in figure 25(a). This is called a venturi tube. Furthermore, it is assumed that
the fluid, under the previously stated assumptions, is flowing in the direction indicated. Stations
1 and 2 have cross-sectional areas A 1 and A2, respectively. Let V1 and V2 be the average flow
speeds at these cross sections (one-dimensional flow). A further assumption is that there are no
leaks in the pipe nor is fluid being pumped in through the sides. The continuity equation states
that the fluid mass passing station 1 per unit time must equal the fluid mass passing station 2 per
unit time. In fact, this "mass flow rate" must be the same value at any cross section examined or
there is an accumulation of mass- "mass creation"- and the steady flow assumption is violated.
Simply stated,

(Mass rate)1 = (Mass rate)2 (1)

where

Mass rate = Density x Area x Velocity (2)

This equation reduces to

pl AlV1 = p2A2V2 (3)

Since the fluid is assumed to be incompressible, p [Greek letter rho] is a constant and
equation (3) reduces to

AlV1 = A2V2 (4)

This is the simple continuity equation for inviscid, incompressible, steady, onedimensional
flow with no leaks. If the flow were viscous, the statement would still be valid as long as average
values of V1 and V2 across the cross section are used.
By rearranging equation (4), one obtains

32
 V2 = (A1/A2)V1 (5)

Since A1 is greater than A2 (see fig. 25(a)), it can be concluded that V 2 is greater than V1.
This is a most important result. It states, under the assumptions made, that the flow speed
increases where the area decreases and the flow speed decreases where the area increases. Figure
25(b) shows this with the longer arrow at the constriction indicating a larger flow speed than at
the ends. In fact, by the continuity equation, the highest speed is reached at the station of smallest
area. This is at the narrowest part of the constriction commonly called the throat of the venturi
tube.
The fact that the product AV remains a constant along a tube of flow allows an
interpretation of the streamline picture. Figure 25(c) shows the streamline pattern in the venturi
tube. In the area of the throat, the streamlines must crowd closer together than in the wide part.
Hence, the distance between streamlines decreases and the fluid speed increases. The conclusion
is that, relatively speaking, widely spaced streamlines indicate regions of low-speed flow and
closely spaced streamlines indicate regions of high-speed flow.

Figure 25. Venturi tube and continuity principle.

33
 Bernoulli's theorem-the conservation of energy.- Assume a fluid flow which, as before, is
inviscid, incompressible, steady, and one dimensional. The energy in the flow is composed of
several energies. The kinetic energy arises because of the directed motion of the fluid; the
pressure energy is due to the random motion within the fluid; and the potential energy is due to
the position of the fluid above some reference level. Bernoulli's theorem is an expression of the
conservation of the total energy; that is, the sum total of these energies in a fluid flow remains a
constant along a streamline. Expressed concisely, the sum of the kinetic energy, pressure energy.
and potential energy remains a constant.
II it is further assumed that the fluid flow is horizontal (as, for example, airflow approaching
an aircraft in level flight), then the potential energy of the flow is a constant. Bernoulli's theorem
reduces to

Kinetic energy + Pressure energy = Constant (6)

where the constant includes the constant value of potential energy. If one considers the energy
per unit volume, one obtains the dimensions of pressure and Bernoulli's theorem may be
expressed in terms of pressure.
The kinetic energy per unit volume is called dynamic pressure q and is determined by q =
1/2pV2 where p and V are, respectively, the fluid flow density and speed at the point in question.
The pressure energy per unit volume (due to random motion within the fluid) is the static
pressure of the fluid and is given the symbol p.
The constant energy per unit volume is called the total pressure pt.
Bernoulli's equation reduces to

Dynamic pressure + Static pressure = Total pressure (7)

or

1/2pV2 + p = pt (8)

For rotational flow the total pressure pt is constant along a streamline but may vary from
streamline to streamline as shown in figure 26(a). In an irrotational flow, the usual case
considered for airflow approaching an aircraft, the total pressure is the same constant value
everywhere as shown in figure 26(b).
Bernoulli's equation states that in a streamline fluid flow, the greater the speed of the flow,
the less the static pressure; and the less the speed of the flow, the greater the static pressure.
There exists a simple exchange between the dynamic and static pressures such that their total
remains the same. As one increases, the other must decrease.
Pressure measurement.- Let us now examine how total, static, and dynamic pressures in a
flow are measured. Figure 27(a) shows the fluid flow about a simple hollow bent tube, called a
pitot tube after its inventor, which is connected to a pressure measurement readout instrument.
The fluid dams up immediately at the tube entrance and comes to rest at the "stagnation point"
while the rest of the fluid divides up to flow around the tube. By Bernoulli's equation the static

34
pressure at the stagnation point is the total pressure since the dynamic pressure reduces to zero
when the flow stagnates. The pitot tube is, therefore, a total-pressure measuring device.
Figure 27(b) shows the fluid flow about another hollow tube except now the end facing the
flow is closed and a number of holes have been drilled into the tube's side. This tube is called a
static tube and may be connected to a pressure measuring readout instrument as before. Except at
the stagnation point, the fluid is parallel to the tube everywhere. The static pressure of the fluid
acts normal to the tube's surface. Since

(a) Rotational flow. Total pressure varies from streamline to streamline.

(b) Irrotational flow. Total pressure same constant value pt,1 everywhere in flow.

Figure 26. Total-pressure variation.

pressure must be continuous, the static pressure normal to the holes is communicated into the
interior of the tube. The static tube, therefore, with the holes parallel to the flow direction, is a
static-pressure measuring device.
Figure 27(c) shows a combined pilot-static tube. When properly connected to opposite ends
of a pressure measuring readout instrument, the difference between total pressure and static
pressure is-measured. By Bernoulli's equation this difference is the dynamic pressure, defined as
1/2pV2. If the fluid density p is known, the fluid flow speed can be calculated. In actual use on
aircraft, the pilot-static tube is connected directly to an airspeed indicator which, by proper

35
gearing, will automatically display the aircraft airspeed to the pilot. The device is sometimes
mounted forward

Figure 27. Pressure measuring devices.

on a boom extending from the airplane nose to insure its measuring, as closely as possible, the
undisturbed approaching flow (also called the free-stream condition).
Returning to the discussion of the venturi tube introduced earlier, the continuity and
Bernoulli equations may be used to describe the static-pressure distribution along the venturi
tube. The static pressure of the undisturbed free-stream fluid flow entering the tube may be used
as a reference value. Any variation of static pressure in the tube then is a greater or lesser value
than the free-stream static pressure. In figure 28 holes have been drilled into the walls of the
venturi tube similar to the static tube of figure 27(b) to measure the static pressure. These holes
are commonly called "static taps" and are connected to a "U-tube manometer" - a tube having a
U-shape within which is a liquid such as colored alcohol. When the static pressure measured at
the static tap equals the free-stream static pressure, the fluid levels in the tube are at some equal
reference level. But static pressures above or below the free-stream pressure are indicated by a
decrease or increase in the level of fluid in the tube.

36
 Figure 28. Venturi tube flow.

Figure 28 shows the complete setup of a venturi tube and a set of manometers and static taps
to measure static pressure. By the continuity equation the speed at station 2 V 2 is greater than
that at station 1 V1 as seen previously-the speed at the throat also is the highest speed achieved in
the venturi tube. By Bernoulli's equation the total pressure p t is constant everywhere in the flow
(assuming irrotational flow). Therefore, one can express the total pressure p t in terms of the static
and dynamic pressures at stations 1 and 2 using equation (8), namely,

1/2plV12 + p1 = 1/2p2V22 + p2 = pt (9)

Since V2 is greater than V1 and p2 = p1 (fluid is incompressible) it follows that p 2 is less than
p1, for as the dynamic pressure, hence speed, increases, the static pressure must decrease to
maintain a constant value of total pressure pt. The block diagrams below the venturi tube show
this interchange of dynamic and static pressures all along the venturi tube. The conclusion drawn
from this is that the static pressure decreases in the region of high-speed flow and increases in
the region of low-speed flow. This is also demonstrated by the liquid levels of the manometers
where as one reaches the throat the liquid level has risen above the reference level and indicates
lower than free- stream static pressure. At the throat this is the minimum static pressure since the
flow speed is the highest.
The airfoil in an ideal fluid.- To supply a point of reference in the discussions to follow of a
real fluid, the following section expands the previous discussion of venturi [38] flow to the ideal
fluid flow past an airfoil. Figure 29(a) shows a "symmetric" (upper and lower surfaces the same)
airfoil operating so that a line drawn through the nose and tail of the airfoil is parallel to the free-
stream direction. The free-stream velocity is denoted by and the free-stream static pressure
by . Following the particle pathline (indicated by the dotted line and equal to a streamline in
this steady flow) which follows the airfoil contour, the velocity decreases from the free- stream
value as one approaches the airfoil nose (points 1 to 2). At the airfoil nose, point 2, the flow
comes to rest (stagnates). From Bernoulli's equation the static pressure at the nose, point 2, is
equal to the total pressure. Moving from the nose up along the front surface of the airfoil (points

37
2 to 3), the velocity increases and the static pressure decreases. By the continuity equation, as
one reaches the thickest point on the airfoil, point 3, the velocity has acquired its highest value
and the static pressure its lowest value.
Beyond this point as one moves along the rear surface of the airfoil, points 3 to 4, the
velocity decreases and the static pressure increases until at the trailing edge, point 4, the flow
comes to rest with the static pressure equal to the total pressure. Beyond the trailing edge the
flow speed increases until the free-stream value is reached and the static pressure returns to free-
stream static pressure. These velocity and static-pressure distributions for the center-line
streamline are shown in figures 29(b) and 29(c).
Note particularly that on the front surfaces of the airfoil (up to the station of maximum
thickness), one has decreasing pressures (a negative pressure gradient) whereas on the rear
surfaces one has increasing pressures (a positive pressure gradient). This relationship will be of
importance in the real fluid case.

Figure 29. Ideal fluid flow about an airfoil.

The lift is defined as the force normal to the free-stream direction and the drag parallel to the
free-stream direction. For a planar airfoil section operating in a perfect fluid, the drag is always
zero no matter what the orientation of the airfoil is. This seemingly defies physical intuition and
is known as D'Alembert's paradox. It is the result of assuming a fluid of zero viscosity. The

38
components of the static-pressure forces parallel to the free-stream direction on the front surface
of the airfoil always exactly balance the components of the pressure forces on the rear surface of
the airfoil. The lift is determined by the static-pressure difference between the upper and lower
surfaces and is zero for this particular case since the pressure distribution is symmetrical. If,
however, the airfoil is tilted at an angle to the free stream, the pressure distribution symmetry
between the upper and lower surfaces no longer exists and a lift force results. This is very
desirable and the main function of the airfoil section.
Air is not a perfect fluid. It possesses viscosity. With slight modification, the continuity and
Bernoulli principles still apply in the real world. The airflow over an airfoil will appear to be
slightly different with an accompanying reduction in lift and the existence of drag in several
forms. The discussions of the past few pages represent basic principles. From this point on, the
inviscid assumption is dropped and a real, viscous flow of air is allowed to exist.

Real Fluid Flow

Laminar and turbulent flow.- There are two different types of real fluid flow: laminar and
turbulent. In laminar flow the fluid moves in layers called laminas. Figure 30(a) shows a laminar
flow, the uniform rectilinear flow, consisting of air moving in straight-line layers (laminas) from
left to right. The laminas may be considered the adjacent streamtubes and then the streamlines
indicate the direction of movement of these fluid layers. Laminar flow need not be in a straight
line. Figure 30(b) shows a small segment of the surface of a curved airfoil. For an ideal fluid the
flow follows the curved surface smoothly, in laminas. Figure 30(c) shows the more complex
flow case for a real fluid to be discussed later. The closer the fluid layers are to the airfoil
surface, the slower they move. However, here also, as indicated by the streamlines, the fluid
layers slide over one another without fluid being exchanged between layers.
In turbulent flow, secondary random motions are superimposed on the principal flow. Figure
30(d) shows a disorganized number of streamlines. They are evidently not fluid layers and there
is an exchange of fluid from one adjacent sector to another. More importantly, there is an
exchange of momentum such that slow moving fluid particles speed up and fast moving particles
give up their momentum to the slower moving particles and slow down themselves. Consider
figure 30(e) which shows the smoke rising from a cigarette. For some distance the smoke rises in
smooth filaments which may wave around but do not lose their identity; this flow is laminar. The
filaments (or streamtubes) suddenly break up into a confused eddying motion some distance
above the cigarette; this flow is turbulent. The transition between laminar and turbulent flow
moves closer to the cigarette when the air in the room is disturbed.
"Another example of a common occurrence of laminar and turbulent flow is the water
faucet. Opened slightly, at low speeds the water flows out in a clear column - laminar flow. But
open the faucet fully and the flow speeds out in a cloudy turbulent column. In a mountain brook
the water may slide over smooth rocks in laminas. In the Colorado River the flow churns
downstream in the confused turbulent rapids. It will be seen that the flow over airfoil surfaces
may assume both a laminar and turbulent characteristic depending upon a number of factors.
In some cases, turbulent flow will appear "naturally" in a laminar flow as in the smoke rising
in the air. In other cases, by causing a disturbance, a laminar flow can be changed to a turbulent
flow. The question arises as to how one can tell whether a flow is to be laminar or turbulent. In

39
1883, Osborne Reynolds introduced a dimensionless parameter which gave a quantitative
indication of the laminar to turbulent transition.

Figure 30. Laminar and turbulent flow.

Reynolds number effects on the flow w field.- In his experiments, Reynolds demonstrated
the fact that under certain circumstances the flow in a tube changes from laminar to turbulent
over a given region of the tube. The experimental setup is illustrated in figure 31(a). A large
water tank had a long tube outlet with a stopcock at the end of the tube to control the flow speed.
The tube was faired smoothly into the tank. A thin filament of colored fluid was injected into the
flow at the mouth.
When the speed of the water flowing through the tube was low, the filament of colored fluid
maintained its identity for the entire length of the tube. (See fig. 31(b).) However, when the flow
speed was high, the filament broke up into the turbulent flow that 'existed through the cross
section. (See fig. 31(c).)

40
 Reynolds defined a dimensionless parameter, which has since been known as the Reynolds
number, to give a quantitative description of the flow. In equation form the Reynolds number R
is

R = pVl/µ (10)

Figure 31. Dependence of flow on Reynolds number. R = (pVl)/µ

where

p density of fluid, kg/m3 [Greek letter rho]

V mean velocity of fluid, m/sec

l characteristic length, m

µ coefficient of viscosity (called simply "viscosity" in the earlier discussion), kg/m-sec

41
 For this setup, Reynolds found, by using water, that below R = 2100 the flow in the pipe was
laminar as evidenced by the distinct colored filament. This value was true regardless of his
varying combinations of values of p , V, l , or µ. A transition between laminar and turbulent flow
occurred for Reynolds numbers between 2100 and 40 000 depending upon how smooth the tube
junction was and how carefully the flow entered the tube. Above R = 40 000 the flow was
always turbulent, as evidenced by the colored fluid filament breaking up quickly. The fact that
the transition Reynolds number (between 2100 and 40 000) was variable indicates the effect that
induced turbulence has on the flow.
The numerical values given for the transition are for this particular experiment since the
characteristic length chosen l is the diameter of the pipe. For an airfoil, l would be the distance
between the leading and trailing edge called the chord length. Additionally, water was used in
the Reynolds experiment whereas air flows about an airfoil. Thus, the transition number between
laminar and turbulent flow would be far different for the case of an airfoil. Typically, airfoils
operate at Reynolds numbers of several million. The general trends, however, are evident. For a
particular body, low Reynolds number flows are laminar and high Reynolds number flows are
mostly turbulent.
The Reynolds number may be viewed another way:

(11)

The viscous forces arise because of the internal friction of the fluid. The inertia forces
represent the fluid's natural resistance to acceleration. In a low Reynolds number flow the inertia
forces are negligible compared with the viscous forces whereas in high Reynolds number flows
the viscous forces are small relative to the inertia forces. An example of a low Reynolds number
flow (called Stoke's flow) is a small steel ball dropped into a container of heavy silicon oil. The
ball falls slowly through the liquid; viscous forces are large. Dust particles settling through the
air are another case of a low Reynolds number flow. These flows are laminar. In a high Reynolds
number flow, such as typically experienced in the flight of aircraft, laminar and turbulent flows
are present. Some very interesting contrasts between the results of low Reynolds number flow
and high Reynolds number flow will be demonstrated shortly.
Surface roughness effects on the flow field. The effect of surface roughness of a body
immersed in a flow field is that it causes the flow near the body to go from laminar to turbulent.
As the surface roughness increases, the point of first occurrence of turbulent flow will move
upstream along the airfoil. Figure 32 illustrates this point. An airfoil surface is shown. In each
succeeding case the degree of surface roughness is increased and the Reynolds number is held
fixed. The flow is seen to go turbulent further upstream in each case. The Reynolds number and
surface roughness are not independent of each other and both contribute to the determination of
the laminar to turbulent transition. A very low Reynolds number flow will be laminar even on a
rough surface and a very high Reynolds number flow will be turbulent even though the surface
of a body is highly polished.
Pressure gradient effects on the flow field.- Another important factor in the transition from
laminar to turbulent flow is the pressure gradient in the flow field. If the static pressure increases
with downstream distance, disturbances in a laminar flow will be amplified and turbulent flow
will result. If the static pressure decreases with downstream distance, disturbances in a laminar
flow will damp out and the flow will tend to remain laminar. Recall that over an airfoil the static
pressure decreased up to the point of maximum thickness. A laminar flow will be encouraged in

42
this region. Beyond the point of maximum thickness (or shoulder of the airfoil) the static
pressure increased. The laminar flow now is hindered and may go turbulent before the trailing
edge.

Figure 32. Surface roughness and flow field. All cases at same Reynolds number.

The boundary layer and skin-friction drag.- The foregoing discussion has provided the
background needed to show how drag is produced on a body immersed in a real fluid flow. An
important aerodynamic force during low-speed subsonic flight is the shear force caused by
viscous flow over the surfaces of the vehicle. This shear force is referred to as the skin-friction
force and is strongly dependent on the factors previously mentioned-Reynolds number, surface
roughness, and pressure gradients. Figure 33 shows that in addition to the pressure forces that act
everywhere normal to a body immersed in a moving fluid, viscous forces are also present. It is
these viscous forces which modify the ideal fluid lift and help create the real fluid drag.

43
 Figure 33. Pressure and viscous forces.

Consider figure 34 which shows a very thin, smooth plate parallel to the approaching flow;
the flow ahead of the leading edge of the plate is a uniform free stream. If the fluid were ideal,
that is, inviscid, the fluid would simply slip over the surface with velocity as shown in figure
34(a). At all points along the surface of the plate, the velocity distribution (that is, the variation
of velocity as one moves perpendicularly away from the surface) would be a uniform constant
value of . No drag would result if the fluid were frictionless (inviscid).
In a real fIuid, however, a very thin film of fluid adheres to the surface. (See fig. 34(b).) This
is the very important no-slip condition. It states that at the surface of a body, point B, the flow
velocity is zero. As one moves away from the body the velocity of the fluid gradually increases
until at some point A the velocity becomes a constant value; in the case of a flat plate this value
is . The layer of fluid where the velocity is changing from zero to a constant value is known
as the boundary layer. Within the boundary layer there are relative velocities between the particle
layers and an internal friction is present. This internal friction extends to the surface of the body.
The cumulative effect of all these friction forces is to produce a drag force on the plate. This drag
force is referred to as skin-friction drag.
Initially, near the leading edge of the plate, one has a laminar flow and the boundary layer
also is steady and layered- hence, a laminar boundary layer. As one moves further downstream,
viscosity continues to act and the laminar boundary layer thickens as more and more fluid is
slowed down by internal friction. Eventually, a point is reached on the plate where the laminar
boundary layer undergoes transition and becomes a turbulent boundary layer. (See fig. 34(b).) As

44
is usual for turbulent flow, there is a random motion in the boundary layer as well as the
downstream directed motion. There is no slip at the surface of the plate. Another important
difference from the laminar boundary layer is the fact that the velocity builds up more quickly as
one moves away from the wall, although the total boundary-layer thickness is greater. This
condition can be seen by comparing the two profiles as shown in figure 34(c). This tendency in a
turbulent boundary layer of the fluid further away from the wall to reenergize the slower moving
fluid near the wall will be shown to produce important consequences.

(a) Inviscid flow along a flat plate.

(b) Viscous flow along a flat plate.
(c) Comparison of laminar and turbulent flow.

Figure 34. Boundary-layer flow in a real fluid.

The Reynolds number has an important effect on the boundary layer. As the Reynolds
number increases (caused by increasing the flow speed and/or decreasing the viscosity), the

45
boundary layer thickens more slowly. However, even though the Reynolds number becomes
large, the velocity at the surface of the body must be zero. Thus, the boundary layer never
disappears.
It is interesting to note that a typical thickness of the boundary layer on an aircraft wing is
generally less than a centimeter. Yet, the velocity must vary from zero at the surface of the wing
to hundreds of m/sec at the outer edge of the boundary layer. It is clearly evident that tremendous
shearing forces (internal friction) must be acting in this region. This gives rise to the skin-friction
drag.
The airfoil in a real fluid. Figure 35 illustrates the real fluid flow over the airfoil surface
originally considered in figure 29. The same free-stream velocity and free-stream static
pressure apply. The flow field ahead of the airfoil is only slightly modified and for all
practical purposes the velocities and static pressures are the same as for the ideal fluid case.
Again a stagnation point occurs at the leading edge of the airfoil and the pressure reaches its
maximum value of pt at this point (total or stagnation pressure). From this point on along the
airfoil, the picture changes.

Figure 35. Real fluid flow about an airfoil. Thickness of boundary layers and wake greatly
exaggerated. Bottom flow along lower surface is the same as on the upper surface.

As noted earlier in the example of the flat plate, a boundary layer begins to form because of
viscosity. This boundary layer is very thin and outside of it the flow acts very much like that of
an ideal fluid. Also, the static pressure acting on the surface of the airfoil is determined by the
static pressure outside the boundary layer. This pressure is transmitted through the boundary
layer to the surface and thus acts as if the boundary layer were not present at all. But the
boundary layer feels this static pressure and will respond to it.
Over the front surface of the airfoil up to the shoulder, an assisting favorable pressure
gradient exists (pressure decreasing with distance downstream). The flow speeds up along the
airfoil. The flow is laminar and a laminar boundary layer is present. This laminar boundary layer
grows in thickness along the airfoil. When the shoulder is reached, however, the fluid particles
are moving slower than in the ideal fluid case. This is an unfavorable condition because the
previous ideal flow just came to rest at the trailing edge. It would appear now, with viscosity
present, that the flow will come to rest at some distance before the trailing edge is reached.
As the flow moves from the shoulder to the rear surface, the static-pressure gradient is
unfavorable (increasing pressure with downstream distance). The fluid particles must push
against both this unfavorable pressure gradient and the viscous forces. At the transition point, the

46
character of the flow changes and the laminar boundary layer quickly becomes a turbulent
boundary layer. This turbulent boundary layer continues to thicken downstream. Pushing against
an unfavorable pressure gradient and viscosity is too much for the flow, and at some point, the
flow stops completely. The boundary layer has stalled short of reaching the trailing edge.
(Remember that the flow reached the trailing edge before stopping in the ideal fluid case.)
This stall point is known as the separation point. All along a line starting from this point
outward into the flow, the flow is stalling. Beyond this line the flow is actually moving back,
upstream toward the nose before turning around. This is a region of eddies and whirlpools and
represents "dead,, air which is disrupting the flow field away from the airfoil. Thus, flow outside
the dead air region is forced to flow away and around it. The region of eddies as shown in figure
35 is called the wake behind the airfoil.

(a) Airfoil upper surface static-pressure distributions.

(b) Ideal fluid airfoil (no pressure drag).
(c) Real fluid airfoil (net pressure drag).
(d) Viscosity effects on an airfoil.

Figure 36. Real fluid effects on an airfoil.

47
 Figure 36(a) compares the ideal fluid case static-pressure distribution at the airfoil surface
and center-line streamline with; the real fluid case. Note that up to the separation point, the
differences are not very large but once separation occurs the pressure field is greatly modified. In
the ideal fluid case the net static-pressure force acting on the front surface of the airfoil (up to the
shoulder) parallel to the free stream exactly opposed and canceled that acting on the rear surfaces
of the airfoil. (See fig. 36(b).) Now, however, in the real fluid case this symmetry and
cancellation of forces is destroyed. The net static-pressure force acting on the front surface
parallel to the free-stream direction now exceeds that acting on the rear surface. The net result is
a drag force due to the asymmetric pressure distribution called pressure drag. (See fig. 36(c).)
This is a drag in addition to the skin-friction drag due to the shearing forces (internal friction) in
the boundary layer. Additionally, the modification of the static-pressure distribution causes a
decrease in the pressure lift from the ideal fluid case.
Figure 36(d) shows figuratively the lift and drag for an airfoil producing lift in both an ideal
and real fluid case. One sees the effects of viscosity - the lift is reduced and a total drag
composed of skin-friction drag and pressure drag is present. Both of these are detrimental effects.
It should be noted, very strongly, that although the previous discussion was limited to an
airfoil section, similar processes are occurring on all the other components of the aircraft to one
degree or another. It is beyond the scope of this text to treat these in detail but the effects will be
noted when the total airplane drag is discussed.
In summarizing this discussion, one observes that the effects of a real fluid flow are the
result of the viscosity of the fluid. The viscosity causes a boundary layer and, hence, a skin-
friction drag. The flow field is disrupted because of viscosity to the extent that a pressure drag
arises. Also, the net pressure lift is reduced. The next section considers the effects of
"streamlining.''
Effects of streamlining.- Figure 37 shows five bodies placed in a real fluid flow of air and
the resultant flow field. Four of the bodies are operating at Reynolds numbers normally
encountered in the flight of subsonic aircraft (R = 10 4 to 105). The fifth body is operating at a
much higher Reynolds number (R = 107).
The flat plate placed broadside to the flow has a large wake with separation points at the
plate edge. A large pressure drag is the result, the skin-friction drag being a relatively small
component. The cylinder. operating at the same Reynolds number, has a smaller wake and the
boundary-layer separation occurs, in this case, before the shoulders of the cylinder. The skin-
friction drag is a little larger in this case than for the plate. but is still smaller than the pressure
drag. Overall, the total drag has been reduced from that of the flat plate; some effects of
streamlining are already evident.
Also, at the same Reynolds number is a streamlined shape. There is almost no boundary-
layer separation and the wake is very small. One may assume then that a streamline shape may
be defined as the absence of 'boundary-layer separation. Operating in the condition shown, the
skin- friction drag now is the dominant component and the pressure drag is very small. Even
more noticeable is the very large reduction in overall drag compared with the cylinder or plate.
This has been accomplished by eliminating the pressure drag since the skin-friction drag has
been increasing only slightly as the bodies became more streamlined. One can explain that the
increase in skin-friction drag is due to the simple fact that the streamlined body has more area
exposed to the flow and thus has a greater area over which the boundary layer may act.

48
 Figure 37. Effects of streamlining at various Reynolds numbers.

Finally, in figure 37 at a Reynolds number of 10 4 is a cylinder approximately 1/10 the

diameter of the streamline shape thickness. Surprisingly it has the same drag as the much larger
streamlined shape. The pressure drag is large because of the turbulent wake. It is not hard to
imagine the reason for the slow speeds of the early biplanes when all the wire bracing used is
considered. A considerably reduced drag could have been realized if the wire were streamlined.
However, the introduction of the monoplane and better structures eliminated the need for this
wire bracing.
The fifth body shown is a cylinder operating in the flow at a much higher Reynolds number
(accomplished by increasing the free-stream velocity). The separation points are downstream of
the shoulders of the cylinder and a much smaller wake is evidenced. This result would lead one
to expect a smaller pressure drag than for the previously discussed cylinder of the same size.
However, the flow speed is higher and the actual drag is much larger. These contradictory facts
may be explained by realizing that actual drag values, under different flow speeds, have been

49
compared. A better measure of the performance is needed. This measure is demonstrated in the
next section to be the nondimensional drag coefficient.
Aerodynamic coefficients.- From everyday experience, consider the factors that determine
the aerodynamic resistance on a body. If one places his hand broadside to a flow outside a car
window at 20 km/hr, little resistance is felt, but if one speeds along at 100 km/hr, the force felt is
considerable. Velocity is one factor that determines the resistance. In fact, considering the flow
problems of subsonic flight (high Reynolds number under relatively small viscosities), the
resistance depends directly on (velocity) times (velocity) or (velocity) 2. In the preceding
example, although the velocity of 100 km/hr is five times that of 20 km/hr, the aerodynamic
resistance is about 25 times as great at the higher velocity.
If one walks along a beach, there is little aerodynamic resistance to doing so. But try to wade
in the water at the same speed. It is considerably more difficult, if not impossible. The density of
water is much greater than the density of air. Density of the fluid represents another determining
factor in the resistance felt by a body.
One more experiment: hold a small piece of cardboard up against a stiff wind. Little
resistance is experienced. Now hold a much larger, similarly shaped sheet of cardboard up
against the same stiff wind. A considerable resistance is felt. Area (or length times length)
exposed to the airflow is another determining factor of resistance.
It is now possible to generalize the discussion by stating that, in the flow of the real fluid,
air, about a body, the aerodynamic resistance is dependent on the size, shape, and attitude of the
body, the properties of the fluid, and the relative velocity between the body and the fluid (air). To
illustrate, consider the lift force defined as the aerodynamic reaction perpendicular to the free-
stream velocity direction. From the previous discussion, lift depends on (size shape attitude fluid
properties. and velocity). For an ideal fluid, the fluid properties (except for density) did not
influence the lift force. For a real fluid, however. viscous, elastic, and turbulent properties are
also important. In addition to the shape and attitude of the body the surface roughness has an
effect on the force. Based on the introductory discussion of this section, it may be demonstrated
that

(12)

where

free-stream fluid density

free-stream velocity

S characteristic body frontal area

l characteristic body length

a [Greek letter alpha] attitude of body

50
 µ coefficient of viscosity

free-stream speed of sound of fluid

(S is a characteristic body frontal area that is usually chosen to be consistent with a series of
comparison experiments. For a cylinder it would be the diameter of the cylinder times its length.
For a wing, however, it is usually taken to be the planform area (chord length times wing span
for a rectangular wing). Thus. it is necessary to check the particular definition of S used for a
body.)

It has previously been shown that the quantity is the Reynolds number or R. Also,

is defined to be the Mach number or M. The Reynolds number is the dimensionless

quantity associated with the fluid viscosity whereas the Mach number is associated with the fluid
compressibility. Surface roughness was shown to have influenced the transition from a laminar
to a turbulent flow. Air turbulence represents the degree of the wake formed past the separation
points. Furthermore, the effects of attitude and shape of a body are lumped together into the
factor. Letting the factor be called K, then,

(13)

The dynamic pressure of a fluid flow was previously defined as 1/2 pV2 so if a value of 1/2
is included in equation (13) and the value of K is doubled to keep the equation the same, 2K may
be replaced by CL. Finally,

(14)

Equation (14) is the fundamental lift formula for usual aircraft flight. C L is known as the
coefficient of lift. The equation states simply that the aerodynamic lift is determined by a
coefficient of lift times the free-stream dynamic pressure times the characteristic body area.
It is very important to realize that the lift coefficient C L is a number dependent upon the
Reynolds number, Mach number, surface roughness, air turbulence, attitude, and body shape. It
is not by any means a constant. C L is generally found by wind-tunnel or flight experiments by
measuring lift and the free-stream conditions and having a knowledge of the body dimensions.
Thus,

(15)

51
 The aerodynamic drag is the aerodynamic resistance parallel to the free-stream direction.
One obtains analogous equations to equations (14) and (15), namely,

(16)

or

(17)

where CD is the drag coefficient, dependent on the previously enumerated parameters.

Figure 38.- Drag coefficients of various bodies.

52
 The moment acting on a body is a measure of the body's tendency to turn about its center of
gravity. This moment represents the resultant aerodynamic force times a moment distance. Let it
be stated that a similar derivation may be applied to the moment equation as used for the lift and
drag equations (14) and (16) such that,

(18)

or

(19)

Cm is the coefficient of moment and an additional characteristic length l is necessary for it to

be dimensionally correct. To reiterate, C L, CD, and Cm are dependent on the Reynolds number,
Mach number surface roughness, air turbulence, attitude, and body shape.
It is now possible to return to the discussion associated with figure 37 and compare the five
bodies by using the force coefficient as a measure of the resistance. The first three bodies
demonstrated the effects of progressively more streamlining. All had the same basic body
dimension d , the same Reynolds number R = 10 5, the same Mach number, and were assumed to
be smooth and alined symmetrically with the flow. The aerodynamic resistance was, therefore,
entirely drag and the drag coefficient C D is a measure of this resistance. By assuming a unit
length for the bodies, the frontal area S is the same for all the bodies as is the dynamic pressure.
By equation (17), CD is then directly proportional to the measured drag. Figure 38 repeats the
results of figure 37 except that now the relative drag force has been replaced by the drag
coefficients CD. At the Reynolds number of 10 5, the CD values for the flat plate, cylinder, and
streamline shape are, respectively, 2.0, 1.2, and 0.12. These values include the combined effects
of the skin-friction drag and pressure drag.
The small cylinder, operating at a Reynolds number of 10 4 with its diameter reduced to one-
tenth the basic dimension of the previous examples, has a C D of 1.2. From equation (16), the
effect of smaller size nullifies the effect of larger C D and the small cylinder and streamline shape
have equivalent aerodynamic drags.
The last cylinder, operating at the higher Reynolds number of 10 7, has a CD of 0.6, that is,
half as large as the cylinders discussed previously. Its aerodynamic drag in figure 37 is large
because has been increased to obtain the higher Reynolds number. The smaller drag
coefficient indicates the effect of the smaller wake and, hence, smaller pressure drag coefficient
component. At high Reynolds numbers, the boundary layer becomes turbulent further upstream
along the cylinder. The turbulence in the boundary layer reenergizes the flow close to the surface
and the fluid drives further along the cylinder against viscous forces and the unfavorable pressure
gradient before stalling. Separation occurs downstream of the shoulders and a smaller wake
results. Compare this condition with the separation and wake at the lower Reynolds number.
Figure 39 is a plot of drag coefficient C D (based on frontal area) against Reynolds number.
The values for each body are shown. Also, the solid line is an experimentally determined curve
of the CD of cylinders tested in wind tunnels. At subcritical Reynolds numbers up to about 10 5,
53
the laminar boundary layer stalls and separates upstream of the shoulders of the cylinder and
produces a very broad wake and high C D values. At supercritical Reynolds numbers from 106
and larger, the laminar boundary layer becomes turbulent and separation is delayed; hence, the
smaller CD values. A rather abrupt transition occurs between Reynolds numbers of 10 5 and 106.
These values are the critical Reynolds numbers.
It is interesting to note that spheres exhibit behavior very similar to that of cylinders. Golf
balls of today are dimpled rather than smooth as they once were, to induce a turbulent boundary
layer and thus decrease their drag coefficient. Much improved driving distances are the result.

Figure 39. Drag coefficients as function of Reynolds number.

The discussion thus far has been rather general and has introduced many important ideas and
principles. Fluid flow behavior has been demonstrated. Numerous references to airfoil or
streamline shapes have been made. Viscous flow of the boundary layer and unsteady flow in the
turbulent wake have been examined. The flow is twodimensional since velocity and other flow
parameters vary normal to the free-stream direction as well as parallel to it. With these ideas in
mind, one may now study aircraft operating in a subsonic flow.

54