ME 5101

Two-Dimensional Incompressible
Potential Flow Aerodynamic Models
Fall 2024
Dr. Thomas Harris
Adjunct Faculty
Department of Mechanical Engineering
01/27/2025 1
 ME 5101 Road Map
Control Volume Analysis
Differential Form of
Aircraft Performance of Conservation of Mass
Governing Equations
& Momentum

3D Potential Flow 2D Potential Flow Incompressible

Aerodynamic Models Aerodynamic Models Potential Flows

2D Inviscid Boundary Layer

Incompressible Laminar
Compressible Transition and
Boundary Layers
Aerodynamics Turbulence

01/27/2025 2
 Overview
• Governing equations and the velocity
potential
• 2D non-lifting flows
• 2D lifting flows
• Vortex panel methods
• Thin airfoil theory
Governing Equations and the
Velocity Potential
 Agenda
Establish the foundation for potential flows.

For irrotational flows (zero vorticity throughout the

flow) there is a scalar function whose gradient is the
velocity vector,
=
where is the velocity potential.
Governing Equations and the Velocity Potential

• Assume the flow can be treated as:

– Steady: properties do not vary with time.
– Incompressible: constant density.
– Inviscid: viscous stresses assumed negligible.
– Uniform freestream flow: upstream properties
are uniform.
– Irrotational: the vorticity is ‘essentially’ zero
everywhere.
Governing Equations and the Velocity Potential

• The goal is to determine the pressure field and

velocity field .
• Far upstream, the uniform conditions are and .
(5.1)
– where is the angle of attack
• The density is constant and given.
• With the previous assumptions, we can determine
and using the conservation of mass and momentum
equations.
 Governing Equations and the Velocity Potential

• For an incompressible flow the continuity equation is

(5.2)
• If the flow has zero vorticity, the velocity vector can
be written as the gradient of a scalar function.
• From vector calculus, a vector field with zero curl can
be written as the gradient of a scalar field. Thus, we
define a scalar field, , as
(5.3)
Governing Equations and the Velocity Potential

• Substituting Equation (5.3) into Equation (5.2)

yields
0 (5.4)
0 (5.5)
Where is called the Laplacian and defined as
(5.6)
 Governing Equations and the Velocity Potential

• The conservation of momentum equation reduces to

the Bernoulli equation:
(5.7)
• The process for determining and then is:
1. Solve 0 for [Equation 5.5]
2. Determine the velocity from [Equation 5.3]
3. Calculate the pressure from Bernoulli’s equation,
[Equation 5.7]
 Boundary Conditions (1)
• At a solid surface, require that the flow is tangent to
the surface. In other words, the velocity component
normal to the wall is zero,
(5.8)
• Substituting the velocity potential…
(5.9)
at a solid surface.
 Boundary Conditions (2)
• In the far field, as , assume that the flow
velocity in the freestream direction returns to
as (5.10)
where / is the unit vector in the direction of
the freestream.
• In terms of the potential,
as (5.11)
 Modeling Approach
• The potential flow modeling approach is based on the
principle of linear superposition.
• Consider two different potentials 1 and 2 both of which
satisfy Laplace’s equation (mass conservation),
1 0 (5.12) 2 0
(5.13)
• Now add these potentials with an arbitrary weighting,
to define a new potential:
(5.14)
with arbitrary constants and
 Modeling Approach
• The new potential function also satisfies mass
conservation:

• This generalizes to an arbitrary number of potentials:

(5.15)
Constant Velocity Potential Contours
𝜑

What streamline pattern corresponds to this velocity potential

distribution? Does the flow enter, exit, or is it tangent to each
boundary?
Corner Flow (Bertin Example 3.3)
Two-Dimensional Nonlifting
Flows
 Cylindrical Coordinate System

• (5.16)
• (5.17)
The unit vectors in the and directions are:
+ (5.18)
+ (5.19)
 Cylindrical Coordinate System
• The radial and circumferential velocity components
are related to and by:
(5.20)
(5.21)
• The gradient operator in cylindrical coordinates can
be applied to the velocity potential to find and .
•
(5.22)
• (5.23)
 Cylindrical Coordinate System
• The divergence and curl of the velocity vector
in cylindrical coordinates are,
(5.24)
(5.25)

• Finally, Laplace’s equation for in cylindrical

coordinates is,
(5.26)
 Point Source Flow

Our first ‘building block’ 2D potential flow is called a point source:

• (5.27)
• (5.28)
• (5.29)
where is a scaling constant called the source strength.
When
 Point Source Flow
• In Cartesian coordinates, the potential and velocity
components for a point source/sink flow are expressed as:

• (5.30)
• (5.31)
• (5.32)
 Flow Over a Rankine Oval (1)
• We will now develop the potential flow over a shape known
as the Rankine Oval. This will be our first potential flow in
which we combine multiple potentials.
• We will combine a freestream flow at zero angle of attack
with two sources. A point source with positive strength will
be located at and a point source with negative strength will
be located at .
• In order to do this, we must translate the point source
potentials from the origin, as they appear in Equations (5.27)-
(5.29).
 Flow Over a Rankine Oval (2)

• Define a coordinate system about a point as shown

in the figure, where
• (5.33)
• (5.34)
 Flow Over a Rankine Oval (3)
• Using this coordinate system, a source of strength 𝛬 located
at point has the following potential and velocity
• (5.35)
• (5.36)
• (5.37)

Note: these radial and circumferential velocity

components are in the respective directions, not the
radial and circumferential directions about the origin!
 Flow Over a Rankine Oval (2)

• Define a coordinate system about a point as shown

in the figure, where
• (5.33)
• (5.34)
 Flow Over a Rankine Oval (4)
• In Cartesian coordinates, the velocity potential
and velocity components for the translated
sources are:
• 2 (5.38)

• w
• (5.39)
(5.40)
 Flow Over a Rankine Oval (5)
which 𝛬/
• Consider the case in

• Nondimensionalize velocities
by and spatial coordinates by ,

(and .
and place the sources at

• The figure indicates the flow is

emitted from the source at ,
and drawn into the source
(sink) at .
Streamlines for sources of strength 𝛬 = +
• The freestream velocity is
added in the next figure. 4 located at (+1,0)
Complete Flow Over a Rankine Oval (6)
In inviscid flow, any portion of a streamline can
be replaced by a solid surface

^
𝑛 ^
𝑛 Rankine Oval
^
𝑛

Source Sink
 Doublet
• A doublet is a source and a sink brought very close together.
• A doublet can be tricky to understand. Interestingly, it has an
analogous concept in electro-magnetic theory: a dipole, a
magnet with ‘north’ and ‘south’ ends brought very close
together.
• If we add a source and a sink with equal strength at the same
location, we simply get:

• Therefore the source and sink must be brought together

carefully so that the product of their strengths () and their
distance apart () remains constant at a prescribed finite value:
 Doublet
The potential function and velocity components for a doublet (in
both cylindrical and Cartesian coordinates) are:
• (5.41)
• (5.42)
• (5.43)

• (5.44)
• (5.45)
Streamlines for a Doublet
 Flow Over a Nonlifting Cylinder
• The flow over a cylinder is obtained by combining a doublet
with a freestream flow.
• First, we need relationships between the doublet strength (),
the freesteam velocity (), and the cylinder radius ().
• The potential and velocity components for this flow are:
• (5.48)
• (5.49)
• (5.50)
 Flow Over a Nonlifting Cylinder
• On the surface of the cylinder, flow tangency requires that
Evaluating at and enforcing gives the doublet strength in terms
of and .

• Therefore, the potential and velocity field for the flow around a
cylinder of radius in a freestream of velocity are:
(5.51)

(5.52)
(5.53)
Streamlines for Cylinder Flow
 Flow Over a Nonlifting Cylinder
• On the surface of the cylinder where , the
velocity components and magnitude are,
• (5.54)
• (5.55)
• (5.56)
 Flow Over a Nonlifting Cylinder
• Apply Bernoulli’s equation to calculate the surface pressure
distribution on the cylinder:

+ (5.57)
• And the corresponding pressure coefficient at the surface is,
(5.58)
 Surface and on a Nonlifting Cylinder

• The maximum speed is at

the top of the cylinder.
• at the stagnation points and
= at the low pressure apex.
Two-Dimensional Lifting Flows
Point Vortex
Our last ‘building block’ 2D potential flow is called a
point vortex:
• (5.59)
• (5.60)
• (5.61)

where 𝛤 is a scaling constant called the circulation of the

vortex [with units (length)2/time]. The streamlines are circles
about the origin. And the velocity becomes infinite as
 Point Vortex

• A point vortex has zero vorticity everywhere, except at the origin, where
it is infinite.
• This is analogous to a source flow where is zero everywhere, except at
the origin, where it is infinite.
• Examine the vorticity of a point vortex flow by applying Stokes’ theorem
to a 2D velocity field…
(5.62)
where is a contour surrounding an area nd the direction of integration
around is taken so that the area is to the left of .
 Point Vortex Summary
❑

∮⃗
𝑉 ∙⃗
𝑑𝑙=− 𝜞 𝐹𝑜𝑟 𝑎 𝑐𝑜𝑛𝑡𝑜𝑢𝑟 𝑠𝑢𝑟𝑟𝑜𝑢𝑛𝑑𝑖𝑛𝑔 𝑡h𝑒𝑜𝑟𝑖𝑔𝑖𝑛
𝐶

• A point vortex flow is irrotational, except at the origin where it is infinite.

• This is analogous to how is infinite at the center of a point source, and
everywhere else is equal to zero.
• Nevertheless, we can still utilize the point vortex flow to model irrotational
flows because the origin is treated as a boundary, and technically not part of
the flow field.
What would result from combining these
potential flows?

Point Vortex

Flow over a Cylinder

 Lifting Flow over a Rotating Cylinder
• Since the vortical flow does not perturb the radial velocity, we
can add a point vortex to the nonlifting cylinder, and the flow
will still be tangent to the surface.
• This flow produces lift!

• The potential and velocity for the lifting cylinder are:

• (5.63)
• (5.64)
• (5.65)
Lifting Flow over a Rotating Cylinder

𝛤 1 𝛤
= =1
2𝜋 𝑉 ∞ 𝑅 2 2𝜋 𝑉 ∞ 𝑅
 Lifting Flow over a Rotating Cylinder
On the cylinder surface, the velocity components and velocity
magnitude are,
• (5.66)
• (5.67)
• (5.68)

Now determine the location of the stagnation points by

determining the angles stag where .
• (5.69)
• (5.70)
 Stagnation Points on a Rotating Cylinder
• There are 2 stagnation points on the surface as long as . (For
higher values, the stagnation point occurs off the surface.)
• Designate the location of the stagnation points as , so /
• For ,

• The stagnation points are located at

, or
 Stagnation Points on a Rotating Cylinder
• There are 2 stagnation points on the surface as long as . (For
higher values, the stagnation point occurs off the surface.)
• Designate the location of the stagnation points as , so /
• Similarly, for ,

• The stagnation points are located at

, or
Lifting Flow over a Rotating Cylinder

𝛤 1 𝛤
= =1
2𝜋 𝑉 ∞ 𝑅 2 2𝜋 𝑉 ∞ 𝑅
 Pressure Coefficient for Lifting Cylinder Flow

• The velocity magnitude on the cylinder surface is

(5.68)
• From Bernoulli’s equation, the surface pressure is

• The pressure coefficient on the surface is,

(5.71)
 Pressure Coefficient for Lifting Cylinder Flow

• The pressure coefficient on the surface is,

(5.71)
• Integrating the pressure distribution yields:
( (5.72)
(5.73)

• and on the cylinder surface are presented in the next chart for
and on the Cylinder Surface

1
 Surface and on a Nonlifting Cylinder

• The maximum speed is at

the top of the cylinder.
• at the stagnation points and
= at the low pressure apex.
 Circulation
• For the lifting flow on a cylinder, the strength of the point
vortex (𝛤 ) is called the circulation of the vortex and is directly
related to the lift.

• Circulation is a more general concept than just the strength of

a point vortex. The general definition of the circulation is,
𝛤 (5.74)
 Kutta-Joukowsky Theorem
• For an incompressible steady two-dimensional flow with
uniform freestream, the lift on a body is related to the
circulation on a contour surrounding the body using the Kutta-
Joukowsky Theorem:
(5.75)

where 𝛤 is the circulation defined by Equation (5.74 ) for a

contour surrounding the body.
𝛤 (5.74)
 Lifting Airfoils and the Kutta Condition
• For any body, there are actually an infinite number of potential
flow solutions that satisfy the boundary conditions. (This is
demonstrated by the cylinder flows for different levels of
circulation.)
• An infinite number of solutions also exist for airfoils, depending
on the circulation.
• It is the flow behavior at the trailing edge that determines the
physically realistic solution. Circulation levels that allow the
flow to wrap around the trailing edge would require the
pressure to be infinitely low due to the vanishing radius of
curvature (and infinitely high velocities).
• In actual physical flows, the flow at a sharp trailing edge leaves
smoothly, with no flow wrapping.
Lifting Airfoils and the Kutta Condition

=0
Lift Generation and Flow Turning

❑
𝐿=−∬ 𝜌 𝑤𝑣 𝑤𝑢𝑤 𝑑𝑆
𝑆𝑤
Lifting Airfoils and the Kutta Condition

= 0.9
Lifting Airfoils and the Kutta Condition

= 1.8
 Lift Coefficient for a Flat Plate
• Exact solutions of the potential flow around airfoils generally
requires conformal mapping techniques. However, it can be
difficult to apply these methods to arbitrary geometries.
Accordingly, approximate methods have also been developed
to model the potential flow about general airfoil shapes.
• Here we introduce the exact solution for the lift variation for a
flat plate, which we will later compare with the approximate
methods.
• The circulation that satisfies the Kutta condition for a flat
plate, with chord length , is
(5.76)
 Lift Coefficient for a Flat Plate
• Using the Kutta-Joukowsky Theorem, the lift generated is,

(5.77)
• The lift coefficient is
(5.78)

• And for small angles of attack, the lift slope is

(5.79)
• Therefore the lift coefficient can be approximated as
(5.80)
Vortex Panel Methods
 Vortex Sheet Model

Vortex sheet on the airfoil surface…

vortex at with strength ds’.

and the infinitesimal velocity contribution at induced by the point

Recognize that the velocity at every point along the airfoil surface is a
function of the entire vortex sheet!
 Vortex Sheet Model
• The vortex sheet is a curve along which an infinite number of
point vortices are placed, with the strength of the vortex at
being Thus is a circulation per unit length.
• The goal then is to determine the distribution such that: 1)
the flow is tangent to the airfoil at a given and 2) satisfies the
Kutta condition.
• Once is determined, we can calculate the velocity field, the
pressure distribution (using Bernoulli’s equation), and the lift
coefficient (using the Kutta-Joukowsky Theorem).
Infinitesimal velocity contribution at due to point vortex at ’ is,
(5.81)
 Review: Point Vortex
Our last ‘building block’ 2D potential flow is called a
point vortex:
• (5.59)
• (5.60)
• (5.61)

where 𝛤 is a scaling constant called the circulation of the

vortex [with units (length)2/time]. The streamlines are circles
about the origin. And the velocity becomes infinite as
Infinitesimal velocity contribution at due to point vortex at ’ is,
(5.81)

At the direction of is perpendicular to ’ and oriented counter-

clockwise, thus
(5.82)
 Vortex Sheet Model
• Combining Equations (5.81) and (5.82), the infinitesimal
velocity contribution at due to point vortex at ’ is:
(5.83)

• The velocity induced at by the entire vortex sheet is then the

integral around the sheet,
(5.84)
 Vortex Sheet Model
• The final velocity field is then obtained by including the
freestream contribution,
(5.85)

• Enforcing flow tangency, for all on the airfoil surface.

(5.86)
 Vortex Sheet Model
• Next we satisfy the Kutta condition at the sharp trailing edge.
To do this, we require the strength of the point vortex at the
trailing edge to be zero.
• Define as the length of the entire surface of the airfoil, then
set the strength (per unit length) of the vortex at the trailing
edge to be the sum of and
(5.87)
• The lift is obtained from the Kutta-Joukowsky Theorem,
Where,
(5.88)
 Linear-Varying Vortex Panel Method

• The vortex sheet model requires the solution of Equation

(5.86), which is an integral equation for .
(5.86)

• Since this equation generally cannot be solved in closed-form

analytically, the standard approach is to solve it numerically.
• This approach is referred to as the linear-varying vortex panel
model.
 Linear-Varying Vortex Panel Method

Panel representation of airfoil with linear-varying vortex sheet on

each panel.
‘Control points’ where flow tangency is enforced are indicated
by a black x, and labeled by the panel number.
 Linear-Varying Vortex Panel Method
• Replace the geometry of the airfoil (and it’s associated vortex
sheet) with a set of panels.
• The panel end points are labeled with the surface distance
and called ‘panel nodes’. Thus panel lies in the range

• Assume the distribution varies linearly along each panel ,

(5.89)
• This linear variation of on each panel gives rise to the term
linear-varying vortex panel.
 Linear-Varying Vortex Panel Method

Panel representation of airfoil with linear-varying vortex sheet on each panel.

• Note: there are panels, panel nodes (at the end of each panel),
and control points
• Flow tangency is enforced at the control points, which are
indicated by a black and labeled according to the panel
number.
 Linear-Varying Vortex Panel Method
• At the trailing edge, the upper and lower surface vortex
strengths and have individual (i.e. distinct) values.
(5.87)

• Thus the total number of variables to describe over the entire

paneled airfoil is Therefore, we need equations to determine
the values of .
• The equations are the flow tangency conditions and the Kutta
condition.
• Flow tangency is enforced at the midpoint of each panel,
which we will refer to as control points.
 Linear-Varying Vortex Panel Method
• The flow tangency condition from Equation (5.86)
(5.86)
• Applying flow tangency at the control point of panel yields:
(5.90)
where the summation from is required to calculate the influence of the
entire vortex sheet (distributed over panels) on the velocity at the
control point of panel

′❑
𝛾(𝑠 )
′ for j=3
𝑟
 Linear-Varying Vortex Panel Method
• Applying flow tangency at the control point of panel yields:
(5.90)
where the summation from is required to calculate the influence of the
entire vortex sheet (distributed over panels) on the velocity at the
control point of panel

• is based on Equation (5.89), and is also a function of

(5.89)
(5.91)
 Linear-Varying Vortex Panel Method
• Accordingly, introduce the following notation,
(5.92)
where and are found by integration and are just functions of the
geometry of panel and the location of the control point
• Equation (5.92) prescribes the velocity vector (component
normal to produced at control point ““ due to the vortex sheet
along panel ““.
• Substituting Equation (5.92) into Equation (5.90) gives,
(5.93)
 Influence Coefficients and Linear System
• The flow tangency conditions and Kutta condition can be
written as a linear system of equations of the form
(5.94)
– where is a vector of length
– is an by matrix
– and is a vector of length

•The flow tangency equations are placed in the first rows of the
matrix, and the Kutta condition is placed in the last row.
 Influence Coefficients and Linear System
• The system of equations has the form:

(5.95)

• Where the entries for are influence coefficients and

represent the influence of on the flow tangency condition at
control point The values of are

(5.96)
 Example Vortex Panel Solution
• The next charts present predictions for the NACA 4412 airfoil
(at ) using the linear-varying vortex panel methodology.
• The charts show the effect of the number of panels used,
from
• The asymptotic behavior (for large ) is approximately Even
the solution comes within of this value.
• Vortex panel methods require very little computation. Even
the solution runs almost instantaneously on a laptop
computer.
Example Vortex Panel Solution
 Example: Vortex Panel Method for Two Airfoils

1. How many unknowns are used to represent the vortex

sheet strength distribution, and what are these unknowns?
2. What equations are used to determine the unknowns?
3. Describe the structure of for the system of equations:
representing this set of unknowns and governing equations.
Matrix Equation for Two Airfoils
Thin Airfoil Theory