AE 3003 Chapter III - Handout #2 Sources, Sinks and Doublets - The Building Blocks of Potential Flow
AE 3003 Chapter III - Handout #2 Sources, Sinks and Doublets - The Building Blocks of Potential Flow
AE 3003 Chapter III - Handout #2 Sources, Sinks and Doublets - The Building Blocks of Potential Flow
0
x 2
y 2
z 2
Or
2 0
(1)
where the operator is called the Laplacian operator. This equation holds for 2-D and
3-D inviscid irrotational flows. If we are only interested in 2-D irrotational inviscid
flows, we may also solve for:
2
2 0
(2)
where is the stream function.
After we have solved for the velocity potential or the stream function, we can compute
the velocities. In a Cartesian coordinate system, for 2-D flows, we will use:
x
y
y
x
(3)
In a polar coordinate system, for 2-D flows we will use:
1
r
r
1
v Tangential velocity
r
r
v r Radial velocity
(4)
In 3-D, the velocities are given only in terms of the velocity potential, as follows:
V
Or ,
u
x
(5)
Once the velocity is known, we can find pressure from the Bernoulli's equation.
In this section, we consider some simple solutions to the Laplace's equation (1 or
2). Since equation 91) and 92) are linear, we can superpose many such simple solutions to
arrive at a more complex flow field. This is like building a complex configuration using
Lego blocks. The individual simple solutions are the individual Lego pieces, which on
their own, are not very interesting. Together, however, they can solve some very
interesting flows, including flow over airfoils and wings.
Building Block #1: 2-D Sources and Sinks: A source is like a lawn sprinkler. It sprays the
water (or air) radially, and equally, in all the directions, at the rate of Q units per unit
time. If this is a sink (e.g. a drain hole on a concrete pavement) the velocity vectors will
still be radial, but directed inwards towards the center. The sign of Q will be positive for a
source, and negative for a sink.
Consider a circle of radius r enclosing this source. Let vr be the radial component of
velocity associated with this source (or sink). Then, form conservation of mass, for a
cylinder of radius r, and unit height perpendicular to the paper,
Q 2r 1 v r
Or ,
vr
Q
2r
(6)
Solving equation (6) for the velocity potential and the stream function we get, for a
source or a sink:
Q
log e r
2
Q
(7)
plus a constant. The constants appear as we integrate the velocity to get the velocity
potential or stream function. Every simple solution we consider will be an analytical
function (like equation 7) plus a constant to be determined later.
Exercise: Verify for yourselves that (7) satisfies Laplace's equation in polar coordinates:
2 1 1 2
2
0
r r r 2 2
r
2 1
1 2
2
0
r 2 r r r 2 2
2
v
u
This is also, by itself, an uninteresting flow. It represents a uniform flow with the
velocity components u and v along the x- and y- axes. The stream function and the
velocity potential associated with this flow are:
Uniform Flow u x v y
Uniform Flow u y v x
(8)
If we use equation (3) on the definitions given in equation (8), we recover the Cartesian
components of velocity. Notice that these functions shown on (8) are simple straight
lines. It is also easy to see that these functions given in equation (8) satisfy the Laplace's
equation.
Superposition of a Source and a Uniform Flow:
Let us try to superpose the uniform flow and the flow field due to a source. All we
have to do is add the flow field given in equations (8) to equation (9). The result is given
below:
Q
log e r u x v y
2
Uniform Flow u y v x
(10)
We can take x and y- derivatives of this flow field to get the velocity field. Note that the
quantity 'r' represents the distance between where the source (xs,ys), and a general point
(x,y) where the distance is being computed. That is,
x x Source 2 y y Source 2
(11)
Thus, the velocity potential in Cartesian form, taking into account where the source has
been placed, is:
Q
2
2
log x x s y y s u x v y
4
u
v
(12)
x xs
Q
u
x 8 x x s 2 y y s 2
y ys
Q
v
y 8 x x s 2 y y s 2
(13)
These velocities may be plugged into the Bernoulli's equation:
p
1 2
1
u v 2 p u 2 v2
2
2
(14)
where p is the pressure value far away from the source.
We can also plot the flow field and the streamlines. The easiest way to accomplish this is
using built-in functions such as the MATLAB function "contour". Here is an example. In
this example, a source of strength Q equal to unity is placed at the origin. The freestream
velcoity is UINF= 1, and VINF=0. Here is the MATLAB script for modeling this flow.
X=-1:.1:1;
Y=X;
[x,y] = meshgrid(X,Y);
Q=1;
UINF=1.;
VINF=0.;
z=Q/(2.*3.14158)*atan2(y,x)+UINF*y-VINF*x;
contour(z,20);
This example produces the contours shown below. Note that this looks like flow
around the nose of a body. This body is called "Rankine's Half-Body."
20
18
16
14
12
10
8
6
4
2
2
10
12
14
16
18
20
Here are the contours from the resulting plot. This flow resembles flow over an
oval shaped object, called Rankines full body. It is similar to the shape made popular
in Ford commercials.
20
18
16
14
12
10
8
6
4
2
2
10
12
14
16
18
20
Doublets:
Doublets are source-sink pairs, initially separated by a distance , which are
brought close together by making the separation distance 0. To keep them from
annihilating each other, their strength Q is progressively increased so that Q times
remains a constant. This constant is given the symbol , called the strength of the
doublet.
We can derive expressions for the stream function (and the velocity potential) for
a doublet from the known expressions for sources and sinks. Consider a source of
strength Q placed on the x-axis at a point A, and a sink of strength Q placed on the xaxis at point B. The points A and B are placed a distance apart. Then, the stream
function at a general point P in the flow field is given by:
Doublet Limit
0
Q
2
2 1
where 1 is the angle formed by the line AP with respect to the x- axis, and 2 is the angle
formed by the line BP with respect to the x- axis. See the figure below.
P
E
Thus, stream function associated with the doublet, in the limit as goes to zero, is given
by:
Doublet
sin
2 r
a2
u y 1 2
r
u y v x
Notice that the stream function is zero on the surface r=a. In other words, r=a, a
cylinder is a streamline. Thus, the superposition of a doublet and a uniform flow, for
some special situations, becomes flow over a circular cylinder. We can plot this function
using MATLAB. We will find that this function does yield flow over a circular cylinder
of radius a.