0% found this document useful (0 votes)
142 views79 pages

Roadmap For 3 Bernoulli's Equation Laplace's Equation

This chapter discusses Bernoulli's equation and potential flow theory. Bernoulli's equation relates pressure and velocity for steady, incompressible, inviscid flow. It is derived from the conservation of momentum equation. Potential flow assumes the flow is irrotational and incompressible, meaning the velocity field can be described by a scalar potential. Laplace's equation then governs the potential. Fundamental solutions to Laplace's equation include uniform flow and source/sink flows, which can be combined using superposition to model more complex flows.

Uploaded by

Dhana Jayan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
142 views79 pages

Roadmap For 3 Bernoulli's Equation Laplace's Equation

This chapter discusses Bernoulli's equation and potential flow theory. Bernoulli's equation relates pressure and velocity for steady, incompressible, inviscid flow. It is derived from the conservation of momentum equation. Potential flow assumes the flow is irrotational and incompressible, meaning the velocity field can be described by a scalar potential. Laplace's equation then governs the potential. Fundamental solutions to Laplace's equation include uniform flow and source/sink flows, which can be combined using superposition to model more complex flows.

Uploaded by

Dhana Jayan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 79

Chapter 3:

Roadmap for 3rd chapter


➢ Bernoulli's Equation
● Applications (Venturi, low speed wind tunnel, pitot tube)
Laplace's Equation

General philosophy and use


● Superposition principle

● Elementary flows (source, doublet, vortex)

● Semi-infinite and oval bodies

● Cylinder (lifting and non-lifting)

● Kutta-Joukowski Theorem

● General non-lifting bodies

● Numerical techniques (panel method)

Objective: to describe potential flow around an arbitrary


body (e.g. airfoil)
3-1
Inviscid Incompressible Flow
Incompressible: ρ=const, low speed, M<0.3.
Inviscid: high Re# and away from solid boundaries.

thin viscous
(and rotational)
region

outer flow:inviscid (and irrotational)


3-2
Bernoulli's Equation
links pressure and velocity for incompressible, inviscid
flow
Local form of momentum conservation:
∂
u 1
 u =− ∇ p f
u⋅∇ 
∂t 
Vector identity: 1

u⋅∇ 
u = ∇ u⋅
u −
u × ∇ ×
u
2

3-3
Applying the vector identity, we get:
∂
u 1 1
 ∇ u⋅
u −  =− ∇ p f
u× 
∂t 2 

Assume steady flow and rearrange:


  
p

1
u − f =
∇ u⋅
2
u ×

Assume a conservative body force: f =∇ H


(H is the body force potential).
Gravity:
3-4

Thus, we get: ∇ p  1 u⋅u  gy =u ×
 2


Let's integrate both sides of this equation along a line:

 
2 2
p 1
∫ ∇   2 u⋅u gy ⋅ds=
 ∫ u ×⋅ 
 ds
1 1

 is aligned with
(1) Along a streamline ds u.

But u ⊥ u × 
  ⇒
(2) Irrotational flow: 
 =0.

3-5
Gradient theorem: ∫ ∇ F ⋅ds=F
 b − F a
C


 p 1
 u⋅
 2  
p 1
u gy −  u⋅
2  2
u gy =0
1

p 1
⇒  
 2
u⋅
u  gy=const

If flow is irrotational ( ), then B is the same constant


throughout the fluid.
Otherwise, B is a different constant along each streamline.
3-6
Neglect the body force:
p 1
 u⋅
u =const
 2

What assumptions did we make in deriving the


Bernoulli's equation?

3-7
Example: wind tunnel contraction
Assume: steady, quasi 1D,
incompressible
inviscid
Find: velocity from
pressure measurements.

3-8
3-9
Incompressible flow in a duct.

3-10
venturi (Venturi tube = tube with a constriction)

Applications:
● carburetors in cars

● spray generation

● airspeed measurements

● liquid (e.g. blood) flow

measurements

Schematic of siphon injector


3-11
Wind tunnels:

3-12
Example: analysis of airplane designs

(CFD, panel methods)


● Solve for the velocity over the body
● Use Bernoulli's equation to calculate the pressure on
the body 1 2 1 2
p  V = p ∞  V ∞
2 2
● Integrate the pressure to find lift
3-13
Stagnation point/stagnation pressure

1 1
p  V = p ∞  V 2∞
2
2 2

3-14
1

2
p 0 = p ∞  V ∞
2

Stagnation pressure – the pressure which would exist


at a point if the fluid there were brought to rest
(neglecting viscous effects)
static pressure – the pressure existing at a point in the
flow
dynamic pressure – the change in the pressure if the
fluid were brought to rest (neglecting viscous
effects).

3-15
Example: Pitot-static probe

1 1
2 2
p1  V 1= p2  V 2
2 2 ⇒

3-16
Pressure coefficient
p − p∞ 1
C p≝ q ∞=  V 2∞
q∞ 2

for incompressible flow (Bernoulli's equation):


1 1
2
p  V = p ∞  V ∞
2 2
2

1
p− p ∞=   V ∞−V 
2 2
2
p − p∞
⇒ C p= q∞ =
For a stagnation point:
C p does not depend on V ∞ , : it does depend on Re,
M, shape and orientation of the body.
3-17
Example 1: jet flow

Where the Bernoulli's equation is applicable?


1 --> 2
1 --> 3
1 --> 4
3-18
Example 2: Find velocity of liquid flowing out of a
tank of height h.

Bernoulli's equation:
1 2 1 2
p 1   V 1  g y1 = p 2   V 2  g y 2
2 2


3-19
Potential Flow
Assume: inviscid, irrotational ( ⇒ potential)
incompressible, 2-D
u =∇ 
  - velocity potential
Conservation of mass for incompressible flow:
∇⋅
u =0 ⇒
2
∇ =0 Laplace's equation
governing equation for potential flow

3-20
Laplace's equation: 2
∇ =0

Cartesian coordinates: Polar coordinates:


2 2 2 2
∂  ∂  ∂  1 ∂ 1 ∂ 
2
 2
=0   2 =0
∂x ∂y ∂r
2
r ∂ r r ∂ 2

(other coordinate systems in the text)

3-21
Stream function. 2-D incompressible:
∂ ∂
u= , v=−
∂y ∂x
irrotational: 
 =∇ ×
u =0

∂ v ∂u
2-D flow: z = − =0
∂x ∂ y

⇒ ∇ =0 2
Laplace's equation
governing equation for stream function
3-22
Note: ∇ 2 =0 and ∇ 2  =0 imply different things.
∇ 2 =0 implies
2
∇  =0 implies

Laplace's operator is linear: n


∇ ∑ a i  i =0
2

2 2 2
∇ 1 =0, ∇  2=0, ... , ∇  n=0
i=1

(if 1,  2, ... ,n are solutions ⇒∑ a  is a solution)


i=1
i i

We can add simple solutions to get more complex ones.

3-23
Boundary conditions for velocity

Infinity BC: Wall BC:

Body is a streamline.
3-24
Fundamental solutions
A. Uniform flow: u =V ∞ ex =
Find the potential  :
∂
u= =V ∞
∂x
∂

v= =0
∂y

3-25
Let's find the streamfuction for uniform flow:
∂
u= =V ∞
∂y
∂

v=− =0
∂x

3-26
B.Source flow: radial flow in 2-D.

u r = f r , u  =0

This flow is irrotational: 


 =∇ ×
u =0
1 ∂ 1 ∂
⇒ z =
r ∂r  r u  −
r ∂
 u r  =0

incompressible: ∇⋅u =0
⇒ 1 ∂ 1 ∂ u
r ∂r
 r ur  
r ∂
=0


3-27
Volumetric flow rate / unit span
2

Q̇=∫  n dS=∫ u r R d =
u⋅
C 0

Q
⇒ c=
2
, Q= source strength

Q
ur = , u  =0
2 r
Also ∇ ×u=0 ⇒

3-28
Streamlines and equipotential lines for source/sink

Q Q
= ln r , = 
2 2

3-29
Q Q Q
ur = , u  =0 = ln r , = 
2 r 2 2
In Cartesian coordinates:

Q Q
= ln  x  y ,
2 2
= tan −1  y / x 
2 2
For a source at x0, y0:
Q x− x 0
u=
2   x− x 0 2  y− y 0 2
Q y− y0
v=
2   x− x 0 2  y− y 0 2

3-30
Uniform flow + source:

Q
velocity potential: =1 2, 1=V ∞ x , 2 = ln r
Q 2
=V ∞ x ln r
2 Cartesian
⇒ =
= Polar
Q
stream function: = 1  2, 1 =V ∞ y ,  2 = 
2
⇒ =

3-31
=
Q 1
u =u1 u2, u1 =V ∞ i , u2 =
Velocity:  er
2 r
Q x
u=V ∞ 2 2
2 x y
Q y
v=
2 x 2  y 2

Pressure: Does p=p1+p2?

What flow does source+uniform


represent?

3-32
Stagnation point: u=v=0

Q y
v= 2
2 x  y 2
=0 ⇒ y sp =

u=V ∞
Q
2
x
2 x y 2
=0 ⇒ x sp =

Q
Dividing streamline: =V ∞ r sin  
2
at the stagnation point: =±
Q Q + upper line
V ∞ r sin  =±
⇒ 2 2 - lower line

3-33
Far downstream:

 0 ⇒
For inviscid flow any streamline can be replaced by a solid
wall or body.
Source + uniform flow can model flow over a semi-infinite
body (Rankine ogive)

Flow from source inside the body stays inside:

x ∞ ⇒
3-34
Rankine Oval (source + sink + uniform flow)

velocity:
Q xa Q x −a
u=V ∞  −
2   xa2  y 2 2  x −a 2 y 2
Q y Q y
v= −
2   xa  y 2  x −a 2 y 2
2 2

stagnation points: (u=v=0)


v=0 ⇒ y sp=0

u=0 ⇒
3-35

x sp=±a 1
Q
aV ∞
Stream function:

3-36
Note: if the body is closed => Sum of sources and
sinks =0. However, simply because sum of sources
and sinks =0 does not mean that a closed body exists.
[necessary but not sufficient condition].

3-37
C. Doublet (dipole)
Let l=2a 0
while
l⋅Q=const =k Q=∞
k is called dipole strength

Velocity potential:
Q
s = ln  x  y = s 0,0
2 2
source at (0,0)
2  s  xa ,0−s  x−a ,0 
d =s  xa , 0− s x −a ,0=2 a
2a
l=2a 0 ⇒
3-38
Finally, we get the potential for dipole as
k x Cartesian
d = 2 2
2 x  y
k cos  polar
d =
2 r
∂ 1 ∂
u = =
Velocity: u r = ∂ r = r ∂
Stream function:

1 ∂
ur = =
r ∂

∂
u =− =
∂r

3-39
Streamlines: =const = 0

k
⇒ 2 2
x  y =−
2  0
y

3-40
Uniform flow + doublet

k cos
velocity potential: =V ∞ x 2 r
stream function: =V ∞ y− k sin  =
2 r

Let R = k
2
⇒ R2
=V ∞ r sin  1− 2 
2 V ∞ r

3-41
2
R
Potential: =V ∞ r cos 1 
2
∂  r
Velocity: u = =
r
∂r
1 ∂
u = =
r ∂
At r=R: ⇒ flow around a cylinder
Stagnation points: u r =0, u =0

On the cylinder:

Where is max velocity?

3-42
Pressure on the cylinder:

 
2
V
C p =1−
V∞ ⇒ C p =1−4 sin 2 
When Cp=0 ( )?
Find Cl, Cd

Is this a reasonable result?

3-43
D' Alembert's Paradox: For incompressible, inviscid, irrotational
(potential) flow drag is equal to zero for any body.

3-44
Vortex. Look at flow with circular streamlines.

u r =0, u ≠0

Incompressible: ∇⋅
u =0

1 ∂ 1 ∂ u
r u r  =0
r ∂r r ∂


Irrotational: 
 =∇ × u =0
1 ∂ 1 ∂ ur
⇒ z =
r ∂r
r u −
r ∂
=0


3-45
Circulation:
2

 =−∮  
u⋅ds=− ∫ u e⋅R d  e=
S 0

 1
⇒ u =−
2 r
 is also called vortex strength

Potential and stream function:


 
= , = ln r
2 2
Compare to source:
Q Q
= ln r , = 
2 2
3-46
Ideal vortex has a singularity at origin:
 1
u =−
2 r
Real vortices have a viscous core:

E.g. Rankine vortex:


r≤r c u =− 2
r
2 r c


rr c u =−
2 r
(end of material for Exam 2)
3-47
*
Aerodynamic Interference - Method of images
Interference: ground / wind tunnel boundary effects. The presence of a
solid wall in potential flow can be modeled using images.

Source over a wall


(e.g., in a wind tunnel).

Wall at y=0.
Source of strength Q at (0, a).

The effect of the wall is the same as that of “image source” of


strength Q at (0,-a).
*
Not in Anderson. See 4.12 “Foundations of Aerodynamics”, Kuethe.
3-48
The resulting velocity 
u =u1 u2 is along the wall
⇒ wall is a streamline
Q
check: = 1  2 = 1 2 
2

at 0,0: 1 =− , 2 = / 2
2 ⇒ =0

Source near a wall is roughly similar to that for an air cushion vehicle in
which a downward-blowing air jet is deflected by the ground or by water
surface.

3-49
Vortex over a wall

image vortex at (0,-a) has


opposite strength

total velocity 
streamline
u =u1 u2 is along the wall ⇒ wall is a

ψ=ψ +ψ = Γ (ln (r )−ln (r ))=


check: 1 2
2π 1 2

3-50
The pressure field can be calculated using Bernoulli's equation
and the resulting ground (wall) reaction force can be calculated
by integration of pressure along y=0.

corner: 3 images parallel walls:

3-51
Lifting flow over a cylinder
non-lifting flow over a cylinder (uniform+doublet) + vortex

 = 1  2
R2 k
 1 =V ∞ r sin 1− 2  , R=
r 2 V ∞

2= ln r C
2

Set C=−

2
ln R ⇒ = 1  2 =

3-52
R2  r
= 1  2 =V ∞ r sin 1− 2  ln
r 2 R
at r=R: r=R is a streamline

2
velocities: u =V cos 1− R 
r ∞ 2
r
R2 
u =−V ∞ sin 1 2 −
r 2 r

find stagnation points:


u r =0 ⇒

3-53
A) r sp =R

u =−2 V ∞ sin − =0
2 R
⇒ sp =

(i) for  4  R V ∞ ⇒

(ii) for  4  R V ∞
(iii) for  =4  R V ∞

3-54
B) sp =± 
2 2
R 
u =∓V ∞ 1 2 − =0
r 2 r
r2  r
⇒ ±
R 2 R V∞ R
2
1=0

[  ]
2
r sp ∣∣ 4  R V∞
⇒ =
R 4  R V∞
1 ± 1−

(i) for ∣∣4  R V ∞


(ii) for ∣∣4  R V ∞

(iii) for ∣∣=4  R V ∞

3-55
In summary:
1.  4  R V ∞ 2.  =4  R V ∞ 3.  4  R V ∞

3-56
Potential flow around a circular cylinder is not unique. To
make flow unique we need to set the circulation.
Let's calculate drag and lift.
On the cylinder (r=R)
R2 R2 
u r =V ∞ cos 1− 2  =0 u =−V ∞ sin 1 2 − =
r r =R r 2  r r=R

V 2
C p =1− =
V∞
2
1
cd =− ∫
2R 0
C p cos R d =0

3-57
Lift: 2
1
cl =− ∫
2R 0
C p sin  R d =
2
 
=− ∫
 R V∞ 0
2
sin  d =
RV ∞
1
L ' =c l  V 2∞ S =
2
The lift per unit span is
proportional to circulation
(known as
Kutta-Joukowski theorem
valid for any 2D body)

3-58
Kutta Condition
Potential flow around lifting cylinder is not unique. Same with
airfoils.

Which picture is physically realistic?


Kutta condition: A body with a sharp TE in motion through a
fluid creates about itself a circulation of sufficient strength to
keep the rear stagnation point at the TE.
In other words: flow leaves TE smoothly.
3-59
2 types of trailing edges:
1. Cusp: upper and lower 2. Finite angle at TE
surface are tangent at TE

Corrolaries:
● The streamline that leaves a sharp TE is an extension of the
bisector of the TE angle
● Near TE the velocities on upper and lower surface are equal
at equal distances (useful for panel methods)
3-60
Kelvin's Theorem – rate of change of circulation
D D
= ∮ 
u⋅ds
Dt Dt C
D D
u  D 
=∮  ⋅ds∮ 
u⋅ ds
Dt C Dt C Dt

Assume inviscid flow. Euler equations (include body force):


D
u 1
=− ∇ p f
Dt 
From the kinematics of a material line:
D 
Dt

ds= ds⋅∇ 
u

3-61
⇒  
u⋅ ds⋅∇   1
u =ds⋅∇  
2
u⋅
u
So,
D 1  ∮∇1 
=∮ [− ∇ p f ]⋅ds u⋅ 
u ⋅ds
Dt C  C 2

Assume incompressible flow and a conservative body force

D p 1
=∮ ∇ [−  H u⋅ 
u ]⋅ds
Dt C  2

Gradient theorem:

⇒ D
Dt
=0 Kelvin's Theorem

3-62
Airfoil acceleration
i) initially at rest

ii) airfoil accelerates

(starting vortex has opposite strength to airfoil)


- opposite happens when airfoil stops
3-63
Numerical methods for potential flow
1. Panel methods:
Non-lifting Lifting

discrete sources discrete sources and vortices

continuous source distribution continuous source and vortex distribution

source panel method (3.17, Anderson) vortex panel method (4.9 Anderson)
+ vortex
3-64
3-D

2. Conformal Mapping.

3. CFD (finite difference/finite element methods)

3-65
Panel methods: fast, but can not be extended to non-linear
problems (viscous, transonic)
2D discrete sources panel method (not in Anderson)
1) solid body is represented by a set of straight panels
2) mid-point of each panel has a source (2D line source) of
unknown strength λ i
3) Find unknown source strength given that flow tangency is
satisfied at the mid-point of each panel.
⃗u⋅⃗ni =0 ∀i=1,... , N

3-66
Example: flow over a bump.

3-67
3-68
2D continuous source panel method (3.17 Anderson)
1. Symmetric bodies, =0 . Assume a source sheet
= s - source strength / unit length (m/s)
s – distance along the surface
r – (local) distance from P

at P:

For a body of arbitrary shape, find

3-69
at P:  ds
d = ln r
2
b
 ds
=∫
⇒ a 2
ln r

For a body of arbitrary shape, find  s.


Approximation: straight panels
assume:  s= i on each panel
N unknowns: 1,  2, ... , N

3-70
Apply flow tangency condition in each control point
N equations
At point P velocity potential due to panel j:
j
⇒   j=
2 j
∫ ln r p d s j
j

because of all panels: N N


j
⇒ P =∑   j =∑ ∫ ln r p d s j
j=1 2  j
j
j =1
where
r P =  x− x j   y− y j 
2 2
j

3-71
If P = i-th panel mid-point xi, yi:
N
j
 xi , yi =∑ ∫ r ij =  x i − x j   y i − y j 
2 2
ln r ij d s j ,
j=1 2 j

The normal velocity of the source panel is: V n= [  xi , yi ]
∂ ni
For j=i (velocity induced by the i-th panel itself) i
V n=
3-72
2
Summing up for the normal velocity at (xi,yi) due to all source panels
and applying flow tangency:
i N  j ∂
V n=  ∑ ∫ [ln r ij ] d s j V ∞ cos  i =0
2 j=1 2  j ∂ n i
or
i N  j
∑ I ij V ∞ cos  i =0
2 j =1 2 


where I ij =∫ [ln r ij ]d s j
j ∂ ni

∂ 1 ∂ r ij
[ ln r ij ]= =
∂ ni r ij ∂ ni

3-73
Finally, we get the system of N equations for N unknowns
(Eq. (3.153) in Anderson).

System is full ⇒ Gaussian elimination ~O(N3)


More panels give higher accuracy

Once source panel strengths i are found, velocity on the


surface of the body can be estimated:
N
j ∂
V i =V ∞ sin  i ∑ ∫ [ln r ij ] d s j
j =1 2 j ∂ s

3-74
Lifting bodies (nonsymmetric flow)
Smith-Hess Method(Moran, Ch. 4)
add a constant (distributed) vortex strength γ
⇒ 1 more unknown
Flow tangency at midpoint of panel i (Eq. (3.153) Anderson)
now becomes:

i N
∑

j
2 j =1 2 
I ij 

2 
J ij V ∞ cos  i =0

where Jij = normal velocity at midpoint of panel i due to a unit


strength constant vortex distributed at panel j.
Jij can be evaluated analytically (depends on geometry).
3-75
Need 1 more equation: Kutta condition
1st panel velocity = N-th panel velocity

Finally, N+1 equations for N+1 unknowns


Other panel methods also exist:
e.g. constant/linearly varying vortices (used in XFOIL)
(1st / 2nd order method) Ch. 4, AAE334/ Kueth & Chow
Cd denotes code accuracy (should --> 0)
For calculation with different N's it is better to plot Cd vs 1/N
(Richardson extrapolation).
3-76
More points (i.e., smaller panels) should be used near LE and
TE (e.g., using a cos θ distribution).

3-77
Summary of inviscid, incompressible flow (Chapter 3)
Bernoulli's equation: p 1
 u⋅u  gy=const
 2
Assumptions: inviscid
steady
incompressible
along streamline/irrotational
no body force/conservative body force
Potential flow: ∇ 2 =∇ 2 =0
boundary conditions: ∂ ∂
u= = =V ∞ at ∞
∂ y ∂x
∂ ∂
v=− = =0
∂x ∂y
u⋅n =0 at body (flow tangency)

3-78
Elementary flows:

3-79

You might also like