(7.

4) Radiation from a Plane Circular Piston

The plane circular piston is of particular interest in acoustics because it is a model for a number of
sources, i.e. loudspeakers, open ended organ pipes, ventilation ducts, and many types of single
element transducers.

a
z
us

Let’s consider a piston of radius a mounted on a flat rigid baffle of infinite extent. We want to look
at the radiating surface of the piston assuming that it moves with a uniform speed
us = U 0 e jω t
normal to the baffle when y 2 + z2 < a .

If we consider a simple source of area dS then we have for the source strength (surface area of
source x velocity amplitude)

dQ = U 0dS

So that the contribution from one baffled simple source is

dQk j( ωt –kr ′)
dp = j ρ 0c e
2π r ′
Oelze ECE/TAM 373 Notes - Chapter 7 pg 14
The total pressure in the disk is found by summing up all of the “simple” sources on the disk

′
ρ0cU 0 k e j(ωt –kr )
2π ∫S r ′
p ( r ,θ, t) = j dS for 0 ≤ σ ≤ a.

Typically, we divide the field from the piston source into the near field and the far field.

Near Field – complicated variation in the pressure with position due to the complex interference
from simple sources (Huygens sources).

An analytic solution is possible only along the axis. The extent of the near field is defined in
terms of this result.

Far Field – Can get an approximate analytic solution everywhere in far field.

I. The Near Field approximations

We’ll consider the case where r is not much greater that a, that is, in the near field (Fresnel Zone).
Here, we can obtain a closed-form solution only for the on-axis (axial) case. We start with

jωρ oU o j (ω t− kR)
dp ( r , t ) = e dS
2π R

where dp (r ,t ) is the axial pressure at an axial distance r due to an annular ring on the source
surface.

R1
a R
p(r,t)
σ
r

You can see from the diagram that there is no dependence on θ when looking at pressure on the
axis. Summing up the annular rings over the radius a

jωρoU o j(ωt −kR )

p (r , t ) = ∫ e dS
S
2π R

where dS = 2πσ dσ for an incremental annulus and R = σ 2 + r 2 .

Therefore,

Oelze ECE/TAM 373 Notes - Chapter 7 pg 15

 jωρ oUo e jω t a
e − jkR a − jkR
j ωt e
p (r , t ) = ∫0 R 2πσ dσ = jωρ U e ∫0 R σ dσ
2π o o

If we note that

d e – jk r +σ σ e – jk r +σ σ e – jkR
2 2 2 2

– = =
dσ jk r +σ
2 2 R

Then

 – jk ( r2 + a2 – r )  e j(ωt – kr) .
p ( r ,0, t ) = ρ0 cU0 1 – e 
 

Pulling e
–j
1
2 (
k r2 + a2 – r ) out of the bracket gives

 j 1 k( r2 + a2 – r ) – e– j12 k( r2+ a2 – r )  e– j12 k( )e j(ωt – kr) .

r2 + a2 – r
p ( r ,0, t ) = ρ0 cU0 e 2 
 
But that just gives us

 – j k( )e j (ωt– kr ) .
1
( )
1
r2 + a2 – r
p ( r ,0, t ) = ρ0 cU0 sin  k r2 + a2 – r  e 2
2 

The pressure magnitude on axis is

1  2  
  a
p ( r ,0) = 2 ρ0cU 0 sin  kr  1 +   –1 
 2  r 


So, what does this equation mean physically about the acoustic pressure along the axis in the near
field?

1  
2
 a maxima for m odd
kr  1 +   – 1 = mπ 2
2  r  minima for m even
 

If we solve for r in terms of m then

rm a mλ
= – for m = 0, 1, 2, …
a mλ 4 a

m = 0 corresponds to r → ∞

Oelze ECE/TAM 373 Notes - Chapter 7 pg 16

 m = 1 ⇒ last axial maximum – This is where the far field equations begin to be valid for
the plane piston source.

r1 a λ or a2
= – r1 ≈ for a > λ
a λ 4a ⇒ λ

0
0 1 2 3 4 5 6 7
r/a
Figure 7.4.2 Axial pressure amplitude for a baffled circular plane piston of
radius a radiating sound of wave number k with ka = 8π. Solid line is
calculated from the exact theory. Dashed line is the far field approximation
extrapolated into the near field. For this case, the far field approximation is
accurate only for distances beyond about seven piston radii.

a2
r1 ≅
λ
r1 a
≅ ≅4 for ka = 8π
a λ
1
Note that for r > ≈ 2r1 the pressure decreases as .
r

*************************** Example 7.2 ***************************

Examples of Axial Pressure Amplitude versus r/a

Using for the baffle source:

1  2  
  a
p ( r ,0) = 2 ρ0cU 0 sin  kr  1 +   –1 
 2  r 


Oelze ECE/TAM 373 Notes - Chapter 7 pg 17

 p ( r, 0)  kr   a2   
The following are examples of = sin   1 +  2  − 1  versus r/a for different values
2 ρo cUo  2   r   
of a/λ (recall that ka = 2π(a/λ).

Axial Pressure Amplitude for radius/wavelength =

0.1
1
0 .9
0 .8
0 .7
0 .6
0 .5
0 .4
0 .3
0 .2
0 .1
0
0 1 2 3 4 5 6 7 8 9 10

r/a
r/ a

Axial PressureAmplitude for radius/wavelength =

Axial PressureAmplitude
0.5 for radius/wavelength =
0.5
1
0 .9
0 .8
0 .7
0 .6
0 .5
0 .4
0 .3
0 .2
0 .1
0
0 1 2 3 4 5 6 7 8 9 10
r/a
r/ a

Oelze ECE/TAM 373 Notes - Chapter 7 pg 18

 Axial PressureAmplitude for radius/wavelength
=1
1
0 .9
0 .8
0 .7
0 .6
0 .5
0 .4
0 .3

0 .2
0 .1
0
0 1 2 3 4 5 6 7 8 9 10
r/a
r/ a

Axial PressureAmplitude for radius/wavelength

= 1.5
1

0 .9
0 .8

0 .7
0 .6

0 .5
0 .4

0 .3
0 .2
0 .1
0
0 1 2 3 4 5 6 7 8 9 10

r/a
r/ a

Oelze ECE/TAM 373 Notes - Chapter 7 pg 19

 Axial PressureAmplitude for radius/wavelength
=2
1
0 .9
0 .8
0 .7
0 .6

0 .5
0 .4
0 .3
0 .2

0 .1
0
0 1 2 3 4 5 6 7 8 9 10
r/a

Axial PressureAmplitude for radius/wavelength

=4
1
0 .9
0 .8
0 .7
0 .6
0 .5
0 .4
0 .3
0 .2
0 .1
0
0 1 2 3 4 5 6 7 8 9 10
r/a

From these pictures what do you think happens to the far field as a/λ gets larger?
How do the number of peaks and the number of multiples a is above the wavelength match up?

******************************************************************

Oelze ECE/TAM 373 Notes - Chapter 7 pg 20

II. The Far Field approximations

The book uses a line source approximation to derive the Far field approximation for the baffled
piston source. We will look at a more traditional way of deriving the far field approximation. Let’s
define the coordinate system we will be working with:
y
p(r, θ,t)
Ψ δ R
r
θ
x

Ψ y
σdΨ

σ a
z
dσ
dS = σdΨdσ

Recall that the contribution from one baffled simple source is

dQk j (ωt – kR)

dp ( r , θ , t ) = jρ 0c e .
2π r ′
where dQ = U 0dS . The incremental pressure, dp, at the field position (r, θ ,t) from the incremental
surface dS is:
U k jωρoU o j( ωt − kR)
dp ( r , θ , t ) = j ρ 0c 0 ej (ω t – kR) dS = e dS
2π r ′ 2π R
or
jωρoU o jω t e − jkR
p ( r ,θ , t ) = e ∫ R dS
2π S :σ ≤ a

To determine R in terms of σ and Ψ:

p(r, θ,t)
R

σ δ r line 2

line 1

Law of Cosines: R 2 = σ2 + r 2 − 2rσ cos δ where cosδ is determined from the Direction Cosines of
lines 1 and 2, that is, cos δ = l1l 2 + m1m2 + n1n2 where the Direction Cosines are determined by the
particular axial component length of a vector divided by the total length of the vector.
Oelze ECE/TAM 373 Notes - Chapter 7 pg 21
 l1 = 0 l2 = cos θ
m1 = cos Ψ m2 = sin θ
n1 = sin Ψ n2 = 0

which yields: cos δ = cos Ψ sin θ .

2 σ2 σ  
R 2 = σ2 + r 2 − 2rσ cos δ = σ 2 + r 2 − 2rσ cos Ψ sin θ = r  1 + 2 − 2   cos Ψ sinθ 
 r r  
σ 2
σ
Because: a<<r, 2 << << 1
r r
 σ 2
σ    σ  
R 2 = r 2 1 + 2 − 2   cos Ψ sin θ  = r 2 1 − 2   cos Ψ sin θ 
 r r   r 
σ   σ  
R = r 1− 2  cos Ψ sinθ ≈ r 1 −   cos Ψ sin θ  = r − σ cos Ψ sin θ .
r   r  
So,
jωρoU o jω t e − jkR jωρ oU o jωt e− jk ( r −σ cos Ψsin θ )
p ( r ,θ , t ) = e ∫ dS = e ∫ dS
2π S :σ ≤ a
R 2 π S :σ ≤a
r − σ cos Ψ sin θ
jωρoUo j( ωt − kr ) e jkσ cos Ψ sinθ
=
2π
e ∫ r − σ cos Ψ sin θ dS
S :σ ≤a

For the far-field approximation: σ cos Ψ sin θ = r in denominator because change in amplitude is
σ2 σ
assumed to be negligible; remember: 2 << << 1 . So,
r r

jωρoU o j(ωt −kr )

p ( r,θ ,t ) = e ∫ e jkσ cos Ψ sinθ dS
2π r S:σ ≤ a

Using for our incremental surface area, dS = σ dσ d Ψ , yields

jωρoU o j(ωt −kr )

p ( r,θ ,t ) = e ∫ e jkσ cos Ψ sinθσ d σ d Ψ
2π r S:σ ≤ a

Limits of integration:
σ : 0→a
Ψ : 0 → 2π
giving
a
 2π jkσ cos Ψ sin θ 
∫ e
jkσ cosΨ cinθ
σ dσ d Ψ = ∫  ∫ e
σ =0  Ψ= 0
d Ψ σ d σ

S :σ ≤ a
2π
Therefore, let’s first look at ∫
Ψ=0
e jkσ cosΨ sinθ d Ψ

Oelze ECE/TAM 373 Notes - Chapter 7 pg 22

Using the general relationship for the Bessel’s function of the first kind of order n
(− j)
n 2π

Jn ( x) = cos ( nΨ ) d Ψ
∫e
jx cosΨ

2π 0
Comparing with the above integral we see that we have a Bessel function of order n = 0, so
2π
1
J o (k σ sinθ ) =
2π ∫0
e jkσ sinθ sosΨ d Ψ

where x = kσ sinθ . Our pressure then is

jωρoU o j( ωt −kr )
a
p (r, θ , t ) =
2π r
e ∫σ = 0{2π J o ( kσ sin θ )}σ d σ
jωρoU o j(ωt −kr )
a
=
r
e ∫ σ J o ( kσ sin θ )dσ
σ =0
a
To evaluate ∫ σ J ( kσ sinθ )d σ , we observe that ∫ xJ ( x ) dx = xJ ( x ) .
σ =0
o o 1
For ∫ xJ ( x ) dx = xJ ( x ) ,
o 1

x dx
we let x = kσ sin θ ⇒ σ = and dx = k sin θ d σ ⇒ dσ = . Our limits of
k sin θ k sin θ
integration become:
σ =0 ⇒ x=0
σ = a ⇒ x = kaSinθ
giving,

ka sinθ
 x   dx 
a
1 ka sin θ
∫ σ J o ( kσ sin θ )d σ = ∫   Jo ( x)   = xJ1 ( x )
 k sin θ   k sin θ  ( k sin θ )
2 0
σ =0 0

a
= J1 ( ka sin θ ) .
k sin θ
Finally our pressure is given by

jωρ oa 2U o j (ωt − kr )  2 J1 ( ka sin θ ) 

p (r, θ , t ) = e  
2r  ka sin θ 
The pressure amplitude will be directional depending on θ . We define the Directional Factor (or
also called the beam pattern) of a source, H ( θ ) by
 2 J ( ka sin θ )  2 J1 ( ka sin θ )
2
I
H (θ ) = H (θ ) =
2
= 1  →
I ref  ka sin θ  ka sin θ

Values for the directional factor are listed in Appendix 6 for a baffled piston.

Let’s look briefly at what the Bessel function (n = 0,1,2) looks like

Oelze ECE/TAM 373 Notes - Chapter 7 pg 23

 1.0
Jo(x)
J1(x)
0.5 J2(x)

Jn(x)
Jn(x)
0

-0.5
2.4 3.8 5.1 5.5 7.0
x = kaSinθ

2J ( x )  2J ( x ) 
2

Looking at the Directivity Factor, H (θ ) = 1 (pressure) and  H (θ )  =  1

2

x  x 
(intensity) where x = ka sin θ
1.0
0.8
2J1(x)/x

0.6
0.4
0.2
0.0
-0.2
3.83 7.02 10.15
x = kaSinθ
We can see that the size of ka is going to determine the number of peaks x can cycle through for
different values of θ . For ka = 1, no zeros will be reached as we cycle through θ .
At θ = 90°, ka sin θ = 1 ,
2 J (1)
H ( 90o ) = 1 = 0.8801 (-1.11 dB)
1

Oelze ECE/TAM 373 Notes - Chapter 7 pg 24

 ka = 1

0.0 0.5 1.0

For ka = 3.83:
−1  3.83 
θ z = sin   = 90
o

 3.83 
−1  1.613741 
θ −3dB = 2sin   = 49.8
o

 3.83 
Why –3 dB?

ka = 3.83

0.0 0.5 1.0

For ka = 6:
−1  3.83 
θ z = sin   = 39.7
o

 6 
−1  1.613741 
θ −3dB = 2sin   = 31.2
o

 6 

Oelze ECE/TAM 373 Notes - Chapter 7 pg 25

The book has a really nice figure for the beam pattern for ka = 10, 10log10 H (θ ) ,
2

Figure 7.4.5 Beam pattern H(θ)

for a circular plane piston of radius
a radiating sound with ka = 10.

What do you expect to happen to the beam pattern as ka → 0?

Oelze ECE/TAM 373 Notes - Chapter 7 pg 26