The Paraxial Approximation

Learning Objectives

Introduction to the paraxial approximation and its

importance in beam modeling
 The Paraxial Approximation
A transducer generates a complex beam of sound. Modeling that
complexity is a challenging task.

The modeling computational burden can be reduced considerably

by introducing approximations. The paraxial approximation is
one of the most useful of those approximations.
 The Paraxial Approximation
The paraxial approximation models the transducer beam as a
quasi-plane wave where most of the sound is propagating in

a given direction with an amplitude profile described by a

diffraction correction term, C(x,y,z,ω).
x
z
v0 C
p = ρ cv0C ( x, y, z , ω ) exp ( ikz − iω t )
 The Paraxial Approximation

To illustrate the paraxial approximation in a simple setting

consider a spherical wave. Suppose that we are only interested
in the spherical wave field in the vicinity of a particular direction
which we will take as the z-axis (i.e. x,y <<z)

z
x

y
r0 exp ( ikr )
p = p0
r
p0 r0 exp ( ikr )
vz ≅ vr ≅
ρc r
 The Paraxial Approximation
Then, since r = x 2 + y2 + z 2
x2 y2
= z 1+ 2 + 2
z z
ρ2
≅z+
2z
where ρ = x2 + y2

we find ⎛ ik ρ 2 ⎞
r0 exp ⎜ ⎟
⎝ 2 z ⎠ exp ikz = p C x, y, z; ω exp ikz
p = p0 ( ) 0 ( ) ( )
z
⎛ ik ρ 2 ⎞
r0 exp ⎜ ⎟
p0 ⎝ 2 z ⎠ exp ikz = p0 C x, y, z; ω exp ikz
vz = ( ) ( ) ( )
ρc z ρc
 The Paraxial Approximation

⎛ ik ρ 2 ⎞
r0 exp ⎜ ⎟
z ⎝ 2 z ⎠ exp ikz = p C x, y, z; ω exp ikz
x p = p0 ( ) 0 ( ) ( )
z
⎛ ik ρ 2 ⎞
r0 exp ⎜ ⎟
p0 ⎝ 2 z ⎠ exp ikz = p0 C x, y, z; ω exp ikz
y vz = ( ) ( ) ( )
ρc z ρc

Note that in obtaining this diffraction correction term we only

approximated the amplitude part of the spherical wave to first
order ( r ~ z) while we approximated the phase to second order.
This is because terms neglected in the phase must not only be small
with respect to the terms retained but also must be small with
respect to 2π if they are to be negligible.
 The Paraxial Approximation

Consider the wave equation

∂2 p ∂2 p ∂2 p 1 ∂2 p
+ 2 + 2 − 2 2 =0
∂x 2
∂y ∂z c ∂t

Let p = P ( x, y, z , ω ) exp [ikz − iω t ]

∂2 P ∂2 P ∂2 P ∂P
Then + 2 + 2 + 2ik =0
∂x 2
∂y ∂z ∂z
∂2 P ∂P ∂ 2 P ∂ 2 P
If we assume << 2ik , 2 , 2
∂z 2
∂z ∂x ∂y
∂2 P ∂2 P ∂P
+ 2 + 2ik =0
∂x 2
∂y ∂z
 ∂2 P ∂2 P ∂P
+ 2 + 2ik =0
∂x 2
∂y ∂z

paraxial wave equation

Our paraxial approximation for a spherical wave

satisfies this paraxial equation exactly:

p0 r0 ⎛ ik ρ 2 ⎞
P= exp ⎜ ⎟
z ⎝ 2 z ⎠
There are other solutions, such as Gaussian waves
that also satisfy this equation and form important
building blocks for modeling ultrasonic transducer
radiation in complex problems
 The Paraxial Approximation
The paraxial approximation allows one to obtain diffraction
correction terms for many practical testing setups

x'
x fluid
z' k1 =ω/cp1
v0
ds

solid dp (x,y,z) k2 = ω/cp2

z or k2 = ω/cs2

v ( x, y, z , z ′, ω ) = v0T12d C ( x, y, z , z ′, ω ) exp ( ik1 z ′ + ik2 z )

velocity plane wave polarization

transmission d or d diffraction
p s correction
coefficient
Limitations of Beam Models based on the Paraxial Approximation

near critical angles

cases where the surface curvature

varies rapidly over the beam width

at locations too close to

a plane transducer or for at high refracted angles
too tightly focused transducers