Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
152 views

1 Example 2: Heat Equation

This document discusses using Fourier transforms to solve partial differential equations describing heat diffusion and pressure distributions. It notes that: 1) The Fourier transform can be used to solve the 1D heat equation, resulting in an expression for the temperature as an inverse Fourier transform. 2) A 2D steady-state pressure problem can be solved using a sine transform, yielding an expression involving hyperbolic sines that must be inverted numerically. 3) Radial problems require transforms like the Hankel transform rather than the Fourier transform due to their coordinate system involving both even and odd derivatives.

Uploaded by

salman
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
152 views

1 Example 2: Heat Equation

This document discusses using Fourier transforms to solve partial differential equations describing heat diffusion and pressure distributions. It notes that: 1) The Fourier transform can be used to solve the 1D heat equation, resulting in an expression for the temperature as an inverse Fourier transform. 2) A 2D steady-state pressure problem can be solved using a sine transform, yielding an expression involving hyperbolic sines that must be inverted numerically. 3) Radial problems require transforms like the Hankel transform rather than the Fourier transform due to their coordinate system involving both even and odd derivatives.

Uploaded by

salman
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

Jim Lambers

ENERGY 281
Spring Quarter 2007-08
Lecture 4 Notes

These notes are based on Rosalind Archer’s PE281 lecture notes, with some
revisions by Jim Lambers.

1 Example 2: Heat Equation


Consider the diffusion of heat in a one-dimensional bar. We’ll consider an
infinitely long bar so the full Fourier transform is required. The governing
equation is:
∂2T ∂T
κ2 2 = (1)
∂x ∂t
The boundary conditions are:

T (x ± ∞, t) = T ′ (x ± ∞, t) = 0 (2)

The initial condition is a prescribed temperature that varies in space:

T (x, t = 0) = T0 (x) (3)

First consider the Fourier transform of the spatial derivatives:


1 ∞
Z
F(f ′ (x)) = √ e−isx f ′ (x) dx
2π −∞
1 1
Z ∞
= √ f (x)e−isx |∞−∞ − √ (−is)f (x)e−isx dx
2π 2π −∞
= (is)F(f (x)) (4)

since f (x) vanishes at ±∞. Similar arguments require f ′ (x) vanishes at


±∞.
The transformed differential equation is

∂ T̂
+ κ2 s2 T̂ = 0 (5)
∂t
which has the solution
2t
T̂ (s, t) = c1 (s)e−(κs) (6)

1
for some function c1 (s). Now consider the initial condition

T̂ (s, t = 0) = T̂0 (s). (7)

It follows from (6) that c1 (s) = T̂0 (s) and therefore


2
T̂ (s, t) = T̂0 (s)e−(κs) t . (8)

Now invert to find T :


1 ∞
Z
2
T (x, t) = √ eisx T̂0 (s)e−(κs) t ds (9)
2π −∞

The transform of T0 is:


1 ∞
Z
T̂0 (s) = √ T0 (λ)e−isλ dλ. (10)
2π −∞

Substituting this transform into (9) yields


1 ∞ isx ∞ −isλ
Z Z
2
T (x, t) = e e T0 (λ)e−(κs) t dλ ds (11)
2π −∞ −∞
1 ∞
Z Z ∞
2
= T0 (λ) e−is(λ−x)−(κs) t ds dλ (12)
2π −∞ −∞
1 (λ−x)2
Z ∞
= √ T0 (λ)e− 4κ2 t dλ. (13)
4κ2 πt −∞

2 Example 3: Elliptic Problem


Consider a steady-state problem in a two-dimensional semi-infinite domain
governed by:
∇2 p = 0 (14)
The boundary conditions are:

p(0, y) = 0 (15)

lim p(x, y) = 0 (16)


x→∞

p(x, 0) = f (x) (17)


p(x, a) = 0 (18)

2
Since the domain is semi-infinite and the pressure is specified on the
boundary, we will use the sine transform to transform the differential equa-
tion: !
∂2p ∂2p
Fs + =0 (19)
∂x2 ∂y 2
∂ 2 p̂
r
2
⇒s p(0, y) − s2 p̂ + 2 = 0 (20)
π ∂y
∂ 2 p̂
⇒ − s2 p̂ = 0 (21)
∂y 2
This equation can be solved for p̂ to give:

p̂(s, y) = c1 (s)cos(isy) + c2 (s)sin(isy) (22)

Now use the boundary conditions to determine c1 (s) and c2 (s):

Fs (p(x, y = 0)) = Fs (f (x)) = F1 (s) (23)

Fs (p(x, y = a)) = 0 (24)


After some algebra we can show:
sinh(s(a − y))
p̂(s, y) = F1 (s) , (25)
sinh(sa)
where
ez − e−z
sinh(z) = = −i sin(iz). (26)
2
Now invert to find p:
r
2 ∞ sinh(s(a − y))
Z
p(x, y) = F1 (s) sin(xs) ds (27)
π 0 sinh(sa)
where r
2 ∞
Z
F1 (s) = f (λ) sin(sλ) dλ (28)
π 0
is the Fourier sine transform of f (x). Substituting F1 (s) into the expression
for p(x, y) gives:
2 ∞Z ∞ sinh(s(a − y))
Z
p(x, y) = f (λ) sin(sλ) sin(sx) dλ ds, (29)
π 0 0 sinh(sa)
which, due to its complexity, will have to be inverted numerically. We will
discuss this later.

3
3 Radial Problems
All of the examples presented so far have been for infinite domains, or semi-
infinite domains with linear boundaries. However, radial problems are often
of more interest to petroleum engineers. Is the Fourier transform helpful in
these cases? Consider the problem

∂ 2 pD 1 ∂pD ∂pD
∂rD 2 + r ∂r =
∂tD
(30)
D D

The sine and cosine transforms won’t work, because the pressure equation
in radial coordinates includes both even and odd orders of derivatives. The
full Fourier transform is a candidate, if we consider applying it to the spatial
variable rD . When transforming the spatial derivatives, we will require the
behavior of the pressure at ±∞. Ideally, the pressure and its first derivative
should vanish at ±∞. This may be the case at r = +∞, but it is much harder
to make that claim at r = −∞. Applying the full Fourier transform in time
is another option. However, transforming the time derivative requires the
behavior of the pressure at t = −∞. This is not such a problem, since it
is likely p = pi would be suitable. However, now the boundary conditions
become time dependent if the flow begins at t = 0.
The Hankel transform
Z ∞
Hν (λ) = rJν (λr)f (r) dr (31)
0
Z ∞
f (r) = λJν (λr)Hν (λ) dλ, (32)
0
where Jν is a Bessel function of the first kind, is better suited to radial
problems. We’ll see Bessel functions later in this course.

4 Inverting Fourier Transforms Numerically


Since the limits of the inversion integral are real, standard numerical inte-
gration routines can be used to evaluate the integral that defines the inverse
Fourier transform. Due to the oscillatory nature of the integrand, however,
many function evaluations may be required, especially for large values of
s. If the Fourier transform is being applied to discrete data instead of a
function, there are formal algorithms that can be used to perform the inver-
sion. We will discuss this further when we learn about the discrete Fourier
transform.

You might also like