Scribd
Upload
188 views

FT Application Solving PDE and ODE

1) The Fourier transform can be used to solve partial differential equations (PDEs) and ordinary differential equations (ODEs) by transforming calculus operations like differentiation and integration into algebraic operations like multiplication and division. 2) Using the Fourier transform method, the solution of an initial-boundary value problem consisting of a PDE and boundary and initial conditions can be obtained. The Fourier sine or cosine transform is used depending on the type of boundary conditions. 3) As an example, the heat equation in an infinite rod is solved using the Fourier transform method. The PDE is transformed into an ODE in Fourier space, which is then solved. The inverse Fourier transform gives the solution to the original problem.

Uploaded by

Gerard Velicaria
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
188 views

FT Application Solving PDE and ODE

1) The Fourier transform can be used to solve partial differential equations (PDEs) and ordinary differential equations (ODEs) by transforming calculus operations like differentiation and integration into algebraic operations like multiplication and division. 2) Using the Fourier transform method, the solution of an initial-boundary value problem consisting of a PDE and boundary and initial conditions can be obtained. The Fourier sine or cosine transform is used depending on the type of boundary conditions. 3) As an example, the heat equation in an infinite rod is solved using the Fourier transform method. The PDE is transformed into an ODE in Fourier space, which is then solved. The inverse Fourier transform gives the solution to the original problem.

Uploaded by

Gerard Velicaria
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 2

Solving PDE and ODE using Fourier Transform Transforming the initial condition, we obtain.

One of the most important application of the


Fourier transform is that it can solve given PDEs
and ODEs. It basically transforms calculus:
differentiation and integration — into algebra: In 𝑣𝑖𝑒𝑤 𝑜𝑓 𝑡ℎ𝑒 𝑟𝑒𝑙𝑎𝑡𝑖𝑜𝑛 𝑈(𝛼, 0) = 𝐹(𝛼)
multiplication and division. 𝑤ℎ𝑖𝑐ℎ 𝑓𝑜𝑙𝑙𝑜𝑤𝑠 𝑓𝑟𝑜𝑚 𝐸𝑞(1) 𝑤𝑒 ℎ𝑎𝑣𝑒 𝑈(𝛼, 𝑡)
2
The solution of an IBVP consisting of a partial = 𝑈(𝛼, 0)𝑒 −𝛼 𝑡
differential equation together with boundary 𝑡ℎ𝑒 𝑖𝑛𝑣𝑒𝑟𝑠𝑒 𝑡𝑟𝑎𝑛𝑠𝑓𝑜𝑟𝑚 𝑖𝑠 𝑔𝑖𝑣𝑒𝑛 𝑏𝑦
and initial conditions can be solved by the 𝑢(𝑥, 𝑡)
∞
Fourier Transform method. If the boundary 1 2
= ∫ [𝐹(𝛼)𝑒 −𝛼 𝑡 ]𝑒 −𝑖𝛼𝑥 𝑑𝛼 … … (2)
conditions are of Dirichlet type where the √2𝜋 −∞
2
function value is prescribed on the boundary, 𝑡ℎ𝑒 𝑖𝑛𝑣𝑒𝑟𝑠𝑒 𝑡𝑟𝑎𝑛𝑠𝑓𝑜𝑟𝑚 𝑜𝑓 𝐺(𝛼)= 𝑒 −𝛼 𝑡 𝑖𝑠
then the Fourier sine transform is used. If the ∞
1 2 𝑡−𝑖𝛼𝑥
boundary conditions are of Neumann type 𝑔(𝑥) = ∫ 𝑒 −𝛼 𝑑𝛼
where the derivative of function is prescribed on √2𝜋 −∞
1 𝑥2
−
boundary, then Fourier cosine Transform is used. = 𝑒 4𝑡
In either case, the Partial Differential Equations √2𝑡
𝑡ℎ𝑢𝑠 𝑏𝑦 𝑐𝑜𝑛𝑣𝑜𝑙𝑢𝑡𝑖𝑜𝑛 𝑡ℎ𝑒𝑜𝑟𝑒𝑚 𝑡ℎ𝑒 𝑟𝑖𝑔ℎ𝑡 𝑠𝑖𝑑𝑒 𝑜𝑓
reduces to an Ordinary Differential Equations in
Fourier Transform which is solved. Then Fourier (2)ℎ𝑎𝑠 𝑡ℎ𝑒 𝑣𝑎𝑙𝑢𝑒
∞
cosine or sine Transform will give the solution to 1
= ∫ 𝑓(𝜉)𝑔(𝑥 − 𝜉)𝑑𝜉
the problem. √2𝜋 −∞
𝑐𝑜𝑛𝑠𝑒𝑞𝑢𝑎𝑛𝑡𝑙𝑦 , 𝑡ℎ𝑒 𝑠𝑜𝑙𝑢𝑡𝑖𝑜𝑛 𝑡𝑎𝑘𝑒𝑠 𝑡ℎ𝑒 𝑓𝑜𝑟𝑚
One example is of the heat equation in an infinite ∞
1 2
rod 𝑢(𝑥, 𝑡) = ∫ 𝑓(𝜉)𝑒 −(𝑥−𝜉) /4𝑡 𝑑𝝃
2√𝜋𝑡 −∞
Heat equation in infinite rod
0, 𝑥 < 0
𝑐𝑜𝑛𝑠𝑖𝑑𝑒𝑟 𝑡ℎ𝑒 𝑐𝑎𝑠𝑒 𝑤ℎ𝑒𝑟𝑒 𝑓(𝑥) = {
𝑎, 𝑥 > 0
𝑇ℎ𝑒𝑛 𝑢(𝑥, 𝑡) 𝑏𝑒𝑐𝑜𝑚𝑒𝑠
∞ (𝑥−𝜉)2
𝑎
𝑢(𝑥, 𝑡) = ∫ 𝑒 − 4𝑡 𝑑𝝃
u(x,t) is bounded the Fourier of u(x,t) is the 2√𝜋𝑡 0
variable x is 𝑤𝑒 𝑖𝑛𝑡𝑟𝑜𝑑𝑢𝑐𝑒 𝑡ℎ𝑒 𝑛𝑒𝑤 𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒 𝜇
1 ∞
𝜉−𝑥
U(α,t)= ∫ 𝑢 (𝑥, 𝑡)𝑒 𝑖𝛼𝑥 𝑑𝑥 = 𝑠𝑜 𝑡ℎ𝑒 𝑖𝑛𝑡𝑒𝑔𝑟𝑎𝑙 𝑡𝑟𝑎𝑛𝑠𝑓𝑜𝑟𝑚𝑠 𝑖𝑛𝑡𝑜
√2𝜋 −∞ 2√𝑡
𝑢𝑥𝑥 = (−𝑖𝛼)2 𝐹 [𝑢] = (−𝑖𝛼)2 𝐹[𝑢] 𝑎 ∞
2
𝑢(𝑥, 𝑡) = ∫ 𝑒 −𝜇 𝑑𝝁
= −𝛼 2 𝑈(𝛼, 𝑡) √𝜋 −
𝑥
2√𝑡
1 ∞
And we have 𝐹[𝑢𝑡 ] = ∫ 𝑢 𝑒 𝑖𝛼𝑥 𝑑𝑥
√2𝜋 −∞ 𝑡
= 𝑈𝑡 0 ∞
𝑎 2 𝑎 2
= ∫ 𝑒 −𝜇 𝑑𝝁 + ∫ 𝑒 −𝜇
√𝜋 𝑥 √𝜋
The heat equation transform into ODE equation −
2√𝑡
0

𝑈𝑡 + 𝛼 2 𝑈 = 0 𝑠𝑜 𝑡ℎ𝑒 𝑠𝑜𝑙𝑢𝑡𝑖𝑜𝑛 𝑜𝑓 𝑤ℎ𝑖𝑐ℎ 𝑖𝑠


2
𝑈(𝛼, 𝑡) = 𝐹(𝛼)𝑒 −𝛼 𝑡 (1)
Since,
∞
𝑎 2 𝑎
∫ 𝑒 −𝜇 𝑑𝜇 =
√𝜋 0 2

𝑥
𝑎 𝑎 2√𝑡 −𝜇2
𝒔𝑜 𝑤𝑒 ℎ𝑎𝑣𝑒 𝑢(𝑥, 𝑡) = + ∫ 𝑒 𝑑𝝁
2 √𝜋 0
𝑎 𝑥
= [1 + erf ( )]
2 2√𝑡

You might also like