FOURIER TRANSFORMS
Solution: Taking Fourier Cosine transform of. = Now taking Inverse Fourier 2.5 Applications of Fourier Transforms to boundary value problems. Partial ...
SEC. 11.8 - Fourier Cosine and Sine Transforms
These transformations are of interest mainly as tools for solving ODEs PDEs
The infinite Fourier transform - Sine and Cosine transform
∫. (b). (. )( ) 2. 2. 2. 2. 2. 0 x dx. x a x b. ∞. +. +. ∫. using Fourier cosine and sine transform. Solution: (i). ( ). (. ) ( ). 0. 2 cos c. F f x. f x.
462 SECTION 11.6 §11.6 Application of Fourier Sine and Cosine
Problems. Fourier sine and cosine transforms are used to solve initial boundary value problems associated with second order partial differential equations on
Chapter 3 - Sine and Cosine Transforms
In spectral analysis of real sequences in solutions of some boundary value problems
STUDY MATERIAL FOR B.SC. MATHEMATICS TRANSFORMS
Problems on Fourier cosine transform: Problem 1: Find the fourier cosine transform of. x>0. Solution: 2 . . = Problem 2: Find the fourier cosine transform of.
MATHEMATICAL PHYSICS UNIT – 5 FOURIER TRANSFORMS
Fourier sine transform and Fourier cosine transform one can solve many important problems of physics with very simple way. • Thus we will learn from this
Fourier Transform Solutions of PDEs In this chapter we show how
Fourier sine/cosine transform when solving. PDEs involving second order ... The techniques for solving these problems are very similar and we only indicate how ...
Chapter 14: Fourier Transforms and Boundary Value Problems in an
٢٦/١٢/٢٠١٣ From the homogeneous boundary conditions we can conclude that the solution u(x
FOURIER TRANSFORMS
Example 4 Show that Fourier sine and cosine transforms of are and respectively. 2.5 Applications of Fourier Transforms to boundary value problems.
The infinite Fourier transform - Sine and Cosine transform
If FC (s) and Gc(s) are the Fourier cosine transforms and Fs(s) and Gs(s) are the PROBLEMS. Page 5. SATHYABAMA UNIVERSITY UNIT III SMT1201 ENGINEERING ...
Chapter 3 - Sine and Cosine Transforms
Fourier Cosine Transforms • Examples on the Use of Some In spectral analysis of real sequences in solutions of some boundary value problems
Fourier Transforms
The examples of a kernel are as a Fourier cosine transform of f(t) ... Solution: The Fourier cosine transform is.
16 integral transforms: fourier transforms and their applications
physics boundary value problems and signal analysis. Fourier cosine or Hankel transform
STUDY MATERIAL FOR B.SC. MATHEMATICS TRANSFORMS
FOURIER SINE AND COSINE TRANSFORMS. 16. III. FINITE FOURIER TRANSFORMS Solution: Given that ... Problems on Fourier cosine transform: Problem 1:.
FOURIER TRANSFORMS
Example 4 Show that Fourier sine and cosine transforms of are and respectively. 2.5 Applications of Fourier Transforms to boundary value problems.
FOURIER TRANSFORMS
Example 4 Show that Fourier sine and cosine transforms of are and respectively. 2.5 Applications of Fourier Transforms to boundary value problems.
Fourier Transform
Fourier Sine and Cosine transform And inverse Fourier cosine transform is given by. ( ) = ? ... 3) Solution for boundary value problems. The Fourier ...
Application of Fourier Sine and Cosine Transforms to Initial
Fourier sine and cosine transforms are used to solve initial boundary value problems associated with second order partial differential equations on the
PDE and Boundary-Value Problems
Winter Term 2014/2015
Lecture 7
Saarland University
24. November 2014
c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 1 / 26Purpose of Lesson
To introduce the idea of
integ raltr ansforms and sho who wthe y transform PDEs innvariables into differential equations inn1 variables.To introduce thesine and cosine tr ansformsand use them to solv e an infinite-diffusion problem.To define theF ourierand in verseF ouriertr ansforms,to illustr ate several useful properties of the Fourier transform and to show how these properties can be used to solve PDEs. c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 2 / 26Integral Transforms (Sine and Cosine Transforms)
General philosophy of transformsObstacleHardproblemEasyproblemSolu4onofeasyproblemSolu4onofhardproblemIntegraltransform(Step1)Inversetransform(Step3)Solvetheeasyproblem(Step2)c
Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 3 / 26Integral Transforms (Sine and Cosine Transforms)
Integral transformation
An integral transformation is a transformation that assigns to one functionf(t)a new functionF(s)by means of a formula like f(t)!F(s) =B Z AK(s;t)f(t)dt
Note that we
st artwith a function of tandend with a function of s.The functionK(s;t)is called thek ernelof tr ansformation. It is the
major ingredient that distinguishes one transform from another; itis chosen so that the transform has certain properties.The limitsAandBalso change from transformation to
transformation. c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 4 / 26Integral Transforms (Sine and Cosine Transforms)
Integral transformation
The general philosophy behind integral transformations is that they eliminate par tialder ivatives with respect to one of the variables; hence, the new equation has one less variable.ExampleIf we apply a transform to the PDE
u t=uxx for the purpose of eliminating the time derivative, then we would arrive at an ODE inx.The transform and its inverse together form what is called a transform pair c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 5 / 26 Integral Transforms (Sine and Cosine Transforms)Sine and Cosine TransformsSine and Cosine transforms
8 >>>>>:F s[f] =F(!) =2 1 Z 0 f(t)sin(!t)dt(Fourier sine transform) F1s[F] =f(t) =1
Z 0F(!)sin(!t)d!(inverse sine transform)
8 >>>>>:F c[f] =F(!) =2 1 Z 0 f(t)cos(!t)dt(Fourier cosine transform) F1c[F] =f(t) =1
Z 0F(!)cos(!t)d!(inverse cosine transform)c
Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 6 / 26 Integral Transforms (Sine and Cosine Transforms)Sine and Cosine TransformsSine and Cosine transforms of derivatives
1F s[f0] =!Fc[f](proved by integration by parts)2F s[f00] =2 !f(0)!2Fs[f]3F c[f0] =2 f(0) +!Fs[f]4F c[f00] =2 !f(0)!2Fc[f]c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 7 / 26 Integral Transforms (Sine and Cosine Transforms)Sine and Cosine TransformsFourier Sine Transform
f(x) =1R0F(!)sin(!x)dxF(!) =2
1 R0f(x)sin(!x)d!0 a F!a 2:e ax2!(a2+!2)3:x 1=2r2 4:H(ax)2
[1cos(!a)]5:x11 c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 8 / 26 Integral Transforms (Sine and Cosine Transforms)Sine and Cosine Transforms Fourier Sine Transform (cont.)
f(x) =1R 0F(!)sin(!x)dxF(!) =2
1 R 0f(x)sin(!x)d!0 x 2+a2e a!7:x x 4+41 2 e!sin(!)8:tan 1ax 1ea!! 9:x2f(x)2
F00(!)c
Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 9 / 26 Integral Transforms (Sine and Cosine Transforms)Sine and Cosine Transforms Fourier Sine Transform (cont.)
f(x) =1R 0F(!)sin(!x)dxF(!) =2
1 R 0f(x)sin(!x)d!0 x2 pa 2 1ea!2!
;a>0Here H(ax) =(1;x6a
0;x>a(Reflected Heaviside function);
erfc(x) =2p1 Z x e t2dt(complimentary-error function):c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 10 / 26 Integral Transforms (Sine and Cosine Transforms)Sine and Cosine Transforms Fourier Cosine Transform
f(x) =1R 0F(!)cos(!x)dxF(!) =2
1 R 0f(x)cos(!x)d!0 a F!a 2:e ax2a(a2+!2)3:x 1=2r2 4:H(ax)2
sin(a!)5:(x)2=c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 11 / 26 Integral Transforms (Sine and Cosine Transforms)Sine and Cosine Transforms Fourier Cosine Transform (cont.)
f(x) =1R 0F(!)cos(!x)dxF(!) =2
1 R 0f(x)cos(!x)d!0 ax21pae!2=(4a)7:sin(ax)xH(a!)8:a x 2+a2e a!9:x2f(x)2 F00(!)c
Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 12 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Solution of an Infinite-Diffusion Problem via the Sine Transform
We now show how the sine transform can solve an important IBVP (the infinite diffusion problem).Problem 7-1 To find the functionu(x;t)that satisfies
PDE:ut=2uxx;0 BC:u(0;t) =A;0 IC:u(x;0) =0;06x61
To solve this, we carry out the following steps.
c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 13 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 1. (Transformation)We transform thex-variablevia the F ouriersine tr ansformso that we get an ODE in time. F s[u] =2 1 Z 0 u(x;t)sin(!;x)dx=:U(!;t) =U(t); F s[ut] =2 1 Z 0 u t(x;t)sin(!;x)dx @@t2 4 2 1 Z 0 u(x;t)sin(!;x)dx3 5 =ddt Fs[u] =ddt
U(t); F s[uxx] =2 !u(0;t)!2Fs[u] =2A! !2U(t):c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 14 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 1. (Transformation)Transformation of the IC provides F s[u(x;0)] =U(0) =0:Substituting all these expressions into our IBVP, we change the original problem into an initial-value problem ODE:U0(t) =2
!2U(t) +2A! IC:U(0) =0:(7:1)c
Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 15 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 2. (Solving the IVP for ODE)To solve (7.1), we could use a variety of elementary techniques from ODEs (integrating factor, homogeneous and particular solution); in any case, the solution is U(t) =2A!
1e!22t
Step 3. (Inverse transform)The last step is to find the inverse transform ofU(t); that is, u(x;t) =F1s[U]: We can either evaluate the inverse transform directly from the integral or else resort to the tables. Using the tables we see that u(x;t) =Aerfcx2pt: c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 16 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Remark
The exact values of the
complimentar y-errorfunction erfc(x) =2p 1 Z x e t2dt can be found in special tables.There is also the so-callederror function erf(x) =2p x Z 0 e t2dterf(x) +erfc(x) =1.The integrals in erf(x)and erfc(x)cannot be integrated by the usual elementary tricks of calculus. c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 17 / 26 The Fourier Transform and Its Application to PDEs
Exponential Fourier transforms:
F[f]F() =1p21
Z 1h f(x)eixdxi (Fourier transform - FT) F 1[F]f(x) =1p21
Z 1h F()eixdi
(inverse FT) c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 18 / 26 The Fourier Transform and Its Application to PDEs
Exponential Fourier transforms:
Remarks
The Fourier transformF()can be acomple xfunction ; for example, the Fourier transform of f(x) =(0;x60 e x;x>0 isF() =1p21i1+2.Not all functions have Fourier transforms; in fact,f(x) =c, sin(x), e x,x2, donot ha veF ouriertr ansforms.Only functions that go to zero sufficiently fast asjxj ! 1have transforms.c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 19 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
Transformation of partial derivatives:
F[ux] =1p21
Z 1 u x(x;t)eixdx=iF[u] F[uxx] =1p21
Z 1 u xx(x;t)eixdx=2F[u] F[ut] =1p21
Z 1 u t(x;t)eixdx=@@tF[u] F[utt] =1p21
Z 1 u tt(x;t)eixdx=@2@t2F[u]c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 20 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
The Fourier transform is a linear transformation; that is F[af+bg] =aF[f] +bF[g]The transform of a product of two functionsf(x)g(x)isnot the product of the individual transforms; that is, F[f(x)g(x)]6=F[f] F[g]Convolution property:
F[fg] =F[f] F[g];
where (fg)(x) =1p21 Z 1 f(x)g()d:c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 21 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
From the convolution property it follows that
fg=F1(F[f] F[g]):Remark Hence, to find
F 1(F[f] F[g]);
all we have to do is find the inverse transform of each f actorto get f andgandthen compute their con volution.c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 22 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Exponential Fourier transform
f(x) =F1[F]F(!) =F[f]1:f (n)(x) (nthderivative)(i!)nF(!)2:f(ax);a>01 a F!a 3:f(xa)e
ia!F(!)4:equotesdbs_dbs12.pdfusesText_18
4:H(ax)2
[1cos(!a)]5:x11 c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 8 / 26 Integral Transforms (Sine and Cosine Transforms)Sine and Cosine TransformsFourier Sine Transform (cont.)
f(x) =1R0F(!)sin(!x)dxF(!) =2
1 R0f(x)sin(!x)d!0 x 2+a2e a!7:x x 4+41 2 e!sin(!)8:tan 1ax 1ea!! 9:x2f(x)2
F00(!)c
Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 9 / 26 Integral Transforms (Sine and Cosine Transforms)Sine and Cosine Transforms Fourier Sine Transform (cont.)
f(x) =1R 0F(!)sin(!x)dxF(!) =2
1 R 0f(x)sin(!x)d!0 x2 pa 2 1ea!2!
;a>0Here H(ax) =(1;x6a
0;x>a(Reflected Heaviside function);
erfc(x) =2p1 Z x e t2dt(complimentary-error function):c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 10 / 26 Integral Transforms (Sine and Cosine Transforms)Sine and Cosine Transforms Fourier Cosine Transform
f(x) =1R 0F(!)cos(!x)dxF(!) =2
1 R 0f(x)cos(!x)d!0 a F!a 2:e ax2a(a2+!2)3:x 1=2r2 4:H(ax)2
sin(a!)5:(x)2=c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 11 / 26 Integral Transforms (Sine and Cosine Transforms)Sine and Cosine Transforms Fourier Cosine Transform (cont.)
f(x) =1R 0F(!)cos(!x)dxF(!) =2
1 R 0f(x)cos(!x)d!0 ax21pae!2=(4a)7:sin(ax)xH(a!)8:a x 2+a2e a!9:x2f(x)2 F00(!)c
Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 12 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Solution of an Infinite-Diffusion Problem via the Sine Transform
We now show how the sine transform can solve an important IBVP (the infinite diffusion problem).Problem 7-1 To find the functionu(x;t)that satisfies
PDE:ut=2uxx;0 BC:u(0;t) =A;0 IC:u(x;0) =0;06x61
To solve this, we carry out the following steps.
c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 13 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 1. (Transformation)We transform thex-variablevia the F ouriersine tr ansformso that we get an ODE in time. F s[u] =2 1 Z 0 u(x;t)sin(!;x)dx=:U(!;t) =U(t); F s[ut] =2 1 Z 0 u t(x;t)sin(!;x)dx @@t2 4 2 1 Z 0 u(x;t)sin(!;x)dx3 5 =ddt Fs[u] =ddt
U(t); F s[uxx] =2 !u(0;t)!2Fs[u] =2A! !2U(t):c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 14 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 1. (Transformation)Transformation of the IC provides F s[u(x;0)] =U(0) =0:Substituting all these expressions into our IBVP, we change the original problem into an initial-value problem ODE:U0(t) =2
!2U(t) +2A! IC:U(0) =0:(7:1)c
Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 15 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 2. (Solving the IVP for ODE)To solve (7.1), we could use a variety of elementary techniques from ODEs (integrating factor, homogeneous and particular solution); in any case, the solution is U(t) =2A!
1e!22t
Step 3. (Inverse transform)The last step is to find the inverse transform ofU(t); that is, u(x;t) =F1s[U]: We can either evaluate the inverse transform directly from the integral or else resort to the tables. Using the tables we see that u(x;t) =Aerfcx2pt: c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 16 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Remark
The exact values of the
complimentar y-errorfunction erfc(x) =2p 1 Z x e t2dt can be found in special tables.There is also the so-callederror function erf(x) =2p x Z 0 e t2dterf(x) +erfc(x) =1.The integrals in erf(x)and erfc(x)cannot be integrated by the usual elementary tricks of calculus. c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 17 / 26 The Fourier Transform and Its Application to PDEs
Exponential Fourier transforms:
F[f]F() =1p21
Z 1h f(x)eixdxi (Fourier transform - FT) F 1[F]f(x) =1p21
Z 1h F()eixdi
(inverse FT) c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 18 / 26 The Fourier Transform and Its Application to PDEs
Exponential Fourier transforms:
Remarks
The Fourier transformF()can be acomple xfunction ; for example, the Fourier transform of f(x) =(0;x60 e x;x>0 isF() =1p21i1+2.Not all functions have Fourier transforms; in fact,f(x) =c, sin(x), e x,x2, donot ha veF ouriertr ansforms.Only functions that go to zero sufficiently fast asjxj ! 1have transforms.c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 19 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
Transformation of partial derivatives:
F[ux] =1p21
Z 1 u x(x;t)eixdx=iF[u] F[uxx] =1p21
Z 1 u xx(x;t)eixdx=2F[u] F[ut] =1p21
Z 1 u t(x;t)eixdx=@@tF[u] F[utt] =1p21
Z 1 u tt(x;t)eixdx=@2@t2F[u]c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 20 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
The Fourier transform is a linear transformation; that is F[af+bg] =aF[f] +bF[g]The transform of a product of two functionsf(x)g(x)isnot the product of the individual transforms; that is, F[f(x)g(x)]6=F[f] F[g]Convolution property:
F[fg] =F[f] F[g];
where (fg)(x) =1p21 Z 1 f(x)g()d:c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 21 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
From the convolution property it follows that
fg=F1(F[f] F[g]):Remark Hence, to find
F 1(F[f] F[g]);
all we have to do is find the inverse transform of each f actorto get f andgandthen compute their con volution.c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 22 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Exponential Fourier transform
f(x) =F1[F]F(!) =F[f]1:f (n)(x) (nthderivative)(i!)nF(!)2:f(ax);a>01 a F!a 3:f(xa)e
ia!F(!)4:equotesdbs_dbs12.pdfusesText_18
9:x2f(x)2
F00(!)c
Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 9 / 26 Integral Transforms (Sine and Cosine Transforms)Sine and Cosine TransformsFourier Sine Transform (cont.)
f(x) =1R0F(!)sin(!x)dxF(!) =2
1 R0f(x)sin(!x)d!0 x2 pa 2 1ea!2!
;a>0Here H(ax) =(1;x6a
0;x>a(Reflected Heaviside function);
erfc(x) =2p1 Z x e t2dt(complimentary-error function):c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 10 / 26 Integral Transforms (Sine and Cosine Transforms)Sine and Cosine Transforms Fourier Cosine Transform
f(x) =1R 0F(!)cos(!x)dxF(!) =2
1 R 0f(x)cos(!x)d!0 a F!a 2:e ax2a(a2+!2)3:x 1=2r2 4:H(ax)2
sin(a!)5:(x)2=c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 11 / 26 Integral Transforms (Sine and Cosine Transforms)Sine and Cosine Transforms Fourier Cosine Transform (cont.)
f(x) =1R 0F(!)cos(!x)dxF(!) =2
1 R 0f(x)cos(!x)d!0 ax21pae!2=(4a)7:sin(ax)xH(a!)8:a x 2+a2e a!9:x2f(x)2 F00(!)c
Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 12 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Solution of an Infinite-Diffusion Problem via the Sine Transform
We now show how the sine transform can solve an important IBVP (the infinite diffusion problem).Problem 7-1 To find the functionu(x;t)that satisfies
PDE:ut=2uxx;0 BC:u(0;t) =A;0 IC:u(x;0) =0;06x61
To solve this, we carry out the following steps.
c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 13 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 1. (Transformation)We transform thex-variablevia the F ouriersine tr ansformso that we get an ODE in time. F s[u] =2 1 Z 0 u(x;t)sin(!;x)dx=:U(!;t) =U(t); F s[ut] =2 1 Z 0 u t(x;t)sin(!;x)dx @@t2 4 2 1 Z 0 u(x;t)sin(!;x)dx3 5 =ddt Fs[u] =ddt
U(t); F s[uxx] =2 !u(0;t)!2Fs[u] =2A! !2U(t):c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 14 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 1. (Transformation)Transformation of the IC provides F s[u(x;0)] =U(0) =0:Substituting all these expressions into our IBVP, we change the original problem into an initial-value problem ODE:U0(t) =2
!2U(t) +2A! IC:U(0) =0:(7:1)c
Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 15 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 2. (Solving the IVP for ODE)To solve (7.1), we could use a variety of elementary techniques from ODEs (integrating factor, homogeneous and particular solution); in any case, the solution is U(t) =2A!
1e!22t
Step 3. (Inverse transform)The last step is to find the inverse transform ofU(t); that is, u(x;t) =F1s[U]: We can either evaluate the inverse transform directly from the integral or else resort to the tables. Using the tables we see that u(x;t) =Aerfcx2pt: c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 16 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Remark
The exact values of the
complimentar y-errorfunction erfc(x) =2p 1 Z x e t2dt can be found in special tables.There is also the so-callederror function erf(x) =2p x Z 0 e t2dterf(x) +erfc(x) =1.The integrals in erf(x)and erfc(x)cannot be integrated by the usual elementary tricks of calculus. c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 17 / 26 The Fourier Transform and Its Application to PDEs
Exponential Fourier transforms:
F[f]F() =1p21
Z 1h f(x)eixdxi (Fourier transform - FT) F 1[F]f(x) =1p21
Z 1h F()eixdi
(inverse FT) c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 18 / 26 The Fourier Transform and Its Application to PDEs
Exponential Fourier transforms:
Remarks
The Fourier transformF()can be acomple xfunction ; for example, the Fourier transform of f(x) =(0;x60 e x;x>0 isF() =1p21i1+2.Not all functions have Fourier transforms; in fact,f(x) =c, sin(x), e x,x2, donot ha veF ouriertr ansforms.Only functions that go to zero sufficiently fast asjxj ! 1have transforms.c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 19 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
Transformation of partial derivatives:
F[ux] =1p21
Z 1 u x(x;t)eixdx=iF[u] F[uxx] =1p21
Z 1 u xx(x;t)eixdx=2F[u] F[ut] =1p21
Z 1 u t(x;t)eixdx=@@tF[u] F[utt] =1p21
Z 1 u tt(x;t)eixdx=@2@t2F[u]c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 20 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
The Fourier transform is a linear transformation; that is F[af+bg] =aF[f] +bF[g]The transform of a product of two functionsf(x)g(x)isnot the product of the individual transforms; that is, F[f(x)g(x)]6=F[f] F[g]Convolution property:
F[fg] =F[f] F[g];
where (fg)(x) =1p21 Z 1 f(x)g()d:c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 21 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
From the convolution property it follows that
fg=F1(F[f] F[g]):Remark Hence, to find
F 1(F[f] F[g]);
all we have to do is find the inverse transform of each f actorto get f andgandthen compute their con volution.c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 22 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Exponential Fourier transform
f(x) =F1[F]F(!) =F[f]1:f (n)(x) (nthderivative)(i!)nF(!)2:f(ax);a>01 a F!a 3:f(xa)e
ia!F(!)4:equotesdbs_dbs12.pdfusesText_18
1ea!2!
;a>0HereH(ax) =(1;x6a
0;x>a(Reflected Heaviside function);
erfc(x) =2p1 Z x e t2dt(complimentary-error function):c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 10 / 26 Integral Transforms (Sine and Cosine Transforms)Sine and Cosine TransformsFourier Cosine Transform
f(x) =1R0F(!)cos(!x)dxF(!) =2
1 R0f(x)cos(!x)d!0 a F!a 2:e ax2a(a2+!2)3:x 1=2r2 4:H(ax)2
sin(a!)5:(x)2=c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 11 / 26 Integral Transforms (Sine and Cosine Transforms)Sine and Cosine Transforms Fourier Cosine Transform (cont.)
f(x) =1R 0F(!)cos(!x)dxF(!) =2
1 R 0f(x)cos(!x)d!0 ax21pae!2=(4a)7:sin(ax)xH(a!)8:a x 2+a2e a!9:x2f(x)2 F00(!)c
Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 12 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Solution of an Infinite-Diffusion Problem via the Sine Transform
We now show how the sine transform can solve an important IBVP (the infinite diffusion problem).Problem 7-1 To find the functionu(x;t)that satisfies
PDE:ut=2uxx;0 BC:u(0;t) =A;0 IC:u(x;0) =0;06x61
To solve this, we carry out the following steps.
c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 13 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 1. (Transformation)We transform thex-variablevia the F ouriersine tr ansformso that we get an ODE in time. F s[u] =2 1 Z 0 u(x;t)sin(!;x)dx=:U(!;t) =U(t); F s[ut] =2 1 Z 0 u t(x;t)sin(!;x)dx @@t2 4 2 1 Z 0 u(x;t)sin(!;x)dx3 5 =ddt Fs[u] =ddt
U(t); F s[uxx] =2 !u(0;t)!2Fs[u] =2A! !2U(t):c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 14 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 1. (Transformation)Transformation of the IC provides F s[u(x;0)] =U(0) =0:Substituting all these expressions into our IBVP, we change the original problem into an initial-value problem ODE:U0(t) =2
!2U(t) +2A! IC:U(0) =0:(7:1)c
Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 15 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 2. (Solving the IVP for ODE)To solve (7.1), we could use a variety of elementary techniques from ODEs (integrating factor, homogeneous and particular solution); in any case, the solution is U(t) =2A!
1e!22t
Step 3. (Inverse transform)The last step is to find the inverse transform ofU(t); that is, u(x;t) =F1s[U]: We can either evaluate the inverse transform directly from the integral or else resort to the tables. Using the tables we see that u(x;t) =Aerfcx2pt: c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 16 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Remark
The exact values of the
complimentar y-errorfunction erfc(x) =2p 1 Z x e t2dt can be found in special tables.There is also the so-callederror function erf(x) =2p x Z 0 e t2dterf(x) +erfc(x) =1.The integrals in erf(x)and erfc(x)cannot be integrated by the usual elementary tricks of calculus. c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 17 / 26 The Fourier Transform and Its Application to PDEs
Exponential Fourier transforms:
F[f]F() =1p21
Z 1h f(x)eixdxi (Fourier transform - FT) F 1[F]f(x) =1p21
Z 1h F()eixdi
(inverse FT) c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 18 / 26 The Fourier Transform and Its Application to PDEs
Exponential Fourier transforms:
Remarks
The Fourier transformF()can be acomple xfunction ; for example, the Fourier transform of f(x) =(0;x60 e x;x>0 isF() =1p21i1+2.Not all functions have Fourier transforms; in fact,f(x) =c, sin(x), e x,x2, donot ha veF ouriertr ansforms.Only functions that go to zero sufficiently fast asjxj ! 1have transforms.c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 19 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
Transformation of partial derivatives:
F[ux] =1p21
Z 1 u x(x;t)eixdx=iF[u] F[uxx] =1p21
Z 1 u xx(x;t)eixdx=2F[u] F[ut] =1p21
Z 1 u t(x;t)eixdx=@@tF[u] F[utt] =1p21
Z 1 u tt(x;t)eixdx=@2@t2F[u]c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 20 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
The Fourier transform is a linear transformation; that is F[af+bg] =aF[f] +bF[g]The transform of a product of two functionsf(x)g(x)isnot the product of the individual transforms; that is, F[f(x)g(x)]6=F[f] F[g]Convolution property:
F[fg] =F[f] F[g];
where (fg)(x) =1p21 Z 1 f(x)g()d:c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 21 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
From the convolution property it follows that
fg=F1(F[f] F[g]):Remark Hence, to find
F 1(F[f] F[g]);
all we have to do is find the inverse transform of each f actorto get f andgandthen compute their con volution.c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 22 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Exponential Fourier transform
f(x) =F1[F]F(!) =F[f]1:f (n)(x) (nthderivative)(i!)nF(!)2:f(ax);a>01 a F!a 3:f(xa)e
ia!F(!)4:equotesdbs_dbs12.pdfusesText_18
4:H(ax)2
sin(a!)5:(x)2=c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 11 / 26 Integral Transforms (Sine and Cosine Transforms)Sine and Cosine TransformsFourier Cosine Transform (cont.)
f(x) =1R0F(!)cos(!x)dxF(!) =2
1 R0f(x)cos(!x)d!0 ax21pae!2=(4a)7:sin(ax)xH(a!)8:a x 2+a2e a!9:x2f(x)2 F00(!)c
Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 12 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Solution of an Infinite-Diffusion Problem via the Sine Transform
We now show how the sine transform can solve an important IBVP (the infinite diffusion problem).Problem 7-1 To find the functionu(x;t)that satisfies
PDE:ut=2uxx;0 BC:u(0;t) =A;0 IC:u(x;0) =0;06x61
To solve this, we carry out the following steps.
c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 13 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 1. (Transformation)We transform thex-variablevia the F ouriersine tr ansformso that we get an ODE in time. F s[u] =2 1 Z 0 u(x;t)sin(!;x)dx=:U(!;t) =U(t); F s[ut] =2 1 Z 0 u t(x;t)sin(!;x)dx @@t2 4 2 1 Z 0 u(x;t)sin(!;x)dx3 5 =ddt Fs[u] =ddt
U(t); F s[uxx] =2 !u(0;t)!2Fs[u] =2A! !2U(t):c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 14 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 1. (Transformation)Transformation of the IC provides F s[u(x;0)] =U(0) =0:Substituting all these expressions into our IBVP, we change the original problem into an initial-value problem ODE:U0(t) =2
!2U(t) +2A! IC:U(0) =0:(7:1)c
Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 15 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 2. (Solving the IVP for ODE)To solve (7.1), we could use a variety of elementary techniques from ODEs (integrating factor, homogeneous and particular solution); in any case, the solution is U(t) =2A!
1e!22t
Step 3. (Inverse transform)The last step is to find the inverse transform ofU(t); that is, u(x;t) =F1s[U]: We can either evaluate the inverse transform directly from the integral or else resort to the tables. Using the tables we see that u(x;t) =Aerfcx2pt: c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 16 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Remark
The exact values of the
complimentar y-errorfunction erfc(x) =2p 1 Z x e t2dt can be found in special tables.There is also the so-callederror function erf(x) =2p x Z 0 e t2dterf(x) +erfc(x) =1.The integrals in erf(x)and erfc(x)cannot be integrated by the usual elementary tricks of calculus. c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 17 / 26 The Fourier Transform and Its Application to PDEs
Exponential Fourier transforms:
F[f]F() =1p21
Z 1h f(x)eixdxi (Fourier transform - FT) F 1[F]f(x) =1p21
Z 1h F()eixdi
(inverse FT) c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 18 / 26 The Fourier Transform and Its Application to PDEs
Exponential Fourier transforms:
Remarks
The Fourier transformF()can be acomple xfunction ; for example, the Fourier transform of f(x) =(0;x60 e x;x>0 isF() =1p21i1+2.Not all functions have Fourier transforms; in fact,f(x) =c, sin(x), e x,x2, donot ha veF ouriertr ansforms.Only functions that go to zero sufficiently fast asjxj ! 1have transforms.c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 19 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
Transformation of partial derivatives:
F[ux] =1p21
Z 1 u x(x;t)eixdx=iF[u] F[uxx] =1p21
Z 1 u xx(x;t)eixdx=2F[u] F[ut] =1p21
Z 1 u t(x;t)eixdx=@@tF[u] F[utt] =1p21
Z 1 u tt(x;t)eixdx=@2@t2F[u]c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 20 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
The Fourier transform is a linear transformation; that is F[af+bg] =aF[f] +bF[g]The transform of a product of two functionsf(x)g(x)isnot the product of the individual transforms; that is, F[f(x)g(x)]6=F[f] F[g]Convolution property:
F[fg] =F[f] F[g];
where (fg)(x) =1p21 Z 1 f(x)g()d:c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 21 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
From the convolution property it follows that
fg=F1(F[f] F[g]):Remark Hence, to find
F 1(F[f] F[g]);
all we have to do is find the inverse transform of each f actorto get f andgandthen compute their con volution.c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 22 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Exponential Fourier transform
f(x) =F1[F]F(!) =F[f]1:f (n)(x) (nthderivative)(i!)nF(!)2:f(ax);a>01 a F!a 3:f(xa)e
ia!F(!)4:equotesdbs_dbs12.pdfusesText_18
F00(!)c
Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 12 / 26Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Solution of an Infinite-Diffusion Problem via the SineTransform
We now show how the sine transform can solve an important IBVP (the infinite diffusion problem).Problem 7-1To find the functionu(x;t)that satisfies
PDE:ut=2uxx;0 BC:u(0;t) =A;0 IC:u(x;0) =0;06x61
To solve this, we carry out the following steps.
c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 13 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 1. (Transformation)We transform thex-variablevia the F ouriersine tr ansformso that we get an ODE in time. F s[u] =2 1 Z 0 u(x;t)sin(!;x)dx=:U(!;t) =U(t); F s[ut] =2 1 Z 0 u t(x;t)sin(!;x)dx @@t2 4 2 1 Z 0 u(x;t)sin(!;x)dx3 5 =ddt Fs[u] =ddt
U(t); F s[uxx] =2 !u(0;t)!2Fs[u] =2A! !2U(t):c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 14 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 1. (Transformation)Transformation of the IC provides F s[u(x;0)] =U(0) =0:Substituting all these expressions into our IBVP, we change the original problem into an initial-value problem ODE:U0(t) =2
!2U(t) +2A! IC:U(0) =0:(7:1)c
Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 15 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 2. (Solving the IVP for ODE)To solve (7.1), we could use a variety of elementary techniques from ODEs (integrating factor, homogeneous and particular solution); in any case, the solution is U(t) =2A!
1e!22t
Step 3. (Inverse transform)The last step is to find the inverse transform ofU(t); that is, u(x;t) =F1s[U]: We can either evaluate the inverse transform directly from the integral or else resort to the tables. Using the tables we see that u(x;t) =Aerfcx2pt: c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 16 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Remark
The exact values of the
complimentar y-errorfunction erfc(x) =2p 1 Z x e t2dt can be found in special tables.There is also the so-callederror function erf(x) =2p x Z 0 e t2dterf(x) +erfc(x) =1.The integrals in erf(x)and erfc(x)cannot be integrated by the usual elementary tricks of calculus. c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 17 / 26 The Fourier Transform and Its Application to PDEs
Exponential Fourier transforms:
F[f]F() =1p21
Z 1h f(x)eixdxi (Fourier transform - FT) F 1[F]f(x) =1p21
Z 1h F()eixdi
(inverse FT) c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 18 / 26 The Fourier Transform and Its Application to PDEs
Exponential Fourier transforms:
Remarks
The Fourier transformF()can be acomple xfunction ; for example, the Fourier transform of f(x) =(0;x60 e x;x>0 isF() =1p21i1+2.Not all functions have Fourier transforms; in fact,f(x) =c, sin(x), e x,x2, donot ha veF ouriertr ansforms.Only functions that go to zero sufficiently fast asjxj ! 1have transforms.c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 19 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
Transformation of partial derivatives:
F[ux] =1p21
Z 1 u x(x;t)eixdx=iF[u] F[uxx] =1p21
Z 1 u xx(x;t)eixdx=2F[u] F[ut] =1p21
Z 1 u t(x;t)eixdx=@@tF[u] F[utt] =1p21
Z 1 u tt(x;t)eixdx=@2@t2F[u]c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 20 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
The Fourier transform is a linear transformation; that is F[af+bg] =aF[f] +bF[g]The transform of a product of two functionsf(x)g(x)isnot the product of the individual transforms; that is, F[f(x)g(x)]6=F[f] F[g]Convolution property:
F[fg] =F[f] F[g];
where (fg)(x) =1p21 Z 1 f(x)g()d:c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 21 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
From the convolution property it follows that
fg=F1(F[f] F[g]):Remark Hence, to find
F 1(F[f] F[g]);
all we have to do is find the inverse transform of each f actorto get f andgandthen compute their con volution.c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 22 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Exponential Fourier transform
f(x) =F1[F]F(!) =F[f]1:f (n)(x) (nthderivative)(i!)nF(!)2:f(ax);a>01 a F!a 3:f(xa)e
ia!F(!)4:equotesdbs_dbs12.pdfusesText_18
BC:u(0;t) =A;0 IC:u(x;0) =0;06x61
To solve this, we carry out the following steps.
c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 13 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 1. (Transformation)We transform thex-variablevia the F ouriersine tr ansformso that we get an ODE in time. F s[u] =2 1 Z 0 u(x;t)sin(!;x)dx=:U(!;t) =U(t); F s[ut] =2 1 Z 0 u t(x;t)sin(!;x)dx @@t2 4 2 1 Z 0 u(x;t)sin(!;x)dx3 5 =ddt Fs[u] =ddt
U(t); F s[uxx] =2 !u(0;t)!2Fs[u] =2A! !2U(t):c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 14 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 1. (Transformation)Transformation of the IC provides F s[u(x;0)] =U(0) =0:Substituting all these expressions into our IBVP, we change the original problem into an initial-value problem ODE:U0(t) =2
!2U(t) +2A! IC:U(0) =0:(7:1)c
Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 15 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 2. (Solving the IVP for ODE)To solve (7.1), we could use a variety of elementary techniques from ODEs (integrating factor, homogeneous and particular solution); in any case, the solution is U(t) =2A!
1e!22t
Step 3. (Inverse transform)The last step is to find the inverse transform ofU(t); that is, u(x;t) =F1s[U]: We can either evaluate the inverse transform directly from the integral or else resort to the tables. Using the tables we see that u(x;t) =Aerfcx2pt: c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 16 / 26 Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Remark
The exact values of the
complimentar y-errorfunction erfc(x) =2p 1 Z x e t2dt can be found in special tables.There is also the so-callederror function erf(x) =2p x Z 0 e t2dterf(x) +erfc(x) =1.The integrals in erf(x)and erfc(x)cannot be integrated by the usual elementary tricks of calculus. c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 17 / 26 The Fourier Transform and Its Application to PDEs
Exponential Fourier transforms:
F[f]F() =1p21
Z 1h f(x)eixdxi (Fourier transform - FT) F 1[F]f(x) =1p21
Z 1h F()eixdi
(inverse FT) c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 18 / 26 The Fourier Transform and Its Application to PDEs
Exponential Fourier transforms:
Remarks
The Fourier transformF()can be acomple xfunction ; for example, the Fourier transform of f(x) =(0;x60 e x;x>0 isF() =1p21i1+2.Not all functions have Fourier transforms; in fact,f(x) =c, sin(x), e x,x2, donot ha veF ouriertr ansforms.Only functions that go to zero sufficiently fast asjxj ! 1have transforms.c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 19 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
Transformation of partial derivatives:
F[ux] =1p21
Z 1 u x(x;t)eixdx=iF[u] F[uxx] =1p21
Z 1 u xx(x;t)eixdx=2F[u] F[ut] =1p21
Z 1 u t(x;t)eixdx=@@tF[u] F[utt] =1p21
Z 1 u tt(x;t)eixdx=@2@t2F[u]c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 20 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
The Fourier transform is a linear transformation; that is F[af+bg] =aF[f] +bF[g]The transform of a product of two functionsf(x)g(x)isnot the product of the individual transforms; that is, F[f(x)g(x)]6=F[f] F[g]Convolution property:
F[fg] =F[f] F[g];
where (fg)(x) =1p21 Z 1 f(x)g()d:c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 21 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Properties of the Fourier Transform:
From the convolution property it follows that
fg=F1(F[f] F[g]):Remark Hence, to find
F 1(F[f] F[g]);
all we have to do is find the inverse transform of each f actorto get f andgandthen compute their con volution.c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 22 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier Transform Exponential Fourier transform
f(x) =F1[F]F(!) =F[f]1:f (n)(x) (nthderivative)(i!)nF(!)2:f(ax);a>01 a F!a 3:f(xa)e
ia!F(!)4:equotesdbs_dbs12.pdfusesText_18
IC:u(x;0) =0;06x61
To solve this, we carry out the following steps.
c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 13 / 26Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 1. (Transformation)We transform thex-variablevia the F ouriersine tr ansformso that we get an ODE in time. F s[u] =2 1 Z 0 u(x;t)sin(!;x)dx=:U(!;t) =U(t); F s[ut] =2 1 Z 0 u t(x;t)sin(!;x)dx @@t2 4 2 1 Z 0 u(x;t)sin(!;x)dx3 5 =ddtFs[u] =ddt
U(t); F s[uxx] =2 !u(0;t)!2Fs[u] =2A! !2U(t):c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 14 / 26Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 1. (Transformation)Transformation of the IC provides F s[u(x;0)] =U(0) =0:Substituting all these expressions into our IBVP, we change the original problem into an initial-value problemODE:U0(t) =2
!2U(t) +2A!IC:U(0) =0:(7:1)c
Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 15 / 26Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Step 2. (Solving the IVP for ODE)To solve (7.1), we could use a variety of elementary techniques from ODEs (integrating factor, homogeneous and particular solution); in any case, the solution isU(t) =2A!
1e!22t
Step 3. (Inverse transform)The last step is to find the inverse transform ofU(t); that is, u(x;t) =F1s[U]: We can either evaluate the inverse transform directly from the integral or else resort to the tables. Using the tables we see that u(x;t) =Aerfcx2pt: c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 16 / 26Integral Transforms (Sine and Cosine Transforms)Solution of an Infinite-Diffusion Problem via the Sine Transform
Remark
The exact values of the
complimentar y-errorfunction erfc(x) =2p 1 Z x e t2dt can be found in special tables.There is also the so-callederror function erf(x) =2p x Z 0 e t2dterf(x) +erfc(x) =1.The integrals in erf(x)and erfc(x)cannot be integrated by the usual elementary tricks of calculus. c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 17 / 26The Fourier Transform and Its Application to PDEs
Exponential Fourier transforms:
F[f]F() =1p21
Z 1h f(x)eixdxi (Fourier transform - FT) F1[F]f(x) =1p21
Z 1hF()eixdi
(inverse FT) c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 18 / 26The Fourier Transform and Its Application to PDEs
Exponential Fourier transforms:
Remarks
The Fourier transformF()can be acomple xfunction ; for example, the Fourier transform of f(x) =(0;x60 e x;x>0 isF() =1p21i1+2.Not all functions have Fourier transforms; in fact,f(x) =c, sin(x), e x,x2, donot ha veF ouriertr ansforms.Only functions that go to zero sufficiently fast asjxj ! 1have transforms.c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 19 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier TransformProperties of the Fourier Transform:
Transformation of partial derivatives:
F[ux] =1p21
Z 1 u x(x;t)eixdx=iF[u]F[uxx] =1p21
Z 1 u xx(x;t)eixdx=2F[u]F[ut] =1p21
Z 1 u t(x;t)eixdx=@@tF[u]F[utt] =1p21
Z 1 u tt(x;t)eixdx=@2@t2F[u]c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 20 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier TransformProperties of the Fourier Transform:
The Fourier transform is a linear transformation; that is F[af+bg] =aF[f] +bF[g]The transform of a product of two functionsf(x)g(x)isnot the product of the individual transforms; that is,F[f(x)g(x)]6=F[f] F[g]Convolution property:
F[fg] =F[f] F[g];
where (fg)(x) =1p21 Z 1 f(x)g()d:c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 21 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier TransformProperties of the Fourier Transform:
From the convolution property it follows that
fg=F1(F[f] F[g]):RemarkHence, to find
F1(F[f] F[g]);
all we have to do is find the inverse transform of each f actorto get f andgandthen compute their con volution.c Daria Apushkinskaya (UdS)PDE and BVP lecture 724. November 2014 22 / 26 The Fourier Transform and Its Application to PDEsProperties of the Fourier TransformExponential Fourier transform
f(x) =F1[F]F(!) =F[f]1:f (n)(x) (nthderivative)(i!)nF(!)2:f(ax);a>01 a F!a3:f(xa)e
ia!F(!)4:equotesdbs_dbs12.pdfusesText_18[PDF] fourier series and fourier transform difference
[PDF] fourier series applications in computer science
[PDF] fourier series applications in electrical engineering
[PDF] fourier series applications in engineering
[PDF] fourier series applications in real life pdf
[PDF] fourier series applications in signal processing
[PDF] fourier series applications in signal processing ppt
[PDF] fourier series boundary conditions
[PDF] fourier series calculator wolfram
[PDF] fourier series code
[PDF] fourier series coefficients online calculator
[PDF] fourier series even square wave
[PDF] fourier series graph examples
[PDF] fourier series graph interactive