• 검색 결과가 없습니다.

11.5. Heat Equation: Solution by Fourier Series

N/A
N/A
Protected

Academic year: 2021

Share "11.5. Heat Equation: Solution by Fourier Series"

Copied!
9
0
0

로드 중.... (전체 텍스트 보기)

전체 글

(1)

11.5. Heat Equation: Solution by Fourier Series

u: temperature, c2: thermal diffusivity ρ: density, σ: specific heat

u(t,x,y,z)

 Simple case: temp. distribution in a long thin bar or wire: u(t,x)

B.C.s: u(0,t) = 0, u(L,t) = 0 for all t I.C.: u(x,0) = f(x) (f(0) = f(L) = 0) - Same procedure as Section 11.3

First Step: Two ODEs let u(x,t) = F(x)G(t)

Second Step: Satisfying the B.C.’s



 

= ρσ

∂ =

∂ K

c u

t c

u 2 2 2

x = 0 x = L

2 2 2

x c u t u

= ∂

∂ Parabolic Eqn. (B2-4AC=0)

0 G p c G

0 F p '' F

2 2 2

= +

= +

) , 2 , 1 n ( 1 B for L x

sin n )

x ( F

0 ) L ( F ) 0 ( F : s ' C . B apply px

sin B px cos A )

x ( F 0

F p '' F

n 2

=  π =

 =

=

=

← +

 =

= +

0 ) t ( G ) L ( F ) t , L ( u

0 ) t ( G ) 0 ( F ) t , 0 ( u

=

=

=

=

(For n = 0 ? F = 0 no interest)

(2)

Third Step: Solution of the Entire Problem - General solution:

I.C.:

Ex.1~3)

Ex. 4) Bar with insulated ends.: B.C.’s: ux(0,t) = F’(0)G(t) = 0, ux(L,t) = F’(L)G(t) = 0 )

, 2 , 1 n ( e

B ) t ( L G

0 cn G

G 2n n   n = n 2nt = 

 

λ = π

= λ

+ λ

) , 2 , 1 n ( e

L x sin n B

) t , x (

u

n n

2nt

= 

 

 π

=

λ

Eigenfunctions with eigenvalues λn

 

 

 λ = π

 

 

 π

=

=  

=

λ

=

L

e cn L x

sin n B

) t , x ( u )

t , x (

u

n

1 n

t n

1 n

n

2 n

) , 2 , 1 n ( L dx

x sin n

) x ( L f

B 2 );

x ( f L x

sin n B

) 0 , x (

u

L

n 0 1

n

n

 = = π = 

 

 π

= 

=

(Fourier sine series)

(all the terms 0 as time increases)

) , 2 , 0 n ( 1 A for L x

cos n )

x ( F

0 ) L ( ' F ) 0 ( ' F : s ' C . B apply px

sin B px cos A )

x ( F 0

F p '' F

n 2

=  π =

 =

=

=

← +

 =

= +

(n includes 0 !!!)

(3)



 

λ = π



 

 π

=

=

 

=

λ

= L

e cn L x

cos n A

) t , x ( u )

t , x (

u n

0 n

t n

0 n

n

2 n

) , 2 , 1 , 0 n L , ( cn

L e x cos n

A )

t , x (

u

n n

2nt

λ

n

= π = 

 

 π

=

λ

) , 2 , 1 n ( L dx

x cos n ) x ( L f

A 2 , dx ) x ( L f

A 1

);

x ( f L x

cos n A

A L x

cos n A

) 0 , x ( u

L n 0

L 0 0

1 n

n 0

0 n

n

= 

= π

=

=



 

 π

+

=



 

 π

=

=

=

(Fourier cosine series)

Steady-State 2D Heat Flow



 

 =

= ∂

∂ + ∂

 ∂



 

∂ + ∂

= ∂

∂ =

∂ 0

t 0 u

y u x

: u state steady

y at u x

c u u t c

u

2 2 2

2 2

2 2

2 2 2

2

Elliptic Eqn. (B2-4AC<0) I.C.:

(4)

From Gn(0) = 0 at y=0:

0 ) a ( F , 0 ) 0 ( F 0

dx kF F d

dy k G d G

1 dx

F d F ) 1

y ( G ) x ( F ) y , x ( u

2 2

2 2 2

2

=

=

= +

=

 =

=

) , 2 , 1 n ( a x

sin n )

x ( F

F = n = π = 

a / y n n a

/ y n n n

2 2

2

e B e

A ) y ( G ) y ( G 0

a G n dy

G

d π π

+

=

 =

=



 

 π

a y sinh n

a A y sinh n A

2 ) y (

Gn = n π = *n π

a x sin n

a y sinh n

A ) y , x (

un = *n π π

=

=

π

= π

=

1 n

* n 1

n

n a

x sin n

a y sinh n

A )

y , x ( u )

y , x ( u

π π π =

π

=

=

a 0

* n 1

n

*

n dx

a x sin n ) x ( a f

2 a

b sinh n

A a ;

x sin n

a b sinh n

A )

b , x ( u

y

x u=f(x)

u=0 u=0 0 u=0

R

a b

Dirichlet Problem in a Rectangle R

B.C.:

(5)

11.6. Heat Equation: Solution by Fourier Integrals and Transforms

- In the case of infinite bars: Fourier series  Fourier integrals

(laterally insulated, infinite ends), u(x,0) = f(x) (-∞ < x < ∞)

Use of Fourier Integrals

- f(x): nonperiodic function (p multiples of a fixed number)  use of Fourier integrals A & B: functions of p, then A(p) & B(p)

Determination of A(p) and B(p) from the Initial Condition

2 2 2

t c u t u

= ∂

0 G p c G

0 F p '' F

2 2 2

= +

= +

 ( )

c p t

t p c

2 2

2 2

e px sin B px cos A )

p

; t , x ( u

e ) t ( G

; px sin B px

cos A )

x ( F

+

=

= +

=

( A ( p ) cos px B ( p ) sin px ) e dp

dp ) p

; t , x ( u )

t , x (

u

c p t

0 0

2

2

= +

=

( )



 

= π

= π



 −

 π

= +

=

 

pvdv sin

) v ( 1 f

) p ( B ,

pvdv cos

) v ( 1 f

) p ( A

dp dv ) pv px

cos(

) v ( 1 f

) x ( f dp px sin ) p ( B px cos ) p ( A )

0 , x (

u 0 0

(6)

dv dp ) pv px

cos(

e )

v ( 1 f

) t , x (

u

0

t p c2 2



∞ −



= π





0∞ − 2 = π b2

s e

ds 2 bs 2 cos

e

 

 = = −

t c 2

v b x

, t p c s

let

2 2 2





− −

= π

0∞ −e cos(pxpv)dp 2c t exp (x4c2vt)2

t p c2 2

t dv c 4

) v x exp (

) v ( t f

c 2 ) 1 t , x (

u

2

2

= π

( let z = ( v x ) /( 2 c t ) ) u ( x , t ) = 1 π

f ( x + 2 cz t ) e

z2

dz

Ex. 1)

Use of Fourier Transforms

Ex. 2) Temperature in the infinite bar

- Fourier transform w.r.t. x and resulting ODE in t

(7)

( let = ( u ) )

t dx uˆ

t ue 2

dx 1 e

2 u ) 1

u (

uˆ w c )

u ( c ) u (

iwx iwx

t t

2 2 xx

2 t

= ∂

= π

= π ℑ

= ℑ

= ℑ

Ex. 3) Convolution method

t w c 2

2

2

e 2

) w ( C ) t , w ( uˆ

uˆ w t c

=

∂ =

= 

= C(w) fˆ(w) )

0 , w ( uˆ : . C .

I uˆ(w,t) = fˆ(w)ec2w2t

dv dw e

e )

v ( 2 f

1

dw e

e ) w ( 2 fˆ

) 1 t , x ( u : Inversion

) wv wx ( i t w c

iwx t w c

2 2

2 2

 





= π

= π

 

= π 

dv

e ) v ( 2 f

) 1 w (

iwv

dv dw ) wv wx

cos(

e )

v ( 1 f

) t , x (

u = π



c2w2t



= π fˆ(w)e e dw 2

) 1 t , x ( u : from

Starting c2w2t iwx



 

use u(x,t) = f *g(x) =

f(p)g(xp)dp =

(w)(w)eiwvdw

t w c2 2

2 e ) 1

w (

= π

(Imaginary part: odd func)

(8)

 

 

 ℑ

= e

 ℑ ( e

) = 2 c t e

= 2 c t 2 π gˆ ( w ) a

2 ) 1

e

(

ax2 w2/4a x2/4c2t 2 c2w2 2





 = tc a

4

1 2

t c 4 / x 2

2

e

2

2 t c 2 ) 1

w ( gˆ of

Inverse

 π

= π

= dp

t c 4

) p x exp (

) p ( t f

c 2 ) 1 x )(

g

* f ( ) t , x (

u

2

2

Ex. 4) Fourier sine transform applied to the heat equation

A laterally insulated bar from x=0 to infinity, (u(x,0) = f(x), u(0,t) = 0)

 

∞ ∞

= π





= π

 =

∂ =

0 0

t w c

s 0 t

w c s

s

s 2 s 2

dw dp wx sin e

wp sin ) p ( 2 f

) t , x ( u

dp wp sin ) p ( 2 f

) w ( fˆ e

) w ( fˆ ) t , w ( uˆ

uˆ w t c

2 2 2

2

) 0 ( 2wf ))

x ( f ( w

)) x ( '' f

( 2 S

S = − ℑ + π

(See Ex. 2 of Sec. 10.10 for the Fourier transform)

(9)

Example) Heating a semi-infinite slab

I.C.: u(x,0) = 0, B.C.’s: u(0,t) = 1, u(∞,t) = 0

2 2

x u t

u

∂ α ∂

∂ =

t 4

x

= α η

= η

=

= η

= η =

η η +

η α

∂ =

∂ η

− η

∂ =

at 0 u , 0 at

1 u : s . C . B with d 0

2 du d

u d

t 4

1 d

u d x

; u d du t 2 1 t

u

2 2

2 2 2

2

) ( erf 1

d ) 2 exp(

1 d

) exp(

d ) exp(

1 ) ( u

C d

) exp(

C u ), exp(

C then

d , let du

0

2

0

2 0

2

0 2

2 1

2 1

η

= η η

π −

− η =

η

η η

− −

= η

+ η η

= η

= η ψ

= ψ

 

η

η

η

Define a new variable:

error function (See Ex. 2 of Sec. 10.10 for the integration of special function)

참조

관련 문서

The purpose of the study was the identification of a simple type of a general solution for the hyperbolic partial differential equation of the

Ø Fourier series: Infinite series designed to represent general periodic functions in terms of simple ones (e.g., sines and cosines).. Ø Fourier series is more general

O’Sullivan, A fast sinc function gridding algorithm for Fourier inversion in computer tomography, IEEE Trans. Pike, Multidimensional Fourier transformation in magnetic

Fourier Series Practise by Matlab... Command

Multiband OFDM UWB의 OFDM 방식의 변조는 DFT(Discrete Fourier Transform)을 이용하는데, 실제 하드웨어 설계에서는 연산량을 줄이기 위해서 FFT(Fast

In fact, they [3] introduced the L 1 -analytic conditional Fourier- Feynman transform and the conditional convolution product over Wiener paths in abstract Wiener space and

Then we study basic properties of the transforms and all differen- tiable one parameter transformation groups induced by the quantum general- ized Fourier-Mehler transform of which

Keywords: double Fourier series, summability of double Fourier series, conver- gence of Fourier series; 이중 푸리에 급수 , 이중 푸리에 급수의 총합가능성 ,