30
Fisika Fisika Matematika Matematika III III Parabolic Partial Differential Equations Parabolic Partial Differential Equations - - Method of Separation of Variables Method of Separation of Variables Irwan Ary Irwan Ary Dharmawan Dharmawan http:// http:// phys.unpad.ac.id/jurusan/staff/dharmawan/kuliah phys.unpad.ac.id/jurusan/staff/dharmawan/kuliah

Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

  • Upload
    others

  • View
    0

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

Fisika Fisika MatematikaMatematika IIIIII

Parabolic Partial Differential EquationsParabolic Partial Differential Equations

-- Method of Separation of VariablesMethod of Separation of Variables

Irwan Ary Irwan Ary DharmawanDharmawan

http://http://phys.unpad.ac.id/jurusan/staff/dharmawan/kuliahphys.unpad.ac.id/jurusan/staff/dharmawan/kuliah

Page 2: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

After this lecture After this lecture ……....

2

2

x

uk

t

u

∂∂=

∂∂

??),( txu

Separation of Variables

With IC and BC

Page 3: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

Linearity ConceptLinearity Concept

�� A linier operator by definition satisfiesA linier operator by definition satisfies

�� ExampleExample

)()()( 22112211 uLcuLcucucL +=+

L

t

uc

t

ucucuc

t ∂∂+

∂∂=+

∂∂ 2

21

12211 )(

22

2

221

2

122112

2

)(x

uc

x

ucucuc

x ∂∂+

∂∂=+

∂∂

Page 4: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

HomogenityHomogenity

�� See the first slide of this lectureSee the first slide of this lecture

Page 5: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

Principle of SuperpositionPrinciple of Superposition

�� If and satisfy a linear homogeneous equation, If and satisfy a linear homogeneous equation,

then an arbitrary linear combination of them, then an arbitrary linear combination of them,

also satisfies the same linear homogeneous equationalso satisfies the same linear homogeneous equation

1u

2211 ucuc +2u

Page 6: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

Heat Equation with Zero Temperatures Heat Equation with Zero Temperatures

at Finite Endsat Finite Ends

�� Consider the linear Consider the linear homegeneoushomegeneous Heat Equation in a oneHeat Equation in a one--

dimensional rod ( ) with constant thermdimensional rod ( ) with constant thermal al

coeficientscoeficients and no sources of thermal energyand no sources of thermal energy

Lx <<0

2

2

x

uk

t

u

∂∂=

∂∂

)()0,( xfxu =

0)(),(

0)(),0(

2

1

====

tTtLu

tTtu

IC

BC

PDE

Page 7: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

Separation of VariablesSeparation of Variables

)()(),( tGxtxu φ=

dt

dGx

t

u)(φ=

∂∂ )(

2

2

2

2

tGdx

d

x

u φ=∂∂

2

2

x

uk

t

u

∂∂=

∂∂

Separation of Variables

Plug in to Heat Eqn.

)()(2

2

tGdx

dk

dt

dGx

φφ =

2

211

dx

d

dt

dG

kG

φφ

=

Equating

λ−= Separation constant

Linear homgn. PDE with

liniear homgn. BC)()0,( xfxu =

0)(),(

0)(),0(

2

1

====

tTtLu

tTtu

Page 8: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

Separation of VariablesSeparation of Variables

λφφ −=2

2

dx

d

λ−=

kGdt

dG λ−=

2

211

dx

d

dt

dG

kG

φφ

=

ktcetG λ−=)(

?

Eigenvalue Problem

Page 9: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

EigenvalueEigenvalue and and EigenfunctionEigenfunction

((λλ>0)>0)

λφφ −=2

2

dx

d 0)0( =φ0)( =Lφ

Boundary Value Problem

General Solution xcxc λλφ sincos 21 +=

Plug into BC

00sin0cos)0( 21 =+= λλφ cc

01 =c

0sincos)( 21 =+= LcLcL λλφ

0sin =Lλ2

=L

nπλ K,3,2,1=n

02 ≡c Trivial Solution

πλ nL =

Page 10: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

EigenvalueEigenvalue and and EigenfunctionEigenfunction

((λλ>0)>0)

2

=L

nπλ K,3,2,1=n

L

xncxcx

πλφ sinsin)( 22 ==

L

xnx

πφ sin)( =

Eigenvalue

Eigenfunction

12 =c

Page 11: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

EigenvalueEigenvalue and and EigenfunctionEigenfunction

((λλ=0)=0)

xcc 21 +=φ

λφφ −=2

2

dx

d

0)0( 1 == cφ 0)( 21 =+= LccLφ02 =c

Trivial Solution

Page 12: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

EigenvalueEigenvalue and and EigenfunctionEigenfunction

((λλ<0)<0)

λφφ −=2

2

dx

d 0)0( =φ0)( =Lφ

Boundary Value Problem

General Solution xcxc λλφ −+−= sinhcosh 43

Plug into BC

00sin0cosh)0( 23 =−+−= λλφ cc

03 =c

0sinh)( 4 =−= LcL λφ

04 ≡c Trivial Solution

Since sinh never zero for a positive

argumen

Page 13: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

SummarySummary

λφφ −=2

2

dx

d0)0( =φ 0)( =Lφ

2

=L

nn

πλ K,3,2,1=nL

xnxn

πφ sin)( =

Page 14: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

Product SolutionProduct Solution

ktcetG λ−=)(

)()(),( tGxtxu φ=

2

=L

nn

πλL

xnxn

πφ sin)( =

tLnkeL

xnBtxu

2)/(sin),( ππ −=

2ccB =

K,3,2,1=n

Initial Condition

Page 15: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

Initial Value Problems (example)Initial Value Problems (example)

2

2

x

uk

t

u

∂∂=

∂∂

tLkeL

xtxu

2)/3(3sin4),( ππ −=

0)(),(

0)(),0(

2

1

====

tTtLu

tTtu

L

xxu

π3sin4)0,( =

Page 16: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

Principle of SuperpositionPrinciple of Superposition

tLnkM

nn e

L

xnBtxu

2)/(

1

sin),( ππ −

=∑=

Muuuu ,,,, 321 K

∑=

=++++M

nnnMM ucucucucuc

1332211 K

IfAre solutions of a linear

homogeneous problem

then

is also a solution

Page 17: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

Initial Condition and Fourier SeriesInitial Condition and Fourier Series

L

xnBxfxu

M

nn

πsin)()0,(

1∑

=

==

FOURIER SERIES

• Any function f(x) (with certain very reasonable restriction, to be discussed

later) can be approximated by a finite linier combination of sin(nπx/L)

• The approximation may not be very good for small M, but gets to be a better

and better approximation as M is increased

• If we consider the limit as M tend to infinity, then not only is *) the best

approximation to f(x) using combination of eigenfunctions, but the resulting

infinite series will converge to f(x)

Page 18: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

Fourier SeriesFourier Series

L

xnBxf

nn

πsin)(

1∑

=

=

Any initial condition f(x) can be written as an infinite linear

combination of sin(nπx/L), known as a type of

Fourier Series

tLnk

nn e

L

xnBtxu

2)/(

1

sin),( ππ −∝

=∑=

What is more important is that we also claim that the

corresponding infinite series is the solutions of our heat

conduction problem

Page 19: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

OrthogonalityOrthogonality of of SinesSines

=≠

=L

nmL

nmdx

L

xm

L

xn

0 2/

0sinsin

ππ

L

xm

L

xnB

L

xmxf

nn

πππsinsinsin)(

1∑

=

=

dxL

xm

L

xnBdx

L

xmxf

L

nn

L

∫∑∫∝

=

=010

sinsinsin)(πππ

dxL

xmBdx

L

xmxf

L

m

L

∫∫ =0

2

0

sinsin)(ππ

Page 20: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

OrthogonalityOrthogonality of of SinesSines

dxL

xmxf

Ldx

L

xm

dxL

xmxf

BL

L

L

m ∫∫

∫==

0

0

2

0 sin)(2

sin

sin)(π

π

π

Page 21: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

Final SummaryFinal Summary

Heat Equation with Zero Temperatures Heat Equation with Zero Temperatures

at Finite Endsat Finite Ends

tLnk

nn e

L

xnBtxu

2)/(

1

sin),( ππ −∝

=∑=

2

2

x

uk

t

u

∂∂=

∂∂ )()0,( xfxu =

0)(),(

0)(),0(

2

1

====

tTtLu

tTtu

IC

BC

PDE

The solution is

dxL

xnxf

LB

L

n ∫=0

sin)(2 π

Page 22: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

Heat Equation in Rod with Insulated EndsHeat Equation in Rod with Insulated Ends

�� Consider the linear Consider the linear homegeneoushomegeneous Heat Equation in a oneHeat Equation in a one--

dimensional rod ( ) with constant thermdimensional rod ( ) with constant thermal al

coeficientscoeficients and no sources of thermal energyand no sources of thermal energy

Lx ≤≤0

2

2

x

uk

t

u

∂∂=

∂∂

)()0,( xfxu =

0),(

0),0(

=∂∂

=∂∂

tLx

u

tx

uIC

BC

PDE

Page 23: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

Separation of VariablesSeparation of Variables

�� By using the same method as previous, we will have By using the same method as previous, we will have

the following resultsthe following results

∑∝

=

−=0

)/( 2

cos),(n

ktLnn e

L

xnAtxu ππ

∫=L

dxxfL

A0

0 )(1

∫=L

m dxL

xmxf

LA

0

cos)(2 π

Page 24: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

EigenvalueEigenvalue and and EigenfunctionEigenfunction

((λλ>0)>0)

λφφ −=2

2

dx

d 0)0( =φ0)( =Lφ

Boundary Value Problem

General Solution xcxc λλφ sincos 21 +=

Plug into BC

00sin0cos)0( 21 =+= λλφ cc

01 =c

0sincos)( 21 =+= LcLcL λλφ

0sin =Lλ2

=L

nπλ K,3,2,1=n

02 ≡c Trivial Solution

πλ nL =

Page 25: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

Heat Equation in a Thin Circular RingHeat Equation in a Thin Circular Ring

Lx

Lx

−==

Lx 20 ≤≤

2

2

x

uk

t

u

∂∂=

∂∂

)()0,( xfxu =

),(),(

),(),(

tLutLu

tLx

utL

x

u

=−∂∂=−

∂∂

IC

BC

PDE

0=x

Page 26: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

EigenvalueEigenvalue and and EigenfunctionEigenfunction ((λλ>0)>0)

λφφ −=2

2

dx

d

)()( LL −= φφ

xcxc λλφ sincos 21 +=Boundary Value Problem

Plug into BC

LcLcLcLc λλλλ sincos)(sin)(cos 2121 +=−+−

0sin2 =Lc λLL λλ sin)(sin −=− LL λλ cos)(cos =−

We obtain

Plug into BC )()( Ldx

dL

dx

d φφ =−

)cossin( 21 xcxcdx

d λλλφ +−=

0sin1 =Lc λλWe obtain

*

**

Page 27: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

EigenvalueEigenvalue and and EigenfunctionEigenfunction ((λλ>0)>0)

0sin2 =Lc λ

0sin1 =Lc λλ0sin =Lλ

2

=L

nπλ

Since there are no additional constraints that c1 and c2 must

satisfy. We say that both sin and cos are eigenfunctions

corresponding to the eigenvalue

,...3,2,1,sin,cos)( == nL

xn

L

xnx

ππφ

Page 28: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

General solution General solution

ktLxneL

xntxu

2)/(cos),( ππ −=

ktLxneL

xntxu

2)/(sin),( ππ −=

∑∑∝

=

−−∝

=

++=1

)/()/(

10

22

sincos),(n

ktLxnn

ktLxn

nn e

L

xnbe

L

xnaatxu ππ ππ

In fact any linear combination of cos nπx/L and sin nπx/L is

an eigenfunctions. There are thus two infinite families of

product solutions of the PDE, n=1,2,3 …

∑∑∝

=

=

++=11

0 sincos)(n

nn

n L

xnb

L

xnaaxf

ππ

Page 29: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

General SolutionGeneral Solution

∫−

=L

L

dxxfL

a )(2

10

∫−

=L

L

m dxL

xnxf

La

πcos)(

1

∫=L

n dxL

xnxf

Lb

0

sin)(1 π

Page 30: Fisika Matematika III · Linearity Concept A linier operator by definition satisfies Example L(c1u1 +c2u2 ) =c1L(u1)+c2L(u2) L t u c t u c u c u c t ∂ ∂ + ∂ ∂ + = ∂ ∂

�� http://phys.unpad.ac.id/staff/irwanhttp://phys.unpad.ac.id/staff/irwan