Documente Academic
Documente Profesional
Documente Cultură
In this lecture we abstract the eigenvalue problems that we have found so useful thus far for solving the PDEs to a
general class of boundary value problems that share a common set of properties. The so-called Sturm-Liouville Problems
define a class of eigenvalue problems, which include many of the previous problems as special cases. The S L Problem
helps to identify those assumptions that are needed to define an eigenvalue problems with the properties that we require.
Key Concepts: Eigenvalue Problems, Sturm-Liouville Boundary Value Problems; Robin Boundary conditions.
Reference Section: Boyce and Di Prima Section 11.1 and 11.2
X 00 + 2 X = 0
n = n
L , n = 1, 2, .. .
=
X(0) = 0 = X(L)
Xn (x) = sin nx
L
II: The Neumann Problem:
X 00 + 2 X = 0
n = n
L , n = 0,
1, 2,. . .
=
0
0
X (0) = 0 = X (L)
Xn (x) = cos nx
L
III: The Periodic Boundary Value Problem:
X 00 + 2 X = 0
n = n
L , n = 0,1, 2, .. .
=
X(L) = 0 = X(L)
X
(x)
1,
cos nx
, sin nx
n
0
0
L
L
X (L) = 0 = X (L)
IV: Mixed Boundary Value Problem A:
(
k =
(2k+1)
,
2L
k =
(2k+1)
,
2L
k = 0, 1, 2,. . .
x
Xn (x) = sin (2k+1)
2L
V: Mixed Boundary Value Problem B:
X 00 + 2 X = 0
X(0) = 0 = X 0 (L)
X 00 + 2 X = 0
X 0 (0) = 0 = X(L)
k = 0, 1, 2,. . .
Xn (x) = cos (2k+1)
x
2L
2
30.2 The regular Sturm-Liouville problem:
Consider the the following two-point boundary value problem
0
p(x)y 0 q(x)y + r(x)y = 0 0 < x < `
1 y(0) + 2 y 0 (0) = 0 1 y(`) + 2 y 0 (`) = 0
(30.1)
0
Ly = ry where Ly = (py 0 ) + qy
SL
(30.2)
Remark 1 Note:
(1) If p = 1, q = 0, r = 1, 1 = 1, 2 = 0, 1 = 1, 2 = 0 we obtain Problem (I) above whereas if p = 1,
q = 0, r = 1, 1 = 0, 2 = 1, 1 = 0, 2 = 1, we obtain Problem (II) above. Notice that the boundary
conditions for these two problems are specified at separate points and are called separated BC. The periodic
BC X(0) = X(2) are not separated so that Problem (III) is not technically a SL Problem.
(2) If p > 0 and r > 0 and ` < then the SL Problem is said to be regular. If p(x) or r(x) is zero for some x or
the domain is [0, ) then the problem is singular.
(3) There is no loss of generality in the form of Ly = (py 0 ) + qy since it is possible to convert a general 2nd order
eigenvalue problem
P (x)y 00 Q(x)y 0 + R(x)y = y
(30.3)
(30.4)
(30.5)
Thus comparing (30.5) and (30.4) we can make the following identifications: p = ZP and p0 = Q p0 =
Q
P0
dx)
0 P + P 0 = Q which is a linear 1st order ODE for with integrating factor exp(
P
P
Q
P0
P
P
h
= 0 Pe
Q
P
dx
i0
=0
Q
P
dx
(30.6)
(30.7)
(0) = 0 =
(30.8)
p(x) = ex
/2
r(x) = ex
Q(x) = x,
R(x) = 1.
(30.9)
/2
Q = x4 ,
Therefore ex
= e
x4 dx
= ex
/5 00
/5
5
y + ex /5 x4 y 0 = ex /5
0
5
5
ex /5 y 0 = ex /5 y.
(30.10)
(30.11)
(30.12)
(b) The eigenfunctions corresponding to different eigenvalues are orthogonal with respect to the weight function r(x):
Z`
r(x)j (x)K (x) dx = 0
j 6= k.
(30.13)
where
f (x)
cn n (x)
n=1
R`
r(x)f (x)n (x) dx
cn
R`
0
r(x)2n (x) dx
(30.14)
4
Example 30.3 Robin Boundary Conditions:
X 00 + X = 0,
= 2
0
X (0) = h1 X(0), X 0 (`) = h2 X(`)
(30.15)
where h1 0 and h2 0.
(30.16)
(30.17)
(30.18)
2
h2
+ h2 sin ` + + cos ` = 0.
h1
h1
(30.19)
(h1 + h2 )
tan(`) =
.
2 h1 h2
(30.20)
Case I: h1 and h2 6= 0
Xn =
n
h1
5
0
n
Xn =
cos n x + sin n x
h1
cos n (` x)
=
sin n `
(30.21)
(30.22)
10
10
Case III: h1
Xn = sin(n x)
2n + 1
n
2
`
(30.23)
n = 0, 1, 2, . . . as n (30.24)
5
0
10
Z` n
o
0
v (pu0 ) + qu dx
(30.25)
Z`
Z`
=
`
vpu0 |0
Z`
=
(30.26)
0
`
vpu0 |0
uqv dx
u pv dx +
`
upv 0 |0
n
o
0
u (pv 0 ) + qv dx
(30.27)
uLv dx.
(30.28)
Z`
Z`
`
Therefore
0
Now suppose that u and v both satisfy the SL boundary conditions. I.E.
1 u(0) + 2 u0 (0) = 0
1 v(0) + 2 v 0 (0) = 0
1 u(`) + 2 u0 (`) = 0
1 v(`) + 2 v 0 (`) = 0
(30.29)
then
Z`
Z`
uLv dx = p(`)u0 (`)v(`) + p(`)u(`)v 0 (`)
vLu dx
0
1
1
= p(`) + u(`)v(`) + u(`) v(`)
2
2
1
1
+p(0) u(0)v(0) u(0) v(0)
2
2
= 0.
Z`
Thus
Z`
vLu dx =
(30.30)
(30.31)
(30.32)
(30.33)
(30.34)
6
Observations:
Z`
Z`
If L and BC are such that
vLu dx =
0
Z`
(30.35)
y)
= (
y , ry) (y, r
Z`
Z`
y (x) dx
= y(x)ry(x) dx y(x)r(x)
0
(30.36)
(30.37)
Z`
= ( )
2
r(x)y(x) dx
(30.38)
is real.
Since r(x)|y(x)| 0 it follows that =
2
(1c) j > 0 provided 1 /2 < 0 1 /2 > 0 and q(x) > 0. Consider Ly = (py 0 ) + qy = ry (SL) and multiply
(SL) by y and integrate from 0 to `:
Z`
Z`
0 0
(y, Ly) =
(py ) y + qy dx =
0
R`
Therefore =
2
r(x) y(x) dx
(30.39)
0
0
(py 0 ) y + qy 2 dx
R`
dx
[py 0 y]`0 +
R`
0
R`
p(y 0 )2 + qy 2 dx
ry 2
(30.40)
dx
2 R`
2
1
+p(`) 12 y(`) p(0)
y(0) + p(y 0 ) + qy 2 dx
2
=
R`
(30.41)
ry 2 dx
(2b) Eigenfunctions corresponding to different eigenvalues are orthogonal. Consider two distinct eigenvalues j 6= k j : Lj = rj j and k : Lk = rk k . Then
0 = (k , Lj ) (j , Lk ) by Lagranges Identity
(30.42)
= (k , rj j ) (j , rk k )
Z`
= (j k ) r(x)k (x)j (x) dx
(30.43)
(30.44)
(30.45)
(3) The eigenfunctions form a complete set: It is difficult to prove the convergence of the eigenfunction series
expansion for f (x) that is piecewise smooth. However, if we assume the expansion converges then it is a simple
X
matter to use orthogonality to determine the coefficients in the expansion: Let f (x) =
cn n (x).
n=1
Z`
f (x)m (x)r(x) dx =
X
n=1
Z`
cn
(30.46)
orthogonality implies
R`
cm =
R`
0
2
r(x) m (x) dx
(30.47)