Partial Differential Equations III/IV
[6pt] Exercise Sheet 5: Solutions
-
1.
Mean-value formula harmonic. Fix . For all
(1) We can parametrise using polar coordinates by
Using this parametrisation we compute
(2) By equation (2) and the Chain Rule
(3) The unit outward-pointing normal to at point is
Taking gives
Using this, we can write equation (3) as
(Divergence Theorem) Differentiating equation (1) with respect to gives . Therefore
for all . There are two ways to reach the punchline from here: Either observe that since
(4) then we must have . (Otherwise, by continuity of , is either strictly positive or strictly negative in for sufficiently small, which contradicts (4).) Alternatively, multiply equation (4) by and take the limit :
since the average of a continuous function over a ball of radius tends to the value of the function at the centre of the ball as .
-
2.
Subharmonic functions.
-
(i)
First we prove the mean-value formula
(5) We can parametrise using polar coordinates by
Using this parametrisation we compute
(6) By equation (6) and the Chain Rule
(7) The unit outward-pointing normal to at point is
Taking gives
Using this, we can write equation (7) as
(Divergence Theorem) since is subharmonic. Therefore and hence if . The mean-value formula (5) follows almost immediately from this:
since the average of a continuous function over a sphere of radius tends to the value of the function at the centre of the sphere as .
Now we prove the second mean-value formula
(8) Using polar coordinates we can write
(9) Observe that is parametrised by , . This parametrisation satisfies . Therefore we can write equation (9) as
as required.
-
(ii)
We prove the strong maximum principle. Let and
Define to be the set of points in where attains its maximum:
Note that is nonempty since .
Let and , i.e, . By part (i)
(10) Therefore the inequality in (10) is an equality,
which means that for all . Hence and so is an open subset of .
The set is the preimage of the closed set under the continuous map and so is closed. Therefore is a closed subset of .
We have shown that is a nonempty open and closed subset of the connected set . Therefore , which implies that constant in , as required. The weak maximum principle follows easily from this (see Q3).
-
(iii)
Subharmonic functions do not satisfy the minimum principle
For example, take , , . Then and so is subharmonic. But the minimum value of is , which is attained at , not on the boundary of .
-
(i)
-
3.
Strong maximum principle weak maximum principle. By the strong maximum principle:
Either: is constant, in which case it is trivial thatOr: is not constant, in which case the strong maximum principle implies that, for all ,
i.e., the maximum of over is not attained in . Since , it follows that
as required.
-
4.
The strong maximum principle is false if is not connected. Simply take , , , and define by
Clearly, and , while is clearly disconnected. Finally, , which is an interior point. Yet, the function is not constant.
-
5.
Minimum principles and an application: Positivity of solutions.
-
(i)
First we state the minimum principles: Let be open, bounded and connected. Let , be harmonic in .
-
(a)
Weak minimum principle: attains its minimum on the boundary of , i.e.,
-
(b)
Strong minimum principle: If attains its minimum in the interior of , then is constant, i.e., if there exists such that
then is constant in .
These can be proved as follows:
-
(a)
Weak minimum principle: Let . Then is harmonic since is harmonic. Therefore by the weak maximum principle
as required.
-
(b)
Strong minimum principle: If attains its minimum at , then the harmonic function attains its maximum at . By the strong maximum principle, is constant. Therefore is constant.
-
(a)
-
(ii)
Since is harmonic it satisfies the strong minimum principle. Therefore:
Either: is constant, in which case for allas the function is continuous up to the boundary.
Or: is not constant, in which case the strong minimum principle implies that, for all ,
In either case for all , as required.
-
(i)
-
6.
Another application of the maximum principle: Bounds on solutions.
-
(a)
Let . We have
We know that is not constant (since on ). Therefore the strong maximum principle for harmonic functions implies that
-
(b)
Let
Then
and therefore is harmonic.
Let . If , then
Letβs find the infimum and supremum of on . We have
Therefore the critical points of are
The maximum and minimum points of on are attained at the critical points of or at the end points . We have
Therefore
as required.
Let . By symmetry of in and ,
Let . Then is harmonic and on . We can decompose as
For ,
Clearly is not constant. Therefore by the strong maximum principle
Therefore
as desired. We have an estimate of correct to two significant figures!
-
(a)
-
7.
Application of the maximum principle for subharmonic functions: Comparison theorems. Let . Then satisfies
By assumption, and so is subharmonic. Therefore it satisfies the maximum principle
Therefore and , as required.
-
8.
Maximum principles for more general elliptic problems.
-
(i)
Consider the one-dimensional steady convection-diffusion equation
where and are constants, . Let . Then
Therefore
for some constant . Hence
for some constant . If , then is constant. Otherwise
and so is strictly increasing if and strictly decreasing if . Therefore attains its maximum and minimum on the boundary of .
-
(ii)
Let satisfy Poissonβs equation
where is a constant. Then is a quadratic polynomial. It is easy to see that if , then satisfies a weak maximum principle, and if , then satisfies a weak minimum principle. See Q2 for the two-dimensional case.
-
(iii)
Consider the equation
with , . The solution has the form
where and are constants. We can assume that and , otherwise the result is obvious. If , then is either increasing (if , ) or decreasing (if , ) and hence attains its maximum and minimum on the boundary of . It follows that . If , then has a unique critical point:
and the critical value of is
If , then is increasing or decreasing on and so as before. Assume that . We consider two case: and .
If , then for all and , which implies that is a local minimum point of . Therefore attains its maximum on the boundary of , as required.
If , then for all and , which implies that is a local maximum point of . Therefore
as required.
If , then the maximum principle does not hold since
for some constants and . For example, take , , , , . Then
-
(i)
-
9.
Maximum principles for 4th-order elliptic PDEs? In general, 4th-order elliptic PDEs do not satisfy a maximum principle. For example, if on , then is a cubic polynomial, which need not attain is maximum or minimum on the boundary of . If on , where is a constant, then is a quartic polynomial, which again need not attain its maximum on the boundary of .
-
10.
Regularity Theorem: Harmonic functions are .
-
(i)
Observe that outside the disc . Therefore supp and so supp. For the rest of the problem it is convenient to write where is defined by
Then . Observe that
by definition of . We compute
(polar coordiates) (back to Cartesian coordinates) -
(ii)
Take . Then
(mean-value formula)
-
(i)
-
11.
analytic. Consider the function defined by
Then is infinitely differentiable but it is not analytic since it does not have a convergent Taylor series expansion about the point :
but is nonzero in any neighbourhood of . In general, nonzero analytic functions cannot have compact support.
-
12.
Non-negative harmonic functions on are constant.
-
(i)
We have
(mean-value formula) (mean-value formula) as required.
-
(ii)
Let . Then
Therefore and hence . We have
-
(iii)
If , then by parts (i) and (ii),
Therefore
Interchanging the roles of and gives
Therefore for all and hence is a constant function.
-
(i)
-
13.
Proof of Liouvilleβs Theorem. Since is bounded, then there exists such that for all . Therefore the harmonic function on . But positive harmonic functions on are constant by Q12. Therefore , and hence , are constant.
-
14.
An application of Liouvilleβs Theorem: βUniquenessβ for Poissonβs equation in . Let where is the fundamental solution of Poissonβs equation in with :
(Recall that is the volume of the unit ball in and hence .) Let be any bounded solution of Poissonβs equation in . Then is a harmonic function since and in . We show that is bounded: Since has compact support, there exists such that supp. In particular, in . Therefore
We just need to show that
is uniformly bounded in . This is more fiddly than you would expect. We consider two cases: and .
If , then (draw a sketch to convince yourself of this) and so
(spherical polar coordinates) If , then for all
and so
Therefore, for all ,
Hence is bounded. Since and are bounded, then is a bounded harmonic function on . By Liouvilleβs Theorem constant. Therefore
as required.
This argument can be extended to for any . It does not work for since is not necessarily bounded in since blows up as . For the case , as , and it converges to sufficiently fast in order for to be bounded.
-
15.
An obstacle to uniqueness for Laplaceβs equation: Unbounded domains.
-
(i)
We can build a nontrivial solution using the fundamental solution of Poissonβs equation:
-
(ii)
Simply take .
-
(i)
-
16.
Eigenvalues of the negative Laplacian. By Exercise Sheet 4, Q11, the eigenvalues are positive, . Therefore we can write each eigenvalue as for some . Then
Recall from ODE theory (see page 21 of the lecture notes) that solutions of this ODE have the form
for some constants . The boundary condition implies that . The boundary condition gives
Since , then . Since , it follows that
Therefore and the eigenfunction-eigenvalue pairs are
In particular, there are countably many eigenvalues.
-
17.
Connection between holomorphic functions and harmonic functions.
Let be a holomorphic (complex analytic) function with real and imaginary parts and :
The Cauchy-Riemann equations are
Therefore
and
Completing the table gives
Harmonic Functions Holomorphic Functions Mean-Value Formula Cauchy Integral Formula Maximum Principle Maximum Modulus Principle Liouvilleβs Theorem Liouvilleβs Theorem See Remark 5.3 in the lecture notes for an explanation of why the Cauchy integral formula implies the mean-value formula.