Partial Differential Equations III/IV
[6pt] Exercise Sheet 2: Solutions
-
1.
Burger’s equation as a model of traffic flow. We will rewrite the following PDE in the form of Burger’s equation:
By expanding the second term on the left-hand side we obtain
This suggest that we should define . Then and
Therefore satisfies Burger’s equation, as required.
-
2.
Revision of ODEs.
-
(i)
Linear first-order ODEs. Multiplying the ODE by the integrating factor gives
as required.
-
(ii)
Separable ODEs. Since , we can divide the ODE by to obtain
Observe that by assumption. Therefore is invertible in a neighbourhood of , and so
as required.
-
(i)
-
3.
The method of characteristics for a linear PDE. Solve the linear first-order PDE
The PDE has the form
with
The value of is prescribed on the line , which we can parametrise by , . Define
Step 1: We need to solve
(1) (2) (3) subject to the initial conditions
(4) (5) (6) Multiply equation (3) by the integrating factor
to obtain
Integrating from to gives
Step 2: We need to invert the map . Setting and solving for and in terms of and gives , . Therefore
Step 3: Finally,
It is easy to check that satisfies the Cauchy problem.
Plotting the characteristics: The characteristics are the curves
which are in fact straight lines. We can write these lines in nonparametric form as . Some representative characteristics are plotted below.
-
4.
The method of characteristics for a linear PDE. We solve the linear first-order PDE
where is a constant.
The PDE has the form
with
The value of is prescribed on the line , which we can parametrise by , . Define
Step 1: We need to solve
(7) (8) (9) subject to the initial conditions
(10) (11) (12) The system of ODEs for and decouples from the ODE for . Let . Multiply equation (7) by , equation (8) by , and add the resulting equations together to get
Therefore is independent of and . Since is constant in , we seek a solution of the form
where is to be determined. Substituting these expressions into equations (7) and (8) gives
It follows that and . The initial conditions , imply that . Therefore
Remark: An alternative way of finding and is to differentiate equation (7) with respect to to get . Now solve this second-order ODE for .
Equations (9) and (12) imply that
Step 2: If , then . It follows that
Since is independent of , we do not need to know in order to find .
Step 3: Define
It is easy to check that satisfies the initial value problem.
-
5.
The method of characteristics for a semilinear PDE. We solve the semilinear first-order PDE
subject to the condition for all . The PDE has the form
with
%
The value of is prescribed on the line , which we can parametrise by , . Define
Now we use the Method of Characteristics:
Step 1: We need to solve
(13) (14) (15) subject to the initial conditions
(16) (17) (18) Equations (13), (16) imply that
Substituting this into equation (14) gives
(19) Equations (19), (17) imply that
Substituting this into equation (18) gives
We solve this using the method of separation:
since .
Let us underline that the definition of is meaningful, one for such values of for which
Let us discuss for which values of can we have
Let us set , then the previous equality reduces to
which clearly has infinitely many solutions, since as . So for these values of for instance, is not well-defined.
Step 2: We need to invert the map . Setting and solving for and in terms of and gives , . Therefore
Step 3: Finally,
By the same arguments as at the end of Step 1, this definition of is meaningful only for points, where the denominator is nonzero. As we have seen there, there are infinitely many points, where this is not the case.
-
6.
The method of characteristics for a quasilinear PDE. Solve the quasilinear first-order PDE
The PDE has the form
with
The value of is prescribed on the line , which we can parametrise by , . Define
Now we use the Method of Characteristics:
Step 1: We need to solve
(20) (21) (22) subject to the initial conditions
(23) (24) (25) Equations (21) and (24) imply that
Differentiating equation (20) with respect to gives
by equations (21) and (22). The linear second-order ODE has solution
where and are constants determined by the initial conditions and
The initial conditions give
Therefore and and
Substituting this into equation (22) yields
By integrating we find that
Step 2: We need to invert the map . Setting and solving for and in terms of and gives
Therefore
Step 3: We have discovered that
Let’s verify that satisfies the Cauchy problem:
-
7.
The method of characteristics for a quasilinear PDE. Find satisfying the quasilinear PDE
and the Cauchy condition . The PDE has the form
with
The value of is prescribed on the line , which we can parametrise by , . Define
Now we use the Method of Characteristics:
Step 1: We need to solve
(26) (27) (28) subject to the initial conditions
(29) (30) (31) Equations (27), (30) imply that
and equations (28), (31) imply that
Substituting this into equation (26) yields
Integrating gives
Step 2: We need to invert the map . Let us notice that this is not possible in neighborhoods of points for which .
We set and try to solve for and in terms of and . We get and
Therefore
and
We notice that these terms are not meaningful for . This is corresponding exactly to the case where (since ).
Step 3: Finally
This expression is meaningful only when . If , then the value of the solution is given by the value of the Cauchy data, for any other point the solution is not defined.
-
8.
The method of characteristics is independent of the parametrisation of . Consider the transport equation
This equation has the form with
The value of is prescribed on the line . We can parametrise by , , . Define
Now we use the Method of Characteristics:
Step 1: We need to solve
(32) (33) (34) subject to the initial conditions
(35) (36) (37) Equation (32) implies that
where is to be determined. Equation (35) implies that . Therefore
Similarly, equations (33) and (36) imply that
and equations (34) and (37) imply that
Step 2: Now we need to invert the function . Setting and solving for and in terms of and gives and . Therefore
Step 3: Finally,
as required.
-
9.
Nonexistence due to a characteristic point.
-
(i)
We can parametrise by , . The point corresponds to parameter since . The vector is the unit outward-pointing normal to at point . We have
In particular, taking gives
Therefore the vector field is tangent to at and hence is a characteristic point. All the other points of are noncharacteristic except for the point .
-
(ii)
Suppose for contradiction that there exists a function satisfying
(38) (39) where is an open set containing . Chose sufficiently small so that the line joining to lies in . This line is tangent to the constant vector field since
Since , then is constant in direction , and hence . In case this doesn’t convince you, a more algebraic argument is as follows:
(Fundamental Theorem of Calculus) (Chain Rule) (equation (38)) which again shows that . This contradicts the Cauchy condition (39) since
(Sketch the graph of the cosine function to convince yourself of this, and recall that .)
Finally, the Cauchy problem has a global solution for any continuously differentiable function satisfying
For example, we could take . Another possibility is to simply take , in which case the solution will be .
-
(i)
-
10.
Nonuniqueness or nonexistence due to a characteristic Cauchy curve. Consider the PDE
(40) -
(i)
The PDE has the form
with
Let . The value of is prescribed on the Cauchy curve , which can be parametrised by , . We can write the Cauchy condition as with . The vector is tangent to at the point . Therefore is normal to at each point and
for all . Therefore no point on the Cauchy curve is noncharacteristic.
Alternative proof: We have
for all , and again we see that no point on the Cauchy curve is noncharacteristic.
The vector field is constant and is tangent to the Cauchy curve at every point. Therefore the flow of the vector field through lies along and the characteristics coincide with , as shown in the figure below. The vector field is also plotted.
-
(ii)
We use the method of characteristics to solve the PDE
where is an arbitrary differentiable function satisfying . As above, the PDE has the form
with
The Cauchy curve is , which we can parametrise by , . Let .
Step 1: We need to solve
(41) (42) (43) subject to the initial conditions
(44) (45) (46) Equations (41), (44) imply that
Equations (42), (45) imply that
Equations (43), (46) imply that
Step 2: We need to invert the map . Setting and solving for and in terms of and gives , . Therefore
- (iii)
-
(i)
-
11.
Obstacles to global existence.
-
(i)
Let satisfy
Here we derive the solution without using the method of characteristics. Since , then is constant in the radial direction. If this isn’t obvious to you, then argue as follows: For all , ,
(Chain Rule) which means that is constant in the radial direction. Therefore
as required.
Let be a non-constant function. Assume for contradiction that is a global solution. By the calculation above for all . Therefore
Therefore for all , which is a contradiction since is non-constant.
If is a constant function, then the global solution is clearly .
-
(ii)
Let and let satisfy
Since , the PDE has the form
with
We can parametrise by , . Define
Step 1: We need to solve
(49) (50) (51) subject to the initial conditions
(52) (53) (54) Equations (49), (52) imply that
Equations (50), (53) imply that
Equations (51), (54) imply that
Step 2: We need to invert the map . Clearly the inverse map is given by
Step 3: Finally,
as required. (It’s actually easier to simply integrate the PDE with respect to , as below, instead of using the method of characteristics.)
Now consider the case . A quick sketch of the vector field shows that each characteristic intersects twice, at and for some . Let be a local solution, which satisfies
Let . Then
as required. Similarly, for ,
as required.
Assume for contraction that is a global solution. Then
Contradiction! From this calculation we see that a necessary condition for a global solution to exist is . In this case the global solution is .
-
(i)
-
12.
Geometric interpretation of the method of characteristics. Let . Then is normal to the solution surface and
Therefore is orthogonal to at every point of the solution surface. Hence is tangent to the solution surface, as required.
-
13.
The method of characteristics for PDEs in independent variables. The transport equation in spatial variables is
where is a constant vector and . The PDE has the form
where is the gradient with respect to , i.e., , and
The value of is prescribed on the –dimensional hypersurface , which we can parametrise by , . Define
Now we apply the Method of Characteristics:
Step 1. We need to solve
subject to the initial conditions
The solution is
Step 2. We need to invert the map . Setting and solving for and in terms of and gives , . Therefore
Step 3. Finally,
as required.
-
14.
Uniqueness theorem for ODEs.
-
(i)
Let be solutions: , , . Then
Taking the dot product of this equation with gives
Therefore by the Cauchy-Schwarz inequality
since is Lipschitz continuous. Define . We have shown that
Rearranging and multiplying by the integrating factor gives
Integrating over the interval yields
But . Therefore and hence for all . Therefore for all , as required.
-
(ii)
Let . By the Mean Value Theorem, there exists such that
Therefore, for all ,
with .
-
(i)
-
15.
Nonuniqueness for ODEs. Consider the separable ODE
Exactly the same method we used in Example 2.3 leads to the solution ,
At , , where is not continuously differentiable. Let . For any , is a solution, where
-
16.
The Eikonal equation.
-
(i)
If , then and hence for . Recall that for . Therefore
for , as required.
-
(ii)
Clearly when , . Therefore if . We compute
Therefore in , except along the lines , as required.
-
(iii)
Define to be the distance from to the boundary of , i.e.,
Clearly when . Assume that is continuously differentiable in a neighbourhood of . Then for all
Let satisfy dist. Then
(55) Observe that
Therefore taking the limit in equation (55) gives
(56) There was nothing special about our choice of , and so this expression also holds with replaced by :
(57) Combining (56) and (57) yields
But this holds for all . Therefore
and , as required.
-
(i)