Scribd
Upload
56 views5 pages

Practice Midterm Solutions

This document contains the solution to a practice midterm exam for a math course. It includes: 1) The solution to a problem solving for characteristic lines of a PDE and expressing the solution in terms of the parameters. 2) A proof showing the equality of two expressions involving distributions and derivatives. 3) A proof that a distribution supported at 0 can be written as a delta function multiplied by a scalar. 4) Solutions to problems involving defining weak solutions to conservation laws and determining the speed of a shock wave. 5) Proofs regarding the time derivative of energy for solutions to a wave equation.

Uploaded by

3dd2ef652ed6f7
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
56 views5 pages

Practice Midterm Solutions

This document contains the solution to a practice midterm exam for a math course. It includes: 1) The solution to a problem solving for characteristic lines of a PDE and expressing the solution in terms of the parameters. 2) A proof showing the equality of two expressions involving distributions and derivatives. 3) A proof that a distribution supported at 0 can be written as a delta function multiplied by a scalar. 4) Solutions to problems involving defining weak solutions to conservation laws and determining the speed of a shock wave. 5) Proofs regarding the time derivative of energy for solutions to a wave equation.

Uploaded by

3dd2ef652ed6f7
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

Solution to Practice Midterm of

Math220/CME303
Qiuye Jia
October 23, 2019

1
The tangent vector field of characteristic lines is: V (x, y) = (1, y).Consider
the characteristic line starting at (0, r), and we use s as the parameter along
the curve. We know:
ẏ = y, ẋ = 1.
Hence we know:
x = s, y = res .
And for v(r, s) = u(x, y), we have:
vs = y 2 = r2 e2s , v(r, 0) = sin r.
which gives:
r2 2s
v(r, s) = sin r + (e − 1).
2
Then use:
s = x, r = ye−s = ye−x ,
we know:
−x y 2 e−2x 2x
u(x, y) = sin(ye ) + (e − 1)
2
y 2 y 2 e−2x
= sin(ye−x ) + − .
2 2
1
2
For the left hand side, we first define the (weak) derivative of ιf , which is a
distribution, to be:

∂xi ιf (ϕ) = −ιf (∂xi ϕ),

for ϕ ∈ Cc∞ (Rn ).


Now we prove that, when we apply both sides to any test function, we get
the same results. For the left hand side, we have:
∫
∂xi ιf (ϕ) = −ιf (∂xi ϕ) = − f ∂xi ϕdx.
Rn

On the other side, we have:


∫
ι∂xi f (ϕ) = ϕ∂xi f dx.
Rn

Now integrate by parts for the right hand side of one of above equations, and
notice that ϕ has compact support, which justfies integration by parts and
make the boundary term vanish, we have:
∫ ∫
ϕ∂xi f dx = − f ∂xi ϕdx,
Rn Rn

which tells us:

ι∂xi f = ∂xi ιf .

3
a
To define a distribution, we only need to define its action on any smooth
compactly supported function. Notice that, for ϕ ∈ Cc∞ (Rn ), gϕ will still be
a member of Cc∞ (Rn ), we define:

gu(ϕ) := u(gϕ).

2
b
We first prove a Lemma: for a smooth function f (0) = 0, then we can find
g ∈ C ∞ (R) such that:

f (x) = xg(x).

The proof relies on the Taylor’s theorem with an integral remainder (see
Vasy’s book, problem 1.2):
∫ 1
f (x) = f (0) + x f ′ (sx)ds.
0
∫1
And now set g(x) = 0
f ′ (sx)ds.

Now pick any fixed ϕ0 ∈ Cc∞ (R), with ϕ0 (0) = 1. Now for any ϕ ∈ Cc∞ (R),
we know that ϕ(x) − ϕ(0)ϕ0 (x) satisfy the condition of our lemma, hence
we can find smooth g(x) to make ϕ(x) − ϕ(0)ϕ0 (x) = xg(x). Since the left
hand side has compact support, so does the right hand side, consequently
g ∈ Cc∞ (R). Now we know:

u(ϕ(·) − ϕ(0)ϕ0 (·)) = u(xg) = xu(g) = 0.

This means:

u(ϕ) = u(ϕ0 )ϕ(0),

or equivalently:

u = u(ϕ0 )δ0 ,

which is the desired result.

4
a
First we rewrite the equation as:
1
ut + ( u2 )x = 0.
2

3
Motivated by integration by parts (i.e. multiply both sides of the PDE by a
test function and then integrate by parts) in the case where u is a smooth
funtion, we define it to be a weak solution if and only if:
∫ ∫
1 2
(uψt + u ψx )dxdt + ϕ(x)ψ(0, x)dx = 0,
Rx ×[0,∞)t 2 Rx

for any ψ ∈ Cc∞ (Rx × [0, ∞)t ).

b
Use the result or the proof (use the divergence theorem) of Problem 3 of
homework 3 (Problem 5.7 in Vasy’s text book), we know the waveshock will
satisfy the Rankine-Hugoniot jump condition, which gives us:
u2− u2+
− = ξ ′ (t)(u− − u+ ),
2 2
where (t, ξ(t)) is the waveshock. This gives us ξ ′ (t) = 12 . Hence the line sep-
arating two regions in which u is a classical solution is x = 12 t. Thus a = 12 .

5
a
Take derivative with respect to t, we have:
∫
′
E (t) = ut utt + ux uxt + uy uyt dxdy
Ω

Since u|∂Ω = 0, take derivative with respect to t, we get:

ut |∂Ω = 0.

Then apply divergence theorem to ut ∇u, which vanishes on ∂Ω, we get:


∫ ∫
∇u · ∇ut dxdy = − ut ∆udxdy,
Ω Ω

4
where all gradients are taken only with respect to (x, y), and ∆ = ∂x2 + ∂y2 .
Substitute back we have:
∫
′
E (t) = ut (utt − ∆u)dxdy = 0,
Ω

where in the last equality we used the given PDE.

b
Consider u0 = u(1) − u(2) . Since u(i) satisfy the same boundary and initial
condition, u0 will satisfy vanishing boundary and initial condition. So, the
energy associated to u0 will be zero when t = 0. According to part (a), and
u0 satisfy the same PDE, we know E(t) associated to u0 will be 0 for all
t ≥ 0. This tells us all derivatives of u0 are zero, hence it is a constant. And
it vanishes when t = 0, so u0 ≡ 0, which means u(1) ≡ u(2) .

You might also like