Tell me more ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Let $\Omega \subseteq \mathbb{R}^n$, $n\geq 2$, be a bounded domain with boundary $\partial \Omega \subseteq C^2,v$ outer unit normal vector on $\partial \Omega$, $h \in L^2(\Omega)$. Let $u \in C\left( \left[ 0,\infty \right);{{L}^{2}}\left( \Omega \right) \right)\cap C^1\left( \left( 0,\infty \right);{{H}^{2}}\left( \Omega \right) \right)$ be a solution to the Dirichlet problem $$ \left\{ \begin{matrix} {{u}_{t}}\left( t,x \right)={{\Delta }_{x}}u\left( t,x \right),t\in \left( 0,\infty \right),\ x\in \Omega \\ u\left( t,x \right)=0,t\in \left( 0,\infty \right),\ x\in \partial \Omega \\ u\left( 0,x \right)=h\left( x \right),\ x\in \Omega \\ \end{matrix} \right.$$

Let $w \in C\left( \left[ 0,\infty \right);{{L}^{2}}\left( \Omega \right) \right)\cap C^1\left( \left( 0,\infty \right);{{H}^{2}}\left( \Omega \right) \right)$ be a solution to the Neumann problem $$\left\{ \begin{matrix} {{w}_{t}}\left( t,x \right)={{\Delta }_{x}}w\left( t,x \right),t\in \left( 0,\infty \right),\ x\in \Omega \\ \frac{\partial w}{\partial v}\left( t,x \right)=0,t\in \left( 0,\infty \right),\ x\in \partial \Omega \\ w\left( 0,x \right)=h\left( x \right),\ x\in \Omega \\ \end{matrix} \right.$$ Show that there are bounded linear operators $ {{E}_{D}}\left( t \right):{{L}^{2}}\left( \Omega \right)\to {{L}^{2}}\left( \Omega \right)$ and $ {{E}_{N}}\left( t \right):{{L}^{2}}\left( \Omega \right)\to {{L}^{2}}\left( \Omega \right)$ such that $u\left( t,\centerdot \right)={{E}_{D}}\left( t \right)h$ and $w\left( t,\centerdot \right)={{E}_{N}}\left( t \right)h,t\in \left[ 0,\infty \right)$ and find their norms. Furthermore show that there are $0\ne {{h}_{D}}\in {{L}^{2}}\left( \Omega \right)$ y $0\ne {{h}_{N}}\in {{L}^{2}}\left( \Omega \right)$ such that ${{E}_{D}}\left( t \right)h_{D}=\left\| {{E}_{D}}\left( t \right) \right\|{{h}_{D}}$ y ${{E}_{N}}\left( t \right)h_{N}=\left\| {{E}_{N}}\left( t \right) \right\|{{h}_{N}}, \forall t\in \left[ 0,\infty \right)$. Describe those elements ${{h}_{D}}$ and ${{h}_{N}}$

Attempt: Let ${{\left\{ {{\lambda }_{n}^D} \right\}}_{n\in \mathbb{N}}}$ be eigenvalues with their respectives eigenfunctions ${{\left\{ {{\phi }_{n}^D} \right\}}_{n\in \mathbb{N}}}$ orthonormal basis of ${{L}^{2}}\left( \Omega \right)$, of the Dirichlet Laplacian. Therefore the solution is $u\left( t,x \right)=\sum\limits_{n\in \mathbb{N}}{{{\left\langle \phi _{n}^{D},h \right\rangle }_{{{L}^{2}}\left( \Omega \right)}}{{e}^{-\lambda _{n}^{D}t}}\phi _{n}^{D}\left( x \right)}$, so that ${{E}_{D}}\left( t \right)\left( f \right)=\sum\limits_{n\in \mathbb{N}}{{{\left\langle \phi _{n}^{D},f \right\rangle }_{{{L}^{2}}\left( \Omega \right)}}{{e}^{-\lambda _{n}^{D}t}}\phi _{n}^{D}\left( x \right)},f\in {{L}^{2}}\left( \Omega \right)$, but I don't know how to find the norm.

share|improve this question
I think also that in your question statement your $u$ and $w$ are not in quite the correct function space. It appears you probably want $u \in C^1([0,\infty);L^2)\cap C^0((0,\infty);H^2)$ instead? – Willie Wong Jun 18 '12 at 12:46

1 Answer

up vote 2 down vote accepted
+100

For the norm part, you have:

$$ \langle E_D(t) f, E_D(t)f\rangle = \sum_{n\in \mathbb{N}} e^{-2\lambda_n^Dt}\langle \phi_n^D,f\rangle^2 $$

by definition. So the norm, being

$$ \sup_{\|f\| = 1} \|E_D(t) f\| $$

is just given by

$$ e^{-\lambda_0^Dt} $$

where $\lambda_0^D$ is the smallest eigenvalue of the Dirichlet Laplacian. Furthermore, $h = \phi_0^D$ the corresponding eigenfunction would be the norm-achieving example desired.


For the Neumann boundary condition case, note that $u \equiv c$ is a solution. It is clear that such a solution maximizes the norm for $E_N(t)$, that is given $h \equiv c$ you have $E_N(t)h = h = \|E_N(t)\|h$.

share|improve this answer
1  
Note that by standard elliptic regularity, the eigenfunctions are automatically $C^\infty$ and hence $H^2$. – Willie Wong Jun 18 '12 at 12:46
Thanks, I didn't think in constant solution – Cristian González Jun 19 '12 at 9:14

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.