Boundary Value Problem in Non-smooth Subsets

44 Views Asked by At

I'm trying to wrap my head around a certain situation involving a boundary value problem. Here's the scoop:

We've got a Lipschitz open and bounded subset $\Omega \subset \mathbb{R}$, with its boundary $\partial\Omega$ composed of two disjoint, open sets $\Gamma_1$ and $\Gamma_2$ (so $\partial\Omega = \Gamma_1 \cup \Gamma_2$).

I'm considering a function $f$ that lies in $L^{2}(\Gamma_2)$ and another function $u$ from $H^{1}(\Omega)$. $u$ happens to be the solution to:

$$ \begin{cases} -\Delta u=0 & \Omega\\ u=0 & \Gamma_1 \\ u=f & \Gamma_2 \end{cases} $$

Let's say we have a non-smooth subset $B$ of $\Omega$. Can we still say that $\Delta u = 0$ applies inside $B$? And if so, in what sense might that be true? The challenge I'm facing is that I don't have access to any regularity theorems.

Any insights or suggestions on how to proceed, given this limitation, would be greatly appreciated!

1

There are 1 best solutions below

0
On BEST ANSWER

Yes, you can say that. Essentially, the what you are trying to show is an interior property, so the boundary information is actually irrelevant. Indeed, that fact that $u\in H^1(\Omega)$ satisfies $\Delta u=0$ in $\Omega$ means that $$ \int_\Omega \nabla u \cdot \nabla \varphi \, dx =0 $$ for all $\varphi \in C^\infty_0(\Omega)$.

Now if $B\subset \Omega$ is open then any $\varphi \in C^\infty_0(B)$ is also in $C^\infty_0(\Omega)$, so $$ \int_B \nabla u \cdot \nabla \varphi \, dx =\int_\Omega \nabla u \cdot \nabla \varphi \, dx =0. $$ Thus, $\Delta u=0$ in $B$ in the weak sense. In fact, this argument still holds true in the very weak sense, that is, we say $u\in L^1_{\mathrm{loc}}(\Omega)$ satisfies $\Delta u=0$ in $\Omega$ in the very weak sense if $$ \int_\Omega u \Delta \varphi \, dx =0 $$ for all $\varphi \in C^\infty_0(\Omega)$.

Also, you mention that 'you don't have access to regularity theorems', but you do! In particular, all the interior regularity theorem still remain valid in you situation, so $u$ is actually smooth (in fact, analytic) in $\Omega$ just not up to the boundary. Since this is the case, the statement also holds in the classical (strong) case too.