I am trying to implement a numerical scheme for a PDE, and it takes as input dirichlet boundary conditions, i.e. $u=0$ on $\partial\Omega$, as well as the hessian of $u$ on the boundary also.

I am handling a non standard boundary condition, called the second boundary condition, where we prescribe $\Omega,\Omega^*\subset\Bbb R^2$, convex, and our requirement is that $\nabla u(\Omega)=\Omega^*$.

It can be shown that this is equivalent to $\nabla u(\partial\Omega)=\partial \Omega^*$.

But say we can find a map $-F:\partial\Omega\to\partial\Omega^*$, of closed form, so we can state:

$\nabla u(x)=-F(x)$ on $\partial\Omega$,

Then taking the divergence of both sides we get:

$-\Delta u=f(x)$, on $\partial\Omega$, where $f(x)=\nabla\cdot F$.

So now we have poissons equation, but my issue is that the PDE is defined on a closed domain, namley $\partial \Omega$.

Is this a pathological problem? since if not, we can find $u$ (with a free constant), and by induce dirichlet boundary conditions.

Thank you for your help in advance.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.