Non-uniqueness in a free boundary problem (Q951990)

The author proves that a result of \textit{J. L. Lewis} and \textit{A. L. Vogel}, from ``Uniqueness in a free boundary problem'', [Commun. Partial Differ. Equations 31, No. 11, 1591--1614 (2006; Zbl 1189.35130)], on uniqueness in a free boundary problem for the \(p\)-Laplace operator is sharp in the case of two dimensions. Namely, for a given compact convex set \(F\subset\mathbb{R}^2\), an open set \(G\supset F\) and \(1<p<\infty\), \(a>0\), if \[ K_p(F,G)=\inf\{\displaystyle\int| \nabla \theta| ^p\,dx,\,\,\,\theta\in C_0^{\infty}(G),\,\,\,\theta=1\,\,\,\text{on}\,\,\,F\}>0, \] then there exists a bounded domain \(D\) which is not convex, a \(p\) harmonic function \(u\) and a corresponding measure \(\mu\) which satisfy \[ \begin{cases}\nabla\cdot(| \nabla u| ^{p-2}\nabla u)=0\,\,\,\text{weakly}\,\,\,\text{in}\,\,\,D\setminus F,\\ u(x)\to 1\,\,\,\text{as}\,\,\,x\to y\in F,\\ u(x)\to 0\,\,\,\text{as}\,\,\,x\to y\in\partial D,\\ \mu=a^{p-1}H^{n-1}\,\,\,\text{on}\,\,\,\partial D.\end{cases} \] Moreover, the above measure \(\mu\) satisfies the condition: for all \(\beta>0\) and \(r_0>0\) there exists \(\tilde r\in (0,r_0)\) such that \(\mu[B(x,\tilde r)]>\beta \tilde r^{n-1}\), where \(B(x,\tilde r)=\{y\in\mathbb{R}^2,\,\,| x-y| <\tilde r\}\).