The behavior of the free boundary near the fixed boundary for a minimization problem (Q859768)

The paper analyzes the behaviour of the free boundary arising from the minimization problem for the functional \[ \int_\Omega | \nabla u(x)| ^2 +q^2(x) \lambda^2(u) \] where \(q(x)\) is a given smooth function, \(q(x)\neq0\) \(\forall x,\) and \[ \lambda^2(u)=\begin{cases} \lambda^2_+ & \text{if}\;u>0,\cr \lambda^2_- & \text{if}\;u<0, \end{cases} \] with \(\lambda^2_+-\lambda^2_-\neq0,\) \(\lambda_+>0,\) \(\lambda_->0.\) Taking \(\Omega\) to be smooth domain and the admissible class of minimizers \(\{u\colon\;u-f\in W^{1,2}_0(\Omega)\}\) with smooth enough \(f,\) the main result of the paper asserts that \textit{at contact point between the fixed and the free boundary where \(f\) and \(\nabla f\) vanish simultaneously, the free boundary approaches the fixed one in a tangential fashion.} This complements the result due to \textit{A. Gurevich} [Commun. Pure Appl. Math. 52, No. 3, 363-403 (1999; Zbl 0928.35213)] which shows that if \(\nabla f=0\) whenever \(f=0,\) then \(f\in \text{Lip\,}(\overline\Omega).\)