Regularizing effect of the interplay between coefficients in some nonlinear Dirichlet problems with distributional data (Q2196667)

The authors study the existence of regular (with respect to the summability or boundedness) weak solutions of the problem \[u\in W_0^{1,2}(\Omega):\ -\mbox{div}(M(x,\nabla u))+a(x)u=-\mbox{div}(F),\] where \(\Omega\) is a bounded set in \({\mathbb R}^N\), \(0\leq a(x)\in L^1(\Omega)\) and \(-\mbox{div}(M(x,\nabla u))\) is a classical nonlinear differential operator, defined by a Carathèodory function \( M(x,\xi)\) satisfying, for some \(0<\alpha\leq\beta\), and for almost every \(x\in\Omega\), \[M(x,\xi)\xi\geq\alpha\vert\xi\vert^2, \ \ \ \vert M(x,\xi)\vert\leq\beta\vert\xi\vert, \ \ \ \forall\xi\in{\mathbb R}^N,\] \[[M(x,\xi)\vert- \vert M(x,\eta)](\xi-\eta)>0, \ \ \ \forall\xi, \eta\in{\mathbb R}^N, \xi\neq\eta .\] The authors prove that the solution \(u\) of the above problem has exponential summability under the only assumption that there exists \(R>0\) such that \(\vert F(x)\vert^2\leq R\, a(x)\); furthermore, they prove the boundedness of \(u\) under the slightly stronger assumption that there exists \(R>0\) such that \(\vert F(x)\vert^p\leq R\, a(x),\) \(p>2\). In order to prove these results they follow an approximation method, that is, they approximate the vector-valued function \(F\) by a sequence of bounded vector-valued functions \(F_n\) for which there is a solution \(u_n\in W_0^{1,2}(\Omega)\cap L^\infty(\Omega)\) of the problem. Since \(u_n\) belongs to \(L^\infty(\Omega)\), this will allow them to use either exponential, or power-like, test functions in the problem.