Generalized quasilinearization and semilinear elliptic boundary value problems (Q1584636)

The paper deals with an extension of the method of quasilinearization (due to R. Bellman and R. Kalaba) and generalized quasilinearization (due to the authors) to classical and weak solvability of boundary value problems for semilinear elliptic equations. More precisely, let \(L\) be the linear uniformly elliptic operator \[ L\equiv -\sum_{i,j=1}^n a_{ij}(x) {{\partial^2}\over {\partial x_i\partial x_j}} +\sum_{i=1}^n b_{i}(x) {{\partial}\over {\partial x_i}} +c(x) \] with \(\alpha\)-Hölder continuous coefficients defined on the closure of a bounded and smooth domain \(\Omega\subset {\mathbb R}^n\) with \(c(x)\geq 0.\) Consider the semilinear boundary value problem \[ \begin{cases} Lu=f(x,u)+g(x,u)\quad & \text{in } \Omega,\\ Bu=\varphi\qquad & \text{on } \partial\Omega,\end{cases}\tag{*} \] where \[ Bu\equiv p(x)u+q(x) {{\partial u}\over {\partial\gamma}}, \] \(\gamma\) is the unit outer normal vector to \(\partial\Omega,\) \(p(x)\) and \(q(x)\) are \(1+\alpha\)-Hölder continuous functions defined on the boundary \(\partial\Omega\) and \(p(x)>0\) (in order to get a non-degenerate problem (*), one should require also \(q(x)\geq 0!\)). Imposing suitable regularity and structure conditions on \(f\) and \(g,\) the authors derive classical solvability of (*) through the method of generalized quasilinearization. The same technique is applied also in proving weak solvability in \(H^1_0(\Omega)\) of the Dirichlet problem \[ \left\{\begin{alignedat}{2} Lu&=f(x,u)+g(x,u) \qquad&&\text{in}\Omega,\\ u&=0\qquad&&\text{ on} \partial\Omega,\end{alignedat}\right. \] for the divergence form operator \[ L\equiv -\sum_{i,j=1}^n {{\partial}\over {\partial x_j}} \left( a_{ij}(x) {{\partial}\over {\partial x_i}} \right)+c(x), \qquad a_{ij},\;c\in L^\infty(\Omega) \] {}.