A duality approach to nonlinear diffusion equations (Q475362)

The paper is concerned with the nonlinear diffusion problem with a monotonically increasing discontinuous nonlinearity, that is, \[ \begin{cases}\frac{\partial y}{\partial t}-\Delta \beta (t,x,y)\ni f,\text{ in }Q:=(0,T)\times \Omega,\\ -\frac{\partial \beta}{\partial \nu}(t,x,y)\ni \alpha \beta (t,x,y),\text{ on }\Sigma:=(0,T)\times \Gamma,\\ y(0,x)=y_0, \text{ in }\Omega,\end{cases}\leqno(1) \] where \(\Omega\) is an open bounded subset of \(\mathbb{R}^N\) (\(N\leq 3\)) with its boundary \(\Gamma\) of class \(C^1\), \(\beta\) has a convex continuous weakly coercive potential \(j\), \(\partial/\partial\nu\) denotes the normal derivative on \(\Gamma\) and \(\alpha\) is positive. The author proves the existence of the generalized (variational) solution of problem \((1)\), provided by a solution to an equivalent minimization problem for a convex functional involving the potential of the nonlinearity and its conjugate. In the case when the potential does not depend on space, the null minimizer in the minimization problem coincides with a weak solution of the nonlinear equation.