A distributed control problem for two coupled fluids in a porous medium (Q2810061)

The author studies a mathematical model in which two non-miscible fluids flow in a porous medium, their temperature presenting a jump on the interface. He considers a bounded and connected domain \(\Omega =\Omega _{1}\cup \Omega _{2}\cup S\) \(\subset \mathbb{R}^{3}\). He assumes that \(S\) is planar and that \(\Omega _{1}\) and \(\Omega _{2}\) are convex and symmetric with respect to \(S\) (he will finally assume that \(S\subset Ox_{2}x_{3}\) and that \(\Omega _{1}=\Omega \cap \{x_{1}<0\}\), \(\Omega _{2}=\Omega \cap \{x_{1}>0\}\)). He considers Darcy's law \(\text{div}(u_{i})=0\), \(\frac{\mu _{i}}{k}u_{i}=-\nabla p_{i}+\rho _{i}f_{i}\) and the heat equation \(\rho _{r}u_{i}\cdot \nabla T_{i}-\theta _{i}\Delta T_{i}=Q_{i}\) in \(\Omega _{i}\), \(i=1,2\), where \(u_{i}\) , \(p_{i}\) and \(T_{i}\), respectively, represent the velocity, the pressure and the temperature of the fluid \(i\), \(\rho _{i}=\rho _{r}(1-\alpha _{i}(T_{i}-T_{r}))\) for some reference \(\rho _{r}\) and \(T_{r}\). On the interface \(S\), he imposes the non-miscibility conditions \(u_{i}\cdot n_{i}=0\) and \(-\theta _{i}\frac{\partial T_{i}}{\partial n_{i}}=\lambda (T_{i}-T_{j})\) , \(i,j=1,2\) \(i\neq j\). The part \(\partial \Omega _{i}\setminus S\) is divided into two pieces. On one piece \(p_{i}=p_{i}^{0}\) and \(T_{i}=\tau _{di}\) are imposed, while on the second piece \(u_{i}\cdot n_{i}=0=\frac{\partial T_{i}}{ \partial n_{i}}\) is imposed.NEWLINENEWLINEThe first main result of the paper proves the existence of a unique solution \((u,T)\) of a variational formulation associated to the problem, under hypotheses on the data. The author proves that this weak solution satisfies \(u_{i}\in (H^{1}(\Omega _{i}))^{3}\) and \( T_{i}\in L^{\infty }(\Omega _{i})\) and the existence of a unique pressure \( p_{i}\in H^{1}(\Omega _{i})\) such that \((u,p,T)\) satisfies the original problem in a weak sense. The main tool is Schauder's fixed-point theorem. The author then studies a distributed control problem for the pressure, introducing the cost functional \(J(P)=\frac{1}{2}\left\| P-p_{d}\right\| _{2}^{2}\). He introduces a variational formulation of the original problem now involving the pressure and he proves the equivalence between the two variational formulations. The second main result of the paper proves the existence of at least one solution to this control problem and the author finally analyzes necessary optimality conditions for this problem.